Khalid Abu Ismail - ESCWA
Racha Ramadan - Cairo University
ERF 24th Annual Conference
The New Normal in the Global Economy: Challenges & Prospects for MENA
July 8-10, 2018
Cairo, Egypt
4. I-The four basic questions
1. What do we know?
2. Why are we doing this report and who are our
audience?
3. What kind of inequality are we interested in and
why?
4. Inequality between whom?
4
5. Yemen
Syria
Zimbabwe
-15.0
-10.0
-5.0
0.0
5.0
10.0
15.0
-10.0 -5.0 0.0 5.0 10.0 15.0 20.0
AverageGDPgrowth2010-2017
Average GDP growth 2000-2009
Nepal
Comorros
20.0
25.0
30.0
35.0
40.0
45.0
50.0
55.0
60.0
65.0
70.0
20.0 30.0 40.0 50.0 60.0 70.0
Gini2010-2017
Gini Index 2000-2010
Q1: What do we know?
Stylized fact 1: Income inequality is structural
5
6. 0
10
20
30
40
50
60
1982
1987
1992
1997
2002
2007
2012
2017
0 20 40 60
Comoros
Djibouti
Egypt
Iraq
Jordan
Lebanon
Mauritania
West Bank…
Tunisia
Yemen, Rep.
Average
Gini Index
2010-2017 2000-2009
Stylized fact 2: Moderate income inequality (and stagnating)
Q1: What do we know?
6
7. Q1: What do we know?
Stylized fact 3: Arab countries are more income-unequal than we think
Inequality in the region is world highest when the income of the top 1% is
factored in (does this make sense?)
22% 23% 23% 25% 26% 26% 26% 27% 27% 28% 32% 32%
26%
60%
0%
10%
20%
30%
40%
50%
60%
70%
Iraq
Algeria
Lebanon
Egypt
Jordan
Tunisia
WestBankandGaza
Syria
Yemen
Iran
Djibouti
Morocco
Average
Worldinequality…
Shareoftopdeclie
7
8. Stylized fact 4: Low extreme poverty but region crosses the
global average at $7.5 per day PL
Q1: What do we know?
Source: World Bank Povcal.net
0.6 2.5
10.9
26.1
37.1
48.7
56.4
62.2
66.5
0
10
20
30
40
50
60
70
80
90
100
0.5 1 1.5 1.9 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10
HeadcountPovertyRates
Poverty Lines in 2011 PPP
East Asia and Pacific Europe and Central Asia
Latin America and the Caribbean Middle East and North Africa
South Asia Sub-Saharan Africa
World Total
8
9. Stylized fact 5: Highest Vulnerability to income poverty
Q1: What do we know?
Source: World Bank Povcal.net
East
Asia
and
Pacific
Europe
and
Central
Asia
Latin
Americ
a and
the
Caribbe
an
Middle
East
and
North
Africa
South
Asia
Sub-
Saharan
Africa
$3.5/$1.9 (1990) 1.4 4.1 2.2 5.1 1.9 1.4
$3.5/$1.9 (2013) 5.8 4.2 2.9 6.9 3.9 1.7
$5.5/$1.9 (1990) 1.5 8.8 3.4 9.4 2.1 1.6
$5.5/$1.9 (2013) 11.7 8.8 6.0 16.0 5.5 2.0
0
2
4
6
8
10
12
14
16
18
RatioofheadcountPovertyat$3.5and
$5.5toHeadcountPovertyat$1.9
$3.5/$1.9 (1990) $3.5/$1.9 (2013) $5.5/$1.9 (1990) $5.5/$1.9 (2013)
9
10. 0
10
20
30
40
50
60
70
80
90
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014
GNI(2011PPP)000s
Very high HDI High HDI Medium HDI Low HDI Acs
Q1: What do we know?
Stylized fact 6: Stagnating growth in per capita terms (000s of
2011 PPPs) and little change in inequality between countries
10
11. Mean years of SchoolingLife expectancy at birth
Q1: What do we know?
Stylized fact 7: Major health and education gains
50
55
60
65
70
75
80
1990
1993
1996
1999
2002
2005
2008
2011
2014
Expectedlifeyears
Very high HDI High HDI
Medium HDI Low HDI
Acs
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
1990
1993
1996
1999
2002
2005
2008
2011
2014
Meanyearsofschooling
Very high HDI High HDI
Medium HDI Low HDI
Acs 11
12. Stylized fact 8: Multidimensional poverty and vulnerability are
also higher than we think when we adjust measure for MICs
Q1: What do we know?
0.2 1.4
20.9
5.5 5.3 11.6
49.7
17.5
1.8 6.3
21.6
7.6
19.2
29.6
23.1
22.6
5.6
16.2
20.4
11.7
27.0
31.0
15.8
25.292.4
76.2
37.0
75.2
48.6
27.8
11.4
34.7
0%
20%
40%
60%
80%
100%
Cluster 1Cluster 2Cluster 3Regional
Average
Cluster1 Cluster2 Cluster3 Regional
Average
Non-Poor
Vulnerable to poverty (deprivation score 20-33.32%)
Poor excluding the severly poor (Deprivation score 33.3-50%)
Severly Poor (deprivation score >50%)
Acute Poverty Poverty
12
13. Stylized fact 9: Deprivation from education is the leading source
of multidimensional poverty
Q1: What do we know?
0%
20%
40%
60%
80%
100%
Jordan
Tunisia
Algeria
Egypt
Iraq
Morocco
Comoros
Yemen
Sudan
Mauritania
%contributiontopoverty
Years of schooling School attendance U5 Mortality Nutrition
Early Preg/FGM Elect Sanitation Water
Roof/Floor Cooking Fuel Overcrowding Assets
13
14. Q1: What do we know?
Stylized fact 10: Glaring subnational inequalities in
multidimensional poverty
2.23
8.35
5.93
2.16
0
1
2
3
4
5
6
7
8
9
Rural/urban Household heads
without
education/highest
education
Bottom wealth
quintile/top wealth
quintile
Household size 8+ /
household size 1-4
Ratio
14
15. Stylized fact 11: Growth without structural transformation (Labour moved
from agriculture and even manufacturing to low value added service sectors)
Source: N. Sarangi, K. Abu-Ismail and V. Gantner (2017). Fiscal Policy and Structural Transformation in the Arab Region: What are the
Pathways? ESCWA Working Paper. E/ESCWA/EDID/ 2017/WP.7.
Q1: What do we know?
15
16. Stylized fact 12: Very low productivity growth
Source: ILO estimates and projections
2.3%
2.9%
0.8%
0.3%
5.0%
2.8%
1.5%
0.7%
2.0%
0.7%
0.4%
-0.8%
4.4%
3.1%
1.0%
0.8%
-2.0%-1.0%0.0% 1.0% 2.0% 3.0% 4.0% 5.0% 6.0%
World
Sub-Saharan Africa
Latin America and the Caribbean
Arab States
Asia and the Pacific
South-Eastern Asia and the Pacific
Europe and Central Asia
Northern, Southern and Western…
2011-2017 2001-2010
Q1:What do we know?
16
17. Stylized fact 13: Low return on education
Q1: What do we know?
Source: Montenegro and Patrinos, 2013 17
18. Q1: What do we know?
Stylized fact 14: An implicit don’t ask and we won’t tax social bargain
0
10
20
30
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
TaxtoGDP(%)
Comoros Djibouti
Egypt Jordan
Lebanon Mauritania
Morocco Sudan
Syrian Arab Republic Tunisia
Yemen, Republic of Arab oil-poor countries
AlgeriaBahrainComoros
Djibouti
Egypt
Iraq
Jordan Kuwait
Lebanon
Libya
Mauritania
Morocco
Oman
Qatar
Saudi
ArabiaSudanSyr…
Tunisia
Un…State of
Palestine
Yemen
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.20 0.40 0.60 0.80 1.00
GovernanceIndex
(Ruleoflaw,VoiceandAccountability)
GNI Index
18
19. Q1: What do we know?
Stylized fact 15: the last 14 stylized facts are changing fast
because of conflict so we need to be careful about stylized facts.
In 2010 I would have predicted Syria to catch up with Lebanon,
Jordan and Tunisia by 2015. Syria is now a LDC. Yemen, which was a
LDC, is now in a state of dire humanitarian crisis with millions facing
famine conditions.
We are now leading the world on refugee shares and military
expenditure.
Etc.
19
20. Conclusion: Two competing narratives affecting our
views on inequality
Narrative 1:
Arab development model was working and people are
better off than their parents and grandparents so even if
inequality is higher (which is not the case) why worry about
it? Thus, inequality not root cause of Arab Spring. Real
problems are with crony capitalism, large public sectors,
state monopolies and education quality.
Policy conclusion: Better doing business indicators, less
state intervention (smaller public sectors, subsidies and
social transfers) and better education quality this will reduce
inequality in opportunities, which is what we should be
concerned with.
21. Arab development model suffers from major structural
weakness as a result of decades of rentier political
economy inducing low HH income growth, high
informality in employment, low productivity. This led to
low extreme poverty but high vulnerability (pressure on
the middle class which is receding since 2010). More
importantly poverty and inequality are much higher when
appropriately measured.
Policy conclusion: Inequality is key and it has to be
understood in a political economy context.
Conclusion: Two competing narratives affecting
our views on inequality
Narrative 2:
22. 1: Normative reasons: equity as end goal.
2: Economic policy: Redistribution positively influences
growth if economy is wage led (Kalecki and Taylor).
3: Social policy: health and education goals at the national
level are often very sensitive to the inclusion of marginalized
and deprived groups (rural areas, particular ethnic groups,
poorest households, etc).
Q2: Why are we doing this report and
who are our audience?
22
23. 1. Growing appetite for the subject after the Arab spring. Growing
recognition that inequality affects political stability and thus
growth sustainability. Policy makers are demanding solutions to
poverty (which necessarily means more work on inequality).
Students are demanding explanations and academics and think
tanks don’t have much to work with.
2. Still many gaps on what we know about the level of inequality in
Arab States Many people are less deprived than their parents in
health and education but inequality may have increased as not all
groups may have improved at the same rate and some groups
may have even deteriorated.
3. In an oil-led rentier political economy inequality analysis is always
key.
Q2: Why are we doing this report and who
are our audience?
23
24. Example: Impact of redistribution on poverty in Egypt
Sc1 (Gini goes to 0.28) Sc2 (Gini goes to 0.4)
58.7
19.4
28.7
5.3
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
Upper Rural All Egypt
59.1
38.3
29.0
23.7
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
Upper Rural All Egypt
Q2: Why are we doing this report and
who are our audience?
25. Q3: What kind of inequality and why?
1. Outcomes versus opportunities?
Both are important (Atkinson, 2015)
2. Unidimensional (money metric) versus
multidimensional (health, education and living conditions)?
We already know quite a lot on money metric inequality
but also data access is problematic
25
26. Q4: Inequality between whom?
1. Rich and poor
2. Spatial (subnational such as rural versus urban).
3. Educated and non-educated household heads.
4. Men and women.
5. Large and small households.
Or between household holding a combination of
these characteristics.
26
28. Data Sources:
24 Household surveys from 2000 to 2015
• The data draws on non-income household surveys from the
Demographic and Health Surveys (DHS), Multiple Indicator Cluster
Surveys (MICS) and Pan Arab Project for Family Health (PAPFAM).
• The surveys collect information on a wide range of monitoring and
impact evaluation indicators in the areas of population, health, and
nutrition and other socio-economic characteristics.
• Although the surveys do not collect information on household
consumption/expenditure, information on the household’s ownership
of selected assets, such as televisions and bicycles; materials used for
housing construction; and types of water access and sanitation facilities
is collected and used to construct the wealth index.
28
30. Availability of Indicators - Health
DZ KM EG IQ JO LY MRT MAR PS SD TN YE
Child Health
Immunization 2003
Mortality Rates 2012 2011 2011 2003 2014
Nutrition Indicator 2012
Living Standards
Overcrowding
(MDG Definition)
Safe Drinking
Water
i i i i
Access to Improved
Sanitation
i i i
Assets (Definition
Arab MPI)
2011
: Available across two points in time. Year: Available in this year only.: iIndicator is only
partly available. Na: not available at all.
30
31. Availability of Indicators – Reproductive
Health
DZ KM EG IQ JO LY MRT MAR PS SD TN YE
Maternal Health
Skilled
attendance at
birth
2011 2014
Antenatal Care i i 2015 2014
FGM na na 2011 na 2014 2011 na na
Age at first
pregnancy
2012
Early Marriage 2012 2011 2014
Anthropometrics
Women
2002 2012 na na na 2003 na na na
: Available across two points in time. Year: Available in this year only.: iIndicator is only
partly available. Na: not available at all.
31
32. Availability of Indicators – Education and
Economic activity
DZ KM EG IQ JO LY MRT MAR PS SD TN YE
Education and Employment/Child Labor
Education Indicators
i i i 2011
Child Labor
2012
na
2007
na 2006 2011 na
Occupation 15+ (all
household members)
na
2002
na
2011
Na na 2002
Work
Status/Occupation
of Women
2012
na na Na na 2002
: Available across two points in time. Year: Available in this year only.: iIndicator is only
partly available. Na: not available at all.
32
33. Harmonization of Surveys &
Availability of the Wealth Index
• The 24 surveys were harmonized to assure comparability of the
indicators across the different survey types, e.g. reference periods for
the maternal health indicators were adjusted, the same reference
population for child nutrition indicators was used, etc.
• A wealth index, i.e. a proxy for the socio-economic status of the
household, is available for 20 surveys. However, as the wealth index
is constructed as a relative index within each country at the time of
the survey, cross-country and trend analysis is limited.
33
34. Methodology:
Inequality of outcome & Inequality of Opportunity
Inequality of outcome:
Analysis of the different indicators in the area of health,
education, and living standard by several measures:
1. Descriptive Statistics (Levels and Trends).
2. Ratios by several socio-economic characteristics like rural/urban,
education of the head of household or wealth quintile
3. Concentration Index.
34
35. Inequality of opportunity:
• An important part of inequality cannot be attributed to differences in
individual efforts or responsibility. On the contrary, it can be directly
ascribed to outside factors such as family background, gender, race
and place of birth. Thus, If people living in the same society do not
have equal access to the same opportunities, it is referred to as
inequality of opportunity.
• Following the analysis of Kraft and Kogali (2014), inequality of
opportunity in health and education will by analysed.
Methodology:
Inequality of outcome & Inequality of Opportunity
35
37. Health Indicator: Stunting
Example of inequality analysis for the indicator stunting:
• Stunting (low height-for-age), reflects chronic
malnourishment.
• Reference population: all children under the age of five.
• Rates of stunting are calculated using the WHO Child
Reference Population.
• Available for all surveys except Tunisia (2001).
37
38. Indicator Stunting: Level and Trend across 12
Arab Countries (national average)
0%
10%
20%
30%
40%
50%
60%
70%
2002/2012
2006/2014
2001/2011
2002/2012
2003/2011
2000/2011
2000/2014
2007/2015
2000/2012
2007/2014
2000/2014
2003/2013
JOR PSE TUN* DZA MAR IRQ EGY RA* MRT COM LBY SDN YEM
PrevalenceofStunting
Stunting (%) Baseline Surveys (2000s) Stunting (%) End-line Surveys (2012s)
38
39. 39
0.00
0.50
1.00
1.50
2.00
2.50
0%
10%
20%
30%
40%
50%
60%
70%
2000 2014 2002 2012 2014 2000 2011 2000 2014 2002 2012 2003 2013 2000 2012 2015 2007 2011 2003 2011
EGY DZA LBY IRQ SDN JOR YEM COM MRT TUN MAR
RatioRuraltoUrban
Prevalencestunting(%)
Rural Urban Ratio Rural/Urban
Indicator Stunting: gaps between rural and
urban areas within each country
40. Indicator Stunting: gaps and ratios between
educated and non-educated heads of household
40
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
0%
10%
20%
30%
40%
50%
60%
70%
2002
2012
2007
2014
2000
2012
2014
2000
2013
2003
2000
2011
2002
2012
2007
2015
2003
2011
2006
2014
2000
2014
DZA LBY COM EGY YEM IRQ JOR MRT MAR PSE SDN
RatioofNon-Educated/Educated
PrevalenceStunting
Education of Head of household (none) Education HoH (12 Years or more)
Ratio of non-educated/educated HoHH
41. Comparing extreme groups: the prevalence of
stunting in the most deprived and most
privileged groups
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
0%
10%
20%
30%
40%
50%
60%
70%
Algeria
2000
Algeria
2012
Egypt
2000
Egypt
2014
Libya
2007*
Libya
2014
Morocco
2003
Morocco
2011
Sudan
2000
Sudan
2014
Ratioofmostdeprived/mostpriviliged
PrevalenceofStunting
Most deprived group Most priviliged group Ratio of most deprived/most priviliged
42. Ratio Analysis – Disparities across several socio-
economic characteristics
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
DZA
2012
COM
2012
EGY
2014
IRQ
2011
JOR
2012
LBY
2014
MRT
2015
MAR
2011
PSE
2014
SDN
2014
YEM
2013
Ratio
Ratio Rural/Urban Ratio Poorest/Richest
Ratio Most Deprived/Priviliged Ratio Non-Educated HoH/Educated
44. Indicator Stunting: Concentration Curve
The concentration curve (CC) ranks the health outcome by
socioeconomic status, in our case by the wealth index. The horizontal axis
of the CC begins with the poorest individual and progresses through the
wealth distribution to the richest individual. This relative ranking of
wealth is then plotted against the cumulative proportion of health
outcome on the vertical axis. The further away the CC from the line of
equality, the more unequal.
44
45. Annual Average Rate of Change (AARC) in the
Concentration Index (Wagstaff, 2005) and in the
Prevalence of Stunting
45
[CELLRANGE]
[C…
[CELLRANGE]
[CELLRANGE]
[CELLRANGE]
[CELLRANGE]
[CELLRANGE]
-6.00%
-5.00%
-4.00%
-3.00%
-2.00%
-1.00%
0.00%
1.00%
-20.00% -15.00% -10.00% -5.00% 0.00% 5.00% 10.00% 15.00%
AARCinstunting
AARC in Concentration Index
More inequality
46. Indicator: Years of Education
For 8 countries, the years of education by each person age 25 and older
could be constructed from the level of education attended and the highest
grade attended for each level across two points in time.
0 2 4 6 8 10 12
2006/2014
2002/2012
2000/2014
2011
2000/2011
2002/2012
2000/2012
2013
2000/2014
2007/2015
2003
PSEJOREGYTUN*IRQDZACOMYEM*SDNMRTMAR*
Average Years of Education achieved
Baseline Survey Endline Survey
47. Indicator Years of Education: Gaps between
rural and urban areas
0
1
1
2
2
3
3
4
0.0
2.0
4.0
6.0
8.0
10.0
12.0
2002 2012 2000 2014 2002 2012 2000 2011 2000 2012 2000 2014 2007 2015
JOR EGY DZA IRQ COM SDN MRT
RatioofUrban/Rural
YearsofEducation
Years of Education Urban Years of Educatio Rural Ratio Urban/Rural
48. Indicator Years of Education: Gaps between
educated and non-educated head of households
0.0
2.0
4.0
6.0
8.0
10.0
12.0
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
2002 2012 2006 2014 2002 2012 2000 2014 2000 2012 2000 2011 2000 2014
JOR PSE DZA EGY COM IRQ SDN
RatioofEducated/non-educatedHoHH
YearsofEducation
Education HoH (12 Years or more) Education of Head of household (none) Ratio Secondary/No Education
49. Indicator Years of Education: Gaps between
extreme groups
0
1
2
3
4
5
6
7
0
2
4
6
8
10
12
14
2014 2002 2012 2006 2014 2000 2012 2000 2014 2011 2000 2012 2013 2000 2014 2007 2003
LIB JOR PSE DZA EGY TUN IRQ COM YEM SUD MRT MAR
RatioofPriviliged/MostDeprived
YearsofEducation
Most Priviliged Most Deprived Ratio priviliged/most deprived
50. Ratio Analysis: Years of Education
0.0
2.0
4.0
6.0
8.0
10.0
12.0
2012 2012 2014 2011 2012 2014 2015 2011 2014 2014
DZA COM EGY IRQ JOR LBY MRT MAR PSE SDN
Ratio
Ratio Urban/Rural Ratio Richest/Poorest Ratio Most Priviliged/Deprived Ratio Educated HoHH/Non Educated
We have a lot of information by countries and through times, so we can do a lot but still miss out on some indicators overtime
Tunisia: only available for 2011
RA* Population Weighted Regional average for 11 countries (excluding Tunisia, the average includes only those countries for which information is available across two points in time) . Population used to weigh the data is based on the population of each country in 2000 and 2014.
They are ranked according to 2012
Tunisia is out of the average
Stunting went down for all countries except Libya , but not all same rate
Most deprived= Prevalence of stunting among children under the age of five that live in rural areas, whose household is in the poorest wealth quintile, household size is bigger than 8 persons, and the head of the household is uneducated.
Most privileged: = Prevalence of stunting among children under the age of five that live in urban areas, whose household is in the richest wealth quintile, whose household size is less than 5 persons and whose head of household has more than 12 years of education.
All reported observations are based on a sample with more than 49 unweighted observations.
Two interesting cases:
Egypt
Sudan !!!
Morocco is counter intuitive
Jordan: increase in inequality over time
Sudan: huge gap between extreme groups
We don’t claim to explain these stylized facts , this what the report will try to explain.
Education of HoHH refers to the educational level of the head of household
Anything above the red line: increase in inequality
Anything below the line decrease in inequality
The red line is “equality” or stagnation
In Egypt: decline in inequality of stunting because the level of stunting increased at the upper quintile
In Yemen, inequality increased, Since 2013 a lot of things changed
In Egypt, it decreases because stunting increases for the richest quintile… Need more investigation!!
Rate of change in Gini against change in stunting
Y axis: Average annual change in deprivation
X : Average annual change in inequality
Egypt: decline in deprivation and decline in inequality
Most countries have decline in deprivation and increasing in income
In Yemen there is 10% increase in income quintile
Countries with asterisk have only information for one point in time
Persisting gaps in most countries, only Jordan and Palestine closed the rural/urban gap over time. In Iraq, the rural/urban gap flipped (in 2011, rural areas show a higher years of schooling)
Education of HoHH refers to the educational level of the head of household
Anything above the red line: increase in inequality
Anything below the line decrease in inequality
The red line is “equality” or stagnation