Poverty: Identification and

Aggregation
Source: Sen, A. (1982). Poverty and famines: an essay on entitlement
and deprivation. Oxford university press.
Commodities and characteristics
• Should basic needs be defined in terms of commodities, or in terms of
'characteristics’ in order to identify poor.
• Wheat, rice, potatoes, etc., are commodities
• calories, protein, vitamins, etc., are characteristics of these
commodities that the consumers seek when they consume them.
• calories are necessary for survival not wheat or rice because calories
can be sourced from other commodities too.
Hybridity of basic/minimum needs vector
• Multiplicity of sources of characteristics is, however, not common across the
needs
• Calories have multiple sources but not shelter.
• Literacy is solely sourced from elementary schooling but there may be other sources
• Often 'basic' or 'minimum' needs are specified in terms of a hybrid vector
• amounts of calories, proteins, housing, schools, hospital beds
• Intermediate cases arise when societies have strong preference of sourcing a
characteristics from a particular commodity
• Calories have to be sourced from rice in many Indian States ‘calories from rice‘
• Community may feel some one is deprived because he/she is unable to access
certain minimum level (carbohydrate) calories from Rice.
• Such conceptual may have important role to play in policy design. (PDS commodities
by states)
Commodity or characteristics
• Dietary habits of population do change but they domonstrate strong ‘staying
power’
• Role of knowledge in reforming/changing ideas of feasible diets may in fact be an
important part of nutritional planning.
• 'I have never tasted it because I don't like it'
• For intercommunity comparisons of poverty, the contrast of characteristics and
commodities may turn out to be significant.
• For example, the ranking of rural living standards in different states in India
changes significantly when the basis of comparison is shifted from command over
commodities to command over characteristics such as calories and protein.
• Due to strong preferences/tastes for certain commodities conversion of these basic needs
into minimum cost diets depends not only on prices but also on consumption habits.
The direct method versus the income method
• Direct Method:
• Identify set of people whose actual consumption baskets happen to leave some basic need
unsatisfied no need of poverty line in this case.
• Income method:
• Step1: calculate the minimum income at which all the specified minimum needs are satisfied.
• Stept2: identify those whose actual incomes fall below that poverty line
• Income method focus on persons ability to meet basic needs
• It controls for individual idiosyncrasies without upsetting the notion of poverty based on
deprivation.
• A rich ascetic who fasts on his expensive bed of nails is poor as per direct method but not as per
income method.
• Another Flip: ability to meet minimum needs without being bothered by tastes, then
one would, of course, set up a cost-minimizing programming problem and simply check
whether someone's income falls short of that minimum cost solution
The direct method versus the income method
• taste constraints that apply broadly to the entire community and
those that essentially reflect individual idiosyncrasies is usually easily
identifiable.
Household size and Calories Per capita (log)

Calorie Engel curves for different

sized households (all adult),
1983, India

Source: Li & Eli (2012)

Household size and Calories Per capita (log)

• Economies of scale in food purchase (bulk discounting) and in

preparation, wastage, collective household models, a large price
elasticity for food, measurement error, intra-household inequality and
what they call “calorie overheads.”
Family Size and Adult equivalence
• The natural unit for data for consumption behaviour is household and
not individual
• Per capita income/consumption Method has problems.
• It Overlooks the economies of large scale that operate for many items of
consumption
• Children's needs may be quite different from those of adults

• Alternative way is to convert each family into a certain number of

'equivalent adults' by the use of some 'equivalence scale',
Family Size and Adult equivalence
• Most conversion rules have problem of arbitrariness.

• Intra-household disparity: maldistribution within the family is also an

important issue.
• One of bases for deriving appropriate equivalence of needs is
nutritional requirement.
no. of consumer units assigned to a person in India Mainhas(1980)
age in completed years
<1 1-3 4-6 7-9 10-12 13-15 16-19 20-39   40-49 50-59 60-69 70+
sex
male 0.43 0.54 0.72 0.87 1.03 0.97 1.02 1.00 0.95 0.90 0.80 0.70

female 0.43 0.54 0.72 0.87 0.93 0.80 0.75 0.71   0.68 0.64 0.51 0.50
Family Size and Adult equivalence
• A second approach is to examine how the people involved regard the
equivalence question themselves.
• How much income they think is needed to make a larger family have the same standard of well-
being as a smaller one

• examine the actual consumption behaviour of families of different size.

• treat some aspect of this behaviour as an indicator of welfare.
• For example, the fraction spent on food has been treated as an indicator of
poverty:
• two families of different size are regarded as having 'equivalent' incomes when
they spend the same proportion of their incomes on food.
• put the same weight on each household, irrespective of size;
• put the same weight on each person, irrespective of the size of the
family to whom they belong; and
• put a weight on the basis of number of equivalent adults in it.
Poverty gaps and relative deprivation
• income short-fall of a person whose income is less than the poverty-line income
can be called his 'income gap'.
• does it make a difference whether or not a person's short-fall is unusually large
compared with those of others?
• Concern over relative deprivation even after a set of minimum needs and a
poverty line have been fixed.
• Consider income transfer from a poor person to another who is richer but still
below the poverty line
• Such a transfer will increase the absolute short-fall of the first person by exactly
the same amount by which the absolute short-fall of person 2 will be reduced.
• magnitudes of absolute deprivation may have to be supplemented by
considerations of relative deprivation.
Poverty gaps and relative deprivation
• Most common measure:
• head-count measure H, given by the proportion of the total population that happens to
be identified as poor
• Another measure that has had a fair amount of currency is the so-called
'poverty gap',
• the aggregate short-fall of income of all the poor from the specified poverty line

• The index can be normalized by being expressed as the percentage short-fall of the
average income of the poor from the poverty line.

• The income-gap ratio I is completely insensitive to transfers of income among the

poor so long as nobody crosses the poverty line by such transfers.
Poverty gaps and relative deprivation
• Income Gap Ratio (IGR) also pays no attention to the number or
proportion of poor people below the poverty line,
• It primarily focuses on the aggregate short-fall, no matter how it is
distributed and among how many.
• HCR pays no attention whatever to the extent of income short-fall of
those who lie below the poverty line.
• It is insensitive to whether someone is just below the line or very far
from it, in acute misery and hunger.
• A transfer of income from a poor person to one who is richer can
never increase the poverty measure HCR -- surely a perverse feature.
Poverty gaps and relative deprivation
• If a unit of income is transferred from a person below the poverty line
to someone who is richer but who still is (and remains) below the
poverty line, then both the measures H and I will remain completely
unaffected.
• Hence any 'combined' measure the two will also remain insensitive to
such a changes or income inequality among the poor in general.
• Obviously, at conceptual level, undeniably aggregate poverty has
increased as a consequence of this transfer in terms of relative
deprivation.
Poverty gaps and relative deprivation
• There is, however, one case in which H & I together may do the job.
• When income of every poor is identical – obviously not a likely event.
• poverty measure P to be thought of as a weighted sum of the
shortfalls of all poor
• Rank of poor in income ladder as weight as weight
• If we think of destitution of poor as some kind of disutility to society
then the weights should be derived as monotone(increasing function)
of rank.
P=H {I + ( 1-I) G}
Poverty gaps and relative deprivation
• A generalised version