OPHI Oxford Poverty & Human Development Initiative Department - - PowerPoint PPT Presentation

OPHI

Oxford Poverty & Human Development Initiative Department of International Development Queen Elizabeth House, University of Oxford hi k www.ophi.org.uk

Multidimensional Poverty from OPHI’s Multidimensional Poverty from OPHI’s perspective S bi Alki J F Sabina Alkire & James Foster CONEVAL, 10 December 2009

Intellectual heritage

• Amartya Sen:

Poverty is multidimensional “There are good reasons for seeing poverty as a deprivation of basic capabilities, rather than merely as low income.” Poverty Measurement: Identification Aggregation (Sen 1976) Aggregation (Sen 1976)

Alkire & Foster 2007

• We constructed a multidimensional poverty

methodology including: methodology including:

– Who is poor?

• Based on Breadth of deprivations

Based on Breadth of deprivations

– How much poverty?

• Headcount (H)

( )

• Breadth (A)
• Overall M0 = H * A
• The first measure (M0) is simple but powerful.
• Other methods reflect Depth Inequality

Other methods reflect Depth, Inequality.

Headcount Ratio (H)

‘Multidimensional headcount ratio’ –

• Very useful partial measure (F&S)

What does it show?

• Percentage of people who are poor.

What does headcount NOT show? What does headcount NOT show?

• Breadth of poverty

CONEVAL d id H i f CONEVAL desiderata – H cannot satisfy.

• Identify the contribution of each dimension
• Axiom

Breadth of Poverty (As)

Average Number of Deprivations among the poor poor.

• Very Useful Partial Measure

What does it show?

• Breadth

What does it not show?

• How many people are poor (headco nt ratio)
• How many people are poor (headcount ratio)

Measure of Intensity

• M0 = H * A

d d

• M0 = Headcount times Breadth
• M0 = Measure of ‘intensity’ (Coneval’s term)
• f poverty in the total population
• Can be broken into H and A
• Tracks poverty over space and time

Tracks poverty over space and time.

• Can be decomposed

A i

• Axioms

M0 is being adopted as MD measure

• Has been applied to data of 25+ countries with

more underway. y

• India

14 African Countries

• China

Brazil

• China

Brazil

• Chile

Mexico

• El Salvador

Uruguay

• Pakistan

Argentina g

• UK?!

M0 is being adopted as MD measure

• Has been presented very widely, in academic

meetings such as AEA, Royal Economic society, g , y y, WIDER; in institutions such as the World Bank and Asian Development Bank; to policy groups, & p ; p y g p , universities.

• Related applications have been used to measure

Related applications have been used to measure governance, child poverty, quality of education, as well as to target social programs. well as to target social programs.

M0 is being adopted as MD measure

• Was adopted by Bhutan as the basis of their

Gross National Happiness Index Gross National Happiness Index

GNH

M0 is being adopted as MD measure

• M0 is being considered at the moment by a

M0 is being considered at the moment by a number of other countries with OPHI.

• Mexico’s experience will be of great interest

OPHI will implement a Multidimensional Poverty Index M0 in over 100 developing and developed countries for the UNDP’s p

2010 Human Development Report

CONEVAL’s 10 methodological criteria:

• Fulfill legal and normative requirements
• Incorporate relevant indicators

Legal Data

• Be applied using information from INEGI
• Identify who in the population is poor

Off l i l f k id if h i Identification

• Offer an analytical framework to identify the regions

and groups who are the most deprived

• Id ntif th

ntrib ti n f h dim n i n t p rt with breadth Decomposable by dimension

• Identify the contribution of each dimension to poverty
• Be decomposable by different population groups
• Reveal the contribution of different administrative

by dimension Decomposable by group

• Reveal the contribution of different administrative

units to national poverty

• Create measures that can be compared across time

by group Comparable p

• Satisfy a set of axiomatic properties

p Rigorous

Calculating the ‘intensity’ of poverty E l L t’ 40% f l d th t i Example: Let’s say 40% of people are poor, and that in average they are deprived in 3.25 out of the 6 social dimensions.

M0 is the grey shaded area = 0.40 x (3.25/6) = 0.22 M0s is the grey shaded area 0.40 x (3.25/6) 0.22

=40%

M0 = 0.40 x (3.25/6)

H=

M0s 0.40 x (3.25/6)

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

M0 rises if a person becomes more deprived

T b j S Su ingreso tá

José Joaquín

Trabaja como jornalero Su casa no tiene agua entubada está por debajo de la línea de bienestar Terminó la primaria

POBREZA MODERADA POBREZA MODERADA

Fuente: estimaciones del CONEVAL con base en el MCS-ENIGH 2008.

M0 also rises if a person is more deprived

k

No I ncome Finished

Housing

José Joaquín

Works as a shepherd

No piped water to house I ncome is below poverty line Finished primary school

Housing now

• ver-

crowded

POBREZA MODERADA POBREZA MODERADA

Fuente: estimaciones del CONEVAL con base en el MCS-ENIGH 2008.

Unlike Headcount, M0 can be broken down by dimension

90 100

Contribución de cada indicador de carencia social a la intensidad de la pobreza multidimensional, México, 2008

70 80 90 40 50 60 10 20 30

Aguascalientes Baja California Baja California Sur Campeche Coahuila Colima Chiapas Chihuahua Distrito Federal Durango Guanajuato Guerrero Hidalgo Jalisco México Michoacán Morelos Nacional Nayarit Nuevo León Oaxaca Puebla Querétaro Quintana Roo San Luis Potosí Sinaloa Sonora Tabasco Tamaulipas Tlaxcala Veracruz Yucatán Zacatecas

Fuente: estimaciones del CONEVAL con base en el MCS 2008.

B

Rezago educativo Acceso a los servicios de salud Acceso a la seguridad social Calidad y espacios de la vivienda Servicios básicos en la vivienda Acceso a la alimentación

Fuente: estimaciones del CONEVAL con base en el MCS-ENIGH 2008.