Global Poverty and the
Multidimensional Poverty Index
(MPI) 2014
Sabina Alkire
STRIVE, November 2014
What is the Global MPI?
• The Global MPI is an internationally comparable index
of acute poverty for 100+ developing countries.
• It was co-designed by OPHI and UNDP’s HDRO,
and was launched in 2010 in the Human Development
Report (HDR), where it has been published since.
• OPHI updated the global MPI in 2011, 2013 and 2014.
• The MPI uses the Alkire-Foster methodology (2011)
• The MPI methodology is being adapted for national
poverty measures – using indicators and specifications
that best reflect each policy context.
MPI
METHODOLOGY
Data: Surveys (Global MPI 2014)Details in: Alkire, Conconi and Seth (2014)
Demographic & Health Surveys (DHS - 52)
Multiple Indicator Cluster Surveys (MICS - 34)
World Health Survey (WHS – 16)
Additionally we used 6 special surveys covering urban Argentina
(ENNyS), Brazil (PNDS), Mexico (ENSANUT), Morocco
(ENNVM/LSMS), Occupied Palestinian Territory (PAPFAM), and
South Africa (NIDS).
Constraints: Data are 2002-2013. Not all have precisely the same
indicators.
Dimensions, Weights, Indicators
Build a deprivation
score for each personNathalie faces multiple deprivations in health
and living standards
Identify who is poor
Global MPI: A person is multidimensionally poor if
they are deprived in 33% or more of the dimensions.
Nathalie’s deprivation score is 67%
Compute the MPI (Alkire-Foster)
The MPI is the product of two components:
1) Incidence ~ the percentage of people who
are poor, H.
2) Intensity ~ the average percentage of
dimensions in which poor people are
deprived A.
MPI = H× A
Alkire and Foster Journal of Public Economics 2011
MPI 2014 FINDINGS
10
Across 108 countries & 5.4 billion people
30% of people are poor
=1.6 billion people
Aggregates
use 2010
population
data
11
Headline results
12
Disaggregated Data
13
Composition of
Poverty
14
Composition by region
The MPI is like a high resolution lens…
Breakdown by Indicator & Region
The MPI is like a high resolution lens…
You can zoom in
The MPI is like a high resolution lens…
You can zoom in
and see more
Of the 1.6 billion MPI poor people, 29% live in Sub-
Saharan Africa, and 52% in South AsiaTotal Population
MPI Poor
East Asia & the Pacific,
34.7%
South Asia, 29.7%
Sub-Saharan Africa, 14.4%
Latin America & Caribbean,
9.5%
Europe & Central Asia,
7.5%
Arab countries, 4.2%
East Asia & the Pacific,
14.6%
South Asia, 52.0%
Sub-Saharan Africa, 28.8%
Latin America & Caribbean,
1.9%
Europe & Central Asia,
0.7%
Arab countries, 2.1%
Most MPI poor people (71%) live in
middle-income countries
2010 Population Data
High Income,
3.4%Low
Income, 13.3%
Lower Middle Income,
44.1%
Upper Middle Income,
39.2%
Total Population by Income Category High
Income, 0.2%
Low Income,
28.9%
Lower Middle Income,
58.3%
Upper Middle Income,
12.7%
MPI Poor Population
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Nig
erE
thio
pia
Mal
iB
urk
ina
Fas
oL
iber
iaG
uin
eaSom
alia
Buru
ndi
Cen
tral
Afr
ican
Rep
ublic
Guin
ea-B
issa
uSen
egal
Co
ngo
, D
emo
crat
ic R
epublic
of
the
Sie
rra
Leo
ne
Ben
inU
gan
da
Mo
zam
biq
ue
Rw
anda
Tim
or-
Les
teM
adag
asca
rM
alaw
iA
fgh
anis
tan
Tan
zan
ia, U
nit
ed R
epub
lic
of
Zam
bia
Ch
adM
auri
tan
iaG
amb
iaC
ote
d'I
vo
ire
Ind
iaY
emen
Ban
glad
esh
To
go
Hai
tiK
enya
Cam
ero
on
Cam
bo
dia
Nep
alP
akis
tan
Nig
eria
Co
ngo
, R
epub
lic
of
Nam
ibia
Zim
bab
we
Les
oth
oSao
To
me
and P
rin
cip
eL
ao P
eop
le's
Dem
ocr
atic
Rep
ub
lic
Gh
ana
Van
uat
uD
jib
outi
Bh
uta
nG
uat
emal
aB
olivia
, P
luri
nat
ion
al S
tate
of
Sw
azilan
dG
abo
nN
icar
agua
Ho
nd
ura
sM
on
golia
Ind
ones
iaP
hilip
pin
esP
arag
uay
Taj
ikis
tan
Ch
ina
Iraq
South
Afr
ica
Mo
rocc
oP
eru
Guya
na
Est
on
iaT
urk
eyE
gyp
tSuri
nam
eT
rinid
ad a
nd T
ob
ago
Syr
ian
Ara
b R
epub
licC
olo
mb
iaSri
Lan
ka
Aze
rbai
jan
Mal
div
esK
yrgy
zsta
nB
eliz
eD
om
inic
an R
epub
lic
Hun
gary
Cro
atia
Vie
t N
amC
zech
Rep
ub
licA
rgen
tin
aM
exic
oB
razi
lJo
rdan
Uzb
ekis
tan
Ecu
ado
rU
kra
ine
Mo
ldo
va,
Rep
ub
lic
of
Uru
guay
Th
aila
nd
Lat
via
Mo
nte
neg
roO
ccup
ied P
ales
tin
ian T
erri
tory
Alb
ania
Russ
ian
Fed
erat
ion
Tun
isia
Geo
rgia
Mac
edo
nia
, T
he
form
er Y
ugo
slav
Rep
ub
lic
of
Un
ited
Ara
b E
mir
ates
Bo
snia
an
d H
erze
go
vin
aA
rmen
iaK
azak
hst
anSer
bia
Bel
arus
Slo
vak
iaSlo
ven
ia
Per
cen
tage
of
the
Po
pula
tio
n
MPI and $1.25/day poverty rates
MPI Poor $1.25 a day
Incidence and Intensity by Country
Namibia
Brazil
Argentina
Indonesia
Guatemala
Ghana
Lao
Nigeria
Tajikistan
ZimbabweCambodia
Nepal
Bangladesh
Gambia
TanzaniaMalawi
Rwanda
Afghanistan
Mozambique
Congo DR
Benin
Burundi
Guinea-Bissau
Liberia
Somalia
Ethiopia Niger
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Ave
rage
Inte
nsi
ty o
f P
ove
rty
(A)
Percentage of People Considered Poor (H)
Poorest Countries, Highest MPI
China
India
The size of the bubbles
is a proportional
representation of the total
number of MPI poor in
each country
MPI varies subnationally too
Namibia
Brazil
Argentina
Indonesia
Guatemala
Ghana
Lao
Nigeria
Tajikistan
ZimbabweCambodia
Nepal
Bangladesh
Gambia
TanzaniaMalawi
Rwanda
Afghanistan
Mozambique
Congo DR
Benin
Burundi
Guinea-Bissau
Liberia
Somalia
Ethiopia Niger
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Ave
rage
Inte
nsi
ty o
f P
ove
rty
(A)
Percentage of People Considered Poor (H)
Poorest Countries, Highest MPI
High Income
Upper-Middle Income
Lower-Middle Income
Low Income
China
India
The size of the bubbles
is a proportional
representation of the total
number of MPI poor in
each country
23
30%
35%
40%
45%
50%
55%
60%
65%
70%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Ave
rage
Inte
nsi
ty o
f P
ove
rty
(A)
Percentage of People Considered Poor (H)
Nigeria
24
Abia
Adamawa
Anambra
Bauchi
Bayelsa
Benue
Borno
Cross River
Ebonyi
Edo
Enugu
FCT (Abuja)
Gombe
Imo
Jigawa
Kaduna
Kano
Katsina
Kebbi
Kogi
Kwara
Lagos
Nasarawa
Niger
OgunOsun
Oyo
Plateau
Sokoto
Taraba
Yobe
Zamfara
30%
35%
40%
45%
50%
55%
60%
65%
70%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Ave
rage
Inte
nsi
ty o
f P
ove
rty
(A)
Percentage of People Considered Poor (H)
Nigeria
MPI also varies greatly across subnational regions
within a country – e.g. Cameroon
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Ave
rage
Inte
nsi
ty o
f P
ove
rty
(A)
Percentage of People Considered Poor (H)
MPI also varies greatly across subnational regions
within a country – e.g. Cameroon
Incidence: 6.5% to 86.7%
Intensity: 36.4% to 62.3%
Cameroon
Adamaoua
Centre (Excluding Yaoundé)
Douala
Est
Extrême-Nord
Littoral (Excluding Douala)
Nord
Nord-Ouest
OuestSud
Sud-Ouest
Yaoundé
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Ave
rage
Inte
nsi
ty o
f P
ove
rty
(A)
Percentage of People Considered Poor (H)
MPI also varies greatly across subnational regions
within a country – e.g. Bangladesh
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Ave
rage
Inte
nsi
ty o
f P
ove
rty
(A)
Percentage of People Considered Poor (H)
MPI also varies greatly across subnational regions
within a country – e.g. BangladeshIncidence: 44.5% to 61.9%
Intensity: 44.5% to 53%
Bangladesh Chittagong
Dhaka
Khulna
Rajshahi
Rangpur
Sylhet
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Ave
rage
Inte
nsi
ty o
f P
ove
rty
(A)
Percentage of People Considered Poor (H)
CHANGES OVER TIME
- Coverage:
- 34 countries, from every region; both LICS and MICS
- 338 sub-national regions
- covering 2.5 billion people
- Rigorously comparable MPI values, with std errors/inference
- Annualized changes compared across countries
30
And how has MPI gone down?
Nepal 2006
Nepal 2011
30%
35%
40%
45%
50%
55%
60%
65%
70%
75%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Ave
rag
e I
nte
nsi
ty o
f P
ove
rty (
A)
Incidence - Percentage of MPI Poor People (H)
How MPI decreased in Nepal 2006-11Reduction in both H and A
Decomposition By Region
(or social group) – shows H, A, inequalities
Subgroup Decompositions
Subgroup Decompositions (regional)
35
How did MPI go
down?
Monitor each
indicator
Indicator Changes by region (Nepal)
-0.11
-0.09
-0.07
-0.05
-0.03
-0.01
0.01
0.03
An
nu
ali
zed
Ab
solu
te C
ha
ng
e
in p
rop
ort
ion
wh
o i
s p
oo
r an
d d
ep
rive
d i
n..
.
Nutrition
Child MortalityYears of SchoolingAttendance
Cooking FuelSanitation
Water
Electricity
Floor
Assets
Disaggregating by subnational region
- Poverty significantly decreased in 208 of the 338 subnational
regions, which house 78% of the poor.
- Ten countries reduced all MPI indicators significantly:
Bolivia, Cambodia, Colombia, the Dominican Republic,
Gabon, India, Indonesia, Mozambique, Nepal, and Rwanda;
- Eight countries reduced poverty in all subnational regions:
Bangladesh (2007-11), Bolivia, Gabon, Ghana, Malawi,
Mozambique, Niger and Rwanda.
- In nine countries the poorest region reduced poverty the
most: Bangladesh (2007-2011), Bolivia, Colombia, Egypt,
Kenya, Malawi, Mozambique, Namibia and Niger.
Adja
Bariba
Dendi
Fon
Yoa and Lopka
Bétamaribe
Peulh
Yoruba
-0.035
-0.030
-0.025
-0.020
-0.015
-0.010
-0.005
0.000
0.005
0.010
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80
An
nu
al
Ab
solu
te C
han
ge i
n M
PI T
Multidimension Poverty Index (MPIT) at initial year
Reduction in
MPI
Size of bubble is proportional to
the number of poor in first year of
the comparison.
Disaggregating by ethnic group - Benin
.
Adja
Bariba
Dendi
Fon
Yoa and Lopka
Bétamaribe
Peulh
Yoruba
-0.035
-0.030
-0.025
-0.020
-0.015
-0.010
-0.005
0.000
0.005
0.010
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80
An
nu
al
Ab
solu
te C
han
ge i
n M
PI T
Multidimension Poverty Index (MPIT) at initial year
Reduction in
MPI
Size of bubble is proportional to
the number of poor in first year of
the comparison.
Disaggregating by ethnic group - Benin
.
Kalenjin
KambaKikuyu
Kisii
Luhya
Luo
Meru
Mijikenda/Swahili
Somali
-0.035
-0.030
-0.025
-0.020
-0.015
-0.010
-0.005
0.000
0.005
0.010
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80
An
nu
al
Ab
solu
te C
han
ge i
n M
PI T
Multidimension Poverty Index (MPIT) at initial year
Reduction in
MPI
Size of bubble is proportional to
the number of poor in first year of
the comparison.
Disaggregating by ethnic group - Kenya
Kalenjin
KambaKikuyu
Kisii
Luhya
Luo
Meru
Mijikenda/Swahili
Somali
-0.035
-0.030
-0.025
-0.020
-0.015
-0.010
-0.005
0.000
0.005
0.010
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80
An
nu
al
Ab
solu
te C
han
ge i
n M
PI T
Multidimension Poverty Index (MPIT) at initial year
Reduction in
MPI
Size of bubble is proportional to
the number of poor in first year of
the comparison.
Disaggregating by ethnic group - Kenya
Poorest ethnic
group reduced
MPI the fastest.
They are catching
up.
MPI vs. Income poverty
-8
-7
-6
-5
-4
-3
-2
-1
0
1
2
MPI Incidence $1.25 Incidence
MPI vs. Income poverty
-8
-7
-6
-5
-4
-3
-2
-1
0
1
2
MPI Incidence $1.25 Incidence
If progress was only measured by reducing income
poverty, the tremendous gains of Rwanda, Ghana,
and Bolivia would have been less visible.
Other Global MPI 2014 Studies
- Rural-Urban poverty: MPI is mainly rural
- Inequality among the poor: How to measure and track it.
- Destitution: a subset of the MPI poor are destitute; who are
they, and how have they changed?
NATIONAL MPIs
MPI: Two kinds ~ both useful
Internationally comparable:- Like $1.25/day poverty measures
- We propose an MPI 2015+, that reflects the SDGs
- A better measure will use better data
- MPPN developed SDG-related survey modules for piloting
MPI: Two kinds ~ both useful
National MPIs: - Vital for policy
- Reflect National Definitions
- Official measures in Mexico, Colombia, Philippines, Bhutan
MPI: Two kinds ~ both useful
Internationally comparable:Example: The Global MPI estimated and analysed by OPHI and published by UNDP’s HDRO can be compared across 108 countries. Facilitates ‘lessons learned’ across countries.
- Like $1.25/day and $2/day poverty measures & MDGs
- Useful for policy analysis, but limited national ownership
Context-Specific: Example: National MPIs reflect national contexts and priorities. They guide policies like targeting and allocation and monitor changes
- Like National income poverty measures
- Useful for policy but can’t be compared internationally
MPI in National Settings
Official National MPIs
Colombia
Mexico
Bhutan
Philippines
Other national applications underway.
Some Policy Applications of MPIs:
• Track poverty over time (official statistics)
• Compare poverty by region, ethnicity, rural/urban
• Monitor indicator changes (measure to manage)
• Coordinate different policy actors
• Target the marginalized
• Geographic targeting
• Household beneficiaries
• Evaluate policy impacts
The Multidimensional Poverty Peer Network
Launched in June 2013 at University of Oxford with:
• President Santos of Colombia
• Ministers from 16 countries in person
• A lecture from Professor Amartya Sen
• Aim: South-South support for National MPIs & an improved Global MPI 2015+ engaging research.
MPPN has 30 countries plus 10 international
agencies in 2014
Supported by the German Federal Ministry for Economic
Cooperation and Development (BMZ)
The MPPN agenda
- Support National MPIs that inform powerful policies
- Suggest an improved Global MPI 2015+ that reflects the SDGs (acute & moderate poverty versions)
- Strengthen the data sources for MPI metrics
Why Measure? Action ‘with vigour’Coordination ~ Policy Design ~ Monitoring ~ Targeting ~ Allocation
“Positive changes have often occurred and
yielded some liberation when the remedying
of ailments has been sought actively and
pursued with vigour”
Jean Dreze and Amartya Sen India: An Uncertain Glory 2013
www.ophi.org.uk/
multidimensional-poverty-index
People and Stories
Policy Briefings
Infographics
Interactive DataBank with Maps
Academic Paper Drafts for Comment
Data Tables
Thanks! From the OPHI-MPI Team
OPHI Research Team: Sabina Alkire (Director), James Foster (Research Fellow), John Hammock (Co-Founder and
Research Associate), Adriana Conconi (coordination MPI 2013/14), José Manuel Roche (coordination MPI 2011, 2013), Maria Emma
Santos (coordination MPI 2010), Mauricio Apablaza, Paola Ballon, Mihika Chatterjee, Bouba Housseini, China Mills, Suman Seth, Ana
Vaz, Gaston Yalonetzky, Diego Zavaleta.
Data analysts and MPI calculation: Ivan Gonzalez de Alba, Aparna John, Usha Kanagaratnam, Saite Lu, Maria
Mancilla Garcia, Christian Oldiges, Felipe Roa-Clavijo and Quang Van Tran.
Special contributions: Ana Vaz (Quality Checks and Changes over Time), Alejandro Olayo-Mendez (Research Assistance),
Putu Natih, John Hammock, Bouba Housseini and Vanita Leah Falao (new Ground Reality Check field material), Suman Seth (design and
programming for Inequality among the Poor and Destitution analyses).
Communication Team: Paddy Coulter (Director of Communications), Emma Feeney (Research Communications Officer),
Heidi Fletcher (Web Manager), Moizza B Sarwar (Research Communications Assistant), and Cameron Thibos (Design Assistant).
Administrative Support: Laura O’Mahony (OPHI Project Coordinator), Natasha Francis (OPHI Project Assistant)
OPHI estimate the MPI, the UNDP Human Development Reports also publish it
and we are grateful to our colleagues in HDRO for their support.