1011* THE ALLEVIATION OF POVERTY UNDER STRUCTURAL ADJUSTMENT Lionel Demery Tony Addison FILE COPY A Wbrld Bank Publication The Alleviation of Poverty under Structural Adjustment Lionel Demery Tony Addison The World Bank Washington, D.C. The authors wish to acknowledge the comments and encouragement of David Beckmann. The views and interpretations in this study are those of the authors and do not necessarily represent the views and policies of the World Bank or of its executive directors or the countries they represent. @ 1987 The International Bank for Reconstruction and Development I THE WORLD BANK 1818 H Street, N.W Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing November 1987 Second printing December 1988 ISBN 0-8213-0956-0 For the past decade and a half the developing countries have been increasingly obliged to adjust their economies to an adverse internation- al economic environment of oil shocks, world recession, and deteriorat- ing terms of trade. These difficulties have been compounded in many cases by inadvisable domestic policies, which have aggravated macro- economic imbalances. Yet concern over the effects of adjustment policies on the welfare of poor or vulnerable groups has only recently become evident. In their attempts to assist the poor under structural adjustment, policymakers have not been particularly well served by theoretical or applied economics. Economic theory has neglected the tradeoffs be- tween macroeconomic balance and distributional balance, and applied analyses have faced methodological difficulties. In this booklet we review a variety of programs that have tried to help the poor under structural adjustment and propose a simple framework in which they can be analyzed. Adjustment policies may adversely affect the poor in two ways. In the short run such policies may reduce the real income and consumption of poverty groups. A World Bank study compared such initial adverse effects to a "crossing of the desert," in which those who are least able to cope with the crossing require some temporary relief to tide them over. Second, even over the long term, some poverty groups may not benefit from the processes set in motion by the adjustment effort. Adjustment policies will shape development and influence the distribu- tion of income for years into the future and will have different effects on the poor. The relationship between adjustment and poverty therefore extends beyond a notion of social cost that features only short-run effects. In this review, we try to give as much attention to the potential for poverty alleviation through adjustment in the longer term as to poverty problems during the transition. Because it is so difficult to measure the net effects of gains and loses among different groups of poor people (Addison and Demery 1985), we do not attempt an overall assessment of the effects of adjustment on poverty. But we do assume that it is possible to identify the particular I 2 THE ALLEVIATION OF POVERTY poverty groups whose living standards are most vulnerable to adjust- ment and, in some cases, to take corrective action. Although our examples are drawn mainly from World Bank experience and documen- tation, we have not tried to assess the overall effects of Bank operations. The experiences reviewed here are meant only as illustrations of the kinds of programs that have been undertaken to assist the poor under structural adjustment. Our purpose is to show that efforts are being made to alleviate the problem and that in many cases World Bank assistance has played an important part. But the examples are not presented as necessarily typical of Bank-supported adjustment programs. Examples are also presented of programs undertaken by other agencies, where these hold important general lessons. Finally, we seek to demonstrate the technical feasibility of an equitable approach to adjustment. Although the thornier issue of political feasibil- ity is outside our scope, we recognize that political constraints are important in circumscribing policy choice, particularly during difficult periods of adjustment. Those groups that are favored by systems of market intervention are certain to exert their political energies to prevent a move toward market liberalization. These pressures will be critical in determining the implementation and pace of adjustment. Analytical Framework This section defines terms and develops a simple analytical framework with which to organize empirical material on the alleviation of poverty under adjustment. The term "adjustment" has come to be used in a loose and sometimes misleading fashion (Killick 1986). It can refer to the objective of adjusting the balance of payments deficit itself, but it has also been used to describe the accommodating changes needed in the domestic economy to correct the external deficit. We shall take the latter definition as the basis of discussion. A distinction is also made between "stabilization" and "structural adjustment." The former refers to short-run policies that realign domestic absorption with domestic supply; the latter, to policies designed to influence the equilibrium configuration itself (Buiter 1986, p. 1) by making structural reallocations and changing some of the ground rules of production and exchange (through institutional reform or privatiza- tion, for example). UNDER STRUCTURAL ADJUSTMENT 3 As our starting point we take Stewart's distinction between the primary and secondary distributions of income. She distinguishes be- tween "primary claims on resources which arise directly out of the productive process of work and accumulation, and secondary claims which result from the transfer of primary claims" (1983, pp. 4-5). This distinction helps to direct our attention to the potential role of adjust- ment in alleviating poverty. Insofar as the poor are economically active, a potential exists for raising their primary incomes through their direct involvement in the process of structural adjustment. Benefits may also extend to some of the nonworking poor who receive transfers from family members who work. But if the poor are not economically active or do not receive family transfers, the only means of aiding them would be through their secondary claims: by increasing their income and consumption transfers. The advantage of a strategy that seeks to alleviate poverty by improving primary claims is that it can be broadly consistent with the objectives of structural adjustment. A strategy that relies on transfer payments, in contrast, is likely to encounter the problem that the resources for such schemes are at their most limited in times of austerity. Furthermore, transfer schemes can divert resources from structural adjustment objectives. If existing transfer schemes are modified in the adjustment process, however, to be better targeted to the poor, then resources might be saved overall and the potential conflict between transfers and fiscal restraint reduced. A guiding principle therefore would be to increase the productive capacity of the poor to the greatest extent possible, taking recourse to transfers only when a strategy of adjusting primary claims is not feasible. In this booklet we distinguish five broad approaches to assisting the poor under adjustment: * Increasing their access to productive assets * Raising their return on assets * Improving their employment opportunities * Ensuring their access to education and health services * Supplementing their resources with transfers. The first four are intended to raise the primary incomes of the poor, while the fifth increases their secondary incomes. Conversely, each approach identifies a way in which adjustment measures may harm the poor. 4 THE ALLEVIATION OF POVERTY Increasing Access to Productive Assets Possession of productive assets plays a crucial role in getting people onto "income escalators," a principal theme of the "redistribution with growth" strategy (Chenery and others 1974, pp. 43-44). The primary claims of the poor can be raised by increasing their access to productive assets. This approach, though not central to a structural adjustment program, might be consistent with its objectives, in that output and productivity may be raised in the process. Raising the Return on Assets The primary claims of the poor can also be increased by raising the return on the assets they hold, especially by increasing output prices and productivity and reducing input prices. This approach may be imple- mented through specific policy interventions or broad measures under structural adjustment to dismantle market distortions. Improving Employment Opportunities If the poor possess few productive assets other than their labor, increasing their wages or access to employment may be the only means of raising primary claims. Measures to increase wages (such as minimum wage legislation), however, may conflict with the objectives of the adjustment effort. Not only would minimum wage legislation usually fail to benefit most of the poor in developing countries, it would also tend to reduce total employment. Thus, the thrust of a policy to raise the labor incomes of the poor during adjustment should be to increase their access to productive employment. Increasing the geographic and occu- pational mobility of the poor may be necessary to ensure that they benefit from the expansion of employment, particularly in sectors producing tradable goods, that may result from the structural adjustment program. In addition to such measures to improve the operation of the labor market, emergency employment schemes may be necessary. Ensuring Access to Education and Health Services The human capital of the poor can be protected by guaranteeing their access to health and education services. Such interventions, by "invest- ing" in the poor, will yield returns in the form of higher productivity and incomes. But they will also confer gains in consumption and thus will partly consist of secondary income claims. UNDER STRUCTURAL ADJUSTMENT 5 Supplementing Resources with Transfers The preceding approaches assist those among the poor who are both economically active and are, or could be, engaged in the production of traded goods. But in many cases the poor are not economically active or are locked in the nontradable sector. Even in working households, some family members-young female children, for example-may not share equally in the family's resources. In such cases, income or consumption transfers may be the only effective means of assistance. Before considering these measures in detail, it is important to emphasize the context that structural adjustment provides for the alleviation of poverty. First, if structural adjustment reduces price distortions, enhances competition, and erodes monopoly profits and economic rents, the incomes of the wealthy are likely to fall and those of the poor to rise. The reduction in market distortions brought about under adjustment may therefore enhance both efficiency and equity. Second, if structural adjustment ultimately promotes economic growth, the alleviation of poverty will undoubtedly be an easier proposition. But unless adjustment strategies are well designed, the benefits of economic growth will bypass some poverty groups. The challenge facing policymakers, especially in countries with chronic poverty, is to direct structural adjustment toward the poor-to identify a strategy of adjustment with equitable growth. Increasing the Access of the Poor to Productive Assets Poor people can share in the process of structural adjustment if their access to productive assets is increased. Policies to this end can help both to alleviate poverty and to achieve structural adjustment because they increase not only consumption but also investment. Our review of this approach highlights land reform as an important means of alleviating rural poverty. In many developing countries poverty is principally a rural problem, and landlessness is a prime cause (Griffen and Khan 1977, Lipton 1985). Improving the availability and terms of credit are also discussed here as important means of increasing the access of the poor to capital assets. Our discussion is confined to material assets; the provision of health and education services, which can be considered as asset transfers of human capital, are discussed later. At first sight land reform may seem to be too administratively 6 THE ALLEVIATION OF POVERTY complex, to be too politically difficult, and to take too much time to fit into a short-term adjustment program. Indeed, few of the World Bank's borrowers have chosen to include land reform in their structural adjustment loans (SALs). Thailand, however, in two recent SALs, attempt- ed a measure of land reform under structural adjustment. This experi- ence, which indicates the concern of both the Bank and the Thai government to reduce the adverse impact of other structural adjustment policies on the poor, will be discussed in detail. Structural Adjustment in Thailand: Land Reform and Poverty Dependent on imported energy, Thailand faced economic difficulties in the early 1980s after the second oil shock. In 1980 the government undertook a program of policy reform to achieve balance of payments viability over the medium term through structural adjustment. Two SALs were disbursed, the first in 1982-83 and the second in 1983-84. Both entailed fiscal policy changes and sectoral reforms. In particular, attempts were made to reorient industrial growth toward the export sector and to reduce energy price distortions. For our purposes the most interesting feature of the Thai adjustment has been the agricultural policy reforms. That program included measures to compensate farmers in the poorer Northeastern region for the adverse effects of structural adjustment on their living standards. This compensation entailed increasing their access to cultivable land. Throughout much of the 1970s the domestic price of rice was controlled through an ad valorem tax on the f.o.b. price, among other policy instruments. Rice policy through the 1970s transferred income from farmers to city dwellers, particularly to urban employers and the government, while it buffered farmers from wide fluctuations in world prices. Variations in the price of rice had only short-run effects on the urban real wage, because increases in the real wage induced rural-urban migration, which soon returned wages to their former levels. The low domestic price of rice did not benefit urban workers over the long term. Rather, urban employers gained greater profits because of the low urban product wage (Bertrand and Squire 1980). But under Thailand's structural adjustment, much of this tax structure has been changed. According to World Bank data, reductions in export duties for rice have lowered the tax burden from more than 30 percent in 1981 to around 4 percent in 1985.The effect of these adjustments on the distribution of income in Thailand cannot be overestimated. Rice not UNDER STRUCTURAL ADJUSTMENT 7 only is an important source of foreign exchange, but also provides half of the calories in the Thai diet. As a wage good, its price is critical in determining real income levels of the bulk of the population, especially the poor. Moreover, rice is the heart of the agriculture sector, accounting for 40 percent of agricultural gross domestic product. Ninety-eight percent of the country's 4 million farming families produce rice. Thus changes in rice pricing and other policy reforms of the structural adjustment programs in Thailand have benefited most farmers, probably at the expense of urban employers. Farms that were net surplus rice producers undoubtedly have benefited. But in a recent assessment Trairatvorakul (1984, p. 23) found that a quarter of Thai rice-producing farms actually consume more rice than they produce and have probably been adversely affected by the increase in the price of rice. Many of these farms were technically illegal, because they were on land classified for forestry, although much of it was suitable for permanent agriculture. About 4 million hectares of this forest area was being cultivated illegally by farmers who had no tenure. Thus, as part of Thailand's structural adjustment, an effort was made to increase the security of tenure of such farmers by reclassifying the land and granting "right to farm" certificates, known locally as STKs. These certificates did not confer ownership on farmers. The program has proceeded patchily. The Land Development Depart- ment has used SAL counterpart funds to accelerate survey work and reclassification of land, but despite these additional resources and strong political support progress has been hampered by the limited capacity of the department. The Forestry Department, with partial funding under the SALs, has proceeded faster in surveying illegal cultivation and in granting right-to-farm certificates. These have been issued to about a third of the million eligible households. A recent study by the Bank (Feder and others, forthcoming) has confirmed doubts about the benefits of STKs. On the basis of sample data, it found that squatter farmers (including those holding STK certificates) had lower productivity, made fewer land improvements, and used fewer inputs than land-owning farmers who were similar in other respects. The reason for the lower output performance of squatter farmers lay in their poor access to institutional credit, which usually requires collateral that only titled landowners can supply. Two conclusions emerge from this study. First, if the squatter farmers of the Northeast are to be assisted 8 THE ALLEVIATION OF POVERTY under structural adjustment, they must be granted full ownership rights and not simply sTK certificates. Second, conferring full ownership on these farmers would promote not only equity, but also structural adjustment: "The results of the study suggest that the provision of full legal ownership to squatters in rural Thailand is a socially beneficial policy, as the productivity gap between squatters and legal owners is substantial" (Feder and others 1986). Although the land reforms introduced under the Thai SALs have been limited, the exercise has encouraged a continuing policy dialogue in the country. In fact the World Bank has recently funded a land policy study (conducted by the Thailand Development Research Institute), which is to translate the findings of recent research on land reform into policy recommendations. Although the political processes behind land reform programs inevitably require a long-run perspective, the initiatives under the SALS seem to have acted as a catalyst for such change. The Thai land reform example illustrates how the objective of increasing output under structural adjustment can also help in alleviating poverty. Similar expectations are currently held in the Philippines, which is revitalizing its land reform program. Policymakers are hopeful that the land reforms will help achieve the twin objectives of reducing rural poverty and hastening economic recovery. The early stages extend existing programs; later stages are expected to encounter more political opposition. The government, however, is faced with serious financial constraints in implementing even the early phases of the program, which require some US$300 million. At present, it is seeking financing from external sources. The Philippines also illustrates how land reform can assist the poor directly affected by a recession. On the island of Negros sugarcane workers and their families have been severely affected by the decline in the world price of sugar. The price fall led to production cutbacks, layoffs, and reductions in already low daily wages. Some unused sugarcane land has been lent to unemployed workers to grow food, and negotiations are under way to make more land available. The new government has announced that estate owners should allocate 10 percent of their land to sugar workers. An Assessment of Land Reform under Adjustment It is difficult to generalize about the feasibility and desirability of increasing the access of poorer groups to productive assets under UNDER STRUCTURAL ADJUSTMENT 9 structural adjustment programs. First, it is not clear that in every case the transfer of these resources is optimal for economic efficiency. The poor are generally located in areas where investment opportunities are limited and rates of return are low. Moreover, land reform measures may disrupt production at a critical time, and governments may be reluctant to meddle with land tenure systems if doing so entails a short-run loss of output. In the Thai case there were clear economic advantages in granting farmers legal rights to land, because returns were higher under cultivation than they would have been under forestry. The main economic problems that have emerged can be attributed to the neglect of complementary inputs and supporting services for farmers. Given that the poor in developing countries derive most of their livelihood from agriculture, land is the most obvious productive asset they require, and land reform the most effective policy instrument to increase their access to it. But land reform has serious political ramifica- tions that call into question its feasibility as a routine element in structural adjustment. The prospect of land reform was very much on the agenda in discussions on structural adjustment between the Bank and the government of Kenya. In the event, it was not taken up. The reasons for the failure to proceed with this reform are undoubtedly political, and most observers are not optimistic that much land reform can be achieved in Kenya in the foreseeable future (House and Killick 1981). In the Thai case the reforms concerned themselves with strengthening the property rights of squatter farmers, rather than the more fundamental and difficult task of redistributing land from one social group to another. But even in the Thai case, where there was a measure of success, reforms did not go far enough. They stopped short of conferring full tenure on poor farmers and appeared simply to satisfy the political objectives of the government. Land reform under the Thai SALs has proceeded only so far as political expediency permitted. A start, however, has been made to protect poor groups who would otherwise suffer under structural adjust- ment. The prospects for building on this initiative remain to be seen. Raising the Return on Assets Held by the Poor Perhaps the most obvious way to involve the poor in structural adjustment and to raise their income in the process is to increase the rate of return on the assets they hold. In many cases the poor possess 10 THE ALLEVIATION OF POVERTY nothing but their labor, so the only way to increase their income is to increase their real wage or employment (this strategy will be discussed later). But there are cases in which the poor hold other productive assets. Raising the return on these assets by increasing output prices and productivity or by reducing input costs can promote both structural adjustment and a reduction in poverty. Such measures offer real pros- pects for improving the well-being of the rural poor in most developing countries, since their activities are generally directed toward the produc- tion of tradables. For the urban poor, who usually find their jobs in the informal economy, there would seem to be fewer opportunities for such a strategy, given the dominance of service activities in urban informal sectors. It is possible, however, that the tendency for the informal sector to produce nontradables is due to previous policies that have favored this emphasis. Under a different policy regime, the informal sector may respond rapidly to incentives to produce tradables. The desirability of removing distortions in product and factor markets is a tenet of structural adjustment. Many of these distortions, such as import controls, confer significant income on the favored few, mainly through the creation of rent income. Insofar as this income is derived from unproductive activities, the distortions transfer income to higher- income groups. The removal of distortions may therefore reallocate resources away from relatively well-off groups and cause increases in productive activity. The distributive effects of imposing and removing market distortions are complex and will differ between the short and long term (Addison and Demery 1986). In the short term, which allows no time for factor mobility, the changes in individual incomes are likely to be greater than in the longer term, which allows the reallocation of both capital and labor. For example, market interventions favoring urban areas will initially inflict greatest harm on rural producers and confer greatest benefit on urban groups. But over the long term such measures will result in rural-urban migration that reduces the income differential caused by the distortions. Similarly, the consequences of removing distortions can be expected to vary over time. Initially, urban groups will suffer the greatest loss of income (possibly the same workers who migrated under the distortion regime, though not necessarily), and rural producers will have the greatest benefit. But factor mobility (such as migration back to rural UNDER STRUCTURAL ADJUSTMENT 11 areas) will tend to reduce these effects. Still, the conclusion is clear: the removal of policy biases, by changing relative prices and the rates of return obtained from various economic activities, is certain to affect incomes and poverty. These effects are beginning to be seen in Sub-Saharan Africa, where structural adjustment efforts in a number of countries have focused on eliminating biases in agricultural policies. Inappropriate policies, espe- cially pricing policies, have long been recognized as one of the reasons for poor performance in the agricultural sector in that region and elsewhere (World Bank 1981, 1984a). We shall review the experience of C6te d'Ivoire, particularly the short-term effects of its stabilization and structural adjustment efforts. The dramatic impact of stabilization and structural adjustment on urban and rural areas in C6te d'Ivoire is not exceptional. In Tanzania during 1980-84 real farm incomes are estimated to have risen by 5 percent, while urban wage earners faced a decline of 50 percent; in Ghana during the same period farm income stagnated, but urban incomes fell by 40 percent in real terms (World Bank 1986a, p. 19). But there are important reasons for examining the C6te d'Ivoire case. First, the data are better than those usually available for Sub-Saharan Africa. Second, the World Bank has already studied this question for Cte d'Ivoire. Third, the experience of C6te d'Ivoire is dramatic and illustra- tive. The changes in both incomes and policy have been significant over the past five years or so. These changes have affected rural areas differently from urban, and some poor groups differently from others. In particular, the C6te d'Ivoire example illustrates how improving incentives in agriculture can also improve the equity of income distribution. Adjustment in Cte d'Ivoire: Prices and Poverty The financial difficulties faced by the government of C6te d'Ivoire in 1980 required an immediate response. With the support of the Interna- tional Monetary Fund (IMF), the government undertook stabilization policies, with fiscal and monetary contraction the most important elements. A measure of public sector restructuring was also initiated. These measures were followed, under two SALs from the World Bank, by a more far-reaching program of policy reforms that required structural adjustments over the medium to long term. The stabilization policies were effective, especially after 1982. A 12 THE ALLEVIATION OF POVERTY government deficit of 12.4 percent of gross domestic product (GDP) in 1982 was transformed into a surplus of 1.5 percent in 1985. Between 1981 and 1984 real GDP declined by 4.0 percent a year, but by 1984 the overall balance of payments moved into surplus, after having reached a deficit of 7.7 percent of GDP in 1980 and 1983. The policy changes affected both the structure of the economy and the distribution of income. The deflationary policies followed by the government from 1980 to 1984 and the drought of 1983-84 make it difficult to attribute observed outcomes to the structural adjustment policies. There can be little doubt, however, that the policy changes (including the stabilization measures) hit the urban formal sector hardest and benefited agriculture the most. Thus the structural adjustment program, by improving agricultural prices, raised rural incomes above what they would otherwise have been and cushioned farmers from the harsh economic realities of the early 1980s. A principal component of adjustment in C6te d'Ivoire was the attempt to restore incentives in agricultural production by raising producer prices in line with world prices. The rural-urban terms of trade rose from 88.5 in 1982 to 100 in 1984, partly as a result of these price reforms. The combination of the stabilization measures and these structural changes significantly altered the distribution of income. The economic contraction throughout most of the 1980s was concentrated in the urban sector, where per capita disposable income declined by an estimated 10.8 percent a year between 1980 and 1984, compared with a reduction of only 1.2 percent for rural areas. Among urban groups, the better-paid suffered the most serious declines in real income. Their income fell by 18.2 percent a year, compared with 10.8 percent for the middle-income and 3.7 percent for the low-income groups. Most of the low-income groups work in the informal sector, which fared better than the formal sector during the demand contraction. By 1984 the ratio of urban to rural income had declined to about 2.5 to 1, compared with 3.5 to I in 1980-and this gain in equity for rural groups was achieved in the face of the 1983-84 drought. By 1985 the ratio had fallen to about 2 to 1. Thus, by restoring agricultural incentives, the government has im- proved the rural-urban terms of trade, reduced rural-urban income inequalities, and established a basis for long-run recovery. The recovery will be built upon a third SAL and the growth-oriented stand-by arrangement recently negotiated with the IMF. UNDER STRUCTURAL ADJUSTMENT 13 Agricultural Prices and Inequality: Some Issues Although there is no doubt that structural adjustment helped maintain farmers' incomes during this difficult period, a number of issues have to be faced in assessing the effectiveness of this approach to increasing the incomes of the poor. Are price reforms enough? Do they increase inequalities among the rural poor? And do they overburden the urban poor? INVESTMENT IN RURAL INFRASTRUCTURE AND SERVICES. Adjusting prices may not be all that is needed to increase the output and incomes of target groups. More often than not, the poor in developing countries are located in areas with little access to roads, transport, communication, agricultural services, marketing facilities, and so on. Improving prices may be a necessary condition for restoring incomes, but not a sufficient one. If farmers cannot get the supplies and services they need, infrastruc- ture investments may be required to give these farmers the capacity to increase output and yields. In a review of the effects of adjustment policies in Sub-Saharan Africa, we have argued that the improvement in agricultural prices is only one (albeit crucial) component of food security (Demery and Addison 1987). Other important components are rural infrastructure and support ser- vices, which have not yet been sufficiently developed to permit rapid growth of agricultural output in the region. In fact, the growth of agricultural output has slowed over time. Bond (1983) attributes this trend to the "unfavorable climate for agricultural investment that has been created, the deterioration of the infrastructure and support services available to the rural farmer, and the growing lack of cash goods to the rural sector." Most observers are conscious of the desperate shortcomings of rural infrastructure in Sub-Saharan Africa. There has been a chronic depletion of the region's transportation, shortages of inputs into sectors servicing agriculture (for example, agroindustries), and severe budgetary constraints affecting agricultural investment. The potentially adverse effects of stabilization programs in reducing investment in such services cannot be denied. The remarkable growth in Indonesian agriculture illustrates how incomes and employment prospects among rural communities can be improved through a combination of increased production incentives and 14 THE ALLEVIATION OF POVERTY infrastructure development. Agricultural growth rates have been high, even by East Asian standards, averaging 4.1 percent a year during 1978-85. Indonesia imported most of its rice in the late 1970s, but now is able to grow enough to be self-sufficient. The increased rates of return in agriculture that have led to such growth can be attributed to two areas of government policy. First, under its adjustment effort the government has raised the producer prices that farmers receive, largely through the devaluations of the rupiah. The recent move away from nontariff trade barriers in the industrial sector is expected to improve agriculture's terms of trade by reducing the cost of inputs into the sector. Second, the government has given priority to expanding rural infrastructure and support services. The greater use of high-yielding varieties, irrigation, and advanced production technolo- gies has increased crop area and yields. This combination of increased incentives and improved infrastructure, in raising agricultural output, has undoubtedly improved rural incomes and employment and raised the living standards of Indonesia's rural poor. But can the restructuring of government investment and the provision of project assistance compensate for the declines in expenditure levels imposed under the constraints of stabilization? In Indonesia the concern of the government to protect investment allocations to agriculture throughout the adjustment period has led to a favorable climate for productivity growth in agriculture. In C6te d'Ivoire the proportion of government investment directed toward agriculture rose from 25 per- cent in 1983 to 35 percent in 1985. But total expenditure has declined dramatically (from CFAF272 billion in 1981 to CFAF88 billion in 1985). Despite the reorientation of government investment, the allocation of resources has fallen, and the longer-run prospects for the growth of agricultural output have been adversely affected. Increasing the return on assets held by the poor therefore may require more than improved price incentives: fundamental improvements in rural infrastructure and agricultural services may be needed, which implies a need for more resources, including development aid, than are presently available. Better targeting of assistance to poor farmers is one way to make more efficient use of available resources. In Zimbabwe, for example, attempts to increase the rate of return on assets held by the rural poor have been directed to increasing agricultural extension services in poorer areas. There are two distinct agricultural sectors in Zimbabwe, the UNDER STRUCTURAL ADJUSTMENT 15 commercial farms, which benefit from better rainfall and use modem inputs, and the traditional smallholder sector in the so-called communal areas. These farms are not so well placed for rainfall, and their access to both modem inputs and marketing facilities is inadequate. The Zimbab- wean government is aware of the difficulties small-scale farmers have in marketing output and obtaining inputs at reasonable cost. It is trying to reach these farmers through government programs, although there are serious problems because their farms are small and geographically dispersed. The agricultural extension service is now mainly directed toward the small-scale farmer in an effort to increase the application of inputs. Moreover, agricultural research (under the direction of the Department of Research and Specialist Services) is beginning to devote most of its effort to the problems of small-scale farming in Zimbabwe. INEQUALITIES AMONG THE RURAL POOR. Another, perhaps more intracta- ble problem with using pricing policy to alleviate poverty arises from the fact that not all farmers benefit from increases in producer prices. As prices rise, inequalities within agriculture may widen, and poor, less- efficient farmers may be left behind. Therefore, although rural-urban income differences may have declined in C6te d'Ivoire as a result of adjustment and the improvement in agricultural prices, it is not clear that rural poverty has been alleviated. Farm incomes tend to be much lower in the savanna of northern C6te d'Ivoire than in the southern forest region. In the savanna the average farm earns a net annual per capita income of CFAF38,000, compared with CFAFI20,000 in the better-placed southern forest region. The main crops grown in the latter region are cocoa, coffee, some annual food crops (rice, maize), and tree crops (oil palm, coconut). In the drier savanna maize is the dominant crop, followed by yams. Cotton is the savanna's principal cash crop, accounting for about 16 percent of the cultivated area. It is not clear how the increases in farmgate prices achieved during the adjustment have affected the relative incomes of these two broad ecological regions. The poorest sections of the rural community produce only food crops. Farm model data show per capita income of only CFAF34,000 in 1985 for savanna farmers who manually cultivated nothing but food crops-half the income earned by those who grew both cotton and food crops (CFAF68,000) and less than a third the 16 THE ALLEVIATION OF POVERTY income of cocoa and coffee farmers in the forest region (CFAF122,000). Unless poor farmers can switch readily to cotton cultivation, there would seem to be little promise for them in the process of price liberalization initiated by the government. In short, we endorse the recommendation of the Food and Agriculture Organization that "measures to implement price policy must be de- signed and administered so as to ensure that small farmers participate and benefit fully" (1985, p. 43). Either price increases should apply to the commodities produced by small farmers or well-targeted policies should encourage small farmers to switch production toward those commodities receiving price support. PROTECTING THE URBAN POOR. Increasing agricultural prices may also have adverse effects on urban poverty and may increase food insecurity among vulnerable urban groups. Much of the leveling of inequality between urban and rural income that has been observed under structural adjustment can be attributed to more rapid declines in urban incomes than in rural incomes. Pinstrup-Anderson (1985) has summarized evi- dence showing how the lowest income deciles suffer most from food price increases, and there is little doubt that the urban poor are the most vulnerable to such price increases. Jamal (1985) has expressed concern for the urban poor in Uganda, who have suffered as a result of price adjustments in the 1980s. The restoration of relative price parities was undoubtedly overdue in Uganda, but wage earners and other fixed income recipients have been hard hit. These have included many of the urban poor, whose food security has noticeably deteriorated. Structural adjustment, by increasing food prices, can affect the urban poor severely and should therefore be accompanied by measures to enhance their access to income-earning activities or by some form of consumption transfer. These measures are discussed later. Improving the Employment Opportunities of the Poor Adjustment often entails job losses in the formal sector, at least over the short term, until growth is resumed. The majority of the poor work in the rural or urban informal sectors and are not engaged in formal employment. But workers affected by these job cuts, if they were not regarded as poor initially, can soon find themselves among the ranks of the "new poor." UNDER STRUCTURAL ADJUSTMENT 17 For labor-abundant countries, reallocating resources in accordance with the country's comparative advantage should maximize employ- ment opportunities over the medium to long run. Eliminating policy- induced distortions in markets, particularly for producer prices, interest rates, and the exchange rate, should also maximize employment, com- pared with the alternative of allowing relative prices to deviate from their market-clearing levels. Despite these benefits of policy reform, however, short-term unemployment can arise for three reasons. First, structural adjustment will, through its effects on relative prices, reallocate resources away from nontradables, as new incentives arise in activities producing tradables. However, this process is not instan- taneous. Expenditure switching can be delayed if the perception of relative price changes is slow. More important, the time taken for production switching will depend on expectations, the speed at which investments in the tradables sector can be implemented, and the availability of supporting infrastructure. Price stickiness, because of price rigidities or imperfect information, is also a source of delay. Furthermore, segmented or underdeveloped capital markets can constrain the oppor- tunities for new investment. In general, declining nontradable sectors can be expected to respond fairly rapidly to adverse market signals, but the expansion of tradable output can proceed only as fast as the expansion of capacity. Consequently, with some enterprises contracting rapidly and others expanding at a slower pace, an employment problem is likely during the transition phase. In some cases it may be several years before labor utilization fully recovers. The impact on poverty caused by this process may therefore not be marginal. Second, the size of the resource constraint and the extent of economic disequilibriums may require rapid and tight restraint of demand to restore macroeconomic balance. Unemployment would then occur if real wages were inflexible downward because of institutional rigidity or informational deficiencies in the labor market. Stabilization therefore raises an additional set of employment problems. Although a devalua- tion combined with microeconomic incentives can promote the restruc- turing of production, demand deflation usually reduces employment faster than structural adjustment generates it. Furthermore, some demand deflation is usually necessary to release resources from nontradable activities in order to expand the production of tradables in line with the incentives created by devaluation. 18 THE ALLEVIATION OF POVERTY Finally, fiscal restraint frequently entails a reduction in public employ- ment. Since government services are mainly nontradable, the reorienta- tion of the economy toward tradables may require the release of resources from government. This release may be desirable on efficiency grounds as well. At such times large numbers of retrenched public workers may appear in the labor market. To the extent that macroeconomic equilibrium can be gained by emphasizing structural adjustment measures and placing less reliance on demand restraint, the problem of unemployed resources can be short in duration. But the problem is complicated by the high unemployment many developing countries suffered before implementing an adjustment program-unemployment resulting from the recession and inappro- priate policies that have made adjustment necessary. Consequently, governments find themselves dealing not only with the unemployment caused by adjustment itself, but also with the unemployment legacy of past events and policies. In addition most developing countries have been forced to deal with new external and environmental shocks after the commencement of their adjustment efforts. These factors will accordingly make resolution of the employment problem a long haul in most countries. Given these constraints, what can policymakers do to minimize the employment problems of adjustment? Essentially the task for policy is to increase the occupational and geographical mobility of the target groups. Many of the poor, with little or no collateral, are often excluded from acting on the new incentives offered by structural adjustment policies. Even those with some resources-such as former public em- ployees-may need assistance. But what kind of mobility should be encouraged? The best policy would be to move the poor and the newly displaced into employment and self-employment within the tradables sector, that is, into activities that produce either exports or import substitutes. This policy would be thrice blessed-it would assist the attainment of macroeconomic balance, generate incomes for the working poor, and allow transfers to be directed to the people most in need. The second-best policy would be to move people into nontradable activities. This would be only twice blessed-it would allow the poor to be self- supporting, but would not assist the resource reallocation to tradables. Exceptions arise if the nontraded activity supported in some measure the production of traded goods, or if the labor reallocation released UNDER STRUCTURAL ADJUSTMENT 19 resources for tradable production by undercutting existing costs in nontradable activities. These themes are discussed through case studies of Bolivia's Emergency Social Fund, Chile's emergency employment programs, and measures to assist retrenched public workers in The Gambia and Guinea-Bissau. The experiences in The Gambia and Guinea-Bissau illustrate how redundant public employees might be enabled to move into the private sector. Bolivia's Emergency Social Fund Bolivia has experienced severe economic difficulties in the 1980s. Large external debts were built up during the 1970s, which yielded major debt- servicing problems as real interest rates rose in the early 1980s. The debt accumulation did not contribute to an adequate development of the export sector. Macroeconomic imbalances were further compounded by fiscal imbalance and overvaluation of the exchange rate. Between 1979 and 1985 per capita GDP fell by nearly one-third, and by early 1985 the inflation rate was running at 28,000 percent a year. Against this background the government of President Victor Paz Estenssoro launched a major adjustment effort upon taking office in August 1985. Strong action was taken to cut the fiscal deficit which, through associated monetary expansion, was the driving force of the hyperinflation; the introduction of an auction system for foreign exchange led to a large effective devaluation; and trade was extensively liberalized. These reforms constitute the cornerstone of a program moving the economy back toward a market-orientated system of incentives. The program succeeded in reducing inflation to an average annual rate of 20 percent by early 1987. Inevitably, the path of policy reform has not been easy. The government had to deal not only with the problems created by previous policies, but also with further shocks, including a precipitous fall in the world price of tin in late 1985 following the collapse of the international tin agreement. In addition the price of Bolivia's natural gas exports has weakened. The elimination of hyperinflation has benefited the poor, who are typically the least able to protect their real incomes in the face of inflation. Economic rents earned by those with access to scarce commod- ities have been rapidly reduced, and the distribution of income is shifting back toward producers of tradable goods. But unemployment remains a problem. The stabilization effort has led to a short-term reduction in 20 THE ALLEVIATION OF POVERTY activity, which combined with the contraction of fin mining has resulted in an increase in numbers looking for work. In response to the social costs of recession and adjustment the government of Bolivia launched an Emergency Social Fund (Fondo Social de Emergencia, or FSE) in 1985. This was reconstituted in November 1986 as an agency reporting directly to the president. The FSE finances projects selected from proposals put forward by municipal- ities, cooperatives, nongovernmental organizations (NGos), and other community organizations. These groups then supervise the work, which is undertaken by local contractors. The projects are concerned with small-scale employment and income generation (which is intended to take 75 percent of the available funding) or with social assistance (which accounts for the remaining 25 percent). The program is financed by the government and bilateral and multilateral donors. The World Bank, through the International Development Association, has provided a credit of about US$10 million, which amounts to 20 percent of the FSE's planned 1987 budget. The FSE is envisaged to have a life of three years (1987-89) to deal expressly with the social costs of the adjustment program. At the end of the three years the activities of the FSE will be absorbed into the responsible ministries. The target groups of FSE projects include unemployed miners, workers in activities dependent on the mining sector, and producers in the informal sector (particularly women). Some projects are directed toward the families of these groups. The selection is not inflexible, and other hardship groups can be considered. Some retrenched public employees may be eligible. Targets for allocating funds among the categories of beneficiaries and the geographical areas will be decided upon after the FSE's first year of operation. A wide variety of projects are to be funded by the FSE. Among employment- and income-generating projects, the following are exam- ples of activities that are either under way or being considered by the FSE during 1987: food production, export crops, seed improvement and distribution, fish farming, reforestation, water supply 'and drainage, erosion control, irrigation works, rural access roads, and municipal improvement. Among social assistance projects are activities in low- income housing, school construction, medical and nutritional assistance, and vocational training. On average FSE projects cost US$100,000; the smallest projects are about US$20,000. UNDER STRUCTURAL ADJUSTMENT 21 In summary, the majority of projects to be supported by the FSE are intended to raise the primary incomes of the poor by offering them opportunities for productive employment and self-employment. Many of these are in activities producing tradable goods or providing essential support to the expansion of tradable output. Furthermore, the social assistance projects will enhance the income-earning capability of the poor by maintaining and raising their human capital (the subject of a later section). Finally, the FSE is a conduit through which World Bank funds can reach NGOs and other community organizations. The FSE thus forms an important link between the Bank and local organizations in their efforts to alleviate poverty. The program has not been in operation long enough to assess its effectiveness in reaching the target groups. It is therefore presented as an example of the type of assistance that can be designed to raise the primary incomes of the poor during periods of adjustment. Chile's Employment Measures After the recession and stabilization of 1974-76, Chile entered a period of export-led growth, encouraged by extensive market liberalization. Nonetheless, the economy came under increasing strain in the late 1970s, particularly after the return to a fixed exchange rate and the indexation of wages in mid-1979. The resultant overvaluation of the peso hindered Chile's adjustment to higher energy prices and to the slump in the world copper price in the early 1980s (which in real terms fell to the level of the 1930s). The economy entered a two-year depression (1982-83) during which real GDP fell by more than 14 percent-Chile's worst economic experience in the last fifty years. A stabilization program was implemented in late 1982 with support from the IMF. Economic recovery began in late 1983, and export volumes, boosted by substantial devaluations during 1984-85, reached record levels. But low prices for Chile's principal exports, the burden of debt repayments, and high unemployment added to the fragility of the post-1983 recovery. Accordingly, a SAL was implemented in October 1985 with the objective of diversifying exports, and the IMF provided further assistance. The recession had a severe impact on jobs, and unemployment peaked at 19.6 percent of the labor force by late 1982, or 33.8 percent if those in emergency work programs are included. Employment loss was 22 THE ALLEVIATION OF POVERTY particularly bad in the construction and manufacturing sectors (with unemployment rates of 50 percent and 26 percent, respectively). Never- theless, the government was able to respond with its emergency employment programs (EEPs), which were first used in the recession of the mid-1970s. These schemes mainly concentrate on the development and maintenance of urban infrastructure and are administered through municipal authorities and financed by central government. Some 240,000 persons were benefiting from the EEPs by the beginning of 1986. The three main schemes are: * The PEM. Introduced in 1975 as a temporary measure, the PEM continued and by 1983 was employing some 200,000 persons (about 40 percent of those then engaged on EEPs). By the early 1980s the PEM had largely degenerated into a transfer mechanism, as work discipline on projects became increasingly lax. Subsequent- ly, the number of beneficiaries has fallen as better-targeted work programs have been introduced and as the employment situation has improved. * The POJH. This scheme began in 1982 and has increasingly taken over from the PEM. It is limited to heads of households and pays a higher stipend than the PEM. The POJH has a higher level of work discipline than the PEM, and there has been much less misuse of the scheme by both workers and employers. * The PIMo. Introduced in 1984, this scheme provides funds to the municipalities for labor-intensive construction. PIMo projects must have an economic rate of return and must meet certain requirements for labor intensity. Stipends under these work programs are low (currently about US$26 a month in PEM and US$40 a month in PoJH). But, while the PEM has only one job category, both the POiH and the PEM have employment structures that allow higher remunerations for skills. In a survey conducted by the University of Chile in October 1983, 70 percent of those employed on work programs were willing to undertake another activity if employment were available. These work schemes serve as an "employer of the last resort" and, by enabling workers to maintain a continuous employment record, improve the chances that they can obtain regular jobs. The main limitation of the EEPs is that most do not encourage the movement of workers into activities that are being UNDER STRUCTURAL ADJUSTMENT 23 promoted by structural adjustment policies. If anything, these interven- tions may be effectively discouraging the mobility of the unemployed. Despite the shortcomings of these work programs, it is likely that they must continue to operate while unemployment remains high. Consequently, the government is taking further steps to raise the efficiency of these schemes. In 1984 a National Employment Commis- sion was established to develop and implement new and more produc- tive employment programs. Nine pilot schemes have been implemented (as offshoots of the POJH program) in fields such as gold prospecting, forestry, health, and education. Additional labor-intensive PIMO projects are also being devised to replace the inefficient PEM program. The Bank has assisted the work of the commission by financing two studies to assess employment policies and cost-effective programs. It is too early to judge whether these new schemes are operating more efficiently than the original ones. Rationalizing Public Employment in The Gambia and Guinea-Bissau The reduction of public employment is frequently necessary to control public expenditures and to raise efficiency. This issue has become especially important in Sub-Saharan Africa, where a number of reform programs (including those of The Gambia, Ghana, Guinea, Guinea- Bissau, Senegal, and Zambia) are now rationalizing the public sector. In such circumstances assistance to retrenched public employees may be warranted to help them move into the private sector. Reform programs under way in The Gambia and Guinea-Bissau provide examples of such assistance. In common with many other African countries, The Gambia has suffered a severe decline in its terms of trade in recent years. By 1983 the terms of trade index for The Gambia was one-third the level of 1976 because of a fall in the world price of groundnuts (the principal export) and a rise in import prices. The impact of external shocks was further exacerbated by low producer prices and an overexpansion of the public sector. The civil service doubled between 1976 and 1980, and public expenditures became seriously out of line with the productive base total incomes from civil service wages and salaries were almost twice the incomes paid to groundnut producers. The outcome of these processes was a 16 percent fall in real per capita incomes during 1975-85. 24 THE ALLEVIATION OF POVERTY In response to growing macroeconomic imbalance, an adjustment program was launched in 1985, with assistance from both the i and the World Bank. The floatation and subsequent depreciation of the dalasi, plus substantial increases in producer prices, have reactivated the economy's export base, and a real GDP growth rate of 5 percent has been achieved in both 1985-86 and 1986-87, after years of stagnation. A principal component of the adjustment effort has been the ration- alization of public employment. The first phase of the program began in 1985 with the retrenchment of about 1,400 public employees, and some 1,200 more workers were made redundant in late 1986. A large number of vacancies have been suppressed. The retrenched workers are almost entirely from the lowest grades of skill, and many held temporary ("unestablished") positions. With the public sector accounting for about two-thirds of modem wage employment, the scale of the redundancies has raised some difficult policy choices. It is inevitable that many of the redundant workers will suffer income losses as they transfer to other sectors. However, given that average urban incomes are four times higher than average rural incomes, the greatest priority must be given to alleviating poverty in rural areas. In this regard the adjustment program, by reducing rural- urban income inequalities, will provide some compensation to the rural poor for their previous neglect. Nevertheless, many of the retrenched workers lack resources, and their real hardship has compelled the government to assist them. Given the severe constraints under which the government is operat- ing, it was decided that the best form of assistance was to help retrenched workers move into self-employment in the private sector. This assistance has been provided through the Indigenous Business Advisory Service (BAs), an organization that promotes small businesses. IBAS provides credit and training to selected former public employees to help them establish viable enterprises. In the first months of operation close to 100 persons have been selected for IBAS assistance, and the first batch of former civil servants is now in business. The main constraint on IBAS in extending assistance is the requirement that applicants must have collateral (the organization had previously run into financial trouble when this requirement was not observed). Conse- quently, although many people have applied, the selection has inevitably been geared to those with the most resources. Furthermore, of those UNDER STRUCTURAL ADJUSTMENT 25 projects selected for assistance, over 60 percent are in retailing, which is a predominantly nontradable sector. Only 20 percent of funded projects are for farming-the priority sector under adjustment. This focus reflects the strong urban bias of most applicants, although iBAs has favored farming projects in its selection procedure. With the policy reforms now generating strong incentives for people to resume farming, it is hoped that more rural-based projects will result. Moreover, the International Labour Office (ILo) is now assisting IBAS to overcome some of its resource and logistical difficulties. Guinea-Bissau faces similar policy problems to those of The Gambia. The government is now adjusting the economy to deal with the cumulative effects of external shocks and market distortions induced by previous policies. In the first phase of a retrenchment 3,000 public workers lost their jobs-about 20 percent of the public sector work force. Most of the retrenched workers are from the lowest skill grades. Although these workers receive severance payments (equal to six months' salary), such extensive redundancies bould carry high social costs. Consequently, the government requested assistance from the ILO to help retrenched workers move into the private sector. The ILo has responded with two projects. The first project consists of resettling some 1,500 retrenched workers (mainly from the ministries of agriculture and construction) in rural cooperatives producing food. The project-which costs US$2.4 mil- lion-includes land reclamation and the provision of infrastructure. Finance has come from the United Nations Development Programme (UNDP), the European Communities (Ec), and some bilateral donors. The second ILO project is a retraining program for those retrenched from the state trading corporations. These corporations have dominated retailing, but are now being privatized, and redundant workers will be helped to move into commerce. The resettlement scheme not only provides employment assistance but also supports the objective of expanding tradable output. Guinea- Bissau has large food deficits, but more than half the potentially fertile area remains uncultivated because investment is slack and producer prices are low. However, the government is rectifying this situation. Recruitment to the resettlement schemes has been oversubscribed, which indicates that workers now perceive farming to be a better source of income than urban employment. 26 THE ALLEVIATION OF POVERTY These examples illustrate two approaches to giving assistance to retrenched workers. First, in Guinea-Bissau the state has established new rural enterprises for the unemployed. Second, in The Gambia the government has offered initial assistance but has left the unemployed to choose new occupations, on the grounds that they may be better able than the state to identify profitable new activities. Unless the schemes of the Guinea-Bissau model are well designed, The Gambia's clearing- house approach may be preferable. Finally, it must be emphasized that neither scheme will be effective in encouraging a return to the rural areas unless adjustment policies raise incomes in agriculture. Other Examples of Employment Assistance These cases are but a sample of a wide range of employment assistance programs being implemented under structural adjustment. Some men- tion should also be made of other examples. In Senegal the ILO, with funding from the UNDP, is currently estab- lishing a project to help retrenched workers start up small-scale enter- prises. The World Bank is providing similar assistance to those made redundant from public sector fertilizer distribution. In Mali the ILO is also providing assistance, while a grant from the U.S. Agency for Interna- tional Development (usAID) is covering some separation payments. The iLo has been involved in parallel exercises in C6te d'Ivoire, Guinea- Conakry, and Niger. Encouraging mobility within the private sector is also important. Some Brazilian states (particularly Sio Paulo) have given modest assis- tance to redundant workers from the private sector, particularly to facilitate their return to agriculture. A number of African countries have schemes in place to assist mobility-for example, those for unemployed urban youths to move back into agriculture. Such assistance can be particularly important in periods of adjustment when the reorganization of production can create special unemployment problems for youths. In such cases assistance to the young may be desirable to allow them to take up the new job opportunities arising from adjustment. In 1982 the Jamaican government established a program to cater to unemployed school-leavers. The Human Employment and Resource Training (HEART) program places school-leavers without a job in designated firms where they receive a weekly stipend for on-the-job training. Some of the young jobless have also been helped (again with a stipend) to enter technical UNDER STRUCTURAL ADJUSTMENT 27 training institutions specially established for the purpose. About 2,000 young Jamaicans have received HEART apprenticeships since 1982, and a further 2,000 have been sent to the training institutions. The HEART program is financed by a levy on firms larger than a certain size (the firms may, as an alternative to the levy, take HEART apprentices). Such schemes are an important policy tool in helping the young, but they must be carefully implemented to ensure that the training provided leads to skills that will be in demand. Inappropriate training could simply exacerbate the problem by mismatching skills to the available demand and by encouraging unrealistic job expectations among the young. The informal sector can also be an important area for policy interven- tion. Bolivia's FSE provides an example of how assistance could be given. A common feature of most adjustments is the absorptive role played by the informal sector. Workers disengaged from formal employment move into informal activities, either tradable or nontradable. This move is often simply a manifestation of low levels of labor utilization during an adjustment phase. In the context of adjustment, assistance should concen- trate on those informal activities producing tradables (Economic Commis- sion for Latin America and the Caribbean 1986, p. 60). The focus should be on small-scale manufactures and nonpersonal services. A variety of ways have now been tested to make technical assistance to the informal sector more cost-effective. For example, techniques (such as lending to groups) have been developed that can reduce the administrative costs of providing credit to microenterprises. The World Bank's experience with assistance to microenterprises is relatively limit- ed, but the Bank is uniquely well-placed to urge policy changes that can encourage microenterprise schemes to grow to a more substantial scale. Such policy changes include the modification of registration and other legal requirements that may be unrealistic for very small enterprises, and the liberalization of credit markets, with special attention to reducing restrictions on lending to small enterprises. An Assessment of Employment Assistance Programs that expand employment and self-employment in the private sector can be necessary during adjustment for a number of reasons. In the first place, urban groups are undoubtedly among the hardest hit by periods of austerity because they are especially exposed to the vagaries of the international economy and to the impact of direct policy 28 THE ALLEVIATION OF POVERTY interventions (including wage freezes and retrenchments). Second, ad- justment requires the reallocation of labor among sectors. By encourag- ing labor mobility, employment assistance programs can further the aims of adjustment, as well as helping the poor. Nevertheless, there is a danger that EEPs may simply be used as transfer mechanisms. Transfers may be justified if the adjustment period is expected to be short, but employment programs are not likely to be the best way of implementing transfers. Finally, a distinction should be made between groups that have benefited from past policies and those that may have been neglected. Recipients of employment assistance, especially retrenched public em- ployees, are generally not among the poorest in the community. There is a real danger-especially in Sub-Saharan Africa-that too much assistance will be given to relatively better-off groups at the expense of the very poor, who predominate in rural areas. Governments must give careful consideration to the tradeoffs involved in allocating resources among competing social groups. Ensuring the Access of the Poor to Education and Health Services Access to services that increase human capital-particularly health and education-is crucial to enabling the poor to rise out of their poverty. Over the long term such access will increase productivity and incomes. Strategies of adjustment with equity must therefore seek to protect social expenditures serving the needs of poor people. At the same time successful adjustment requires the protection of public expenditures in directly productive sectors and on supporting infrastructure. Consequently, whenever resources are constrained, some tradeoff may be implied between the respective requirements of alleviat- ing poverty and promoting adjustment. This tradeoff can be reduced by improving the efficiency and cost- effectiveness of all public expenditures. The need to restructure public expenditures under adjustment programs can be an opportunity for governments to evaluate their social programs. Higher priority can be given to services for the poor-primary health, primary education, rural water supplies, sanitation, and low-cost housing. Social expenditures can be targeted to those most in need, and the burden of economies can be UNDER STRUCTURAL ADJUSTMENT 29 placed on the relatively wealthy who have access to private health and education facilities. These issues are discussed through two case studies. First, Indonesia provides an example of a government that has increased social expen- ditures during the course of an adjustment program. Second, Brazil provides a case in which poverty-focused social services have been maintained despite successive stabilization programs. Increasing Social Expenditures in Indonesia In 1983 the government of Indonesia began a broad adjustment and stabilization effort in response to the weakening of its oil revenues. The rupiah was devalued, and an austerity budget was implemented for fiscal 1983-84 to rephase public investments, cut subsidies, and restructure the tax system. The fiscal stance was loosened toward the end of the period (1984-85) as the effects of the devaluation spurred strong growth in nonoil exports. The program was largely successful in achieving the required reduction in the current account deficit. The growth rate fell over the period, but employment was protected by the expansion of nontraditional exports and agriculture. High labor mobility between agriculture and nonagricultural sectors and public investments in labor- intensive infrastructure development also helped. With respect to the alleviation of poverty, the most important aspect of the 1983-85 adjustment was the increase in social expenditures. Overall public expenditure restraint was coordinated with the formula- tion of the fourth five-year development plan (REPELITA IV, covering 1984-85 through 1988-89). The plan shifts investment priorities in favor of the social sectors, with housing, education, and health all receiving substantial increases in their shares. The shares of some directly productive sectors are reduced (manufacturing, mining, and transport), while the shares of agriculture and infrastructure rise (but by less than the increase in social investment's share). The World Bank supported this shift in priorities, which is justified by Indonesia's limited development of social infrastructure compared with the regional average and by the increased role of private investment in the directly productive sectors. The Bank's support for the sectoral priorities of REPELITA IV was demonstrated in its recasting in 1985 of the development budget in the light of a further deterioration in oil revenues. The Bank prepared an indicative public investment program, 30 THE ALLEVIATION OF POVERTY which resulted in an investment level for REPELITA IV about one-third lower than that originally planned. But, while the indicative plan reduces the planned total development budget, housing, health, and education all increase their budget shares compared with REPELITA III. Each of these social priorities will be considered in turn. HOUSING. REPELITA IV incorporates an ambitious housing program for low-income families. The construction target is 300,000 housing units, of which half will be low-cost units. About half of the units are being built by private developers with government loans, while construction of the remainder is by public institutions. Although financial liberaliza- tion has increased mortgage interest rates, it is expected that the lowest rates will be applied to cheaper housing units, thus providing a subsidy to lower-income purchasers. Although state housing schemes have also given priority to housing units for poor households, evidence suggests that most units have actually been allocated to civil servants and others earning higher incomes. This outcome has been attributed to biases in selection criteria, and the Bank has advised that the selection procedures be clarified. Consequently the actual benefits that low-income families will get from the new housing schemes are not yet known. HEALTH. The increased share of health services in planned development expenditures includes a shift in priorities from community health care to disease control and nutrition. Although community health programs remain important, the phase of rapid expansion of these services is now complete. More resources therefore have been allocated to immuniza- tion of children and to family nutrition programs. Although the poor benefited from the emphasis on rural community medicine under REPELITA III, they are likely to benefit more under the current plan as a result of the increased priority given to controlling communicable diseases and reducing nutritional deficiencies. These benefits are partly offset by the reduction in the subsidy on drugs at rural clinics and by the increase in user charges. But the impact of these offsetting factors has so far been low. Moreover, the government is encouraging the use of private health services, which will allow a concentration of public resources on preventive medicine and the needs of the poorest, where private services are weakest. UNDER STRUCTURAL ADJUSTMENT 31 EDUCATION. Under the current plan the government is committed to improving the quality of primary education to poor families (enrollment is now nearly 100 percent) and to special groups such as the handi- capped. The Bank has also suggested that grants should be provided to private schools willing to establish themselves in rural areas and poor urban neighborhoods. Given the success of the primary education program in achieving nearly universal enrollment, resources have been switched to secondary and university education. This switch will mainly benefit relatively better-off families, although as more poor children complete basic schooling they too will be more likely to take up postprimary education. The collapse of oil prices in 1986 has meant that the government has had to make further adjustments, including further fiscal restraint, devaluation of the rupiah, and relaxation of import licensing-all measures that put greater pressure on resources for the social sector. Although the government is maintaining the priority assigned to social expenditures, it is too early to judge how these further austerity measures will affect services to the poor. Protecting Social Expenditures in Brazil In 1981 the government of Brazil began a stabilization effort to correct an overexpansionary fiscal and monetary policy and in response to a deterioration in the terms of trade and a sharp increase in debt repayments. A tight demand squeeze was applied, and a substantial devaluation was undertaken in 1983. The IMF supported the adjustment effort during 1983-85. The stabilization program cut the current account deficit in half between 1982 and 1983, and a trade surplus was achieved in 1984 through strong export growth and import compression. How- ever, the recessionary impact of external shocks plus demand restraint resulted in a fall in per capita GDP of more than 14 percent between 1981 and 1983, thus ending a remarkably long period of sustained growth. The World Bank estimates that low-income groups experienced a fall in real incomes of 20-30 percent during 1981-83. Average rural wages fell by 32 percent in real terms-the impact of recession being exacerbated by rapid growth of the labor supply. Urban real wages also fell, particularly for those with few skills. Unskilled construction workers in the S5o Paulo area, for example, lost about 28 percent of their real wages during 1981-84. Other groups of unskilled workers experienced similar hardship. Real 32 THE ALLEVIATION OF POVERTY incomes contracted in the informal sector in line with the recession in formal activity. Moreover, the unemployed were absorbed into the informal sector and provided more competition for the available business. Evidence at hand suggests that the acute fall in real incomes adversely affected nutrition and health. For example, in 1983 there was a reversal of the long-term downward trend in infant mortality, although drought in the North-east region was also a contributory factor. In addition the recession exacerbated an outbreak of child diseases in S5o Paulo in 1984. However, the impact of the recession on welfare would certainly have been greater if certain social services, with high benefits to poor families, had not continued to operate. Real expenditures on these services were either maintained or expanded. Thus spending on primary health care and basic nutrition programs increased by 73 percent, and payment-free medical consultations by 22 percent between 1980 and 1983. Overall, nutrition programs increased their food distribution by 40 percent. These programs continued to expand despite general expenditure restraint-federal expenditures on health fell by 16 percent during 1980-83, with a large fall (22 percent) in curative health care. The cut in curative health care released some resources for the primary health and nutrition programs. In addition, the expansion of feeding programs, communicable disease control, and free drug distribution was helped by a special tax (FINSOCIAL) earmarked for the social sectors. In summary, although aggregate public health expenditure was cut, there was a substantial increase in the real per capita resources committed to programs benefiting the poor. Programs providing water and basic sanitation also accelerated their expansion during 1981-83. The National Housing Bank enlarged its low-cost housing schemes and introduced more progressive tariffs for household services. The World Bank's Special Action Program, adopted in 1983, helped Brazil continue its Bank-financed projects, including some in the social sector, by increasing the Bank's share in the financing of on-going projects. Recovery began in 1984, when GDP grew 4.5 percent in real terms, and was boosted by substantial growth following devaluation and by buoyant markets in the United States. Real GDP growth increased to 8.3 percent in 1985, assisted by a relaxation of demand restraint measures. However, an acceleration of inflation led the government to tighten monetary policy toward the end of 1985. UNDER STRUCTURAL ADJUSTMENT 33 The government of President Jose Sarney (elected in March 1985) entered office with a committment to coordinate adjustment with the expansion of measures to alleviate poverty. Subsequently the Cruzado plan was unveiled in early 1986, with the aim of breaking the inflationary momentum generated by indexation and price expectations. The pro- gram involved the introduction of a new currency, a temporary price freeze, and an end to the indexation of most financial assets. The new social objectives are set out in the target plan (Plano de Metas), announced in July 1986, which focuses on improving health, nutrition, and housing for low-income Brazilians. The health strategy emphasizes strengthening the endemic disease and nutrition programs and expanding and decentralizing health pro- grams to poorer areas, especially the North-east region. This strategy is intended to enhance and extend the coverage of basic health services to the rural population and includes a program of comprehensive health care for women and children. The expansion of the nutrition programs includes a substantial increase in the targeted milk schemes. The World Bank supports the government's social priorities and has recommended ways to further improve the cost-effectiveness of social programs. The Bank is expanding its already large project commitments in the areas of primary health care, low-cost housing, water, sanitation, and integrated rural development; these projects are designed to assist low-income groups, especially in the most disadvantaged regions-the North-east and North-west. In summary, the Brazilian government has taken steps to implement social programs orientated toward the poor during the recent adjust- ment period. However, the objectives of its adjustment program have proved difficult to meet. Thus, although the Cruzado plan had some initial success in reducing inflationary expectations, it was gradually undermined by insufficient demand restraint. With a price freeze in place, mounting shortages began to develop throughout 1986, as consumer demand ran ahead of supply. The price freeze was accord- ingly abandoned in early 1987, and monetary and fiscal policy was tightened. These developments imply that harder policy choices may have to be made concerning social expenditures. Consequently, the government will have to exercise great care in devising cost-effective social programs within tighter resource constraints. 34 THE ALLEVIATION OF POVERTY Some Conclusions on Social Expenditures The principle that the market should increasingly be allowed to take up the needs of better-off groups deserves to be further explored in the context of adjustment (World Bank 1987). Chile's experience is relevant in this respect, since the Chilean government has had some success in targeting its social expenditures to the poorest families by shifting resources to primary education, primary health, nutrition, and basic sanitation. The market has been allowed to take up the supply of social services to the relatively better off, and public investment in facilities catering to the lower middle classes has fallen considerably. Moreover, Chile has raised the efficiency of its delivery of social service and reduced its costs (World Bank 1986c, p. 20). Maintaining and improving social programs targeted to poor groups can generate large gains in productivity by raising the rate of return on human capital. In Indonesia, the Bank's analysis of the costs and benefits of treating iron deficiency anemia indicates that the productivity of workers who received iron supplements for a two-month period in- creased by 15-25 percent (Basta and Churchill 1974). The productivity increase was estimated to be worth $260 for each dollar invested in iron supplementation. Examples from other developing countries generally show similarly high rates of return. Furthermore, such productivity gains, by raising the output of tradable goods (especially food), can contribute to the attainment of adjustment objectives. Supplementing the Resources of the Poor with Transfers Much of the impact of adjustment on the poor depends on whether programs concentrate on structural adjustment or stabilization. Greater opportunities are available to involve the poor in the expansion of tradable output if the external payments problem can be resolved through structural adjustment than if the payments problem and the resource constraint are so severe that they limit the adjustment effort to stabilization measures. In adjustment programs where stabilization must dominate, the need for transfers to assist the poor is correspondingly greater than in programs able to emphasize structural adjustment. Even when structural adjustment raises the productive potential of the poor, however, transfers may still be needed to ensure their welfare. UNDER STRUCTURAL ADJUSTMENT 35 Many poor families work in sectors that do not produce tradable goods (such as services in the urban informal sector). Even if they have productive assets and produce tradable goods, they are often locked into poverty by poor health and limited education. In addition, assistance to help raise their productivity usually takes some time to yield its benefits. Consequently, the poor are likely to realize the benefits of structural adjustment only over the medium to long term, while in the short term they may become even more impoverished by the measures necessary for stabilization. Furthermore, many members of poor families are not economically active (for example, young children or the aged), and they may be isolated from family transfers because of discrimination (for example, against young girls in some cultures). Their problems cannot be readily or speedily alleviated through changes in primary income claims, and income or consumption transfers may be the only effective means of providing support to them. Although the need for transfers may be greater under stabilization programs than under structural adjustment, the requirement to restrain public expenditures may also be greater. The inconsistency between increasing transfers to the poor and cutting government budgets must be resolved by restructuring such transfers to maximize the benefits to those in greatest need, while reducing the benefits to the better off. Thus adjustment can act as a catalyst in encouraging governments to reform assistance programs-just as it does in the case of social programs-to get the maximum mileage from limited allocations. This section focuses on food transfers, one of the most important ways to protect the welfare of the poor. The case of Jamaica illustrates the coordination of food transfers with adjustment, and other examples of transfer programs are also noted. Protecting Nutrition in Jamaica In the 1980s Jamaica has made substantial efforts to adjust the economy to a large terms of trade deterioration (including a slump in world bauxite prices) and the consequences of past policy distortions. A stabilization program supported by the IMF began in March 1981, and a number of SALS were provided during 1982-84 to assist export expansion and diversification. Substantial bilateral assistance was re- ceived, and extensive commercial borrowing was undertaken. 36 THE ALLEVIATION OF POVERTY Despite the policy reforms, including deregulation and public sector rationalization, the stabilization targets were not fully achieved during 1982-84. The large capital inflow in 1982 proved a mixed blessing in delaying the full correction of the overvaluation of the Jamaican dollar, despite the implementation of exchange rate reforms at the beginning of the program. GDP growth was nil in 1982, and only 1.8 percent in 1983, although 4-5 percent annual growth had been forecast at the time of the program's formulation. The shortfall reflected the very cautious expansion of private sector activity and the severe impact of the international recession on the bauxite and alumina markets. In response to these adverse developments a new adjustment effort began in late 1983. A major depreciation of the Jamaican dollar was undertaken from late 1983 through 1984, and tight expenditure controls and new revenue measures halved the central government deficit between 1983-84 and 1984-85. These measures succeeded in cutting the current account deficit by 30 percent in 1984, and GDP registered a small positive growth rate. The Jamaican adjustment was therefore achieved through stabilization rather than structural changes. Because there was little opportunity for increasing the primary incomes of the poor, transfers were the only means of helping them cope with adjustment. Adjustment in Jamaica has entailed serious social costs. Jamaica has a high dependency on imported food, partly because previous agricultural policies created disincentives to farming. Consequently, devaluation had strong effects on domestic food prices, which increased by 12 percent in 1983 and 20-25 percent in 1984. At the same time real purchasing power among low-income families declined as unemployment increased and real wages fell sharply. In a policy statement of May 1984 the government recognized that its adjustment measures, while necessary, posed a serious threat to the welfare of poor groups. Studies undertaken by the World Bank indicated that the adjustment program was having a severe effect on the poor and that there was little prospect of a reduction of unemployment in the short term. For these reasons, in 1984 the government initiated the Food Aid Program (FAP), which consists of a child feeding program operating through schools, a food supplemen- tation program for pregnant women and young children, and assistance for the very poor through a food stamp scheme. Estimates of size of the target groups are taken from the FAP. UNDER STRUCTURAL ADJUSTMENT 37 THE SCHOOL FEEDING PROGRAM. About 500,000 children (four years old or more) are officially covered by this scheme, compared with 160,000 children under a previous school feeding scheme. Unlike previous schemes, the current program is intended to reach all schools. In the expanded program, children are provided with a special lunch for the school year of 190 days. The lunch is estimated to provide about 30 percent of the children's energy (calorie) needs and to satisfy daily protein needs. In the original plan a nominal charge for the lunch was envisaged. There remain doubts, however, about how well targeted such a program is to the children of poor households, since children at all income levels benefit from the scheme. Moreover, not all poor children attend school. FOOD SUPPLEMENTATION THROUGH HEALTH CLINICS. This program pro- vides skim milk and cornmeal to pregnant and nursing women as well as children up to three years old. The target group consists of 200,000 persons, which more than doubles the numbers covered in a previous scheme. Nutritional support under the latter was low, and food supplies were irregular. The new program seeks to correct these weaknesses. At the time of the program's implementation it was estimated that children up to three years of age would be provided with 43-71 percent of the recommended energy (calorie) intake and more than 100 percent of the recommended protein intake (malnourished children receive larger allo- cations). Pregnant and nursing women would receive more than 21 percent of the recommended calorie intake and 48-58 percent of the recommended protein intake. FOOD STAMPS FOR THE ELDERLY AND VERY POOR. This program targets a group of 200,000 persons consisting of the elderly and a proportion of those officially classified as indigent. The latter group consists of the unemployed, the underemployed, and those who are unemployable through disablement. This target group contains about 40 percent of persons with a monthly income of J$50 or less, who number about 500,000 (a 1984 estimate). The criteria for selection into the group are not clear from the FAP. Those eligible for food stamps are registered through the Public Assistance and Poor Relief Records systems and receive stamps each month worth J$20. This program is estimated by the government to provide some 21 percent of the energy requirements 38 THE ALLEVIATION OF POVERTY of adult men and women and about 40-62 percent of their protein requirements. But these may well be overestimates in the light of inflation. Initially it was envisaged that food stamps would be used only for basic foodstuffs (such as rice), but currently they may be used for any food item except alcohol. The stamps have a high acceptance rate among shopkeepers. The program's overall cost was initially expected to be J$141 million (US$35.3 million) for the first fiscal year (1984-85); food imports were to be the main cost at 55.3 percent of the total. However, the food component cost has been met by food aid from USAID, the EC, the World Food Program, and the Italian government. The reduction of global food subsidies has also released public resources for these better-targeted programs. The previously large subsidy on imported rice disproportion- ately benefited better-off families and acted as a disincentive to domestic food production. The FAP dearly affords some protection to the nutritional status of the poor. But its target group (some 900,000 people) is large in relation to Jamaica's total population (2.3 million). The food supplementation schemes organized through the schools and clinics benefit many people who are not poor. Nonetheless, the school feeding program is probably the least costly method of reaching poor children. And the clinic program entails a substantial amount of self-selection, since women from better-off families are less likely to use the public health system. The food supplementation program for the elderly and very poor is the most narrowly targeted scheme, but the size of the monthly food stamp allocation is small in relation to the monthly basket of basic foods. The World Bank estimates that the value of food stamps would have to be at least doubled to J$40 a month to guarantee an adequate nutritional intake. Given the available data on these schemes, we are unable to assess the extent to which better-off groups benefit from the provisions. Serious questions must therefore remain over the size and distribution of the benefits. Other Examples of Targeted Food Assistance Although we have selected Jamaica to illustrate the use of consumption transfers under adjustment, other examples must be noted. One of the earliest examples of moving from food subsidies to targeted interven- tions is that of Sri Lanka, where subsidies were phased out and replaced UNDER STRUCTURAL ADJUSTMENT 39 by food stamps from 1979 onward. Although this change generated large savings, the government failed to maintain the real value of the food stamps. The government is now restoring the real value and refining the selection criteria to reach more of the poor. In Argentina a national food program (PAN) is currently in operation to offset some of the negative welfare effects of the adjustment effort. The PAN is administered by the Ministry of Health and Social Affairs, and its main component is the periodic distribution of food packages. It currently covers some 5.5 million people (out of a population of 30 million) and provides about 30 percent of the needs of an average family of four (World Bank 1986c, p. 18). Packages are distributed to poor families identified in two PAN census surveys (the latest being September 1985). In Chile a national nutrition program (PNAC), which is run by the Ministry of Health, distributes enriched food through health clinics to pregnant women, mothers, and young children (World Bank 1986c, p. 23). The distribution of food is monthly, and at the same time women are given health checks and the youngest children are weighed for signs of malnutrition (which is then dealt with by giving them additional food). The PNAC has a very high coverage of children (80 percent of those under six years and 92 percent of those under two), while most pregnant women benefit from the health services. The PNAC is consid- ered to be a cost-effective means of generating significant improvements in nutrition and health and an important contributor to the decline in the infant mortality rate despite the current austerity. Chile has been able to expand this program by restructuring its social expenditure programs (as mentioned earlier) and by making substantial economies in programs that largely serve middle- and higher-income groups. We have already cited the case of the Filipino island of Negros, where the government has allocated land to sugar workers whose earnings have been severely affected by the fall in the world price of sugar. Malnutrition was widespread even before the current crisis, and the resilience of children to further nutritional shortfalls was low. In 1985 the government, with assistance from the United Nations Children's Fund (UNICEF) and Oxfam, implemented a nutrition program that pro- vides a high-protein meal each day to some 85,000 children. Food for work programs (with assistance from the ILO) are also under way. Food programs that have been implemented during droughts can be 40 THE ALLEVIATION OF POVERTY transferred to situations of more general austerity. For example, in Zimbabwe a National Child Supplementary Feeding Program was initiated by nongovernmental organizations in 1980 when drought affected Southern Africa. The program was later taken over by the Ministry of Health, and its coverage expanded from 70,000 to 270,000 children by 1983 (UNICEF 1985, p. 34). It provided supplementary food (equivalent to about half the daily energy requirement) to malnourished children under the age of five. Although this program was implemented to deal with the effects of drought, its later phase also coincided with the beginning of an adjustment effort launched in 1983. Finally, the Bank's Tamil Nadu (India) nutrition program also holds important lessons (Berg 1985, p. 33; Berg 1987). Although the project was launched to deal with pervasive malnutrition, it also helped to offset the nutritional effects of drought and economic difficulties in the 1980s. In neighboring areas that were not covered by the program, malnutrition increased from 16 to 30 percent. Programs that respond to environmen- tal shocks (as well as more general situations of malnutrition) demon- strate potential ways to deal with the welfare effects of shocks emanat- ing from the international economy. Some Conclusions on Targeting Transfers Transfers can protect the poor under adjustment, but ideally they should be a supplementary component, rather than the main element, of policies to alleviate poverty. Increasing primary income entitlements remains the best policy, but malnourished people are poorly equipped to take up the new opportunities afforded by structural adjustment (World Bank 1986b, p. 6). Targeted transfer programs have a particularly valuable role to play in offsetting the fall in poor people's nutritional status that often accompanies the removal of food subsidies. Many developing countries have now been forced to cut back on marketwide subsidies, because of both their negative effects on producer prices (and thus overall food security) and their escalating budgetary costs. Moreover, such subsidies often benefit the better-off groups dispro- portionately or are a relatively ineffective way of reaching the most needy. For example, in the case of Morocco a Bank mission in 1986 calculated that only 16 percent of the total food subsidy benefited the bottom 30 percent of the expenditure group. Similar examples can be found in many other countries (Berg 1981, p. 41; Berg 1987). In this UNDER STRUCTURAL ADJUSTMENT 41 regard the Bank has advised governments on how to maintain and improve nutritional levels while phasing out marketwide subsidies. In Morocco the Bank has recommended a series of direct food assistance programs, the marketing of new low-cost foods with high nutritional content, and cost-reduction measures in food production and distribu- tion. Such compensatory programs can do much to reduce political opposition to ending subsidies and raising food prices (Mateus 1983, p. 45). Although targeted food programs have lower budgetary costs than marketwide food subsidies (for the same level of nutritional support), they have higher delivery costs than subsidies. The administrative difficulties in implementing targeted programs are well known, and improving institutional ways to identify vulnerable families and deliver targeted foods is essential (Baum and Tolbert 1985, p. 222). In this regard, national and international nongovernmental organizations can perform a valuable function, given their expertise in reaching vulnerable groups. Donors can also be mobilized to provide food aid as nutritional support during the adjustment process, as has been done in the case of Jamaica's food plan. Such assistance has generally been given in response to environmental shocks, but it can also play a useful role in adjusting to macroeconomic difficulties. Cooperation between the World Food Program and the World Bank in this area has already been undertaken. But the pitfalls of food aid are well known, and such assistance must be handled carefully in order to avoid disincentives to the recovery of domestic food production. Adjustment often entails moves to raise food prices in order to restore producer incentives. But, substantial food aid may-by increasing food supply and thus reducing food prices-undermine the recovery of domestic food production. The general principle that food aid should meet transitory nutritional shortfalls, rather than provide long-term assistance, thus gains special relevance under adjustment (Stewart 1986, p. 83). Consequently, food aid needs to be well targeted and ideally should complement measures designed to raise primary incomes (for instance, through food for work). With the implementation of adjust- ment policies and the consequent expansion of food output, it may be strongly advisable to switch from food aid to financial aid for purchases of local foods, which can then be transferred to the poor. Such financial 42 THE ALLEVIATION OF POVERTY assistance has a double benefit-it increases food consumption security and promotes food supply security. Although this section has focused on food transfers, the possibility of cash transfers under adjustment must be raised. Cash transfers can be more cost-effective than food transfers (for example, in reaching remote rural people) and, by raising the demand for food, can positively affect producer incentives. However, cash transfers are most difficult to finance in times of fiscal austerity (unless they offer savings compared with subsidies), although they can be provided by external donors. More fundamentally, it is difficult to ensure that cash transfers are spent on meeting basic needs or that they reach the most vulnerable family members. In this regard food stamps represent a useful compromise between food and cash transfers in reaching some (but not all) groups. One use of cash transfers has been to maintain poor families on their land in the face of drought, as UNICEF's scheme in Ethiopia has done (UNICEF 1985, p. 32). In the context of adjustment such cash transfers (or indeed food transfers) could be used to encourage the poor to move for example, providing transfers to urban families who take up farming, the transfer giving support while they become established. Transfers therefore have a role in encouraging the kinds of labor market mobility discussed earlier. In the final analysis the choice of transfer method will depend on country specifics. Policy packages applied within Sub-Saharan Africa will inevitably differ from those applicable to Asia or Latin America (World Bank 1986b, p. 40). Concluding Observations This study has identified a number of examples in which concerns about the alleviation of poverty have been given some practical expression in adjustment programs. Our intention has been to show what can be done. In many of these cases government concern over this issue has received the support and finance of the World Bank-through its structural adjustment lending (in the case of land reform in Thailand) and other program or project lending (in the cases of Bolivia, Brazil, and Indonesia). In other instances the Bank has encouraged measures to cushion the impact of adjustment on the poor in its economic analysis and poverty dialogue (for example, in the cases of Jamaica's nutrition programs and Chile's emergency employment schemes). UNDER STRUCTURAL ADJUSTMENT 43 Because most of these examples are of very recent origin, it is difficult to assess how effective they have been in helping the poor. More information is needed on how the different interventions and the adjustment process itself have affected different groups of the poor. Nevertheless, some guiding principles emerge from our review. First, one striking feature of the illustrations compiled here is that they are invariably orderly or planned structural adjustment programs that have often been implemented with Bank assistance (Bolivia, C6te d'Ivoire, The Gambia, Indonesia, Jamaica, and Thailand). A point that needs underscoring is that some form of adjustment is inevitable, orderly or not. Disorderly adjustment, with little preparation and planning, is usually unable to take account of the problems of vulnerable groups. Furthermore, the problems of the poor are likely to be exacerbated if measures to deal with the underlying causes of economic imbalance are postponed or if too much reliance is placed on further extension of trade and price controls. Second, income distributions will inevitably change during adjust- ment. These changes are the principal incentives for resource realloca- tion, and without them the objectives of adjustment will not be realized. Interventions to protect the poor should guard against removing these incentives. The challenge for policymakers is to increase the participa- tion of the poor in activities that are expanding under adjustment. Conversely, if the poor are in activities that are contracting, interven- tions should help them move to more profitable sectors. The examples of employment interventions in Bolivia and Chile offer an interesting contrast on this point. In the latter case, most of the emergency employment schemes do not encourage the movement of the unem- ployed into expanding activities-rather the reverse. Bolivia's social fund offers the potential for raising labor mobility in accordance with the objectives of the overall adjustment effort. Third, the study highlights the advantages of interventions that increase the primary income claims of the poor. Such an approach minimizes the potential conflict between poverty alleviation and adjust- ment and maximizes the potential benefit of transfer payments by enabling them to be devoted to the most vulnerable of the poor. A question that remains, however, is how rapidly primary incomes can be improved in helping the poor "across the desert." Finally, adjustment can act as a catalyst for policymakers to examine 44 THE ALLEVIATION OF POVERTY carefully the costs and benefits of their programs for the poor. For example, the inclusion of land reform in the Thai SALs has prompted a continuing policy debate on the possibilities of further action. A number of govemnments under the disciplines of adjustment are now restructur- ing their social expenditures to raise the benefits to the poor. These changes, desirable in themselves, have in many cases been promoted by adjustment efforts. We have concentrated on the technical feasibility of adjustment with equity, but the issues of political feasibility are as important, if not more so. If development has been characterized by pervasive interventionism, then the political problems of implementing equitable adjustment are likely to be severe. In such cases the income losses may be rapid to the minority who previously benefited from market distortions, while the diffusion of gains may be slow to the rest of the population. Conse- quently, the minority are likely to marshall political support to reverse the reform process before the larger constituency is aware of its potential gains. In order to mobilize political support for reform, there may be some merit in giving priority to measures that clearly assist the less well- off majority relatively quickly. Measures intended to protect vulnerable groups may also add to the ethical legitimacy and political acceptability of adjustment programs. But such political considerations are beyond the scope of this report. Our argument is that the potential technical conflict between adjustment and poverty need not be as severe as some suppose. Bibliography Addison, Tony, and Lionel Demery. 1985. Macro-Economic Stabilisation, Income Distribution and Poverty; A Preliminary Survey. ODI Working Paper 15. London: Overseas Development Institute. Addison, Tony, and Lionel Demery. 1986. The Impact of Liberalisafion on Growth and Equity. ILO World Employment Programme Working Paper 2-46/WP4. Geneva: International Labour Office. Basta, S. S., and Anthony Churchill. 1974. Iron Deficiency Anemia and Productivity of Adult Males in Indonesia. World Bank Staff Working Paper 175. Washington, D.C. UNDER STRUCTURAL ADJUSTMENT 45 Baum, Warren C., and Stokes M. Tolbert. 1985. Investing in Development: Lessons of World Bank Experience. Oxford: Oxford University Press. Berg, Alan. 1981. Malnourished People: A Policy View. Washington, D.C.: World Bank. Berg, Alan. 1985. "Improving Nutrition: The Bank's Experience." Finance and Development 22, no.2 (June): 32-35. Berg, Alan. 1987. Malnutrition: What Can Be Done? Baltimore, Md.: Johns Hopkins University Press. Bertrand, Trent, and Lyn Squire. 1980. "The Relevance of the Dual Economy Model: A Case Study of Thailand." Oxford Economic Papers 23, no.3 (November). Bond, M. E. 1983. "Agricultural Response to Prices in Sub-Saharan African Countries." IMF Staff Papers 30, no. 4 (December): 703-26. Buiter, W. H. 1986. Macroeconomic Responses by Developing Countries to Changes in External Economic Conditions. Centre for Economic Policy Research Discussion Paper 93. London. Chenery, Hollis, and others. 1974. Redistribution with Growth. Oxford: Oxford University Press. Demery, Lionel, and Tony Addison. 1987. "Food Insecurity and Adjust- ment Policies in Sub-Saharan Africa: A Review of the Evidence." Development Policy Review 5, no.2 (June): 177-98. Economic Commission for Latin America and the Caribbean. 1986. "The Economic Crisis: Policies for Adjustment: Stabilization and Growth." Report prepared for the Twenty-First Session, Mexico City, April. Feder, Gershon, and others. 1986. "Land Ownership Security, Farm Productivity, and Land Policies in Rural Thailand." World Bank, Wash- ington, D.C. This report will be published in 1987 as Feder and others, Land Policies and Farm Productivity in Thailand. Baltimore, Md.: Johns Hopkins University Press. Food and Agriculture Organization. 1985. Agricultural Price Policies. Pro- ceedings of the FAO Conference, Twenty-Third Session, Rome, November. Ghai, Dharam, and Lawrence Smith. 1983. Food Policy and Equity in Sub- 46 THE ALLEVIATION OF POVERTY Saharan Africa. Working Paper WEP 10-6/WP 55. Geneva: International Labour Office, World Employment Programme. Griffen, Keith, and A. R. Khan, eds. 1977. Poverty and Landlessness in Rural Asia. Geneva: International Labour Office. House, W. J., and Tony Killick. 1981. "Inequality and Poverty in the Rural Economy, and the Influence of Some Aspects of Policy." In Tony Killick, ed. Papers on the Kenyan Economy, pp. 157-79. Nairobi: Heinemann. Jamal, Vali. 1985. Structural Adjustment and Food Insecurity in Uganda. Working Paper WEP 10-6/WP 73. Geneva: International Labour Office, Rural Employment Policy Research Programme. Killick, Tony. 1986. "Balance of Payments Adjustment and Developing Countries: Some Oitstanding Issues." In M. Posner, ed. Problems of International Money, 1972-85, pp. 64-90. Washington D.C.: International Monetary Fund and Overseas Development Institute. Lipton, Michael. 1985. Land Assets and Rural Poverty. World Bank Staff Working Paper 744. Washington, D.C. Mateus, Abel. 1983. Targeting Food Subsidies for the Needy: The Use of Cost-Benefit Analysis and Institutional Design. World Bank Staff Working Paper 617. Washington, D.C. Pinstrup-Anderson, Per. 1985. "Food Prices and the Poor in Developing Countries." European Review of Agricultural Economics 12, nos. 1-2: 69-85. Stewart, Frances, ed. 1983. Work, Income and Inequality: Payment Systems in the Third World. London: Macmillan. Stewart, Frances. 1986. "Food Aid: Pitfalls and Potential." In United Nations Children's Fund and World Food Program. Food Aid and the Well- Being of Children in the Developing World, pp. 67-84. New York. Trairatvorakul, P. 1984. The Effects on Income Distribution and Nutrition of Alternative Rice Price Policies In Thailand. Research Report 46. Washington D.C.: International Food Policy Research Institute. United Nations Childrens's Fund (UNICEF). 1985. Within Human Reach: A Future for Africa's Children. New York. UNDER STRUCTURAL ADJUSTMENT 47 World Bank. 1981. Accelerated Development in Sub-Saharan Africa: An Agenda for Action. Washington, D.C. World Bank. 1984a. Toward Sustained Development in Sub-Saharan Africa: A Joint Program of Action. Washington, D.C. World Bank. 1984b. Thailand: Pricing and Marketing Policy for Intensifica- tion of Rice Agriculture. Washington, D.C. World Bank. 1986a. Financing Adjustment with Growth in Sub-Saharan Africa, 1986-90. Washington, D.C. World Bank. 1986b. Poverty and Hunger Issues and Options for Food Security in Developing Countries. Washington, D.C. World Bank. 1986c. Poverty in Latin America, The Impact of Depression. Washington, D.C. World Bank. 1987. Financing Health Services in Developing Countries: An Agenda for Reform. Washington, D.C. The World Bank Headquarters 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. Telephone: (202) 477-1234 Telex: RCA 248423 WORLDBK WUI 64145 WORLDBANK Cable address: INTBAFRAD WSHINGTON DC European Office 66, avenue d'Itna 75116 Paris, ftance Telephone (1) 47.23.54.21 Telex: 842-620628 Tokyo Office Kokusal Building 1-1, Marunouchi 3-chome Chiyoda-ku, Tokyo 100, Japan Telephone: (03) 214-5001 Telex: 781-26838 ISBM 0-8213-0956-0