Chapter2 PDF

13
Chapter II
Poverty: the official numbers
Monitoring and reporting on the levels, patterns and trends of poverty have
become a standard part of anti-poverty programme design and assessment.
With the steady internationalization of the poverty agenda, development or-
ganizations, both multilateral and bilateral, have demanded a template for
regular reporting, and new concepts, definitions, data sets and instruments
have been generated to meet this demand. Every major development organiza-
tion produces its own report card, often ranking countries in terms of their
performance. Special interest usually attaches to the annual Human Develop-
ment Reports of the United Nations Development Programme (UNDP) and,
of late, the Millennium Development Goals progress reports; however, it is
perhaps the reports of the World Bank on the incidence of poverty based on
the dollar-a-day criterion that generate the greatest interest and commentary
in the development community. Statistics have an awesome power, and these
global accounting exercises present statistical data to journalists, researchers,
practitioners and activists as irrefutable facts. What, then, are those ostensible
facts? The present chapter provides a summary of the currently most influ-
ential versions, largely associated with the World Bank’s dollar-a-day poverty
estimates.
Global poverty trends: 1981-20051

Poverty is most often measured in monetary terms, captured by levels of in-
come or consumption per capita or per household. The commitment made in
the Millennium Development Goals to eradicate absolute poverty by halving
the number of people living on less than US$ 1.25 dollar a day represents the
most publicized example of an income-focused approach to poverty.
Based on this measure, the last 20 years have seen significant reductions in
the depth and severity of extreme poverty in the developing world.2 In absolute
terms, extreme income poverty has fallen substantially, with the number of
1 The present chapter uses the revised series of country-level poverty data issued by the
World Bank in August 2008 following the findings of the 2005 International Comparison
Program. These data are available on PovcalNet, a web-based interactive research tool
which can be used to replicate Bank poverty estimates and test alternative assumptions
regarding, inter alia, the poverty line or country groupings. Despite many criticisms, the
Bank’s approach remains highly influential, and provides the prevailing benchmark for
discussions of the extent and trends of poverty globally, including in the United Nations
system. Hence, what we think we know continues to rely heavily on the accuracy of the
poverty estimates generated by the Bank.
2 For definitions of terms, see annex II.1.
14 Rethinking Poverty
people living on less than $1.25 a day having declined from a high of 1.9 bil-
lion in 1981 to a low of 1.4 billion in 2005. In relative terms, the proportion
of people living in extreme poverty dropped from 52.0 to 25.7 per cent during
this period (Chen and Ravallion, 2008).3
Notwithstanding the continued growth in the world’s population, the
absolute number of people living in extreme poverty has fallen, regardless of
whether the poverty-line income threshold is set at $1.25 or raised to $2 or
$2.50 per day (figure II.1). This has occurred in the midst of an expanding
global economy, which has resulted, on average, in higher per capita incomes
in both developed and developing countries (Sachs, 2008; United Nations,
2005a). Since the 1960s, gross domestic product (GDP) in low-income coun-
tries has grown at an average of 4.1 per cent per annum, while GDP in middle-
Figure II.1
World population and number of people living in poverty, 1981-2005
7,000
6,000
Population (millions)
5,000
4,000
3,000
2,000
1,000
1981 1984 1990 1996 2005
World population Population living on less than $2.50 a day
Population of less developed regions Population living on less than $2.00 a day
Population living on less than $1.25 a day
Sources: United Nations, Department of Economic and Social Affairs, Population Division; and World
Bank, Development Research Group (2009).
3 For many developing countries, the estimation of current and past poverty levels is a
complex task given the general lack of reliable data. In these countries, information on the
depth, severity and duration of poverty may be limited, and available information may
be unreliable. To address some of these data gaps, the research programme of the World
Bank Poverty and Inequality Group has been engaged in improving current data as well as
methods and tools for poverty and inequality analysis through, inter alia, producing new
household-level data (notably through the Group’s Living Standards Measurement Study),
monitoring poverty and inequality using household-level data, developing more reliable
“poverty maps”, and rolling out computational tools such as ADePT and PovCalNet (see
http://go.worldbank.org/NT2A1XUWP0).
Poverty: the official numbers 15
and high-income countries has grown at an average of 4.2 and 3.2 per cent per
annum, respectively (Soubbotina, 2004).
By 2050, the world’s population is projected to surpass 9 billion, with
developing countries accounting for most of the 2.3 billion increase. The popu-
lation of the developing world is expected to rise from 5.6 billion in 2009 to
7.9 billion in 2050. In contrast, the population of the developed regions is
expected to increase slightly, from 1.23 billion to 1.28 billion (United Nations,
Department of Economics and Social Affairs, Population Division, 2009). The
continued rapid increase in the population of developing countries highlights
the importance of having appropriate policies designed to promote the sus-
tained economic growth and structural transformation of their economies
so as to ensure durable poverty reduction. Although the income-based (per
capita) conventional poverty measure is sensitive to population growth, careful
analysis does not provide any support for the Malthusian claim that poverty
can be attributed to population growth in excess of output growth, especially
food production. Instead, the demographic transitions experienced by a wide
variety of societies suggest that family sizes tend to decline with higher in-
comes and greater economic security. Conversely, poor families tend to have
more children in the hope of increasing contributions to household income as
well as of ensuring continued economic security as parents age (Leibenstein,
1957; Mamdani, 1972; Robbins, 1999).
As can be seen from figure II.2.A, faster rates of decline in the number of
people living on less than $1.25 a day occurred between 1999 and 2005. A sig-
nificant proportion of this decline can be largely attributed to the rise in living
standards in East Asia and the Pacific which accompanied explosive economic
growth, particularly in China. Other regions of the world also experienced
a decline in the incidence of poverty, with the exception of Eastern Europe
and Central Asia, where the proportion of people living on less than $1.25 a
day increased from 1.7 to 3.7 per cent between 1981 and 2005 (figure II.2.B).
While this declining trend in poverty levels is welcome, it is also important to
point out that poverty rates remain unacceptably high in sub-Saharan Africa
and South Asia.
Figure II.3 presents global poverty trends with and without some major
countries and regions, thereby illustrating the role that these countries and
regions play in shaping the global trends. The absolute global poverty level in
2005 was about 1.4 billion; however when China is excluded from the analysis,
poverty increased from 1.1 billion in 1981 to about 1.3 billion in 1999, before
declining to about 1.2 billion in 2005 (figure II.3.A). However, if sub-Saharan
Africa is left out, the number of people living on less than $1.25 a day falls
precipitously, from 1.7 billion in 1981 to 986 million in 2005. Without the
rapidly growing developing economies of Brazil, the Russian Federation, India
and China, the absolute number of people living in extreme poverty actually
went up, from 619 million in 1981 to about 699 million in 2005. However,
in terms of incidence, poverty levels declined across all regions (figure II.3.B).
Figure II.2
Global and regional trends in extreme poverty, 1981-2005
A. Number of people living on less than $1.25 a day

2,000
1,600 548
Number of people (millions)
579
589
1,200
596
800 1,071 873

635
316
400
298 383 388

212
0
1981 1990 1999 2005
Middle East and Latin America East Asia and
Northern Africa and the Caribbean the Pacific
Europe and South Asia Sub-Saharan Africa
Central Asia
B. Proportion of the population living on less than $1.25 a day

90
80
70
60
Percentage
50
40
30
20
10
0
1981 1984 1987 1990 1993 1996 1999 2002 2005
Total Latin America and the Caribbean
East Asia and the Pacific Middle East and Northern Africa
South Asia Europe and Central Asia
Sub-Saharan Africa
Source: World Bank, Development Research Group (2009).

Figure II.3
Poverty trends over time, with and without major countries and regions, 1981-2005
A. Number of people living on less than $1.25 a day

2,000
1,600
1,200
800
400
0
1981 1984 1987 1990 1993 1996 1999 2002 2005
Developing world
Developing world, excluding sub-Saharan Africa
Developing world, excluding China
Developing world, excluding China and India
Developing world, excluding Brazil, China, India and Russian Federation
B. Proportion of the population living on less than $1.25 a day

60
50
Percentage
40
30
20
1981 1984 1987 1990 1993 1996 1999 2002 2005
Developing world
Developing world, excluding sub-Saharan Africa
Developing world, excluding China
Developing world, excluding China and India
Developing world, excluding Brazil, China, India and Russian Federation
With sub-Saharan Africa excluded, the incidence of poverty in the develop-

ing world declined by 31 percentage points between 1981 and 2005, while a
developing world without China saw the incidence of poverty drop from about
40 to 28 per cent during this period.
The dramatic drop in poverty levels has been attributed to, inter alia,
improved rates of economic performance and higher wages, as well as the
provision of social protection systems. Unfortunately, in the context of the
current global economic and financial crisis, which is slowing down rates of
economic growth, faster rates of population growth in developing countries
without commensurate increases in productive employment and with a lack
of comprehensive social protection have exacerbated the declines in real per
capita incomes and have thus contributed to a rise in the number of poor
persons.
The contraction in the global economy has already resulted in massive job
losses, with the global unemployment rate having increased from 5.7 to 6 per
cent between 2007 and 2008 (International Labour Organization, 2009a).
Data on recent global employment trends released by the International Labour
Organization (ILO) also show that based on growth projections, the number
of unemployed persons globally could rise by 20 million in 2009 as a result
of the economic crisis (International Labour Organization, 2009a). It is also
estimated that the number of the unemployed could rise by 50 million if the
global economic outlook worsens to the point where increases in unemploy-
ment match the magnitudes witnessed in the 1990s. This will bring the global
unemployment rate to above 7 per cent. The number of the working poor is
also likely to rise, particularly in emerging and developing countries where
growth was primarily export-led. Besides posing serious challenges with re-
spect to Government social spending and social stability, given the long lag in
employment recovery, this increase will in all likelihood erase progress made
over the last decade in reducing extreme poverty through the generation of
decent work opportunities for all.4
Regional shares and trends

Although global trends are informative, they tend to conceal significant di-
versity in poverty levels across regions. The transformative power of sustained
economic growth combined with urbanization trends has propelled some
countries to middle-income status (particularly those in East Asia) while a
few others are poised to join the high-income status group. For other regions,
the lack of strong and sustained economic growth and persisting income in-
4 Employment recovery generally lags from four to five years behind output recovery. His-
torical evidence culled by Reinhart and Rogoff (2009), based on 14 cases, suggests that it
takes 4.8 years for the unemployment rate to revert to its pre-crisis level, even though GDP
returns to its pre-crisis level in 1.9 years.
equalities have meant that the number of poor people continues to grow. Con-
sequently, the spatial distribution of poor people within and across regions
is changing. Before examining what is happening to levels of poverty across
regions, it is important to determine where poor people live. One way to do
this is to compare the share of poor people living in each region.
As shown in table II.1, the distribution of extremely poor people (defined
as those living on less than $1.25 a day) across developing regions has changed
significantly since 1981 when East Asia and the Pacific had the highest share
of the number of poor people. It is now South Asia and sub-Saharan Africa
that have the highest shares. In 1981, China and other East Asian countries
accounted for 57 per cent of extremely poor people in the world. However,
over a span of less than 25 years, the East Asian and Pacific region managed
to reduce its global share of extremely poor people to about 23 per cent by
2005. In contrast, the share of the world’s extremely poor people increased
in South Asia, from 29 per cent in 1981 to 43 per cent in 2005. The share of
poor people in sub-Saharan Africa more than doubled over the same period,
having gone from 11 per cent in 1981 to 28 per cent in 2005. These changes
are partly accounted for by high rates of population growth in the absence of
strong economic and productive employment growth, as well as by the failure
in both regions to achieve significant structural change.
In other words, the changing picture of the distribution of poverty across
the regions reflects the broad nature of their economic performances. In the
case of East Asia and the Pacific, there is little doubt that strong economic
Table II.1
Regional shares in number of people in the world
living on less than $1.25 a day, 1981-2005 (percentage)
Region 1981 1984 1987 1990 1993 1996 1999 2002 2005
East Asia
and the Pacific 56.50 52.39 47.81 48.16 47.09 37.57 37.44 31.61 22.97
Eastern Europe
and Central Asia 0.37 0.32 0.28 0.50 1.12 1.32 1.43 1.35 1.26
Latin America
and the Caribbean 2.21 2.89 3.04 2.37 2.33 3.15 3.23 3.64 3.35
Middle East
and Northern Africa 0.72 0.64 0.69 0.53 0.55 0.64 0.68 0.64 0.80
South Asia 28.91 30.28 33.09 31.94 31.17 35.89 34.72 38.42 43.26
Sub-Saharan
Africa 11.27 13.48 15.09 16.49 17.74 21.43 22.50 24.33 28.37
Total (per cent) 100 100 100 100 100 100 100 100 100
Total number
of poor (millions) 1 896.2 1 808.2 1 720 1 813.4 1 794.9 1 656.2 1 696.2 1 603.1 1 376.7
growth and structural change, particularly in China, have been important

contributors to the phenomenal decrease in the region’s share of people living
below the $1.25-a-day poverty line. Africa’s recent growth surge, particularly
since 2002 (Economic Commission for Africa, 2008), had offered some hope
of reducing levels of extreme poverty. However, the recent global financial and
economic crisis, which came on the heels of food and energy price hikes, is
set to reverse recent gains (United Nations, 2009b). Moreover, Africa’s recent
growth surge had been driven by commodity exports which did not induce
much structural change. Instead, it reinforced Africa’s narrow export base;
hence, its growth could not be sustained.
Table II.2 reveals that East Asia and the Pacific is the only region that has
already attained the Millennium Development Goal of halving poverty by
2015. Other regions on track to meet this goal are Eastern Europe and Central
Asia, Latin America and the Caribbean, and the Middle East and Northern
Africa. In contrast, halving poverty remains a major challenge in South Asia
and sub-Saharan Africa. These two subregions actually saw a significant in-
crease in the shares of people living on less than $1.25 a day over the period
1990-2005.
Intraregional poverty trends

It is important to remember that all poverty is local. Hence, efforts to reduce
poverty tend to be vigorously waged first and foremost at the country level.
Analysis of intraregional poverty can therefore reveal considerable diversity in
respect of the patterns and trends in poverty rates across countries that help
to spur countries and their development partners to action. The present sec-
tion will therefore provide a detailed description of poverty trends by income
at intraregional levels. Such a focus will help illuminate the discussion and
understanding of poverty patterns and trends that followed the publication of
a World Bank report (Chen and Ravallion, 2008) whose purpose was to show
that “the developing world is poorer than we thought, but no less successful in
the fight against poverty”.
Sub-Saharan Africa
Although the absolute number of people living in extreme poverty has been on
the rise in sub-Saharan Africa, the incidence of poverty fell marginally, from
54 to 51 per cent between 1981 and 2005 (table II.3), after having risen briefly
to 59 per cent in 1996. However, this regional trend disguises large country
differences. For instance, in 1981, the proportion of people living on less than
$1.25 a day had varied from a low of 3.6 per cent in Gabon to a high of 89.9 per
cent in Swaziland. This pattern persisted into 2005, with the proportion in
extreme poverty ranging from a low of 4.8 per cent in Gabon to a high of
86.1 per cent in Liberia. These differences are strongly correlated with differ-
ences in respect of both economic growth and the severity of income inequal-
Table II.2
Progress made in reducing poverty by half at the regional level,
over the period 1990-2005
East Asia Eastern Latin America Middle East Sub-

and Europe and and the and Northern Saharan
the Pacific Central Asia Caribbean Africa South Asia Africa
Percentage living on less than $1.25 a day
2005 16.8 3.7 8.2 3.6 40.3 50.9

1999 35.5 5.1 10.9 4.2 44.1 58.4
1990 54.7 2.0 11.3 4.3 51.7 57.6
2015 target 27.4 1.0 5.7 2.2 25.9 28.8
Change needed to
a
achieve the target –2.7 –2.6 –1.4 –14.5 –22.1
Annual rate of change (percentage)
1990-2005 –7.6 4.2 –2.1 –1.2 –1.6 –0.8

1990-1999 –4.7 11.0 –0.4 –0.2 –1.7 0.2
1999-2005 –11.7 –5.2 –4.6 –2.6 –1.5 –2.3
Rate needed to
achieve target
a
from 2005 level –12.3 –3.7 –5.0 –4.4 –5.5
Percentage point change
1990-1999 –19.2 3.1 –0.4 –0.1 –7.6 0.8

1999-2005 –18.7 –1.4 –2.7 –0.6 –3.8 –7.5

a By 2005, the region had achieved the 2015 target of halving poverty, using 1990 as the baseline.
ity, especially in countries of Southern Africa. For example, the Gini coefficient
was above 50 per cent in all five Southern African countries in 2005, with Na-
mibia registering the world’s highest level of income inequality. Table II.3 also
shows that only seven sub-Saharan African countries had poverty rates below
25 per cent in 2005,5 up from two (Côte d’Ivoire and Gabon) in 1981. In gen-
eral, countries with extremely high poverty levels also trail behind in respect
of a number of demographic and social indicators such as life expectancy at
birth, infant mortality, and children’s school enrolment and completion rates.
Further insights into country-level poverty trends can be gained by ex-
amining progress made in meeting the Millennium Development Goal target
of halving extreme poverty rates by 2015. Among the 19 African countries
with extremely high poverty levels (that is, those where more than half of the
5 Botswana, Cape Verde, Côte d’Ivoire, Gabon, Kenya, Mauritania and South Africa.
Table II.3
Proportion of the population living on less than $1.25 a day
in countries of sub-Saharan Africa, 1981, 1990 and 2005,
and the change needed to reach the 2015 target (percentage)
Annual rate
Proportion Change of change
living on less than needed needed
$1.25 a day Annual rate to achieve to achieve
2015 of change the target the target
Country 1981 1990 2005 target (1990-2005) (2005-2015) (2005-2015)
Countries with the proportion of the poor above 50 per cent in 2005
Liberia 74.5 83.2 86.1 41.6 0.2 –44.5 –7.3

United Republic
of Tanzania 65.4 70.3 82.4 35.2 1.1 –47.2 –8.5
Burundi 85.2 84.5 81.3 42.3 –0.3 –39.0 –6.5
Rwanda 66.5 70.5 74.4 35.2 0.4 –39.2 –7.5
Malawi 85.3 90.5 73.9 45.2 –1.4 –28.7 –4.9
Guinea 86.6 92.6 69.8 46.3 –1.9 –23.5 –4.1
Mozambique 60.8 84.0 68.2 42.0 –1.4 –26.2 –4.8
Madagascar 85.9 68.4 67.8 34.2 –0.1 –33.6 –6.8
Niger 57.4 65.0 65.9 32.5 0.1 –33.4 –7.1
Central African
Republic 79.5 80.8 64.4 40.4 –1.5 –24.0 –4.7
Zambia 53.6 62.8 64.3 31.4 0.2 –32.9 –7.2
Swaziland 89.9 83.7 62.4 41.8 –2.0 –20.6 –4.0
Nigeria 47.2 49.1 62.4 24.5 1.6 –37.9 –9.3
Democratic Republic
of Congo 31.9 80.2 59.2 40.1 –2.0 –19.1 –3.9
Chad 61.1 55.9 58.7 27.9 0.3 –30.8 –7.4
Burkina Faso 73.0 61.9 55.0 30.9 –0.8 –24.1 –5.8
Congo, Republic of 61.9 66.3 54.1 33.2 –1.4 –20.9 –4.9
Uganda 66.2 68.7 51.5 34.3 –1.9 –17.2 –4.1
Mali 81.5 85.2 51.4 42.6 –3.4 –8.8 –1.9
Countries with the proportion of the poor at or below 50 per cent in 2005
Benin 53.5 66.0 50.0 32.9 –1.9 –17.1 –4.2
Sierra Leone 59.4 63.1 49.9 31.6 –1.6 –18.3 –4.6
Comoros 53.3 51.4 46.1 25.7 –0.7 –20.4 –5.8
Namibia 45.4 43.3 43.8 21.6 0.1 –22.2 –7.1
Guinea-Bissau 31.2 41.3 42.5 20.7 0.2 –21.8 –7.2
Angola 63.6 47.3 42.5 23.7 –0.7 –18.8 –5.8
Annual rate
Country 1981 1990 2005 target (1990-2005) (2005-2015) (2005-2015)
Ethiopia 66.2 65.9 39.0 32.9 –3.5 –6.1 –1.7

Togo 35.2 33.8 38.7 16.9 0.9 –21.8 –8.3
Lesotho 44.1 57.7 38.7 28.8 –2.7 –9.9 –3.0
Senegal 68.3 65.8 33.5 32.9 –4.5 –0.6 –0.2
a a
Gambia 64.3 67.9 31.3 33.9 –5.2
Ghana 48.9 50.7 30.0 25.3 –3.5 –4.7 –1.7
Cameroon 43.0 45.7 27.5 22.8 –3.4 –4.7 –1.9
Botswana 35.6 25.6 23.1 12.8 –0.7 –10.3 –5.9
South Africa 34.7 22.1 20.6 11.0 –0.5 –9.6 –6.3
Côte d’Ivoire 6.7 17.3 20.4 8.7 1.1 –11.7 –8.5
Kenya 38.4 35.4 19.7 17.7 –3.9 –2.0 –1.1
Cape Verde 52.3 36.0 18.4 17.9 –4.5 –0.5 –0.3
a a
Mauritania 38.9 45.9 13.4 22.9 –8.2
Gabon 3.6 1.9 4.8 1.0 6.2 –3.8 –14.9
Total 53.7 57.9 51.2 28.8 –0.8 –22.4 –5.5

a By 2005, the country had achieved the 2015 target of halving poverty, using 1990 as the baseline.
national population live on less than $1.25 a day), 12 countries managed to

reduce poverty levels between 1990 and 2005. Although these declines are
encouraging, they still leave a large proportion of the total population living
in extreme poverty, and all 19 countries face major challenges in meeting the
2015 Millennium Development Goal target. For instance, in Liberia, poverty
levels increased between 1990 and 2005; moreover, the estimated extreme
poverty headcount (86 per cent) would need to be cut by half in order for the
2015 target (41.6) to be reached; that is, Liberia will have to reduce poverty
at an annual rate of 7 per cent from now to 2015 to achieve the target. Huge
challenges face all the sub-Saharan African countries that have extremely high
levels of poverty, including Africa’s most populous country, Nigeria, which
saw its level of poverty increase by 20 per cent between 1990 and 1999 before
recording a modest 7 per cent decrease between 1999 and 2005. Furthermore,
income inequality has grown in the largest countries of sub-Saharan Africa,
including South Africa, the Democratic Republic of the Congo, the United
Republic of Tanzania and Kenya.
Among countries with lower levels of poverty (at or below 50 per cent),
only the Gambia and Mauritania succeeded in halving the 1990 levels of pov-
erty by 2005. On the other hand, a few other countries, Cape Verde, Senegal
and Kenya, seem to be on track to meet the Millennium Development Goal
target by 2015: they need to reduce the poverty levels that prevailed in 2005
by less than 2 per cent in order to meet the 2015 target. Overall, however, the
goal of eradicating extreme poverty will continue to be a major challenge in a
large number of African countries and huge efforts will continue to be required
to halve extreme poverty by 2015.
East Asia and the Pacific

The East Asian and Pacific region has recorded some of the fastest declines in
levels of extreme poverty in the world. The proportion of people with income
levels below the $1.25 poverty line declined from 67 per cent in 1981 to about
9 per cent in 2005 (table II.4), translating as the movement of more than
755 million people out of extreme poverty in about 25 years. A decline of this
magnitude in less than a quarter-century is a historic first. However, although
this regional picture is outstanding, it is also equally important to acknowledge
huge intraregional differences in levels of absolute poverty. As table II.4 shows,
all East Asian and Pacific countries recorded significant declines in poverty
with the exception of Papua New Guinea, where poverty remained largely
unchanged between 1981 and 2005.
As already mentioned, the most impressive reductions occurred in China.
Since 1979, China’s economy has experienced high and sustained growth fol-
lowing successful land and economic reforms. Over the last 25 years, China
successfully moved from agriculture to manufacturing activities and saw an
annual trend of real gross domestic product (GDP) growth of just under 10 per
cent (Ghosh, 2008a). China’s manufacturing sector doubled its share of the
workforce and tripled its share of output. The rate of extreme poverty at the
national level declined from a high of 84 per cent in 1981 to a low of 16 per
cent in 2005. The decline occurred in both rural and urban areas: the rural
poverty rate fell from 94 to 26 per cent and the urban poverty rate fell from
45 to less than 2 per cent during this period.
Although China’s poverty eradication efforts overshadow in their scale the
progress made by other countries in the region, a number of other countries
with large populations have also successfully moved millions of people out
of poverty—namely, Indonesia, the Philippines, Thailand, Malaysia and Viet
Nam. They have been able to capitalize effectively on the opportunities offered
by strong growth to significantly reduce extreme poverty. Indeed, the big three
success stories—China, Indonesia and Viet Nam—together account for an
absolute drop in the numbers of the extreme poor of 718 million. Without the
progress of these countries, the global poverty record would have appeared far
less successful.
Despite this tremendous effort, the region is still home to about 316 mil-
lion people living on less than $1.25 day. Compared with those of neighbour-
ing countries, poverty rates remain very high in Cambodia (40 per cent),
Timor-Leste (44 per cent) and the Lao People’s Democratic Republic (36 per
cent). Poverty has remained high in these countries partly because economic
growth rates have been comparatively low (Economic and Social Commission
for Asia and the Pacific, 2007).
Yet, although progress has been uneven across the region, almost all coun-
tries are on track to meet the 2015 target. The prospects for halving poverty
remain a major challenge, however, for Papua New Guinea, the Philippines and
Timor-Leste. In the Philippines, for instance, even if the incidence of poverty
declined from 30 per cent in 1990 to 23 per cent in 2005, at an annual rate of
about 2 per cent, the country will need to reduce the number of people living in
extreme poverty at an annual rate of 4 per cent in order to reach its 2015 target.
Table II.4
in countries of East Asia and the Pacific, 1981, 1990 and 2005,
Annual rate
living on less than needed to needed to
$1.25 a day Annual rate achieve achieve
Country 1981 1990 2005 target (1990-2005) (2005-2015) (2005-2015)
Timor-Leste 82.1 71.3 43.6 35.7 –3.3 –7.9 –2.0

Cambodia 86.1 77.3 40.2 38.7 –4.4 –1.5 –0.4
Lao People's
Democratic Republic 88.0 65.9 35.7 32.9 –4.1 –2.8 –0.8
Papua New Guinea 29.5 43.0 29.7 21.5 –2.5 –8.2 –3.2
a a
Indonesia (rural) 73.8 57.1 24.0 28.6 –5.8
a a
Indonesia (urban) 63.8 47.8 18.7 23.9 –6.3
Viet Nam 90.4 34.2 22.8 17.1 –2.7 –5.7 –2.9
Philippines 31.4 29.7 22.6 14.8 –1.8 –7.8 –4.2
Mongolia 62.4 34.9 22.4 17.5 –3.0 –4.9 –2.5
a a
China 84.0 60.2 15.9 30.1 –8.9
a a
China (rural) 94.1 74.1 26.1 37.0 –6.7
a a
China (urban) 44.5 23.4 1.7 11.7 –16.0
a a
Malaysia 3.8 1.9 0.5 1.0 –8.9
a a
Thailand 21.9 9.4 0.4 4.7 –21.0
Total 66.8 39.1 9.3 24.4 –7.6

South Asia
South Asia is the developing subregion with the largest number of poor peo-
ple: 43 per cent of the developing world’s 1.4 billion poor people live in South
Asian countries. The absolute number of people living in extreme poverty in-
creased from 548.3 million to 595.6 million between 1981 and 2005. Rates
of population growth in these countries have remained high and have led to
an enlargement of both the total population as well as the numbers living in
extreme poverty. In recent years, economic growth has been relatively high
in the three largest countries in the region, India, Bangladesh and Pakistan,
which recorded annual rates of growth of GDP per capita above 5 per cent in
2000-2006.6 As a result, the subregion saw the proportion of those living in
extreme poverty decline in relative terms, from a high of 59 per cent in 1981 to
40 per cent in 2005 (table II.5). However, such growth has not been sufficiently
inclusive and pro-poor to reduce the absolute number of persons living in pov-
erty. Income inequalities have grown steadily in India since the early 1980s, in
both urban and rural areas. The same pattern can be observed in Bangladesh.
South Asian countries have been unable to generate sufficient decent work op-
portunities to lift working poor people out of poverty. The structural change of
the subcontinental economies has also been slow; for example, manufacturing
accounts for about 17 per cent of GDP in Bangladesh and for about 28 per
cent in India and Pakistan, as opposed to close to 35 per cent in China (World
Bank, 2008c).
The headcount index declined in almost all countries with data on income
poverty, with the exception of Bangladesh, where the estimated proportion of
people living below the $1.25 a day poverty line increased from 44 per cent in
1981 to 51 per cent in 2005. In India alone, the poverty headcount fell by 18
percentage points, from 60 per cent in 1981 to 42 per cent in 2005. Pakistan
also experienced a decline in the headcount index from 73 to 23 per cent dur-
ing the same period.7 Yet, table II.5 shows that, in terms of progress in meeting
the Millennium Development Goal target of halving extreme poverty by 2015,
6 Growth rates calculated based on World Bank data of GDP per capita in purchasing power
parities (PPPs), as shown in the United Nations Key Global Indicators database (http://
data.un.org) (accessed 8 June 2009).
7 The sharp decline in the poverty headcount in Pakistan potentially highlights the prob-
lems with data on and measurement of poverty. The ups and downs reflected in the poverty
data collected during the 1990s are questionable. It is possible that the data from this
period are not comparable. The variation could also be a result of sensitivity in respect
of the poverty lines. Alternative sources of information report different poverty trends;
for instance, the latest Human Development Report for Pakistan reports an increase in
poverty during the 1990s, while a report by the Asian Development Bank cites several
studies that showed a trend for the 1990s that was the reverse of the one reported by the
World Bank (see http://hdr.undp.org/en/reports/nationalreports/asiathepacific/pakistan/
and http://www.adb.org/documents/reports/poverty_pak/chapter_2.pdf). Data prob-
lems may also be responsible for a reported rise in poverty in Bangladesh, which appears
counter-intuitive, given the rise in GDP per capita.
Table II.5
in countries of South Asia, 1981, 1990 and 2005,
Annual rate
Country 1981 1990 2005 target (1990-2005) (2005-2015) (2005-2015)
Nepal — 77 54.7 38.5 –2.3 –16.2 –3.5

Bangladesh 44.2 49.9 50.5 24.9 0.1 –25.6 –7.1
India 59.8 51.3 41.6 25.7 –1.4 –15.9 –4.8
India (rural) 62.5 53.9 43.8 27.0 –1.4 –16.9 –4.7
India (urban) 51 43.5 36.2 21.8 –1.2 –14.4 –5.0
Bhutan 47.4 51 26.8 25.5 –4.3 –1.3 –0.5
a a
Pakistan 72.9 58.5 22.6 29.3 –6.3
Sri Lanka 31 15 10.3 7.5 –2.5 –2.8 –3.2
Total 59.4 51.7 40.3 25.9 –1.6 –14.5 –4.4

several countries in the region, including Bangladesh, India, Nepal and Sri
Lanka, will need higher rates of poverty reduction than recorded since 1990 if
they are to meet the 2015 target.
Latin America and the Caribbean

Over the last 25 years, Latin America and the Caribbean has had mixed results
in eradicating poverty. While poverty declined in most countries, levels of
poverty went up in the Plurinational State of Bolivia, Guyana, Haiti, Panama,
Paraguay, Peru and the Bolivarian Republic of Venezuela and in urban areas
of Argentina8 (table II.6). These disparities are to a large extent a reflection of
the huge inequalities in the distribution of income across the region and within
countries. For example, 12 out of 23 countries in the world with Gini coeffi-
cients above 50 per cent in 2005 were in Latin America. It is estimated that, in
Latin American and Caribbean countries, the per capita income of households
in the tenth decile is about 17 times greater than that of the poorest 40 per cent
of households (Economic Commission for Latin America and the Caribbean,
2008). Nonetheless, poverty levels have declined at the regional level owing to
strong per capita GDP growth, averaging over 3 per cent per annum between
8 Poverty data are not available for rural areas of Argentina.

Table II.6
in countries of Latin America and the Caribbean, 1981, 1990 and 2005,
Annual rate
Country 1981 1990 2005 target (1990-2005) (2005-2015) (2005-2015)
Haiti 54.6 56.8 58.0 28.4 0.1 –29.6 –7.1

Honduras — 43.5 22.2 21.8 –4.5 –0.4 –0.2
Bolivia (Plurinational
State of) 2.0 4.0 19.6 2.0 10.6 –17.6 –22.8
Saint Lucia 24.0 26.8 17.8 13.4 –2.7 –4.4 –2.8
a a
Nicaragua 21.0 39.5 15.8 19.8 –6.1
Suriname 17.2 18.6 14.2 9.3 –1.8 –4.9 –4.2
El Salvador 14.8 15.9 13.5 8.0 –1.1 –5.6 –5.3
a a
Guatemala 46.7 37.2 12.1 18.6 –7.5
Venezuela (Bolivarian
Republic of) 6.2 3.1 10.0 1.6 7.8 –8.5 –18.6
Ecuador 11.1 14.2 9.8 7.1 –2.5 –2.7 –3.2
Paraguay 4.8 5.9 9.3 3.0 3.0 –6.4 –11.5
Panama 6.0 15.8 9.2 7.9 –3.6 –1.3 –1.5
Peru 1.0 1.3 8.2 0.7 12.3 –7.6 –25.3
Brazil 17.1 15.5 7.8 7.8 –4.6 0.0 0.0
Guyana 3.1 8.4 7.3 4.2 –0.9 –3.1 –5.5
a a
Dominican Republic 16.6 14.9 5.0 7.5 –7.3
a a
Argentina (urban) 0.0 0.4 4.5 0.2 16.1
a a
Costa Rica 21.4 9.2 2.4 4.6 –9.0
a a
Mexico 9.8 5.4 1.7 2.7 –7.7
a a
Chile 6.3 4.4 0.7 2.2 –12.3
a a
Trinidad and Tobago 0.0 4.4 0.5 2.2 –14.5
Jamaica 5.6 0.2 0.2 0.1 0.0 –0.1 –6.9
Total 11.5 9.8 8.4 5.7 –2.1 –2.6 –3.7

2003 and 2007. Such growth, the highest the region has experienced since
the 1970s, resulted in an increase in the average labour income of the poorest
(Economic Commission for Latin America and the Caribbean, 2009).
This trend in poverty is consistent with the findings of the Economic

Commission for Latin American and the Caribbean (ECLAC), based on a
series of household surveys in 18 countries.9 The most recent figures from these
surveys show that poverty has continued on a downward trend. According to
ECLAC, about 34 per cent of the population of Latin America and the Car-
ibbean were living in poverty in 2007, among whom 13 per cent were living
in extreme poverty. In absolute numbers, 184 million were considered poor,
among whom 68 million were living in extreme poverty (Economic Commis-
sion for Latin America and the Caribbean, 2008).
According to World Bank estimates, in 2005, most countries in the region
were on track to halve poverty rates by 2015. This was accounted for largely
by the fact that poverty levels had been very low in 1990, which is the base
year for measuring progress made by countries towards the achievement of
the Millennium Development Goals. However, for a number of countries, the
target remains a major challenge. For Haiti to halve its poverty rate by 2015, it
will have to reduce poverty levels at an annual rate of 7.1 per cent from 2005
onward. The country’s poverty levels have practically remained unchanged
since 1981. For the Plurinational State of Bolivia and the Bolivarian Republic
of Venezuela, two countries that saw sharp increases in the incidence of pov-
erty between 1990 and 2005 and increases in income disparities during the
same period that were among the world’s highest, much higher annual rates
of poverty reduction are required even though their poverty levels are much
lower than that of Haiti.
Middle East and Northern Africa

The Middle East and Northern Africa region has managed to reduce both the
incidence of poverty and the absolute number of people living in extreme pov-
erty despite poor economic performance in the last two decades. The incidence
of poverty in the region is the lowest in the developing world. It dropped from
7.9 per cent in 1981 to 3.6 per cent in 2005 (table II.7). In absolute terms, the
number of poor people has declined from 13.7 million to 11 million. Unlike
other middle-income countries, the countries of the Middle East and Northern
Africa have been very successful in reducing extreme poverty owing in part
to improvements in the health and education levels of the general population,
as well as to the availability of extensive food and energy subsidies in several
countries. For example, between 1980 and 2000, the regional child mortality
rate plunged from 138 per thousand live births to 47 per thousand, the average
years of schooling per person over age 15 rose from 2.6 to 5.5 years, and life
expectancy at birth increased by 10 years, from 58 to 68 years (Iqbal, 2006).
Gains of this magnitude within the social dimensions of development which
9 The findings of these surveys are reported on a regular basis in the issues of the Social
Panorama of Latin America, published by the Economic Commission for Latin America
and the Caribbean.
enhance human capabilities are known to contribute to reductions in poverty

at the household level even when per capita incomes stagnate.
While poverty rates are low at the regional level, poverty levels and trends
differ across countries. For example, table II.7 shows that, while 12 per cent
of Egypt’s population had lived below the $1.25 poverty line in 1981, the
incidence of poverty dropped to 2 per cent in 2005. In contrast, poverty rates
increased sharply in Djibouti (from 6.1 to 18.6 per cent) and Yemen (from 9.1
to 17.5 per cent) over the same period, giving them the highest poverty rates
in the region. Poverty levels are lower in oil-rich Gulf countries which are
able to use the vast wealth derived from oil and gas to subsidize consumption
goods and social services for their citizens (Iqbal, 2006). Net oil importers in
the Middle East and Northern Africa region were impacted negatively by the
recent increases in the prices of energy and food. This created fiscal burdens
for Governments, increased production costs for small businesses, and reduced
the food intake of poor families.
Despite the decline in poverty at the regional level, fighting poverty is still
a major concern for many countries in the region. Although per capita income
is high as a result of high prices of oil and gas, not all segments of society have
benefited. Egypt, the Islamic Republic of Iran, Jordan and Tunisia have al-
Table II.7
in countries of the Middle East and Northern Africa, 1981, 1990 and 2005,
Annual rate
Country 1981 1990 2005 target (1990-2005) (2005-2015) (2005-2015)
Djibouti 6.1 1.8 18.6 0.9 15.6 –17.7 –30.3

Yemen 9.1 4.9 17.5 2.5 8.5 –15.1 –19.7
Algeria 3.8 6.2 4.3 3.1 –2.4 –1.2 –3.3
Morocco 10.4 2.5 3.0 1.3 1.2 –1.8 –8.8
a a
Egypt 12.0 4.5 2.0 2.3 –5.4
Islamic Republic
a a
of Iran 4.0 3.9 1.5 2.0 –6.4
a a
Tunisia 9.7 5.9 1.0 3.0 –11.8
a a
Jordan 0.0 2.8 0.4 1.4 –13.0
Total 7.9 4.3 3.6 2.2 –1.2 –1.4 –5.0

ready cut by more than half the poverty rates that prevailed in 1990, although
some increases did occur in Egypt and the Islamic Republic of Iran between
1999 and 2005. According to the World Bank (2004), Egypt and the Islamic
Republic of Iran managed to lower levels of poverty by considerably reducing
poverty in urban areas and, in the case of the latter, sharp declines in urban
poverty managed to offset an increase in rural poverty.
Eastern Europe and Central Asia

Given the very low levels of absolute poverty in Eastern Europe, it is more illu-
minating to use an absolute poverty line of $2 a day, considering the cost of the
heating and warm clothing that are required in this region (Alam and others,
2005). According to this measure, there had been 21.7 million poor people in
Eastern Europe in 1981. This figure dropped to 5 million in 2005, attesting to
a significant decline in poverty in the region. On the other hand, in Central
Asia, the number of poor according to the $2-a-day measure increased from
13.2 million in 1981 to 36.1 million in 2005. However, unlike Eastern Europe,
Central Asia shares many of the characteristics of other developing regions
and hence $1.25 a day perhaps represents a more appropriate poverty line for
Central Asia. Application of this measure shows that the number in absolute
poverty in Central Asia increased more than 4 times, from 3.7 million in 1981
to 16.1 million in 2005. The difference between the poverty levels in Central
Asia according to the two poverty lines ($2 a day and $1.25 a day)—20 million
people in 2005—can be taken as a rough measure of vulnerability.
Another feature of the region is the widening of intercountry disparities in
extreme poverty (table II.8). The highest levels of absolute poverty are found in
Central Asian countries such as Georgia, Kyrgyzstan, Tajikistan, Turkmeni-
stan and Uzbekistan, countries whose economies were centrally planned dur-
ing the Soviet period and hence which shared many structural similarities.
Differences in levels of absolute poverty in these countries were considerably
lower in the early 1980s largely because of the significant resource transfers,
including guaranteed employment, subsidies and social safety nets, that these
countries received from the central budget. In contrast, the Eastern European
countries, although also under the Soviet influence, were primarily responsible
for meeting their own budgetary needs. The fact that, in a large majority of
those countries, extreme absolute poverty was very low during the period 1981-
2005 has been attributed to the growth in wages and employment opportuni-
ties as well as adequate social transfers (Alam and others, 2005). In contrast,
levels of absolute poverty actually increased in a number of Central Asian
countries. Consequently, these countries face a substantially bigger challenge
in respect of halving poverty by 2015.
In large measure, this sharp increase in absolute poverty is associated
with the collapse of the Soviet Union and the transition to market economies
in the early 1990s, which witnessed large declines in real output and high
Table II.8
in countries of Eastern Europe and Central Asia, 1981, 1990 and 2005,
Annual rate
Country 1981 1990 2005 target (1990-2005) (2005-2015) (2005-2015)
Uzbekistan 0.0 4.9 38.8 2.5 13.8 –36.4 –27.6

Kyrgyzstan 0.0 4.8 21.8 2.4 10.1 –19.4 –22.1
Tajikistan 1.4 1.5 21.5 0.8 17.8 –20.8 –33.6
Georgia 2.5 2.9 13.4 1.5 10.2 –12.0 –22.2
a a
Turkmenistan 21.9 34.2 11.7 17.1 –7.2
Republic
of Moldova 20.2 15.2 8.1 7.6 –4.2 –0.5 –0.6
Armenia 0.9 6.3 4.7 3.2 –2.0 –1.6 –4.0
Turkey 4.5 1.5 2.7 0.8 3.9 –2.0 –12.8
Kazakhstan 0.0 0.5 1.2 0.3 5.8 –1.0 –15.7
Albania 0.2 0.9 0.9 0.5 0.0 –0.5 –6.9
Romania 0.0 0.0 0.8 0.0 –0.8 –0.1
Lithuania 0.0 0.0 0.4 0.0 –0.4 –0.1
The former
Yugoslav Republic
of Macedonia 0.0 0.0 0.3 0.0 –0.3 0.0
Bosnia and
Herzegovina 0.0 0.0 0.2 0.0 –0.2 0.0
a a
Russian Federation 0.7 1.4 0.2 0.7 –13.0
a a
Poland 0.0 1.3 0.1 0.7 –17.1
a a
Ukraine 3.0 1.2 0.1 0.6 –16.6
a a
Azerbaijan 13.4 16.1 0.0 8.1
a a
Belarus 0.0 0.0 0.0 0.0
a a
Bulgaria 0.0 0.0 0.0 0.0
a a
Croatia 0.0 0.0 0.0 0.0
a a
Czech Republic 0.0 0.0 0.0 0.0
a a
Estonia 0.0 0.0 0.0 0.0
a a
Hungary 0.0 0.0 0.0 0.0
a a
Latvia 0.0 0.0 0.0 0.0
a a
Slovakia 0.0 0.0 0.0 0.0
Annual rate
Country 1981 1990 2005 target (1990-2005) (2005-2015) (2005-2015)
a a
Slovenia 0.0 0.0 0.0 0.0
Total 1.7 2.0 3.7 1.0 4.2 –2.7 –12.3

inflation. During the transition, the political, economic and social institu-
tions in these formerly centrally planned economies underwent major changes
which affected the distribution of public and private resources, both across
and within countries. In particular, public social services crumbled in most
of these countries (Bandara, Malik and Gherman, 2004), which contributed
to the rise in poverty and inequalities as well as greater regional disparities,
especially between rural and urban areas (Cukrowski, 2006; Anderson and
Pomfret, 2004). On average, within-country income inequalities rose faster
in this region than in any other between the early 1980s and the late 1990s.
In particular, successor republics of the former Soviet Union such as Turk-
menistan, the Russian Federation and Uzbekistan, as well as the three Baltic
States, saw their Gini coefficient increase by more than 10 points between
1981 and 1999.
Least developed countries

Further insight and nuance into poverty patterns and trends around the world
can be gleaned from an examination of the situation in the least developed
countries. This group of countries10 is home to 750 million people, or 12 per
cent of the world’s population. It is claimed that the economic growth pros-
pects of these countries have been undermined by their geography, with 28 of
them being landlocked or small island States (Gallup, Sachs and Mellinger,
1998; Collier, 2007). Controlling for economic policies and institutions, some
researchers contend that the location and climate of the continent of Africa
have had a negative impact on its income levels and growth. In particular, they
note that these geographical factors affect growth through their impact on ag-
ricultural productivity, transport costs and a debilitating disease burden. In ad-
dition, landlocked countries also tend to be held hostages by their neighbours
if the latter have poor transport links to the coast; consequently, landlocked
countries find it more difficult to reap the benefits of globalization inasmuch
as they are hamstrung in their ability to export commodities or any manufac-
10 The current list of the least developed countries comprises 49 countries: 33 in Africa, 15 in
Asia and the Pacific, and 1 in Latin America.
tured products (Collier, 2007). Countries with access to the sea therefore tend
to have higher incomes than their landlocked counterparts because they have
better and cheaper access to global markets.
Although 38 per cent of the people in “bottom billion societies”11 live in
countries that have no sea access, serious research has questioned the find-
ings of studies that give prominence to geography and climate. For example,
Nordhous and Chen (2009) found that a substantial part of the “latitude ef-
fect” (distance from the equator) does not reflect geophysical variables such as
climate, elevation, distance from coastlines and rivers, and similar factors: vari-
ables other than purely geographical ones are responsible for much of the poor
economic performance of low-latitude regions. In this regard, an International
Monetary Fund (IMF) study (Hernández-Catá, 2000) has raised questions
about the methodological soundness of the influential study by Bloom and
Sachs (1998) which relied only on cross-country data and hence may have
picked up fixed effects specific to Africa other than those related to geogra-
phy or climate. According to Hernández-Catá, if landlockedness had been a
growth-inhibiting factor, then the economies of Switzerland and the Czech
Republic would have been given a very low probability of success starting from
the seventeenth century. Even in Africa, landlocked Bostwana grew impres-
sively in the 1990s, and a tropical climate has not hampered growth in Thai-
land, Malaysia and Indonesia and several southern states of the United States
of America. While recognizing the disadvantages of geography faced by many
African countries, then World Bank economist Benno Ndulu (2006, pp. 215-
216) has made the following point:
The most important message I am trying to convey in this paper is that
offsetting natural or geographical disadvantages is a choice for which pub-
lic action is important. Malaria can be eradicated, and it was in many areas
where it was once preponderant. Fragmentation can be overcome through
integration and deliberate effort to offset its negative effects. For exam-
ple, Tanzania was able to overcome the potential of high ethnolinguistic
fractionalization through a deliberate effort to create national unity and a
single language among 132 tribes. Remoteness, likewise, can be overcome
and distance can be bridged through improvements in infrastructure.
Botswana’s experience perhaps best embodies all aspects of this message.
In other words, what matters for sustained economic growth and poverty re-
duction is the nature of public policy and action.
Figure II.4 shows the absolute number of poor people and the incidence of
poverty in heavily indebted poor countries and the least developed countries.
In absolute terms, the number of people living on less than $1.25 a day in both
11 These are the people living in failing States (a group of about 50) who are dropping fur-
ther and further behind the majority of the world’s people, often falling into an absolute
decline in living standards (Collier, 2007).
Figure II.4
Poverty in heavily indebted poor countries and least developed countries, 1981-2005
400 80
350 70
300 60
250 50
Percentage
200 40
150 30
100 20
50 10
0 0
1981 1984 1987 1990 1993 1996 1999 2002 2005
Left axis data:
Heavily indebted poor countries
Least developed countries
Least developed countries (Africa)
Least developed countries (Asia and the Pacific)
Right axis data:
Heavily indebted poor countries (percentage)
Least developed countries (percentage)
Least developed countries (Africa) (percentage)
Least developed countries (Asia and the Pacific) (percentage)
groups of countries increased between 1981 and 2002, and declined, slowly,
only between 2002 and 2005, the number being larger in the least developed
countries of Africa than in the least developed countries of Asia and the Pacific.
However, in relative terms, the proportion of people living in extreme poverty
has been declining since 1990 in both the heavily indebted poor countries
(HIPC) and in the least developed countries.
Figure II.5 shows that, among all landlocked developing countries and
small island developing States, African landlocked developing countries have
the highest proportion of people living in extreme poverty, followed by Asian
landlocked developing countries. However, the numbers of people living in
extreme poverty have been on the decline since the early 1990s.
In terms of overall poverty reduction efforts, all least developed countries
face a major hurdle (table II.9). In order for all least developed countries to
reduce the 1990 poverty headcount of 67.9 per cent to the 2015 target of
33.9 per cent, they will have to significantly accelerate the pace of poverty re-
duction efforts. Starting from 2005, least developed countries will have had to
Figure II.5
Poverty in landlocked developing countries and small island
developing States, 1981-2005
160 80
140 70
120 60
100 50
Percentage
80 40
60 30
40 20
20 10
0 0
1981 1984 1987 1990 1993 1996 1999 2002 2005
Left axis data:
Landlocked developing countries
Landlocked developing countries (Africa)
Landlocked developing countries (Asia)
Small island developing States
Right axis data:
Landlocked developing countries (percentage)
Landlocked developing countries (Africa) (percentage)
Landlocked developing countries (Asia) (percentage)
Small island developing States (percentage)
maintain an annual reduction rate of 4.7 per cent to achieve the 2015 target.
Small island developing States will also have to redouble their efforts, given
the slackened pace of poverty reduction they experienced between 1999 and
2005, which was considerably slower than that experienced between 1990
and 1999. Given the trends as shown in table II.9, the least developed coun-
tries, landlocked developing countries, highly indebted poor countries and
small island developing States will not meet the Millennium Development
Goal target of halving poverty by 2015.
Countries of the European Union (EU) and other countries of

the Organization for Economic Cooperation and Development (OECD)
The World Bank does not track progress on reducing poverty in developed
countries, even though poverty is a major public policy concern in those coun-
tries. European Union (EU) members and the United States of America have a
long-standing tradition and practice of collecting national poverty data as well
as releasing official poverty estimates. The definition of poverty used for public
Table II.9
in least developed countries, landlocked developing countries, heavily indebted
poor countries and small island developing States, 1990, 1999 and 2005,
Annual rate
Country 1990 1999 2005 target (1990-2005) (2005-2015) (2005-2015)
All least developed

countries 67.9 62.3 54.3 33.9 –1.5 –20.4 –4.7
African least
developed countries 70.2 65.9 56.9 35.1 –1.4 –21.8 –4.8
Asia and the Pacific
least developed
countries 56.6 42.7 38.5 28.3 –2.6 –10.2 –3.1
Heavily indebted
poor countries 63.8 57.5 48.9 31.9 –1.8 –17.0 –4.3
All landlocked
developing countries 49.1 50.7 42.8 24.5 –0.9 –18.3 –5.6
African landlocked
developing countries 69 63.1 52.7 34.5 –1.8 –18.2 –4.2
Asian landlocked
developing countries 27.7 36.6 33.1 13.9 1.2 –19.2 –8.7
Small islands
developing States 32.4 27.7 27.5 16.2 –1.1 –11.3 –5.3
policy purposes and in public discourse is quite different from that used by the
Bank. The prime concern is the standard of living relative to other people in
the country; hence, poverty is a relative concept in the developed world. In the
present section, therefore, the poverty estimates used are not comparable to
those for developing countries as published by the World Bank.
Over the long run, there have been modest changes in overall poverty
indicators in EU and other countries of the Organization for Economic Co-
operation and Development (OECD), with levels of poverty growing in the
recent past. For instance, with the poverty threshold defined as 60 per cent
of a country’s median income, in 2006, 72 million people in the EU were
at risk of falling into poverty; and in 2001, more than half of all people in
low-income households in the EU lived with the persistent risk of falling into
poverty. In addition, it is estimated that one in five people in Europe lives in
substandard housing and 10 per cent live in households where no one works
(Commission of the European Communities, 2007).
Differences in poverty rates across Europe and North America are gener-
ally small. During the 1990s, poverty rates were highest in the United States,
the United Kingdom of Great Britain and Northern Ireland, Ireland, Italy
and Greece. As regards more recent trends, figure II.6 provides relative pov-
erty estimates for various OECD countries based on the 40, 50 and 60 per
cent median household disposable income levels. The graph shows large dis-
parities across countries in the share of people with incomes less than 40,
Figure II.6
Relative poverty rates for different income thresholds in OECD countries, mid-2000s
Denmark
Sweden
Czech Republic
Austria
Norway
France
Iceland
Hungary
Finland
Netherlands
Luxembourg
Slovakia
United Kingdom
Switzerland
Belgium
OECD-30
New Zealand*
Germany
Italy
Canada
Australia
Greece
Portugal
Spain
Poland
Republic of Korea
Ireland
Japan
United States
Turkey
Mexico
0 10 20 30 40 50 60
Percentage
Poverty rates:
40 per cent 50 per cent 60 per cent
Source: OECD (2008a), computations from an OECD income distribution questionnaire.

Note: Poverty rates are defined as the share of individuals with equivalized disposable income that is
less than 40, 50 and 60 per cent of the median for the entire population. Countries are ranked, from top
to bottom, in increasing order of income poverty rates at the 50 per cent median threshold. The income
concept used is that of household disposable income adjusted for household size.
* Poverty rates based on a 40 per cent threshold are not available for New Zealand.
50 or 60 per cent of the median income for the entire population. Relative
poverty rates are lowest in Denmark, Sweden and the Czech Republic and
highest in Mexico, the United States and Turkey. Cross-country differences
in the mid-2000s range between 5 and 18 per cent when the income thresh-
old is set at 50 per cent of the median, and between 11 and 25 per cent when
the threshold is set at 60 per cent of the median (Organization for Economic
Cooperation and Development, 2008a).
Although levels of poverty are generally low in OECD countries, the
structure of poverty has shifted and has led to higher poverty risks among cer-
tain groups, particularly unemployed single parents and younger age groups
(Förster, 2004). Poverty among young adults and families with children in-
creased over the past 20 years as well. By 2005, the poverty rate for children
and that for young adults were about 25 per cent above the total average,
whereas they were close to and below that average, respectively, in 1985. Pov-
erty rates are three times higher than the average among households with
children; for single-parent families, they exceed 40 per cent in one third of
OECD countries. In contrast, poverty among older persons has fallen (Or-
ganization for Economic Cooperation and Development, 2008a; European
Commission, Directorate-General for Employment, Social Affairs and Equal
Opportunities, 2008).
Some of the cross-country differences in the levels of poverty are ac-
counted for by the nature and extent of public transfers and direct taxes that
are aimed at reducing income poverty. The recent report Growing Unequal
Income Distribution and Poverty in OECD Countries (Organization for Eco-
nomic Cooperation and Development, 2008a) points to growing inequality
and poverty in two thirds of OECD countries and finds that the economic
growth of recent decades has largely benefited the rich more than the poor.
Across OECD countries, the income of the richest 10 per cent of people is
nearly nine times that of the poorest 10 per cent. In Mexico, Turkey, Portugal,
Poland and the United States, the income gap is well above the OECD aver-
age. In Canada, Finland, Germany, Italy, Norway and the United States, the
gap has also increased between the rich and the middle-class. The report also
notes that countries with a wide distribution of income tend to have more
widespread income poverty. It points out that the rise in cash-income poverty
from the mid-1980s to the mid-1990s had been offset by increased government
redistribution through public expenditures; however, between the mid-1990s
and the mid-2000s, the redistributive effect of transfers and taxes slackened,
leading to higher poverty rates based on disposable incomes.
Diversity within countries: poverty in China and India

The above discussion has focused on poverty trends at the global, regional
and intraregional levels. Although such a focus is important, it is equally im-
portant to point out that spatial and inter-group disparities in poverty exist
at the country level. For instance, the poor tend to be heavily concentrated
in rural areas as well as in areas with limited access to public assets such as
roads, schools and hospitals. In most countries, welfare disparities are re-
flected in the persistent gaps in living standards between rural and urban
areas. Understanding and addressing these intracountry regional disparities
are important in many ways. First, while some of these countries are on track
to meet the first target of the Millennium Development Goals, namely to
halve the proportion of people living on less than $1.25 a day at the national
level, rural and remote areas, with their significant levels of extreme poverty,
may still be far behind. Left unchecked, these disparities will likely worsen
horizontal inequalities, that is, inequalities in respect of economic and politi-
cal resources that exist among ethnic or religious groups (Brown, Stewart and
Langer, 2007; Stewart, 2002).
In regions such as Asia and the Pacific, widening disparities between the
well off and the poor and vulnerable groups are a major concern. The latter are
falling further behind their urban counterparts (Economic and Social Com-
mission for Asia and the Pacific, 2008b). To illustrate this point in detail, the
present section will highlight the importance of regional variations by con-
trasting two countries, China and India, both large countries in terms of both
geographical and population size and both regarded as quite successful in hav-
ing reduced poverty at the aggregate level. As a result, however, of divergent
regional patterns of economic growth and social provisions, one finds sharp
differences in levels of living standards across provinces or states, as well as
between rural and urban areas (Ravallion and Jalan, 1999).
For instance, in China, the slower pace of income growth in the central
and western regions compared with the eastern coastal region has widened
the intraregional income gap. This gap is related to structural changes in out-
put and employment: the coastal regions have provided more opportunities
for non-agricultural employment and income. By contrast, the distribution of
agricultural income across regions has been more equal, reflecting the better
quality of control over agricultural land. Notwithstanding China’s substan-
tial improvement in poverty reduction since 1978, new forms of poverty have
arisen. This is accounted for by the deteriorating quality of growth in terms of
its employment-generation potential and an increase in the degree of inequality
(Hu, Hu and Chang, 2003). Trends in poverty have also been closely linked
with trends in employment. In rural areas, slow growth in the agricultural
sector resulted in almost stagnant employment after the mid-1990s. Rising un-
employment had been a major driver of urban poverty in the post-1985 phase,
a dynamic further strengthened by migration from rural areas.
It is known that reforms in China adversely impacted urban poverty by
generating unemployment through the restructuring of the State-owned sector
within a context where the social security system was weak or absent. It has
been argued that urban poverty is closely associated with inability to find work,
and that the increase in urban unemployment as a result of market-oriented

reforms and withdrawal of financial support for ailing State enterprises had
been a prime cause of the increase in urban poverty (Bouche and others, 2004).
Prior to the restructuring of the State-owned enterprises, there was no
great variation in urban poverty among regions owing to guaranteed employ-
ment and the ubiquitous urban welfare system. This regional pattern changed
when market-oriented reforms led to significant closures of State-owned enter-
prises, privatization and large-scale layoffs of workers, and a weakened urban
social welfare system. Changes in the regional distribution of urban poverty
were highly correlated with the original structure of industry and with regional
economic growth. The incidence of urban poverty has tended to be higher in
those regions where the heavy industries—set up during the era of central
planning—were earlier concentrated and lower in the towns and cities of the
south-eastern coast which have experienced more dynamic growth. However,
overall, the prevalence of poverty is much higher in rural areas (26 per cent in
2005) than in urban areas (estimated at under 2 per cent in 2005), although
income inequalities have grown faster in urban areas: the Gini coefficient in-
creased from 26 to 35 per cent in urban areas and from 31 to 36 per cent in
rural areas between 1990 and 2005.
Similar regional differences in levels of living standards have also been
noted in India. While India’s recent economic growth experience has been less
spectacular than that of China, it has still been extremely impressive measured
against that of most other developing countries in the same period and in
comparison with its own past. Real GDP growth rates rose to a higher level
over the last two decades and increases in per capita income were even more
marked because of the falling rate of population growth. Official estimates of
the extent of poverty, that is, the headcount ratio below the official poverty
line, provide some food for speculation in respect of the slower rate of poverty
reduction in the recent period of fast economic growth: poverty has been de-
clining continuously in both rural and urban areas since the early 1970s, but
between 1973-1974 and 2004-2005, the proportion of people living below the
poverty line declined fastest in rural areas, from 56.4 to 28.3 per cent, while in
urban areas the poverty rate declined from 49.2 to 25.7 per cent (India, Press
Information Bureau, 2007; Sharma, 2004).
Levels of poverty have also varied significantly at the state level. The share
of the total number of poor in the southern states of Andhra Pradesh, Kar-
nataka, Kerala and Tamil Nadu decreased from 18 per cent in 1993-1994 to
15 per cent in 1999-2000. In contrast, the share in the total number of poor
in the states of Bihar, Orissa, Madhya Pradesh, Uttar Pradesh and West Ben-
gal jumped from 57 to 63 per cent during the same period (Sharma, 2004).
Therefore, although there has been a steady decline in the incidence of poverty
in India, the efforts of the Government have not resulted in a uniform impact
across regions. There remain regions where the poverty is still deep and severe
and hence they require greater attention.
These two countries demonstrate that, in spite of a country’s strong

growth, rural areas and other depressed regions often face bigger poverty re-
duction challenges. In several countries, the rate of rural urban or interregional
convergence has declined over time as a result of widening income inequalities
and pro-urban industrial and public investments policies. Hence, spatial pov-
erty differentials are likely to persist into the foreseeable future.
Annex II.1
Poverty: indicators and their relationship
The poverty headcount index is the percentage of the population living in house-
holds with consumption or income per person below a commonly agreed pov-
erty line. Trends in the poverty headcount index are determined by trends in
the number of poor persons (the numerator) and by population trends (the
denominator). If the growth (or decline) in the number of poor persons is
proportional to total population growth (or decline), the poverty headcount
index will remain constant. The headcount index will grow if the number of
poor persons grows faster than the total population. Similarly, the headcount
index will decline if the number of poor persons grows more slowly than the
total population. Therefore, when the poverty rate (headcount ratio) falls, this
means not that the total or absolute number of poor has declined but, simply,
that the rate of growth of the number of poor persons is lower than the rate of
growth of the total population.
The absolute number of poor persons by region given by the World Bank is
obtained by applying the estimated headcount index to the population of each
region, under the assumption that the estimated regional headcount index ap-
plies to countries with no data.
Income inequality, or the extent to which income is distributed in a more or
less equitable manner, is measured using various summary indices. The most
well-known is the Gini coefficient, a ratio with values between 0 and 1 (or
between 0 and 100 per cent). A low Gini coefficient indicates a more equal in-
come distribution and a high coefficient indicates a more unequal distribution.
Another widely used series of indices compare the income of a given percen
tage of the richest population (most often the top 10 or 20 per cent) with that
of the total population or of the bottom 10 or 20 per cent. Because different
summary measures are sensitive to different parts of the income distribution,
income inequality rankings depend on the specific measure used.
Poverty trends are arithmetically related to trends of economic growth
per capita (mean income) and income distribution. The figure shows that an
overall change in the proportion of poor persons can be decomposed into a
growth component (area 1), resulting from higher economic growth per capita,
holding distribution constant, and a distribution component (area 2), resulting
from a more equal distribution, holding economic growth constant. According
to this simple arithmetic identity, poverty reduction will be faster when the
growth of per capita income is higher and/or when income distribution is im-
proving. Since income distribution is far from equal in most developing coun-
tries, significant reductions in poverty are possible if distribution improves.
Similarly, for any given growth rate of income per capita, poverty reduction
will occur faster if incomes are more equally distributed.
The impact of these phenomena partly depends on the initial level of
income, inequality and population growth; in highly unequal or very poor
countries, an initial change in income levels or income distribution has a much
stronger impact than in richer, less unequal countries. Empirically, their effects
on poverty differ significantly across countries—even among countries with
similar levels of income. Clearly, political, social and economic factors other
than income per capita, income distribution and population growth are at play.
Decomposition of poverty reduction into growth and composition effects

Share of population
Poverty line
Economy moves from initial income

distribution to final income distribution.
The new position has a higher mean
income and distribution is more equal.
Poverty declines from the full white tinted
area to the hatched white area.
The decline in poverty can be decomposed
into two effects:
1. Growth effect: poverty reduction due to
higher mean income holding income
Initial mean income
distribution unchanged (old distribution

Final mean income
shifting into the right) and

2. Distribution effect: poverty reduction
due to improved distribution holding
mean income unchanged.
1
2
Income ($ per day) log scale

Initial income distribution
Final income distribution
Source: Khan (2009).

Chapter2 PDF

Uploaded by

Chapter2 PDF

Uploaded by

13

Global poverty trends: 1981-20051

A. Number of people living on less than $1.25 a day

800 1,071 873

298 383 388

B. Proportion of the population living on less than $1.25 a day

Source: World Bank, Development Research Group (2009).

A. Number of people living on less than $1.25 a day

B. Proportion of the population living on less than $1.25 a day

With sub-Saharan Africa excluded, the incidence of poverty in the develop-

Regional shares and trends

growth and structural change, particularly in China, have been important

Intraregional poverty trends

East Asia Eastern Latin America Middle East Sub-

Percentage living on less than $1.25 a day

2005 16.8 3.7 8.2 3.6 40.3 50.9

Annual rate of change (percentage)

1990-2005 –7.6 4.2 –2.1 –1.2 –1.6 –0.8

Percentage point change

1990-1999 –19.2 3.1 –0.4 –0.1 –7.6 0.8

Source: World Bank, Development Research Group (2009).

Liberia 74.5 83.2 86.1 41.6 0.2 –44.5 –7.3

Ethiopia 66.2 65.9 39.0 32.9 –3.5 –6.1 –1.7

Source: World Bank, Development Research Group (2009).

national population live on less than $1.25 a day), 12 countries managed to

East Asia and the Pacific

Timor-Leste 82.1 71.3 43.6 35.7 –3.3 –7.9 –2.0

Source: World Bank, Development Research Group (2009).

Nepal — 77 54.7 38.5 –2.3 –16.2 –3.5

Source: World Bank, Development Research Group (2009).

Latin America and the Caribbean

8 Poverty data are not available for rural areas of Argentina.

Haiti 54.6 56.8 58.0 28.4 0.1 –29.6 –7.1

Source: World Bank, Development Research Group (2009).

This trend in poverty is consistent with the findings of the Economic

Middle East and Northern Africa

enhance human capabilities are known to contribute to reductions in poverty

Djibouti 6.1 1.8 18.6 0.9 15.6 –17.7 –30.3

Source: World Bank, Development Research Group (2009).

Eastern Europe and Central Asia

Uzbekistan 0.0 4.9 38.8 2.5 13.8 –36.4 –27.6

Source: World Bank, Development Research Group (2009).

Least developed countries

Source: World Bank, Development Research Group (2009).

Countries of the European Union (EU) and other countries of

All least developed

Source: World Bank, Development Research Group (2009).

Source: OECD (2008a), computations from an OECD income distribution questionnaire.

Diversity within countries: poverty in China and India

and that the increase in urban unemployment as a result of market-oriented

These two countries demonstrate that, in spite of a country’s strong

Decomposition of poverty reduction into growth and composition effects

Economy moves from initial income

distribution unchanged (old distribution

shifting into the right) and

Income ($ per day) log scale

Source: Khan (2009).

You might also like