Growth and Structural Change
Author: Hollis B. Chenery
Publication Date:
01 Sep 1971
eISBN:
9781616353049
Language:
English
Keywords:
FD; F&D; import substitution; special drawing rights; import; economic development; distribution issue; natural resources; external finance; savings rate; Exports; Personal income; Capital inflows; Development strategy; national income
Download PDF (2.5 MB)

An understanding of the relationship between overall growth and structural change is central to any choice of priorities in development policy. The author summarizes results from several econometric studies of structural change and the source of economic growth, and identifies the principal strategies that have led to successful development in the past 20 years.

Abstract

An understanding of the relationship between overall growth and structural change is central to any choice of priorities in development policy. The author summarizes results from several econometric studies of structural change and the source of economic growth, and identifies the principal strategies that have led to successful development in the past 20 years.

Hollis B. Chenery

Two international groups—the Pearson Commission and the United Nations Committee for Development Planning (Tinbergen Committee)—have recently published blueprints for a more effective effort in international development.1 While these reports differ in scope, they agree on growth targets of at least 6 per cent for the developing countries and on the importance of substantial changes in their economic structures.

Although the relationships between the growth of gross national product (GNP) and structural change are very complex, they may be regarded as interconnected features of the process of development. The economic structure affects the rate of growth that can be achieved over any period of time, and changes in this structure are necessary to sustain continued growth. A development strategy must, therefore, encompass both the mobilization of resources and their reallocation to new uses.

THE NATURE OF STRUCTURAL CHANGE

Empirical studies have shown that rising income is associated with systematic variations in almost every feature of the economic structure—for example, rates of saving and investment, the types of commodity produced and consumed, the patterns of trade, and the use of capital and labor. These normal patterns of development can be traced to similarities among societies in tastes, access to technology and to international trade, and common elements of development policy. Because of these common elements, we can combine the statistics from a large number of countries to measure the normal variation with income level and to identify the effects of other factors such as size, resources, and capital inflow.

There are many familiar examples of the effects of differences in economic structure on a country’s rate of development and need for external capital. Small countries rely much more than large ones on international trade, it affects their growth more and they receive larger capital inflows per capita from abroad. Countries with abundant agricultural or mineral resources to export need aid mainly to supplement domestic savings, while those less favored usually require aid mainly to supplement export earnings. Predominantly agricultural countries find it harder to raise growth rates to 6 per cent than do those in which the share of agriculture in total production has already fallen. Such structural features need to be taken into account in analyzing development targets and strategies.

Table 1 lists some measures of structural characteristics that affect long-term growth and may in turn be affected by it. The table shows the normal variations of each element with income, based on observations for about 100 countries over the postwar period. The method of statistical analysis allows for differences in size and trends over time.

TABLE 1.

NORMAL VARIATIONS IN ECONOMIC STRUCTURE WITH LEVEL OF DEVELOPMENT

Level of GNP per capita (in 1964 U.S. dollars)

article image

Levels for 50, 100, and 1,000 have been adjusted proportionately to total 100 per cent.

All values are computed from multiple regression for a sample of about 100 countries 1950-65. In Chenery, Elkington, and Sims, “A Uniform Analysis of Development Patterns,” (1970), the values shown apply to a country of 10 million population in 1960. Underlying data from the IBRD, World Tables, December 1968.

Growth theories generally emphasize the processes of accumulation (Group I): increasing physical capital and skills. To continue growing, however, a country must not only increase its resources but must also shift them according to changing patterns of demand and opportunities for trade. In fact, many countries find the problems of accumulation easier than the task of transforming the productive structure to balance supply and demand in each sector without excessive strain on the balance of payments. Groups II-IV in Table 1 measure this transformation by changes in the normal composition of production, labor use, and trade at each income level.

My subject here is the structural changes that take place in the course of transforming an underdeveloped country (per capita income below $100) into a developed one (over $1,000). It is convenient to divide this transitional range into an early stage of development ($50 to $300) and a later stage ($300 to $1,000).

Accumulation

In this intercountry comparison, gross savings (line 1) rises from 10 per cent of GNP at the lowest incomes to over 20 per cent for developed countries. The transfer of capital from abroad normally adds 2½ to 3 per cent to domestic resources, so that investment (line 2) normally rises in the early stages from 12 per cent to 20 per cent. Government tax revenue (line 3) increases even more—from 10 per cent to 20 per cent of GNP—between $50 and $300 per capita—and goes on rising. Growing public revenue promotes development through increased current expenditures on education, health, and other productive services as well as through increased investment.

The normal increase in skills can be measured by expenditures on education, by school enrollment, or by literacy. All show that education is now stressed early in development; with per capita income even below $300 a few countries such as Korea and the Philippines have educational attainments approaching the advanced countries. The normal values given in lines 4 and 5 are also among the few structural elements that show a rising trend at lower income levels since World War II. While there is ample evidence that rapid growth requires increased education and skills, statistical estimates of this need are not so accurate as those for physical capital.

As developing countries reach the middle-income level of $300, we find them typically investing one and a half times as much of their national income and educating three times as large a proportion of their children as the poorest countries. Since output can be increased without using appreciably more capital and skills per unit, these higher accumulation rates also lead to higher growth rates if the resources are properly used. Middle-income countries can therefore reasonably be expected to grow more rapidly and can set higher targets for themselves than very poor countries.

A03ufig01
A03ufig01

ECONOMIC STRUCTURE: NORMAL COMPOSITION OF OUTPUT AND OF LABOR FORCE AT DIFFERENT LEVELS OF GNP PER CAPITA

Citation: Finance & Development 8, 003; 10.5089/9781616353049.022.A003

Source: Lines 6 - 12 of Table 1

Transformation

The normal transformation of the economy is shown in Table 1 by the changing composition of output and the labor force. Deviation from these normal patterns is made possible primarily through international trade, which enables the composition of output in small countries to differ substantially from that of domestic demand. On the other hand, large countries typically have much smaller proportional trade and are more closely restricted to balanced growth.2

The existing composition of demand and output also helps to determine the overall growth rate. Growth in the poorest countries is dominated by the potential of agriculture (typically more than half their output), which, more than other sectors, is limited by the possible rate of adoption of new techniques, demand patterns, and organization. Sustained agricultural growth rates much above 5 per cent are rare, whereas in industry 10 per cent growth is quite common. This difference is partly explained by the smaller proportion of increased income that is devoted to increased food consumption, so that the share of manufactured goods in total demand rises. There also appears to be a supply limit to agricultural growth, set by the knowledge and business ability of its many small productive units. Ivory Coast, indeed, appears to be the only country that, starting with an agricultural sector of more than 50 per cent of the gross domestic product (GDP), has sustained a 6 per cent growth rate of its national income for as much as a decade.

As industry and utilities approach agriculture in size, countries become capable of more rapid growth, particularly while domestic manufacturing is substituting for imports. This transformation becomes possible only after the share of food in a rising total expenditure has declined, since poor countries can rarely exchange manufactures economically for imported food.

An example illustrates some of the effects of changing output structure on the potential growth rate of total production. Assuming growth potential of a favorable 4.5 per cent in primary production and 9.5 per cent in industry and utilities over the income range from $50 to $300 per capita, the changing shares of industry, utilities, and primary production shown in Table 1 would increase the growth potential from 5.4 per cent to 7 per cent. Although accelerating growth in this way also requires increased accumulation and expanding trade, enough countries have done it successfully to suggest that growth targets can be increased significantly between the very poor and middle-income levels.

Population, Employment, and Urbanization

The extent of the movement of labor out of agriculture and into industry and services is even more pronounced than the change in the composition of production. The differences in labor productivity between agriculture and other sectors of the economy are most pronounced in the poorest countries (below $100), as shown by the fact that 70 per cent of their labor force is occupied in primary production but produces only 50 per cent of GDP. The lower productivity of labor in agriculture tends to persist in most countries because the economy does not expand fast enough to absorb the accelerated population growth in other sectors. It is only in the later stages of development that surplus labor is removed from agriculture and the productivity of labor in different sectors becomes more equal.

Urbanization is a natural consequence of the changed composition of demand and production and the rising productivity of agriculture. In recent years, however, the acceleration of population growth in almost all developing countries has tended to produce a more rapid migration to cities than could be absorbed by the demands for labor in industry and services.

Rapid urbanization is the dominant feature of structural change in the middle-income levels, with the population of cities doubling every 10 or 12 years. By the time a country achieves a per capita income of $600, half its population lives in cities and the rate of urbanization tends to slow down. Below this level of income, however, it has proved very difficult to prevent urban unemployment from rising unless growth rates of GDP have attained the very high levels of 7 per cent or more.

The aspect of postwar growth patterns that shows the most pronounced shift from earlier periods is, of course, the reduction in the death rates of poor countries. This dramatic success of public health technology has not yet been matched by any significant fall in birth rates at lower-income levels; it is only above $500 per capita income that the rate of population growth typically falls below 2 per cent a year. Several of the structural changes already discussed—particularly education and urbanization—contribute to this desired result.

A03ufig02

NORMAL VARIATIONS IN RESOURCE ACCUMULATION

(In US Dollars)

Citation: Finance & Development 8, 003; 10.5089/9781616353049.022.A003

Source: Table 1. Lines 1, 2, 3 and 5.

SOURCES OF GROWTH

For some time economists have tried to measure the sources of growth in advanced countries and to identify the separate contributions of increases in capital, labor force, skills, and technological change. Similar statistical methods have recently been applied to determine the sources of growth in a number of less developed countries and to compare them with the advanced countries.

While the statistical results are by no means conclusive as yet, they do suggest several significant differences between the determinants of growth in advanced countries on the one hand and less developed countries on the other. In the advanced countries, the principal findings are that (1) growth of capital accounts for less than a quarter of GDP growth and (2) improved resource allocation and technological change typically account for half of the total growth of 4 to 5 per cent.

The main assumption underlying this methodology is that the relative contributions of labor and capital to growth are measured by their shares of GDP, which is more dubious when applied to less developed countries. Most studies of these countries have concluded that, as a result of surplus labor, the contribution of a unit of labor to development is less than its share of income. When allowance is made for this disequilibrium condition, capital accumulation has typically been shown to account for 40 to 50 per cent of total growth in developing countries.

Since capital accumulation is more important, the residual growth unrelated to increased inputs is correspondingly less significant in low-income countries than in ones with higher incomes. These results, although subject to further confirmation, should discourage the uncritical application to the developing countries of conclusions from Europe and the United States that growth can be secured primarily by technological change. The only studies attributing more than 2 percentage points of growth to technological changes relate to Israel, Japan, and Yugoslavia, where half of an overall growth of 8 to 10 per cent is attributable to capital formation.

Another type of statistical analysis has tried to relate growth rates to structural differences among countries. In an intercountry study we can allow for the possibility of balance of payments disequilibrium and the difficulties of changing the composition of output. The main weakness of intercountry estimation is that it assumes that the explanatory factors affect each country in the same way and hence only measure an average effect of each factor.

Several intercountry studies confirm that there are significant differences in measured sources of growth between advanced and less developed countries. At low-income levels, an increase in capital inflow has a marked effect on growth, which depends on the contribution of the inflow to import availability as well as on its contribution to increased investment. This finding is borne out in the comparison of successful development strategies in the next section. Even when investment rates are held constant, there is a tendency for growth rates to increase up to an income level of about $300 but to decline again at high-income levels as population growth is reduced. Finally, the growth of the labor force contributes much less to income growth in poor countries than in advanced ones, reflecting the tendency to surplus labor in the earlier stages of development.

SUCCESSFUL DEVELOPMENT STRATEGIES

Because of the interdependence of the various aspects of structural change, we cannot hope to explain development very adequately as the sum of separate influences, which is the hypothesis underlying most econometric studies. I will therefore supplement these conclusions by examining the development strategies of rapidly growing countries to try to identify the kinds of policy that have led to successful growth.

From 75 countries with per capita incomes of less than $600 (in 1964 prices) for which the World Bank has compiled fairly complete statistics for most of 1950-69, I have chosen 29 that have sustained relatively high growth rates throughout, making some allowance for countries starting with a very large agricultural sector. The average growth rates of 26 of them were at least 5.5 per cent for the period covered; Pakistan, Malaysia, and Peru were added as borderline cases.3

My. analysis of the sources of sustained growth concentrates on three aspects displayed in Table 2: (1) the volume and financing of investment as between domestic savings and external sources (columns 8-11); (2) the role of trade and natural resources (columns 13-16); and (3) the pace of transformation of the economy, as evidenced by the lead or lag of industrialization compared with the normal patterns described in section I (columns 17-20). The last two characteristics were measured by the deviation of a country’s trade and production patterns from the normal for its income level and size.

TABLE 2.

A CLASSIFICATION OF DEVELOPMENT STRATEGIES OF HIGH GROWTH COUNTRIES

article image
article image
Sources: IBRD and Chenery, Elkington, and Sims (1970).

Colunms 17 and 19: 1950 except 1960.

Columns 17 and 19: 1950 except 1955.

Countries having more than 50 per cent of GNP from agriculture in 1950.

Singapore, 1951-60. Goods only.

The basic data are those compiled and circulated in January 1971 by the Economic Program Department, Socio-Economic Data Division, IBRD, under the title World Tables. References are to that volume unless otherwise noted. The data came from the UN. IBRD Country Reports, and national sources. The sample includes all countries having adequate data and a 5.5 per cent or more GDP growth rate plus several borderline cases (see text).Col. 2: 1950 GNP per capita (1960 for Ivory Coast) in 1964 U.S. dollars from Table IV, Col. 17.Col.3: 1960 (midyear) Population, except Jordan, Greece, and Yugoslavia 1961 (Table II, Col. 1).Cols. 4 and 5: Annual Average Growth Rate of Total Gross Domestic Product: at constant market prices whenever possible, otherwise at constant factor cost (Table 1, Col. 2). Data available only to 1967 for Jordan and Nigeria, 1968 for Trinidad and Tobago, Iran, Pakistan, Sudan, and Bulgaria.Col. 6: Annual Average Growth Rate of Population 1951-69 (Table I, Col. 1).Col. 7: Increase in per capita GNP 1951-69 (Table IV, Col. 17).Cols. 8 and 9: Investment Ratios, averages of 1950 and 1955 and of 1960, 1965, 1968, and 1969, respectively, from Table IV, Col. 1 (Gross Domestic Product) and Col. 2 (Gross Domestic Investment).Cols. 10 and 11: Balance of Payments Current Deficit 1951-60 and 1961-68 or 1961-69 as available, from Table V, Col. 8.Col. 12: Ratios for each period were obtained by dividing the gross addition to capital stock by the increase in GDP and the two ratios averaged (Table III, Col. 16). For several countries only one figure for Gross Marginal Capital-Output Ratio was available.Col. 13: Primary Exports/GNP. Per cent primary of total value of exports, derived from UN Yearbook of International Trade Statistics. Primary exports were defined as Food (0). Unmanufactured tobacco leaf (121), Inedible (2 up to 266 Synthetic Fibers), Crude or partly refined oil (33), Natural gas (341.1), Oils and fats (4), Wild animals (941). Per cent exports of GDP from IBRD, Table III, Col. 4. Figures for Iran, Iraq, Trinidad and Tobago, and Venezuela revised to include refined petroleum and petroleum products and Zambia to include copper products.Col. 14: M = manufacturing orientation, B = balanced, P = primary orientation, according to the deviation from the normal proportion of manufactured goods and primary products in exports, as measured in Chenery and Taylor (1968).Cols. 15 and 16: Annual Average Growth Rate of Exports of Goods and Services (Table I, Col. 8) defined to exclude factor and transfer payments to and from abroad.Cols. 17 and 19: Ratio of actual industrial and primary productions to the normal level for a country of the same income a head and size (from Chenery, Elkington. and Sims, 1970).Col. 18: Annual Average Growth Rate of Manufacturing Production 1951-69 from Table I, Col. 7. Generally computed from country indices of manufacturing production published by UN Statistical Office.Col. 20: Annual Average Growth Rate of Primary Production 1950-69. Primary is defined as agriculture plus mining. Per cent shares of GDP were obtained from Table IV, Cols. 6 and 7, and applied against the IBRD total GNP (Table IV, Col. 16).

To evaluate the role of external resources we need to know how far investment in these countries was financed by external capital, whether this external dependence was reduced over time, and how the productive structure was adapted to maintain sectoral balance. Table 2 gives several measures of structural change relevant to these processes.

A preliminary analysis of these data shows quite different patterns of resource use as between those countries (mainly small) with ample supplies of foreign exchange from aid or primary exports and those (mainly large) that lack such external sources. Hence the strategies of successful growth are grouped into four types:

A. High Capital Inflow: More than 30 per cent of investment financed by aid or other foreign sources for at least the first decade.

B. High Primary Exports: Primary exports at least 50 per cent and usually 100 per cent above normal levels. High primary exports seem more important in characterizing this pattern than a high capital inflow, although they sometimes go together.

C. Moderate Capital Inflow: Countries with 10-30 per cent of their investment externally financed during most of the period.

D. Low External Dependence: Countries having none of these sources.

High Capital Inflow

The seven countries that achieved high growth rates using substantial amounts of external capital received support for reasons that are largely political; all these countries, except Korea, are small. They typically started the period in political difficulties and with unfavorable natural resources. Capital, mainly public, of some 5 to 10 per cent of GNP has flowed in for substantial periods. In all, investment in the second decade has substantially increased and dependence on foreign capital diminished while shifting to private sources.

A high development assistance strategy essentially permits a country to expand its most readily growing sectors without worry about short-term balance of payments problems. Generally aid has substituted initially for agricultural production, but both industry and primary production have grown quite rapidly. The normal stage of specialization in primary exports is thus bypassed and manufactured exports (or tourism) develop instead.

Success in a high-aid strategy normally requires a substantial reduction in the dependence on capital inflows after a decade or so. All but Puerto Rico and Jordan have achieved this result by high growth rates of both saving and exports, and all seven countries have grown very rapidly in the second decade.

The high-aid strategy has not always been so successful, but considering the unsettled political and economic conditions in which it has typically been attempted it has turned out remarkably well. Other high-aid countries (and their decade growth rates) are Cyprus (4.5, 4.7), Tunisia (“not available,” 4.4), and Bolivia (“not available,” 5.1). However, the total population of Group A countries is only 50 million, and no large country is now likely to receive such substantial resources from abroad.

High Primary Exports

Development led and financed by primary exports is the traditional colonial strategy and one followed by many noncolonial countries blessed with rich natural resources. Ten countries with a total population of 85 million followed it successfully over the past 20 years. Their natural resources attract private investment, but after several decades amortization and profit remittances typically make the net inflow zero or negative.

Although, like aid recipients, the primary exporters have relatively high imports, their productive structure is strikingly different. Industry is typically lower than normal and primary production significantly higher. Even in the successful countries the growth rate is closely linked to export growth and recently in a few favored cases to tourism.

Sound development by this route eventually requires a country to become less dependent on primary exports and shift toward a more balanced production and trade structure. Successful countries such as Thailand, Trinidad and Tobago, and Iran, have—like high-aid recipients—achieved this reduced dependence by rapid industrialization without lowering the growth rate. A smooth transition is fairly exceptional, however, since most countries delay the needed structural change until demand for their primary exports declines, as in prewar Argentina, Brazil, and Chile and postwar Colombia, Uruguay, and Ghana.

Moderate Capital Inflow

Seven countries have achieved substantial growth with only moderate primary exports and dependence on external capital. This strategy is typical of medium-sized and large countries with primary exports less than 15 per cent of GNP and trade affecting the productive structure less than in the smaller export-oriented countries. Since large countries’ import needs are less, moderate aid may still finance a substantial proportion of them, thus preventing a balance of payments bottleneck from inhibiting their initial accelerated growth. Most countries in this group previously depended heavily on primary exports but have industrialized sufficiently to achieve an average productive structure for their level of income.

Success in moderate-aid countries has been similar to that in high-aid ones, but less spectacular. Both savings rates and exports have normally risen in the second decade to adequate levels for fairly rapid growth. However, the poorer countries in this category will probably still need external finance for more than a fifth of their imports to sustain 6 per cent growth rates.

Since few countries have the option of high-aid or high-primary-export strategies, most have attempted the development strategy outlined here. Medium and small economies depending on primary exports such as Ceylon, Chile, Colombia, Ghana, and Uruguay have found the subsequent transformation to a more balanced structure particularly difficult. Manufacturing offers limited scope for economical import substitution in small countries, so continued growth requires either a shift to manufactured exports or development of new primary exports. In the past few years, Colombia has been actively promoting nontraditional exports of both types.

Low External Dependence

Relatively few countries have been successful in economic development with low levels of both primary exports and external capital. This strategy requires the early development of manufactured exports to cover minimum import needs, a hard task for very poor countries. This more self-sufficient pattern typically requires relatively more than average capital per unit of output. Rapid growth therefore demands high investment rates.

The most successful countries employing this strategy, Japan and Yugoslavia, have achieved high growth rates through very high domestic savings and investment. Japan has long since ceased to be “underdeveloped” but Yugoslavia provides something like a model of the possibilities—and problems—of such development.

Exports in Group D countries have grown very rapidly except in Brazil. Argentina, Burma, and India also attempted self-sufficient development but with notable lack of success because of inadequate stress (until recently) on exports or agricultural development. Aid donors however share responsibility for India’s failure to achieve more rapid growth: they provided considerably less capital per capita than they did to other countries with comparable performances.4

These four development strategies are distinguished primarily by different priorities in shifting resources and so transforming the economy to meet the needs of sustained growth. In the accumulation of capital and skills, successful development patterns look more alike: virtually everywhere rapid growth and a rising savings rate go together, so that either capital inflow can be reduced or the growth rate further accelerated. The experience of slower-growing countries is very different, suggesting that the greater profitability of investment at higher growth rates significantly affects the supply of savings. Experience to date, therefore, shows increased self-reliance accompanying rapid growth; only for slow-growing countries does dependence on aid tend to be perpetuated.

In deciding where to give aid, an important consideration is the attractiveness of rapidly growing countries to private investors. Rapid export growth is the main support of a country’s foreign debt service capacity. Private capital and hard loans therefore have increasingly replaced concessional lending to several countries (the Republic of China, Korea, Thailand) which a decade earlier interested few foreign investors.

OBJECTIVES OF DEVELOPMENT

While economic development can be measured in terms of inputs and outputs, its aim is a transformation of the economy and the society, and the only true measure of success is the degree to which this transformation is achieved. Rates of progress toward this goal largely depend on the starting point. It appears to be as difficult for the poorest countries to achieve 5 per cent growth in GNP as for the $300 country to achieve 7 per cent, and the 6 per cent target for all developing countries might be better conceived in these terms. It may indeed turn out that in more primitive societies growth rates in excess of 5-6 per cent produce social stresses (at present little understood) that are not conducive to sustained growth. If so we should seek not a maximum but an optimum rate of development at each stage.

Possibilities of Accelerated Growth

On the whole, achieving high growth rates has turned out to be easier than was imagined 10 or 15 years ago. Developing countries have already succeeded in devoting notably higher proportions of their resources to education. Over the observed range, capital requirements per unit of output tend to fall (not rise) with rising growth rates. Growth of 7 per cent is frequently achieved by developing countries investing 20 per cent (gross) of GNP or less. Since raising the growth rate usually helps to raise the savings rate, self-sustaining growth becomes easier to finance once it gets under way.

Unlike accumulation, transformation of the economic structure has on the whole been harder than was anticipated. In comparing the countries that did not qualify as “successes” for Table 2 to those that did, I would judge that balance of payments difficulties had caused more failures among countries where some development has taken place than insufficient investment or shortage of skills, although this proposition is difficult to demonstrate statistically.

Aid and primary exports have permitted many countries to succeed by first developing industrial capacity and skills in a rapidly growing domestic market and then going on to manufactured exports. Frequently, the less successful countries could not take the steps needed to shift output from domestic to foreign markets.

Growth vs. Distribution?

It has become almost a cliché that “growth is not enough” and that welfare increases for the poorer groups in the society must be given more importance as development objectives. Since poor countries can transfer little income through fiscal measures, the chief possibility for improving the lot of the poor lies in expanding the quality and quantity of employment provided by the economic system. The income distribution issue then reduces to a question of the nature of the trade-off between growth and employment.

In a number of countries, important programs have been undertaken with a view to improving the quality or quantity of employment opportunities. The most important of these have consisted of agrarian reform programs, including land redistribution, land reorganization, resettlement and colonization schemes. In some countries there have been major programs to improve conditions in the poorer regions (e.g., Northeast Brazil). A few others have tried large-scale employment-creation schemes in conservation, reforestation, and other rural works. There have been relatively few deliberate attempts to encourage the development of more labor-intensive types of production in manufacturing, construction, or other activities. From the evidence available to date, it is difficult to draw any firm conclusions about the extent to which these programs have succeeded in their income redistribution and employment objectives.

While employment is stressed in almost every development plan and international discussion, yet there is surprisingly little evidence on the extent of this tradeoff in practice. The countries where the lowest income groups have most clearly benefited from development in the past 20 years are probably those with the highest growth rates. Given the policy instruments and administration capacity of the less developed countries, I would judge that the employment increases generated by high growth are the most reliable means for maximizing the welfare of the lower-income groups.5 Even the more optimistic 6-7 per cent growth targets of the Committee for Development Planning, however, will hardly prevent unemployment from growing in many countries; greater attention is also needed to the possibilities of substituting labor for capital.

The Allocation of Aid

From the records of the 13 countries in Group A and C of Table 2, it seems clear that where aid has led to rapid growth and higher savings, measuring its ratio to investment underestimates its importance. Conversely, its importance is easily overstated in countries where growth has not been accelerated and aid merely substitutes for domestic savings. The basic test of aid effectiveness therefore is accelerated growth.

When total public assistance is not expanding, more penetrating assessment of performance is needed for allocation of aid. In general, it should be presumed that aid is given to accelerate growth, so that the highest amounts per capita should go to countries with per capita incomes of from $100-300. Over $300, exports and private investment should be able to finance imports, and adequate domestic savings are mainly a question of government policy.

An earlier version entitled “Targets for Development” was presented to the Columbia University Conference on International Economic Development in February 1970. It is published in The Widening Gap, edited by Barbara Ward (Columbia University Press, 1971). I am indebted to Hazel Elkington for research assistance and to Thomas Silcock for editorial help.

1

Partners in Development, Report of the Commission on International Development (1969); United Nations Economic and Social Council, Committee for Development Planning (1970).

2

Chenery, Hollis B. and L. Taylor, “Development Patterns: Among Countries and Over Time,” Review of Economics and Statistics, Cambridge, Mass., U.S.A. (November 1968).

3

Other countries that appear to have had growth rates of this magnitude, but are omitted for lack of comparable data are Albania, North Korea, Libyan Arab Republic, Rumania, Saudi Arabia, and North Viet-Nam. Czechoslovakia, Eastern Germany, and Italy are also omitted since they were already relatively developed in 1950.

4

India’s Second and Third Five Year Plans used a moderate aid strategy; increasingly poor prospects for external capital forced the Fourth Plan (1969-74) to stress more self-sufficiency. Its large size makes India one of the least-favored countries (See Partners in Development, p. 298). Indeed India is perhaps the outstanding example of aid shortages hampering growth. The Indian Government could, however, have done better by favoring agriculture and exports much earlier.

5

Experience of communist and capitalist countries alike in the postwar decades seems to support this generalization. Cuba is perhaps the most notable case in which the poor have gained in employment and welfare with little growth in average per capita income.

Other IMF Content

Other Publishers

Cover Finance & Development
Finance & Development, September 1971
Author: International Monetary Fund. External Relations Dept.
Volume/Issue:
Volume 8: Issue 003
Publisher:
International Monetary Fund
ISBN:
9781616353049
ISSN:
0015-1947
Pages:
67
DOI:
https://doi.org/10.5089/9781616353049.022

ECONOMIC STRUCTURE: NORMAL COMPOSITION OF OUTPUT AND OF LABOR FORCE AT DIFFERENT LEVELS OF GNP PER CAPITA

ECONOMIC STRUCTURE: NORMAL COMPOSITION OF OUTPUT AND OF LABOR FORCE AT DIFFERENT LEVELS OF GNP PER CAPITA

NORMAL VARIATIONS IN RESOURCE ACCUMULATION

(In US Dollars)