Frances Stewart

8/8/2019 Frances Stewart

1/32

1

Alternative approaches to

conceptualising and measuringpovertyFrances Stewart


2/32

2

Overwhelming agreement that poverty

reduction is THE objective of development

World Bank: Poverty reduction is the benchmark

against which we should be judged. (Preston,

carried on by Wolfensohn PRSPs etc.) Millennium goals: halving world poverty by 2015.

DFID all projects and research directed towards

poverty reduction.

Wasnt always so

(growth/BN/stabilisation/adjustment)


3/32

3

But do we know what poverty

means? And if not does it matter? Aim to consider meaning of poverty, looking at

four alternative approaches to conceptualisation

and measurement. Consider whether differences and problems in

each matter in terms of measurement, targetingand policies.

Draw heavily on QEH project (Barbara Harriss-White, Caterina Ruggeri Laderchi, Ruhi Saith,Susana Franco, and partners in India, Dr. Shariff(India) and Dr.Vasquez (Peru).


4/32

4

Approaches to defining poverty Seems obvious: poverty is deprivation.

Yet deprivation of WHAT?

By how much?

Over what time period?

Whose? A country, region, village, family

individual. 4 approaches take different views on these issues,

and each has problems of interpretation.


5/32

5

Four approaches Monetary income: conventional definition

most commonly applied.

Capability approach (Sen; HDR)

Social exclusion: adopted in Europe,

spreading to developing countries.

Participatory approaches (Chambers)


6/32

6

Some common problems

Space:

money income/utility; what people can do/be; whatpeople do/are; how they perceive their condition?

Categories of life:

material/spiritual/cultural/political.

Universality same definition for all societies? Monetary and SE

devised for developed countries.Capabilities/participatory for developing. Can they betransferred?

Objective/subjective

Whose judgement? External/internal

Unit: individual/family/group

Multidimensionality: how to add up.

Time


7/32

7

The poverty line Absolute/relative. Related to space question.

Is a relative line non-arbitrary? Relative to whom?

What is justification for absolute line: Is there a

discontinuity between poor/non-poor in behaviour

or outcomes? Or do we impose it? Nutrition

approach/ BN approach/human rights approach.

The depth of poverty and different measures one

line or two?


8/32

8

The monetary approach Utilitarian basis: poverty is deprivation of money

income = utility.

In principle includes all income in money metric.In practice omits social goods. Imputation ofsubsistence.

Consumption v. income.

Poverty lines relative or nutrition based. National v. international lines.

Poverty attributed to individuals but measured byfamily income/consumption.


9/32

9

Some problems Translating household income into individual

poverty: Allocation of income within the family

Needs of different family members

Economies of scale of consumption.

Selecting a poverty line(s) Ravallion: the minimum cost of the poverty level of utility.

Generally nutrition based.

Szekeley et al : poverty rate varies between 13% and 66% in 17Latin American countries, according calculation of equivalence scales;

assumptions of economies of scale; methods for treating missing; formisreporting.

Measures arbitrary essentially depending ondecisions of those measuring poverty.


10/32


11/32

11

Capability approach Sens capability approach to development the

objective is to enhance individuals capabilities to

be or do different things. Poverty is the failure of some basic capability to

function

basic capabilities are intended to separate out the

ability to satisfy certain elementary and cruciallyimportant functionings

Monetary income is a means only (and only one ofseveral).


12/32

12

Problems in interpreting capability

approach

What are basic capabilities? Not defined by Sen, though education, health etc, often

mentioned. Deliberate to allow for differences across cultures.

Numerous attempts to define them: Nussbaum (overlappingconsensus); Doyal and Gough (avoid serious harm);

Can one measure capabilities? In practice functionings.I.e. similar to BN approach.

Cut-off rate? Controversy, e.g. over nutrition; educationlevel. Arbitrary.

HPI1: life expectancy below 40, adult illiteracy, and average of notusing improved water sources and under five mortality.

HP2: life expectancy below 60, adult functional illiteracy, the long-term unemployment rate, and the population below an incomepoverty line of 50% of median disposable household income.

Adding up issue


13/32

13

Social exclusion Mon and cap. Are individualistic; dont look at

group situation; nor dynamics.

SE is intended to capture structural features ofpoverty.

Process of exclusion.

Atkinson: relativity; agency; and dynamics

Developed in Europe: includes unemployment;lack of social insurance; lack of housing; lowmonetary incomes.


14/32

14

Applying SE to developing

countries Relative to whose standards?

Developed country standards are not

normal. The norm may be one of

deprivation.

Exclusion can be part of inclusion (caste

system).

Inevitably society specific and arbitrary.


15/32

15

Practical applications

India: exclusion from health services, education,housing, water supply, sanitation and social security.Very large numbers socially excluded.

Venezuela: study first defines social and political

rights and then defines SE as not having these rights. Tanzania study - certain very poor urban occupations

and the rural landless excluded.

Studies of Cameroon and Thailand define ethnic

minorities as excluded; plus poorly educated farmers,informal sector workers and the homeless.

Peru: exclusion from political, cultural and economic

processes.


16/32

16

Advantages and disadvantages of

SE approach Disadvantages

Problems of interpretation of dimensions

Identifying breaks Adding up

Advantages

Points to groups and structural characteristics Agency including those who exclude

Redistributive

Begins to look at causes


17/32

17

Participatory Mon, cap., SE all impose EXTERNAL values.

Participatory approach gets poor themselves to

define what is poverty and who are poor.

Uses wealth of methodologies introduced by

Chambers -- mapping; modelling; seasonal

calendars; wealth and well-being ranking.

Large variety of methods can be used flexibly,

according to the situation.


18/32

18

Some elements in a Zambian Participatory

assessment

ISSUES METHODS

Perceptions and indicators of wealth,

well being and poverty

Wealth/Well-being grouping

Social mapping

Semi-structured mapping

Communal assets of rural communities Resource mapping

Focus group discussion

Institutional diagramming

Assets of rural households Wealth ranking/grouping

Livelihood analysis

Coping strategies in times of crisis

Livelihood analysis Semi-structured interviews

Ranking exercises

Community based support mechanisms Institutional mapping

Semi-structured interviews


19/32

19

Advantages/disadvantages

Picks up peoples ownperceptions: but could besubject to Sens two neglects.

Could provide answers to value judgements neededin other approaches

Samples rarely representative. Can omit poorest. Group often define others as poor.

Underrepresent women in group and in voice in group.

Not representative of country as a whole.

Group can be divided (women/men): who decides?

Organiser imposes structure on questions andreplies.


20/32

20

Clearly no unique and objective

of poverty: does it matter? Partly depends on whether approaches in

practice pick up same people. Empirical

issue.

But also approach tends to dictate policy

approach and targetting methods.


21/32

2

1

Empirical question: international

comparisons At macro level clear that monetary and capability

poverty give different country rankings Precise

differences depending on interpretations of each. The Spearman rank correlation coefficient

between national poverty lines and the HPI is

0.54;

the correlation with the internationally set lines of

those falling below $1 or or $2 a day is much

higher at 0.707 and 0.779 respectively.


22/32

22

Graph 1

Relationship between HPI and monetary poverty

(international, $1 a day)

0

10

20

30

40

50

60

70

80

0 10 20 30 40 50 60 70

HPI

%p

opu

lationbelow

$1p

erday


23/32

23

Gr 2

l ti i t v rt I x

t r v rt ( ti l v rt li )

0

10

20

30

40

50

60

70

80

0 10 20 30 40 50 60 70

Human

v rt Inde

x (HP

I)

el

natinal

vertl

ine


24/32

24

Empirical comparisons: national and micro

studies, India and Peru

Broad questions:

Do levels of monetary poverty parallel levelsof poverty in other dimensions?

Are the findings an artefact of the monetarypoverty threshold used?

Use of national surveys and specificmicro surveys


25/32


26/32

26

Individual overlaps in types of poverty

Capability poverty

measured as

Education Nutrition/health

child adults child adults

% of CA-poor not

in monetary

poverty

India 43 60 53 63

Peru 32 37 21 55

% of monetary poor

not CA-poor

India 65 38 53 91

Peru 93 73 66 94


27/32

27

Altering monetary threshholds would not

greatly alter the results.

India: 33% of the richest tenth of the population were illiterate (compared

with 64% among the lowest decile).

Among those with incomes as high as the 7th monetary decile more

than 50%

are poor in either education or health. Proportion of health poor in the highest decile is quite similar to that

in the lowest decile.

Peru: 12% of top decile are education poor among adults, and 5% among

children compared with 32% in the lowest decile for adults and9% for children.

incidence of child undernutrition is 5% for the top decile of moneyincomes compared with 9% for the lowest decile.

Changing the cut-off line for monetary poverty would not

eliminate poor overlaps in either country.


28/32

28

Participatory measure of poverty

Peru:

Rural wealth ranking : Half pop. classified differently.

15% of population mon. non-poor identified as poor 34% of population identified as mon. poor but wealth non-

poor.

Urban wealth ranking: 47% of pop classified

differently.

31% wealth poor were mon non-poor

16% mon poor classifed as wealth non-poor.


29/32

29

Focus group discussions

Characteristics of poverty:

India, both urban and rural groups considered

completion of secondary school as the minimum

desirable level of education; and that a chronic health

condition, or a serious one requiring hospitalisation,

should be considered as ill-health.

Adopting these criteria would considerably increase CA-

poverty compared with our estimates.

In Peru, multiple dimensions of poverty were notedincluding capabilities, material needs, and insecurity.

Were not able to identify SE dimensions. Social boycott

suggested in India (mixed marriage/leprosy)


30/32

30

It does matter! So approach and measurement does make a

difference empirically (supported by work

in Chile and Vietnam)

Implications for targeting and policy


31/32

31

Targeting criteria Major errors of both types if use one criteria

(monetary) to reach other types of poverty.

Targeting errors apply at macro as well as

micro level. I.e. aid distribution.

Need to decide what type of poverty

reduction is objective, therefore who ispoor, and target accordingly.


32/32

32

Policy implications Concept of poverty important determinant of policy

approach.

Mon poverty suggests solution is generation of moneyincomes. Ca approach that this is a means only.

Ca approach more emphasis on social goods provision.

Both concerned with absolute poverty, so policy responseis to raise incomes/ca generally (growth is good for thepoor)

But SE suggests redistribution essential aspect. Unlikely

that growth can EVER eliminate SE withoutredistribution.

Mon and Ca individualistic (though CA more women-focussed). SE and participatory focus on groupcharacteristics

.

Frances Stewart

Documents