Cover Image
close this bookOvercoming Global Hunger (WB)
close this folderSession two - macroeconomic reform: its impact on poverty and hunger.
View the documentMacroeconomic reform: its impact on poverty and hunger
View the documentDiscussant remarks
View the documentWorkshop spokesperson remarks
View the documentWorkshop spokesperson remarks
View the documentFloor discussion
View the documentSpecial address: the scope for public action to reduce chronic hunger

Macroeconomic reform: its impact on poverty and hunger

Nancy Birdsall

My former colleagues in the World Bank have given me a tough topic and a challenging audience. I am delighted to have the opportunity to address a group among whom are many with on-the-ground experience in working in and managing programs to combat hunger. I see it as an opportunity to try to affect somewhat how you think about the problem we all face.

I want to convince you of the centrality of economy wide reforms for the reduction of poverty and the defeat of hunger. You and your constituents need to be involved in and engaged with efforts of the World Bank, my bank (the Inter-American Development Bank), and other multilateral institutions in encouraging and assisting governments in the difficult task of reforming their economies.

I start from the premise that chronic hunger is not a food production problem, but a poverty problem. A strategy to combat chronic hunger bools down to a strategy to combat poverty.

I will talk about three points. Let me state them briefly and then discuss each one.

· First, the importance of shared growth to the reduction of poverty. By shared growth I mean growth that benefits everyone, that lifts all boats, including the boats of the non elites. To reduce poverty, economic managers in most poor countries (and officials of the multilateral agencies that work with them, and members of the community of nongovernmental organizations) must worry not only about growth itself, but about the pattern of growth, and must aim for shared growth.

· Second, that structural adjustment can and has reduced poverty, including in the short run. Adjustment is not only necessary, but is at times a major contributor to reduced poverty and hunger

· Third, that in reducing poverty and eliminating hunger, transfers and other compensatory programs can help, but are not a substitute for adjustment and shared growth.

Background: Poverty and Hunger

The relationship between poverty and hunger is clear: poverty is the root cause of chronic hunger, and hunger probably contributes to poverty, for example, by reducing the amount of energy available for manual labor. As most of you know, South Asia and Sub-Saharan Africa, which have the highest percentage of poor, also have the highest percentage of hungry people The relationship over time also holds. In Latin America and Sub-Saharan Africa poverty increased between 1985 and 1990, and these regions also experienced a rise (or no change) in hunger. In East Asia and South Asia, where poverty decreased, so did hunger.

Poverty, in turn, and therefore hunger, are in part a function of country income per capita. Cross country statistical analysis indicates, not surprisingly, a close relationship between average per capita income and average per capita consumption of calories It is also true that no matter what the level of average income, growth of average income is associated with growth of average per capita consumption.

More interesting is that much of what is not explained in the statistical analysis is due to differences. among countries in the distribution of income for any given average income, countries with less equal income distributions have more poor people, who are more likely to be hungry.

Managing Shared Growth: Lessons from East Asia

This brings me to my first point: managing shared growth. Income distribution within a country is the result of initial conditions at some past time and economic and social policies since then Policies of shared growth over time can make a difference to income distribution, and through distribution to poverty and hunger.

Why doesn't growth in some countries lift all boats? Why did growth reduce poverty so much more in Indonesia than in Brazil? I believe we should be more cynical in thinking about this question. Let us star. from the premise that where the elites have political as well as economic power, in e absence of managed efforts government policies are likely to favor the elites. Often such policies contravene the market A prime example is import substitution policy. Import substitution policy helps industrialists and those who get privileged access to foreign exchange, and via overvalued exchange rates permits cheap imports of consumer goods for the urban middle class, but it reduces the local currency proceeds of the exports of rural farmers and reduces demand for the labor of poor urban workers Given the powerful combination of political and economic power among elite groups in many developing countries, including in the poorest, a trickle down laissez faire approach to growth is simply not likely to reach the non elites.

To offset this tendency, a policy stance we can label shared growth is critical. Active management to reach non-elites Noms key to the phenomenon of shared growth described ire recent work by the World Bank on East Asia 2 In Hong Kong, Indonesia, the postwar Republic of Korea, Malaysia, Singapore, Taiwan (China), and Thailand new regimes faced a formidable challenge from the communists, either externally, internally, or both The need to build legitimacy with the working class in both rural and urban areas was no doubt pressing To ensure the continuing support of the working class non-elites, leaders aggressively used a variety of mechanisms: strong support for public education at the primary and eventually at the secondary, level in all these countries; land reform in Korea and Taiwan (China); housing programs in Hong Kong and Singapore; massive investment in rural infrastructure in Indonesia; and credit and export guarantee programs for small and medium-size enterprises in Taiwan (China), and since the 1980s in Korea (where the shares of small and medium size enterprises in manufacturing employment and value added had risen to 51 and 35 percent, respectively, by 1988).

In addition, these countries avoided the false conflation of social policy with labor policy and worker entitlements. In Korea and Singapore, and to some extent in Malaysia, the government actively managed the labor movement to suppress radical political activity, and in Thailand and Indonesia labor unions were routinely suppressed out of fear of communism As a result of rapid growth and rising productivity, wages rose, generally in line with market forces, but not ahead of market forces The governments did not allow the mechanisms that elsewhere permitted a group of privileged workers in the formal sector labor force to reap huge rents through unionized pressure or interest group politics to be created. This suppression of labor rights had high costs, of course, in terms of loss of human rights The lesson is not necessarily that such rights should be suppressed, but that wages above market rates, need to be avoided if the benefits of growth are to be broadly shared.

Three aspects of shared growth in East Asia seem to have been key first, that there was active management (usually to supplement, catalyze, or strengthen market forces not to override the market); second, that the emphasis was not on direct transfers, but on improving opportunities (this was not a populist approach); and third, that the distinct) on was not between the poor and the rest, but between the elites (which elsewhere in the developing world include privileged public sector and public enterprise labor as well as industrialists and large landowners) and the rest In East Asia, the elites were also attended to, but not to the exclusion of the non-elites.

The programs noted above were not necessarily the core economic ingredients of shared growth, however. Three ingredients of the East Asian growth experience were fundamental to shared growth and the rapid reduction in poverty: a relatively level playing field for agriculture, an export push that created high demand for labor, and a universalist approach to public social and infrastructure investments.

In the fast-growing economies of East Asia direct and indirect taxation on agriculture has been lower than elsewhere in the developing world. In the last three decades, many governments in other regions favored manufacturing and hurt agriculture by overvaluing their currencies and protecting domestic industries that manufactured agricultural inputs and the goods rural households The exchange rate that resulted from restrictions on manufactured imports reduced the domestic currency proceeds of agricultural exports Industrial taxation acted as a hidden tax on agriculture, raising the price of agricultural inputs to subsidize industry Figure 1 compares the taxation of agriculture in selected East Asian countries Taxation has been higher for three decades in Pakistan, the Philippines, and Sri Lanka than in Korea (where agriculture is now protected) or Malaysia Thailand's level of agricultural taxation was relatively high in the 1960s and 1970s, but was reduced in the 1980s, while taxation in the Philippines rose.

Low taxation of agriculture has been associated with high growth. If we compare agricultural growth rates across regions for the period 1965 to 1988, we find that agricultural income and productivity have grown fastest m East Asia, one percentage point or more every year for twenty five years, than in other regions Increases in productivity contributed to and reinforced relatively slower rates of growth of employment in agriculture, as workers moved into even more rapidly expanding manufacturing and other employment. Other factors, of course, also contributed to success in agriculture in East Asia, but the policy of not penalizing the farmers was clearly key.

The second key ingredient of shared growth was East Asia's export push. Complemented by the rapid growth of agricultural output, the export push meant the demand for labor was high, the result not only of rapid growth of agriculture and manufacturing, but of relatively labor-intensive growth The export push strategy, including use of the exchange rate as a commercial tool and the need for global competitiveness, reinforced the tendency to rely heavily on labor, the most plentiful factor of production, and to minimize the antilabor bias associated with protecting capital-intensive industries. Equally important for shared growth, demand for unskilled labor has been relatively high, in Korea in the earlier decades, and more recently in Malaysia and now in Indonesia and Thailand. Combined with a rapidly increasing supply of skilled labor brought about by a rapid expansion of education, the result has been a relatively small wage gap between skilled and unskilled labor, which in itself contributed to a better distribution of income and a greater sense of shared growth.

Figure 1. Intervention and growth in the agricultural sector selected East Asian countries and decades.

Figure 1. Intervention and growth in the agricultural sector selected East Asian countries and decades.

Finally, East Asia is noteworthy for its universalist or saturation approach to public investment in infrastructure and in basic social programs. Table 1 shows the much Lower disparity in public investments in water and sanitation between rural and urban areas in East Asia compared to other regions (top panel) and the smaller differences in access to electricity (bottom panel). This lack of urban bias is an excellent proxy for the universalist emphasis in East Asia. In the case of electricity we can compare Thailand, which in 1984 had 78 percent of its urban population and 40 percent of its rural population benefiting from electricity with Brazil, which in 1981 had more than 95 percent of its urban population benefiting from electricity, but less than 20 percent of its rural population.

Also extraordinary has been the emphasis on education and health services. In the [ace of scarce resources, the universalist emphasis induced governments to concentrate their resources at the lower, more basic levels of services, where unit costs are lower and where the poor are more likely to benefit. For example, developing countries in general have a similar average percentage share of spending on education as East Asia, roughly 3.4 percent of GDP, but in East Asia most of that public spending has gone to primary and secondary education, with about 15 percent going to university education, compared to 24 percent in Latin America and South Asia, and at times even more in Africa.

Table 1. Comparison of rural and urban public investment, selected economies and years.

Rural urban disparities in access to services, 1987-90 (100 = rural-urban parity)




Korea Republic of









Other Asian countries



Latin America



Sub-Saharan Africa



Percentage of population served by electricity




Indonesia. 1984



Malaysia, 1983



Thailand, 1984



Other Asian countries

Bangladesh 1981



India 1981



Sri Lanka. 1982



Latin America

Argentina, 1987



Bolivia, 1981



Brazil. 1981



Fast and West Africa

Cd’Ivoire, 1981



Liberia. 1982



Senegal. 1982



Source: World Bank The cast Asian Miracle.

Korea and Venezuela provide an extreme example of the contrast. In 1985 public expenditure on education in Venezuela was higher as a percentage of GDP than in Korea: 4.3 percent compared to 3 percent. However, more than 40 percent of that public education spending went to higher education in Venezuela compared to 10 percent in Korea. As a result, more than two times as much in GDP terms was available for spending on primary and secondary education in Korea. This saturation approach at the lower level in Korea brought girls into school with all the benefits for children's health and education that are associated with mother's education.

Ironically, total enrollment rates at the tertiary level are higher in East Asia than elsewhere as the private sector has expanded to absorb the high demand generated by the much larger pools of graduates of good quality public secondary schools.

Structural Adjustment Can Reduce Poverty.

Let me go to my second point, about the benefits of structural adjustment for reducing poverty. For many countries a period of structural adjustment is a necessary first step on the path toward economic growth. People often hold this adjustment, associated with recession and cuts in public social sector spending, responsible for expanding and worsening conditions of poverty. However, we must ask whether or not the poor or would have benefited from less adjustment. The answer is almost certainly not. In Lima, Peru, the poor suffered disproportionately between 1985 and 1990, when President Garcia attempted to reactivate the economy by stimulating consumer demand, raising minimum wages, reducing taxes, and postponing external debt payments.

After an initial spurt of growth, the lack of adjustment led to an estimated 55 percent decline in average consumption, and a 60 percent decline among the poorest 20 percent of Poverty increased from less than 1 percent of the population to mot-e than 17 percent.

In Latin America throughout much of the 1980s, lack of adjustment was associated with high and rising inflation. While inflation stabilization usually means more unemployment and lower real wages in the short run, accelerating inflation also reduces real wages and increases poverty. Using data for 1977' to 1989 for seven Latin American countries, Cardoso found that real wages fell by about 14 percent when inflation doubled. She concludes that while there were years in which real wages declined in response to stabilization policy, the decline caused by a lack of policy and the resulting high inflation totally dominated the scene between 1977 and 1989. This is particularly significant given the findings of Morley and Alvarez. In their study of the 1989 recession in Argentina, they conclude that the decline in real wages caused by inflation was one of the most significant contributors (much more so than unemployment) to the rise in poverty during this period. Similarly, in a subsequent study of Brazil, Cardoso reports that real wage losses in the 1980s hurt all groups including the poor [they hurt the middle class the most), and that increases in the minimum wage worsened inequality, probably by reducing formal sector employment. By 1974-91 reducing the rapid inflation that erodes real wages, adjustment policies may actually prevent a worsening of poverty and hunger.

Figure 1 Premium on the parallel market for foreign exchange.

Lack of adjustment is also associated with import substitution facilitated by trade protection and overvalued exchange rates. The combination tends to lead to capital-intensive industrialization, which limits labor demand and weakens the incentives for the productivity gain that help drive up wages. If labor demand is weak, there is little hope for improvements in employment or the wages of the poor. As I mentioned in invoking the experience of East Asia, growth in labor demand is a critical component of shared growth that benefits all groups.

Unfortunately the best evidence for the effects of lack of adjustment on the poor and on poverty and hunger comes from Africa. Africa's poor growth performance in the 1980s and the failure to reduce poverty were caused by many factors, including severe terms of trade shocks and low initial levels of human capital. However, lack of adjustment also played a role: exchange rates were consistently overvalued (in countries with flexible rates), resulting in high black market premiums, and trade regimes were highly distorted with high non-tariff barriers (see figures 2 and 3). High black market premiums reflected governments' efforts to defend overvalued exchange rates by restricting access to foreign exchange. It was the poor who suffered the most from resulting higher parallel market prices for imported goods and from lack of access to highly rationed investment and consumer goods.

Of course, structural adjustment is almost always recessionary in the short run, implying employment and wage declines and reductions in public services. The poor may be affected by these changes, and even if they are affected proportionately less than other groups, they can least afford any losses. However, initial conditions in the world's poorest countries, where chronic hunger is great, and the nature of adjustments needed in those countries imply that the poor lose relatively less than we might think, and can have offsetting gains as a result of adjustment reforms Let me suggest at least four ways in which adjustment can help the poor.

Figure 3. Outward orientation of selected country group, 1965-85.

First, adjustment policies, especially devaluation and real depreciation of the exchange rate, raise the producer price of agricultural tradables. Among economies with the largest numbers of poor and hungry are those with large agricultural sectors producing tradables, and within economies, many of the poorest houses holds are in the agricultural sector Morley of the Inter-American Development Bank points out that although recessions in Latin America hurt the poor, in at least one case, Venezuela from 1981 to 1986, rural poverty declined during a recession, and that during the adjustment-led recoveries the decline in rural poverty has generally been greater than the decline in urban poverty. They also note that in Costa Rica, even though per capita income in 1987 was still below its pre-adjustment 1980 level, poverty had declined because the rural poor gained from adjustment reforms.

Second, elimination of subsidies to owners of capital through directed credit programs and through preferential access to rationed foreign exchange means the poor may gain access to services and goods they could not obtain, and may enjoy lower prices for those services and goods. In Africa, where rationing was most severe, the urban poor benefited quickly from the elimination of such subsidies.

Third, stabilization and adjustment reduce inflation, which, as I mentioned earlier, is associated with declining real wages, and especially hits the urban poor.

Fourth, in Africa, deregulation of the public sector monopolies of marketing boards raised rural incomes by eliminating the large taxes on producers extracted by the marketing boards At the same time, consumer prices of food staples have not necessarily risen, because limited supply kept parallel market prices so high in the pre-adjustment period. In Tanzania, real consumer prices for maize, rice, and beans all fell sharply between 1985 and 1987, when food crop marketing was first being liberalized, and even in 1992 were below the levels of the early 1980s. At the same time, the rural poor—the poorest in most countries and the most likely to suffer from chronic hunger—are relatively insulated from rises in the official prices of some consumer goods following depreciations, if only because they often rely heavily on nontraded food goods for their own consumption and because they had no access to cheap, but rationed, food imports before.

Finally, one effect of adjustment, often decried, is probably neutral in its effects on the poor. Cutbacks in spending on social programs do not hurt the poor, at least in the poorest economies, for a simple reason: the poor benefited relatively little from them in the first place.

None of this is to say that we should be sanguine about the design of adjustment programs Adjustment programs can and should be better designed so as to maximize their effects in reducing poverty. Where general food and energy subsidies are cut, the resulting fiscal resources can at least in part be directed toward targeted food programs for the poor. Where social sector spending is cut, the cuts should come from reductions in the types of social services that benefit the rich, and from reductions of redundant and politicized employment rather than from reductions in expenditures on drugs, fuel, books, and the other complementary inputs that ensure adequate quality for the poor and others. More emphasis on reducing deficits by enforcing broadly based and progressive taxes and by eliminating exemptions and loopholes that favor elites and less emphasis on cutting public investment in rural roads and schools is clearly in order.

These aspects of better design imply strength the adjustment process, not weakening it. Think of adjustment as a difficult political and social process of taking away privileges enjoyed by the elite, such as special access to economic favors and to social subsidies. In this light, better design of adjustment simply becomes another aspect of what I have called shared growth.

Transfers Cannot Substitute for Shared Growth and Adjustment

Finally, I come to my third point. During the adjustment process, transfers and other compensatory programs to redistribute income to the poor can also make sense and should certainly be tried, especially because even a small drop in income during a recession can take the poor across the line into chronic hunger. However, transfers, targeted employment, and other short-run safety net programs cannot substitute for a political shift to shared growth and the development of a new social and political consensus In the short run, such programs are easy to mount because they do not require contesting the economic position of the elites, especially given that international resources can usually be galvanized for these attractive clearly pro-poor programs. (In fact they are not easy to mount all that quickly.) However, without a new consensus, widely shared, they will be difficult to sustain fiscally, and thus politically, and to the extent they compete with scarce local administrative, technical, and political resources for implementing more universal programs, they present some tradeoff for those concerned with a frontal attack on the causes of poverty.


Let me conclude by restating my main objective: to convince you that you cannot ignore the need for economic and adjustment reforms if you want to eliminate hunger. Transfers and targeted programs can help, but they are not a substitute for the fundamental challenge of developing a new consensus around the idea of shared growth and the need for structural adjustment.

Reasonable people with reasonable skepticism will ask: how can we propagate the idea of shared growth in settings where it hasn't taken hold? After all, in East Asia, shared growth emerged as a reaction to specific events after the war.

I don't pretend it is an easy task, but I do see reasons for optimism. In Latin America, for example, the recent efforts of government leaders in Chile, Costa Rica, Mexico, and Peru represent the beginnings of a new consensus about the need to ensure that all groups benefit from growth To build on those efforts and to extend them to other regions, the international financial institutions can focus more on the need to develop some social consensus. A technocratic approach is not enough And community and nongovernmental organization groups can focus more on the reality that economic reforms, including structural adjustment reforms, are politically difficult to undertake precisely because they take away the elites privileges, and recognizing this, might take up anew the mandate to represent the interests of the poor in pushing for such reforms precisely because the poor stand to gain from them.


1. The views expressed here do not necessarily reflect those of the Inter-American Development Bank or of its members. A note on which these remarks are based, including technical references and sources, is available from the author.

2. World sank' The East Asian Miracle: Economic Growth and Public Policy especially chapter 4 (Washington, D.C Oxford University Press, 1993).

3 Paul Glewe and Gillette Hall, The Poor in Latin America during Adjustment: A Case Study Of Peru, 198590, World Bank Living Standards Measurement Study, Working Paper No. 86 (Washington, D C: World Bank 1992).

4. Eliana Cardoso, "The Macroeconomics of Poverty in Latin America" (Boston: Tufts University and National Bureau of Economic Research, 1992).

5. Samuel A. Morley and C Alvarez, "Recession and the Growth of Poverty in Argentina" (Washington, D C.: InterAmerican Development Bank, 1991). Processed.

6 Eliana Cardoso, "Cyclical Vacations of Earnings Inequality in Brazil," Revista de Economia Politica (forthcoming).

7 Samuel A Morley, Structural Adjustment and the Determinants of Poverty in Latin America (Washington, D C: Vanderbilt University and Inter-American Development Bank, revised March 1993). Processed.