GLOBAL ECONOMIC PROSPECTS AND POLICIES

CHAPTER IV HOW CAN THE POOREST COUNTRIES CATCH UP?

P
rogress in raising real incomes and allevi-
ating 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 Figure 4.1. Advanced and Developing Economies:
Per Capita Income1
countries in the Middle East and the Western (Thousands of U.S. dollars at 1996 prices)
Hemisphere, real incomes have risen, but at a
East Asia, which includes China and has the largest population, and the newly
slower pace than in industrial countries (Figure industrialized economies are the only country groups that are rapidly converging with
4.1). Sustained and rapid improvements in rela- the industrial countries. The thickness of the bars reflects the population in each region.

tive income positions—convergence—have only 1970 1985 1998 [ ] Average growth,

occurred among the developing countries of east 1970–98 (percent)
Asia (including China), the newly industrialized
economies, and a few smaller countries in other Africa [–0.2]
regions. Average per capita incomes in regions
with the largest populations (as represented by
the thickness of the bars in Figure 4.1) remain
East Asia [5.6]
well below those in the industrial and newly in-
dustrialized 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 South Asia [2.5]

in reducing the share of the world population liv-

ing in poverty. This represents both huge Middle East and Europe [0.9]
amounts of unnecessary human suffering and an
Western Hemisphere [1.5]
enormous squandering of human potential.
Newly industrialized economies [5.6]
Recognizing these divergences in income lev-
els and rates of growth, and in particular the un- Industrial countries [2.2]
acceptably high levels of poverty that persist in
many countries, the Interim and Development
Committees of the IMF and World Bank placed Memorandum
renewed emphasis in September 1999 on Heavily indebted poor countries [0.0]
poverty reduction and on strengthening the
0 5 10 15 20 25
links between debt relief and poverty reduction.
Subsequently, the Executive Board of the IMF re- 1Converted into U.S. dollars using purchasing-power-parity (PPP) conversion rates.

focused 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

113
 CHAPTER IV HOW CAN THE POOREST COUNTRIES CATCH UP?

Box 4.1. Poverty Reduction Strategy Papers

In September 1999, the IMF and the World facility—formerly the Enhanced Structural
Bank endorsed a new approach to enhance the Adjustment Facility—with the Poverty Reduction
focus on poverty reduction in programs sup- and Growth Facility.
ported by concessional assistance, and to It is envisaged that countries will prepare PRSPs
strengthen the link between debt relief and on a three-year cycle, with progress reports in in-
poverty reduction. Underpinning the new tervening years. The PRSP will diagnose poverty
approach, each eligible country will prepare a in the country and describe the poor and their
Poverty Reduction Strategy Paper (PRSP). main characteristics. The PRSP will also quantify
A PRSP will outline a country’s anti-poverty the resources needed for various poverty reduc-
strategy over the medium and long term. To fos- tion programs and incorporate them in a sustain-
ter ownership, the PRSP will be drawn up by the able fiscal and macroeconomic framework, tak-
government after broad-based consultations ing account of the availability of non-inflationary
with stakeholders, including representatives of financing. Work toward the preparation of
civil society and development partners, and with PRSPs has begun in a number of countries (e.g.,
assistance from World Bank and IMF staff. Long- Bolivia, Mozambique, and Uganda).
term poverty reduction goals will be translated Since the design of robust poverty reduction
into annual targets for intermediate indicators strategies can take time, a phased introduction
(i.e., primary school enrollment, immunization of a full-fledged PRSP is foreseen. Some govern-
rates, etc.) to facilitate shorter-term program ments will thus prepare an interim PRSP, stating
monitoring. The PRSP is expected to become a their commitment to poverty reduction and lay-
key instrument for a country’s relations with the ing out the principal elements of their strategy,
donor community and civil society and will pro- the timetable to complete the PRSP, and a de-
vide the basis for debt relief under the HIPC scription of the consultative process through
Initiative and all World Bank/IMF concessional which the PRSP will be formulated. The three-
lending operations. Reflecting the new policies year macroeconomic program supported by IMF
and the central focus on poverty reduction, and World Bank lending could then be revised
the IMF has replaced its concessional lending to reflect the PRSP when one is completed.

structural soundness by also stressing the impor- fronts, as emphasized in the World Bank’s
tance of country ownership of the poverty reduc- Comprehensive Development Framework.1
tion strategy. The growing concern for country Previous issues of the World Economic Outlook
ownership, including through the involvement have discussed the experiences of the successfully
of civil society, is intended to reduce the risk of converging countries extensively. This chapter, in
slippages in implementation as the countries contrast, investigates the main impediments to
themselves take greater responsibility for the de- growth in the developing countries that have
sign and success of their economic plans. The failed to prosper and where poverty rates remain
enhanced amount of debt relief under the high. The large number of countries in this
Heavily Indebted Poor Countries (HIPC) group seems to suggest either that the conven-
Initiative is intended to release resources for tional growth strategy is not being implemented
poverty reduction, increase incentives for re- forcefully enough or that the strategy has been
forms, and remove a deterrent for both domes- overlooking critical obstacles to development.
tic and foreign investors. Of course, a successful However, the bulk of development research re-
development strategy requires progress on many veals neither a unique set of preconditions that

1See Entering the 21st Century: World Development Report, 1999–2000 (Washington: World Bank, 1999), p. 21.

114
 INCOME GROWTH AND POVERTY REDUCTION: THE RECENT HISTORY

are always present during economic takeoff nor it is important to remember that in countries
an easily identified set of impediments that have such as China and India, which have been grow-
prevented poor countries from achieving sus- ing quickly and seen poverty rates fall, the num-
tained growth. There is no single formula for ber of poor remains high. Some middle-income
kick-starting growth, and it is more likely that the countries, for example in Latin America, not
explanation for the unsatisfactory performance only have pockets of absolute poverty, but also
of many developing countries lies in the interplay significant relative poverty.3 In all of these cases,
of economic and political factors that vary by poverty alleviation remains important, and most
country. Nevertheless, experience in the success- of the policy considerations discussed in this
ful developing countries clearly points to macro- chapter also apply.
economic stability, sound institutional arrange-
ments, and openness to trade as factors that are
conducive to, or at least associated with, high sus- Income Growth and Poverty Reduction:
tainable growth.2 Experience in the poorest The Recent History
countries highlights poor education and health, Human living conditions have improved
ineffective governance, weak rule of law, and war greatly over the past 100 years, brought on by un-
as frequent impediments to prosperity. precedented technological and economic trans-
A theme of the chapter is the plight of poor, formations. Global output almost tripled in the
low-growth countries and within this group the first half of the twentieth century and increased
heavily indebted countries, about which there ninefold in the second half, greatly outpacing
is growing consensus that unsustainable debt has population growth. Life expectancy, education,
become a critical barrier to future growth and and other indicators of well-being also improved,
poverty reduction. External debt levels, despite particularly in poor countries.4 This chapter
years of rescheduling often at concessionary looks at trends over 1970–98, the longest period
terms, have become unsustainable in many for which more comprehensive data are available
cases. Without the efforts of the international for most developing countries. (Long-run per-
community to reduce this burden substantially, spectives are discussed further in Chapter V.)
there is little hope for significant improvement Rising global prosperity, however, has not ben-
in living conditions, as debt overhang saps eco- efited all countries and regions, and the global
nomic incentives to reform and grow. To be ef- distribution of income—measured by average in-
fective, however, debt relief must be accompa- comes across countries—remains very skewed.
nied by domestic policy reforms to address the This failure to converge can be seen in longer-
root causes of much of the initial debt buildup. term growth rates: 75 percent of developing
The chapter deals less with poverty problems countries recorded slower per capita income
in middle-income countries and low-income growth than in the industrial countries over the
countries that are growing rapidly. Nevertheless, past three decades (Figure 4.2, upper panel).5

2See, 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).
3Absolute poverty refers to the number of individuals living in poverty conditions, often defined in terms of internation-

ally comparable monetary measures, while relative poverty refers to income differences within a country.
4D. Gale Johnson, “Population, Food, and Knowledge,” American Economic Review, Vol. 90 (March 2000), pp. 1–14.
5Developing 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.

115
 CHAPTER IV HOW CAN THE POOREST COUNTRIES CATCH UP?

Per capita income fell in 32 countries in the

sample, while only seven developing countries
grew fast enough to reduce substantially the in-
come gap with—and rapidly converge toward—
the industrial countries as a group.6
The picture is more encouraging when
progress is assessed based on both per capita in-
Figure 4.2. Developing Countries: Grouped by Per Capita
Income Growth, 1970–981 come growth and the size of a country’s popula-
tion (Figure 4.2, lower panel). Less than 10 per-
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
cent of the developing world’s population live in
all developing countries), although this group of developing countries comprises about countries where average income declined, while
30 percent of the total in terms of population.
70 percent live in countries where per capita in-
Number of Countries come growth exceeded that in the industrial
countries. This more positive outcome mainly re-
flects strong economic growth in China in par-
Negative growth
ticular and also in India, which together account
for about 50 percent of the population in devel-
oping countries and which had per capita
Slow growth
growth rates of 7 percent and 2#/4 percent, re-
spectively, over the 1970–98 period (Box 4.2).7
A more positive picture also emerges when ex-
Fast convergence amining the growth performance over a shorter
time frame. During 1993–98, 14 developing
Slow convergence countries converged rapidly toward the indus-
trial countries—double the number between
Share of Population
1970 and 1998—despite the problems encoun-
tered by many countries because of financial
Slow growth crises in Asia, Russia, and Latin America.
Conversely, per capita income fell in 23 coun-
Slow convergence
tries (compared to 32 during the longer
period).8 The improved performance in these
Negative growth countries partly reflects stronger domestic poli-

6Fast-growing or rapidly converging countries are de-

fined for this analysis as those countries with annual per

capita income growth of more than 3#/4 percent. This cut-
off was chosen for tractability and is similar to the cutoff
chosen in Lant Prichett, “Divergence, Big Time,” Journal
Fast convergence
of Economic Perspectives, Vol. 11 (Summer 1997), pp. 3–17.
Prichett finds 11 countries (both developing and ad-
1Slow-growth countries have annual per capita income growth rates between 0 and 2 vanced) in which growth exceeded 4.2 percent over the
percent; slow-convergence countries, between 2 and 3#/4 percent; and fast-convergence 1960–90 period. The relatively larger number of fast-grow-
countries, over 3#/4 percent. ing countries in that study mainly reflects the inclusion of
advanced economies.
7It is important to note, however, that regional growth

rates have varied considerably within China and India.

8Interestingly, 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.

116
 INCOME GROWTH AND POVERTY REDUCTION: THE RECENT HISTORY

Box 4.2. India: Reinvigorating the Reform Process

India has been among the fastest-growing external debt (especially short-term), a surge in
economies in the world over the last two world oil prices, and a sharp decline in remit-
decades, and has achieved trend improvements tances from Indian workers in the Middle East.
in growth, literacy, mortality, and poverty rates As capital flight accelerated and official reserves
(see the figure, upper panels). In recent years, were rapidly depleted, the Indian government
deft handling of monetary policy has helped entered into a Stand-By Arrangement with the
India to successfully weather the Asian crisis, IMF and embarked on a program of fiscal and
while maintaining low inflation and a comfort- structural reforms.
able external position. Yet despite these gains, Corrective policy measures were successful in
poverty rates remain high, with more than a restoring macroeconomic stability. The central
third of the population still living below the offi- government deficit was brought down from 8
cial poverty line.1 This uneven progress raises percent of GDP before the crisis to 4#/4 percent
questions about the impact of the economic and in 1996/97,2 through tax reforms, cuts in subsi-
structural reforms implemented since the mid- dies, and reductions in defense and other ex-
1980s on growth in India, and what more can be penditures (see the figure, lower panels). The
done to make greater inroads into poverty lower deficit, in turn, reduced financing that
reduction. had to be provided by the central bank, and
In the three decades following independence wholesale price inflation declined from a precri-
in 1947, growth in India was stifled by a high de- sis level of almost 14 percent to nearly 6 percent
gree of government planning and regulation, by 1996/97.
with per capita GDP rising by only 1!/2 percent In addition, important structural reforms
per annum (see the first table). Industrial con- were introduced. Industrial licensing and invest-
trols were pervasive, and restrictions on private ment approval procedures were liberalized, and
credit, the role of the public enterprise sector, the number of industries reserved for the public
and subsidy programs increased throughout the sector was reduced. External sector reforms in-
period. Strict controls on foreign direct invest- cluded a reduction in the import-weighted tariff
ment, an import licensing system, and—from rate from 87 percent in 1990/91 to 25 percent
the 1970s—high tariff rates further limited the by 1996/97, easing of import licensing require-
economy’s growth potential. ments, relaxation of controls on foreign direct
The liberalization of import and industrial and portfolio investment, and greater exchange
controls in 1985 and improved agricultural per- rate flexibility. Financial sector measures in-
formance spurred an acceleration of real per cluded interest rate liberalization, strengthened
capita GDP growth to an average rate of 3#/4 per- prudential norms and supervision, the introduc-
cent in the 1980s. However, this expansion also tion of greater competition into the banking sys-
reflected other developments—increased fiscal tem, and improvements to the operation of capi-
stimulus and a debt-financed consumption and tal markets.
investment boom—which became unsustainable In response to the government’s policy pack-
toward the end of the decade. age, the recovery from the 1991 crisis was rapid.
A balance of payments crisis ensued in 1991, Private investment rates rose sharply, and real
reflecting the deteriorating fiscal position, rising per capita GDP growth increased to more than
6 percent by 1995/96. Significant improve-
ments in productivity were also achieved—
1The World Bank’s World Development Report
as evidenced by increased total factor produc-
1999/2000 suggests an even more severe poverty prob-
tivity growth at both the aggregate and firm lev-
lem, 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 liv-
ing on less than $2 per day. 2The fiscal year runs from April through March.

117
 CHAPTER IV HOW CAN THE POOREST COUNTRIES CATCH UP?

Box 4.2 (continued)

India: Economic Indicators1

10 80
Real Per Capita GDP Growth Poverty Rate
(percent) (percent of population below
Trend the official poverty line)

5 60

0 40

–5 20

–10 0
1951 61 71 81 91 99 1973 78 83 88 93

12 100
Public Sector Deficit Public Sector Debt
(percent of GDP) (percent of GDP)
Consolidated2

Consolidated2
9 75

6 Central government 50

Central government

3 25

0 0
1950 60 70 80 90 99 1980 85 90 95 99

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.

els3 and by declining incremental capital-output However, per capita growth has slowed
ratios, particularly in the services sector (see more recently, averaging closer to 4 percent
the first table). between 1997/98 and 1999/00 compared
with 4#/4 percent between 1992/93 and
3See World Bank, India: Policies to Reduce Poverty and

Accelerate Sustainable Development, Report No. 19471-IN

(2000); and P. Krishna and D. Mitra, “Trade Growth: New Evidence from India,” Journal of
Liberalization, Market Discipline and Productivity Development Economics, Vol. 56 (1998), pp. 447–62.

118
 INCOME GROWTH AND POVERTY REDUCTION: THE RECENT HISTORY

1996/97.4 To some extent, this reflected the Expenditure and Sectoral Components of Growth1
completion of cyclical catch-up following the (Average annual percent, unless otherwise noted)
1991 balance of payments crisis, as well as the 1951–79 1980–90 1992–96 1997–992
adverse impact of the 1997 regional crisis and
Real per capita GDP
agricultural supply shocks. In addition, though, growth3 1.5 3.8 4.7 4.1
economic performance appears to have been Real GDP growth3 3.7 5.9 6.7 5.8
adversely affected by a reversal of fiscal adjust-
Contribution to growth,
ment, infrastructure bottlenecks, and delays in by expenditure item:
implementing structural reforms. Increases in Private consumption 2.4 3.8 3.9 2.5
civil service wages and subsidies, as well as ris- Public consumption 0.4 0.8 0.5 1.4
Gross fixed investment 0.8 1.5 1.9 1.2
ing debt service, pushed up the fiscal deficit Private investment ... 0.8 1.8 0.9
and resulted in higher real interest rates. These Public investment ... 0.6 0.1 0.3
higher rates, combined with banks’ efforts to Net exports4 ... 0.1 0.1 0.6
improve their balance sheets, slowed credit Contribution to growth,
growth. Infrastructure constraints also contin- by sector:
Public 1.1 1.7 2.8 5.2
ued to bind, as the earlier fiscal consolidation Private 2.2 4.2 3.8 0.7
had relied too heavily on reductions in public Contribution to growth,
investment. Consequently, the contribution of by sector:
private investment to growth fell by half from Agriculture 1.1 1.6 1.4 0.5
Industry 1.0 1.7 2.0 1.5
earlier in the decade, and measured productiv- Services 1.4 2.5 3.2 3.8
ity growth, particularly in the industrial sector,
ICORs, by sector:5
deteriorated (see the first table). Overall ... 4.2 4.1 4.8
Moreover, the poverty rate remains very high, Agriculture ... 2.0 1.5 2.6
and the impressive rate of decline from the mid- Industry ... 5.7 6.8 10.7
Services ... 4.0 2.9 2.1
1970s through the 1980s may have slowed.5 This
outcome partly reflects the relatively poor per- Sources: Central Statistical Organisation (CSO), National Accounts
Statistics.
formance of the agricultural sector during the 1Averages computed over fiscal years beginning in April.

1990s, since some 70 percent of the labor force 21999 figures on GDP and sectoral production are CSO Advance
Estimates; annual population growth assumed constant at 1.7 percent;
still relies on the land for its livelihood.6 While average contribution of expenditure categories and private and public
adverse supply shocks played a role, the lack of production computed over 1997–1998.
3Measured at market prices; base year is 1980 for data until 1993,
agricultural reform also contributed to low in-
and 1993 thereafter.
vestment rates and productivity in this sector. In 4Includes statistical discrepancy.
5The incremental capital output ratio (ICOR) is the ratio of the invest-
addition, the scope for mobility of low-skilled la-
ment rate to the GDP growth rate; a falling ICOR over time therefore in-
bor out of the agricultural sector has likely been dicates improved capital productivity.
limited by the absence of robust and sustained

growth in the industrial sector and by the rela-

4The projection for growth in 1999/00 is based on tively larger contribution of the higher-skilled
Central Statistical Organization Advance Estimates. service sector to GDP growth.
5Unfortunately, the most recent official poverty statis-
What measures need to be implemented to
tics are for 1993/94, and do not fully capture the impact
of the rapid growth in the post-crisis period. However, sustain high growth rates in all sectors and
unofficial figures suggest that the poverty rate may not achieve more substantial progress toward poverty
have declined appreciably during the last decade. alleviation?7 As discussed in the text—and quan-
6See S. Tendulkar, “Indian Economic Policy

Reforms and Poverty: An Assessment,” in India’s

Economic Reforms and Development: Essays for Manmohan 7Most analysts believe that a significant reduction in

Singh, ed. by I.J. Ahluwalia and I.M.D. Little (Delhi: the poverty rate would require an annual real GDP
Oxford University Press, 1998). growth rate of at least 7 percent (5 percent in per

119
 CHAPTER IV HOW CAN THE POOREST COUNTRIES CATCH UP?

Box 4.2 (concluded)

Explaining India’s Relative Growth Performance

Estimated Difference in
Average Value (percent)
_______________________ Contribution to Growth Rates1
India East Asia (percentage points)
Factors contributing to growth during 1970–95
Investment/GDP 21.9 29.6 –1.2
Net FDI/GDP 0.1 2.5 –0.8
Trade/GDP 4.5 113.5 –1.1
Government consumption/GDP 10.3 10.4 0.0
Secondary school enrollment rate 35.2 50.5 –0.3
CPI inflation rate 8.8 8.4 0.0
Convergence effect and other factors ... ... 0.2
Real per capita GDP growth (1970–95 average) 2.4 5.7 –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).
1Calculated as the estimated coefficient times the difference in the independent variable value (India less East Asia). Reported differ-
ences in growth rates of real per capita GDP are actuals.

tified in the Indian context in the second table vorable current economic situation and the sig-
—faster growth would require durable fiscal con- nificant majority enjoyed by the ruling coalition.
solidation to raise national saving and crowd-in Second, with the consolidated public sector
private investment spending; further liberaliza- deficit rising again and the public sector debt
tion of foreign trade and investment flows; and stock close to 80 percent of GDP, fiscal sustain-
additional reforms to labor markets and in the ability is a serious concern.8 Third, India is com-
agricultural, industrial, and financial sectors to mitted to trade liberalization measures under
promote greater efficiency and export competi- the World Trade Organization, including the re-
tiveness. These reforms need to include removal moval of all quantitative restrictions by 2001. For
of domestic pricing distortions, improvements to India to achieve the maximum benefits from a
bankruptcy procedures, and an easing of restric- more liberal trade system, the structural impedi-
tions on firm and farm size and regulations that ments affecting domestic producers must be ad-
make it difficult to shed labor (and therefore im- dressed in the interim.
pede job creation). Fiscal priorities also need to Encouragingly, the new government has taken
be redirected toward investment in human and a number of initiatives that suggest a strength-
physical capital. ened commitment to structural reform, includ-
There is broad agreement in India that fur- ing liberalization of the insurance sector, auto-
ther reforms are needed—the experience of the matic clearance for foreign direct investment in
early 1990s has demonstrated the potential ben- many sectors, and a landmark agreement on
efits of reform, and consistent views on many of state sales tax rationalization. At the same time,
these key issues emerged from the major parties however, the budget introduced in February
during the October 1999 election. Several fac- 2000 targets only modest deficit reduction in
tors argue for translating this consensus into the coming fiscal year, and a clearly defined
swift action. First, the establishment of a bold agenda for reform has yet to be established.
agenda would be facilitated by the relatively fa- Hence, critical and difficult challenges remain
to be addressed.
capita terms) on a sustained basis. The government’s
Ninth Five-Year Plan (1997–2002), which targeted an 8See P. Reynolds, “Fiscal Adjustment and Growth

average real GDP growth rate of 7 percent, projected Prospects in India,” in India: Selected Issues, IMF Staff
that the official poverty rate would be reduced by 11 Country Reports (Washington: International
percentage points by the end of the plan period. Monetary Fund, forthcoming).

120
 INCOME GROWTH AND POVERTY REDUCTION: THE RECENT HISTORY

cies and for some of these countries, a more sup-

portive 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 coun- Figure 4.3. Developing Countries: Population Living
tries in the rapidly converging category and the Below $1 per Day
smaller number of countries in which per capita (Percent of total population)

income fell (which in part reflects a pickup in 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,
growth performance in Africa over this period) to a lesser extent, in south Asia. The thickness of the bars reflects the total population
is encouraging. But progress in alleviating in each region (except for the total).

poverty remains fragile and inadequate. 1987 1993 19981

The percentage of the population in develop-
ing countries living under the $1 per day line has
Africa
declined from 30 percent to 24 percent during
the past decade (largely reflecting poverty reduc-
tion in east Asia and to a lesser extent in south
Asia), but the incidence of poverty remains unac-
ceptably high (Figure 4.3).9 The number of peo- East Asia 2

ple 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 de-
veloping countries). Moreover, progress in
South Asia 3
poverty reduction has been uneven across geo-
graphic regions. There has been no progress in
reducing poverty rates in Africa and in the
Middle East and Europe
Western Hemisphere, and poverty rates actually
increased in the transition countries. Western Hemisphere
It should be noted that poverty is a multidi-
mensional phenomenon, reflecting not only ma- Countries in transition

terial deprivation but also, for example, lack of

security and access to basic services including Memorandum
Total
health, education, and sanitation. Health meas-
0 10 20 30 40 50
ures (for example, life expectancies) have con-
Source: Martin Ravallion and Shaohua Chen, “Global Poverty Measures 1987–98 and
verged more rapidly across countries than aver- Projections for the Future,” forthcoming.
age incomes, in part because of the diffusion of 1Estimated.
2East Asia comprises the developing countries in Asia except those in south Asia, as
medical technology. This aspect of globalization
defined in footnote 3.
has unambiguously helped the poorest coun- 3South Asia comprises Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan, and

tries. Health and other dimensions of poverty Sri Lanka.

are not captured in the monetary measures of

poverty discussed in this chapter. See Box 5.1 for
an alternative measure of economic progress

9The pattern of poverty rates is similar when the

poverty line is defined as $2 per day.

121
 CHAPTER IV HOW CAN THE POOREST COUNTRIES CATCH UP?

that combines income, health, and education The primary thrust of these studies is to identify
into a single summary index. the conditions—some of which are policy re-
Finally, it is important to recognize the bene- lated—that catalyze growth by promoting invest-
fits of policy reforms and other accomplishments ment in physical and human capital, improving
in the countries that have grown rapidly both the efficiency of production (including the allo-
over the full 30-year period under review and cation of resources), and encouraging techno-
more recently. However, it is equally important logical progress.
to recognize the risks of slippage in the reform Unfortunately, the empirical studies generally
process or other setbacks in these very successful provide only limited guidance and do not iden-
countries, especially the larger ones. Further, the tify a single set of variables that strongly promote
1997–98 financial crises demonstrated that se- or inhibit growth.11 For many of the concepts
vere recessions in a small number of countries tested, the correlation with growth is often not
can reverse some of the past progress in poverty robust to small changes in the variables being
reduction, while producing significant, adverse studied (that is, the conditioning set of other fac-
spillover effects on poverty levels in trading part- tors that may help explain differences in
ners. Moreover, some have argued that over the growth).12 In general, factors that are consid-
short run economic cycles have an asymmetric ered robustly and positively correlated to growth
effect on poverty—that is, poverty increases include the share of investment in GDP, school
more in an economic contraction than it de- enrollment, health indicators, openness to trade
creases in an equivalent expansion.10 (which can be considered a proxy for outward
orientation), and the share of nonprimary goods
in total exports. Factors that are robustly related
Explanations for Diverging Performance: to poor or negative growth include weak institu-
Analytical Complications tional structures, measures of political instability
In the main, the empirical literature on (such as the absence of civil liberties), weak rule
growth and development attempts to explain the of law, wars, and market distortions.
observed differences in growth and prosperity The relevance and proper interpretation of
across countries by “conditional convergence.” these cross-sectional results have been called
That is, each country’s economic growth rate de- into question by a number of economists:
pends on a number of underlying conditions, in- • First, the observation that a variable is cor-
cluding national endowments, preferences, related with growth does not mean that it
macroeconomic and domestic stability, institu- helps cause growth, and therefore, an ob-
tional and social structures, outward orientation, served relationship might provide only lim-
and the state of financial market development. ited insight for policymakers.13 For exam-

10See “External Shocks, Financial Crises, and Poverty in Developing Countries,” Global Economic Prospects 2000

(Washington: The World Bank, 2000).

11Robert J. Barro and Xavier Sala-i-Martin, Economic Growth (New York: McGraw-Hill, 1995), provide a review of the theo-

retical and empirical literature.

12Ross 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 invest-
ment 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 Regres-
sions,” NBER Working Paper 6252 (Cambridge, Massachusetts: National Bureau of Economic Research, 1997), however, ar-
gue that extreme-bounds analysis is too strong a test and find about two dozen robust variables (including regional and reli-
gion variables) by examining the cumulative distributions of coefficient estimates from a large number of regressions.
13Instrumental variables are used in some of these cross-sectional studies to control for endogeneity. However, it is possi-

ble 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.

122
 OBSTACLES TO GROWTH

ple, while a higher investment share is cor- interactions between growth and other vari-
related with (and is often thought necessary ables are significant—that good policies
for) faster growth, this correlation may re- tend to be mutually reinforcing and that
sult because faster growth induces more in- policy complementarities are important.15
vestment (reverse causation) or because • Fourth, cross-sectional studies generally as-
some other (omitted) factor induces both sume that the growth process is the same in
faster growth and more investment. rich and poor countries. However, it has
Determining the causal relationship matters been argued that growth involves at least
for policy advice. Simply increasing invest- two dimensions: technological progress and
ment (without concern for the efficiency of catch-up.16 Factors that may best promote
that investment) may not be sufficient to technological progress (most relevant for
promote growth. advanced economies) may not necessarily
• Second, these results do not fully explain be the most optimal for low-income devel-
the growth performance of many individual oping countries where takeoff and catch-up
economies (including many of the poor are most important.
countries with negative per capita real
growth over long time periods) or regions
of the world, and indeed may explain well Obstacles to Growth
only the growth performance of industrial Overall, factors related to growth can be
countries and a small set of developing grouped in six areas that, broadly speaking, in-
countries. For example, empirical studies fluence capital and the labor force, and the effi-
have found that regional dummy variables cient use of their services in production. Each of
are needed to explain the growth perform- these six areas is reviewed below, tying together
ance of countries in sub-Saharan Africa, theory and evidence from the literature and data
Latin America, and east Asia.14 analysis tailored to the question of uncovering
• Third, most cross-sectional analyses assume impediments to growth in the poorest countries.
linear relationships between growth and the In order to avoid some of the pitfalls of cross-
explanatory factors and miss important in- sectional studies, particularly the assumption of
teractions among factors. There are com- the homogeneity of countries at different stages
pelling reasons to believe that some rela- of the growth process, the analysis in the rest of
tionships are nonlinear and that this section examines countries grouped by in-

14See 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 coun-
tries, 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.
15Chapter 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 coun-
tries with good fiscal, monetary, and trade policies.
16See 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.

123
 CHAPTER IV HOW CAN THE POOREST COUNTRIES CATCH UP?

Table 4.1. Low- and Middle-Income Countries: Average Per Capita Income Growth, 1970–98
Low-Income Countries Middle-Income Countries
Negative Per Capita Growth (Growth rates below 0 percent)
Angola –1.9 Madagascar –2.0 Algeria –0.2 Namibia –0.6
Burkina Faso –0.5 Mali –0.5 Djibouti –4.3 South Africa –0.1
Burundi –0.3 Nicaragua –2.4 Lebanon –2.8 Trinidad and Tobago –0.3
Central African Rep. –0.3 Niger –0.9 Libya –1.3 Venezuela –0.3
Comoros –0.2 Rwanda –1.3
Congo, Dem. Rep. of –4.3 São Tomé and Príncipe –0.4
Côte d’Ivoire 0.0 Senegal –0.4
Ethiopia –0.1 Sierra Leone –2.5
Ghana –0.6 Solomon Islands –0.4
Guinea-Bissau –0.1 Somalia –0.9
Haiti –0.7 Togo –1.1
Liberia –2.0 Zambia –2.2

Slow Per Capita Growth (Growth rates of 0–2 percent)

Bangladesh 1.1 Malawi 0.7 Argentina 0.7 Jamaica 0.1
Benin 0.1 Mauritania 1.3 Bahrain 0.5 Jordan 0.4
Cameroon 0.2 Mozambique 0.9 Bolivia 0.5 Mexico 1.6
Chad 0.1 Myanmar 1.9 Cape Verde 0.5 Morocco 1.9
Congo, Rep. of 1.2 Nepal 0.9 Colombia 2.0 Panama 1.5
Gambia, The 1.5 Nigeria 0.1 Costa Rica 1.5 Papua New Guinea 0.5
Guinea 1.2 Sudan 1.1 Ecuador 2.0 Peru 0.1
Honduras 0.5 Tanzania 0.1 El Salvador 0.7 Philippines 1.0
Kenya 1.0 Uganda 0.1 Fiji 1.5 Suriname 1.5
Lesotho 0.8 Zimbabwe 0.3 Guatemala 0.9 Swaziland 1.4
Guyana 0.6 Uruguay 1.9
Iran, I.R. of 0.3

Slow Per Capita Convergence (Growth rates of 2–3#/4 percent)

Bhutan 3.3 Pakistan 2.2 Brazil 2.3 Sri Lanka 3.2
Cambodia 2.7 Vietnam 3.4 Chile 2.5 Syrian Arab Rep. 2.5
India 2.7 Yemen, Republic of 2.1 Dominican Republic 2.5 Tunisia 3.0
Lao P.D.R. 2.8 Egypt 2.6 Turkey 2.7
Paraguay 2.1

Fast Per Capita Convergence (Growth rates above 3#/4 percent)

China 6.9 Indonesia 3.9 Botswana 7.3 Mauritius 4.6
Korea 6.0 Thailand 4.5
Malaysia 4.3

come level and growth performance. Countries (Table 4.1).17 These subgroups comprise coun-
are divided by income level according to the cur- tries where per capita income is declining or re-
rent World Bank classification of low-, middle-, gressing in absolute terms, those with slow (less
and high-income countries. Developing coun- than 2 percent) per capita growth that are re-
tries in the low- and middle-income categories gressing relative to the industrial countries,
are then separated into eight groups (four for those that are slowly converging with the indus-
each income level) depending on their per trial countries (up to 3#/4 percent growth), and
capita growth rates over the period 1970–98 those that are rapidly converging.

17As 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.

124
 OBSTACLES TO GROWTH

The classification is intended for analytic pur- percent attributed to improved (total factor)
poses only and will be used to identify average productivity.19
characteristics of low-income countries and con- Not surprisingly, the low- and middle-income
trast these with higher-income and faster-grow- countries with declining or slowly rising per
ing economies. Clearly, the makeup of the capita income had on average lower investment
growth subgroups could change somewhat, for and saving rates than their faster-growing coun-
example, if average growth rates were calculated terparts in recent years, confirming the impor-
over a different time period. The relatively long tance of capital accumulation in the growth
1970–98 period was chosen to reduce the poten- process (Figure 4.4). Causality is difficult to in-
tial impact of business cycle effects, with the risk fer, however, because investment and saving
that some countries that have experienced rates were not substantially different, on average,
higher growth rates more recently would be in- across groups during the early 1970s except for
correctly placed in one of the lower-growth perhaps the fastest-growing economies. Even in
groups.18 It also needs to be recognized that in this latter group of countries, investment rates
sorting countries by current income, the low-in- rose only after the growth takeoff. 20 In other
come group will naturally tend to have more words, it is far from obvious that high initial in-
countries with disappointing growth over the vestment and saving rates are preconditions for
past than will the middle-income group. In other growth. It may indeed be that higher investment
words, the two dimensions of the classification and saving rates result because of higher growth
are not independent. or that other factors cause both growth and
investment.
Low levels of schooling or investment in hu-
The Role of Investment, Saving, Human Capital, man capital may be impediments to growth and
and Productivity also delay takeoff. Secondary school enrollment
It is well established that the accumulation of rates in the 1970s were substantially lower on av-
physical and human capital and advances in erage in nonrapidly converging, low-income
production efficiencies and technology lead to countries than in the middle-income countries
higher per capita income. Studies have typically (Table 4.2).21 Moreover, the fastest-growing, low-
found that approximately 60–70 percent of per and middle-income countries also experienced
capita growth in developing countries reflects larger improvements in enrollments rates than
increases in physical capital and another the other developing countries did between
10–20 percent is due to increases in education 1975 and 1995. Although it is possible that
and human capital with the remaining 10–30 growth induces more education as demand in-

18For 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.
19See, 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.
20Christopher 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 rela-
tionship 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 sav-
ing—in other words, investment and saving rates are endogenous.
21Literacy rates and primary school enrollment rates show a similar pattern.

125
 CHAPTER IV HOW CAN THE POOREST COUNTRIES CATCH UP?

creases 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.
Basic education, including training, can con-
tribute directly to a country’s potential for
growth by raising the skill level of the workforce.
Figure 4.4. Developing Countries: Investment and Saving1 In addition, because physical and human capital
(Percent of GDP)
are often complementary, education can also
Investment and saving rates were higher on average in countries with faster per capita raise growth indirectly by inducing greater in-
income growth, although causality is hard to infer.
vestment.22 While increased schooling and train-
Negative Slow Slow Fast
growth growth convergence convergence
ing alone may not be sufficient to boost growth,
particularly when economic opportunities to use
Low-Income Countries Middle-Income Countries
the acquired skills are missing, improving educa-
Investment
40 40 tion will be an important part of a sustainable
growth and poverty reduction strategy for devel-
35 35
oping countries. It therefore makes sense for
30 30 countries to shift resources toward basic educa-
tion and for the donor community to emphasize
25 25
education as a high priority.
20 20 Another obstacle to a productive workforce
(and society) is inadequate health care. As with
15 15
school enrollment rates, life expectancy rates at
10 10 birth were substantially lower on average in
1970 75 80 85 90 95 98 1970 75 80 85 90 95 98
nonrapidly-converging, low-income countries
Saving
than in the middle-income countries in the
40 40 1970s (see Table 4.2), and other health indica-
tors show a similar pattern. Even though these
30 30 health indicators have improved over time in
most developing countries, they remain rela-
tively bad in many low-income countries—for
20 20
example, average life expectancy is still below
55 years for the negative- and slow-growth, low-
10 10 income countries—representing an enormous
loss in potential human capital. In addition,
0 0 progress in improving life expectancy rates has
1970 75 80 85 90 95 98 1970 75 80 85 90 95 98
slowed in some countries mainly because of the
1Unweighted
devastating effects of the AIDS epidemic. The
cross-country averages. For definitions of country groups, see Table 4.1.
repercussions from the epidemic (as well as
from other infectious diseases) will have long-
term effects on the economic performance of

22See 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).

126
 OBSTACLES TO GROWTH

Table 4.2. Developing and Advanced Economies: Education and Health Indicators1
Secondary School Enrollment Rate2
_________________________________ Life Expectancy at Birth3
_______________________________
Country Groups 1975 1995 Increase 1970 1997 Increase
Low-income countries
Negative growth 11 19 8 43 50 7
Slow growth 11 21 10 44 53 8
Slow convergence 17 41 24 46 59 13
Fast convergence 34 67 34 55 67 13
Middle-income countries
Negative growth 37 60 23 55 66 11
Slow growth 40 60 20 59 69 10
Slow convergence 35 61 26 59 70 11
Fast convergence 36 68 32 59 66 8
Memorandum
Industrial countries 80 115 35 72 77 6
Other high-income, non-oil countries4 56 86 30 70 77 7
Middle- and high-income oil countries5 29 67 38 53 67 14
Source: World Bank, World Development Indicators.
1Unweighted cross-country averages. For definitions of country groups, see Table 4.1.
2Gross rates, in percent.
3Years.
4Cyprus, Hong Kong SAR, Israel, Singapore, and Taiwan Province of China
5Equatorial Guinea, Gabon, Kuwait, Oman, Qatar, Saudi Arabia, and United Arab Emirates.

some poor countries, particularly in sub- these countries to better use their limited re-
Saharan Africa.23 sources. To the extent that this and other re-
Inefficient investment has also been a hin- source reallocations can be accomplished rela-
drance for many countries, although, again, tively quickly, countries could begin to grow
causality is difficult to infer. Not surprisingly, in without immediate increases in saving and
the developing countries with declining per investment.
capita growth during the last three decades, the The challenge then is to fashion the appropri-
incremental output-capital ratio (the inverse of ate mix of policies and conditions for poor
the incremental capital-output ratio), which is a countries that removes impediments to the accu-
very rough proxy for the productivity of invest- mulation and efficient allocation of saving and
ment, was lower on average than in the coun- investment (including in human capital) and al-
tries that were growing (Figure 4.5). Estimates of lows growth to take off. Once growth has accel-
total factor productivity growth, which are avail- erated, there are many indications that a virtu-
able for only a subset of the countries under re- ous and mutually reinforcing cycle is possible as
view, also confirm that resources were not used growth may further increase saving, investment,
as efficiently in many of the negative-growth and productivity. The next sections will review
countries as in other developing countries.24 some of the main obstacles to increasing saving,
Although a difficult task that needs to be ad- investment, and efficiency. These are generally
dressed through a variety of reforms depending considered to include macroeconomic instabil-
on country-specific circumstances, increasing ity, unsupportive institutions, inward-oriented
productivity and allocative efficiency will allow and protectionist policies, poverty, income in-

23It is estimated that 33!/2 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).
24See Bosworth and Collins, “Economic Growth in East Asia.”

127
 CHAPTER IV HOW CAN THE POOREST COUNTRIES CATCH UP?

equality, and (particularly in recent years) unsus-

tainable external debt.

Macroeconomic Instability
Uncertainty caused by macroeconomic insta-
bility—such as high inflation, volatile and over-
Figure 4.5. Developing Countries: Per Capita GDP Growth and valued exchange rates, or excessive fiscal
Incremental Output-Capital Ratio1 deficits—can significantly distort economic deci-
Productivity, as proxied by the incremental output-capital ratio, was substantially lower sions and thereby reduce capital accumulation,
on average in developing countries with negative growth than in other countries during
the past three decades. hamper the efficient allocation of resources, and
slow growth. Empirical studies have found, for
Negative Slow Slow Fast example, that high inflation rates, which are of-
growth growth convergence convergence
ten also accompanied by more variable inflation
Low-Income Countries Middle-Income Countries and relative price changes, have a negative im-
Per Capita GDP Growth
(percent change) pact on growth.25 Because the effect appears to
15 15
be nonlinear, though, there is some disagree-
ment over the precise threshold above which
10 10
higher inflation becomes a detriment.
Nevertheless, recent evidence indicates that in-
5 5
flation rates above the single digits have adverse
0 0
implications for growth.
Inappropriate exchange rate regimes, gener-
–5 –5 ally in conjunction with high and variable infla-
tion rates, often lead to overvalued exchange
–10 –10 rates. These can impede the proper allocation of
1970 75 80 85 90 95 98 1970 75 80 85 90 95 98
resources between the production of tradable
Incremental Output-Capital Ratio 2 and nontradable goods and may also deter in-
0.6 0.6 ward foreign capital while encouraging capital
flight. Persistently large fiscal deficits also inhibit
0.4 0.4 growth through several mechanisms. As dis-
cussed in previous issues of the World Economic
0.2 0.2
Outlook, fiscal deficits tend to crowd out private
investment (or lead to higher inflation if the
deficits are monetized) and inhibit financial
0.0 0.0
market development.26 In addition, they may be
indicative of excessive government intervention
–0.2 –0.2 in the economy, although this is difficult to
1970 75 80 85 90 95 98 1970 75 80 85 90 95 98

1Unweighted cross-country averages. For definitions of country groups, see Table 4.1.
2Inverse of five-year moving average of incremental capital-output ratio.
25See, 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.
26See, for example, Chapter IV, “Fiscal Policy Issues in

Developing Countries,” in the May 1996 World Economic

Outlook.

128
 OBSTACLES TO GROWTH

Table 4.3. Developing and Advanced Economies: Macroeconomic Stability1

(Percent during 1970–98, unless otherwise noted)
Inflation Inflation Current Black Market
Country Groups Average Standard Deviation Fiscal Deficit2 Account Deficit2 Premium3
Low-income countries
Negative growth 26.4 26.3 6.2 6.5 386.4
Slow growth 16.1 12.4 4.2 8.6 214.9
Slow convergence 20.5 19.0 6.9 3.1 112.2
Fast convergence 9.4 7.7 0.7 0.7 23.6
Middle-income countries
Negative growth 15.1 10.6 4.6 3.1 49.1
Slow growth 25.4 21.6 4.9 4.5 59.3
Slow convergence 32.3 23.6 4.7 3.5 45.0
Fast convergence 8.4 5.3 2.3 1.7 7.2
Memorandum
Industrial countries 6.5 3.9 3.2 1.0 0.8
Other high-income, non-oil countries4 13.0 11.7 1.8 0.4 4.1
Middle- and high-income oil countries5 7.1 7.9 1.7 –2.3 0.5
1Unweighted cross-country averages. For definitions of country groups, see Table 4.1.
2Percent of GDP.
3From World Bank, World Development Indicators. Percent difference between the official and market exchange rate.
4Cyprus, Hong Kong SAR, Israel, Singapore, and Taiwan Province of China.
5Equatorial Guinea, Gabon, Kuwait, Oman, Qatar, Saudi Arabia, and United Arab Emirates.

quantify when analyzing large groups of coun- aligned exchange rates. It should be noted that
tries. Expectations about the financing needs as- the black market premium, the difference be-
sociated with large fiscal deficits also create un- tween the official and market exchange rates
certainty about future taxes and future inflation and an indicator of the extent of exchange rate
and deter planning for investment. overvaluation, was higher on average for devel-
Empirical studies have shown that fiscal oping countries that grew more slowly—particu-
deficits and overvalued exchange rates often larly among the low-income countries. The
have a negative impact on growth. Among low- strong correlation between this premium and
and middle-income countries, other than the growth underscores the problems caused by in-
rapidly converging countries, there is little or no appropriate exchange rate regimes, including
apparent relationship between growth or in- excessively high costs for investment goods and
come and average inflation, its variance, average consequently lower investment.
fiscal deficits, or average current account deficits
(Table 4.3). Nevertheless, the fastest-growing,
low- and middle-income countries on average Institutions and the Role of Government
generally had much lower and less variable infla- Economic uncertainty increases when wars,
tion and lower fiscal and current account military coups, political instability, and corrup-
deficits than other developing countries (al- tion are widespread or when basic institutional
though this may partly be an outcome of structures such as property rights, the rule of
stronger growth). This finding highlights that law, and those underpinning sound financial in-
beyond a certain threshold high inflation and stitutions are inadequate.27 In most countries
fiscal and current account deficits may be obsta- plagued by these problems, firms and people
cles to growth. Moreover, large fiscal deficits and face constraints on saving, investing, efficiently
high and variable inflation can lead to mis- allocating resources, and profiting from legal

27The negative economic consequences of war are well recognized. Thucydides, a historian in ancient Greece, com-

mented on this link in regard to the Peloponnesian war.

129
 CHAPTER IV HOW CAN THE POOREST COUNTRIES CATCH UP?

Table 4.4. Developing and Advanced Economies: Political Stability and Institutions1
(Average 1984–98)
Political and Government Contract
Country Groups Social Stability2 Stability3 Law and Order4 Security5
Low-income countries
Negative growth 5.0 4.2 3.9 4.1
Slow growth 5.3 4.9 4.6 5.0
Slow convergence 6.0 5.0 4.6 5.7
Fast convergence 7.0 5.9 5.6 6.8
Middle-income countries
Negative growth 5.7 5.5 5.4 6.0
Slow growth 5.8 5.0 4.5 6.0
Slow convergence 5.5 5.4 5.3 6.3
Fast convergence 7.8 6.0 7.0 7.8
Memorandum
Industrial countries 8.9 6.6 9.4 9.2
Other high-income, non-oil countries6 8.0 6.5 7.6 8.6
Middle- and high-income oil countries7 7.0 5.8 6.1 6.4
Source: International Country Risk Guide (published by the Political Risk Services Group).
1Unweighted cross-country averages. Data are normalized from 0 to 10: higher is better quality. For definitions of country groups, see Table 4.1.
2Political stability, as measured by the lack of political violence and its actual or potential impact on governance (“Internal Conflict” in the source).
3Government’s ability to carry out declared programs, and to stay in office (“Government Stability” in the source).
4Strength and impartiality of the legal system and general observance of the law (“Law and Order” in the source).
5Absence of risk of contract repudiation (“Contract Viability” in the source. Data through 1997).
6Cyprus, Hong Kong SAR, Israel, Singapore, and Taiwan Province of China.
7Equatorial Guinea, Gabon, Kuwait, Oman, Qatar, Saudi Arabia, and United Arab Emirates.

economic activities. While there are exceptions Corruption also hinders growth by distorting
to every rule, these economies will regress.28 incentives, as government officials and favored
It is hardly surprising that armed conflicts and private individuals receive a larger share of pub-
civil wars—as well as the turmoil associated with lic benefits or bear a lower share of the cost of
political instability—sharply increase uncertainty public goods. Empirical studies have found that
and diminish the expected profitability of invest- corruption lowers private investment, distorts the
ments. Political instability also discourages long- composition of public expenditures toward areas
term investment projects because of the risk of where collecting bribes are easier (for example,
policy reversals. Studies have found that political, toward excessive and inefficient physical public
social, and government instability raised invest- investments and away from education), and
ment risk in the poor-growth countries of Africa tends to reduce government revenue because it
and therefore was a major disincentive for for- contributes to tax evasion, improper tax exemp-
eign investors.29 For the broader group of devel- tions, or weak tax administration, thereby limit-
oping countries, these factors are also negatively ing the ability of the government to provide criti-
correlated with growth performance, particularly cal services.31 There is evidence as well that
for the low-income countries (Table 4.4).30 corruption increases income inequality and

28Empirical 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).
29See Paul Collier and Catherine Pattillo, Investment and Risk in Africa (London: MacMillan Press, 1999), Chapter 1.
30Similar results are found for indices of civil war.
31See 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).

130
 OBSTACLES TO GROWTH

poverty by reducing the progressivity of the tax Table 4.5. Developing and Advanced Economies:
system, the level and effectiveness of social Financial Maturity and Deepening1
spending, and the formation of human capital, (Percent of GDP)
and by perpetuating an unequal distribution of Broad Money
___________________________
asset ownership and unequal access to 1970 1998 Increase
education.32
Low-income countries
The lack of a strong and impartial judicial sys- Negative growth 20.5 22.4 1.9
tem, including the inadequate observance of Slow growth 29.5 23.5 –6.0
Slow convergence 22.9 35.9 12.9
laws and the inability to enforce laws, also signifi- Fast convergence 17.9 94.6 76.7
cantly increases economic and social uncertainty.
Middle-income countries
In particular, inadequate property rights, such as Negative growth 31.8 60.8 29.1
the risk of contract repudiation or of expropria- Slow growth 29.9 48.6 18.7
Slow convergence 23.9 46.8 22.9
tion without adequate compensation, substan- Fast convergence 30.4 75.8 45.4
tially increase the risks to entrepreneurship. Memorandum
Indicators of these risks appear largely related to Industrial countries 58.2 74.6 16.4
Other high-income,
growth both for low- and middle-income coun- non-oil countries2 51.0 125.2 74.2
tries (see the last two columns in Table 4.4). Middle- and high-income
Beyond maintaining peace and providing oil countries3 11.9 49.1 37.3
1Unweighted cross-country averages. For definitions of country
adequate institutional structures such as a sound
groups, see Table 4.1.
judicial system, the government needs to provide 2Cyprus, Hong Kong SAR, Israel, Singapore, and Taiwan Province

social or public goods, including some basic serv- of China.

3Equatorial Guinea, Gabon, Kuwait, Oman, Qatar, Saudi Arabia,
ices, education, health care, and when resources and United Arab Emirates.
are available, social safety nets. Better education,
family planning, and health care—especially for
women—can not only improve living conditions, Liberalization, along with proper supervision
but can also enhance the productivity of the la- and regulation, is probably most important in the
bor force and reduce population growth. This, in financial sector. The financial sector plays a pri-
turn, may promote per capita income growth mary role in intermediating saving and efficiently
particularly when government and household re- allocating investment. The financial system also
sources are being strained by a rapidly growing plays a crucial role in smoothing the effect of
population. The government may also need to shocks both on consumption and on investment.
make direct public investments to provide infra- On average, low-income countries had a low level
structure, when this is lacking or inadequate. In of financial sophistication, as proxied by the ratio
general, though, the government’s role in the of broad money to GDP, and slower-growing
economy needs to be limited, particularly in ar- countries had lower rates of financial deepening,
eas of the economy where the private sector can as proxied by the change in ratio of broad money
efficiently provide goods and services and where to GDP, although for the latter causality is diffi-
markets are not distorted. Excessive intervention cult to infer (Table 4.5). Cross-country empirical
tends to foster corruption, can strain govern- studies also confirm that slow growth is related to
ment budgets, and is prone to distort price sig- financial sector immaturity.34 Governments,
nals and the efficient allocation of resources.33 therefore, need to foster a competitive and effi-

32Sanjeev Gupta, Hamid Davoodi, and Rosa Alonso-Terme, “Does Corruption Affect Income Inequality and Poverty?”

IMF Working Paper 98/76 (Washington: International Monetary Fund, 1998).

33See Tanzi and Davoodi, “Corruption, Public Investment, and Growth.”
34See Ross Levine, “Financial Development and Economic Growth: Views and Agenda,” Journal of Economic Literature, Vol.

35 (June 1997), pp. 688–726.

131
 CHAPTER IV HOW CAN THE POOREST COUNTRIES CATCH UP?

Table 4.6. Developing and Advanced Economies: Globalization and Trade Openness1
(Percent of Purchasing Power Parity Adjusted GDP)
Exports
___________________________________ Openness (Imports and Exports)
__________________________________
Country Groups 1970–74 1994–98 Increase 1970–74 1994–98 Increase

Low-income countries
Negative growth 10.6 7.4 –3.2 22.9 18.0 –5.0
Slow growth 10.5 8.4 –2.1 23.9 20.6 –3.4
Slow convergence 4.8 8.1 3.3 16.4 19.5 3.0
Fast convergence 4.3 6.8 2.5 8.2 13.4 5.2
Middle-income countries
Negative growth 16.4 15.2 –1.2 31.4 33.2 1.8
Slow growth 17.8 18.1 0.3 36.4 38.5 2.1
Slow convergence 8.4 10.9 2.5 18.4 23.4 5.1
Fast convergence 18.2 28.1 9.9 38.5 55.5 16.9
Memorandum
Industrial countries 20.5 36.9 16.4 44.2 74.5 30.3
Other high-income, non-oil countries2 50.7 81.0 30.3 106.5 160.7 54.2
Middle- and high-income oil countries3 32.4 40.5 8.1 48.2 76.9 28.7
1Unweighted cross-country averages. For definitions of country groups, see Table 4.1. To account for the relative differences in prices, GDP is
measured in U.S. dollars based on purchasing power parity adjusted exchange rates. Differences across groups when GDP is measured with the
bilateral exchange rate are similar, although not as strong.
2Cyprus, Hong Kong SAR, Israel, Singapore, and Taiwan Province of China.
3Equatorial Guinea, Gabon, Kuwait, Oman, Qatar, Saudi Arabia, and United Arab Emirates.

cient financial sector, including a thorough su- tensive industries, which provide employment
pervisory and regulatory system.35 for the poor. When economies of scale are pres-
ent, inward-oriented countries also cannot bene-
fit from larger markets and the opportunities for
Inward-Oriented Policies greater specialization provided by trade. In addi-
Developing countries with inward-oriented, tion, domestic prices are less likely to reflect
protectionist economic policies (such as in world prices when a country maintains trade
much of Africa and Latin America during the barriers, thereby leading to a worse allocation of
1970s and 1980s) have suffered from poor resources. The lack of exposure to competition
growth rates, while those with outward-oriented from foreign sources can shield production in
policies (such as in east Asia) have prospered. A the short run but at the cost of reducing incen-
closed economy hinders growth through a vari- tives and opportunities for domestic producers
ety of channels. First and foremost, countries to innovate and improve productivity.
that have not adopted an orientation toward Many empirical studies have found a robust
open trade cannot exploit their comparative ad- and positive impact of increasing trade on
vantages in production. This may hurt the poor growth and income (even while attempting to
because import-competing industries tend to be control for reverse causation).36 For the groups
capital intensive and thus, without outward ori- of countries under review, the poorer and
entation, production moves away from labor-in- slower-growing countries were on average less

35See Manuel Guitián, “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.
36See, 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.

132
 OBSTACLES TO GROWTH

Table 4.7. Developing and Advanced Economies: Globalization and Trade Volumes and Prices1
(Annual Percent Change 1970–98, unless noted otherwise)
Volumes
____________________________________________ Prices
___________________________
Partner country Real export Gain in Goods terms Real nonfuel
Country Groups output growth growth market share2 of trade commodities
Low-income countries
Negative growth 3.3 2.2 –2.8 –1.0 –1.7
Slow growth 3.4 4.3 –0.8 –1.0 –1.6
Slow convergence 4.4 6.2 –0.4 0.4 –2.2
Fast convergence 4.5 11.1 4.2 1.9 –1.7
Middle-income countries
Negative growth 3.2 0.7 –4.1 0.9 –1.6
Slow growth 3.5 3.0 –2.2 0.1 –1.9
Slow convergence 3.2 6.9 2.2 –0.2 –2.0
Fast convergence 4.1 10.3 4.1 0.4 –1.7
Memorandum
Industrial countries 3.2 5.6 0.9 –0.6 –1.4
Other high-income, non-oil countries3 4.2 9.4 3.1 –0.8 –5.0
Middle- and high-income oil countries4 4.5 3.1 –3.7 2.8 –0.6
1Unweighted cross-country averages. For definitions of country groups, see Table 4.1.
2Difference between real export growth and partner-country import growth, which is proxied assuming an income elasticity for imports of 1.5,
consistent with world income and trading volume growth during 1970–98.
3Cyprus, Hong Kong SAR, Israel, Singapore, and Taiwan Province of China.
4Equatorial Guinea, Gabon, Kuwait, Oman, Qatar, Saudi Arabia, and United Arab Emirates.

open compared to the faster-growing, middle-in- in recent years was made in Latin America (in
come countries, and generally faster growth ac- conjunction with structural reforms following
companied increasing openness (Table 4.6).37 In the debt crisis) and eastern and central Europe,
addition, faster-growing, low- and middle-income followed by southeast Asia, which was already rel-
countries benefited from rapid growth in their atively open to trade. Trade reforms began
trading partners (Table 4.7). This can be ex- somewhat late in Africa, and many of these were
plained by groups of countries or regions, such in conjunction with IMF-supported structural ad-
as east Asia, growing together and expanding in- justment programs.
traregional trade. An alternative explanation is Few industrial countries have allowed develop-
that faster-growing countries were better able to ing countries substantially unimpaired or unlim-
take advantage of the opportunities provided by ited access to their markets on a unilateral basis.
increasing trade and globalization and to gain Since the Uruguay Round, progress in expand-
market share because of policies and conditions ing market access has been largely confined to
that were more supportive of export expansion. regional and bilateral trade arrangements, such
The evidence supports the latter explanation as those negotiated between the European
(see Table 4.7). Union and various developing country groups,
Many developing countries are already ad- including its neighbors and former colonies.
vanced in their efforts to open their economies These relatively recent agreements, though wel-
to trade.38 Unilateral liberalization began in the come, have tended to benefit selected developing
1980s and early 1990s as many countries shifted countries and not necessarily the poorest ones.
away from inward-looking policies. Most progress The present system of trade preferences excludes

37It 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 pre-
dicted trade flows or volumes. See Harrison, “Openness and Growth.”
38See Chapter V, “Trends and Issues in the Global Trading System,” in the October 1999 World Economic Outlook.

133
 CHAPTER IV HOW CAN THE POOREST COUNTRIES CATCH UP?

a number of “sensitive products” in precisely vide access to new technologies and production
those sectors—primarily agriculture, textiles, and processes through imported capital goods and
footwear—where many poor countries have the management (or human capital).
greatest potential to expand and diversify their During the last three decades, many develop-
exports. Moreover, the complexity, imperma- ing countries (particularly the fast-growing, mid-
nence, and lack of transparency in these arrange- dle-income countries) have benefited from sig-
ments have discouraged the desired response in nificant private capital flows and foreign direct
investment and trade. A bolder and more coher- investment. Exposure to capital flows, however,
ent approach to liberalization by the industrial also makes countries vulnerable to externall gen-
countries would be needed to spur development erated financial shocks and increases the need
and promote their integration into the world for strong domestic and external macroeco-
trading system. Such an approach would ideally nomic and financial fundamentals (including
combine the provision of duty-free and quota- adequate foreign reserves).40 For many low- and
free market access for products originating in de- some middle-income countries, therefore, full
veloping countries with a reduction in agricul- liberalization of their capital account may need
tural subsidies in the advanced economies. to wait until these countries are better able to
Although trade can create opportunities, it manage external risks. This includes the ability
can also create difficulties for developing coun- to pursue monetary and exchange rate policies
tries as they are exposed to externally generated that are consistent with a liberalized environ-
shocks. Terms-of-trade shocks may have been a ment. When liberalization does occur, measures
particularly important hindrance in the low-in- need to be carefully sequenced.
come countries with negative or low growth It is important to dispel the notion that it is
where terms of trade have fallen by 1 percent an- no longer possible for poor countries to benefit
nually on average since 1970 (see Table 4.7). from globalization. Some have argued that glob-
Most other developing countries, however, saw alization now offers few opportunities, stemming
flat or even rising terms of trade. A key reason from the belief that an outward-oriented strategy
why some low-income countries had falling based on exporting labor-intensive goods (and
terms of trade is that primary products are a taking advantage of relatively lower labor costs)
larger share of their exports and nonfuel com- has become much more difficult because large
modity prices fell in real terms.39 The impact of countries such as China already dominate export
future commodity price declines on these markets in these goods. Essentially, the argu-
economies is likely to diminish automatically as ment is that it is too late for those who have not
they grow and become less dependent on pri- yet started. However, the same argument was
mary products for export revenue. This process, made in east Asia when first Japan and then
however, can be expected to be slow. Hong Kong SAR, Korea, Singapore, and Taiwan
Openness to capital flows holds risks for devel- Province of China exported labor-intensive
oping countries, but at the same time it allows goods. This did not prevent China, Indonesia,
countries to meet their financing needs when in- Malaysia, and Thailand from following suit later
vestment requirements exceed domestic saving. when the other east Asian countries shifted to
Moreover, private capital flows, particularly in less labor-intensive exports as relative wages rose.
the form of foreign direct investment, also pro- Moreover, as incomes increased in the region,

39Jeffrey 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.
40For 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.

134
 OBSTACLES TO GROWTH

Table 4.8. Developing Countries: Poverty and Poverty Reduction1

(Percent of population)
Below $1 per Day
____________________________________ Below $2 per Day
_____________________________________
Country Groups 1987 19982 Reduction 1987 19982 Reduction
Low-income countries
Negative growth 40.7 41.6 –0.9 77.7 77.3 0.4
Slow growth 44.4 43.6 0.8 88.6 86.9 1.7
Slow convergence 47.7 42.3 5.4 85.2 70.9 14.3
Fast convergence 28.2 16.4 11.8 69.2 50.9 18.3
Middle-income countries
Negative growth 7.0 8.6 –1.6 32.1 34.4 –2.3
Slow growth 12.9 13.8 –0.9 36.8 36.3 0.5
Slow convergence 12.0 9.0 3.0 34.4 22.3 12.1
Fast convergence 10.7 2.7 8.0 37.0 20.3 16.7
Source: Martin Ravallion and Shaohua Chen, “Global Poverty Measures.”
1Based on survey data. It is important to note that income growth measures that underly the survey poverty data differ from national accounts
income growth measures, which are used to determine the country groups.
2Estimated.

intraregional trade expanded rapidly, creating unclear. In a subsistence economy, it is unlikely

new markets for exports. that saving and investment (including in human
capital) will be sufficient to promote growth since
income is required to provide basic necessities.
Poverty and Income Inequality High poverty rates can, therefore, be impediments
Monetary measures of poverty reduction, not to growth. The observation that income is more
very surprisingly, appear to be correlated with in- equitably distributed in the high-growth, countries
come growth (Table 4.8). One recent study con- of east Asia compared to some lower- growth
firms this relationship and finds that a 10 per- Western Hemisphere countries also has led to
cent rise in per capita income is correlated with a speculation that income inequality hampers
10 percent increase in income among the poor- growth. However, both the theoretical relationship
est quintile.41 Other studies correlate changes in between growth and income inequality and the
aggregate income and poverty rates directly and empirical evidence are ambiguous.43 If individual
generally suggest that a 1 percent increase in av- saving rates rise with the level of income, a less-
erage per capita income or consumption is asso- even income distribution may actually result in a
ciated with a reduction in poverty rates by up to higher aggregate saving rate. However, significant
3!/2 percent and also find that poverty rates fall, income inequality also implies that low-income
almost always, with growth in average living stan- groups may be unable to acquire the skills neces-
dards and rise with contraction.42 sary to benefit from economic opportunities, re-
Although poverty reduction and economic sulting in a potential loss of human capital.
growth are correlated, the causal relationship be- Irrespective of the direct economic links be-
tween poverty or income inequality and growth is tween income inequality and growth, there are

41Michael Roemer and Mary Kay Gugerty, “Does Economic Growth Reduce Poverty?” HIID Technical Paper (Cambridge,

Massachusetts: Harvard Institute for International Development, March 1997).

42Martin 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.
43Some 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 cross-
country differences in the measurement of income inequality.

135
 CHAPTER IV HOW CAN THE POOREST COUNTRIES CATCH UP?

many other reasons to believe that the persist- trast, have access to private markets so that for-
ence of large income divergences and high eign capital flows, in the form of both debt and
poverty rates, even among many middle-income equity, can fill the financing gap and provide the
countries, can be a powerful obstacle to stronger resources to spur growth. Equity is an important
economic growth. In addition to the waste of hu- component of financing from the perspective of
man potential, widespread poverty and large dif- recipient countries, not only because risk is
ferences in income and wealth can be a source shared by foreign stakeholders, but also because
of social unrest and political instability. This in it often comes in the form of foreign direct in-
turn can prevent the establishment and mainte- vestment that tends to bring in new technology
nance of adequate institutional structures and and physical and human capital, including man-
policies, and eventually lead to an unstable eco- agement expertise. However, because interna-
nomic environment that will adversely affect pri- tional capital markets may be imperfect and the
vate saving and investment decisions and deter acquisition of timely information on investment
foreign investors. projects by perspective investors costly, countries
In some cases, policies that promote growth often rely more heavily on borrowing than eq-
may not promote immediate poverty reduction. uity finance. There is no evidence that moderate
Some studies have found evidence that policies external indebtedness hurts growth, but when
such as fiscal stabilization and trade liberaliza- indebtedness rises to the point that debt service
tion, which are both essential for sustained becomes onerous, growth prospects are substan-
growth, may raise poverty rates in the short tially damaged.
term.44 This may call for complementary reforms External debt can become an impediment to
to strengthen social safety nets. In contrast, re- growth for a developing country when funds are
ducing high inflation rates is almost always bene- not used for productive investments that allow
ficial for the poor. In any case, the short-run neg- the country to service its debt on time. One sign
ative impact, if any, of necessary adjustment of an excessive debt burden is the regular need
measures is typically modest compared with the for comprehensive rescheduling or the pro-
long-run gains to the poor from the additional tracted buildup of external arrears, as has been
growth that results from these policies. observed in many low-income countries.
Excessive levels of external debt are likely to re-
duce incentives for a government to undertake
The Role of Debt appropriate reforms and businesses to invest be-
As discussed above, access to external finance cause of the real possibility that a significant
is vital for many developing countries because share of the returns from these activities will
domestic saving is usually insufficient to meet in- need to be transferred to foreign creditors to re-
vestment needs. In low-income countries, domes- pay the outstanding debt.45 Moreover, high debt
tic saving is limited by factors such as poverty levels create moral hazard: a government may
and underdeveloped financial markets. These delay reforms needed to reduce the debt burden
countries typically do not have access to private because it expects debt relief in the future.
finance and must rely on official lending (and
aid). Many middle-income countries, by con- ***

44For 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).
45See 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).

136
 DEBT BURDEN AND DEBT RELIEF

Clearly, there are many possible explanations Almost all of them (38 countries) are low-in-
for the failure of the poorest countries to catch come countries, and about half of them (21
up. Many of these reasons are interrelated and countries) experienced negative per capita in-
contribute to a vicious circle. While it may be come growth over the past 30 years.
difficult to identify unambiguously the most crit- Worldwide events in the 1970s and at the be-
ical factors, there are many shortcomings that ginning of the 1980s were major contributors to
appear to be contributing to, or at least be asso- the debt buildup in both the HIPCs and some
ciated with, persistently inadequate growth. This middle-income countries.46 Oil price shocks and
calls for a broad-based, sustained effort if greater the rise in industrial country interest rates gen-
progress is to be achieved. This effort, first and erated balance of payment pressures for many
foremost, will need to be undertaken by the developing countries. These factors were exacer-
poorest countries themselves: without the bated by stagnant or contracting export rev-
strongest commitment on the part of their lead- enues, caused mainly by the subsequent world
ers and elected bodies, supported by society at recession and declining commodity prices, and
large, success will be elusive. But the poorest by the appreciation of the U.S. dollar, which in-
countries cannot succeed without support from flated the domestic value of dollar-denominated
the international community. The advanced debt and service payments and depressed dollar
economies, in particular, will need to increase commodity prices.
levels of foreign aid and assistance to countries Domestic factors also played a large role in
that have strengthened their own efforts to alle- the debt buildup. (Compare the performance of
viate poverty. The international community also both the high-debt countries and “successful ad-
needs to liberalize fully trade in products in justers” to the low-debt countries in Figure 4.6.
which the poorest countries have a comparative See figure footnotes 1–3 for group definitions.)
advantage, and the debt burden also needs to be Countries with low saving and large current ac-
addressed. count deficits were unable to withstand the im-
pact of these external shocks and continued to
borrow heavily as their reliance on external fi-
Debt Burden and Debt Relief nancing grew. Because low export and fiscal rev-
Over the past 30 years, the total external debt enue shares of GDP made it difficult for both
of developing countries has risen sharply from the private and the public sectors to set aside the
$90 billion in 1970 (or 15 percent of GDP) to al- resources required to service old and new debt,
most $2,000 billion in 1998 (37 percent of GDP; external debt quickly rose to unsustainable lev-
Appendix Table 38). Debt burdens vary across els. The drag on domestic resources, coupled
countries, and although the debt profile for with already low private and public investment,
some countries has improved substantially since further constrained the potential for growth and
the debt crisis in the 1980s, for many low-income exports, in some cases inducing a vicious cycle of
countries the picture is bleak. At the turn of the unsustainable indebtedness and low growth.
millennium, up to about 40 of the poorest devel- International private investors, increasingly
oping countries, the Heavily Indebted Poor alarmed by the financial conditions of debtor
Countries (HIPCs), still have unsustainable debt countries, became more resistant to granting
burdens even after large-scale and persistent fi- new lending or debt rollovers, further exacerbat-
nancial assistance provided by official donors. ing the difficulties of these countries. In order to

46For 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).

137
 CHAPTER IV HOW CAN THE POOREST COUNTRIES CATCH UP?

alleviate such difficulties, official creditors ini-

tially provided nonconcessional rescheduling of
debt-service payments through the Paris Club,
under the presumption that the debt crisis was
temporary and liquidity related. This nonconces-
sional rescheduling, although granting tempo-
rary cash-flow relief, contributed to increasing
the outstanding stock of debt.47
In the 1980s, it became apparent that the debt
Figure 4.6. Developing Countries: Economic Indicators crisis involved more than temporary liquidity
(Percent of GDP unless otherwise noted) problems and actions beyond cash-flow resched-
uling would be needed. The international finan-
High-debt countries rank below low-debt countries in all macroeconomic and
structural indicators. The successfully adjusting countries reduced fiscal deficits and cial community therefore launched what would
increased saving and exports.
become a series of assistance initiatives aimed at
High-debt Low-debt Successful reducing the value of future obligations and re-
countries1 countries 2 adjusters 3
payments. In association with adjustment pro-
1000 6000 grams supported by multilateral institutions,
Debt Real PPP GDP Per Capita
(percent of exports) (1996 dollars)
5000 middle-income countries with excessive debt
800
arranged market-related debt reduction deals
4000 with their creditors under the Brady Plan, while
600
3000
poor countries received concessional reschedul-
ing from official creditors and new lending on
400
2000 increasingly concessional terms. The assistance
200
schemes were often complemented by, if not
1000
conditional on, macroeconomic and structural
0 0 reforms undertaken by the recipient countries.
1971 76 81 86 91 96 1971 76 81 86 91 96
These concerted efforts, by both the interna-
40
Saving Investment
40 tional community and recipient countries,
proved to be effective in helping many develop-
30 30
ing countries adjust their resource balances, re-
duce outstanding debt ratios to sustainable lev-
els, and resume steady output growth. The
20 20
“successful adjusters” in Figure 4.6 were mostly
in the middle-income group and had access to
10 10 international financial markets. As a group, they
drastically reduced the public sector deficit,
0 0
which helped to increase aggregate savings and
1971 76 81 86 91 96 1971 76 81 86 91 96 reduce the trade deficit. At the same time, coun-
tries initiated macroeconomic stabilization, pri-
vatization, and market liberalization programs,
which all contributed to the resumption of

47See 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
[Link]/external/np/HIPC.

138
 DEBT BURDEN AND DEBT RELIEF

steady growth, although in many cases the ad-

justment 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 tradi-
tional debt-relief mechanisms, official assistance,
and adjustment policies and their implementa-
tion were insufficient to reduce excessive indebt-
edness. Their economic performance has lagged
and debt ratios have remained high, despite a se- Figure 4.6 (continued )
ries of Paris Club Initiatives (concessional
High-debt Low-debt Successful
reschedulings known as Toronto, London, and countries1 countries 2 adjusters 3
Naples terms), and additional action by a num- 50 50
Exports Fiscal Revenues
ber of non-Paris Club bilateral and commercial
creditors, aimed both at providing immediate 40 40
cash-flow relief on the payments falling due and
slowing the growth of the debt stock itself. The 30 30
costs of these early debt-relief initiatives for low-
income countries have been estimated to be at 20 20

least $30 billion, and possibly much more.48

10 10
Many of these poor countries had limited suc-
cess in improving their debt-burden indicators
0 0
and rekindling output growth because they were 1971 76 81 86 91 96 1971 76 81 86 91 96
unable to provide a stable macroeconomic envi- 4 4
Current Account Fiscal Balance
ronment, as demonstrated by persistently high
fiscal deficits and increasing inflation. (See the
0 0
group of high-debt countries in Figure 4.6.)
Accordingly, aggregate saving fell further, and
profitable investment opportunities were not ex- –4 –4
ploited. High debt-service payments continued
to exacerbate already wide current account –8 –8
deficits or forced countries to run arrears. By
the mid-1990s outstanding debt of these coun-
–12 –12
tries remained high at around 600 percent of ex- 1971 76 81 86 91 96 1971 76 81 86 91 96
ports on average (compared to about 150 per-
cent 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

48Daseking and Powell, “From Toronto Terms to the

HIPC Initiative,” review debt assistance provided to low-in-

come countries and provide estimates of the associated
costs.

139
 CHAPTER IV HOW CAN THE POOREST COUNTRIES CATCH UP?

external indebtedness. Conditions and policies

prior to the debt crisis may have played an im-
portant 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
Figure 4.6 (concluded ) for the former group while low private saving
may have been a larger factor for the latter. To
High-debt Low-debt Successful
countries1 countries 2 adjusters 3 the extent that government policies can more
easily address public saving (by reducing the fis-
60 60
Inflation Openness 4 cal deficit) than private saving, adjustment may
(percent change
50 in CPI) 50 have been easier for the countries that started
with higher public sector deficits. Also, low pri-
40 40
vate saving in the unsuccessful countries may be
30 30 partly a reflection of much lower average per
capita income in this group.
20 20

10 10
The Enhanced HIPC Initiative: The Road
0
1971 76 81 86 91 96 1971 76 81 86 91 96
0 to Sustainability?
In September 1996, the IMF and the World
50 25
Financial Maturity 5 Official Aid Bank jointly launched the Initiative for Heavily
40
Indebted Poor Countries, or HIPC Initiative. It
20
was based on concerted financial assistance from
30 15 the international community (in addition to tra-
ditional debt-rescheduling mechanisms) and was
20 10 available to countries that demonstrated a
record of successful macroeconomic and struc-
10 5 tural adjustment and met other eligibility crite-
ria. By 1999, seven countries had qualified for as-
0 0
1971 76 81 86 91 96 1971 76 81 86 91 96 sistance under the Initiative.49
Sources: IMF, World Economic Outlook; and World Bank, World Development
A central objective of the HIPC Initiative is to
Indicators. reduce external debt of the qualifying countries
1HIPC countries plus countries whose debt-to-export ratio was above 200 percent
to sustainable levels. In theory, a debt burden is
in 1996–97.
2Countries whose debt-to-export ratio was always below 200 percent in the sustainable if the debtor country can be expected
1971–97 period (based on three-year averages). to meet its current and future debt-service obli-
3Countries whose debt-to-export ratio went temporarily above 200 percent during

the 1971–97 period. gations without recourse to debt relief or similar

4Average of imports and exports of goods and services as a percentage of GDP.
assistance. In practice, however, identifying a
5Ratio of M2 to GDP (in percent).
maximum or target level of debt that can be ex-

49See David Andrews and others, “Debt Relief for Low-

Income Countries: The Enhanced HIPC Initiative,”

Pamphlet Series No. 51 (Washington: International Mone-
tary 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 [Link]/external/np/HIPC.

140
 THE ENHANCED HIPC INITIATIVE: THE ROAD TO SUSTAINABILITY?

Table 4.9. Heavily Indebted Poor Countries (HIPCs): Selected Debt Indicators 1
(Billions of dollars, at the year prior to the decision point, unless otherwise indicated)
Debt in Net Present Value
__________________________________________________________________________
Estimated Estimated
Average Average Target Target HIPC HIPC
Nominal ratio to ratio to (at 150 (at 250 assistance assistance
debt
_______ exports revenue percent of percent of at decision in end-1999
Level Level (percent) (percent) exports) revenue) point2 terms3
Group 14 41.1 22.5 242 323 ... ... 7.7 7.7
Qualifying under exports criteria 21.8 11.1 288 297 5.9 ... 5.2 5.2
Qualifying under fiscal criteria 19.4 11.3 150 375 ... 8.9 2.4 2.4
Group 25 66.1 31.6 349 521 ... ... 14.7 13.9
Qualifying under exports criteria 53.4 22.4 365 509 12.4 ... 10.0 9.4
Qualifying under fiscal criteria 12.7 9.3 283 566 ... 4.5 4.8 4.5
Other HIPCs6 81.8 35.9 275 426 ... ... 8.0 6.6
Qualifying under exports criteria 34.6 12.7 415 615 6.7 ... 6.0 4.8
Qualifying under fiscal criteria 6.0 4.2 205 421 ... 2.2 2.0 1.8
Deemed sustainable 41.3 19.0 99 145 ... ... ... ...
Total 189.1 90.0 297 440 ... ... 30.4 28.2
1For details on the HIPC Initiative and the net present value (NPV) debt ratios, see [Link] For
details on groupings and costing, see [Link]
2Difference between NPV debt level and NPV debt targets.
3Data may differ from previous column due to base year employed in NPV calculation.
4Countries that are being reassessed under the enhanced HIPC Initiative: Benin, Bolivia, Burkina Faso, Côte d’Ivoire, Guyana, Mali,
Mozambique, Senegal, and Uganda. The assumed decision point of each country is 1999.
5Countries that are expected to qualify for assistance by the end of 2000 under the enhanced framework of the HIPC Initiative: Cameroon,
Chad, Ethiopia, Guinea, Guinea-Bissau, Honduras, Lao PDR, Malawi, Mauritania, Nicaragua, Niger, Rwanda, Sierra Leone, Tanzania, and Zambia.
The assumed decision point of each country is 2000.
6Angola, Burundi, Central African Republic, Republic of Congo, Kenya, Madagascar, Myanmar, Saõ Tomé and Príncipe, Togo, Vietnam, and
Yemen. The assumed decision point of each country ranges from 1999 to 2003.

pected to be sustained is far more difficult be- Sustainable debt levels are typically expressed
cause it involves the uncertainties of projecting in terms of net present value, a concept that is
future earnings that would be used to service the estimated as the stream of future scheduled
debt. For a nation, exports are typically the ap- debt-service flows discounted to today’s value at
propriate measure of earnings as these provide market interest rates. The net present value of
the necessary foreign currency for external debt debt can differ from contract amounts (or nomi-
service, but for some countries, government rev- nal face value) when loans are made at conces-
enue may be appropriate (for example, when the sional, below market, interest rates or when debt
debt is mostly public and when tax revenue is a has been rescheduled at concessional rates.
small proportion of output). Estimating future Thus, the difference between contract amounts
export earnings (or output and tax revenue) is (Table 4.9, first column) and the estimated net
always difficult due to unforeseen factors such as present value (second column) provides a rough
shocks; these estimates are especially uncertain indication of the concessional component of
for heavily indebted countries because the debt lending and of debt relief associated with tradi-
level itself can affect output through disincen- tional rescheduling techniques.50 For all HIPCs,
tives for policymakers and investors. this difference amounts to about $100 billion. It

50Fluctuations 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.

141
 CHAPTER IV HOW CAN THE POOREST COUNTRIES CATCH UP?

does not reflect the impact of debt forgiveness, percent of exports or, for those eligible under
which lowers debt both in nominal and in net the fiscal window, 250 percent of government
present value terms. revenue, whichever provides greater debt relief
The HIPC Initiative targets the net present to the recipient country. Deeper debt relief is
value of debt expressed as a percentage of ex- likely to have several advantages. First, it is hoped
ports or, in some cases, government revenue. that additional debt reduction will free resources
These targets were established based on past for poverty reduction and other important objec-
experience of countries that have successfully tives set in the country’s Poverty Reduction
avoided debt servicing problems.51 On the basis Strategy Paper. Second, the more ambitious tar-
of these targets, debt relief was already ex- gets of the enhanced HIPC Initiative will provide
tended or committed to seven qualifying coun- a greater safety margin for the achievement of
tries, while the debt burden of two other coun- debt sustainability and increase the chances of a
tries was deemed sustainable after traditional permanent exit from the need for debt resched-
debt-relief mechanisms. For all of these nine uling. This in turn will improve the chances of
countries, the debt in net present value terms future economic success by strengthening incen-
was reduced to $22!/2 billion, or 240 percent tives for economic reform and private invest-
of exports in 1998 (see Table 4.9, Group 1), ment and reducing moral hazard.
and compares to a total nominal debt level of Under the enhanced HIPC Initiative, all the
about $41 billion. The difference reflects the countries in Group 1 may be reassessed for fur-
impact of past concessional lending and debt ther assistance in light of the new sustainability
restructuring. targets. Five countries are expected to qualify un-
In September 1999, after a review of the initial der the export criteria and three under the fiscal
implementation phase, it was decided to enhance criteria. These countries are expected to receive
the HIPC Initiative to provide deeper, faster, and debt relief of approximately $8 billion (in net
broader debt relief, and to accelerate poverty re- present value terms in 1999). Countries that are
duction (Box 4.3). The enhanced Initiative speci- expected to qualify for assistance by the end of
fies lower sustainability targets, more flexible eli- 2000 (Group 2 in Table 4.9) have higher average
gibility criteria, front-loaded debt relief, and a debt in net present value terms than the coun-
specific link between debt relief and poverty re- tries that received assistance under the original
duction. Debt relief under the enhanced HIPC HIPC Initiative (Group 1). For the Group 2
Initiative would be tied to the recipient country’s countries, past concessional arrangements and
adoption and implementation of a poverty reduc- the full use of traditional debt-relief mechanisms
tion strategy (see Box 4.1). would lower the total debt burden from a nomi-
In particular, debt-sustainability targets were nal value of $66 billion to a net present value
lowered as a direct way of providing deeper debt measure of about $32 billion (in 1999 terms),
relief for qualifying countries and to broaden the but before HIPC relief. Under current plans, and
list of countries that could potentially qualify for assuming all countries in the group can proceed
assistance. Thus, debt relief will now be aimed at with debt relief as expected, debt in present value
reducing the net present value of debt to 150 terms would be halved, to about $17 billion.52

51The 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 concen-
tration 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).
52See Andrews and others, “Debt Relief for Low-Income Countries: The Enhanced HIPC Initiative” and

[Link]/external/np/HIPC/cost4/[Link].

142
 THE ENHANCED HIPC INITIATIVE: THE ROAD TO SUSTAINABILITY?

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

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

1Currently,
Debt relief for HIPCs could provide addi-
there are 40 HIPC-eligible countries, of
which 8 countries have reached the decision point—
tional resources for anti-poverty programs. In
i.e., the point when a country’s qualification for HIPC the HIPCs for which data are available, debt
assistance is determined (Bolivia, Burkina Faso, Côte service paid and the stock of debt as a share of
d’Ivoire, Guyana, Mali, Mauritania, Mozambique, and GDP are higher than in other poor countries el-
Uganda), and 4 countries have reached the comple- igible for concessional lending under the PRGF.
tion point—i.e, the point when additional measures
are taken to assist a country to reach a sustainable
Although country experiences vary consider-
debt level (Bolivia, Guyana, Mozambique, and ably, on average, HIPCs allocate slightly more
Uganda). In addition, Bolivia and Uganda have budgetary resources to debt service than to ed-
reached their second decision points. Benin and ucation and health-care taken together, both in
Senegal reached decision points, were determined un- terms of GDP as well as total government ex-
der the Initiative to face sustainable debt burdens af-
ter traditional debt relief, and therefore did not re-
penditures (about 5 percent of GDP and 20
ceive assistance. There are 40 other PRGF-eligible percent of total government expenditures, re-
countries in addition to HIPCs. spectively). Total debt relief under the Heavily

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

(In units as indicated; latest year for which data are available)1
Education Spending2
_____________________ Health Spending2
____________________ Debt Service3
____________________
In percent In percent In percent
of total of total of total Debt Stock
In percent government In percent government In percent government In Percent Number of
of GDP expenditures of GDP expenditures of GDP expenditures of GDP Countries4
HIPCs5 3.3 13.2 1.6 6.3 5.1 19.8 117.1 30
of which program countries 3.4 13.5 1.7 6.5 5.1 20.4 130.5 28
Non-HIPC PRGF-eligible
countries6 4.6 15.3 2.5 8.0 2.9 11.2 56.6 20
of which program countries 3.9 15.4 1.8 7.3 3.0 11.6 58.6 13
Sources: World Bank; national authorities; and IMF staff estimates.
1For most countries, the latest year for which data are available is 1997 for debt service and 1998 for health and education spending.
2In general, data on local government spending and in-kind donor contributions are not available, thereby understating total public spending.
3World Bank Global Development Finance (GDF) estimates of debt service paid, which may be lower than debt service due. However, caution
should be exercised in interpreting these ratios due to the misclassification of debt service between cash and accrual.
4Sample size may vary across categories.
5Excludes Nigeria.
6Excludes transition economies (Albania, Armenia, Azerbaijan, Bosnia and Herzegovina, Georgia, the Kyrgyz Republic, the former Yugoslav
Republic of Macedonia, Moldova, Mongolia, and Tajikistan) and includes Nigeria.

143
 CHAPTER IV HOW CAN THE POOREST COUNTRIES CATCH UP?

Box 4.3 (concluded)

spending on education and health care alone

Social Spending and Social Indicators, 1985–981 provides only a partial view of the relationship
(Annual percent change) between international official resource flows
HIPCs Non-HIPC PRGF-eligible
and spending on poverty reduction programs,
countries2 because public spending on poverty reduction
8 is not confined to education and health care
Education
programs. Public spending on, for instance,
rural roads, sanitation, and access to safe water
6
could contribute as much, or more, to poverty
reduction. Gross external financing flows
4
(new loans and grants, for instance) are not
necessarily independent of debt-service pay-
2 ments, and these flows are typically higher
than either social spending or debt-service
0 payments in HIPCs. For countries to reap full
Real per capita Gross primary Gross secondary Persistence to benefits from debt relief, efforts are needed to
spending enrollment enrollment grade five
improve the efficiency of existing, as well as
additional, public outlays made possible by debt
10
Health Care relief. This will require action on at least three
fronts.
8
In HIPCs, the allocation of budgetary re-
6 sources within the social sectors (e.g., between
primary and tertiary education) is typically
4 skewed toward services that are less demanded
by the poor, such as tertiary education and
2 curative, rather than preventive, health care.
For instance, data for selected HIPCs show that
0
Real per capita Immunization Access to safe Births attended 14 percent of total spending on public educa-
spending DPT water by skilled staff tion and 13 percent of health care outlays, on
average, accrue to the poorest fifth of the
Sources: IMF staff estimates; national authorities; and World Bank, World population (lowest quintile), compared with
Development Indicators.
1Average annual change between first year since 1985 and the most recent 30 percent for the richest quintile for both
year for which data are available. spending categories (see the second figure).3
2Excludes transition economies: Albania, Armenia, Azerbaijan, Bosnia and
Furthermore, public spending on primary
Herzegovina, Georgia, the Kyrgyz Republic, the former Yugoslav Republic of
Macedonia, Moldova, Mongolia, and Tajikistan.
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
Indebted Poor Country (HIPC) Initiative is
estimated at $28.2 billion (in end-1999 present 3Hamid Davoodi and Sawitree Sachjapinan, “How

value terms), including debt forgiveness and Useful Are Benefit Incidence Studies?” IMF Working
Paper (Washington: International Monetary Fund,
concessional debt service.2 However, compar-
forthcoming).
ing data on debt service paid and public 4In a smaller sample of HIPCs for which informa-

tion is available, public spending on health favors the

2Information is available via the Internet at rural population over the more prosperous urban
[Link]/external/np/vc/1999/[Link]. population.

144
 THE ENHANCED HIPC INITIATIVE: THE ROAD TO SUSTAINABILITY?

share of wages and salaries in social sector

allocations have weakened the link between Benefit Incidence of Public Spending on Education
public expenditures and improvements in social and Health Care in Heavily Indebted Poor Countries,
indicators.5 Early 1990s1
(Percent of total spending)
The capacity to formulate and execute the
budget in a transparent manner is critical for Bottom income Top income
quintile quintile
anti-poverty programs. In this context, the use 35
of poverty funds for channeling debt relief to
30
the poor has been widely advocated. Such
funds should be subject to the same level of 25

scrutiny and oversight as other government 20

spending, including transparency and accounta- 15
bility both to parliament and the executive.
10
Anti-poverty programs should be integrated
into the budget to prevent leakages, duplica- 5
tion, and implementation of relatively low-prior- 0
ity projects. All education (9) Primary education (10) All health (8)
Effective monitoring of debt relief for poverty
reduction requires timely and adequate data.6 To
Rural Urban 2
this end, coordinated technical assistance from 70
international organizations and donors, as well
60
as increased efforts by HIPCs themselves, will be
crucial.7 At present, coverage of public spending 50
data is not comprehensive in many countries, 40
and typically excludes spending by local govern- 30
ments and in-kind donor contributions. Public
spending data become available with a lag, which 20

for some countries can be as long as two to three 10

years. Moreover, virtually no HIPC-eligible coun- 0
try has consistent annual series for public expen- All education (5) All health (4)
diture allocations within the education and
Source: Davoodi and Sachjapinan, “How Useful are Benefit Incidence
health sectors (e.g., distinguishing between out-
Studies?”
lays on primary and tertiary education, preven- 1Numbers 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 samples.

5Vito Tanzi, “Corruption Around the World: Causes,

Consequences, Scope, and Cures,” IMF Staff Papers,

Vol. 45 (December 1998), pp. 559–94.
6Sanjeev Gupta and others, Social Issues in IMF-

Supported Programs, IMF Occasional Paper No. 191

tive and curative health care, and wage and non-
(Washington: International Monetary Fund, 2000). wage items). Social indicators are produced in-
7Currently, the IMF, the Organization for Economic frequently, and typically become available only
Cooperation and Development, the United Nations, every five years. Also, data for many important
and the World Bank are collaborating in the indicators are derived from models rather than
Partnership in Statistics for Development in the
Twenty-First Century Consortium (PARIS21) to sup-
from actual observations. Up-to-date information
port statistical capacity building in countries preparing on poverty indicators, including the nature and
poverty reduction programs. locus of poverty, is often lacking.

145
 CHAPTER IV HOW CAN THE POOREST COUNTRIES CATCH UP?

Table 4.10. Heavily Indebted Poor Countries (HIPCs): Economic Indicators

(Percent of GDP unless otherwise indicated)
Developing Countries
Group 11
__________________________ Group 22
___________________________ excluding HIPCs
___________________________
1990–94 1995–97 1998–99 1990–94 1995–97 1998–99 1990–94 1995–97 1998–99
Real per capita GDP3 0.4 3.4 2.5 –2.1 1.6 –1.1 1.6 1.7 0.2
Saving 7.4 13.2 14.8 9.0 8.5 9.2 19.7 21.8 20.4
Total investment 18.6 21.1 22.7 16.4 15.5 17.8 24.2 26.3 24.9
Private investment 10.2 12.1 14.0 8.5 7.2 8.0 ... 16.7 16.8
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 services 28.4 31.4 31.0 20.2 23.2 23.8 39.8 41.0 40.9
Fiscal revenues4 21.6 22.1 21.9 17.2 18.8 18.4 28.5 27.7 27.1
CPI Inflation3 15.1 9.1 3.7 31.0 15.7 11.8 19.7 10.2 7.5
Openness5 33.5 36.0 35.9 26.1 29.3 30.6 42.7 43.0 43.0
Broad money 26.6 28.2 28.8 19.1 18.8 18.9 49.1 49.4 ...
Net private capital flows 1.6 3.7 3.7 –2.8 1.1 1.3 4.1 3.2 3.0
Official assistance6 12.5 2.1 5.3 14.8 12.0 11.7 ... 3.6 3.4
Sources: World Bank, World Development Indicators; and IMF staff estimates.
1Countries for which assistance has been extended (Uganda, Bolivia, Guyana, Mozambique), committed (Burkina Faso, Côte d’Ivoire, Mali), or
deemed not required (Benin, Senegal) under the initial framework of the HIPC Initiative.
2Countries that are expected to qualify for assistance by the end of 2000 under the enhanced framework of the HIPC Initiative: Cameroon,
Chad, Ethiopia, Guinea, Guinea-Bissau, Honduras, Lao PDR, Malawi, Mauritania, Nicaragua, Niger, Rwanda, Sierra Leone, Tanzania, and Zambia.
3Percent change.
4Including grants.
5Exports plus imports.
6Official lending and transfers.

Debt-relief estimates for other HIPCs in Table 4.9 of private and public investment, and stronger
are very preliminary. private capital inflows. The Group 2 countries
To qualify for HIPC assistance, countries need that are expected to qualify for HIPC assistance
to implement a program of macroeconomic and in the course of 2000 also recorded improve-
structural reform and demonstrate a track ments in these economic variables, but their situ-
record of good performance because experience ations today remain substantially worse than
suggests that progress in these areas is a neces- those in Group 1 in 1995–97. This lack of suffi-
sary precondition for successful debt reduc- cient progress indicates that many countries now
tion.53 An improved track record from 1995 on- anticipating HIPC debt relief will need firm im-
ward is evident in the countries that received plementation of policy reforms to qualify for
relief under the original HIPC Initiative (Group assistance.
1 in Table 4.10). While it is difficult to deter-
mine causality, better policies and implementa-
tion are evident in lower fiscal deficits and infla- The New Approach to Poverty Reduction
tion. These and other policy improvements have Poverty reduction has been and will need to
contributed to a stronger resource balance and be for a long time among the highest priorities
ability to service debt (a rise in domestic savings for domestic policymakers, the donor commu-
and an improving trade deficit, associated with nity, and international institutions. The develop-
larger export and fiscal revenues), higher levels ment strategies followed until now have worked

53William 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.”

146
 THE NEW APPROACH TO POVERTY REDUCTION

for some countries, but for far too many, poverty ital. Each country will need to decide on how
remains prevalent and a stronger approach is best to provide these fundamentals within its de-
needed. velopment strategy.
There are, nevertheless, grounds for opti- The enhanced development strategy frame-
mism. First, the industrial economies have work stresses three elements. First, each country
demonstrated that long periods of sustained will formulate a poverty reduction strategy to
growth are possible and that living standards can elaborate its plans for development and poverty
improve dramatically over a few decades. reduction in a multi-year framework. These
Second, the newly industrialized economies, al- strategies will be prepared by the country itself,
though small in number, have shown that rapid with the participation of civil society, in order to
catch-up is possible with growth rates even increase the level of national awareness of the is-
higher than those observed during the periods sues and strategy, and with it the shared commit-
of the strongest growth in the industrial coun- ment to implementing the agreed reforms.
tries, especially when the familiar pillars of sus- Greater ownership by the domestic authorities
tainable growth—macroeconomic stability, and civil society is expected to lead to stronger
sound institutions, and free trade—are pursued. policy implementation. Development partners,
Third, a few low-income countries—in particular the World Bank, and the IMF will provide a
China and to a lesser extent India with their broad range of assistance. The poverty reduction
large populations—have also achieved long peri- strategy, when broadly endorsed by the
ods of sustained growth at a rapid pace. Taken Executive Boards of the IMF and the World
together, the successes of these groups of coun- Bank, will form the basis for concessional lend-
tries show that countries with different charac- ing from both institutions. It is hoped that other
teristics can achieve significant increases in per development partners will also link their finan-
capita income. Finally, and more recently, suc- cial and technical support to these poverty re-
cessful implementation of growth-oriented ad- duction strategies.
justment policies and reform is clearly paying off Second, poverty reduction is now placed at
in some low- and middle-income developing the center of programs supported by the IMF’s
countries that have witnessed falling inflation facility for concessional lending. This follows
and stronger rates of economic growth in the from the recognition that poverty reduction it-
1990s. More needs to be done, of course, to en- self is a primary objective, but also the belief
sure that these improvements endure. that by jump-starting poverty reduction, coun-
Still, while the fundamentals of economic tries can begin a virtuous circle of domestic sav-
growth are reasonably clear, there is no unique ing and investment leading to growth and fur-
formula to achieve it in all countries. Solutions ther poverty reduction. Consistent with the
will vary and depend on country-specific institu- change in objectives and practices, the IMF’s fa-
tions, customs, and economic conditions. cility for concessional lending—formerly the
Experience identifies a few key conditions that, Enhanced Structural Adjustment Facility—has
if lacking or absent, can be an impediment to been renamed the Poverty Reduction and
takeoff and sustained growth. Among these are Growth Facility.
an incentive structure, the rule of law, and a The third element of the strategy is to bring
level of governance that allow individuals to enhanced debt relief to poor countries where
save, invest, and ultimately benefit from these debt levels are unsustainable. Without debt re-
endeavors. The enhancement of public health duction through traditional means and the
and education standards are also important for HIPC Initiative, the incentive problems associ-
sustainable growth, as they not only directly con- ated with unsustainable debt burdens will not be
tribute to human well-being but also allow for an overcome, and policy reforms and private invest-
efficient accumulation of productive human cap- ment increases may not be forthcoming. The

147
 CHAPTER IV HOW CAN THE POOREST COUNTRIES CATCH UP?

cost to debtor countries of any resulting delay none can afford to miss. The HIPC Initiative is
would come in forgone output. The experience no panacea for all poverty and economic prob-
over the past 30 years during which per capita lems in these countries, however, and its goals
income growth stagnated in many of these coun- can only be achieved with continued hard work
tries demonstrates the magnitude of the poten- by domestic and international participants. The
tial loss. For the donor countries, the cost of in- international community needs to increase levels
adequate debt reduction at this stage is the of foreign aid and ensure that aid promotes re-
possibility of more debt problems and further forms and poverty alleviation. Major efforts also
rescheduling down the road that could come at are needed to reform trade policies that ad-
an even greater cost to taxpayers. versely affect the poorest countries. The poorest
countries themselves will need to persevere with
***
macroeconomic and structural reform, with em-
The deepening of debt relief in the enhanced phasis on providing an environment conducive to
HIPC Initiative clearly moves in the right direc- private saving and investment decisions, includ-
tion and adds an extra margin of safety that debt ing better governance, public sector reform, and
burdens will be brought to sustainable levels. The market liberalization. Their responsibility to stay
enhanced HIPC Initiative is an opportunity that the course is a sine qua non part of the strategy.

148