Yukon Huang and Peter Nicholas
The economic turbulence of the late 1970s and early 1980s has taken its toll in the developing world. Most developing countries have lost a decade or more of economic progress. Many have suffered a severe deterioration in social conditions, with increased malnutrition and urban unemployment, growing numbers of absolute poor (living below minimum requirements of food and shelter), and a slowing or even a reversal of improvements in health. The suffering has been particularly evident in two groups of countries. In low-income Africa severe drought exacerbated the effects of worsening terms of trade, reduced capital inflows, and domestic policies. GDP growth almost came to a standstill by 1983. Although growth has since resumed, per capita incomes have continued to fall in the face of rising populations. In the highly indebted, middle-income countries, a decline in the flow of commercial lending and reduced access to the markets of industrialized countries have compounded the problems emerging out of past policy mistakes and the international recession. Incomes in many countries in the latter group have declined precipitously.
Reduced economic growth and weakened financial positions have adversely affected social programs for the poor. Real per capita spending on health and education is, in many cases, a quarter to one half of what it was a decade earlier. In many African countries the prices of basic foodstuffs have gone up as drought and inappropriate past policies have constrained supply even though governments have liberalized prices and reduced subsidies. In highly indebted countries, unemployment has increased as firms face material shortages, declining demand, and financial difficulties. Devaluations and wage restraints have pushed down real wages for organized labor. Meanwhile, inflation has resurged, particularly in Latin America, fed by deficit financing.
Against this background, questions have been asked about the impact of adjustment programs on the poor and about the ability of governments and international agencies to protect the poor in the process of adjustment.
The social costs
What are the social costs of adjustment? Three types of transitional costs can be identified. First, adjustment measures designed to balance aggregate demand and supply usually, though not inevitably, depress output, employment, and consumption. These costs are typical of a recession. Second, the changes in the structure of incentives stimulate the reallocation of resources and hence benefits between sectors and activities. Businessmen and employees who previously benefited from subsidies and other forms of protection from market forces are likely to suffer substantial declines in income and wealth while those in stimulated activities should benefit. Third, lags and difficulties in moving productive resources into alternative uses in response to changes in relative prices may add to costs initially.
These transitional costs are largely unavoidable but country experiences suggest that the transitional costs of an orderly adjustment are likely to be smaller and the longer run benefits larger than that of an ad hoc, disorderly process. The costs associated with well planned adjustment programs are outweighed by the long-term benefits of the more rapid and viable growth that results.
Any assessment of the social costs of adjustment programs must evaluate the effects on the welfare of the poor—the bottom 30-40 percent of the per capita income and consumption ladder in a population. This group includes poor farmers, unskilled or low paid laborers, and women and children in low-income households. It excludes traders, mid-level government employees, or relatively well-paid workers in hitherto protected industries. Such an assessment must concentrate on the immediately measurable effects on the poor of adjustment policies.
On the basis of Bank experience with such programs, a number of conclusions can be drawn about the consequences of policy measures in adjustment programs.
•Theimpactofaspecificadjustmentpolicyonthepoordependsonacountry’seconomicstructure. Such policy changes as devaluations and reforms of the tariff regime affect low-income groups differently, depending on a country’s economic and social structure, and in some circumstances, on how the government designs and implements these measures. The effects of a devaluation on the poor, for example, depend on the labor intensity of production in the tradable and nontradable sectors, on the importance of the affected products in the consumption basket of the poor, and on the lags in relocating workers and raw materials. If small-scale farmers (for example, of jute in Bangladesh) produce most of the exports then devaluation will benefit them. But if the export goods are primarily staples that the poor consume (for example, rice in Thailand), devaluation will hurt the poor (see box).
•Thetransitionalcostsofadjustmentmayhelessinlow-incomeAfricathaninthehighlyindebted, middle-incomecountries, particularlythoseinLatinAmerica. In Africa the poor are usually concentrated in the rural areas. Those employed in government, industry, and urban services are usually in the higher income groups. In this setting, a devaluation coupled with trade liberalization and unshackling of agricultural markets should improve the average income of the poor and increase efficiency. The losers are likely to be higher income (mainly urban) groups who had previously benefited from priority access to foreign exchange and protected markets. In Latin America, such adjustment measures are likely to impose greater social costs because land ownership is more concentrated and because many of the poor are low-wage urban workers and thus more vulnerable to higher food prices.
•Manyadjustmentpoliciespromotebothefficiencyandequity, particularlyinthelongrun. This is often the case, for example, with reforms of social services. By shifting priorities and imposing selective user charges, governments can improve the efficiency of social programs while continuing to help the poor. Thus in the health sector, shifting priorities in favor of preventive rather than curative services and toward expansion of rural rather than urban facilities can increase social rates of return as well as promote equity (see “Financing of Health Services in LDCs” by John Akin and Nancy Birdsall in this issue). User charges for curative care can help to expand such programs and to subsidize preventive services. In education, charges can be increased for higher education and the extra resources used to expand primary education to benefit the poor. Education loans and scholarships based on income levels can maintain access to higher education for the truly needy.
•Somepolicyreformsmayaffectthepoorestgroupsonlymarginally. Many of the rural poor in Africa and South Asia are largely untouched by the tax system. Their cash incomes fall below exemption levels for income taxes, and they consume mainly domestic goods that are generally not taxed. The more important consequences of tax reforms for the poor probably are efficient production and increases in employment opportunities over the long term. Liberalization of financial markets also may not directly affect the poor.
•Anearlyandorderlyadjustmentproducesfewersocialcoststhanadelayedandhaphazardresponse. The experience of the Republic of Korea provides a good example. Following the external shocks of the late 1970s, economic growth turned negative, inflation shot up, and the current account moved from surplus to a substantial deficit. Despite this, the country’s prompt and well-planned adjustment—which produced growth averaging 8 percent in 1981-85—enabled the authorities to preserve and even increase their social achievements. In the health sector, for example, subsidized medical programs for low-income groups have grown since their initiation in 1980 to cover nearly 10 percent of the population, and health insurance has increased four-fold in six years to cover 40 percent of the population. In the short run, strategies which concentrate almost exclusively on helping the poor but neglect adjustment might reduce absolute poverty even more quickly. But if such strategies lead to stagnant growth as evidenced in Tanzania (before 1986) and Ghana (before 1983), poverty-oriented programs soon become vulnerable to fiscal pressures.
•Simplyrevivinggrowth, thoughnecessary, isnotenoughtoprotectthepoorduringanadjustment. Both Brazil and Malawi, for example, were unable to prevent some deterioration in social conditions for the poor even with a degree of success in restoring growth. It is difficult, however, to know whether this deterioration was due to the earlier recession or the ineffectiveness or absence of social programs to protect the poor during the subsequent adjustment. Countries which do not have a development strategy that is already oriented toward the poor may not have the institutions or processes available to protect the poor during adjustment.
•Indesigningadjustmentpackages, thereareoftentrade-offsbetweenshort-andlong-termeconomicobjectivesandbetweenvarioussocialobjectives. First, and most significantly, adjustment programs may entail some reduction in short-run growth as a price for achieving higher growth later. One major study for semi-industrialized countries suggests that the growth rate can be increased by two to three percent when the trade bias and payments regime are successfully altered. The short-term costs, according to the same study, are probably a decrease in the growth rate of one to two percent for a year or eighteen months, and are thus greatly exceeded by the later benefits. Second, various combinations of policies can, in principle, achieve a given balance of payments objective, each with a different trade-off between inflation control and short-run unemployment. Many Latin American countries, for example, have borrowed excessively so they can postpone adjustment, and have often chosen policies that minimized the immediate adverse employment consequences of postponing adjustment at the expense of inflation and future unemployment. Finally, the manner in which a policy is implemented can involve trade-offs between efficiency and welfare. Sudden and substantial price increases or exchange-rate adjustments, for example, may help producers and lend more credibility to the reform process, but may create transitional costs for poor consumers.
•Amelioratingsocialcostsrequiresadequateexternalfinancing. A growth-oriented adjustment program cannot be implemented without adequate external resources. Under current conditions, this implies a reversal of recent declines in resource flows. Increased resource flows would reduce the need to slow down the economy while encouraging growth-oriented policies which typically take more time to have an impact. While a few developing countries have continued to receive high levels of external support, overall new lending has fallen sharply. During 1981-84, net public capital flows to Sub-Saharan Africa fell from $7 billion to less than $3 billion. During the same period, public and private net capital flows to the highly indebted, middle-income countries plummeted from $43 billion to $14 billion.
The Bank’s role
The Bank’s early support for economic adjustment focused almost exclusively on improving the efficiency of resource allocation and strengthening institutions—with a view to restoring growth and improving the balance of payments in adjusting economies. This support reflected the judgment that the quick restoration of growth was the most effective way to help the poor. More recently, with the experience of earlier adjustment efforts and as sustained growth has proved more elusive than first expected, the Bank has broadened its efforts to help adjusting countries cope directly with the social effects of adjustment. These efforts are in addition to and distinct from the Bank’s continuing efforts to address the long-term problem of poverty through investments which increase the assets of the poor and the rate of return on those assets. Such poverty-oriented lending by the Bank goes for health, nutrition, population, primary education, urban development, small-scale enterprises, water and sewerage, and many agricultural projects; it comprises a substantial segment of the Bank’s lending other than for adjustment programs.
The Bank is strengthening its efforts to help governments ameliorate the adverse effects of adjustment by: (1) improving the efficiency and poverty orientation of social expenditures, particularly in health and education; and (2) developing cost-effective compensatory programs in the areas of nutrition and employment in coordination with other agencies and nongovernmental organizations.
Through reviews of public expenditure and in its lending operations, the Bank is assisting governments to restructure social expenditures in favor of the poor. Public expenditure reviews for Brazil, Ghana, Nigeria, and Zambia have recommended redirecting health and education programs towards low-income groups. Where appropriate, the Bank is following up these reviews with support for reform in both the health and education sectors through its sector and adjustment lending.
The Bank is also increasing support for programs to reduce the social costs of adjustment. As discussed in a number of recent articles in Finance & Development (those in this issue, and, “Improving nutrition: the Bank’s experience” by Alan Berg, June 1985, and “Food security and poverty” by Shlomo Reutlinger, December 1985), a growing body of experience in the Bank and in other agencies shows that properly targeted compensatory programs—to improve nutrition and reduce unemployment by facilitating the redeployment of labor—are feasible in widely different situations. Well-targeted programs of nutrition and supplementary feeding can be an important way of reducing social costs. So can food-for-work schemes and, in some cases, cash employment schemes of the kind used in Chile. In considering compensatory schemes, governments need to consider their costs, both budgetary and economic. In times of fiscal austerity, the budgetary costs may involve restructuring rather than expanding programs, to reach the poor. The economic cost includes possible efficiency losses resulting from distortions introduced by compensatory programs; for example subsidy or income supplement schemes may discourage rather than promote a shift of resources into new activities.
The Bank is strengthening its cooperation with other international agencies, such as the World Food Programme and the US Agency for International Development, on using food aid to reduce adjustment costs for the urban and rural poor, and with agencies such as the International Labour Organization on employment programs. Increased coordination with Unicef is helping some governments expand low-cost programs of mother-and-child health care. The Bank is also increasing its cooperation with nongovernmental organizations, who often have the flexibility, local knowledge, and staff commitment to administer social programs more efficiently than many official agencies.
Impact of higher crop and fertilizer prices
Many governments, although committed to supporting rapid growth of agriculture, have paradoxically established a complex set of policies, notably low crop procurement prices, which have a strong anti-agricultural bias. Simultaneously, governments often subsidize food, ostensibly to help the poor. But low procurement prices end up reducing the incomes of farmers, who in many cases are much poorer than those who benefit most from the subsidies.
What happens to the poor when procurement prices are raised to stimulate agricultural production? The immediate impact on the poor is ambiguous. Various poor groups in society—farmers and laborers in the rural sector and the urban poor—are affected differently. Further, the impact of a rise in the prices of nonfood crops differs from that of an increase in food crop prices. The income gains to poor producers of higher food crop prices are counterbalanced by the losses to poor consumers who have to pay more for their food. Surplus farmers clearly benefit while deficit food producers suffer; a steep increase in food prices may even force the latter group to sell off assets, including land. Subsistence farmers are largely unaffected, because they neither buy nor sell significant amounts of food at the new prices. By contrast, nonfood cash crops are seldom purchased by the very poor, so the welfare effects of a price increase for these crops are confined to income gains to producers.
The higher purchase prices for food which normally result from raising producer prices inevitably hurt poor landless laborers and the urban poor, at least in the short run. The long-run impact, however, is less clear. For marginal laborers, employment opportunities may eventually increase as agriculture expands. This increase may be accentuated—or tempered—by such factors as the labor intensity of different crops and the choice of agricultural inputs. For the urban poor, long-run benefits would include greater food availability and (if agricultural prosperity reduces rural-urban migration) a brake on the decline in real incomes in the urban informal sector. The net effect on poverty would obviously vary according to the distribution of land, the composition of the labor force, and other structural economic factors.
Fertilizer subsidies are among the most widespread although most of the theoretical arguments support such input subsidies as a temporary measure, and even then only in special cases. What then are the effects on the poor of reducing these subsidies? Because fertilizer provides higher incremental yields on irrigated land, the generally better-off farmers with irrigated land tend to benefit most from fertilizer subsidies. The poor, on the other hand, lose from an emphasis on fertilizer-intensive crops; the demand for labor falls as fertilizers are substituted for labor, reducing incomes of the landless. Meanwhile, relatively less expensive staple crops such as coarse cereals (which are not fertilizer-intensive) are discouraged. As the availability of such crops decreases, the cereal consumption of the rural poor falls even more.