can the mdgs provide a pathway to social justice? the challenge of intersecting inequalities

Can the MDGs provide a pathwayto social justice? The challenge

of intersecting inequalitiesNaila Kabeer

Institute of Development Studies

Acknowledgements 2

List of abbreviations 2

List of figures, tables and boxes 3

Foreword 5

Executive summary 6

1 Introduction: the fundamental values of the Millennium Declaration 11

2 Intersecting inequalities and social exclusion 132.1 Relevance to the MDG agenda 132.2 The regional history and geography of social exclusion 14

3 Intersecting inequalities and the MDGs: the regional picture 163.1 Progress on the MDGs in Latin America 173.2 Social exclusion and the MDGs in Latin America 173.3 Progress on the MDGs in Asia 213.4 Social exclusion and the MDGs in Asia 233.5 Progress on the MDGs in sub-Saharan Africa 263.6 Social exclusion and the MDGs in sub-Saharan Africa 26

4 The intersecting dynamics of inequality: why social exclusion persists 304.1 The cultural dynamics of exclusion 304.2 The economic dynamics of exclusion: asset inequalities 324.3 The economic dynamics of exclusion: disadvantageous livelihoods 334.4 The dynamics of exclusion in service provision: access and quality 354.5 The political dynamics of exclusion 38

5 The MDGs as a pathway to social justice: equalising life chances 395.1 Responsive states, active citizens: towards a new social contract 395.2 Strengthening information policies to tackle exclusion 425.3 Macroeconomic policies and redistributive growth 435.4 Progressive fiscal policies 445.5 Legal policies and affirmative action 455.6 Land reform 475.7 Promoting livelihoods, decent work and access to credit 485.8 Investing in infrastructure and area development 505.9 Extending basic services to all groups 515.10 Inclusive social protection 54

6 Conclusion: key concerns and principles 57

Further information 60

Endnotes 61

Can the MDGs provide a pathway to social justice? 1

Contents

Copyright © 2010by the United Nations Development Programme1 UN Plaza, New York, NY 10017, USA

All rights reserved. No part of this publication may bereproduced, stored in a retrieval system or transmitted,in any form or by any means, electronic, mechanical,photocopying, recording or otherwise, without priorpermission.

The analysis and policy recommendations of this reportdo not necessarily reflect the views of the MDGAchievement Fund or the United Nations DevelopmentProgramme. The report is an independent publicationcommissioned by the MDG Achievement Fund and is thefruit of a collaborative effort with the Institute ofDevelopment Studies, Sussex.

ISBN: 978 1 85864 947 1

About the authorNaila Kabeer is Professor of Gender andDevelopment at the School of Oriental andAfrican Studies, University of London.

AcknowledgementsWe wish to thank Professor Jo Beall (Universityof Cape Town/London School of Economics),Professor Ashwini Deshpande (Delhi School ofEconomics, University of Delhi) and ProfessorEdward Telles (Princeton University), whocontributed regional background papers for thisreport. We also wish to thank Gabriele Koehlerand Professor Hilary Standing for providinginputs on key themes, and Selim Jahan fromUNDP and Minar Pimple from the UNMillennium Campaign for their comments.Excellent contributions were made by theresearch team: Peroline Ainsworth (research co-ordinator), Luisa Natali, Karin Seyfert,Devika Lal, Harsh Malhotra, Line FalkenbergOllestad and Smriti Sharma.

We also benefited from extremely useful andconstructive comments from our readers group:Dr Sarah Cook (UNRISD), Jan Vandemoortele(formerly with the UN; now independentresearcher and writer), Kevin Watkins(UNESCO) and Alexander Cobham (ChristianAid). Special thanks to Layla Saad from the UNMDG Achievement Fund for her role inoverseeing the process and her valuable inputsthroughout. Thanks finally to Kathryn O’Neillfor editorial work and to Barbara Cheney fordesigning the report.

Cover Kuala Lumpur, Malaysia: the Petronas TwinTowers and the luxury Radisson Hotel rise behindslum housing in the centre of the city. Photographer: Mark Henley/Panos Pictures

List of abbreviations

ADB Asian Development BankDHS Demographic and Health SurveyECLAC Economic Commission for Latin America and the

CaribbeanFGM female genital mutilationGDP gross domestic productGNP gross national productHDI human development indexIADB Inter-American Development BankIDP internally displaced personILO International Labour OrganizationINDISCO The Interregional Programme to Support Self-

Reliance of Indigenous and Tribal Peoples through Cooperatives and Self-Help Organizations

LAC Latin America and the CaribbeanMDGs Millennium Development GoalsMPCE monthly per capita expenditureNAFTA North American Free Trade AgreementNGO non-governmental organisationODA Official Development AssistanceODI Overseas Development InstituteOECD Organisation for Economic Co-operation and

DevelopmentPPP purchasing power paritySSA sub-Saharan AfricaUN ESCAP United Nations Economic and Social Commission for

Asia and the PacificUNESCO United Nations Educational, Scientific and Cultural

OrganizationUNRISD United Nations Research Institute for Social

DevelopmentWHO World Health Organization


List of figuresFigure 3.1 Global progress on achieving the MDGs 16

Figure 3.2 Ratio of indigenous to non-indigenous infant mortality rate, 2000–02, selected Latin American countries 18

Figure 3.3 Average number of years of education for indigenous people aged 17–22, selected countries 18

Figure 3.4 Latin America (14 countries): extreme poverty rates among indigenous peoplesand Afro-descendants, as a multiple of the rate for the rest of the population (dollar-a-day line) 19

Figure 3.5 Under-five mortality rate by wealth quintile and the ratio of poorest to richestquintile 21

Figure 3.6 Nepal: poverty Head Count Rate (HCR) trends by social groups 21

Figure 3.7 Percentage of adult population who have not completed any education, Vietnam 23

Figure 3.8 Rural China: distribution of poverty among ethnic minorities and the majoritypopulation, 2002 24

Figure 3.9 Western region of rural China: distribution of poverty among ethnic minoritiesand the majority population, 2002 24

Figure 3.10 Malaysia: infant mortality rate by ethnic group, 1970–2006 25

Figure 3.11 Malaysia: poverty Head Count Rate (HCR) trends by ethnicity 25

Figure 3.12 Prevalence of child malnutrition by residence, Nigeria 27

Figure 4.1 The Nepal caste pyramid according to the Muluki Ain (1854) 30

Figure 4.2 Nepal: composite empowerment and inclusion index (CEI) bygender/caste/ethnicity 31

Figure 4.3 Brazil: pay levels by gender, ethnic group and years of schooling, 1992 and 2002 32

Figure 4.4 Pastoralists facing extreme education deprivation 35

Figure 4.5 Nepal: ethnic/caste and gender representation in Parliament, 1959–99 36

Figure 5.1 Maternal mortality rates, selected states in Mexico 54

List of tablesTable 3.1 Absolute and relative overall progress of the MDGs: top 20 achievers 16

Table 3.2 Institutional coverage in municipalities where indigenous people are a majority, minority or moderate proportion of the population, Bolivia, 2006 17

Table 3.3 Poverty incidence (MDG 1) by household per capita income in selected Latin American countries 18

Table 3.4 Primary school enrolment rates in selected Latin American countries 19

Table 3.5 Comparing years in education of indigenous and non-indigenous groups in Bolivia 19

Table 3.6 Education indicators by region, Mexico, 2005 19

Table 3.7 Brazil: percentage of the population living below the extreme poverty line by race 20

Table 3.8 Brazil: distribution of the poorest 10% and the richest 1% by race (2005) 20

List of figures, tables and boxes

4 List of figures, tables and boxes

Table 3.9 Comparison of child malnutrition: eastern and western China (%) 21

Table 3.10 Poverty incidence for different social groups, selected states in India 22

Table 3.11 India: under-five mortality rate for different social groups, 1992/93 to 2005/06 23

Table 3.12 Nepal: early childhood mortality rates, 2006 23

Table 3.13 Access to public services among different groups in Vietnam 24

Table 3.14 Nigeria: childhood mortality rates by residence 27

Table 3.15 Socioeconomic indicators for selected regions in Ghana (%) 27

Table 3.16 South Africa: individual level poverty by race and gender (poverty line R515 per capita per month) 28

Table 3.17 Estimates of annual per capita personal income by race group in 2000 28

Table 3.18 Relative per capita personal incomes by race group (% of white level) 28

Table 3.19 Nigeria: states with the highest and lowest shares of poverty, 2006 29

Table 4.1 Time to reach nearest health facility in north and south Ghana (% of respondents) 34

Table 5.1 Collective land rights in new constitutions and agrarian codes 49

List of boxesBox 2.1 Intersecting inequalities and the persistence of social exclusion 15

Box 3.1 Institutional coverage of deliveries in Bolivia 17

Box 3.2 Distribution of income and extreme poverty in Brazil 20

Box 3.3 Poverty and social inequality in India 22

Box 3.4 Child health inequalities in India and Nepal 23

Box 3.5 Intersecting ethnic and spatial inequalities in Vietnam and China 24

Box 3.6 The intersection of spatial, ethnic and religious inequalities in Ghana and Nigeria 27

Box 3.7 Focus on South Africa 28

Box 3.8 Ethnic inequalities in child mortality in Africa 29

Box 4.1 The caste system in Nepal 30

Box 4.2 Why social exclusion persists: internalising inferiority 31

Box 4.3 Social exclusion and violent crime: life in the favelas of Brazil 32

Box 4.4 The Kamaiya system in Nepal 33

Box 4.5 Intersecting ethnic, religious and political inequalities in Sudan 37

Box 5.1 Making everyone count: collecting census data on excluded groups in Latin America 41

Box 5.2 The Right to Information in India 43

Box 5.3 Progressive hydrocarbon taxation in Bolivia 45

Box 5.4 Using equity-weighted transfer rules to spend more on excluded groups 45

Box 5.5 Affirmative action in Nepal 46

Box 5.6 Affirmative action in Malaysia 46

Box 5.7 Tackling institutionalised racism in Brazil 47

Box 5.8 Land rights campaigns in India 48

Box 5.9 Informal workers organise to claim rights 49

Box 5.10 Mobilising around maternal mortality in Mexico 53

Box 5.11 Reducing maternal mortality in Mexico 54


In September 2010, ten years after signing theMillennium Declaration and committing tothe Millennium Development Goals (MDGs),

world leaders will convene at the United NationsGeneral Assembly to take stock of how far theyhave come in delivering on their promises toeradicate extreme poverty and hunger, and ensureadequate education, health and gender equalityfor all. The gains that have been made with regardto these goals, as well as in reducing child andmaternal mortality and halting the spread ofdiseases, are undeniable. Some of the poorestcountries in the world have shown remarkableprogress with scarce resources, demonstrating thatpolitical will is one of the primary factorscontributing to development success.

This report on the MDGs and social justiceargues that despite these gains, the focus onaggregate progress and the use of nationalaverages to measure countries’ performancedisguises a picture of uneven achievement that ischaracterised by deep disparities between socialgroups. In every country in every region, peopleare being excluded from the opportunity to playan active role in social and economicdevelopment on the mere basis of their race,ethnicity, religion, gender and, often, location.

With the clock ticking rapidly to 2015, thisreport comes at a critical juncture for the MDGs.It aims to put the challenge of social inequalityonto the 2010 MDG Review Summit agenda andto demonstrate that equity is just as important as

aggregate progress if we want the MDGs to besustainable and address the root causes ofinsecurity and instability.

It also demonstrates that social exclusion andits resulting inequalities are not only rooted inthe denial of people’s social and economic rights,but are also inextricably linked to the lack ofvoice and participation of marginalisedpopulations, thereby placing civil and politicalrights on a par with social and economic rights.It asserts that fostering the growth of democraticinstitutions like the media and the rule of law isinstrumental in tackling exclusion and the denialof rights, while pushing for greater transparencyand accountability on behalf of governmentstowards their citizens.

This report renews the call for social justicemade by the Millennium Summit in the year 2000.It outlines policy recommendations andinterventions that can be integrated into 2010–15MDG Action Plans to tackle inequalities so thatdevelopment benefits all groups in society. It callson governments of rich countries as well as poorcountries, and on citizens and civil societyorganisations to put the issue of social exclusionand inequality firmly at the centre of the fightagainst poverty and all efforts to achieve theMillennium Development Goals.

Layla SaadAdvocacy AdvisorUN MDG Achievement Fund

Foreword

6 Executive summary

At the start of the 21st century, 189 worldleaders signed the Millennium Declarationand pledged to achieve the Millennium

Development Goals (MDGs) – a set of eightbenchmarks to eradicate extreme poverty andhunger, achieve universal primary education,promote gender equality and empower women,reduce child mortality, improve maternal health,combat major diseases, ensure environmentalsustainability and develop a global partnershipfor development by the year 2015. Thiscommitment provided a ringing call to action tothe international community to concentrateefforts on freeing men, women and children fromthe abject and dehumanising condition ofpoverty. The Declaration was based on a set offundamental values that included freedom,equality, tolerance and solidarity, whichtogether spell out a firm commitment to socialjustice as the guiding spirit of all efforts.

Unfortunately, the social justice agenda was lostin the process by which the Declaration wastranslated into an agenda for action centred on theeight MDGs and their targets and indicators tomonitor progress. While the MDGs acknowledgethe multiple dimensions of poverty, they pay verylittle attention to inequality or social injustice. Theyare largely formulated in terms which measure‘average’ progress in relation to the goals. Whilethese measures capture overall progress at global ornational level, they do not indicate whether suchprogress has been broad-based or equitable.

The starting point for this report is thatinequalities matter for the achievement of theMDGs. Inequalities matter at the macroeconomiclevel because they slow down the pace at which agiven rate of growth translates into povertyreduction. They also matter for society at largebecause they generate high levels of socialtensions, crime and conflict, with adverse effectsfor human wellbeing and progress. In addition,inequalities ensure that the poorer sections of theworld’s population are generally bypassed by‘average’ rates of progress on the MDGs.

This report argues that it is the socially excludedsections of the poor who are systematically left outor left behind from their countries’ progress. Theirexcluded status is the product of multiple andintersecting inequalities. Along with the economicdeficits generally associated with poverty, excludedgroups face additional discrimination on the basis ofsocially marginalised identities such as race,ethnicity, caste and sometimes religion or language,with gender cutting across these various groups.Socially excluded groups suffer from spatialinequalities; they tend to be concentrated indisadvantaged locations – remote and challengingrural terrains or overcrowded slumneighbourhoods. Social, economic and spatialinequalities in turn contribute to political exclusion;such groups are generally denied voice and influencein collective decisions that affect their lives.

It is the intersecting – and mutuallyreinforcing – nature of these inequalities thatmakes socially excluded groups harder to reachthan other sections of the poor. It ensures thatthey do not benefit to the same extent, if theybenefit at all, from national progress on theMDGs. It gives their poverty an enduring quality,lasting over a lifetime and often over generations.

This report brings together evidence fromLatin America, South/Southeast Asia and sub-Saharan Africa to demonstrate the relationshipbetween social exclusion and the MDGs and todeepen understanding of its underlyingdynamics. It focuses particularly on thosecountries for which data are available on howsocially excluded groups are faring in relation toMDG-related objectives concerned with extremepoverty, children’s education and maternal andchild health. These data support the maincontention of this report: the persistence ofintersecting inequalities undermines progress onthe MDGs and betrays the promise of socialjustice contained in the Millennium Declaration.

Latin America continues to be among themost unequal regions in the world. Theseinequalities have strong ethnic, racial and spatialdimensions. Extreme poverty is much higheramong indigenous and Afro-descendantpopulations compared with the white/Latinopopulation in most countries.

Asia, as a region, is on track to meet the targeton reducing extreme poverty, primarily due to

Executive summary

The persistence of intersecting inequalitiesundermines progress on the MDGs and betrays thepromise of social justice contained in the MillenniumDeclaration


the extraordinary performance of China in recentyears. However, this has been accompanied byincreasing income inequalities in many of theregion’s countries, including China. It is alsoevident from the data that the decline in povertyhas been uneven across different social groups:ethnic minorities remain over-represented in thepoorest strata. Caste and religious differences arealso reflected in unequal health, education andpoverty outcomes.

Sub-Saharan Africa, as a region, is most offtrack in relation to the MDGs, but some countriesin this region are among the most successful in thedeveloping context when progress is measured inabsolute terms. However, data on how sociallyexcluded groups are faring in relation to theMDGs are, with the exception of South Africa, notreadily available. Data for the countries covered inthe report point to spatial, religious and ethnicvariations in terms of progress on MDGs.

The evidence on the relationship betweensocial exclusion and the MDGs reviewed for thereport allows for two broad generalisations. Thefirst is that the intersecting and mutuallyreinforcing nature of the inequalities which giverise to social exclusion helps to explain itspersistence over time and its apparent resistance tothe forces of change. The second is that theseinequalities are not immutable; while progress onthe MDGs in relation to excluded groups has beenextremely slow and uneven, there is, nevertheless,evidence that it has occurred.

There are policy lessons to be learned fromboth these generalisations. Understanding theintersecting dynamics that reproduce socialexclusion over time points to the domains in whichpolicy must operate and the underlying causalitiesthat it must address if progress on the MDGs is tobe accelerated. Understanding the dynamics ofchange, and the role of policy in making changehappen, provides lessons on what has worked indifferent contexts and how this can be adapted toaddress similar problems in other contexts.

The cultural dynamics of social exclusion – thesystems of norms and beliefs that define somegroups as inferior to others – do not merelypromote discriminatory behaviour towardssubordinated groups. Persistent reminders of theirinferiority erode the self-confidence and sense ofself-worth of members of these groups and theirbelief in the possibility of escaping poverty. Highlevels of substance abuse, criminality and conflictare often consequences of such frustration.

The economic dynamics of social exclusionoperate through inequalities in the distributionof productive assets and livelihood opportunities.These economic dynamics can operate through

explicit norms such as legal restrictions on landownership for specific social groups, and theirconfinement to predetermined, and generallystigmatised, occupations. Or they can operatethrough the discriminatory behaviour of others,making it harder for excluded groups to translatethe assets at their disposal into sustainedpathways out of poverty.

The policy dynamics of social exclusion reflectthe availability, affordability and terms ofprovision of the basic services essential to theachievement of the social MDGs. There is a veryclear spatial dimension to this dynamic: areas inwhich marginalised groups are located tend to bethe most underserved by essential services.Financial constraints also constrain access. Thiswas dramatically illustrated by the fall inutilisation of health and educational serviceswhen user fees were imposed as part of structuraladjustment programmes.

There are also problems of quality andrelevance of services. The failure to reflect theneeds, priorities and constraints of marginalisedgroups in the provision of health and education,the language used to communicate, the behaviourdisplayed towards members of these groups – allof these shortfalls in the quality of service help toexplain the poor pace of progress on the healthand education objectives of the MDG agenda.

The political dynamics of social exclusionoperate through the denial of the opportunity tosocially excluded groups to participate in thecollective decision-making processes of theircommunity and society. Formal democraticprocesses are unlikely to be sufficient to overcomehistorically entrenched exclusions, particularly

Vietnam: minority hill-tribe woman with corn from her field,pauses on the road to Dien Bien Phu, near the Laos border.Tanks and other ordnance from the famous 1954 battle still litterthe landscape. Photographer: Chris Stowers/Panos Pictures

8 Executive summary

since these groups are often minorities in theirsociety. The denial of political voice and influenceto socially excluded groups in a society has oftenbeen a major driver of violent conflict, as thesegroups have no other forum to express their voice.

The intersecting dynamics of social exclusion,and their resilience over time, suggest thatpiecemeal policy efforts in relation to the MDGsare unlikely to have a discernible impact. Thereport makes a number of general policyrecommendations for tackling inequality andsocial exclusion and then considers examples ofmore MDG-specific interventions which haveworked in different contexts.

At the most general level, the construction ofa more inclusive social contract betweenresponsive states and active citizens offers themost enabling governance framework for theholistic approaches needed to tackle socialexclusion. Such a contract would incorporatecommitment to transparency, accountability,democratic participation and civic values as thebasic foundation of the relationship between astate and all its citizens. These commitmentswould be given operational form in differentpolicy interventions to address social exclusion.

Given their isolation from the mainstream oftheir society, the organisation of socially excludedgroups and their mobilisation around self-identified needs, interests and priorities is acritical precondition for their participation in theconstruction of a new social contract. Suchorganisations can be built as a deliberate aspectof policy through the setting up of citizens’

committees, user groups of various kinds andconsultative exercises. Or they may be built frombelow through the self-organisation of excludedgroups or through the support of civil societyintermediaries.

A comprehensive policy on information isnecessary if social exclusion is to be tackledthrough the MDGs. States need to collectappropriately disaggregated data on all theircitizens in order to track their progress – orfailure to progress – as a result of developmentefforts. Such data need to be made widelyavailable, to be subject to critical analysis andused to inform future policy directions as well ascivil society advocacy efforts. Informationpolicies should also produce information forexcluded groups so that they are moreknowledgeable about their entitlements andrights, and more confident about claiming them.

Different channels of information should alsobe used to address discriminatory attitudes andbehaviour among the general public and topromote the values of tolerance, respect andsolidarity. The media can be a powerful ally inthe fight against discrimination and should bemobilised to educate, inform and entertain inways that help to break down the barriersbetween excluded groups and the rest of society.

It is evident that recent patterns of economicgrowth have not always reduced, and have, in fact,exacerbated economic inequality and socialexclusion. Macroeconomic frameworks whichpromote broad-based growth and a generalexpansion of opportunities are much more likely toreach marginalised sections of society. At the sametime, growth on its own is unlikely to overcomethe barriers that have held them back in the past.

Inclusive patterns of growth will need to besupplemented by redistributive policies that canserve to directly address the intersecting dynamicsof social exclusion. Fiscal policies are an importantinstrument for redistribution. In terms ofdomestic revenue mobilisation, the redistributivepotential of fiscal policies will depend on whichsectors, groups and activities are taxed andwhether taxes are direct or indirect, progressive orregressive. Taxation policies are likely to havegreatest relevance to economic MDG objectives, asthey have a direct bearing on livelihood activities.On the expenditure side, the allocation of budgetsto different sectors and services and thedistribution of social transfers and subsidies willhave a direct bearing on the financing of socialservices and social protection measures of greatestrelevance to poor and excluded groups.

Legislation against discrimination is animportant signal of state commitment to social

Colombian boy from the mountainous indigenous region ofPaletará peers through the window. The community has longbeen fighting for their rights to land and livelihood.Photographer: Luis Alfonso Ortega Fernandez


justice and it can provide advocate groups with thesupport they need to take action. However, wheresocial inequalities are deeply entrenched, morepositive action may also be necessary. Manycountries have sought to take affirmative actionon behalf of excluded groups in order to breakwith past patterns and to set new precedents forthe future. Affirmative action may take the formof reserved places for members of excludedgroups in economic, political and educationalsystems or they may operate through recognitionof the explicit rights of minority groups.

Economic exclusion has been addressedthrough a variety of means intended to strengthenthe resource base of excluded groups. Landreform and land titling programmes areimportant routes for transferring assets toexcluded groups. Labour regulations provideformal protection against the exploitation ofvulnerable workers, but they are most likely to beenforced when workers are sufficiently organisedto exert the necessary pressure. Microfinance hashelped to overcome some of the limits of formalfinancial services, but has not proved effective inreaching the most marginalised or promotinggraduation out of poverty. Microfinance serviceprovision needs to be supplemented with othersupportive services as well as pathways into themainstream financial sector.

The spatial concentration of many sociallyexcluded groups lends itself to area-baseddevelopment. Economies of scale may allowvarious aspects of spatial disadvantage to beaddressed simultaneously and cost-effectively.These include transport and communication,water and sanitation, social service infrastructureand the provision of services.

Improving the outreach, quality and relevance ofbasic social services is a critical precondition forachieving the social MDGs. The abolition of userfees, and social transfers to offset costs or provideincentives to take up services (children’s education,maternal healthcare), are powerful mechanisms toaddress this. The report points to variousexamples of how services can be made morerelevant and user-friendly to men, women andchildren from excluded groups, includingrecruiting and training staff from excludedcommunities, relying on female staff in gender-segregated societies, ensuring services are providedin languages understood by those who need them,the use of mobile schools and clinics. What thesevarious policy options share in common is thatthey seek to address some aspect of the underlyingdynamics of social exclusion.

Social protection measures have assumedincreasing importance in promoting livelihoods and

hence improving the resilience of poor people inthe face of crisis. Those found to be mostbeneficial to excluded groups include conditionaland unconditional transfers, often targeted tochildren, the elderly or the poorest sections ofsociety. The conditionality associated withtransfers may relate to work obligations, toencouraging women to make use of maternityservices or to ensuring children go to school andfor health checkups. Social transfers clearly have aredistributive effect, but where they are providedon a transparent, regular and predictable basis,they have also been found to have importantdevelopmental impacts. These include theincreased capacity to participate in local labourmarkets, to upgrade skills and knowledge, toinvest in productive assets, to access credit, alongwith multiplier effects in the local economy.

There is sufficient evidence from across theworld that addressing social exclusion through theMDGs can be achieved, but not through a‘business-as-usual’ approach. The experiences ofprogress reported by different countries andanalysis of the policies that have worked suggest anumber of key principles that can serve as thebasis for broader strategies to address exclusion.

The importance of a rights-based frameworkSocial exclusion entails the denial of fullpersonhood and full citizenship to certain groupson the basis of who they are, where they live orwhat they believe. Efforts to overcome exclusionshould be located within the broad structure ofrights, including the Universal Declaration of

Niger: The Fulani are nomadic people who have travelled andsettled across the African continent. Many Fulani in Niger live astheir forefathers did, maintaining their traditional culture.Photographer: Giacomo Pirozzi/Panos Pictures

10 Executive summary

Human Rights and the international conventionson civil and political rights and on social,economic and cultural rights. These provide aninternationally agreed basis on which to tacklethe violation of the basic rights of such groups.

Building tolerance and solidarity amongcitizensWhile responsive states and actively engagedcitizens are essential to building a moredemocratic society, relations between citizens areas important as the relationship between statesand citizens. Social exclusion is partlyperpetuated by the discrimination that citizenspractise against each other. Legislating against thepractice of discrimination, along with educating,informing and challenging long-held prejudices,is an essential aspect of building democracy.

Getting the right balance between equalityand difference Broad-based or universal policies to promoteequality of opportunity, and special provision forthose who have been systematically excluded inthe past, can work successfully in tandem.Universalist approaches are essential to building asense of social solidarity and citizenship,particularly critical for excluded groups. At thesame time, the fact that it is their ‘difference’from the rest of the poor that has led sociallyexcluded groups to be left behind or locked outof processes of growth and development suggestsstrong grounds for plurality and diversity withinuniversal frameworks of provision.

Beyond amelioration to transformationIt is quite possible to meet the basic needs ofpoor and marginalised groups withoutstrengthening their capacity to do so themselves,thereby leaving their longer-term vulnerabilityintact. Addressing the root causes of socialexclusion and adopting integrated andtransformative approaches are necessary in orderto shatter the resilience of social exclusion.

Group-based exclusion requires group-basedsolutionsRelated to this is the inadequacy of policies thattarget individuals or households in tacklingproblems that are essentially collective andgroup-based. Individual solutions may leave

marginalised groups more isolated andimpoverished than before. The need for morecollective approaches introduces another routethrough which ‘difference’ may have to be builtinto the design of policies.

A new social contract for an interconnectedworld: states, citizens and MDG 8 The problems of poverty and social exclusion arenot purely national in their causes orconsequences. They are the product of structuralinequalities at the global level. This is notadequately acknowledged in the MDGs. WhileMDGs 1 to 7 concern objectives to be achievedby developing countries, MDG 8 encompassesthe relationships between developed anddeveloping countries. Not only does it fail toaddress the unequal nature of these relationships– as manifested in aid, trade and debt – but it isalso the only MDG with no targets or indicatorsto monitor any form of progress.

It is extremely unlikely that developingcountries will be able to achieve growth,prosperity and social justice without greaterattention to building greater solidarity, a genuinepartnership of equals, at the global level. Thisshould constitute the central platform for thepost-2015 successor to the MDGs. However, inthe run-up to 2015, rich countries shouldprioritise a number of actions to accelerate thepace of progress:

Honour the commitment to increase OfficialDevelopment Assistance (ODA) to 0.7 percent of gross national product (GNP).Recognise and support the role of civil societyorganisations in mobilising excluded groups,holding governments to account, andtransforming power relationships.Provide support for a universal socialprotection floor that allows marginalisedgroups to cope with risk, invest in their futureand lobby on their own behalf.Promote fairer trade relations, with specialattention to the needs of the working poor inthe informal economy.Hold corporations and the private sector toaccount for socially responsible investments.Strengthen government capacity andcommitment to the international humanrights framework.


The Millennium Declaration was signed in 2000by 189 of the world’s leaders on behalf of theinternational community. It represents a

promise to coordinate and accelerate efforts to ‘freeour fellow men, women and children from the abjectand dehumanizing conditions of extreme poverty’ by2015.1 It is based on a set of fundamental values thatinclude freedom, equality, tolerance and solidarity.Together, these spell out a firm commitment to socialjustice as the guiding spirit of the Declaration.

Unfortunately, this commitment to social justicewas not carried over into the eight MillenniumDevelopment Goals (MDGs), which sought to giveactionable content to the Declaration through anumber of objectives, targets and indicators. TheMDG agenda has been welcomed for its integratedand multi-dimensional approach to the challenge ofpoverty reduction – one that addresses incomedeprivation, deficits in human capabilities, and lack ofaccess to basic services such as health and education.Its major limitation is the failure to incorporateconcrete measures on equality and social justice.

‘The integrated treatment of multipledisadvantages – the “more than poverty”element of social exclusion – is already amajor element of the development agenda.The Millennium Development Goalsdemonstrate a recognition that a wholepackage of areas of progress is required, andthat poverty, although of crucial importance,cannot be tackled in isolation…’2

The measures used to monitor progress onthe MDGs are couched in terms of national

‘averages’ and ‘proportions’. For example, two ofthe indicators for measuring progress on the keygoal of eradicating extreme poverty (MDG 1) arehalving, between 1990 and 2015, the proportionof people who live on less than $1 a day andthose who suffer from hunger. Such indicatorscapture overall progress at global, regional andnational levels, but they do not show whether thisprogress has been equitable. Only one indicatorunder MDG 1 touches on inequality: the share ofthe national income that goes to the poorestincome quintile. But this is also the measure thatfeatures least frequently in MDG reports.3

MDG 3 deals explicitly with gender equality (ineducation, paid work and political participation)but does not address how the most deprivedwomen might fare within overall progress onthese indicators.

This report takes, as its starting point, thegrowing body of evidence that inequality imposescosts on society. It imposes costs at themacroeconomic level, because it slows down thepace at which a given rate of economic growthtranslates into poverty reduction.4 And it imposescosts at the societal level, because of its impacts interms of social tensions, crime, violence and

Introduction: the fundamental values ofthe Millennium Declaration 1

‘Men and women have the right to live their lives and raise theirchildren in dignity, free from hunger and from the fear of violence,oppression or injustice...’ (Millennium Declaration 2000, para. 6).Photographer: Barbara Cheney

Social justice and the Millennium Declaration

‘Men and women have the right to live their lives andraise their children in dignity, free from hunger and fromthe fear of violence, oppression or injustice… The equalrights and opportunities of women and men must beassured… Those who suffer or who benefit least deservehelp from those who benefit most… Human beings mustrespect one another, in all their diversity of belief, cultureand language. Differences within and between societiesshould be neither feared nor repressed, but cherished asa precious asset of humanity.’

(Millennium Declaration 2000, para. 6)

12 Introduction: the fundamental values of the Millennium Declaration

conflict, and their knock-on effects on investmentsin both human and economic capital.5

The costs of inequality mean that the poorestsections of the world’s population are generallybypassed by ‘average’ rates of progress on theMDGs. This report contends that deprivation isnot randomly distributed. Rather, itdisproportionately affects certain groups that facediscrimination on the basis of their socialidentities. These groups are harder to reach thanthe rest of the poor population, giving theirpoverty an enduring quality and underminingthe pace of progress on the MDGs.

‘Sometimes poverty arises when people haveno access to existing resources because of whothey are, what they believe or where they live.Discrimination may cause poverty, just aspoverty may cause discrimination…’6

This report aims to bring together evidence insupport of this argument and to deepen ourunderstanding of the processes that perpetuatethe systematic exclusion of certain social groups.While it recognises that social exclusion is evidentin wealthy OECD (Organisation for EconomicCo-operation and Development) countries as wellas in the transitional economies, our focus is ondeveloping regions, specifically Latin America,sub-Saharan Africa (SSA) and South/SoutheastAsia. Progress on the MDGs within these regionsand countries has been far from uniform. Therehas been more progress in some countries thanothers, and on some goals than others. We willanalyse examples of such progress in order to

draw out lessons for promoting the MDGs as apathway to social justice.7

Unfortunately, the problems of socialexclusion in relation to certain groups manifestthemselves in a dearth of data about theirsituation. There are very few references to theeffects of intersecting inequalities in most of theMDG reports. A great deal of the discussion inthis report is, therefore, necessarily dictated bythe availability of data. We focus, first, oncountries that show marked evidence of thecategory of inequalities that we are interested in.Second, we focus on those countries for which wehave disaggregated data that reveal the extent ofthese inequalities. These are not necessarily thecountries with the most marked forms ofinequality. Indeed, some of the most unequalcountries have very little data because they areaffected by conflict (or have been until recently).

Third, even where countries have disaggregateddata, they may not give uniform coverage to all theMDGs. As a result, we will be focusing primarilyon the MDGs dealing with income poverty, healthand education, the most consistently addressed inthe disaggregated data. Despite these limitations,the countries selected are sufficiently diverse andoffer sufficiently robust evidence on patterns ofexclusion to support the key contention of thisreport: the persistence of intersecting inequalitiesundermines progress on the MDGs and betrays thepromise of social justice contained in theMillennium Declaration.

‘Data from wealthy OECD countries showthat regardless of levels of per capita GNP(gross national product), higher levels ofinequality within a society are associated withhigher levels of crime, violence, obesity,mental illness, prison populations, teenagepregnancies and anxiety. Equality, in otherwords, is better for everyone.’8

The report is structured as follows. Section 2discusses the phenomenon of intersectinginequalities, the exclusions they give rise to, andtheir relevance to the MDGs. Section 3 trackspatterns of equality in the Latin America, sub-Saharan Africa and South/Southeast Asia regionsin relation to selected MDGs. Section 4 examinessome of the factors which explain the persistenceof these inequalities, while Section 5 focuses onthe policies and interventions that hold out thepromise of transformative change. Finally,Section 6 draws out some broad principles forthe promotion of equality, including the values oftolerance and solidarity spelt out in theMillennium Declaration.

Two Mauritanian girls look shyly at the camera in their homevillage in the sparsely populated desert area between Senegaland Mauritania. Photographer: Layla Saad, MDG Achievement Fund


2.1 Relevance to the MDG agenda‘Human diversity is no secondarycomplication (to be ignored or introduced‘later on’): it is a fundamental aspect of ourinterest in inequality.’9

As the key indicators to measure progress forMDG 1 suggest, poverty has been largelyunderstood in terms of individual deprivation. Theconventional approach to the measurement ofpoverty is based on the economic characteristics ofindividuals or individual households – primarilytheir income or wealth. The subsequent ranking ofhouseholds by these characteristics gives rise towhat has been described as a ‘vertical’ model ofinequality.10 The focus on the national incomeshare of the poorest 20 per cent (the poorestquintile) of the income distribution that featuresin MDG 1 is derived from such a model.

The literature on social exclusion, on theother hand, takes group-based disadvantages (or‘horizontal’ inequalities11) as its entry point intothe discussion of inequality. These disadvantagesare products of social hierarchies which definecertain groups as inferior to others on the basisof their identity. Such hierarchies are createdthrough cultural norms and practices which serveto disparage, stereotype, exclude, ridicule anddemean certain social groups, denying them fullpersonhood and equal rights to participate in theeconomic, social and political life of their society.

‘[Social exclusion is] the process throughwhich individuals or groups are wholly orpartially excluded from full participation inthe society in which they live.’12

The identities in question may take differentforms in different societies, but the moreenduring forms of poverty in most contexts areassociated with identities that are ascribed frombirth, such as race, caste, ethnicity – as well asreligious beliefs when they are in a minority.Gender cuts across these inherited identities, sothat within most groups, women are subordinateto men. However, unlike most sociallysubordinate groups, women are distributed fairlyevenly across different economic classes, so thatgender on its own does not constitute a markerof poverty. It is the intersection of gender with

economic and other disadvantages that explainsthe intensified nature of the disadvantages facedby poorer women and girls.

Identity-based inequality intersects with otherforms of inequality to define social exclusion. Itsintersection with economic inequalities – the factthat socially excluded groups face particularbarriers in gaining access to resources andopportunities – means that those most likely tobe left behind in national progress on the MDGsare disproportionately drawn from ethnic andreligious minorities, from racially disadvantagedgroups and from the lowest castes. Women andgirls from these groups are frequently at a greaterdisadvantage.

Social exclusion also frequently entails spatialinequalities, consigning culturally devalued andeconomically impoverished groups to inhabitphysically deprived spaces. Spatial disadvantagein rural areas tends to be associated withremoteness, ecological vulnerability, low levels ofagricultural productivity, a meagre resource base

Intersecting inequalitiesand social exclusion 2

A child being weighed by the village health volunteer while thedistrict nurse supervises in the Tigray region of Ethiopia.Photographer: Sophie de Caen, MDG Achievement Fund

14 Introduction: the fundamental values of the Millennium Declaration

and weak integration into the national economy.In urban contexts, it is associated with poor-quality housing and sanitation, inadequateservices and infrastructure and, very frequently,with subcultures of criminality, violence, drugdependence and squalor.13

Finally, there is an inter-temporal dimension tosocial exclusion. The intersecting inequalities thatconstitute social exclusion, and the cumulativedisadvantage they entail, explain why it persistsover time (see Box 2.1). Social exclusion deniesthe groups who experience it the ‘normal’ escaperoutes out of poverty, reinforcing inequalities overpeople’s lifetimes, and often over generations.

2.2 The regional history and geography ofsocial exclusion The exclusion of certain groups on the basis ofsocially ascribed identity generally has deep rootsin a country’s history. In some countries, it is tiedup with experiences of colonisation – frequentlyaccompanied by the genocide of indigenouspopulations – and with slavery, war and conflict,as well as the continued practice ofdiscrimination into the present day. In others, itis linked to concepts of superiority-inferioritydrawn from religious or cultural beliefs.

In the Latin American context, race and

ethnicity, involving categories of white,indigenous, black and mixed race, are the keymarkers of social exclusion. There are more than50 million indigenous peoples and more than120 million individuals of African descent (Afro-descendants) in Latin America and the Caribbean(LAC), comprising around 33 per cent of thepopulation.14 These groups are not uniformlydistributed across the region. For instance, morethan 25 per cent of the population of Bolivia,Guatemala, Peru and Ecuador are indigenous,while more than 25 per cent of the population inPanama, Brazil, Nicaragua and almost all theCaribbean countries are Afro-descendants. Theymake up just 6 per cent in Uruguay and have anegligible presence in Argentina.

There is a spatial dimension to social exclusionwithin countries as well.15 In each country exceptBrazil, over 45 per cent of the indigenous or Afro-descendant population live in rural areas.16

Indigenous groups tend to live in the more remoteand hard-to-reach parts of their countries, oftenpushed out of more productive areas by non-indigenous groups.17 Afro-descendants live mainlyalong coastal areas of the Pacific and AtlanticOceans and the Caribbean Sea, and in urban areas.

Across the Asia region, social exclusion reflectsethnic, tribal and indigenous identities, but religion

Yunnan Province, China: members of the Bai ethnic minority, who traditionally wear blue clothing. Photographer: Jeremy Horner/Panos Pictures


and language also feature in some countries. As inLatin America, indigenous ethnic minorities inAsia are often located in remote geographical areas.And as in Latin America, this location has notalways been a matter of choice. The mountainranges that stretch from Afghanistan to the Gulf ofTonkin have long been a refuge for indigenouscommunities who have, for various reasons,occupied a marginal position in relation to thedominant majorities in the valleys and plains.

Ethnic and indigenous groups make up around8 per cent of the population in China, 10 per centin Vietnam, 8 per cent in India and 37 per cent inNepal. They are to be found in the poorest areas:rural areas of China’s western region; the remote,usually upland, mountainous areas of northernand central Vietnam; the hilly and forested regionsof India, Bangladesh and Nepal.

In India, spatial concentration means thatindigenous minorities or Adivasis comprise amuch higher proportion of the population insome states than others. More than 80 per cent ofindigenous minorities are found in nine states incentral-western India: Orissa, Madhya Pradesh,Chhattisgarh, Jharkhand, Maharashtra, Gujarat,Rajasthan, Andhra Pradesh and West Bengal. Mostof the rest are in the north east of the country.

In addition, India and Nepal, withpredominantly Hindu populations, arecharacterised by a caste hierarchy that goes backmany centuries. Caste is closely intertwined withan occupational hierarchy that assigns menialand ritually polluting occupations to the lowest‘untouchable’ castes or Dalits. Dalit groupscomprise 17 per cent of the population in India.They tend to be more geographically dispersedthan indigenous groups, but around 80 per centlive in rural areas.18 Nepal’s population is dividedbetween a Hindu majority (58 per cent as of2001), Janajatis (indigenous minorities) (37 percent), and religious minorities, mainly Muslims(around 4 per cent).19 Around 12 per cent of thepopulation is made up of Dalits.

Sub-Saharan Africa is home to more than2,000 distinct ethnic groups characterised bydifferent language, culture and traditions, and,sometimes, religious beliefs. Ethnic groups inAfrica vary in size from millions of people to a fewhundred thousand, and are often associated with aspecific territory. Much of the region continues to

suffer from the carving up of the continent bycolonial powers in 1884 with scant regard forexisting social, political, ethnic and linguisticcontours. As a result, the political geography of theregion, which has more countries than any otherregion of the world, has long been characterisedby regional and civil armed conflicts. Added tothese deep historic divisions is a physicalenvironment that is characterised by longdistances and low population density.

In countries like South Africa and Zimbabwe,where colonial powers had a strong presence,intersecting inequalities have a strong racialdimension. More pervasive across the sub-continent are social cleavages associated withethnicity, frequently reinforced by geographicallocation and distance from main urban centres.Religion itself may not necessarily be a driver ofexclusion unless it intersects with ethnic orregional disparities.

Social exclusion also frequently entails spatialinequalities, consigning culturally devalued andeconomically impoverished groups to inhabitphysically deprived spaces

Women selling clothes in Soweto on the outskirtsof Johannesburg. Photographer: Barbara Cheney

Box 2.1 Intersecting inequalities and the persistence ofsocial exclusion

Social exclusion is the product of intersecting inequalities:

The cultural devaluation of certain groups, deeming theminferior to others in ways which undermine their sense ofself-worth and agencyTheir disadvantaged position in the distribution of valuedresources, services and opportunities within their societyTheir location in places where their efforts yield poorer returnsThe denial of voice and influence in the decisions that affecttheir lives and their communitiesThe mutual interaction between these inequalities, whichleads to their persistence over time.

16 Intersecting inequalities and the MDGs: the regional picture

Aglobal overview of progress on the MDGsshows that the world is closer to meetingsome goals and targets than others (see

Figure 3.1). It is on track in relation to extremepoverty, gender parity in primary and secondaryeducation enrolment and access to safe drinkingwater. But there is still some way to go to achievegoals and targets relating to universal primaryeducation completion rates, access to sanitation,and the proportion of people living in hunger. Itis most seriously off track for meeting theobjectives relating to under-five child mortalityand maternal mortality.20 The MDGs discussed inthis report thus include one on which there hasbeen considerable progress (MDG 1: eradicatingextreme poverty and hunger), one where there issome way to go (MDG 2: achieving universalprimary education) and two that remain a majorchallenge (MDG 4: reducing child mortality, andMDG 5: improving maternal health).

The global picture is useful in drawingattention to some of the more intractabledimensions of poverty facing the worldcommunity. A more disaggregated picture isnecessary to establish how different countries,and socially excluded groups within them, arefaring. As we have noted, measuring ‘average’progress at the national level tends to concealmajor inequalities at the sub-national level.

It is also important to note that even thoughcountries may be off track in terms of progressrelative to global MDG targets, they havenonetheless made considerable progress inabsolute terms, measured as overall progress.This is particularly the case for sub-SaharanAfrica, where a recurring theme in policydiscussions is that it will be ‘off target’ for everyone of the MDGs by 2015.21 Comparing overallabsolute progress on the MDG goals with relativeprogress in relation to MDG targets puts this‘Afro-pessimism’ in perspective.22 It reveals that:

11 African countries featured among the 20top countries in terms of absolute overallprogress on the MDGs but only 3 countriesfeatured in terms of relative progress6 Asian countries featured among the 20 topperformers in terms of absolute progress and4 featured in terms of relative progress3 Latin American countries featured amongthe top 20 performers in terms of absoluteprogress and 9 featured in terms of relativeprogress.

Intersecting inequalities and the MDGs:the regional picture3

0

20

40

60

80

100

MDG 1

.aex

trem

e p

over

ty

MDG 1

.chu

nger

MDG 2

prim

ary

com

plet

ion

rate

MDG 3

gend

er p

arity

(prim

ary)

MDG 3

gend

er p

arity

(sec

onda

ry)

MDG 4

child

mor

talit

yun

der fiv

e

MDG 5

mat

erna

lm

orta

lity

MDG 7

.cac

cess

to

safe

wat

er

MDG 7

.cac

cess

to

sani

tation

◆ distance to goal achieved◆ distance to goal to be on track to achieve the target by 2015

Figure 3.1 Global progress on achieving theMDGs

Source World Bank (2010) Global Monitoring Report:The MDGs After the Crisis

Table 3.1 Absolute and relative overall progress onthe MDGs: top 20 achievers

Top 20 Top 20Absolute progress Progress relative to MDG targets

Benin EcuadorMali ChinaEthiopia ThailandGambia BrazilMalawi EgyptVietnam VietnamUganda HondurasNepal BelizeIndia NicaraguaCambodia ArmeniaBangladesh KazakhstanHonduras Sri LankaMauritania CubaGhana MexicoChina El SalvadorBurkina Faso BeninRwanda ChileNicaragua MalawiGuatemala GambiaTogo Guatemala

Source ODI (2010)22


3.1 Progress on the MDGs in Latin AmericaLatin America, as a region, falls into the middle-income category. Historically, it has been one ofthe most unequal regions in the world, with thepoorest income quintile earning around 3 percent of the total regional income.23 But progress isevident. Poverty declined from around 44 percent in 2002 to 33 per cent in 2008, whileextreme poverty declined from 19 to 13 percent.24 Most countries in the region are on trackto meet the goal of halving extreme poverty by2015, but a few of the least developed countrieslag behind: Bolivia, Haiti, Guatemala, Honduras,Nicaragua and Paraguay.

There are grounds for cautious optimism withregard to equity of progress on the MDGs, asthere is evidence of a marked decline in incomeinequality between 2000 and 2007 in 12 out of 17countries for which data are available.25 Theseinclude many of the countries with a highpercentage of socially excluded groups, such asEcuador, Brazil, Bolivia and Mexico.

3.2 Social exclusion and the MDGs in LatinAmericaHistorically, Latin America has been one of the mostunequal regions of the world... but there aregrounds for cautious optimism with regard toequity of progress on the MDGs.

‘The concepts of social exclusion and inclusionemphasize how the benefits of development,social interaction networks, and politicalparticipation are distributed inequitably… Socialexclusion manifests itself in Latin America mostclearly in persistent unequal income distribution,which gives rise to poverty that is worse than theregion’s level of development would suggest.’26

Latin America is one of the few developingregions that has systematically collected andsynthesised data disaggregated by ethnicity inrelation to the MDGs. These allow some degreeof comparison across the region, from whichsome general patterns can be discerned:

While moderate and extreme poverty havedecreased, extreme poverty remains higheramong indigenous people and Afro-descendantsthan among the white population in manycountries (see Table 3.3 on page 18 andFigure 3.4 on page 19).Ethnic identity remains associated withinequalities in enrolment rates at all levels ofeducation, although there is some narrowing ofethnic disparities in many countries for primaryeducation (see Table 3.4 on page 19). The

broad-based expansion of basic education isone of the key factors behind the decline inincome inequality in the region.27 In somecountries, though, group-based inequalitiesremain high (between indigenous and non-indigenous groups in Bolivia, for example, seeTable 3.5 on page 19).Children from indigenous groups are morelikely to die than those from non-indigenousgroups (see Figure 3.2 on page 18). Accordingto the most recent Demographic and HealthSurvey (DHS) data from Bolivia, Ecuador,Guatemala, Peru and Brazil, children ofindigenous origin were also between 1.6 and2.5 times more likely to be undernourishedthan children of non-indigenous origin.Gender and ethnic inequalities intersect witheconomic status to keep indigenous women fromlow-income households at the bottom of thehierarchy in education (see Figure 3.3 on page18) and health. Maternal mortality is generallymuch higher among indigenous communities,particularly those in remote areas, where it canbe two or three times the national average. Forinstance, in Ecuador in 2003, the nationalmaternal mortality rate was 74.3 per 100,000live births, whereas it was 250 per 100,000among remote indigenous communities.28

Table 3.2 Institutional coverage in municipalities whereindigenous people are a majority, minority or moderateproportion of the population, Bolivia, 2006

Percentage Minority Moderately Majorityof deliveries indigenous indigenous indigenousthat occurred (under 33%) (33–66%) (over 66%)in a health facility

No % No % No %

<50% 22 44.9 14 26.4 116 52.5≥50% <70% 11 22.4 19 35.8 58 26.2>70% 16 32.7 20 37.7 47 21.3

Source Adapted from UDAPE/CIMDM (2006) Progreso de losObjetivos de Desarrollo del Milenio, Cuarto Informe asociados alDesarrollo humano, La Paz, Programa de las Naciones Unidas parael Desarrollo (PNUD)

Box 3.1 Institutional coverage of deliveries inBolivia

In Bolivia, estimated to have one of the highestmaternal mortality ratios in Latin America, a highproportion of deliveries do not take place withininstitutions. Table 3.2 shows that less than half ofdeliveries happen in healthcare facilities in 52.5per cent of municipalities classified as ‘majorityindigenous’.


Table 3.3 Poverty incidence (MDG 1) by household per capita income in selected Latin American countries

Country Year 1 USD per day Official extreme poverty line

White Non-white Total White Non-white Total

Bolivia 1997 14.8 45.3 28.7 27.8 57.5 41.32002 16.6 37.1 27.7 25.5 45.8 36.6

Brazil 1995 4.7 12.7 8.3 17.7 40.9 28.22002 5.1 10.6 7.6 17.7 38.0 27.0

Chile 2000 3.4 9.4 3.5 4.5 3.1 4.5Colombia 1999 15.2 24.1 16.2 20.5 31.0 21.7Ecuador 1998 29.0 61.0 30.3 16.0 38.9 16.9Guatemala 2002 11.4 31.5 20.0 14.3 39.1 24.9Mexico 1992 10.7 53.3 12.8 No figures available

2002 12.3 40.9 13.9 14.5 47.9 16.4Nicaragua 1998 24.5 46.0 24.9 17.0 37.7 17.5

2001 15.0 35.0 15.8 14.5 33.9 15.2Panama 2002 9.3 54.7 12.4 19.2 74.4 23.0Paraguay 1995 2.7 24.3 11.8 4.8 24.5 13.1

2001 2.2 17.3 9.9 5.0 24.9 15.2Peru 2001 14.6 26.9 19.5 18.7 33.5 24.6

Source Busso et al. (2005)14

0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

22.6*50.3* 36.9*

75.9* 37.1*54.6*

29*

82.2*59.3*

54.1*

Chile

(200

2)

Gua

tem

ala

(200

2)

Hon

dura

s(2

001)

Bol

ivia

(200

1)

Bra

zil

(200

0)

Mex

ico

(200

0)

Cost

a Ric

a(2

000)

Para

guay

(200

2)

Ecua

dor

(200

1)

Pana

ma

(200

0)

Rat

io

Figure 3.2 Ratio of indigenous to non-indigenous infantmortality rate, 2000–02, selected Latin American countries

*Indigenous infant mortality rateSource Del Popolo, F. and Oyarce, A. (2005) ‘América Latina,Población Indigena: Perfil Sociodemigráfico en el Marco de laConferencia Internacional sobre la Población y el Desarollo y delas Metas del Milenio’, Notas de Población no. 79, Santiago deChile: CELADE

Bolivia Peru Colombia Guatemala0

2

4

6

8

10

Aver

age

year

s of

edu

cation

Country average

Indigenous female

Poor, indigenous female

Indigenous

Figure 3.3 Average number of years of education forindigenous people aged 17–22, selected countries

Source UNESCO (2010)107

Spatial inequalities intersect with ethnicity. Table 3.6(page 19) demonstrates the intersection betweenspatial inequality, ethnicity and education in Mexico.The poorest southern states (for example, Chiapas,Oaxaca and Guerrero), home to around 75 per centof the country’s indigenous peoples, report far lowerlevels of education. These states also reportedmaternal mortality ratios that were higher than the

national average.29 In Guerrero, for example, the ratewas 128.2 deaths per 100,000 live births – nearly fivetimes the rate in the more developed northern stateof Nuevo León (26.9 per 100,000).30 There has,however, been an overall decline in maternalmortality rates which these states, with the exceptionof Guerrero, participated in, which is discussedfurther below.


0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0Co

sta

Ric

a(2

001)

Hai

ti(2

001)

Colo

mbi

a(1

999)

Hon

dura

s(2

003)

Peru

(200

1)

Ecua

dor

(199

8)

Bra

zil

(200

2)

Bol

ivia

(200

2)

Nic

arag

ua(2

001)

Gua

tem

ala

(200

2)

Chile

(200

2)

Mex

ico

(200

2)

Pana

ma

(200

2)

Para

guay

(200

1)s u ) r ) l ) a ) a ) a ) e ) o a y )

Extrem

e po

verty

rate

s am

ong

indi

geno

us p

eopl

es a

ndAfro-

desc

enda

nts/

extrem

e ra

te for

res

t of

pop

ulat

ion

Figure 3.4 Latin America (14 countries): extreme poverty rates among indigenous peoples and Afro-descendants, as a multiple of the rate for the rest of the population (dollar-a-day line)

Source Machinea J.L.; Bárcena, A. and León, A. (2005) The Millennium Development Goals: A Latin American andCaribbean Perspective, Santiago, Chile: Comisión Económica para América Latina y el Caribe

Table 3.4 Primary school enrolment rates in selectedLatin American countries

Country Year White Non-white Total

Bolivia 1997 0.95 0.90 0.932002 0.95 0.93 0.94

Brazil 1995 0.93 0.87 0.902002 0.97 0.95 0.96

Chile 2000 0.99 0.98 0.99Colombia 1999 0.90 0.89 0.90Ecuador 1998 0.92 0.87 0.92Guatemala 2002 0.88 0.77 0.83Mexico 1992 0.94 0.81 0.93

2002 0.97 0.98 0.97Nicaragua 1998 0.84 0.82 0.84

2001 0.89 0.81 0.89Panama 2002 0.98 0.85 0.96Paraguay 1995 0.92 0.86 0.89

2001 0.95 0.91 0.93Peru 2001 0.96 0.97 0.97

Source Busso et al. (2005)14

Table 3.5 Comparing years in education of indigenous andnon-indigenous groups in Bolivia

Average years of education 2000 2007 (population 15+) (provisional)

National 7.7 8.7Indigenous 6.0 7.2Non-indigenous 9.7 10.5

Source Data from Unidad de Análisis de Políticas Sociales yEconómicas (UDAPE) www.udape.gov.bo/, accessed 30 June 2010

Table 3.6 Education indicators by region, Mexico, 2005

Fewer than 4 yearsof education

(%)

National average 11

Disadvantaged southern statesChiapas 26Guerrero 19Michoacán 14Oaxaca 16Veracruz 20

Selected northern and central statesBaja California 6Distrito Federal 3Mexico 6

Data for population aged 17–22.Source UNESCO (2010)107

Historically, Latin America has been one of themost unequal regions of the world

Table 3.7 Brazil: percentage of the population living below theextreme poverty line by race

1993 2005

Afro-descendants 15 6Whites 4.9 2.5

Table 3.8 Brazil: distribution of the poorest 10% and therichest 1% by race (2005)

Poorest 10% Richest 1%

Whites 26.5 88.4Afro-descendants 73.5 11.6

Source Adapted from IPEA (2007) Millennium DevelopmentGoals: National Monitoring Report/Coordination, Institute forApplied Economic Research and the Secretariat for StrategicPlanning and Investment

Box 3.2 Distribution of income and extremepoverty in Brazil

Brazil experienced a steep decline in the level ofextreme poverty among Afro-descendants between1993 and 2005. But Afro-descendants still comprise74 per cent of households in the bottom incomedecile (while white people comprise just 27 percent), and only 12 per cent of the top decile (whitepeople comprise 88 per cent). The intersection ofethnic and spatial inequalities is very marked inBrazil: eight of the ten poorest states are in thenorth east, which has the highest concentrations ofAfro-descendants, while three of the four states inthe south east are among the five richest in thecountry. In addition, the intersection of ethnicity,gender and class inequalities means that whitemen generally earn the highest wages for any givenlevel of education, while black women earn theleast (see Figure 4.3 on page 32). Race, gender andlocation were among the key predictors of povertyin 1981, and remained so in 2001.


Mexico City: Alfredo (26) sells candy at a busy intersection to support his family. He has four children andlives in Mansion Mazahua, a crumbling colonial building which is home to 42 families of MazahuaIndians. Forced to migrate to the city in search of a livelihood, they live in cramped conditions in theearthquake-damaged building. Photographer: Aubrey Wade/Panos Pictures


3.3 Progress on the MDGs in Asia The Millennium Development Goals Report 2010suggests that the Asia-Pacific region as a whole ison track to achieve the target on extreme poverty,despite the setbacks resulting from the globaleconomic crisis. This is primarily due to theextraordinary performance of China. Manycountries in the region have achieved majorreductions in income poverty at the aggregatelevel as a result of strong economic growth, butsome have seen a significant increase in incomedisparities. The poorest 20 per cent of theregion’s population have seen their share of thenational income drop steeply. Between 1990 and2004, it fell from 7.2 per cent to 6.7 per cent inSouth Asia, and from 7.1 to 4.5 per cent in EastAsia. Out of 20 countries, 14 saw their Giniindex, which measures how unequal a country is,rise, while only 6 saw it fall.

Growing inequalities within the region havefocused attention on the need for inclusive growth.The 2007 UN ESCAP (Economic and SocialCommission for Asia and the Pacific) report on theMDGs attempts to examine progress ‘beyondnational averages’.31 Figure 3.5 illustrates theimplications of rising income inequality for childmortality. It shows the ratio of under-five mortalityrates between the poorest and richest quintiles inselected countries. In all the countries included, thepoorest 20 per cent of the population account formore, sometimes considerably more, than 20 percent of child mortality.

However, the ESCAP report only concernsincome, gender and rural–urban inequalities,testifying to the dearth of available data that aredisaggregated by ethnicity.

3.4

3.2

2.6

3.1

3.5

1.82.4

1.4

2.0

1.7

2.1

1.9

1.5

0 20 40 60 80 100Per cent

Turkmenistan

Nepal

Armenia

Bangladesh

Kyrgyzstan

Uzbekistan

Cambodia

Kazakhstan

Indonesia

India

Turkey

Philippines

Vietnam

low second third fourth high

Source ESCAP/ADB/UNDP (2007)31

0

10

20

30

40

50

60

70

Nepal Newar Taraimiddle caste

Brahman/Chhetri

Muslim Other Hill Janajati Tarai Janajati DalitsNeepalepal NNNewarNNewar TaraiTarai Braahma /n/ahhman MuMuslimuslilim OOtherOther Hill JanajatiJanajjat Tarai JanajatiJJan jaja DalitslDaliits

◆ 1995/6 ◆ 2003/04

Figure 3.6 Nepal: poverty Head Count Rate (HCR) trends by social groups

Source Constructed using 2005 data from Nepal Central Bureau of Statistics, cited in the Nepal HumanDevelopment Report (2009)

Table 3.9 Comparison of child malnutrition: eastern andwestern China (%)

Year Eastern Western provinces provinces

1992 Underweight 15.1 21.4Stunting 33.6 42.7




Source Ministry of Foreign Affairs of the People’s Republic of Chinaand United Nations System in China (2008) China's ProgressTowards the Millennium Development Goals: 2008 Report

Figure 3.5 Under-five mortality rate by wealth quintile andthe ratio of poorest to richest quintile


Table 3.10 Poverty incidence for different social groups, selected states in India

States Total Hindus Muslims All othersUrban All Hindus SCs/STs Other Hindus

2004– 1993–4 2004– 1993–4 2004– 1993–4 2004– 1993–4 2004– 1993–4 2004– 1993–405 05 05 05 05 05

All India 29 33 27 31 46 51 22 26 44 47 16 23Bihar 42 34 38 31 70 52 33 26 57 46 4 29Assam 7 8 5 6 7 14 5 5 13 22 4 0Andhra Pradesh 34 38 32 37 48 47 27 35 49 49 22 26Gujarat 18 28 16 25 21 42 15 22 34 47 2 23Rajasthan 27 31 26 28 48 47 18 23 39 56 7 14Tamil Nadu 28 40 29 40 50 57 25 36 24 46 21 34Orissa 50 41 49 40 74 57 42 35 51 68 51 24

Source Government of India (2006) Social, Economic and Educational Status of the Muslim Community in India, SacharCommittee Report, http://zakatindia.org/Files/Sachar%20Report%20(Full).pdf, accessed 15 August 2010

Box 3.3 Poverty and social inequality in India

India is one of the many countries in the region that has experienced economic growth accompanied bypoverty reduction, but enduring and sometimes rising inequality. As Table 3.10 shows, poverty has beenfalling between 1993 and 2004 for all social groups, but has been systematically higher among Dalit(scheduled caste, or SC) and Adivasi (scheduled tribe, or ST) groups, and, to a lesser extent, the Muslimminority. The geographical concentration of Adivasis in certain states of India, such as Bihar and Orissa,means that incidence of poverty in these states is much higher than the national level. Caste, ethnicityand geographical inequalities reinforce one another and create pockets of extreme poverty impactingcertain groups: for example scheduled castes and tribes living in Orissa have a poverty incidence of 74compared to ‘other Hindus’ (non-SC or ST) living in Assam state with a poverty incidence of only 5.

Dalit boys work feeding granite blocks into stone-crushing machines by a quarry. Photographer: Mark Henley/Panos Pictures


3.4 Social exclusion and the MDGs in Asia‘Developing Asia is now embracing inclusivegrowth as a key development goal because ofrapidly rising income and non-incomeinequalities, leading to increasing concernsthat the benefits of Asia’s spectacular GDPgrowth are not being equitably shared, whichcould derail the growth process itself…Inclusive growth is growth with equalopportunity that emphasizes both the creationof and equal access to opportunities for all…the latter [inequalities due to differences inindividual circumstances] is often reflective ofsocial exclusion associated with market,institutional, and policy failures.’32

There are no systematic data on identity andgroup-based inequalities that span the whole ofthe South and Southeast Asia region. We cantherefore only elicit trends and patterns from thecountry-level data available. It must beemphasised that Asian countries have very diverseand sometimes contrasting experiences. However,there are some general observations worth noting:

Caste, ethnic and religious inequalities areevident in education and health outcomes. InIndia, Dalits (scheduled castes, SC) and Adivasis(scheduled tribes, ST) have lower educationlevels. In Nepal, despite improvements in overallliteracy rates from 23 per cent in 1981 to 54 percent in 2001, 52 per cent of Hill Dalits, 47 percent of Tarai Dalits, 48 per cent of Muslims and

Table 3.11 India: under-five mortality rate for different socialgroups, 1992/93 to 2005/06

Under-five mortalitySC ST OBC Others Total

1992/3 149.1 135.2 NA 111.5 109.31998/9 119.3 126.6 103.1 82.6 94.92005/06 88.1 95.7 72.8 59.2 74.3

Source National Family Health Survey, 1, 2 and 3, India

Table 3.12 Nepal: early childhood mortality rates, 2006

Caste/ethnicity and Infant Under-fiveregional identity mortality mortality

Brahman/Chhetri 59 76Tarai/Madhesi other castes 64 86Dalit 68 90Newar 36 43Janajati 59 80Muslim 68All hill/mountain groups 58 77All Tarai/Madhesi groups 65 84All Nepal 55 68

Source Bennett, L.; Dahal D. and Govindasamy, P. (2008) Caste,Ethnic and Regional Identity in Nepal: Further Analysis of the2006 Nepal Demographic and Health Survey, Calverton,Maryland, USA: Macro International Inc.

0

10

20

30

40

50

60

70

80Kinh and Chinese Ethnic minorities• ◆

1993 1998 2002 2004 2006

◆

◆

◆

◆◆

•

•

•

• •

Figure 3.7 Percentage of adult population who havenot completed any education, Vietnam

Although educational outcomes are improving for allgroups in Vietnam, the gap between the Kinh and Chinesemajority and ethnic minorities persist.Source Vietnam Development Report 2008: SocialProtection, Joint Donor Report to the Vietnam ConsultativeGroup Meeting Hanoi, 6–7 December, 2007

Box 3.4 Child health inequalities in India andNepal

In India, there have been reductions in under-fivemortality across all social groups (Table 3.11). Butscheduled castes (SC), scheduled tribes (ST) and‘other backward castes’ (OBC) continue to suffermuch higher than average rates of under-fivemortality. The gap between rates of mortalityamong scheduled tribes and ‘others’ increasedbetween 1992 and 2006.

Nepal’s steady progress in reducing under-fivemortality has been accompanied by wideninginequalities: the ratio of under-five mortality ratesbetween the poorest and richest quintiles rosefrom 2.8 in 1997 to 3.4 in 2002. Only 23 per cent ofchildren from the majority Kinh/Chinesecommunity were underweight (‘wasted’) comparedwith 34 per cent of children from ethnic minoritiesin the northern mountains, and 45 per cent in thecentral highlands and coastal areas.

Table 3.12 shows that the Newars (an indigenouscaste group) had the lowest rates of childmortality, followed by the Brahman/Chhetri castes.Dalit and Janajati groups reported the highest ratesof child mortality.

Box 3.5 Intersecting ethnic and spatial inequalities in Vietnam and China

There is a marked spatial dimension to ethnic disparities in Vietnam. Ethnic minorities have to travelfurther to get to a school or marketplace, are further away from all-weather roads and are less likely tohave access to improved water and sanitation facilities. This partly explains why educational attainmentsand health indicators remain lower among ethnic minority groups, although they have been improving.Ethnic disadvantage varies according to location; ethnic minorities living in the lowlands have seen rapidpoverty reduction, while those in the northern mountains, the central highlands and the south and northcentral coasts remain in extreme poverty.

In China, ethnic minorities, who are largely concentrated in the western region, remain at a cleardisadvantage; they comprise 8.4 per cent of the population, but accounted for 46 per cent of people livingin extreme poverty in 2003. It is likely that the association between ethnicity and location is driving ethnicdifferentials in income. Figure 3.9 shows that within the western region, there are no significantdifferences in poverty rates of ethnic minority and majority groups. This is in contrast to the aggregatedpoverty distributions shown in Figure 3.8. Regression results confirm that differences in location, size ofhousehold, the head of household’s age and education, and party membership explained much of theethnic differences in poverty.

Table 3.13 Access to public services among different groups in Vietnam

Travel time Travel time Within 2 km of Access to Access toto lower- to market an all-weather improved improved secondary (minutes) road (% of water sources sanitationschool population (% of (% of(minutes) population population

2006Kinh and Chinese 12 9 91 87 43Ethnic minorities 25 32 80 55 7Vietnam 15 13 90 82 38

Source Vietnam Development Report 2008: Social Protection, Joint Donor Report to the VietnamConsultative Group Meeting Hanoi, 6–7 December, 2007

0

5

200 300 400 500 600 627 700 800 878 900 1,000

10

15

20

25

30

• ••

•• •

•

•

••

•

◆

◆◆

◆

◆◆◆

◆◆◆◆

Perc

enta

ge o

f pe

rson

s

Minority in rural China Majority in rural China• ◆

Per capita income (yuan)

0

5

200 300 400 500 600 627 700 800 878 900 1,000

10

15

20

25

30

•◆

Perc

enta

ge o

f pe

rson

s

West minority West majority•

••

•

•

•

•

••

•

•

◆

Per capita income (yuan)

◆ ◆◆

◆◆

◆

◆

◆

◆

◆

Figure 3.8 Rural China: distribution of poverty amongethnic minorities and the majority population, 2002

Figure 3.9 Western region of rural China: distribution ofpoverty among ethnic minorities and the majoritypopulation, 2002

Source Gustafsson, B. and Sai, D. (2009) ‘Temporary and Persistent Poverty among Ethnic Minorities and the Majorityin Rural China’, Review of Income and Wealth, Series 55, Special Issue 1: 588–606



30 per cent of Hill Janajatis had never been toschool, compared with just 12 per cent of theupper castes.33 Similar patterns can be seen inchild mortality rates (see Box 3.4 on page 23).Rapid economic growth and declining povertylevels in the region have not reducedinequalities. In some countries, inequalitiesbetween social groups have increased. This isoften because the pace of poverty reduction hasbeen slowest among socially excluded groups.In India, poverty remains higher among Dalitand Adivasi groups despite overall povertydecline (see Box 3.3 on page 22). Nepal hasseen a steep increase in inequality over the pastdecade. The overall decline in poverty between1995 and 2003 varied from 46 per cent for theupper caste Brahman/Chhetri groups, to 10 percent for Janajatis living in the hills and 6 percent for Muslims (see Figure 3.6 on page 21).In Vietnam, poverty among ethnic minoritieshas declined at an average rate of 2.6 per cent ayear over the last decade compared with 3.4 forthe majority Kinh/Chinese community. InChina, one in three people lived in poverty in1990, but with rapid rates of growth, this hasdeclined to less than one in ten.34 However, thishas been accompanied by growing incomeinequalities, reflected in a rise in its Gini indexfrom 35.5 in 1993 to 41.5 in 2005.35

Geographical location reinforces ethnicinequalities. Available data suggest that thelocation of certain social groups is central totheir failure to benefit from economic growthand development (see Box 3.5 on Vietnam andChina). China has managed to substantiallyreduce its maternal mortality ratio between1996 and 2005, from 64 per 100,000 live birthsto 48,36 but it remains much higher in thepoorer, rural western provinces. And within thewestern region, 87 per cent of maternal deathswere among ethnic minority groups;37 this maypartly be explained by the fact that women inthese groups have less access to health services.Rates of decline in child malnutrition between1992 and 2005 have also been much slower inChina’s western provinces, with large disparitiesremaining (see Table 3.9 on page 21).

Focus on MalaysiaMalaysia’s experience stands in contrast to thetrend of growing inequality across the Asia region,as it has succeeded in reducing ethnic inequalities.According to national measures, poverty decreasedby 78 per cent between 1990 and 2007. Accordingto dollar-a-day estimates, poverty incidence iscurrently around 2 per cent. Poverty has alwaysbeen higher among the majority ethnic

community, Bumiputera (‘sons of the soil’), andthis is still the case, but it has actually declinedmore rapidly among this group (see Figure 3.11).As a result, differences in poverty rates betweenthe Bumiputera, the ethnic Indian and the ethnicChinese populations have decreased. There hasbeen a decline in Malaysia’s Gini index from 49 in1984 to 38 in 2004, while the poorest quintile’sshare of national income increased from 4.5 percent in 1992 to 6.4 per cent in 2004.

Malaysia has also achieved progress on theother human development indicators.38 Infantmortality rates declined steadily between 1970and 2006 for all ethnic groups, and there hasbeen a decline in ethnic differentials (seeFigure 3.10). Nevertheless, the Bumiputeracontinue to have higher infant mortality rates.

0

10

5

15

25

35

45

1970

20

30

40

50

BumiputeraIndianChinese

’75 ’80 ’85 ’90 ’95 2000 ’06

Infa

nt m

orta

lity

rate

(pe

r 1,

000

live

births

)

Figure 3.10 Malaysia: infant mortality rate by ethnicgroup, 1970–2006

◆

0

5

1985 1990 1995 2000 2005 2010

10

15

20

25

•◆

BumiputeraIndianChineseAll•

◆

•

• ••

•

•

•

◆

◆

◆

◆

◆◆

▼

✖

▼

▼

▼

▼

▼

▼

▼

✖

✖✖✖

✖

✖

✖

✖

▼

Figure 3.11 Malaysia: poverty Head Count Rate (HCR)trends by ethnicity

Source Malaysia Department of Statistics, www.statistics.gov.my

Source http://chengfansoon.com/ as of 23 May 2010

3.5 Progress on the MDGs in sub-Saharan AfricaSub-Saharan Africa has the highest rates of povertyin the world.39 It is also, after Latin America, theregion with some of the most unequal countriesin the world.40 Regional conflicts, the legacy ofstructural adjustment and the HIV and AIDSpandemic have left large swathes of the continentpoorer than ever, with a population-weightedpoverty gap of 42 per cent for SSA as a whole,including South Africa.41 The proportion ofpeople living in extreme poverty has not declinedmuch from its 1980 level, but this masksprotracted deterioration during the 1980s andthe first half of the 1990s, as well as markedimprovements since the late 1990s. The proportionof people living in poverty fell by around 7percentage points between 1996 and 2004.42

Regional aggregates, however, disguisesignificant progress at country level. As notedearlier, some SSA countries are among the bestperformers in terms of absolute progress on theMDGs. While only one country featured amongthe top ten performers for progress relative toglobal MDG targets, six are among the top ten interms of absolute overall progress, which hastranslated into real changes in the lives ofmillions of people.

The region also has the highest levels of childmortality in the world, with rates highest in West

and Central Africa.43 Across the region, while ratesare not declining fast enough to reach the MDG by2015,44 the rate of reduction from 2000 to 2010 hasimproved compared with 1990 to 2000.45 Theregion also has the highest maternal mortality ratiosin the world, accounting for more than half of allmaternal deaths globally. There are not enough datato monitor progress on this indicator.46

3.6 Social exclusion and the MDGs in sub-Saharan Africa

‘Massive poverty and obscene inequality aresuch terrible scourges of our times – times inwhich the world boasts breathtaking advancesin science, technology, industry and wealthaccumulation – that they have to rankalongside slavery and apartheid as social evils.’(Nelson Mandela, ‘Make Poverty History’speech in Trafalgar Square, London, 2005)

With the notable exception of South Africa, dataon the relationship between ethnic identity andpoverty in Africa are not consistently availableand are rarely up to date. In some countries,region or location is used as a proxy for ethnicity.For this reason, it is important to be cautiouswhen discussing group-based inequalities.However, available data on selected countriessuggest the following:


Namibia: Ella, a washlady from the minority Herero ethnic group. Photographer: Barbara Cheney


Ethnicity is linked to poverty, health, andeducation outcomes. A study conducted in the1990s using data for 11 SSA countriesconfirmed that ethnic inequalities in childmortality persist, and have been increasing (seeBox 3.8 on page 29).47 Other studies haverevealed that ethnicity is linked to differenthealth and education outcomes in a number ofAfrican contexts. In Kenya, less dominantethnic groups fall well behind in terms ofimmunisation levels, and ethnicity dramaticallystratifies under-five mortality rates. Among theMijikenda/Swahili ethnic groups, for example,27 per cent of births have a skilled attendantpresent, compared with 71 per cent ofKikuyu.48 In Nigeria, child mortality rates arelower for the Igbo and Yoruba ethnic groups.This may be the result of higher educationalattainment among Yoruba and Igbo women,but ‘it may also signal privileged access of thesepowerful groups to healthcare, as well asreflecting different cultural practices and beliefstowards child and maternal health acrossethnic groups’.49

Spatial inequalities between developed and lessdeveloped regions within countries are central toand reinforce inequalities between social groups.Data often focus on regional differences, whichin many cases overlap with ethnic and religiousidentities. In Nigeria, the northern,predominantly Hausa–Fulani states have higherlevels of poverty than the southern,predominantly Yoruba and Igbo states (seeTable 3.19 on page 29). Child mortality ratesare also considerably higher in northern areas

(see Table 3.14)50 and maternal mortality ratesare almost ten times higher in the north thanin the south. This largely reflects differentialaccess to healthcare: for example, in the northwest, only 25 per cent of pregnant women use

Overall Urban Rural Northeast Northwest Southeast Southwest Central

0

10

20

30

40

50

60

Overall Urban Rural Northeast Northwest Southeast Southwest Central

■ Stunted ■■ Severely stunted■ Underweight ■ Severely underweight■ Wasted ■ Severely wasted

Perc

enta

ge o

f m

alno

uris

hed

child

ren

Figure 3.12 Prevalence of child malnutrition by residence,Nigeria

Table 3.14 Nigeria: childhood mortality rates by residence

Zone Child mortality

North Central 62Northeast 126Northwest 139Southeast 64South South 58Southwest 32

Estimates are for deaths per 1,000 children age 12–59 months.Source National Population Commission (NPC) [Nigeria] and ICFMacro (2009) Nigeria Demographic and Health Survey 2008,Abuja, Nigeria: National Population Commission and ICF Macro

Table 3.15 Socioeconomic indicators for selected regionsin Ghana (%)

Incidence Literacy: % Access to of poverty literate health (1999) (1998) services (1997)

Greater Accra 5 76 78Ashanti 28 64 43Northern 69 13 18Upper West 84 20 8Upper East 88 20 20National 40 51 37

Source Adapted from Langer et al. (2007)51

Source Omilola, B. (2010) ‘Patterns and Trends of Child andMaternal Nutrition Inequalities in Nigeria’, IFPRI DiscussionPaper 00968

Box 3.6 The intersection of spatial, ethnic andreligious inequalities in Ghana and Nigeria

The importance of indigeneity in the Nigerian Federation,which confers affirmative action in the public sector andcitizenship rights, has deterred internal migration to theextent that less than 5 per cent of people from ethnicgroups concentrated in the north live in the south. Ethnicand religious divides therefore tend to coincide with andreinforce geographical ones. In Nigeria, the population inthe less developed northern regions is predominantlyMuslim and from the Hausa-Fulani ethnic group. Thepopulation in the richer southern regions is largelyChristian, Yoruba and Igbo being the main ethnic groups.

Ghana, unlike Nigeria, has high rates of internalmigration, which has resulted in considerable mixing ofethnic groups: 35 per cent of northern ethnic Ghanaianslive in the south. As a result, data from 2008 on thedistribution of wealth show a marked spatial rather thanethnic pattern to inequality.

Source: Mancini (2009)49

Box 3.7 Focus on South Africa

South Africa represents an extreme case of intersecting inequalities given its ‘infamous history of highinequality with an overbearing racial stamp’.58 This, ironically, is why it has the most comprehensive dataon intersecting inequalities in the region. The post-apartheid government’s determination to monitorprogress in tracking inequalities means that the data are also up to date.

The country is well on track to meet all the MDGs and has already met a number of them.59 At the sametime, deeply entrenched inequalities persist. While poverty has fallen during the post-apartheid period, itremains acute for the African and coloured populations. Income inequalities that were already substantialappear to have increased further, from a ratio of around 66 in 1993 to 70 in 2008.60 The majority Africanpopulation was, and has remained, at the bottom of the income hierarchy. In fact, its position has slightlydeteriorated: Africans earned 13.0 per cent of white income in 2008, compared with 15.9 per cent in 1995.

The impact of gender and race on poverty is illustrated by data for 1993, 2000 and 2008 (Table 3.16).While for each of these years, the incidence and share of poverty was higher for Africans than the rest ofthe population, both incidence and share of poverty was consistently higher for African women than men. A similar pattern prevailed in the coloured population, although differentials were less marked.

Equalising educational opportunities has been a major thrust of the post-apartheid government and,by 2006, more than 98 per cent of children were in primary education – an increase of 1 per cent since2002. However, racial differences persist: for instance, 86 per cent of white 9th graders in 2002 hadreached grade 12 by 2005, while only 29 per cent of Africans and 42 per cent of coloureds had done so.61

Infant and under-five mortality rates have declined between 1998 and 2003, from 45.4 to 42.6 per1,000 live births and from 59.4 to 57.6 respectively.62 But infant mortality risk is four times higher amongblack African children than white children, even after controlling for demographic factors such as themother’s age and the timing and number of births.63 There has also been a decline in maternal mortalityrates from 150 to 124 per 100,000 live births between 1998 and 2002, although this is still consideredhigh for a middle-income country like South Africa.64

Table 3.16 South Africa: individual level poverty by race and gender (poverty line R515 per capita per month)

Population Head count Poverty share1993 2000 2008 1993 2000 2008 1993 2000 2008

African female 0.40 0.41 0.42 0.72 0.66 0.68 0.51 0.50 0.52African male 0.36 0.38 0.38 0.66 0.61 0.60 0.42 0.43 0.41Coloured female 0.04 0.05 0.05 0.32 0.32 0.36 0.02 0.03 0.03Coloured male 0.04 0.04 0.04 0.29 0.30 0.35 0.02 0.02 0.03Indian/Asian female 0.01 0.01 0.01 0.12 0.11 0.11 0.00 0.00 0.00Indian/Asian male 0.01 0.01 0.01 0.12 0.09 0.19 0.00 0.00 0.00White female 0.06 0.05 0.05 0.05 0.06 0.04 0.01 0.01 0.00White male 0.06 0.05 0.04 0.06 0.08 0.03 0.01 0.01 0.00

Source Leibbrandt et al. (2010)58


Table 3.17 Estimates of annual per capita personal incomeby race group in 2000

Year White Coloured Asian African Average

Per capita income in constant 2000 Rands:1917 13,069 2,875 2,894 1,184 3,9461936 19,212 3,000 4,443 1,462 5,3591956 30,494 5,158 6,668 2,627 8,5411975 49,877 9,688 12,687 4,289 12,6961995 48,387 9,668 23,424 6,525 12,5722000 56,179 12,911 23,025 8,926 16,2202008 75,297 16,567 51,457 9,790 17,475


Table 3.18 Relative per capita personal incomes by racegroup (% of white level)

Year White Coloured Asian African Average

Relative per capita personal incomes (% of white level):1917 100 22.0 22.1 9.1 30.21936 100 15.6 23.1 7.6 27.91956 100 16.9 21.9 8.6 28.01975 100 19.4 25.4 8.6 25.51995 100 20.0 48.4 13.5 26.02000 100 23.0 41.0 15.9 28.92008 100 22.0 60.0 13.0 23.2



clinics, compared with 85 per cent in the southeast.51 The interaction between ethnicity,gender and location means that Hausa femalesfrom poor rural households have the lowestlevels of education. In Ghana, the north hasalso fared worse than the south historically.Table 3.15 (on page 27) shows that in 1999 inthe Upper East region, the poverty rate wasaround three times higher than in the Ashantiregion, some 17 times higher than in GreaterAccra, and more than twice the nationalaverage. Literacy, primary school enrolmentand access to electricity and health serviceswere also very low in the northern regions. InKenya, the North Eastern regions, home to theSomali and Swahili (largely pastoralist) groups,have much higher unemployment than otherprovinces. And while almost every child inCentral province attends primary school, onlyone in three does in the North Easternprovince.52 Literacy rates in Ethiopia also varyenormously by region; in rural areas, the

highest self-reported literacy rate is observed inTigray (36 per cent) compared with just 8 percent in the Somali region.53

Inter-ethnic and regional conflicts can befuelled by and exacerbate socioeconomicinequalities. When inequalities in wealth,access to services or political power coincidewith group differences, ethnicity may becomean axis around which disadvantaged groupsmobilise, sometimes violently.54 In Côted’Ivoire, ongoing civil war, which is partlyfuelled by the cultural and politicalmarginalisation of mainly Muslimnortherners, has led to a sharp decrease ineconomic growth and increase in povertylevels. The north–south divide is likely to beincreasing – particularly as state educationand health services are no longer supplied inthe rebel controlled north.55 The issue ofexclusion and conflict is discussed further inSection 4.5.

Box 3.8 Ethnic inequalities in child mortality in Africa

A World Health Organization (WHO) study using surveydata from 11 countries in Africa (Central African Republic,Côte d’Ivoire, Ghana, Kenya, Mali, Namibia, Niger,Rwanda, Senegal, Uganda, and Zambia) revealed thatethnic inequality in child mortality spread in sub-Saharan Africa through the 1980s and 90s. Children ofAshanti women in Ghana were at that time about 20 percent less likely to die than other Ghanaian children. InNiger, where child mortality levels are among the highestin the world, the mortality rate around 1990 for under-fives was 242 deaths per 1,000 in Djerma, compared to353 per 1,000 elsewhere. A range of other socioeconomicinequalities between ethnic groups were also identified.In most countries there are large disparities in schoolingbetween ethnic groups. In Ghana and Uganda, 84 percent of Ashanti and 95 per cent of Baganda women

received some education, compared with 60 per cent ofall others. Location of residence and differential use ofchild health services appeared to be linked to childmortality inequalities; particularly striking is the highlydisparate uptake of immunisation services betweenethnic groups. This suggests that, along with policies toreduce economic disparities among ethnic groups, ‘childsurvival efforts in African countries should pay specialattention to disadvantaged ethnic groups and thelocations in which they are concentrated’.56 We lack datato see whether the efforts of the past ten years havereduced these inequalities; however, the study’sconclusion is echoed in a recent WHO Bulletin, whichpointed out that, in many cases, ‘ethnic, educational andregional variations were more pronounced than thedisparities by wealth level’.57

Source: Brockerhoff and Hewett (2000)47

Table 3.19 Nigeria: states with the highest and lowest shares of poverty, 2006

10 states with highest incidence of poverty 10 states with lowest incidence of poverty

State Geographical location % State Geographical location %

Jigawa North 95.0 Bayelsa South 20.0Kebbi Northwest 89.7 Anambra South 20.1Kogi Central 88.6 Abia South 22.3Bauchi North 86.3 Oyo Southwest 24.1Kwara West 85.2 Imo South 27.4Yobe Northeast 83.3 Rivers South 29.1Zamfara Northwest 80.9 Enugu South 31.1Gombe Northeast 77.0 Ogun Southwest 31.7Sokoto Northwest 76.8 Osun Southwest 32.4Adamawa East 71.1 Edo South 33.1

Source Adapted from Langer et al. (2007)51

The preceding discussion makes two generalpoints about the relationship betweenprogress on the MDGs and social justice.

First, the intersecting and mutually reinforcinginequalities that lead to social exclusion aredeeply entrenched in the historical structures andeveryday practices of societies. This makes themremarkably resistant to change. Second, despitethis apparent intransigence, these inequalities arenot, in fact, immutable. Change is evident inevery region, though more rapidly in some

contexts than others, and more rapidly in relationto some MDGs than others. This should guardagainst any pessimism about future progress.

4.1 The cultural dynamics of exclusion Processes of cultural devaluation are keymechanisms through which social exclusion isperpetuated over time. The effects of exclusioncan work in silent and invisible ways that cannevertheless have a profound impact on thosewho are excluded. Or they may work in ways thatare visible and noisy, with negative spillovereffects within the larger society.

Cultural devaluation is sometimes embeddedin law (see Box 4.1 on Nepal). However, whilelegal protections are important to tackle socialexclusion, they are not sufficient. The IndianConstitution (1949), for example, acknowledgesthe need to counter the historical disadvantageexperienced by Dalit and Adivasi groups.Nevertheless, the everyday practice ofdiscrimination continues to undermine the self-worth and agency of excluded groups.

In Nepal, a recent attempt to measure theextent of empowerment and inclusion by socialgroup, using a variety of indicators, found thatthe upper castes scored twice as high as Dalitswith regard to knowledge about rights andprocedures, confidence in accessing services,exercising rights, social networks and localpolitical influence. While 90 per cent of the uppercaste groups had never faced any restrictions orintimidation (the 10 per cent exceptions werewomen), 100 per cent of Dalit respondents hadexperienced some degree of restriction onentering certain public spaces, and 20 per centreported high levels of harassment, intimidationand restriction. In addition, women from theupper castes scored higher than women fromDalit and Janajati groups. Not surprisingly, uppercastes scored significantly higher (46) on acomposite index made up of these indicators,with Janajati groups scoring 36, and Dalits 25.Combining caste and gender placed Dalit womenat the bottom of the social hierarchy (Figure 4.2).

Caste and ethnic identity accounted for 33 percent of variation in scores, while genderaccounted for 7 per cent. Of particular interest topolicymakers was the finding that ten years ofeducation was associated with a 19 per cent

BrahmanChhetri

High caste NewariTagadhari ‘Twice Born’

Matwali ‘Liquor-drinking’UnenslavableEnslavable

Pani Na Chaine ‘Water-unacceptable’Muslims & foreigners

Acchut ‘Untouchable’

‘Low caste’

Non-caste Janajati/indigenous peoples

‘High caste’

‘Pure’

‘Impure’

Dalit

Figure 4.1 The Nepal caste pyramid according to the Muluki Ain(1854)

The area showing the different groups in the triangle does notrepresent population size. Dark red shows the Hindu caste groups.Source World Bank/DFID (2006) Unequal Citizens: Gender, Casteand Ethnic Exclusion in Nepal

The intersecting dynamics of inequality:why social exclusion persists4

30 The intersecting dynamics of inequality: why social exclusion persists

Box 4.1 The caste system in Nepal

The 1854 Muluki Ain (Country Code) in Nepalformalised the caste system into law, grantingdifferent social groups differential privileges andobligations according to their position in the castehierarchy. The characteristics associated with eachgroup signalled the respect – or lack of – they wereaccorded by their society. The ‘pure’ Brahmans/Chhetris were at the top of the caste pyramid; Janajati(indigenous peoples) were underneath, followed bythe low caste or ‘impure’ Pani Na Chaine, with Dalits(untouchables) at the bottom of the pyramid.

Nepal’s new Constitution (1990) declares that allcitizens are equal ‘irrespective of religion, race,gender, caste or ideology’. But inherited inequalitiescontinue to affect the lives and life chances of thosegroups at the lower end of the hierarchy.


increase in the score and that membership of alocal development association led to a 5 per centincrease. Other positive factors included contactwith the local women’s development office,holding office in an association, and exposure tothe media.

The Latinobarometer carries out regularsurveys in 17 countries in the Latin Americanregion. The 2000 round included a questionabout which groups were most discriminatedagainst in their country. Most answers pointed tothree groups: Afro-descendants, indigenouspeople, and ‘the poor’. Respondents in countrieswith large proportions of Afro-descendants andindigenous people – Bolivia, Brazil, Ecuador,Guatemala and Panama – were most likely toname one of these three groups. In Argentinaand Uruguay, countries with a morehomogenous ethnic composition, the poor wereperceived to be the most discriminated againstgroup. What is telling is that those who believedthat ethnic groups were most discriminatedagainst believed that they faced suchdiscrimination in all spheres of life: at work, inschool, in political parties, and in the justicesystem (police and courts).66

Evidence from some countries highlights thefrequently documented and mutually reinforcinglink between social exclusion and harmfuladdictions. In Sri Lanka, for example, restrictedoptions for progress in life may explain the highlevels of alcoholism among the predominantlyIndian Tamil labourers in the plantation sector,one of the ‘silent’ consequences of exclusion. This

was in turn seen by the labouring community asa major cause of their poverty, indebtedness andsocial stigma. In South Africa, higher levels ofalcoholism and substance abuse in the colouredpopulation (both men and women) suggest thatthe experience of exclusion is not a uniformone.67 The DHS (2003) reported that 9.1 per centof adolescent women in the coloured populationgroup reported harmful levels of drinking in thepast 12 months compared with just 1.1 per centof black African adolescent women. WesternCape is reported to have one of the highest ratesof Foetal Alcohol Spectrum Disorders (FASD) inthe world. This may account for the higher levelsof infant and child mortality among the colouredpopulation noted in the recent DHS.68

Intersecting inequalities are also oftenassociated with higher levels of crime and drug-

related violence (see Box 4.3 on page 32). It hasbeen estimated that in urban Colombia, crimeswere most likely to be committed by people fromhouseholds with a per capita income below 80per cent of the national average. Surveys fromurban South Africa show that patterns of crimevary by class and race. While the wealthierquintiles were more likely to report crimeswithout injury, it was the poorest andpredominantly black population in informalsettlements and townships who suffered thebrunt of violent crime.69

Group violence, riots and civil war are someof the more extreme consequences of socialexclusion. Studies of conflicts in different regionsof the world point to the recurring significance ofgroup-based disadvantage as a factor.70 Social

Processes of cultural devaluation are keymechanisms through which social exclusion isperpetuated over time

Figure 4.2 Nepal: composite empowerment and inclusionindex (CEI) by gender/caste/ethnicity

Source Bennett, L. et al. (2006) ‘Negotiating Social Change:Gender Caste and Ethnic Dimensions of Empowerment andSocial Inclusion in Rural Nepal’, in R. Alsop, M. Bertelsen and J. Holland (eds), Empowerment in Practice: From Analysis toImplementation, Washington D.C.: World Bank

Box 4.2 Why social exclusion persists: internalisinginferiority

A telling experiment from India found that Dalit school childrenperformed as well as children from other caste groups in solvingpuzzles in exchange for payment when caste identities wereconcealed. They performed less well when their caste identitieswere revealed, and even less well when the children playing thegame were segregated into separate caste groups.65 If such afinding can be generalised, it suggests that the ‘internalisation’of ascribed inferiority has a powerful effect on the capacity ofexcluded groups to respond to economic opportunities.

Source: Hoff and Pandey (2004)65

All groups

Janajatis

Dalits

Tarai middlecastes

BCN

0.0 0.1 0.2 0.3Index

0.4 0.5

◆ Male◆ Female


exclusion does not inevitably lead to conflict, butit significantly increases its likelihood. A reportby the Indian government into the long-standingNaxalite insurgency, which affects 125 districtsspread over 12 states, found that the movement’smain support comes from scheduled castes andscheduled tribes. It also found that while thereare many districts with high concentrations ofDalits and Adivasis which did not have a Naxalitepresence, it was generally the case that areas ofNaxalite influence had a higher than averageproportion of Dalits and Adivasis in theirpopulations. The issue of social exclusion andconflict is discussed in greater detail in Section 4.5,‘The political dynamics of exclusion’.

4.2 The economic dynamics of exclusion:asset inequalitiesThe poverty of socially excluded groups isfrequently mediated by cultural norms andpractices. In some cases, excluded groups are notpermitted to own or buy land because of whothey are. This has long been the case for the‘untouchable’ castes in India and Nepal, and eventoday, the vast majority of people belonging tothese castes are landless.

In other cases, ethnicity differentiates theamount and quality of land people own. In Peruand Ecuador, for instance, indigenous groups hadlandholdings that were between two and eighttimes smaller than those of non-indigenousgroups.71 And only 13 per cent of all irrigatedland in Ecuador was in the hands of indigenousfarmers. In Vietnam, ethnic minorities are heavilydependent on farming and most own some land.However, their land is on sloping terrain andyields just one crop a year. The majority Kinhgroup is more likely to own irrigated land andperennial crop land.

Indigenous people are more likely to bedispossessed of their land because theircustomary tenure systems may not be recognisedby law. Non-recognition of customary landarrangements for forest dwellers and uplandpeople has been a major factor in theirimpoverishment.72 Forest departments havetraditionally held police and judicial powers, inaddition to administrative powers, to enforcetight state controls over forest lands. This hasresulted in forest dwellers being treated ascriminals or squatters on their own land; in some

Figure 4.3 Brazil: pay levels by gender, ethnic group and years of schooling, 1992 and 2002

Source Machinea J.L.; Bárcena, A. and León, A. (2005), The Millennium Development Goals: A Latin American andCaribbean Perspective, Santiago de Chile: United Nations

Box 4.3 Social exclusion and violent crime: life in thefavelas of Brazil

Studies carried out in 1969 and 2002 in the same favelas(slums) in Rio de Janeiro revealed that in 1969, what peoplefeared most was forcible relocation by the housing authorities ofthe military dictatorship. By 2002, their biggest fear was dyingin the crossfire between police and drug dealers or gangs.

Sixteen per cent of those interviewed in 1969 said thatcrime and violence were what they disliked most about livingin the city, while in 2002, 60 per cent gave this answer. Whilethe population of the favelas was, and continues to be, raciallymixed, black people are less likely to leave: half of the blackpeople interviewed in 1969 still lived in the favelas, comparedwith just a third of the white people.

Source: Perlman, J.E. (2004) ‘Marginality: From Myth to Realityin the favelas of Rio de Janeiro, 1969–2002’, in A. Roy and N.AlSayyad (eds), Urban Informality: Transnational Perspectivesfrom the Middle East, Latin America, and South Asia, Lanham,MD: Lexington Books

500

0

1,000

1,500

2,000

2,500

3,000

3,500

0 to 3 4 to 7 8 to 10 11 to 14 15 to more 0 to 3 4 to 7 8 to 10 11 to 14 15 to more

White women Black women White men Black men


countries (Thailand, for example), forest dwellersare not recognised as citizens.73

In some countries, indigenous people havebeen dispossessed or displaced by large-scalemining ventures, the expansion of the agriculturalfrontier, and other infrastructure projects fromwhich they cannot expect to benefit.74

The greater poverty of excluded groups maynot be simply a matter of fewer assets but also oflower returns on their assets. One study fromIndia75 decomposed the differences in the povertyrates of scheduled castes, scheduled tribes andothers, and found that half of the differencecould be attributed to group characteristics suchas education, occupation, demography andlocation. The rest could be attributed to the effectthat these characteristics had on the probabilityof being poor. Another study emphasised theimportance of location as well as education indifferentiating returns by ethnic identity, while itwas largely differences in returns in relation toeducation that mattered for caste.76

4.3 The economic dynamics of exclusion:disadvantageous livelihoodsThe nature of livelihoods may also be a basis forsocial exclusion and a barrier to the reduction ofextreme poverty. In many parts of the world,indigenous groups pursue a nomadic or semi-nomadic way of life that is regarded as ‘inferior’ bythe rest of society. In Thailand, for instance, the

Box 4.4 The Kamaiya system in Nepal

The Kamaiya system is a bonded agricultural labour systempractised in the western lowlands of Nepal. It is almost entirelyconcentrated among the Tharus, a low caste ethnic group. AKamaiya binds himself and his family to cultivate a landowner’sland in return for an annual payment. While in principle thissystem is based on a voluntary contract, in practice, theKamaiya often becomes deeply dependent on the landowner.He lives in a hut on the land, which means the landowner isalso landlord. As the debts are inherited from father to son,some Kamaiya families are tied to the same landlord forgenerations. What differentiates this from other bonded laboursystems, though, is that if the Kamaiya is unable to repay hisloans at the end of the contract period, the landlord canexchange his Kamaiyas with another landlord who is willing torepay the Kamaiya’s debt. This sales transaction makes thissystem similar to slavery. Surveys have estimated there areabout 15,000 Kamaiya households, with 46 per cent ofKamaiyas being homeless. Besides adults, there are roughly13,000 children working under this system. The Government ofNepal abolished the Kamaiya system in 2000, but thousands offormer bonded labourers are still living from hand to mouthwith no means to support themselves or their families.

See ‘Bonded Labour In Nepal’, The South Asian, 18 April 2005,accessed 4 August 2010 at www.thesouthasian.org/blog/archives/2005/bonded_labour_in_nepal.html

Sudan, near Gereida, South Darfur: a nomadic family travellingby camel. Eleven months ago the village of Kulaykili wasattacked and destroyed by Janjaweed militias. The blackpopulation, mainly farmers, have fled and nomads of Arab originhave moved in. Photographer: Sven Torfinn/Panos Pictures


reliance of the Hmong, Akha, Lahu and Lisu onswidden agriculture in upland forested areas isperceived to signify their ‘uncivilised’ way of life, incontrast to the settled wet rice cultivation practisedby lowland villagers. In East Africa and the Horn ofAfrica, the pastoralist livelihoods of some ethnicgroups have long been the basis of their exclusionfrom mainstream development efforts.

In other cases, social exclusion is associatedwith the cultural assignment of excluded groupsto the worst-paid and most demeaning jobs. TheHindu caste system in India and Nepal assignsthe lowest castes to the removal of night soil,sweeping, garbage collection and other jobs that

are considered to be menial, degrading and dirty.More generally, a combination of discrimination

and lower levels of educational attainment keepsocially excluded groups confined to lower-paidwork and more exploitative working conditions.In Vietnam, for instance, migrants from ethnicminorities earn half as much as those from theKinh majority, are far less likely to have a workcontract, and are far less likely to receive help infinding a job.77 In Peru, white workers were morelikely to be found in the higher hourly incomequintiles than indigenous workers, and weremore likely to be professionals, technicians andexecutive staff. In South Africa and Brazil, Afro-descendant and indigenous populations reportedhigher levels of informal employment. Moreover,in every employment category, hourly earningswere highest for white workers and lowest forblack workers.78

The interaction between gender and ethnicinequalities generally places women from ethnicminorities at the bottom of the income hierarchy:for example, indigenous and Afro-descendantwomen in Bolivia, Brazil, Guatemala and Peruwere more likely to earn ‘poverty wages’ ($1 anhour, purchasing parity power (PPP) adjusted)than either men from their ethnic group or men

Table 4.1 Time to reach nearest health facility in north andsouth Ghana (% of respondents)

Time to reach nearest health National Average Average facility average South North

30 minutes or less:Rural 52.9 45.5 15.1Urban 52.9 79.3 51.4

Source Core Welfare Indicators Questionnaire (CWIQ) Survey,1997 Main Report, Ghana Statistical Service, Accra Ghana, citedin Langer et al. (2007)51

Indigenous women and children from Colombia gather to discuss issues of food security, nutrition and health in theircommunities. Photographer: David Parra


and women from the rest of the population. InBrazil, Afro-descendant women earned the least,while white men earned the most for each level ofeducation (Figure 4.3 on page 32).

Lack of access to financial services, or accesson extremely usurious terms, has been a majorconstraint for poor and excluded groupseverywhere. In India, Pakistan and Nepal, theprevalence of bonded labour is a stark indicatorof the unfavourable terms on which such groupsobtain credit (see Box 4.4 on page 33, on Nepal’sKamaiya system). They are disproportionatelydrawn from lower caste and ethnic minoritygroups. Such forms of indebtedness tend to serveas a mechanism for the inter-generationaltransmission of poverty, since the children ofbonded labourers often become bondedlabourers themselves.79

In Latin America, only a small fraction ofindigenous households were found to have accessto formal (or informal) credit; non-indigenoushouseholds had two to three times more access tocredit.80 Evidence from Peru suggests thateducation increases the likelihood of indigenouspeople obtaining credit, mainly through betteraccess to information. But most indigenous peoplelack education. In rural Ecuador, more indigenousbusiness owners are deterred from seeking a loandue to high interest rates than non-indigenousbusiness owners – 36 per cent and 23 per centrespectively. Further, interest rates for formalcredit were higher for indigenous businesses andfarms.81 In Vietnam, access to formal credit does

not vary much by ethnic group. But ethnicminorities tend to utilise smaller loan facilities(tied to specific activities) provided by theVietnam Bank for Social Policies, while themajority group are more likely to take advantageof the larger loans offered by the Vietnam Bankfor Agriculture and Rural Development.

4.4 The dynamics of exclusion in serviceprovision: access and qualityThe poor, more than any other group, rely onbasic public services to meet their needs for healthand education. The failure of such services toaddress their needs is a major factor in explainingthe uneven pace of progress on relevant MDGs.Unequal spatial distribution of services, and thecosts, quality and relevance of the services onoffer, are some of the aspects of this failure.

In India, the most important source ofvariation in the per capita state provision ofdoctors, nurses and teachers in rural districts isreligion and caste; the higher the percentage ofDalits and Muslims in the district population, thelower the provision of medical and educationalservices.82 Nigeria reports large disparities inantenatal care between the north and the south.83

Only 4 per cent of women in the north eastreceived care from a doctor compared with 52per cent of women in the south west, while only8.4 per cent of mothers in the north westdelivered in a health facility compared with 73.9in the south east. Women from the north citedlack of money, distance to facilities and concern

Country average

Pastoralist group

0 20 40 60 80 100

57

26

17 85

929

51

853960 88

53 96

94

68

91

80

7143

17

7

33

0 20 40 60 80 100

40

35

17 78

629

51

8361

60 90

42 8787

79

84

535041

30

9

38

Extreme education poverty: population aged17 to 22 with fewer than 2 years of education (%)

% out of primary school

Ethiopia:

Afar/related

m

f

m

f

m

f

m

f

m

f

Kenya:

Somalia

Uganda: Karamojong

Benin:

Peul/related

Nigeria:

Fulani

Senegal:

Poular

Figure 4.4 Pastoralists facing extreme education deprivationPercentage of population aged 17–22 with fewer than two years of education and percentage of childrenof primary school age not attending primary school, by gender and membership of selected pastoralistgroups, latest available year

Source UNESCO (2010)107


that there were no drugs available as the mainreasons they did not deliver in a health facility.84

In Ghana, too, there are differences between thenorth and the south in the utilisation ofhealthcare facilities, albeit less pronounced; onein four children in the Northern region aredelivered at a health facility compared with fourout of five in the Greater Accra region.85

In Vietnam, compared to a national figure of17 per cent in 2002, 33 per cent of women in thenorth east, 65 per cent in the north west and 40per cent in the central highlands gave birth withno assistance from qualified health workers.86

Reluctance to seek help from male health workersand the difficulties of recruiting women to workin remote mountainous areas, together withfinancial constraints and the distances involved,were largely responsible for these outcomes.

Financial constraints are also a majorconsideration for poor people. The imposition ofuser fees for social services by large numbers ofAfrican countries in the context of structuraladjustment programmes in the 1980s and 1990sled to a dramatic decline in the use of educationand health services, while in the case of health, thefrequency of catastrophic expenditure increased.87

Finally, the quality of service provisionmatters. A study of skilled attendance forchildbirth in rural Tanzania found that ethnicity,education and household assets were importantpredictors of service uptake. Social positioning,past experience, entitlement, shame, and self-identity reinforced some women’s preference todeliver at home. Some had experiencedsubstandard treatment, or been turned awayfrom health facilities and felt humiliated.

‘Health facilities were public spaces and whatoccurred there became a matter of publicknowledge; social standing was diminishedwhen poverty was made evident. Otherfeatures that could put women at risk formistreatment were ethnicity or family. Astigma on Pogoro and other groups (typicallyenacted by persons from more urban areassuch as facility providers) increasedsusceptibility to substandard treatment.’(Spangler et al. 2010: 7)88

In Latin America, belonging to an indigenousgroup or being monolingual in an indigenouslanguage constitutes a barrier to access tohealthcare in at least five of the region’s countries(Guatemala, Bolivia, Ecuador, Paraguay andPeru). In Colombia, racial and ethnic disparitiesin health status and access to healthcare werelargely explained by differences in socioeconomiccharacteristics, employment status, type of joband geographical location. In Brazil, utilisation ofmaternity-related services was found to be relatedto education and household resources, as well aslocation. Households in rural areas and in thepoorer north and north east were less likely tomake use of such services. Social exclusion thusworked through the interaction between race,poverty and location.

In South Asia, religion and caste affect theuptake of maternal health and delivery services,along with household wealth status and women’seducation.89 Acts of discrimination against Dalitsare reported in the public health services, alongwith prejudice towards religious minorities. Thisincludes avoidance by health workers,

10

0

20

30

40

50

60

70

1959

Bahun-Chhetri Dalit Janajati Newar Madhesi Women

1967 1978 1981 1986 1991 1994 1999

◆

◆

◆◆

◆

◆◆

◆ ◆

••••••••

✖

✱

▼■

▼ ▼▼

✱

✱ ✱ ✱ ✱ ✱

✱ ✖

✖

✖

✖

✖

✖✖ ✖

•

Perc

enta

ge

■ ■ ■ ■ ■ ■ ■ ■

▼ ▼

▼▼

▼

✱

Figure 4.5 Nepal: ethnic/caste and gender representation in Parliament, 1959–99

Source Lawoti, M. (2002) ‘Defining Minorities in Nepal’, Nepali Journal of Contemporary Studies 2.1: 20–52


particularly paramedics and nursing staff, ofphysical contact with Dalits and reluctance tovisit Dalit households. In turn, both real andanticipated discriminatory behaviour on the partof health workers deters Dalits from using healthproviders, particularly for services that involvephysical contact, such as giving birth.90

Teachers in India are also predominantlyupper caste and bring their caste prejudices intothe classroom. Dalit children are expected to runerrands and are assigned menial tasks such assweeping and cleaning the classrooms. Higherrates of teacher absenteeism were reported inareas where children were mainly from Dalit andtribal communities. In West Bengal, for example,teacher absenteeism was 75 per cent in suchschools compared with 33 per cent elsewhere.Such treatment has particularly negative effectsbecause Dalit children are likely to be first-generation learners.

In sub-Saharan Africa, pastoralist livelihoods areclosely associated with specific ethnic groups. Theseare among the most excluded from educationservices. In many cases, national education systemshave failed to offer relevant curricula, provideappropriate textbooks and respond to therealities of pastoralist livelihoods, which involve

children travelling for long periods to tend cattle.National educational curricula can reinforce

exclusion in other ways. Classical educationalmodels assume the existence of a monolingualand mono-cultural student based on the idea ofhomogeneity of beliefs, culture and language. InMexico, indigenous children fared worse withinthe educational system, particularly those in

Box 4.5 Intersecting ethnic, religious and politicalinequalities in Sudan

In Sudan, intersecting inequalities underpin poverty and poorlife chances. Around 50 per cent of the population in thenorthern region live in poverty, compared with more than 90per cent of those in the south; and 62 per cent of children inthe north are enrolled in primary education compared with 20per cent in the south. Maternal mortality is more than threetimes higher in the south than the north. This has led to thelongest civil war in Africa. The death toll from the conflict,which began in the early years of independence, is estimatedto be 5 million. A peace agreement signed in 2005 includedprovisions to address the south’s political exclusion and thesocioeconomic inequalities between north and south, butfundamental grievances have not been addressed.

Source: Swearingen, M. (2010) ‘Group Inequality and Conflict:Some Insights for Peacebuilding’, Peace Brief 28, WashingtonD.C.: United States Institute of Peace

Children in the Saloum Delta in Senegal hold up their catch. Communities in this area depend largely onthe rich marine and estuary life for their livelihoods. Photographer: Sebastian


disadvantaged rural areas.91 However, bilingualindigenous children (who could speak Spanish aswell as their own language) fared almost as wellas non-indigenous children, and considerablybetter than indigenous children who could notspeak Spanish. Thus, language barriers operatewithin the educational system in ways that maynot be easy to observe.

4.5 The political dynamics of exclusion When group-based inequalities are reproduced inthe exercise of political power and access to publicinstitutions, they undermine the confidence thatdifferent sectors of the population have in thegovernment’s ability to rule fairly. Excluded groupsare often minorities, and there is often littleincentive for political parties to take their interestsinto account.92 When the economic prospects ofsuch groups are undermined by unevendevelopment, when differential access to essentialservices persists, and when political opportunitiesfor voice and influence are denied, grievancesemerge and often spill over into conflict (see Box4.5 on page 37, on Sudan). Ethnic or groupconflicts are almost universally associated withpolitical inequalities, and frequently related toeducational and income inequality.93

In Mexico, the Zapatista uprising can be tracedto a long history of exploitation, exclusion andinjustices towards the indigenous population, andis rooted in the intersecting inequalitiesexperienced by indigenous people, based on ethnicidentity, location and poverty. The proportion ofthe indigenous population in the state of Chiapas,where the movement began, was over three timesthat of Mexico as a whole: the proportion ofpeople on incomes below the minimum wage wasnearly three times that of Mexico as a whole, whilethe proportion of people on high incomes was lessthan half the national rate.94 Illiteracy in Chiapas istwice the national rate, and indigenous people hadsubstantially lower school attendance and incomesthan the rest of the state’s population. They hadbeen excluded from land reform efforts andconsigned to poor and ecologically vulnerableland. For years, organised resistance in the statewas met by political repression. The adoption ofthe North American Free Trade Agreement(NAFTA) after a long period of worseningconditions as a result of neo-liberal policiesprovided a major impetus to the uprising.

In Nepal, Brahmans and Chhetris maintained

around 60 per cent presence in the legislatureright through the period from 1959 to 1999, thelast decade of which included the first ten yearsof multiparty democracy in the country (seeFigure 4.5 on page 36). Dalits were almostentirely absent right through this period, whilewomen were hardly represented, regardless ofcaste. Civil service positions were alsodisproportionately drawn from Brahman/Chhetri groups (83 per cent) and the uppercastes held virtually all positions in thejudiciary.95 Poverty, caste and ethnic identitycombined with persistent political exclusion,even after the introduction of democracy, togive rise to several years of civil conflict by theend of the 1990s.

In Uganda, the main economic, cultural andpolitical divisions are between the largely Bantu-speaking population in the centre–south and thenon-Bantu speaking population in the north. Thecentre–south has consistently performed betterthan the more peripheral areas of the country,especially the north, and received the lion’s shareof public resources. The north is markedlypoorer than the centre–south; average incomesare broadly twice as low and social servicessubstantially worse. However, northern ethnicgroups are in the minority, and can only attainpower through alliances, vote-rigging or violence.This explains why the first 25 years of thecountry’s 40 years of violent conflict was mainlyinstigated by northern politicians in poweragainst southern privileged elites.96

All post-conflict societies suffer fromdiminished resources. Eight of the ten countrieswith the worst human development index (HDI),and eight out of ten countries with the lowestgross national product (GNP) per capita, havehad major civil wars in the recent past, withcausality working in both directions.97 These arelikely to feature prominently among countriesthat are most off track in relation to the MDGs,although they are also characterised by poor datafor monitoring progress.

When group-based inequalities are reproducedin the exercise of political power and access topublic institutions, they undermine theconfidence that different sectors of thepopulation have in the government’s ability torule fairly


‘We are committed to making the right todevelopment a reality for everyone and tofreeing the entire human race from want…’(Millennium Declaration 2000, para. 11)

The Millennium Declaration held out thepromise of social justice, in a world where theright to development would be a reality foreveryone. But this promise evaporated as thevalues and principles of the Declaration weretranslated into the eight MDGs, and thecorresponding targets and indicators. The failureto retain an explicit commitment to equality,tolerance and solidarity in the formulation of theMDG agenda has led to an uneven pace ofprogress on achievements, with persistinginequalities between different social groups.Unless the MDGs are adapted to the realities ofintersecting inequalities and social exclusionwithin the different regions, they may not onlyfail to provide a pathway to a more just society,but may even exacerbate existing inequalities.Using national averages to measure progressencourages going for the ‘low hanging fruit’ –that is, helping those who find it easiest tograduate out of poverty.98

The rest of this section deals with policies andinterventions that have been found to work indifferent contexts as the basis for crafting moreequitable approaches to the MDGs. The first setof recommendations deal with the general policyenvironment. The first of these deals with thepolitical dimension of change, the missingelement in the MDGs. It highlights the need towork towards new social contracts between statesand their citizens in order to lay the foundationsfor more democratic and inclusive societies. Atthe heart of such contracts are responsive andaccountable states, on the one hand, andresponsible and active citizens, on the other. Suchcontracts do not come into existence by fiat. Theyhave to be constructed over time through mutualinteractions between state and citizens. Therecommendations contained in the rest of thissection can be seen as part of this interactiveprocess. They will both help to strengthenaccountability relationships between states andcitizens as well as to pave the way for furtherpolicies to build a socially just society.

Other more general recommendations dealwith the need for a stronger knowledge base and

public information about social exclusion in orderto develop well-grounded policy responses as wellas the importance of enabling macroeconomicframeworks to pursue such policies. These generalrecommendations are followed by more specificones focusing on the poverty-related MDGs andthen the social MDGs.

5.1 Responsive states, active citizens: towardsa new social contract

‘We resolve therefore… to work collectively formore inclusive political processes, allowinggenuine participation by all citizens in all ourcountries.’(Millennium Declaration 2000, para. 25)

While the MDGs focus attention on what statesmust do (deliver education and health services,promote livelihoods), the political commitmentsof the Millennium Declaration provided a clearmessage about how they should do it: throughinclusive political processes and genuineparticipation.99 However, this was yet anothermessage that was lost in the Declaration’stranslation into key development goals.

The focus on politics is rooted in recognition ofthe power dynamics that reproduce social exclusionover time. Socially excluded groups have had littlevoice or influence in shaping the policies that affecttheir lives and life chances, and hence remain on themargins of development processes. If the MDGs are

The MDGs as a pathway to social justice:equalising life chances 5

Nepal, Karnali Zone: school children learning to write.Photographer: Peter Barker/Panos Pictures

40 The MDGs as a pathway to social justice: equalising life chances

to deliver on the promise of inclusive politics,systematic attention has to be given to expandingthe democratic space for interaction between thestate and its most excluded citizens.

The key role of the stateExperience has made it clear that the state has acritical role to play in promoting the MDGs as apathway to social justice. The private sectorundoubtedly has an important contribution tomake because of its central role as an engine ofgrowth. But driven as it is by profitconsiderations, it is unlikely to take a lead inpromoting social justice. Civil society is alsoindispensable because of its ability to mobiliseagainst injustice and hold the state and privatesector accountable. But it, too, represents specificinterests and cannot speak for society as a whole.However flawed the state might be, it is the onlyinstitution that has a mandate to respond to claimsfor social justice by all its citizens.

Many contemporary examples of progress indriving social transformation and reversinghistorical inequalities have been led by the statealthough in most cases, citizen mobilisation hashelped to drive the agenda. Examples include

South Africa, China and Malaysia. In LatinAmerica, the reversal of long-standinginequalities has been largely effected by leftistgovernments, some of them led by people fromethnic minority communities who havethemselves benefited from the expansion ofeducational opportunities in recent decades.100

Decentralisation and democratisation of localgovernment

‘To obtain the benefits of active citizenengagement across the whole MDGframework, both for people themselves andthrough more effective policymaking, thereare some key accountability building blocksthat must be in place. These includesubstantial freedom of information,transparent budgets, participatorypolicymaking and the political “space” toallow civil society organisations to operatewithout draconian restrictions. Thesemeasures can ensure that women and men inpoverty – whose views are too oftenoverlooked – are able to direct the anti-poverty plans in poor countries.’101

‘We resolve therefore… to work collectively for more inclusive political processes, allowing genuineparticipation by all citizens in all our countries.’ (Millennium Declaration 2000, para. 25) Photographer: Barbara Cheney


Some of the building blocks necessary to buildmore responsive states have been touched uponin the preceding discussion; they include freedomof the press, the right to information, andtransparent or participatory budgeting anddecentralised decision-making. There has been anincreasing emphasis on decentralisation, bringingdecision-making closer to disadvantaged groups, asa key route to greater state responsiveness. As thepart of government closest to the people, thelocal state certainly has the potential to promotemore effective state–society relations. But thereare no a priori reasons why more localised formsof governance should be more democratic orinclusive than other levels.102 Competing interestsclustered around local resources and power basescan act as much to exclude as to include, andpower exercised at the local level can be moreconcentrated, elitist and sometimes moreruthlessly applied than at the centre.103

Decentralisation is therefore a measure thatshould be recommended with caution and notput forward as a panacea. Decentralisation worksbest in the context of a strong, central state with theconfidence to devolve resources as well asresponsibilities, and when the local state encountersa vibrant and well-organised civil society that canengage effectively with policymakers. It must,therefore, be accompanied by efforts todemocratise local government, while recognisingthat on technical and political grounds, certainproblems must still be addressed at higher levels.It must also be accompanied by efforts to build thecapacity of all sections of local society to engagewith decision-making processes at different levels ofgovernment and to hold governments accountable.

Increasing the capacity and political agency ofexcluded groupsThe active engagement of socially excluded groupscan be initiated by state action. In 1996, thegovernment of Kerala launched the People’sCampaign for Decentralised Planning. It aimedto devolve significant resources and authority tothe panchayats (village councils) andmunicipalities, and also mandated villageassemblies and citizen committees to plan andbudget local development expenditures. Nearlyone in four households attended villageassemblies in the first two years of the campaign,and the assemblies continue to draw largenumbers. Many citizens have received training inplanning and budgeting.

In addition, the redesign of institutionalincentives and new mobilisation efforts meantthat women accounted for 40 per cent of theparticipants in village assemblies – much higher

than elsewhere in India – while the participationof Dalits has exceeded their representation in thepopulation.104 A large survey of key respondentsfound that ‘disadvantaged groups’ were the mainbeneficiaries of targeted schemes. It also led tothe widespread view that elected representativeshad become more responsive to the needs oflocal people.

Capacity for political participation can also bebuilt by engaging communities in the design andmanagement of community development projects,building in mechanisms to reach excluded groups.The Kecamatan Development Program inIndonesia is an example of a successful attemptto increase the political agency of marginalisedgroups through community-driven development.It allocates grant money at sub-district level, forwhich several groups of villagers (two of whommust be women) compete for funds on the basisof presentation of a formal sub-project proposal.Procedures, institutions and norms are largelydecentralised, focusing on joint problem solving,inviting broad public participation and scrutiny,and occurring in a more or less continuous andinstitutionalised way. A recent study found thatthe programme helps marginalised groups tocultivate access to more constructive spaces andprocedures for addressing project and non-project conflicts. Results include not only basicfacilities but also a new style of grouprepresentation.

Finally, the capacity for collective action can bebuilt from below. Mobilisation by marginalisedgroups has been promoted though a wide varietyof channels, including social movements, NGOs,faith-based organisations, women’s groups andtrade unions, and self-organisation by the groupsin question. Various examples of these efforts –and their relevance to the MDGs – have beentouched on at different stages of this report.While not all of these organisations are equallyinclusive or effective, many have acted as a

Box 5.1 Making everyone count: collecting census dataon excluded groups in Latin America

The Inter-American Development Bank (IADB) is supportingregional initiatives to incorporate race and ethnicity intonational census reports, civil registries, and household surveysthrough its Everyone Counts programme. In 2001, the IADB,with the Economic Commission for Latin America and theCaribbean (ECLAC) and the World Bank, launched the Programfor the Improvement of Surveys and the Measurement of LivingConditions in Latin America and the Caribbean to improvesurvey and measurement tools to assess living conditions. As aresult of these initiatives, most countries in the region nowcollect data on the indigenous population, and 9 out of 19 LatinAmerican countries collect census data on Afro-descendants.


powerful force for change on a range ofinequalities. In India, for example, civil societygroups, in collaboration with progressivepolitical parties, succeeded in getting the Indiangovernment to recognise the right toinformation, the right to food and the right towork. These rights have then been used by civilsociety to improve the implementation of publicpolicy and to hold service providers accountable(see Box 5.2).

5.2 Strengthening information policies totackle exclusion

‘Strategies [to combat social exclusion] aredependent on the attitudes society as a wholeadopts towards those it excludes. Very often,the first response… is denial and concealmentof its existence… One of the first conditionsfor the formulation of strategies to addressexclusion is… to bring it to the surface, makeit visible and give it recognition.’105

A first step in efforts to tackle social exclusion is toacknowledge that the problem exists; the secondstep is to act on the information. Unlessgovernments know who is being left out of

progress on the MDGs, and why, they cannotaddress the problem. This report has highlightedthe difficulties of obtaining accurate andconsistent information on excluded groups. Thisabsence of data often arises from governmentsfailing to recognise the significance of socialexclusion. In other cases, it may result from adeliberate policy not to acknowledge inequality,as in the case of Myanmar.

There may be other reasons too. The post-genocide government in Rwanda, for instance,has pursued a policy of promoting national unityand reconciliation by highlighting factors thatunify Rwandans, and controlling factors that maybe ‘divisive’, which includes banning any referenceto ethnicity. This makes it extremely difficult totrack progress on the MDGs in relation to theHutu, Tutsi and Twa populations, or to monitorthe differential impact of policies on thesegroups. In other contexts, data are specificallydisaggregated by ethnicity or race, as in SouthAfrica, where the post-apartheid government hasretained reference to ‘historically disadvantagedgroups’ specifically in order to track progress onsocial, economic and political indicators.

International agencies and governments needto improve methods to collect and disaggregatedata so that they can monitor whether excludedgroups are benefiting from progress in relation tothe MDGs. The level of disaggregation is key.Equity-adjusted indicators can play an importantrole in tracking progress. National averages giveequal weight to all quintiles in measuringprogress. Adjusting national statistics fordisparities can be done by weighting quintile-specific values to accord greater importance toprogress in the lower quintiles.106 This can bedone not only for income but also for education,health and the other MDGs. The UNESCOreport, Reaching the Marginalized, which isconcerned with the education levels of sociallyexcluded groups, monitors progress based onequity-adjusted educational indicators.107

Disaggregated data must then be analysed toestablish the extent and causes of inequality andthe impact of policy on excluded groups. Datashould also be disseminated to the relevantconstituencies, including civil society groups, toenable them to fulfil their lobbying role. Forinstance, a recent paper in The Lancet has calledfor more detailed analysis to understand why anumber of countries, such as Egypt and Rwanda,have had considerable success in reducingmaternal mortality – one of the MDGs that isproving more difficult to achieve.108 Without suchanalysis, governments will not find it easy tomake progress on this MDG, and civil society

Bamako, Mali: a meeting at the Centre for Treatment, Activitiesand Counselling for People living with HIV/AIDS (CESAC), whichis the main organisation working with AIDS patients in Mali.Photographer: William Daniels/Panos Pictures


groups will not be able to lobby for the necessarychanges in policy.

A broader set of information policies will beneeded to change public attitudes and perceptions.Much of the discrimination associated withsocial exclusion revolves around, and isreproduced by, the general public’s prejudicesand preconceptions about socially excludedgroups. Research and statistical documentation,along with the education system, the media andpublic campaigns, can help to influence publicdiscourse.

The media should be mobilised to educate,inform and entertain in ways which break downsome of the barriers that separate sociallyexcluded groups from the rest of society. A freepress is one of the cornerstones of democracy,and the media play a critical role indisseminating information, raising awareness andadvocating for policies that can bring aboutchange. The power of cable and satellitetelevision has begun to transform genderrelations in the South Asia context, challengingthe acceptability of domestic violence,questioning the devaluation of women andincreasing their capacity for agency.109

Information policies for excluded groups are asimportant as information policies about them.Research into why certain groups systematicallyfail to utilise social services or claim theirentitlements stresses a lack of awareness amongthose groups about availability and terms ofaccess. The importance of the right to informationis being increasingly acknowledged in legislation incountries like India and Bangladesh, and in draftlegislation form in Ghana. Knowledge aboutrights, entitlements and the rules that governservice provision can build the awareness ofexcluded groups and strengthen their ability toclaim such rights and services. It can also resultin desirable policy outcomes: for instance, theThai government’s mass media campaign, urgingrespect for women and safer sexual practices iscredited with reducing the incidence of AIDS to apoint where the country can now consider afiscally viable treatment programme.110

5.3 Macroeconomic policies and redistributivegrowth A country’s national macroeconomic frameworkprovides the broader context in which policies toachieve the MDGs are pursued. It influences thepace and pattern of economic growth as well asthe distribution of resources and opportunitiesbetween different productive sectors and sectionsof the population. Economic growth can providean enabling environment for promoting the

MDGs as a pathway to social justice, but only if itgenerates an adequate level and equitabledistribution of resources and opportunities.

However, the neo-liberal macroeconomicframework that has shaped the pace and pattern ofgrowth in most countries in recent years has donelittle to challenge the intransigent inequalities thatgive rise to social exclusion. Greater attention toredistribution is necessary. Redistribution can occurthrough growth. Broad-based, employment-centredpatterns of growth that bring about majorstructural transformations of the economyprovide favourable conditions for addressing socialexclusion. Investments in agricultural productivitywill help large sections of the marginalised poorwho continue to rely on agriculture for theirlivelihoods. But they have to be accompanied bythe expansion of employment opportunities

Box 5.2 The Right to Information in India

In 1996 the National Campaign for People’s Right toInformation (NCPRI) was launched in India, spurred by anetwork of grass-roots organisations campaigning around localdevelopment issues. Largely thanks to their efforts, a far-reaching Right to Information (RTI) Act was implemented in2005. This Act requires all central, state and local governmentinstitutions to meet public demands for information and hasbecome an important tool for transparency and accountabilityin the day-to-day functioning of government.111

The right to information and collective action has been used toensure the proper implementation of government programmesaimed at assisting the most vulnerable and disenfranchisedpeople, including the Public Distribution System (PDS) in Delhi.The PDS provides essential commodities – basic food grains,sugar and kerosene – at affordable prices to poor, marginalisedurban dwellers. Officially, PDS cardholders are entitled to claimthese at PDS outlets each month – in practice, however, many areturned away from outlets, told that the food grain had run out orsimply refused their quota. A complaint from a woman in a low-income settlement that she had not received her entitled foodgrains for three months prompted a local campaigningorganisation, Parivartan, to use the Right to Information toinvestigate the PDS operation. Scrutiny of the local PDS recordsrevealed that 87 per cent of PDS wheat and 94 per cent of rice hadbeen sold on the open market. The complainant herself had beenlisted by the shop owner as receiving her quota; a thumb print byher name in lieu of a signature.

Parivartan mobilised the media and rallied local residents touse their RTI to check PDS records for all the shops in the area.Over 300 people publicly filed applications to view theirrecords, despite threats of violence and bribery. The result wasa marked change in shopkeepers’ attitudes. Following the 2005RTI Act, Parivartan has been pushing for institutionalisation ofpublic scrutiny of PDS records. This localised small-scalecampaign shows the potential the RTI has for opening updialogue and negotiation, giving voice to the mostmarginalised, and enabling bottom-up accountability.

See Pande, S. (2008) ‘The Right to Information and societalaccountability: the case of the Delhi PDS Campaign’, IDS Bulletin 38


outside agriculture and the pathways that willenable these groups to take advantage of them.

‘Sustained and equitable growth based ondynamic structural economic change isnecessary for making substantial progress inreducing poverty. It also enables fasterprogress towards the other MillenniumDevelopment Goals. While economic growthis necessary, it is not sufficient for progress inreducing poverty.’112

However, even broad-based growth is unlikelyto be sufficient on its own to break historicallyestablished cycles of inequality. Pro-poor growthhas to be combined with redistributive policies ifthe MDGs are to benefit excluded groups. TheLatin American experience provides one exampleof a policy approach that combines ‘growth withprudent redistribution’.113 Brazil has achieved asignificant reduction in historically establishedpatterns of inequality, despite moderate levels ofgrowth, because of the state playing a proactiverole in redistributive policies.114

‘Latin American countries which havemanaged to reduce inequality have continuedto emphasize orthodox objectives:macroeconomic stability, fiscal prudence andthe preservation of free trade and financialflows. Where they have made a break withpast orthodoxy is in their reliance onmanaged exchange rates, greater mobilisationof domestic resources, neutral/countercyclicalfiscal policies and an active role for the statein the field of labour and social policies.’115

5.4 Progressive fiscal policiesGovernment budgets are a particularly powerfulmacroeconomic policy instrument for ensuringthat redistributive growth leads to greater, moreequitable progress on the MDGs. Taxation hasparticular relevance for addressing economicinequalities, while public expenditure can be usedto redress social inequalities.116 Both the levelsand composition of these instruments willdetermine their effectiveness. Developingcountries generally do not devote enough effortto mobilising domestic tax revenues comparedwith wealthier countries, and consequentlyremain far more dependent on external finance.Developing country governments shouldprioritise strengthening their tax systems – forexample, investing in administrative capacity,closing tax loopholes and expanding the tax base.

From a redistributive perspective, thecomposition of taxation matters. Direct taxes on

income, profits, property and capital gains holdthe greatest promise for economic redistributionand poverty reduction. Exempting all propertybelow a certain value from taxation or using aprogressive rate of income tax are obvious meansfor realising redistributive potential. Indirecttaxes such as value-added taxes, can beprogressive if there are exemptions for basicnecessities, such as food, and higher rates forluxury items. Fuel taxes can be differentiated toimpose higher taxes on fuels used by the wealthy.Where socially excluded groups aregeographically concentrated, the regional balanceof taxation could be altered.

The level and composition of publicexpenditure will have a particular bearing on the

Almost half of Peru’s population is of indigenousorigin, the great proportion of which live inpoverty. These groups have consistently loweraccess to health, education and livelihoodopportunities, setting them at a clear disadvantagefrom non-indigenous populations. Photographer: Jorge Villar Fernandez


social MDGs. While the overall level of publicexpenditure represents the government’s capacityfor influencing the MDGs, it is the compositionof this expenditure that will make a difference tosocially excluded groups. Reaching sociallyexcluded groups will require higher levels ofspending than for the rest of the population.

In addition, fiscal mechanisms can be designedto direct resources to excluded groups and regions.Indonesia has used the elimination of a subsidyon fuel, which benefited the more affluent insociety, to finance a targeted cash transferprogramme for the poor. Mongolia introduced awindfall profits tax in 2006 in order to benefitfrom historically high copper and gold prices toenable it to meet its social commitments.However, it was forced to rescind the tax in theface of fierce opposition from the miningcorporations. The Bolivian government has sofar had more success with its direct hydrocarbontax.

Governments can use equity-weighted transferrules to achieve more redistributive publicexpenditure patterns. Successful examples comefrom India and Brazil (see Box 5.4).

Governments should respond positively to thegrowing momentum for greater transparency inbudgetary processes, facilitating analysis of howexpenditure affects different social groups. Over thepast decade, initiatives to promote gender-responsive budget analysis, for example, havebeen introduced in more than 60 countries. Thisenables an assessment of the extent to whichpublic finance management is gender-sensitiveand whether policy commitments to genderequality are matched with resource allocations.

Greater transparency of budgetary processes canalso be promoted through greater publicparticipation. Participatory budgeting (PB) seeksto encourage active citizenship and hence ensuremore equitable public spending and greatergovernment transparency and accountability.Pioneered by the Workers’ Party in the Braziliancity of Porto Allegre, PB allows citizens to have adirect say in how municipal budgets are spent.There are many initiatives underway indeveloping regions based on the PB principle ofcitizen participation in decisions about allocationof public finances.

Studies have found that the PB process, wheneffectively implemented, can bolster social capital,trust and mutual cooperation,118 encouragecommunities to become conscious of new ways tochallenge poverty and inequality,119 promotegreater transparency and reduce clientelism,accelerate redistribution of public resources, andlead to more intelligent civic investment.

5.5 Legal policies and affirmative action Constitutions and legal systems provide theoverarching structure for outlawingdiscrimination on the grounds of race, gender,ethnicity and religious beliefs. However, legalequality has to be backed by explicit provisions forenforcement if it is to translate into practice. Anumber of countries have adopted explicitaffirmative action policies to deal withintersecting inequalities. India has perhaps one ofthe longest track records in reserving quotas formembers of tribal and Dalit communities inpublic education and employment, as well aspolitical quotas for women and these groups inlocal government.120

Other countries have opted for affirmativeaction in response to social mobilisation bydisadvantaged groups, which has often spilledover into conflict. Post-apartheid South Africa’sConstitution provides a strong framework forpromoting equality, including economic andsocial rights. It has also introduced affirmativeaction in public and private sectors thatprioritises race and gender. Post-conflict Nepalhas taken a series of affirmative action measures

Box 5.3 Progressive hydrocarbon taxation in Bolivia

In Bolivia, the government has introduced a number of fiscaltransfers based on a direct hydrocarbon tax. Two of these aredirectly redistributive. One is the Juancito Pinto programme,which targets districts with high dropout rates and low schoolattendance. The other provides minimum income support. Therest of the tax goes in block grants to sub-national governmentsand is not particularly redistributive, since it favours better-offgas-producing areas.117

Box 5.4 Using equity-weighted transfer rules to spendmore on excluded groups

In India, until 2007, district population size was the maincriterion for estimating need. Now, a new formula has beenadopted which attaches more weight to social indicators indistrict-level indices of need. While differences in educationspending per child in deprived and advanced districts werenegligible in 2005/06, by 2008/09, districts in the bottomquartile of the education development index were receivingtwice as much per child as those in the top quartile.

In Brazil, the federal government uses two levers to equalisepublic social spending. First, it sets regulatory standards toestablish national benchmarks for per capita spending. Thenorms are weighted for equity, favouring rural over urban areasand indigenous minorities over the rest of the population.Second, states whose tax revenues leave them below thestipulated threshold are eligible for complementary federalfinancing. In 2008, nine states in the poorest north and northeast regions received this financing.

Source: UNESCO (2010)107


to ensure that the exclusions of the past are notreproduced in the present.

Malaysia is frequently held up as a country inwhich affirmative action policies seem to haveworked (see Box 5.6), and there has been asystematic decline in ethnic disparities over time.But it is important to note that its successstemmed from the simultaneous pursuit ofaffirmative action and structural economicchange.121 Had growth been the sole focus, onlythose with productive resources would havegained, and thus social inequalities may haveincreased, resulting in renewed conflict. At thesame time, had the state only focused onredistributing existing wealth, excessive burdenon the minorities would have renderedaffirmative action politically and economicallyunfeasible.

The use of affirmative action by aneconomically disadvantaged but politicallydominant ethnic majority was not so successfulin the Sri Lankan case. While income disparitiesbetween the Sri Lankan Tamil and Sinhalacommunities were not large, the former hadenjoyed privileged access to education andgovernment jobs since colonial times. Thereplacement of English with Sinhalese as theofficial language, the use of educational quotas

and regional investment policies achieved thegoal of addressing Sinhala disadvantage, but thedrastic reversal of group inequalities left the SriLankan Tamils feeling that they could neverexpect equality of treatment. The conflict wasspurred by the apparent lack of any politicalavenues through which to resolve theirgrievances.122

The constitutional reforms that took place inLatin America in the 1980s were in response tothe growing tide of social mobilisation againstrepressive military governments. The activeparticipation of indigenous and blackorganisations within these movements led to thegreater incorporation of anti-discriminationstrategies in the campaigns of mainstreampolitical parties. The constitutional reformsrecognised the multi-ethnic nature of thesesocieties and, in some cases, acknowledged thespecific rights of indigenous peoples.

Colombia, Peru and Venezuela have politicalquotas for ethnic minorities, and specifically forindigenous peoples.123 A number of universities inBrazil have introduced quotas for Afro-descendants. Local governments, such as those inBrazil and Uruguay, have set up specialdepartments for Afro-descendant issues.124 Noneof these countries has designed political quotasfor ethnic minority women, whose representationin the national legislatures of Latin Americacontinues to be especially low. However, allcountries in Latin America and the Caribbeannow have a ministry or department dedicated towomen’s empowerment, and several have awomen’s ombudsperson or someone within thatdepartment dedicated to gender issues.125 Peruhas set up a human rights ombudsman to helpenforce the rights of indigenous people.

Debates continue about the desirability ofaffirmative action. Such policies may meet withvigorous opposition from previously privilegedgroups, leading to the outbreak of conflict. And ifthey are not time-bound, they can serve toreinforce the salience of identity as a factor inpublic life. At the same time, proponentsemphasise that the greater representation ofhistorically disadvantaged groups in public lifecould not have come about without such action.In India, where it has been in operation forseveral decades, quotas have succeeded increating a middle class of Dalits, something thatis missing in neighbouring Nepal.

Box 5.5 Affirmative action in Nepal

The post-conflict Interim Parliament in Nepal has stipulatedthat approximately one-third of all parliamentarians arewomen, and each ethnic group and the hitherto unrepresentedDalit caste are represented in proportion to their shares in thetotal population. Proportional representation was introducedfor the first time, with minority inclusion quotas alongside thetraditional first-past-the-post system (240 members wereelected via first-past-the-post, whereas the other 335 weredirectly elected through a proportional representation system).The parties fielding candidates for the proportional race wereobliged to put forward candidates on closed lists in accordancewith their share of the population (the source is the electorallaw). That, combined with selection of minority groupcandidates for the first-past-the-post race on the list of theCommunist Party of Nepal (Maoist), made the ConstituentAssembly the most inclusive state institution in Nepal’s history.

Box 5.6 Affirmative action in Malaysia

When Malaysia became independent, the majority Bumiputeracommunity were at severe economic disadvantage vis-à-vis theChinese minority. Broadly democratic institutions soon gavethe Bumiputera an entrenched political majority. Serious ethnicriots in 1969 led to the adoption of the New Economic Policy,which had a strong focus on correcting economic imbalances. Itintroduced quotas, targets and affirmative action with respectto education, land ownership, public service employment andownership of quoted companies.


5.6 Land reformThe MDG objectives on extreme poverty and thelivelihoods of the poor require policy action onthe economic front. Governments need to play anactive role in promoting the fairer distribution ofproductive assets in favour of poor and excludedpeople – for instance, through the reform of propertyrights. Land is critical in rural economies becauseit is a precondition for access to water, grazingrights, residential security and other resources.There is, of course, a natural limit to what can beachieved through radical land redistribution indensely populated countries like India. However,even where there is not enough land to benefit allsections of the poor, land reform can have theeffect of damping down the monopoly bargainingpower of large landlords. But as land reform islikely to be a long-term and challenging process,other measures should also be explored.

Land legislation should be revised to securelonger-term tenancy arrangements, and resolutionof disputes regarding interpretation andenforcement of land rental arrangements. Wheretenant protection has been vigorouslyimplemented by the state, as in West Bengal, it hasled to a rise in productivity.126 Elsewhere, poor and

marginalised groups have had to engage incollective action to press for recognition of theirclaims.

There is also a need to recognise differentlandholding patterns among indigenous people. Thenew Land Law in Vietnam, for instance, providesfor land allocation practices that accommodatecommunal land use patterns and also joint titling

Box 5.7 Tackling institutionalised racism in Brazil

In Brazil, the growing power of Afro-Brazilian groups within theWorkers’ Party contributed to the open promotion of the goal ofracial equality. In the mid-1990s, President Cardoso publiclyacknowledged that racism existed in the country, opening theway for public policies to deal with the problem.

Led by Lula da Silva, the Workers’ Party came to power in2002 on a platform of ‘Brazil without racism’. It stated thatpolicies for the promotion of equality of opportunities andtreatment should not be limited to the activity of isolatedagencies within the government, but should be theresponsibility of government as a whole in accordance with thelegal provisions.

A Special Secretariat for the Promotion of Racial Equalitywas set up in 2003 as an advisory body to the Presidency withthe mission of promoting racial equality and protecting therights of ethnic and racial groups.

Mato Grosso, Brazil: the Movimento dos Sem Terra (MST – Brazil's Landless Workers Movement) camp on theBordolandia farm, located between the municipalities of Bom Jesus do Araguaia and Serra Nova Dourada.Photographer: Eduardo Martino/Panos Pictures


to include women. The Sarawak Land Code grantsthe native customary land regime a special statusin recognition of the link between indigenousidentity and indigenous land. However,government attempts to modernise nativeagriculture, without due attention to differinglocal circumstances, have left the Dayaks, theindigenous people of Sarawak, worse off. Properlydesigned approaches to land rights requireguidelines for implementation that ensure effectiveprotection of the land rights of indigenous people.

In Latin America, indigenous groups havebeen active in demanding recognition ofindigenous territories and collective land rights.A review of the new constitutions and agrariancodes put in place in a number of LatinAmerican countries since the late 1980s showsthat they have made considerable gains.127 Landregularisation programmes for indigenous people

have also been used throughout the region,though they vary widely in their effectiveness.

Programmes for Afro-descendant ruralcommunities, modelled largely on the recognitionof indigenous communities’ collective rights, havealso been implemented in Colombia, Brazil andEcuador. Since 1993, Ley (Law) 70 has recognisedAfro-Colombian communities, seeking to protecttheir territories and strengthen their rights. TheBrazilian government has recognised these landssince the 1988 Constitution, but actual titling andthus legal recognition and support was stalleduntil 2001, with the beginnings of affirmativeaction.128 Ecuador passed a similar law in 2004,with policies and strategies for the developmentof Afro-Ecuadorian communities.129

Latin America has also made considerableprogress in the legal recognition of women’s landrights. This has been achieved through acombination of women’s own mobilising efforts,the transition to democracy in many countries inthe region, and the impetus provided by theConvention on the Elimination of All Forms ofDiscrimination against Women (CEDAW).

5.7 Promoting livelihoods, decent work andaccess to creditMDG 1 – reducing extreme poverty – wasamended in 2005 to include access to ‘productiveand decent employment’, since it is obvious thatfor most people, poverty cannot be structurallyaddressed without also creating opportunities fordecent work. Generating sustainable employmentis ultimately a matter for the macroeconomicpolicy framework. However, governments can takeaction to improve the equitable functioning oflabour markets with sector-wide minimum wageprovisions and the promotion of core labourstandards across all sectors.

While many countries have broad civilprotection against discrimination, as well asspecific labour legislation, implementation andenforcement is generally weak.130 The relevantministries and monitoring bodies are poorlyfunded and have limited expertise. The use ofcivil suits to compel enforcement or tocompensate victims of discrimination is notwidespread in most developing countries. In anycase, in many developing countries, most poorpeople work in the informal economy, puttingthem outside the purview of labour regulation.

Organisations for informal workers provideimportant models for creating dialogue betweenemployers, informal workers and governments inways that more traditional trade unions have oftenfailed to do. They have also succeeded in gettinglegal protection for their members, despite their

Box 5.8 Land rights campaigns in India

Ekta Parishad, a federation of community-based organisationsin India, campaigns on land rights across eight states. Itbelieves that granting rights to small pieces of land for thelandless poor, thereby reducing their dependence on casualwage labour, is critical to breaking the cycle of poverty.However, even where the poor have legal entitlement to land,they are not able to claim it.

In December 2005, an assembly of representatives oflandless groups agreed a set of demands on land andlivelihoods. If the government of India would not meet thosedemands by October 2007, 25,000 landless people wouldmarch (padyatra) to Delhi to highlight their plight.

They developed three specific demands around land reformin order to mobilise people for action to create fairer policiesfor the most vulnerable groups:

Establish a National Land Authority to provide a clearstatement of land utilisation in India, identify the landsavailable for redistribution and strengthen pro-poor lawsEstablish fast-track courts to settle past and future conflictsrelated to land Establish a single window system so that farmers canresolve land issues easily and freely, without wasting time,money and energy.

In October 2007, 25,000 landless people, many of themAdivasis and Dalits, gathered to walk together to claim theirrights. They walked almost 350km from Gwalior in MadhyaPradesh to Delhi, marching for almost a month. Theyhighlighted their plight from state to state, raising public andmedia interest in their cause.

Faced with growing public attention and pressure, thegovernment agreed to speak to the representatives. The Ministerof Rural Development came to the assembly ground andannounced that the government would agree to their demands. ANational Land Reform Commission was created to developpolicies on land reform. It is now headed by the Prime Minister.

Source: Christian Aid (forthcoming)99


informal status. Development organisations haveworked with dispossessed tribal groups in Orissa,India, who had been forced into bonded labour,to take up resources offered under governmentrehabilitation schemes and to organise themselvesto bargain with their landlords. Sankalp, in UttarPradesh, supports the self-help organisation oftribal mine workers in forming worker co-operatives and applying for mine leases as a toolagainst child and bonded labour. Socialmobilisation efforts in these areas have resulted inchildren being sent to school and adults startingup their own literacy classes. Self-helporganisations have also encouraged these groupsto get more involved in the running of localschools in order to counter teacher absenteeism.

Promoting the skills and productivity ofmarginalised groups is essential if they are to escapepoverty. The IADB has been funding ‘demand-driven’ training and employment programmes ina number of Latin American countries to addresspoor labour market performance among ‘low-income, high-risk’ youth. These programmesconsist of short-term traineeships orientedtowards basic work skills and practices, followedby an internship, which allows participants togain valuable experience in a formal sector job.Evaluations report that, while their impacts varyacross countries, women and younger cohortshave benefited in terms of increased likelihood ofemployment. There was also some evidence of apositive impact in terms of quality ofemployment among those who found jobs. Thefirst job experience has been found to besignificant to longer-term job performance andearnings potential.131

The outreach and design of agriculturalextension services could be considerably

improved to better meet the needs of indigenousfarmers. Women farmers in particular arepersistently excluded from such services.132

Research has endorsed the broad-based publicprovision of agricultural extension services butsuggests that some degree of targeting may benecessary to tailor services to local needs.133

However, geographical targeting only partlyaddresses the problem if extension services aredisproportionately taken up by majority groupsliving in areas with ethnic minorityconcentrations. Specific interventions andprogrammes targeted geographically at poor areasneed to be appropriately tailored to the problemsand needs of ethnic minority households.

Microfinance initiatives have made someheadway in delivering financial services to the poor.Many non-governmental organisations and evengovernments (for example, India) have beeninvolved in pro-poor lending programmes.

Table 5.1 Collective land rights in new constitutions and agrarian codes

Recognition of indigenous land claims Recognition of Recognition of Possibility ofCountry Constitution Code collective customary law privatising

indigenous lands collective land

Bolivia 1994 1996 Yes Yes NoBrazil 1998 No No No NoChile No Yes (1993) No No Yes (1979); No (1993)Colombia 1991 1994 Yes Yes NoCosta Rica No – – – –Ecuador 1998 1994 Yes Yes Yes (1994); No (1998)El Salvador No – – – –Guatemala 1998 – Yes Yes NoHonduras No 1992 Yes No –Mexico 1992 – Yes Partial YesNicaragua 1987 – Yes Yes NoPeru 1993 1995 Yes Yes Yes

Source Deere and León (2001)127

Box 5.9 Informal workers organise to claim rights

The Self Employed Women’s Association in India has beenworking with women workers in the informal economy – bothwaged and self-employed – since the 1970s. It combines a tradeunion approach to engage in collective bargaining on behalf ofits members with a co-operative approach to promote theirlivelihood strategies.

It has more than 700,000 members spread across a numberof states in India and lobbies the government to win legalrecognition of members’ rights, including minimum wagelegislation. It played an important role in the government’sadoption of the Social Security for the Unorganised Sector Act.It has also been active at the international level and is one ofthe founding members of a global network of women workers,Women in Informal Employment: Globalizing and Organizing(WIEGO), which collects data, conducts research and engages inadvocacy on behalf of its members. They were an active forcebehind the adoption of the ILO Convention for Homeworkers.


However, the vast majority of them areconcentrated in Asia. In the South Asian context,group-based lending of the kind pioneered byGrameen Bank in Bangladesh, and the self-helpgroup model pioneered by organisations likeMYRADA (Mysore Resettlement andDevelopment Agency) and PRADAN(Professional Assistance for Development Action)in India, have been found to be more successfulin reaching poor and excluded groups thanindividual lending.134 One study in India foundthat the poor comprised 43 per cent of borrowersunder a group-based approach compared with 15per cent under individual lending approaches.Thirty-five per cent of their members were fromDalit and Adivasi groups, compared with 32 percent for individual lending programmes. Theyalso performed considerably better at reachingwomen: 95 per cent, compared with just 35 percent for individual lending programmes.

The kind of group-based approach used canalso make a considerable difference to outcomes.Studies from South Asia suggest that organisationswhich specialise in the minimalist microfinanceapproach do very little to empower their clients,often women, in terms of participation in theirwider community or in politics.135 In theBangladesh context, it has been found thatsavings-led group strategies have been moresuccessful in generating impacts in communityand political participation than mainstreamcredit-led approaches.136 It is important to combinemicrofinance provision with a range of supportiveservices, including livelihoods training and marketaccess, if clients are to graduate out of poverty.

5.8 Investing in infrastructure and areadevelopmentThe spatial concentration of poverty and socialexclusion means that certain areas of a countryface intersecting development deficits – in termsof infrastructure, services, markets – which setthem apart from the rest of the country. Theseareas may be remote rural locations or urbanslum neighbourhoods. National, sector-specificor targeted group approaches are unlikely tohave much impact on their situation. A morecomprehensive area development approach will benecessary to deal with the structural causes ofdisadvantage as well as tackling immediateproblems. These can encompass roads, publictransportation and household utilities such aswater, sanitation and electricity and socialservices.

The area development approach is moreeffective because it allows governments to actsimultaneously on a number of inequalities andto benefit from the resulting synergies. Forinstance, the construction of transportinfrastructure not only contributes to economicactivities and poverty reduction but alsoimproves the access of women and children toschools and health services. A decentralisedapproach and attention to local governancebecomes increasingly important for area-baseddevelopment, given the evident limitations ofcentralised, top-down systems of planning,budgeting, implementation and administration.

The Northern Mountains CommunityDevelopment Project in Vietnam is a multi-sectoral decentralised project to assist 1 millionrural poor people, 85 per cent from ethnicminority groups, in one of the country’s poorestregions. The project used a participatoryapproach to build infrastructure and socialservices in the region: this included schools,commune health stations, roads, improved watersupply, new irrigation schemes and agriculturalextensions services. In addition, evaluations of thefirst phase suggest that average household incomein the area more than doubled between 2004 and2007.137 The project is now in its second phase.

The proposed Mexico Indigenous People’sDevelopment Project is another example of area-based development directed at socially excludedgroups. This will involve infrastructure sub-projects in water, sanitation, electrification andrural roads. In addition, indigenous people’sparticipation will be built into the project with aview to ‘building social capital’ for development.

Indigenous groups have often been subject toforcible and disruptive resettlement to make wayfor large-scale infrastructure projects that they

A young Colombian girl from the mainly Afro-descendant area ofQuibdó in Colombia’s Chocó Department. The MDG Fundprogramme on nutrition and food security is working withschools to ensure children are aware of, and have access to, ahealthy diet. Photographer: David Parra


have not been consulted about and are unlikely tobenefit from. Indigenous people represent aspecial case for resettlement because of their deepattachment to ancestral lands. In many cases,smaller projects can yield higher returns. However,large infrastructure projects are likely to beundertaken if their gains are considered tooutweigh their costs. If resettlement is deemednecessary, it needs to be designed in directconsultation with those affected and should beplanned to improve their living standards, physicalsecurity, productive capacity and income levels.They need to be offered a direct andcommensurate share of the benefits of the mainproject. In irrigation projects, they can receive titledeeds and actual use of irrigated lands as part oftheir compensation. In some countries, indigenousgroups have been trained as forest managementexperts and forest guards, and they continue tolive in the forest protection zone created with theconstruction of a hydroelectric complex.

5.9 Extending basic services to all groupsLabour remains the most abundant asset of poorand socially excluded groups. Investments in theirhuman capital and human capabilities have beenfound to contribute to people’s survival andwellbeing, as well as providing an importantescape route out of poverty and social exclusion.Such investments are central to the social MDGs.

Providing physical accessMaking social services more inclusive will requireaction on the various constraints that exclude poorpeople. Physical access is the first policy challenge.Directing services to areas that are under-servedby service providers can be combined with astrong element of demand mobilisation todetermine the shape and form of services. InIndia, the Educational Guarantee Scheme inMadhya Pradesh guarantees state provision of aprimary school to children in areas where there isno such facility within a kilometre, within 90days of receiving a demand from the community.Eligible communities must have at least 40learners in the 6–14 age group, but in tribal areas,only 25 learners are necessary. Evaluationssuggest that tribal areas have benefiteddisproportionately from this provision and thatthere has been a sharp reduction in the numbersof out-of-school children in these communities.

In terms of healthcare, governments have tofind ways of connecting excluded groups withqualified professionals, particularly in rural areas.Gonoshasthya Kendra, a health-based NGO inBangladesh, adopted the barefoot doctor modelfrom China to train female paramedics to

provide healthcare outreach to the poor. It wasextremely successful in bringing down maternalmortality rates in the areas it served. InIndonesia, doctors had to complete compulsoryservice in health centres for five years, withshorter periods for more remote areas, beforethey could obtain a lucrative civil service post.This increased the number of doctors in healthcentres by an average of 97 per cent from 1985 to1994, with gains of more than 200 per cent forremote rural areas.

Malaysia and Sri Lanka have achievedsubstantial reductions in maternal mortality ratesby making competent professional midwives andsupervisory nurse-midwives widely available inrural areas to assist in deliveries at home or insmall rural hospitals. In Afghanistan, rovingextension clinics visit sparsely populated areas toprovide care locally or help with transportationto better facilities when required. Transportationor transport subsidies are used to enableemergency visits to hospital in Bangladesh.

In China, there have been substantialreductions in maternal mortality in the poorer,rural western provinces as a result of targetedgovernment interventions.138 The mostoutstanding performance was in Yunnan, one ofthe poorest and most remote provinces, wherethe maternal mortality ratio declined from 149 in1989 to 101 per 100,000 live births in 1998.139

Making healthcare affordableAffordability is the second policy challenge. In thelight of the dramatic impact of the imposition ofuser fees, a number of African countries have nowabolished fees or phased them out, replacing the

India, Jaipur district, Rajasthan: in the village of Dhani Bagaria a Dalit woman works in front of her home putting down a newlayer of mud as a young girl gathers firewood. Photographer: Mikkel Ostergaard/Panos Pictures


lost income from pooled funds (e.g. publicsubsidies, insurance and pre-payment schemes).This has resulted in substantial increases inservice utilisation, particularly by the poor. InUganda, for instance, the abolition of user fees ledto a sudden, dramatic and sustained uptake ofhealth services.140 Similarly positive responses werereported in relation to the abolition of school feesin Kenya and Malawi. The pooling of resourceshas the potential to build solidarity across thesystem, allowing some degree of cross-subsidy.

‘Pre-payment and pooling institutionalizessolidarity between the rich and the less well-off, and between the healthy and the sick. Itlifts barriers to the uptake of services andreduces the risk that people will incurcatastrophic expenses when they are sick.Finally, it provides the means to re-invest inthe availability, range and quality of services.’141

Making services relevant and representativeHowever, the increased demand for schoolingand health services has drawn attention to the

importance of investing in the quality andrelevance of services, particularly from the pointof view of socially excluded groups. Recruitingand training service providers from thesecommunities is an important means of signallingstate commitment to more inclusive services andhelps bridge the social distance betweenmainstream services and excluded groups. It alsohelps to build parental involvement andcommunity support. Mongolia has experimentedwith the establishment of pre-school units usingthe traditional gers (a type of yurt) as trainingcentres during the summer. Teachers are nomads,moving with their families and stock, togetherwith a group of households involved in pre-school education programmes. In India, theRural Litigation and Entitlement Kendra (RLEK),together with the Van Gujjars (indigenous forestnomads) in Uttaranchal, have devised aninnovative approach whereby local teachersmigrate with the community and implement aculturally adapted curriculum.

Reliance on women community healthworkers in countries where there are restrictions

A social worker in the northern Tigray region of Ethiopia gathers data and information on communityhealth conditions. The role of rural social workers is critical to ensuring that all citizens are accounted forin national interventions to achieve the MDGs. Photographer: Sophie de Caen, MDG Achievement Fund


on women’s mobility in the public domain havemade a considerable difference to maternalhealth. Initiatives in Pakistan, Nepal andBangladesh have raised awareness of healthconditions, addressed taboos and created demandfor reproductive health services, enabling womento claim their health rights. In Nepal, which hadextremely high maternal and child mortalityrates, women health workers have beeninstrumental in improving health outcomesbecause of their proximity to the community.Health services offered in local languages or bywomen doctors have also proven to be effective.

Public action and community mobilisationcan play an important role in extending servicesto socially excluded groups. In Mexico, forexample, mobilisation by women’s reproductivehealth groups in collaboration with a researchinstitute that focused on budgetary analysishighlighted long-standing inequalities in federalsupport to the poorest states, which had thehighest rates of maternal mortality. In Bolivia,India and Nepal, community mobilisationthrough participatory women’s groups helped toimprove birth outcomes in poor ruralcommunities. Initiatives included setting upwomen’s organisations, developing women’s skillsin identifying and prioritising problems, andtraining community members in safe birthingtechniques.

In Bolivia, community mobilisation led to adecline in perinatal mortality and an increase inthe proportion of women receiving prenatal careand starting to breastfeed on the first day afterbirth. In Nepal, the same approach led to areduction in neonatal mortality and an evenlarger and statistically more significant effect onmaternal mortality rates. In the Indian context,where the intervention was carried out in twostates with high proportions of Adivasi groups, itwas associated with a dramatic fall in neonatalmortality rates.

Extending access to educationIn the field of education, nurturing womenteachers in Pakistan and Afghanistan has hadsome success, such as chaperoning girls to schooland back, assuring their physical security but alsoaccommodating conservative notions of‘propriety’.142 For older girls, school sanitation(including running water and privacy) is anissue. Enclosed toilets can help address parents’notions that their menstruating daughters shouldnot attend school. In caste-divided communities,sharing a water tap at school or eating mealstogether can help tackle discriminatory attitudeslearned at home or in the wider community.143

There is now increasing recognition of theneed for bilingual education for minority groups,initially providing instruction in the mothertongue, and gradually moving on to themainstream language. Studies suggest that theimplementation of this education model canimprove the performance of minority languagegroups, increasing enrolment rates, improvingeducational results and reducing gender gaps inschooling, thus increasing the likelihood oflearning the mainstream language.

The Brazilian government has created a law(10.639) to teach about the history of Africa andAfro-Brazilians. It has also begun a national schooltextbook programme to substitute books depictingracist stereotypes with those that promote thediversity of Brazilian society. In India, schoolbook

Box 5.10 Mobilising around maternal mortality in Mexico

During the 1990s, women’s groups in Mexico came togetheraround the issue of maternal mortality, supported byinternational initiatives, including the MDGs. In 2001, thegovernment launched Arranque Parejo en la Vida (A Fair Start inLife, or APV), a programme focused on women during theirchildbearing years and children under two years old, as part ofits commitments to meeting the MDGs. However, maternalmortality rates among indigenous groups remained high.

In 2002, Fundar, a research institution that specialised inbudgetary analysis, was asked to do an assessment ofgovernment policy around maternal health. They formed acoalition with a diverse group of women’s organisations,including those with expertise in reproductive health, and inmedia relations. One important partner organisation was K’inalAntzetik (‘Land of Women’ in the Mayan language).

The coalition’s research found that rates of maternalmortality varied widely between states, and that public fundswere not being spent in areas with the highest rates. Even aftersuccessfully campaigning for greater allocation to APV in the2003 budget, they found huge inequalities in expenditureacross states. For example, in 2003, Chiapas (with a maternalmortality rate of 93.2 per 100,000 live births) was allocated980,192 pesos for APV, while Nuevo León (with a rate of 28.5)received 23,968,069 pesos.

In addition, allocated funds were not being spent; inChiapas, which received its highest ever budget allocation in2003, only 45 per cent of the resources had been spent. Thecoalition used their research findings to campaign for policyimprovements, and the results were publicised through mediacampaigns and public meetings.

The success of the coalition’s campaign was, in part, linkedto its international element; the fact that the United Nationsrepeatedly endorsed the reduction of maternal mortality, andthat the Mexican government was a signatory to severalinternational initiatives, gave the cause both social andpolitical legitimacy, giving the coalition a moral lever in itsengagement with government. It helped place the issue of safedelivery firmly on the policy agenda and in 2009 and 2010, anumber of states have seen a dramatic decrease in maternalmortality rates (see Box 5.11).

Source: Layton et al. (2007)29


examples of famous Dalit personalities aregradually being recognised as a tool to create prideand address prejudice. Behaviour-changecommunication in general is an under-utilisedtool that could be used to great effect, notably inschools, health institutions, and public offices.

5.10 Inclusive social protection Social protection has emerged as a criticalcomponent of development policy in response tothe recurring financial crises that have

accompanied the process of globalisation. Socialprotection policies for the poor include socialtransfers (in cash and in kind) as well as publicworks programmes, school-feeding programmesand community-based insurance schemes. Theyare now estimated to reach over 150 millionhouseholds in poor countries and benefit aroundhalf a billion people.144

The emerging literature on social protectionemphasises its redistributive potential and itssynergy with development objectives.145 Its

Box 5.11 Reducing maternal mortality in Mexico

Figure 5.1 shows a significant reduction of maternalmortality in the state of Chiapas, Mexico, between 2009 and2010. This has been spurred by several state-level policies:

Working with indigenous midwivesIn Chiapas, 56 per cent of all pregnancies/deliveries arehandled by (mainly indigenous) midwives. The collaborationwith indigenous midwives has been critical in reducingmaternal mortality: without their inclusion in interventions,the likelihood of indigenous women using the healthservices is very remote.

1,000 of the most active midwives received capacitybuilding focused on various aspects of reproductive healthsuch as clean delivery and how to detect ‘alarm signs’. These women were given an essential midwifery tool kitwith stethoscope, gloves and special scissors for cuttingthe umbilical cord and cell phones connected free ofcharge to their closest hospital and clinic.A new ‘Hospital of Cultures’ has been opened offeringregular training that integrates scientific and traditionalknowledge. This initiative has led doctors to accept andwork with the traditional indigenous practice of verticaldelivery.

17 ‘Motherhood Houses’, staffed by both doctors andindigenous midwives, have been established. Instead ofreturning to their villages, women can stay there whenclose to delivery, to ensure they can return to thehospital in time.

Working with community networksOver the past five years, networks of health committeeshave been established with the participation of the mayorsof the 118 municipalities. These committees know aboutpregnancies in their area.

Some 3,000 rural health auxiliaries, who are elected bytheir communities and have already been active for someyears, have received training and health kits. Theyidentify pregnant women in the community and givethem the necessary information for a healthy pregnancy.Links are maintained with local NGOs to ensure thatinformation and advice is consistent.

Working with hospitalsFifty per cent of hospital admissions are for obstetric care.Between 2009 and 2010, over US$50 million was invested inhospital infrastructure.

17 new hospitals have been built, each with 60 beds,some of which are reserved for women. Hospitals will besupported by smaller clinics, Basicos Communitarios,who will deal with less urgent cases and follow up.138 health centres have opened across the state, eachwith a live-in doctor, Monday to Friday.To support hospital transfer, the state government haspurchased an ambulance for each municipality. Statehelicopters may also on occasion do emergency rescuesin rural areas.

Working at the national levelThe Health Secretary has coordinated with the System forthe Integral Development of the Family (Dif) to ensurethat, in indigenous areas, all pregnant women receive anutritional package with vitamins, iron and othersupplements. This has helped to improve the health ofnewborn babies and to reduce child mortality.There is a national agreement on ‘Agreement of UrgentObstetric Care’ which obliges all clinics to attend tourgent pregnancy cases free of charge regardless ofhealth insurance coverage.This has been supported by the national programme‘Health Pregnancy’, started in 2009, which allows womento deliver at a hospital free of charge, if they come forpre-natal care.

Source Interview with Secretary of Health, State of Chiapas,Mexico (James Gómez Montes), 31 July 2010

0

20

40

60

80

100

Vera Cruz Oaxaca Mexico State Chiapas

◆ Maternal Mortality Rate 2008◆ Maternal Mortality Rate 2009

Figure 5.1 Maternal mortality rates, selected states inMexico

Source Subsecretaria de Prevencion de la Salud (2010)‘Strategia Integral para Acelerar la reduccion de laMortalidad Materna en Mexico’, power point presentation


redistributive potential rests on its success inreaching poor and marginalised groups, as well asdesign mechanisms to guard against leakage andhigh administration costs. Where they have beenwell designed, social protection measures havehelped to make development efforts moreinclusive on a variety of fronts. Indeed, as long asthey are well targeted towards the very poor,social protection measures are likely to benefitsocially excluded groups because they comprise adisproportionate share of the poor.

Targeted schemes can be effective in reachingsocially excluded groupsThis is the case, for instance, with Mexico’sconditional cash transfer programme,Oportunidades, where mothers in low-incomehouseholds receive monthly cash transfers oncondition that their children attend school andhealth clinics. As a well-targeted programme, it hasbenefited indigenous peoples disproportionately,although it is still not reaching some of the poorestindigenous communities who need it most but arehardest to reach. Indigenous children in families orcommunities that are receiving the transfers are aslikely as non-indigenous children to remain inschool, and the programme has been successful inclosing gender and ethnic disparities in education.

In Brazil, 21.7 per cent of Afro-descendantfamilies (based on the recipient race) received theBolsa Familia (family grant) compared with 8.7per cent of white families, which reflects theracial differences in levels of extreme poverty.146

Unconditional cash transfers in the form of thechild support grant in South Africa, and socialpensions in Brazil, South Africa, Namibia,Lesotho and Botswana, have also effectivelytargeted poor and excluded groups. Indeed, theold age pension in South Africa is regarded as themost effective social programme in targeting andreaching economically vulnerable groups, and isthe primary source of income for older peoplewho would otherwise be living in abject poverty.

The newly enacted National EmploymentGuarantee Scheme in India is a self-targetedprogramme with strong inclusive elements builtinto it. It requires that a third of jobs should bereserved for women and crèches provided wherethere are more than a certain number of womenon a scheme. It also stipulates that some of theinfrastructure projects should be used to promoteland improvement and other assets for sociallymarginalised groups. Early evaluations confirmhigh levels of participation by women, Dalits andAdivasis. The scheme works best in areas wherethere are civil society organisations able to mobilisethose groups who are entitled to participate.

Positive direct and knock-on effectsWhat is important to note in the emergingliterature on social protection is not only theability of well-designed programmes to reachsome of the most excluded groups, but also theirpotential to generate a range of positive direct andknock-on effects. This broader developmentalimpact helps to offset some of the concernsaround the affordability of redistributive policies.

For instance, the use of public works to buildlocal infrastructure has been found to generatestrong multiplier effects in the local economy andin terms of access to social services. Pensions forolder people in South Africa and Namibia havehad the knock-on effect of improving children’snutrition and education, as well as female labourforce participation among African households.147

The regularity of the pension has also eased accessto credit and provided a stimulus to local trade.

Social protection schemes also have thepotential to address other aspects of socialexclusion. In Andhra Pradesh, India, thewillingness of the state government to engage withwomen’s self-help groups in managing a nationallymandated school-feeding programme ensured thatmany lower caste women were employed to workas cooks, and that many of the children whobenefited were from the lowest castes.

In rural Malawi, a cash transfer programmedirected to women through bank accounts issuedthem with smart cards containing their personaland bank account details. This achieved a degreeof financial inclusion for thousands of ruralfamilies who had previously been excluded from

Afro-descendant children from the coastal areas of Colombia’sChocó Department where infant mortality rates are double thenational average, with 60 per cent of children aged 1–4 sufferingfrom anaemia. The Chocó has the lowest social indicators and ishome to thousands of internally displaced people due toColombia’s national conflict. Photographer: David Parra


financial services. Evaluations also revealed thatthe women who received identity documentsand/or smart cards felt strongly empowered bythe legal recognition that these documentsrepresented. In focus group discussions, severalwomen stated passionately that, before theproject, it was as if they did not exist in the eyesof the state, but now that they had their ‘papers’,they had an identity and the government couldno longer ignore them.148

Similarly, the citizenship potential of the BolsaFamilia149 is illustrated by the views of femalebeneficiaries drawn from ‘a Brazil that is poor,non-white and with scarce social mobilityopportunities’. For these women, receipt of thetransfers constituted their first contact with thestate. The need to obtain documents such asbirth certificates and identity cards in order toregister and apply for the Bolsa Familia broughthome to them their membership of a larger socialentity beyond their immediate community. Asone of the managers put it, it ‘caused a revolution,an overturning of awareness of the social spacethat they can seek to belong to…’.150

The contribution that social protection

schemes can make to social cohesion has alsobeen explored through studies of cash transfers inthe post-conflict context of Colombia.151 Here,women’s regular participation in various activitiesrequired by the programme (meetings, trainingcourses, visiting health centres) contributed tobuilding trust within the community, as expressedby their willingness to invest in public goods thatwould benefit all members.

Several states have now institutionalised socialprotection as a basic right. In some countries,such as Thailand, this has taken the form ofuniversal provision: what started out as the ‘30baht scheme’ to provide access to healthcare forinformal workers had expanded into universalcoverage by 2002, the culmination of a 27-yearprocess which began with a low-income schemefor the poor. Elsewhere, it has taken the form of a‘social guarantee’ approach, which seeks toincorporate the commitment to access, quality,continuous revision and participation andredress within service provision.152 The Regime ofExplicit Guarantees in Health in Chile is anexample of an approach that has been conceivedwithin such a framework.

An afterschool centre in the township of Katatura, Windhoek, Namibia. Photographer: Barbara Cheney


‘We resolve therefore to respect fully and upholdthe Universal Declaration of Human Rights…’(Millennium Declaration 2000, para. 25)

Ethnically diverse societies do not have to beethnically divided societies. Cultural or religiousdifferences only harden into exclusion andconflict when there is systematic discrimination –when people are denied resources, recognitionand representation because of who they are, whatthey believe or where they live. The persistence ofhistorically established patterns of exclusion andthe conflicts that they generate can give rise to adeep sense of despair and hopelessness. However,there are enough examples of progress, many ofwhich have been touched on in this report, tosuggest that change is possible if we learn thelessons of history. A range of practical policiesand interventions has brought about change indifferent parts of the world. This leads us tosuggest a number of key concerns, principles andrecommendations that can provide the basis forcontinued efforts to tackle social exclusion.

Using a rights framework to tackle socialexclusionSocial exclusion entails the denial of fullpersonhood and full citizenship to excludedgroups. The first principle, therefore, relates tothe importance of locating efforts to overcomesocial exclusion within the broad frameworkprovided by the International Bill of Humanrights. This includes the Universal Declaration ofHuman Rights and the two Conventions adoptedon the basis of that Declaration, the InternationalConvention on Civil and Political Rights and theInternational Covenant of Economic, Social andCultural Rights. Although these provide a highlyabstract normative framework, they uphold theimportance and indivisibility of civil, political,economic, social and cultural rights and enshrinethe commitment to equal worth and dignity ofall human beings that is the underlying rationalefor tackling social exclusion.

How this is translated into actionable policiesmay involve many different elements: theconstitutional enshrinement of rights, the use ofthe law, the provision of social guarantees, thestrengthening of inclusive democratic process,and the systematic right to information and

transparency. Other human rights conventions,in particular, the Convention on the Eliminationof All Forms of Discrimination Against Women,which requires states to report on progress inpromoting gender equality, help to buttress theUniversal Declaration.

Engaging citizens in changeA second principle stems from concerns relating tothe question of citizenship. A great deal of theliterature revolves around a vertical model ofcitizenship – one that focuses on democratisingrelationships between states and their citizens. Butthe social exclusion literature brings home the needto complement this with a horizontal view ofcitizenship – one that stresses that the relationshipbetween citizens is at least as important.153 It is notthe state alone that is responsible for tackling thebarriers of discrimination. The agenda for changemust also focus on citizens themselves, particularlythose whose attitudes and prejudices serve toperpetuate discrimination on an everyday basis.The principles of tolerance and solidarity spelt outin the Millennium Declaration are, in other words,essential in any agenda to promote equality.

Getting the right balance between equality anddifferenceThe third principle relates to the challenge ofbalancing equality and difference. In other words,to what extent can broad-based or universalpolicies to promote equality of opportunity becombined with making special provision forthose who have been systematically excluded inthe past? While these are often treated asmutually incompatible approaches, they can, infact, work successfully in tandem. Universalistapproaches are essential to building a sense ofsocial solidarity and citizenship, particularlycritical for excluded groups. Universal coveragealso gives privileged groups more of a stake inpolicy outcomes, a greater willingness tocontribute to them, and hence the greaterpossibility of redistribution through cross-subsidies from the affluent to the poor.

At the same time, the fact that it is their‘difference’ from the rest of the poor that has ledsocially excluded groups to be left behind orlocked out of processes of growth anddevelopment suggests that ‘universality’ should

Conclusion: key concerns and principles 6

58 Conclusion: key concerns and principles

not be taken to imply ‘uniformity’. There arestrong grounds for plurality and diversity withinuniversal frameworks of provision.

To date, targeted programmes have generallyserved as a means of compensating forgovernment weakness in delivering universalservices rather than as a means of addressingexclusion. The achievement of ‘inclusion bydesign’154 will need careful thought, with attentionto building in incentives to promote suchoutcomes. Targeted programmes may end upmarginalising some groups just as much aspoorly implemented universal approaches if theyrely on ‘labelling’ practices that reinforcestigma.155 In the final analysis, the greater the

participation of excluded groups in the design ofprogrammes and in the political decision-makingprocesses that affect their lives, the less necessaryit may be to address their problems in isolationfrom the rest of the population.

Achieving transformative changeA fourth principle is the need to go beyondameliorative approaches that address thesymptoms of the problem to transformativeapproaches that address its root causes. It is quitepossible to meet the basic needs of poor andmarginalised groups without strengthening theircapacity to do so themselves, thereby leavingtheir longer-term vulnerability intact. Addressing

Women in a desert village between Mauritania and Senegal play the traditional rhythms and dance to themusic that is an integral part of their culture. Photographer: Layla Saad, MDG Achievement Fund


the causes of social exclusion means breaking theprocesses by which disadvantage is reproducedover time and across generations. There is scopehere to explore multi-pronged approaches thataddress different dimensions of disadvantage.

Group-based exclusion requires group-basedsolutionsA fifth principle stems from the inadequacy ofpolicies that target individuals or households intackling problems that are essentially collectiveand group-based. In fact, individual solutionsmay leave marginalised groups more isolated andimpoverished than before. The need for morecollective approaches introduces another routethrough which ‘difference’ may have to be builtinto the design of policies. This report hasprovided a variety of examples that emphasise thepower of numbers and the importance of solidarity.

A new social contract for an interconnectedworld: states, citizens and global governance The MDGs have provided a major impetus forcoordinating national and international efforts toreduce poverty and promote human development.They have helped to mainstream the fight againstpoverty in the policies, plans and programmesimplemented across different regions of the world.But they fell short of the social justice agenda speltout by the Millennium Declaration. This reporthas examined the structural factors that give riseto deeply entrenched and intersecting inequalitiesand the persistence of social exclusion. It has putforward a number of strategic policy options thatcan help to transform the MDGs into a pathwayfor social justice.

These are not options to be pursued inisolation. The multidimensional nature ofpoverty and social exclusion requires suchoptions to be pursued as part of a broaderagenda of social transformation. This calls for anew social contract between states and citizens.The report has touched on various policies andinterventions that can contribute to buildingmore responsive states and active citizens.

What has been left out of the discussion is therole of the international community, which is alsocritical for the more equitable achievement of theMDGs. The problems of poverty and social exclusionare not purely national in their causes or theirconsequences. They are the product of structuralinequalities at the global level. Yet this is notacknowledged in the MDGs. While MDGs 1 to 7concern objectives to be achieved by developingcountries, MDG 8 encompasses the relationships

between developed and developing countries.156 Itconcerns global partnerships: betweengovernments (including multilateral institutions),the corporate sector and global civil societyorganisations. Not only does it fail to address theunequal nature of these relationships – asmanifested in aid, trade and debt – but it is also theonly MDG with no targets or indicators to monitorany form of progress. MDG 8 has been criticisedfor singularly lacking ‘vision, bark and bite’.157

It is extremely unlikely that developingcountries will be able to achieve growth, prosperityand social justice without greater attention tobuilding greater solidarity, a genuine partnership ofequals, at the global level. This should constitutethe central platform for the post-2015 successor tothe MDGs. However, in the run-up to 2015, richcountries should prioritise a number of actions toaccelerate the pace of progress:

Honour the commitment to increase OfficialDevelopment Assistance (ODA) to 0.7 percent of GNP.

Recognise and support the role of civil societyorganisations in mobilising excluded groups,holding governments to account, andtransforming power relationships.

Provide support for a universal socialprotection floor that allows marginalisedgroups to cope with risk and invest in theirfuture.

Promote fairer trade relations, with specialattention to the needs of poor farmers, tradersand workers, particularly those in theinformal economy.

Hold corporations and the private sector toaccount for socially responsible investments.

Strengthen government capacity andcommitment to the international humanrights framework.

‘Responsibility for managing worldwide economic and socialdevelopment, as well as threats to international peace andsecurity, must be shared among the nations of the world andshould be exercised multilaterally. As the most universal andmost representative organization in the world, the UnitedNations must play the central role.’

(Millennium Declaration, para. 6)

60 Further information

Project informationThis report has been financed by the MDGAchievement Fund and produced by the Instituteof Development Studies, Sussex, in the beliefthat shedding light on the systematic injusticesfaced by women, men, young people andchildren around the world can ignite ourconviction to embark on a collective fightagainst poverty and for the equitableachievement of the Millennium DevelopmentGoals. It has been prepared in the lead-up to theMDG Review Summit to be held at the UNGeneral Assembly in September 2010 in order toplace the issue of social exclusion andinequalities firmly on the agenda. It outlines aset of policy recommendations and interventionsthat governments in poor and middle-incomecountries can use to tackle social exclusion andinequality in their efforts to achieve the MDGs.The report provides both a substantive input aswell as an advocacy tool for the Summit and itsfollow-up in the years up to 2015.

Further information

The United Nations MDG Achievement Fund

The United Nations (UN) MDG Achievement Fund was created inDecember 2006 with a generous contribution from thegovernment of Spain in fulfilment of its aid commitments inrelation to MDG 8 (developing a global partnership fordevelopment). With almost $700 million invested in 128programmes in 49 countries around the world, the MDGAchievement Fund is currently the largest global fund dedicatedto achieving the MDGs. We work closely with nationalgovernments, citizens and civil society organisations to tacklepoverty and inequalities and speed up progress on the MDGs.The MDG Fund represents a unique initiative of the UnitedNations that brings together more than 22 UN agencies andprogrammes, building on the strength of each to delivereffective multisectoral interventions that improve the lives ofpoor and marginalised citizens.

www.mdgfund.org For more information contact:Layla SaadAdvocacy Advisor, MDG Achievement [email protected]


1 United Nations Millennium Declaration, 8 September2000, UN General Assembly: para. 11

2 Hobcraft, J. (2007) Child Development, the Life Course,and Social Exclusion: Are the Frameworks Used in the UKRelevant for Developing Countries?, Chronic PovertyResearch Centre Working Paper 72: 8

3 Vandemoortele, J. (2009) ‘The MDG Conundrum:Meeting the Targets Without Missing the Point’,Development Policy Review 27.4: 355–71

4 United Nations Department of Social Affairs (2005)The Inequality Predicament. Report on the World SocialSituation, New York: United Nations DevelopmentProgramme; Son, H.H. and Kakwani, N. (2004)Economic Growth and Poverty Reduction: InitialConditions Matter, Working Paper No. 2, Brasilia:International Poverty Centre; Birdsall, N. and Londono,J.L. (1997) ‘Asset Inequality Matters: An Assessment ofthe World Bank’s Approach to Poverty Reduction’,American Economic Review 87.2: 32–37

5 Montalvo, R.G. and Reynal-Querol, M. (2005) ‘EthnicPolarization, Potential Conflict and Civil Wars’,American Economic Review 95.3: 796–816; Thorbecke,E. and Charumilind, C. (2002) ‘Economic Inequalityand its Socioeconomic Impact’, World Development30.9: 1477–95

6 UN Committee on Economic, Social and CulturalRights (2001) Prevention of armed conflict: report of theSecretary-General, A/55/985_S/2001/574, cited in UNDepartment of Economic and Social Affairs (2005) TheInequality Predicament. Report on the World SocialSituation 2005, New York: United Nations: 17

7 Or are averse to publishing data: Myanmar is a strikingexample. As is Rwanda, where the post-genocidegovernment has pursued a policy of promoting nationalunity and reconciliation by highlighting factors thatunify Rwandans and controlling factors that may be‘divisive’. This includes banning any reference toethnicity.

8 Wilkinson, R. and Pickett, K. (2010) The Spirit Level:Why Equality is Better for Everyone, London: PenguinBooks Ltd. For discussion in the OECD context also seeUN ECE (2010) Draft Report on the MDGs in Europeand Central Asia

9 Sen, A. (1992) Inequality Rexamined, Oxford: OxfordUniversity Press: xi

10 Stewart, F. (2002) Horizontal Inequalities: A NeglectedDimension of Development, Working Paper 81,University of Oxford, Queen Elizabeth House

11 While the concepts of social exclusion and horizontalinequalities have much in common, they describedifferent phenomena. Horizontal inequalities cut acrossrich as well as poor groups; thus, a multi-ethnic societymay have dominant and minority ethnic groups in richas well as poor strata. Social exclusion, on the otherhand, refers to the intersection of poverty with identity-based inequalities and the various disadvantages thataccompany them. Kabeer, N. (2000) ‘Social Exclusion,Poverty and Discrimination: Towards an AnalyticalFramework’, IDS Bulletin 31.4: 83–97

12 European Foundation for the Improvement of Livingand Working Conditions (1995) Public Welfare Servicesand Social Exclusion: The Development of Consumer

Oriented Initiatives in the European Union, Dublin:The Foundation, cited in A. De Haan and S. Maxwell(1998) ‘Poverty and Social Exclusion in North andSouth’, IDS Bulletin 29.1: 1–9

13 Beall, J. (2002) ‘Globalisation and Social Exclusion inCities: Framing the Debate with Lessons from Africaand Asia’, Environment and Urbanisation 14.1: 41–51

14 Busso, M.; Cicowiez, M. and Gasparini, L. (2005)‘Ethnicity and the Millennium Development Goals inLatin America’, Buenos Aires: Centro de EstudiosDistributivos, Laborales y Sociales (CEDLAS)

15 Busso et al. (2005)16 Busso et al. (2005)17 Patrinos, H. and Skoufias, E. (2007) Conference Edition:

Economic Opportunities for Indigenous Peoples in LatinAmerica, Washington D.C.: World Bank

18 Kabeer, N. (2009) ‘Poverty, Social Exclusion and theMDGs: The Challenge of “Durable Inequalities” in theAsian Context’, IDS Bulletin 37.3: 64–78

19 Bennett, L. (2005) ‘Gender, Caste and Ethnic Exclusionin Nepal: Following the Policy Process from Analysis toAction’, draft working paper produced for the WorldBank conference, New Frontiers of Social Policy:Development in a Globalizing World, World Bank

20 It must be noted that the World Health Organization(WHO) cautions about the reliability of maternalmortality data in most developing countries, in particularabout international comparisons over time, because ofdifferences in methods used to calculate this measure.

21 Easterly, W. (2007) How the Millennium DevelopmentGoals are Unfair to Africa, Global Economy andDevelopment Working Paper 14, Washington D.C.:Brookings Institution

22 Overseas Development Institute (ODI) (2010)Millennium Development Goals Report Card Learningfrom Progress, www.odi.org.uk/resources/download/4908.pdf, accessed 25 July 2010

23 IADB (2009) The Millennium Development Goals inLatin America and the Caribbean

24 Economic Commission for Latin America and theCaribbean (ECLAC) (2008) Panorama Social de AméricaLatina 2008, Santiago, Chile: United Nations, cited inThe Inter-American Dialogue (2009) How Poor andUnequal is Latin America and the Caribbean?, PolicyBrief No. 1, www.thedialogue.org/PublicationFiles/Social%20Policy%20Brief%20No%201%20-%20Poverty%20and%20Inequality%20in%20LAC.pdf,accessed 15 June 2010

25 López-Calva, L.F. and Lustig, N. (2010) DecliningInequality in Latin America: A Decade of Progress?,Baltimore, MD: Brookings Institution Press and theUnited Nations Development Programme

26 Ocampo, J. (2004) ‘Economic Development and SocialInclusion’, in M. Buvinic and J. Mazza (eds), SocialInclusion and Economic Development in Latin America,Inter-American Development Bank (IADB)

27 López-Calva, L.F. and Lustig, N. (2010)28 Montenegro, R.A. and Stephens, C. (2006) ‘Indigenous

Health in Latin America and the Caribbean’, The Lancet367: 1863

29 Layton, M.; Carrete, B.; Terrazas, I. and Rodríguez, A.(2010) ‘Reducing Maternal Mortality in Mexico:

Endnotes

62 Endnotes

Building Vertical Alliances for Change’, in J. Gaventaand R. McGee (eds), Citizen Action and National PolicyReform: Making Change Happen, London: Zed Books

30 Layton, M.; Carrete, B.; Terrazas, I. and Rodríguez, A.(2007) ‘Mexico Case Study: Civil Society and theStruggle to Reduce Maternal Mortality’, CitizenshipDRC Research Report, Brighton: IDS

31 United Nations Economic and Social Commission forAsia and the Pacific (ESCAP), Asian Development Bank(ADB) and United Nations Development Programme(UNDP) (2007) The Millennium Development Goals:Progress in Asia and the Pacific 2007

32 Ali, I. and Zhuang, J. (2007) Inclusive Growth Toward aProsperous Asia: Policy Implications, ERD Working Paper97: 19, Asian Development Bank (ADB)

33 National Planning Commission Secretariat, His Majesty’sGovernment of Nepal (2004) Nepal Living StandardsSurvey 2003/04, Kathmandu: Central Bureau of Statistics

34 ESCAP et al. (2007)35 Ravallion, M. (2009) A Comparative Perspective on

Poverty Reduction in Brazil, China and India, PolicyResearch Working Paper 5080, The World BankDevelopment Research Group

36 Koblinsky, M. (ed.) (2003) Reducing Maternal Mortality:Lessons from Bolivia, China, Egypt, Honduras, Indonesia,Jamaica, and Zimbabwe, Human DevelopmentNetwork, Health, Nutrition and Population Series,Washington D.C.: The World Bank. Liang, J.; Zhu, J.;Dai, L.; Li, X.; Li, M. and Wang, Y. (2010) ‘MaternalMortality in China, 1996–2005’, International Journal ofGynecology & Obstetrics, in press, corrected proof,available online 14 May

37 Du, Q.; Nass, O.; Bergsjo, P. and Nirmal Kumar, B.(2009) ‘Determinants for High Maternal Mortality inMultiethnic Populations in Western China’, Health Carefor Women International 30.11: 957–70

38 UNRISD (forthcoming) Flagship Report 2010:Combating Poverty and Inequality

39 Chen and Ravallion (2008)40 Milanovic, B. (2003) Is Inequality in Africa Really

Different? Policy Research Working Paper Series 3169,Washington D.C.: World Bank

41 Barrett, C.; Carter, M. and Little, P. (eds) (2006)‘Understanding and Reducing Persistent Poverty inAfrica’, Journal of Development Studies 42.2: 168

42 World Bank (2007) Global Monitoring Report 2007:Millennium Development Goals – Confronting theChallenges of Gender Equality and Fragile States,Washington D.C.: The International Bank forReconstruction and Development/The World Bank

43 You, D.; Wardlaw, T.; Salama, P. and Jones, G. (2010)‘Levels and Trends in Under-5 Mortality, 1990–2008’,The Lancet 375.9709: 100–3

44 You et al. (2010)45 Rajaratnam, J.; Marcus, J.; Flaxman, A.; Wang, H.;

Levin-Rector, A.; Dwyer, L.; Costa, M.; Lopez, A. andMurray, C. (2010) ‘Neonatal, Postneonatal, Childhood,and Under-5 Mortality for 187 Countries, 1970–2010: ASystematic Analysis of Progress Towards MillenniumDevelopment Goal 4’, The Lancet 375.9730: 1988–2008

46 UN Economic Commission for Africa (2010) AssessingProgress in Africa Towards the Millennium DevelopmentGoals 2010, E/ECA/COE/29/15 AU/CAMEF/EXP/15(V)

47 Brockerhoff, M. and Hewett, P. (2000) ‘Inequality of ChildMortality among Ethnic Groups in Sub-Saharan Africa’,Bulletin of the World Health Organization 78.1: 30–41

48 Wirth, M.; Balk, D.; Delamonica, E.; Storeygard, A.;Sacks, E. and Minujin, A. (forthcoming) ‘Setting theStage for Equity-Sensitive Monitoring of the Maternaland Child Health MDGs’, Bulletin of the World HealthOrganization, www.ciesin.columbia.edu/repository/

povmap/analysis/bulletin_who_preprint.pdf, accessed14 June 2010

49 Mancini, L. (2009) Comparative Trends in Ethno-Regional Inequalities in Ghana and Nigeria: Evidencefrom Demographic and Health Surveys, Oxford: CRISE,Working Paper 72: 18

50 Langer, A.; Mustapha, A.R. and Stewart, F. (2009)‘Diversity and Discord: Ethnicity, HorizontalInequalities and Conflict in Ghana and Nigeria’, Journalof International Development 21.4: 9; 477–82

51 Langer, A.; Mustapha, A. and Stewart, F. (2007)‘Horizontal Inequalities in Nigeria, Ghana and Côted’Ivoire: Issues and Policies’, CRISE Working Paper 45

52 Odhiambo, A. (2004) Pulling Apart: Facts and Figures onInequality In Kenya, Society for InternationalDevelopment. For more on educational inequalities inKenya see Alywy, A. and Schech, S. (2004) ‘EthnicInequalities in Education in Kenya’, InternationalEducation Journal 5.2

53 United Nations (2010) ‘Ethiopia: MDGs Country Report2010: Trends and Prospects for Meeting MDGs by 2015’,draft for discussion, UN Country Office Ethiopia

54 Bangura, Y. (2006) ‘Ethnic Inequalities in the PublicSector: A Comparative Analysis’, Development andChange 37.2

55 Langer et al. (2007) 56 Brockerhoff and Hewett (2000): 38 57 Wirth et al. (2010): 9 58 Leibbrandt, M.; Woolard, I.; Finn, A. and Argent, J.

(2010) Trends in South African Income Distribution andPoverty Since the Fall of Apartheid, OECD Social,Employment and Migration Working Paper 101, Paris:Organisation for Economic Co-operation andDevelopment (OECD)

59 UNESCO (2005) South Africa: Millennium DevelopmentGoals Country Report 2005,http://planipolis.iiep.unesco.org/upload/South%20Africa/South%20Africa%20MDG%20Report.pdf, accessed 29 June 2010

60 Leibbrandt et al. (2010)61 Lam, D.; Ardington, C. and Leibbrandt, M. (2006)

‘Progress Through School in Urban South Africa:Evidence from Panel Data’, paper prepared forsubmission to the DPRU/TIPS conference, Acceleratedand Shared Growth in South Africa: Determinants,Constraints and Opportunities

62 UNESCO (2007) South Africa Millennium DevelopmentGoals Mid-Term Country Report, 2007,http://planipolis.iiep.unesco.org/upload/South%20Africa/South_Africa_MDG_midterm.pdf, accessed 29June 2010

63 Burgard, S. (2005) ‘Race and Pregnancy-related Care inBrazil and South Africa’, Social Science and Medicine 59:1127–46

64 UNESCO (2007) 65 Hoff, K. and Pandey, P. (2004) Belief Systems and

Durable Inequities: An Experimental Investigation ofIndian Caste, World Bank Policy Research WorkingPaper 3351, Washington D.C.: World Bank

66 Behrman, J.; Gaviria, A. and Szekely, M. (2003) Who’s Inand Who’s Out? Social Exclusion in Latin America, NewYork: IADB

67 Harker, N.; Kader, R.; Myers, B.; Fakier, N.; Parry C.;Flisher, A.; Peltzer, K.; Ramlagan, S. and Davids, A.(2008) ‘Substance Abuse Trends in the Western Cape: AReview of Studies Conducted Since 2000’,commissioned by the Department of SocialDevelopment, Cape Town

68 Department of Health, Medical Research Council,OrcMacro. (2007) South Africa Demographic and HealthSurvey 2003, Pretoria: Department of Health


69 Louw, A.; Shaw, M.; Camerer, L. and Robertshaw, R.(1998) Crime in Johannesburg: Results of a City VictimStudy, Institute for Security Studies, ISS MonographSeries 18

70 Kanbur, R. and Lustig, N. (1999) ‘Why is InequalityBack on the Agenda?’, paper prepared for the AnnualWorld Bank Conference on Development Economics,World Bank, Washington D.C., 28–30 April. See alsoBangura, Y. (2006) and Stewart, F. (2010) HorizontalInequalities and Conflict: Understanding Group Violencein Multiethnic Societies, Basingstoke: Macmillan

71 Patrinos, H. and Skoufias, E. (2007) EconomicOpportunities for Indigenous Peoples in Latin America,Washington D.C.: The International Bank forReconstruction and Development/The World Bank

72 Tomei, M. (2005) Indigenous and Tribal Peoples: AnEthnic Audit of Selected Poverty Reduction StrategyPapers, Geneva: International Labour Organization(ILO)

73 Vandergeest, P. (2003) ‘Racialization and Citizenship inThai Forest Politics’, Society & Natural Resources 16:19–37

74 Mazza, J. (2004) ‘Social Inclusion, Labor Markets andHuman Capital in Latin America’, in M. Buvinic and J.Mazza (eds), Social Inclusion and Economic Developmentin Latin America, Inter-American Development Bank(IADB): 179–200

75 Gang, I.N.; Sen, K. and Yun, M.S. (2002) Caste,Ethnicity and Poverty in Rural India, IZA DiscussionPaper 629, Bonn: Institute for the Study of Labor

76 Kijima, Y. (2006) ‘Caste and Tribe Inequality: Evidencefrom India, 1983–1999’, Economic Development andCultural Change 54.2: 369–404

77 Deshingkar, P. (2006) ‘Internal Migration, Poverty andDevelopment in Asia: Including the Excluded ThroughPartnerships and Improved Governance’, backgroundpaper prepared for Asia 2015 Conference, 2006

78 Heintz, J. (2008) Employment, Informality and Poverty:An Empirical Overview of Six Countries with a Focus onGender and Race, draft paper, UNRISD

79 Thorat, S. and Lee, J. (2005) ‘Caste Discrimination andFood Security Programmes’, Economic and PoliticalWeekly 40.39

80 Patrinos and Skoufias (2007)81 Patrinos and Skoufias (2007)82 Betancourt, R. and Gleason, S. (2000) ‘The Allocation of

Publicly-Provided Goods to Rural Households in India:On Some Consequences of Caste, Religion andDemocracy’, World Development 28.12: 2169–82

83 National Population Commission (NPC) [Nigeria] andICF Macro (2009) Nigeria Demographic and HealthSurvey 2008, Abuja, Nigeria: National PopulationCommission and ICF Macro

84 Langer et al. (2007)85 Ghana Statistical Service (GSS), Ghana Health Service

(GHS), and ICF Macro (2009) Ghana Demographic andHealth Survey 2008, Accra, Ghana: GSS, GHS, and ICFMacro

86 Duong, D.V.; Binns, C.W. and Lee, A.H. (2004)‘Utilization of Delivery Services at the Primary HealthCare Level in Rural Vietnam’, Social Science & Medicine59.12: 2585–95

87 WHO (2008) The World Health Report 2008: PrimaryHealth Care (Now More Than Ever), New York: WorldHealth Organization

88 Spangler, S.A. and Bloom, S.S. (2010) ‘Use ofBiomedical Obstetric Care in Rural Tanzania: The Roleof Social and Material Inequalities’, Social Science &Medicine, in press, corrected proof: 1–9; 7

89 Matthews, Z. (2002) ‘Maternal Mortality and Poverty’,Centre for Reproductive Health, DFID Resource Centrefor Sexual and Reproductive Health, www.jsieurope.org/

docs/maternal_mortality_and_poverty.pdf, accessed10 June 2010

90 Center for Human Rights and Global Justice/HumanRights Watch (HRW) (2007) ‘Caste DiscriminationAgainst Dalits or So-called Untouchables in India’,paper presented at the Seventieth Session of theCommittee on the Elimination of RacialDiscrimination

91 Parker, S.; Rubalcava, L. and Teruel, G. (2003) ‘LanguageBarriers and Schooling Inequality of the Indigenous inMexico’, in J. Behrman et al. (2003): 145–178

92 Bangura, Y. (2006): 299–32893 Stewart (2000)94 Stewart, F. (2002) Horizontal Inequalities: A Neglected

Dimension of Development, Working Paper 81, Oxford:Queen Elizabeth House

95 Lawoti, M. (2005) Towards a Democratic Nepal: InclusivePolitical Institutions for a Multicultural Society, NewDelhi: Sage

96 Stewart (2010)97 Stewart (2000)98 Barrientos, A. and Shepherd, A. (2003) ‘Chronic Poverty

and Social Protection’, paper prepared for the CPRCConference on Chronic Poverty

99 Christian Aid (forthcoming, September 2010), draft ofunpublished, currently untitled report, on the subject ofending poverty

100 Cornia (2010)101 Christian Aid (forthcoming)102 Heller, P. (2001) ‘Moving the State: The Politics of

Democratic Decentralisation in Kerala, South Africa,and Porto Alegre’, Politics & Society 29.1: 131–63

103 Beall, J. (2009)104 World Bank (2006) World Development Report 2006:

Equity and Development, Washington D.C.: World Bankand Oxford University Press

105 Estivill, J. (2003) Concepts and Strategies for CombatingSocial Exclusion: An Overview, Geneva: InternationalLabour Office

106 Vandermootele, J. (2009) ‘Taking the MDGs Beyond2015: Hasten Slowly’, paper commissioned for theHigh-Level Policy Forum, After 2015: Promoting Pro-Poor Policy after the MDGs, organised byDSA/EADI/Action Aid, Brussels, June 2009

107 UNESCO (2010) Reaching the Marginalised,http://unesdoc.unesco.org/images/0018/001866/186606e.pdf, accessed 4 August 2010

108 Horton, R. (2010) ‘Maternal Mortality: Surprise, Hope,and Urgent Action’, The Lancet 375.9726: 1581–2

109 Jensen, R. and Oster, E. (2007) The Power of TV: CableTelevision and Women’s Status in India, NBER WorkingPaper 13305, Cambridge, MA: National Bureau ofEconomic Research. Also, Kabeer, N.; Mahmud, S. andTasneem, S. (forthcoming) Paid Work and Pathways ofWomen’s Empowerment: Preliminary Findings from aSurvey in Rural Bangladesh, Bangladesh: BangladeshDevelopment Institute

110 www.who.int/inf-new/aids1.htm, accessed 4 August 2010111 Baviskar, A. (2008) Winning the Right to Information

Campaign in India, IDS Research Summary, Brighton:IDS

112 UN (2010) Keeping the Promise: A Forward-lookingReview to Promote an Agreed Action Agenda to Achievethe MDGs by 2015, report of the Secretary Generalwww.un.org/millenniumgoals/pdf/sgreport_draft.pdf,accessed 4 August 2010

113 Cornia, G.A. (2010) ‘Income Distribution Under LatinAmerica’s New Left Regimes’, Journal of HumanDevelopment and Capabilities 11.1: 85–114

114 Ravallion (2009) 115 Cornia (2010)116 Stewart and Venugopal (2009)

64 Endnotes

117 UNESCO (2010)118 Souza, C. (2001) ‘Participatory Budgeting in Brazilian

Cities: Limits and Possibilities in Building DemocraticInstitutions’, Environment and Urbanization 13.1: 159–184

119 Goldsmith, W. and Vainer C. (2001) ‘ParticipatoryBudgeting and Power Politics in Porto Alegre’, LandLines 13.1

120 ODI (2006) ‘Affirmative Action’, India Inter-regionalFacility Brief 14, London: Overseas DevelopmentInstitute

121 Jomo, K.S. (2006) Growth with Equity in East Asia?,Department of Economic and Social Affairs WorkingPaper 33, ST/ESA/2006/DWP/33; Kreuzer, P. (2006)Democracy, Diversity and Conflict: Managing EthnicDivisions in the Philippines and Malaysia, CornellUniversity Peace Studies Program, Occasional Paper 30-4

122 Stewart (2002) 123 Inter-American Development Bank (2007) ‘Mujeres

Indígenas en la Política: Superando Obstáculos yCreando Oportunidades’ in Ideas en Marcha: Igualdadde Género en el Desarollo, Septiembre, Número 3

124 Marquez, G.; Chong, A.; Duryea, S.; Mazza J. and Ñopo,H. (eds) (2007) Outsiders? The Changing Patterns ofExclusion in Latin America and the Caribbean, 2008report, Washington D.C.: Inter-American DevelopmentBank; Telles, Edward E. (2004) Race in Another America:The Significance of Skin Color in Brazil, Princeton, NJ:Princeton University Press

125 Marquez et al. 2007; Buvinic and Roza (2004)126 World Bank, World Development Report 2006: Equity

and Development127 Deere, C. and León, M. (2001) Empowering Women:

Land and Property Rights in Latin America, Pittsburgh:University of Pittsburgh Press

128 Telles (2004) 129 Marquez et al. (2007) 130 Buvinic and Mazza (2004) 131 Ibarran, P. and Shady, D.R. (2008) Evaluating the Impact

of Job-training Programs in Latin America: Evidence fromIDB-Funded Operations, Washington D.C.: Inter-American Development Bank

132 World Bank/FAO/IFAD (2009) Gender in AgricultureSource Book

133 Van de Walle, D. and Gunewardena, D. (2001) ‘Sourcesof Ethnic Inequality in Viet Nam’, Journal ofDevelopment Economics, 65: 177–207

134 EDA Rural Systems (2004) ‘The Maturing of IndianMicrofinance. A Longitudinal Study’, report preparedfor SIDBI Foundation for Micro-credit, New Delhi

135 Kabeer, N. (2005) ‘Chapter 4: Direct Social Impacts forthe Millennium Development Goals’ and ‘Chapter 5:Wider Impacts: Social Exclusion and Citizenship’, in J.Copestake et al. (eds) Money with a Mission, Vol. 1,Bourton-on-Dunsmore: ITDG Publishing

136 Kabeer, N.S. Mahmud and Castro, J.G.I. (2010) NGOs’Strategies and the Challenge of Development andDemocracy in Bangladesh, IDS Working Paper 343,Brighton: Institute of Development Studies

137 www.mekongeconomics.com/Document/Publications/2006/MKE%20Northern%20Mountains%20Baseline%20Survey.pdf

138 Liang et al. (2010)139 Koblinsky (2003)140 Yates, R. (2007) The Impact of Abolishing User Fees in

Africa: Recent Developments in Six African Countries,London: UK Department for International Development

141 World Health Organization (2008) Primary HealthCare: Now More Than Ever, Geneva: WHO: 27

142 UNICEF Pakistan (2007) ‘Mobility Support for FemaleTeachers in NWFP (Pakistan)’, www.unicef.org/rosa;Afghanistan Girls’ Education Partnership (2007),‘Situation of Girls’ Education in Afghanistan’, Kabul:Ministry of Education

143 Thorat, S. and Lee, J. (2005) ‘Caste Discrimination andFood Security Programmes’, Economic and PoliticalWeekly 40.39

144 Barrientos, A. and Hulme, D. (2009) ‘Social Protectionfor the Poor and Poorest in Developing Countries:Reflections on a Quiet Revolution’, Oxford DevelopmentStudies 37.4: 439–56

145 As well as its relevance to rights-based approaches topoverty reduction. See Devereux, S. and Sabates-Wheeler, R. (2004) Transformative Social Protection, IDSWorking Paper 232, Brighton: IDS; Cook, S. andKabeer, N. (2010) Social Protection as DevelopmentPolicy; Asian Perspectives, New Delhi: Routledge;Barrientos and Hulme (2009). For discussion ofevidence for these arguments see Kabeer, N. (2010)Gender and Social Protection Strategies for the InformalEconomy, New Delhi: Routledge; also Kabeer, N. (2009)Scoping Study on Social Protection: Evidence on Impactsand Future Research Directions (downloadable fromwww.research4development.info/PDF/Outputs/Misc_HumanSec/Social_protection_scoping_study_NK_09Final.pdf

146 Carvano, L. and Paixão, M. (eds) (2008) Relatório Anualdas Desigualdades Raciais no Brasil: 2007–2008, Rio deJaneiro: Garamond

147 Kabeer, N. (2010) Gender and Social Protection Strategiesin the Informal Economy, New Delhi: Routledge

148 Devereux, S. et al. (2007) An Evaluation of ConcernWorldwide’s Dowa Emergency Cash Transfer Project(DECT) in Malawi, 2006/07, www.concern.net/documents

149 Suarez, M. et al. (2006) ‘The Bolsa Família Programmeand the Tackling of Gender Inequalities’, report to theBrazilian Ministry of Social Development and FightAgainst Hunger (MDS) and DFID, Brasilia: 5

150 Suarez et al. (2006): 58151 Attanasio, O.; Pellarano, L. and Polania, S. (2008)

‘Building Trust: Conditional Cash Transfers and SocialCapital’, EWP08/02

152 Gacitúa-Marió, E.; Norton A. and Georgieva, S. (2009)Building Equality and Opportunity Through SocialGuarantees: New Approaches to Public Policy and theRealization of Rights, Washington D.C.: TheInternational Bank for Reconstruction andDevelopment/The World Bank

153 Kabeer, N. (2005) ‘The Search for Inclusive Citizenship:Meanings and Expressions in an Interconnected World’,in N. Kabeer (ed.) Inclusive Citizenship: Meanings andExpressions, London: Zed Press

154 Mackintosh, M. and Tibandebage, P. (2002) ‘Inclusionby Design? Rethinking Health Care Market Regulationin the Tanzanian Context’, Journal of DevelopmentStudies 39.1: 1–20

155 Beall, J. (1997) ‘Introduction’, in J. Beall (ed.), A City forAll: Valuing Difference and Working with Diversity,London: Zed Books

156 Manning, R. (2009) Using Indicators to EncourageDevelopment: Lessons from the MDGs, DIIS Report2009:01, Copenhagen: Danish Institute for InternationalStudies

157 Saith, A. (2007) ‘From Universal Values to MDGs: Lost inTranslation’, Development and Change 37.6: 1167–1199:1186

Institute of Development Studiesat the University of SussexBrighton BN1 9RE UKwww.ids.ac.uk

MDG Achievement Fund SecretariatOne United Nations Plaza, DC-1-2160

New York, NY USA 10017www.mdgfund.org

can the mdgs provide a pathway to social justice? the challenge of intersecting inequalities

Business

development report

social justice

social commission

social exclusion persists304

social exclusion132

economic dynamics of

new social contract

escap united nations