Chapter 2 Addressing Inequality of Opportunity in Asia and ... › sites › default › d8files › 04Chapter2.pdfdistribution lie in the work of John Rawls and Amartya Sen. Rawls

INEQUALITY IN ASIA AND THE PACIFIC24

Chapter 2

Addressing Inequality ofOpportunity in Asia andthe Pacific

25INEQUALITY IN ASIA AND THE PACIFIC

While the aggregate costs of inequality ofoutcome can be high, the impact ofinequality is perhaps more corrosive at theindividual or the household level. So far, research andheadlines on inequality have emphasized the mostglaring contrasts: the lavish lives of billionaires comparedwith the uncertainty, stagnant wages and exploitationthat are often experienced by the poorest people insociety. Global analysts have tended to overlook theimpact of income and wealth inequality on accessingbasic opportunities, including quality education andhealth care, meaningful work and decent livingconditions.

Inequality in access to opportunities has gone particularlyunnoticed in the fast-growing economies of the Asia-Pacific region, where many people have been able toimprove their quality of life. There are more schools andhealth clinics than ever, water supplies and treatmentworks have come on stream and electricity grids andtelecommunications systems have sprouted. TheMillennium Development Goals (MDGs) played their partin broadening access to essential services, but many ofthese achievements were not been evenly distributed.

Inclusion is at the core of the 2030 Agenda forSustainable Development, reflected in the pledge toleave no one behind and in the vision of a “just,equitable, tolerant, open and socially inclusive world inwhich the needs of the most vulnerable are met.” Thischapter goes beyond the debate of an ideal level ofincome or wealth and examines the extent to whichpeople from different circumstances and backgroundsacross the Asia-Pacific region have equal chances to fulfiltheir potential.

The chapter begins by exploring the levels of inequalityof opportunity people in Asia and the Pacific face onaggregate levels and how each of the opportunitiespresented is linked to the Agenda for SustainableDevelopment. It measures inequality of opportunityusing the dissimilarity index (D-index), which allows acomparison of inequality levels among countries as wellas a further decomposition of the observed inequalityinto those circumstances that contribute mostly to it.

The analysis then delves deeper to determine thosehouseholds and individuals that lack access toopportunities. Using an algorithm that produces country-and opportunity-specific classification trees for21 countries in the Asia-Pacific region, it reveals thecircumstances shared by those most disadvantaged andthe most advantaged groups in each country. Thechapter concludes by discussing trends and whetherpolicies have been effective in influencing access toopportunity.


2.1 WHAT DOES INEQUALITY OFOPPORTUNITY MEAN IN THE CONTEXTOF THE 2030 AGENDA?

Inequality of opportunity is concerned with access tokey dimensions necessary for meeting aspirationsregarding quality of life. It has economic dimensions(e.g. unequal access to decent work, financial services,land ownership, etc.), social dimensions (e.g. unequalaccess to health care, education, nutrition, etc.) andenvironmental dimensions (e.g. unequal access to water,sanitation, clean fuels, electricity, access to land andnatural resources, etc.).

To gain an understanding of inequality of opportunityacross the Asia-Pacific region, 14 categories or indicatorshave been selected that encompass basic, yet criticalopportunities for individuals and households. The eightindicators of opportunities for individuals are: secondaryeducation, higher education, modern contraceptives,professional help during childbirth, decent work and(absence of) stunting, wasting and overweight amongchildren. The five indicators of opportunities forhouseholds are: access to basic drinking water, basicsanitation, electricity and clean fuels, and ownership ofa bank account. To summarize the opportunities forhouseholds, an additional indicator combines these fivecategories.

Inequality of opportunity is here defined as the gaps inaccess to each of these opportunities that depend oncircumstances beyond a person’s control. The conceptof inequality of opportunity has previously been usedto distinguish between personal responsibility andcircumstances for economic outcomes in life. Thephilosophical foundations of this approach to incomedistribution lie in the work of John Rawls and AmartyaSen. Rawls was among the first modern politicalphilosophers who articulated the importance ofbalancing personal liberties with distributive justice andfair options for all, arguing that public policy choicesshould focus on raising the welfare of the poorestpeople.1 Rawls argued that a set of primary goodsshould be made available for everyone, so that she orhe would be able to realize their life plan. Sen, later,argued that inequality could be re-examined from theperspective of human capability, looking at the meansrather than the ends of development, since withoutequal opportunity, equitable outcomes could not besecured.

Focusing on inequality of opportunity also serves asa reminder that inequality is not a static phenomenon,but rather is transmitted to children, creatingintergenerational inequality of opportunity traps thatreproduce and magnify income inequality.2 Inequality ofopportunity therefore combines issues of both equity

and efficiency. The equity argument calls for levellingthe playing field, in accordance with internationalagreements and with established human rights. Theefficiency argument motivates policymaking thatequalizes opportunities at the top level, meaning thateveryone should have access, rather than reducingaccess for those who already have it.3

Target 10.3 of SDG 10 calls for ensuring equalopportunity and reducing inequalities of outcome,including by eliminating discriminatory laws, policies andpractices and promoting appropriate legislation, policiesand action. Officially reported numbers of incidents ofdiscrimination may underestimate the real experience ofthose most marginalized. Reported data may also notreveal the daily experience of women, men and childrenwho lack access to basic opportunities because ofentrenched poverty and institutional failure.

Target 10.3, however, is not the only one in the 2030Agenda that relates to inequality of opportunity. Theindictors of opportunities presented in this chapter referto specific SDGs and most of them also directly respondto established indicators. Of them, five are examined indetail: secondary educational attainment, stuntingamong children (used as a proxy for adequate nutrition),professional help during childbirth, full-time employment(used as a proxy for decent work in developing countrycontexts) and a group of basic household services(Table 2.1). To identify the patterns of advantage ordisadvantage, we have referred to data available inDemographic and Health Surveys (DHS) and MultipleIndicator Cluster Surveys (MICS) and the latest WorldGallup Survey.

2.2 WHY INEQUALITY IN ACCESS TOOPPORTUNITIES MATTERS

Equity in opportunity can be described as a level playingfield on which all households enjoy the same access tobasic services, such as clean water, sanitation, electricityand clean fuels; where all children have adequatenutritious food and complete education; where everyonehas access to health care services when needed, ataffordable prices; and where those who want to workcan find a decent job. These rights are enshrined invarious Conventions of the United Nations, including theConvention on the Rights of the Child, the Conventionon the Rights of Persons with Disabilities and, certainly,the Universal Declaration of Human Rights. They are alsoenshrined in Constitutions and other legislation acrossthe region and are what drove leaders of 193 countriesto adopt the SDGs in 2015.

Although the Asia-Pacific region has experiencedsignificant advances in many development indicators, theplaying field is not levelled. Enrolment rates in primary


Note: For all opportunities, except No. 8 (full-time employment) the data used are from DHS and MICS, earliest and latest surveys available (36 surveys in total), while foropportunity No. 8 the data used are from the Gallup World Poll, latest year.Note 2: In orange are those opportunities that are the focus of this chapter.Note 3:For opportunity 8, which is based on a different survey, the person’s place in the wealth distribution was not used for the analysis. Marital status (divorced, married,separated, single) was used as an additional circumstance (not as a replacement for wealth).

Table 2.1 Links between opportunities, circumstances and the SDGs

Opportunity Circumstances used to determine groups of the furthest behind/ahead Closest SDG indicator reference

Wealth: Residence: Education: Sex: Children:Age:

OpportunitySurvey Reference group

Bottom 40- Urban- No/Primary- Male- Yes-No,15-24,

Related SDG Indicatorused in survey

Top 60* Rural Secondery-Higher Female Number25-49,50-64

1 Secondary DHS/ Household Wealth Residence n/a Woman/ n/a n/a 4.1.1 Proportion of children and young people:education MICS member Man (a) in grades 2/3; (b) at the end of primary; and

aged 20-35 (c) at the end of lower secondary achieving atleast a minimum proficiency level in (i) readingand (ii) mathematics, by sex

2 Higher education DHS/ Household Wealth Residence n/a Woman/ n/a n/a 4.3.1 Participation rate of youth and adults inMICS member Man formal and non-formal education and training in

aged 25-35 the previous 12 months, by sex

3 Stunting DHS/ Child aged 0-5 Wealth Residence Mother’s Boy/Girl Number of n/a 2.2.1 Prevalence of stunting (height for age


Inequality in access to health care is also damaging fromboth an equity and an efficiency perspective. Yet, thereis ample evidence of persisting health-care-relatedinequalities across the Asia-Pacific region. One exampleis the proportion of births attended by skilled personnel,a critical factor for reducing neonatal and maternalmortality. The divide in skilled birth attendance betweenrich and poor segments within many countries isenormous, although there is a much lower variationamong wealthier groups between countries (Figure 2.2).The richest citizens in all countries, with the exceptionof Timor-Leste, enjoy a similarly high level of access toskilled personnel when giving birth. Conversely, thepoorest citizens have the lowest level of access to skilledpersonnel when giving birth. Armenia, Kazakhstan andUzbekistan are exceptions, where nearly all births areattended by skilled personnel.

Inequality in access to health care often has long-termhealth implications for women and children, with anegative impact on educational attainment and futurelabour force participation rates. Improvements inreproductive health services are associated with reducedfertility rates among poorer women, which, in turn, notonly increase their chances of survival, but may alsoboost their earning potential.5 Inequality in access tohealth care can also have a broader economic impact.As health-care costs and out-of-pocket health-related

expenditures are particularly high in the Asia-Pacificregion, families tend to save to guard against the riskof unexpected or unplanned needs.6 This higher savingsrate can result in lower domestic consumption andslower economic growth.

Being excluded from basic household services can exacta high cost. For example, contaminated water and poorsanitation cause diarrhoea that, if untreated, can leadto long-term cognitive and developmental impacts andeven death through dehydration. The World HealthOrganization (WHO) has estimated that, in 2012,diarrhoea resulting from a lack of clean water and poorsanitation was responsible for more than 800,000deaths globally and for 1.5 per cent of the global burdenof disease.7 A lack of sanitation costs the world anestimated US$260 billion every year in terms of lowerproductivity, sickness and loss of revenue fromunrealized investment in sectors such as tourism.8

While substantial progress has been made across theAsia-Pacific region in the past two decades, access toimproved sanitation facilities remains low in rural areasof several countries. Fewer than 20 per cent of PapuaNew Guinea’s citizens, for example, have accessto improved sanitation facilities, and the figure islower than 40 per cent in Afghanistan, India, Kiribati,Papua New Guinea, Solomon Islands and Timor-Leste(Figure 2.3).

Figure 2.1 Secondary school attendance gaps in Asia-Pacific countries, latest year

Source: ESCAP, 2017. Sustainable Social Development in Asia and the Pacific: Towards a people-centred transformation.Note: Secondary net attendance ratio data were disaggregated by wealth quintiles and location of residence. For countries in the Asia and Pacific region,the most recent data were used.

Total Urban Rural Poorest Richest

Van

uat

uSo

lom

on

Isla

nd

sA

fgh

anis

tan

Paki

stan

Lao

PD

RC

amb

od

iaTi

mo

r-Le

ste

Ban

gla

des

hIn

dia

Bh

uta

nM

yan

mar

Kir

ibat

iSa

mo

aN

epal

Phili

pp

ines

Mal

div

esN

auru

Turk

eyIn

do

nes

iaTh

aila

nd

Ton

ga

Kaz

akh

stan

Aze

rbai

jan

Geo

rgia

Tajik

ista

nC

hin

aV

iet

Nam

Arm

enia

Kyrg

yzst

anM

on

go

liaTu

rkm

enis

tan

100

90

80

70

60

50

40

30

20

10

0

Net

att

end

ance

in s

eco

nd

ary

edu

cati

on

(%)


Figure 2.2 Births attended by skilled personnel in Asia-Pacific countries, by wealth quintile

Source: ESCAP, 2017. Sustainable Social Development in Asia and the Pacific: Towards a people-centred transformation.Note: Data refers to the most recent year between 2003 and 2014.

Shar

e o

f bir

ths

(%)

100

80

60

40

20

0Q1 (poorest) Q2 Q3 Q4 Q5 (richest)

ARM/UZB/KAZ ARM/UZB/KAZ ARM/UZB/KAZ ARM/UZB/KAZ ARM/UZB/KAZ

IND LAOVUT

TLSNPL IND

LAO

NPLIND

LAO

VUT

TLS

NPL

IND

LAO

VUT

TLS

NPL

IND

LAO

VUT

TLS

NPL

VUT

TLS

Figure 2.3 Availability of improved sanitation facilities, Asia-Pacific region

Source: ESCAP, 2017. Sustainable Social Development in Asia and the Pacific: Towards a people-centred transformation.Note: Data refers to the most recent year between 2003 and 2014.

Shar

e o

f po

pu

lati

on

(%)

100

80

60

40

20

0

Total Urban Rural

Pap

ua

New

Gu

inea

Solo

mo

n Is

lan

ds

Afg

han

ista

nIn

dia

Kir

ibat

iTi

mo

r-Le

ste

Cam

bo

dia

Nep

alB

hu

tan

Mic

ron

esia

(Fed

erat

ed S

tate

s o

f)Va

nu

atu

Mo

ng

olia

Ban

gla

des

hIn

do

nes

iaPa

kist

anLa

o P

DR

Ru

ssia

n F

eder

atio

nPh

ilip

pin

esC

hin

aM

arsh

all I

slan

ds

Vie

t N

amM

yan

mar

Dem

ocr

atic

Peo

ple

’s R

epu

blic

of K

ore

aG

eorg

iaA

zerb

aija

nA

rmen

iaIr

an (I

slam

ic R

epu

blic

of)

Ton

ga

Fiji

Sam

oa

Thai

lan

dKy

rgyz

stan

Turk

eyTa

jikis

tan

Sri L

anka

Mal

aysi

aK

azak

hst

anM

ald

ives

Niu

eA

ust

ralia

Jap

an

Inequality in access to clean energy, i.e. electricity andclean fuels, also weighs on the efficiency of an economy.Households without electricity, for example, devote lesstime to study, work or leisure, which can result in feweropportunities for career development and earningpotential. As a result, inequality in access to electricityand clean fuels can create and reinforce inequality gapsin skills and productivity. It also perpetuates disparities

in health outcomes among and within countries.Burning dirty fuels affects air quality in homes and inthe community. Globally, indoor air pollution causesmore than 4 million deaths per year, more than half ofwhich occur in China and India alone.9 Despite economicprogress and greater awareness, close to half of allpeople in Asia and the Pacific still rely on traditional andinefficient fuels for cooking and heating.10 There are still


marked gaps in access to electricity, notably in ruralareas of Papua New Guinea, Solomon Islands, Timor-Leste and Vanuatu (Figure 2.4).

Inequalities in access to services such as an electricitysupply, clean cooking fuels, drinking water andsanitation disproportionately affect women, who bearthe brunt of household work and caretaker tasks.11 Theyalso suffer more from the health consequencesassociated with indoor air pollution, while foregoing theopportunity to earn their own income.12

2.3 WHY AVERAGE PROGRESS IS NOT ENOUGH

The evidence is clear that a rising tide has so far failedto lift all boats. The remainder of this chapter will analysedata from three types of household surveys:Demographic and Health Surveys (DHS), MultipleIndicator Cluster Surveys (MICS) and the latest WorldGallup Survey, to better understand the circumstancesof those left behind in Asia-Pacific countries.

The D-index is a useful tool for measuring thedistribution of access to a certain opportunity acrosssocieties. It is calculated for 14 individual or household-based indicators of opportunities critical for humanwellbeing (Table 2.1): 1) attainment of secondaryeducation for 20-35 year-olds; 2) attainment of highereducation for 25-35 year-olds; 3) prevalence of stunting(0-5 year-olds); 4) prevalence of wasting (0-5 year-olds);5) prevalence of overweight (0-5 year-olds); 6) women’saccess to modern contraception; 7) women’s access to

professional help during childbirth; 8) access tofull-time employment;13 and 9) household’s accessto safe drinking water; 10) household’s access tobasic sanitation; 11) household’s access to electricity;12) household’s access to clean fuels; 13) household’sownership of a bank account; and 14) household’saccess to all of the basic services opportunities forhouseholds (9-13), or “multiple deprivation”.

The D-index measures how all population groups farein terms of access to a certain opportunity. For example,two countries with identical secondary educationattainment rates may have a different D-index if thedistribution of attainment in one country excludescertain groups. Like the Gini coefficient, the D-indextakes values from 0 to 1, 0 meaning no inequality, and1 maximum inequality (Annex 2.1). Unlike the Ginicoefficient, the ideal level of a D-index is 0, wherebyeveryone has access to an opportunity.

The highest D-indices are found in South and South-West Asia, followed closely by South-East Asia(Figure 2.5). In both subregions, the opportunities thatstand out as most unequally distributed are access toclean fuels, higher and secondary education andownership of a bank account. South-East Asia also hasthe highest average D-index for access to full-timeemployment, although there is no significant variationacross subregions. In East and North-East Asia, data forthe majority of opportunities are only available forMongolia, with safe sanitation the most unequallydistributed opportunity, followed by access to clean

Figure 2.4 Access to electricity, Asia-Pacific region

Source: ESCAP based on World Bank, Sustainable Energy for All Database, SE4ALL Global Tracking Framework, 2017.

100

80

60

40

20

0

Shar

e o

f po

pu

lati

on

(%)

Pap

ua

New

Gu

inea

Van

uat

uSo

lom

on

Isla

nd

sTi

mo

r-Le

ste

Kir

ibat

iM

yan

mar

Cam

bo

dia

Ban

gla

des

hM

icro

nes

ia (F

eder

ated

Sta

tes

of)

Lao

PD

RIn

dia

Nep

alW

orl

dM

on

go

liaPh

ilip

pin

esA

fgh

anis

tan

Mar

shal

l Isl

and

sSr

i Lan

kaTo

ng

aIn

do

nes

iaPa

kist

anSa

mo

aTu

valu

Vie

t N

amIr

an (I

slam

ic R

epu

blic

of)

Kyrg

yzst

anPa

lau

Tajik

ista

nU

zbek

ista

nTu

rkm

enis

tan

Turk

eyTh

aila

nd

Sin

gap

ore

Russ

ian

Fed

erat

ion

New

Zea

lan

dN

ew C

aled

on

iaM

ald

ives

Mac

ao, C

hin

aRe

pu

blic

of K

ore

aK

azak

hst

anJa

pan

Ho

ng

Ko

ng,

Ch

ina

Fiji

Ch

ina

Bru

nei

Dar

uss

alam

Bh

uta

nA

zerb

aija

nA

ust

ralia

Arm

enia

Total Urban Rural


Figure 2.5 D-indices for 10 basic opportunities, by subregion

Source: ESCAP calculations using data from the latest Gallup, DHS and MICS surveys for countries in Asia-Pacific.Note: The closest a country is the to centre of the graph, the higher the D-index and higher the inequality. The furthest away from the centre, the lowerthe D-index and lower the inequality.Note 2: This figure only depicts selected average subregional D-Indices (covering 10 out of 14 opportunities) for clarity.

South and South-West Asia (SSWA)

North and Central Asia (NCA)

South-East Asia (SEA)

East and North-East Asia (Mongolia)

Pacific (Vanuatu)

Electricity

Clean Fuel

Bank Account

Clean Water

Basic Sanitation

Non-Stunting

Modern Contraception

SecondaryEducation

HigherEducation

Full-timeemployment

lowerinequality

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

fuels. In the Pacific, data were only available forVanuatu, where access to electricity and clean fuels wereparticularly unequally distributed.

Zooming further into each individual country, thesubregional messages are repeated. The most unequalopportunities are higher educational attainment andaccess to clean fuels, followed by ownership of a bankaccount (Table 2.2). The countries where inequality inaccess is large in a wide range of opportunities are:Afghanistan, Bhutan, Lao People’s Democratic Republic,Pakistan and Timor-Leste.

Lao People’s Democratic Republic has the highestinequality of all countries in three opportunities:professional help in childbirth, secondary and highereducational attainment. Timor-Leste also tops theinequality list for three opportunities: access to cleanfuels, ownership of a bank account and access tomodern contraceptives. Afghanistan exhibits the highestinequality in terms of access to full-time employmentand in access to clean water. Pakistan experiences thehighest inequality in terms of children’s nutrition(prevalence of stunting and wasting).

On the other hand, almost all North and Central-Asiancountries have low inequality in access to opportunities,thanks to a tradition of a large state that ensuresuniversal provision of basic services. Kazakhstan has the

lowest inequality in three opportunities (basic sanitation,non-stunted children and secondary education), andbelow-average inequality in all other opportunities.Turkmenistan is in a similar category, with the lowestinequality in terms of electricity and clean fuels access,as well as professional help during childbirth.

Averaging the D-indices for individuals and householdsby country confirms the patterns described earlier, butalso highlights which countries have relatively higherinequality across all opportunities (Figure 2.6). In additionto Afghanistan, Lao People’s Democratic Republic,Pakistan and Timor-Leste, which had the highestinequality in individual opportunities, Cambodia,Myanmar and Vanuatu also appear as particularlyunequal across the board of opportunities. At the otherend of the scale, Maldives and Thailand stand out,together with several North and Central Asian countriesas having achieved a relatively equal distribution ofopportunities across various population groups for mostopportunities. In the middle of the distribution are someof the region’s most rapidly developing countries,including India, Indonesia, Philippines and Viet Nam.India, in particular, made tremendous progress over thepast few years in achieving almost universal accessto financial services for all households (see Box 4.2in chapter 4), as well as in increasing women’s accessto professional help during childbirth (see Figure 2.17).


Source: ESCAP calculations using data from the latest DHS and MICS surveys for countries in Asia-Pacific. Additional countries from the Gallup WorldPoll are not included as D-index was available for one opportunity only, “Full-time employment.”Note: The green light is given to the values that are in the lowest third (from zero to 33th percentile), yellow to the middle third (33-67th) and red forthe highest third. The split into percentiles is done based on all opportunities together, hence most of the Asia-Pacific countries listed here belong to thelowest third for child nutrition and to the highest third for education and employment. Additionally, the best and the worst performer in each opportunityare highlighted with green/red shading.

Table 2.2 Calculated D-indices for all opportunities, Asia-Pacific countries

Calculated D-indices

Afghanistan ArmeniaBangladeshBhutanCambodiaIndiaIndonesiaKazakhstanKyrgyzstanLao PDRMaldivesMongoliaMyanmarPakistanPhilippinesTajikistanThailandTimor-LesteTurkmenistan VanuatuViet Nam

Household-based Individual-based

EnergyFinancial inclusion WASH

Multiple Child Nutrition (0-5 years) Women’s Health(15-49 years)

Education Employment

Country Electricity Clean fuels

Bank account

Basic drinking

water

Basic sanitation

Multiple deprivation

Not stunted

Not wasted

Not overweight

Professional help in

childbirth

Modern con-

traception

Secondary education

Higher education

Full-time employment

Figure 2.6 Average D-indices in Asia-Pacific countries, grouped by subregion

Source: ESCAP calculations using data from the latest DHS and MICS surveys for countries in Asia-Pacific.

0.400.350.300.250.200.150.100.05

0

Ave

rage

D-in

dex

Afg

hani

stan

Bang

lad

esh

Paki

stan

Bhut

an

Ind

ia

Mal

div

es

Tim

or-L

este

Lao

PDR

Cam

bod

ia

Mya

nmar

Ind

ones

ia

Phili

pp

ines

Vie

t Nam

Thai

land

Vanu

atu

Tajik

ista

n

Arm

enia

Kyrg

yzst

an

Turk

men

ista

n

Kaza

khst

an

Mon

golia

SSWA SEA NCAPacific ENEA

2.4 WHAT DRIVES INEQUALITY OFOPPORTUNITY…

The contribution of each of the circumstances toinequality measured in terms of the D-index can beestimated using a methodology called the Shapleydecomposition (Annex 2.2). The decomposition resultsshow that different circumstances weigh differently inshaping inequality for each opportunity and country,although common threads can be found. This sectionfirst reviews the drivers of inequality in three keyopportunities: secondary educational attainment, accessto adequate nutrition among children and access to

decent work. It then takes a bird’s-eye view to spot themost important drivers of inequality across all countriesand opportunities.14 Identifying these common driversreveals not only that inequality of opportunity is tightlylinked with inequality of outcome (wealth and income),but also that it is easily transmitted across generations.

2.4.1 …in stunting

Stunting in children is associated with poorer schoolperformance and lower future earnings potential. Thecircumstances that underpin observed inequality instunting levels among children vary.15 In 9 out of 18


countries, the child’s household wealth status is whatdetermines most of the inequality (Figure 2.7). Amongthose countries, Lao People’s Democratic Republic andPakistan stand out as the countries with the highestinequality, as measured by the D-index. The second mostimportant circumstance is a mother’s education, which

is driving most of the inequality in 5 out of the 18countries. In three countries, Armenia, Kyrgyzstan andTurkmenistan, inequality is mostly associated with thesize of the household, and less so by wealth, residence,or the child’s sex.

Figure 2.7 Inequality in adequate nutrition among children (non-stunted) and its decomposition forselected countries, grouped by the most important circumstance in shaping inequality, latest year

Source: ESCAP, using data from the latest DHS and MICS surveys for Asia-Pacific countries.

Paki

stan

Lao

PD

R

Vie

t N

am

Ban

gla

des

h

Ind

ia

Cam

bo

dia

Tajik

ista

n

Thai

lan

d

Mo

ng

olia

Van

uat

u

Tim

or-

Lest

e

Mya

nm

ar

Bh

uta

n

Mal

div

es

Kaz

akh

stan

Arm

enia

Kyrg

yzst

an

Turk

men

ista

n

Wealth Residence Child’s sex (boy) Mother’s education No. of children under the age of 5

0.14

0.12

0.10

0.08

0.06

0.04

0.02

0Dec

omp

osit

ion

of D

-ind

ex

No. of children under the age

of 5

Mother’s educationWealth Child’s sex

These results are also confirmed through country-specificlogistic regressions (Figure 2.8, Panel 1).16 A mother’seducation is prominent in determining her child’snutrition status in most South Asian countries, includingBangladesh, Bhutan and Pakistan, indicating thatwomen have a strong role in determining their children’snutrition status, despite marginalization and persistentinequalities in other aspects of social and economic life.A child whose mother has completed secondaryeducation has between 20 per cent (Timor-Leste) and110 per cent (Pakistan) higher chances of being non-stunted. Investing in the education of girls, particularlyin South Asian countries, could therefore help disruptthe cycle of intergenerational disadvantage that istransmitted across generations through inadequatenutrition.

Family (household) wealth is also startlingly importantin shaping a child’s risk of being stunted. In Pakistan,children from households in the top 60 per cent ofthe wealth distribution are half as likely to be stunted(Figure 2.8, Panel 2). In most of the remaining countries,belonging to the top 60 reduces the risk of childrenbeing stunted by between 20 and 40 per cent, asignificant impact.

2.4.2 …in education

Wealth is, in turn, a critical factor for accessingsecondary education in 12 out of 21 countries (Figure2.9). The contribution of wealth in shaping inequality isshown by the light-shaded colour in each bar. Wealthmakes up for a higher share of the D-index in countriesmarked in green, which includes very different countries:from Bangladesh and Pakistan, with higher inequality,to Armenia and Kazakhstan, with much lower inequalitylevels. The importance of an individual’s wealth level indriving inequality in education also emphasizes thevicious cycle between inequality of outcome andinequality of opportunity, whereby poorer young menand women join the labour force with less formaleducation and possibly fewer skills.

The second most prevalent circumstance is residence ina rural area, highlighting the urban-rural divide in theavailability of quality schools and opportunities in 7 outof 21 countries. In two countries, Afghanistan andTajikistan, gender matters most. A closer look at thegroups of those being left behind (Annex 2.3: Who arethe furthest behind?) confirms that in both countrieswomen are mostly excluded from secondary education.


baseline: child in a bottom 40household

baseline: motherhas noeducation

Timor

-Leste

Myan

mar

Kaza

khsta

n

Maldi

ves

Tajik

istan

Viet N

am

Bhut

an

Thail

and

Turk

men

istan

Bang

lades

hInd

ia

Cam

bodia

Mong

olia

Lao P

DR

Pakis

tan

Timor

-Leste

Cam

bodia

Maldi

ves

Myan

mar

India

Mong

olia

Bang

lades

h

Arm

enia

Bhut

an

Thail

and

Viet N

am

Lao P

DR

Pakis

tan

120

100

80

60

40

20

0

Perc

enta

ge

Perc

enta

ge

0-10-20-30-40-50-60-70-80-90

-100

Figure 2.8 Impact of mother’s education and household wealth in reducing stunting among children,selected Asia-Pacific countries

Source: DHS and MICS surveys, latest data.Note: Results are based on country-specific logistic regressions. Only countries with statistically significant coefficients and odds-ratios are shown.

Panel 2: The lower odds of having a stunted child for top 60 households,compared to bottom 40 households

Panel 1: The higher odds of having a healthy (non-stunted) child for mothers with completedsecondary education, compared to mothers with no education

Figure 2.9 Inequality in secondary educational attainment among 20-35-year olds and its decomposition,countries grouped by most important circumstance, latest year

Source: ESCAP calculations, using data from DHS and MICS surveys for Asia-Pacific countries.

Lao

PD

R

Mal

div

es

Cam

bo

dia

Mya

nm

ar

Bh

uta

n

Tim

or-

Lest

e

Kyrg

yzst

an

Afg

han

ista

n

Tajik

ista

n

Ban

gla

des

h

Paki

stan

Ind

ia

Van

uat

u

Vie

t N

am

Ind

on

esia

Mo

ng

olia

Thai

lan

d

Phili

pp

ines

Turk

men

ista

n

Arm

enia

Kaz

akh

stan

Residence Gender Wealth

Residence Gender Wealth

0.400.350.300.250.200.150.100.05

0Dec

omp

osit

ion

of D

-ind

ex


While these findings do not indicate causality (living inrural areas does not cause people to drop out ofsecondary education), the strong association indicatesan underlying relationship worth exploring. Indeed,regression analysis shows that in Afghanistan and inTajikistan women have up to a 60 per cent less chancethan men to complete secondary education (Figure 2.10,Panel 2).17 The chances of women completing secondaryeducation are also lower in the less developed South-East Asian countries and in most of South Asia. Overall,however, the gender impact is mixed. In many North andCentral Asian countries, as well as in some South-EastAsian countries (Myanmar, Philippines and Thailand),women have higher chances of completing secondaryand higher education than men, all else being equal. InMongolia, their chances are twice as high as for men.

The impact of rural residence, on the other hand, alwaysgoes in one direction, limiting the chances of individualscompleting secondary or higher education. In themajority of countries, residing in a rural area is associatedwith lower chances of obtaining a secondary educationby 50 per cent or more. In certain South-East Asiancountries, including Cambodia, Lao People’s DemocraticRepublic, Myanmar and Timor-Leste, the chances ofa rural resident completing a secondary or highereducation are up to 80 per cent lower than those of anurban resident. The impact of residence is lesspronounced in upper-middle income countries, such asThailand and Kazakhstan, but also in Indonesia, a resultthat could be attributed to decentralization andprioritization of investments in schools in rural areas inthat country.

Figure 2.10 Impact of gender and residence in completing secondary and higher education, selected Asia-Pacific countries

Source: Latest DHS and MICS surveys.Note: Results are based on country-specific logistic regressions. Only countries with statistically significant coefficients and odds-ratios are shown.

Panel 2: The lower odds of rural residents completing secondary andhigher education, compared to urban residents

Panel 1: The odds of women completing secondary and higher education,compared with men

Mon

golia

Kaza

khsta

n

Kyrg

yzsta

n

Mya

nmar

Thail

and

Philip

pine

s

Indo

nesia

Turk

men

istan

Indi

a

Bang

lades

h

Pakis

tan

Lao P

DR

Vanu

atu

Cam

bodi

a

Tajik

istan

Bhut

an

Timor

-Leste

Afgh

anist

an

150

90

120

30

60

0

-30

-60

-90

Perc

enta

ge

0

-30

-20

-10

-50

-40

-60

-70

-80

-90

Perc

enta

ge

Men’s odds are higher

Women’s odds are higher

males

urban

Secondary education Higher education

Secondary education Higher education

Thail

and

Indi

a

Kaza

khsta

n

Indo

nesia

Bang

lades

h

Mon

golia

Viet

Nam

Tajik

istan

Kyrg

yzsta

n

Bhut

an

Vanu

atu

Pakis

tan

Afgh

anist

an

Turk

men

istan

Timor

-Leste

Cam

bodi

a

Mald

ives

Mya

nmar

Lao P

DR


2.4.3 …in decent work

Beyond adequate nutrition and 12 completed years ofeducation, the most direct determinant of futureoutcomes is the opportunity to access decent work.Decent work is characterized by four main components:employment, social protection, rights at work and socialdialogue. While it is not possible to review all componentsthrough available survey data, a proxy used in thisanalysis is being a full-time employee for a company oremployer. Given the large scale of labour marketinformality and underemployment in most developingcountries in the region, access to full-time employmentis used to proxy the conditions of decent work.

Among all circumstances, gender explains the bulk ofinequality in access to full-time employment morefrequently than any other factor, including education(Figure 2.11). It is the most prominent circumstance in10 out of the 33 countries studied, including Nepal andthe Republic of Korea. The second most importantcircumstance is the level of education, followed by theage group. Age group matters most in several moreadvanced countries such as New Zealand, the RussianFederation and Singapore, but also in Cambodia, India,the Islamic Republic of Iran and Pakistan. Marital statusis generally less associated with high inequality, exceptin the cases of Australia and Kazakhstan.

Figure 2.11 Inequality in access to full-time employment and its decomposition in selected countries,grouped by the most important circumstance, latest year

Source: ESCAP calculations, prepared with the help of ILO and using data from the Gallup World Poll.Note: In more advanced countries, being in full-time employment could also reflect personal choice to work part-time. For the definition of full-timeemployment, see relevant endnote (No.13).

0.60

0.50

0.40

0.30

0.20

0.10

0Dec

omp

osit

ion

of D

-ind

ex

Afg

hani

stan

Lao

PDR

Mya

nmar

Tajik

ista

nV

iet N

amTh

aila

ndM

ongo

liaC

hina

Turk

men

ista

n

Kyrg

yzst

anA

rmen

iaG

eorg

iaA

zerb

aija

nBa

ngla

des

h

Nep

alBh

utan

Phili

pp

ines

Uzb

ekis

tan

Ind

ones

iaSr

i Lan

kaJa

pan

Mal

aysi

aTu

rkey

Rep

ublic

of K

orea

Iran

(Isl

amic

Rep

ublic

of)

Cam

bod

iaPa

kist

anIn

dia

New

Zea

land

Sing

apor

eRu

ssia

n Fe

der

atio

n

Aus

tral

iaKa

zakh

stan

Marital status

AgeGenderResidenceEducation

Gender Age Marital status Have children Education Residence

These results are confirmed through regression analysisof survey data.18 Completing higher education increasesthe odds of being a full-time employee by a staggering11 times (1,200 per cent) in Afghanistan, 9 times inArmenia and 7 in Azerbaijan. Even in countries wherethe impact is smaller, like in Indonesia, the RussianFederation and Thailand, having completed highereducation still doubles the chances of being a full-timeemployee (Figure 2.12, Panel 1). The scale of importanceof higher education in creating an advantage in the

labour market is beyond any other seen in the regressionanalyses conducted for this report.

Women are less likely than men to be in full-timeemployment in all countries studied apart from RussianFederation. In Nepal, women are almost 80 per cent lesslikely to be in full-time work than their malecounterparts, all else being equal – the largest gap inthe region – followed by Indonesia, Philippines andAfghanistan (Figure 2.12, Panel 2).


baseline:individualswith no education

Indo

nesia

Russ

ian Fe

dera

tion

Thail

and

Singa

pore

Turk

ey

Philip

pine

s

Uzbe

kista

n

Cam

bodi

a

Iran

(Islam

ic Re

publ

ic of

)

Kaza

khsta

n

Viet

Nam

Chin

a

Repu

blic

of Ko

rea

Lao

PDR

Nepa

l

Sri L

anka

Tajik

istan

Mon

golia

Azer

baija

n

Arm

enia

Afgh

anist

an

baseline: males

Perc

enta

ge

40

20

-80

0

-20

-40

-60

-100

Nepa

l

Indo

nesia

Philip

pine

s

Afgh

anist

anJa

pan

Uzbe

kista

n

Sri L

anka

Bhut

an

Turk

ey

Turk

men

istan

Thail

and

Bang

lades

h

New

Zeala

nd

Mya

nmar

Chin

a

Viet

Nam

Lao

PDR

Cam

bodi

aIn

dia

Azer

baija

n

Mala

ysia

Russ

ian Fe

dera

tion

Perc

enta

ge

1 200

1 000

200

800

600

400

0

2.4.4 Overall determinants

These decompositions point to the important linksbetween a mother’s education, children’s nutrition,school completion and employment prospects,particularly for the region’s developing countries. Thesepatterns are repeated across all opportunities studiedand sketch an image of the following four mostimportant drivers of inequality in access to opportunitiesin Asia and the Pacific:

• Education has a prominent role in shapinginequality in access to all opportunities.Education, when viewed as a “circumstance,”matters in different ways depending on theopportunity: for children’s nutrition, it is theeducation of a mother; and for securing full-timeemployment, it is the individual’s own level of

education. The highest education level in thehousehold is also important for determining accessto all basic household-level services, but mostlyassociated with ownership of a bank account.19

Given that basic literacy is necessary for accessing,understanding and operating banking services, thisassociation is not surprising.

• The rural-urban divide is behind much of theobserved inequality in access to opportunities.Together with education, the rural-urban divide isamong the most prevalent circumstances indetermining inequality in access to varioushousehold-based opportunities, particularly basicwater and sanitation, electricity and clean fuels, butalso individual-based opportunities like secondaryand higher education attainment. Interestingly,across all household-related opportunities, countries

Figure 2.12 Impact of education and gender in getting full-time employment

Source: Results are based on country-specific logistic regressions. Only countries with statistically significant coefficients and odds-ratios are shown.Analysis conducted with the help of ILO and using data from the Gallup World Poll.Note on Panel 2: In developed countries, the level of full-time employment may reflect a personal choice, rather than an access issue.

Panel 1: The higher odds for individuals with completed higher education to bein full-time employment, compared with those with no education

Panel 2: The odds of women being in full-time employment,compared with men


with higher D-indices (hence higher inequality) arealso those where the rural-urban divide is mostimportant.

• Gender is an important determinant of inequalityin education and full-time employment. Beinga woman explains the bulk of inequality in accessto full-time employment more frequently than anyother factor, including education. The impact ofbeing a woman or a man with respect to secondaryand higher educational attainment is interestingbecause it goes both ways, depending on thecountry.

• Wealth is overall the most common driver ofinequality of opportunity in all countries. It is themost important circumstance with respect toinequality in secondary and higher educationattainment, stunting levels, but also in access tomost household-related opportunities.20 Whilebeing a proxy for many social, economic andenvironmental disadvantages, its importance indetermining inequality of opportunity is striking andconfirms the expectation and intuition thatdisadvantages are intertwined. The prominent roleof wealth in shaping these inequalities furtheremphasizes the intergenerational inequality trap,where inequality of outcome (wealth) has a directbearing on inequality of opportunity, transmittedacross generations.

2.5 WINNERS AND LOSERS – IDENTIFYINGTHOSE FURTHEST BEHIND

Knowing that inequality of opportunity is broadlyassociated with these four circumstances opens the doorto deeper exploration of the data to see exactly whichgroups are the most marginalized and which groupshave benefitted most from development. Identifyingthese two sets of groups could help policymakers betterfocus policy and programmes to tackle inequality,particularly with regards to the provision of basic services.

Using the classification tree approach, a methodologycommonly used in data mining and popular in machinelearning, this section identifies the commoncircumstances shared by those who are most likely tolack access to the selected opportunities.21 In this newmethodological approach, an algorithm splits the valuefor each variable (access rate to an opportunity) intosignificantly different population groups based on sharedpredetermined circumstances.

These circumstances vary by opportunity, followingdifferent paths for household-based opportunities andindividual-based opportunities. In each iteration, theclassification tree ascertains groups that are most and

least advantaged. The final groups could, for example,share the circumstances of belonging to the bottom40 per cent of the wealth distribution and residing ina rural area. The circumstances used for identifyingthose furthest ahead or behind for each opportunity aresummarized in Table 2.1 and are broadly the same asthose used in the decomposition analysis of key drivers(previous section).

To illustrate how the classification tree identifies themost disadvantaged or advantaged groups, the exampleof access to professional help during childbirth in LaoPeople’s Democratic Republic is used (Figure 2.13). Theclassification tree starts at the average access rate of42 per cent. The algorithm determines that the first splitinto branches should be wealth, specifically where in thewealth distribution a woman belongs: the top 60 percent or the bottom 40 per cent. Women belonging tothe top 60 per cent group have 65 per cent access rateto professional help in childbirth, compared with only17 per cent for those in the bottom 40 group.

In the same example, the algorithm determines a secondsplit for the less advantaged (bottom 40 group) aroundthe number of children a woman has had. For their firstchildbirth, one in four women in the bottom 40 groupuses professional help. That rate falls to one in nine forsubsequent childbirths or for women with more thanone child. The rate of access to professional help alsovaries for women with more than one child: only onein ten women with no education get professional help,while one in eight of those with completed primary,secondary or higher education do. Among the womenbelonging to the top 60 group, the only further split isbased on education. Half of the women with primaryor no education access professional help, compared witheight out of ten of those with secondary or highereducation.

The group with the highest access to professional helpin childbirth is women with secondary or highereducation in households belonging to the top 60 of thewealth distribution. They have an access rate of 82 percent and represent 26 per cent of Laotian women inunion who have given birth in the past five years.Conversely, only one in ten women in the bottom40 group with no education and two or more childrenunder 5 years of age use professional help duringchildbirth. The total gap between the groups withthe highest and the lowest access is a staggering72 percentage points.

The uniqueness of the classification tree approach is thatit becomes very clear where policies should, or shouldnot, be focused to reach those furthest behind.Repeating and summarizing classification tree results for21 countries is visualized in Figure 2.14. The upper lines


Top 60Size: 51%

Access: 65%

Secondary or H

igher educatio

n

None or Primary education

Population Size: 100%Access rate: 42%

Less than two children under 5

Bottom 40Size: 49%

Access: 17%

Primary, Secondary or Higher

education

Two or more children under 5

Size: 28%Access: 12%

No education

Acc

ess

to p

rofe

ssio

nal

hel

p in

ch

ildb

irth

(%)

90

80

70

60

50

40

30

20

10

0

Women with Secondary or Higher education in Top 60 households


Women with no or Primary Education in Top 60 households


Women with less than two children under 5 in Bottom 40 households


Women with Primary, Secondary, or Higher education and two or more

children under 5 in Bottom 40 households


Women with no education and two or more children under 5 in

Bottom 40 householdsSize: 16%

Access: 10%

Figure 2.13 Classification tree highlighting differences in women’s access to professional help inchildbirth in Lao People’s Democratic Republic, 2011 (15-49 years of age)

Source: ESCAP calculations, using data from the latest DHS and MICS surveys for countries in Asia-Pacific.

Figure 2.14 Access to professional help during childbirth, Asia-Pacific countries, latest year

Source: ESCAP calculations, using data from the latest DHS and MICS surveys for countries in Asia-Pacific, latest years.Note: Data for Myanmar not available for this indicator.

Arm

enia

Kyrg

yzsta

n

Thail

and

Turk

men

istan

Kaza

khsta

n

Viet

Nam

Cam

bodi

a

Tajik

istan

Indi

a

Philip

pine

s

Vanu

atu

Bhut

an

Pakis

tan

Afgh

anist

an

Bang

lades

h

Lao

PDR

Timor

-Les

te

Mald

ives

Indo

nesia

Mon

golia

100

80

60

40

20

0

Acc

ess

rate

(%)

Average access level Group access rate (highest) Group access rate (lowest)


of each bar represent the access rate of the mostadvantaged groups in each country. The lower linesrepresent those with lowest access rates and the middleline shows the average access rate, by which countriesare also sorted. Countries in North and Central Asia andThailand fare the best with almost universal access andno substantial gaps between population groups. Bycontrast, Indonesia, the Maldives and Timor-Leste havethe lowest average access of below 40 per cent. Thelargest gaps are not found in the countries with thelowest access but in Lao People’s Democratic Republic(72 percentage points), Bhutan and Bangladesh(64 percentage points).

Overall, wealth and education levels strongly impactaccess to professional care during childbirth, wherewomen from the bottom 40 with lower levels ofeducation appear frequently among the mostdisadvantaged. In some countries age also matters, witholder women being less likely to obtain skilled personnelattendance during childbirth. Lastly, having morechildren also plays a role, suggesting either lack ofresources or awareness in the household of theimportance of professional attendance for all births.Annex 2.3 lists the circumstances of groups with lowestaccess rates.22

Identifying the furthest behind in all opportunities foreach of the 21 countries (33 countries for access to full-time employment) generates over 500 classification trees

(like the one in Figure 2.13). Each tree reveals anindividual, community- or country-based story, ofsuccess among urban educated elites, of catching-upamong rural communities through education, but alsoof marginalization, mostly in remote, minoritycommunities. The more nuanced, country-based storiesneed to be explored by policymakers and researchersworking in specific sectors in individual countries.

Summarizing the findings from the classification treesfor all opportunities, however, yields some generalpatterns (Table 2.3). The most common sharedcircumstance of the most disadvantaged households andindividuals is a low level of education (primary or below).The second most common circumstance is belonging tothe poorest 40 per cent of the national wealthdistribution. Households in rural areas are also morelikely to be in the most marginalized groups with loweraccess to basic services. Women are more likely to bein the furthest behind groups, as are younger peopleand those over 50 years of age.

On the contrary, the profiles of the most advantagedgroups in terms of access to basic household servicesis, expectedly, belonging to the richest 60 per cent of thedistribution, having a family member with at leastsecondary education in the household and living in urbanareas. For individuals, the most common circumstance isagain being among the wealthiest 60 per cent, havingsecondary or higher education and being male.

Table 2.3 Shared circumstances of the worst-off and best-off groups in access to opportunities

Common Circumstances: HOUSEHOLDS THAT ARE…

FURTHEST BEHIND FURTHEST AHEADCircumstances Count (times) Circumstances Count (times)

Lower and primary education 130 Top 60 80Bottom 40 107 Secondary and higher education 73Rural 43 Urban 69

Common Circumstances: INDIVIDUALS WHO ARE…

FURTHEST BEHIND FURTHEST AHEADCircumstances Count (times) Circumstances Count (times)

Bottom 40 of wealth distribution 80 Top 60 of wealth distribution 69Lower and primary education 74 Secondary and higher education 53Female 63 Male 50Living in a rural area 42 Living in an urban area 46Age 15-24 33 Age 25-49 28Male 16 Female 17Age 50-64 14 Age 15-24 9

Source: ESCAP calculations using data from the latest DHS and MICS surveys for countries in Asia-Pacific, latest years


2.6 PROGRESS OVER TIME: WHY POLICYMATTERS

With an overall increase in access to opportunities inrecent years, the groups of households and individualsthat are furthest behind have experienced someprogress. Comparing results from earlier and latersurveys reveals that the most marginalized groupsrepresent smaller sections of society but also that theiraccess rates to most opportunities have improved.

This analysis isolates four core opportunities to reviewprogress over time: secondary educational attainment(Figure 2.15), child stunting (Figure 2.16), access toprofessional help during childbirth (Figure 2.17), and full-time employment (Figure 2.18). It finds that in somecountries, despite economic growth and improvementsin average access, sizeable groups are being excluded.Countries that increased their investment in socialprotection, particularly in education and health care,were more successful in closing the gaps compared withthose that did not.

In all countries except Kyrgyzstan, Lao People’sDemocratic Republic and Turkmenistan, averageattainment rates for secondary education increased inthe period between the two surveys (Figure 2.15).However, it is only in Kazakhstan, the Philippines andThailand that the distance of the most marginalizedgroup from the average fell. In the remaining 12countries, the gap, in percentage point (pp) differencefrom the mean, grew between the two surveys in time;revealing exclusion of certain groups from the countries’general upward trends.

In the three countries where the gaps from the meanclosed, education had been a strong policy priority. InThailand, the introduction of the 1999 NationalEducation Act guarantees nine years of compulsoryeducation (from six to 15 years of age) and 12 years offree basic education. The free education policy wasextended to 15 years in 2009. Kazakhstan’s Constitutionof 1995 also states that the citizens shall be guaranteedfree secondary education in state educationalestablishments and that secondary education isobligatory. In the Philippines, social protectionprogrammes have prioritized education, at both theprimary and secondary levels.23

Progress has been more equitably shared with respectto children’s nutrition (Figure 2.16). Average stuntingrates fell in all countries except Armenia and Thailand.Thailand also saw an increase in the average rateof overweight children, from 10 per cent in 2005 to12 per cent in 2012.24 This finding for Thailand, anupper-middle income country, suggests that ensuringchildren’s access to the right nutrition is a complexeconomic, social and cultural issue. In Bangladesh, LaoPeople’s Democratic Republic, Pakistan and Thailand thegaps of the most disadvantaged groups (the groups withthe highest prevalence of stunting) from the averageincreased, suggesting that some children were being leftbehind from overall progress.

The most successful countries in reducing stunting ratesfor all were Cambodia, Kazakhstan, Kyrgyzstan,Mongolia and Turkmenistan. In Mongolia, the universaland unconditional Child Money Programme (initially atargeted and conditional programme) is considered to

Figure 2.15 Distance between the worst-off groups and average attainment of secondary education forindividuals 20 to 35 years of age, earliest and latest

Source: ESCAP calculations, using data from the latest DHS and MICS surveys for countries in Asia-Pacific.Note: The disadvantaged groups may not have the exact same composition in both surveys. However, the most disadvantaged groups always representat least 10 per cent of the population and have at least one common circumstance.Note 2: pp stands for percentage points.

100

80

60

40

20

0

Att

ain

men

t ra

tes

(Per

cen

tag

e o

f p

op

ula

tio

n a

ges

20-

35)

2003 2014 2010 2015 2000 2014 1991 2013 2000 2011 2006 2016 2000 2013 2003 2012 2005 2012 2006 2015 2000 2013 1998 2013 1997 2012 2000 2010 2003 2015

Cambodia Afghanistan Bangladesh Pakistan Lao PDR India Viet Nam Indonesia Thailand Turkmenistan Mongolia Philippines Kyrgyzstan Armenia Kazakhstan

-5 pp

-13 pp-10 pp

-14 pp -17 pp-17 pp

-6 pp

-28 pp -22 pp

-19 pp-16 pp

-24 pp

-18 pp

-26 pp

-23 pp

-27 pp

-27 pp

-25 pp

-6 pp-8 pp

-30 pp-48 pp

-35 pp

-33 pp

-1 pp -7 pp

-9 pp-10 pp -6 pp

-5 pp

Average attainment rate Attainment rate of the worst-off group


Figure 2.16 Distance between the worst-off groups and the average in stunting for children 0 to 5 yearsof age, earliest and latest


80

40

60

20

0

Stu

nti

ng

rat

e (P

erce

nta

ge

of c

hild

ren

)

2000 2014 2006 2016 2000 2014 2000 2011 1997 2012 2000 2013 2000 2010 2005 2012 2006 2015 2006 2015Bangladesh India Cambodia Lao PDR Kyrgyzstan Mongolia Armenia Thailand Turkmenistan Kazakhstan

8 pp

13 pp

11 pp

14 pp

8 pp

8 pp

7 pp

18 pp

15 pp

6 pp

8 pp

11 pp

13 pp 4 pp

4 pp

7 pp 4 pp 4 pp

2 pp3 pp

Average stunting rate Stunting rate of the worst-off group

Figure 2.17 Distance between the worst-off groups and the average in access to professional help duringchildbirth for women aged 15 to 49, Asia-Pacific countries, earliest and latest


2003 2015 1997 2012 2000 2010 2005 2012 2000 2013 2000 2013 1998 2013 2006 2016 2000 2014 2010 2015 1991 2013 2003 2012 2000 2011 2000 2014

Kazakhstan Kyrgyzstan Armenia Thailand Mongolia Viet Nam Philippines India Cambodia Afghanistan Pakistan Indonesia Lao PDR Bangladesh

100

40

80

60

20

0Pro

fess

ion

al h

elp

acc

ess

rate

(Per

cen

tag

e)

-0 pp -1 pp-0 pp

-4 pp-0 pp

-4 pp -5 pp-1 pp

-2 pp-1 pp

-20 pp

-30 pp

-40 pp

-32 pp

-17 pp

-35 pp

-21 pp

-19 pp-20 pp

-24 pp

-14 pp

-26 pp

-28 pp

-17 pp

-18 pp

-32 pp

-10 pp

-24 pp

Average professional help access rate Professional help access rate of the worst-off group

have had an important contribution to this development.In Cambodia, public health initiatives focusing onincreasing the interval between births and reducing useof tobacco during pregnancy have contributed to thereduction in stunting.25

Average rates of access to professional help duringchildbirth have increased in most countries (Figure 2.17).However, in most countries in South-East and South andSouth-West Asia, the distance between groups with thelowest access and the average increased. On the other

hand, Armenia, India, Kyrgyzstan, the Philippines andViet Nam saw impressive increases, both in terms ofaverage access and in closing the gaps with the mostmarginalized groups.

Viet Nam, for example, prioritized reducing maternalmortality (MDG 5) through the Strategy for Protectionand Care of the People’s Health 2001-2010, as well asthe Reproductive Health Strategy 2001-2010. In thePhilippines, the Philhealth programme introduced in1997 was designed to provide access to health care for


the underprivileged, the sick, older persons, persons withdisabilities and women and children. The continuousexpansion and strengthening of Philhealth may havecontributed to the observed progress.26

Based on alternative data sources, access to a decentjob reveals a disconnect between overall employmentgrowth and decent job growth. Behind this disconnectis the nature of economic growth.27 As populationscontinue to expand in many countries in the region,creation of decent jobs has failed to meet the risingnumbers of new labour market entrants. With noalternative, people are forced to accept whatever jobsare available.

The extent of vulnerable employment is illustrated forthe period 2000-2015 in Figure 2.18. In countries above

the diagonal line, such as Fiji, Papua New Guinea andSri Lanka, vulnerable employment increased faster thanoverall employment. In countries below the diagonalline, but above the horizontal dotted line, which is themajority of the region’s developing countries, overallemployment increased faster than vulnerableemployment, indicating a falling share of vulnerablejobs. Nevertheless, there was still an increase in theabsolute number of vulnerable workers. In Afghanistan,Bhutan and Pakistan, for example, the overallemployment increase of 60 to 80 per cent wasaccompanied by a 50 per cent increase in vulnerableemployment. Only in a few countries, located below thedotted line, did the absolute number of vulnerableworkers fall — as in China, the Russian Federation andsome OECD members.28

Figure 2.18 Change in total employment and in vulnerable employment, 2000-2015

Source: ESCAP calculation using ILO (2017), KILM (9th edition).

Vu

lner

able

Em

plo

ymen

t G

row

th (P

erce

nta

ge)

100

40

80

60

20

0

-20

-40

Employment Growth (%)

-40 -20 0 20 40 60 80 100

BTN

CHN

FJI LAO

JPN

GEO

HKG

MMR

IDN IND

GUM

THA

KOR

PRK

WSM TUR

LKA

TLS

TON

RUS

KGZ

NZL KAZ

AUS

BGD

PYFARM

MAC

TJK

PNG

VNMBRN

TKM

NPL

NCLSLB

MYS

IRNVUT

PAK

SGP

PHLAZE

MNGKHM UZB

AFG

MDV

While not all workers in the informal economy are poor,there is a frequent overlap; trapped in hazardous, low-paid jobs without any protection or security, theseworkers have few opportunities to escape poverty.29

Meanwhile, wages in Asia-Pacific are growing fasterthan in any other region, and grew by 4 per cent in2015, suggesting a widening gap between thosebenefiting from economic growth and productivityincreases and those left behind.30

2.7 CONCLUSIONS AND RECOMMENDATIONS

Positive progress in reducing the gaps between thosefurthest behind and an average household is mostly seenin countries that have prioritized investments in thesocial sector, including through social protection. The

level of economic development, whether a country is inthe low, lower-middle or upper-middle income bracket,has a bearing on inequality of opportunity, but not asmuch as its social development policies.

The superior performance of many lower-middle incomecountries from North and Central Asia, but also Bhutan,India, Mongolia, the Philippines and Viet Nam,particularly in education and health, point to thatfinding. For household-based opportunities, prioritizationof investments in basic water and sanitation, energy andfinancial services has been stronger in upper-middleincome countries. However, certain lower-middleincome countries also stand out, including most Northand Central Asian countries, India, Viet Nam and somePacific Islands.


These findings suggest that certain policies andinstitutions can help close the gaps in terms of accessto opportunities.

Broaden social protection coverage

• Social protection policies are key to reducinginequality, increasing prosperity, resilience andempowerment and ensuring that “no one is leftbehind”. Expanding social protection to all supportslow-income families through cash transfers, or otherincome-support mechanisms with strong multipliereffects on the economy as these groups tend tospend the extra income on domestic goods andservices. It also insures against risks such as disasters,illness and unemployment, impacts of which can belife-threatening, particularly for vulnerable groupswith no financial reserves.

Make education affordable, accessible, and relevantfor all

• A well-educated population is fundamental for allspheres of development. National education systemsshould therefore encourage and facilitate highereducation attainment and at the minimum improvesecondary completion rates for all populationgroups. This is particularly important for those livingin rural areas. The quality of education also needsto be strengthened by investing in teachers’education and training, school equipment andinfrastructure. It is finally critical that current curriculacorrespond to future labour market needs andsmoothen the school-to-work transition.

Ensure that health-care services are affordable,accessible and universal for life

• Access to affordable essential health care is centralto leading healthy lives and a key determinant ofequality. Poor access to affordable health-careservices, often combined with material deprivationand social exclusion, creates or perpetuatesinequality traps. As a core component of buildingnational social protection floors, countries need toinvest in universal access to a nationally defined setof goods and services, constituting essential healthcare, including maternity care, that meets the criteriaof availability, accessibility, affordability, acceptabilityand quality. As health challenges vary throughoutthe life cycle, services need to cater to the health-care needs of all ages and all parts of the countryincluding rural areas. Health care should also becomplemented by access to other services requiredto sustain the basic living conditions for good health,such as sufficient nutrition, clean drinking water,sanitation, electricity and clean fuels, as well as basicshelter.

Protect and promote the rights of women

• Women and girls are excluded from mainstreamdevelopment more often than men and boys. It istherefore paramount that their rights andparticipation be placed at the centre of allpolicymaking. Gender equality is not only afundamental principle of human rights, but is alsoa vital component to effectively meet future needsand challenges in Asia and the Pacific. Public policiesshould uphold and mainstream gender equality inall spheres of life.

Closing rural-urban gaps in public service delivery

• Physical access and mobility constraints compoundinequality of opportunities. For example, access tohealth care, education and decent jobs in rural areasare often constrained by a lack of infrastructure,including transport connections. Removing thesebottlenecks can also encourage labour mobility andcreate opportunities for income-generating activities.

Improve effective service delivery

• Strong political commitment, broad public supportas well as capable and accountable institutionsgoverned by transparent regulatory frameworks areprerequisites for effective service delivery. Ineffectiveadministration, weak rule of law, corruption, andlack of regulatory frameworks influence operationalcapacity to generate change and disproportionallyharm the poorest and most vulnerable segments ofsociety.31 Simply allocating more public resourceswithout reforming governing principles maytherefore not have the desired impact.

Encourage multi-sectorial and multi-stakeholdercollaboration

• To reach population groups at the highest risk ofbeing left behind, policy reforms need to beunderpinned by multisectoral and multi-stakeholderinvolvement at all stages, from development anddesign to implementation and monitoring. Given thediversity of circumstances impacting individual andhousehold decisions and opportunities, suchinvolvement and coordination are imperative forcreating opportunities and incentives for households.

Improve the quality of services and opportunitiesprovided

• An underexplored area is the importance of ensuringthe quality of services. Even when education andhealth services are publicly provided, they may notbe of adequate quality, pushing wealthier individualsto seek private options. Those who can afford to payprivately for better health-care or education serviceswill do so. Those who cannot are left with no option,


but the publicly available service and they may haveto settle for a lower-quality job, a disadvantagedlocation for their home and the prospect of uncleanfuels and poor sanitation solutions. Inequality ofopportunity, and gaps in the quality of opportunitiesafforded, often result from income and wealthinequality and become a driving force forintergenerational inequality and for trapping peopleand communities in a vicious cycle of persistingpoverty and exclusion.

Bolster capabilities to understand inequality ofopportunity through disaggregated data analysis

• To identify those at risk of being left behind and todirect policymaking at certain population groups,national data collection needs to allow for betterdisaggregation. Additional research also needs tocapture how unequal opportunities impact individualaspirations and household possibilities and whycertain individuals and households may, for example,take their children out of school or continue usingunclean energy options, while other, sometimesneighbouring households will not. This chapter hasused innovative analytical methods to analyseavailable surveys. However, the number of countrieswith available surveys were limited and the surveysdid not provide answers to important questions,such as the quality of education or the perception

of inequality. With the availability of more and betterdata, countries will be better placed to takeadvantage of the wide array of analytical toolsavailable to them.

Inequality of opportunity and gaps in quality are notlimited to services provided by the State. It expands todaily choices around what transportation means to use,what phone device to buy and what news sources torely on. Increasingly, these services are provided by theprivate sector. As chapter 4 will show, the incredibletechnological progress that has underpinned growthover the past decades has afforded people in Asia andthe Pacific a vast choice of products and services tochoose from. Yet, what is affordable for those earningaround US$1.90 a day is not comparable with what theelites or the growing middle class across the region canenjoy.

Before exploring the interaction of technology withinequality, chapter 3 analyses in more depth inequalitiesin the quality of the environment people live in or accessto meet basic needs and generate livelihoods. Itdescribes how disadvantaged groups are oftendisproportionately exposed to the hazards ofenvironmental degradation and less able to protectthemselves and recover from various environmentalimpacts.

ENDNOTES1 See Ravallion, (2016).2 United Nations Development Programme (UNDP) (2013).3 Ravallion, (2016).4 United Nations, Economic and Social Commission for Asia and the Pacific (ESCAP) (2018a) (forthcoming).5 Canning and Schultz (2012).6 United Nations, Economic and Social Commission for Asia and the Pacific (ESCAP) (2018d) (forthcoming).7 See World Health Organization (WHO) (2018), available at: http://www.who.int/gho/phe/water_sanitation/burden_text/en/8 See World Bank (2013), available at: http://www.worldbank.org/en/news/feature/2013/08/30/whats-a-toilet-worth-infographic9 United Nations, Economic and Social Commission for Asia and the Pacific (ESCAP) (2017a).10 United Nations, Economic and Social Commission for Asia and the Pacific (ESCAP) (2017a) Available from https://www.adb.org/sectors/energy/issues/access-energy.11 Habtezion, (2013).12 Sovacool, (2013).13 Access to full-time employment is used as a proxy for decent work in this paper. Decent work is not easy to measure. Toidentify gaps in the labour force in terms of access to decent work, this analysis uses the “employed full-time for an employerindex” in the Gallup survey as a proxy. In practice, this index is a subset of the ILO’s ‘non-vulnerable’ employment classification,which includes wage and salaried workers together with employers. The ‘vulnerable,’ on the other hand, are own accountworkers and contributing family members. For more information, please see United Nations, Economic and Social Commissionfor Asia and the Pacific (ESCAP) (2018b) (forthcoming).14 The decomposition of other opportunities can be found in a series of policy papers available here: http://www.unescap.org/our-work/social-development/poverty-and-inequality/resources


15 The D-index for stunting measures the inequality in distribution of non-stunted children among all groups of children.16 United Nations, Economic and Social Commission for Asia and the Pacific (ESCAP) (2018c) (forthcoming).17 United Nations, Economic and Social Commission for Asia and the Pacific (ESCAP) (2018a) (forthcoming).18 United Nations, Economic and Social Commission for Asia and the Pacific (ESCAP) (2018b) (forthcoming).19 Data on the D-index decomposition is available upon request.20 The wealth index, which has been used for this analysis, is a composite index reflecting a household’s cumulative livingstandard that is developed by the DHS and MICS researchers and combines a wide range of household assets and characteristics.21 A full description of the methodology and algorithm is available in the Annex of the relevant Policy Paper series, such asESCAP (2018a).22 The Annex 2.3 tables do not show the composition of the most advantage group (highest attainment rate), but thisinformation is available upon request.23 United Nations, Economic and Social Commission for Asia and the Pacific (ESCAP) (2018e) http://www.socialprotection-toolbox.org/practice/philippines-conditional-cash-transfer-families24 United Nations, Economic and Social Commission for Asia and the Pacific (ESCAP) (2018c) (forthcoming).25 Ikeda et.al (2013).26 United Nations, Economic and Social Commission for Asia and the Pacific (ESCAP) (2018e). http://www.socialprotection-toolbox.org/practice/philippines-health-care-programme27 International Labour Organization (ILO) (2009). http://www.ilo.org/wcmsp5/groups/public/@dgreports/@dcomm/documents/publication/wcms_115087.pdf28 United Nations, Economic and Social Commission for Asia and the Pacific (ESCAP) (2017b).29 International Labour Organization (ILO) http://www.ilo.org/wcmsp5/groups/public/—ed_emp/—emp_policy/documents/publication/wcms_212689.pdf30 International Labour Organization (2016).31 United Nations, Department of Economic and Social Affairs (DESA), United Nations Development Programme (UNDP), UnitedNations Educational, Scientific and Cultural Organization (UNESCO) (2012).

Documents

Chapter 2 Addressing Inequality of Opportunity in Asia and ... › sites › default › d8files › 04Chapter2.pdfdistribution lie in the work of John Rawls and Amartya Sen. Rawls