Musalia Mudavadi

The Bretton Woods conference of 1944 focused on the problems to be encountered in postwar reconstruction, including projects for development and rehabilitation, as well as short-term solutions to temporary balance of payments disequilibria. New problems have now been identified, many of which relate to development in its market-oriented context and the requirements brought about by change in emerging economies, rather than the need to re-establish order in mature ones. The concept of poverty has changed from being a problem of an enclave to being the problem of the nation as a whole. Solutions are no longer, therefore, to be found in simple, targeted interventions. They depend more on augmenting a nation’s resource base than on its redistribution.

Not so long ago, discussions of development occupied the minds of those managing the World Bank and the Fund; then the pendulum swung, and the conscience of those institutions was disturbed by the plea that development must have a human face. Moreover, faith in trickle-down effects and balanced growth declined, and governments were called on to identify those who were bypassed by prosperous change. Women, the rural landless, the urban poor, and a whole multitude of vulnerable groups became the focus of special programs.

To reduce the incidence of poverty on a sustainable basis requires the vertical mobility of many of these groups, not just an alleviation of suffering, which may merely be a temporary palliative. Unfortunately, some changes, such as those relating to the status of women, do not immediately relieve poverty and, by diverting resources, may temporarily worsen it.

Change is always painful to some. Not all have the ability to voice their distress to sympathetic ears; some may easily gain attention, and others may be able to express their pleas for lost privileges. It is always difficult, therefore, to set about long-term poverty reduction where the method of intervention is either an appeal to efficiency, equity, or charity. It is my hope that today’s discussion will not just bring to light the causes of poverty but will also explore what strategies are feasible for eradicating it in the economies of the third world. We must not launch into programs that are unsustainable, because the frustration of failure and of falling back into poverty creates far greater distress.

Widjojo Nitisastro

Let me first say how honored I feel to be invited to speak on the topic of sustainable poverty reduction at this conference on the occasion of the fiftieth anniversary of the World Bank and the International Monetary Fund. I would like to express my good wishes to the IMF and the World Bank Group on this most important event. The choice of sustainable poverty reduction as a topic for discussion reflects the commitment of the two institutions to support developing countries in their endeavors to reduce poverty and improve living standards.

As to poverty reduction, I am familiar only with the experience of Indonesia. Therefore, my contribution to this conference will be limited to poverty reduction in Indonesia. Let me start with a few statistics.

The 1994 Annual Report of the World Bank indicated that

Among regions, East Asia and Pacific stands out as the one that has made the most impressive gains in poverty reduction, as well as being the fastest growing The most impressive gains have been made in Indonesia, where the percentage share of absolute poverty has fallen from 60 to 15 percent of the total population over the period [1970–1990], and China, where it fell from 33 to 10 percent.

The analysis of poverty over time is difficult and controversial. Nonetheless, estimates on poverty in Indonesia have reached a fair degree of consensus: poverty has fallen rapidly over the past 25 years. The proportion of the population below the poverty line fell from about 60 percent in 1970 to about 29 percent in 1980 and to about 14 percent in 1993. The absolute numbers of those in poverty fell from about 70 million in 1970 to about 26 million in 1993. Total population in 1993 was almost 190 million. More than 110 million live on the island of Java, which has an average density of more than 800 people per square kilometer. Of the 26 million below the poverty line, 15 million, or 58 percent, are in Java.

What explains the sharp drop in poverty in Indonesia? The most important factor seems to be sustained rapid economic growth, which was broadly based and labor intensive. The effects of this growth were reinforced by an array of policies that improved the health and education of the poor, reduced population growth to manageable levels, and provided infrastructure. In economic terms, the rate and pattern of growth generated a strong demand for labor, while the policies in education, health, and infrastructure enabled the poor to take advantage of this demand to improve their incomes.

Let us first discuss the growth process. Between 1970 and 1993, real GDP increased by about 6.5 percent annually. With population growing by about 2.2 percent on average, per capita GDP grew by over 4 percent a year. Of key importance to poverty reduction in the 1970s and early 1980s was the high rate of growth in the agricultural sector, on which most of the population and the poor depended. Production of rice, the most important crop, grew by nearly 5.3 percent a year between 1971 and 1983.

The sources of rapid growth in rice production have been a combination of the rapid spread of irrigation, the provision of key inputs, and the spread of high-yielding varieties. At the same time, investment in rural infrastructure, as well as price policy, public procurement, and price stabilization, increased the level and stability of the prices received by the farmer. This early emphasis on agriculture played a decisive role in breaking the downward cycle of poverty, population growth, and environmental degradation.

In the second half of the 1980s and the 1990s, a different process became important in generating high growth and reducing poverty—the rapid growth of exports of labor-intensive manufactures, which generated employment growth in manufacturing of about 7 percent a year after 1985. In addition, the growth in manufacturing employment was accompanied by rapid growth of construction and employment in the construction sector. During the three-year period from 1984 to 1987, Indonesia’s total outstanding debt increased from $31.2 billion to $50.2 billion, an increase of 60 percent. Its annual debt-service payment increased from $4.2 billion in 1984 to $6.9 billion in 1987, an increase of 64 percent.

The very sharp increase in both total debt outstanding and annual debt servicing was mainly due to the realignment of world currencies, in particular the rapid depreciation of the U.S. dollar against the Japanese yen and other major currencies. However, a substantial proportion of Indonesia’s debts was denominated in yen and currencies other than the U.S. dollar. It has been estimated by the IMF that more than 80 percent of the increase in the stock of outstanding debt and about two thirds of the increase in debt-service payments were attributable to the valuation effects of the depreciation of the U.S. dollar.

The impact of the sharp depreciation of the U.S. dollar on Indonesia’s development was even worse because of its timing. It coincided with the sharp decline in the international price of oil in 1986. Indonesia’s development was thus receiving a double blow: the sharp fall in oil prices and, at the same time, the sharp depreciation of the U.S. dollar. As a response to the new challenges, Indonesia embarked on a comprehensive set of economic reforms, policy adjustments, and restructuring. These measures included the postponement of major capital-intensive projects, devaluations, flexible exchange rate management, prudent fiscal and monetary policies, tax reforms, financial sector reforms, waves of deregulation of foreign trade and industry, the rollback of non-tariff barriers, tariff reform, improvement of the climate for investment, development of the stock exchange, and an all-out effort to capture foreign markets for non-oil exports.

All these measures, which were implemented in a consistent manner, resulted in a substantial growth of exports, employment, and investments. Instead of capital flight, the country had a vigorous capital inflow.

An issue that attracted wide-ranging interest was whether all those economic reforms, policy adjustment, and restructuring did not result in an ever-increasing burden for the poor. A number of studies carried out to measure the incidence of poverty and its trend concluded that the incidence of poverty declined from 22 percent in 1984 to about 18 percent in 1987, implying an absolute decline in the number of poor Indonesians from about 35 million in 1984 to about 30 million in 1987. The success in reducing poverty during the difficult adjustment period in the 1980s was due to three factors:

• First, the development efforts before the adjustment period were directed at establishing a strong rural economy and establishing an extensive network of social and physical infrastructure, such as primary schools leading to universal primary education; integrated centers for health, nutrition, and family planning; and rural networks of roads, irrigation facilities, and support for flood control.

• Second, the economic reform and policy adjustments introduced contained elements geared toward sustaining progress on poverty reduction. Budgetary expenditures in poverty-related sectors—such as agriculture, human resource development, and transfers to regional governments—were protected relative to other sectors.

• Third, the combination of trade and industrial deregulation and real exchange rate adjustments led to a rapid recovery in investment and employment in manufacturing and agriculture.

These policy changes were supported by the World Bank, the IMF, and bilateral donors. The inflow of funds from these sources cushioned the fall in oil prices, allowing government spending and aggregate demand to be maintained while avoiding high levels of inflation.

In summary, rapid, sustained, and labor-intensive growth was a major factor in reducing poverty in Indonesia. This close linkage between growth and poverty reduction is reflected in Indonesia’s particular growth process: initial rapid growth in agriculture followed by rapid growth in labor-intensive manufactures—and in the policy packages that generated it.

Let me now turn to another important factor accounting for poverty reduction in Indonesia: the slowdown in population growth. Annual population growth rates in Indonesia fell from 2.5 percent in 1970 to about 1.7 percent at present. This drop was the result of a pronounced reduction in fertility rates at the same time as mortality rates fell rapidly. A successful family planning program, as well as rapid educational and employment gains among women, played a crucial role in reducing population growth. Indonesia’s total fertility rate was nearly halved in two decades—from 5.5 in 1970 to 3.0 in 1990. Child survival rates improved, owing to the provision of health care, especially through programs such as universal child immunization, improved nutrition, and rapid gains in female education. The gross enrollment rate for females in primary education increased from about 73 percent in 1970 to 114 percent in 1990,¹ and in secondary education from 16 percent in 1970 to 45 percent in 1990. The effect of these factors on reducing population growth is now well recognized.

In sum, as I have tried to illustrate, Indonesia was able to reduce poverty rapidly, first, through sustained, broad-based and labor-intensive growth based on rapid growth of agriculture, and then through rapid growth of labor-intensive manufacturing exports. Second, the poor were able to participate in that growth because of substantial improvements in education and health and investments in infrastructure. Third, population growth fell sharply.

What lessons can be drawn from Indonesia’s experience in achieving a rapid reduction of poverty? First and foremost is the need for a true commitment to achieving it. It has to be a true commitment as opposed to adhering passively to possible fashions in the development debate. Such a commitment has to be translated into operational policies and programs to be implemented in a consistent manner. These policies and programs have to be internally designed and self-imposed rather than being parts of conditions attached to loans or grants. The test of a true commitment arrives when the availability of resources is rapidly declining: whether to forgo other claims or to yield to pressures and sacrifice the poverty-reduction programs.

Another critical requirement is the development of capabilities, both individual and institutional, to identify problems and opportunities and to design and implement poverty-reduction policies and programs. However, poverty reduction should not be postponed until such capabilities are completely developed.

The Indonesian experience with poverty reduction also shows clearly the important role of broadly based labor-intensive economic growth, together with the rapid growth of primary education and the effective delivery of health care and family planning services. It follows that the pattern of growth pursued in a country is of great importance for achieving the objective of poverty reduction. Broad-based labor-intensive growth in Indonesia was primary in agriculture during the 1970s and shifted toward labor-intensive manufactured exports in the 1980s and 1990s.

Education and health are ends in themselves, but they also have a great effect on the pace of labor-intensive growth of the economy. In particular, the education of women is of great social and economic benefit.

Finally, what can or should the international community do to assist developing countries in achieving poverty reduction? Needless to say, poverty reduction is the responsibility and, therefore, the homework of the individual country concerned. However, to be successful in carrying out a poverty-reduction program, there is a serious need for an enabling external environment—a world economic environment that is supportive. The following are some examples.

Volatile exchange rates between the major currencies can have a devastating impact on a developing country’s economy. Thus, the rapid depreciation of the U.S. dollar against the yen could have wiped out Indonesia’s endeavors in poverty reduction. The two economic superpowers in the Pacific certainly did not intend to harm another country in the Pacific, but in trying to find solutions to their problems, they did not take into account the impact of their actions on the developing countries. The establishment of a stable and predictable exchange rate system can be achieved only by the major developed countries.

Another important enabling external environment is related to market access. Manufactured exports of developing countries, such as textiles and garments, are facing nontariff barriers in accordance with the Multifiber Arrangement (MFA). Everybody is happy that the Uruguay Round was agreed upon by all countries, but the MFA will remain in force for another ten years. We have to wait ten years before there will be real market access for our textiles.

The burden of debt repayment can be very severe for developing countries. Let me describe Indonesia’s experience with debt. When there was a change in government in 1966, Indonesia had a huge debt in relation to exports, reserves, and GNP. At that time, Indonesia had arrears, and what it did, of course, just as any other developing country, was to go to the Paris Club and negotiate a rescheduling. We came out with an agreement, namely, rescheduling for eight years with three years’ grace period. Following that agreement, we had to go to the Paris Club for another rescheduling, and then again. In addition, we had to negotiate at that time with 22 creditor countries. There was complete uncertainty about what would happen the following year. That was very much in our mind. If you ask people working in a country to take care of poverty problems while worrying about the payment of debt the next year, I think that is too much to ask.

What happened next was that the Paris Club agreed to ask a third party to study the problem of Indonesia’s debt and come up with a proposal for a solution. After some time, we agreed to ask the late Dr. Hermann Abs of the Deutsche Bank—which at that time was not one of Indonesia’s creditors—to make a study of Indonesia’s debt problem for the Paris Club and for Indonesia. In his work, Dr. Abs was assisted by many able officials from the World Bank and the Fund.

Dr. Abs came up with three proposals. The first was that the settlement of the debt should be a once-and-for-all settlement that did not have to be repeated every year. Once done, that was it; no more negotiations. The second principle was that Indonesia must be able to pay; the settlement was to make Indonesia creditworthy so that it could have normal financial relations. Creditworthiness was the objective. And third, there should be a nondiscriminatory treatment of all debt.

Indonesia was lucky because the Paris Club countries agreed at that time to these three proposals. Unfortunately, the Paris Club said that this was a unique case that could not be used as a precedent for other developing countries.

This history is just to give you an example of the big difference it made to Indonesia to have its debt settled—and settled in a way that can be called a final settlement. Because of that, we were able to concentrate all of our efforts on the development of the country and on poverty alleviation. The settlement of Indonesia’s debt on the basis of the principles put forward by Dr. Hermann Abs (once-and-for-all settlement, creditworthiness as an objective, and the principle of nondiscrimination) could be considered more generally for the settlement of debts of other developing countries.

Albert Fishlow

Income distribution entered the postwar discussion of economic development relatively late. Up to the 1960s, much of the focus was on industrialization and the need for capital accumulation as a motor for more rapid expansion. It was an era, moreover, of large flows of external assistance, much of it to the European economy, in the midst of global division between East and West. Indeed, not until the mid-1950s did the World Bank turn its focus to the developing countries, and then only gradually. And already by that time, the flows of assistance from industrial countries, primarily the United States, were beginning to diminish proportionally.

Moreover, that early period saw the end of colonization and the beginning of new national units throughout Asia and Africa. For them, the focus became economic expansion as a means of achieving political identity. Rapid population growth, associated with great declines in mortality, focused attention on the acceleration of production required to keep pace. Overall, the period saw much higher rates of aggregate expansion than had been experienced previously. And there was even acceleration: per capita growth increased from 2.4 percent a year in the 1950s to 3.5 percent in 1965–70. Indeed, in this latter period, growth rates were not only positive in every region, but the lowest, for sub-Saharan Africa, amounted to an annual increase of 2.3 percent per capita.

In the late 1960s and early 1970s, in the context of positive achievement, there emerged a new interest in income distribution. On the academic side, a new series of studies called attention to continuing significant inequality and the failure of widespread economic growth to remedy the problem (see Adelman and Robinson, 1989). On the practical side, Robert McNamara’s interest in the subject stimulated the World Bank study by Chenery and others (1974), an important effort to define new policies and strategies that could be effective in combining increased equity with improved aggregate performance.

In the event, distributional issues, soon extending into new attention to basic needs—a policy approach emanating from the World Bank—had only a brief moment on center stage. Most important, the rise in oil prices in 1973 set in motion a sequence of excessive dependence by developing countries on foreign debt, initially quite cheap but soon impossible to sustain, which dominated attention in the 1980s and beyond. Adjustment dominated. Suddenly, the IMF became a central element in providing assistance to developing countries, something it had not done in great measure before. The balance of payments became the determining factor in policy Countries were necessarily focused not only on getting their macroeconomic policies right but also on a whole range of related issues. These involved questions like internal wage and interest rate policy, government subsidies, and, inevitably, the role of the state in guiding the development process.

Growth, or lack thereof, increasingly became the burning issue among developing countries. Since the 1980s, we have seen wider differences in regional performance than at any other time in the postwar period. Africa and Latin America have experienced stagnation while Asia has moved rapidly ahead. Yet in the midst of more widespread democratization than has occurred at any previous time, increasing concerns about the fate of the poor and their future become virtually inevitable. With renewed global expansion, as finally now seems the case, the question of income equality and what can be done will surely again command attention.

In this paper, I briefly take up three central questions: the state of income inequality currently; policy measures that can be adopted to help the evolution of the poor more generally; and the state of recent research on the relevance of inequality to the growth process. I conclude with some final observations.

The Record

The World Development Report 1994 of the World Bank reports in its statistical annex, as it has for many years, recent data accumulated on the measurement of income inequality (Table 1).² These data reveal that shares of income received by the top 10 percent of the population in developing countries vary from more than 50 percent in Brazil to about 25 percent, shared by a number of countries; the inclusion of Eastern Europe, showing the effects of socialist equalization, brings that minimum closer to 20 percent. The range among developed countries is much narrower, and concentrates at less than 25 percent; income inequality is obviously smaller.

Table 1.

Income Distribution

(Percent share of income or consumption)

Source: World Bank (1994), Table 30.

¹Brazil, Botswana, Malaysia, Venezuela, Hungary, Mexico, and Korea are considered “upper-middle-income” economies; remaining countries in this group are considered to be “lower-middle-income” economies.

Table 1.

Income Distribution

(Percent share of income or consumption)

	Year	Lowest 20 percent	Second quintile	Third quintile	Fourth quintile	Highest 20 percent	Highest 10 percent
Low-income economies
Ethiopia	1981–82	8.6	12.7	16.4	21.1	41.3	27.5
Tanzania	1991	2.4	5.7	10.4	18.7	62.7	46.5
Nepal	1984–85	9.1	12.9	16.7	21.8	39.5	25.0
Uganda	1989–90	8.5	12.1	16.0	21.5	41.9	27.2
Bangladesh	1988–89	9.5	13.4	17.0	21.6	38.6	24.6
Guinea-Bissau	1991	2.1	6.5	12.0	20.6	58.9	42.4
Rwanda	1983–85	9.7	13.1	16.7	21.6	38.9	24.6
India	1989–90	8.8	12.5	16.2	21.3	41.3	27.1
Kenya	1992	3.4	6.7	10.7	17.3	61.8	47.9
Pakistan	1991	8.4	12.9	16.9	22.2	39.7	25.2
Ghana	1988–89	7.0	11.3	15.8	21.8	44.1	29.0
China	1990	6.4	11.0	16.4	24.4	41.8	24.6
Mauritania	1987–88	3.5	10.7	16.2	23.3	46.3	30.2
Sri Lanka	1990	8.9	13.1	16.9	21.7	39.3	25.2
Zimbabwe	1990–91	4.0	6.3	10.0	17.4	62.3	46.9
Honduras	1989	2.7	6.0	10.2	17.6	63.5	47.9
Lesotho	1986–87	2.9	6.4	11.3	19.5	60.0	43.6
Indonesia	1990	8.7	12.1	15.9	21.1	42.3	27.9
Zambia	1991	5.6	9.6	14.2	21.0	49.7	34.2
Middle-income economies1
Côte d’Ivoire	1988	7.3	11.9	16.3	22.3	42.2	26.9
Bolivia	1990–91	5.6	9.7	14.5	22.0	48.2	31.7
Philippines	1988	6.5	10.1	14.4	21.2	47.8	32.1
Senegal	1991–92	3.5	7.0	11.6	19.3	58.6	42.8
Peru	1985–86	4.9	9.2	13.7	21.0	51.4	35.4
Guatemala	1989	2.1	5.8	10.5	18.6	63.0	46.6
Morocco	1990–91	6.6	10.5	15.0	21.7	46.3	30.5
Dominican Republic	1989	4.2	7.9	12.5	19.7	55.6	39.6
Jordan	1991	6.5	10.3	14.6	20.9	47.7	32.6
Bulgaria	1992	10.4	13.9	17.3	22.2	36.2	21.9
Colombia	1991	3.6	7.6	12.6	20.4	55.8	39.5
Jamaica	1990	6.0	9.9	14.5	21.3	48.4	32.6
Tunisia	1990	5.9	10.4	15.3	22.1	46.3	30.7
Algeria	1988	6.9	11.0	14.9	20.7	46.5	31.7
Thailand	1988	6.1	9.4	13.5	20.3	50.7	35.3
Poland	1989	9.2	13.8	17.9	23.0	36.1	21.6
Costa Rica	1989	4.0	9.1	14.3	21.9	50.8	34.1
Panama	1989	2.0	6.3	11.6	20.3	59.8	42.1
Chile	1989	3.7	6.8	10.3	16.2	62.9	48.9
Brazil	1989	2.1	4.9	8.9	16.8	67.5	51.3
Botswana	1985–86	3.6	6.9	11.4	19.2	58.9	42.9
Malaysia	1989	4.6	8.3	13.0	20.4	53.7	37.9
Venezuela	1989	4.8	9.5	14.4	21.9	49.5	33.2
Hungary	1989	10.9	14.8	18.0	22.0	34.4	20.8
Mexico	1984	4.1	7.8	12.3	19.9	55.9	39.5
Korea, Rep. of	1988	7.4	12.3	16.3	21.8	42.2	27.6
High-income economies
New Zealand	1981–82	5.1	10.8	16.2	23.2	44.7	28.7
Israel	1979	6.0	12.1	17.8	24.5	39.6	23.5
Spain	1988	8.3	13.7	18.1	23.4	36.6	21.8
Hong Kong	1980	5.4	10.8	15.2	21.6	47.0	31.3
Singapore	1982–83	5.1	9.9	14.6	21.4	48.9	33.5
Australia	1985	4.4	11.1	17.5	24.8	42.2	25.8
United Kingdom	1988	4.6	10.0	16.8	24.3	44.3	27.8
Italy	1986	6.8	12.0	16.7	23.5	41.0	25.3
Netherlands	1988	8.2	13.1	18.1	23.7	36.9	21.9
Canada	1987	5.7	11.8	17.7	24.6	40.2	24.1
Belgium	1978–79	7.9	13.7	18.6	23.8	36.0	21.5
Finland	1981	6.3	12.1	18.4	25.5	37.6	21.7
France	1989	5.6	11.8	17.2	23.5	41.9	26.1
Germany	1988	7.0	11.8	17.1	23.9	40.3	24.4
United States	1985	4.7	11.0	17.4	25.0	41.9	25.0
Norway	1979	6.2	12.8	18.9	25.3	36.7	21.2
Denmark	1981	5.4	12.0	18.4	25.6	38.6	22.3
Sweden	1981	8.0	13.2	17.4	24.5	36.9	20.8
Japan	1979	8.7	13.2	17.5	23.1	37.5	22.4
Switzerland	1982	5.2	11.7	16.4	22.1	44.6	29.8

Source: World Bank (1994), Table 30.

¹Brazil, Botswana, Malaysia, Venezuela, Hungary, Mexico, and Korea are considered “upper-middle-income” economies; remaining countries in this group are considered to be “lower-middle-income” economies.

View Table

Table 1.

Income Distribution

(Percent share of income or consumption)

	Year	Lowest 20 percent	Second quintile	Third quintile	Fourth quintile	Highest 20 percent	Highest 10 percent
Low-income economies
Ethiopia	1981–82	8.6	12.7	16.4	21.1	41.3	27.5
Tanzania	1991	2.4	5.7	10.4	18.7	62.7	46.5
Nepal	1984–85	9.1	12.9	16.7	21.8	39.5	25.0
Uganda	1989–90	8.5	12.1	16.0	21.5	41.9	27.2
Bangladesh	1988–89	9.5	13.4	17.0	21.6	38.6	24.6
Guinea-Bissau	1991	2.1	6.5	12.0	20.6	58.9	42.4
Rwanda	1983–85	9.7	13.1	16.7	21.6	38.9	24.6
India	1989–90	8.8	12.5	16.2	21.3	41.3	27.1
Kenya	1992	3.4	6.7	10.7	17.3	61.8	47.9
Pakistan	1991	8.4	12.9	16.9	22.2	39.7	25.2
Ghana	1988–89	7.0	11.3	15.8	21.8	44.1	29.0
China	1990	6.4	11.0	16.4	24.4	41.8	24.6
Mauritania	1987–88	3.5	10.7	16.2	23.3	46.3	30.2
Sri Lanka	1990	8.9	13.1	16.9	21.7	39.3	25.2
Zimbabwe	1990–91	4.0	6.3	10.0	17.4	62.3	46.9
Honduras	1989	2.7	6.0	10.2	17.6	63.5	47.9
Lesotho	1986–87	2.9	6.4	11.3	19.5	60.0	43.6
Indonesia	1990	8.7	12.1	15.9	21.1	42.3	27.9
Zambia	1991	5.6	9.6	14.2	21.0	49.7	34.2
Middle-income economies1
Côte d’Ivoire	1988	7.3	11.9	16.3	22.3	42.2	26.9
Bolivia	1990–91	5.6	9.7	14.5	22.0	48.2	31.7
Philippines	1988	6.5	10.1	14.4	21.2	47.8	32.1
Senegal	1991–92	3.5	7.0	11.6	19.3	58.6	42.8
Peru	1985–86	4.9	9.2	13.7	21.0	51.4	35.4
Guatemala	1989	2.1	5.8	10.5	18.6	63.0	46.6
Morocco	1990–91	6.6	10.5	15.0	21.7	46.3	30.5
Dominican Republic	1989	4.2	7.9	12.5	19.7	55.6	39.6
Jordan	1991	6.5	10.3	14.6	20.9	47.7	32.6
Bulgaria	1992	10.4	13.9	17.3	22.2	36.2	21.9
Colombia	1991	3.6	7.6	12.6	20.4	55.8	39.5
Jamaica	1990	6.0	9.9	14.5	21.3	48.4	32.6
Tunisia	1990	5.9	10.4	15.3	22.1	46.3	30.7
Algeria	1988	6.9	11.0	14.9	20.7	46.5	31.7
Thailand	1988	6.1	9.4	13.5	20.3	50.7	35.3
Poland	1989	9.2	13.8	17.9	23.0	36.1	21.6
Costa Rica	1989	4.0	9.1	14.3	21.9	50.8	34.1
Panama	1989	2.0	6.3	11.6	20.3	59.8	42.1
Chile	1989	3.7	6.8	10.3	16.2	62.9	48.9
Brazil	1989	2.1	4.9	8.9	16.8	67.5	51.3
Botswana	1985–86	3.6	6.9	11.4	19.2	58.9	42.9
Malaysia	1989	4.6	8.3	13.0	20.4	53.7	37.9
Venezuela	1989	4.8	9.5	14.4	21.9	49.5	33.2
Hungary	1989	10.9	14.8	18.0	22.0	34.4	20.8
Mexico	1984	4.1	7.8	12.3	19.9	55.9	39.5
Korea, Rep. of	1988	7.4	12.3	16.3	21.8	42.2	27.6
High-income economies
New Zealand	1981–82	5.1	10.8	16.2	23.2	44.7	28.7
Israel	1979	6.0	12.1	17.8	24.5	39.6	23.5
Spain	1988	8.3	13.7	18.1	23.4	36.6	21.8
Hong Kong	1980	5.4	10.8	15.2	21.6	47.0	31.3
Singapore	1982–83	5.1	9.9	14.6	21.4	48.9	33.5
Australia	1985	4.4	11.1	17.5	24.8	42.2	25.8
United Kingdom	1988	4.6	10.0	16.8	24.3	44.3	27.8
Italy	1986	6.8	12.0	16.7	23.5	41.0	25.3
Netherlands	1988	8.2	13.1	18.1	23.7	36.9	21.9
Canada	1987	5.7	11.8	17.7	24.6	40.2	24.1
Belgium	1978–79	7.9	13.7	18.6	23.8	36.0	21.5
Finland	1981	6.3	12.1	18.4	25.5	37.6	21.7
France	1989	5.6	11.8	17.2	23.5	41.9	26.1
Germany	1988	7.0	11.8	17.1	23.9	40.3	24.4
United States	1985	4.7	11.0	17.4	25.0	41.9	25.0
Norway	1979	6.2	12.8	18.9	25.3	36.7	21.2
Denmark	1981	5.4	12.0	18.4	25.6	38.6	22.3
Sweden	1981	8.0	13.2	17.4	24.5	36.9	20.8
Japan	1979	8.7	13.2	17.5	23.1	37.5	22.4
Switzerland	1982	5.2	11.7	16.4	22.1	44.6	29.8

Source: World Bank (1994), Table 30.

¹Brazil, Botswana, Malaysia, Venezuela, Hungary, Mexico, and Korea are considered “upper-middle-income” economies; remaining countries in this group are considered to be “lower-middle-income” economies.

What is apparent from these numbers, now available more accurately and widely than before, is that the simple U-hypothesis advanced by Simon Kuznets does not seem to hold as universally as was first thought (see Kuznets, 1955; and Robinson, 1976). That conception, in which low-income and high-income countries would share lower levels of inequality owing to greater homogeneity of their labor forces first in agriculture and then in nonagricultural activities, has been carefully re-examined in recent years.³

Over the range of developing countries, there seems to be little regular variation with income level. Thus, a simple relationship trying to fit a parabola to the income share of the bottom 40 percent to level of income, whether as estimated by Summers and Heston (see, for example, Summers and Heston, 1991) or using more conventional values, does not work for recent periods. This is true both for a sample of countries similar to that used by Ahluwalia (1976) and a broader one that incorporates more recent entries. Anand and Kanbur (1993) utilize a series of different measures of inequality and find, first, that these different indices yield widely disparate turning points; second, that the statistical results indicated by a strict theoretical derivation are generally poor, especially when developing countries alone are considered, and that implied restrictions on the arithmetic value of coefficients are not satisfied; and third, that the change in relationship between means of agricultural and urban labor forces imposes new and still more rigorous limitations on the data.

The problem exposed, however, seems less with the inherent consistency of the initial Kuznets hypothesis than with the reality of substantial intervention in the economic system. Although not contemplated in the original formulation, policy in individual countries does differ widely with respect to the incidence of land redistribution, as well as with utilization of the rural labor force. It also varies in the urban areas, where the levels of employment and wage rates, and of capital intensity and profitability, are much affected by the industrial strategies pursued. Moreover, for most of the developing countries chronicled, the share of the population engaged in agriculture already has fallen significantly in recent years, suggesting that many should be on an equalizing course already.

If one goes on to more complex statistical relationships that introduce a number of other relevant variables to the analysis, inequality seems again to vary parabolically with income. I provide some indication of these findings in Table 2. What these relationships also show is clear indication of the much higher inequality in Latin America, once that regional distinction is introduced; the importance of secondary school attendance, past population growth, and share of income generated by agriculture as determining variables; and, both for developing countries alone and for a broader sample, including the developed countries, some tendency for income distribution to improve at higher levels of per capita income.⁴ The important point is that inequality is not merely a random variable.

Table 2.

Regression Estimates of Inequality¹

Sources: World Bank, World Development Report 1994; Social Indicators of Development, 1994; and World Tables, 1994.

¹t-values are in parentheses.

²Includes 64 countries.

³Includes 45 countries.

⁴Logarithm of the level of Summers-Heston income.

⁵Average percent of age group enrolled in secondary education, 1965–90.

⁶Average percent of aggregate income generated by agriculture, 1977–92.

⁷Average annual population growth, 1977–92.

⁸Dummy variable to represent inclusion in Latin America.

Table 2.

Regression Estimates of Inequality¹

	All Countries2		Developing Countries3
	Income Share of Bottom 40 percent	Theil Index of Inequality	Income Share of Bottom 40 percent	Theil Index of Inequality
Constant	−121.5	696.7	−111.5	479.8
	(3.8)	(8.1)	(2.1)	(5.7)
Income4	32.5	−82.5	29.3	−39.5
	(4.2)	(4.0)	(2.0)	(1.6)
Income squared4	−1.91	4.59	−1.67	2.68
	(4.2)	(3.7)	(1.7)	(1.4)
Education5	0.03	−0.09	0.04	−0.23
	(1.0)	(1.1)	(.8)	(1.8)
Agriculture6	0.22	−4.4	0.23	−0.53
	(3.8)	(1.4)	(3.5)	(2.6)
Population growth7	−1.32	2.26	−1.14	3.72
	(2.1)	(1.4)	(1.4)	(1.6)
Latin America8	−6.2	18.5	−6.4	15.3
	(5.3)	(5.8)	(4.3)	(3.6)
R squared	0.55	0.61	0.53	0.41

Sources: World Bank, World Development Report 1994; Social Indicators of Development, 1994; and World Tables, 1994.

¹t-values are in parentheses.

²Includes 64 countries.

³Includes 45 countries.

⁴Logarithm of the level of Summers-Heston income.

⁵Average percent of age group enrolled in secondary education, 1965–90.

⁶Average percent of aggregate income generated by agriculture, 1977–92.

⁷Average annual population growth, 1977–92.

⁸Dummy variable to represent inclusion in Latin America.

View Table

Table 2.

Regression Estimates of Inequality¹

	All Countries2		Developing Countries3
	Income Share of Bottom 40 percent	Theil Index of Inequality	Income Share of Bottom 40 percent	Theil Index of Inequality
Constant	−121.5	696.7	−111.5	479.8
	(3.8)	(8.1)	(2.1)	(5.7)
Income4	32.5	−82.5	29.3	−39.5
	(4.2)	(4.0)	(2.0)	(1.6)
Income squared4	−1.91	4.59	−1.67	2.68
	(4.2)	(3.7)	(1.7)	(1.4)
Education5	0.03	−0.09	0.04	−0.23
	(1.0)	(1.1)	(.8)	(1.8)
Agriculture6	0.22	−4.4	0.23	−0.53
	(3.8)	(1.4)	(3.5)	(2.6)
Population growth7	−1.32	2.26	−1.14	3.72
	(2.1)	(1.4)	(1.4)	(1.6)
Latin America8	−6.2	18.5	−6.4	15.3
	(5.3)	(5.8)	(4.3)	(3.6)
R squared	0.55	0.61	0.53	0.41

Sources: World Bank, World Development Report 1994; Social Indicators of Development, 1994; and World Tables, 1994.

¹t-values are in parentheses.

²Includes 64 countries.

³Includes 45 countries.

⁴Logarithm of the level of Summers-Heston income.

⁵Average percent of age group enrolled in secondary education, 1965–90.

⁶Average percent of aggregate income generated by agriculture, 1977–92.

⁷Average annual population growth, 1977–92.

⁸Dummy variable to represent inclusion in Latin America.

Clearly, the transformation of economies from low levels of income can create inherent tendencies toward higher levels of inequality, but what the record obviously suggests is that policies can make a palpable difference in averting them. Equally, and much more disturbing, is the continuing unaltered concentration of income in several countries despite intervening rises in income; interestingly, that leads to higher parabolic turning points in recent estimates than was characteristic of the earlier statistical analyses. That is particularly true of inequality in Latin America.

New and more vigorous efforts are necessary in the laggards, and at an early juncture, lest the democratic trend of the past decade be reversed. Perhaps the most important two interventions, whose economic effects are clearly visible, are those relating to physical asset structure and to rapid accumulation of human capital.

Of the possibilities associated with redistribution of wealth, land stands out as clearly the most potent variable. Land reform, where used, has been a powerful factor in affecting income distribution. The Republic of Korea and Taiwan Province of China stand out as the two most obvious cases where an initial commitment to equalization played a major role; the implications of that experience are currently very much being re-examined in an attempt to find methods to generalize it (see World Bank, 1993; but also Fishlow and others, 1994, for a critical evaluation). But in addition to those two cases, the experiences of China and Viet Nam, now showing signs of quite rapid and vigorous expansion, should not be forgotten. In those countries, an imposed equality generated relatively little growth until combined with greater personal incentives and opportunities for productivity improvement. Strict communist doctrine did not work, but its initial redistributive measures served as a powerful equalizing device that those countries subsequently could draw upon to their advantage.

Where there is a large rural population, productivity and equality are both served by such efforts. There is little doubt that the high rates of inequality found in Latin America trace back to the initially much higher inequality of land distribution in the region dating from the nineteenth century and, in some cases, even before. That inequality preceded the later commitment to import-substituting industrialization. Also, there can be little question, despite the lesser enthusiasm now available, that fairer allocation of this basic asset can serve as an important stimulant to broadly based development. For many reasons, there is a seeming reluctance to act; land reform seems to await war or revolution for its imposition. Yet where land is poorly utilized and its productivity can be much enhanced, redistribution—even with the payment of current capital value—can be quite economical.

What must be stressed, however, are the essential additional public inputs required for any distributive policy to be effective. Reallocation alone will not work. Only if viewed as a means toward the end of greater efficiency and output can any effort be successful. Too frequently, reform initiatives have encountered significant opposition from landholders reluctant to give up their command over resources, even when poorly utilized.

Human capital is a second area in which public actions can make a large difference. And it is one to which developing countries have made significant commitments over the postwar period. One need only cite the record from 1960 to 1991. For low-income countries, enrollment in primary school has increased from 54 percent to 101 percent (see also footnote 2, above); for secondary school, from 14 percent to 41 percent; and for tertiary, from 2 percent to 5 percent. For middle-income countries, enrollment has risen from 81 percent to 104 percent; from 17 percent to 55 percent; and from 4 percent to 18 percent, respectively (World Bank, 1994). Many countries have substantially increased their expenditures in this area.

There is good reason for these actions, in two dimensions. First, there is a quite high social return from the investment. The World Bank finds rates in excess of 25 percent for primary education, 15–18 percent for secondary, and 13–16 percent for tertiary. There is even a positive relationship with years of schooling: East Asia, with an average of six years, yields a higher increase in GDP for additional investment than do other regions (Psacharopoulos, 1985). What this reinforces is the importance of achieving adequate minimal training: many students in rural areas, through grade repetition and lack of completion, wind up little better for it. Note as well that female education regularly has an indirect, negative feedback on fertility, as well as positive effects on family health and nutrition.

The second benefit from education is in the realm of income distribution. For a large variety of countries, where some type of decomposition of the sources of inequality has been undertaken, variation in education typically shows up as the principal explanatory variable. Its range of importance extends from 10 percent to more than 20 percent of observed inequality. Those who are better schooled regularly receive higher incomes, holding constant age, sector, etc. Typically, at the bottom of the income distribution, education is more important: one recent study cites those with no schooling as having a 56 percent probability of being in the lowest 20 percent, whereas those with university training had only a 4 percent chance (Behrman, 1993, p. 196). Undoubtedly, education is partially serving as a proxy for other associated factors, such as family characteristics, school quality, and labor market characteristics. However, its repeated relevance, even discounting some of the gross effect, gives it a crucial role as a policy contributing to greater distributional equality.

More active participation by government is necessary, particularly at local levels, where efforts must be undertaken to ensure attendance and maintenance of school quality. But beyond this, curricular reform in many instances is badly required. It does little good to provide a standard training that leads simply to grade repetition rather than the acquisition of skills. Whatever their many deficiencies, moreover, the record of the former communist countries stands out in the speed with which they achieved universal education. It remains as a task to show that market-oriented economies can do as well.

Beyond education itself, there is a whole range of ancillary activities whose effects are similar in improving the quality of labor: these include nutrition, health, and other social investments. Indeed, for very low-income countries, the returns from such outlays may be greater than those received from education, exactly because the latter is insufficient to have a direct impact. Countries that have made early and substantial commitments to universal access to this human capital have been rewarded by lower inequality as growth proceeds; in turn, the social returns from expanded investment in education have permitted simultaneous gains in real product.

There is also a generational phenomenon. Today’s outlays will not have an immediate and observable impact on output, but one that is delayed. That makes it easier to postpone these outlays when fiscal stringency is necessary. One of the great tragedies of the 1980s in Africa and Latin America is not merely the poor economic performance that occurred but also the simultaneous reductions in government expenditures that were necessarily imposed. In many ways, it was much easier to cut back in areas that had initially failed to be universal in scope. This, in turn, meant shortfalls in reaching the very groups initially excluded from adequate coverage. In most cases, private outlays for education, health services, etc., continued, thus extending initial divisions between those who were better off and those with limited initial access. These kinds of delayed response are still to be felt in future decades, even as growth recovers. Not only has there been a “lost decade” of growth, but potentially a “lost generation” deprived of skills and any capacity to acquire them.

There is now greater realism in understanding that major efforts to achieve better distribution of income cannot rely on populist promises of higher wages and extensive direct economic controls and regulations. These efforts are temporary at best. The evidence of failure—in Chile in the 1970s and Peru and Nicaragua in the 1980s—is quite marked. Such paths of initially improved distribution come from arbitrary gains in income, gains that subsequently evaporate as inevitable adjustment occurs. Even without such extremes, policies adopted that rely heavily on subsidies, whether to agriculture or industry, frequently translate into efforts that distort incentives and benefit well-placed groups; the multiplicity of interventions ironically cancel out any real allocation effect that had been anticipated.

Indeed, the triumph of the market is now widely recognized, even where it had once been resisted. But conversely, the recognition that markets should play a wider role in the development process is still consistent with important areas of public responsibility. A danger is that the fear of excessive and inefficient government intervention excludes productive investment that can yield growth as well as improve income distribution. That would be a great tragedy in the midst of the present policy convergence. The commitment must be not merely to ensure current increases in product but also to guarantee them in the future through a strategy of continuing public investment and taxation.

Scope for progressive taxation is necessarily limited when one must simultaneously advance private savings and investment. One cannot assess positive gains from redistribution without paying attention to potential negative consequences. But equally, one cannot ignore possibilities for utilizing enhanced revenues productively. The real challenge for the future is creating an efficient public presence. Nowhere is it more important than in the realm of income distribution.

New Theoretical Advances

Some of the gains achieved in recent years in the analysis of growth have come from applying new models. Most prominent among them has been the theory of endogenous growth. One part of this neo-Schumpeterian literature starts from external increasing returns that drive continuing expansion; others have emphasized perfect competition, but permitting unlimited accumulation of capital with positive productivity; still a third has gone on to introduce either improved or new goods, invoking fixed costs as the limiting factor; and yet another group has introduced freer trade as a mechanism for growth, either through technological progress directly or indirectly by allowing for specialization in intermediate goods.

Some of these efforts have led to empirical estimation. Thus, models in which education, for example, plays a key role in explaining extended growth for a cross-section of countries have become common. Others have related trade policy to long-run growth in a similar fashion.⁵

A second and even newer element in the recent literature has been the explicit addition of political factors as important causal factors. The essence of such political economy is to recognize the important interaction between a willingness to accept less today in return for more tomorrow: democratic voters may be willing to forgo populism now in favor of increased investment and future growth. There is thus a combination of endogenous growth with endogenous policy in such efforts.

The key result so far obtained, exemplified by recent articles by Persson and Tabellini (1994) and by Alesina and Rodrik (1994), is that inequality is negatively related to growth. In the former piece, the effects of equality on growth are not only statistically significant but also quantitatively important. Thus an increase in equality with a standard deviation of 1, changing the share of income of the middle quintile of the income distribution by about 3 percent, is capable of increasing growth by ½ of 1 percentage point. Moreover, and powerfully, this relationship seems to hold only for those countries that follow democratic policies; for those that do not—some 40 percent of the sample—the effect of inequality is not statistically different from zero. In the latter study, the Gini coefficient for a larger number of countries is employed as a key variable and turns out to be statistically significant. A reduction in the Gini coefficient with a standard deviation of 1 increases growth by more than 1 percentage point. But Alesina and Rodrik stress that they find no difference between democracies and nondemocracies.

As of yet, these findings, and still others in a similar vein, must still be regarded as tentative. The income distribution data in both studies relate to an early period, the 1950s through the early 1970s, and therefore are clearly doubtful, although Alesina and Rodrik do experiment with a slightly later set of observations. For a somewhat different sample of countries and slightly longer period, for example, as illustrated in Table 3, I find evidence of no statistical significance for inequality, especially when a dummy is introduced for the Latin American observations.

Table 3.

Determinants of Rate of Growth of Product¹

¹t-values are in parentheses.

²The second regression excludes Germany, Kenya, and Peru.

³The second regression includes Kenya and Peru.

⁴Dummy variable to represent inclusion in Latin America.

Table 3.

Determinants of Rate of Growth of Product¹

	1976–90		1960–80				1960–90
	All countries	Developing countries	All countries2		Developing countries3		All countries2		Developing countries3
Number of countries	63	44	46	43	29	27	46	43	29	27
Constant	−3.16	−2.41	1.37	2.86	3.27	3.11	2.40	2.87	3.46	3.22
	(1.1)	(0.85)	(0.88)	(1.87)	(2.01)	(1.79)	(1.74)	(1.92)	(2.01)	(1.77)
Initial real income	−0.00018	−0.00052	−0.00028	−0.00034	−0.00063	−0.00065	−0.00030	−0.00032	−0.00059	−0.00062
	(1.32)	(1.63)	(2.40)	(3.15)	(1.94)	(1.91)	(2.89)	(3.07)	(1.71)	(1.73)
Initial primary enrollment	0.045	0.048	0.039	0.038	0.026	0.027	0.036	0.036	0.027	0.028
	(2.91)	(2.75)	(3.34)	(3.50)	(2.15)	(2.09)	(3.45)	(3.37)	(2.13)	(2.10)
Share of income in middle quintile	−0.013	0.052	....	....	....	....	....	....	....	....
	(.09)	(.31)
Gini coefficient	....	....	−1.50	−3.91	−3.44	−3.12	−3.77	−4.45	−4.74	−4.26
			(0.55)	(1.40)	(1.20)	(0.97)	(1.55)	(1.63)	(1.56)	(1.26)
Latin America4	−2.47	−1.47	−0.84	−0.76	0.28	0.35	−1.58	−1.49	−0.67	−0.56
	(2.94)	(1.35)	(1.32)	(1.26)	(0.37)	(0.45)	(2.78)	(2.53)	(0.86)	(0.68)
R squared	0.15	0.16	0.16	0.23	0.13	0.12	0.29	0.29	0.21	0.18

¹t-values are in parentheses.

²The second regression excludes Germany, Kenya, and Peru.

³The second regression includes Kenya and Peru.

⁴Dummy variable to represent inclusion in Latin America.

View Table

Table 3.

Determinants of Rate of Growth of Product¹

	1976–90		1960–80				1960–90
	All countries	Developing countries	All countries2		Developing countries3		All countries2		Developing countries3
Number of countries	63	44	46	43	29	27	46	43	29	27
Constant	−3.16	−2.41	1.37	2.86	3.27	3.11	2.40	2.87	3.46	3.22
	(1.1)	(0.85)	(0.88)	(1.87)	(2.01)	(1.79)	(1.74)	(1.92)	(2.01)	(1.77)
Initial real income	−0.00018	−0.00052	−0.00028	−0.00034	−0.00063	−0.00065	−0.00030	−0.00032	−0.00059	−0.00062
	(1.32)	(1.63)	(2.40)	(3.15)	(1.94)	(1.91)	(2.89)	(3.07)	(1.71)	(1.73)
Initial primary enrollment	0.045	0.048	0.039	0.038	0.026	0.027	0.036	0.036	0.027	0.028
	(2.91)	(2.75)	(3.34)	(3.50)	(2.15)	(2.09)	(3.45)	(3.37)	(2.13)	(2.10)
Share of income in middle quintile	−0.013	0.052	....	....	....	....	....	....	....	....
	(.09)	(.31)
Gini coefficient	....	....	−1.50	−3.91	−3.44	−3.12	−3.77	−4.45	−4.74	−4.26
			(0.55)	(1.40)	(1.20)	(0.97)	(1.55)	(1.63)	(1.56)	(1.26)
Latin America4	−2.47	−1.47	−0.84	−0.76	0.28	0.35	−1.58	−1.49	−0.67	−0.56
	(2.94)	(1.35)	(1.32)	(1.26)	(0.37)	(0.45)	(2.78)	(2.53)	(0.86)	(0.68)
R squared	0.15	0.16	0.16	0.23	0.13	0.12	0.29	0.29	0.21	0.18

¹t-values are in parentheses.

²The second regression excludes Germany, Kenya, and Peru.

³The second regression includes Kenya and Peru.

⁴Dummy variable to represent inclusion in Latin America.

Table 3 utilizes observations both about the middle quintile of the distribution, as used by Persson and Tabellini, and the Gini coefficient chosen by Alesina and Rodrik. I have used more recent data to test the former, on the grounds that the income distribution seems to have shown little systematic change over a period as short as two decades; for the latter, I have used initial inequality dating from the 1960s and early 1970s but extended the analysis to a longer time frame.

In each of the three exercises, neither measure of inequality, the share of the income in the middle quintile, nor the Gini coefficient comes out as a statistically significant variable. This is true when all countries are utilized or developing countries only. Moreover, the conclusion is unaffected whether one uses the Gini coefficients selected by Alesina and Rodrik for three countries—Germany, Kenya, and Peru—where there are large differences between potential candidates, or excludes them altogether. The latter regressions involve fewer countries and are reported as well.

Additional conceptual issues about this new approach can also be posed. The mechanisms through which actual redistributive policies operated are not specified. The large place for developed countries in the sample may make the findings for developing countries more dubious. The implicit assumption of equilibrium can hardly hold in an interval in which the debt crisis had already begun and growth was significantly affected. And the model is still not dynamic and really only an imperfect model of reality.

Nonetheless, the Persson and Tabellini results, but not those of Alesina and Rodrik, contribute affirmatively to the notion that efforts combining greater equality with democracy may also yield higher rates of economic performance. They do so in this case because high income inequality motivates populist policies that do not contribute to the higher rates of investment required to ensure adequate growth. In the end, that conclusion conforms to good policy sense rather than being dependent on scientific demonstration. Let me cite the World Bank’s World Development Report 1991 (p. 139):

…efforts to improve equity can sit comfortably within reform programs aimed at promoting growth. It is clear, however, that market-distorting and overzealous redistribution can quickly pose over-whelming financial problems …. Also, crude transfers through market-distorting interventions almost always end up worsening the distribution of income rather than improving it.

The difficulty comes in the practical implementation.

Fazle Hasan Abed

We are gathered here to celebrate the completion of 50 years of the IMF and the World Bank. I feel greatly honored at having been invited to speak today, but first, my warm felicitations to the Bank and the IMF on this occasion. Celebrations are indeed in order today!

Bangladesh, my country, has nearly 120 million people living in an area of 144,000 square kilometers. This makes it one of the most densely populated countries in the world, a country characterized by pervasive poverty. Bangladesh’s GNP per capita is only $220, life expectancy is a low 57 years, and the literacy rate is still a miserable 30 percent.

Yet in this ethnically homogeneous and geographically compact country, as a recent UNICEF report⁶ says, “this grim portrait is slowly being redrawn.” The report notes that per capita income has grown by almost 2 percent a year over the last decade in Bangladesh. Recounting the nation’s progress, it acknowledges that Bangladesh’s fourth five-year development plan incorporates most of the goals adopted at the world summit for children. It mentions that 80 percent of the rural residents are now within 150 meters of a source of safe drinking water, which, according to the report, is a feat unmatched by many richer nations. The report finds that noticeable progress has been made in the field of family planning. The contraceptive prevalence rate has risen from 3 percent in the early 1970s to 40 percent in the 1990s, and, in two decades, the country’s total fertility rate has fallen from 7 births to 4.2 births per woman. In short, the report compliments the progress that is today noticeable in various sectors in Bangladesh. In all this, I dare say, both the IMF and the World Bank, in spite of the reservations one may have about their intervention in certain areas, have contributed in an increasingly effective manner. As these two institutions mature over the years and as they become enriched by their interactions with developmental problems, our faith in their playing an even more positive role in the transformation of the developing world takes firmer root.

At the Bangladesh Rural Advancement Committee (BRAC), we have a conviction that the main thrust for development of a country must come from within, where the poor and the deprived are empowered to alleviate their poverty and, as a result, that of the nation. But, then, we remain mindful that the world community is inseparably linked today through interdependent economies and shared sociopolitical views. Together, we confront such negative factors as malnutrition, illiteracy, the AIDS pandemic, and environmental degradation. For better or for worse, we share planet Earth together, and only through joint action will we overcome global problems. We therefore believe that cumulative actions, by both individual nations and the comity of nations, by institutions national and international, and by organizations great and small, can only accelerate the progress of humanity as we pass into the twenty-first century.

BRAC, a Bangladeshi nongovernmental organization, is but a small stitch in the huge tapestry of the saga of human development. It is to this that I have committed my life, and it is of its hopes, its endeavors, and its possibilities in attaining sustainable poverty alleviation that I should briefly like to speak this morning.

BRAC’s birth in 1972 followed that of Bangladesh. Its activities were then mainly directed at providing basic relief to a people ravaged by war. It soon became apparent that poverty was the biggest problem facing the newly emerged nation. BRAC thus set about designing a strategy to group the landless for their self-sustaining personal development through training, credit, and income generation, as well as through education for their children. Improving the health and nutritional status of women and children and developing and strengthening the capacity of communities to sustain primary health care activity also gradually became part of BRAC’s development strategy. Today, BRAC is the largest national nongovernmental organization in the South, implementing several development programs all over Bangladesh, and employing over 12,000 full-time staff and 33,000 part-time functionaries.

Poverty alleviation, in our view, calls for a holistic approach. Poverty is a complex syndrome that manifests itself in many different ways. It is not only the lack of income that makes one poor but also the lack of good health, education, gender equity, access to resources, and an enabling environment.

It has been our policy to ensure effective participation of the people so that they may acquire the capacity to change their lives. With that end in view, we have given the highest emphasis to organization building, functional education, and training in human development. Our program beneficiaries, now numbering well over 1 million, organize themselves into cooperative groups at the village level. Members of the group undergo a course of functional education aimed at raising their awareness and at developing a sense of self-worth.

A village organization’s initial task is to mobilize local and external resources to generate income and benefits for its members. Fish farming is developed in previously unused bodies of water; roadside verges, unutilized homesteads, and common lands are planted with trees and vegetables to meet the domestic needs of members. The members then turn to other resources in their locality, and their demand for the services of agricultural extension agents, health centers, and veterinary service centers for livestock is intensified.

Generating capital through savings is an important precondition for sustaining a village organization. BRAC group members are required to contribute a small sum weekly to a group savings fund. BRAC’s revolving loan fund provides individual and collective loans to members participating in group activities and making regular savings contributions. A total of $120 million has been disbursed to date as small credits to group members, and their own fund—generated through savings—equals 40 percent of the loan fund outstanding with them.

Time and again, the poor have proved their creditworthiness. The loan recovery rate has been higher than 98 percent. The village organizations are gaining new confidence and strength, and they are venturing into new enterprises to create a more secure future for themselves and their families.

To improve income-generation opportunities for the poor, BRAC has moved into a number of new areas. One such area has been poultry development. We have trained 27,000 poultry vaccinators, 1 per village, who have been linked with the livestock offices. With a regular supply of vaccines from the Government, they provide vaccinations in their villages for a small fee, which has reduced the poultry mortality level. With the assurance of a regular vaccination service, large numbers of poor women have been encouraged to take up poultry rearing as a promising source of income; some 450,000 women have been trained. Hybrid varieties of birds with higher egg-laying capacity are also being introduced. We supply 500,000 day-old chicks every month to chick rearers who, in time, sell these to village poultry rearers. The livestock department, the vaccinators, the chick rearers, and the poultry farmers all form a supportive infrastructure for poultry development.

BRAC, as a leading organization, came to realize that in spite of all our efforts to empower the poor, the bottom 10 percent of the population was not being effectively reached. This, we felt, called for a separate strategy. The poorest among the poor were unable to create self-employment under the prevailing conditions. Employment opportunities needed to be created for them. Something more was needed, and this took us into new ventures. One intervention to reach that 10 percent is a program under which we have already planted 21 million mulberry trees all over the country. This also forms a part of our program target: to raise 40 million trees by 1997. This program is also an essential component in the development of BRAC’s sericulture program, aimed at raising Bangladesh’s silk production by the turn of the century from the current annual level of 60 tons to 1,200 tons and at creating half a million jobs for the most needy rural women.

In response to unacceptably high infant and child mortality rates in Bangladesh, BRAC took up a challenge in 1979—the International Year of the Child—to reduce childhood mortality from diarrhea by half. The efficacy of oral rehydration therapy to combat dehydration was well known among medical scientists. But how would a mother in a remote village in Bangladesh be informed that a solution of salt, sugar, and water, mixed in an appropriate measure, could save her child from dehydration and death? BRAC decided to go nationwide to show mothers face-to-face how to measure and mix an oral rehydration solution with home-based ingredients. It took as many as 1,300 staff ten years to cover 13 million mothers in rural Bangladesh. I am happy to say that oral rehydration therapy is now universally known in rural Bangladesh, and its use has risen to more than 60 percent. The result is that deaths caused by diarrhea have indeed been significantly reduced. Our success with oral rehydration therapy further made us aware of the need to change the national health statistics of the country and inspired BRAC to design a health program evolving through several distinct phases. BRAC’s current Women’s Health and Development Program, involving about 12.5 million people, integrates the various aspects of a community-based program in health, family planning, and nutrition.

Education is a basic element in generating what has been described as “social capital.” The number of unserved children in Bangladesh runs into the millions. Among them, the number of girls is disproportionately high. There had to be a simpler way of tackling the problem of illiteracy! How could we hope to bring this massive population of nonschooled children to school? The simple answer was to situate schools near children’s homes and to appoint a teacher from the immediate locality. Keeping the class size to a maximum of 33 helped the para-teachers with no previous teaching experience, and enrolling 70 percent female students in the school was our natural response to the appalling state of female education in the country. Sufficient parent involvement and participation strengthens the flexible design of the program, especially in setting school timing according to familial needs. Parents also form school management committees in every school, and all parents must meet with the school teacher every month to be informed of their child’s progress. The school curriculum is specially designed to be relevant to the needs of rural children and is under continual revision based on feedback from the field. Working with simple program design, regular school visits by school supervisors, and intensive training and refresher courses for teachers, BRAC has been able to extend the nonformal primary education from 22 schools in 1985 to 24,000 schools today, reaching more than 750,000 children.

Access and equity have been long-standing concerns of education policymakers and practitioners. BRAC’s nonformal primary education undoubtedly provides reasons for optimism—a hope that this may be a way to begin creating the social capital that we so seriously lack and so desperately need.

The limits of government development programs in bringing about a significant reduction in poverty, hunger, and illiteracy in most of the developing countries of Asia, Africa, and Latin America have provided impetus to involvement by nongovernmental organizations in national development. With their grassroots orientation and participatory approach, these organizations have been mobilizing and organizing the poor and the powerless to participate in the development process as subjects, rather than as the passive recipients of fragmented inputs delivered by the state.

Nongovernmental organizations’ relations with their respective governments have, in many cases, stood on sensitive ground. I think a lasting answer to this problem lies in the realization by all concerned that these organizations are there to supplement and not supplant the development efforts that national governments undertake. Governments, however, believing in results achieved through privatization of development efforts, will only find it helpful to support and nurture programs undertaken by the organizations.

The World Bank’s advocacy of structural adjustments calls for a comment. We think that the time has now come for the World Bank to consider making the network of nongovernmental organizations an element of that structural adjustment process. We believe that an authentic reform must direct resources to the rural areas, empower the poor, and democratize the purchasing power. If that is so, nongovernmental organizations can be effective partners in this task, if only because the successful ones among them have over the years developed mechanisms to reach the poorer sections of the community.

There is another point on which I should like to make my concern known. It is on the idea of growth, as promoted by the World Bank. In principle, no one can seriously dispute the need for growth, but to what extent it is “the necessary and sufficient” condition for poverty reduction needs, I think, to be examined carefully. We would do well to remember that growth-based development in East Asian countries has been successful in reducing poverty only in societies in which egalitarian and agrarian reforms preceded the present growth-oriented model of development. Those reforms in the countries of East Asia destroyed the power of the landed elite and thus created the ground for sustainable poverty reduction. In countries such as mine, not having gone through a meaningful land reform, access to the shared growth by the disadvantaged sections of the population remains limited, and it is this access that nongovernmental organizations like BRAC are engaged in developing. The World Bank and the IMF’s advocacy of structural adjustment and policy reform should be aimed at facilitating this process.

Before I conclude, I wish to extend my appreciation to the organizers of this conference and would once again like to congratulate the World Bank and the IMF on the completion of their five decades of active support for the alleviation of poverty around the globe.