Chap11 PDF
Chap11 PDF
While economic growth is essential for development, this is hardly an end in itself. In order to derive
benefits of growth, it is important to recognize the interdependence between social and economic
policies and promote their integration. Unfortunately social sector development has been an area of
neglect by successive governments over a long period. In fact, this underdevelopment of human
capital is a serious concern facing Pakistan at present.
In this backdrop, this new chapter in SBP Annual Report provides a brief update on socio-economic
developments in Pakistan with a focus on poverty, demography, employment, literacy, education and
health.
11.1 Poverty
Poverty is the result of economic, social, and
Figure 11.1: He ad Count Ratio
political processes that interact with and T otal Rural Urban Gap
reinforce each other in ways that can
35
accentuate the state of deprivation in which
30
poor people live. Hence, any assessment of
25
poverty encompasses a wide range of inter-
percent
Notwithstanding the issues regarding lack of consistency in definition and estimation methods for
poverty measurement, there is a broad agreement that the incidence of poverty in Pakistan has
increased considerably during the decade of 1990s, particularly between FY97 and FY99 (see Figure
11.1).3 This finding also receives considerable support from various social and human development
indicators, which are showing a dismal picture, both in absolute and relative terms (discussed later).
Focusing on FY97-99, this period is marked by severe fiscal crunch and slowdown in economic
growth. While limited fiscal resources constrained the government from undertaking pro-poor
initiatives, the economic slowdown capped any notable rise in the per capita income (see Figure
11.2). Given the kind of shocks the economy has been facing since FY99, the prospects for any
1
In a broader context, poverty is also related to institutional factors like inability of poor to access public goods and services,
which determine the overall human development.
2
The Planning Commission has defined poverty line estimates at 2,350 calories per adult equivalent per day, which in
monetary terms is equal to Rs 670 per capita per month.
3
The head count ratio (defined as the number of poor below the poverty line) is derived from Household Income
Expenditure Survey (HIES) published by the Federal Bureau of Statistics (FBS), which is available latest for FY99.
State Bank of Pakistan Annual Report FY02
significant poverty reduction have considerably Figure 11.2: Growth in Pe r Capita Income
mitigated.4 Although the development 3.0
expenditures have risen recently following the 2.5
take-off of PRGF, the poverty reduction 2.0
1.5
initiatives would yield definite results after a 1.0
percent
time lag. 0.5
0.0
Looking at Figure 11.1, the poverty trends -0.5
-1.0
were strikingly different in rural and urban -1.5
areas during early 1990s: while the poverty in -2.0
urban areas was declining, the rural areas were
FY93
FY94
FY95
FY96
FY97
FY98
FY99
FY00
FY01
FY02
facing a sharp rise in poverty incidence, thus
leading to a notable jump in the rural-urban
Figure 11.3: Agriculture Growth & Rural Poverty
poverty gap during FY94. The steep rise in
Head count ratio (rural) - LHS
rural poverty appears unusual, as agricultural Agricultural growth (one-period lag) - RHS
sector witnessed an average annual growth of 40 16
3.6 percent during FY91-94. This seeming
35 12
anomaly can be explained by the large
variations in the agricultural growth (see 30 8
percent
percent
Figure 11.3). To explain it further, the 25 4
resulting contraction in rural incomes might 20 0
have forced households to use up their assets as
15 -4
well in order to meet current expenses. Since
asset accumulation and poverty reduction are 10 -8
FY90
FY91
FY92
FY93
FY94
FY95
FY96
FY97
FY98
FY99
closely related, any cut on asset holdings
negatively impacts household welfare.
Interestingly, the subsequent growth in the agriculture sector coincides with the decline in rural
poverty during FY94-97. Afterwards, though the poverty has been growing both in rural and urban
areas, the incidence is marginally higher in urban household, thus resulting into slower narrowing of
the gap (see Figure 11.1).
The persistently higher rural poverty is also Table 11.1: Farm Classification by Size - 1990
attributable to the highly skewed pattern of land percent
holdings. While more than one-half of the rural Hectares Number Farm
population in Pakistan is landless, 93 percent of of farms area
total farms (covering over 60 percent of area) Less than 5 81 39
was smaller than the subsistence land holding 5 and under 10 12 22
of 10 hectares (see Table 11.1). The land 10 and under 20 5 16
ownership is also important in determining the 20 and under 60 2 14
access to credit from the formal banking More than 60 -- 10
system. Source: Pakistan Statistical Year Book 2002 by FBS
In addition, the nature of contract between landowner and cultivator also influences the economic
status of the rural population. For example, a more traditional contract based on `sharecropping',
(whereby the landowner and the cultivator share the cost of the inputs and the revenues from the
harvest) is chosen when the leasee is poor and cannot raise the funds to pay the amount for lease and
inputs beforehand. In such a case, the risks for the cultivator are lower, but so are the revenues.
Furthermore, low level of mechanization does not allow significant increases in value addition and a
4
Since FY99, the economy has suffered the implication of SBA (fiscal squeeze and sharp changes in exchange rate/interest
rates), severe drought, and the September 11 shock on investment climate.
200
Socio-economic Update
The changes in poverty incidence are also Figure 11.4: Income Inequality & Economic Growth
explained by trends in income inequality that Rural Urban GDP growth
largely follow the economic growth rates (see 10
Figure 11.4). While the overall income
8
distribution improved slightly during FY91-99
period, the inequality in urban household is 6
percent
consistently higher than in rural areas, except
4
for FY97; the gap has widened considerably in
FY99. 2
0
As discussed earlier, the poverty is a much FY85 FY91 FY93 FY94 FY97 FY99
broader concept than represented by Income inequality: ratio of households holding highest 20
consumption and income measures alone. In percent income to lowest 20 percent income.
particular, the notion of poverty also considers
deficiency in social and human development as Table 11.2: Poverty of Opportunity Index (POPI)
well as institutional growth. Looking from this percent
perspective, the broad definition of poverty Poverty of opportunity in POPI
POPI
provides some understanding on the widening Health Education Income Male Female Gap
social gap between urban and rural areas, and 1970 55 77 40 61 56 67 100
between genders, which is considered as one of 1975 49 74 35 58 52 64 102
the main reasons hindering growth and poverty 1980 46 73 38 56 51 62 101
reduction. Although it is difficult to estimate a 1985 42 67 25 51 46 59 107
single index of poverty that includes all of its 1990 36 62 20 46 41 56 114
dimensions, the Mahbub-ul-Haq Centre for 1995 30 58 30 44 37 52 116
Human Development (MHCHD) has come up Source: Mahbub-ul-Haq Centre for Human Development, 1999
with a composite index of poverty in Pakistan
that incorporates deprivation in health and Table 11.3: Trends in POPI for Rural and Urban Areas
education along with the income poverty.5 Poverty of opportunity
Gap
Urban Rural
It is evident from Table 11.2 that income 1986 40 61 100
poverty is increasing during 1990s. However, 1988 37 55 99
the overall poverty of opportunity index (POPI) 1993 35 48 92
is falling largely due to significant Source: Mahbub-ul-Haq Centre for Human Development, 1999
improvements in health. On the other hand, the
poverty of education opportunities is showing a more gradual decline. Like income-based measure,
POPI is also showing widening gender-gap since 1980. The rural-urban gap has narrowed since 1986,
which is in contrast to earlier findings based on
income poverty (compare Figure 11.1 with Table 11.4: Pakistan’s Ranking in the World Population
Table 11.3). 1950-55 2000-05 2045-50
Population size1 13 7 4
11.2 Demography Population increase 9 3 2
There is a broad consensus that the rapid World annual births 9 3 3
1
population growth and poverty reinforce each : Data pertains to 1950, 2000 and 2050 respectively
Source: World Population Prospects: the 2000 Revision, Vol. III,
other (i.e., high fertility causes poverty, which United Nation Population Division
5
The POPI is a composite of deprivation in health, education and income. The POPI for health is a weighted average of
three variables: the percentage of people not expected to survive to age 40; the percentage of people without access to safe
water; and percentage of malnourished children under age 5. The POPI for education includes two variables: the percentage
of illiterate adults and the percentage of primary school-age children who are out of school. The income deprivation is based
on estimates from work conducted by Shahid Javed Burki.
201
State Bank of Pakistan Annual Report FY02
in turn, contributes to the higher fertility). In this context, the low fertility turns up as one of the keys
to reducing poverty incidence.
However, in the case of Pakistan, despite definite evidences of fertility decline in 1990s, the
population growth rate is still very high in relative terms.6 To put this into perspective, in terms of
population size, Pakistan is the 7th most populous country in the world and the 4th in Asia. It is
expected that by 2050, Pakistan’s population will rank fourth in the world after India, China and USA
(see Table 11.4). Even considering the
projected world population increase during Table 11.5: Intercensal Growth Rates in Pakistan
2000-2005, Pakistan ranks third in absolute Annual growth
Population rate
numbers after India and China.
1941 28,282 1.9
1951 33,816 1.8
A further analysis of demographic increase for
1961 42,978 2.4
Pakistan indicates a long history of high
1972 65,321 3.6
population growth rates (see Table 11.5). The
1981 84,253 3.0
expansion during 1960s and 1970s is attributable
1998 130,580 2.6
to the lagged impact of the rapid decline in crude
death rate observed during 1950s and 1960s,
which was not followed by any fall in crude birth rate in these decades (see Figure 11.5).7 In fact, the
total fertility rate (TFR) was relatively higher in 1970s compared to the previous decade (see Table
11.6).8
Table 11.6: Trends in Total Fertility Rates Depicted by Various Data Sources
1960s 1970s 1980s 1990s
Source TFR Source TFR Source TFR Source TFR
PGE 1962-65 (CD) 7.95 PLM 1975-79 6.50 PCPS 1984-85 6.0 PCPS 1994-95 5.6
PGE 1962-65 (LR) 6.09 PLM 1970-75 7.10 PDS 1984-88 6.9 PFFPS 1992-96 5.4
NIS 1968-69 5.02 PFS 1970-74 6.28 PDHS 1986-91 5.5 PIHS 1994-96 4.5
PGS 1968-71 6.04 PFS 1965-69 7.07 PIHS 1987-91 6.3
PGS 1976-79 6.90
Average 6.3 6.8 6.2 5.1
Source: ‘Fertility in Pakistan: Past, Present and Future’ by Zeba A. Sathar
PGE (1962-65) CD: Population Growth Experiment Rates Adjusted by Chandra-Deeming Formula
PGE (1962-65) LR: Population Growth Experiment - Rates Based on Longitudinal Registration
NIS (1968-69): National Impact Survey
PGS (1968-71): Population Growth Survey, 1968, 1969, 1971
PLM (1970-74 & 1975-79): Pakistan Labor Force and Migration Survey 1979
PFS (1965-69 & 1970-74): Pakistan Fertility Survey 1975
PGS (1976-79): Population Growth Survey, 1976, 1977, 1978 and 1979
PCPS (1984-85): Pakistan Contraceptive Prevalence Survey 1984-85
PDS (1984-88): Pakistan Demographic Surveys 1984, 1985, 1986, 1987 and 1988
PDHS (1986-91): Pakistan Demographic and Health Survey 1990-91
PIHS (1987-91): Pakistan Integrated Household Survey 1991
PCPS 1994-95: Pakistan Contraceptive Prevalence Survey
PFFPS 1996-97: Pakistan Fertility and Family Planning Survey
PIHS 1994-96: Pakistan Integrated Household Survey
6
According to current estimates, Pakistan has a total population of 145.96 million, with intercensal growth rate recording a
considerable fall from 3.0 to 2.6 percent during 1981-98 period.
7
Crude birth rate is calculated as the number of births per 1000 population, whereas crude death rate is calculated as the
number of deaths per 1000 population. The Figure 11.5 is based on numbers available in various issues of Pakistan
Economic Survey.
8
Total fertility rate indicates the number of children per woman during her reproductive span of life.
202
Socio-economic Update
Figure 11.5: Demographic Transition in Pakistan Figure 11.6: Singulate Me an Age at Marriage
Male Female
Crude birth rate Crude death rate
60 29
27
50
per 1000 population
25
40
23
year
30
21
20
19
10
17
0
15
1901 1921 1941 1961 1981 2001
1951 1961 1972 1981 1991 1997 2001
Subsequently, the TFR showed a marginal decline during 1980s, and a more distinct fall for 1990s.
The recent evidence on fertility transition is also supported by rising age of females at marriage (see
Figure 11.6).9 Consequently, the population growth rate that remained at fairly high level of 3.0
percent during 1972-81, fell to 2.6 percent during 1981-98. Since the growth is an average of
intercensal period, this suggests that the population growth during the latter part would be still lower.
Despite this falling population growth rate, the pace and size of fertility transition are very slow
compared to other developing countries. As shown in Table 11.7, Pakistan’s performance is very
poor in terms of improvement in total fertility rate per decade. This also points out the fact that the
decline in birth and death rates is not necessarily a function of economic development. Indeed, the
effectiveness of population policies is also a major determinant of falling birth and death rates. In this
regard, the examples of Sri Lanka and Bangladesh are notable where the total fertility rate declined
rapidly despite the fact that these countries have not yet achieved significant levels of economic
development.
This slower pace of fertility transition has clear implications for the age structure of the population.
As evident from Figure 11.7, around 43 percent of the population is under 15 years of age, whereas
the adolescent population (age 15-24) is around 20 percent. This means that a sizeable cohort will
very soon be reaching childbearing age. At the same time, the large female population in the age
group of 15-44 is also adding to the built-in inertia of population growth that may not allow reduction
in fertility rate to immediately translate into equivalent fall in birth rates. It may be noted that the
9
The singulate mean age at marriage (the average number of years lived in single state) is based on census for years 1951,
1961, 1972 and 1981; Pakistan Demographic Housing Survey for 1991; Pakistan Fertility and Family Planning Survey for
1997; and Pakistan Reproductive Health and family Planning Survey for 2001.
203
State Bank of Pakistan Annual Report FY02
1981 1998
70--74 70--74
60--64 60--64
Female
Female
Male
50--54
Male
50--54
40--44 40--44
30--34 30--34
20--24 20--24
10--14 10--14
0--4 0--4
20 15 10 5 0 0 5 10 15 20 20 15 10 5 0 0 5 10 15 20
------ in percent ------ ------ in percent ------
longer the fertility rate remains at high levels, the more extended will be the population growth
momentum. In other words, even if the fertility rate reaches the replacement level, this will take
further several decades to attain population stability (where birth rate balances death rate).
According to the projections provided by the Ministry of Population Welfare, if the fertility rate
declines to replacement level of 2.1 children per woman by the year 2023, Pakistan’s population will
reach 204 million (see Table 11.8). However, due to momentum alone, population will continue to
grow for another 3-4 decades. Thus, the attainment of population stabilization is not possible before
40-50 years now onward. Indeed, any shortcoming in the actual fertility rate from the targeted level
will further stretch the projected timeline.
The slower pace of fertility transition has strong Table 11.8: Population Estimates for 2023
implications for economic and social
Total fertility rate Population
development. In particular, this results into children per woman million
relatively higher dependency ratio. According to
2.6 217
1998 Census, only 32 percent of the total
2.4 212
population lies in the working age group (25-
2.1 204
59), which is almost unchanged since 1981. The
Source: Interim Population Sector Perspective Plan 2012,
resulting high dependency ratio not only The Ministry of Population Welfare
constraints the saving capacity of average
households but it also has strong implications for their consumption pattern and overall quality of life.
Finally, the rapid urbanization during the last two decades is a remarkable change in the population
structure. During 1981-98, the urban population has grown at an annual rate of 3.5 percent, which is
higher than the total population growth of 2.6 percent during the same period. Apart from
urbanization related problems, this also has some positive implications for overall population growth,
as fertility rates are found to be considerably lower in major urban areas.10
11.3 Employment
The employment profile of a country is determined by an interaction of demographic, economic,
social and political factors. In Pakistan, despite a discernible fall in the intercensal growth rate,
population pressures continue to impact negatively on the employment. In addition, relatively lower
economic growth during the last ten years affected the employment situation.
10
According to the Pakistan Demographic Survey (1992), the fertility rates for urban and rural population were 6.2 and 7.3
children per woman respectively – a differential of 1.1 child per woman. However, Pakistan Fertility and Family Planning
Survey (1996-97) shows a higher differential of 2 children per woman, with fertility rates for urban and rural population at
3.8 and 5.8 respectively.
204
Socio-economic Update
The rate of unemployment in Pakistan is estimated with the help of Labour Force Survey (LFS),
which was conducted in FY00 by the Federal Bureau of Statistics (FBS). According to the survey,
“The unemployment situation has shown an increase from 5.9 percent in FY98 to 7.8 percent in
FY00. This increase has been observed for both males and females. Females are more unemployed
(17.3 percent) compared to males (6.1 percent). Further by treating the employed persons, who
worked less than 15 hours during the reference week, as unemployed, the unemployment rate jumps
to 9 percent”.
Table 11.9: Unemployment and Investment
Although, estimates for the unemployment rate Rate of unemployment Investment to
in the current (and recent years) will not be GDP ratio
available till the next Labour Force Survey is FY90 3.1 18.9
conducted by FBS, it can be safely conjectured FY91 6.3 19.0
that the significant declines in this rate would FY92 5.9 20.3
remain unlikely unless the economic growth FY93 4.7 20.8
and investment scenario changes significantly. FY94 4.8 19.6
Table 11.9 shows the rates of unemployment FY95 5.4 18.6
for the years in which LFS were conducted FY96 - 19.0
along with the investment to GDP ratios for the FY97 6.1 17.9
past ten years. An inverse relationship between FY98 5.9 17.7
unemployment rates and investment to GDP FY99 - 15.6
ratios is also observable in Figure 11.8. FY00 7.8 16.0
FY01 - 15.9
While there is a considerable misunderstanding FY02 - 13.9
regarding estimates of unemployment rate,
Chart 11.1 helps in clarifying some of the Figure 11.8: Unemployme nt and Inve stment
underlying concepts. The size of the labour Unemployment rate Investment to GDP ratio
force is assessed by estimating the Currently 25
Active Population (labour force) that considers 20
all persons (ten years age and above) fulfilling
percent
FY92
FY94
FY96
FY98
FY00
FY02
rate.
• In Pakistan, the participation rate in economic activities has decreased from 29.4 percent in FY98
to 29.0 percent in FY00. The decrease has been observed for male as well as females.
• The structure of employment by industry shows that the share of agriculture sector has increased
from 47.2 percent in FY98 to 48.4 percent in FY00 which is an year of exceptional growth. This
may be related to a disguised unemployment in agriculture sector. Persons who are not getting
employment in the non-agriculture sectors are accommodated by the growing agriculture sector.
11
This definition is fully in accordance with the guidelines of International Labour Orgnization regarding compilation of
labour statistics.
205
State Bank of Pakistan Annual Report FY02
Population
Voluntarily
unemployed
Unemployed Employed
Too young to Those without work but Those who have worked at
work currently available and least one hour during the
seeking work reference period on the basis
of “paid -employed” or “self -
employed”
Living on charity
Paid-employed Self-employed
Disabled
Under-employed
Retired or old
School going
• Further, within the non-agriculture sector, two-thirds (65.8 percent) of the employed persons were
engaged in informal sector. Informal sector in rural areas has accommodated relatively more
employed persons (68.0 percent) compared to urban areas (63.8 percent). This shows that the
formal sector is unable to create enough jobs consistent with the labour supply.
• Parallel to the increase in agriculture sector, farming activities have also shown an upward trend
from 39.9 percent in FY98 to 40.6 percent in FY00. In addition to other reasons, the disguised
unemployment factor may have also contributed to its size, as persons who are not getting
employment in other sectors are engaged in farming activities.
• The employment status indicates that self-employed category has increased from 41.5 percent in
FY98 to 42.2 percent in FY00. Similarly the employees category have also shown an increasing
206
Socio-economic Update
trend. In contrast, the unpaid family helpers and employers categories have experienced a
downward trend.
11.4 Literacy
According to 1998 Census, the literacy rate in Table 11.10: Literacy Rate
Pakistan increased for both men and women Male Female Overall
during intercensal period (see Table 11.10). Overall 1998 56.5 32.6 45.0
1981 35.0 16.0 26.2
However, there is still wide disparity between
Rural 1998 47.4 20.8 34.4
male and female, and between rural and urban 1981 26.2 7.3 17.3
populations. Urban 1998 72.6 55.6 64.7
1981 55.3 37.3 47.1
The analysis of the literacy age profile for Source: Pakistan Statistical Year Book 2002
urban population suggests higher literacy rates
for younger group. Interestingly, the gender disparity almost disappears for this age group as the
female literacy rate is increasing at a much faster pace. In the case of rural population, though the
literacy rate is increasing, the gender-gap remains still very high for all age brackets (see Figure 11.9
& Figure 11.10).
Figure 11.9: Literacy Age Profile (Urban) - 1998 Figure 11.10: Literacy Age Profile (Rural) - 1998
T otal Male Female T otal Male Female
90 90
80 80
70 70
60 60
percent
percent
50 50
40 40
30 30
20 20
10 10
0 0
55+
55+
45-54
35-44
25-34
20-24
15-19
10-14
45-54
35-44
25-34
20-24
15-19
10-14
Looking from a different perspective, the difference in the male literacy rate between rural and urban
population is falling with lower age. However, disparity in the female literacy rates is widening with
lower age groups (compare Figure 11.9 with Figure 11.10).
Rural
Urban
Urban
Overall
Overall
12
The primary gross enrollment ratio is the number of students enrolled at the specified level (regardless of their age) as
percent of population of official school age for that level.
207
State Bank of Pakistan Annual Report FY02
showing an improvement both in rural and urban areas, and thereby leading to decline in the gender
gap.
Table 11.11: Education Indicators for Selected Countries
The inter-country comparison of selected Gross
indicators in the education sector reflects very Adult literacy primary
Expenditures ratio (age 15 and enrollment
poor standing for Pakistan (see Table 11.11). as % of GNP above) ratio1
This shows that the education sector is
1985-87 1995-97 1985 2000 1970 1997
generally accorded low priority in economic Pakistan 3.1 2.7 31.4 43.2 40 74
policy decision despite its important role in Malaysia 6.9 4.9 76.4 87.5
economic growth and poverty reduction. The Thailand 3.4 4.8 90.3 95.5
situation is more disturbing as the lower quality Sri Lanka 2.7 3.4 87.1 93.3 99 109
of education and mismatch of acquired skills China 2.3 2.3 71.9 84.1
with the market demand lead to lower returns Indonesia 1.0 1.4 74.7 86.9
on investment in human capital. This, in turn, India 3.9 3.2 45.2 57.2 73 100
weakens the role of education as a catalyst in Bangladesh 1.4 2.2 32.0 41.3 54 92
poverty reduction efforts. The poor education Sources: Human Development Report 2002, UNDP
also drags country’s ability to adopt 1
: Human Development in South Asia 2001, Globalization and Human
technological innovation, and in turn Development, Mahbub ul Haq Human Development Centre, Oxford
University Press.
integrating with the global economies.
11.6 Health
Health is also closely related to the social sector performance. In this regard, infant mortality rate, life
expectancy at birth, access to basic health services, and expenditure on health services are important
indicators.
Infant mortality rate (IMR) reflects the quality of health services provided in the country. As shown
in the Table 11.12, IMR declined from 107.7 in
1990 to 85.0 in 2002,but it is still very high on Table 11.12: Indicators of Health
account of unhygienic livings, infectious 1990 2002
diseases, malnutrition, and lack of education Infant mortality rate (per 1,000 live births) 107.7 85.0
and information of childcare on the part of Life expectancy at birth (years) 57.7 63.6
mothers. Expenditure on health (as percent of GNP) 0.8 0.7
Mortality (per 1,000) 1231 110.3
1
Life expectancy at birth is about 63.6 years in : Mortality rate pertains to the year 1996.
Pakistan in year 2002, although it has
Figure 11.12: Health Expenditure
improved from year1990 (see Table 11.12), 0.9
but still lower than the average of 67.3 years 0.8
for countries at medium level of human 0.7
as percent of GNP
208
Socio-economic Update
The low budgetary allocation explains much of Table 11.13: Indicators of Health
the poor performance in the health sector. As Life Infant mortality Under 5-years
shown Figure 11.12, expenditures on health are expectancy (per 1000 live mortality
(years) births) (per 1000)
unchanged for the last five years. 1990 2000 1970 2000 1970 2000
Pakistan 57.7 60.0 118 95 183 110
In comparison to other regional and developing Malaysia 70.1 72.5 46 9 63 9
countries, the Pakistan’s performance in the Thailand 66.1 70.2 74 30 102 29
selected health indicators is not satisfactory Sri Lanka 70.9 72.1 65 17 100 19
(see Table 11.13). China 70.1 70.5 85 38 120 40
Indonesia 61.5 66.2 104 40 172 48
India 59.1 63.3 130 69 206 96
Bangladesh 51.8 59.4 148 79 239 82
Source: Human Development Report 2002, UNDP
209