International Monetary Fund. Research Dept.
Published Date:
April 2000
  • ShareShare
Show Summary Details

Progress in raising real incomes and alleviating poverty has been disappointingly slow in many developing countries, and the relative gap between the richest and poorest countries has continued to widen. In Africa, the level of real per capita income today is lower than it was 30 years ago. In developing countries in the Middle East and the Western Hemisphere, real incomes have risen, but at a slower pace than in industrial countries (Figure 4.1). Sustained and rapid improvements in relative income positions—convergence—have only occurred among the developing countries of east Asia (including China), the newly industrialized economies, and a few smaller countries in other regions. Average per capita incomes in regions with the largest populations (as represented by the thickness of the bars in Figure 4.1) remain well below those in the industrial and newly industrialized economies. Overall, the number of very poor (those living on less than $1 per day) has remained roughly unchanged over the past decade, and only limited progress has been made in reducing the share of the world population living in poverty. This represents both huge amounts of unnecessary human suffering and an enormous squandering of human potential.

Figure 4.1.Advanced and Developing Economies: Per Capita Income1

(Thousands of U.S. dollars at 1996 prices)

East Asia, which includes China and has the largest population, and the newly industrialized economies are the only country groups that are rapidly converging with the industrial countries. The thickness of the bars reflects the population in each region.

1 Converted into U.S. dollars using purchasing-power-parity (PPP) conversion rates.

Recognizing these divergences in income levels and rates of growth, and in particular the unacceptably high levels of poverty that persist in many countries, the Interim and Development Committees of the IMF and World Bank placed renewed emphasis in September 1999 on poverty reduction and on strengthening the links between debt relief and poverty reduction. Subsequently, the Executive Board of the IMF refocused the objective of the IMF’s concessional lending to put greater explicit emphasis on poverty reduction (Box 4.1).

The new approach to poverty reduction, now in the initial implementation stages, builds on the traditional emphasis on macroeconomic and structural soundness by also stressing the importance of country ownership of the poverty reduction strategy. The growing concern for country ownership, including through the involvement of civil society, is intended to reduce the risk of slippages in implementation as the countries themselves take greater responsibility for the design and success of their economic plans. The enhanced amount of debt relief under the Heavily Indebted Poor Countries (HIPC) Initiative is intended to release resources for poverty reduction, increase incentives for reforms, and remove a deterrent for both domestic and foreign investors. Of course, a successful development strategy requires progress on many fronts, as emphasized in the World Bank’s Comprehensive Development Framework.1

Box 4.1.Poverty Reduction Strategy Papers

In September 1999, the IMF and the World Bank endorsed a new approach to enhance the focus on poverty reduction in programs supported by concessional assistance, and to strengthen the link between debt relief and poverty reduction. Underpinning the new approach, each eligible country will prepare a Poverty Reduction Strategy Paper (PRSP).

A PRSP will outline a country’s anti-poverty strategy over the medium and long term. To foster ownership, the PRSP will be drawn up by the government after broad-based consultations with stakeholders, including representatives of civil society and development partners, and with assistance from World Bank and IMF staff. Long-term poverty reduction goals will be translated into annual targets for intermediate indicators (i.e., primary school enrollment, immunization rates, etc.) to facilitate shorter-term program monitoring. The PRSP is expected to become a key instrument for a country’s relations with the donor community and civil society and will provide the basis for debt relief under the HIPC Initiative and all World Bank/IMF concessional lending operations. Reflecting the new policies and the central focus on poverty reduction, the IMF has replaced its concessional lending facility—formerly the Enhanced Structural Adjustment Facility—with the Poverty Reduction and Growth Facility.

It is envisaged that countries will prepare PRSPs on a three-year cycle, with progress reports in intervening years. The PRSP will diagnose poverty in the country and describe the poor and their main characteristics. The PRSP will also quantify the resources needed for various poverty reduction programs and incorporate them in a sustainable fiscal and macroeconomic framework, taking account of the availability of non-inflationary financing. Work toward the preparation of PRSPs has begun in a number of countries (e.g., Bolivia, Mozambique, and Uganda).

Since the design of robust poverty reduction strategies can take time, a phased introduction of a full-fledged PRSP is foreseen. Some governments will thus prepare an interim PRSP, stating their commitment to poverty reduction and laying out the principal elements of their strategy, the timetable to complete the PRSP, and a description of the consultative process through which the PRSP will be formulated. The threeyear macroeconomic program supported by IMF and World Bank lending could then be revised to reflect the PRSP when one is completed.

Previous issues of the World Economic Outlook have discussed the experiences of the successfully converging countries extensively. This chapter, in contrast, investigates the main impediments to growth in the developing countries that have failed to prosper and where poverty rates remain high. The large number of countries in this group seems to suggest either that the conventional growth strategy is not being implemented forcefully enough or that the strategy has been overlooking critical obstacles to development. However, the bulk of development research reveals neither a unique set of preconditions that are always present during economic takeoff nor an easily identified set of impediments that have prevented poor countries from achieving sustained growth. There is no single formula for kick-starting growth, and it is more likely that the explanation for the unsatisfactory performance of many developing countries lies in the interplay of economic and political factors that vary by country. Nevertheless, experience in the successful developing countries clearly points to macroeconomic stability, sound institutional arrangements, and openness to trade as factors that are conducive to, or at least associated with, high sustainable growth.2 Experience in the poorest countries highlights poor education and health, ineffective governance, weak rule of law, and war as frequent impediments to prosperity.

A theme of the chapter is the plight of poor, low-growth countries and within this group the heavily indebted countries, about which there is growing consensus that unsustainable debt has become a critical barrier to future growth and poverty reduction. External debt levels, despite years of rescheduling often at concessionary terms, have become unsustainable in many cases. Without the efforts of the international community to reduce this burden substantially, there is little hope for significant improvement in living conditions, as debt overhang saps economic incentives to reform and grow. To be effective, however, debt relief must be accompanied by domestic policy reforms to address the root causes of much of the initial debt buildup.

The chapter deals less with poverty problems in middle-income countries and low-income countries that are growing rapidly. Nevertheless, it is important to remember that in countries such as China and India, which have been growing quickly and seen poverty rates fall, the number of poor remains high. Some middle-income countries, for example in Latin America, not only have pockets of absolute poverty, but also significant relative poverty.3 In all of these cases, poverty alleviation remains important, and most of the policy considerations discussed in this chapter also apply.

Income Growth and Poverty Reduction: The Recent History

Human living conditions have improved greatly over the past 100 years, brought on by unprecedented technological and economic transformations. Global output almost tripled in the first half of the twentieth century and increased ninefold in the second half, greatly outpacing population growth. Life expectancy, education, and other indicators of well-being also improved, particularly in poor countries.4 This chapter looks at trends over 1970–98, the longest period for which more comprehensive data are available for most developing countries. (Long-run perspectives are discussed further in Chapter V.)

Rising global prosperity, however, has not benefited all countries and regions, and the global distribution of income—measured by average incomes across countries—remains very skewed. This failure to converge can be seen in longerterm growth rates: 75 percent of developing countries recorded slower per capita income growth than in the industrial countries over the past three decades (Figure 4.2, upper panel).5 Per capita income fell in 32 countries in the sample, while only seven developing countries grew fast enough to reduce substantially the income gap with—and rapidly converge toward—the industrial countries as a group.6

Figure 4.2.Developing Countries: Grouped by Per Capita Income Growth, 1970–981

The number of developing countries with slower per capita income growth than that of the industrial countries (negative growth and slow growth) is large (over 75 percent of all developing countries), although this group of developing countries comprises about 30 percent of the total in terms of population.

1 Slow-growth countries have annual per capita income growth rates between 0 and 2 percent; slow-convergence countries, between 2 and 3¾ percent; and fast-convergence countries, over 3¾ percent.

The picture is more encouraging when progress is assessed based on both per capita income growth and the size of a country’s population (Figure 4.2, lower panel). Less than 10 percent of the developing world’s population live in countries where average income declined, while 70 percent live in countries where per capita income growth exceeded that in the industrial countries. This more positive outcome mainly reflects strong economic growth in China in particular and also in India, which together account for about 50 percent of the population in developing countries and which had per capita growth rates of 7 percent and 2 ¾ percent, respectively, over the 1970–98 period (Box 4.2).7

A more positive picture also emerges when examining the growth performance over a shorter time frame. During 1993–98, 14 developing countries converged rapidly toward the industrial countries—double the number between 1970 and 1998—despite the problems encountered by many countries because of financial crises in Asia, Russia, and Latin America. Conversely, per capita income fell in 23 countries (compared to 32 during the longer period).8 The improved performance in these countries partly reflects stronger domestic policies and for some of these countries, a more supportive external environment (for example, through concessional lending programs or more open trade). The better performance, however, in some instances also reflects cyclical factors, such as improved harvests or recoveries from wars or recessions. The larger number of countries in the rapidly converging category and the smaller number of countries in which per capita income fell (which in part reflects a pickup in growth performance in Africa over this period) is encouraging. But progress in alleviating poverty remains fragile and inadequate.

Box 4.2.India: Reinvigorating the Reform Process

India has been among the fastest-growing economies in the world over the last two decades, and has achieved trend improvements in growth, literacy, mortality, and poverty rates (see the figure, upper panels). In recent years, deft handling of monetary policy has helped India to successfully weather the Asian crisis, while maintaining low inflation and a comfortable external position. Yet despite these gains, poverty rates remain high, with more than a third of the population still living below the official poverty line.1 This uneven progress raises questions about the impact of the economic and structural reforms implemented since the mid- 1980s on growth in India, and what more can be done to make greater inroads into poverty reduction.

In the three decades following independence in 1947, growth in India was stifled by a high degree of government planning and regulation, with per capita GDP rising by only 1½ percent per annum (see the first table). Industrial controls were pervasive, and restrictions on private credit, the role of the public enterprise sector, and subsidy programs increased throughout the period. Strict controls on foreign direct investment, an import licensing system, and—from the 1970s—high tariff rates further limited the economy’s growth potential.

The liberalization of import and industrial controls in 1985 and improved agricultural performance spurred an acceleration of real per capita GDP growth to an average rate of 3¾ percent in the 1980s. However, this expansion also reflected other developments—increased fiscal stimulus and a debt-financed consumption and investment boom—which became unsustainable toward the end of the decade.

A balance of payments crisis ensued in 1991, reflecting the deteriorating fiscal position, rising external debt (especially short-term), a surge in world oil prices, and a sharp decline in remittances from Indian workers in the Middle East. As capital flight accelerated and official reserves were rapidly depleted, the Indian government entered into a Stand-By Arrangement with the IMF and embarked on a program of fiscal and structural reforms.

Corrective policy measures were successful in restoring macroeconomic stability. The central government deficit was brought down from 8 percent of GDP before the crisis to 4¾ percent in 1996/97,2 through tax reforms, cuts in subsidies, and reductions in defense and other expenditures (see the figure, lower panels). The lower deficit, in turn, reduced financing that had to be provided by the central bank, and wholesale price inflation declined from a precrisis level of almost 14 percent to nearly 6 percent by 1996/97.

In addition, important structural reforms were introduced. Industrial licensing and investment approval procedures were liberalized, and the number of industries reserved for the public sector was reduced. External sector reforms included a reduction in the import-weighted tariff rate from 87 percent in 1990/91 to 25 percent by 1996/97, easing of import licensing requirements, relaxation of controls on foreign direct and portfolio investment, and greater exchange rate flexibility. Financial sector measures included interest rate liberalization, strengthened prudential norms and supervision, the introduction of greater competition into the banking system, and improvements to the operation of capital markets.

In response to the government’s policy package, the recovery from the 1991 crisis was rapid. Private investment rates rose sharply, and real per capita GDP growth increased to more than 6 percent by 1995/96. Significant improvements in productivity were also achieved—as evidenced by increased total factor productivity growth at both the aggregate and firm levels3 and by declining incremental capital-output ratios, particularly in the services sector (see the first table).

India: Economic Indicators1

Sources: Central Statistical Organization; Government of India Planning Commission; Union Budget documents; Reserve Bank of India; Public Enterprises Survey; Government Finance Statistics; and IMF staff estimates and projections.

1Data shown are for fiscal years, which begin in April.

2Consolidated public sector comprises the central and state governments, central public enterprises, and the accounts of the Oil Coordinating Committee.

However, per capita growth has slowed more recently, averaging closer to 4 percent between 1997/98 and 1999/00 compared with 4¾ percent between 1992/93 and 1996/97.4 To some extent, this reflected the completion of cyclical catch-up following the 1991 balance of payments crisis, as well as the adverse impact of the 1997 regional crisis and agricultural supply shocks. In addition, though, economic performance appears to have been adversely affected by a reversal of fiscal adjustment, infrastructure bottlenecks, and delays in implementing structural reforms. Increases in civil service wages and subsidies, as well as rising debt service, pushed up the fiscal deficit and resulted in higher real interest rates. These higher rates, combined with banks’ efforts to improve their balance sheets, slowed credit growth. Infrastructure constraints also continued to bind, as the earlier fiscal consolidation had relied too heavily on reductions in public investment. Consequently, the contribution of private investment to growth fell by half from earlier in the decade, and measured productivity growth, particularly in the industrial sector, deteriorated (see the first table).

Expenditure and Sectoral Components of Growth1(Average annual percent, unless otherwise noted)
Real per capita GDP growth31.
Real GDP growth33.
Contribution to growth, by expenditure item:
Private consumption2.
Public consumption0.
Gross fixed investment0.
Private investment0.81.80.9
Public investment0.60.10.3
Net exports40.10.10.6
Contribution to growth, by sector
Contribution to growth, by sector:
ICORs, by sector:5
Sources: Central Statistical Organisation (CSO), National Accounts Statistics.

Moreover, the poverty rate remains very high, and the impressive rate of decline from the mid- 1970s through the 1980s may have slowed.5 This outcome partly reflects the relatively poor performance of the agricultural sector during the 1990s, since some 70 percent of the labor force still relies on the land for its livelihood.6 While adverse supply shocks played a role, the lack of agricultural reform also contributed to low investment rates and productivity in this sector. In addition, the scope for mobility of low-skilled labor out of the agricultural sector has likely been limited by the absence of robust and sustained growth in the industrial sector and by the relatively larger contribution of the higher-skilled service sector to GDP growth.

What measures need to be implemented to sustain high growth rates in all sectors and achieve more substantial progress toward poverty alleviation?7 As discussed in the text—and quantified in the Indian context in the second table—faster growth would require durable fiscal consolidation to raise national saving and crowd-in private investment spending; further liberalization of foreign trade and investment flows; and additional reforms to labor markets and in the agricultural, industrial, and financial sectors to promote greater efficiency and export competitiveness. These reforms need to include removal of domestic pricing distortions, improvements to bankruptcy procedures, and an easing of restrictions on firm and farm size and regulations that make it difficult to shed labor (and therefore impede job creation). Fiscal priorities also need to be redirected toward investment in human and physical capital.

Explaining India’s Relative Growth Performance
Average Value (percent)Estimated Difference in Contribution to Growth Rates1 (percentage points)
IndiaEast Asia
Factors contributing to growth during 1970-95
Net FDI/GDP0.12.5-0.8
Government consumption/GDP10.310.40.0
Secondary school enrollment rate35.250.5-0.3
CPI inflation rate8.88.40.0
Convergence effect and other factors0.2
Real per capita GDP growth (1970–95 average)2.45.7-3.3
Source: P. Kongsamut and A. Vamvakidis, “Economic Growth,” Chapter 2 in Philippines: Toward Sustainable and Rapid Growth, IMF Occasional Paper No. 187 (Washington: International Monetary Fund, 2000).

There is broad agreement in India that further reforms are needed—the experience of the early 1990s has demonstrated the potential benefits of reform, and consistent views on many of these key issues emerged from the major parties during the October 1999 election. Several factors argue for translating this consensus into swift action. First, the establishment of a bold agenda would be facilitated by the relatively favorable current economic situation and the significant majority enjoyed by the ruling coalition. Second, with the consolidated public sector deficit rising again and the public sector debt stock close to 80 percent of GDP, fiscal sustainability is a serious concern.8 Third, India is committed to trade liberalization measures under the World Trade Organization, including the removal of all quantitative restrictions by 2001. For India to achieve the maximum benefits from a more liberal trade system, the structural impediments affecting domestic producers must be addressed in the interim.

Encouragingly, the new government has taken a number of initiatives that suggest a strengthened commitment to structural reform, including liberalization of the insurance sector, automatic clearance for foreign direct investment in many sectors, and a landmark agreement on state sales tax rationalization. At the same time, however, the budget introduced in February 2000 targets only modest deficit reduction in the coming fiscal year, and a clearly defined agenda for reform has yet to be established. Hence, critical and difficult challenges remain to be addressed.

The World Bank’s World Development Report 1999/2000 suggests an even more severe poverty problem, with almost half of the population in 1994 living on less than $1 per day (on a purchasing power parity adjusted basis) and seven-eighths of the population living on less than $2 per day.

The fiscal year runs from April through March.

See World Bank, India: Policies to Reduce Poverty and Accelerate Sustainable Development, Report No. 19471-IN (2000); and P. Krishna and D. Mitra, “Trade Liberalization, Market Discipline and Productivity Growth: New Evidence from India,” Journal of Development Economics, Vol. 56 (1998), pp. 447–62.

The projection for growth in 1999/00 is based on Central Statistical Organization Advance Estimates.

Unfortunately, the most recent official poverty statistics are for 1993/94, and do not fully capture the impact of the rapid growth in the post-crisis period. However, unofficial figures suggest that the poverty rate may not have declined appreciably during the last decade.

See S. Tendulkar, “Indian Economic Policy Reforms and Poverty: An Assessment,” in India’s Economic Reforms and Development: Essays for Manmohan Singh, ed. by I.J. Ahluwalia and I.M.D. Little (Delhi: Oxford University Press, 1998).

Most analysts believe that a significant reduction in the poverty rate would require an annual real GDP growth rate of at least 7 percent (5 percent in per capital terms) on a sustained basis. The goverment’s Ninth Five-Year Plan (1997–2002), which targeted an average real GDP growth rate of 7 percent, projected that the official poverty rate would be reduced by 11 percentage points by the end of the plan period.

See P. Reynolds, “Fiscal Adjustment and Growth Prospects in India,” in India: Selected Issues, IMF Staff Country Reports (Washington: International Monetary Fund, forthcoming).

The percentage of the population in developing countries living under the $1 per day line has declined from 30 percent to 24 percent during the past decade (largely reflecting poverty reduction in east Asia and to a lesser extent in south Asia), but the incidence of poverty remains unacceptably high (Figure 4.3).9 The number of people living on incomes below $1 per day has stayed roughly constant at 1.2–1.3 billion because declines in poverty rates have been broadly offset by population growth (1.9 percent a year in developing countries). Moreover, progress in poverty reduction has been uneven across geographic regions. There has been no progress in reducing poverty rates in Africa and in the Western Hemisphere, and poverty rates actually increased in the transition countries.

Figure 4.3.Developing Countries: Population Living Below $1 per Day

(Percent of total population)

Poverty rates worldwide fell slightly during the past decade. Progress across regions, however, was uneven with, poverty rates decreasing substantially only in east Asia and, to a lesser extent, in south Asia. The thickness of the bars reflects the total population in each region (except for the total).

Source: Martin Ravallion and Shaohua Chen, “Global Poverty Measures 1987–98 and Projections for the Future,” forthcoming.


2East Asia comprises the developing countries in Asia except those in south Asia, as defined in footnote 3.

3South Asia comprises Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan, and Sri Lanka.

It should be noted that poverty is a multidimensional phenomenon, reflecting not only material deprivation but also, for example, lack of security and access to basic services including health, education, and sanitation. Health measures (for example, life expectancies) have converged more rapidly across countries than average incomes, in part because of the diffusion of medical technology. This aspect of globalization has unambiguously helped the poorest countries. Health and other dimensions of poverty are not captured in the monetary measures of poverty discussed in this chapter. See Box 5.1 fo an alternative measure of economic progress that combines income, health, and education into a single summary index.

Finally, it is important to recognize the benefits of policy reforms and other accomplishments in the countries that have grown rapidly both over the full 30-year period under review and more recently. However, it is equally important to recognize the risks of slippage in the reform process or other setbacks in these very successful countries, especially the larger ones. Further, the 1997–98 financial crises demonstrated that severe recessions in a small number of countries can reverse some of the past progress in poverty reduction, while producing significant, adverse spillover effects on poverty levels in trading partners. Moreover, some have argued that over the short run economic cycles have an asymmetric effect on poverty—that is, poverty increases more in an economic contraction than it decreases in an equivalent expansion.10

Explanations for Diverging Performance: Analytical Complications

In the main, the empirical literature on growth and development attempts to explain the observed differences in growth and prosperity across countries by “conditional convergence.” That is, each country’s economic growth rate depends on a number of underlying conditions, including national endowments, preferences, macroeconomic and domestic stability, institutional and social structures, outward orientation, and the state of financial market development. The primary thrust of these studies is to identify the conditions—some of which are policy related—that catalyze growth by promoting investment in physical and human capital, improving the efficiency of production (including the allocation of resources), and encouraging technological progress.

Unfortunately, the empirical studies generally provide only limited guidance and do not identify a single set of variables that strongly promote or inhibit growth.11 For many of the concepts tested, the correlation with growth is often not robust to small changes in the variables being studied (that is, the conditioning set of other factors that may help explain differences in growth).12 In general, factors that are considered robustly and positively correlated to growth include the share of investment in GDP, school enrollment, health indicators, openness to trade (which can be considered a proxy for outward orientation), and the share of nonprimary goods in total exports. Factors that are robustly related to poor or negative growth include weak institutional structures, measures of political instability (such as the absence of civil liberties), weak rule of law, wars, and market distortions.

The relevance and proper interpretation of these cross-sectional results have been called into question by a number of economists:

  • First, the observation that a variable is correlated with growth does not mean that it helps cause growth, and therefore, an observed relationship might provide only limited insight for policymakers.13 For example, while a higher investment share is correlated with (and is often thought necessary for) faster growth, this correlation may result because faster growth induces more investment (reverse causation) or because some other (omitted) factor induces both faster growth and more investment. Determining the causal relationship matters for policy advice. Simply increasing investment (without concern for the efficiency of that investment) may not be sufficient to promote growth.

  • Second, these results do not fully explain the growth performance of many individual economies (including many of the poor countries with negative per capita real growth over long time periods) or regions of the world, and indeed may explain well only the growth performance of industrial countries and a small set of developing countries. For example, empirical studies have found that regional dummy variables are needed to explain the growth performance of countries in sub-Saharan Africa, Latin America, and east Asia.14

  • Third, most cross-sectional analyses assume linear relationships between growth and the explanatory factors and miss important interactions among factors. There are compelling reasons to believe that some relationships are nonlinear and that interactions between growth and other variables are significant—that good policies tend to be mutually reinforcing and that policy complementarities are important.15

  • Fourth, cross-sectional studies generally assume that the growth process is the same in rich and poor countries. However, it has been argued that growth involves at least two dimensions: technological progress and catch-up.16 Factors that may best promote technological progress (most relevant for advanced economies) may not necessarily be the most optimal for low-income developing countries where takeoff and catch-up are most important.

Obstacles to Growth

Overall, factors related to growth can be grouped in six areas that, broadly speaking, influence capital and the labor force, and the efficient use of their services in production. Each of these six areas is reviewed below, tying together theory and evidence from the literature and data analysis tailored to the question of uncovering impediments to growth in the poorest countries.

In order to avoid some of the pitfalls of crosssectional studies, particularly the assumption of the homogeneity of countries at different stages of the growth process, the analysis in the rest of this section examines countries grouped by income level and growth performance. Countries are divided by income level according to the current World Bank classification of low-, middle-, and high-income countries. Developing countries in the low- and middle-income categories are then separated into eight groups (four for each income level) depending on their per capita growth rates over the period 1970–98 (Table 4.1).17 These subgroups comprise countries where per capita income is declining or regressing in absolute terms, those with slow (less than 2 percent) per capita growth that are regressing relative to the industrial countries, those that are slowly converging with the industrial countries (up to 3¾ percent growth), and those that are rapidly converging.

Table 4.1.Low- and Middle-Income Countries: Average Per Capita Income Growth, 1970–98
Low-Income CountriesMiddle-income Countries
Negative Per Capita Growth (Growth rates below 0 percent)
Burkina Faso-0.5Mali-0.5Djibouti-4.3South Africa-0.1
Burundi-0.3Nicaragua-2.4Lebanon-2.6Trinidad and Tobago-0.3
Central African Rep.-0.3Niger-0.9Libya-1.3Venezuela-0.3
Congo, Dem, Rep. of-4.3São Tomé and Principe-0.4
Côte d’Ivoire0.0Senegal-0.4
Ethiopia-0.1Sierra Leone-2.5
Ghana-0.6Solomon Islands-0.4
Slow Per Capita Growth (Growth rates of 0–2 percent)
Chad0.1Myanmar1.9Cape Verde0.5Morocco1.9
Congo, Rep. of1.2Nepal0.9Colombia2.0Panama1.5
Gambia, The1.5Nigeria0.1Costa Rica1.5Papua New Guinea0.5
Honduras0.5Tanzania0.1EI Salvador0.7Philippines1.0
Iran, I.R. of0.3
Slow Per Capita Convergence (Growth rates of 2–3 percent)
Bhutan3.3Pakistan2.2Brazil2.3Sri Lanka3.2
Cambodia2.7Vietnam3.4Chile2.5Syrian Arab Rep.2.5
india2.7Yemen, Republic of2.1Dominican Republic2.5Tunisia3.0
Lao P.D.R.2.8Egypt2.6Turkey2.7
Fast Per Capita Convergence (Growth rates above 3¾ percent)

The classification is intended for analytic purposes only and will be used to identify average characteristics of low-income countries and contrast these with higher-income and faster-growing economies. Clearly, the makeup of the growth subgroups could change somewhat, for example, if average growth rates were calculated over a different time period. The relatively long 1970–98 period was chosen to reduce the potential impact of business cycle effects, with the risk that some countries that have experienced higher growth rates more recently would be incorrectly placed in one of the lower-growth groups.18 It also needs to be recognized that in sorting countries by current income, the low-income group will naturally tend to have more countries with disappointing growth over the past than will the middle-income group. In other words, the two dimensions of the classification are not independent.

The Role of Investment, Saving, Human Capital, and Productivity

It is well established that the accumulation of physical and human capital and advances in production efficiencies and technology lead to higher per capita income. Studies have typically found that approximately 60–70 percent of per capita growth in developing countries reflects increases in physical capital and another 10–20 percent is due to increases in education and human capital with the remaining 10–30 percent attributed to improved (total factor) productivity.19

Not surprisingly, the low- and middle-income countries with declining or slowly rising per capita income had on average lower investment and saving rates than their faster-growing counterparts in recent years, confirming the importance of capital accumulation in the growth process (Figure 4.4). Causality is difficult to infer, however, because investment and saving rates were not substantially different, on average, across groups during the early 1970s except for perhaps the fastest-growing economies. Even in this latter group of countries, investment rates rose only after the growth takeoff.20 In other words, it is far from obvious that high initial investment and saving rates are preconditions for growth. It may indeed be that higher investment and saving rates result because of higher growth or that other factors cause both growth and investment.

Figure 4.4.Developing Countries: Investment and Saving1

(Percent of GDP)

Investment and saving rates were higher on average in countries with faster per capita income growth, although causality is hard to infer.

1Unweighted cross-country averages. For definitions of country groups, see Table 4.1.

Low levels of schooling or investment in human capital may be impediments to growth and also delay takeoff. Secondary school enrollment rates in the 1970s were substantially lower on average in nonrapidly converging, low-income countries than in the middle-income countries (Table 4.2).21 Moreover, the fastest-growing, low- and middle-income countries also experienced larger improvements in enrollments rates than the other developing countries did between 1975 and 1995. Although it is possible that growth induces more education as demand increases with income, it is noteworthy that among the low-income countries enrollment levels in the 1970s were highest (and similar to the levels in the middle-income countries) in the countries that subsequently grew the fastest.

Table 4.2.Developing and Advanced Economies: Education and Health Indicators1
Secondary School Enrollment Rate2Life Expectancy at Birth3
Country Groups19751995Increase19701997Increase
Law-Income countries
Negative growth1119843507
Slow growth11211044538
Slow convergence174124465913
Fast convergence346734556713
Middle-income countries
Negative growth376023556611
Slow growth406020596910
Slow convergence356126597011
Fast convergence36683259668
Industrial countries801153572776
Other high-income, non-oil countries456863070777
Middle- and high-income oil countries5296738536714
Source: World Bank, World Development Indicators.

Basic education, including training, can contribute directly to a country’s potential for growth by raising the skill level of the workforce. In addition, because physical and human capital are often complementary, education can also raise growth indirectly by inducing greater investment.22 While increased schooling and training alone may not be sufficient to boost growth, particularly when economic opportunities to use the acquired skills are missing, improving education will be an important part of a sustainable growth and poverty reduction strategy for developing countries. It therefore makes sense for countries to shift resources toward basic education and for the donor community to emphasize education as a high priority.

Another obstacle to a productive workforce (and society) is inadequate health care. As with school enrollment rates, life expectancy rates at birth were substantially lower on average in nonrapidly-converging, low-income countries than in the middle-income countries in the 1970s (see Table 4.2), and other health indicators show a similar pattern. Even though these health indicators have improved over time in most developing countries, they remain relatively bad in many low-income countries—for example, average life expectancy is still below 55 years for the negative- and slow-growth, low-income countries—representing an enormous loss in potential human capital. In addition, progress in improving life expectancy rates has slowed in some countries mainly because of the devastating effects of the AIDS epidemic. The repercussions from the epidemic (as well as from other infectious diseases) will have long-term effects on the economic performance of some poor countries, particularly in sub-Saharan Africa.23

Inefficient investment has also been a hindrance for many countries, although, again, causality is difficult to infer. Not surprisingly, in the developing countries with declining per capita growth during the last three decades, the incremental output-capital ratio (the inverse of the incremental capital-output ratio), which is a very rough proxy for the productivity of investment, was lower on average than in the countries that were growing (Figure 4.5). Estimates of total factor productivity growth, which are available for only a subset of the countries under review, also confirm that resources were not used as efficiently in many of the negative-growth countries as in other developing countries.24 Although a difficult task that needs to be addressed through a variety of reforms depending on country-specific circumstances, increasing productivity and allocative efficiency will allow these countries to better use their limited resources. To the extent that this and other resource reallocations can be accomplished relatively quickly, countries could begin to grow without immediate increases in saving and investment.

Figure 4.5.Developing Countries: Per Capita GDP Growth and Incremental Output-Capital Ratio1

Productivity, as proxied by the incremental output-capital ratio, was substantially lower on average in developing countries with negative growth than in other countries during the past three decades.

1 Unweighted cross-country averages. For definitions of country groups, see Table 4.1.

2Inverse of five-year moving average of incremental capital-output ratio.

The challenge then is to fashion the appropriate mix of policies and conditions for poor countries that removes impediments to the accumulation and efficient allocation of saving and investment (including in human capital) and allows growth to take off. Once growth has accelerated, there are many indications that a virtuous and mutually reinforcing cycle is possible as growth may further increase saving, investment, and productivity. The next sections will review some of the main obstacles to increasing saving, investment, and efficiency. These are generally considered to include macroeconomic instability, unsupportive institutions, inward-oriented and protectionist policies, poverty, income inequality, and (particularly in recent years) unsustainable external debt.

Macroeconomic Instability

Uncertainty caused by macroeconomic instability—such as high inflation, volatile and overvalued exchange rates, or excessive fiscal deficits—can significantly distort economic decisions and thereby reduce capital accumulation, hamper the efficient allocation of resources, and slow growth. Empirical studies have found, for example, that high inflation rates, which are often also accompanied by more variable inflation and relative price changes, have a negative impact on growth.25 Because the effect appears to be nonlinear, though, there is some disagreement over the precise threshold above which higher inflation becomes a detriment. Nevertheless, recent evidence indicates that inflation rates above the single digits have adverse implications for growth.

Inappropriate exchange rate regimes, generally in conjunction with high and variable inflation rates, often lead to overvalued exchange rates. These can impede the proper allocation of resources between the production of tradable and nontradable goods and may also deter inward foreign capital while encouraging capital flight. Persistently large fiscal deficits also inhibit growth through several mechanisms. As discussed in previous issues of the World Economic Outlook, fiscal deficits tend to crowd out private investment (or lead to higher inflation if the deficits are monetized) and inhibit financial market development.26 In addition, they may be indicative of excessive government intervention in the economy, although this is difficult to quantify when analyzing large groups of countries. Expectations about the financing needs associated with large fiscal deficits also create uncertainty about future taxes and future inflation and deter planning for investment.

Empirical studies have shown that fiscal deficits and overvalued exchange rates often have a negative impact on growth. Among low- and middle-income countries, other than the rapidly converging countries, there is little or no apparent relationship between growth or income and average inflation, its variance, average fiscal deficits, or average current account deficits (Table 4.3). Nevertheless, the fastest-growing, low- and middle-income countries on average generally had much lower and less variable inflation and lower fiscal and current account deficits than other developing countries (although this may partly be an outcome of stronger growth). This finding highlights that beyond a certain threshold high inflation and fiscal and current account deficits may be obstacles to growth. Moreover, large fiscal deficits and high and variable inflation can lead to misaligned exchange rates. It should be noted that the black market premium, the difference between the official and market exchange rates and an indicator of the extent of exchange rate overvaluation, was higher on average for developing countries that grew more slowly—particularly among the low-income countries. The strong correlation between this premium and growth underscores the problems caused by inappropriate exchange rate regimes, including excessively high costs for investment goods and consequently lower investment.

Table 4.3.Developing and Advanced Economies: Macroeconomic Stability1(Percent during 1970–98, unless otherwise noted)
Country GroupsInflation AverageInflation Standard DeviationFiscal Deficit2Current Account Deficit2Black Market Premium3
Law-Income countries
Negative growth26.426.36.26.5386.4
Slow growth16.
Slow convergence20.519.06.93.1112.2
Fast convergence9.
Middle-Income countries
Negative growth15.
Slow growth25.421.64.94.559.3
Slow convergence32.323.64.73.545.0
Fast convergence8.
Industrial countries6.
Other high-income, non-oil countries413.
Middle- and high-income oil countries57.17.91.7-2.30.5

Institutions and the Role of Government

Economic uncertainty increases when wars, military coups, political instability, and corruption are widespread or when basic institutional structures such as property rights, the rule of law, and those underpinning sound financial institutions are inadequate.27 In most countries plagued by these problems, firms and people face constraints on saving, investing, efficiently allocating resources, and profiting from legal economic activities. While there are exceptions to every rule, these economies will regress.28

It is hardly surprising that armed conflicts and civil wars—as well as the turmoil associated with political instability—sharply increase uncertainty and diminish the expected profitability of investments. Political instability also discourages long-term investment projects because of the risk of policy reversals. Studies have found that political, social, and government instability raised investment risk in the poor-growth countries of Africa and therefore was a major disincentive for foreign investors.29 For the broader group of developing countries, these factors are also negatively correlated with growth performance, particularly for the low-income countries (Table 4.4).30

Table 4.4.Developing and Advanced Economies: Political Stability and Institutions1(Average 1984–98)
Country GroupsPolitical and Social Stability2Government Stability3Law and Order4Contract Security5
Low-Income countries
Negative growth5.
Slow growth5.
Slow convergence6.
Fast convergence7.
Middle-income countries
Negative growth5.
Slow growth5.
Slow convergence5.
Fast convergence7.
Industrial countries8.
Other high-income, non-oil countries68.
Middle- and high-income oil countries77.
Source: International Country Risk Guide (published by the Political Risk Services Group).

Corruption also hinders growth by distorting incentives, as government officials and favored private individuals receive a larger share of public benefits or bear a lower share of the cost of public goods. Empirical studies have found that corruption lowers private investment, distorts the composition of public expenditures toward areas where collecting bribes are easier (for example, toward excessive and inefficient physical public investments and away from education), and tends to reduce government revenue because it contributes to tax evasion, improper tax exemptions, or weak tax administration, thereby limiting the ability of the government to provide critical services.31 There is evidence as well that corruption increases income inequality and poverty by reducing the progressivity of the tax system, the level and effectiveness of social spending, and the formation of human capital, and by perpetuating an unequal distribution of asset ownership and unequal access to education.32

The lack of a strong and impartial judicial system, including the inadequate observance of laws and the inability to enforce laws, also significantly increases economic and social uncertainty. In particular, inadequate property rights, such as the risk of contract repudiation or of expropriation without adequate compensation, substantially increase the risks to entrepreneurship. Indicators of these risks appear largely related to growth both for low- and middle-income countries (see the last two columns in Table 4.4).

Beyond maintaining peace and providing adequate institutional structures such as a sound judicial system, the government needs to provide social or public goods, including some basic services, education, health care, and when resources are available, social safety nets. Better education, family planning, and health care—especially for women—can not only improve living conditions, but can also enhance the productivity of the labor force and reduce population growth. This, in turn, may promote per capita income growth particularly when government and household resources are being strained by a rapidly growing population. The government may also need to make direct public investments to provide infrastructure, when this is lacking or inadequate. In general, though, the government’s role in the economy needs to be limited, particularly in areas of the economy where the private sector can efficiently provide goods and services and where markets are not distorted. Excessive intervention tends to foster corruption, can strain government budgets, and is prone to distort price signals and the efficient allocation of resources.33

Liberalization, along with proper supervision and regulation, is probably most important in the financial sector. The financial sector plays a primary role in intermediating saving and efficiently allocating investment. The financial system also plays a crucial role in smoothing the effect of shocks both on consumption and on investment. On average, low-income countries had a low level of financial sophistication, as proxied by the ratio of broad money to GDP, and slower-growing countries had lower rates of financial deepening, as proxied by the change in ratio of broad money to GDP, although for the latter causality is difficult to infer (Table 4.5). Cross-country empirical studies also confirm that slow growth is related to financial sector immaturity.34 Governments, therefore, need to foster a competitive and efficient financial sector, including a thorough supervisory and regulatory system.35

Table 4.5.Developing and Advanced Economies: Financial Maturity and Deepening1(Percent of GDP)
Broad Money
Low-income countries
Negative growth20.522.41.9
Slow growth29.523.5-6.0
Slow convergence22.935.912.9
Fast convergence17.994.676.7
Middle-Income countries
Negative growth31.860.829.1
Slow growth29.948.618.7
Slow convergence23.946.822.9
Fast convergence30.475.845.4
Industrial countries58.274.616.4
Other high-income, non-oil countries251.0125.274.2
Middle- and high-income oil countries311.949.137.3

Inward-Oriented Policies

Developing countries with inward-oriented, protectionist economic policies (such as in much of Africa and Latin America during the 1970s and 1980s) have suffered from poor growth rates, while those with outward-oriented policies (such as in east Asia) have prospered. A closed economy hinders growth through a variety of channels. First and foremost, countries that have not adopted an orientation toward open trade cannot exploit their comparative advantages in production. This may hurt the poor because import-competing industries tend to be capital intensive and thus, without outward orientation, production moves away from labor-intensive industries, which provide employment for the poor. When economies of scale are present, inward-oriented countries also cannot benefit from larger markets and the opportunities for greater specialization provided by trade. In addition, domestic prices are less likely to reflect world prices when a country maintains trade barriers, thereby leading to a worse allocation of resources. The lack of exposure to competition from foreign sources can shield production in the short run but at the cost of reducing incentives and opportunities for domestic producers to innovate and improve productivity.

Many empirical studies have found a robust and positive impact of increasing trade on growth and income (even while attempting to control for reverse causation).36 For the groups of countries under review, the poorer and slower-growing countries were on average less open compared to the faster-growing, middle-income countries, and generally faster growth accompanied increasing openness (Table 4.6).37 In addition, faster-growing, low- and middle-income countries benefited from rapid growth in their trading partners (Table 4.7). This can be explained by groups of countries or regions, such as east Asia, growing together and expanding intraregional trade. An alternative explanation is that faster-growing countries were better able to take advantage of the opportunities provided by increasing trade and globalization and to gain market share because of policies and conditions that were more supportive of export expansion. The evidence supports the latter explanation (see Table 4.7).

Table 4.6.Developing and Advanced Economies: Globalization and Trade Openness1(Percent of Purchasing Power Parity Adjusted GDP)
ExportsOpenness (Imports and Exports)
Country Groups1970–741994–98Increase1970–741994–98Increase
Low-income countries
Negative growth10.67.4-3.222.918.0-5.0
Slow growth10.58.4-2.123.920.6-3.4
Slow convergence4.88.13.316.419.53.0
Fast convergence4.
Middle-income countries
Negative growth16.415.2-1.231.433.2: 3
Slow growth17.818.10.336.438.52.1
Slow convergence8.410.92.518.423.45.1
Fast convergence18.228.19.938.555.516.9
Industrial countries20.536.916.444.274.530.3
Other high-income, non-oil countries250.781.030.3106.5160.754.2
Middle- and high-income oil countries332.440.
Table 4.7.Developing and Advanced Economies: Globalization and Trade Volumes and Prices1(Annual Percent Change 1970–98, unless noted otherwise)
Country GroupsPartner country output growthReal export growthGain in market share2Goods terms of tradeReal nonfuel commodities
Low-Income countries
Negative growth3.32.2-2.8-1.0-1.7
Slow growth3.44.3-0.8-1.0-1.6
Slow convergence4.46.2-0.40.4-2.2
Fast convergence4.511.14.21.9-1.7
Middle-Income countries
Negative growth3.20.7-4.10.9-1.6
Slow growth3.53.0-2.20.1-1.9
Slow convergence3.26.92.2-0.2-2.0
Fast convergence4.110.3410.4-1.7
Industrial countries3.25.60.9-0.6-1.4
Other high-income, non-oil countries34.29.43.1-0.8-5.0
Middle- and high-income oil countries44.53.1-3.72.8-0.6

Many developing countries are already advanced in their efforts to open their economies to trade.38 Unilateral liberalization began in the 1980s and early 1990s as many countries shifted away from inward-looking policies. Most progress in recent years was made in Latin America (in conjunction with structural reforms following the debt crisis) and eastern and central Europe, followed by southeast Asia, which was already relatively open to trade. Trade reforms began somewhat late in Africa, and many of these were in conjunction with IMF-supported structural adjustment programs.

Few industrial countries have allowed developing countries substantially unimpaired or unlimited access to their markets on a unilateral basis. Since the Uruguay Round, progress in expanding market access has been largely confined to regional and bilateral trade arrangements, such as those negotiated between the European Union and various developing country groups, including its neighbors and former colonies. These relatively recent agreements, though welcome, have tended to benefit selected developing countries and not necessarily the poorest ones. The present system of trade preferences excludes a number of “sensitive products” in precisely those sectors—primarily agriculture, textiles, and footwear—where many poor countries have the greatest potential to expand and diversify their exports. Moreover, the complexity, impermanence, and lack of transparency in these arrangements have discouraged the desired response in investment and trade. A bolder and more coherent approach to liberalization by the industrial countries would be needed to spur development and promote their integration into the world trading system. Such an approach would ideally combine the provision of duty-free and quotafree market access for products originating in developing countries with a reduction in agricultural subsidies in the advanced economies.

Although trade can create opportunities, it can also create difficulties for developing countries as they are exposed to externally generated shocks. Terms-of-trade shocks may have been a particularly important hindrance in the low-income countries with negative or low growth where terms of trade have fallen by 1 percent annually on average since 1970 (see Table 4.7). Most other developing countries, however, saw flat or even rising terms of trade. A key reason why some low-income countries had falling terms of trade is that primary products are a larger share of their exports and nonfuel commodity prices fell in real terms.39 The impact of future commodity price declines on these economies is likely to diminish automatically as they grow and become less dependent on primary products for export revenue. This process, however, can be expected to be slow.

Openness to capital flows holds risks for developing countries, but at the same time it allows countries to meet their financing needs when investment requirements exceed domestic saving. Moreover, private capital flows, particularly in the form of foreign direct investment, also provide access to new technologies and production processes through imported capital goods and management (or human capital).

During the last three decades, many developing countries (particularly the fast-growing, middle- income countries) have benefited from significant private capital flows and foreign direct investment. Exposure to capital flows, however, also makes countries vulnerable to externall generated financial shocks and increases the need for strong domestic and external macroeconomic and financial fundamentals (including adequate foreign reserves).40 For many low- and some middle-income countries, therefore, full liberalization of their capital account may need to wait until these countries are better able to manage external risks. This includes the ability to pursue monetary and exchange rate policies that are consistent with a liberalized environment. When liberalization does occur, measures need to be carefully sequenced.

It is important to dispel the notion that it is no longer possible for poor countries to benefit from globalization. Some have argued that globalization now offers few opportunities, stemming from the belief that an outward-oriented strategy based on exporting labor-intensive goods (and taking advantage of relatively lower labor costs) has become much more difficult because large countries such as China already dominate export markets in these goods. Essentially, the argument is that it is too late for those who have not yet started. However, the same argument was made in east Asia when first Japan and then Hong Kong SAR, Korea, Singapore, and Taiwan Province of China exported labor-intensive goods. This did not prevent China, Indonesia, Malaysia, and Thailand from following suit later when the other east Asian countries shifted to less labor-intensive exports as relative wages rose. Moreover, as incomes increased in the region, intraregional trade expanded rapidly, creating new markets for exports.

Poverty and Income Inequality

Monetary measures of poverty reduction, not very surprisingly, appear to be correlated with income growth (Table 4.8). One recent study confirms this relationship and finds that a 10 percent rise in per capita income is correlated with a 10 percent increase in income among the poorest quintile.41 Other studies correlate changes in aggregate income and poverty rates directly and generally suggest that a 1 percent increase in average per capita income or consumption is associated with a reduction in poverty rates by up to 3½ percent and also find that poverty rates fall, almost always, with growth in average living standards and rise with contraction.42

Table 4.8.Developing Countries: Poverty and Poverty Reduction1(Percent of population)
Below $1 per DayBelow $2 per Day
Country Groups198719982Reduction198719982Reduction
Low-income countries
Negative growth40.741.6-0.977.777.30.4
Slow growth44.443.60.888.636.91.7
Slow convergence47.742.35.485.270.914.3
Fast convergence28.216411.869.250.918.3
Middle-income countries
Negative growth7.08.6-1.632.134.4-2.3
Slow growth12.913.8-0.936.836.30.5
Slow convergence12.
Fast convergence10.
Source: Martin Ravallion and Shaohua Chen, “Global Poverty Measures.”

Although poverty reduction and economic growth are correlated, the causal relationship between poverty or income inequality and growth is unclear. In a subsistence economy, it is unlikely that saving and investment (including in human capital) will be sufficient to promote growth since income is required to provide basic necessities. High poverty rates can, therefore, be impediments to growth. The observation that income is more equitably distributed in the high-growth, countries of east Asia compared to some lower- growth Western Hemisphere countries also has led to speculation that income inequality hampers growth. However, both the theoretical relationship between growth and income inequality and the empirical evidence are ambiguous.43 If individual saving rates rise with the level of income, a less-even income distribution may actually result in a higher aggregate saving rate. However, significant income inequality also implies that low-income groups may be unable to acquire the skills necessary to benefit from economic opportunities, resulting in a potential loss of human capital.

Irrespective of the direct economic links between income inequality and growth, there are many other reasons to believe that the persistence of large income divergences and high poverty rates, even among many middle-income countries, can be a powerful obstacle to stronger economic growth. In addition to the waste of human potential, widespread poverty and large differences in income and wealth can be a source of social unrest and political instability. This in turn can prevent the establishment and maintenance of adequate institutional structures and policies, and eventually lead to an unstable economic environment that will adversely affect private saving and investment decisions and deter foreign investors.

In some cases, policies that promote growth may not promote immediate poverty reduction. Some studies have found evidence that policies such as fiscal stabilization and trade liberalization, which are both essential for sustained growth, may raise poverty rates in the short term.44 This may call for complementary reforms to strengthen social safety nets. In contrast, reducing high inflation rates is almost always beneficial for the poor. In any case, the short-run negative impact, if any, of necessary adjustment measures is typically modest compared with the long-run gains to the poor from the additional growth that results from these policies.

The Role of Debt

As discussed above, access to external finance is vital for many developing countries because domestic saving is usually insufficient to meet investment needs. In low-income countries, domestic saving is limited by factors such as poverty and underdeveloped financial markets. These countries typically do not have access to private finance and must rely on official lending (and aid). Many middle-income countries, by contrast, have access to private markets so that foreign capital flows, in the form of both debt and equity, can fill the financing gap and provide the resources to spur growth. Equity is an important component of financing from the perspective of recipient countries, not only because risk is shared by foreign stakeholders, but also because it often comes in the form of foreign direct investment that tends to bring in new technology and physical and human capital, including management expertise. However, because international capital markets may be imperfect and the acquisition of timely information on investment projects by perspective investors costly, countries often rely more heavily on borrowing than equity finance. There is no evidence that moderate external indebtedness hurts growth, but when indebtedness rises to the point that debt service becomes onerous, growth prospects are substantially damaged.

External debt can become an impediment to growth for a developing country when funds are not used for productive investments that allow the country to service its debt on time. One sign of an excessive debt burden is the regular need for comprehensive rescheduling or the protracted buildup of external arrears, as has been observed in many low-income countries.

Excessive levels of external debt are likely to reduce incentives for a government to undertake appropriate reforms and businesses to invest because of the real possibility that a significant share of the returns from these activities will need to be transferred to foreign creditors to repay the outstanding debt.45 Moreover, high debt levels create moral hazard: a government may delay reforms needed to reduce the debt burden because it expects debt relief in the future.

* * *

Clearly, there are many possible explanations for the failure of the poorest countries to catch up. Many of these reasons are interrelated and contribute to a vicious circle. While it may be difficult to identify unambiguously the most critical factors, there are many shortcomings that appear to be contributing to, or at least be associated with, persistently inadequate growth. This calls for a broad-based, sustained effort if greater progress is to be achieved. This effort, first and foremost, will need to be undertaken by the poorest countries themselves: without the strongest commitment on the part of their leaders and elected bodies, supported by society at large, success will be elusive. But the poorest countries cannot succeed without support from the international community. The advanced economies, in particular, will need to increase levels of foreign aid and assistance to countries that have strengthened their own efforts to alleviate poverty. The international community also needs to liberalize fully trade in products in which the poorest countries have a comparative advantage, and the debt burden also needs to be addressed.

Debt Burden and Debt Relief

Over the past 30 years, the total external debt of developing countries has risen sharply from $90 billion in 1970 (or 15 percent of GDP) to almost $2,000 billion in 1998 (37 percent of GDP; Appendix Table 38). Debt burdens vary across countries, and although the debt profile for some countries has improved substantially since the debt crisis in the 1980s, for many low-income countries the picture is bleak. At the turn of the millennium, up to about 40 of the poorest developing countries, the Heavily Indebted Poor Countries (HIPCs), still have unsustainable debt burdens even after large-scale and persistent financial assistance provided by official donors. Almost all of them (38 countries) are low-income countries, and about half of them (21 countries) experienced negative per capita income growth over the past 30 years.

Worldwide events in the 1970s and at the beginning of the 1980s were major contributors to the debt buildup in both the HIPCs and some middle-income countries.46 Oil price shocks and the rise in industrial country interest rates generated balance of payment pressures for many developing countries. These factors were exacerbated by stagnant or contracting export revenues, caused mainly by the subsequent world recession and declining commodity prices, and by the appreciation of the U.S. dollar, which inflated the domestic value of dollar-denominated debt and service payments and depressed dollar commodity prices.

Domestic factors also played a large role in the debt buildup. (Compare the performance of both the high-debt countries and “successful adjusters” to the low-debt countries in Figure 4.6. See figure footnotes 1–3 for group definitions.) Countries with low saving and large current account deficits were unable to withstand the impact of these external shocks and continued to borrow heavily as their reliance on external financing grew. Because low export and fiscal revenue shares of GDP made it difficult for both the private and the public sectors to set aside the resources required to service old and new debt, external debt quickly rose to unsustainable levels. The drag on domestic resources, coupled with already low private and public investment, further constrained the potential for growth and exports, in some cases inducing a vicious cycle of unsustainable indebtedness and low growth.

Figure 4.6.Developing Countries: Economic Indicators

(Percent of GDP unless otherwise noted)

High-debt countries rank below low-debt countries in all macroeconomic and structural indicators. The successfully adjusting countries reduced fiscal deficits and increased saving and exports.

Sources: IMF, World Economic Outlook; and World Bank, World Development Indicators.

1HIPC countries plus countries whose debt-to-export ratio was above 200 percent in 1996–97.

2Countries whose debt-to-export ratio was always below 200 percent in the 1971–97 period (based on three-year averages).

3Countries whose debt-to-export ratio went temporarily above 200 percent during the 1971–97 period.

4Average of imports and exports of goods and services as a percentage of GDP.

5Ratio of M2 to GDP (in percent).

International private investors, increasingly alarmed by the financial conditions of debtor countries, became more resistant to granting new lending or debt rollovers, further exacerbating the difficulties of these countries. In order to alleviate such difficulties, official creditors initially provided nonconcessional rescheduling of debt-service payments through the Paris Club, under the presumption that the debt crisis was temporary and liquidity related. This nonconcessional rescheduling, although granting temporary cash-flow relief, contributed to increasing the outstanding stock of debt.47

In the 1980s, it became apparent that the debt crisis involved more than temporary liquidity problems and actions beyond cash-flow rescheduling would be needed. The international financial community therefore launched what would become a series of assistance initiatives aimed at reducing the value of future obligations and repayments. In association with adjustment programs supported by multilateral institutions, middle-income countries with excessive debt arranged market-related debt reduction deals with their creditors under the Brady Plan, while poor countries received concessional rescheduling from official creditors and new lending on increasingly concessional terms. The assistance schemes were often complemented by, if not conditional on, macroeconomic and structural reforms undertaken by the recipient countries.

These concerted efforts, by both the international community and recipient countries, proved to be effective in helping many developing countries adjust their resource balances, reduce outstanding debt ratios to sustainable levels, and resume steady output growth. The “successful adjusters” in Figure 4.6 were mostly in the middle-income group and had access to international financial markets. As a group, they drastically reduced the public sector deficit, which helped to increase aggregate savings and reduce the trade deficit. At the same time, countries initiated macroeconomic stabilization, privatization, and market liberalization programs, which all contributed to the resumption of steady growth, although in many cases the adjustment process lasted longer than anticipated. As a result, and with the global recovery, exports began to rise and debt ratios started decreasing.

However, for many low-income countries that did not have access to external private financing, it became evident by the mid-1990s that traditional debt-relief mechanisms, official assistance, and adjustment policies and their implementation were insufficient to reduce excessive indebtedness. Their economic performance has lagged and debt ratios have remained high, despite a series of Paris Club Initiatives (concessional reschedulings known as Toronto, London, and Naples terms), and additional action by a number of non-Paris Club bilateral and commercial creditors, aimed both at providing immediate cash-flow relief on the payments falling due and slowing the growth of the debt stock itself. The costs of these early debt-relief initiatives for low-income countries have been estimated to be at least $30 billion, and possibly much more.48

Many of these poor countries had limited success in improving their debt-burden indicators and rekindling output growth because they were unable to provide a stable macroeconomic environment, as demonstrated by persistently high fiscal deficits and increasing inflation. (See the group of high-debt countries in Figure 4.6.) Accordingly, aggregate saving fell further, and profitable investment opportunities were not exploited. High debt-service payments continued to exacerbate already wide current account deficits or forced countries to run arrears. By the mid-1990s outstanding debt of these countries remained high at around 600 percent of exports on average (compared to about 150 percent for the successfully-adjusting countries), despite large flows of official assistance.

Inappropriate choice or implementation of adjustment policies was not the only reason why some countries were unsuccessful in reducing external indebtedness. Conditions and policies prior to the debt crisis may have played an important role. “Successful adjusters” had larger fiscal deficits on average during the initial phase of the debt buildup than high-debt countries had, suggesting that low public saving was a larger factor in the rapid accumulation of debt for the former group while low private saving may have been a larger factor for the latter. To the extent that government policies can more easily address public saving (by reducing the fiscal deficit) than private saving, adjustment may have been easier for the countries that started with higher public sector deficits. Also, low private saving in the unsuccessful countries may be partly a reflection of much lower average per capita income in this group.

The Enhanced HIPC Initiative: The Road to Sustainability?

In September 1996, the IMF and the World Bank jointly launched the Initiative for Heavily Indebted Poor Countries, or HIPC Initiative. It was based on concerted financial assistance from the international community (in addition to traditional debt-rescheduling mechanisms) and was available to countries that demonstrated a record of successful macroeconomic and structural adjustment and met other eligibility criteria. By 1999, seven countries had qualified for assistance under the Initiative.49

A central objective of the HIPC Initiative is to reduce external debt of the qualifying countries to sustainable levels. In theory, a debt burden is sustainable if the debtor country can be expected to meet its current and future debt-service obligations without recourse to debt relief or similar assistance. In practice, however, identifying a maximum or target level of debt that can be expected to be sustained is far more difficult because it involves the uncertainties of projecting future earnings that would be used to service the debt. For a nation, exports are typically the appropriate measure of earnings as these provide the necessary foreign currency for external debt service, but for some countries, government revenue may be appropriate (for example, when the debt is mostly public and when tax revenue is a small proportion of output). Estimating future export earnings (or output and tax revenue) is always difficult due to unforeseen factors such as shocks; these estimates are especially uncertain for heavily indebted countries because the debt level itself can affect output through disincentives for policymakers and investors.

Sustainable debt levels are typically expressed in terms of net present value, a concept that is estimated as the stream of future scheduled debt-service flows discounted to today’s value at market interest rates. The net present value of debt can differ from contract amounts (or nominal face value) when loans are made at concessional, below market, interest rates or when debt has been rescheduled at concessional rates. Thus, the difference between contract amounts (Table 4.9, first column) and the estimated net present value (second column) provides a rough indication of the concessional component of lending and of debt relief associated with traditional rescheduling techniques.50 For all HIPCs, this difference amounts to about $100 billion. It does not reflect the impact of debt forgiveness, which lowers debt both in nominal and in net present value terms.

Table 4.9.Heavily Indebted Poor Countries (HIPCs): Selected Debt Indicators1(Billions of dollars, at the year prior to the decision point, unless otherwise indicated)
Debt in Net Present Value
Nominal debtAverage ratio to exports (percent)Average ratio to revenue (percent)Target (at 150 percent of exports)Target (at 250 percent of revenue)Estimated HIPC assistance at decision point2Estimated HIPC assistance in end-1999 terms3
Group 1441.122.52423237.77.7
Qualifying under exports criteria21.811.12882975.95.25.2
Qualifying under fiscal criteria19.411.31503758.92.42.4
Group 2566.131.634952114.713.9
Qualifying under exports criteria53.422.436550912.410.09.4
Qualifying under fiscal criteria12.79.32835664.54.84.5
Other HIPCs681.836.92754268.06.6
Qualifying under exports criteria34.612.74156156.76.04.8
Qualifying under fiscal criteria6.04.22054212.22.01.8
Deemed sustainable41.319.099145

The HIPC Initiative targets the net present value of debt expressed as a percentage of exports or, in some cases, government revenue. These targets were established based on past experience of countries that have successfully avoided debt servicing problems.51 On the basis of these targets, debt relief was already extended or committed to seven qualifying countries, while the debt burden of two other countries was deemed sustainable after traditional debt-relief mechanisms. For all of these nine countries, the debt in net present value terms was reduced to $22½ billion, or 240 percent of exports in 1998 (see Table 4.9, Group 1), and compares to a total nominal debt level of about $41 billion. The difference reflects the impact of past concessional lending and debt restructuring.

In September 1999, after a review of the initial implementation phase, it was decided to enhance the HIPC Initiative to provide deeper, faster, and broader debt relief, and to accelerate poverty reduction (Box 4.3). The enhanced Initiative specifies lower sustainability targets, more flexible eligibility criteria, front-loaded debt relief, and a specific link between debt relief and poverty reduction. Debt relief under the enhanced HIPC Initiative would be tied to the recipient country’s adoption and implementation of a poverty reduction strategy (see Box 4.1).

In particular, debt-sustainability targets were lowered as a direct way of providing deeper debt relief for qualifying countries and to broaden the list of countries that could potentially qualify for assistance. Thus, debt relief will now be aimed at reducing the net present value of debt to 150 percent of exports or, for those eligible under the fiscal window, 250 percent of government revenue, whichever provides greater debt relief to the recipient country. Deeper debt relief is likely to have several advantages. First, it is hoped that additional debt reduction will free resources for poverty reduction and other important objectives set in the country’s Poverty Reduction Strategy Paper. Second, the more ambitious targets of the enhanced HIPC Initiative will provide a greater safety margin for the achievement of debt sustainability and increase the chances of a permanent exit from the need for debt rescheduling. This in turn will improve the chances of future economic success by strengthening incentives for economic reform and private investment and reducing moral hazard.

Under the enhanced HIPC Initiative, all the countries in Group 1 may be reassessed for further assistance in light of the new sustainability targets. Five countries are expected to qualify under the export criteria and three under the fiscal criteria. These countries are expected to receive debt relief of approximately $8 billion (in net present value terms in 1999). Countries that are expected to qualify for assistance by the end of 2000 (Group 2 in Table 4.9) have higher average debt in net present value terms than the countries that received assistance under the original HIPC Initiative (Group 1). For the Group 2 countries, past concessional arrangements and the full use of traditional debt-relief mechanisms would lower the total debt burden from a nominal value of $66 billion to a net present value measure of about $32 billion (in 1999 terms), but before HIPC relief. Under current plans, and assuming all countries in the group can proceed with debt relief as expected, debt in present value terms would be halved, to about $17 billion.52 Debt-relief estimates for other HIPCs in Table 4.9 are very preliminary.

Box 4.3.Social Spending, Poverty Reduction, and Debt Relief in Heavily Indebted Poor Countries

Heavily Indebted Poor Countries (HIPCs) are characterized by low income, high ratios of debt to exports, a heavy debt-service burden, and poor social indicators. On average, public spending on education and health care is lower in HIPCs than in other countries eligible for concessional assistance under the Poverty Reduction and Growth Facility (PRGF), both in relation to GDP and total government expenditures (Table B4.3).1 The increase in education and health care spending and improvements in key health and education indicators during 1985–98 have been generally lower in HIPCs than in other PRGF-eligible countries (see the first figure). Poverty is a multidimensional phenomenon, reflecting not only material deprivation but also, for example, lack of access to basic services. For targeted spending to have a broad impact on poverty, the benefits from improved basic social services have to be accompanied by greater income-earning opportunities for the poor, of the type generally produced by robust economic growth.

Debt Relief and Poverty Reduction

Debt relief for HIPCs could provide additional resources for anti-poverty programs. In the HIPCs for which data are available, debt service paid and the stock of debt as a share of GDP are higher than in other poor countries eligible for concessional lending under the PRGF. Although country experiences vary considerably, on average, HIPCs allocate slightly more budgetary resources to debt service than to education and health-care taken together, both in terms of GDP as well as total government expenditures (about 5 percent of GDP and 20 percent of total government expenditures, respectively). Total debt relief under the Heavily Indebted Poor Country (HIPC) Initiative is estimated at $28.2 billion (in end-1999 present value terms), including debt forgiveness and concessional debt service.2 However, comparing data on debt service paid and public spending on education and health care alone provides only a partial view of the relationship between international official resource flows and spending on poverty reduction programs, because public spending on poverty reduction is not confined to education and health care programs. Public spending on, for instance, rural roads, sanitation, and access to safe water could contribute as much, or more, to poverty reduction. Gross external financing flows (new loans and grants, for instance) are not necessarily independent of debt-service payments, and these flows are typically higher than either social spending or debt-service payments in HIPCs. For countries to reap full benefits from debt relief, efforts are needed to improve the efficiency of existing, as well as additional, public outlays made possible by debt relief. This will require action on at least three fronts.

Social Spending and Indebtedness: HIPC and Non-HIPC PRGF-Eligible Countries

(In units as indicated; latest year for which data are available)1

Education Spending2Health Spending2Debt Service3
In percent of GDPIn percent of total government expendituresIn percent of GDPIn percent of total government expendituresIn percent of GDPIn percent of total government expendituresDebt Stock In Percent of GDPNumber of Countries4
of which program countries3.413.
Non-HIPC PRGF-eligible countries64.615.
of which program countries3.915.
Sources: World Bank; national authorities; and IMF staff estimates.

Social Spending and Social Indicators, 1985–981

(Annual percent change)

Sources: IMF staff estimates; national authorities; and World Bank, World Development Indicators.

1 Average annual change between first year since 1985 and the most recent year for which data are available.

2Excludes transition economies: Albania, Armenia, Azerbaijan, Bosnia and Herzegovina, Georgia, the Kyrgyz Republic, the former Yugoslav Republic of Macedonia, Moldova, Mongolia, and Tajikistan.

In HIPCs, the allocation of budgetary resources within the social sectors (e.g., between primary and tertiary education) is typically skewed toward services that are less demanded by the poor, such as tertiary education and curative, rather than preventive, health care. For instance, data for selected HIPCs show that 14 percent of total spending on public education and 13 percent of health care outlays, on average, accrue to the poorest fifth of the population (lowest quintile), compared with 30 percent for the richest quintile for both spending categories (see the second figure).3 Furthermore, public spending on primary and secondary education has disproportionately favored the more prosperous urban population over the rural poor.4 In addition, corruption, poor targeting of social spending, and a high share of wages and salaries in social sector allocations have weakened the link between public expenditures and improvements in social indicators.5

Benefit Incidence of Public Spending on Education and Health Care in Heavily Indebted Poor Countries, Early 1990s1

(Percent of total spending)

Source: Davoodi and Sachjapinan, “How Useful are Benefit Incidence Studies?”

1 Numbers of countries are in parentheses; the latest year for which data are available.

2The share of the population living in urban areas averages one-third for the countries in the education and health care sample

The capacity to formulate and execute the budget in a transparent manner is critical for anti-poverty programs. In this context, the use of poverty funds for channeling debt relief to the poor has been widely advocated. Such funds should be subject to the same level of scrutiny and oversight as other government spending, including transparency and accountability both to parliament and the executive. Anti-poverty programs should be integrated into the budget to prevent leakages, duplication, and implementation of relatively low-priority projects.

Effective monitoring of debt relief for poverty reduction requires timely and adequate data.6 To this end, coordinated technical assistance from international organizations and donors, as well as increased efforts by HIPCs themselves, will be crucial.7 At present, coverage of public spending data is not comprehensive in many countries, and typically excludes spending by local governments and in-kind donor contributions. Public spending data become available with a lag, which for some countries can be as long as two to three years. Moreover, virtually no HIPC-eligible country has consistent annual series for public expenditure allocations within the education and health sectors (e.g., distinguishing between outlays on primary and tertiary education, preventive and curative health care, and wage and nonwage items). Social indicators are produced infrequently, and typically become available only every five years. Also, data for many important indicators are derived from models rather than from actual observations. Up-to-date information on poverty indicators, including the nature and locus of poverty, is often lacking.

Currently, there are 40 HIPC-eligible countries, of which 8 countries have reached the decision point—i.e., the point when a country’s qualification for HIPC assistance is determined (Bolivia, Burkina Faso, Cote d’Ivoire, Guyana, Mali, Mauritania, Mozambique, and Uganda), and 4 countries have reached the completion point—i.e, the point when additional measures are taken to assist a country to reach a sustainable debt level (Bolivia, Guyana, Mozambique, and Uganda). In addition, Bolivia and Uganda have reached their second decision points. Benin and Senegal reached decision points, were determined under the Initiative to face sustainable debt burdens after traditional debt relief, and therefore did not receive assistance. There are 40 other PRGF-eligible countries in addition to HIPCs.

Information is available via the Internet at

Hamid Davoodi and Sawitree Sachjapinan, “How Useful Are Benefit Incidence Studies?” IMF Working Paper (Washington: International Monetary Fund, forthcoming).

In a smaller sample of HIPCs for which information is available, public spending on health favors the rural population over the more prosperous urban population.

Vito Tanzi, “Corruption Around the World: Causes, Consequences, Scope, and Cures,” IMF Staff Papers, Vol. 45 (December 1998), pp. 559–94.

Sanjeev Gupta and others, Social Issues in IMF-Supported Programs, IMF Occasional Paper No. 191 (Washington: International Monetary Fund, 2000).

Currently, the IMF, the Organization for Economic Cooperation and Development, the United Nations, and the World Bank are collaborating in the Partnership in Statistics for Development in the Twenty-First Century Consortium (PARIS21) to support statistical capacity building in countries preparing poverty reduction programs.

To qualify for HIPC assistance, countries need to implement a program of macroeconomic and structural reform and demonstrate a track record of good performance because experience suggests that progress in these areas is a necessary precondition for successful debt reduction.53 An improved track record from 1995 onward is evident in the countries that received relief under the original HIPC Initiative (Group 1 in Table 4.10). While it is difficult to determine causality, better policies and implementation are evident in lower fiscal deficits and inflation. These and other policy improvements have contributed to a stronger resource balance and ability to service debt (a rise in domestic savings and an improving trade deficit, associated with larger export and fiscal revenues), higher levels of private and public investment, and stronger private capital inflows. The Group 2 countries that are expected to qualify for HIPC assistance in the course of 2000 also recorded improvements in these economic variables, but their situations today remain substantially worse than those in Group 1 in 1995–97. This lack of sufficient progress indicates that many countries now anticipating HIPC debt relief will need firm implementation of policy reforms to qualify for assistance.

Table 4.10.Heavily Indebted Poor Countries (HIPCs): Economic Indicators(Percent of GDP unless otherwise indicated)
Group 11Group 22Developing Countries excluding HIPCs
Real per capita GDP30.43.42.5-2.11.6-
Total investment18.621.122.716.415.517.824.226.324.9
Private investment10.
Fiscal deficit4-5.3-2.1-2.3-6.8-3.2-4.0-4.1-3.1-3.7
Current account-9.3-8.5-8.8-9.6-8.7-10.2-4.6-3.5-4.0
Exports of goods and services28.431.431.
Fiscal revenues421.622.121.917.218.818.428.527.727.1
CPI Inflation315.19.13.731.015.711.819.710.27.5
Broad money26.628.228.819.118.818.949.149.4
Met private capital flows1.63.73.7-
Official assistance612.52.15.314.812.
Sources: World Bank, World Development Indicators; and IMF staff estimates.

The New Approach to Poverty Reduction

Poverty reduction has been and will need to be for a long time among the highest priorities for domestic policymakers, the donor community, and international institutions. The development strategies followed until now have worked for some countries, but for far too many, poverty remains prevalent and a stronger approach is needed.

There are, nevertheless, grounds for optimism. First, the industrial economies have demonstrated that long periods of sustained growth are possible and that living standards can improve dramatically over a few decades. Second, the newly industrialized economies, although small in number, have shown that rapid catch-up is possible with growth rates even higher than those observed during the periods of the strongest growth in the industrial countries, especially when the familiar pillars of sustainable growth—macroeconomic stability, sound institutions, and free trade—are pursued. Third, a few low-income countries—in particular China and to a lesser extent India with their large populations—have also achieved long periods of sustained growth at a rapid pace. Taken together, the successes of these groups of countries show that countries with different characteristics can achieve significant increases in per capita income. Finally, and more recently, successful implementation of growth-oriented adjustment policies and reform is clearly paying off in some low- and middle-income developing countries that have witnessed falling inflation and stronger rates of economic growth in the 1990s. More needs to be done, of course, to ensure that these improvements endure.

Still, while the fundamentals of economic growth are reasonably clear, there is no unique formula to achieve it in all countries. Solutions will vary and depend on country-specific institutions, customs, and economic conditions. Experience identifies a few key conditions that, if lacking or absent, can be an impediment to takeoff and sustained growth. Among these are an incentive structure, the rule of law, and a level of governance that allow individuals to save, invest, and ultimately benefit from these endeavors. The enhancement of public health and education standards are also important for sustainable growth, as they not only directly contribute to human well-being but also allow for an efficient accumulation of productive human capital. Each country will need to decide on how best to provide these fundamentals within its development strategy.

The enhanced development strategy framework stresses three elements. First, each country will formulate a poverty reduction strategy to elaborate its plans for development and poverty reduction in a multi-year framework. These strategies will be prepared by the country itself, with the participation of civil society, in order to increase the level of national awareness of the issues and strategy, and with it the shared commitment to implementing the agreed reforms. Greater ownership by the domestic authorities and civil society is expected to lead to stronger policy implementation. Development partners, the World Bank, and the IMF will provide a broad range of assistance. The poverty reduction strategy, when broadly endorsed by the Executive Boards of the IMF and the World Bank, will form the basis for concessional lending from both institutions. It is hoped that other development partners will also link their financial and technical support to these poverty reduction strategies.

Second, poverty reduction is now placed at the center of programs supported by the IMF’s facility for concessional lending. This follows from the recognition that poverty reduction itself is a primary objective, but also the belief that by jump-starting poverty reduction, countries can begin a virtuous circle of domestic saving and investment leading to growth and further poverty reduction. Consistent with the change in objectives and practices, the IMF’s facility for concessional lending—formerly the Enhanced Structural Adjustment Facility—has been renamed the Poverty Reduction and Growth Facility.

The third element of the strategy is to bring enhanced debt relief to poor countries where debt levels are unsustainable. Without debt reduction through traditional means and the HIPC Initiative, the incentive problems associated with unsustainable debt burdens will not be overcome, and policy reforms and private investment increases may not be forthcoming. The cost to debtor countries of any resulting delay would come in forgone output. The experience over the past 30 years during which per capita income growth stagnated in many of these countries demonstrates the magnitude of the potential loss. For the donor countries, the cost of inadequate debt reduction at this stage is the possibility of more debt problems and further rescheduling down the road that could come at an even greater cost to taxpayers.

* * *

The deepening of debt relief in the enhanced HIPC Initiative clearly moves in the right direction and adds an extra margin of safety that debt burdens will be brought to sustainable levels. The enhanced HIPC Initiative is an opportunity that none can afford to miss. The HIPC Initiative is no panacea for all poverty and economic problems in these countries, however, and its goals can only be achieved with continued hard work by domestic and international participants. The international community needs to increase levels of foreign aid and ensure that aid promotes reforms and poverty alleviation. Major efforts also are needed to reform trade policies that adversely affect the poorest countries. The poorest countries themselves will need to persevere with macroeconomic and structural reform, with emphasis on providing an environment conducive to private saving and investment decisions, including better governance, public sector reform, and market liberalization. Their responsibility to stay the course is a sine qua non part of the strategy.

See Entering the 21st Century: World Development Report, 1999–2000 (Washington: World Bank, 1999), p. 21.

See, for example, the May 1997 World Economic Outlook. For a discussion of complementarities across policies, see Robert F. Wescott and Jahangir Aziz, “Policy Complementarities and the Washington Consensus,” IMF Working Paper 97/118 (Washington: International Monetary Fund, 1997), and Craig Burnside and David Dollar, “Aid, Policies, and Growth,” Policy Research Working Paper 1777 (Washington: World Bank, 1997).

Absolute poverty refers to the number of individuals living in poverty conditions, often defined in terms of internationally comparable monetary measures, while relative poverty refers to income differences within a country.

D. Gale Johnson, “Population, Food, and Knowledge,” American Economic Review, Vol. 90 (March 2000), pp. 1-14.

Developing countries in this chapter are those classified as low income (1998 GNP per capita of $760 or less, calculated using the World Bank Atlas method) or middle income ($761 to $9,860) by the World Bank in Entering the 21st Century: World Development Report, 1999–2000. Countries heavily dependent on oil exports (Equatorial Guinea, Gabon, Oman, and Saudi Arabia), countries with populations less than 400,000, and countries in transition are excluded from the analysis. Korea, which is an advanced country in the World Economic Outlook classification, is considered a middle-income country in the World Bank classification and is included as a developing country in the analysis below.

Fast-growing or rapidly converging countries are defined for this analysis as those countries with annual per capita income growth of more than 3¾ percent. This cutoff was chosen for tractability and is similar to the cutoff chosen in Lant Prichett, “Divergence, Big Time,” Journal of Economic Perspectives, Vol. 11 (Summer 1997), pp. 3–17. Prichett finds 11 countries (both developing and advanced) in which growth exceeded 4.2 percent over the 1960–90 period. The relatively larger number of fast-growing countries in that study mainly reflects the inclusion of advanced economies.

It is important to note, however, that regional growth rates have varied considerably within China and India.

Interestingly, of the 14 rapidly converging countries, only one (China) also converged rapidly during the 1970–98 period, while two others had negative growth during the longer period, five had slow growth, and six converged slowly.

The pattern of poverty rates is similar when the poverty line is defined as $2 per day.

See “External Shocks, Financial Crises, and Poverty in Developing Countries,” Global Economic Prospects 2000 (Washington: The World Bank, 2000).

Robert J. Barro and Xavier Sala-i-Martin, Economic Growth (New York: McGraw-Hill, 1995), provide a review of the theoretical and empirical literature.

Ross Levine and David Renelt, “A Sensitivity Analysis of Cross-Country Growth Regressions,” The American Economic Review, Vol. 82 (September 1992), pp. 942–63, using extreme-bounds analysis, find that only one variable, the share of investment in GDP, is robust to changes in specification (although the trade-to-GDP ratio is also robust if investment is excluded in the conditioning information set). Xavier Sala-i-Martin, “I Just Ran Two Million Regressions,” The American Economic Association Papers and Proceedings, Vol. 86 (May 1997), pp. 178–83, and Xavier Sala-i-Martin, “I Just Ran Four Million Regressions,” NBER Working Paper 6252 (Cambridge, Massachusetts: National Bureau of Economic Research, 1997), however, argue that extreme-bounds analysis is too strong a test and find about two dozen robust variables (including regional and religion variables) by examining the cumulative distributions of coefficient estimates from a large number of regressions.

Instrumental variables are used in some of these cross-sectional studies to control for endogeneity. However, it is possible to find plausible reasons why these instruments are not truly exogenous because explanations for growth are so broad. Granger-causality tests, which assume only weak exogeneity, generally also provide mixed results.

See Sala-i-Martin, “I Just Ran Four Million Regressions,” which finds that regional dummies for sub-Saharan Africa and Latin America are negatively correlated with growth. The study also notes that the fraction of the population that is Buddhist or follows Confuscianism, which the author interprets to be similar to a dummy variable for the east Asian countries, is positively correlated with growth. It should be noted, however, that cross-country studies based on African countries alone find results similar to the more comprehensive studies. See, for example, Dhaneshwar Gupta and Michael T. Hadjimichael, “Growth in sub-Saharan Africa,” IMF Staff Papers, Vol. 43 (September 1996), pp. 605–34, and Dani Rodrik, “Trade Policy and Economic Performance in sub-Saharan Africa,” NBER Working Paper 6562 (Cambridge, Massachusetts: National Bureau of Economic Research, 1998). For an analysis focused on sub-Saharan Africa, see Chapter VI, “Growth in sub-Saharan Africa: Performance, Impediments, and Policy Requirements,” in the October 1999 World Economic Outlook.

Chapter IV, “Globalization and the Opportunities for Developing Countries,” in the May 1997 World Economic Outlook argues that macroeconomic stability, openness to trade, and limited government intervention in economic activity are all necessary conditions for growth and together these conditions substantially increase the probability of achieving fast growth. Burnside and Dollar, “Aid, Policies, and Growth,” conclude that aid has a positive impact on growth only in countries with good fiscal, monetary, and trade policies.

See Danny T. Quah, “Twin Peaks: Growth and Convergence in Models of Distribution Dynamics,” The Economic Journal, Vol. 106 (July 1996), pp. 1045–55. Steven N. Durlauf and Danny T. Quah, “The New Empirics of Economic Growth,” NBER Working Paper 6422 (Cambridge, Massachusetts: National Bureau of Economic Research, 1998) review recent studies using panel-data and distribution-dynamics econometric techniques that relax some of the assumptions about the homogeneity of countries and the growth process.

As mentioned above, countries heavily dependent on oil exports (Equatorial Guinea, Gabon, Oman, and Saudi Arabia), countries with populations less than 400,000, and countries in transition are excluded from the analysis, and Korea, which is an advanced country in the World Economic Outlook classification, is included as a developing country in this analysis because it is a middle-income country in the World Bank classification.

For example, if growth subgroups are calculated based on average annual growth rates during 1985–98, 35 countries (or approximately a third of the total) would change groups, but the number of countries in each group would remain largely the same and average characteristics (and hence conclusions) would not change substantially. It is worth noting, however, that the number of rapidly converging countries would increase by three compared to the classification based on 1970–98. The rapidly converging, low-income countries would include India, Mozambique, and Vietnam (but exclude Indonesia) in addition to China. The rapidly converging, middle-income countries would include Chile in addition to Botswana, Korea, Malaysia, Mauritius, and Thailand.

See, for example, Barry P. Bosworth and Susan M. Collins, “Economic Growth in East Asia: Accumulation Versus Assimilation,” Brookings Papers on Economic Activity: 2, Brookings Institution (1996), pp. 135-203.

Christopher D. Carroll and David N. Weil, “Saving and Growth: A Reinterpretation,” NBER Working Paper 4470 (Cambridge, Massachusetts: National Bureau of Economic Research, 1993) and Patricia Reynolds, “Does Growth Cause Saving and Investment?” (unpublished; Washington: International Monetary Fund, January 2000) examine the causal relationship between growth and investment or saving. The authors find that they cannot reject the possibility that causation runs from growth to investment or to saving or that some other (omitted) factor causes both growth and investment or saving—in other words, investment and saving rates are endogenous.

Literacy rates and primary school enrollment rates show a similar pattern.

See Per Krussell, Lee E. Ohanian, Jose-Victor Rios- Rull, and Giovanni L. Violante, “Capital-Skill Complementarity and Inequality,” Federal Reserve Bank of Minneapolis Staff Report 239 (Minneapolis, Minnesota: Federal Reserve Bank of Minneapolis 1997).

It is estimated that 33½ million people worldwide were infected with AIDS or the HIV virus by the end of 1999. About 32 million of these people were in developing countries—over 23 million in sub-Saharan Africa alone. This represents about 1 percent of the developing country workforce (adults aged 15 to 49 years), but 8 percent in sub-Saharan Africa. See UNAIDS and World Health Organization, “AIDS Epidemic Update” (Geneva: 1999).

See Bosworth and Collins, “Economic Growth in East Asia.”

See, for example, Michael Sarel, “Nonlinear Effects of Inflation on Economic Growth,” IMF Working Paper 95/56 (Washington: International Monetary Fund, 1995), and Atish Ghosh and Steven Phillips, “Warning: Inflation May Be Harmful to Your Growth,” IMF Staff Papers, Vol. 45 (December 1998), pp. 672–710.

See, for example, Chapter IV, “Fiscal Policy Issues in Developing Countries,” in the May 1996 World Economic Outlook.

The negative economic consequences of war are well recognized. Thucydides, a historian in ancient Greece, commented on this link in regard to the Peloponnesian war.

Empirical studies generally confirm that economic growth is hampered under these conditions. See, for example, Robert J. Barro, “Determinants of Economic Growth: A Cross-Country Empirical Study,” NBER Working Paper 5698 (Cambridge, Massachusetts: National Bureau of Economic Research, 1996).

See Paul Collier and Catherine Pattillo, Investment and Risk in Africa (London: MacMillan Press, 1999), Chapter I.

Similar results are found for indices of civil war.

See Paulo Mauro, “Corruption and Growth,” Quarterly Journal of Economics, Vol. 110, (August 1995), pp. 681–712; Paulo Mauro, ‘The Effects of Corruption on Growth, Investment, and Government Expenditure: A Cross-Country Analysis,” in Corruption and the Global Economy, ed. by Kimberly Ann Elliott (Washington: Institute for International Economics, 1997); and Vito Tanzi and Hamid Davoodi, “Corruption, Public Investment, and Growth,” IMF Working Paper 97/139 (Washington: International Monetary Fund, 1997).

Sanjeev Gupta, Hamid Davoodi, and Rosa Alonso-Terme, “Does Corruption Affect Income Inequality and Poverty?” IMF Working Paper 98/76 (Washington: International Monetary Fund, 1998).

See Tanzi and Davoodi, “Corruption, Public Investment, and Growth.”

See Ross Levine, “Financial Development and Economic Growth: Views and Agenda,” Journal of Economic Literature, Vol. 35 (June 1997), pp. 688–726.

See Manuel Guitian, “Banking Soundness: The Other Dimension of Monetary Policy,” in Banking Soundness and Monetary Policy, ed. by Charles Enoch and John H. Green (Washington: International Monetary Fund, 1997), pp. 41–62.

See, for example, Levine and Renelt, “A Sensitivity Analysis of Cross-Country Growth Regressions,” and Ann Harrison, “Openness and Growth: A Time-series, Cross-Country Analysis of Developing Countries,” Journal of Development Economics, Vol. 48 (March 1996), pp. 419–47. Using instrumental variables (to control for endogeneity), Jeffrey A. Frankel and David Romer, “Does Trade Cause Growth?” The American Economic Review, Vol. 89 (June 1999), pp. 379–99, find that trade has a quantitatively large and robust positive effect on income.

It is difficult to find an adequate measure for the openness to trade—which should ideally measure how open markets are to foreign competition. Proxies for openness that have been used include tariffs, nontariff barriers, effective rates of protection, trade liberalization, relative prices, import penetration, export intensity, and deviations of actual from predicted trade flows or volumes. See Harrison, “Openness and Growth”.

See Chapter V, “Trends and Issues in the Global Trading System,” in the October 1999 World Economic Outlook.

Jeffrey Sachs and Andrew Warner, “Natural Resource Abundance and Economic Growth,” NBER Working Paper 5398 (Cambridge, Massachusetts: National Bureau of Economic Research, 1995) find a negative relationship between growth and a high ratio of natural resource exports to GDP.

For a more extensive discussion see Chapter IV, “Financial Crises: Characteristics and Indicators of Vulnerability,” in the May 1998 World Economic Outlook, and Chapter III, “International Financial Contagion,” in the May 1999 World Economic Outlook.

Michael Roemer and Mary Kay Gugerty, “Does Economic Growth Reduce Poverty?” HIID Technical Paper (Cambridge, Massachusetts: Harvard Institute for International Development, March 1997).

Martin Ravallion and Shoahua Chen, “What Can New Survey Data Tell Us about Recent Changes in Income Distribution and Poverty,” The World Bank Economic Review, Vol. 11 (1997), pp. 357-82.

Some studies have found a statistically significant negative relationship between income inequality and growth while others have found a positive one. Overall, the relationship is not robust for a broad panel of countries. See, for example, Robert J. Barro, “Inequality, Growth, and Investment,” NBER Working Paper 7038 (Cambridge, Massachusetts: National Bureau of Economic Research, 1999). In part, the lack of a robust result may reflect limited data availability and crosscountry differences in the measurement of income inequality.

For example, see Michael Bruno, Martin Ravallion, and Lyn Squire, “Equity and Growth in Developing Countries: Old and New Perspectives on the Policy Issues,” in Income Distribution and High-Quality Growth, ed. by Vito Tanzi and Ke-young Chu (Cambridge, Massachusetts: MIT Press, 1998).

See Michael P. Dooley, “A Note on Debt Reduction and Economic Efficiency,” IMF Working Paper 90/36 (Washington: International Monetary Fund, 1990). See also Ibrahim A. Elbadawi, Benno J. Ndulu, and Njuguna Ndung’u, “Debt Overhang and Economic Growth in sub-Saharan Africa,” ed. by Zubair Iqbal and Ravi Kanbur, External Finance for Low- Income Countries, (Washington: IMF Institute, 1997).

For a more detailed discussion, see, for example, Chapter VI in the April 1986 World Economic Outlook, and Chapter IV in the April 1989 World Economic Outlook. For specific examples, see Ray Brooks and others, “External Debt Histories of Ten Low-Income Developing Countries: Lessons from their Experiences,” IMF Working Paper 98/72 (Washington: International Monetary Fund, 1998).

See Christina Daseking and Robert Powell, “From Toronto Terms to the HIPC Initiative: A Brief History of Debt Relief for the Low-Income Countries,” IMF Working Paper 99/142 (Washington: International Monetary Fund, 1999). See also the IMF’s HIPC website at

Daseking and Powell, “From Toronto Terms to the HIPC Initiative,” review debt assistance provided to low-income countries and provide estimates of the associated costs.

See David Andrews and others, “Debt Relief for Low- Income Countries: The Enhanced HIPC Initiative,” Pamphlet Series No. 51 (Washington: International Monetary Fund, 1999). An evaluation of the effects of the initial Initiative is provided by Stijn Claessen and others, “HIPC Debt: A Review of the Issues,” Journal of African Economies, Vol. 2 (1997). See also

Fluctuations in market interest rates complicate the measurement exercise and introduce an element of uncertainty in the estimated net present value of the debt. If the true level of sustainable debt were known with certainty, creditors could provide debt relief to this point without economic loss because debt above the sustainable level could not be repaid and therefore would have no value. Therefore, although debt relief can have a budgetary cost to the creditor of its full amount, it entails an economic cost only to the extent that it goes beyond the sustainable level.

The targets in the initial framework were net present value of the debt-to-export ratio of 200–250 percent, of the debt service-to-export ratio of 20–25 percent, and of the debt-to-fiscal revenues ratio of 280 percent. The specific sustainability targets for each country were to be based on an assessment of country-specific “vulnerability factors” (such as the concentration and variability of exports, the ratio of debt to GDP, the resource gap, the level of international reserves, and the burden of private sector debt). The empirical relevance of the various debt ratios in the assessment of debt sustainability is investigated by Daniel Cohen, “Growth and External Debt: A New Perspective on the African and Latin American Tragedies,” CEPR Discussion Paper No. 1753 (London: Center for Economic Policy Research, 1997).

See Andrews and others, “Debt Relief for Low-Income Countries: The Enhanced HIPC Initiative” and

William Easterly, “How Did Highly Indebted Poor Countries Become Highly Indebted? Reviewing Two Decades of Debt Relief (unpublished; Washington: World Bank, 1997) argues that when countries do not pursue macroeconomic and structural reform policies, they will tend to respond to debt relief by borrowing more or decreasing domestic assets, so as to leave practically unaltered the sustainability of their external liabilities. See also Burnside and Dollar, “Aid, Policies, and Growth.”

    Other Resources Citing This Publication