Abstract
Experience from the Bank’s lending for structural reforms
Since the beginning of the 1980s, economic difficulties have led many developing countries to implement adjustment programs to correct policy distortions and structural imbalances. In the early years, little attention was given to the possible adverse impact of adjustment programs on the poor. This was to a large extent justified by the expectation that adjustment programs would facilitate growth, and in the process reduce poverty. With the passing of time, however, it became evident that the economic problems were more severe than expected, that adjustment was slower than anticipated, and that many countries, particularly in Sub-Saharan Africa and Latin America, were experiencing prolonged periods of economic stagnation and deteriorating social conditions for some groups.
This led the international development community to renew its attention to poverty reduction and emphasize the need for policies and programs to improve incomes and standards of living of the poor during periods of limited growth. This focus underscored the importance of protecting the welfare of those who may be hurt by adjustment programs and of the chronically poor (not necessarily the same group) for the success, and the political viability, of the adjustment process itself.
The Bank’s efforts
The nature and extent of the impact of adjustment programs on the poor has been hard to establish empirically. This is largely because of the difficulty in distinguishing the direct costs of adjustment from the costs of unfavorable exogenous shocks and inappropriate past policies.
Recent studies carried out by the World Bank (see box) have not confirmed that adjustment programs are systematically associated with worsening poverty and social indicators. One study, for example, concludes that available evidence shows that changes in living conditions of the general population in the short run do not appear to be related to adjustment lending. But the analysis is based on aggregate data and, therefore, has not established the extent to which particular groups may have been adversely affected.
It is possible that some adjustment measures have an adverse short-run impact on specific groups of the poor. The adverse impact may result from reductions in public expenditures, increases in the prices of goods and services consumed by the poor, and declines in employment or real wages in sectors in which they work. Nevertheless, a clear lesson from experience is that not adjusting is likely to reduce living standards and that an orderly adjustment process designed to promote growth and efficient resource allocation is indispensable for improving the longer-term position of the poor.
The Bank, in concert with other multilateral agencies, such as the IMF, and bilateral development agencies, has assisted countries to adapt to changing circumstances, increase economic opportunities for the poor, and mitigate any adverse short-run impacts of adjustment programs on the poor. Major interventions have included targeted projects aimed at specific groups (such as employment, nutrition, credit, training, resettlement, and severance pay schemes) and social sector reforms to increase the efficiency and access of the poor to basic services. Labor market reforms and targeted subsidies have also been supported in some cases. Relatively less attention has been given, however, to analyzing the mix, sequencing, and timing of macroeconomic policies that promote a pattern of growth that is most conducive to long-run poverty reduction, as well as to the overall composition, incidence, and effectiveness of public expenditures and their implications for poverty reduction.
This article is based on Ribe et al, “How Adjustment Programs Can Help the Poor,” Discussion Paper No. 71, The World Bank, 1990; World Bank, “Adjustment Lending: An Evaluation of 10 Years of Experience,” Policy Research Series No. 1, 1988; and World Bank, “Report on Adjustment Lending II: Policies for the Recovery of Growth,” forthcoming; and the World Development Report, 1990. It also draws on several internal sources and from work for the Social Dimensions of Adjustment project undertaken by UNDP and the World Bank.
Social sector expenditure reforms. A number of Bank-supported structural adjustment programs have attempted to protect the poor from short-term costs and enhance the beneficial long-term social impact of the adjustment programs by:
• maintaining or increasing expenditures on basic health and education, while overall public expenditures are cut. In Mauritania, for example, budget allocations to these sectors were given priority.
• redirecting expenditures toward the poor mainly through better targeting of resources toward primary education and basic health. For example, in Niger, the share of primary education was increased, although the total education expenditure was reduced and in Guinea, public resources were being redirected to rural health care and primary education.
• improving the efficiency of social sector expenditures. For example, in Sao Tome & Principe, the efficiency of the delivery systems in health and education sectors were to be improved.
Chile and Indonesia provide good examples of social expenditure reforms in times of economic adjustment. In Chile, health and nutrition interventions contributed to a decline in infant and child mortality rates despite a 5 percent fall in public expenditures during the period of recession and adjustment. In Indonesia, overall development expenditures were reduced during 1982-87, but the share of expenditures on human resources increased (from 16 to 24 percent).
In some cases, comprehensive sectoral reforms have been initiated under Bank-supported sectoral loans, such as the ones in Argentina, Mali, Morocco, and Senegal. Typically, these programs are intended to improve services to the disadvantaged groups (e.g., improving basic schooling in rural areas); reduce unit costs (e.g., introducing double school shifts); build institutional capacity (e.g., training personnel); and improve quality (e.g., strengthening school curricula). Under the Primary Education Development Credit in Senegal, the sectoral adjustment component included specific actions to improve service quality and teacher performance.
Labor market reforms. Labor market regulations (such as minimum wages, job security, and social security regulations) are intended to raise worker welfare and security. But instead they tend to raise labor costs and reduce employment in the formal sector. Relatively little attention has been given in adjustment programs to remove such rigidities. Burundi and Senegal are among the few countries where attempts have been made at increasing labor market flexibility by allowing temporary labor contracts and eliminating requirements for government authorization to dismiss employees.
Targeting subsidies. Marketwide subsidies, which reduce the retail price of a good sold through regular market channels, have been a common and effective means of raising the real incomes of the poor. But such subsidies can be inefficient because they usually cover commodities whose consumption increases with income, hence primarily benefiting the better-off segments of society. Many reform programs have included a reduction in such general subsidies as part of an effort to reduce deficits. Quite naturally, some subsidy reductions have caused considerable hardship for the poor—particularly the urban poor.
Subsidies aimed at specific groups, however, by reducing the price of a good for individuals based on some indicator of need, can be both more effective in reaching the poor and less expensive for the government. Political and administrative constraints, however, have made such targeting difficult in practice and only a few countries have been able to introduce such schemes successfully. Nevertheless, experience has shown that targeting can be facilitated by selecting low-income neighborhoods (as in Mexico and the Philippines), readily identifiable groups (such as malnourished children and mothers), and commodities that are consumed mainly by the poor (such as cassava or certain grades of wheat and flour).
Programs aimed at specific groups affected by adjustment measures represent an attractive option for addressing the possible adverse social impact of macroeconomic reforms.
In a few countries, governments undertaking adjustment programs have improved the implementation of such schemes to protect the poor. For example, in Tunisia, subsidies on commodities that constitute a small proportion of expenditures by the poor (milk, sugar, and vegetable oil for soap-making) were reduced earlier than subsidies on wheat and cooking oil, which are widely used by the poor. Marketwide subsidies in Mexico have been replaced by targeted milk distribution, subsidies on commodities consumed mainly by the poor in selected geographic areas, and school feeding programs. Further, the removal of fertilizer subsidies was coordinated with changes in agricultural producer prices to protect poor farmers. Recently, in Venezuela, the Government has developed, with the help of the private sector, differentiated products to be marketed solely in poor areas at subsidized prices.
Implementing targeted projects
Programs aimed at specific groups affected by adjustment measures represent an attractive option for addressing the possible adverse social impact of macroeconomic reforms. This is, first, because they are limited (“targeted”) to a narrow and clearly deserving group of beneficiaries and second, because they attempt to be highly cost-effective and efficient.
Targeted projects can include short- or long-run interventions. There is a clear rationale for short-term targeted projects to compensate groups adversely affected during the transitional period. Long-run projects generally require more time and may best be addressed in the context of the Bank’s overall lending program to a particular country. However, since the adjustment period can prove to be quite long, short-run measures may sometimes be more successful if they initiate longer-term action. This is being attempted in Madagascar, Guinea-Bissau, and Chad. In Guinea-Bissau, for example, a drug distribution program is being carried out within the framework of a separate, larger program to reform the drug distribution system.
Since 1986, several targeted projects have been designed in countries undertaking adjustment. These have included mainly special employment programs, nutrition projects, redeployment programs (including training and credit projects), resettlement projects for displaced urban workers, and multisector programs. In a few cases, adjustment programs have supported population projects (Madagascar and Cameroon), housing projects (Guinea and Guinea-Bissau), and urban sanitation projects (Uganda, Guinea-Bissau, and Chad).
Special employment programs. These programs have been an effective way to reach the poor and provide a safety net for the unemployed in times of economic recession in Latin America and during the lean season in South Asia. While their main objective is to provide employment and basic security, employment programs can also help build and maintain local infrastructure.
Experience shows that these programs are generally more successful in reaching the poor when the wages offered are less than the market wage. In Chile, the government-financed Emergency Employment Program (PEM) offered one-fourth to one-half of the minimum wage and succeeded in ensuring that two-thirds of its beneficiaries were from the poorest 20 percent of the population. By contrast, when market wages were paid in Bolivia, only less than half of the workers employed by the Emergency Social Fund (ESF) came from the poorest 40 percent of Bolivian households. Another lesson of experience is that when governments lack the institutional capacity to organize such programs aimed at the poor, mobilizing local groups and working with nongovernmental organizations (NGOs) can help increase administrative efficiency and ensure that benefits reach the needy. In Ghana and Bolivia, for example, the Bank has helped expand employment programs carried out by local NGOs.
Nutrition projects. Some adjusting countries have supported programs to compensate losses in the purchasing power of poor households and protect groups that are likely to suffer most—the urban poor, women, and children. By 1988, the ESF in Bolivia had approved 14 nutrition projects, including school feeding, mothers’ clubs, and soup kitchens for adults and children. They benefited some 42,000 persons.
Targeting nutrition projects is difficult, but it can be facilitated by involving agencies that work at the grassroots level. This is being done in a number of countries, such as Chad, where the health component of the Social Development Action Program includes a nutrition program for women and children, which is being carried out by local social centers. Delivering nutrition supplements through the basic health systems can also be an effective mechanism, but requires a fairly advanced health sector. The distribution of milk and other food supplements by local health centers in Chile has resulted in major improvements in targeting and monitoring of maternal and child nutrition.
Redeployment program. The issue of unemployment stemming from fiscal retrenchment has been addressed rather prominently in Bank-supported adjustment programs. Projects have been designed to compensate dismissed workers and provide them severance payments, special credit, and retraining. For instance, compensation payments and bonuses were granted in Ghana, Guinea, Madagascar, and Togo. These programs are meant to protect the standards of living of retrenched workers during the period of transition while they find other jobs and sometimes to facilitate their entrance into new occupations or to self employment.
The use of severance payments may often be superior to special credit and retraining schemes (such as those designed in Ghana, Guinea, Mauritania, and Senegal), which have been generally more difficult to implement successfully. Subsidized credit schemes have generally failed and those that have succeeded rely on large organizational resources which are unlikely to work in the relatively short timeframe of adjustment programs. New training programs are also difficult to organize and should focus on improving existing training systems (as was done in Mexico) and on increasing the flexibility of the labor force, rather than on preparing displaced workers for specific new jobs.
Resettlement projects. Some adjusting countries have attempted to shift the newly unemployed urban workers toward agriculture. In Ghana, for instance, the resettlement project is intended to provide food aid and extension services to enable retrenched workers to move into unutilized but arable fields. In the Central African Republic, Ghana, and Guinea-Bissau, the public sector retrenchment was helped by external assistance specifically for this purpose.
Multisector programs. In some countries, targeted projects have been combined into multisector compensatory programs, including short-term interventions to provide relief (such as public works and feeding programs) and a range of developmental programs (ranging from urban sanitation to provision of textbooks). The most advanced and best-known Bank-supported programs are the ones in Bolivia, Ghana, and Madagascar. Others are being carried out or prepared in Benin, Burundi, Cameroon, Guinea-Bissau, Chad, Guinea, Senegal, Mauritania, Zaïre, and Uganda. They usually have multi-donor financing and combine public sector, private, and NGO implementation.
The Bolivian ESF was initiated in 1987 and had disbursed about $135 million by early 1990. It was originally designed as a three year program, with funds from a consortium of external donors, including the Bank, to assist those affected by the prolonged recession and the adjustment program that followed—primarily displaced miners and public sector employees. The ESF was managed by a small and select group independent from other government agencies and funded demand-driven infrastructure projects implemented by private subcontractors and local groups.
The ESF was effective in disbursing external resources quickly to small local groups and NGOs with low administrative costs. But it did not reach many of its original target groups. In practice, many miners found employment elsewhere and ESF projects employed mainly unskilled construction workers, who were likely to come from low-income families, but were neither the poorest groups nor the most adversely affected by the adjustment program. The program, now in its third phase, has become an integral part of the country’s long-term development program, aiming to reach the poorest groups in isolated regions.
The implementation of other multisector compensatory programs has been limited. Delays have been caused by the need for extensive coordination among several government agencies, NGOs, and external donors; by weak institutional arrangements; by difficulties in funding some of the sub-projects; and by political problems. In some cases, however, separate components, such as the public works scheme in Ghana, are being successfully implemented.
Some of the more recent reform packages, for example the Social Action Programs prepared by the Social Dimensions of Adjustment program in Sub-Saharan Africa, are more integrated with national long-term programs (see article by Michel Noel and Ismail Serageldin). It is still too early to assess the effectiveness of multisectoral programs and, clearly, a great deal of additional work and experimentation are needed. This is an area where many questions remain, particularly regarding the more effective institutional arrangements and the appropriate mix of short-term and longer-term developmental programs.