5 The Establishment of Social Safety Nets and the Implications for Macroeconomic and Structural Adjustment


My presentation will touch on four basic issues as they affect low-income groups in developing and transforming economies: the need for macroeconomic and structural adjustment; the implications of adjustment; the mix of adjustment policies and social safety nets that can moderate the short-run negative impact of adjustment; and finally, the implications of social safety nets for macroeconomic imbalances and adjustment.

Ke-young Chu

My presentation will touch on four basic issues as they affect low-income groups in developing and transforming economies: the need for macroeconomic and structural adjustment; the implications of adjustment; the mix of adjustment policies and social safety nets that can moderate the short-run negative impact of adjustment; and finally, the implications of social safety nets for macroeconomic imbalances and adjustment.

The Need for Adjustment

High and volatile inflation harms the poor in both the long and short run. In the long run, it creates an environment detrimental to productive investment, sustained economic growth, and poverty reduction. In the short run, the poor frequently see their incomes and few assets eroded by inflation and are unable to protect them because of a lack of access to inflation hedges. Moreover, governments often try to stifle high inflation by controlling certain prices, and the resulting combination of inflation and price controls can be even worse for the poor. And while relatively high-income, well-connected individuals may enjoy easy access to a variety of goods at low official prices, as explained below, the poor may have to rely on parallel markets for even essential goods and pay far higher prices. For these reasons, low inflation and liberal pricing not only help to bring about sustained economic growth and to reduce poverty over time but also may help the poor more immediately.

Distorted relative prices can also negatively affect low-income groups. In many countries, governments use administrative controls to keep the prices of essential commodities (e.g., foodstuffs) at unrealistically low levels. Proponents of these controls argue that the main objective of using them is to maintain low food prices for the poor. However, food prices are often so low that farmers have no incentive to increase production. In fact, the prices may be so low that the foodstuffs intended for human consumption are used to feed livestock or, in some cases, are smuggled to neighboring countries where prices are higher. Consumers also do not have much incentive to use foodstuffs efficiently. Moreover, because those who are well off in general consume more food than those with little money, subsidized food prices provide more benefits per person to the well-off than to the poor, in absolute amounts. In addition, controlled prices often cause the emergence of parallel or black markets. Thus, despite its efforts to hold down the cost of basic necessities for low-income groups, the government may find that many households are paying a higher price for food than they would in a decontrolled market.

To maintain low consumer prices, governments either limit producer prices (thus effectively taxing farmers) or incur large budgetary subsidies for consumers. Because low producer prices reduce production incentive, the government may also find that scarce foreign exchange must be used to import food to satisfy demand. In any case, consumer subsidies drive up the fiscal deficit, reducing the availability of resources for productive investment and growth and putting pressure on the overall price level through bank financing of the deficit.

Adjustment and the Poor

In discussing the effects of adjustment on the poor, it is necessary to distinguish between short-run and long-run implications. In the short run, the fiscal adjustment aimed at establishing conditions for sustained economic growth (and, through growth, poverty alleviation) requires higher taxes or lower public expenditures. While it would be ideal to reduce nonessential expenditures, it is not always possible, and cuts may have to be made in social or capital spending. (Some of these cuts may not harm the poor, to the extent that prereform social programs are inefficient or poorly targeted.) It is also important to consider how the poor might suffer in the absence of adjustment. Certain adjustment measures can have short-run negative effects on low-income groups, but in the absence of such measures, these groups might suffer even more over the long term.

As a result of the various policy measures included in economic adjustment programs, some groups gain and some lose. For example, when food prices are raised, domestic net producers of foodstuffs—often small farmers, many of whom are poor—are likely to gain, unless the government has been maintaining dual food prices (realistic producer prices and unrealistically low consumer prices) before the adjustment. If the government has been maintaining dual prices, raising the consumer prices of foodstuffs increases the prices for consumers and reduces budgetary subsidies. However, if the government has been financing food subsidies by taxing farmers (i.e., by maintaining low producer prices), then raising food prices for both producers and consumers benefits farmers but has negative effects on net consumers of foodstuffs.

When relative prices change—as a result of exchange rate devaluation, for example—domestic resources tend to be reallocated from the nontradable to the tradable goods sectors. In turn, workers in the tradable goods sectors will tend to have better employment opportunities and higher incomes over time, while workers in the nontradable goods sectors are likely to face declining job opportunities and incomes. It is clear that not all groups lose, and if those who do—either through unemployment or a decline in real income—fall into poverty, they should receive help from the government. However, it must be recognized that because of the nature of adjustment, changes have to take place in the relative income positions of different groups and that these shifts are an important means of providing incentives to reallocate resources (including human resources) from less efficient to more efficient sectors. When the losers are particularly vulnerable, these negative effects become important social issues.

The best policy for a given magnitude of adjustment is to design a broad mix of adjustment policies that will minimize negative effects on low-income groups. In reducing the fiscal deficit, the government can either increase taxes or reduce spending. In raising taxes, the government can use different mixes of tax instruments, ranging from sales taxes or excises (which often have significant negative implications for some poor groups) to income taxes (which can be designed to be progressive, although limited administrative capacity can severely constrain the implementation of a progressive income tax system). Broadening the tax base with relatively low rates is likely to be more equitable than increasing taxes by boosting rates on narrow-based taxes such as excises. In the case of expenditures, improving expenditure efficiency is as necessary as protecting certain essential social expenditures.

These are policy mix questions, questions of how to minimize the negative economic effects resulting from adjustment policy choices. However, even if the government makes every effort to minimize these negative effects, there will still be declines in real incomes for some groups, including those with low incomes. Therefore, the adjustment programs have to include certain additional measures–referred to here as social safety nets—to mitigate these negative effects.

Moderating the Losses

There are two sets of circumstances in which social safety nets are a particularly important issue. One develops in the presence of the large consumer subsidies, typically including food subsidies, that exist in virtually all reforming countries. For example, in 1993, Uzbekistan maintained budgetary subsidies equivalent to 8 percent of GDP; the administered price of a loaf of bread was less than half of its production cost. Subsidies accounted for 8.4 percent of GDP in Armenia and 14.3 percent of GDP in Belarus. In reducing the budgetary burden and reforming the pricing system, such countries face a number of critical issues: the extent of changes in real prices of foodstuffs and other essential commodities, the speed with which these changes should be implemented, and their effects on household expenditures. In considering these issues, the government must decide how best to mitigate negative effects. In this way, subsidy reform gives rise to the need for a social safety net. The other situation develops when unemployment rises on account of structural adjustments such as public sector retrenchment programs and enterprise reform. A social safety net can mitigate the negative impact of such an increase.

As regards consumer subsidies, one important issue is often to make subsidies more targeted. Rather than eliminating subsidies right away, governments can revamp them so that they benefit only those in need. Practical approaches include targeting subsidies either to certain geographical regions (e.g., urban areas) or to certain socioeconomic groups (e.g., children, the elderly, and the unemployed); a mix can also be used. Alternatively, the government can phase out subsidies by increasing consumer prices to cover production costs and mitigate the effects of the price increase on low-income households by providing targeted cash transfers.

In terms of unemployment, adjusting countries face various problems. In countries of the former Soviet Union, newly established employment funds confront financial and administrative constraints. Declining real wages and increasing unemployment have reduced the wage bills on which the payroll tax that finances unemployment benefits is based. Moreover, newly established employment offices have only limited capacity to deal with the increasing number of unemployed and underemployed workers. Many developing countries either have unemployment benefits with severely limited coverage or do not have any formal social security arrangement at all. Moreover, the social implications of adjustment are imposing increasing financial and administrative strains on the formal arrangements that do exist. The question then is how to soften these strains and address the needs of the people who are outside the formal social security system. In some cases, public works programs can be useful. A well-designed public works program with low wages can be effective as a self-targeted mechanism, thus eliminating the need for elaborate means testing. People given the opportunity to work elsewhere at reasonable wages will not seek employment in a public works program, because the wages these programs offer are very low.

The choice of instruments to be used in specific contexts and the mix depends on the adjustment measures, the composition of vulnerable groups, existing social policy arrangements, and administrative capacity.

Social Safety Nets and Macroeconomic Imbalances

Adjusting economies need to find the financial resources for social safety nets. It is often argued that social safety nets cannot be integrated into a reform program because of the financial constraint. It is also argued that what really matters is achieving economically efficient adjustment; if a large macroeconomic imbalance is a source of economic ills, then reducing this imbalance in the most efficient possible manner is an important policy objective. However, these arguments do not recognize that macroeco-nomic and structural policies and their distributional implications are practically inseparable. The real issue is how to achieve economically efficient and socially sustainable adjustment. Moreover, establishing social safety nets does not necessarily mean having to search for additional resources. It often means reforming inefficient and unsustainable existing social programs and integrating them into adjustment programs. In such cases, establishing social safety nets means reducing the fiscal imbalance. Subsidies can be reduced by 50 percent or more through improved targeting, and various social benefits, such as pensions and child allowances, can be more effectively targeted to the truly vulnerable. Thus, economic and social objectives are not necessarily in conflict; in fact, they often reinforce each other.

Robert Liebenthal

My talk follows quite well from what Ke-young Chu has just said. For the last year or so, I have been working with the countries of the former Soviet Union (FSU), specifically with the issues related to establishing or strengthening social safety nets in those countries. Thus, my comments will focus on only the FSU countries. But I want to add that having this diverse seminar group as an audience has forced me to think more carefully than I have before about the ways in which the countries of the FSU differ from other countries, both developed and developing.

First, I will present some background on IMF and World Bank involvement with the FSU countries. Then I will describe some characteristics specific to these countries that affect not only social safety nets but the economies for which these nets are designed. Next, I will outline three different concepts of the social safety net, each of which is potentially of considerable importance to the FSU countries, and the kind of policy strategies necessary for implementing them. By way of conclusion, I will briefly set out some of the key issues the FSU countries confront at the moment.

The World Bank’s and IMF’s involvement in structural reform in the FSU countries began in late 1991. At that time, the two organizations and the governments involved saw social safety nets as a top priority in the reform process. Social safety nets were given the same emphasis as other major issues such as macroeco-nomic stabilization, enterprise and financial sector reform, and the improvement of price and trading arrangements.

While it is interesting, in the context of what Mr. Chu has just said, to consider whether the IMF and the World Bank would have given the same emphasis to such matters ten years ago, the fact is that they did not. World Bank and IMF staff have since learned a great deal in the last decade about handling social policies during transition periods and the importance of social safety nets, and this knowledge was brought to bear on the task of analyzing the FSU countries and their needs. Over the last 18 months, the World Bank has completed social sector studies in Ukraine, the Kyrgyz Republic, Russia, and Belarus. The Russian Federation has received a loan for employment services and social protection, and a lending pipeline now exists for FSU countries that includes around 12 human resources projects of various kinds, covering either social safety nets, training schemes, or health-related projects. IMF and World Bank staff, in collaboration with colleagues in the affected countries, are developing a better understanding of how to deal with some of the problems of transition economies.

Special Characteristics of FSU Countries

Certain features of the FSU economies make these countries rather different from others in terms of the way social policy should be handled during the transition. The first singular characteristic is the unemployment situation. Unemployment as it is known in the West, and even as it is known in developing countries, simply did not exist—at least officially—in the centrally planned economies. The prospect of joblessness was held off by underemployment and overstaffing at enterprises; in short, the enterprises absorbed the unemployed or unemployable. As a result, the obvious task of social safety nets in the FSU, as Mr. Chu mentioned, is not only to provide financial assistance but to help people develop the job-seeking skills they have never needed before.

Second, the history of centrally planned economies was officially one of price stability. It is common knowledge that suppressed inflation existed and that price stability did not represent economic realities. But the impact on daily life of the shift from stable prices to near-hyperinflation, and the subsequent problems people have faced, cannot be underestimated. The situation is certainly extreme when compared with the experiences of other countries, with the possible exception of some Latin American countries.

A third interesting feature is that income transfers—that is, pensions, allowances, and subsidies of various kinds—totaled around 40-50 percent of GDP in most of the FSU countries in 1991, when the transition began. Some newly independent countries made adjustments in these payments that raised the costs considerably. These figures are similar to, or even higher than, those in Organization for Economic Cooperation and Development countries and certainly far above those in most developing countries.

A fourth characteristic, at least for the last two years or so, has been the sharp decline in output. Output has dropped in Russia by some 30 percent since 1990; in Latvia it has fallen by around 50 percent. In fact, in all the FSU countries, the decline in output has been of devastating proportions, of a kind not seen anywhere in the world since the Great Depression. However, welfare has declined less, since much of the lost output has been in the military sector or other areas that do not really affect personal consumption.

Fifth, the FSU countries have huge state enterprise sectors. In effect, almost all businesses are in the public sector, and all require restructuring and privatization. Moreover, under the former Soviet system, state enterprises supplied much more than wages. They were also responsible for providing many social services—child care, health benefits, and even hospitals—for their employees.

The sixth notable feature has been the absence of a small-scale, privately owned service and trading sector during the start-up of reforms. Such “informal” sectors are often an important part of the social safety net in other countries. In most of the developing world, people do not necessarily enter open unemployment the way they would in developed countries. Rather, they depend on extended families, private transfers, and other informal coping mechanisms. Such mechanisms were not available in the FSU when the reform process started.

Last, but not least, the whole process of reform has been and continues to be politically traumatic. For the governments themselves, decision making is complicated, in large part because of the difficulty of achieving a consensus among different parts of the political system—a problem that has been exacerbated by falling living standards, economic uncertainties, and social pressures. And it is fair to say that a degree of institutional paralysis exists when it comes to dealing with some problems.

Three Types of Social Safety Nets

Let me turn quickly to the three concepts of the social safety net with which I will be concerned. The most narrow concept assumes that the social safety net provides only the means for dealing with the transitional social costs of adjustment, such as unemployment and inflation, by meeting immediate needs for income support. The second, broader concept holds that the social safety net or social policies are part of the system for supporting market reforms. Measures such as proactive labor market policies, including training; improvements to the housing market, which are extraordinarily important in most FSU countries, where housing shortages have been a major constraint on the more efficient use of labor; tax policies; wage subsidies; and public works programs all support the reform process. These measures also help make the labor market function more effectively, enabling workers to move in order to meet the needs of those parts of the economy experiencing growth.

The third and broadest concept of the social safety net envisions the system delivering comprehensive human resource services, including health care and education. These services are important not only as welfare measures, but also as contributions to the overall well-being and productivity of the economy.

Policy Approaches

What kinds of policy approaches are useful in FSU countries? The first step in implementing social policy measures is to identify the poor, in part by establishing a reasonable “poverty line” that can be defended but is relatively austere. The next step is to target the cash benefit system to people below this line, reducing or even eliminating benefits for those well above it and tightening the system so that it cannot be abused. Obviously a state-financed system should not provide benefits to people who really do not need them. Getting the benefits to the right people is the essential first element in targeting.

The IMF and World Bank recommend a number of specific policy measures in FSU countries. One of the most important is raising the retirement age. Pensions in the FSU countries account for around 7 percent of GDP. The retirement age for women is 55, for men 60—an obvious inequity—and both these ages are below standard retirement age in many Western countries. Even industrial countries are currently trying to raise the standard retirement age because of the costliness of pension schemes and the problems of intergenerational transfers. Countries of the FSU must confront these issues, too.

Means-tested social assistance, such as the targeted public works projects mentioned by Mr. Chu, is a benefit of last resort. However, in many Western countries, some kind of social assistance is available to those who fall through the net—that is, who fail to qualify for other cash benefits. The idea behind means-tested assistance is that anybody who is poor and can demonstrate it according to a predetermined standard is entitled to receive some sort of help from the state. Because much of the unemployment in FSU countries is likely to be long-term structural joblessness, and because groups other than the unemployed are experiencing poverty, such assistance is needed alongside other benefits such as unemployment compensation and training schemes. Moreover, this type of benefit is an effective substitute for existing programs that are more expensive but less effective in dealing with poverty, such as subsidies.

Other strategic measures that relate directly to the broad-based concepts of the social safety net described earlier include supporting effective labor market policies. These measures would support activities to assist people in learning to conduct a job search; help deal with the fallout from mass layoffs; and promote training and other interventions to assist people in physically moving to obtain work or in learning another occupation. The result would be a move toward a flexible, responsive labor market.

Finally, there are the issues of restructuring in the health and education sectors. Both sectors make heavy demands on public resources in the FSU. They are the areas in which the public sector has traditionally been involved, and there are many ways the resulting expenditures (on health and education, for example) can be made more effective. They should not necessarily be cut back—in fact, there may be underinvestment in health and education in some FSU countries—but the return that can be achieved on such expenditures needs to be improved. The overall goal is to reduce the total fiscal cost of the publicly financed social welfare or social support system in order to promote fiscal sustainability and greater macroeconomic stabilization. At the same time, social programs can be targeted in the ways mentioned earlier, so that they become more effective in reducing poverty. The task is not an easy one, politically or otherwise.

Four Pivotal Issues

Let me conclude by setting out four of the most important issues that have emerged during the efforts to create social safety nets in FSU countries.

The first is employment. The experience of the East European economies in transition has shown that at some point unemployment will start to grow rapidly. Open unemployment in the FSU is still relatively low; the highest figures are now in the Baltics, where joblessness stands at about 4-5 percent (still well below the levels in Western Europe and the United States). The rate is sure to climb, but how high, and when, and in what context are open questions. How to prepare for it and what priority to give it are even more problematic and uncertain issues.

Second, there is a serious problem in the enterprise sector. As I mentioned before, enterprises in FSU countries have traditionally provided many social services. As enterprises are reformed, restructured, privatized, and subjected to hard budget constraints, these services will come under considerable strain. Some enterprises will decide to maintain them, but some firms will try to shift the responsibility to, for example, local authorities. The possibility that the local authorities might not be able to deliver these services then effectively becomes an important potential deterrent to enterprise reform. To move the enterprise sector toward a market economy, something must be done about social services, or the groups that will suffer most from reform—trade unions, retirees, and so forth—will resist change altogether.

A third dilemma is that more than output and social welfare is threatened in these countries; much of the rest of the system is also endangered. One especially good example is the system of vocational training institutions, which in Russia alone constitutes an extremely large industry. Vocational training accounts for around 3 million students, young adults currently being trained for occupations that, for the most part, are unlikely to be useful in the future. Any system that employs 3 million young adults and is on the verge of collapse is a major social problem.

Finally, there is the problem of the working poor. A large number of people formally employed in the enterprise sector are technically poor—that is, according to the best available information, their incomes are below any reasonable estimate of a poverty line. These are the people who are likely to move into unemployment relatively soon, as labor market pressure continues to manifest itself. In the meantime, these workers are poor even though they are employed, and they make up a surprisingly large group.

These are the pressure points and the priorities. Which should be dealt with first? Should they all be confronted at once? Should they be prioritized in turn? Deciding what to do is a difficult problem. And one thing that makes it so hard to solve these dilemmas is another dilemma within the social welfare system itself: the extent to which the FSU countries maintain income support through unemployment benefits and other sorts of assistance, as opposed to trying to maintain levels of essential social services.

The final point is that the responsibility for delivering social assistance and social services in the FSU countries has now devolved largely on local authorities. The problem thus becomes how a central government can ensure that a social safety net exists at the local level, where it really counts.


Steven Dunaway

My talk will address the ways certain issues relating to social safety nets affect operational aspects of the IMF’s work. Let me start by noting that the participants in this seminar represent the full range of IMF member countries—that is, industrial, developing, and transition economies. While there are significant differences in the basic conditions of each group, across all of them, social safety nets have become an important part of the economic policy debate.

The basic rationale for adjustment and reform is that they will ultimately allow an economy to use its scarce available resources more efficiently, increasing the benefits to society over time. But the idea of “adjustment” implies the need for sacrifice and the imposition of a burden on the economy. Likewise, structural reforms, while ultimately improving an economy’s supply capacity and raising the sustainable level of income growth, may also impose a burden, because reallocating resources to more productive activities takes time and can entail significant costs. Social safety nets are intended to ease the burden that adjustment or structural reform imposes on individuals and to limit disruptions to the overall economy. However, it is essential that the safety nets not impede the adjustment process too much. In a broad sense, IMF lending programs themselves can be viewed as safety nets, because they aim to cushion the burden of adjustment and limit the disruptive effects on the international trade and payments system. The conditionality that comes with these loans acts as a means of ensuring that adjustment actually takes place.

The IMF touches on issues relating to social safety nets in its relations with all member countries, either in the context of lending programs, as mentioned, or through the surveillance process and Article IV consultations. There is no standard institutional formula for creating an “ideal” social safety net; as Ke-young Chu mentioned, the most appropriate structure depends on the economic situation, social conditions, and cultural heritage.

In analyzing members’ needs, IMF staff focus on both the types of social assistance needed and alternative means of providing them. These analyses tend to be based on five principal considerations.

  • The assistance should be carefully targeted so that it reaches the intended recipients and does not have broader, often unintended effects.

  • The cost of assistance in monetary terms should be as transparent as possible, because officials must know the potential costs in order to make effective decisions.

  • Any distortions the assistance might cause should be understood and minimized.

  • The costs and benefits of alternative means of providing the same assistance must be considered in order to minimize costs and maximize benefits.

  • The implications of the assistance for overall economic efficiency have to be recognized.

In industrial countries, most issues relating to social safety nets involve the fiscal position and economic efficiency. For example, New Zealand introduced reforms in its social welfare schemes in late 1990, both as part of its efforts to reduce the budget deficit and as a complement to labor market reforms. The new reforms emphasized more careful targeting of unemployment benefits and ways of reducing disincentives to work. Eligibility criteria were tightened, the replacement ratio was lowered, and greater incentives were provided for individuals to find new work. New Zealand sought to achieve a better balance between the need for adjustment in the labor market and the need to cushion the adjustment burden associated with unemployment.

In Canada, a supplemental regional unemployment benefits scheme was established to cushion the adjustment burden in regions of the country that were experiencing economic difficulties arising from the decline of major industries. However, the scheme was systematically exploited through various informal job-sharing arrangements that allowed individuals to maintain eligibility for benefits and avoid relocating. In the end, the net result was to inhibit adjustment by reducing incentives to reallocate resources.

In developing countries, social safety nets tend to be less well established, and related issues tend to involve building appropriate schemes. In general, subsidies provided by the government tend to raise important issues. An example would be the commodity price stabilization funds in Papua New Guinea. The sharp drop in world prices for major agricultural commodities that began in the late 1980s has had a profound effect on the country’s agricultural sector. Among producers, the plantations have suffered the most, and their problems, in turn, have disrupted other parts of the economy—especially the financial system, which holds substantial claims on the plantations. Because they tend to have lower costs and greater flexibility, small producers have been less severely affected by the drop in prices.

Price stabilization funds for major tree crops (coffee, cocoa, coconuts, and palm oil) have been a feature of Papua New Guinea’s economic system since the country gained its independence. They have generally functioned well, smoothing the fluctuations in the domestic prices of these crops that arise from sharp increases or declines in world market prices, thus limiting unnecessary adjustments in the agricultural sector. In the past, the funds have been largely self-financing, but the sharp sustained fall in world prices over the last few years has depleted their resources. Subsequently, government resources have been required to continue the price supports. While the supports can be justified on economic grounds as necessary to keep the production base from eroding, and on social grounds as essential to preventing a sharp drop in employment (especially important in a country with substantial unemployment and problems with law and order), persistence of the supports at a high level impedes adjustment. The result is that output and employment have been largely maintained, but only at a substantial budgetary cost.

The transition economies have a combination of problems, some similar to those in industrial countries, others similar to those in developing countries, and some unique. Like many industrial countries, some transition economies have comprehensive social welfare systems that may inhibit work incentives and reduce economic efficiency. At the same time, like developing countries (primarily the more advanced), the transition countries need to develop certain aspects of the social safety net that have not been necessary before, such as unemployment benefits. The absence of these aspects of social safety nets could impede the transition process.

At the same time, some rather unique institutional arrangements have evolved in these countries. Robert Liebenthal pointed out one of the most important: the number of social welfare functions taken on by state-owned enterprises, including the provision of health care, housing, and education. This situation has created a major impediment to enterprise reform and makes the question of how to separate these responsibilities from the enterprises a key issue. Complicating the situation is the fact that the governments in these countries do not currently have the resources to take on these social welfare functions themselves.

Summary of Discussion

The discussion began with participants raising the point that social safety nets are often seen as appendages to the main business of reform and are often introduced late in the reform process. Many governments are concerned about the potential high fiscal costs of safety nets, and the ministries charged with implementing these programs are often administratively weak. There is a clear contradiction between trying to improve the fiscal balance and simultaneously introducing substantial new expenditure programs linked to social protection. IMF and World Bank speakers suggested that the contradiction is more apparent than real, since in most developing and reforming economies there is often considerable scope for tax reforms or reductions in government expenditures that could make some room for increased social spending. It was also noted that in some of its fast-disbursing adjustment loans, the World Bank has limited the restrictions on the use of counterpart funds so that some of this money can be used for social services. The World Bank has also begun offering specific funds for social welfare programs, such as employment services in East European countries.

Participants also observed that nongovernmental organizations (NGOs) are often asked to take on a larger part in social welfare activities. How can their work be integrated with that of governments? IMF speakers argued in favor of a significant role for NGOs and did not think new government initiatives in the social sectors should be allowed to crowd out NGO activities. Robert Liebenthal agreed but emphasized that approaching NGOs only when there is an immediate crisis to contend with is not a good idea. Governments, and for that matter the IMF and the World Bank, could try harder to find ways of working with these organizations on a long-term basis.

A question was raised about the distributional judgments the IMF makes in designing its programs. IMF speakers indicated a preference for leaving as many of the distributional judgments as possible to the governments and political processes of the countries concerned, in accordance with IMF policies. Nonetheless, if certain vulnerable groups are clearly harmed as a direct consequence of an IMF-supported adjustment program, then the IMF can seek to include measures in the program to mitigate the damage. However, according to the IMF speakers, a proactive approach on distributional issues as such is somewhat outside the institution’s scope.

This stance provoked mild criticism of the IMF for its apparent timidity in the area of social safety nets. On occasion it appears that the IMF recognizes the merit of such schemes without being completely sure why it is getting involved with them. The legitimacy of IMF involvement can be defended easily enough, given the institution’s mandate to support sustainable adjustment programs in countries facing balance of payments problems. A poor distributional outcome from a particular reform program would place the program’s sustainability in serious doubt.

Mr. Liebenthal reminded the seminar of the detailed policy statements made on the subject of poverty in the World Bank’s 1990 World Development Report. The report noted that promoting a healthy and growing economy is the best way to reduce poverty. Considering the volume of World Bank lending currently devoted to human resources, Mr. Liebenthal felt that it was inappropriate to describe the World Bank’s approach as “timid.” He also observed that the mandate to help reduce poverty does not necessarily have implications for the pattern of income distribution in a particular country.

The IMF speakers argued that there had been a very considerable change in the IMF’s attitude on this matter in the last five years. The institution is now making a more systematic attempt to address the problems of the poor. However, there is great resistance to extending conditionality to cover such issues, or indeed any issues that veer from the IMF’s macroeconomic focus. Several participants applauded this commitment not to extend the scope of conditionality.

There was also some discussion of the short- and long-term aspects of the slowly improving social conditions in reforming economies. It was suggested that if there is not a significant private sector response to reforms within a few years, these economies are likely to face serious difficulties. The income growth and tax revenues the private sector generates are what will promote economic security in the long run. Many of the short-term interventions, important as they are, are primarily palliatives, not solutions.

Finally, it was noted that the nature and outcomes of social programs are much less precise than those of narrowly defined economic reforms. Because of this lack of clarity, and because of uncertainty in their own countries about how to integrate social programs and broader reform measures, some participants hoped the World Bank and IMF would develop more expertise in this area. However, most participants recognized that, difficult as the social welfare aspect of reform may be, it is ultimately the responsibility of the individual countries.

Proceedings of the Seminar Coordination of Structural Reform and Macroeconomic Stabilization, Washington, D.C., June 17-26, 1993