Chapter

IV. Globalization and the Opportunities for Developing Countries

Author(s):
International Monetary Fund. Research Dept.
Published Date:
May 1997
Share
  • ShareShare
Show Summary Details

Since the mid-1980s, the pace of globalization in the world economy has quickened considerably. World trade has increased nearly twice as fast as world GDP, financial markets in many countries have been liberalized rapidly, and capital flows to many developing countries have accelerated. Clearly, some economies have benefited from globalization enormously. With this World Economic Outlook, Hong Kong, Korea, Singapore, and Taiwan Province of China (together with Israel) are moved from the developing country group to the new advanced economy group; they vividly demonstrate the great successes that can be achieved when policies take advantage of these forces. But what do the pressures of globalization mean for economic performance and policy orientations in developing countries more generally? Do larger trade flows and more liberal financial markets benefit all countries equally, or are some economies better positioned to reap the gains than others? What do closer trade and financial linkages suggest for the cross-country income convergence process? Are there successive groups of developing countries following in the footsteps of the Asian success cases or not? Are there particular policies that can help countries enjoy the benefits of globalization, enhance economic performance, and reduce the danger of becoming marginalized? These issues are the focus of this chapter.

The chapter begins with a review of the key changes that are taking place in developing countries as the world becomes more integrated—in trade, in financial markets, and in the movement of people. Patterns of growth performance are then highlighted to examine the degree to which per capita incomes have been converging on those in the advanced economies. Many developing economies have shown substantial gains in living standards, and real per capita incomes on average have roughly doubled over the past thirty years. This average gain, however, is no greater than that achieved by the advanced economies, so that on average there has been no convergence of per capita income levels between the two groups of countries; in fact, in absolute terms there has been a widening. Moreover, evidence to be presented below suggests that there has been an increasing polarization among the developing countries. Highly successful developing countries, such as Chile, Malaysia, and Thailand, have been converging toward advanced economy per capita income levels quite rapidly, but many poor countries have been falling relatively farther behind. The reasons for this bipolar pattern and the factors that are associated with faster growth and convergence are then analyzed.

A key lesson seems to be that the pressures of globalization, especially in the past decade or so, have served to accentuate the benefits of good policies and the costs of bad policies. Countries that align themselves with the forces of globalization and embrace the reforms needed to do so, liberalizing markets and pursuing disciplined macroeconomic policies, are likely to put themselves on a path of convergence with the advanced economies, following the successful Asian newly industrialized economies (NIEs). These countries may expect to benefit from trade, gain global market share, and be increasingly rewarded with larger private capital flows. Countries that do not adopt such policies are likely to face declining shares of world trade and private capital flows, and to find themselves falling behind in relative terms.

The analysis then looks at what factors seem to be necessary and sufficient for faster per capita income growth. A main finding is that there are important policy complementarities. It is not just one type of policy that is needed, such as openness to trade, but rather a comprehensive set of policies and reforms that are mutually reinforcing. Finally the chapter looks at the problems of countries that seem in danger of being marginalized and suggests what policies might help put them on paths of higher growth and eventual convergence with more successful countries.

Forces of Integration

Changing Trade Linkages

A striking feature of the growth in world trade and capital flows over the past decade has been the heightened involvement of developing countries. Developing countries not only increased their share of world trade from 23 percent in 1985 to 29 percent in 1995, but they also deepened and diversified their trade linkages. Interdeveloping country trade increased from 31 percent of total developing country trade in 1985 to 37 percent in 1995. Between 1985 and 1995, the share of manufactured products in these countries’ exports increased from 47 percent to 83 percent, which reflects the industrialization process they have been undergoing (Table 16). Despite these overall encouraging developments, there have been wide disparities among the developing countries (Chart 33). Except for countries in Asia and some in Latin America, integration with the world economy has been rather slow. Africa’s share of world trade has continuously declined since the late 1960s, while for the major oil producing countries the share has fallen dramatically since oil prices and revenues peaked in the early 1980s.

Table 16.Advanced Economies Versus Developing Countries Including Newly Industrialized Economies: Diversification of Exports(In percent of merchandise imports or exports)
Advanced Economics (Excluding Newly Industrialized Economies)Developing Countries Plus Newly Industrialized Economics
ImportsExportsImportsExports
197519851995197519851995197519851995197519851995
Nonfuel primary products10.26.85.27.15.64.25.76.15.010.17.45.7
Fuel26.022.48.45.98.93.815.919.97.261.445.411.2
Manufacturers63.870.886.487.085.592.078.474.087.828.247.283.0

Chart 33.Developing Countries and Asian Newly Industrialized Economies (NIEs): Trade1

(In percent of total world trade)

While the shares of world exports of the Asian newly industrialized economics and the rapidly industrializing economics have increased in the past decade, the shares of most other developing country regions have been roughly flat or have declined.

1 Excluding Cyprus and Malta.

2 Excluding Iraq.

3 Excluding major oil exporters.

4 Consists of Chile, Indonesia, Malaysia, and Thailand.

5 Excluding Asian newly industrialized economies, Indonesia. Malaysia, and Thailand.

6 Excluding major oil exporters and Chile.

The expansion, diversification, and deepening of developing countries’ trade linkages have to a large extent been the result of significant changes in trade and exchange regimes. Statist and inward-looking policies of protectionism and import-substitution increasingly have been abandoned in favor of more outward-looking and open policies; trade and exchange regimes have been liberalized, with tariff and nontariff barriers lowered significantly. On the basis of a fairly restrictive definition of openness adopted in one study, 33 developing countries switched from relatively closed to open trade regimes between 1985 and 1995.53 Moreover, many developing countries have committed themselves to further reductions of tariffs and nontariff barriers in the multilateral context of the Uruguay Round. The participation by developing countries in regional trading arrangements, which may entail risks of trade diversion as well as benefits of trade creation, has also increased in the last decade or so.54

More Interconnected Capital Markets

Developing countries are also becoming increasingly integrated with the global financial system. Net private capital flows to developing countries (excluding the Asian NIEs) averaged about $150 billion a year over 1993–96 and almost hit $200 billion in 1996—nearly a sixfold increase from the average annual inflow over 1983–89. In fact, capital flows to one country, China, were larger in 1996 than they were to all developing countries as recently as 1989. These capital inflows roughly doubled in relation to developing country GDP between 1985 and 1996. Unlike in the 1970s and early 1980s when most capital flows represented bank lending, the largest flows in recent years have been in equity and portfolio investments (Chart 34). Such private capital flows rose from a low of ½ of 1 percent of developing country GDP in 1983–89 to 2–4 percent of GDP in each of the years 1994–96. Foreign direct investment has posted the largest rise. This has flowed overwhelmingly toward the emerging market countries that have been experiencing relatively fast economic growth. Asian developing countries received almost twice the net private capital inflows as a percentage of their GDP that African countries received over 1990–96 (Chart 35). Liberalization of financial markets in both recipient and source countries has helped to spur this growing capital market integration. Successful developing countries increasingly have lifted controls on cross-border flows, especially on capital inflows, and removed restrictions on payments for current account transactions. The number of developing countries accepting the obligations to maintain current account convertibility of their currencies under the IMF’s Article VIII has increased from 41 in 1985 to 99 today. With China’s acceptance of Article VIII in late 1996, the proportion of developing country trade carried out under current account convertibility has increased from around 30 percent in 1985 to nearly 70 percent in 1997 (Chart 36). Impressive growth performance and an improved track record in terms of macroeconomic stability by many developing countries, and emerging market countries in particular, also have promoted capital market integration by making these markets more attractive to investors from advanced economies wishing to diversity their portfolios.

Chart 34.Developing Countries: Net Private Capital Flows1

(In percent of GDP)

Overall, capital flows to developing countries have rebounded sharply in the 1990s from the depressed levels of the 1980s. Direct investment has led the way.

1 Excludes major oil exporters. Because of data limitations, these data may include some official flows. Data for 1994 exclude Brazil.

Chart 35.Developing Countries: Net Private Capital Flows, 1990–961

(Annual average; in billions of U.S. dollars)

In proportion to GDP, capital flows to Asia have been running at twice the rate of flows to Africa.

1 Excludes Brazil in 1994.

2 Comprises Argentina, Brazil, Chile, China, Colombia, India, Indonesia, Malaysia, Mexico, Pakistan, Peru, the Philippines, South Africa, Thailand, Turkey, and Venezuela.

3 Comprises Afghanistan, Bangladesh, Benin, Bhutan, Botswana, Burkina Faso, Burundi, Cambodia, Cape Verde, Central African Republic, Chad, Comoros, Djibouti, Equatorial Guinea, Ethiopia, the Gambia, Guinea, Guinea-Bissau, Haiti, Kiribati, Lao People’s Democratic Republic, Lesotho, Liberia, Madagascar, Malawi, Maldives, Mali, Mauritania, Mozambique, Myanmar, Nepal, Niger, Rwanda, Sâo Tomé and Príncipe, Sierra Leone, Solomon Islands, Somalia, Sudan, Tanzania, Togo, Uganda, Vanuatu, Western Samoa, Republic of Yemen, Zaïre, and Zambia.

Chart 36.Advanced, Developing, and Transition Economies: Current Account Convertibility1

(In percent)

The pace of liberalization of exchange regimes in developing economies has quickened in recent years.

1 Percent of advanced, developing, and transition economies that have accepted Article VIII of the IMF’s Articles of Agreement; economics are weighted by their 1990–95 share of aggregate exports of all advanced, all developing, or all transition economies. As of March 31, 1997, a total of 138 countries had accepted Article VIII.

Changes in Employment and Relative Wages

As discussed in Chapter III, relative changes in employment and wages of skilled and unskilled labor in the advanced economies do not appear to have been closely related to increased trade or capital mobility. Instead, studies generally attribute the bulk of the decline in employment or relative wages of unskilled workers in advanced economies to a natural development progression as economies mature. Economic development has typically involved relative shifts of resources and output from agriculture to unskilled labor-intensive manufacturing, to high value-added manufacturing and services. Thus, the shift of employment away from manufacturing appears largely to reflect the forces of technological progress and capital deepening, rather than international trade pressures.55 If advanced economies have flexible labor markets and good adjustment mechanisms, it can be viewed as natural and highly beneficial for them to shift their production to sectors with higher value added per unit of input than low value-added manufacturing.

Similar forces are at work in the developing countries themselves. Many of the developing countries that have integrated into the world economy have, for example, seen their highly skilled workers shift toward their tradable goods sectors, while their unskilled workers have shifted to nontradable sectors, such as construction and transportation.56 This might imply temporary increases in unemployment as these economies adjust to the demands of closer integration into global markets. The rise in developing country incomes, in turn, helps to provide a growing market for some of the high-value-added output of the industrial countries. Increased North-South trade and integration may therefore be expected to lead to increasing prosperity in advanced and developing economies alike as both groups move up to the production of higher-value goods and services. Employment in the advanced economies can remain at high levels as the demand for services increases, provided labor markets are flexible, and employment in developing countries can increase as people shift out of the informal sector into the formal sector. However, the pressures of technology that tend to cause the relative incomes of certain groups of unskilled workers in both country groups todecline must be recognized and addressed by policies in such areas as social safety nets, and education and training.57

Changes in the Movement of People

The flow of people across national borders has also increased as the world economy has become more interconnected, although the flow remains relatively small. In 1990, roughly 120 million people were living in countries in which they were not born, up from 75 million in 1965. The share of the world’s labor force that is foreign born increased by roughly half between 1965 and 1990. Although the largest portion of labor migration is from developing country to developing country, flows from developing countries to industrial countries have accelerated the most over the past two decades. As incomes in poor countries increase and as the wage differential between an advanced economy receiving immigrants and a poor country experiencing net emigration drops to less than about four to one, migration from the latter to the former tends to diminish.58

Labor flows almost certainly would have been greater without the surge in international trade already described. Some researchers have identified a humpbacked relationship between migration and trade, whereby trade helps to establish contacts, information networks, and channels that can lead to a temporary swelling of migration from poor countries to rich countries over the short and intermediate run.59 Over the longer term, however, trade substitutes for the physical movement of capital and labor.

The free flow of factors, including labor, has many economic benefits—the main one being that it helps to maximize global output, promoting efficiency in both labor-supplying and labor-receiving countries. Countries receiving immigrant workers will find that certain production bottlenecks are opened up, which can reduce inflationary pressures, and that aggregate supply is boosted. Countries that receive foreign business managers or technical experts to help develop or manage enterprises are likely to find that the productivity in such facilities improves. Labor-exporting countries are likely to receive foreign resources in the form of worker remittances, which are estimated in one study to have totaled over $70 billion globally in 1995, or other flows.60 These flows have been particularly important for certain countries, including Bangladesh, Pakistan, and the Philippines. Such countries may also find that their overseas workers acquire valuable skills that benefit the domestic economy when the workers return home.

Both groups of countries, however, have concerns about migration. Labor-receiving countries are concerned, for example, that an influx of unskilled workers will reduce wages or employment opportunities, or both, for native unskilled workers. As indicated, however, the forces of technological change are likely to play a bigger role, through deindustrialization and changes in the structure of demand for labor. Meanwhile, labor-exporting countries are often concerned that the loss of human capital, especially skilled labor—including the “brain drain”—may not be compensated by the flow of remittances from workers abroad, with a substantial portion of foreign labor earnings not repatriated or invested in unproductive domestic assets, such as real estate.61

With regard to likely future trends in migration, the economic forces that influence the desire of people to move, including the growing per capita income gaps between successful and unsuccessful countries, might seem to point to the likelihood of a substantial increase in the potential supply of migrants. On the other hand, the downward pressure on low-skilled wages in the advanced economies and the trend toward tighter immigration policies in many countries may tend to limit both legal and clandestine labor flows.

Implications for Relative Income Patterns and Convergence

How have these forces of integration affected cross-country growth and income patterns? In absolute terms, living standards as measured by real per capita incomes have risen substantially in most developing countries over the past thirty years. This is shown in Chart 37, where movements in absolute real per capita income are shown, measured in terms of average 1995 per capita GDP in the industrial countries. Even excluding the successful Asian NIEs, developing countries as a group more than doubled their real per capita income between 1965 and 1995, in line with the industrial countries. Most developing countries experienced substantial economic progress over the period. The gains have been nothing less than spectacular in some countries. Korea, for example, experienced almost a tenfold rise in per capita income between 1965 and 1995, while Thailand saw a fivefold increase, and Malaysia a fourfold rise. In the developing countries of the Western Hemisphere average per capita incomes doubled between 1965 and 1980 before stagnating over the next 15 years, a period much of which was dominated by the debt crisis and its aftermath.

Chart 37.Developing Countries and Asian Newly Industrialized Economies (NIEs): Real Per Capita Income1

(In percent of 1995 industrial country per capita GDP: purchasing power parity terms)

Per capita incomes in most developing country groups have increased since 1965, but progress has been far from uniform.

1 Excluding Cyprus and Malta.

2 Excluding major oil exporters.

3 Consists of Chile, Indonesia, Malaysia, and Thailand.

4 Excluding major oil exporters and Chile.

5 Excluding Iraq.

6 Excluding Asian newly industrialized economies, China, Indonesia, Malaysia, and Thailand.

While the success stories illustrate that dramatic improvements in living standards are possible, many countries regrettably are not realizing their potential. In relative terms, most developing countries have failed to raise their per capita incomes toward those of the industrial countries (Chart 38). In fact, Asia is the only major region to have registered significant relative progress, in the sense of having achieved significant convergence toward industrial country living standards. The four Asian NIEs increased per capita incomes from 18 percent of the industrial country level in 1965 to 66 percent in 1995. For other Asian economies, the gap has also been reduced, with the fastest progress in the 1985–95 period. However, among the groups of countries in the Western Hemisphere, the Middle East and North Africa (MENA) region, and Africa, the gaps have widened since 1965 and especially since the mid-1970s. Western Hemisphere countries, for example, which had almost double the average NIE level of per capita income in 1965, saw the gap between their income level and that of the industrial countries gradually widen after the debt crisis of the 1980s. The average per capita income level of African countries fell in relative terms from 14 percent of the industrial country level in 1965 to just 7 percent in 1995. Africa and Asia roughly exchanged relative positions in this 30-year period. These regional developments in relative income performance seem to parallel the patterns of integration proxied, for example, by shares of world trade (Chart 33).

Chart 38.Developing Countries and Asian Newly Industrialized Economies (NIEs): Relative Economic Performance1

(In percent of current industrial country per capita GDP; purchasing power parity terms)

With the exception of the Asian newly industrialized economies, China, and a group of four industrializing economies, most country groups have not experienced convergence toward per capita incomes in the industrial countries.

1 Excluding Cyprus and Malta.

2 Excluding Iraq.

3 Excluding major oil exporters and Chile.

4 Excluding major oil exporters.

5 Consists of Chile, Indonesia, Malaysia, and Thailand.

6 Excluding Asian newly industrialized economies, China, Indonesia, Malaysia, and Thailand.

There has also been a sharp decline in upward mobility of developing countries within the international distribution of average per capita incomes and an increased tendency for countries to become polarized into high- and low-income clusters. Using the average per capita incomes of developing countries each year to define five income brackets—a lowest quintile for income levels from zero to 20 percent of the richest developing country level, a second quintile for income levels from 20 percent to 40 percent, and so forth—reveals an interesting profile. Of the 108 non-oil-producing developing countries for which data are available, 52 were in the lowest-income quintile in 1965, but the number had increased to 84 countries by 1995 (Table 17). Meanwhile, the number of developing countries in the middle-income categories fell rapidly. In 1965, 49 of these countries had incomes in the second and third income quintiles (between 20 percent and 60 percent of the richest developing country income level), but the number had dipped dramatically to just 21 countries by 1995.62 Simply put, over the past thirty years the vast majority of developing countries—84 out of 108—have either stayed in the lowest-income quintile or fallen into that quintile from a relatively higher position. Moreover, there are now fewer middle-income developing countries, and upward mobility of countries seems to have fallen over time. While there was some tendency for countries to move to higher brackets and to progress relative to the advanced economies over the 1965–75 period, the forces of polarization seem to have become stronger since the early 1980s.

Table 17Developing Countries and Asian Newly Industrialized Economies: Increased Polarization and Reduced Mobility in Cross-Country Relative Income1

(Per capita income in purchasing power partly terms: income distribution is in quintiles)

Countries that started out in the lowest income quintile have generally remained in the lowest quintile.

1965–75
Final position in 1975 income distribution2Number of

Countries
FirstSecondThirdFourthFifth
Initial

relative

position in

1965 income

distribution
First46652
Second423734
Third76215
Fourth22
Fifth145
Number of countries50291478108
1975–85
Final position in 1985 income distribution2Number of

Countries
FirstSecondThirdFourthFifth
Initial

relative

position in

1975 income

distribution
First5050
Second20929
Third1II214
Fourth527
Fifth628
Number of countries71251002108
1985–95
Final position in 1995 income distribution2Number of

Countries
FirstSecondThirdFourthFifth
Initial

relative

position in

1985 income

distribution
First7171
Second1311125
Third63110
Fourth0
Fifth22
Number of countries8417412108
1965–95
Final position in 1995 income distribution2Number of

Countries
FirstSecondThirdFourthFifth
Initial

relative

position in

1965 income

distribution
First501152
Second276134
Third761115
Fourth112
Fifth325
Number of countries8417412108

Excluding major oil exporters, Malta, and Cyprus.

The figure in each cell is the number of countries whose relative position in the initial and terminal year was in the income brackets corresponding to the row and column of that cell. For example, for the period 1965–95, the first row of numbers show that out of 52 countries that were in the bottom one-fifth of the income distribution in 1965, 50 remained in the bottom one-fifth, 1 country moved to the second quintile and 1 country to the third quintile in 1995. Similarly, the numbers in the first column show that out of the 84 countries in the bottom one-fifth of the income distribution in 1995, 50 were in the first quintile, and that 27 and 7 moved down from the second and third quintiles, respectively, or the 1965 income distribution. For a similar analysis, see V. V. Chari, Patrick J. Kehoe, and Ellen R. McGratten, The Poverty of Nations: A Quantitative Exploration, Staff Report No. 204, Federal Reserve Bank of Minneapolis (January 1996).

Excluding major oil exporters, Malta, and Cyprus.

The figure in each cell is the number of countries whose relative position in the initial and terminal year was in the income brackets corresponding to the row and column of that cell. For example, for the period 1965–95, the first row of numbers show that out of 52 countries that were in the bottom one-fifth of the income distribution in 1965, 50 remained in the bottom one-fifth, 1 country moved to the second quintile and 1 country to the third quintile in 1995. Similarly, the numbers in the first column show that out of the 84 countries in the bottom one-fifth of the income distribution in 1995, 50 were in the first quintile, and that 27 and 7 moved down from the second and third quintiles, respectively, or the 1965 income distribution. For a similar analysis, see V. V. Chari, Patrick J. Kehoe, and Ellen R. McGratten, The Poverty of Nations: A Quantitative Exploration, Staff Report No. 204, Federal Reserve Bank of Minneapolis (January 1996).

Income convergence requires that poor countries have faster per capita income growth than rich countries. Given the trends just discussed, it should not be surprising that there is little evidence of such income convergence between developing countries and advanced economies over recent decades. Chart 39 shows a scatter diagram of average per capita income growth rates over the period 1965–95 plotted against initial (1965) per capita income levels for both advanced and developing countries. If incomes in these countries were tending to converge toward some global average, there would be few data points in the northeast and southwest quadrants—the richest countries would not be growing faster than average, and the poorest countries would not be growing more slowly than average. However, as the diagram makes clear, there has been no such convergence trend. Even excluding the advanced economies from the diagram, to test whether there has been convergence just among the developing countries and newly industrialized economies themselves, suggests no such tendency (Chart 40).

Chart 39.Advanced and Developing Economies: Convergence in Per Capita Income, 1965–951

(In purchasing power parity terms)

The fact that many economies are in the southwest quadrant—lower than average per capita income in 1965 and slower than average growth over 1965–95—suggests a lack of convergence.

1 Excluding Iraq, Kuwait, Lebanon, and Qatar; as a percent of industrial country per capita income in 1995. A necessary condition for convergence is that economics be concentrated in the upper left and lower right-hand quadrants.

Chart 40.Developing Countries and Asian Newly Industrialized Economies: Convergence in Absolute Income, 1965–951

(In purchasing power parity terms)

Even among just the developing countries and the newly industrialized economies there is no evidence of convergence over 1965–95.

1 Excluding Cyprus, Lebanon, Malta, and the oil exporting countries; as a percent of industrial country per capita income in 1995. A necessary condition Tor convergence is that countries are concentrated in the upper left and lower right-hand quandrants.

This lack of cross-country income convergence may be surprising because there are many reasons to expect a converging pattern, especially in a more open and integrated world economy. First, there are wide technology gaps between advanced economies and developing countries, giving the latter great potential for technological catch-up. With open trade and liberal financial markets, poorer countries should be able to benefit from technology spillovers, such as through the stock of knowledge embedded in imported capital goods. Second, capital-to-labor ratios in developing countries are lower than in advanced economies, and other things equal this relative scarcity of capital might be expected to make the return from investment in the former higher than in the latter. In a world in which capital is free to flow in search of highest returns, there are therefore grounds for expecting that it might increasingly flow to developing countries where it can help boost income growth. These forces should work to promote productivity and income growth in developing countries and should therefore increase the likelihood of convergence.

So, given the empirical evidence, what remains of income convergence? The data appear to show that there is a tendency for countries to converge to long-term per capita income levels that are determined by their own policies and resources. In cross-country analyses of growth, factors that have been found to be important in contributing positively to long-run potential per capita income include the skill level of the workforce, the absence of distortions affecting investment decisions, the degree of openness of the economy, macroeconomic stability, and freedom from political and civil unrest. A country’s rate of convergence then depends upon these factors and the gap between its initial and potential income levels. The larger the gap, the faster the rate of growth, but the longer the time taken to converge.63 China, for example, would take about 16 years to cut in half its current income gap with the advanced economies if it maintained its 10 percent a year real per capita growth rate of recent years (Table 18). Chile, even though it has been growing at less than half the rate of China, would halve its gap with the advanced economies in only ten years if it maintained its recent growth rate, because it has a higher current level of income. These two cases illustrate the point that although most developing countries are not converging toward advanced economy levels of income, there are some cases where growth conditions and policies are highly favorable, and where progress toward convergence can be achieved in a relatively short period of time.

Table 18.Developing Countries: Convergence and Growth in Selected Countries1(in percent)
Relative Per Capita IncomeAverage Growth Rate of Relative Per Capita Income,Average Rate of Convergence,Average Growth Rate of Relative Per Capita Income,Average Rate of Convergence,Years to Close Half the Gap in 1995 at 1990–95Implied Rate ofImplied Relative Per Capita Income After Halving of 1995 Gap
198519951985–951985–9521990–951990–952Growth RateConvergence2
Chile33.646.73.31.34.51.8102.673.3
Indonesia13.718.83.20.55.00.8231.759.4
Malaysia39.948.01.90.85.22.283.174.0
Thailand20.536.15.81.66.01.8113.068.0
Argentina33.531.4-0.6-0.22.80.8261.365.7
China7.313.36.20.69.51.0162.756.6
India35.86.91.70.11.90.11120.453.4
Sri Lanka14.515.80.90.12.811.4460.957.9
Uganda3.56.97.20.34.00.2520.953.5
Uruguay32.839.31.80.62.00.7291.069.6
Bangladesh6.66.90.40.031.50.11410.353.4
Vietnam0.70.92.70.024.80.04870.650.4
Memorandum
Asian newly industrialized economies456.486.74.72.94.63.4

All growth rates, rates of convergence, and relative per capita income levels are in relation to the average of advanced economies (excluding the newly industrialized economies). For example, the relative per capita income for any country (shown in the first two columns) is the ratio (in percent) of its per capita income to the average per capita income of advanced economies (excluding the newly industrialized economies). Income refers to GDP in U.S. dollars, based on purchasing power parity exchange rates. The simulations, summarized in the last three columns, assume that the industrial countries as a group maintain their current average rate of growth.

The rate of convergence is defined as the percentage of the gap below the average per capita income of advanced economies (excluding the newly industrialized economies) that is reduced per year.

For India, real per capita GDP growth in 1995 and 1996 has been significantly higher than the average for 1990–95. The average growth rate of relative per capita income in 1995–96 was 3 percent. If India maintains this growth rate then the number of years required to close half its gap with the advanced economies (excluding the newly industrialized economies) will be lowered to 69 years from 112 years as shown in the table.

In 1995. Hong Kong’s relative income was 114.6 percent. Korea’s 54.6 percent. Singapore’s 105.6 percent, and Taiwan Province of China’s 72.1 percent of the average per capita GDP of advanced economies (excluding the newly industrialized economies).

All growth rates, rates of convergence, and relative per capita income levels are in relation to the average of advanced economies (excluding the newly industrialized economies). For example, the relative per capita income for any country (shown in the first two columns) is the ratio (in percent) of its per capita income to the average per capita income of advanced economies (excluding the newly industrialized economies). Income refers to GDP in U.S. dollars, based on purchasing power parity exchange rates. The simulations, summarized in the last three columns, assume that the industrial countries as a group maintain their current average rate of growth.

The rate of convergence is defined as the percentage of the gap below the average per capita income of advanced economies (excluding the newly industrialized economies) that is reduced per year.

For India, real per capita GDP growth in 1995 and 1996 has been significantly higher than the average for 1990–95. The average growth rate of relative per capita income in 1995–96 was 3 percent. If India maintains this growth rate then the number of years required to close half its gap with the advanced economies (excluding the newly industrialized economies) will be lowered to 69 years from 112 years as shown in the table.

In 1995. Hong Kong’s relative income was 114.6 percent. Korea’s 54.6 percent. Singapore’s 105.6 percent, and Taiwan Province of China’s 72.1 percent of the average per capita GDP of advanced economies (excluding the newly industrialized economies).

Policies to Boost Growth and Promote Convergence

In the light of growth patterns that show many developing countries diverging away from advanced economy per capita income levels over recent decades, the question of how developing countries’ growth performance could be improved gains urgency. What are the sources of economic growth and what policies could make a difference in whether a country converges toward high income levels? In simplest terms, economic growth springs from the accumulation of physical and human capital, labor, and advances in production technology (total factor productivity). Although views differ on the relative importance of these factors (see Box 9), for most developing countries conventional growth-accounting studies show that the accumulation of factors, especially physical capital, has accounted for the greater part of output growth. Recent estimates suggest that in the period 1960–92 roughly 60 to 70 percent of growth in per capita incomes was due to increases in physical capital per worker, while education contributed about 15 to 20 percent, and total factor productivity accounted for the remainder.64 A comparison of fast- and slow-growing developing countries over the periods 1965–85 and 1985–95 shows that the shares of both investment and saving in GDP have been significantly higher for the first group (Table 19). It appears therefore that policies that raise the rates of investment and saving can play a crucial role in raising growth, if the investment is productive.65 This section considers what role policies can play in boosting capital accumulation and total factor productivity.

Table 19.Developing Countries and Asian Newly Industrialized Economies: Policies and Economic Performance1
Low Growth2Medium GrowthHigh Growth
1970–841985–951970–841985–951970–841985–95
Initial conditions
GDP per capita in initial year31,6972,1852,2662,1881,7762,734
Human capital42.23.33.23.83.55.4
Macro conditions
Saving517.816.518.519.226.031.4
Investment519.019.422.121.127.431.9
Inflation rate per year (median)11.014.110.911.111.37.8
Fiscal conditions
Fiscal balances5-5.7-5.6-4.2-3.3-2.0-2.4
Government expenditure519.925.019.620.019.518.9
Government revenue514.219.415.416.717.616.5
Monetary conditions
Money + quasi-money533.038.428.736.425.664.9
Quasi-money516.524.313.822.910.036.6
Bank credit to the private sector520.425.418.531.021.063.1
International
Net private capital flows620.211.812.919.966.968.3
Balance on current account5-1.0-2.6-3.7-1.4-1.90.3
Exports511.317.214.917.218.233.0
Imports512.217.717.418.119.532.4

Excludes major oil exporting countries, Cyprus, and Malta.

Low growth is defined as per capita real income growth of less than one-half of 1 percent a year, which is roughly the mean growth rate minus one-half of the standard deviation of growth in the sample for the specified period. Correspondingly, high growth refers to rates above the mean plus one-half of the standard deviation (2.9 percent).

Group average in U.S. dollar terms, using purchasing power parity weights.

Average schooling years in population aged 15 and over. See Robert J. Barro and Jong-Wha Lee, “International Measures of Schooling Years and Schooling Quality,” American Economic Review, Papers and Proceedings, Vol. 86 (May 1996), pp. 218–23.

In percent of GDP.

In percent of total private capital flow to developing countries. Excludes Asian newly industrialized economies.

Excludes major oil exporting countries, Cyprus, and Malta.

Low growth is defined as per capita real income growth of less than one-half of 1 percent a year, which is roughly the mean growth rate minus one-half of the standard deviation of growth in the sample for the specified period. Correspondingly, high growth refers to rates above the mean plus one-half of the standard deviation (2.9 percent).

Group average in U.S. dollar terms, using purchasing power parity weights.

Average schooling years in population aged 15 and over. See Robert J. Barro and Jong-Wha Lee, “International Measures of Schooling Years and Schooling Quality,” American Economic Review, Papers and Proceedings, Vol. 86 (May 1996), pp. 218–23.

In percent of GDP.

In percent of total private capital flow to developing countries. Excludes Asian newly industrialized economies.

Macroeconomic Stability

By reducing uncertainty, macroeconomic stability allows investment and saving decisions to be made in a manner consistent with underlying economic fundamentals, thereby promoting an efficient allocation of resources. Macroeconomic stability also boosts confidence, which can encourage domestic investment and the inflow of foreign capital. Over 1985–95, median inflation was about 8 percent a year and fiscal deficits about 2 percent of GDP among the fastest-growing developing countries, while in the slowest-growing economies median inflation was about 14 percent a year and budget deficits averaged about 6 percent of GDP (Table 19). Empirical studies using large samples of country experiences suggest that the effect of inflation on growth becomes increasingly negative as inflation rises through some range, which some researchers put in the neighborhood of 8 percent a year, and that the relationship may be nonlinear.66 During episodes of low-to-moderate inflation, the effect of marginally higher or lower inflation on growth may be small, but high rates of inflation tend to have significant, negative growth effects. Large and persistent budget deficits also may slow growth, because they tend to reduce the supply of loanable funds for the private sector and crowd out private investment. Among many examples, an extreme case is Argentina, where growth rebounded from −6 percent in 1989 to nearly 9 percent in 1993 during a period when annual inflation was reduced from a rate close to 5,000 percent to 18 percent. Although it is difficult to establish from the data a close relationship between large fiscal deficits and low growth, many countries that have experienced low growth have also had large fiscal deficits. Countries such as Chile and Uganda, where strong fiscal adjustments were undertaken during the 1980s to promote macroeconomic stability, saw output shift to markedly steeper paths.67

Box 9.Measuring Productivity Gains in East Asian Economies

In recent years, many economists have attempted to identify how much of the rapid economic growth in the East Asian region has been due to productivity growth and how much to the growth of factor inputs, These growth-accounting exercises have aimed to unearth the process underlying the impressive success stories of such economies as Hong Kong, Korea, Singapore, and Taiwan Province of China since the 1970s and Indonesia, Malaysia, and Thailand more recently. Identifying the growth process is important not only in evaluating the role played by policies but also in assessing the growth prospects of these economies and the lessons for others. However, largely as a result of the wide variety of accounting methodologies and empirical techniques, which are subject to a high degree of arbitrariness, little consensus has emerged regarding the relative importance of productivity growth vis-à-vis resource accumulation in accounting for the growth rates. While some studies find that almost all the growth in these countries can be attributed to unusually high rates of resource mobilization, leaving little to be explained by gains in productivity, others conclude that increases in productivity have been as high as 4 percent a year, thus accounting for a large proportion of output growth (see table).

To those who find little evidence of significant growth in factor productivity, the sustained high rates of output growth in the East Asian region are explained by the large increases in the use of inputs during the period.1 They are more pessimistic about the prospects for continued growth in these countries at the high rates witnessed in the last two decades, since that would require continued high rates of resource mobilization, including the maintenance of domestic saving at over 30 percent of GDP and the availability of increasingly skilled labor. Others, however, are more optimistic about the future economic performance of these countries, given the evidence of productivity growth that they find, since if this continues, relatively high rates of output growth can be sustained even with lower rates of factor accumulation.

In general, a growth-accounting exercise deducts from output growth measured, for example, as the annual change in real GDP, a weighted average of the changes in aggregate physical and human capital and labor inputs and then interprets the residual as the growth of total factor productivity (TFP). There are two sets of problems associated with this approach that are at the root of the large variations in the estimates of productivity growth found in the various studies. Measures of the stock of aggregate physical capital are not only difficult to obtain but are also unreliable; they are generally constructed from historical investment data, using simplifying and to some extent arbitrary assumptions about the quality and depreciation of capital. Proxies for human capital, such as the average number of years of schooling of the labor force, the proportion of the labor force with higher education, and so on, are not only largely arbitrary measures but they also fail fully to reflect cross-country differences in the quality of education. Similar data-related issues arise in measuring employment. Moreover, there is little uniformity in the classification and collection of such data across countries.

The second type of problem arises in determining the appropriate weight to attach to the growth of a factor in assessing its contribution to overall output growth. In principle, the weight should be the elasticity of output with respect to the factor concerned—that is, the proportionate increase in output when an additional unit of the particular factor is used in production. But this cannot be measured directly, and estimating it is not straightforward. For example, estimation using regression techniques generally suffers from the assumption that the weights are constant over the estimation period. In fact, however, the relative importance of different factors may well change over time, especially when an economy undergoes rapid transformation, as has been the case in the East Asian region. An alternative and more frequently used approach has been to approximate the output elasticity of a factor by its share in national income. This approximation is valid if factors are paid their marginal products; but this requires that there are no increasing returns to scale or externalities in the use of any factor, that technological progress is not embodied in factor inputs, and that all input and product markets are perfectly competitive. In fact, however, many industries are characterized by increasing returns, and factors such as human capital generate strong externalities. Furthermore, technological progress is often embodied in new inputs of capital and labor,2 while the large profit markups3 observed in some product markets do not support the assumption of perfect competition. In the presence of such markups in product markets, productivity growth, computed as the residual component of growth after the contributions of the different factors have been accounted for, can be overestimated. Moreover, empirical evidence indicates that the residual component of growth is also correlated with demand-side variables, such as monetary and fiscal policies,4 so that interpreting the residual as purely productivity growth may be misleading. Also crucial is the time period for which the analysis is conducted. During periods of rapid growth, the average rate of TFP increase has been disproportionately higher than during low growth periods.5 In sum, given the large degree of arbitrariness in measuring inputs and their effect on output and given the differences in the time periods for which the accounting exercises have been performed, it is not surprising that estimates of TFP growth for countries in East Asia vary enormously (see table).

Selected Developing Countries and Newly Industrialized Economies in Asia: Estimates of Total Factor Productivity Growth(in percent a year)
Bosworth and CollinsBosworth and Collins
Young (1995)(1996)(1996)Sarel (1995)Sarel (1996)
1966–901960–941984–941975–901979–96
Hong Kong2.33.8
Korea1.71.52.13.1
Singapore0.21.53.11.92.5
Taiwan Province of China2.62.02.83.5
Indonesia0.80.90.9
Malaysia0.91.42.0
Philippines0.4-0.9-0.9
Thailand1.83.32.0
Sources: Alwyn Young, “The Tyranny of Numbers: Confronting the Statistical Realities of the East Asian Growth Experience,” Quarterly Journal of Economics, Vol. 110 (August 1995), pp. 641–80; Barry Bosworth and Susan M. Collins, “Economic Growth in East Asia: Accumulation Versus Assimilation,” Brookings Papers on Economic Activity: 2 (1996), pp. 135–203; Michael Sarel, “Growth in East Asia: What We Can and What We Cannot Infer From It,” IMF Working Paper 95/98 (September 1995); and Michael Sarel, “Growth and Productivity in ASEAN Economies.” paper presented at the Conference on “Macroeconomic Issues Facing ASEAN Countries,” held in Jakarta, Indonesia on November 6–8, 1996.
Sources: Alwyn Young, “The Tyranny of Numbers: Confronting the Statistical Realities of the East Asian Growth Experience,” Quarterly Journal of Economics, Vol. 110 (August 1995), pp. 641–80; Barry Bosworth and Susan M. Collins, “Economic Growth in East Asia: Accumulation Versus Assimilation,” Brookings Papers on Economic Activity: 2 (1996), pp. 135–203; Michael Sarel, “Growth in East Asia: What We Can and What We Cannot Infer From It,” IMF Working Paper 95/98 (September 1995); and Michael Sarel, “Growth and Productivity in ASEAN Economies.” paper presented at the Conference on “Macroeconomic Issues Facing ASEAN Countries,” held in Jakarta, Indonesia on November 6–8, 1996.

The low rates of TFP growth in the fast-growing East Asian economies found in some studies contrast with the finding that it has been the main contributor to the output growth of the industrial countries in recent decades.6 However, it may not be surprising if the relative importance of the different sources of growth changes in the process of economic development. Economic growth in the nineteenth century in the United States appears to have been largely due to increases in inputs rather than productivity growth7 as was the growth of the Japanese economy between the Meiji Restoration and World War I. The experiences of the more advanced economies suggest that the accumulation of physical capital is an important source of growth in the early stages of economic development, but that once a relatively high level of capital intensity (capital-to-labor ratio) is reached, technological progress takes over as the principal source of growth.8 Capital intensities in the economies of the East Asian region, including even the newly industrialized economies, are still significantly lower than they were in the industrial countries in the early 1960s. Given these relatively low capital-to-labor ratios, there is still plenty of room for input-based growth among fast-growing countries in the region. Consequently, even if productivity gains may not have contributed as much to growth as some studies suggest, the future growth prospects of the East Asian economics remain bright, and both their rapid rates of accumulation of capital and the high levels of efficiency with which it has been allocated remain impressive accomplishments.

1See Paul Krugman, “The Myth of Asia’s Miracle,” Foreign Affairs, Vol. 73 (November–December 1994), pp. 62–78.2See Paul M. Romer, “Crazy Explanations for the Productivity Slowdown,” NBER Macroeconomics Annual, 1987, ed. by Stanley Fischer (Cambridge, Massachusetts: MIT Press, 1987), pp. 163–203.3There have been few studies that estimate markups for industries in the East Asian region. In industrial countries, which generally have a more competitive environment, in almost all industries markups are positive and large, for example, see Joaquim O. Martins, Stefano Scarpetta, and Dirk Pilat, “Mark-Up Ratios in Manufacturing industries: Estimates for 14 OECD Countries,” OECD Working Paper No. 162 (Paris: OECD, 1996).4See, for example, Charles L. Evans, “Productivity Shocks and Real Business Cycles,” Journal of Monetary Economics, Vol. 29 (April 1992), pp. 191–208.5Sec Arnold C. Harberger, “Reflections on Economic Growth in Asia and the Pacific,” Journal of Asian Economies, Vol. 7, No. 3 (1996), pp. 365–92.6See Barry Bosworth. Susan M. Collins, and Yu-Chin Chen, “Accounting for Differences in Economic Growth,” Brookings Discussion Papers on International Economics, No. 115 (October 1995), pp. 1–630.7Moses Abramovitz and Paul A. David, “Reinterpreting Economic Growth: Parables and Realities.” American Economic Review, Vol. 63 (1973), pp. 428–39.8See Lawrence J. Lau, “The Sources of East Asian Economic Growth,” Stanford University Working Paper (Palo Alto, California: Stanford University: November 1996).

Openness

Policies toward foreign trade are among the more important factors promoting economic growth and convergence in developing countries.68 With open trade, domestic prices reflect world prices, thereby promoting the efficient allocation of resources. Open trade and capital account policies not only allow a country to exploit its comparative advantages in production, but they also promote the importation of lowest-cost products, often with embedded advanced technology. Trade also allows a country to employ a larger variety of intermediate goods and capital equipment that enhance the productivity of its own resources. Such spillovers of advanced technology into developing countries provide a key mechanism for productivity catch-up with advanced economies.69 The strong correlation between policies fostering trade openness and fast economic growth is evident in Table 19. Over the period 1985–95, the developing countries that achieved the fastest economic growth were the countries that as a group had the highest ratios of imports and exports to GDP. Medium- and low-growth countries had import and export ratios that were roughly half as large as the fast-growing countries. Chart 41 shows that the group of countries that substantially liberalized their trade over the period 1988–92 experienced a sharp pickup in both exports and imports, and a noticeable increase in their absolute income levels.

Chart 41.Selected Economies: Exports, Imports, and Per Capita Real Income1

(Simple group averages; in percent of GDP, unless otherwise noted)

A group of countries that liberalized trade over the period 1988–92 has experienced a sharp pickup in both exports and imports and an increase in per capita income levels.

1 Comprises Argentina, Brazil, China, India, Indonesia, Kenya, Mexico, the Philippines, Sri Lanka, Turkey, Uganda, and Uruguay. Among the countries that undertook major trade reforms in the period 1988–92, these were the 12 largest in terms of GDP.

2 Percent of 1995 industrial country per capita GDP (in purchasing power parity terms).

Role of State-Owned Enterprises

Excessive state intervention in the economy limits the role of the private sector in economic activity, and administrative controls in product and financial markets distort resource allocation. The result is often low investment and growth, and poor investment quality. Among many developing countries, the direct involvement of the state in economic activity is large and widespread, with state-owned enterprises (SOEs) having monopoly rights in a large number of sectors, including manufacturing and the financial sector (see Box 10). While SOEs in developing countries generated 11 percent of GDP on average in the period 1978–91, in industrial countries their involvement was limited to about 5 percent. In countries such as Ethiopia, Somalia, Sri Lanka, and Tanzania, the SOE share of manufacturing has been in excess of 30 percent.

While the public sector’s share of the economy has been large in many of the poorer developing countries, many of the more successful developing countries also have had large shares of manufacturing under state control. However, there has been a significant reduction in the involvement of state-owned enterprises over the years in the more successful cases. For example, in Taiwan Province of China, the share of such enterprises in manufacturing output fell from 56 percent in 1952 to about 21 percent in 1970, and by 1990 was down to 11 percent.70 The SOE share of manufacturing output in China also has been dropping rapidly. However, the SOE share of financial resources remains large in China and could complicate macroeconomic policy and impede growth unless it is reduced quickly. Often state-owned enterprises are operated inefficiently, and despite their monopoly status they tend either to make low profits or to run persistent and large losses that burden government budgets. In India during 1989–94, for example, after-tax profits as a percent of total sales were roughly four times higher among private sector companies than in state-owned enterprises in comparable sectors.71 Market-oriented structural reforms that limit state intervention to areas of genuine market failure (such as health, education, and infrastructure) and improve the efficiency of government may be expected to boost growth by reducing distortions and encouraging greater private sector participation.

Box 10.Stabilization and Reform of Formerly Centrally Planned Developing Economies in East Asia

Several countries in East Asia, including Cambodia, China, the Lao People’s Democratic Republic, and Vietnam, while undergoing development, are also in the process of moving from central planning toward economic systems based largely on market principles. 1 The experience of these countries differs substantially from that of the group of transition countries considered in Chapter V, reflecting both differences in the structures of their economies and in the nature of the process of transformation being undertaken.2

Though these countries differ greatly in size of population, they share many similarities in terms of economic structure, notably a population that remains largely rural and predominantly employed in agriculture. In Cambodia, the Lao People’s Democratic Republic, and Vietnam, although the share of industry is increasing, the economies remain dominated by a family-based, rather than collectivized, agricultural sector (see table). The private sector remained active throughout the period of central planning in these three countries, and state-owned industries generally accounted for a relatively small part of the economy. This contrasts with the transition countries of eastern Europe and the former Soviet Union, where urban-based industries typically employed the majority of workers and accounted for the largest part of output, and where state-owned enterprises constituted a dominant share of the economy before the transition. This is true even for Mongolia, which as a result of tight economic alignment with the former Soviet Union is characterized by an urbanized population and an industrial structure more similar to the transition countries than to other newly emerging market economies in Asia. In China, although the economy until recently remained largely rural and the workforce concentrated in agriculture, central planning was more deeply embedded than in the other East Asian countries, with virtually no private enterprise before the initiation of reforms.

The particular structural characteristics of the East Asian economies have led to substantial differences in the evolution of their economies as compared with the countries in transition. Most striking has been the absence of sharp output drops of the kind suffered by the transition countries, although growth did slow at the start of reforms in each country and briefly turned negative in the Lao People’s Democratic Republic. Output is now growing at annual rates of 7–10 percent in all four countries. The absence of large output drops reflects, to some extent, the low initial level of output in the East Asian countries, partly as a result of war and civil conflict. But it is also attributable partly to the less pervasive nature of the initial economic distortions and hence of the transformation. In the East Asian economies, the transformation has consisted primarily of relatively narrowly focused economic reforms without accompanying political and social changes, and partly as a result there has been much less uncertainty and short-term instability than in many of the transition countries. The smaller share of industry, particularly in terms of the labor force, and the relative absence of a capital-intensive manufacturing base further meant that the East Asian countries did not face the challenge of replacing a massive amount of capital suddenly made obsolete by the end of central planning. Instead, the transformation has involved the establishment of conditions that would permit a shift from agriculture to industry.

These countries were also less closely integrated into the trading arrangements of the former Council for Mutual Economic Assistance (CMEA), and thus did not suffer to the same degree as the other members from its collapse, although Cambodia, the Lao People’s Democratic Republic, and Vietnam all experienced both losses of financial assistance from the former Soviet bloc and declines in their terms of trade as prices of imports from the transition countries, such as energy and raw materials, adjusted upward to market levels. This limited degree of integration into the trading system of the other former centrally planned economies is another factor setting the East Asian countries apart from the transition economies of central Asia, which were far more severely disrupted by the breakdown of intraregional flows of raw materials with the collapse of central planning. The East Asian countries also benefited to some degree from their geographical proximity to the newly industrialized and other fast-growing economies of the region, which has led to investment in joint ventures and wholly owned subsidiaries formed to take advantage of expanding markets and low wages.

The process of reform began in China in 1978, in the Lao People’s Democratic Republic in 1979 with comprehensive reforms starting in 1986, in Vietnam in 1985 with comprehensive reform starting in 1989, and in Cambodia in the early 1990s. At the start of their reforms, Cambodia, Vietnam, and the Lao People’s Democratic Republic had only recently emerged from isolation or war. Initial reforms in the East Asian countries broadly followed the Chinese model in being principally aimed at boosting productivity, particularly in agriculture where reforms such as the undoing of collectivization proceeded most rapidly. At the same time, nonstate enterprises were allowed to expand by absorbing surplus labor from agriculture, while other reforms including the liberalization of regulations on joint ventures and foreign direct investment have been introduced more gradually. Although productivity rose in agriculture as a result of the reforms, the striking difference between the share of the labor force in agriculture and the share of GDP produced by agriculture highlights the very low productivity levels in that sector, and thus the scope for continued industrialization. This remains particularly evident for China despite its head start in implementing reforms, possibly because its economy had been the most centralized, with the most distorted structure of production and prices.

Comparison Between Developing Countries in East Asia Moving Away from Central Planning and Selected Transition Countries in Eastern Europe and the Former Soviet Union
199319901994
Share of GDP inShare of Labor Force inPercent Urban
AgricultureIndustryAgricultureIndustryPopulation
Developing countries moving away from central planning
Cambodia5174720
China2048721529
Lao People’s Democratic Republic511878621
Vietnam2928711421
Selected transition countries
Albania4013552337
Czech Republic63573565
Moldova3548333051
Mongolia2146322260
Russia951144273
Ukraine3547204070
Sources: World Bank, World Development Report, 1996 and Social Indicators of Development.
Sources: World Bank, World Development Report, 1996 and Social Indicators of Development.

A two-track system of prices has emerged in these countries, with controls on a narrow range of essential items, but prices of most goods (accounting for 80 to upward of 90 percent of prices at the retail level) now determined by market forces. In some countries, even before price liberalization, extensive informal markets fed by remittances from expatriates abroad and diverted export proceeds acted to reduce the price distortions of central planning.

In China, growth has risen significantly since the start of the reform progress, as increased productivity in agriculture and the employment of surplus rural labor in the nonstate manufacturing sector made possible fairly large increases in output. Reforms in agriculture that allowed farmers to benefit from harvests above specified levels met with an immediate supply response, though less progress has been made in boosting output in largely industrial state-owned enterprises, for which soft budget constraints and complex tax regulations continue to stifle incentives to raise productivity. Instead, the nonstate sector grew rapidly, particularly in coastal regions that benefited from foreign direct investment, which began in 1984 and surged in the early 1990s. Subsequent reforms have included trade liberalization, financial market reform including the establishment of offices by foreign banks, and exchange system liberalizations that culminated in convertibility for current account transactions in December 1996.

In Cambodia, the Lao People’s Democratic Republic, and Vietnam, reforms were undertaken during periods in which monetization of large fiscal deficits led to sharp spikes in inflation, accompanied by extensive dollarization and widening gaps between official and informal exchange rates. Initial stabilization policies in all three countries comprised a mix of flexible exchange rates, trade liberalization, high interest rates, and reductions in government expenditures, particularly through cuts in subsidies to consumers and restrictions on soft credits to state-owned enterprises. Stabilizations from high inflation were largely money based, with steep expenditure cuts allowing reduced monetary expansion. Given the lack of resources, maintaining a stable exchange rate was not feasible at the start of the reforms; instead, official exchange rates were adjusted with increased frequency to track rates determined in parallel markets. Despite cuts in government expenditures, the boost in productivity that resulted from the reforms, particularly in the agricultural sector, served to maintain positive growth both before and throughout the macroeconomic stabilizations, while employment in the private industrial and service sectors rose rapidly during the early transition period and helped absorb workers released from the public sector.

Further market-oriented reforms are required to ensure continued growth that matches the potential of these countries. Following the success of their initial stabilizations, the newly emerging market economies of East Asia have each had to deal with the overheating problems that often accompany strong growth, including those associated with substantial inflows of foreign capital. Particularly important challenges for these countries are enhancing their infrastructure and developing modern banking systems, both of which will promote increased investment. Closer integration into the global economic and financial system will also promote further growth, including through increased trade.

1Myanmar has also taken some steps to reverse central planning; however, these reforms have so far been partial and have failed to achieve a fundamental transformation in the economic system.2See John R. Dodsworth, Ajai Chopra, Chi D. Pham, and Hisanobu Shishido, “Macroeconomic Experiences of the Transition Economies in Indochina,” IMF Working Paper 96/112 (October 1996) for an in-depth review of stabilizations in Vietnam, Cambodia, and the Lao People’s Democratic Republic.; Eduardo Borensztein and Jonathan D. Ostry, “Accounting for China’s Growth Performance,” American Economic Review, Vol. 86 (May 1996), pp. 224–28 for a discussion of the effects of reforms on China; and Richard W.T. Pomfret, Asian Economies in Transition: Reforming Centrally Planned Economies (Cheltenham: Edward Elgar, 1996) for a broad overview of reforms across Asian countries.

Financial Liberalization

The mobilization of savings and their efficient allocation among competing investment projects require stable financial markets with well-designed instruments. Studies using long-run data have found a stable and positive correlation between growth and indicators of financial development, and also between the initial level of financial sector maturity and subsequent growth.72 Countries that suffer from low growth have typically been characterized by a significantly lower level of financial development, as indicated by the ratio of broad money to GDP, than high-growth economies (Table 19). Although the positive relationship between financial development and economic growth is likely to be the result of two-way causation, and while there have been cases where countries have experienced rapid development with only modest degrees of financial openness, limited development of financial markets and institutions is likely to be a severe obstacle to overall economic development.73 Also, it is becoming increasingly clear that financial liberalization needs to be accompanied by strengthened regulation and supervision of financial institutions for an economy to derive the greatest benefits.

Governance

How effective an economic reform package is in delivering higher growth and long-run prosperity often depends on the quality of governance in an economy, Definition of the concept of governance is far from straightforward, and judgments about its quality are subjective, but there are several facets on which there is some agreement. In many developing countries, a lack of transparency and accountability in public policymaking and excessive government intervention and regulation of economic activities have invited widespread rent-seeking behavior and corruption. Not only do weak governance and corruption tend to lower government tax revenue74 and thereby both contribute to fiscal imbalances and reduce critical public investment in areas such as health and education, but they also deter both domestic and foreign direct investment. Inadequate protection of private property rights and a weak rule of law also have long been held to be critical obstacles to growth.75 Reduced state intervention in economic affairs and greater transparency in regulatory policies can limit rent seeking and corruption and allow governments to focus on their essential tasks—of maintaining order and justice, of allocating public resources to priority uses, including investment in health and education, of helping to maintain macro-economic stability, and of providing cost-effective and well-targeted social safety nets. While it is difficult to quantify the efficiency of governance and the extent of corruption, some studies based on rudimentary and somewhat subjective indicators suggest that weakness in these areas can have significant and lasting negative effects on growth.76 Other studies using different proxies for the effectiveness of government, such as the quality of the bureaucracy (including its degree of autonomy from political pressure), expropriation risk, and so forth, have found little effect on growth.77

Human Resource Policies: Education and Population Growth

Investment in education and human capital leads to the acquisition of skills that raise efficiency and make more widespread the use of existing technology, and also promotes new technological development. Reflecting such investment, the level of initial human capital in high-growth countries has been significantly higher than in less-successful countries (Table 19). More formal analysis also finds that the initial level of education, especially at the primary level, is an important determinant of subsequent growth.78 However, few studies have found evidence of a strong positive impact of changes in the level of education on growth.79 Higher education has been found to have a relatively strong positive impact on growth, and public spending on education as a share of GDP has been found to be strongly and positively related to growth. These findings suggest that the quality of education is important. Also, social factors, such as the nature of institutions in society, affect the pace of economic development in a country. The initial level of social development, of which the level of human capital is an important component, is also significantly related to subsequent growth in per capita income and productivity.80

Among many of the poorer countries, high growth rates of population have undermined efforts to increase the average levels of education and health. Although rapid population growth increases the labor force and raises the output capacity of an economy, cross-country analyses of long-run data have shown that population growth has a negative effect on the growth of per capita income.81 In many countries, despite attempts to contain the growth of population, which have varied according to national culture and values, progress has been slow. In these cases, long-run economic and social development and the alleviation of poverty hinge critically on greater success in slowing the rate of population growth.

Economic Convergence and the Importance of Policy Complementarities

While policies in all the areas discussed above can help promote the accumulation of physical and human capital and the development of technology, and thus help determine the growth performance of a country, a few key policy areas seem particularly important. To study the effects of some of these policies on growth outcomes, a data set for 110 developing countries was compiled for the period 1985–95—a time period when the global integration process described earlier was in full swing. Data were collected on trade openness: the degree of macroeconomic stability, proxied by the standard deviation of the rate of inflation; and the degree of government intervention in the economy, proxied by the share of government spending in GDP, which although not the best measure of government intervention suffers less than others from data-related problems. Countries were scored as “high,” “medium,” or “low” in each of these categories with cutoff points determined by statistical criteria, though necessarily with some degree of arbitrariness.82 Developing countries with average 1985–95 real per capita GDP growth rates of 2.9 percent or higher were classified as high-growth countries (e.g., Chile, Thailand, Uganda), those with growth rates between 0.5 percent and 2.9 percent a year were classified as medium-growth countries (e.g., Colombia, Morocco, Pakistan), and those with growth rates lower than 0.5 percent a year were classified as low-growth countries (e.g., Cameroon, Ecuador, Zambia) (Table 20). Altogether 28 percent of all developing countries were classified as high-growth countries, while 44 percent fell into the low-growth category.

Table 20.Developing Countries Including Asian Newly Industrialized Economies: Relationship Between Policies and Growth, 1985–95, and Conditional Probabilities of Success1(In percentage points)
High

Growth
Medium

Growth
Low

Growth
Number of

Countries

in Sample
Percentage Distribution
Percentage of countries282844110
Conditioning policies
High openness241194137
High macroeconomic stability341322741
Small government430333743
High openness with at least one other category being low and in most one other being medium2525508
High macroeconomic stability with at least one other category being low and at most one other being medium60405
Small government with at least one other category being low and at most one other being medium21334624
Low openness with at least one other category being high2 and the other at least medium19433821
Low macroeconomic stability with at least one other category being high and the other at least medium1001
Large government with at least one other category being high and the other at least medium25175812
Policy combination with at least two categories being high2 and the other medium57192421
Source: Jahangir Aziz and Robert Wescott, “The Washington Consensus and Policy Complementarities in Developing Countries,” IMF Working Paper (forthcoming).

High-growth countries were the ones with average per capita real growth rates of 2.9 percent or higher over the period 1985–95, while the low-growth countries were the ones with average growth rates below ½ of 1 percent. For all variables, the high and low cutoff points were determined as mean plus one-half standard deviation and mean minus one-half standard deviation of the respective distributions.

Openness is measured by the ratio of total foreign trade to GDP. A high degree of openness is defined as the case where the average of exports and imports as a percent of GDP is higher than 45 percent, while in the case of low openness it is lower than 27 percent.

Macroeconomic stability is measured by the standard deviation of the rate of inflation in the period. Low macroeconomic stability was defined as having a standard deviation of inflation higher than 19 while high stability required a standard deviation of less than 5.

Size of government is measured by the ratio of government expenditure to GDP. A large government is categorized as one with public expenditure of the central government above 38 percent of GDP, and a small government is one with less than 23 percent of GDP.

Source: Jahangir Aziz and Robert Wescott, “The Washington Consensus and Policy Complementarities in Developing Countries,” IMF Working Paper (forthcoming).

High-growth countries were the ones with average per capita real growth rates of 2.9 percent or higher over the period 1985–95, while the low-growth countries were the ones with average growth rates below ½ of 1 percent. For all variables, the high and low cutoff points were determined as mean plus one-half standard deviation and mean minus one-half standard deviation of the respective distributions.

Openness is measured by the ratio of total foreign trade to GDP. A high degree of openness is defined as the case where the average of exports and imports as a percent of GDP is higher than 45 percent, while in the case of low openness it is lower than 27 percent.

Macroeconomic stability is measured by the standard deviation of the rate of inflation in the period. Low macroeconomic stability was defined as having a standard deviation of inflation higher than 19 while high stability required a standard deviation of less than 5.

Size of government is measured by the ratio of government expenditure to GDP. A large government is categorized as one with public expenditure of the central government above 38 percent of GDP, and a small government is one with less than 23 percent of GDP.

Three general policy combination clusters emerged. Approximately one-fifth of all developing countries in the data set had closed economies, poor macroeconomic stability, and a large government—all the presumed least favorable policies. Another fifth of the countries were moderately open, with moderate macroeconomic stability and a medium-sized government. Another fifth of the countries can be described as highly open with moderate to high macroeconomic stability and a small to medium-sized government—the presumed best policies. As expected, there was found to be a strong overall correlation between policies and income growth performance: countries with open trade positions, a stable macroeconomy, and a relatively small government tended to show better growth outcomes than countries that were less open, less stable, and had larger governments.

Perhaps the most interesting finding is that not one of these desirable policies by itself seems to have been sufficient to ensure that a country had high growth. That is, good performance in one category, but mediocre or poor performance in the other two categories, appears to have been a recipe for low growth. For example, among countries with the most open trade stances, but with only low or medium macro stability and large or medium-sized governments, only about one-fourth experienced high growth and about one-half low growth. Likewise, of the 24 countries that had the smallest size of government, but low or medium openness and stability, only about one-fifth achieved high growth, and about one-half posted low growth. Along a single policy dimension, the probability of failure in the case of high openness was just as high as in the case of a small government (Table 20).

The analysis also suggests that poor performance in one policy area can hold an economy back, even if other policies are favorable. To illustrate this necessity of comprehensively good policies, consider the countries that had medium or high macro stability and small or medium-sized governments, but low openness: only one in five had fast growth, and twice that fraction—two in five—had slow growth. Or, of countries with medium or high openness and stability, but an undesirably large government, fewer than one-third experienced fast growth, while nearly two out of three experienced low growth.

The key lesson that emerges is that no policy by itself is sufficient for fast growth, and that at least a moderate degree of policy success is necessary in several areas to support fast growth.83 That is, good policies tend to be mutually reinforcing and policy complementarities are important. For example, in a relatively open economy, financial markets may punish bad macroeconomic policies and reward good policies more vigorously than they would in a closed economy. To illustrate the importance of these complementarities, the developing countries with either medium or high degrees of success in all three policy areas had a three in five chance of achieving fast growth over 1985–95, and a better than three in four probability of either medium or fast growth (17 out of 21 such countries in the sample). Fewer than one out of four of these countries had low growth.

This simple analysis does not suggest an ironclad relationship between good policies and good growth outcomes, and certainly there are exceptions to the rule reflecting the influence of many other factors (social and institutional factors, resource endowments, and so forth) that also exert a strong influence on growth. Uruguay, for example, had low trade openness and only moderate macroeconomic stability over 1985–95, and yet experienced relatively fast economic growth. Other countries, such as Botswana, have experienced relatively fast growth despite having a large government sector. Rather this analysis suggests that for most of the 110 developing countries in the data set over 1985–95, the goal of fast income growth was most likely to be achieved by pursuing market-oriented policies (trade openness and small to medium-sized governments) in an environment of macroeconomic stability.

Reaping Gains from Globalization and Avoiding Marginalization

For countries with relatively strong fundamentals and the types of policies that the above analysis suggests are conducive to growth, openness has helped to speed up the convergence process. Malaysia and Thailand are examples of countries with these characteristics. The policy challenge for these countries is to safeguard their gains by maintaining a market-oriented policy stance, maintaining macroeconomic stability, and improving infrastructure and the supply of skilled labor so as to ease supply constraints in the economy. Overheating pressures often have been a consequence of strong capital inflows, and experience shows that although capital inflows can supplement domestic saving and contribute to strong economic performance, without appropriate policies or in the presence of large exogenous shocks such flows can raise a country’s vulnerability to external and domestic financial disturbances. This was seen in the Mexican crisis, and it underscores again the importance of maintaining macro-economic and financial stability, with a sustainable balance of payments. In addition, a scarcity of skilled labor can boost wage growth, erode external competitiveness, and fuel inflationary pressures. This suggests that labor market reforms may also be necessary to ease such capacity constraints.

Countries should also aim to let domestic investors diversify their portfolios internationally, to reduce their risks and also to help prevent price bubbles in real estate and other domestic asset markets. In this respect, a gradual and cautious removal of capital controls within the framework of policies to promote a sound domestic banking system, along with an exchange rate policy that permits an appropriate degree of flexibility, will lessen the burden on fiscal adjustment and provide a better balance among policy instruments through a more developed financial sector. Additionally, many countries, such as China, Thailand, and Malaysia, are experiencing rapidly growing demands for transportation and other public facilities. Private sector participation in these areas, as employed in Malaysia (port facilities), the Philippines (power supply), and Chile (public utilities), can supplement government efforts to alleviate supply bottlenecks without burdening public finances excessively.

Policies fur Avoiding Marginalization

Many of the countries near the bottom of the world’s distribution of per capita incomes face difficult conditions, such as low stocks of human capital, poor resource bases, and political instability—including civil wars and regional conflicts that have acted to deter investment and growth. Many of these countries also suffer from high levels of public debt, including external debt, that was accumulated over years of poor fiscal management, commodity price shocks, macroeconomic instability, and poor governance. Still, a number of developing countries have overcome such obstacles, and there has been a resurgence of GDP growth in many countries where macroeconomic and structural reforms have been undertaken since the early 1990s. Between 1990 and 1995, the number of countries in sub-Saharan Africa with rates of real GDP growth greater than 4 percent increased steadily from 14 to 25, just as the number of countries with negative growth declined from 18 to 9. Uganda, which has implemented far-reaching reforms since the late 1980s, has been closing the gap with advanced economy income levels in the 1990s, and India has experienced average growth of 7 percent in 1995 and 1996, reflecting the effects of the liberalization program that started in 1992. Vietnam had an average real per capita income in 1990 of less than 1 percent of advanced economy levels but has been growing at more than 7 percent a year in real per capita terms aided by continued macroeconomic and structural reforms. Still, given the low levels of per capita income in such developing countries, relatively high growth rates will need to be sustained for many years to close the gap with the advanced economies (see Table 18).

As the above analysis of policy complementarities suggests, a successful strategy for growth requires openness toward international trade, macroeconomic stability, and limited government intervention in the economy.84 Protectionist trade policies, such as high tariffs and widespread nontariff barriers, clearly have obstructed the integration of many countries into the world economy. In sub-Saharan African countries, for example, tariffs average about 27 percent compared with 15 percent among the East Asian countries, and the average nontariff barrier coverage ratio is many times higher than in the world’s fastest-growing developing country group. At least partly because of such policies, the sub-Saharan African share of world trade fell from roughly 3 percent in the mid-1950s to just over 1 percent in 1995. And in recent years, this region has been attracting only about 3 percent of the total foreign direct investment in developing countries. With the phased reductions in nontariff barriers under the Uruguay Round agreements, many of the low-income countries will also need to improve the competitiveness of their exports, which have enjoyed preferential treatment in the past.

In addition to becoming more open, poor countries also need to reform government operations. These countries have great needs in the areas of health, education, and infrastructure, but government spending has been channeled too heavily into defense (at least until recently), into subsidies for loss-making and inefficient state enterprises, and into inefficient public administration. Spending needs to be rechanneled into more socially beneficial uses, and especially where international aid is available, project implementation needs to be improved. Governments in these countries also need to reform their revenue systems. In many countries, practices of arbitrary exemptions and weak enforcement have, in effect, led to the imposition of high tax rates on narrow tax bases. The result is that sectors that are subject to taxes have a strong incentive to evade them. To finance large fiscal deficits, many of these governments have resorted to financial repression, which has thwarted the development of financial markets, and to direct monetization, which has fueled inflation; and many have borrowed heavily in their small domestic capital markets, crowding out private investment and increasing public debt. These policies have fueled inflation and increased macroeconomic instability.

Many poor countries have accumulated large stocks of external debt, including debt owed to multilateral agencies. To address the debt problems of the heavily indebted poor countries, a joint initiative has been launched by the IMF and the World Bank that will provide special assistance to the countries that have followed sound policies but for whom traditional debt-relief mechanisms have failed to secure a sustainable external position (see Chapter II).85 As many poor and formerly poor developing countries have demonstrated, the mutually reinforcing benefits of a return to external viability, greater international openness, domestic macroeconomic stability, and good governance with priority-based government spending can generate higher growth rates and rapid convergence.

53

See Jeffrey D. Sachs and Andrew M. Warner, “Economic Convergence and Economic Policies,” NBER Working Paper No. 5039 (Cambridge, Massachusetts: National Bureau of Economic Research. February 1995). They deem a country’s trade regime to be closed if it has any one of the following characteristics: (1) nontariff barriers covering 40 percent or more of total trade. (2) average tariff rates of 40 percent or more, (3) a black market exchange rate in which the domestic currency is depreciated 20 percent or more relative to the official exchange rate, (4) a socialist economic system, or (5) a state monopoly on major exports.

54

See Richard Harmsen and Michael Leidy, “Regional Trading Arrangements,” in International Trade Policies: The Uruguay Round and Beyond, Vol. II, Background Papers, by Naheed Kirmani and others (IMF, 1994).

55

See, for example, Rowthorn and Ramaswamy, “Deindusirialization: Causes and Implications.”

56

See “Workers in an Integrated World,” World Development Report (Washington: World Bank, 1995).

57

See Donald J. Robbins, “Evidence on Trade and Wages in the Developing World,” OECD Development Centre Technical Paper No. 119 (December 1996). The author argues that contrary to the predictions of the factor price equalization theorem, the wages of unskilled workers in a number of developing countries have been falling in relative terms, probably because of the forces of technology.

58

See Thomas Straubhaar, On the Economics of International Labor Migration (Bern: Stuttgart: Paul Haupt, 1988).

59

See Phillip Martin, “Economic Aspects of International Migration” (unpublished: IMF, Research Department, December 1996).

60

The estimate of worker remittances is from Martin, “Economic Aspects.”

61

For an analysis of the effects of human capital flight on growth, see Nadeem Ul Haque and Se-Jik Kim, “‘Human Capital Flight’: Impact of Migration on Income and Growth,” Staff Papers, IMF, Vol. 42 (September 1995), pp. 577–607.

62

This thinning of the middle range of the developing country income distribution leads to a global income distribution that appears to be characterized by two large clusters at each end, which has been termed the “twin peaks” phenomenon. For example, sec Danny T. Quah, “Twin Peaks: Growth and Convergence in Models of Distribution Dynamics.” Economic Journal, Vol. 106 (July 1996), pp. 1045–55.

63

After correcting for different long-run potential income levels resulting from differences in policies and resources, initially poorer countries do tend to grow faster than relatively richer countries. This tendency has been termed “conditional convergence” to signify the dependence on policies. For further details, see the box “Economic Convergence,” in the October 1994 World Economic Outlook, pp. 94–95.

64

See Barry Bosworth, Susan M. Collins, and Yu-chin Chen, “Accounting for Differences in Economic Growth,” Brookings Institution, Discussion Paper, No. 115 (Washington: Brookings Institution. December 1995).

65

Since the growth rate of output is necessarily equal to the investment-output ratio divided by the incremental capital-output ratio (ICOR), higher investment-output ratios will be associated with faster output growth unless the ICOR is at least commensurately higher (marginal productivity of capital lower). This relationship indicates also that raising the investment ratio may not boost growth if the capital stock added has a low productivity.

Ross Levine and David Renelt, “A Sensitivity Analysis of Cross-country Growth Regressions,” American Economic Review, Vol. 82 (September 1992), pp. 942–63, find that among a variety of economic policy, political, and national indicators, only the share of investment in GDP turns out to have a positive and robust correlation with growth.

66

See, for example, the October 1996 World Economic Outlook, pp. 120–22, and Michael Sarel, “Nonlinear Effects of Inflation on Economic Growth,” Staff Papers, IMF, Vol. 43 (March 1996), pp. 199–215. Also see Michael Bruno and William Easterly. “Inflation Crises and Lima-Run Growth” World Bank Working Paper No. 1517 (Washington: World Bank, September 1995).

67

See May and October 1996 issues of the World Economic Outlook for a fuller discussion of the consequences of fiscal imbalances and inflation, respectively.

68

Sachs and Warner, using their own openness indicator, claim that openness is the single most important factor in bringing about convergence. Dan Ben-David and Atiqur Rahman in “Technological Convergence and International Trade,” Centre for Economic Policy Research Discussion Paper No. 1359 (London: CEPR, March 1996), find that among the richest 25 countries in the world there is significant evidence of absolute convergence within each country’s main trading partners’ group.

69

See David T. Coe, Elhanan Helpman, and Alexander W, Hoffmaister, “North-South R&D Spillovers,” Economic Journal, Vol. 107 (January 1997), pp. 134–49.

70

See James A. Schmitt Jr., “The Role Played by Public Enterprises: How- Much Does It Differ Across Countries?” Quarterly Review, Federal Reserve Bank of Minneapolis, Vol. 20 (Spring 1996), pp. 2–15.

71

See Omkar Goswami, “Whither Corporate Sector Reforms in India?” paper presented at the seminar on “Putting India on a High-Growth Path: The Macroeconomic Strategy and Key Structural Reform” organized by the IMF, and held in Washington on March 6, 1996.

72

See Ross Levine, “Financial Development and Economic Growth: Views and Agenda,” World Bank Working Paper No. 1678 (Washington: World Bank, October 1996).

73

Recent issues of the World Economic Outlook, especially the October 1996 World Economic Outlook, have discussed in detail the importance of financial development to the growth process.

74

See Nadeem Ul Haque and Rama Sahay, “Do Government Wage Cuts Close Budget Deficits? A Conceptual Framework for Developing Countries and Transition Economies,” IMF Working Paper 96/19 (February 1996).

75

This argument was made in 1776 by Adam Smith in The Wealth of Nations (New York: The Modern Library, 1937), p. 862, and was recently stressed by Sachs and Warner, “Economic Convergence.”

76

See Paolo Mauro, “The Effects of Corruption on Growth, Investment, and Government Expenditure,” IMF Working Paper 96/9K (September 1996).

77

For further details, see Robert J. Barro and Xavier Sala-i-Martin, Economic Growth (New York: McGraw Hill 1995), pp. 439–40.

78

Barro and Sala-i-Martin, Economic Growth, pp. 436.

79

See, for example, Lant Pritchelt, “Where Has All the Education Gone?” World Bank Policy Research Working Paper No. 1581 (Washington: World Bank, March 1996).

80

See Jonathan Temple and Paul Johnson. “Social Capability and Economic Development.” Nuffield College Working Paper (Oxford. England: Nuffield College, July 1996). They use the Adelnum-Morris index, which includes measures of the size of the agricultural sector, the extent of urbanization, social mobility, literacy, and mass communication to proxy the level of social development.

81

For example, Levine and Renelt, “Sensitivity Analysis” report that a 1 percentage point increase in the growth rate of population reduces per capita GDP growth by roughly ½ of 1 percentage point. However, population growth and fertility may themselves be responsive to the rate of output growth.

82

For all variables, the high category was defined as the mean value plus one-half of the standard deviation, or higher; the low category was the mean minus one-half of the standard deviation or lower: and the medium category included all values in between.

83

This conclusion differs from that drawn by Sachs and Warner, “Economic Convergence.” These authors test for the effectiveness of various policies in promoting higher-than-average economic growth and conclude that an open trade stance and protection of private property rights together are sufficient for fast growth.

84

William Easterly and Ross Levine, “America’s Growth Tragedy: A Retrospective. 1960–89,” World Bank Working Paper No. 1503 (Washington: World Bank, August 1995), however, find that in addition to government intervention, political instability and spillovers between neighboring countries’ economic performances are significant in explaining low growth.

85

For further details on the Highly Indebted Poor Countries (HIPC) Initiative and the debt burden of these countries, see the October 1996 World Economic Outlook, pp. 74–76.

    Other Resources Citing This Publication