POVERTY IN INDIA

IWE
2016-17
ANUP SINHA
 COVERAGE
Introduction
Concepts of Poverty and Poverty Line
Measurement of Poverty
Trends in Poverty over Time
Variations across States and Social Groups
Inequality: Concept and Measurement
Some Policy Issues
 GENERAL
One fifth of the worlds people live on less than $ 1
a day, and 44% of them are in South Asia
26 percent of India is below the poverty line
This is happening in mainly in rural areas of India
POVERTY IN THE STATES OF INDIA
One half of Indias poor is located the three states
of Uttar Pradesh, Bihar, and Madhya Pradesh
Maharashtra, West Bengal and Orissa account for
22.5% of poverty
 FEMALE LITERACY AND INFANT
MORTALITY RATES
Lack of food and health care due to low
income/assets is associated with the higher
probability of a new born child dying between birth
and the age of one
The High Female illiteracy rate has a major impact
on IMR
If more women were literate the IMR would be
much improved
 WHY IS THIS HAPPENING
Even though Indias economy is growing there
wealth distribution is alarmingly uneven
1/4 of the nation's population earns less than the
government-specified $0.40/day
Unemployment and underemployment
Over-reliance on agriculture
High population growth rate
Lack of basic capabilities in terms of health and
education
 CULTURAL REASONS
The Caste and Religion prevent poor people from
educational, ownership, and employment
opportunities
 WHAT IS BEING DONE
Microfinance( very small loans) has helped India a
lot
There are multiple organizations to help them and
keep their agriculture going and seek alternative
livelihoods
Large number of pro-poor specifically targeted
government programmes both at state as well as
central levels e.g MGNREGA
 INTRODUCTION
Indias economic structure has changed dramatically
over last 5-6 decades; among the most dynamic
economies recently.
Benefits of growth not widely spread to various
sections in society, reached only marginally to low
income groups.
Similar experience of other countries too.
Question then arose: Can we guarantee to all at least a
minimum level of living necessary for physical and
social development of a person?
Absolute poverty literature grew out of this question.
 WHY ESTIMATE POVERTY?
Poverty estimates are vital input to design, monitor and
implement appropriate anti-poverty policies.

Analysis of poverty profiles by regions, socio-economic

groups
Determinants - factors affecting poverty
Relative effects of factors affecting poverty
Allocation of resources to different regions and to
various poverty reduction programs

Precise estimates of poverty neither easy nor universally

acceptable. Yet, can act as a broad and reasonably good
policy guide.
Intellectual genesis of poverty very old
Adam Smith, Ricardo, Marx: subsistence wage concept

An early empirical work by Dadabhai Naoroji, 1901

Estimated an income level necessary for the bare wants

of a human being, to keep him in ordinary good health
and decency. Estimated cost of food, clothing, hut, oil
for lamp, barber and domestic utensils to arrive at
subsistence per head.

In the absence of income distribution data, Naoroji

compared computed subsistence level with per capita
production to draw attention to mass poverty.
Poverty is multidimensional
Deprivation in income, illiteracy, malnutrition,
mortality, morbidity, access to water and sanitation,
vulnerability to economic shocks.

Income deprivation is linked in many cases to other

forms of deprivation, but do not always move together
with others.

This discussion focuses on Income poverty.

 MEASUREMENT OF POVERTY
(PERCENTAGE OF POOR)
Two basic ingredients in measuring poverty:

(1)Poverty Line: definition of threshold income

or consumption level

(2)Data on size distribution of income or

consumption (collected by a sample survey
representative of the population)
 POVERTY LINE (PL): ABSOLUTE VS.
RELATIVE

Relative PL defined in relative terms with reference to

level of living of another person; or, in relation to an
income distribution parameter.
Examples: 50% of mean income or median, mean minus
one standard deviation.

Absolute PL refers to a threshold income (consumption)

level defined in absolute terms. Persons below a pre-
defined threshold income are called poor.
 INDIAN POVERTY LINE
A minimum level of living necessary for physical and social
development of a person.
Estimated as: total consumption expenditure level that
meets energy (calorie) need of an average person.
PL comprises of both food and non-food components of
consumption.
Considers non-food expenditure actually incurred
corresponding to this total expenditure.
Difficult to consider minimum non-food needs entirely on
an objective basis
RELATIONSHIP BETWEEN CALORIE INTAKE AND
PER CAPITA EXPENDITURE

3500
Per Capita Calorie Intake per day

3000
2500
2000
1500
1000
500
0
0 100 200 300 400 500 600 700 800 900 1000
Per Capita Consumption Expenditure per Month
(Rupees)
AN EXAMPLE OF SIZE DISTRIBUTION OF CONSUMPTION EXPENDITURE

MPCE %Population
0-150 3.2
150-200 4.0
200-250 6.5
250-300 8.6
300-340 10.0 (half of 10% are below poverty line 320)
340-400 11.3
400-450 8.6
450-500 9.2
500-550 9.3
550-650 11.4
650-800 8.9
800-1000 5.0
Above 1000 4.0
All classes 100.0
MPCE: Monthly Per Capita Consumption Expenditure

Poverty Line: Rs. 320 per capita per month

HCR= 3.2+4.0+6.5+8.6+5.0 = 27.3%
 INCIDENCE OF POVERTY VS. UNDER-NUTRITION
Classification of Population by Poverty Line and
Calorie Norm - Rural India, 1977-78

Below Above Total

Poverty Poverty
Line Line
Below Calorie 45.32 12.47 57.79
Norm
Above Calorie 12.31 29.21 42.21
Norm
Total 57.63 42.37 100.00
Source: Government of India (1993): Report of Expert Group.
 OFFICIAL PL IN INDIA
Originally estimated for 1973-74: Rs 49 and 56 for rural
and urban areas respectively.
Updated using an appropriate price index (CPIAL for
rural India, CPIIW for urban).
A monthly per capita consumption expenditure of Rs.
356 and 539 for rural and urban areas respectively for
2004-05.
More than a quarter of Indias population remain
below PL in 2004-05.

28.3% Rural 25.7% Urban 27.5% Total

Absolute no.: 302 million in 2004-05

 POVERTY IN INDIA: CHANGES OVER TIME
70
65
Rural HCR
60
% population below PL

Urban HCR
55
50
45
40
35
30
25
20
1960

1963

1966

1969

1972

1975

1978

1981

1984

1987

1990

1993

1996

1999

2002

2005
Up to mid-1970s fluctuations with cycles
Since mid-1970s continuous fall
Except a few years immediately after start of reforms
(early 1990s)
Controversies around estimates for 1999-2000 (under
estimates poverty)
POVERTY HEAD COUNT RATIO: MAJOR INDIAN STATES
POVERTY BY SOCIAL GROUPS: RURAL 2004-05
States ST SC OBC OTHERS
Andhra Pradesh 30.5 15.4 9.5 4.1
Assam 14.1 27.7 18.8 25.4
Bihar 53.3 64 37.8 26.6
Chhattisgarh 54.7 32.7 33.9 29.2
Delhi 0.0 0.0 0.0 10.6
Gujarat 34.7 21.8 19.1 4.8
Haryana 0.0 26.8 13.9 4.2
Himachal Pradesh 14.9 19.6 9.1 6.4
Jammu & Kashmir 8.8 5.2 10.0 3.3
Jharkhand 54.2 57.9 40.2 37.1
Karnataka 23.5 31.8 20.9 13.8
Kerala 44.3 21.6 13.7 6.6
Madhya Pradesh 58.6 42.8 29.6 13.4
Maharashtra 56.6 44.8 23.9 18.9
Orissa 75.6 50.2 36.9 23.4
Punjab 30.7 14.6 10.6 2.2
Rajasthan 32.6 28.7 13.1 8.2
Tamil Nadu 32.1 31.2 19.8 19.1
Uttar Pradesh 32.4 44.8 32.9 19.7
Uttarakhand 43.2 54.2 44.8 33.5
West Bengal 42.4 29.5 18.3 27.5
All India 47.2 36.8 26.7 16.1
POVERTY BY SOCIAL GROUPS: URBAN 2004-05
States ST SC OBC OTHERS
Andhra Pradesh 50 39.9 28.9 20.6
Assam 4.8 8.6 8.6 4.2
Bihar 57.2 67.2 41.4 18.3
Chhattisgarh 41.0 52.0 52.7 21.4
Delhi 9.4 35.8 18.3 6.4
Gujarat 21.4 16 22.9 7.0
Haryana 4.6 33.4 22.5 5.9
Himachal Pradesh 2.4 5.6 10.1 2.0
Jammu & Kashmir 0.0 13.7 4.8 7.8
Jharkhand 45.1 47.2 19.1 9.2
Karnataka 58.3 50.6 39.1 20.3
Kerala 19.2 32.5 24.3 7.8
Madhya Pradesh 44.7 67.3 55.5 20.8
Maharashtra 40.4 43.2 35.6 26.8
Orissa 61.8 72.6 50.2 28.9
Punjab 2.1 16.1 8.4 2.9
Rajasthan 24.1 52.1 35.6 20.7
Tamil Nadu 32.5 40.2 20.9 6.5
Uttar Pradesh 37.4 44.9 36.6 19.2
Uttarakhand 64.4 65.7 46.5 25.5
West Bengal 25.7 28.5 10.4 13.0
All India 33.3 39.9 31.4 16.0
 POVERTY MEASURES
Head Count Ratio (HCR), Poverty Gap
(PG) and Squared Poverty Gap (SPG)
m
HCR
n

z yi
m
PG 1
n i 1
(
z
)

z yi
m 2

n
SPG 1
i 1 z
m= no. of poor population, n = total population,
z= poverty line, yi =income of i-th person
 ALTERNATIVE POVERTY MEASURES
Head Count Ratio (HCR): proportion of total population that
falls below poverty threshold income or expenditure. Based
on either national PL or dollar-a-day PL.
Poverty Gap Index (PGI): unlike HCR, it gives us a sense of how
poor the poor are. It is equivalent to income gap below PL per
head of total population, and expressed as a percentage of the
poverty line.
Squared Poverty Gap index (SPG): Adds the dimension of
inequality among the poor to the poverty gap index. For a
given value of the PGI, population with greater dispersion of
income among poor indicates a higher value for the SPG.
INCIDENCE OF POVERTY AFFECTED BY TWO FACTORS:

(1)Growth in average income (2)Distribution.

Poverty reduction fast when average income rises and

inequality falls.

Fluctuations in poverty incidence till early 1970s primarily

due to slow per capita income growth.

Incidence of poverty started to fall after mid-1970s when

there was marked acceleration in per capita GDP growth
rate to above 3 per cent.
 AVERAGE ANNUAL GROWTH RATES: REAL GDP
1951-2 to 1981-82 1991-92 2000-01 2002-03 to
1980-81 to 1990- to 1999- to 2006- 2006-07
91 2000 07 (Tenth Plan
Period)
Agriculture 2.6 3.8 3.0 2.5 2.2
Industry 5.3 7.0 5.7 7.8 9.1
Service 4.6 6.7 7.9 8.5 9.4
GDP (total) 3.6 5.6 5.8 6.9 7.6
Per Capita GDP 1.4 3.4 3.6 5.2 6.0
Neglect of agriculture after economic reforms even as
overall economic growth accelerated
AVERAGE ANNUAL GROWTH RATE IN PER CAPITA GSDP
ARRANGED BY 1993-94 PER CAPITA GSDP

16000 6.0
14000
5.0
12000
4.0
10000
8000 3.0
6000
2.0
4000
1.0
2000
0 0.0
Tamil Nadu

Jammu and Kashmir

Bihar
Punjab

Kerala

Orissa
Haryana

Karnataka

Rajasthan
Maharashtra

Uttar Pradesh
Andhra Pradesh

Madhya Pradesh
Himachal Pradesh

Assam
Gujarat

West Bengal

Per capita Income 1993-94 Growth Rate 1993-2004

COEFFICIENT OF VARIATION IN PER CAPITA GSDP
AMONG 16 MAJOR STATES

0.4000
0.3900
0.3800
0.3700
0.3600
0.3500
0.3400
0.3300
0.3200
0.3100
1993- 1994- 1995- 1996- 1997- 1998- 1999- 2000- 2001- 2002- 2003- 2004-
94 95 96 97 98 99 00 01 02 03 04 05
 URBAN-RURAL DIFFERENCES IN MEAN CONSUMPTION
EXPENDITURE
States Urban MPCE as % of Rural MPCE

1993-94 2004-05
Andhra Pradesh 141.5 173.9
Assam 177.9 194.8
Bihar 142.9 166.9
Chhatishgarh 180.6 232.9
Gujarat 149.8 187.1
Haryana 123.1 132.3
Himachal Pradesh 212.8 174.2
Jharkhand 190.7 232.0
Karnataka 157.2 203.3
Kerala 126.7 127.4
Madhya Pradesh 155.7 205.9
Maharashtra 194.1 202.1
Orissa 183.2 189.7
Punjab 118.0 156.6
Rajasthan 132.0 163.1
Tamil Nadu 149.0 179.4
Uttar Pradesh 141.2 151.2
Uttaranchal 166.7 158.5
West Bengal 169.9 200.0
All India 163.0 188.2
Poverty depends on per capita household income which
in turn affected by employment, wage rate, land
productivity, industrialization, expansion of service
sector and other general growth and distribution
factors
Special role of
per capita agricultural income
Employment and real wage rate
Inflation rate and relative food prices
Government expenditure
Per capita development expenditure
Social sector expenditure
Indian growth process since 1950s more or less
distribution neutral till 1980s.
Importance of a critical minimum steady growth in per
capita income for poverty reduction.
Inequality increased in recent years after reforms.
Income elasticity of poverty has fallen.
A given growth will be associated with more limited
gains for the poor
Higher growth might more than compensate the adverse
effect if fall in elasticity is small.
Reasons for weak participation of poor: limited access
to education, land, credit; low agricultural growth,
underdeveloped infrastructure
 DEMOGRAPHIC DIVIDEND
As fertility drops, ratio of workers to non-workers
rises.
Provides an window of opportunity provided
potential workers acquire skills and find
productive employment
About a fourth of poverty reduction could be
attributed to demographic factors in India
Right economic policies critical, otherwise the
scenario could turn out to be demographic liability
Dividend for 2-3 decades only since proportion of
older population would eventually increase
increasing dependency ratio again
 LONG TERM SCENARIO FOR POVERTY
Long term growth prospects fairly optimistic: India likely to
continue among the fasted growing economies, BRIC to
dominate world economy
India might surpass Japan and Germany in terms of total size
of the economy, yet its per capita income would be less than
world average for a long time
Poverty could be reduced faster provided inequality is under
control, labour intensive activities must grow, removal of
rigidities in land and labour market critical for reallocation of
resources
Government can afford to devote more resources for poverty
removal programmes: wage employment (NREGA) or self
employment type (SJSY).
 RISING INEQUALITIES AND
GLOBALIZATION

SOME DATA AND A SIMPLE ILLUSTRATIVE MODEL

 RANKING OF THE 10 BIGGEST COUNTRIES IN
2003
GDP billion 1990 Population Per capita GDP
PPP$ million 1990 PPP$
China 6,188 1,288 4,803
USA 8,431 290 29,037
India 2,267 1,050 2,160
Japan 2,699 127 21,218
Germany 1,577 82 19,144
France 1,316 60 21,861
UK 1,281 60 21,310
Russia 914 145 6,323
Indonesia 763 214 3,555
Brazil 1,013 182 5,563
 RANKING OF THE 10 BIGGEST COUNTRIES IN
2030
GDP billion Population Per capita GDP
1990 PPP$ million 1990 PPP$
China 22,983 1,458 15,763
USA 16,662 364 45,774
India 10,074 1,421 7,089
Japan 3,488 116 30,072
Germany 2,406 80 30,179
France 2,171 63 34,462
UK 2,150 64 33,593
Russia 2,017 126 16,007
Indonesia 1,973 285 6,924
Brazil 1,853 223 8,316
 RICH & POOR NATIONS
Rich Nations Poor Nations
Population (billions) 1 2.3
GDP per capita (1990 PPP$) $30,000 $2,100
Human Development Index high low
Annual population growth rate(%):1966-2004 0.8 2.4
Annual growth rate of GDP per capita(%):1966-
2004 2.4 1.8
Total Fertility Rate 1.08 3.7
Adult Literacy(%) >95 58
Index of Government Corruption low high
Life Expectancy at birth (yrs) 78 58
Under 5 mortality (per 1000) 7 120
Rural Population(% of total population) 10 70
Agriculture's share in GDP(%) 5 25

Source: World Development Indicators (World Bank,2005)

 SOME FACTS ABOUT INEQUALITY
In 2002 the ten richest people in the world had a
net worth of $217 billion.
They had made money by software innovation, oil,
retailing scale economies, investment luck and
inheritance
Even if they invested their wealth very carelessly,
they would get at least 5%, that is $10.5 billion
Tanzania, that year, had a GDP of $10.15 billion.
Population 35 million!
 The richest country Norway had a per capita income
of $43000 as against the poorest (Ethiopia and
Burundi) at $90
Even if PPP corrections are made, a person picked in
Norway would likely to be 60 times more rich than
a randomly chosen person from Burundi
Even if global inequality is falling the LEVEL of
inequality is staggering and unacceptable to most
people
 WHAT DO EMPIRICAL STUDIES REVEAL?

Generally shows increase in both inter-country and

intra-country inequality
There does remain problems of measurement
population weighted or unweighted data to be used
Population weighted means all Indians (total
population) are assumed to earn the per capita
income of India and then compute the Gini
Coefficient for the world
Unweighted means each country is treated as one
person with the per capita income of the country
 Both measures are obviously imperfect
Unweighted data inequality across countries rising
Weighted data inequality falling very slowly, but
picking up a little bit from the 1990s
Gini measures worsening for most emerging
economies
Ideal is to rank all individuals globally and then see
the trend
However, whichever way we measure the absolute
amount is likely to be morally unacceptable and
politically dangerous
 THE QUINTILE AXIOM

What should governments do?

Maximize the per capita income?
Maximize the income of the poorest 20%
Both are arbitrary measures, but quintile axiom
focuses more on inequality and the poor
As normative target it is close to the Rawlsian
maximin strategy
 LEVELS OF GDP PER CAPITA, 1500-
1998, (IN 1990 PPP DOLLARS)
1500 1700 1913 1998
U.S.A 400 527 5,301 27,331
Sweden 695 977 3,096 18,685
U.K. 714 1,250 4,921 18,714
Japan 500 570 1,387 20,413
India 550 550 673 1,746
China 600 600 552 3,117
Africa 400 400 585 1,368
Ratio of Richest 1.8:1 3.1:1 9.4:1 20:1
to Poorest
 QUINTILE INCOMES OF NATIONS, 2002
% of
Quintile
Per Capita Income
P.C.
Country Income Accruing to
Income
USD, PPP Poorest
USD , PPP
20%
(1) (2) (3)
Norway 36,690 9.6 17,611
USA 36,110 5.4 9,750
Japan 27,380 10.6 14,511
Korea, South 16,960 7.9 6,699
South Africa 9,810 2.0 981

Contd
 QUINTILE INCOMES OF NATIONS, 2002
Country (1) (2) (3)
Malaysia 8,500 4.4 1,870

Russian Federation 8,080 4.9 1,980

Peru 4,880 2.9 708

China 4,520 4.7 1,062
Guatemala 4,030 2.6 524
India 2,650 8.9 1,179
Bangladesh 1,770 9.0 797
Sierra Leone 500 1.1 28
 One measure of aggregate welfare due to Sen is
(1-G) where is the per capita income of the
economy and G the Gini coefficient.
The quintile axiom is similar to this measure, except
that it is weighted by the per capita income of the
poorest 20%
A SIMPLE ILLUSTRATIVE MODEL - KAUSHIK
BASU
1 country: 2 kinds of people, productive p, and
unproductive u. The unproductive live off the
productive, in the sense that income created by the
productive creates a positive externality for the
unproductive.
Labour is the only factor of production
The economy is closed, and the government pursues
a tax-subsidy policy to affect income re-distribution
 Total population is n = p + u
n>0 and p, u are non-negative
h is effort with h [0,1] and t is the proportional
income tax rate t [0,1]
The effort h is negatively related to t
h = (1-t)
 Y = Ah with A>0, Y is the pre-tax income of each
productive person
y = ah with A>a>0 is the pre-tax income of each
unproductive person
Post tax-subsidy incomes are :
Y*(t) = (1-t)Y and y*(t) = y + ptY/u
Using Y and y and h from above we get:
Y*(t) = (1-t)2A
y*(t) = (1-t)(a + pAt/u)
 dY*/dt = -2(1-t)
Since (1-t) cannot be negative with t lying between
0 and 1, the slope of Y*(t) is negative and declining
as t increase. Hence the blue locus in the diagram
dy*/dt = pA/u 2pAt/u a = 0
Second derivative is negative. Hence solving the
above for t: t=1/2 au/2pA is maximum
Hence the red locus in the diagram
 Economic intuition behind falling Y*(t):
As t rises the post tax income of the productive
people fall, disposable income falling, and falling
effort h, reinforcing one another
Economic intuition behind shape of y*(t):
As t rises tax collection rises and hence subsidy
rises. But as t rises h falls, hence y falls. At some
time the tax collection itself will fall as productive
people work less intensively. Hence subsidy also
falls. After a sufficiently high t (t*), the fall in h and
the fall in subsidy both ensure that y*(t) falls
 A POSSIBLE CONFIGURATION
Incomes

At t Y*(t) = y*(t)
Y*(t*)

Y*(t)
As long as
y*(t*) t<t the
government
y*(t) will focus
on u.
a Rawlsian
Objective

t
0 1
t* t
 The government desires to maximize the income of
the lowest fifth of the population
Quintile Axiom: u/n 1/5 and p/n 1/5
t* = arg max y*(t) and in the diagram
t* is < t
Government objective implies:
Max. with respect to t min.{ y*(t), Y*(t) }
The solution value is t*
 Solve for t* and t
t* = 1/2 - au/2pA
t = u(A - a) / A(u + p)
Consider a special numerical example:
a=1 A=4 and u/p=2
then it can be seen t* = 1/4 and t = 1/2
Y*(t*) = 9/4 and y*(t*) = 9/8
 If the government wants to pursue the quintile
axiom objective, then
some people are twice as rich as other people
but this is also the poverty minimizing level of
inequality
Solution:
Y* = 9/4 and y = 9/8
 If the government wants to follow an objective of
egalitarianism in the strict sense of zero inequality:
then Y*(t) = y*(t) = 1
by choosing a t=1/2 that is a tax rate of 50%
everybody is poorer, including the poor
 If government wishes to maximize per capita
income:
Max. w.r.t. t [Y*(t)p + y*(t)u] / (u+p)
taking population to be constant,
Max w.r.t. t (1-t)2Ap + (1 - t )(au + Atp)
Note in the above expression, as t falls, per
capita income rises
so solution is to set t=0 that is the absence of tax
Solution: Y = 4 and y = 1
Incomes Maximize No inequality Maximize
of p and u Q t=1/2 Per Capita
income Income
t= t=0
Y* 9/4 1 4

y* 9/8 1 1
 OPEN ECONOMY
MIGRATION OF LABOUR ALLOWED

Labour is the only factor of production.

Workers choose the maximum post tax-subsidy
income
Each government can control migration
If all countries have the same tax-subsidy rates then
migration is 0, everybody stays where they are
 If boundaries were exogenously closed and each
government pursued the quintile axiom objective,
then each government would choose a t* of 25%
If migration was allowed then t* =1/4 can no longer
an equilibrium
To see this, suppose one government unilaterally
reduced t, then all p from all other countries will
want to migrate to the tax reducing country
 In the face of this demand for migration, the
government allows some p to come in, but does not
allow a single u to enter
In that nation income goes up because it has more
p. Also, note that subsidy per u goes up, and hence
the country is unambiguously better off
So all countries will want to reduce t
 It can be seen that there does not exist any t>0 that
can be a Nash equilibrium
Therefore t = 0 is the only equilibrium with Y=4 and
y=1
Note, despite each governments objective to follow
the quintile axiom objective, it APPEARS that each
is following the maximization of per capita income
with no concern for poverty or inequality
Globalization increases inequality!
 Kaushik Basu calls this real tax competition
Globalization compromises a nations policy efficacy
Each government wants t>0 therefore t=1/4 is
preferred to t=0
Yet, all are compelled to have t=0
Global coordination of national policies required?
 A GLOBAL SOLUTION?
Policy Options: Kaushik Basu discusses one method
by which all workers have a stake in all corporate
profits
Inter-country transfers would be required.
In the labour market there are winners and losers in
all countries
The pattern is not country specific
It is capital vis--vis labour (class conflict to class
compromise)
Institutions of global governance
Can you think of any workable global tax-transfer
schemes or safety nets to address this problem?