Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized

WORLD

-7-7
~
ATTACKING

7
REPORT
DEVELOPMENT

22684
2

September 2000
POVERTY

I
WORLD DDEVELOPMENT
REPORT /2ool

ATTACKING POVERTY
Cover Art

Voyageto the New World, 1999. Mixed media on wood, by Manuel Cunjama,
Mexico. The images used to open the chapters are also taken from this painting and
from three others by Cunjami: Dialogue with the Universe,Magic Kite, and Looking
for the CosmicBalance.

Manuel Cunjama was born in 1971 in Tuxtla Gutierrez, Chiapas. He began exhibiting
his work in Chiapas in 1992. Currently, Ana Quijano Gallery in Mexico City is repre-
senting Cunjamr in traveling exhibits in the United States. Cunjama's work is included
in the collection of the World Bank art program.

In my work I seek and try to incorporatenative symbolsand elementsthat would take me

toprehispanic issuesand all that I considerusefluto this effect I show it in my work.

I use the sun, the moon, the night, and the universeas symbolsrepresentingthe worryand
anxiety of the human beingeforthe wholenesssurrounding him: The infinite.
-Manuel Cunjama

The World Bank Art Program makes particular efforts to identify artists from develop-
ing nations and make their work available to a wider audience. The art program orga-
nizes exhibits, educational and cultural partnerships, competitions, artists' projects, and
site-specific installations.
WORLD DEVELOPMENT
REPORT 2 a/2001
a _~~~~~~~~~~~~~~~~~~~~~E

ATTACKING POVERTY

Published for the WORLD BANK

OXFORD UNIVERSITY PRESS

Oxford UniversityPress

OXFORD NEX' YORK ATHENS AUCKLAND BANGKOK

BOGOTA BUENOS AIRES CALCUTTA CAPETOWN CHENNAI
DAR ES SALA,AM DELHI FLORENCE HONG KONG ISTANBUL
KARACHI KUALALUMPUR MADRID MELBOURNE MEXICO
CITY MUMBAI NAIROBI PARIS SAO PAULO SINGAPORE
TAIPEI TOKYO TORONTO WVARSAW
and associatedcompanies in
BERLIN IBADAN

©D2001 The InternationalBank for Reconstructionand

Development/ The World Bank
1818 H Street, N.W., Washington, D.C. 20433, U.S.A.

Cover designand chapter openersby Tomoko Hirata.

Publishedby Oxford UniversityPress,Inc.

200 Madison Avenue,New York,N.Y. 10016

Oxford is a registeredtrademark of Oxford UniversityPress.

All rights reserved.No part of this publication may be

reproduced,stored in a retrievalsystem,or transmitted, in any
form or by any means, electronic,mechanical,photocopying,
recording,or otherwise,without the prior permission of Oxford
UniversityPress.

Manufactured in the United Statesof America

First printing September2000

This volume is a product of the staffof the World Bank, and the
judgments made herein do not necessarilyreflectthe viewsof its
Boardof ExecutiveDirectors or the countries they represent.
The World Bank does not guaranteethe accuracyof the data
included in this publicationand accepts no responsibility
whatsoeverfor any consequenceof their use. The boundaries,
colors, denominations,and other information shown on any
map in this volume do not imply on the part of the World Bank
any judgment on the legalstatus of any territory or the
endorsementor acceptanceof such boundaries.

ISBN 0- 19-521598-2 clothbound

ISBN 0-19-521129-4 paperback
ISSN 0163-5085

Text printed on recycledpaper that conformsto the American

Standard for Permanenceof Paper for Printed Library Material
Z39.48-1984.

IlPoverty
amid plenty is the world'sgreatestchal-
lenge.We at the Bankhavemadeit our missionto fight
povertywith passionand professionalism, puttingit at
the centerof all the work we do. And we haverecog-
nizedthat successfuldevelopmentrequiresa compre-
hensive,multifaceted, and properlyintegratedmandate.
This report seeksto expandthe understandingof
povertyand its causesand sets out actionsto createa
worldfreeof povertyin allits dimensions.It both builds
on our past thinking and strategyand substantially
broadensand deepenswhat wejudge to be necessary
to meetthe challengeof reducingpoverty.It arguesthat
major reductionsin human deprivationare indeed
possible,and that the forcesof globalintegrationand
technologicaladvancecan and must be harnessedto
servethe interestsof poor people.Whetherthisoccurs
willdependon howmarkets,institutions,and societies
function-and on the choicesfor publicaction, glob-
ally,nationally,and locally.
The report accepts the now establishedview of
povertyasencompassing not onlylowincomeand con-
sumption but also low achievement in education,
health, nutrition, and other areasof human develop-
ment. And basedon what peoplesay povertymeans
Foreword
to them, it expandsthis definitionto includepower-
lessnessand voicelessness, and vulnerabilityand fear.
These dimensions of human deprivation emerged
forcefullyfrom our Voices ofthe Poorstudy,conducted
as backgroundfor the report, which systematically
soughtthe viewsof morethan 60,000menandwomen
livingin povertyin 60 countries.
These differentdimensionsof povertyinteract in
important ways.So do interventionsto improvethe
well-beingof poor people.Increasingeducationleads
to betterhealthoutcomes.Improvinghealthincreases
income-earningpotential.Providingsafetynetsallows
poor peopleto engagein higher-risk,higher-returnac-
tivities.Andeliminatingdiscrimination againstwomen,
ethnic minorities, and other disadvantagedgroups
both directlyimprovestheir well-beingand enhances
their abilityto increasetheir incomes.
The 20th century saw great progressin reducing
povertyand improvingwell-being.In the past four
decadeslife expectancyin the developingworld in-
creased20 yearson average,the infantmortalityratefell
morethanhalf,and fertilityratesdeclinedbyalmosthalf
In the pasttwo decadesnet primaryschoolenrollment
in developing countriesincreasedby 13percent.Between
v
VI WORLD DEVELOPMENT REPORT 2000/2001
1965and 1998averageincomesmorethandoubledin de-
velopingcountries,and in 1990-98 alonethe numberof
peoplein extremepovertyfellby 78 million.
But at the start of a new century,poverty remainsa
globalproblemof hugeproportions.Of the world's6 bil-
lion people,2.8 billionliveon lessthan $2 a day,and 1.2
billionon lessthan $1 a day. Sixinfantsof every 100 do
not seetheir firstbirthday,and 8 do not surviveto their
fifth. Of thosewho do reachschoolage,9 boysin 100,
and 14 girls,do not go to primaryschool.
These broadtrendsconcealextraordinarydiversityin
experiencein differentparts of the world-and large
variationsamongregions,with someseeingadvances,and
otherssetbacks,in crucialnonincomemeasuresof poverty.
Wideningglobaldisparitieshaveincreasedthe senseof
deprivationand injusticefor many.And socialmobility
and equalopportunity remainalienconceptsforfar too
many people.
Future demographicchangeswill add to the chal-
lengewefacein further reducingpoverty.In the next25
yearsroughly2 billionpeoplewillbe addedto the world's
population-almost allof them (anestimated97 percent)
in developingcountries, putting tremendouspressure
on thesesocieties.Clearly,much must be done to reduce
povertyin its multipledimensionsand to promotehuman
freedom,today and in the yearsahead.
While current and future challengesremain daunt-
ing, we enter the new millenniumwith a better under-
standingof development. We havelearnedthat traditional
elementsof strategiesto fostergrowth-macroeconomic
stabilityand market-friendlyreforms-are essentialfor
reducing poverty.But we now alsorecognizethe need
for much more emphasison layingthe institutionaland
socialfoundationsfor the developmentprocessand on
managingvulnerabilityand encouragingparticipationto
ensure inclusivegrowth.And while domesticaction is
critical,we havealso learnedthat global developments
exerta potent influenceon processesof changeat national
and local levels-and that global action is central to
povertyreduction.We havetakena freshlookat our work
through the ComprehensiveDevelopmentFramework,
whichconvergeswith the viewsand findingsof this re-
port.
Basedon its analysisof ideasand experience,this re-
port recommendsactionsin threeareas:
Promotingopportunity.Expandingeconomicoppor-
tunity for poor peopleby stimulatingoverallgrowth
and by buildingup their assets(suchas land and ed-
ucation) and increasingthe returns on these assets,
through a combination of market and nonmarket
actions.
X Facilitatingempowerment: Makingstate institutions
more accountable and responsiveto poor people,
strengtheningthe participationof poor peoplein po-
litical processesand local decisionmaking,and re-
movingthe socialbarriersthat resultfromdistinctions
of gender,ethnicity,race,religion,and socialstatus.
Enhancingsecurity: Reducingpoorpeople'svulnerability
to ill health, economicshocks,crop failure,policy-
induceddislocations,natural disasters,and violence,
as well as helping them cope with adverse shocks
whentheyoccur.A bigpart of thisis ensuringthat ef-
fectivesafetynetsare in place to mitigatethe impact
of personaland nationalcalamities.
Advances in the three areas are fundamentally
complementary-each is important in its own rightand
each enhancesthe others. Drawingon this framework,
countriesneed to developtheir own poverty reduction
strategies,in a manner consistentwith preservationof
culture.Decisionson prioritiesmust be madeat the na-
tional level, reflecting national priorities. But action
must alsotakeplacewith localleadershipand ownership,
reflectinglocal realities.There is no simple, universal
blueprint.
Action at the local and nationallevelsis not enough,
however.The evidenceof the past decadevividlyreveals
the importanceof globalaction, both to ensurethat the
opportunitiesfrom globalintegrationand technological
advancebenefitpoor peopleand to managethe risksof
insecurity and exclusionthat may result from global
change.Fiveactionsare key:
Promotingglobalfinancialstabilityand openingthe
marketsof rich countriesto the agriculturalgoods,
manufactures,and servicesof poor countries.
Bridging the digital and knowledge divides, thus
bringing technology and information to people
throughout the world.
t Providingfinancialand nonfinancialresourcesforin-
ternationalpublicgoods,especiallymedicaland agri-
cultural research.
v Increasingaid and debt reliefto help countriestake
actionsto endpoverty,withina comprehensive frame-
work that puts countries themselves-not external
aid agencies-at the center of the designof develop-
ment strategyand ensuresthat externalresourcesare
used effectivelyto support the reductionof poverty.
 FOREWORD VI:[

* Givinga voiceto poor countriesand poor peoplein viewedin thisreportshowthat thereisnowsufficientun-

globalforums,includingthrough internationallinks
with organizationsof poor people.
Public action must be driven by a commitment to
povertyreduction.Thepublicandprivatesectorsmustwork
together-along with civilsociety-both withinand be-
tweencountries.Whilewehavemuch to learn,andwhile
the worldcontinuesto changerapidly,the experiencesre-
derstanding to makeactionsto reducepovertytrulyeffective.
Weare livingin a time in whichthe effortsand issuessur-
roundingpovertyreductionaresubjectto greatscrutiny;
In the aftermathof protestsand in the midst of contro-
versy,this reportoffersrealsubstanceto the publicdebate
and bringsthe dialogueto the foreground,whereindeed
the goalof a world withoutpovertybelongs.

JamesD. Wolfensohn
President
The World Bank
August2000

Thisreporthasbeenpreparedbya teamledbyRaviKanbur(director oftheteamuntilMay2000)andNoraLustig(deputy di-

rectoruntilMay2000anddirector since).MonicaDasGupta,Christiaan Grootaert,
VictoriaKwakwa, Christina
Malmberg Calvo,
andKevnMorrison servedasfull-timeteammembers. Othercoreteammembers includedAliceSindzingre,
Michael
Woolcock,
andZainalYusofMajorcontributions tochaptersweremadebyHomiKharas, AartKraay,PeterLanjouw,andGiovanna Pren-
nushiaswellasbyBenuBidani,WilliamEasterly, EnriqueFlores,Hlene Grandvoinnet, RichardNewfarmer, Gi-TaikOh,and
MattiaRomani.Michael Walton,in hiscapacityasdirector,PovertyReduction, workedcloselywiththeteamthroughout the
process.ShankaChakraborty andShahinYaqubassisted theteam.InternsfromtheWashington Centerprovidedothervaluable
assistance.
Theworkwascarriedour underthe generaldirectionofJozefRitzenandJosephE. Stiglitzand,in thefinalstages,
Nicholas Stern.TheWorldBankDevelopment DataGroupwasresponsible fortheSelected
WorldDevelopment Indicators.
BruceRoss-Larson andMetadeCoquereaumont weretheprincipal editorsof thereport.
The teamwasadvisedbyAnthonyAtkinson,AnthonyBebbington,NancyBirdsall,FrancoisBourguignon, Angus
Deaton,Alaindejanvry,YujiroHayami,EmmanuelJimenez,GrzegorzKolodko,MichaelLipton,LantPritchett,Mar-
tin Ravallion,AmarryaSen,LynSquire,T. N. Srinivasan,andMarianoTommasi.DeepaNarayanled the Voices ofthe
Poorstudy.VinodThomasled the Qualityof Growthstudy,whichcomplements the workof thisreport.ArielFiszbein
ledthe organization of the consultationprocess.Muchinsightwasgainedfromthe papersand discussions at the Sum-
merWorkshopheldin Washington,D.C., on 6-1 0 July 1999,aswellasfromthe backgroundpaperspreparedforthis
report.The authors,commentators, and participants,as wellas the backgroundpapers,arelistedin the bibliographic
note.Manyothersinsideandoutsidethe WorldBankprovidedhelpfulcommentsandothercontributions; theirnames
arealsolistedin the bibliographic note.
Awiderangeofconsultations withacademics, grassrootsleaders,NGOs,privatesectorrepresentatives,
andpolicymakers
wereundertakenforthisreport,frominitialoutlineto finaldraft.Thanksgoto all thepublicinstitutionsandcivilsociety
organizationsthathelpedmaketheseconsultations possible,aswellasthe organizersandparticipants,
whosecontributions
greatlyenrichedthisreport.Particular thanksgoto the BrettonWoodsProjectandtheNewPolicyInstitutein theUnited
Kingdom,whichmoderatedthe electronicdiscussion on theWebdraft,andto allwhosubmittedcomments:424 contri-
butionswerepostedfrom44countries(44percentfromdeveloping countries).Consultations
wereheldin Argentina(Buenos
Aires),Bangladesh (Dhaka),Canada(Ottawa),Chile(Santiago), Denmark(Copenhagen), Egypt(Cairo),Ethiopia(Addis
Ababa),France(Paris),Germany(Berlin),Guatemala(Antigua), Hungary(Budapest), India(Ahmedabad, NewDelhi),
Japan(Tokyo),Malaysia(KualaLumpur),Morocco(Marrakech), Russia(Moscow), Senegal(Dakar),SouthAfricaJo-
hannesburg), Sweden(Stockholm), theUnitedKingdom(Brighton, London),the UnitedStates(Boston,NewYork,Wash-
Vill WORLD DEVELOPMENT REPORT 2000/2001

ington, D.C.), and Vietnam(Hanoi). Consultationswerealsoheld at and in conjlunctionwith the InternationalMonetary

Fund, the regionaldevelopmentbanks,and severalUnited Nations organizations.
The preparationof backgroundpapersand the conveningof severalworkshopswere supportedby the governmentsof
Canada, Denmark,France,Germany,Japan, the Netherlands,Sweden,Switzerland,the United Kingdom,and the United
States; the MacArthurFoundation;Cornell University;and the DevelopmentPolicy Forum of the German Foundation
for International Development.The Voicesof the Poorstudy was supported by the MacArthur Foundation, Cornell
University,the U.K. Departmentfor InternationalDevelopment,and the SwedishInternationalDevelopmentCoopera-
tion Agency;and severalof the country studieswere partiallyor fullyfinancedby the NGOs that conductedthe research.
RebeccaSuguiservedas executiveassistantto the team, and Maribel Flewitt,Shannon Hendrickson,Khin-U Khine,
RudeewanLaohakittikul,JimenaLuna, Nelly Obias,Gracie Ochieng,LeilaSearch,and RobertSimmsas team assistants.
Maria D. Amealservedas administrativeofficer.Nacer Megherbiand Edith Thomas providedtechnicalsupport.
Book copyeditingand production were by FionaBlackshaw,Garrett Cruce, Terry Fischer,Wendy Guyette,Daphne
Levitas,Molly Lohman,Megan Klose,JessicaSaval,and Alison Strong of CommunicationsDevelopment.Book design,
editing,and production weremanagedbyjamila Abdelghani,CatherineHudson, BrettKravitz,Nancy Lammers,Brenda
Mejia,Randi Park and Betty Sun of the of the World Bank'sOffice of the Publisher.ArtemisZenetou, managerand cu-
rator of the World BankArt Program,and her staffwere instrumentalin the useof the artworkin the book. Tomoko Hi-
rata designedthe coverand the chapter openers.
RaviKanburis T. H. LeeProfessorof WorldAffairsin the DepartmentofAgricultural,Resource,and Managerial
Economics,
CornellUniversity,Ithaca,NewYork.Nora Lustigis senior advisorand chiefof the Povertyand InequalityAdvisoryUnit
of the Inter-AmericanDevelopmentBank. KevinMorrisonis a fellowat the OverseasDevelopmentCouncil.AliceSitidz-
ingre is a researcherat the Centre Nationalde la RechercheScientifique(CNRS)in Parisand an associateresearcherat the
Centre d'Etude d'AfriqueNoire (CEAN)in Bordeaux.ZainalYusofis deputy directorgeneralof the Institute of Strategic
and InternationalStudies(ISIS)in KualaLumpur, Malaysia.The World Bank is gratefulto the six institutionsfor mak-
ing possibletheir participationin the preparationof the report.
The manycontributorsto the report may not endorseeveryviewpointor assertionin it, and they bear no responsibil-
ity for any errorsthat remain.
 Contents
Overview Attacking Poverty:Opportunity, Empowverment, arid Secuirity I
Basrabai's story 2
Poverty in an unequal world 3
A strategy for poverty reduction 6
From strategy to action 7
Working together to fight poverty 12

Part I Framework
Chapter1 The Natureand Evolutionof Poverty 15
Poverty'smany dimensions 15
Measuring poverty in its multiple dimensions 16
The evolution of poverty 21

Chapter 2 Causesof Poverty and a Framework for Action 31

The causes of poverty 34
A framework for action 37

Part 11Opportunity
Chapter 3 Growth, Inequality, and Poverty 4
Economic growth and poverty reduction 46
What drives economic growth? 49
Why are similar rates of growth associated with different rates of poverty reduction? 52
Economic growth and nonincome poverty 57

Chapter 4 Making Markets Work Better for Poor Pople 6o

Have market reforms delivered growth? 62
Have market reforms delivered benefits to poor people? 64
Making markets do more for poor people 72

Chapter5 ExpandingPoor People'sAssets arideTackSiq Inequaboes 77

Assets and their synergies 77
Public action to facilitate the accumulation of assets 78
Redistributing public spending 80
Institutional reforms for effectivedelivery: governance, markets, and competition 85
Participation: choice, monitoring, and accountability 88
Complementarities in public action 93
lx
X WORLD DEVELOPMENT REPORT 2000/2001

Part 1II Empowerment

Chapter 6 Making State institutions More Responsiv/eto Poor People 99
Pubtic administration and poverty reduction 99
Poor people and the rule of law 102
How can decentralization be made pro-poor? 106
The politics of poverty reduction: pro-poor coalitions 108
Political regimes and poverty 112

Chapter 7 Removing Social Barriers and BuiiId'ng Sci_ai Institutions 117

Gender discrimination and poverty 117
Social stratification and poverty 123
Social fragmentation and conflict 126
Building social institutions and social capital 128

Part IV Security
Chapter 8 Helpirg Poor PeopIe Manage Risk 135
A typology of risks 136
The nature and magnitude of vulnerability 138
Responses to risk by households and communities 140
Policy responses for improving risk management 146

C'haptexi 9 Managing Eco,nomic Crises anda?Jaturai Disstefs 161

Preventing and coping with economic crises 161
Reducing vulnerability to natural disasters 170

Part V InternationalActions
' if',aessng
1 Global Forces for Poor PeoplF 179
Expanding market access in high-income countries 179
Reducing the risk of economic crises 180
Producing pro-poor international public goods 181
Ensuring a voice for poor people in global forums 185

Chapler 1RIeforrwng i
Development Cooperation to Attack Poverty 189
Making aid more effective in reducing poverty 192
Relieving the debt burden of poor countries 200

Bibfinqraphic Note 205

Se-ecleri VVoridDevelopment Indicators 267

5oxe$
I The voices of the poor 3
2 A better world for all: international development goals 5
1.1 Poverty in the voices of poor people 16
1.2 Measuring income poverty: 1899 and 1998 17
 CONTENTS Xi

1.3 Measuring vulnerability 20

1.4 Measuring voice and power using participatory methods 21
1.5 Measuring governance: participatory methods and cross-country surveys 21
1.6 Multidimensionality: dealing with aggregation 22
1.7 Uganda's poverty reduction goals 22
1.8 Tracking poverty in India during the 1990s 26
2.1 On interacting with state institutions: the voices of the poor 36
2.2 Preventing famines: the local press matters 40
2.3 Attacking poverty in Vietnam 41
3.1 Population, growth, and poverty 49
3.2 How war devastates poor people 50
3.3 Divergence and worldwide income inequality 51
3.4 Inequality trends and poverty reduction 53
3.5 What makes growth pro-poor in India? 54
3.6 Complex patterns of distributional change in three economies 55
3.7 Diversification and migration in rural China 56
3.8 Redistribution can be good for efficiency 57
4.1 The Washington consensus (3
4.2 Why do reforms sometimes fail? 65
4.3 Agricultural reforms in Chile and China help small farmers 68
4.4 Land markets and poor peasants in Mexico 68
4.5 Listening to farmers in Zambia 69
4.6 Attacking poverty with information 73
5.1 Interactions between human and physical capital 78
5.2 Links between the environment and health 78
5.3 Win-win policies in the health sector 79
5.4 Locked out by health and education fees 83
5.5 Mexico's Progresa:paying parents to send children to school 84
5.6 Some general principles on how to design subsidies 85
5.7 Effective public-private partnership in immunization 87
5.8 Local entrepreneurs increase access to telecommunications services 87
5.9 Local participation in Nicaragua's decentralized education system 89
5.10 Single-sectorand multisector arrangements for improving rural roads in Zambia 91
5.11 Rejuvenating India's forests through joint action 92
5.12 A new approach to land reform in Brazil 94
5.13 West African businesses pioneer water and sanitation services for the urban poor 95
5.14 Sustainable water and sanitation for Brazil's urban poor 96
6.1 Poor people are often harassed by public officials 100
6.2 Lawlessnesscontributes to poverty 103
6.3 Legal service organizations help poor people gain access to the protections of the legal system 105
6.4 Community monitoring can reduce environmental pollution 107
6.5 National coalitions against communicable diseasesin the West 109
6.6 NGOs can help mobilize and empower communities 110
6.7 Politics and poverty in OECD countries 1 14
6.8 The evolution of civil society and state reform in Mexico 1.15
7.1 Making land titling less gender biased in Latin America 121
7.2 Using subsidies to close gender gaps in education 122
Xii WORLD DEVELOPMENT REPORT 2000/200I

Boxes{continued)
7.3 Toward stronger female voice in policymaking: women's budget initiatives in southern Africa 122
7.4 Using development programs to break the power of agrarian elites:
a case study from eastern Uttar Pradesh, India 123
7.5 Discrimination is psychologicallydevastating 125
7.6 Using affirmative action against caste-based discrimination in India 126
7.7 Ethnic divisions and civil conflict 127
7.8 How does social capital affect development? 129
7.9 The federation of comedoresin Peru: the creation of linking social capital 130
7.10 Mobilizing and creating social capital in development projects 131
8.1 Poor people's exposure to risk 137
8.2 AIDS and poverty 139
8.3 Some key terms: risk, risk exposure, and vulnerability 139
8.4 Managing risk: the modular approach to social safety nets 147
8.5 Is targeting by the community a good idea? 150
8.6 Insurance options for the informal sector 151
8.7 Two universal health insurance systems: Costa Rica and Singapore 153
8.8 Social pensions in Chile and Namibia 154
8.9 Principles of successfulworkfare programs 156
8.10 The Eritrean Community Development Fund 157
8.11 The Self-Employed Women's Association of India 157
9.1 Providing social protection in response to crisis in the Republic of Korea 167
9.2 Public expenditure reviews to assessthe impact of fiscal retrenchment on poor people 168
9.3 Protecting poor people during fiscal adjustment: Peru's FiscalPrudence and Transparency Law 169
9.4 Mitigation is the cornerstone of emergency management in the United States 173
9.5 Mitigating the risks of natural catastrophes: lessons from the 1999 earthquakes in Turkey 173
9.6 Turning reconstruction into risk mitigation with the help of a local NGO 174
9.77 Mitigating risk with catastrophe bonds 174
9.8 Sharing the costs of catastrophes: the Mexican fund for natural disasters 175
9.9 Involving communitities in postdisaster reconstruction: lessons
from the Maharashtra Emergency Earthquake Rehabilitation Program 176
10.1 A successstory: the fight against river blindness in Africa 182
10.2 Research, maize, and pigs in rural Guizhou 184
10.3 Most biotechnology patents are private 185
10.4 The Global Environment Facility: a model for developing country participation 187
11.1 Learning about the consultative process through the Comprehensive Development Framework 195
11.2 The new poverty reduction strategy initiative 195
11.3 Sectorwide development cooperation 197
11.4 The common pool for development cooperation 197
11.5 Assessingcountry policies and institutions 198
11.6 How aid can help in countries with a weak policy environment 199
11.7 The Enhanced Heavily Indebted Poor Countries Debt Relief Initiative 201
11.8 How debt relief fits into a poverty reduction strategy: Uganda's Poverty Action Fund 204

Figures
1 Where the developing world's poor live 4
2 Where poverty has fallen, and where it has not 4
3 Infant mortality rates vary widely across the world 4
 CONTENTS XIII

1.1 Povertyin the developingworld is shiftingtoward SouthAsiaand Sub-SaharanAfrica 24

2.1 Povertyshowsa stronglink with economiccontractionsand expansions 35
3.1 In general,the wealthiera country,the lowerthe incidenceof poverty 46
3.2 Health indicatorsimproveas incomesrise 47
3.3 Povertytrends trackedgrowthtrendsin the 1980sand 1990s 47
3.4 Economicgrowthwasa forcefor povertyreductionin the 1980sand 1990s ... 48
3.5 Inequalityvaried widelyin the 1980sand 1990sbut showedno systematicassociationwith growth 53
3.6 Initial inequalitiesinfluencethe paceof povertyreduction 5,4
3.7 Acrosscountries,the ratio of femaleto maleliteracyriseswith per capitaincome 5l8
4.1 Indicationsof successfulpolicyreformsin the developingworld 63
4.2 Reformsdeliveredgrowthin LatinAmerica,althoughthe gainsvaried 64
4.3 The gap betweenskilledand unskilledwageswidenedin Mexico 71
6.1 Merit-basedrecruitmentin governmentis associatedwith lesscorruptionand bureaucraticdelay 101
6.2 Corruption is a regressivetax 102
6.3 Betterrule of lawis associatedwith higherper capitaincome 103
6.4 Decentralizationlowersthe costof raisingthe income of poor peoplein SouthAfrica 106
6.5 State-communitycoalitionscan foster rapid developmentand better servicedelivery 112
6.6 Good politicaland administrativeinstitutionsgo hand in hand with economicgrowth 112
7.1 Closingthe gendergap in schoolingmore rapidlywould boost economicgrowth 119
7.2 Trendsin femaleeducationand life expectancyreflectincreasingequalitybetweenwomen and men 120
7.3 Minoritygroupsin Vietnamhavelessaccessto servicesthan nonminorities 124
7.4 Ethnic diversityis associatedwith violencewhereinstitutionalqualityis low 128
8.1 Privatetransfersrepresenta largeshareof the incomeof the poor 144
8.2 Central governmentspendingon socialsecurityvaried greatlyin 1995 148
9.1 Developingcountriesbore the brunt of natural disastersin 1990-98 171
10.1 High-incomecountriesprotect manufacturingand agriculture 180
10.2 The burden of HIV/AIDSis heavilyconcentratedin Sub-SaharanAfrica 132
11.1 While donor countries'economiesgrewafter 1992,their developmentassistanceshrank 190
11.2 With the exceptionof Europeand Central Asia,the regionaldistribution
of officialdevelopmentassistanceremainedroughlyconstant ... 190
11.3 . . . while receiptsfell after 1992-93 in all regionsbut LatinAmericaand the Caribbean 191
11.4 Aid doesnot go only to poor countries 196
11.5 NGOs channeledsome $10 billionto developingcountriesin 1998,
about half of it from the officialsector 200
11.6 As per capitaincome in the heavilyindebtedpoor countries
has gone down, debt has gone up-and viceversa 201
11.7 Concessionaltransferslargelycompensatefor negativenet transfersof nonconcessionalresources 202

Tables
1.1 Incomepovertyby region,selectedyears,1987-98 23
1.2 Relativeincomepovertyby region,selectedyears,1987-98 24
1.3 Incomepovertyin sevenAfricancountries,variousyears 25
4.1 Impact of reformson agriculturalprices,output, and productivityin sevencountries 67
5.1 Publicspendingon educationby incomequintile in selecteddevelopingcountries,variousyears 80
5.2 Publicspendingon health by income quintilein selecteddevelopingcountries,variousyears 81
8.1 Main sourcesof risk 136
8.2 Shocksfacedby ruralhouseholdsin Ethiopia 140
8.3 Mechanismsfor managingrisk 141
XiV WORLD DEVELOPMENT REPORT 2000/2001
8.4 Income diversification among African farmers
8.5 Private cash and in-kind transfers for poor households
9.1 Effect of economic crises on incidence of poverty in selected countries
9.2 Social impacts of economic crises in selected countries
Definitionsand data notes
The countriesincludedin regionalgroupingsin this report
arelisted in the Classificationof Economiestableat the end
of the SelectedWorld DevelopmentIndicators.The income
groupingsin the main text of the reportare basedon 1998
GNP per capitaestimates,whilethosein the SelectedWorld
DevelopmentIndicatorsarebasedon 1999 GNP per capita
estimates(seethe Classificationof Economiestable). The
thresholdsfor the incomeclassificationsand the differences
betweenthe 1998 and 1999 classifications of countriesmay
be found in the introductionto the SelectedWorld Devel-
opment Indicators.Group averagesreported in the figures
and tables of the main text are unweightedaveragesof the
countriesin the group unlessnoted to the contrary.
In additionto the changesin incomeclassifications,re-
cent revisionsto data inthe SelectedWorld DevelopmentIn-
dicatorsmay result in differenceswith the data in the main
text(seethe introductionto the SelectedWorld Development
Indicators).
The use of the word countriesto referto economiesim-
pliesno judgmentby the WorldBankaboutthe legalor other
statusof a territory.The term developingcountriesincludes
low- and middle-incomeeconomiesand thus may include
economiesin transitionfrom central planning, as a matter
of convenience.The termsdeveloped, industri4 or richcoun-
142
144
163
164
tries may be used as a matter of convenienceto denote the
high-incomeeconomies.
Aggregatepovertymeasuresin the reportare often based
on the "$1 a day" povertyline. This lineis equal to $1.08 a
day in 1993purchasingpowerparity terms (for furtherex-
planationseebox 1.2 in chapter 1).
Dollar figuresare current U.S. dollars,unlessotherwise
specified.Billion means 1,000 million;trillionmeans 1,000
billion.
The followingabbreviationsare used:
AIDS Acquiredimmunedeficiencysyndrome
GATT GeneralAgreementon Tariffsand Trade
GDP Grossdomesticproduct
GNP Grossnationalproduct
HIPC Heavilyindebted poor country
HIV Human immunodeficiencyvirus
NGO Nongovernmentalorganization
OECD Organisationfor EconomicCo-operation
and Development
PPP Purchasingpowerparity
SEWA Self-EmployedWomen's Association
WTO World Trade Organization

Attacking
Opportunity,
Empowerment,
il. oorpeoplelivewithout fundamentalfreedoms
of action and choice that the better-off take for
granted.' They often lack adequate food and shelter,
education and health, deprivations that keep them
from leading the kind of life that everyone values.They
also face extreme vulnerability to ill health, economic
dislocation, and natural disasters. And they are often
exposed to ill treatment by institutions of the state and
society and are powerless to influence key decisions
affecting their lives. These are all dimensions of
poverty.
The experience of multiple deprivations is intense
and painful. Poor people's description of what living
in poverty means bearseloquent testimony to their pain
(box 1). For those who live in poverty, escaping it can
seem impossible. But it is not impossible. The story
of Basrabai-the chair of a local council in an Indian
village-illustrates both the many facetsof poverty and
the potential for action (see page 2).
OVERVIEW
Poverty:
and Security
Basrabai's
storyservesasa backdropto the exploration
of the nature and causesof poverty and of what can be
done. Poverty is the result of economic, political, and
social processes that interact with each other and fre-
quently reinforceeach other in ways that exacerbatethe
deprivationin which poor people live.Meager assets,in-
accessiblemarkets,and scarcejob opportunitieslock peo-
ple in material poverty. That is why promoting
opportunity-by stimulating economic growth, mak-
ing markets work better for poor people, and building
up their assets-is key to reducing poverty.
But this is only part of the story. In a world where
political power is unequally distributed and ofien
mimics the distribution of economic power, the way
state institutions operate may be particularly unfa-
vorable to poor people. For example, poor people fre-
quendy do not receivethe benefits of public investment
in education and health. And they are often the vic-
tims of corruption and arbitrariness on the part of the
I
2 WORLD DEVELOPMENT REPORT 2000/200I
Basrabailivesin Mohadi,a village500kilometersfrom Ahmed-
abad, in the Indianstate of Gujarat,on the shoresof the Ara-
bian Sea2 She is the first woman to be sarpanchof the
panchayat-chairof the ocalcouncil-as a resultof constitutional
amendmentsthat reservea thirdof localcouncilseatsandathird
of headshipsfor women.
Arriving in hervillageafter a longdrive,we crosseda small
sea inlet on a roadimpassableat high tide. The first building
we saw was a recentlybuilt concretestructure-the primary
school. In last year's cyclone,the worst in living memory,as
the villagers'straw huts were blown away, they took shelter
in the only stabie structure in the village-the school.When
the cyclone relief operation arrived, the villagers asked for
more concrete buildings,and the village now has about a
dozen of them.
We arrivedat Basrabai'sone-roomconcretehouse,nextto
a straw hut.After the usualgreetings,talkturnedto the school.
Since it was a weekday,we wonderedif we could sit in on a
class.Basrabaiinformedus that the master (the teacher)was
not there and hadnot beenthere for a while. In fact, he came
only oncea month, if that. Protectedby the district education
officer, he did pretty muchwhat he pleased.
The mastercamethe nextday.Wordhadgottento him that
the villagehadvisitors.Hecameinto Basrabai' s house,andwe
begantalkingabout the schoolandthe children.Believingthe
educatedgueststo be kindredspirits,he launchedinto a litany
of his troublesandthe difficultiesof teachingthe children.He
referredto them as jong/ee-from the jungle."
This was too much for Meeraiben,a memberof the Self-
EmployedWomen'sAssociation(SEWA),who hadarrangedour
visit. Shepointedout that his salarywas 6,000rupeesa month
(morethan sixtimes the IndianpovertyIne) ina securejob and
that his responsibilitywasat leastto show upfor work.The par-
ents wantedtheir childrento learnto readandwrite, even if at-
tendingschoolmeantthat the boyscould not helptheir fathers
with fishingandthe girlscouldnot helptheir mothersfetch water
andwood andwork in the fields.
Laterin the evenng Basrabai conductedthe villagemeeting.
Therewere two maintopics.Thefirst was compensation for the
cyclone:despitethe greatfanfarewith whichreliefschemeshad
beenannounced in the statecapital,localdeliveryleft muchto be
desiredandlocalofficialswere unresponsive. SEWAorganizers
tookdownthe namesof thosewho hadnot yetreceivedthe com-
pensationto whichtheywere entitled,andit wasagreedthatthey
andBasrabai would meetwith localofficalsthe followingweek.
Thesecondissuewas afishingbanthatthe governmenthad
imposedon coastalwatersto protectfish stocks.It was the big
trawlersthat were responsiblefor the overfishing,but the small
fishersseemedto bepayingthe price.Thebigtrawlerscouldcon-
tinue to fish as longas theypaidthe rightofficials.
lnthemiddleofthemeetingacommotionoccurredattheside.
Basrabai's brotherhadbeengoredinthe facewhiletryingto sep-
aratetwo fightingcows.Withoutimmediatetreatmentthe wound
was boundto becomeinfected.But it was lateat night,andthe
nearestdoctorwas in the next big settlement, 10 kilometers
away.Normally,this wouldhavemadeimmediatetreatmentim-
possible.As it happened,however,ourJeepwasthereandcould
take Basrabai's brotherto the doctor.
Duringour stay we also sawthe craft work that the village
women havebeendoingfor generations.Demandfor their tra-
ditionalembroidered andtie-dyedproductsis high,thanksto the
internationallove affairwith things Indianand the rediscovery
by the growingIndianmiddle classof its roots. But the traders
get awaywith offeringverylow pricesbecauseof the women's
isolation.
Thenationalandstategovernmentshavecountlessschemes
to supporttraditionalcrafts,noneveryeffective.SoSEWAis step-
pingin to organizethe home-based craftworkersandto provide
directaccessto internationalmarkets.Onepieceof embroidery
we lookedat wouldfetch 150rupeesinthe international market,
60 rupeesin governmentoutlets,and20 rupeesfrom traders.
Onthelastdayof ourstaywe wentto Basrabai's field,anhour's
walkfromher house.Therisksof agriculturewere plainlyvisible.
The lackof rainhadleft the groundhardanddry. If it didn't rain
in the nextfew days,her milletcrop would be lost, andwith it
her outlayto a hiredtractordriverto till herfield,an investment
made possibleby the sale of her crafts. When we met her in
Ahmedabad dayslater,it still hadnot rained.
The interactionswith Basrabaiandthe manythousandsof
poorpeopleconsultedin preparingthis reportbringto the fore
recurrent-and familiar-themes. Poorpeoplementionthe lack
of income-earning opportunities, the poorlinkswith markets,and
the failureof state institutionsto respondto their needs.They
mentioninsecurity,suchas health risks,the risk of beingout
of work, andthe agriculturalrisks that makeany gainsalways
fragile.Everywhere-from the villagesin Indiato the favelasof
RiodeJaneiro,the shantytownsoutsideJohannesburg, andthe
farms in Uzbekistan-the storiesbringforward similarissues.
Buttalkingto Basrabai andotherpoorpeoplealsorevealswhat
is possible.Althoughlocalofficialsandstate structuresare still
not accountableto Basrabaiandhervillage,an explicitaffirma-
tiveactionpolicyallowedBasrabai'selectionas sarpanch, show-
ing what can be donethrough state action.And SEWAshows
how poorpeoplecan makea differenceif they organizethem-
selvesto defendtheir rights,take advantageof marketoppor-
tunities,andprotect themselvesfrom risks.
 ATTACKING POVERTY:
Box 1
The voices of the poor
The Voicesof the Poorstudy,basedon realitiesof morethan
60,000poorwomenandmenin 60countries,wasconducted as
background for WorldDevelopment Report2000/2001. It consists
of two parts:areviewof recentparticipatory
povertystudiesin 50
countriesinvolvingabout40,000poorpeople,anda newcompar-
ativestudyin 1999in23countriesengaging about20,000poorpeo-
ple. Thestudyshowsthat poorpeopleareactiveagentsin their
lives,but are oftenpowerlessto influencethe socialandeco-
nomicfactorsthatdeterminetheirwell-being.
The followingquotationsarean illustrationof what livingin
povertymeans.
Don't askme whatpoverty is becauseyou havemet it outside
my house.Lookat the houseandcountthe numberof holes.
Look at the utensils and the clothesI am wearing.Look at
everythingand write what you see.Whatyou seeis poverty,
-Poor man,Kenya
Certainlyour farmingis little; all the products,things bought
from stores,are expensive;it is hard to live, we work and
earn little money,buy few things or products;products are
Source:Narayan,Chambers,Shah,and Petesch 2000; Narayan,Patel,Schafft, Rademacher,and Koch-Schulte2000.
state. Poverty outcomes are also gready affected by social
norms, values, and customary practices that, within the
family, the community, or the market, lead to exclusion
of women, ethnic and racial groups, or the socially dis-
advantaged. That is why facilitating the empowerment
of poor people-by making state and social institutions
more responsiveto them-is also key to reducing poverty.
Vulnerability to external and largely uncontrollable
events-illness, violence, economic shocks, bad weather,
natural disasters-reinforces poor people's sense of ill-
being, exacerbatestheir materialpoverty,and weakenstheir
bargaining position. That is why enhancing security-
by reducing the risk of such events as wars, disease, eco-
nomic crises, and natural disasters-is key to reducing
poverty.And so is reducing poor people's vulnerability to
risksand putting in place mechanisms to help them cope
with adverse shocks.
Povertyin an unequalworld
The world has deep poverty amid plenty. Of the world's
6 billionpeople, 2.8 billion-almost half-live on lessthan
$2 a day, and 1.2 billion-a fifth-live on less than $1 a
day,with 44 percent living in South Asia (figure 1). In rich
countries fewerthan 1 child in 100 does not reach its fifth
OPPORTUNITY, EMPOWERMENT, AND SECURITY 3
scarce,there is no money and we feel poor. If there were
money ...
-From a discussion
groupofpoormenandwomen,Ecuador
We facea calamitywhenmy husbandfallsill. Ourlife comes
to a halt until he recoversand goesbackto work.
-Poor woman,ZawyetSultan,Egypt
Povertyis humiliation,the senseof beingdependentonthem,
and of being forcedto acceptrudeness,insults,and indiffer-
encewhen we seekhelp.
-Poor woman,Latvia
At first I wasafraidofeveryoneand everything:my husband,
the villagesarpanch,thepolice.TodayI fear noone.I havemy
own bank account,I am the leaderof my village's savings
group ... I tell my sistersaboutour movement.And we have
a 40,000-strong
union in thedistrict
-From a discusswon
group
of poormenandwomen,India
birthday,while in the poorest countries as many as a fifth
of children do not. And while in rich countries fewerthan
5 percent of all children under five are malnourished, in
poor countries as many as 50 percent are.
This destitution persists even though human concii-
tions have improved more in the past century than in the
rest of history-global wealth, global connections, and
technologicalcapabilitieshave never been greater. But the
distribution of these global gains is extraordinarily un-
equal. The average income in the richest 20 countries is
37 times the average in the poorest 20-a gap that has
doubled in the past 40 years. And the experience in dif-
ferent parts of the world has been very diverse (figure 2;
see also table 1.1 in chapter 1). In East Asia the number
of people living on less than $1 a day fell from around
420 million to around 280 million between 1987 and
1998-even after the setbacks of the financial crisis.3
Yetin Latin America, South Asia,and Sub-SaharanAfrSica
the numbers ofpoor people have been rising. And in the
countries of Europe and Central Asia in transition to mar-
ket economies, the number of people living on less than
$1 a day rose more than twentyfold.4
There have also been major advances and serious set-
backs in crucial nonincome measuresof poverty.India lhas
seen marked progressin girlsattending school, and in the
4 WORLD DEVELOPMENT REPORT 2000/2001

Figure 1 Figure 2
Where the developingworld's poor live Where poverty has fallen, and where it has not

Distributionof populationlivingon lessthan S1a day, Changein numberof peoplelivingon lessthan S1a day,
199811.2billion) 1987-98
Europeand CentralAsia Middle Eastand North Africa 100 Millions
2.0%/ 0.5%
Latin America and 75
the Caribbean 50

25
EastAsia
and Pacific South Asia 0
23.2% 43.5% -25

-50
-75

-100
Su -Saharan -125
Africa
24.3% -150
EastAsia Europe Latin Middle South Sub-
Source:WorldBank2000s. and and AmericaEastand Asia Saharan
Pacific Central and the North Africa
Asia Caribbean Africa

most advancedstate, Kerala,life expectancyis greaterthan Source:WorldBank 2000s.

in other places with many times the level of income
(such as Washington, D.C.). Yetin countries at the cen-
ter of the HIV/AIDS epidemic in Africa,such as Botswana
and Zimbabwe, one in four adults is infected, AIDS or- Figure 3
phans are becoming an overwhelming burden on both Infant mortality rates vary widely acrossthe
traditional and formal support mechanisms, and all the world
gains in life expectancy since the middle of the 20th
centurywill soon bee wiped out. The varying
u.Te ayn infant
nfn mor-o-Per Infantmortalityrate,1998
1,000live births
taltty rates across the world-Sub-Saharan Africa'sis 15 100
times that of high-income countries-give an idea of 90
this widely differing experience (figure 3). 80
Experiences are also vastly different at subnational 70
levelsand for ethnic minorities and women. Different re-
gions in countries benefit to very different extents from 60
growth. In Mexico, for example, total poverty fell- 50
though modestly-in the early 1990s, but rose in the 40
poorer Southeast. Inequalities also exist across different 30
ethnicgroupsin many countries. In some African coun- 20
tries infant mortality ratesare loweramong politicallypow- 10
erful ethnic groups, and in Latin American countries 0
10
indigenous groups often have less than three-quarters East Europe Latin Middle South Sub- High-
the schooling on averageof nonindigenous groups. And Asia and AmericaEastand Asia Saharan income
be disadvantaged Iandthn In Central and the North Africa countries
women continue to be more disadvantaged than men. In Pacific Asia Caribbean Africa
South Asia women have only about half as many years
of education as men, and female enrollment rates at the Source: World Bank2000s.

secondary level are only two-thirds the male rates.

 ATTACKING POVERTY: OPPORTUNITY, EMPOWERMENT, AND SECURITY ';

Box2
A better world for all: internationaldevelopment goals

Reducethe proportion of People living on less than development

Thegoalsfor international
people living in extreme $t a day (%) 30 addressthat mostcompellingof human
W poverty by half between Poress 1990-98 desires-aworldfreeof povertyandfree
1990 and 2015 20 of the miserythatpovertybreeds.
an
Eachof the sevengoalsaddresses
Average path to goal''*- ., aspectof poverty.Theyshouldbeviewed
1990 2015 togetherbecausetheyare mutuallyre-
Higherschoolenrollments,
inforcing. es-
Enrollall childrenin primary Net primary enrollment rate (%) peciallyfor girls, reduce povertyand
school by 2015 100 mortality.Better basichealth care in-
_ Average path to goal -- ' creasesenrollment andreducespoverty.
''i P'rogress 1990-98 75 Manypoorpeopleearntheir livingfrom
the environment. Soprogressis needed
50 oneachof the sevengoals.
1990 2015 In the past decadeon averagethe
world hasnot beenon trackto achieve
Makeprogress toward Ratioof girlsto boysin primary the goals.Butprogressinsomecountries
genderequalityand and secondary school (%) 1 and regionsshowswhat can be done.
_ empowering women °__
,__________

by eliminating gender Average path to go,al.-.-.--

disparities in primary and
secondaryeducationby 2005
Reduceinfantandchild
mortality rates by two-third
w between 1990 and 201 5
- Chinareduced itsnumberin poverty from
Progres 1990 98 75 360 millionin 1990to about210million
in 1998.Mauritiuscut itsmilitarybudget
50 and investedheavilyin healthandedu-
19902005 cation.TodayallMauritians haveaccess
to sanitation,98 percentto safewater,
Under-5 mortality rate and 97 percent of birthsareattendedby
(per 1,000 live births) loO skilledhealthstaff. And many LatinAmer-
_ ________
Progress1990-98 ican countriesmoved muchcloserto
" genderequalityin education.
ah to goa - -
Avrg
Themessage:if somecountriescan
0 makegreat progresstoward reducing
povertyin itsmanyforms,otherscanas

O
1990 2015

well.But conflictis reversinggainsinso-

Reduce maternal
mortality Births attended by skilled health cial developmentin manycountriesin
ratios by three-quarters
a%) personnel 10d Sub-SaharanAfrica. The spread of
between 1990 and 201 5 Averagepathtogoal - - HIV/AIDSis impoverishingindividuals,
-,.--''--- - - 50 families,andcommunitieson all conti-
Progress 1988-98 nents.Andsustained economic growth-
0 that vital component for long-run
1988 2015 reductionsin poverty-still eludeshalf
the world'scountries.Formorethan30
~ Provide accessfor all who Contraceptive prevalencerate have
of them,realpercapitaincomes
ET;_ need reproductive health
_ services by 2015 80 fallenoverthe past35years.Andwhere
Progress 1993-98 there is growth, it needsto be spread
moreequally.
600 Thegoalscanbemet-with a combi-
50 nationof effectivedomesticandinterna-

O
1993 1998 tionalactions.

Implement national Countries

with environmental
strategiesfor sustainable strategies
1Ml
_ developmentby 2005 so as so
to reverse the lossof _ _8_
environmental resourcesby Progres 25
2015

1984 1997

countries
Note:Dataareforlow-andmiddle-income except whichreferto allcountries.
strategies,
for thoseonenvironmental
Source:IMF,OECD, andWorldBank
UnitedNations, 2000(www.paris2l.org/betterworldh.
6 WORLD DEVELOPMENT REPORT 2000/2001
Faced with this picture of global poverty and in-
equality, the international community has set itself sev-
eral goals for the opening years of the century, based on
discussions at various United Nations conferencesin the
1990s (box 2). These international development goals,
most for 2015, include reducing income poverty and
human deprivation in many dimensions (the bench-
marks are figures for 1990):
* Reduce by half the proportion of people living in ex-
treme income poverty (living on less than $1 a day).
* Ensure universal primary education.
* Eliminate gender disparity in primary and secondary
education (by 2005).
* Reduceinfant and childmortality by two-thirds.
• Reduce maternal mortality by three-quarters.
* Ensure universal accessto reproductive health services.
* Implement national strategies for sustainable devel-
opment in every country by 2005, so as to reversethe
loss of environmental resources by 2015.
These will have to be achieved in a world whose popu-
lation will grow by some 2 billion in the next 25 years,with
97 percent of that increasein developing countries. Stud-
ies of what must be done to achievethese goals revealthe
magnitude of the challenge.For example, cutting income
poverty by half between 1990 and 2015 would require a
compound rate of decline of 2.7 percent a year over those
25 years.The World Bank's latest estimates indicate a re-
duction of approximately1.7 percent a year between 1990
and 1998. Much of the slow progressobservedin some re-
gions is due to low or negative growth. In some casesris-
ing inequalitycompounded this effect;this was particularly
so in some countries in the former SovietUnion. The cur-
rent pace of educationalenrollmentis unlikelyto bring uni-
versalprimary education, especiallyin Sub-SaharanAfrica.
Reducing infant mortality rates by two-thirds between
1990 and 2015 would have required a 30 percent decline
between 1990 and 1998, fargreaterthan the 10 percent de-
veloping countries experienced.In some parts of Sub-Sa-
haran Africa infant mortality is actually on the rise,partly
as a result of the AIDS epidemic.And maternal mortality
ratios are declining too slowlyto meet the goals.
Attaining the international development goals will
require actions to spur economic growth and reduce in-
come inequality, but even equitable growth will not be
enough to achievethe goalsfor health and education. Re-
ducing infant and child mortality rates by two-thirds
depends on halting the spread of HIV/AIDS, increasing
the capacity of developingcountries' health systemsto de-
liver more health services, and ensuring that technolog-
ical progress in the medical field spills over to benefit the
developing world.5 And meeting the gender equality
goals in education will require specific policy measures
to address the cultural, social, and economic barriers
that prevent girls from attending school.6 Furthermore,
actions to ensure greater environmental sustainability
will be crucial in augmenting the assets availableto poor
people and in reducing the long-term incidence of
poverty.7 These actions will all interact to push toward
the achievement of the goals. Hence the need for a
broader, more comprehensive strategy to fight poverty.
A strategy for poverty reduction
The approach to reducing poverty has evolved over the
past 50 years in response to deepening understanding of
the complexity of development. In the 1950s and 1960s
many viewedlarge investments in physical capital and in-
frastructure as the primary means of development.
In the 1970s awarenessgrew that physical capital was
not enough, and that at least as important were health and
education. WorldDevelopmentReport1980 articulatedthis
understandingand argued that improvementsin health and
education were important not only in their own right but
also to promote growth in the incomes of poor people.
The 1980ssaw another shift of emphasisfollowingthe
debt crisis and global recession and the contrasting ex-
periencesof East Asia and Latin America, South Asia,and
Sub-Saharan Africa. Emphasis was placed on improving
economic management and allowing greater play for
market forces. WorldDevelopmentReport 1990: Poverty
proposed a two-part strategy: promoting labor-intensive
growth through economic openness and investment in
infrastructure and providing basic services to poor peo-
ple in health and education.
In the 1990s governance and institutions moved to-
ward center stage-as did issuesof vulnerabilityat the local
and national levels. This report builds on the earlier
strategies in the light of the cumulative evidence and ex-
perience of the past decade-and in the light of the
changed global context. It proposes a strategy for at-
tacking poverty in three ways: promoting opportunity,
facilitating empowerment, and enhancing security.
* Promotingopportunity.Poor people consistently em-
phasize the centrality of material opportunities. This
means jobs, credit, roads, electricity,markets for their
produce, and the schools, water,sanitation, and health
 ATTACKING POVERTY:
servicesthat underpin the health and skillsessentialfor
work. Overall economic growth is crucial for generat-
ing opportunity. So is the pattern or quality of growth.
Market reforms can be central in expanding opportu-
nities for poor people, but reforms need to reflectlocal
institutionaland structuralconditions.And mechanisms
need to be in placeto createnew opportunitiesand com-
pensate the potential losers in transitions. In societies
with high inequality, greater equity is particularly im-
portant for rapid progressin reducing poverty.This re-
quires action by the state to support the buildup of
human, land, and infrastructure assets that poor peo-
ple own or to which they have access.
Facilitatingempowerment.The choice and implemen-
F ple can help to ensure service deliveryand policy choices
tation of public actions that are responsiveto the needs
of poor people depend on the interaction of political,
social,and other institutional processes.Accessto mar-
ket opportunities and to public sector servicesis often
strongly influenced by state and social institutions,
which must be responsiveand accountableto poor peo-
ple.Achieving access,responsibility,and accountability
is intrinsicallypolitical and requires active collabora-
tion among poor people, the middle class,and other
groups in society.Activecollaborationcan be greatlyfa-
cilitated by changesin governancethat make public ad-
ministration,legalinstitutions,and publicservicedelivery
more efficient and accountable to all citizens-and by
strengtheningthe participationof poor peoplein political
processesand local decisionmaking.Also important is
removingthe socialand institutional barriers that result
from distinctions of gender,ethnicity, and socialstatus.
Sound and responsiveinstitutions are not only impor-
tant to benefit the poor but are also fundamental to the
overallgrowth process.
* Enhancing security. Reducing vulnerability-to
economic shocks, natural disasters, ill health, dis-
ability, and personal violence-is an intrinsic part of
enhancing well-being and encourages investment in
human capital and in higher-risk, higher-return ac-
tivities.This requires effectivenational action to man-
age the risk of economywide shocks and effective
mechanisms to reduce the risks faced by poor people,
including health- and weather-related risks. It also re-
quires building the assets of poor people, diversify-
ing household activities, and providing a range of
insurance mechanisms to cope with adverse shocks-
from public work to stay-in-school programs and
health insurance.
OPPORTUNITY, EMPOWERMENT, AND SECURITY 7
There is no hierarchyof importance. The elements are
deeplycomplementary.Each part of the strategyaffectsurn-
derlying causesof poverty addressed by the other two. For
example,promoting opportunity through assetsand mar-
ket accessincreasesthe independence of poor people and
thus empowers them by strengthening their bargaining
position relativeto state and society.It also enhances secu-
rity,since an adequate stock of assetsis a buffer against ad-
verseshocks.Similarly,strengtheningdemocraticinstitutions
and empoweringwomen and disadvantagedethnicand racial
groups-say, by eliminating legal discrimination against
them-expand the economicopportunitiesfor the poor and
sociallyexcluded.Strengtheningorganizationsof poor peo-
responsiveto the needs of poor people and can reduce cor-
ruption and arbitrarinessin state actionsaswell.And if poor
people do more in monitoring and controllingthe local de-
livery of social services,public spending is more likely ito
help them during crises.Finally,helping poor people cope
with shocks and manage risksputs them in a better posi-
tion to take advantage of emerging market opportunities.
That is why this report advocates a comprehensive ap-
proach to attacking poverty.
From strategy to action
There is no simple, universalblueprint for implementing
this strategy.Developing countries need to prepare their
own mix of policiesto reduce poverty,reflecting national
priorities and local realities. Choices will depend on tlhe
economic, sociopolitical, structural, and cultural context
of individual countries-indeed, individual communities.
While this report proposes a more comprehensive ap-
proach,prioritieswillhave to be set in individualcasesbased
on resourcesand what is institutionally feasible. Progress
in reducing some aspects of deprivation is possible even
if other aspects remain unchanged. For example, inex-
pensive oral rehydration campaigns can significantly re-
duce infant mortality, even if incomes of poor people do
not change.8 But actions will generallybe necessaryin all
three clusters-opportunity, empowerment,and security--
because of the complementarities among the three.
The actions of developedcountries and multilateralor-
ganizationswill be crucial.Many forcesaffectingpoor peo-
ple'slivesare beyond their influenceor control. Developing
countries cannot on their own produce such things as iin-
ternationalfinancialstability,major advancesin health and
agricultural research,and international trading opportu-
8 WORLD DEVELOPMENT REPORT 2000/2001
nities. Actions by the international community and de-
velopment cooperation will continue to be essential.
Here are the suggested areas for action, first national
and then international.
Opportunity
The core policies and institutions for creating more op-
portunities involve complementary actions to stimulate
overallgrowth, make markets work for poor people, and
build their assets-including addressing deep-seated in-
equalities in the distribution of such endowments as
education.
Encouragingeffectiveprivateinvestment.Investment and
technological innovation are the main drivers of growth
in jobs and labor incomes.Fosteringprivateinvestment re-
quires reducing risk for private investors-through stable
fiscal and monetary policy, stable investment regimes,
sound financial systems,and a clear and transparent busi-
nessenvironment. But it also involvesensuring the rule of
law and taking measures to fight corruption-tackling
businessenvironments based on kickbacks, subsidies for
large investors, specialdeals, and favored monopolies.
Special measures are frequently essential to ensure
that microenterprises and small businesses, which are
often particularly vulnerable to bureaucratic harassment
and the buying of privilege by the well-connected, can
participate effectivelyin markets. Such measures include
ensuring accessto creditby promoting financialdeepening
and reducing the sources of market failure; lowering the
transactionscostsof reachingexport markets by expanding
access to Internet technology, organizing export fairs,
and providing training in modern business practices;
and building feeder roads to reduce physical barriers.
Creating a sound business environment for poor house-
holds and small firms may also involve deregulation and
complementaryinstitutionalreform,for example,reducing
restrictionson the informal sector,especiallythose affecting
women, and tackling land tenure or registryinadequacies
that discourage small investments.
Private investment will have to be complemented by
public investment to enhance competitiveness and cre-
ate new market opportunities. Particularly important is
complementary public investment in expanding infra-
structure and communications and upgrading the skills
of the labor force.
Expandinginto internationalnarkets.International mar-
kets offera huge opportunity for job and income growth-
in agriculture,industry,and services.Allcountries that have
had major reductions in income poverty havemade use of
international trade. But opening to trade can create losers
as well as winners, and it will yieldsubstantialbenefits only
when countries have the infrastructure and institutions to
underpin a strong supply response. Thus the opening
needs to be well designed,with specialattention to coun-
try specificsand to institutionaland other bottlenecks.The
sequencing of policiesshould encourage job creation and
manage job destruction. A more pro-poor liberalizationis
not necessarilya slower one; moving fast can create more
opportunities for the poor.And explicitpoliciesshould off-
set transitory costs for poor people, as the grants for small
Mexican maize producers did in the wake of the North
American FreeTrade Agreement (NAFTA).
The opening of the capital account has to be managed
prudently-in step with domestic financial sector
development-to reduce the risk of high volatility in
capital flows. Long-term direct investment can bring
positive externalities, such as knowledge transfer, but
short-term flows can bring negative externalities, partic-
ularly volatility. Policies need to address them separately.
Building the assetsof poor people. Creating human,
physical,natural, and financialassetsthat poor people own
or can use requires actions on three fronts. First, increase
the focus of public spending on poor people in particu-
lar, expanding the supply of basic social and economic
services and relaxing constraints on the demand side
(through, for example, scholarships for poor children).
Second, ensure good quality servicedelivery through in-
stitutional action involvingsound governanceand the use
of markets and multiple agents. This can imply both re-
forming public delivery, as in education, or privatizing
in a fashion that ensuresexpansion of servicesto poor peo-
ple, as often makes sense in urban water and sanitation.
Third, ensure the participation of poor communities
and households in choosing and implementing services
and monitoring them to keep providers accountable.
This has been tried in projects in El Salvador,Tunisia, and
Uganda. Programs to build the assets of poor people
include broad-basedexpansion of schooling with parental
and community involvement, stay-in-school programs
(such as those in Bangladesh,Brazil,Mexico, and Poland),
nutrition programs, mother and child health programs,
vaccinations and other health interventions, and
community-based schemes to protect water resources
and other elements of the natural environment.
There are powerful complementaritiesbetween actions
in different areas. Because of close linkages between
 ATTACKING POVERTY:
human and physical assets,for example, improving poor
people's access to energy or transport can increase their
access and returns to education. And improving the en-
vironment can have significant effects on poverty. This
is well documented in terms of the substantial gains in
health from reduced air and water pollution-which
have a major influence on some of the most important
diseasesof poor people, including diarrheal problems of
children and respiratory infections.
Addressingassetinequalitiesacrossgender,ethnic, racial,
and socialdivides. Special action is required in many so-
cieties to tackle sociallybased asset inequalities. Although
political and social difficulties often obstruct change,
there are many examplesof mechanisms that work, using
a mix of public spending, institutional change, and par-
ticipation. One is negotiated land reform, backed by
public action to support small farmers, as in Northeast
Brazil and the Philippines. Another is getting girls into
school, such as by offering cash or food for schooling, as
in Bangladesh, Brazil, and Mexico, and hiring more fe-
male teachers, as in Pakistan. A third is support for mi-
crocredit schemes for poor women.
Gettinginfiastructureandknowledgetopoorareas-rural
and urban.Specialaction is alsoneededin poor areas,where
a combination of asset deprivations-including at the
community or regional level-can diminish the material
prospectsfor poor people.Tacklingthis again requirespub-
lic support and a range of institutional and participatory
approaches. It requires providing social and economic
infrastructure in poor, remote areas, including transport,
telecommunications,schools,health services,and electricity,
as in China's poor areas programs. It also requires broad-
based provision of basic urban services in slums, within
an overallurban strategy.Alsoimportant is expanding ac-
cess to information for poor villages,to allow them to par-
ticipate in markets and to monitor local government.
Empowerment
The potential for economic growth and poverty reduc-
tion is heavily influenced by state and social institutions.
Action to improve the functioning of state and social in-
stitutions improves both growth and equity by reducing
bureaucraticand socialconstraintsto economic action and
upward mobility. However, devising and implementing
these changes require strong political will, especially
when the changes fundamentally challenge social values
or entrenched interests.Governments can do much to in-
fluence public debate to increase awareness of the soci-
OPPORTUNITY, EMPOWERMENT, AND SECURITY 9
etal benefits of pro-poor public action and build politi-
cal support for such action.
Laying the political and legalbasisfor inclusive devel-
opment. State institutions need to be open and account-
able to all. This means having transparent institutions,
with democratic and participatory mechanisms for mak-
ing decisions and monitoring their implementation,
backed up by legal systems that foster economic growth
and promote legal equity. Since poor people lack the re-
sources and the information to access the legal system,
measures such as legal aid and dissemination of infor-
mation on legalprocedures-for example, by the Ain-O-
Salish Kendra (ASK) organization in Bangladesh-are
especiallypowerful instruments for creating more inclu-
sive and accountable legal systems.
Creatingpublic administrationsthatfoster growth and
equity.Public administrations that implement policiesef-
ficiently and without corruption or harassment improve
service deliveryby the public sector and facilitategrowth
of the private sector.Appropriate performance incentives
are needed to make public administrations accountable
and responsiveto users.Accessto information such as bud-
gets, participatory budget mechanisms, and performance
rating of public services all enhance citizens' capacity to
shape and monitor public sector performance while re-
ducing opportunities and scope for corruption. Reform-
ing public administrations and other agencies such as the
police to increasetheir accountability and responsiveness
to poor people can have a major impact on their dailylives.
Promotinginclusivedecentralizationand communityde-
velopment.Decentralizationcan bring serviceagenciescloser
to poor communities and poor people, potentially en-
hancing people's control of the servicesto which they are
entitled.This will requirethe strengtheningof localcapacity
and devolution of financialresources.It is also necessaryto
have measures to avoid capture by local elites. Decentral-
izationneedsto be combinedwith effectivemechanismsfor
popular participationand citizenmonitoring of government
agencies. Examples include decentralization that fosters
community-driven choices for resource use and project
implementation.There is alsoa rangeof options for involving
communities and households in sectoral activities-sach
as parental involvement in schooling and users associa-
tions in water supply and irrigation.
Promotinggenderequity.Unequal gender relations are
part of the broader issue of social inequities based on so-
cietal norms and values.But gender equalityis of such per-
vasivesignificancethat it deserves extra emphasis.While
10 WORLD DEVELOPMENT REPORT 2000/200I
patternsof genderinequityvarygreatlyacrosssocieties,in
almostallcountriesa majorityof womenand girlsaredis-
advantagedin termsoftheirrelativepowerandcontrolover
materialresources(in most countriesland titlesare vested
in men),and theyoften facemore severeinsecurities(for
example,afterthe deathof their husband).Poorwomen
are thus doublydisadvantaged. Moreover,the lackof au-
tonomy of womenhas significantnegativeconsequences
for the educationand healthof children.
Greater gender equity is desirablein its own right
and forits instrumentalsocialand economicbenefitsfor
povertyreduction.There has beenprogress-for exam-
ple, in educationand health-but much more needsto
be done. Experienceindicatesthat a mix of political,
legal,anddirectpublicactionis required.Thirty-twocoun-
tries,from Argentinato India,havemeasuresto promote
women'srepresentationin localand nationalassemblies,
and this is alreadytransformingwomen'sabilityto par-
ticipate in public life and decisionmaking.Somecoun-
triesare correctinggenderbiasesin the law,asin the 1994
Colombian AgrarianLaw. Use of public resourcesto
subsidizegirls' educationhas been shownto pay off in
Bangladeshand Pakistan.A range of measuresin pro-
ductiveactivities,notablymicrofinanceand farmingin-
puts, haveproduced documentedbenefitsin terms of
increasedyields(in Kenya,forexample)and increasedau-
tonomy for womenand better nutritionalstatusof chil-
dren (in Bangladeshand in virtuallyeverysettingwhere
this issuehas beenexamined).
Tacklingsocialbarriers.
Socialstructuresand institutions
form the frameworkfor economicand politicalrelations
and shapemany of the dynamicsthat createand sustain
poverty-or alleviateit. Socialstructuresthat are exclu-
sionaryandinequitable,suchasclassstratificationor gen-
der divisions,are majorobstaclesto the upwardmobility
of poorpeople.Governmentscan helpby fosteringdebate
overexclusionarypracticesor areasof stigmaand by sup-
portingthe engagementand participationof groupsrep-
resenting the sociallyexcluded. Groups facing active
discriminationcan be helpedby selectiveaffirmativeac-
tion policies.Socialfragmentationcan be mitigatedby
bringinggroupstogetherin formaland informalforums
and channelingtheir energiesinto politicalprocessesin-
steadofopenconflict.Otheractionscouldincluderemoving
ethnic, racial,and genderbiasin legislationand the oper-
ation of legalsystemsand encouragingthe representation
and voiceof womenand disadvantagedethnicand racial
groupsin communityand nationalorganizations.
Supportingpoorpeoples socialcapital.Socialnorms and
networksare a keyform of capitalthat peoplecan use to
move out of poverty.Thus it is important to work with
and support networksof poor people and to enhance
their potential by linkingthem to intermediaryorgani-
zations,broadermarkets,and publicinstitutions.Doing
this alsorequiresimprovingthe legal,regulatory,andin-
stitutional environmentsfor groups representingpoor
people.Sincepoorpeopleusuallyorganizeat the locallevel,
actionswillalsobe neededto strengthentheir capacityto
influencepolicyat the state and national levels,such as
by linkinglocalorganizationsto widerorganizations.
Security
Achievinggreatersecurityrequiresa heightenedfocuson
howinsecurityaffectsthe livesandprospectsof poorpeo-
ple. It alsotakes a mix of measuresto deal with econo-
mywide or regionwiderisksand to help poor people
copewith individualadverseshocks.
Formulatinga modularapproachto helpingpoorpeo-
ple managerisk. Different interventions-at the com-
munity,market,and state levels-are neededto address
differentrisksand differentsegmentsof the population.
A mixof interventionsmaybe neededto supportthe man-
agementof risksfor communitiesand households,de-
pendingon the typeof risk and the institutionalcapacity
of the country.Microinsurance programscancomplement
microcreditprograms for poor women, built around
their organizations,as in the schemesSEWAruns in
India for women in the informal sector. Public work
schemescan expand in response to local or national
shocks.Food transferprogramsand socialfunds to help
financeprojectsidentifiedby communitiescan alsobe
effectivein copingwith disaster.
Developing nationalprogramstoprevent,preparefor, and
respondto macroshocks-financialand natural.Economy-
wideshocksare often the hardestfor poor communities
and householdsto copewith, especiallywhenthe shocks
are repeated,deep,or persistent.To managethe riskof fi-
nancialand termsof tradeshocks,soundmacroeconomic
policyand robust financialsystemsare fundamental.But
they haveto be complementedby prudent management
of the openingof the capitalaccount,to reducethe risk
ofvolatileshort-runflows.Specialmeasuresarealsoneeded
to ensurethat spendingon programsimportantto poor
people-social programsand targetedtransfers-doesnot
fallduringa recession,especiallyrelativeto the risingneed.
Equallyimportant,countercyclical safetynetsshould be
 ATTACKING POVERTY:
permanentand readyto be deployedwhen countriesare
hit by a shock.These and other actionscan alsohelp in
copingwith naturalshocks."Calamityfunds"canfinance
reliefeffortsfollowingnaturaldisastersand supportnew
technologyand trainingfor betterrisk assessment.Mak-
ing investmentsand insurancearrangementsin normal
timescan reducepersonalcostswhena disasteroccurs.
Designing nationalsystemsofsocialriskmanagement that
arealsopro-growth.There is demandacrossthe worldfor
nationalsystemsofsocialriskmanagement.The challenge
is to designthem so that they do not undercutcompet-
itivenessand so that poor peoplebenefit.Someexamples:
systemsthat both provide insurancefor the nonpoor
andincludesocialpensionsfor the poor,as in Chile;health
insurancethat protects againstcatastrophicillnessthat
couldwipeout a family'sassets,as in CostaRica;and un-
employmentinsuranceand assistancethat do not com-
promisethe incentiveto work. To gain the full benefits
of such schemes,however,economiesneed the institu-
tional capacityto managethem effectively.
Addressingcivil conflict.Civil conflict is devastating
for poor people:the bulk of conflictsare in poor coun-
triesand mostare civilwars-more than 85 percentof all
conflictswere fought within country borders between
1987and 1997.In additionto the directlossof life,they
wreaksocialandeconomichavocandcreatea terriblelegacy
of psychologicaland socialtrauma. Child soldiersare
often recruitedto fight-as in SierraLeone-and many
more children suffer the loss of family,disruption of
schooling,and psychological scarsthat permanentlydi-
minsh their prospects.
Whileit is immenselyimportantto sustainthe focuson
rebuildingsocietiesafter conflict,such as in Cambodia
and Rwanda,it is equallyurgentto takemeasuresto pre-
ventconflict.Thereissomeevidencethatstrengthening plu-
ralistinstitutions-supporting minorityrightsandproviding
the institutionalbasisfor peacefulconflictresolution-
has a significantinfluence.Alsoimportant for averting
conflictareeffortsto getdifferentgroupsto interactthrough
moreinclusiveand participatorypoliticalinstitutionsand
throughcivilinstitutions.Asnotedbelow,internationalac-
tion to reduceaccessto the resourcesto financeconflictand
to reduceinternationaltradein armamentsis alsoneces-
sary.If countriescanget onto a pathof inclusiveeconomic
development, theyhavethepotentialto shiftfroma vicious
to a virtuouscycle.Violentconflictconstitutesone of the
most urgentand intractableareasforactionaffectingsome
of the poorestpeoplein the world.
OPPORTUNITY, EMPOWERMENT, AND SECURITY II
TacklingtheHIVIAIDSepidemic.HIV/AIDSisalready
one of the mostimportantsourcesof insecurityin severely
afflictedcountriesin Africa.Whilethe immediate,dev-
astatingeffectsare at the individualand householdlevel,
the consequencesare much broader, from intolerable
strains on traditional child fostering mechanisms to
extremepressureson healthsystemsand lossof produc-
tive labor affecting whole communitiesand nations.
Morethan 34 millionpeopleare infectedwith HIV (90
percentin the developingworld),with 5 millionmore
infectedeachyear.Morethan 18millionpeoplehaveal-
readydied of AIDS-relatedillness.Action at the inter-
nationallevelto developan AIDS vaccineis crucialfor
the future, but differingexperiencesshowthat what will
reallymakea differencenow is effectiveleadershipand
societalchangeto preventthe spreadof HIV and carefor
those already infected. This can involve confronting
taboosaboutsexuality,targetinginformationand support
to high-riskgroups such as prostitutes, and providing
compassionatecare for AIDS sufferers.Brazil,Senegal,
Thailand,andUgandaallillustratewhatcan bedonewhen
there is a will to act decisively.
International actions
Actionat nationaland locallevelswilloftennot beenougJh
forrapid povertyreduction.There aremanyareasthat re-
quireinternationalaction-especiallybyindustrialcoun-
tries-to ensure gains to poor countries and to poor
peoplewithin the developingworld.An increasedfocus
on debt reliefand the associatedmoveto makedevelop-
ment cooperationthrough aid more effectiveare part of
the story.Of equalimportanceare actionsin other areas--
trade,vaccines,closingof the digitaland knowledgedi-
vides-that can enhancethe opportunity,empowerment,
and securityof poor people.
Opportunity.Within a rule-basedtradingsystem,in-
dustrialcountriescould expandopportunitiesby open-
ing their marketsmore completelyto importsfrom poor
countries,especiallyin agriculture,labor-intensiveman-
ufactures,and services.It has beenestimatedthat OECD
tariffs and subsidiescause annual lossesin welfareof
almost$20 billionin developingcountries,equivalentto
about 40 percentof aid in 1998.Many developingcoun-
triesfeelthat whiletheyareliberalizingtheirtraderegimes,
keydimensionsof the traderegimesof richcountriesare
putting them at a disadvantage.Furthermore,donor
countriescould strengthendevelopingcountries'ability
to pursue povertyreduction,by increasingaid flowsto
I2 WORLD DEVELOPMENT REPORT ZOOO/200I
countrieswith a soundpolicyenvironmentsupportiveof
povertyreductionandby financingthe EnhancedHeavily
Indebted Poor Countries Debt Relief Initiative with
funds additionalto aid budgets.
Empowerment.Global action can empower poor
peopleandpoor countriesin nationalandglobalforums.
Aid shouldbe deliveredin waysthat ensuregreaterown-
ershipby recipientcountries,and it should go increas-
ingly to country-driven, results-oriented poverty
reduction programs,developedwith the effectiveen-
gagementof civilsocietyand privatesectoragents.Poor
people and poor countriesshould havegreatervoicein
international forums, to ensure that international
priorities,agreements,and standards-such as in trade
and intellectualpropertyrights-reflect their needsand
interests.
The internationalfinancialinstitutionsand other in-
ternationalorganizationsshouldcontinuetheir effortsto
ensurefull transparencyin their strategiesand actions-
and open,regulardialoguewith civilsocietyorganizations,
particularlythoserepresentingpoorpeople.International
organizationsshould also support the ongoing global
coalitionsof poor peopleso that theymayinformglobal
debates.Actionsby multinationalcorporations,such as
adhering to ethical investmentpracticesand adopting
labor codes,can alsoempowerpoor groups.
Security.Actionsare alsoneededto reducerisksfrom
adverseinternationalforces.Jointlywith governments and
the private sector, the international financial institu-
tions must strengthenthe internationalfinancialarchi-
tectureand improveits managementto lesseneconomic
volatility,whichcan be devastatingfor poor people.In-
dustrialcountry governments,often in cooperationwith
the private sector,should alsoprovide more supportfor
international public goods-for developing and dis-
tributing vaccines for HIV/AIDS, tuberculosis, and
malariaand forproducingand disseminatingagricultural
advances for tropical and semiarid conditions.
Internationalactionto protect the environmentcan re-
duce the harmfuleffectsof environmentaldegradation,
which can be severein some poor countries.And the
international community should seek to stem armed
conflict-which affectspoorpeoplethe most-by taking
measuresto reducethe internationalarmstrade,promote
peace,andsupportphysicaland socialreconstructionafter
conflictsend.
Working together to fight poverty
The strategyin this report recognizesthat poverty is
more than inadequateincomeor humandevelopment-
it is alsovulnerabilityand a lackofvoice,power,and rep-
resentation.With this multidimensionalviewof poverty
comesgreatercomplexityin povertyreductionstrategies,
becausemorefactors-such associaland culturalforces-
need to be taken into account.
The way to deal with this complexity is through
empowermentand participation-local, national, and
international. National governmentsshould be fully
accountableto their citizenryfor the developmentpath
theypursue.Participatorymechanismscan providevoice
to women and men, especiallythose from poor and
excludedsegmentsof society.The designof decentralized
agenciesand servicesneeds to reflectlocal conditions,
socialstructures,and cultural norms and heritage.And
internationalinstitutionsshouldlistento-and promote-
the interestsof poor people.The poor are the main ac-
tors in the fightagainstpoverty.Andtheymustbe brought
centerstagein designing,implementing,and monitoring
antipovertystrategies.
There is an important role in this for rich countries
and internationalorganizations.If a developingcoun-
try has a coherentand effectivehomegrownprogramof
povertyreduction,it should receivestrong support-to
bring health and education to its people, to remove
want and vulnerability.At the same time globalforces
needto be harnessedforpoor peopleand poor countries,
so that they are not left behind by scientificand med-
ical advances.Promotingglobal financialand environ-
mental stability-and loweringmarket barriersto the
productsand servicesof poorcountries-should be a core
part of the strategy.
A divergentworld?Or an inclusiveone?Aworld with
poverty?Or a worldfreeof poverty?Simultaneousactions
to expandopportunity,empowerment,and securitycan
createa new dynamicfor changethat will makeit pos-
sibleto tacklehumandeprivationand createjust societies
that arealsocompetitiveandproductive.If the developing
world and the internationalcommunitywork together
to combinethis insightwith real resources,both finan-
cialand thoseembodiedin peopleandinstitutions-their
experience, knowledge, and imagination-the 21st
centurywillseerapid progressin the fightto end poverty.
 PART I

Framework

The Nature and Evolution
I overtyis pronounceddeprivationin well-being.
But what preciselyis deprivation?The voicesof poor
people bear eloquenttestimony to its meaning(box
1.1). To be poor is to be hungry,to lack shelter and
clothing,to be sickand not caredfor, to be illiterate
and not schooled.Butforpoor people,livingin poverty
ismorethan this.Poorpeopleareparticularly vulnerable
to adverseeventsoutsidetheir control.They are often
treatedbadlybythe institutionsof stateand societyand
excludedfrom voiceand powerin those institutions.
Poverty's many dimensions
This reportacceptsthe nowtraditionalviewof poverty
(reflected,for example,in WorldDevelopment Report
1990) as encompassingnot only materialdeprivation
(measuredbyan appropriateconceptofincomeor con-
sumption)but alsolowachievementsin educationand
health.Lowlevelsof educationand healthare of con-
cernin theirownright,but theymeritspecialattention
when they accompanymaterialdeprivation.This re-
port alsobroadensthe notionof povertyto includevul-
nerabiityand exposureto risk-and voicelessness and
powerlessness. Alltheseformsof deprivationseverelyre-
CHAPTER 1
of Poverty
strictwhatAmartyaSencallsthe "capabilities thata per-
son has, that is, the substantivefreedomshe or she en-
joysto leadthe kind of life he or shevalues."1
This broaderapproachto deprivation,by givinga
bettercharacterization of the experienceof poverty,in-
creasesour understandingof its causes.This deepertin-
derstandingbringsto the foremore areasof actionand
policyon the povertyreductionagenda(chapter2).
Anotherimportantreasonfor consideringa broader
range of dimensions-and hence a broader range of
policies-is that the differentaspectsof povertyinteract
and reinforceone another in important ways(chap-
ter 2). This means that policiesdo more than simply
add up. Improving health outcomes not only im-
proveswell-beingbut alsoincreasesincome-earning po-
tential. Increasing education not only improves
well-being-it alsoleadsto betterhealthoutcomesand
to higherincomes.Providingprotectionforpoor peo-
ple (reducingvulnerabilityin dealingwith risk)not only
makesthem feel lessvulnerable-it alsoallowsthem
to take advantageof higher-risk,higher-returnop-
portunities.Increasingpoor people'svoiceand par-
ticipationnot onlyaddressestheirsenseof exdusion--it
alsoleadsto better targetingof health and education
15
I6 WORLD DEVELOPMENT REPORT 2000/2001
Poverty in the voices of poor people
Poor people in 60 countries were asked to analyze and share
their ideasof well-being(agood experienceof life) and "ill-being"
(a bad experience of life).
Well-being was variously described as happiness, harmony,
peace,freedom from anxiety, and peaceof mind. In Russiapeo-
ple say, "Well-being is a life free from daily worries about lack of
money." In Bangladesh,"to havea lifefree from anxiety." In Brazil,
"not having to go through so many rough spots."
Peopledescribe ill-beingas lack of materialthings, as bad ex-
periences, and as bad feelings about oneself. A group of young
men in Jamaicarankslackof self-confidenceas the secondbiggest
impact of poverty: "Poverty means we don't believe in self, we
hardlytravel out of the community-so frustrated, just lockedup
in a house all day."
Although the nature of ill-beingand poverty varies among lo-
cations and people-something that policy responsesmust take
into account-there is a strikingcommonalityacrosscountries.Not
surprising, material well-being turns out to be very important.
Lack of food, shelter, and clothing is mentioned everywhere as
critical. In Kenya a man says: "Don't ask me what poverty is be-
cause you have met it outside my house. Look at the houseand
count the number of holes. Look at my utensils and the clothes I
am wearing. Lookat everythingand write what you see. What you
see is poverty."
Source.Narayan,
Chambers,
Shah,andPetesch2000;Narayan,
Patel,Schafft,Rademacher,
andKoch-Schulte
services to their needs. Understanding these comple-
mentarities is essential for designing and implement-
ing programs and projects that help people escape
poverty.
r vt'S
frmffi (A
tXi119nproach
- a.r1i.rc,,',
'l.~
Measuring poverty permits an overview of poverty that
goes beyond individual experiences. It aids the formula-
tion and testing of hypotheses on the causes of poverty.
It presents an aggregate view of poverty over time. And
it enablesa government, or the international community,
to set itselfmeasurabletargets for judging actions.In what
follows, the chapter discussesthe measurement of income
poverty and the indicators of education and health-and
then turns to vulnerability and voicelessness.
Using monetary income or consumption to identify and
measure poverty has a long tradition. Though separated
Alongside the material, physical well-being features promi-
nently in the characterizationsof poverty. And the two meld to-
getherwhen lackof food leadsto ill health-or when ill healthleads
to an inability to earnincome. Peoplespeakabout the importance
of looking well fed. In Ethiopiapoor people say, "We are skinny,"
"We are deprived and pale," and speak of life that "makes you
older than your age."
Security of income is also closely tied to health. But insecu-
rity extends beyond ill health. Crime and violence are often men-
tioned by poor people. In Ethiopiawomen say, "We live hour to
hour," worrying aboutwhether it will rain.An Argentinesays, "You
have work, and you are fine. If not, you starve. That's how it is."
Twosocialaspectsof ill-beingandpovertyalsoemerged.Formany
poorpeople,well-beingmeansthe freedom of choiceandactionand
the powerto controlone'slife. A youngwomanin Jamaicasaysthat
povertyis "like living in jail, living in bondage,waiting to be free."
Linked to these feelings are definitions of well-beingas social
well-being and comments on the stigma of poverty. As an old
woman in Bulgariasays, "To be well means to see your grand-
children happyand well dressed and to know that your children
have settled down; to be able to give them food and money
whenever they come to see you, and not to ask them for helpand
money." A Somaliproverbcapturesthe other side:"Prolongedsick-
ness and persistent poverty cause peopleto hate you."
2000.
by a century, Seebohm Rowntree'sclassicstudy of poverty
in the English city ofYork in 1899 and the World Bank's
current estimates of global income poverty share a com-
mon approach and a common method (box 1.2). Based
on household income and expenditure surveys, the ap-
has become the workhorseof quantitativepoverty
analysis and policy discourse. It has several strengths.
Because it is based on nationally representative samples,
it allows inferences about the conditions and evolution
of poverty at the national level. Moreover, since house-
hold surveyscollectinformation beyond monetary income
or consumption, the approach makes it possible to ob-
tain a broader picture of well-being and poverty, inves-
tigate the relationships among different dimensions of
poverty, and test hypotheses on the likely impact of pol-
icy interventions.
Povertymeasuresbasedon income or consumption are
not problem free. Surveydesign variesbetween countries
and over time, often making comparisons difficult. For
example, some countries ask respondents about their
food spending over the past month, while others do so
 THE NATURE AND EVOLUTION OF POVERTY ]7

Box 1.2
Measuring income poverty: 1899 and 1998

In a classicstudy first publishedin 1901, SeebohmRowntree cal- tional incomebasedon nationalaccountsdata.Thisprocedure,un-

culatedthat 10 percentof the populationof the Englishcity of York changed from past exercises,scales back income to obtaincon-
in 1899was living in poverty (below minimum needed expendi- sumption but leavesthe distribution unchanged.
tures). As we enter the next century, the World Bank calculates Prices.To compare consumption levels across countries, es-
that a fourth of the populationof the developingworld-about 1.2 timatesof pricelevelsare needed,andthe World Bank'spurchasing
billion people-is livingin poverty(below $1 a day).Thesetwo cal- power parity (PPP)estimatesfor 1993were used. Theseestimates
culationsof income poverty are separatedby a century and have are basedon new price data generatedby the InternationalCom-
very differentcoverage.Nevertheless,the basicconceptsandmeth- parison Program(ICP),which now covers 110 countries, upfrom
ods they embody have strong similarities. 64 in 1985, and a more comprehensiveset of commodities.
Povertylines.The 1990calculationsof the internationalpoverty
Rowntree'sapproach lines had to be updatedusing 1993 price data and the 1993 PPP
Rowntree's method was to conduct a survey covering nearly estimates. In 1990nationalpovertylinesfor 33 countrieswere con-
every working-classfamily in York to collect information on earn- verted into 1985PPPprices, and the most typical line among the
ings and expenditures.He then defined poverty as a level of total low-income countries for which poverty lines were availablewas
earningsinsufficientto obtainthe minimumnecessitiesfor the main- selected. In 1999the same lines were converted into 1993PPP
tenanceof "merely physicalefficiency," including food, rent, and prices, andthe new line was obtainedas the medianof the 10 low-
other items. He calculated that for a family of five-a father, est povertylines.Thatline isequalto $1.08a dayin 1993PPPterms
mother, and three children-the minimum weekly expenditureto (referredto as "$1 a day" in the text). This line has a similar pur-
maintainphysicalefficiencywas 21 shillings,8 pence;he proposed chasing power to the $1 a day line in 1985PPPprices, in terms
other amounts for families of different size and composition. of the commandover domesticgoods.The upper poverty line (re-
Comparingthese poverty lines with family earnings,he arrivedat ferred to as "$2 a day") was calculatedby doubling the amount
his poverty estimate. of the lower poverty line, as in 1990, reflecting povertylines more
commonly used in lower-middle-incomecountries.
The World Bank'sapproach Estimates for 7998.To obtain consumption levels for 1998
TheWorld Bankhas beenestimatingglobalincomepovertyfigures where surveydata were not yet available,estimated growth rates
since 1990.The latest round of estimation, in October 1999,used of per capitaprivateconsumptionfrom nationalaccountsstatistics
new samplesurveydata andpriceinformationto obtaincomparable were used to updateconsumptiondatafrom the latest surveyyear
figures for 1987, 1990,1993,1996,and 1998(thefigures for 1998 to 1998.Thismeant assumingthat the distributionof consumption
are preliminaryestimates).Themethod is the same as in past es- did not changefrom the time of the last survey to 1998.The per
timates (World Bank 1990,1996d). capita private consumption growth rates came from estimates
Consumption.Poverty estimates are based on consumption basedon the model used for other World Bank forecasts (World
or incomedata collected through householdsurveys. Datafor 96 Bank 1999j).Surveyswere availablefor 1997or 1998only for Be-
countries, from a total of 265 nationallyrepresentative surveys, larus,China,India,Jordan, Latvia,Nigeria,Pakistan,Panama,Rus-
correspondingto 88 percent of the developingworld's people,are sia,Thailand,andYemen. Sothe 1998figuresshouldbe considered
now available,upfrom only 22 countriesin 1990.Of particularnote tentative, andtrends should be interpretedcautiously,particularly
is the increase in the share of people covered in Africa from 66 in light of the controversysurroundingIndiandata (seebox 1.8later
to 73 percent, a result of extensive efforts to improve household in the chapter).
data in the region. Country-specificpovertylines.The$1 and $2 a day povertyes-
Consumption is conventionallyviewed as the preferred wel- timates described here are useful only as indicators of global
fare indicator, for practicalreasonsof reliabilityand becausecon- progress, not to assess progress at the country level or to guide
sumption is thought to better capture long-runwelfare levelsthan country policy and programformulation.Country-specificpoverty
currentincome.Where survey data were availableon incomesbut lines, reflecting what it means to be poor in each country's situ-
not on consumption, consumptionwas estimated by multiplying ation and not affected by international price comparisons, are
all incomes by the shareof aggregateprivateconsumption in na- used in country-levelanalysis.

Source:ChenandRavallion
2000.

for the past week. One-month recall data tend to result
in higher poverty estimates than one-week recall data.
Converting the information on income or consumption
collected in household surveys into measures of well-
being requiresmany assumptions,such as in deciding how
to treat measurement errors and how to allow for house-
hold size and composition in converting household data
into measures for individuals. Poverty estimates are very
sensitive to these assumptions (see, for example, the dis-
cussion in box 1.8, later in the chapter).2
Moreover,income or consumption data collectedat the
household level have a basic shortcoming: they cannot re-
i8 WORLD DEVELOPMENT REPORT 2000/2001
veal inequality within the household, so they can under-
state overall inequality and poverty. One study that dis-
aggregatedhouseholdconsumption by individualmembers
found that relying only on household information could
lead to an understatement of inequality and poverty by
more than 25 percent.3 In particular, the conventional
household survey approach does not allow direct mea-
surement of income or consumption poverty among
women. That is one reason why data on education and
health, which can be collected at the individual level,are
so valuable-they allow a gender-disaggregatedperspec-
tive on key dimensions of poverty.
A key building block in developing income and con-
sumption measures of poverty is the poverty line-the
critical cutoff in income or consumption below which an
individual or household is determined to be poor.The in-
ternationally comparable lines are useful for producing
globalaggregatesof poverty (seebox 1.2). In principle,they
test for the ability to purchase a basket of commodities
that is roughly similar across the world. But such a uni-
versal line is generally not suitable for the analysis of
poverty within a country. For that purpose, a country-
specificpoverty line needsto be constructed, reflectingthe
country'seconomicand social circumstances.Similarly,the
poverty line may need to be adjusted for different areas
(such as urban and rural) within the country if prices or
accessto goods and servicesdiffers.4 The construction of
country profiles based on these country-specific poverty
lines is now common practice.
Once a poverty line has been specified, it remains to
be decided how to assessthe extent of poverty in a par-
ticular setting. The most straightforward way to measure
poverty is to calculate the percentage of the population
with income or consumption levelsbelow the poverty line.
This "headcount" measure is by far the most commonly
calculated measure of poverty. But it has decided disad-
vantages. It fails to reflect the fact that among poor peo-
ple there may be wide differences in income levels,with
some people located just below the poverty line and oth-
ers experiencing far greater shortfalls. Policymakersseek-
ing to make the largest possibleimpact on the headcount
measure might be tempted to direct their poverty allevi-
ation resources to those closest to the poverty line (and
therefore least poor).
Other poverty measures,which take into account the
distanceof poor people from the poverty line (the poverty
gap) and the degreeof income inequalityamong poor peo-
ple (the squared poverty gap), can be readily calculated.
In comparing poverty estimates across countries or over
time, it is important to check the extent to which con-
clusions vary with the selection of poverty measure.5
Health and education
Measuring deprivation in the dimensions of health and
education has a tradition that can be traced back to such
classicaleconomists as Malthus, Ricardo, and Marx. De-
spite Rowntree's primarily income-based approach to
measuring poverty, he devoted an entire chapter of his
study to the relation of poverty to health and went on to
argue that the death rate is the best instrument for mea-
suring the variations in the physical well-being of peo-
ple.6 Classifyinghis sampleinto three groups ranging from
poorest to richest, he found that the mortality rate was
more than twice as high among the very poor as among
the best paid sections of the working classes.Calculating
infant mortality, he found that in the poorest areas one
child out of every four born dies before the age of 12
months. According to this argument, mortality could be
used as an indicator both of consumption poverty and
of ill-being in a broader sense.
The tradition of measuring deprivation in health and
education is well reflected in the international develop-
ment goals (seebox 2 in the overview). But data on these
nonincome indicators have their own problems. For ex-
ample, infant and under-five mortality rates derived
mostly from census and survey information are available
for most countries only at periodic intervals.7 A complete
vital registration system would be the best source for
mortality data, but such a systemexists in only a few de-
veloping countries. For the period between censuses or
surveys, estimates of vital rates are derived by interpola-
tion and extrapolation based on observed trends and
models, such as life tables that estimate survival from one
year to the next. Infant mortality rates are available for
most developing countries for only one year since 1990,
and the year differs because surveys are conducted at
different times. The data situation is even worse for life
expectancy, because it is often not measured directly
Education data are also far from satisfactory. The
most commonly availableindicator, the grossprimary en-
rollment rate, suffers from serious conceptual short-
comings. The greatest is that school enrollment is only
a proxy for actual school attendance. Moreover,the gross
primary enrollment rate can rise if grade repetitions in-
crease.The much-preferred net primary enrollment rate
(showingthe ratio of enrolledprimary-school-agechildren

to all primary-school-age
children)is availableforonly
around 50 developingcountriesfor 1990-97-not
enoughto makereliableaggregations byregion.Anum-
ber of ongoingsurveyinitiatives,
however,areimprov-
ing the quantityand qualityof data on health and
education.
Vulnerability
In thedimensions ofincomeandhealth,vulnerability is
theriskthata householdorindividual willexperience an
episodeof incomeor healthpovertyovertime.Butvul-
nerabilityalsomeanstheprobability of beingexposedto
a numberof otherrisks(violence,crime,naturaldisas-
ters,beingpulledout of school).
Measuring vulnerability
isespeciallydifficult:
sincethe
conceptisdynamic,it cannotbe measured merelyby ob-
servinghouseholdsonce.Only with householdpanel
data-that is, householdsurveysthat followthe same
households overseveral years-canthebasicinformation
begatheredto captureandquantifythevolatility andvul-
nerability thatpoorhouseholds sayissoimportant. More-
over,people'smovementsin and out of povertyare
informative aboutvulnerabilityonlyafterthefact.Thechal-
lengeisto findindicators of vulnerabilitythatcaniden-
tifyat-riskhouseholds and populations beforehand.
Manyindicatorsofvulnerability havebeenproposed
overthe years,but thereis now a growingconsensus
that it is neitherfeasiblenor desirableto capturevul-
nerability in a singleindicator.
If thegovernment provides
an effective workfare program,forexample,households
maydolessthantheyotherwisewouldto diversify their
incomeor buildup theirassets.Similarly, a household
that ispartofa reliablenetworkofmutualsupportmay
seelessneedfor largebufferstocksof foodor cattle.So
a vulnerability measure basedsolelyonhousehold assets-
or on incomeanditssources-maynot reflectthehouse-
hold'strueexposureto risk(box1.3).
Voicelessness
and powerlessness
Voicelessness
and powerlessness canbe measuredusing
a combination ofparticipatorymethods(box1.4),polls,
and nationalsurveyson qualitativevariables
suchasthe
extentofcivilandpoliticalliberties(box1.5).However,
measuringthesedimensions ofpovertyin anaccurate,ro-
bust, and consistentwayso that comparisonscan be
madeacrosscountriesand overtimewillrequirecon-
siderable
additionaleffortson both themethodological
anddata-gathering fronts.
THE NATURE AND EVOLUTION OF POVERTY 19
Multidimensionalityand measuringprogress
Definingpovertyas multidimensional raisesthe ques-
tionofhowto measureoverallpovertyandhowto corn-
pareachievements in the differentdimensions.One
dimension mightmovein a differentdirectionfroman-
other.Healthcouldimprovewhileincomeworsens.Or
an individualmightbe "incomepoor"but not "health
poor."Oronecountrymightshowgreaterimprovement
inhealththanin vulnerability-while anothershowsthe
converse.
Thisbringsto theforetherelativevalueof thediffer-
entdimensions: howmuchincomearepeoplewillingto
giveupfor,say,a unitofimprovement inhealthorin voice?
In otherwords,whatweightscanbe assigned to thedif-
ferentdimensions to allowcomparisons acrosscountries,
households, orindividuals andovertime?Therearenoeasy
answers.
Oneapproachto addressing comparability isto define
a multidimensional welfarefunctionor a composite
index.Analternativeis to defineaspooranybodywho
ispoorin anyoneof thedimensions-withoutattempt-
ing to estimatetradeoffsamongthe dimensions-or
anybodywhoispoorin alldimensions, andto definethe
intensityofpovertyaccordingly (box1.6).Thisreportdoes
not try to definea composite indexor to measuretrade-
offsamongdimensions. Instead,it focuseson depriva-
tion in differentdimensionsand, in particular,on the
multipledeprivations experienced by the income-poor.
Thisisa necessary firststepin developing a comprehensive
multidimensional framework.
Howshouldindicators beselected to monitorprogress?
Theinternational development goalsarea goodstarting
point.Butin practice,thesegoalswillhaveto be adapted
(bylengthening or shorteningthe timespan,forexam-
ple) and modified(increasingthe numberof dimen-
sions),dependingon context.The specificgoalswil
haveto emergefroma participatory processin whichgov-
ernmentsandcivilsocietyagreeonpriorities. Thisprocess
isalreadyunderwayin manycountries,andmultilateral
organizations arehelpingwithresources andtechnical as-
sistance(box1.7).8
Investingin measurementand monitoring
Measurements ofpovertythusmustcovermanydimen-
sions.Sofar,theincomeandconsumption dimension has
receivedmostattention.Thanksto effortsoverthepast
20yearsbysuchinternationalagencies
astheUnitedNa-
tions,the WorldBank,and the regionaldevelopment
20 WORLD DEVELOPMENT REPORT 2000/2001

Box 1.3
Measuringvulnerability

Sincevulnerability is a dynamicconcept,itsmeasurement centers sessmentof vulnerability shouldbebasednotonlyonthe ob-

on thevariability of incomeor consumption or onthevariability of servedtransfersbut also on the household'sexpectation
other dimensionsof well-being,suchas healthor housing.In aboutthe assistance it will receivein a crisis.It is this expec-
muchof the literatureon risk,this variability is measuredby the tationthat determinesthe household'sdecisionsabouten-
standarddeviationor coefficientof variationof incomeor con- gagingin other risk managementactivities.Unfortunately,
sumption.Fromthe perspective of poorpeople,this measureis household surveysrarelyincludedirectinformation onnetworks
flawedinseveralways: or on expectations of assistance.
* It givesequalweightto upwardanddownwardfluctuations, * Participation in the formalsafetynet.A household's vulnera-
Yet poor peopleare concernedprimarilywith downward bility is reducedif it is entitledto socialassistance,unem-
fluctuations. ploymentinsurance,pensions,and other publiclyprovided
* It hasnotimedimension. Given10fluctuations, thecoefficient transfers-andif it can benefitfromworkfareprograms,so-
of variationis thesamewhethergoodandbadyearsalternate cialfunds,andsimilarmechanisms. Soinformation onsuchpro-
or fivebadyearsarefollowedbyfivegoodones.Yetbunched gramsandtheirrulesof eligibilityisalsoimportantinassessing
downward fluctuationsaremoredifficultfor poorpeopleto cope vulnerability andriskexposure.
with. * Accessto creditmarkets.Similarly,a household's vulnerabil-
* A scenario withmanysmallandonelargefluctuation mayyield ity is reducedif it has accessto credit for consumption
thesamecoefficientof variationasascenariowithequalmod- smoothing.
eratefluctuations.Yet poorpeoplearelikelyto be hurtmore Clearly,assessingvulnerabilityis morecomplexthan mea-
by the first scenario. suringpovertyat a pointin time.The lengthof time overwhich
Thecoefficientof variationis,moreover,a measureafterthe vulnerabilityis to be assessedis of greatimportanceandmay
fact. Neededare indicatorsthat makeit possibleto assessa well differ acrosspeopleandcircumstances. Conventional an-
household's riskexposurebeforehand-information bothon the nualmeasuresof incomeor consumptionmay often be too
household andon itslinksto informalnetworksandformalsafety long. Furthermore,measuringvulnerabilityrequiresdata on
nets: householdassets(physical,human,andsocialcapital)in com-
* Physicalassets.A household'sphysicalassets-those that binationwith dataonformalsafetynets,the functioningof mar-
canbe soldto compensatefor temporarylossof income- kets,andthe economicpoliciesthat determinea household's
are a measureof its capacityto self-insure.What matters opportunityset andthe rangeof activitiesit canpursueto man-
is not justthe totalvalueof the assets,butalsotheir liquidity. agerisk. Manyof today'shouseholdsurveysdonot providethe
Thus knowledgeof the functioningof asset marketsis neededinformation.
needed to determine the usefulnessof the assets as Cross-sectional surveysneedto expandtheir standardex-
insurance. pendituremodulesbyaddingquestionsonassets,linkswith net-
* Humancapital.Households with limitededucationtendto be works,perceptionsof sourcesof emergencyassistance,and
moresubjectto incomefluctuationsandlessableto manage participation informalsafetynets.OneWorldBanksurveyhastaken
risk-for example,throughaccessto creditor multiplein- a stepinthisdirection: the recentLocal-Level InstitutionsSurveys
comesources. combineassetdatawith detailedquestionson households' links
* Incomediversification. Theextentof diversificationof income with localassociations.SomeLivingStandardsMeasurement
sourceshasoftenbeenusedto assessvulnerability. In rural Surveyshavealsobegunto incorporate moduleson socialcapi-
settingsanalystsmightlookat nonfarmincome,whichtends tal.Ultimately,suchenrichedcross-sectional surveysneedto be
to fluctuatelessthanfarmincome,thusproviding a measure combinedwith panelsurveys,monitoringthe samehouseholds
of protection againstweather-related risks.But incomediver- overtime,to allowdirectobservation of howhouseholds dealwith
sificationcanbea misleading indicator
of riskexposure. A sin- shocks.
gle low-riskactivitycouldbe preferable to multiplehigh-risk Vulnerability to nonincome riskscanbemeasured bythe preva-
activitiesthatarestronglycovariant.Somorediversification is lenceof theserisks(crime,naturaldisasters, andso on)in special
notnecessarily lessrisky.Diversification
needsto beevaluated modulesof household surveys. Aprogramsponsored jointlybythe
in the contextof the household's overallriskstrategy. Inter-American Development Bank,WorldBank,and Economic
* Linksto networks.Family-based networks,occupation-based Commission for LatinAmericaandthe Caribbean, the Program for
groupsof mutualhelp,rotatingsavingsandcreditgroups,and the Improvement of Surveys andthe Measurement of LivingCon-
othergroupsor associations to whicha household belongs- ditions(knownas Mecovifor its Spanish acronym) is incorporat-
all partof the household's socialcapital-canbe a sourceof ing such modulesin specificcountriesin Latin America(the
transfersin cashor kind in the eventof a calamity.An as- questionnaire canbefoundin IDB2000).
Source:Dercon1999;GroshandGlewwe2000;Holzmann
andJorgensen
1999;IDB2000;SinhaandLipton1999;WorldBank1998t.
 THE NATURE AND EVOLUTION OF POVERTY 21

Box 1.4 need to be supported. But the efforts need to go much

Measuringvoice and power using participatory farther than this, focusing on improving information on
methods education and health indicators. The Demographic and
Health Surveys need to be continued and expanded. As
IntheVoices ofthePoorstudy,insmallgroupdiscussions,poor important are effortsto expandand improvethe verysniall
peoplediscussed important
therangeof institutions intheirdaily . . .
of vulnerability and on voicel.s.
livesandthenidentifiedthecriteriathatwere importantinrat-
ing institutions.
Oncecriteriawereidentifiedandagreedon, ness and powerlessness.
groupsratedinstitutions
onthesecriteria usingpebbles,
beans,
or otherlocal material.Characteristicsincludedtrust, participa- The evolution of poverty
tion,accountability,
abilityto buildunity,responsiveness,
respect,
fairness
andcaring,andlistening andloving.Poorpeopledefined
thesecriteria
inclear
andsimpletermsbefore scoring
institutions. What are the magnitudes and patterns of poverty in the
developing world? How has poverty evolved over the
past decade?The answers to these questions are impor-
Box 1.5 tant in framing the challenge of attacking poverty.
methods
participatory
Measuringgovernance: The restof thischapter describesglobaltrendsin the
and cross-countrysurveys income (consumption), education, and health dimensions
of poverty and shows the large diversity of outcomes-
Cancountrywideinformationonvoiceandparticipation be across dimensions, regions, countries, communities,
obtainedsystematically to assesstheir rolein development g
andto comparecountries? A recentstudybroughttogether h a
adatabase covering178countries to assessthe widerissue mance reflect differences in growth, in the distribution
of governance, with voiceandaccountability measuredby of assets, in the quality and responsiveness of state insti-
indicators of civilliberties,politicalrights,the transparency tutions, in the degree of inclusiveness in societies (lower
of thelegalsystem,andthe existence of independent media. social barriers for women, ethnic minorities, and the so-
Thedatacamefromtwo typesof sources: pollsof experts
onthecountryor region(includingagencies specializing inrisk cially disadvantaged more generally), and in how coln-
rating,opinionsurveys, analysis)
andpolitical andcross-country tries and people manage risks.
surveys of residents byinternationalorganizations andNGOs. Highlighting the diversity in outcomes is important
Indicatorsfromthetwotypesof datatendto corTelate strongly, for at least two reasons. It allows the identification of
increasing confidence inthe results.Thestudyfoundastrong
positiveassociationbetweenvoiceandaccountability and successes and failures in poverty reduction and thereby
five otherclustersof governance indicators andthreedevel- enhances the understanding of what causes poverty
opmentoutcomes: percapitaincome, infantmortality, andadult and how best to reduce it. And it brings to the fore the
literacy(Kaufmann, Kraay,andZoido-Lobaton 19991. fact that aggregate trends can hide significant differences
Thestudyalsohighlighted majorweaknesses in existing
databases on voice,empowerment, andgovernance. Mar- in poverty outcomes-for different ethnic groups, re-
ginsof errorinthe resultsarewide.Significant investment gions, and sectors within a country, for example. Aware-
is neededin developing andundertaking surveys, with com- ness of these differences will help policymakers set
parablemethodsacrosscountries,to collectdataonthisim- priorities, concentrating actions where they are most
portantdimension of povertyandwell-being. National surveys
onvoiceandempowerment wouldcomplement participatory needed.
assessments. In designing the surveys,carewouldhaveto
betakento ensurethat theyarecapable of capturingdiffer- Global and regional patterns: income poverty
encesbyregion, gender,ethnicity, andsoon.Suchdifferences and social indicators
areimportant notjustinmaterialpovertybutinvoiceandem- Between 1987 and 1998 the share of the population in
powerment as well. Bten18 n 98tesaeo h ouao
developing and transition economies living on less than
$1 a day fell from 28 percent to 24 percent (table 1.1).
banks, 85 percent of the developing world's population This decline is below the rate needed to meet the inter-
lives in countries with at least two household income or national development goal of reducing extreme income
expenditure surveys. These surveys need to be improved poverty by half by 2015 (see box 2 in the overview).
greatly and made more accessible to the public. Efforts Because of population growth, the number of people
such as the Living Standards Measurement Surveys at the in poverty hardly changed. But there are large regional
World Bank and Mecovi 9 in Latin America (see box 1.3) variations in performance. East Asia and the Middle East
22 WORLD DEVELOPMENT REPORT 2000/2001

Box 1.6
Multidimensionality:dealingwith aggregation

Thereareseveralpossibleapproaches to aggregatingmeasures areasin figure).Thismethodaddsvaluebecauseit goesbeyond

of the differentdimensions of povertyandwell-being. income.But it can be criticizedbecauseit would imply,for ex-
Welfarefunction. A welfarefunctionapproach includesvarious ample,thatapersonwhohasveryhighincomebutis uneducated
dimensions of well-beinganddefinespoorpeopleasallindividuals is poor.Analternativeis to countas pooreverybody who is poor
belowa specified minimumlevelof totalwelfare(Tsui1995,1997; in all dimensions(seedarkshadedareain figure).In bothcases
Bourguignon andChakravarty 1998). Thewelfarefunction approach the complicationsof makingcomparisonsremainwhen one
allowsfor tradeoffs,usingindividuals' own choicesfor comparing wantsto measurenot only the extentbut alsothe intensityof
situations andfor assessing how muchimprovement is neededin povertyof individualswith multipledeprivations
or with depriva-
onedimension to maintainwelfareif anotherdimensionworsens. tionsin differentdimensions.
Thedifficultyisfindinga suitablewelfarefunctionfor comparisons
betweennonmarket elementsof individual welfare.Whileusinga
moneymetricandtotalexpenditure isappropriatefor assessinghow
manyadditional eggsor applesa personwouldhaveto consume Alternativeaggregationrulesto measure
to acceptlessrice,it is lessreliablefor suchimportant dimensions the multipledimensionsof poverty
of welfareassocialexclusion andpoliticalvoicelessness.Moreover,
choosing appropriate "weights"to formasingleaggregate of these H
nonmarket elementsof individualwelfarefromexistingdatahasso
far provedto be aninsurmountable challenge.
Composite index. An alternativeto using weights estimated
frompeople'sobservedchoicesis to simplyimposeweights,as
a simplistic,special-case application
of the welfarefunctionap-
proach.Therehavebeenseveralwell-knownefforts,suchasthe
physicalqualityof life indexIcombiningthe literacyrate,the in-
fant mortalityrate, andlife expectancy;Morris 1979)andthe H
human development index(UNDP1999a). Whileeasyto use,these
indexesdo not reallysolvethe intractable weightingproblembe-
causetheyassignarbitrary(usuallyequal)weightsto eachcom-
ponent(Ravallion 1997b). Y Income
Altemative aggregationrules. If the objective is to measure Note:H is the thresholddefiningthe health-poor,
and Ythat
the numberof poorpeople,anotherpossibilityis to countaspoor defining
the income-poor.
everybody
who ispoorinanyoneof the dimensions (seeallshaded

Box 1.7
Uganda's poverty reduction goals

Therecentpovertyreductionstrategypaperfor Ugandapresents * CuttingHIVprevalence by 35 percent.

a clearstatementof the povertyreductiongoalsthatthe govern- * Reducingthe incidenceof stuntingto 28 percent.
ment has set. The goals focus on reducingabsoluteincome * Reducingtotalfertilityto 5.4 birthsperwoman.
povertyto 10percentby2017andachievinguniversal primaryen- The povertyreductionstrategypaperoutlinesthe govern-
rollment(alongwith higherprimarycompletionratesandeduca- ment'sapproach to achievingthesegoals,withwell-developed in-
tionalachievement) by 2004-05.Thegovernment alsosetaseries terventionsinfourbroadareas:creatingaframeworkfor economic
of otherhumandevelopment goalsfor 2004-05: growthandtransformation, ensuringgoodgovernance andsecurity,
* Reducing the under-five
mortalityrateto nomorethan103per directlyincreasing
theabilityof poorpeopleto raisetheirincomes,
1,000livebirths. anddirectlyimprovingthe qualityof life of poorpeople.

Source:IDA2000.

and North Africahave reduced their numbers in poverty- over the decade, from 474 million to 522 million, even
EastAsia dramaticallyso. But in all other regions the num- though the share of people in poverty fell from 45 per-
ber of people living on less than $1 a day has risen. In cent to 40 percent. In Latin America and the Caribbean
South Asia, for example, the number of poor people rose the number of poor people rose by about 20 percent.
 THE NATURE AND EVOLUTION OF POVERTY 23

Table 1.1
Income poverty by region, selectedyears, 1987-98

Population
coveredby
at least Peopleliving on lessthan$1 a day
onesurvey (millions)
Region (percent) 1987 1990 1993 1996 19988
East Asia and Pacific 90.8 417.5 452.4 431.9 265.1 278.3
Excluding China 71.1 114.1 92.0 83.5 55.1 65.1
Europe and Central Asia 81.7 1.1 7.1 18.3 23.8 24.0
Latin America and the Caribbean 88.0 63.7 73.8 70.8 76.0 78.2
Middle East and North Africa 52.5 9.3 5.7 5.0 5.0 5.5
South Asia 97.9 474.4 495.1 505.1 531.7 522.0
Sub-SaharanAfrica 72.9 217.2 242.3 273.3 289.0 290.9
Total 88.1 1.183.2 1,276.4 1,304.3 1,190.6 1,198.9
Excluding China 84.2 879.8 915.9 955.9 980.5 985.7

Shareof populationliving on lessthan $1 a day

(percent)
Region 1987 1990 1993 1996 1998a
East Asia and Pacific 26.6 27.6 25.2 14.9 15.3
Excluding China 23.9 18.5 15.9 10.0 11.3
Europe and Central Asia 0.2 1.6 4.0 5.1 5.1
Latin America and the Caribbean 15.3 16.8 15.3 15.6 15.16
Middle East and North Africa 4.3 2.4 1.9 1.8 1.9
South Asia 44.9 44.0 42.4 42.3 40.0
Sub-SaharanAfrica 46.6 47.7 49.7 48.5 46.:3
Total 28.3 29.0 28.1 24.5 24.0
Excluding China 28.5 28.1 27.7 27.0 26.2
Note: The poverty line is $1.08 a day at 1993 PPP. Poverty estimates are based on income or consumption data from the countries in each
region for which at least one survey was available during 198S-98. Where survey years do not coincide with the years in the table, the
estimates were adjusted using the closest available survey and applying the consumption growth rate from national accounts. Using the
assumption that the sample of countries covered by surveys is representative of the region as a whole, the number of poor people was then
estimated by region. This assumption is obviously less robust in the regions with the lowest survey coverage. For further details on data and
methodology see Chen and Ravallion (2000).
a. Preliminary.
Source: World Bank 2000s.

Two regions fared particularly badly. In Europe and
Central Asia the number in poverty soared from 1.1 mil-
lion to 24 million. In Sub-Saharan Africa the number of
poor people increased from an already high 217 million
to 291 million over the same period, leaving almost half
the residents of that continent poor.
These variations in regional performance are leading
to a shift in the geographical distribution of poverty. In
1998 South Asia and Sub-Saharan Africa accounted
for around 70 percent of the population living on less
than $1 a day, up 10 percentage points from 1987 (fig-
ure1.1).
While these numbers provide a sense of broad trends,
they should be treated with caution in light of the short-
comings of the data mentioned above and the fact that
figures for 1998 are tentative because of the limited num-
ber of surveys available (see box 1.2).
Relativepoverty.The poverty estimates in table 1.1 are
based on a poverty line that reflectswhat it means to be
poor in the world's poorest countries (see box 1.2). f his
definition judges poverty by standards common in South
Asia and much of Sub-Saharan Africa, regardless of the
region for which poverty is being measured. An alterna-
tive definition of poverty-expounded by the British so-
ciologist Peter Townsend, among others-is a lack ofthe
resourcesrequired to participate in activities and to enjoy
living standards that are customary or widely accepted in
the society in which poverty is being measured."0
Table 1.2 presentsestimatesof poverty basedon a com-
bination of absolute and relative poverty concepts. The
24 WORLD DEVELOPMENT REPORT 2000/2001

Figure 1.1
Poverty in the developingworld is shiftingtoward South Asia and Sub-SaharanAfrica

Distribution of population living on less than S1 a day

Sub-Saharan
Africa ~ Sub-Saharan
- East Asia and
18.4% Africa Pacific
East Asia and
Pacific _
35.3% Europe and
CentralAsia 2.0%

LatinAmerica and
Europe and the Caribbean 6.5%
Central Asia0.1%
South Asia MiddleEast and
40.1% LatinAmerica North Africa
and the Caribbean 5.4% South Asia 0.5%
MiddleEast and 43.5%
North Africa
1987 0.8% 1998

Source:Chenand Ravailion2000.

poverty estimates are based on the same data and proce-
dures as those in table 1.1, but a different poverty line.
A country-specific poverty line was used, equal to one-
third of a country's average consumption level in 1993
at 1993 purchasing power parity (PPP), if that figure is
higher than the $1 a day poverty line. Otherwise, the $1
a day line was used. The effectof using a relative poverty
line-instead of the $1 a day poverty line-is that poverty
is now much higher in regions with higher average con-
sumption. It is also higher in regions with greater in-
equality. In Latin America, for example, where roughly
15 percent of the population was below the $1 a day
poverty line, more than 50 percent of the population was
under the relative poverty line. Similarly,in the Middle
East and North Africa and in Europe and Central Asia
poverty estimates are much higher by the relativepoverty
criterion. But the time trends remain unchanged."l
Socialindicators.Socialindicators in developing coun-
tries have improved on averageoverthe past three decades.
For example, infant mortality rates fell from 107 per
1,000 live births in 1970 to 59 in 1998. But the decline
between 1990 and 1998 was only 10 percent, while

Table 1.2
Relative income poverty by region, selected years, 1987-98

Regionalaverage Shareof populationlivingon lessthan

povertyline one-thirdof averagenationalconsumption for 1993
(1993 PPP (percent)
Region dollarsa day) 1987 1990 1993 1996 1998a
East Asia and Pacific 1.3 33.0 33.7 29.8 19.0 19.6
Excluding China 1.9 45.1 38.7 30.8 23.2 24.6
Europe and Central Asia 2.7 7.5 16.2 25.3 26.1 25.6
Latin America and the Caribbean 3.3 50.2 51.5 51.1 52.0 51.4
Middle East and North Africa 1.8 18.9 14.5 13.6 11.4 10.8
South Asia 1.1 45.2 44.2 42.5 42.5 40.2
Sub-SaharanAfrica 1.3 51.1 52.1 54.0 52.8 50.5
Total 1.6 36.3 37.4 36.7 32.8 32.1
Excluding China 1.8 39.3 39.5 39.3 38.1 37.0
Note:See text for a definitionof the poverty line.
a. Preliminary.
Source:Chen and Ravallion2000.
 THE NATURE AND EVOLUTION OF POVERTY 25

meeting the international development goal would have ica show that between 1989 and 1996 the incidence
required 30 percent. of poverty fell in Brazil, Chile, the Dominican Re-
Theseaggregatefiguresmaskwideregionaldisparities. public, and Honduras-and rose in Mexico and
Life expectancy in Sub-Saharan Africa in 1997 was still Republica Bolivariana de Venezuela. 1 5 In another
only 52 years-13 years less than the developing world group of countries for which only urban surveys were
average-and 25 years-a full generation-less than the available, poverty rose in Ecuador, stayed nearly un-
OECD average. One of the main causes is the still un- changed in Uruguay, and fell in Argentina, Bolivia,
acceptably high infant mortality rate in Sub-Saharan Colombia,' 6 and Paraguay.
Africa, 90 per 1,000 live births. The rate is also very In East Asia poverty trends in the 1990s were influ-
high in South Asia (77). Those levels are a far cry from enced by the impact of the recent economic crisis. In-
the OECD average of 6 per 1,000. The AIDS crisis has donesia, the Republic of Korea, and Thailand all suffered
aggravatedthe situation, leading to risinginfant mortality increases in poverty, though to differing degrees (see
in severalAfrican countries. Between 1990 and 1997 the chapter 9 ).17 In Indonesia one recent study estimated that
infant mortality rate rose from 62 to 74 in Kenyaand from poverty increased from around 11 percent in February
52 to 69 in Zimbabwe. Maternal mortality also remains 1996 to 18-20 percent in February 1999. Since then,
exceptionally high in the region: of the 12 countries in poverty appears to have declined considerably, though
the world with rates exceeding 1,000 deaths per 100,000
live births, 10 are in Sub-Saharan Africa. Table 1.3
Regional differencesare equally obvious in education Income poverty in seven African countries,
indicators. South Asia improved its gross primary en- various years
rollment rate from 77 percent to more than 100 percent Shareof population
in 1982-96. But Sub-Saharan Africa'srate remained un- below the national
changed at 74 percent (between 1982 and 1993 it actu- povertylineb
ally declined). Other education indicators confirm the Country (percent)
importanceof regionaldifferences.Almost the entire de- and perioda Area Year 1 Year 2
cline in the illiteracyrate in the developingworld has been Burkina Faso Rural 51.1 50.7
1994,1998 Urban 10.4 15.8
in East Asia. By contrast, the number of illiterate people Total 44.5 45.3
increased by 17 million in South Asia and by 3 million Ghana Rural 45.8 36.2
in Sub-Saharan Africa. 12 Sub-Saharan Africa also has the 1991/92, Urban 15.3 14.5
lowest net primary enrollment rate. 1998/99 Total 35.7 29.4
Mauritania Rural 72.1 58.9
Variations in poverty across countries 1987, 1996 Urban 43.5 19.0
Detailed studies using national income poverty lines and Total 59.5 41.3
national-level social indicators show equally large varia- Nigeria Rural 45.1 67.8
tionsipoveryperfrmance
across 1996 Urban 29.6 57.5
tions in poverty performance across countris w1992,
countries within Total 42.8 65.6
each region. Uganda Rural 59.4 48.2
In Europe and Central Asia the proportion of the 1992, 1997 Urban 29.4 16.3
population living on less than $2 a day (at 1996 PPP) Total 55.6 44.0
ranges from less than 5 percent in Belarus, Bulgaria, Zambia Rural 79.6 74.9
Estonia, Hungary, Lithuania, Poland, and Ukraine to 1991, 1996 Urban 31.0 34.0
19 percent in Russia, 49 percent in the Kyrgyz Re- Total 57.0 60.0
public, and 68 percent in Tajikistan. 13 Among seven Zimbabwe Rural 51.5 62.8
African countries with data spanning the 1990s, four 1991,1996 Urban 6.2 14.9
(BurkinaFasouNigeriat Zambiasand Zimbabwe) ex- Total 37.5 47.2
(Burkina
Nigeria,'Zambia,
F'so, and Zimbabwe) ex- a. Thedatesin this columncorrespondto year 1 andyear 2.
perienced an increase in poverty, matching the re- b. Nutrition-basedpovertylines. Comparisons
between
gional pattern gional
for thepatern
for te decade, while three
decade,while thre (Ghana,
(Ghana, countries
are not1999;
Source:Demery valid.GhanaStatisticalService1998.
Mauritania, and Uganda) had a decline (table 1.3).14S
Available national poverty estimates for Latin Amer-
26 WORLD DEVELOPMENT REPORT 2000/200I
it is still substantially higher than precrisis levels.'8
Trends in China in 1996-98 are sensitive to the choice
of welfare measure. Income poverty measures based on
the $1 a day or national poverty line show continued de-
cline. But a consumption-based poverty measure shows
a stalling in poverty reduction between 1996 and 1998,
suggestingthat poor households, especiallyin rural areas,
have been saving an increasingly large share of their in-
comes.19 The most recent data for Vietnam show that
between 1993 and 1998 the incidence of poverty, based
on a national poverty line, fell from 58 percent to 37
percent.20
Poverty reduction also varied in South Asia in the
1990s. Bangladeshturned in a good performance despite
its worst floods in living memory, with GDP growth of
4.5 percent in 1998-99, thanks to a bumper ricecrop after
the floods.The concerted reliefefforts by the government,
NGOs, and donors-and the ongoing food-for-work
programs-limited the loss of life and the impact of the
floods on poverty. Pakistan and Sri Lanka made little or
Box 1.8
Tracking poverty in India during the 1990s
Recent datafrom India'sNational SampleSurveys(NSS)suggest
that the pace of poverty reduction slowed in the 1990s, particu-
larly in ruralareas.This occurredagainsta backdropof strong eco-
nomic growth (GDPgrowth of 6.1 percent a year during 1990-98),
according to the nationalaccounts (NAS).There are signs of ris-
ing inequalitynationallyin the NSS data, due in large part to ris-
ing averageconsumption in urban areas relative to rural areas,
though with some signs of higher inequalityin urbanareas. How-
ever,an importantfactor in the slow rateof poverty reductionwas
slow growth in averageconsumption, as measured by the NSS.
Closerexaminationshows that NSSconsumptionis anincreas-
inglysmallerfractionof privateconsumptionas estimatedin the NAS.
NSSconsumptionhasdeclinedrelativeto NASconsumptionduring
the pastthree decades;the two were much closerin the 1950sand
1960s(Mukherjeeand Chatterjee1974).If the averageconsumption
figures from the NSSare replacedby the averageconsumptionfig-
ures from the NAS,and everybody'sconsumptionis adjustedpro-
portionately, poverty would show a downward trend during the
1990s(asfound by Bhalla2000).
But comparingNSSand NAS data is a complexmatter, involv-
ing differencesin coverage,recallbiasesin the NSS,price imputa-
tions (for example, for home-producedconsumption and in-kind
wages in the NSSand for nonmarketedoutput inthe NAS),and sam-
plingand nonsamplingerrorsin both.Thus,without examiningwhy
the differencesbetweenthe two havewidened, adjustingthe NSS
meanupwardto equalthe NAS meanwould be an arguableproce-
dure.Foronething, it is not clearwhy the averageconsumptiondata
from the NSSwould be wrong but not the inequalitydata,the as-
no progress in poverty reduction in the 1990s.21 For
India, there is an ongoing debate on the accuracy of the
statistics. It provides a telling example of how difficult it
is to track poverty over time, even within countries
(box 1.8).
Variations in poverty within countries
Country aggregatesof differentdimensions of poverty pro-
vide a useful overview of performance. But they hide as
much as they reveal.There are distinct patterns of poverty
within countries, and different groups within a country
can become better or worse off.
Poverty in different areas within a country can-and
does-move in different directions. In Burkina Fasoand
Zambia rural poverty fell and urban poverty rose, but the
urban rise dominated and overall poverty rose (see table
1.3).22 In Mexico, while overall poverty declined-
though modestly-between 1989 and 1994, there were
large variations across regions within the country.23 In
China rapid income growth has been accompaniedby ris-
sumption made when everybody'sconsumption is adjusted pro-
portionately.For example, it cannot be ruled out a priori that non-
responseand nonsamplingerrors in measuringconsumptionmay
differamongincomegroups.Also,Visania 12000)findsthe differences
betweentheNSSandNAStobeconsiderablylessifoneweekrather
than one month is used in the NSSas the referenceperiodfor con-
sumption.Srinivasan(2000)presentsa detaileddiscussionof these
issues (Srinivasan and Bardhan1974present earlierdiscussionsof
these issues.)
There is also evidencethat part of the observedtrend in rural
poverty in the earlier part of the 1990smay result from using in-
adequateprice deflatorsfor ruralareas.As a result, "it is likely that
the decline in ruralpoverty rates has been understated in the of-
ficial poverty counts. Indeed, we are led to suggest as a working
hypothesisthat, between1987-88and 1993-94,therewas no great
differencein the rate of decline of urban andrural poverty,at least
according to the headcountmeasure" (Deatonand Tarozzi1999,
pp. 34-35).
It is plausible that the NSS-basedpoverty numbers are un-
derestimatingthe rateof poverty reductionin India.The issues in-
volved are important not only because of the Indian poverty
figures' weight in global poverty trends, but aiso because similar
problems are likely to arise elsewhere. India has a stronger sta-
tistical tradition than most poor countries. And it is not simply a
matter of getting accurateestimates of poverty. Suchsurveys are
a key resource for identifying the characteristicsof poor people
and thus are a vital input for focusing policy. Researchin this area
is a high priority.

ing inequalitybetween urban and rural areas and between
provinces.2 4
Poverty tends to be associated with the distance from
citiesand the coast,as in China, Vietnam, and Latin Amer-
ica.25In China many of the poor reside in mountainous
counties and townships. In Peru two-thirds of rural
households in the poorest quintile are in the mountain
region,while fewerthan a tenth are in the coastalregion.26
In Thailand the incidence of poverty in the rural north-
east was almost twice the national average in 1992, and
although only a third of the population livesthere, it ac-
counted for 56 percent of all poor.
Differences in health and education between
low- and high-income households
Social indicators in many countries remain much worse
for the income-poor than for the income-nonpoor-
often by huge margins.In Mali the differencein child mor-
tality rates between the richest and poorest households
is equal to the average gain in child mortality rates
recorded over the past 30 years.27 In South Africa the
under-fivemortality rate for the poorest 20 percent is twice
as high as the rate for the richest 20 percent, and in
Northeast and Southeast Brazil, three times as high.
The picture is the same for malnutrition. A study of
19 countries found that stunting (lowheight for age-an
indicator of long-term malnutrition), wasting (lowweight
for height-an indicator of short-term malnutrition),
and being underweight (low weight for age) are higher
among poor people in almost all countries.2" But the dif-
ferencesbetween poor and nonpoor tend to be smaller in
countries with high averagerates of malnutrition.2 9
The incidence of many illnesses, especially commu-
nicable diseases,is higher for poor people, while their ac-
cess to health care is typicallyless. In India the prevalence
of tuberculosisis more than four times as high in the poor-
est fifth of the population as in the richest, and the preva-
lence of malaria more than three times as high.30 In 10
developing countries between 1992 and 1997, only 41
percent of poor people suffering from acute respiratory
infections were treated in a health facility, compared
with 59 percent of the nonpoor. In the same period only
22 percent of births among the poorest 20 percent of peo-
ple were attended by medically trained staff, compared
with 76 percent among the richest 20 percent.31 Al-
though HIV/AIDS initiallyaffected the poor and the rich
almost equally, recent evidence indicates that new infec-
tions occur disproportionately among poor people.
THE NATURE AND EVOLUTION OF POVERTY 27
Similar disparities show up in access to schooling
and in educational achievement. In some poor countries
most children from the poorest households have no
schoolingat all. A study of Demographic and Health Sur-
vey data found 12 countries in which more than half the
15- to 19-year-olds in the poorest 40 percent of house-
holds had zero years of schooling: Bangladesh,India, Mto-
rocco, Pakistan, and eight countries in Sub-Sahatan
Africa. In contrast, the median number of years corm-
pleted by 15- to 19-year-olds in the richest 20 percent
of households was 10 in India and 8 in Morocco. In other
countries the gap in educational achievement was much
smaller: one year in Kenya, two in Ghana and Tanzania,
and three in Indonesia and Uganda.32 In Mexico aver-
ageschooling was less than 3 yearsfor the poorest 20 per-
cent in rural areas and 12 years for the richest 20 percent
in urban areas.
Primary enrollment rates show similar gaps. The en-
rollment rate for 6- to 14-year-olds is 52 percentage
points lower for the poorest households than for lthe
richest households in Senegal,36 percentage points lower
in Zambia, and 19 percentagepoints lowerin Ghana. The
gaps are also large in North Africa (63 percentage points
in Morocco) and South Asia (49 percentage points in
Pakistan).3 3
Within-country differences in social indicators also
exist between urban and rural areas, across regions, and
across socioeconomic classes. In China there has been
a widening rural-urban gap in health status and health
care use. While the rural population's use of hospital ser-
vices declined 10 percent between 1985 and 1993, the
urban population's increased by 13 percent.3 4 In Rus-
sia the increase in mortality during the transition has
been concentrated among younger males, and stunting
of children, relatively high for an industrialized coun-
try, has been most prevalent in rural areas and among
poor people.3 5
Gender disparities
One of the key variations within a country is the differ-
ent achievement of women and men. The allocation of
resourceswithin households varies depending on the age
and gender of the household member. But estimating the
number of poor men and women independently is dif-
ficult, if not impossible, because consumption data are
collected at the household level.36 Even so, available
health and education data indicate that women are often
disadvantaged.
28 WORLD DEVELOPMENT REPORT 2000/2001
A recentstudyof 41 countriesshowsthat femaledis-
advantage,definedas the gap betweenmaleand female
primaryenrollmentrates,variesenormously.In Benin,
Nepal,and Pakistanthe male-femalegapin the primary
enrollmentrate is more than 20 percentagepoints, and
in Morocco,18.But in Brazil,Indonesia,Kenya,Mada-
gascar,the Philippines,and Zambia the enrollmentrates
of boysand girlsare almost the same.3 7 The gendergap
in educationis often lowerforthe richesthousholdsand
highestfor the pooresthouseholds.In India the gender
gap in enrollmentratesis 4.7 percentagepoints forchil-
dren fromthe wealthiest20 percentof households,com-
pared with 11 percentagepoints for childrenfrom the
poorest20 percentof households. 38
Disparities by caste, ethnicity,
and indigenous status
There mayalsobe groupsthat faceparticularsocialbar-
riers.Disadvantagedin many developingand developed
countries and transition economies,ethnic minorities
and racial groups often face higher poverty. 3 9 The in-
digenouspopulationshavea much higherincidenceofin-
comepovertyin a sampleof LatinAmericancountriesfor
which data are available. 4 0 Schoolingattainments for
these disadvantagedgroupsare alsolowerthan for other
groups.Theindigenousgroupsin Guatemalahave1.8years
of schooling,and the nonindigenous4.9 years.4' In Peru
indigenouspeoplewere40 percentmorelikelyto be poor
thannonindigenousgroupsin 1994and 50 percentmore
likelyin 199742In rural Guatemalachildrenof indige-
nousmothersaremorelikelythan thoseof nonindigenous
mothersto be stunted.4 3 In the inner citiesof the United
Stateswhitemarriedcoupleshavean incidenceof poverty
of 5.3 percent, while black or Hispanic single-mother
householdshavean incidenceof more than 45 percent. 44
Evidencefor India showsthat scheduledcastesand
scheduledtribes face a higher risk of poverty. 4 5 These
are among the structural poor who not only lack eco-
nomic resourcesbut whosepoverty is stronglylinked
to social identity, as determined mainly by caste.4 6
They also have worse social indicators. Among rural
scheduled caste women in India the literacy rate was
19 percent in 1991, half that for the country, and
among scheduled caste men, 46 percent, compared
with 64 percent for the country.47 When severaldis-
advantagesare combined-being a woman from a so-
cially excluded group in a backward region-the
situation is worse. In Uttar Pradesh, one of India's
poorest states, only 8 percent of rural scheduledcaste
women are literate, a third the rate for rural women in
Uttar Pradesh. But new researchsuggeststhat literacy
rates of rural scheduled caste women are on the rise
acrossIndia. Although only 31 percent of rural sched-
uled casteor scheduledtribe girls in the primaryschool
age group were enrolledin schoolin 1986-87, 53 per-
cent were by 1995-96.48
Volatility at the household level
Studiesof income povertychangesfor the same house-
holds over time showsignificantmovementin and out
of poverty.Whilesomegroupsarechronicallybelowthe
povertyline, other groupsfacea high risk of fallinginto
povertysome of the time. Studiesfor China, Ethiopia,
Russia,and Zimbabwefind that the "alwayspoor" group
is smallerthan the "sometimespoor" group.49 However,
these resultsshouldbe treatedwith cautionbecauseob-
servedchangesreflectmeasurementerrorsaswellas real
changes. 50
One immediate question is whether some types of
householdsare morelikelyto sufferfromchronic(rather
than transitory)poverty.The answerdiffersfrom coun-
try to country,but assetholdingsoften play a keyrole.
In Chinaa lackof physicalcapitalisa determinantof both
chronic and transitorypoverty,but householdsizeand
educationof the headof householddeterminethe like-
lihood of chronicbut not of transitorypoverty. 5'
In the transitioneconomiesofEuropeand CentralAsia
economicmobilityhas increased,but chronicpovertyis
emergingas a keyissue. 52 Whethera householdjoins the
ranks of the new poor or the new rich depends very
much on its characteristics,especiallyits links with the
labor market. The transition has increasedthe disad-
vantageof "oldpoor" (pensioners, familieswithlargenum-
bersof children,and single-parentfamilies)and givenrise
to "newpoor" (long-termunemployed,agriculturalwork-
ers,youngpeoplein searchof their firstjob,and refugees
5 3 In Poland the chronically
displacedby civil conflict).
poor constitute a distinct segment of the population.
Largerhouseholds,thoseworkingon farms,and house-
holdsdependenton socialwelfareare mostat riskof stay-
ing poor.5 4 Russiahas seen the emergenceof new poor
during the transition. In the early 1990snew groupsof
poor formedas a result of the erosionof realwagesand
pensionsandthe impactof unemployment, 5 5 andpoverty
is becominglongerin termandmoreresistantto economic
recovery.56

* *
This chapter has shown that progress in income
povertyreductionand humandevelopmentvarieswidely
acrossregions,countries,and areaswithin countries.It
has alsoshown the existenceof significantgaps in per-
formanceby gender,ethnicity,race,and socialstatus.
Much of the differencein performanceacrossregions
and countriescan be attributedto differencesin economic
growth(chapter3).The growthcollapsesin many coun-
tries in Africa and the former Soviet Union had a
devastatingimpacton poverty.The economywidecrises
and natural disastersin EastAsia,Latin America,Sub-
SaharanAfrica,and Europeand CentralAsiaalsoled to
important setbacksin povertyreduction (chapter9). By
contrast,the spectaculargrowthperformancein China
resultedin a sharpdrop in incomepoverty.In the restof
EastAsia,despitethe financialcrisis,steadygrowthrates
also translatedinto significantlylowerpoverty over the
1990s.
But the initialinequalitiesand the pattern of growth
also account for the differences in performance in
povertyreductionin its multipledimensionsassomege-
ographic areas and social groups are left behind. In
somecasesinitialdifferencesinclude unequal accessto
assets,markets, and infrastructureand an unevendis-
tributionof skills(chapters3, 4, and 5).The differences
in health and education among and within countries,
THE NATURE AND EVOLUTION OF POVERTY 2')
for example,also reflectthe extent to which state in-
stitutions are responsiveand accountableto poor peo-
ple (chapter 6). In other cases social barriers linked
with gender,ethnicity,race, and socialstatus help per-
petuateincomepovertyand low levelsof healthand edL-
ucation amongthe sociallydisadvantaged(chapter7).
Policybiasesagainstlabor-intensivesectorssuch as agri-
culture and light manufacturingat the national (chap-
ter 4) or international(chapter10) leveland skill-biased
technologicalchange(chapter4) can result in lowerre-
ductions in income poverty at similar growth rates.
This chapter has also noted that there can be large
volatilityin incomesof households.This brings to the
forethe importanceof understandingthe sourcesof risk
that householdsfaceand the mechanismsbestsuited to
managingthose risks(chapters8 and 9).
Finally,this chapterhas arguedthat the experienceof
povertygoesbeyondmaterialdeprivationand lowlevels
of health and education.The inabilityto influencetlhe
decisionsthat affectone'slife, ill treatment by state in-
stitutions,and the impedimentscreatedby socialbarri-
ers and norms are alsodimensionsof ill-being.Another
is vulnerabilityto adverseshocks,natural disasters,dis-
ease,and personalviolence.This broaderconceptionof
povertyleadsto a deeperunderstandingof its causesand
a broader range of actions for attacking it. These are
outlinedin chapter2 anddevelopedin moredetailin sub-
sequent chapters.
 CHAPTER 2

Causes of Poverty
and a Framework
for Action
That reportalsonoted that theseeffortshad to be com-
FromWorldDevelopment Report1990... plementedby safetynetsforpeopleexposedto shocks
- Labor-intensive
growth and unableto benefitfromthe strategy.Butsafetynets
* Broadprovisionof socialservices wereclearlyseen as playinga supportingrole for the
to WorldDevelopment
... Report2000/2001 two mainpartsof the strategy.
Opportunity The 1990report'sframeworkforactionwasderived
* Empowerment from its conceptof poverty,its analysisof the causes
* Security of poverty,the experienceof the 1970sand 1980s,and
the state of the world economy at the end of the
1980s.It viewedpovertyas lowconsumptionand low
achievement in education and health. Economic
A decade ago WorldDevelopmentReport1990 pre- development-brought about essentiallyby liberaliz-
sented a two-partstrategyfor povertyreduction: ingtradeand markets,investingin infrastructure,and
providingbasic socialservicesto poor people,to in-
Countriesthat have beenmostsuccessfulin attacking creasetheir human capital-was seen as key to re-
povertyhaveencouraged a patternof growththat makes ducingpoverty.
efficientuse of laborand haveinvestedin the human The experiencethat definedthe 1990report,from
capitalof thepoor.Bothelementsare essential.Thefirst its vantageof 1989,wasthe contrastingexperiencein
providesthe poor with opportunities to use their most the 1970sand 1980sof EastAsia,wherepovertyhad
abundant asset-labor. The second improves their fallensharply,and of Africa,LatinAmerica,and South
immediatewell-being and increases
theircapacityto take Asia,wherepovertyhad declinedlessor evenrisen.Why
advantageof the newly createdpossibilities.Together, did Indonesia outperform Brazilin the 1970s and
theycanimprovethe livesof mostof the world'spoor. 1980sin reducingincomeandnonincome(education
-World Bank1990(p.51) and health) poverty?The answerwas labor-intensive
31
32 WORLD DEVELOPMENT REPORT 2000/200I

growth and broad provisionof socialservices-the report's
two-part strategv.
This report uses new evidence and multidisciplinary
thinking that together broaden the choices for devel-
opment action to reduce poverty in its multiple di-
mensions. The evidence confirms that economywide
growth improves the incomes of poor people-and in
the longer run reduces nonincome poverty. And ex-
panding the human capabilities of poor people re-
mains central in any poverty reduction strategy, both
for the intrinsic value of such capabilities as health
and education and for their instrumental contribu-
tion to other dimensions of well-being, including
income.
But the experiences of the 1990s show that:
* Growth cannot be switched on or off at will. Mar-
ket reforms can indeed boost growth and help poor
people, but they can also be a source of dislocation.
The effects of market reforms are complex, deeply
linked to institutions and to political and social
structures. The experience of transition, especiallyin
the countries of the former Soviet Union, vividly il-
lustrates that market reforms in the absence of ef-
fectivedomestic institutions can fail to delivergrowth
and poverty reduction. Furthermore, there is evi-
dence that technological change in the past decade
has been increasingly biased toward skills. So in con-
trast to what was expected and needed, the pattern
of growth in developing countries is not necessarily
intensive in unskilled labor.
* The emphasis on social services for building human
capital was perhaps too optimistic about the institu-
tional, social, and political realities of public action.
Public investment in basic education and health in de-
veloping countries has been rising-though in some
countries more slowly than GDP, suggesting a pos-
sible lack of commitment to expanding social services.
In many countries social spending is regressive.More-
over, such investment has been less effective than ex-
pected, in part because of serious problems in quality
and in responsiveness to poor people's needs-
institutional failures highlighted in the Voicesof the
Poor study and other research. But there have been
successeseven in seeminglydifficult conditions of low
resources. Experience and research show that effec-
tivenessin servicedeliveryis highly dependent on local
institutional capabilities, market structure, and pat-
terns of political influence.
* There is a powerful case for bringing vulnerability
and its management to center stage. Participatory
poverty work underlines the importance of vulnera-
bility to economic, health, and personal shocks. So do
the financial crisesof the 1990s-not leastin EastAsia,
the shining example of success in development and
poverty reduction-and the sequence of devastating
natural disasters.

BRAZIL

Brazilhasmade impressiveimprovements
in social indicators. Net enrollment in pri-
maryeducationincreasedfrom 88.2 percent
in 1992to 97.1 percentin 1997.Infantmor-
tality fell from 62 per 1,000live births in the
mid-1980s to 38 in the mid-1990s. And
much urban infrastructure helps poor peo-
ple. New programs guarantee minimum
per capitaspendingfor basichealthcare and
minimum per student spending in primary
schools. Innovative action to get children
intoschoolincludesthe BolsaEscola,which
gives poor families grants if their children
go to school.
Despite the advances,the inequalities
in health and education remaingreat, with
the poorest fifth of the populationhaving
three years of education,and the top fifth
more than nine years. The income-poor
still leaveschool with skills inadequatefor
a middle-income countryintegratedwith the
global economy. And reducing income
poverty hasproved difficult. Indeed, in the
unstable macroeconomicenvironment of
the 1980s and early 1990s, poverty rose.
Two recentevents confirm that the groups
most vulnerableto economicinsecurityare
those with the highestincidenceof poverty.
Drought in the Northeast hit poor rural
workers severely,and the ripple effects of
the EastAsiancrisis, though more benign
than expected, reduced the income of
workers with the least education.
Some illustrative priorities for action:
job growth throughproductiveinvestment
and prudentmacroeconomicmanagement
is clearly centralto increasingincome op-
portunities. But unlessstructural inequali-
ties are tackled effectively, the gains for
poor people will be modest. To reduce
structural inequalities,a large land reform
programis underway, andthere have been
promising experiments in negotiated land
reform in the Northeast. In the ongoing
education effort the next steps will proba-
bly requireeven broader,deeper,and more
participatory reforms-many of these are
now under discussion.The government is
also continuing to ease the constraints of
constitutionalentitlements,which limit the
room for maneuveron public spending. Fi-
nally, successfulcommunity-drivendevel-
opment approaches-in urban upgrading,
small-farm investments, and community
health agents-show what is possible
when there is anempowering mobilization
of citizens.
 CAUSES OF POVERTY AND A FRAMEWORK FOR ACTION 3

* Inequality is back on the agenda-in the realm of
ideas and experience and in the political discourse of
many developing (and developed) countries. New
work shows the importance of gender, ethnic, and
racial inequality as a dimension-and a cause-of
poverty. Social, economic, and ethnic divisions are
often sources of weak or failed development. In the
extreme, vicious cycles of social division and failed
development erupt into internal conflict, as in Bosnia
and Herzegovina and Sierra Leone, with devastating
consequences for people.
* The global forces of integration, communication, and
technological advancehave proceeded apace, bringing
significantadvancesto some. But they have passed oth-
ers by. Private capital flows now dominate official
flowsin the world, but they reinforcepositiveeconomic
developments, either neglecting or punishing coun-
tries with weak economic conditions.
The new evidence and broader thinking do not negate
earlier strategies-such as that of WorldDevelopmentRe-
port 1990. But they do show the need to broaden the
agenda.Attacking povertyrequiresactions beyond the eco-
nomic domain. And public action has to go beyond in-
vesting in social services and removing antilabor biases
in government interventions in the economy.
Acknowledgingthe need for a broader agenda, this re-
port proposes a general framework for action in three
equally important areas:
* Promoting opportunity:expanding economic oppor-
tunity for poor people by stimulating overall growth
and by building up their assets and increasing the re-
turns on these assets, through a combination of mar-
ket and nonmarket actions (part II).
* Facilitating empowerment: making state institutions
more accountable and responsive to poor people,
strengthening the participation of poor people in po-
litical processes and local decisionmaking, and re-
moving the social barriers that result from distinctions
of gender, ethnicity, race, and social status (part III).
* Enhancingsecurity: reducing poor people'svulnerabiliry
to ill health, economic shocks, policy-induced
dislocations, natural disasters, and violence, as well as
helping them cope with adverse shocks when they
occur (part IV).
Opportunity, empowerment, and security have in-
trinsic value for poor people. And given the important
complementarities among them, an effectivepoverty re-
duction strategy will require action on all three fronts, by
the full range of agents in society-government, civil
society, the private sector, and poor people themselves.
Actions cannot be confined to individual countries
in the developing world. Harnessingglobal forcesin favor
of poor countries and poor people will be essential. Ac-
tions are needed to promote global financial stability--
and to ensure that poor countries are not left behind by
advances in technology and in scientific and medical re-
search. The markets of rich countries must be opened
to the products of poor countries, and aid and debt re-
lief must be increased to help poor people help them-
selves. And poor countries and poor people need to be
given a voice and influence in international forums
(part V).

CHINA

Chinastandsoutforitsextraordinary decline areas,butmicroevidencesuggeststhatvii- slowdown, insecurityinareasdependent on

in incomepovertyand its highlevelsof
educationandhealth.Butit hasalsohadsig-
nificantincreasesin inequality-between
townandcountry, andbetweenthe coastal
areasandinlandChina,withthepoor,semi-
aridareas of inlandChinaparticipating lit-
tie ingrowth.
Formalstructuresof securityareintran-
sition,and thereare deepconcernsabout
the lessdynamicparts of urbanChina,
whichareexperiencing the beginnings
of
a majorshakeoutin stateenterpriseand
governmentemployment.Formalprovi-
sionof securitywasalwaysweakerinrural
lagemechanisms continuetoprovidehigh nowinefficient state productioncouldbe
levelsoffoodsecuritythroughlandalloca- acute. Second,the smoothintegration of
tion processes-confirmedas politically Chinaintothe globaltradingsystemwillbe
popularin villagesby democraticvotes. keyinlockinginreformsandensuringeco-
Ensuringthe voiceofthenewpoorinurban nomicstability andsteadyjobgrowth.But
areasandthoseleftbehindininlandChina ifnew opportunities are notto widendis-
willbe importantinguidingaction. parities,thiswillhavetobe accompanied by
Threeareasfor actionare illustrative. greateremphasisonbuilding the assetsof
rapidgrowththroughhigh peoplelivinginthe poorerareas.Third,con-
First,maintaining
nonstateinvestment is crucialif thereisto tinuingarea-basedintegratedruraldevel-
bea smoothprocessof job destructionin opmentactivitiesin poorareasof inland
inefficient
stateactivities
andsmoothreform Chinaand,more generally,balancingin-
of the socialprotectionarrangements for vestmentacrossgeographic areasneedto
state employees.If there is a sustained beimportantpartsof anyoverallstrategy.
34 WORLD DEVELOPMENT REPORT 2000/200I

The causes of poverty dependson a legalstructurethat definesand enforcespri-

One route for investigatingthe causesof povertyis to ex-
amine the dimensionshighlightedby poor people:
* Lackof incomeand assetsto attain basicnecessities-
food,shelter,clothing,and acceptablelevelsof health
and education.
* Senseof voicelessness and powerlessness in the insti-
tutions of stateand society.
* Vulnerabilityto adverseshocks,linked to an inabil-
ity to copewith them.
To understandthe determinantsof povertyin all its
dimensions,it helps to think in terms of people'sassets,
the returns to (or productivityof) these assets,and the
volatilityof returns.These assetsare of severalkinds:
* Humanassets,suchas the capacityforbasiclabor,skills,
and good health.
* Naturalassets,suchasland.
* Physicalassets,such as accessto infrastructure.
* Financialassets,such as savingsand accessto credit.
* Socialassets,such as networksof contactsand recip-
rocalobligationsthatcanbecalledon in timeofneed,
and politicalinfluenceoverresources.
The returnsto theseassetsdependon accessto mar-
ketsandalltheglobal,national,and localinfluences
on
returnsin thesemarkets.But returnsdependnot just on
the behaviorof markets,but alsoon the performanceof
institutionsof state and society.Underlyingassetown-
ershipandreturnsto assetsarenot onlyeconomicbut also
fundamentalpoliticaland socialforces.Accessto assets
vate property rightsor on customarynorms that define
commonpropertyresources.Accessmayalsobe affected
byimplicitor explicitdiscriminationon the basisof gen-
der,ethnicity,race, or socialstatus.And both accessto
assetsand returns to assetsare affectedby public policy
and stateinterventions,which are shaped by the politi-
calinfluenceof differentgroups.
Alsoimportant is the volatilityof returns.Volatility
resultsfrom market fluctuations,weather conditions,
and, in some societies,turbulent politicalconditions.
Volatilityaffectsnot only returns, but alsothe valueof
assets,as shocksunderminehealth, destroynatural and
physicalassets,or depletesavings.
Lack of income and assets
If you have a job at all now, you're overworkedand
underpaid.
-- VUUngI'omain fi7omDiffi-rozgra6d
Bkdgarzia
Some have land, but they can't buy fertilizer; if some
work as weavers,they aren't wellpaid; if someworkfor
daily wages,they aren't paid a just wage.
--(.ackchiquel
hIdian,(GuatemlaL
Poor people consistentlyemphasize the centrality of
work to improvingtheir lives.A countty'soverallwealth
is an important influence on this: as countries grow
richer,so on averagedo poor people in those countries,

INDIA

Indiasuffersseveredeprivationsin edu- sial becauseof measurementproblems in mostof India.In areaswith deepdepri-

cationandhealth-especiallyinthe North,
wherecaste,class,andgenderinequities
areparticularlystrong.In studiesin Bihar
andUttar Pradesh poorwomenandmen
emphasized theirextremevulnerabilityand
the ineffectivenessof state institutions,
fromschoolsto police.
In the past,povertyreductionin India
laggedbehindthatof EastAsiabecause of
slower growth and significantly less
progressinpromotingmasseducation and
basic health. More recently,however,
growthhas accelerated and povertyhas
fallen,althoughtheactualimpactof growth
on povertyreductionremainscontrover-
(seebox 1.8). vationin healthandeducation, the devel-
Therearealsomarked differenceswithin opmentof socialinfrastructure is critical.
India-withthe South,particularly thestate Expanding educationand healthservices
of Kerala,having sharplybettereducation and will requirethatstategovernments reverse
health.Keralahaslifeexpectancies greater thedeterioration intheirfiscalpositions, as
than thosein Washington,D.C.,despite subsidies to the loss-making powersector
vastlylowerincome levels.
Theeffectiveness crowdout spendingin the socialsectors.
of publicactioninKerala hasbeenattributed The higher spending will need to be
to its strongtraditionof politicalandsocial matched by betterservice provision.Thiswill
mobilization. requiredeepimprovements ingovernance,
Whatarethe priorities foractionin India? oftenweakestinIndia'spoorestregions, and
Accelerated povertyreductionwill require in combatingteacherabsenteeism. Also
fastergrowth,whichin turn demandslib- neededismoreequitable serviceprovision,
eralization,
especially inagriculture,andbet- whichwill requireempowering womenand
ter provisionof infrastructure,
sorelylacking membersof lowercastes.
 CAUSES OF POVERTY AND A FRAMEWORK FOR ACTION 35

with the main mechanism being better-paid work. Figure 2.1

With economic growth, income poverty falls; with Poverty shows a strong link with economic
economic contraction, income poverty rises (figure contractionsand expansions
2.1). Some countries in EastAsia sustainedper capita Annualchangein povertyrate
GDP growth rates of 4-5 percent over four decades, Percent
with massive improvements in living standards and in 15
health and education for poor people and for everyone
else. Other countries, most in Africa, registered nega- 10
tive growth or no growth at all over the same period,
delivering no improvements even in average living
standards.
While economic growth is systematically associated .
with poverty reduction, the rate at which growth trans- 0
lates into lowerpoverty depends on the initial level of in-
equality in the distribution of income and how that -5

distribution changes over time. Growth-and its effec-

tivenessin reducing poverty-also depends on sound, sta-
ble governance.So confronting socioeconomicinequalities
and building sound institutions can be important both
forprovidinga sociallysustainablebasisforoverallgrowth
and forensuringthat poorpeoplegainsubstantiallyfrom
that growth.
Voicelessnessand powerlessness-
the institutional basis ofpoverty
Those materially deprived feelacutely their lack of voice,
power, and independence (see box 1.1 in chapter 1).
This helplessness subjects them to rudeness, humilia-
tion, shame, inhumane treatment, and exploitation at the
10
-10 - o 5 10
Percent
Note:Thedata referto 150country-levelspells of povertychange
and meanincomechange from povertysurveysin 1980-98. Each
point representsone-fourthof the sample,orderedfrom strongest
contractionto strongestexpansion.
Source:WorldBankstaffestimates basedonChenandRavallion (20001.
hands of the institutions of state and society (box 2.1).
Absence of the rule of law, lack of protection against vi-
olence, extortion and intimidation, and lack of civilityand
predictabilityin interactionswith publicofficials-all these

JORDAN

Poverty and inequalityin Jordan increased disabled,microcredit,and healthinsurance ries.Its 1998servicedeliverysurveyreflects

at the end of the 1980s as a result of a
macroeconomicshock. But between 1992
and 1997Jordan reducedincome poverty,
despite lagging-at times even negative-
per capita GDP growth rates. The expla-
nation lies in a decline in inequalitydue in
part to the phaseout of regressive food
subsidies coupled with expansion of the
governmentsafety net (WorldBank1999q).
To sustainthese gains,it is importantto im-
prove growth-to make social spending
more affordable and to directly expandop-
portunities for poor people.
Governmentassistanceis impressive-
targeted cash transfers favoring female
and elderly heads of household and the
benefits. But by focusing on the always
poor ratherthan the sometimes poor,gov-
ernment programsfail to address the shal-
lowness of poverty in Jordan. So the
vulnerabilityof the poor and the near-poor
to outsideshocksis high,thoughtempered
by substantial nongovernmentaland reli-
giouscharitable activity that complements
strong family and community networks.
Thatvulnerabilitycanbe addressedby com-
munity-basedpublic work programsoffer-
ing low-wage jobs and by unemployment
insuranceand assistance.TheNationalAid
Fund (in charge of the government safety
net) could identify other means of assis-
tance by soliciting ideas from beneficia-
dissatisfaction among beneficiaries,who
complainof proceduraldifficulties and ob-
stacles, benefits canceled without verifi-
cation,andinadequateassistance.Thefund
is acting on some of these findings, mov-
ing it closerto true accountabilityand to em-
powering poor people.
Jordanians stand to gain much from
the long-awaited Arab-Israelipeace divi-
dend. But for the gains to materialize,
Jordanians must have the tools that ed-
ucation provides. So continued emphasis
on ensuring access for poor people to
publicly provided basic education services
is critical both for the short and the long
run.
36 WORLD DEVELOPMENT REPORT 2000/2001
Box2.1
On interactingwith stateinstitutions:
the voicesof the poor
Althoughthere are pocketsof excellence,poorpeoplein the Voices
of the Poor study generally rated state institutions low on fair-
ness, honesty,relevance,effectiveness, responsiveness,and ac-
countability. Nevertheless,they view government agencies as
havinganimportant role in their lives,and they havea clearpicture
of the qualitiesthey would like to see in the institutionswith which
they interact.
In Indiathe characteristicsof credit institutionscan deter poor
people from seeking loans. Poor people in many regions also
report widespread corruption in health care systems. But when
facing serious health conditions, they feel they have no choice
but to comply with demands for bribes. In Macedonia people
conclude, "nobody wants you to come with empty hands."
The behavior of health care providers becomes another de-
terrent to those needing health services. In Tanzaniain many
areas, men, women, and youth stated over and over that they
are treated like animals, worse than dogs. They report that
even before they could explain their symptoms, they would be
Source: Narayan,Chambers,Shah, and Petesch2000; Narayan,Patel,Schafft, Rademacher,and Koch-Schulte2000.
place a large burden on poor people. They are prevented
from taking advantage of new economic opportunities or
engaging in activities outside their immediate zone ofse-
curity. Threats ofphysical force or arbitrary bureaucratic
power make it difficult for them to engage in public af-
fairs, to make their interests known, and to have them
taken into account. And unaccountable and unrespon-
sive state institutions are among the causes of relatively
slow progress in expanding the human assets of poor
people.
In agrarian societies poor people's lack of assets and
income-earning opportunities ties them to rich land-
owners in patron-client relationships. And for women,
a lack of savings and assets precludes a more indepen-
dent role in decisionmaking in the household and the
community.
Social norms and barriers can alsocontribute to voice-
lessnessand powerlessness.While local cultures have in-
trinsic value, they can sometimes be inimical to reducing
human deprivation. Pervasivein almost all societies is in-
equality between men and women. Poor women are dis-
criminated against in the household and in land, labor,
and credit markets. This both causes poverty and un-
dercuts development-for women's agency is a powerful
source of human gains, especiallyfor children. Discrim-
ination based on ethnicity, religiousbeliefs, social status,
and race has similar effects.
shouted at, told they smelled bad and were lazy and good-for-
nothing.
Poorpeople in many countries spoke about being kept end-
lesslywaiting while the rich went to the headof the queue.
In Europeand CentralAsia pensionerstrying to collect their
meager pensions experience endless red tape, rude and unre-
sponsiveofficials, and withholding of information.Poorpeople in
the region criticize mayors and local authorities for their arbi-
trariness,inefficiency, and often corruption (thoughthere are no-
table exceptions).
Poorpeople hungerfor institutionsthat are fair, polite, honest,
listening, trustworthy, and neither corrupt nor corrupting.A poor
woman in VilaJunqueira, Brazil,summed it all up:
An institution should not discriminate against people because
they are not well dressed or because they are black. If you wear
a suit you are treated as sir; if you are wearing sandals they
send you away.
Vulner ability
Threeyearsago it was a ve?y badyear. Theflood washed
away all of our crops, and there was a lot of hunger
a,-ound here, to the point that nmanypeople actually died
of hunger. They must have been at least a dozen, mostly
children and old people. NVobodycould help them. Their
relatives in the village had no food either; nobody had
enough food for his own children, let alone for the
children of his brother or- cousin. Andfew had a richer
relative somewhere el-e IVho could help.
----lw 'iThV , Lic,
Vulnerability is a constant companion of material and
human deprivation, given the circumstances of the poor
and the near-poor. They live and farm on marginal
lands with uncertain rainfall. They live in crowded
urban settlements where heavy rains can wipe out their
homes. They have precarious employment, in the for-
mal or informal sector. They are at higher risk of dis-
eases such as malaria and tuberculosis. They are at risk
of arbitrary arrest and ill treatment at the hands of local
authorities. And they-women in particular-are at
risk of being socially excluded and victims of violence
and crime.
The risks that poor people face as a result of their cir-
cumstances are the cause of their vulnerability. But the
 CAUSES OF POVERTY AND A FRAMEWORK FOR ACTION 3,7

deeper cause is the inability to reduce or mitigate risk or
copewith shocks-a causethat both draws from and feeds
into the causes of other dimensions of povertv. Low lev-
els of physical,natural, and financialassetsmake poor peo-
ple especiallyvulnerable to negative shocks-those with
more assets can weather these shocks as long as they are
temporary.Lackof adequateassetscan set up a vicious
downward spiralin which actions to cope in the short term
worsen deprivation in the long term. Pulling children out
of school to earn extra income during an economic cri-
sis. Depleting natural resources beyond the sustainable
level. Making quick sales of land or livestock at desper-
ately low prices. Lowering nutritional intake below the
levels necessary to sustain health (chapters 8 and 9).
Another underlying cause of vulnerability is the in-
ability of the state or community to develop mecha-
nisms to reduce or mitigatethe risksthat poor people face.
Irrigation, infrastructure, public health interventions,
honest police and a fair legal system, public work schemes
in times of stress, microcredit to tide people through the
aftermath of an adverse shock, social networks of support
and insurance, famine relief in extreme circumstances-
all reduce vulnerability for poor people. The diverse
cross-country experience with each of these mechanisms
can help in developing actions to addressvulnerability in
specific circumstances.
Poor people also are exposed to risks beyond their
community-those affecting the economy, the environ-
ment, and the society in which they live. Civil conflict
and wars, economic crises,and natural disasters affectnot
only their current living standards but also their ability
to escape poverty. And to the extent that global forces-
such as volatile capital flows, global climate change, and
arms sales-are the causes of shocks and disruptions iin
poor countries, the inability or unwillingness of the
global community to address them increases the vulner-
ability of poor people (chapter 10).
A framework for action
What framework for action is needed to effectivelyreduce
poverty in all its dimensions? National economic devel-
opment is central to success in poverty reduction. But
poverty is an outcome of more than economic processes.
It is an outcome of economic,social,and politicalprocesses
that interact with and reinforce each other in ways that
can worsen or ease the deprivation poor people face every
day. To attack poverty requires promoting opportunity,
facilitating empowerment, and enhancing security--
with actions at local, national, and global levels.Making
progress on all three fronts can generate the dynamics for
sustainable poverty reduction.
The areas for action illustrate the complexity of de-
velopment. How can priorities be decided in practice?
Do all actions have to be carried out in all three areas?
Both the strategic approach and the areas of suggested
action are only a guide. Actual priorities and actions need
to be worked out in each country's economic, sociopo-
litical, structural, and cultural context-indeed, each
community's. But even though choices depend on local
conditions, it generally is necessary to consider scope for
action in all three areas-opportunity, empowerment, and
security-because of their crucial complementarities.
The country examples in this chapter illustrate how Ito

RUSSIAN FEDERATION

Likeother countriesof the formerSoviet of elitecaptureofthestate.Asthe newoli- closely

Union, Russiahashada dramaticriseinboth
povertyandinequalityanda worseningof
adultmortality.
TheRussianpeoplehaveex-
periencedlargeincreasesin insecurity-
throughmacroeconomic the loss
volatility,
of oldjob-related
formsofsecurity,andthe
sharpriseinviolence-andoftenacutepsy-
chological
stressfromthe risein poverty.
Whilethe electoralprocesshas beenim-
portantin empoweringthe citizenry,this
hasbeenoffsetbythe profound feelingsof
disempowerment stemmingfromthe new
sourcesof insecurityandby the problems
garchshavealsocapturedprivatized assets
and resourcerents,the riseininequality is
the productnotof the market-oriented re-
formsthemselves butoftheinteractions be-
tween the reformsand the politicaland
institutional
structuresduringthe transition
process.
Whatarethe prioritiesfor actionto re-
duce poverty?Fundamental to improving
the overallenvironmentis reducingthe
elite's captureof the stateat the national
level,includingthroughfurthermarketre-
formsto deconcentrate economicpower.
Today'sstructuralinequality, linked
tothe political
structure,runsthe riskof be-
comingdeeplyembedded,ifit has notal-
readybecomeso. Dealingwithassociated
issuesof governanceis likelyto be a pre-
requisite
toreducedmacroeconomic volatil-
ityanda businessenvironment thatfosters
the investmentneededto counterthe ex-
traordinarycollapsein formalsectorjobs.
It is alsoa prerequisite
to pro-poorbudget
allocations, backedbydecentralization and
participatory engagement to fostergreater
accountabilityandresponsiveness in service
provision.
38 WORLD DEVELOPMENT REPORT 2000/2001

identify priorities and areas of action in country-specific
poverty reduction strategies.
Opportunity
Growth is essential for expanding economic opportunity
for poor people-though this is only the beginning of the
story of public action (chapter 3). The question is how
to achieve rapid, sustainable, pro-poor growth. A busi-
ness environment conducive to private investment and
technological innovation is necessary, as is political and
social stability to underpin public and private invest-
ment. And asset and social inequalities directly affect
both the pace of growth and the distribution of its ben-
efits.The distribution of growth benefits matters, not least
because distributional conflict can undermine the stability
needed for overall growth.
Markets are central to the lives of poor people (chap-
ter 4). The evidence shows that on average countries
that are open to international trade and have sound
monetary and fiscal policy and well-developed financial
markets enjoy higher growth. Where market-friendly re-
forms have been successfullyimplemented, on average
stagnation has ended and growth resumed. But at times
reforms to build markets fail entirely. The impact of
market reforms on economic performance and inequal-
ity depends on institutional and structural conditions, in-
cluding the comparative advantage of countries and
patterns of asset ownership. And the impact of market
reforms differs for different groups in an economy-
there are winners and losers, and the losers can include
poor people. The design and sequencing of such reforms
thus need to take account of localconditions and the likely
effects on poor people. This does not necessarily mean
going slow: rapid reforms can be important to bring
gains to poor people and to break down monopoly priv-
ilegesfor the rich. Adverseeffectsof reforms on poor peo-
ple can be compensated for by action in other areas,
such as safety nets to ease the transition costs.
Relativelyneglected is market reform targeted to poor
people. Such reformsmay havea differentfocus than other
reforms-eliminating or simplifying the regulations af-
fecting microenterprises and small and medium-size
firms, strengthening registries to allow small producers
to use land as collateral, or developing the policy frame-
work for small-scale insurance.
Key in expanding economic opportunities for poor
people is to help build up their assets(chapter 5). Human
capabilities such as health and education are of intrinsic
value, but also have powerful instrumental effects on
material well-being. Also important to the material
prospects of poor people is ownership of-or accessto-
land, infrastructure, and financial services.And social as-
sets, including social networks, often also play an
instrumental role.
A range of actions can support poor people in ex-
panding their assets.The state-because of its power to
raise revenues and use them as an instrument of
redistribution-has a central role, especially in provid-
ing basic social servicesand infrastructure. Where access
to land is highly unequal, there is a social and economic
case for negotiated land reforms. For many services the
state's role in provision can be complemented by market
mechanisms, civil society, and the private sector, in-
creasing the benefits to poor people. And for local ser-

SIERRA LEONE

Sierra Leone, by the latest price-adjusted
measures, is the poorest country in the
world. But this statement fails to convey
the true depth of human deprivation in
that nation. The people of Sierra Leone
remain caught in a tragic conflict-one
that has taken a terrible toll through lost
lives, rape, mutilation, and the psycholog-
ical harm to boys abducted into the army
and militias.
Work on the sources of conflict in de-
velopingcountries suggests that material
povertyandweak democraticstructuresin-
teract with ethnic and other social divides
to cause internal strife. The effects of
conflict-destruction of fragile institutions
of governance, flight of skills, personal
losses, and social wounds that could take
generationsto heal-create a vicious cycle
of continued poverty and strife.
SierraLeone has a desperate needfor
assets-human, physical,and social-and
for greater market opportunities. And the
personal insecurity is unimaginable. But
there can be no progress without mecha-
nisms to resolve the social conflict, deal
with citizens' disempowerment by those
with guns, and re-create the institutions
for mourning and for managing psycho-
logical losses.
Internationalaction will be important.
Oncethereisa basisfor somedevelopment,
concerted external support will be crucial.
TheEnhancedHeavilyIndebtedPoorCoun-
tries Debt Relief Initiative provides transi-
tional support to postconflict societiesfor
economicreconstruction.Much more chal-
lengingwill be the delicatetask of socialand
institutional reconstruction.
 CAUSES OF POVERTY AND A FRAMEWORK FOR ACTION 39

vice delivery,engagingpoor people and communities The burden of petty corruptionfallsdisproportionately

can havea powerfulimpacton effectiveness.
Empowerment
Empowerment meansenhancingthe capacityofpoorpeo-
ple to influencethe stateinstitutionsthat affecttheirlives,
bystrengtheningtheir participationin politicalprocesses
and local decisionmaking.And it means removingthe
barriers-political, legal,and social-that work against
particulargroupsand buildingthe assetsof poor people
to enablethem to engageeffectivelyin markets.
Expandingeconomicopportunitiesforpoor peoplein-
deedcontributesto their empowerment.But effortsare
neededto makestate and socialinstitutionswork in the
interestsof poor people-to makethempro-poor(chap-
ter 6). Formaldemocraticprocessesare part of empow-
erment.Asimportantarethe mechanismsthroughwhich
everydaystate interventionshelp or hurt poor people.
Here, more detailedprocessesof accountabilitycome
into play-mobilizing poor peoplein their own organi-
zationsto hold stateinstitutionsaccountableand ensur-
ing the rule of lawin their dailylives.
Empoweringpoor peopleispart of the broaderagenda
of soundgovernanceand accountabilityof state institu-
tions to theircitizens.Nationalempowermentof citizens
canhaveimportantindirecteffectson poor people,by in-
fluencingthe qualityand paceofeconomicand socialde-
velopment.Butthe outcomeforpoorpeopledependson
the politicaland socialstructureswithin a society.Gov-
ernmentsareoftenmoreresponsive to the concernsof elites
than to the needsof poor groups.So the extentto which
the concernsof nonpoor and poor groupscoincidewill
frequentlydeterminewhethergovernanceis pro-poor.
Improvinggovernancealsorequiresbuildingadmin-
istrativeand regulatorycapacityand reducingcorruption.
on poor people,whogenerallyhavecommoncausewith
an anticorruptionagenda.
Social interactions between individuals and comr-
munitiesalso havean important influenceon poverty
outcomes.Culture'spart in the developmentprocessis
complex. The beliefs and practices that are part of
localculturecan be a sourceof sustainabledevelopment.
But customary practices and discrimination on the
basisof gender, ethnicity, race, religion, or social sta-
tus can alsobe a sourceof inequalityin manycountries.
Removingdiscrimination and managing these divi-
sionscan help reduce poverty.Confronting genderin-
equities is a fundamental part of this, with direct
benefitsfor women (and men) and instrumentaleffects
on growth and development. Recent evidenceshows
that greater gender equity is associated with faster
growth (chapter 7).
Security
Enhancingsecurityforpoor peoplemeansreducingtheir
vulnerabilityto such risksasill health,economicshocks,
and naturaldisastersand helpingthemcopewith adverse
shockswhen they do occur (chapters8 and 9).
Povertyreduction strategiescan lessenthe vulnera-
bilityof poor householdsthrough a rangeof approaches
that can reduce volatility,provide the means for poor
peopleto managerisk themselves,and strengthenmar-
ket or publicinstitutionsforrisk management.The tasks
include preventingor managingshocksat the national
and regionallevel-such as economicdownturns and
natural disasters-and minimizingthe impacton poor
people when they do occur.
Supportingthe rangeof assetsof poorpeople-human,
natural, physical,financial,and social-can help them

UGANDA

Havingemergedfroma periodof destruc- poverty,impressiveeffortstowarduniver- countabilityandparticipation

tiveconflictjustovera decade ago,Uganda
suffersdeeppovertyin manydimensions.
Butit alsoshowswhatanimmenselypoor
Sub-Saharan Africancountrycanachieve.
Thefirst countryto receiveenhanced
debt relief on the basisof its povertyre-
ductionstrategy,Uganda standsout for its
steadygrowthin the 1990s.It alsostands
out for significantreductionsin income
salprimaryeducation, anda majoreffortto
ensuretransparent, poverty-focused bud-
gets,bothcentrallyandlocally.Oneof its
mainvulnerabilitiesis health.HIV/AIDS hit
Uganda early:a tenthof adultsarenow in-
fected,andAIDSorphans arestrainingtra-
ditionalsystemsof fosteringchildren.
Threeareas are prioritiesfor future
action.Consolidatinganddeepening theac-
in resource
al-
locationandstrengthening centralandlocal
state institutions,to providea basisfor
soundlocal investmentprogramsin so-
cialandphysicalcapital.Tacklingthe per-
ceivedrisks inthe businessenvironment
so that job-creatinggrowthcan takeoff.
Andfurtheringcurrenteffortsto stopthe
spreadof HIV/AIDS andsuchdiseasesas
tuberculosis.
40 WORLD DEVELOPMENT REPORT 2000/200I
manage the risks they face. And supporting the institu-
tions that help poor people manage risk can enable them
to pursue the higher-risk, higher-return activitiesthat can
lift them out of poverty.Improvingrisk managementin-
stitutions should thus be a permanent feature of poverty
reduction strategies.A modular approach ISneeded, with
reduction stratgies. A modula
differentschemesto cover different types of risk and dif-
ferent groups of the population. The tools include health
insurance,old age assistanceand pensions, unemployment
insurance, workfare programs, social funds, microfi-
nance programs, and cashtransfers. Safety nets should be
designed to support immediate consumption needs-and
to protect the accumulation of human, physical, and so-
cial assets by poor people.
In addressingrisk and vulnerability,the issueonce again
is whether public interventions and institutions work
well-and in the interests of poor people. Famines are a
constant threat in many parts of the world,yet some coun-
tries have been able to avoid mass deaths. In the 20th cen-
tury no democratic country with a free press and a free
political opposition ever experienced famine (box 2.2).
Access to information and participation can reduce
.Separating
vulnerabiliy.
vulnerability.
Interconnections at local and national levels
Just as the dimensionsand causesof povertyare interlinked,
the areas for action are interconnected.
so Action to ex-
50teaesfrato r mecnetd clnt x by the local electorate.
pand opportunity is itself a potent source of empowerment,
in a deep, intrinsic sense with respect to basic human ca-
pabilities, but also instrumentally-for as the asset base,
incomes, and market opportunities of poor people increase,
so will their potential political and social influence. Im-
proving material conditions is also instrumental in en-
hancing security: adverse shocks have lower costs when a
person is above the margin of bare survival,and assetsare
at the heart of people's risk management strategies. Em-
powerment is fundamental in determining action in mar-
ket reforms and the expansion of assets that affect the
pattern of material opportunities and in shaping the de-
sign of policies and institutions that help poor and non-
poor people manage the risks they face. Finally,reducing
vulnerability, with all its debilitating consequences, is
central to improving material well-being (or preventing
reversals) and empowering poor people and communities.
Inter-national actions
With globalforceshaving central-and probably rising-
importance, actions at the local and national level are not
Box 2.2
Preventingfamines: the local press matters
Faminesareoftenthe result of crises that affect agricultural
production-floods or droughts.Howquicklygovernments
respondto such crisesdependson manyfactors.Afunda-
rsodt uhcie eed nmn atr.Afna
mental one is the level of democracy and the extent to
whichpoliticiansare held accountablefor the efficiencyof
reliefprograms.A recent study in Indiashows that the dis-
tributionof newspaperscanplay a big part.
India has a relatively free press, with only 2 percent of
newspapers controlled directly by central or state govern-
ments. The study looked at the interaction between gov-
ernmentresponsivenesstofloodsanddroughts(measured
in public relief funds) and the circulationof newspapers
across Indian states. The hypothesis: an informed population
can linkinefficiencyto a particular politicianand therefore elicit
a greaterresponseto crises.
The results confirmedthe hypothesis:for a given shock
(droughtor flood), higher newspaper circulationleads to
greaterpublicfooddistributionor reliefspending.A 10 per-
cent drop in food production due to a crisis is associated with
a 1 percent increase inpublicfood distributionin the states
withthe median newspapercirculationper capitabut a more
than 2 percent increase in states in the 75th percentile of
newspaper circulation.
newspapers by language yields an interest-
ing result. Among three types of papers-those in Hindi, Eng-
lish, and local languages-only those in local languages
seem to enhance governmentresponsivenessto crisis.The
state governments' responses to localcrises are thus very
sensitive to the distributionof local newspapers, typicallyread
Source: BesleyandBurgess2000.
enough. Global economic advance, access to interna-
tional markets, global financial stability, and technolog-
ical advances in health, agriculture, and communications
are all crucial determinants of poverty reduction (chap-
ter 10). International cooperation is thus needed to re-
duce industrial countries' protectionism and avert global
financial volatility. And the growing importance of such
international public goods as agricultural and medical re-
search calls for a shift in the focus of development co-
operation. Furthermore, because of the importance of
international actions in poverty reduction, the voices of
poor countries and poor people should be strengthened
in international forums.
Country-focused aid programs remain essential-to
help countries implement poverty reduction strategies
that empower poor people, enhance their security, and
expand their opportunities (chapter 11). Aid should be
 CAUSES
Box2.3
Attackingpovertyin Vietnam
Vietnamhas madestrikingprogressagainstpoverty,reducing
the shareof itspopulationin incomepovertyfrom 58 percentto
37 percentin 1993-98.A recentanalysisby the Poverty Working
Group,with membersfrom government,donors,andNGOs,
foundthat:
* Themainengineof the rapidpovertyreductionwasreform.
Especiallyimportantwere the land reformsthat Vietnam
launchedin the mid-1980s,whichcreatedenormousoppor-
tunitiesfor peopleto improvetheir livesandlivelihoods.
* Despitethe gains,poorpeopleexpressed a senseof voice-
lessnessandpowerlessness. Participatorypovertyassess-
ments (donejointly with Oxfam,Actionaid,and Savethe
Children) foundthatpeoplewere hungryfor a two-wayflow
of information-fromthe government to them aboutthe na-
tureandtimingof publicpoliciesandprogramsaffectingtheir
livesandfromthemto the government to influence
thosepoli-
ciesandprograms.
* Povertyremainsdeepandwidespread-andthe gainsfrag-
ile. Millionsof peopleare still vulnerableto poverty.Illness,
thedeathof a familymember,andnaturaldisasters(flooding,
Source: WorldBank1999bb.
directed to countries with high levels of poverty. But
that should be only part of the criteria for allocating
aid. Also essential is having the right policy and insti-
tutional framework in place to make poverty reduction
a success.
Debt relief for the world's poorest nations, the heav-
ily indebted poor countries, has been the most prominent
issue in development cooperation in recent years.This re-
port recognizes that debt reduction must play a central
part in an overall strategy for attacking poverty.
The chapters that follow outline different sets of ac-
tions to consider in devisinga poverty reduction strategy.
The priorities cannot be set in the abstract. They must
fit the context-and reflect a broad national consensus.
Recent experiencein Vietnam showshow this processcan
be set in motion (box 2.3).
OF POVERTY AND A FRAMEWORK FOR ACTION 41
drought)remainever-presentthreats.Women,ethnic mi-
norities,andunregisteredurbanmigrantsremainespecially
disadvantaged.
Theanalysis-whichinvolveda widerangeof stakeholders-
identifiedthreeareasof policyaction:
* Launcha newroundof reformsthatunleashthe dynamism of
theprivatesectorandcreateopportunities for employment and
productivity
growth,sothat incomesriseandpoorpeopleare
ableto escapepoverty.
* Implementthe Grassroots Democracy Decree, whichaimsto
empowerpeopleby authorizingtheir directparticipation in
localdecisionmaking andimprovinglocalgovernance.
* Strengthen safetynetsandtargetedprograms,suchas the
HungerEradication andPovertyReduction Program,to reduce
thevulnerability
of poorpeopleto risks(illness,poorharvests).
Thereportonthe studyhasbeendisseminated widelyin Viet-
nam,includingto all450membersof the NationalAssemblyand
to all61provinces.
Theprimeministerhasaskedthe Poverty Work-
ingGroupto translatethe findingsinto a comprehensive poverty
reductionstrategyfor Vietnambeforethe endof 2000.
This chapter has presented an overall framework for
actions in three areas-opportunity, empowerment, and
security-to reduce poverty in its different dimensions.
Actions need to be taken by the full range of agents in
society-poor people, government, the private sector,
and civilsocietyorganizations-and at local,national, and
global levels. The country examples in this chapter il-
lustrate three fundamental points:
* Actions affecting opportunity, empowerment, and se-
curity are interconnected-there is no hierarchy,
and advances in one area generally depend on gains
in the others.
* In all cases the social, political, and institutional in-
derpinnings for action are offundamental importance.
* Context matters.While it generallyis alwaysdesirableto
take or sustainaction in all three areas,the designof ac-
tion, and the agentsthat matter,depend on the economic,
social,and politicalconditions prevailingin the country.
 PART Il

Opportunity

Growth, Inequality,
,A _ s countries become richer, on averagethe in-
cidence of income poverty falls. Other indicators of
well-being, such as average levels of education and
health, tend to improve as well. For these reasons,
economic growth is a powerful force for poverty re-
duction. This observation is not the end of the story,
for it raises the questions of what causes economic
growth and why countries with similar rates of eco-
nomic growth can havevery different rates of poverty
reduction.
Until the mid- 18th century improvements in liv-
ing standardsworldwide were barely perceptible. Most
societieswere resigned to poverty as an inescapablefact
of life.' As late as 1820 per capita incomes were quite
similar around the world-and very low, ranging
from around $500 in China and South Asia to
$1,000-i ,500 in the richest countries of Europe.2
Roughly three-quarters of the world's people lived
on less than $1 a day.3
The onset of modern economic development
opened the possibility that growth could significantly
improve the living standards of poor people-and
everyone else. Over the next two centuries per capita
incomes in the richest countries of Europe increased
CHAPTER 3i
and Poverty
more than tenfold in real terms, in China more than
fourfold, and in South Asia threefold. The conse-
quences for poverty have been dramatic. In the rich
countries of Europe the fraction of the population liv-
ing on less than $1 a day has fallen to nil. In China,
where growth was slower, less than 20 percent of the
population now lives on less than $1 a day. In South
Asia, where growth was slower still, around 40 per-
cent of the population does. Today roughly a fifth of
the world's people fall below this austere income
threshold.
But differences in rates of economic growth, and
in the rates at which that growth translates into
poverty reduction, are not the consequence of sim-
pie choices. Countries do not choose to have slow
growth or to undergo painful crises.Nor do they sim-
ply choose how equitable growth will be. Instead, the
patterns of growth, the changes in the distribution
of income and opportunities, and the rates of poverty
reduction reflect a complex set of interactions among
the policies, institutions, history, and geography of
countries. Understanding the forces underlying coun-
tries' disparate growth experiences, and the mecha-
nisms through which this growth has reached poor
45
46 WORLD DEVELOPMENT REPORT 2000/2001

people, is essential for formulating poverty reduction Figure 3.1

strategies. In general,the wealthiera country,the lower
This chapter takes up these issuesin turn. It first doc- the incidenceof poverty
uments the strong links between economic growth and
Averageannualpercapitaconsumption
of poorestfifth
the income and nonincome dimensions of poverty. It of population
next turns to the policiesand institutions that under- 1993u.s. dollars (PPP);log scale
pin growth and provide the basis for poverty reduction. 10,000
It recognizes that there are substantial deviations from
these general relationships reflecting the wide diversity
of country experience-and that these deviations reflect
a further set of interrelationships between distribu- 1
tional outcomes, policies, and institutions. It therefore 100.
discusses how cross-country differences in the poverty-
growth nexus are a consequence of initial inequalities
in the distribution of income and opportunities
-and of changes in the distribution of income that .
occur with growth. These inequalities themselves reflect
an array of factors, which in turn have consequences for 100 o

economic growth. Last, the chapter explores the inter- 100 1,000 10,000
Average annual per capita consumption
actions between growth and two nonincome dimensions 1993U.S. dollars (PPP); logscale
of poverty-health and education.
Share of population living on less than $1 a day
Percent
Economic growth and poverty 80
reduction
70 0

Today close to a fifth of the people in the world sur- 60

vive on less than $1 a day. The incidence of this de- 50
privation varies greatly across countries. Not surprising, .
the richer the country, the higher the average con- 40
sumption of the poorest fifth of its population-and 30

the smaller on average the fraction living on less than 30

$1 a day (figure 3.1). There are also significant varia- 20 *

tions around this relationship. Countries with the same 10*

average consumption have quite different proportions
of the population living on less than $1 a day, reflect-
ing substantial differences in inequality across countries.
Education and health indicators are also better on av-
erage for richer countries. In rich countries fewer than
1 child in 100 does not reach its fifth birthday, while
in the poorest countries as many as a fifth of children
do not (figure 3.2). Similarly, in the poorest countries
as many as half of children under five are malnourished
-in rich countries fewer than 5 percent. Again, how-
ever, there can be striking deviations from the average.
For example, the United States is vastly richer than
China and India, but the life expectancy of African
Americans is about the same as that in China and in
some states in India.4
.
0 1,000 10,000
Averageannualper capitaconsumption
1993U.S. dollars (PPP),log scale
Note: The datacover65 developing countries and refer to various
yearsin the 1990s.
Source:WorldBankstaff estimatesbasedon datafrom Chenand
Ravallion(2000).
[Poverty is] . . low salaries and lack ofjobs. And it's also
not having medicine, food, and clothes.
-- f-ron: a discusalongroiqp,Brazil
Still,the stark differencesin poverty outcomes between
rich and poor countries point to the central role of eco-
 GROWTH, INEQUALITY, AND POVERTY 417

Figure 3.2 Figure 3.3

Health indicatorsimprove as incomesrise Poverty trends tracked growth trends in the
1980sand 1990s
Under-fivemortalityrate, 1995 -
Per 1,000
live births Averageannualgrowthin per capitaconsumption
of
300 poorestfifth of population
* Percent
250 40 *
ecce 30 C
200 20 -3

1* . *: 20ercen .

* c *t
l evelopent
*omlc poverty reauctlo. 1 nese Percent10
50 r ct
ce
c ~~~~~~~-20

Pero
capta GDP,1996 -40
U.S.dollars
(PPP);
logscale -40 -30 -20 -10 0 10 20 30 40
Source:
WorldBankdata. Average
annualgrowthin percapitaconsumption
Percent
Average
annualgrowthinshareofpopulation
livingon
pdevelopment
nomh c in poverty reduction. These airrer- lesst
h pens erally reflect cross-country differences in eco- 80
nomc lgrowth over the very long run. But the benefits 60 °
of growth in reducing income poverty can also be seen
over shorter periods. Chapter 1 discusses the highly 40
variable evolution of income poverty across countries in 20
the past two decades. Differences in economic growthco- _
across countries account for much of this variation: as 0ug
in the very long run, growth in the 1980s and 199aOs was -20
a powerful force for reducing income poverty. On aver-
age, growth in the consumption of the poorest fifth of
the population tracked economic growth one-for-oneover A60
this period (figure 3.3). In the vast majority of cases
growth led to rising consumption in the poorest fifth of -840 -30 -20 -10 0 10 20 30 40
the population, while economic decline led to falling Averageannualgrowth in percapita
consumption
consumption. Percent
The pattern is similar for the share of people livingon Note:Thedatacover65developingcountries.
lessUthan
of pove y.
Con than
less O averselylowory
,
1 a day.
average, very adtiv ronl per-
additonal erS- World
Source:
Ravallion Bank
(2000). staffestimates
basedon datafromChenand
centage point of growth in average household con-
sumption reduces that shareby about 2 percent. Although
the deviations from this average relationship show that sulting from the collapseof the state, natural disaster,war,
in some countries growth is associated with much more or economic crisis can have a devastating impact on poor
poverty reduction than in others, the relationship high- people.
lights the importance of economic growth for improv- As chapter 1 shows, national poverty figureshide much
ing the incomes of poor people and for moving people variation in outcomes within countries. But just as cross-
out of poverty. Conversely, low or negative growth re- country differences in economic growth do much to cx-
48 WORLD DEVELOPMENT REPORT 2000/2001

plain cross-country differences in poverty outcomes, re-
gional and subregional growth does much to explain sub-
national poverty outcomes. World regions, countries, and
provinces within countries have grown at very different
rates (figure 3.4). Where growth has occurred, it has been
an important source of poverty reduction, and where it
has not, poverty has often stagnated. Understanding why
countries and regions have had such disparate growth ex-
periences,and how this growth reachespoor people, is es-
sential for formulating poverty reduction strategies.

Figure3.4
Economicgrowth was a force for poverty reductionin the 1980sand 1990s ...

... acrossregionsof the world ... ... acrosscountriesin EastAsia...

Averageannualchangein incidenceof poverty Averageannualchangein incidenceof poverty
Percent Percent
12 5
* Europeand i
CentralAsia
8 0_
Mongolia Philppines

4 \ | 5 tLaoPDR * Vietnam * China

\ Latin America andi\
\the Caribbean MalaysIa\
o ---------
* SouthAsia -10 Indon

-4 , \East Asia -15

and Pacific

-8 -20
* Middle East
and North Africa Thailand.
-12 -25
-5 0 5 10 -5 0 5 10
Averageannualgrowth in per capitaGDP Averageannualgrowth in per capitaGOP
Percent Percent

... across regions in the Philippines ... ... and across provinces in southern Luzon

Averageannualchangein incidenceof poverty Averageannualchangein incidenceof poverty

Percent Percent
" '.Autonomous 20 Mindoro
Region *Caraga Region Occidental
in Muslim 15
Mindanao-,
O -Eastern Visayas------------
Mindanao
ICentral
-2 CentralLuzon* * Ilocos 10
Cagayan.*
F * Bicol
4 Mindanao Csestern i
C Laguna \
Southern Luzon* Cavite. * Ara
-6 Western C vte* a
-~~~~~~~~~ ~~~Visay as * 0
Cordillera Southern 0
-8
t Administrative Mindanao lan Romblon
Region Northern Marinduque *
-10 Mindanao Mindoro Oriental- Q
Metro Manila . Batangaso
-12 -10
-3 -2 -1 0 1 2 3 4 5 -5 -4 -3 -2 -1 0 1 2 3 4 5
Averageannualgrowthin per capitaconsumption Averageannualgrowth in per capitaconsumption
Percent Percent

Note:Theincidenceof povertyisthe shareof the populationlivingon lessthan $1a day.

Source:Topleft panel,WorldBankdata;top right panel,WorldBank1998f;bottomtwo panels,WorldBank2000o.
 GROWTH, INEQUALITY, AND POVERTY 49

What drives economic growth? Someeconomicpolicies-suchasopennessto inter-

Understandingthe policiesand institutionsthat lead
to sustainedand sustainableeconomicgrowthis a first
step in developingstrategiesfor improvingthe lot of
poorpeople.Widedivergences in growthreflecttheout-
comeofinteractions amongcountries'initialconditions,
their institutions,their policychoices,the external
shocksthey receive,and no smallmeasureof good
luck.
Thereisevidencethat growthdependson education
and lifeexpectancy,particularly
atlowerincomes. 5 For
example,it hasbeenshownthatfemaleliteracyandgirls'
educationaregoodforoveralleconomicgrowth. 6 There
is alsosomeevidencethat rapidpopulationgrowthis
negativelyassociated withper capitaGDPgrowthand
that the changingagestructureof the populationcan
alsoaffectgrowth(box3.1).7
nationaltrade,soundmonetaryand fiscalpolicies(re-
flectedin moderatebudgetdeficitsandtheabsence ofhigh
inflation),
a well-developed financialsystem,anda modL-
eratelysizedgovernment-arealsostronglyconducive to
economicgrowth. 8 Aidcanboostgrowthif suchpolicies
arein place,but not if theyareabsent. 9 Bothdomestic
andexternalshocksmatteraswell.Notsurprising, wars,
civilunrest,andnaturaldisastersalllowergrowthrates
(box3.2).Lessdramatically, sodo macroeconomic volatil-
ity,adversetermsof tradeshocks,and slowergrowth
amongtradingpartners.' 0 Poorlysequenced and badly
implementedreformscan lead to suddenreversalsin
capitalflowsor othermacroeconomic disruptions,also
slowing growth(chapter4).Thesecollapses in growthcan
be particularlydevastatingfor poorpeople,whohave
weakersupportmechanisms and generallyleada more
precariouslifethanthebetter-off(chapter9).

Box3.1
Population,growth,and poverty

Manystudieshavedocumented thatas countries becomericher, the potentialof a growingworkforcewasdueto a varietyof fac-

bothfertilityandmortalitydeclineonaverage,with reductionsin tors,including strongeducational attainmentandasupportive pol-
mortalitytypicallyprecedingreductionsin fertility.' The interac- icyandinstitutionalenvironment.In other regionsof the world,
tionsbetweenthis demographic transitionandeconomicdevel- notablyLatinAmerica,a similarchangeinthe composition of the
opmentarecomplex. Theyhavefueledheateddebateat leastsince population occurredwithouta comparable growthbenefit.This
1798,when ThomasMalthusarguedthatsince"food is neces- failureis disappointing sincethe demographic "bonus"of alarger
saryto the existenceof man"and"the passionbetweenthesexes workforceis temporary andis followedby a periodof higherold
is necessaryandwill remainnearlyin its presentstate" (1985, agedependency ratesthat placegreaterdemandson the social
p. 70),populationgrowthwouldinevitablyleadto animbalance securityinstitutionsthat providesupportfor the elderly.
betweenpeopleandavailable resources. Second,there is evidencethat better educationis associ-
3
Malthus'sgrimpredictiononthe effectsof population growth atedwith highercontraceptive useandlowerfertility. Thisevi-
on economicdevelopment failedto materialize-sincethe turn dencemay reflect a varietyof mechanisms.More education
of the 19thcenturythe world's populationhas increasedmore expandseconomic opportunitiesfor womenandso canraisethe
than fivefold,andthanksto improvements in technologyof all opportunity costof havingmorechildren (Becker1960).Infantmor-
kinds,percapitaincomeshaveincreased by evengreatermulti- tality is often lowerin familiesin whichwomenare betteredu-
ples.The linksbetweendemographic changeanddevelopment cated,and so fewer birthsare requiredto achievea desired
aremoresubtlethanthis.Twoissuesarenoteworthy: theeffects numberof children.And bettereducationcan improvethe ef-
of changesinthe agestructureof the population inducedbythis fectivenessof contraceptive use,Investments in improvingpoor
demographic transition,andthe links betweeninvestmentsin people'saccessto education andhealthcanthereforehavea dou-
healthandeducation, growth,anddemographic outcomes. ble impact.Theseinvestmentshave beenshownto improve
First,in manycountriessharpdeclinesinfertility havebeen growthandreducepovertydirectly.Tothe extentthatthey are
followedbysharpincreases inthe working-age shareof the pop- associated with lowerfertilityandpopulationgrowth,theycanalso
ulation.In somecountries,notablyin EastAsia,the increasein contribute to avirtuouscircleof improvedmaternal healthandbet-
the numberof workersper capitawas accompanied by faster ter investment in children's
healthandeducation, whichreinforce
2
growthin GDPpercapita. Thesecountries'successin tapping thesegains.
1. SeeLivi-Bacci
(1997)for a historical
surveyandBirdsall(forthcoming) for a modernreviewof theliteratureon demography andeconomics.
2. Forexample,Young11995) providesa carefulassessment of the contributionof a growinglaborforceandgreaterparticipation ratesto therapid
growthin percapitaGDPobservedin fourAsianeconomies.
3. Schultz(1994)provides
cross-country evidenceonthelinksbetweenfemaleeducation andfertility.SeeFeyisetan andAinsworth(1996)for micro-
economic evidenceoneducationandcontraceptive useandAinsworth,Beegle,andNyamete(19961 oneducation andfertility.PritchettandSummers
(1994)providea morecautiousassessment of the magnitude
of the effectof contraceptiveavailability
onfertility.
50 WORLD DEVELOPMENT REPORT 2000/200I
Box 3.2
How war devastates poor people
Warsare devastatingwhereverthey occur.Sincethey occur
disproportionatelyin poorcountries,the devastationfalls dis-
proportionately
onthe world'spoorpeople(seefigure).Morewars
arenow civil.During1987-97morethan85 percentof conflicts
were foughtwithin nationalborders(14 were in Africa, 14 in
Asia,1 in Europe).Tragically,
90 percentof war deathsarenot
military(Pottebaum 1999).In Cambodia1.7 millionpeopledied
in 20yearsof fightingandpoliticalmassmurder-amongthem,
mostof the country'sdoctors,lawyers,andteachers.Civilianvic-
tims are also singledout becauseof their ethnic identity:as
manyas 800,000TutsisandmoderateHutuswere killedby ex-
tremistHutusin Rwandain 1994.
Conflict is overwhelmingly concentrated in poor countries
Shareof incidentsof civil warand strife,1990-95
Percent
100
80
60
40
20
0
EastAsia Europe LatinAmerica MiddleEast
and and andthe and
Pacific CentralAsia Caribbean NorthAfrica
Source: USAID, OFDA 1999.
Institutional factors are also important for growth. For
example, there is evidence that strong rule of law and the
absence of corruption contribute to growth-by pro-
viding a fair, rule-based environment in which firms and
households can investand grow."t Strong institutions can
also havepowerful indirect benefits.For example,adjusting
to adverse shocks often requires painful but necessary
changes in domestic economic policies. In countries
where conflicts between competing interests are pro-
nounced, and the institutions to resolve these conflicts
are weak, recovery from shocks is often slower than it is
where these institutions are strong."2
Similarly,there is growing evidence that ethnic frag-
mentation has adverse effects on growth. Ethnically frag-
mented countries and regions within countries tend to
Norarechildrenexempt,for theyareoftenrecruitedto fight.
Childrenluckyenoughto surviveaconflictbeardeeppsycholog-
icalscars.Theyalsopayaheavypricefor theirabandoned school-
ingin permanently diminished economicopportunities.
Wars cripple economiesby destroyingphysical,human,
andsocialcapital-reducinginvestment,divertingpublicspend-
ing from productiveactivities,anddriving highlyskilledwork-
ers to emigrate.In civil war a country'spercapitaoutput falls
an averageof more than 2 percenta year relativeto what it
wouldhavebeenwithoutconflict.In moresevereandprotracted
wars,the economicandhumancostsareevengreater(Collier
1999b).
SouthAsia Sub-Saharan Developing High-income
Africa countries countries
provide fewer-and poorer quality-public goods, es-
pecially education. Such areas are also more prone to vi-
olent ethnic conflict. Institutions that guarantee minority
rights and provide opportunities to resolveconflicts have
been shown to offset the side effects of polarized societies
(chapter 7).13
Other exogenous factors, such as geography and ini-
tial incomes, matter as well.There is some evidence that
geographic characteristics affect growth-for example,
a remote or landlocked location acts as a drag on
growth.t 4 On average, initially poor countries have
grown more slowly than rich countries, so that the gap
between rich and poor countries has widened (box 3.3).
However, there is strong evidence that, controlling for
some of the factors mentioned above, growth is faster in

Box 3.3
Divergenceand worldwide income inequality
Giventhe importance of growth for poverty reduction,the failure
of growth to take root in some of the poorest countries with the
highest incidence of poverty is particularly disappointing. One
symptom of this failure is the widening gap in averageincomes
between the richestand poorestcountries. In 1960per capitaGDP
in the richest20countrieswas 18 timesthat in the poorest20 coun-
tries. By 1995this gap had widened to 37 times, a phenomenon
often referred to as divergence(seeleft-hand panel of figure).
Such figures indicate that income inequality between coun-
tries has increased sharply over the past 40 years. What has
happened to worldwide inequality between individuals?Trends
in worldwide inequality between individuals reflect trends in
both inequality between countries and inequality between in-
dividuals within countries. The contribution of inequality be-
tween countries depends on differences in country growth
performance and country size: rapid growth in a few large
and initially poor countries can offset the disequalizing effect
of slow growth in other poor countries. In China, for example,
rapid growth from a very low base has helped a fifth of the
world's population halve the gap in average per capita in-
comes with the world as a whole, significantly reducing world-
wide inequality between individuals. In contrast, the 20 poorest
countries in the world in 1960 accounted for only about 5 per-
cent of the world's population,and so their failure to grow, while
Wideninggapsbetweenrich andpoorcountriesaccountfor muchof the increasein worldwideincome
inequalityacrossindividualsoverthe past 40 years
PercapitaGOP
1985 U.S. dollars (PPP)
20,000
15,000
Richest
10,000 20 countries
5,000
Poorest
20 countries
1960 1995
Note:Theleft panelrefersto population-weighted
averagesof per capitaGDPin the indicatedgroups,basedona sampleof 123countrieswith
completedataon per capitaGDPoverthe period1960-95.Chinaisexcludedfrom the poorest20 in 1960.TheTheilindexis a measureof
incomeinequality;highervaluesindicatehigherinequality.
Source:SummersandHeston1991;WorldBankdata;BourguignonandMorrisson1999;Milanovic1999.
GROWTH, INEQUALITY, AND POVERTY 5I
disappointing, contributed less to worldwide inequality be-
tween individuals.
Income inequality within countries shows less pronounced
trends: in some countries inequalityhas increased,while in oth-
ers it has fallen. Recent studies have found that across countries
increases and decreasesin inequality are roughly equally likely
(Deiningerand Squire 1996b).Again, however, country size mat-
ters: changes in inequalityin populous countries such as China,
India, or Indonesiawill contribute more to changes in worldwide
inequality between individuals than will changes occurring in
small countries.
Trends in worldwide income inequality between individuals
reflect both these factors, with the between-countrycomponent
typically more important than the within-country component. In
light of the difficulties with measuringincome described in chap-
ter 1, it is not surprising that estimates of worldwide inequality
between individualsare subject to substantial margins of error.
But available estimates indicate that there have been some in-
creases in worldwide inequality between individuals in past
decades (see right-handpanel of figure). While the size of these
increases depends on the methodology used and the period
considered, the evidence suggests that the increases in world-
wide inequalityin recent yearsare small relativeto the much larger
increases that occurred during the 19th century.
Incomeinequalityamongindividuals
Theit index
1.00
Bourguignon
and Morrisson1999
0.75
Milanovic1999
0.50
0.25
1800 1850 1900 1950 2000
52 WORLD DEVELOPMENT REPORT 2000/200I
countries that are initially poor. This relationship may
not be linear, with higher growth kicking in only after
countries reach a threshold level of income. This raises
the possibility of poverty traps at very low levels of de-
velopment.15 Finally, initial inequality can influence
later growth, with implications for how growth translates
into poverty reduction. This important issue is discussed
in the following section.
What determines the sustainability of growth? In ad-
dition to the policy, institutional, and geographic factors
mentioned above, a further important consideration is
whether or not growth is accompanied by environmen-
tal degradation, which can in turn undermine growth.16
Environmental degradation can exact a heavy toll on the
economy through poor health and reduced agricultural
productivity. For example, heavy reliance on coal with-
out effectivecontrolson particulate,sulfur,and other emis-
sions can cause high rates of lung disease, and sulfur
emissionslead to acid rain, which reducesagriculturalpro-
ductivity.17 In the long run especially, attending to the
quality of the environment and the efficiencyof resource
use is likely to boost investment, accumulation, and
growth. Rapid growth and environmental protection
can go together-because new additions to industrial ca-
pacity can take advantage of cleaner technologies and ac-
18
celeratethe replacementofhigh-pollution technologies.
Water is lift, and because we have no water, lift is
miserable.
-4---j-l i Ji:ifsiwlY,a,A.a2il
Why are similarrates of growth
associatedwith differentrates
of poverty reduction?
The general relationship between economic growth and
poverty reduction is clear. But there are also significant
differences across countries and over time in how much
poverty reduction occurs at a given rate of economic
growth. The bottom panel of figure 3.3 shows that there
can be large variation in poverty reduction for the same
growth rate in per capita consumption (though extreme
values should be considered outliers). What explains
these large differences?For a given rate ofgrowth, the ex-
tent of poverty reduction depends on how the distribu-
tion of income changes with growth and on initial
inequalities in income, assets, and access to opportuni-
ties that allow poor people to share in growth.
Changes in the distribution of income
How growth affects poverty depends on how the addi-
tional income generated by growth is distributed within
a country. If economic growth is accompanied by an in-
crease in the share of income earned by the poorest,
incomes of poor people will rise faster than average in-
comes. Similarly,if economic growth is accompanied by
a decline in this share, growth in the incomes of poor
people will lag behind growth in averageincomes.
The same is true for poverty rates. For a given rate of
economic growth, poverty will fall faster in countries
where the distribution of income becomesmore equal than
in countries where it becomes less equal. For example, in
Uganda growth with rising equality delivered strong
poverty reduction, while in Bangladesh rising inequality
tempered the poverty reduction from growth (box 3.4).
Another example is Morocco, where the number of poor
people increasedby more than 50 percent between 1990
and 1998, mainly becauseof declining real per capita pri-
vate consumption (-1.4 percent a year). In urban areas
the increase in poverty was dampened by a decline in in-
equality, while in rural areas rising inequality reinforced
the increase in poverty.1 9
Does growth itself lead to systematic increases or de-
creases in income inequality? Do the policies and insti-
tutions that contribute to higher growth increase or
decrease inequality? Does the regional or sectoral com-
position of growth affect changes in income inequality?
To answer these questions the chapter first looks at the
available cross-country evidence-and then turns to
more detailed country-specific evidence, which high-
lights the fact that changes in income inequality are often
driven by a complex array of opposing forces.
Many studies show that on average there is no sys-
tematic relationship across countries between growth
and summary statistics of income inequality such as the
Gini coefficient (figure 3.5).2° While this average rela-
tionship is of interest, so are the substantial deviations
around it.
The differencesin inequality at a given rate of growth
could reflect the fact that the combination of policiesand
institutions that led to this growth differed across
countries-and that these differences in policies matter
for income distribution. But at the aggregate cross-
country level, there is not much evidence that this is the
case.A recent study of growth and poverty reduction in
a sample of 80 industrial and developing countries found
that macroeconomic policies such as a stable monetary
 GROWTH, INEQUALITY, AND POVERTY 53

Box3.4 Figure3.5
Inequalitytrendsand povertyreduction Inequalityvariedwidelyinthe 1980sand 1990s
butshowednosystematic association
with growth
In Ugandagrowthwith risingequalitydelivered
strongpovertyreduction... Average
annual
growthinGinicoefficient
After decadesof war andeconomiccollapse,growth re- 20en
coveredinUganda in the 1990s,averaging morethan5per- . .
centayear.Injustsixyears(1992-98) the shareof Ugandans t *X ;* '
in povertyfell from56 percentto 44 percent.Thebenefits * .
of growthwere sharedby all incomegroups,by ruraland 0 .
urbanhouseholds, andby nearlyalleconomic sectors.Real .,. ** *
percapitaconsumption rosefor alldecilesof the population, * *:
implyinga reduction in povertyregardless of thepovertyline. - 0* * **e
Modestreductions in incomeinequalitymadegrowthes- .20
pecially effectivein reducing poverty,
withthe Ginicoefficient 5

fallingfrom0.36to 0.34duringthe fiveyears.Livingstandards

improved moreamongpoorerhouseholds. Consumption (per -40
adultequivalent) rose27percentforthe poorestdecile,com-
paredwith 15percentfor households in the richestdecile.
Amongcashcropproducers-especially coffeefarmers,ini-
tiallyas pooras the averageUgandan-povertyfell more 60
-20 -10 0 10 20
thantwice asfast asfor the countryasa whole.
Average annualgrowthinpercapitaconsumption
... whilein Bangladesh risinginequalitytemperedthe Percent
povertyreduction fromgrowth Note:Thedatacover65developingcountries.
In Bangladesh percapitaGDPgrewat about2 percentayear Source:WorldBankstaffestimates
based
ondatafromChenand
duringthe1990s,andpovertydeclined quiteslowly.Between Ravallion12000).
1983and1996the shareof peoplein extremepovertyfell _
from40.9 percentto 35.6percent-andthe sharein mod-
erate poverty from 58.5 percentto 53.1 percent.Rural
povertyin particularremainsveryhigh. vors growth and may lower icome iequality by im-
Whythe slow decline?Partof the answerlies in rising proving accessto credit.22
inequality,in both urbanand ruralsectors,especiallybe-
tween1992and1996,whenthe Ginicoefficientrosefrom When I retired, I had 20,000 rubles in my savings
0.26to 0.31.Depending onthe povertymeasureused,a fifth account. . . . But what the government did with it-the
to athirdof thepotentialpovertyreductionfromgrowthmay
havebeenlost becauseof higherinequality.If inequality government we truistedwith our money! They re-indexed
hadnot increased, the povertyratewouldhavebeenabout savings so that iniflation ate it! That money is now not
7-10percentage pointslowerin 1995-96thanit actually
was. enough for hread and water.
Thehigherinequalityin Bangladesh doesnot implythat -From a discussion.noup, IJkrainie
growthshouldnotbepursued. Tothe contrary,
fastergrowth
is neededto reducepovertyfaster,becausegrowth'snetef-
fect onpovertyreductionispositive.Alsorequiredareefforts Another possible explanation for the lack of associa-
to limit risinginequalityandto ensurethat growthreaches tion between growth and inequality is that countries with
ruralareas,wheremanyof the country'spoorpeoplelive. similar overall growth rates couldexperiencevery differ-

Source:
Appleton
andothers1999;
Wodon
1997,1999,
2000c.
policy, opennessto international trade, and a moderate-
size government raise the incomes of poor people as
much as average incomes.2 1 In other words, these poli-
cies did not systematically affect income distribution.
Other policies, such as stabilization from high
inflation, may even disproportionately favor poor peo-
ple (chapter 9). And greater financial development fa-
ent changesin income distribution becauseof differences
in the regional and sectoral composition of growth. If
growth bypassespoor regionsand poor people cannot eas-
ily migrate to regions where opportunities areexpanding,
growth can lead to rising inequality. If growth is concen-
trated in sectorsfrom which poor people are more likely
to derive their income, such asagriculture, growth can be
associatedwith declining income inequality.
In China much of the sharp increase in income in-
equality between the mid-1980s and mid-1990s reflects
the much swifter growth in urban areas relative to rural
54 WORLD DEVELOPMENT REPORT 2000/200I
areas. 23 India's states tell a similar story ofthe importance
of rural growth in poverty reduction (box 3.5). So does
Indonesia.24 A study of 38 developing countries found that
the variation in inequality reflects the abundance of arable
land, the prevalence of smallholder farming, and the pro-
ductivity of agriculture. 2 5 These findings underscore the
importance of removing policy biases against agriculture
for generating more equitable growth (chapter 4).
Cross-country evidence can take us only so far in un-
derstanding the factors underlying changes in the distri-
bution of income that make growth more or less pro-poor.
Careful country-specific analyses paint a more nuanced
picture, highlighting a complex set of reinforcing and
countervailing forces. These include changes in the dis-
tribution of education, changes in the returns to educa-
tion, labor market choices, and demographic changes (box
3.6). Those changes are the result of:
Box 3.5
What makes growth pro-poor in India?
Consistent with cross-country evidence for developing
countries, consumptionpoverty in Indiahas fallenwith the
growth in mean householdconsumption.Moreover,the re-
gional and sectoral composition of growth affects the na-
tionalrate of poverty reduction,withfar stronger responses
to ruraleconomic growth than to urban.And within rural
areas growth inagricultureand services has been particu-
larlyeffective in poverty reduction,while industrialgrowth
has not. hasnot.
In rural India higher agriculturalproductivityis crucial
for pro-poor economic growth. Data spanning 1958-94
show that higherreal wages and higherfarm yields raised
average livingstandards and did not affect income distrib-
ution.The result: less absolute poverty.
Theeffectivenessof nonfarmgrowth inreducingpoverty
has varied widely across states, reflecting systematic dif-
ferences in initialconditions.In states with low farm pro-
ductivity,low rurallivingstandards relativeto urban areas,
and poor basic education, poor people were less able to
participatein the growth of the nonfarmsector. The role of
initialliteracyis notable:morethan halfthe differencebetween
the elasticity of poverty to nonfarm output for Bihar(the
state with the lowest elasticityin India)and that for Kerala
(thehighest)isattributableto Kerala'ssubstantiallyhigherini-
tial literacyrate. Women's literacyis a slightlymore signifi-
cant predictorof growth's contributionto povertyreduction
than men's literacy.
For poor people to participatefullyin India's economic
growth,agriculture,infrastructure,and social spending (es-
peciallyin laggingruralareas) need to be higherpriorities.
Source:Ravallionand Datt1996, 1999.
9 Market forces, such as changes in the demand for
labor.
* Policies, such as public investment in education.
* Social forces, such as higher participation of women
in the labor force or changes in practices discriminating
against women and ethnic minorities.
A Institutional forces, such as changes in legal restrictions
on the ownership of property by women or ethnic
groups.
Not every increase in income inequality should be seen
as a negative outcome. As economies develop, income
inequality can rise because the labor force shifts from agri-
culture to more productive activities. For example, if
wages are lower in agriculture than in industry and ser-
vices and the labor force shifts toward those two sectors,
many summary statistics, especially those sensitive to
changes at the bottom end of the income distribution,
will show increases in inequality despite an overall de-
cline in poverty. These trends should not be seen as neg-
ative if:
i The incomes at the bottom rise or at least do not fall.
* Thedevelopmentprocessexpandsopportunitiesforall.
Figure 3.6
Initial inequalitiesinfluencethe pace of poverty
reduction
Averageannualreductionin incidenceof poverty
associated
with 1 percentincreasein averageper capita
~~~~~~~~~~~~consumption
Percent
3.0
2.5
2.0
I
1.5
1.0
1.0
0.5
0
0.2 0.4 0.6
InitialGini coefficient
Note:Thedatacover65 developingcountriesin the 1980sand
1990s.Theincidenceof povertyisthe shareof the populationliving
on lessthan $1 a day.
Source:World Bankstaff estimates based on the methodology of
Ravallion11997a)and data from Chen and Ravailion(2000).
 GROWTH, INEQUALITY, AND POVERTY 55

Box3.6
Complexpatterns of distributionalchangein three economies

Observedchangesinthe distribution of incomereflectacomplex changesinthe distributionof educationdidnot offsetthis.While

arrayof factors-amongthem,changesin the distributionof as- educationalattainmentrosefasterfor the lesseducated, the re-
sets,changes inthe returnsto theseassets,labormarketchoices, turnsto highereducationwere sufficientlyhighthat the addi-
anddemographics. Brazil,Mexico,andTaiwan,China,showhow tionalearningsdueto greatereducationdisproportionately
favored
theseforcescan reinforceandoffsetoneanotherto resultin in- the more educated.Superimposed on this were importantre-
equalitythatis respectively
lowered,increased, andunchanged. gionaleffects,withwideningrural-urban
realwagedifferences con-
tributingsubstantially
to inequality,
despitesomeconvergence of
Brazil-inequalitylowered urbanandruralreturnsto educationandexperience.
Incomeinequality declinedin Brazilbetween1976and1996,with
the Ginicoefficientfallingfrom0.62to 0.59.Duringthe samepe- Taiwan,China-inequalityunchanged
riodthe returns to education becamemoreunequal: bothwageearn- Notedfor its low andstablelevelof inequality, Taiwan,China,has
ers andself-employed workerswith moreeducationsawlarger hadaGinicoefficient of about0.30forthepast30years. As inBrazil,
increases inearnings thantheirless-educated counterparts,even this outcomereflectsa varietyof opposing forces.Despitea rapid
aftercontrolling for ageandgender.Therewerenochangesinthe increaseintheirsupply, more-educated workerssawlargerincreases
returnsto experienceandonlysmalldeclinesin the paygapbe- inearnings thanless-educated workers.Thiswasmorethanoffset
tweenmenandwomen,so overallearnings inequalityincreased. by greaterequalityin thedistributionof educationandgreaterlabor
Thisdisequalizing effectwasmorethanoffsetby threefactors: marketparticipation by women.Thepatternof taxesandtransfers
X Thedistribution of education becamemoreequal. wasalsoequalizing, with the effectthatthe distribution
of individ-
* Averageeducational attainmentrosefrom3.8to 5.9yearsof ualincomebecamemoreequal.Interestingly, however,incomein-
schooling, andhigherlevelsof schooling (particularly
forwomen) equalityatthehousehold levelincreased,asmanyof thenewfemale
contributed to a noticeable reductioninfamilysize,withtheav- entrantsto the laborforcecamefrominitiallybetter-offhouseholds.
eragehousehold fallingfrom 4.3to 3.5 members.Sincefam-
ily sizefell morefor poorerhouseholds, inequalityfell.
* Inequality inthe returnsto characteristics otherthaneducation
seemsto havefallen,suggesting a reductionin labormarket
Theseexamples showthatsimpletrendsinsummary measures
segmentation during1976-96anda possible declineinregional
of incomeinequalitycan disguisemajorstructuralforces.Some
inequalities. of them,suchaschangesinthe distributionof education, canbe
influencedby policy-though this takestime. Others,suchas
Mexico-inequalityincreased changes inthe returnsto education,
reflectprimarily
marketforces
Mexico'sGinicoefficientrosesharplybetween1984and1994, andarelessamenable to directpolicyinterventions.
AndasTai-
from0.49to 0.55.As in the previoustwo examples, changesin wan,China,shows,taxandtransferpoliciescancounterincreases
the returnsto educationwerea stronglydisequalizing
force.But in primaryincomeinequality.

Source:
ForBrazil,
Ferreira
andPaes
deBarros
(1999b);
forMexico,
Legovini,
Bouillon,
andLustig(1999);
andforTaiwan,
China,
Bourguignon,
Fournier,
andGurgand
(1998).

* The observed trends are not the result of dysfunctional
forces such as discrimination.
u The number of poor people falls.
Initial inequality and poverty reduction
Even when the distribution of income itself does not
change with growth, countrieswith similarratesof growth
can have very different poverty outcomes, depending on
their initialinequality.Other things being the same,growth
leads to lesspoverty reduction in unequal societiesthan in
egalitarian ones. If poor people get a small share of exist-
ing income and if inequality is unchanged, they will also
get a small share of the new income generated by growth,
muting the effectsof growth on poverty.Evidenceconfirms
this: when initial inequality is low,growth reducespoveirty
nearlytwiceas much as when inequalityis high (figure3.6).
Initial inequality in income is not the whole story--
for inequality in other dimensions matters too. The sen-
sitivity of poverty to growth depends a great deal on
initial inequality in poor people's accessto opportunities
to share in this growth. If disparities in educational at-
tainment mirror disparities in income, poor people may
not have the skills to find employment in dynamic and
growing sectors of the economy. This effect is coin-
pounded by gender inequality in access to education
(chapter 7). In addition, if fixed costs or overt policybar-
riers hinder movement from remote, rural, and eco-
nomicallydepressedregionsto more vibrant urban centers,
56 WORLD DEVELOPMENT REPORT 2000/ZOOI
poor people will be less likely to take advantage of op-
portunities to migrate (box 3.7).
They have alwnys excluded uis Mlavas, theY have
discriiminated against us. They cut down the tree, bitt
*forgot to pll down the roots. That tree is nowsprouting.
g ilOigI8-.Uli.
l 'n1/'~ ( -'I
If social inequities-such as caste systems or discrim-
ination against indigenouspeoples-confine members
inationagainstindigenouspeoplescon
emembefor
of disadvantaged groups to employment in stagnant sec-
tors, poor people will benefit less from growth (chapter
7). Or if ethnic discrimination in the marketplace leads
to different returns to the same level of education, growth
will be lesseffectivein reducing poverty for the group dis-
criminated against. A study in Latin America found that
in several countries differences in earnings between in-
digenous and nonindigenous people cannot be explained
by differencesin skillsor experience, suggestingthat dis-
crimination in the labor market may be to blame.2 6 These
results bring to the fore the importance of eliminating so-
cial barriersfor women, ethnic minorities,and sociallydis-
advantaged groups in making growth broad based.
Initial inequality and growth
High initial inequality reduces the poverty impact of a
given rate of economic growth. It can also undermine
poverty reduction by lowering overall economic growth.
Early thinking on the effects of inequality on growth sug-
gested that greaterinequality might be good for growth-
for example, by redistributing income to the rich, who
save, from the poor, who do not. This view implied a
tradeoff-more growth could be bought for the price of
more inequality, with ambiguous effects on poor people.
More recent thinking-and empirical evidence-
weaken the case for such a tradeoff: lower inequality can
increase efficiency and economic growth through a va-
riety of channels. Unequal societiesare more prone to dif-
ficulties in collective action, possibly reflected in
dysfunctional institutions, political instability, a propen-
sity for populist redistributive policies, or greater volatil-
ity in policies-all of which can lower growth.And to the
extent that inequality in income or assets coexists with
imperfect credit markets, poor people may be unable to
invest in their human and physical capital, with adverse
consequences for long-run growth.
The effects of inequality on growth have been sub-
jected to considerable empirical scrutiny. Evidence on
Box 3.7
Diversificationand migration in rural China
Forruralagricultural
households in China,opportunities
for
off-farmemploymenthavebeen an importantsourceof
growthin incomes.Theseopportunities canbeequalizingor
disequalizing.
Tothe extentthatdiversification intononfarm
;ilemployment
, reflectsa pull factor-higher returnsoff the
farm-diversification
canbedisequalizing asricherandbetter-
educatedworkerstakeadvantage of theseopportunities.
To
theextentthatdiversification
reflects
a survivalmechanism
thepoorest,itcanbe equalizing.
Evidence fromfourprovincesinChinasuggeststhatthe
pullfactorhasbeenmoreimportantthanthe survivalmech-
anism,withaccessto nonfarmemployment accountingfor
a nsing inruralareasbetween1985
shareofincomeinequality
and1990.Evidencealso suggests thateventhe modest gap
(byinternational
standards)betweenfemaleandmaleedu-
cationalattainmentexacerbatesthese trends,with less-
educatedwomenlesslikelyto findoff-farmemployment. In
contrast,migration
hashad equalizing
effectsonincome.Sur-
veydatafromthefourprovincesshowthatprivatetransfers
migrants'remittances)
reflecting
(largely havebeenan equal-
izingforce.
Source: World Bank1997b.
the impact of inequality in assets-and gender
inequality-is generally clearest.A recent study of sugar
cooperatives in India found that those that are most un-
equal (in land ownership among cooperative members)
are the least productive.77 Various studies have also
found an adverse effect of land inequality on growth.28
A study in China found that living in a high-inequality
area reduced growth rates at the farm household level,
controlling for a household's human and physical cap-
ital.2 9 Other studies have found evidence of a link be-
tween education and gender inequality and growth.30
In contrast, evidence on the effect of initial income in-
equality on subsequent growth is more mixed. Some
studies have found negativeeffects.3 1 Others have found
positive effects.32 Still others have found different effects
over different ranges.33
Theseresultsopen the possibilitythat policiesto improve
the distribution of income and assets can have a double
benefit-by increasinggrowth and by increasingthe share
of growth that accruesto poor people.This is not to say that
everypro-equity policywill have such desiredeffects.If the
reduction in inequality comes at the expense of the other
factors conducive to growth (discussedin the early part of
this chapter), the gains from redistribution can vanish. Ex-

Box 3.8
Redistributioncan be good for efficiency
Redistribution need not compromiseefficiencyandgrowth.Insev-
eral instances redistributivepoliciescan increase asset accumu-
lationby poor people-while improvingefficiencyand growth.A
few recent studies illustrate the possibilitiesfor win-win out-
comes, further strengtheningthe case for redistribution.
Landreformis a classicexample of a redistributivepolicy.Op-
eration Barga,a tenancy reform in the Indianstate of West Ben-
gal in the late 1970s and early1980s,is one of the few examples
of large-scaletransfers of property rights not accompanied by
majorsocialupheaval.The operationwas associated with an 18
percent increase in agriculturaloutput in the state (Banerjee,
Gertler,and Ghatak1998).
Redistributioncan also be a source of efficiencygainsif trans-
fers to poor people improvetheir human capital.Publicprovision
of infrastructuretargeted to poor peopleis an importantexample.
Massiveprimaryschoolconstruction(61,000newschoolsbuiltand
staffed in five years) under Indonesia's INPRES(presidential in-
structions)program,the mainmechanismfor redistributing the gain
fromthe oil boom in Indonesia,substantiallyincreasededucation
and income.Theprimaryschool graduationrate rose 12 percent,
and male wages 5 percent (Duflo2000b).
propriation of assets on a grand scale can lead to political
upheaval and violent conflict, undermining growth. And
sometimes attempts to redistribute income can reduce in-
centives to save, invest, and work. But there are a number
of win-win possibilities (box 3.8). Policies should focus on
building up the human capital and physical assets of poor
people by judiciously using the redistributive power of
government spending and, for example, market-based and
other forms of land reforms (chapter 5; box 5.12).
Economic growth and nonincome
poverty
Just as income poverty declines as average incomes in-
crease, so does nonincome poverty, such as in health and
education. Just as with income poverty, there are signif-
icant deviations around these general relationships: coun-
tries and regions with similar per capita incomes can
have quite different outcomes in nonincome poverty as
well. And just as with income poverty, these deviations
reflect a wide array of forces-including initial inequal-
ity, the effectivenessof public interventions, and the level
of development. Conversely, there is strong evidence
that better health and education outcomes contribute to
faster economic growth.
GROWTH, INEQUALITY, AND POVERTY 'j7
Universalpolicies(suchas pricingof governmentservices)can
have redistributiveand efficiencyeffects as well.Abolishingsec-
ondary school fees in Taiwan, China, in 1968 and introducing
compulsoryeducationbenefited poorer childrenmore than richer
children (Spohr 2000). It also substantially increased school
attainment(0.4year for males)and laborforce participation,trans-
latinginto higherearnings (Clarkand Hsieh1999).
Directincomeredistribution(throughcash transfers)is rare in
developingcountries.A concern is that cash may not be spent in
the most efficiency-enhancing ways. InSouth Africaat the end of
apartheid,the smallpensionprogramwas dramaticallyexpanded
for the black population.In 1993the pensionamounted to twice
the medianincomeforblacksin ruralareas (Caseand Deaton1998).
When the pension was received by the maternalgrandmothers
of girls,it had largeeffects on nutrition-halvingthe gap in height
betweenthese girlsand those of the same age inthe UnitedStates
(Duflo2000a).Other studies haveshown, however,that the pen-
sion, when receivedby an elderlywoman, also led to a reduction
in prime-agemale laborsupply(Bertrand,Miller,and Mullainathan
1999). The results suggest that cash transfers can-but may
not-lead to efficiencygains.
Across countries, and across individuals within coun-
tries, there are strong correlations between health and ed-
ucation outcomes and incomes. Richer countries arLd
richer individuals within countries have lower rates of mor-
tality and malnutrition. 3 4 Within and between counI-
tries both the quantity and the quality of education
improve with income-although quality is difficult t-o
measure. 35 Disparities in educational attainment also
decline with income.
These strong correlations reflect reinforcing causal
effects from higher income to better health and educa-
tion outcomes-and from better health and education
to higher income. For individuals, this is not surprising.
Ill health and malnutrition reduce productivity and time
spent working, effects that vary with the level of educa-
tion. For example, a study of Brazilian men showed that
adult height was strongly associated with wages-and that
wages increased faster with height among individuals
with some (as opposed to no) education.3 6 Conversely,
individuals with higher incomes can better afford to in-
vest in health and education.37 Many studies document
the positive effects of parental education on children's
health and education.
Similar patterns hold for countries, with positive ef-
fects of higher per capita income on infant mortality.3 8
58 WORLD DEVELOPMENT REPORT 2000/200I
Other studies have documentedthe benefitsof lower
mortality for faster growth, with most of the growth
payoffat lowlevelsof income. 39 Andwehavealreadyseen
the evidencethat bettereducationoutcomesleadto faster
growth.
Moreover,thereis someevidencethat theserelation-
ships are not linear,with strongerincreasesin healthas-
sociatedwith growth in poorer countries and regions.
Fairly small differencesin economicgrowth rates can
thus have large impacts on human developmentout-
comesin such countries.One studyestimatedthat had
growthrates in the developingworld (excludingChina
and India) been as high in the 1980sas they werein the
1960sand 1970s,656,000deathscouldhavebeenaverted
during the 1980samongchildrenunder five.4 0
Thesereinforcingeffectsfrom humandevelopmentto
economicdevelopmentand backsuggestthe possibility
of viciousand virtuouscircles.Poorcountriesand poor
people can be locked in a viciouscircle,as low human
developmentdiminisheseconomicopportunities,mak-
ing it more difficultto investin healthand education.In
contrast,well-targetedpublicinterventionsin healthand
educationcan contribute to a virtuouscircle of greater
economicopportunitiesgeneratingresourcesfor further
investments(chapter5).
The considerablevariations in country experience
around these generalrelationshipsagainreflecta com-
binationof factors.One isinequalityin income."Wehave
seen that the effectsof incomeon healthare most pro-
nounced at low levelsof income.This impliesthat the
same rate of economicgrowth can havevery different
health and educationoutcomes,dependingon the ini-
tial distributionof incomeand on howthat distribution
changeswith growth. In particular,growth accompa-
nied by a reduction in inequalityis more likelyto lead
to better health outcomes.
Research has found evidencethat the correlation
acrosscountriesbetweenaveragehealth indicatorsand
average income vanishesafter controlling for differ-
ences in the incidenceof income poverty and in pub-
lic spending.4 2 The same research has found that
cross-country differences in public health spending
matter more to the health of the income-poorthan to
others: the nonpoor are better able to protect their
health from lowerpublic spending. These resultssug-
gest that growth improvesaveragehealth attainments
through its ability to reduce income poverty and per-
mit more pro-poor social spending.
Figure3.7
Acrosscountries, the ratioof femaleto male
literacyriseswith percapitaincome
Ratioof femaleto male literacyrate
Average, 1990-98
120
0 .
100
0 .
60 *. t ** S
e .*
40 *
5.
20
0
100 1,000 10,000
Per capita GDP
1995 U.S. dollars (PPP);log scale
Source: World Bank data.
Nonincomeinequalitiesmatter as well.Discrimina-
tion by genderand ethnicity-in the allocationof pub-
licspendingfor educationand healthor in the operation
of educationand health facilities-can lead to differ-
encesin educationand healthachievements. Genderdis-
parities in educational attainment are especially
pronouncedin poorcountries(figure3.7). In the Indian
stateof Kerala-whichhas a longhistoryof equitablegen-
derrelations-educationandmortalitydifferlittlebetween
men and women.But in such statesas Uttar Pradesh-
wheregenderdiscriminationis high-the femaleliteracy
rate is lessthan half the male rate and the femaleto male
ratioin the populationis a disturbing87.9to 100.43Cross-
country studies havealsoidentifiedgeographicfactors,
ethnic fragmentation,and especiallyfemaleeducational
attainment as important in explaining differencesin
health outcomesat a givenincome.44Finally,the qual-
ity and quantityof publicspendingmatteraswell,though
the sizeof the impacton poor peopledependsgreatlyon
supportivepoliciesand institutions(chapter5).
* *
This chapter has shown the importance of growth
for poverty reduction, particularly for income and
 GROWTH, INEQUALITY, AND POVERTY 'j9

human development. It has also shown how low and
declining inequality enhances the impact of growth on
poverty. Growth can be made more equitable by re-
ducing inequality in access to assets and opportunities.
This requires opening market opportunities to poor peo-
ple and building up their assets. It also requires mak-
ing state institutions work better for poor people, re-
moving social barriers, and supporting poor people's
organizations. These issues are taken up in subsequent
chapters.

Making Markets
Wlork Better
for Poor People
.. ~l..arkets matterfor the poor becausepoor
peoplerelyon formaland informalmarketsto sell
theirlaborandproducts,to financeinvestment, andto
insureagainstrisks.Well-functioning marketsareim-
portantin generating growthand expandingoppor-
tunitiesforpoorpeople.Thatiswhymarket-friendly
reformshavebeenpromotedby international donors
and by developing countrygovernments, especially
thosedemocratically elected.
Butto developmarketsandtheinstitutions thatsup-
portthemisdifficultandtakestime.Attimes,reforms
to buildmarketsfailentirely. Whentheysucceed,they
frequentlyimposecostson specificgroupsin society.
Whenthelosersfromreforms indudepoorpeople,who
areparticularlyvulnerable to shocks,countrieshavea
special obligation
to easetheburdenofreform. Andeven
whenmarketswork,societies haveto helppoorpeo-
ple overcomethe obstaclesthat preventthemfrom
freelyandfairlyparticipating in markets.
In the 1950sand 1960smanyofthoseshapingpol-
icybelievedthat economicdevelopment and poverty
CHAPTER 4
reductionrequiredactiveparticipationofthestateand
protectionoflocalindustry.Thisinward-looking,state-
leddevelopment pathwasadoptedbya widearrayof
countriesthroughouttheworld,withvaryingdegrees
of success.Manycountriesadoptedprotectionism,
government controlof investment, andstatemonop-
oliesin keysectors.In countriessuchas Indiathis
strategyresultedin persistently
slowgrowth.In other
countries,particularlyin LatinAmerica,thisstrategy
initially
delivered
stronggrowththroughthe1960s,but
growtheventually falteredas countrieswerebuffeted
by oil shocksin the 1970sand thedebtcrisisof the
1980s.Andin Chinain thelate1970stherewasa grad-
ual realization
that the economy,especially the agri-
culturalsector,hadnotrealizeditsfullpotentialunder
heavystatecontrol.
The increasingdisenchantmentwith inward-
looking,state-leddevelopment led nationalgovern-
ments to implementreformsthat replacedstate
interventionin marketswithprivateincentives, pub-
licownershipwithprivateownership,andprotection
61
62 WORLD DEVELOPMENT REPORT 2000/2001
of domestic industrieswith competition from foreignpro-
ducers and investors.2 Where such market-friendlyreforms
have been successfully implemented, on average eco-
nomic stagnation has ended and growth has resumed.
But in some cases reforms were not successfullyim-
plemented, often with particularly severe consequences
for poor people.The broad diversityof failedreforms does
not lend itself to easygeneralization.3Some reforms pro-
ceeded too quickly and failed for want of supporting in-
stitutions. Others proceeded too slowlyand were captured
and undermined by special interests. Yet others were im-
posed by government elites and foreign donors and
foundered for lack of strong domestic leadership and a
broad-based commitment to reform.4
The debate about reforms is therefore not over a
choice between reforms or no reforms: the absence of re-
forms to develop vibrant, competitive markets and cre-
ate strong institutions condemns countries to continued
stagnation and decline. Nor is the debate over a simplistic
dichotomy between gradualism and shock therapy: re-
forms can proceed either too quickly or too slowlyto suc-
ceed. Rather, the debate is on how reforms to build
markets can be designed and implemented in a way that
is measured and tailored to the economic, social,and po-
litical circumstances of a country.5
Inevitably, market-oriented reforms have different ef-
fects on different segments of society. Every reform pro-
gram has its winners and losers, and poor people may be
found in either group. The particular vulnerabilityof poor
people demands a careful assessmentof the likelypoverty
impact and the implementation of appropriate compen-
sating policies.6 It also callsfor carefulconsiderationof the
pace of reformsin the light of the likelyeffectson poor peo-
ple. Experienceshowsthat direct dialogue with poor peo-
ple can be particularly effectivein informing this process.
Even when markets function, they do not alwaysserve
poor people as well as they could. Physicalaccess to mar-
kets can be difficult for poor people living in remote
areas. Regulatory barriers often stifleeconomic activity in
sectors and regions where poor people are likely to seek
jobs. And access to some markets, especiallyfor financial
services,can be difficult for poor people since they often
engage in small transactions,which traditional market par-
ticipantsfind unprofitable or insignificant. Investments
in infrastructure, lighter regulatory burdens, and innov-
ative approaches to improving accessto financial markets
can therefore do much to ensure that the benefits of mar-
kets are shared by poor people.
This chapter addresses these issues in turn. It first
considers the widely varying experience of countries that
implemented market-oriented reforms over the past 20
years,highlighting both successstories and the severecon-
sequences of failed reforms. It then illustrates the com-
plex effectsthat market reformshave on poor people, with
examples from three areas: agriculture, fiscal policy, and
trade. Last,it discusseshow lightening the regulatory bur-
den, promoting core labor standards, and expanding
microfinance can be beneficial in improving the terms on
which poor people participate in markets.
Have market reforms delivered
growth?
In the 1980s and 1990s much of the developing world
moved towardimplementing market-friendlyreforms.The
motivation for reforms and their scope and pace varied
widely.In China, for example,the "householdresponsibility
system" replaced communal farming and created new
incentives for rural householdsto produce, invest, and in-
novate. These reforms were provoked neither by macro-
economic crisis nor by ideologicalepiphany; rather, they
reflecteda growing realizationthat China'sagriculturalpo-
tential was not being fulfilled.These initial agriculturalre-
forms were followed by the introduction of market
mechanisms throughout the economy. In other countries
macroeconomiccrisesprovided the catalystfor reform: in
Mexico, for example, the debt crisis of the 1980s was fol-
lowed by the introduction of wide-ranging economic re-
forms. And in the countries of Eastern Europe and the
former Soviet Union the political transition precipitated a
dramatic progresstoward markets that succeeded as spec-
tacularly in some countries as it failedin others.7
As a result of this move toward reforms the economic
landscape in many developing countries-but not all-
has been significantly altered. Government involvement
in economic activity has been scaled back. Domestic
markets are more open to international trade and capi-
tal flows. Revised tax codes are in place. And generally
markets, not governments, determine prices, output,
and the allocation of resources. Many-but not all-of
these reforms reflected the principles of the so-called
Washington consensus, which laid out 10 policy priori-
ties that were adopted in different combinations by many
countries (box 4.1).
Given the wide diversity of reforms implemented by
different countries at different times and under different
 MAKING MARKETS WORK BETTER FOR POOR PEOPLE 63

Box 4.1 Figure 4.1

The Washington consensus Indicationsof successfulpolicy reformsin the
developingworld
The Washingtonconsensusof market-friendlyreformsrefers
to the following 10 objectives of policy: Medianannualinflationrate
* Fiscaldiscipline. 25 Percent
* Redirection of public expenditure toward education,
health,and infrastructure investment.
* Tax reform-broadening the tax baseand cutting marginal 20
tax rates.
* Interestratesthat are marketdeterminedandpositive(but
moderate) in realterms. 15
* Competitive
exchange
rates.
* Tradeliberalization-replacement of quantitativerestric-
tions with low and uniform tariffs. 10
* Opennessto foreign direct investment.
* Privatizationof state enterprises.
* Deregulation-abolishment of regulations that impede 5
entryor restrict competition,except for those justified on
safety, environmental,and consumerprotectiongrounds,
and prudential oversight of financial institutions. 0
I Legal security for property rights.

Source:
Williamson1993. Medianblackmarketpremiumonforeignexchange
Percent
50

circumstances, summarizing overallprogress is difficult.

Nevertheless, encouraging indicators are clear (figure
4.1). For example, typical inflation rates in developing
countries fell from around 15 percent in the early 1980s 30
to 7 percent in 1997, indicatinga broad trend towardmore
disciplined monetary policy. More important, many 20
countries have escaped the scourge of chronic bouts of
high inflation and hyperinflation. The black market pre-
mium on foreign exchange-a sure indicator of unreal- 10
istic and nonmarket exchangerates-fell from 25 percent
for a typicaldeveloping country in the mid- 1980s to only 0
1960 1970 1980 1990 1997
5 percent in the late 1990s.
Reducing barriers to international trade and capital Note: Data are for all developing countries. The band around the
movements has been a central part of many reform pro- is the 95 percent
central relationship confidenceinterval.Dataon
blackmarketpremiumsfor 1994and 1995are unavailable.
grams. In Latin America average tariffs were reduced Source:Easterly2000b.
from 50 percent in 1985 to 10 percent in 1996, and max-
imum tariffs fell from an averageof 84 percent to just 41
percent.8 By 1996 nontariffbarriers affected only 6 per- on average, countries with market-friendly policies such
cent of imports, down from 38 percent before reform.9 as openness to international trade, disciplined monetary
Reforms have also been widespread in other areas, such and fiscal policy, and well-developed financial markets
as liberalizing investment regulations, reducing or elim- enjoy better long-run growth performance than countries
inating a large assortment of subsidies to bring down fis- where such policies are absent (chapter 3).
cal deficits, and privatizing many state enterprises. Only There is alsoevidence that reforms that move countries
in labor markets have reforms generally been slow.iO closer to such market-friendlypolicies also contribute to
Have these reforms delivered the expected growth better growth performance in the medium term. Cross-
payoff?A large empirical literature has documented that, country studies of the impact of reforms typically either
64 WORLD DEVELOPMENT REPORT 2000/200I

Figure 4.2 predict cross-country differences in economic perfor-

Reforms deliveredgrowth in Latin America, mance. But many developing countries were buffeted by
although the gainsvaried large external shocks.World interest rates rose sharply,in-
Additional per capita growth due to reforms in the 1990s creasingthe burden of debt serviceobligations.Growth
Percentage points in the industrialcountries slowed,loweringgrowth in their
8 Easterly,Loayza, developingcountry tradingpartners.In somecasesthese
7 and Montiel 1997 shockserodedthe benefitsof reformsthat werebeingim-
6 Lora and plemented concurrently.
Barrera 1997 At times, however,reform programs have failedto de-
5 Fernandez Ar as liver as much as expected-and at times reformshave
4 andMontiel1997 r failedentirely.Considerwhat went wrong in EastAsia,
3 countries of the former SovietUnion, and Africa(box 4.2).
The grimlessonsof thesefailures,and the heavyburdens

I IU
2
2 l rl l I
1 l
o_ _i z _^ L^ - ---*that
-1.
Argentina Brazil Costa Mexico Peru
Rica
Source: As noted in the figure.
compare the performance of countries before and after re-
forms or else examine whether changes in measures of re-
forms explain changes in growth rates. Reforms are
measured indirectly as changes in such variablesas trade
volumes, tariff rates, inflation rates, or budget deficits.
Suchstudiesoften find a strong growthpayofffrom reforms.
Figure4.2 summarizesthe results of three such studies for
Latin America, which found a significantgrowth impact
of reforms. Similar studies of the transition economiesof
Eastern Europe and the former Soviet Union, where suc-
cess in implementing market reforms has varied widely,
found that countries that implemented reforms forcefully
and early(and enjoyedfavorableinitialconditions)achieved
1
stronger growth than reform laggards." A 1999 study of
India finds that the states that implemented reforms saw
faster growth and stronger improvements in education
and primary health care than those that did not."2
This does not mean that the developing world as a
whole enjoyed rapid growth as a result of reforms in the
1980s and 1990s. Indeed, growth in the developing
world has been disappointing, with the typicalcountry
registering negligible growth.
A recent study argues that this disappointing growth
should not be attributed to the failure of reforms.13 De-
spite slowoverallgrowth, the study found that differences
in indicators of market-friendly policies continued to
nf they placed on poor people, underline the importance of
a measuredand realisticapproachto reformsto ensure
4
areattained.1
theirobjectives
A note of caution on the future of reforms. In many
Latin cases the reforms discussed above are straightforward
America "first-generation"reforms,such as stabilizingfrom high
inflation, moderating chronic budget deficits, and dis-
mantling the most egregious trade barriers. Consolidat-
ing the gains from these reforms often requiresinstitution
building in much more difficult areas, such as develop-
ing an independent judiciary, creating independent and
effective regulatory agencies,and instilling professional-
ism in the public sector.Such "second-generation"reforms
are not only much more complex and take much more
time-they are also often likely to be opposed by pow-
erful and entrenched interests.15 This is not to say that
such second-generation reforms should be postponed-
precisely because they take time to bear fruit, it is im-
portant to embark on them as promptly as possible.
In sum, market-oriented reforms have been wide-
spread though uneven throughout the developing world.
On averagethey have deliveredlowerinflation and higher
growth, both powerful forcesfor reducingincome poverty.
But reforms can also go awry, with painful consequences
for poor people. Lackof supporting institutions, mistakes
in sequencing reforms, and the capture of the reform
process by powerful individuals or groups lie at the bot-
tom of most failed reforms.
Have market reforms delivered
benefits to poor people?
Even when market-friendlyreforms have succeededin de-
livering growth, the effects on the incomes of poor peo-
ple have varied. This reflects both initial inequalities in
 MAKING MARKETS WORK BETTER FOR POOR PEOPLE 6(5

Box 4.2
Why do reforms sometimes fail?

Reforms can go awry when supportiveinstitutions are absentor nopoly rights and to bias the awarding and pricing of public con-
powerful individualsor groups manipulatethe results. tracts. State capture runs counter to the premises of a free and
fair competitive market economy-and contributes to increasing
Incomplete financial sectorreforms contributedto the inequality.State capture is also widespread. In severalcountries
East Asian crisis of the former SovietUnionmorethan 30 percent of firms surveyed
During the 1990sseveralemerging economies in EastAsia liber- in a business environmentsurvey reportedthat they hadsuffered
alizedtheir domestic financial markets and lifted capitalaccount as a result of successful state capture by their competitors (Hell-
restrictions.In the Republicof KoreaandThailandespecially,a surge man and others 2000).
of capital inflows, often through newly formed nonbank financial Market economies cannot function well where the institu-
institutions, placed heavyfinancial stresses on banks. Prudential tional and incentiveenvironment permits such corruptionto flour-
regulationof banksand nonbankfinancial institutionsdid not keep ish. Worse, countries mayfall into viciouscircles,with incomplete
pacewith these developments,andtherewas rapidgrowthin often- reforms creating new incentives for corruption. Fighting the cor-
unhedgedshort-termforeign currencyliabilities.Suddenexchange rosiveeffects of state capture requiresmuch deeperinstitutional
ratefluctuations in the summer of 1997wreaked havoc on these development-in the organization of the political system, the
foreign currency exposures, contributing to the depth of the en- checks and balancesamong core state institutions, and the rela-
suing crisis (World Bank 1998f). tionships between state and firms and between state and civil
This experience matches a broader pattern emerging from society.
cross-countryanalysis:financialreforms unaccompaniedby ade-
quate supervisory institutions are a significant determinant of Inadequatepublic investment and excessivebureaucracy
bankingcrisesworldwide (Demirgug-Kuntand Detragiache1998). have underminedmarket reformsin Sub-SaharanAfrica
At the root of the 1995Mexican peso crisis were inadequaciesin SeveralAfricancountries havefailedto grow since the mid-1980s,
the bankprivatizationprocessand in financialliberalization(Lustig when, with the support of internationalfinancial institutions,they
1998).These experiencesdo not invalidatethe importance of re- beganimplementingmarket reforms, especiallyin agriculture.The
forms in developing financialmarkets. In fact, the effective inter- results havebeen less than spectacular,in partdue to inadequate
mediation of savingsto productiveinvestment was a contributing publicinvestment and persistent red tape (World Bank2000b).
factor in EastAsia's remarkabledevelopmentsuccess,a success African farmers, like those in other parts of the world, re-
that dwarfs the setbacks of the recent crisis. But incautiousand spond vigorouslyto price and nonprice incentives. But if public
excessivelyrapidreforms can culminate in crises. infrastructure-such as roadsto remote agriculturalareas-is un-
developed or underdeveloped,the impactof pricingandmarketing
Grand corruptionsubverted reformsin countriesof the reforms on output is muted. Inadequate infrastructure affects
former Soviet Union other sectors as well. Business surveys carried out in a number
The state steals from us all the time, so deceiving the state is of African countries in 1996-97 consistently point to the poor
not a sin, quality of infrastructure services as a critical barrier to expansion
-From a discussiongroup, Ukraine into labor-intensiveexportsin responseto trade reform. In Uganda
transport and other costs increasedthe cost of capital goods by
What kind of government do we have?One hand g~ivesand the almost half. And in Zimbabwe poortransport services mean that
other takes away! delivery of inputs is unreliable,forcing firms to hold large inven-
-From a discussiongroup, Ukraine tories despite high interest rates.
These difficulties have been compounded by a lack of im-
In the countries of the former Soviet Union market reforms and provementin transparencyand accountability.Although legaland
perceptions of corruption are inextricably intertwined (see, for regulatory changes are often integral parts of reform packages,
example,Narayan,Patel,Schafft, Rademacher,and Koch-Schulte their implementation is often flawed or half-hearted.As a result,
2000).This is understandable:most of these countries score very regulatory barriers to competition remain serious obstacles,
poorlyin cross-countrycomparisonsof corruption,and encounters and corruption, red tape, and lack of transparency continue to
with corruption are dispiritinglyfrequent for many firms and indi- impede trade and investment by raising costs. Business surveys
viduals. Corruption has coincided with worse macroeconomic often also identify corruption and bureaucraticred tape as bar-
performanceand deeper output declines as these countries have riers to business expansion and diversification in severalAfrican
wrestled with the transition to a market economy. countries. For example, it can take more than a week for inter-
A particularlyperniciousform of corruption is "state capture," mediate inputs to clear customs on the Ugandanborder, and de-
referringto the abilityof firms and powerful individualsto influence lays of more than a day are routine at customs checkpoints in
the formation of new laws and regulationsto their own advantage. southern Africa. These obstacles are symptomatic of larger in-
This may involve manipulatingthe judicial, executive,and legisla- stitutional failures that policymakersmust address if reforms are
tive branchesof governmentto obtainspecial privilegesand mo- to be effective.

U
66 WORLD DEVELOPMENT REPORT 2000/2001
income and opportunity, and the effects of reforms on
growth and inequality. What has actually happened?
And what can be learned from this experience with mar-
ket-friendly reform?
Cross-country evidence suggeststhat macroeconomic
reforms on average have had little effect on income dis-
tribution. For example, recent studies have examined
the impact of market-friendlypolicies-such as openness
to international trade, low inflation, a moderate-sizegov-
ernment, and strong rule of law-on the incomes of
poor people in a large cross-country sample. The find-
ings: these policies on average benefit poor people as
much as anyone else.16 Some policies, notably stabiliza-
tion from high inflation, may even benefit poor people
more than others. This outcome is consistent with sur-
veyevidenceshowing that poor people are more likelyto
single out high inflation as a pressing concern.
Where reforms have adverse distributional effects,
these are generally small compared with the growth ben-
efits that reforms deliver, especially over periods of sev-
eral yearsor more.17 So the macroeconomicevidencedoes
not suggestthat the benefits of reform have bypassedpoor
people-nor even that the benefits only gradually"trickle
down" to them. Instead, it suggestsa pattern in which all
income groups on average benefit equally from reforms.
Even among the countries of the former socialist bloc,
where reforms have often gone awry, inequality increased
least in countries that successfullyimplemented reforms.
It increasedmost in countries that introduced reformsonly
partially or not at all.' 8
This kind of cross-country evidence provides only a
partial picture of the effects of reforms on poor people.
The same reforms may have very different effects in dif-
ferent countries, and so such average results provide
only a rough guide to the likely future impact of reforms
in a particular country. Furthermore, even when re-
forms on average have no effect on aggregateincome in-
equality, there will still be winners and losers from
reform. And when the main effects of reforms are on the
provision of public goods such as health, education, or
infrastructure, it may take time before the effects on in-
come distribution and human development outcomes are
felt. Detailed case studies of reforms in specificcountries
shed light on some of the complexities of reform. While
it is as difficult to generalize from an individual coun-
try'sexperienceas it is to generalizefrom an averagecross-
country relationship, both types of evidence provide
useful insights into the effects of reforms.
Not surprising, case studies of reform episodes show
that market-friendly reforms have uneven costs and
benefits-especially in the near term-with the costs
concentrated on particular groups and the benefits spread
broadly over the economy as a whole. Costs and bene-
fits can also be distributed unevenly over time. For ex-
ample, trade liberalization can lead quickly to reductions
in employment in previouslyprotected sectors,but it may
take time for affectedworkersto developthe skillsrequired
to take advantage of growing opportunities in other sec-
tors. In Hungary the averageduration of unemployment
for those laid off from state enterprises between 1990 and
1992 was more than four years.'9
Our leaders announced a transition to new market
relationsand then left us to the merry offate.
--Fr ln a discluiOng0roup, (;coigya
On the whole these costs do not negate the benefits of
the reforms discussedabove. But they do point to the im-
portance of social policiesto easethe burdens that reforms
impose (seechapter 8). This is particularlyso for poor peo-
ple,whoseassets,particularlythe human capitalof their chil-
dren, can be irreversiblyaffectedby even short-term costs.
The costsalso remind us that successor failure is not mea-
sured only by changes in averageincomes.Surveyevidence
from Latin America indicates that reforms can be unpop-
ular if they are associatedwith the perception-and often
the reality-of greaterrisk and uncertainty.20
Who wins?And who loses?The winners are often those
in rural areas, those in countries where the enabling envi-
ronment for the private sector is strong and private sector
capacityto seizenew opportunities is good, those with the
skillsto be absorbedinto new activities,and those who are
geographicallymobile and willing to look for work in new
occupationsand sectors.The losershaveoften beenin urban
areas(whereserviceshavebeenhit), in governmentjobs,and
in jobs where protected insiders once earned more than
marketwageswould support. The losersmight alsoindude
the unskilled,the immobile,and thosewithout accessto the
new market opportunities-because they lack human cap-
ital,accessto land or credit,or infrastructureconnectingfar-
flungareas.The losersmay alsoindude otherwiseviablefirms
hit by economic crisesnot of their own making.
As the state sector contracts, employmentopportunities
are evaporating.
-From a a'iscussion
group,Ukraine
 MAKING MARKETS WORK BETTER FOR POOR PEOPLE 67

Table4.1
Impactof reformson agricultural
prices,output,and productivityinsevencountries
Percentagechange,five-year postreform period compared with five-year prereform period

Agricultural
productivity
RealGDP growth
Real growth rate (percentage
agricultural Real (percentage Agricultural point
Country prices exchangeratea point change) output change)
Chile 120 105 2.8 40 8.2
Ghana 5 230 3.9 50 12.2
Hungary -10 -23 .. -15 25.4
Indonesia 20 75 -0.6 42 2.3
Madagascar 11 94 2.0 15 2.9
Mexico -24 22 -3.7 14 1.3
New Zealand -31 -2 0.4 5 0.8
..Notavailable.
a. Anincrease
indicates
depreciation.
Source:Meerman1997.

Since poor people are representedamong both the win-
ners and the losers described here, there can be no gen-
eral lesson that reforms are good (or bad) for all poor
people all the time. But examples of reforms in three
areas-agriculture, fiscal policy, and trade-yield im-
portant insights into what determines success and fail-
ure, how reforms affect poor people, and whether it is
possible to mitigate the adverse effects on losers.
Agriculture
Under inward-oriented models of development the
structure of tariffs and nontariff barriers and often the
exchange rate were biased against agriculture. Market-
oriented reforms that reduced this antiagriculture bias-
and dismantled various forms of state intervention
(price supports, input and credit subsidies, support for
marketing products)-have generally increased agri-
cultural growth. Policy reforms such as privatization, re-
duced regulation, and trade and price liberalization
have had a positive impact for many countries.21 Agri-
cultural output and productivity growth have generally
risen in the postreform period, sometimes substantially
(table 4.1). Because many poor people are small agri-
cultural producers, they have benefited directly from
these reforms. Case studies of Chile, China, Ghana,2 2
Uganda, and Vietnam show that reforms have helped
raise producer prices for small farmers by eliminating
marketing boards, changing real exchange rates through
broader economic reforms, lowering tariffs, and elim-
inating quotas (box 4.3).
As chapter 5 discusses,accessto land playsan important
part in poverty reduction. Better access to land, accom-
panied by access to such assets as credit and infrastruc-
ture, can improve the productivity of land and labor f'or
poor people. Thus liberalizing land markets has large
potential benefits. Evidence from Mexico, for example,
indicates that land market reforms expanded small farm-
ers' access to land through the rental market (box 4.4).
Beyond these direct benefits,growth in agricultural in-
comes appears to have been particularly effective at re-
ducing rural poverty because of demand spillovers to
local markets in which the nonfarm rural poor havea large
stake. Rural construction, personal services,simple man-
ufacturing, and repair have been major channels through
which poor people haveshared in agriculturalbooms, even
when they havenot been direct beneficiariesof higher crop
prices. In Ghana the big beneficiaries of reform-cocoa
producers-constitute lessthan 8 percent of the poor, yet
rural poverty fell sharply.
All our problems derive from lack of land. If we have
enough land we will be able to produce enough tofeed our
househokls, build houses, and train our children.
-Poor man,Nigeria
Another exampleof the indirect benefits of market re-
forms comes from cotton smallholders in Zimbabwe.23
Before the reforms the Cotton Marketing Board used its
power as the sole buyer to impose low producer priceson
farmers to subsidizethe textileindustry. Large farmers di-
68 WORLD DEVELOPMENT REPORT 2000/2001
Box 4.3
Agriculturalreforms in Chile and China help small farmers
Chiledramatically illustrateshow incompletereformscan harm
agriculture-andhowcompleted reformscanhavelargebenefits.
Themilitarygovernment thattookpowerin 1973implemented a hectareover those30 years.State-imposed
sustainedprogramof policyreform.Agriculturalproductionin-
creasedby a quarterin 1974,but then stagnated through1983,
thanksto the uncertaintyover future policiesandthe incom-
pletenessof reforms.In 1978-82eliminationof creditandinput
subsidies andappreciation inthe realexchange ratehit agriculture
hard,whiledelaysin implementing reformsin land,labor,andwater
rightsmarketspreventedaneffectiveresponse(Valdes1994).
In 1984anaggressive devaluation andcompletionof reforms
ledto a vigorousrecovery.Thesectorresponded strongly.Agri-
culturallaborforceparticipation quicklyrose-from a low of 14
percentof the total laborforceto more than 19 percent,sub-
stantiallyhigherthanat anytime in the previousdecade.Agri-
culturalgrowthincreasedfrom0.2 percenta yearin 1960-74to
4.9 percentin 1974-90.Greaterlandproductivitywas a major
factor.
China'sagricultural liberalizationledto a swift response. Be-
forethe reformsin 1979,Chinahadgoodroadsandirrigationin-
frastructure,excellenttechnicalpackages for grainsand other
crops,andeffectiveapplication of fertilizerandother inputs.Be-
tweenthe 1940srevolutionandthe 1970s,irrigation capacityhad
Source:Meerman1997.
Box4.4
Land markets and poor peasantsin Mexico
Liberalization
of landrentalandreorganization
of thepropertyrights thelandrentalmarketcouldbeoffsetbyreducedaccessto credit
systeminthe ejidos(communal lands)in1992formedthebackbone for the land-poor,
of structuralreforms to transformthe Mexicaneconomy. Liberal-
izing land marketsand better definingand enforcing landproperty
rights were expectedto drasticallyreducethe costs of transactions
in both landand credit markets, improvingaccessto landandcredit
for poor, small-scale(andperhapsmore efficient) producers.
Policies aimed at activatingland rental markets would bene-
fit the landlessand the land-poorby increasingtheir accessto land
throughrental and sharecroppingtransactions. Butwith both land not worsened,land-poorfarmersmighthave benefitedeven more
and credit markets being liberalized,
the easing of restrictionson from landmarketliberalization.
Source:Olinto,Davis,andDeininger
1999.
versifiedinto unregulated crops, such as horticulture and
tobacco. After the reforms cotton prices rose. In absolute
termsthe gainswouldbe greaterforlargerfarmers,simply
becausethey produce more cotton. But therehavebeen par-
ticulargainsfor smaliholders,as newlyprivatizedcotton buy-
ers have chosen to compete with one another in part by
providingnew extensionand input servicesto smaliholders.
morethandoubledandfertilizerproduction hadincreased signif-
icantly.Butarablelandpercapitadeclinedfromabout0.2 to 0.1
croppingpatterns
forcedmost cultivationinto rice and other cereals.Collective
farmshadto fulfill grainquotasfor deliveryto the cities,andthe
nationalgrainmarket wasfragmented into30self-sufficientregions.
Startingin 1979,familyfarmingthroughthe "householdre-
sponsibilitysystem"sweptthecountry,replacing communal farm-
ing.Althoughinitiallyfarmerswerestillobligedto delivergrainat
low prices,theywereotherwisepermittedto producewhatthey
wishedat mostlymarketprices.Commerce in ruralareasandbe-
tweenfarmandcity,previously repressed bythestate,wasallowed
to flourish.Ratherthanhaveself-sufficient provinces, the stateen-
couragedregionalandnationalmarkets.Effectivedemandin-
creasedrapidlyfor high-value products(vegetables, fruits,meat,
fish,eggs)thathadbeenrepressed bytheearlierstate-directed poli-
cies.China'speasantfarmers-skilled,hardworking,andstrongly
motivated-responded to the new opportunities with greatvigor
andlaunched fiveyearsof thefastestsustained agricultural growth
everrecorded anywhere. Between1978and1984netagricultural
outputincreased 7.7percentannually andgrainoutput4.8percent
(Lin1995).Withthe vastmajorityof China'spoorpeoplein rural
areas,the incidence of povertyfell dramatically.
whowerelessableto uselandas collateral.
This
could have shifted the benefits from smallerto largerholders.
Theincreasedsupplyof landfrom largefarmersin the rentalmar-
ket allowed the rural poor a small but statistically significant in-
creasein accessto land.After controllingfor the greateraccessto
credit, it appears that large farmers increased their demand for
landrentals. But small farmers appearto have increasedtheir de-
mandfor landeven more,suggestingthat hadtheir accessto credit
' u io k thbe ab, is A 'nerous, bhutwhat is the incentive
! X, mOo/lfe
wo rhb the fiiutni1 7 n1eedsif there are nio
,i,A- ' ;'Ouad to gft protdi'c to the mnarket?
.. ,; u ,. 'U) '"'l
Market-friendly reforms have also sometimes hurt
the rural poor. In some countries financial reforms tight-
 MAKING
Box 4.5
Listeningto farmers in Zambia
Since1991Zambiahas radicallychangedthe policyand institutional
environmentfor agriculture.With liberalizationand privatization,
private suppliers have replaced state agriculturalservices for
credit,inputs, and marketing.
Using participatoryrural appraisals and beneficiaryassess-
ments, the WorldBank-assisted AgriculturalSector Investment
Program has established systematic and regular feedbackbe-
tween policymakers,service providers, and those affected by
programs. Talkingto farmers has helped policymakersunder-
stand the farmers'resourceconstraints,servicedeliveryproblems,
and strategies for dealingwith the vicissitudesof transition.
Participatoryassessments alsoexaminelocalperceptionsof the
effectivenessof agriculturalinfrastructureandservices.These con-
sultationsrevealedthat agriculturalcreditand marketing,now han-
dledbythe privatesector,were unevenandunpredictable-because
of poor infrastructure,lackof capacity,and inadequate enforce-
mentmechanisms.Publicextensionandanimalhealthservices,suf-
Source:WorldBank
1998a.
ened credit and closed rural bank branches, reducing
the availability of credit.2 4 And in some cases research,
data collection, reporting, and quality monitoring dis-
appeared after the abolition of state enterprises and mar-
keting boards. In Cameroon the marketing boardhad been
maintaining rural roads, but the responsibility was not
reassigned after the reforms. In Zambia remote farmers
had been implicitly subsidized by a uniform pricing pol-
icy that did not take into account transport costs, while
small farmerswithout storagefaciities were implicitlysub-
sidized by prices held constant across seasons. After the
reforms market forces eliminated the implicit subsidies,
and transport infrastructure deteriorated significantly,
leaving many farmers worse off.
The converse of the gain to small producers from rel-
ative price shifts is the cost to poor urban dwellers.Take
Ghana. The rural sector gained from higher export
prices and greater rural demand as cocoa farmers spent
their windfall, but urban residents grew poorer. Living
standards in Accra deteriorated in 1988-92, even while
conditions were improving elsewhere in the country.
Poor and middle-income urban residents suffered from
higher food prices. Moreover, the dismantling of the
old export marketing system removed an important
source of public revenues that was not quickly re-
placed. This led to higher inflation and public sector
retrenchment, with the costs of both felt most by urban
residents.
MARKETS WORK BETTER FOR POOR PEOPLE 69
feringfrom staffshortages and lackof operatingfunds,transport,
andequipment,were also failingto respondwell to farmers'needs.
Farmers want better infrastructure (especially roads and
bridges)and more effectiveregulationof the privatesector.They
also want more informationon markets for agriculturalproducts
andeasier access to more flexibleand responsivecredit facilities.
Andthey want adviceon subsistence cropsand storage methods,
whichthey prefer to get throughgroup extension.
Talkingto farmers also identifiedways to help those who are
economically vulnerable take part in agricultural markets-
extendingmicrocredit,promotinglocalseed productionsystems,
and offering research and extension services for subsistence
crops and low-inputagriculture.To create the localorganizational
basis for participatoryextension and economicallyviablejoint
activities-such as marketing,localfinancialservices, and cattle
dipping-support needs to go to producerassociations,service-
providingNGOs,and other organizationsactive in communities.
These examples suggest at least two lessons. The first
is simple: reforms can benefit poor people but also hurt
them. Listeningto stakeholdersthrough participatory pol-
icymaking can do much to identify and avoid unin-
tended consequences for poor people (box 4.5). Second,
when reforms leave an institutional vacuum, perfor-
mance suffers.As with other reforms, agricultural mar-
ket liberalization without the proper institutional
framework will not deliver the expected results-and
could have serious consequences for poor people.
Fiscal policy
In many countries fiscal reforms to strengthen revenue
collection capacity and control unsustainable spending
have been a central element of broader reform programs.
Sinceraising revenuestakes time, fiscalreformsoften show
up first in spending cuts. When those cuts are felt in so-
cial sectors and in subsidies, they can hurt poor people.
As chapter 5 discusses,there is evidence that the intro-
duction of user fees in health services hurts the poor
more than the rich. In Madagascar the real incomes of
poor households in the capital city declined substan-
tially when food prices were decontrolled.2 5 But elimi-
nating subsidies does not always hurt the poor. A study
in Guinea and Mozambique found that eliminating food
subsidies did not hurt poor people because the subsidies
had not reached them in the first place.2 6 The lesson is
clear:loweroverallsubsidiesneed not be inconsistentwith
70 WORLD DEVELOPMENT REPORT 2000/2001
helpingpoor people,if the subsidiesare better targeted
or replacedby other forms of assistance.
In the 1990sgovernmentsin EasternEuropeand the
formerSovietUnion introduceda rapidphaseoutof util-
ity subsidiesacrossthe board.The urgencywas dictated
by the need to reduceunsustainablefiscaldeficits.This
had a hugeimpacton the welfareof allfamilies,especially
poor families.In Ukraine household energytariffsin-
creasedfour- to twelvefold(in real terms)between1992
and 1995,while averagehouseholdincomedropped to
lessthan halfits prereformlevel.Tohelp cushionthe im-
pact, a cap of 20 percent of familyincome wasput on
what householdspay in utility billsand rent. The state
budgetis supposedto payany billsin excessof thislimit
(althougharrearsin paymentscontinueto be aproblem).
In Moldovathe averagewinter heating billwould have
exceeded60 percentof the income(cashand in kind) of
a typicalfamilyof four in the lowestfifthof the income
distributionlivingin a smallapartment.Awarethat this
wasunsustainable,the governmenteventuallyintroduced
mechanismsto subsidizefamilies,rangingfrom tolerat-
ing nonpaymentto establishingdifferenttariffratesfor
poor families.27
Experiencein the countriesof the formerSovietUnion
alsoshowsthat fiscaladjustmentcould havebeen done
in different,and much more pro-poor,ways.For exam-
ple, beforethe politicaltransitionthe ratioof healthand
educationpersonneland facilitiesto the total population
wasaboveOECD standards.Duringthe 1990spublicrev-
enuesand expendituresfell as a shareof GDP.And since
GDP collapsedas well, government spending in real
terms felldramatically.Ratherthan downsizepersonnel,
rationalizefacilities,
and institutesomecost-recovery mea-
sures,governmentsallowedreal public sectorwages to
erode,and spendingon maintenanceand materialinputs
collapsed.Publicsectorwageswere often in arrears,and
publicemployees respondedto theirpersonalfinancialpres-
suresby demandingunder-the-tablepaymentsfor pub-
lic services,somethingpoor peoplecouldill afford. 28
Revenue-raising measures,such as a growingreliance
on valueaddedtaxes,can alsohurt poorpeopleifnot im-
plementedcarefully.Strongefficiencyargumentsforvalue
addedtaxationarebeingheededthroughoutthe developing
world. But introducingsuch taxes can haveeither pro-
gressiveor regressive effects.If valueadded taxesreplace
progressiveincometaxesor if poor peopleeitheravoided
or did not qualifyfor other taxes,such reformsare re-
gressive.Pakistan'sintroduction of a value added tax
shiftedthe burdenof taxationtowardthe poor:the taxbur-
denon the richestincomegroupdeclined4.3 percentage
points,whilethat on the poorestgrouprose10.3percentage
2 9 In contrast, when tax reformsreduce the re-
points.
liance on inflationaryfinance,they can be progressive
becauseofthe heavyburdenhigh inflationplaceson poor
people.Moreover,most of the redistributivepower of
publicfinancelieson the expendituresideratherthan the
revenueside.Therefore,evena slightlyregressive tax re-
formcan haveprogressive resultsifthe additionalrevenue
is devotedto expenditurestargetedtowardpoor people.
Trade
Tradereforms-reducing tariffsand nontariffbarriers-
have had profound effectsin many developingcoun-
tries. As chapter 3 discusses,there is now substantial
evidencethat opentraderegimessupportgrowthand de-
velopmentand that movingtowardan open regimeand
its attendant benefitsis the reasonfor tradereform.But
the consequences forpoorpeopledependcruciallyon how
tradeliberalizationaffectsthe demand for their greatest
asset:their (oftenunskilled)labor.Furthermore,tradere-
forms in the developingworld have not alwaysbeen
matchedby complementaryreformsby rich countries,
wherethe remainingprotectionimposesa heavyburden
on the developingworld (chapter10).
The initialpush for trade liberalizationas an instru-
ment for povertyreductionwas influencedby a narrow
readingof predictionsfrom tradetheory:removingtrade
barriersin developingcountrieswould increasedemand
fortheirabundantlow-skilledlaborand expandunskilled
employmentand earnings.Not onlywouldtradeliberal-
izationraiseaverageincomes-it wasalsoexpectedto be
particularlypro-poor through this effect on unskilled
labor.The evidenceshowsthat the actualresultsin the past
15 yearshavebeenmixed.Tradereformshavedelivered
growth,and thus povertyreduction-but their distribu-
tional effectshavebeen more complex.Carefulanalysis
suggeststhree mainfactorsat work.
First,in somecountriestraderestrictionshad benefited
poor peoplebyartificiallyraisingthe pricesof the goods
theyproduced.In suchcasesit is not surprisingthat trade
liberalizationwould hurt poor people. For example,a
studyof Mexicofound that the wagesof unskilledlabor
relativeto those of skilledlabor declinedover 1986-90,
and that about a quarter of the declinewas from the re-
duction in tariffsand the eliminationof import license
requirements(figure4.3).3°The authorsexplainthis ap-
 MAKING
parent anomaly by noting that Mexico, despite its com-
parative advantage in low-skilledindustry, had protected
labor-intensive sectors-such as textiles and clothing-
before adopting trade reforms. Supporting the incomes
of the unskilled through trade barriers is very inefficient.
Support can often be given in other ways at lower social
cost, although designing and implementing such better-
targetedprogramstake time. But it is not surprisingin these
circumstances that trade liberalization-unaccompanied
by compensatory programs-would hurt poor people. In
some other countries, however,the pattern was different:
urban manufacturing workers protected by trade barriers
were more skilled and lesslikely to be poor.
Second, some countries that liberalizedtrade were not
particularlyabundant in unskilledlabor.In Africaand Latin
America land is relativelyabundant, and in Eastern Eu-
rope skilled labor is plentiful. Although this does not de-
tract from the efficiencyand growth arguments for trade
reform, it does call into question the earlier presumption
that trade reform might also deliver equalizing effects by
raising the demand for unskilled labor. But in countries
where unskilled labor is abundant, such as Bangladesh,
China, and Vietnam, the gains from integrating into the
world economy can be significant for unskilled labor,
Third, trade reformswere often accompanied by other
developments that were disequalizing rather than equal-
izing. In many developing countries that opened to trade,
as in many industrial countries, the wagesof skilled work-
ers have grown faster than those of unskilled workers. In
the United Statesthe wagesof unskilledworkershave fallen
in real terms by 20 percent since the 1970s, despite rapid
growth in the overalleconomy.31 Studies for countries as
diverseas Chile, Colombia, Mexico,Turkey,and Venezuela
show a similar phenomenon-premiums paid to skilled
workers have increased in all these countries.32
Is trade the culprit behind this widening inequality?
The balance of evidence suggeststhat it is not. More im-
portant has been technological change favoring workers
with better education and skills, sometimes in the form
of imported foreign technologies.This can be seen from
severalpiecesof evidence. Even though the relativewages
of skilled workers have risen in many countries, there has
also been a shift toward greater employment of skilled
workers-contrary to what simple trade theory would pre-
dict. This shifthas been pervasiveacrossindustries-again
contrary to simple trade models, which would have pre-
dicted increases in some sectors and declines in others.
And there is evidencethat the pattern of shiftstoward more
MARKETS WORK BETTER FOR POOR PEOPLE 71
Figure 4.3
The gap between skilledand unskilledwages
widened in Mexico
130
120 Skilled
110
100
90
80 Unskilled
70
1984 1990 1995
Source: Lustig 1998.
skill-intensive employment in the industrial world in
the 1970s and 1980s is being matched by a similar, later
shift in the developing world.3 3
This is of course not to say that technological change
should be avoided because it hurts poor people. On the
contrary, technological change is a fundamental deter-
minant of growth and rising living standards, powerful
forces for poverty reduction. Instead, the importance of
a risingrelativedemand for skills points to the need to in-
vest in the skills of poor people, to enable them to take
advantage of the new opportunities that technological
change brings.
Private sector response
These examplesof agricultural, fiscal,and trade reforms
show that reforms can have complex distributional out-
comes.But remember that the objectiveof market-friendly
reforms-a vibrant and dynamic private sector-can be
one of the most effectiveantidotes to the costs of reforrn.
New job creation, technological change that raises labor
productivityand wages,and institutions that ensure equal
opportunities for gaining accessto the new jobs do much
to ensure that the benefits from reform are widely shared.
Fortunately,a strong private responseappears to be the
general experience in developing countries after reform,
especiallywhen labor market regulations are not onerous
and do not inhibit adjustment.34 A retrospectivestudy of
72 WORLD DEVELOPMENT REPORT 2000/2001
tradeliberalizations
foundthat in 12 of 13caseswheredata
were availableformal manufacturingemploymentin-
creasedwithin one year after liberalizationwas com-
pleted.3 5 The exception was Chile, where increased
employmentin agricultureoffsetthe declinein manu-
facturingemployment.In Estoniaaflexiblelabormarket
createdmanynewjobs,leadingto minimalunemployment
despitethe intensejob destructionand laborturnoveras-
sociatedwith reform. 3 6 In Panamaunemploymentfell
steeplyafterliberalization.
In SouthAsiagrowthin the for-
mal manufacturingsectoracceleratedfrom 3.8 percenta
year to 9.4 percentafter liberalization,as many workers
pulled out of informal sector employment. 3 7 And in
Africathe micro and smallenterprisesectoris the most
dynamicin fiveeconomiesconsideredin a recentstudy.
Annualemploymentgrowthwasstrongin theseenterprises
afterreform,andnewenterprises startedup at a highrate.38
Making markets do more
for poor people
Evenwheremarket-friendly reformshavetakenhold,there
is much that countriescando to improvethe benefitsthat
marketsoffer to the poor. To reachpoor people,many
reformsneedto be accompaniedby institutionalsupport,
investment in infrastructure,and complementary re-
forms at the micro level.The incentivesfor policymak-
ers to undertake such reforms are small because the
marketsinvolvingpoor people are typicallysmall. So
the reformsget little attention, eventhough theycan be
powerfulforcesforpovertyreduction.But increasingac-
cessto productiveassetsand lighteningand improving
regulationcando much to involvepoorpeoplemore di-
rectlyin markets.3 9 New technologiescan help as well,
especiallyinformation technologies that break down
someof thebarriersofphysicalremotenessthat manypoor
peopleface(box 4.6).
The potentialof reformsto improveaccessto markets
forpoorpeoplecan be seenfromexamplesin threeareas:
liftingthe heavyhand of regulation,especially
on the small
businessesthat often providethe poorwith employment;
promotingcorelaborstandards;and improvingaccessto
financial markets for the poor, especially through
microfinance.
Lightening the regulatory burden
Compliancewith regulationsimposesfixed costs that
are particularlyonerous for small firms. Carefullyre-
viewingregulationsand exploringpossibilitiesfor more
flexiblerequirementscan easethe burden. In Chile the
governmentrecentlysimplifiedthe duty-drawbacksys-
tem to reducethe administrativecostsforsmallfirms.In
Boliviaparts of the taxsystemweredrasticallysimplified
for smallfirms.4 0 In the Philippinestherearemuch lower
minimumcapitalrequirementsforsmallthrift and rural
banks than for commercialbanks.
In contrast,in Indonesiaofficialand unofficiallevies
areestimatedto raisethe costsfacedbysmalland medium-
sizeenterprisesby as much as 30 percent.4 1 In somesec-
tors smallenterpriseshave to secureas many as eight
licenses-some of which have identicalfunctions but
are issuedby differentagencies.Obtaininglicensestakes
solongand proceduresaresocomplicatedthat somebusi-
nessownerschoose to operate illegally. 42 In the Indian
state of Gujarat licensingrequirementsfor gum collec-
tors are a barrierthat hinderwomen'scollectorgroups. 43
Reformsto reduceleviesand to simplifyand shorten li-
censingand entry proceduresfor smalland medium-size
enterprisescould easethis burden.
Given the opportunity,small and medium-sizeen-
terprisesmightservesomesegmentsof the marketsnor-
mallythought of as natural monopolies.In many urban
areas in Africa and Latin Americasmall independent
water providersbring basicwater servicesat lowcost to
poor marginalcommunities.Smallenterpriseshavealso
been effectivein solid waste management. 4 4 But they
often face barriers-such as requirementsfor experi-
ence,complexor expensiveproceduresfor registration
and tendering,and noncompetitivebehaviorin markets.
Removing these constraints could allow small and
medium-sizeenterprises to expand their activitiesin
thisarea,increasingemploymentopportunitiesforlow-
incomegroupswhileexpandingaccessto servicesforpoor
communities.
Betterregulationdoesnot alwaysmeanlessregulation.
Take the privatizarion of gas, water, electricity,and
telecommunicationsutilitiesin Argentinain the early
1990s.Privatization improvedperformance, and poorpeo-
ple, as directconsumers,benefitedalongwith the restof
the economy-and morethan proportionately forgasand
electricity,majorcomponentsof their consumptionbas-
ket. But becauseprivatizedutilitiesare often monopolies,
appropriateregulatoryinstitutionswereessentialto fair
pricing.The new regulationsto ensurethat utilityprices
yieldedonly normal ratesof return had important indi-
rect benefitsforpoor people,by encouraginginvestment
 MAKING
Box 4.6
Attacking poverty with information
Virtual Souk expands market accessfor artisansin the
Middle Eastand North Africa
FadmaAoubaida,a Moroccanweaverfrom Taliouineand a mother
of seven-with the money she eaMed from selling her products
on the VirtualSouk-repaired her roof and started building an in-
door latrine,one of the few in her village.lj]aAittalblhsen, another
woman artisan in Morocco, spent her profits to buy cement and
windows for her house. With future profits, she wants to buy a
truck to transport rugs from her village to the market or buy bi-
cycles that women can ride.
-BBC OnlineNews, 14 October 1999
Artisans in the Middle Eastand North Africa have always crafted
high-qualityproducts using traditional techniques and ancestral
know-how. But shrinking local markets and difficulties in gainang
accessto more lucrativenationalandinternationalmarketsare lead-
ing to a gradual disappearanceof culturally rich crafts-and with
them an important source of income for poor people.
The VirtualSouk is buckingthis trend. Since 1997this Internet-
basedmarketplacehasbeenprovidingdirect accessto international
markets for several hundred artisansfrom Egypt, Lebanon,Mo-
rocco,andTunisia,manyof them women.Thenetworkisexpanding
to other countries in the region, and there is demandto adaptthe
conceptto EastAsia and Latin America.
Onlinesales soaredtenfold between the first and last quarters
of 1999,reachingmarkets aroundthe world, includingcountries in
Europeand NorthAmericaandas far as Australia,Japan,andSouth
Africa.Participatingartisansreceive65-80 percentof the proceeds,
a much largermarginthan throughtraditionalchannels.And the gains
are more than simply financial.Throughthe Virtual Souk, artisans
Source:FortheVirtualSouk,seewww.peopbnk/vsouk/;
for the Grameen
Telecomcellularphoneprogram,
and job creation throughout the economy. One study
found that these indirect gainsfor poor people-reflecting
the power of appropriate regulation-were five times as
large as the direct gains from lower utility prices and bet-
ter service.45
An appropriate and generallylighter regulatory frame-
work in labor markets could also benefit poor people. In
general, excessivelyburdensome labor market regula-
tions can limit job creation and thus opportunities for poor
people to productivelyemploy one of their most important
assets-their labor. These constraints are especially im-
portant when reformsin other areascreate temporary em-
ployment dislocations. But the benefits of deregulating
labor markets should not be overstated. Often labor mar-
ket regulations are not well enforced, especiallyin the in-
formal sector, so relaxingthem would have little effecton
employment opportunities for poor people.
MARKETS WORK BETTER FOR POOR PEOPLE 73
gain accessto opportunitiesfor empowerment,capacitybuilding,
incomegeneration-and for the use of their skills with dignity.
Cellular phonetechnology gives bargainingpower to
w
I always sell eggs to middlemen. In the past, whatever prices
they offered, I accepted because I had no idea about the
going prices of eggs.... Last week, the middleman came ...
and desired to pay me 12taka per hali [four units]... Keep-
ing him waiting, I rushed to check the prices through the Vil-
lage Phone. Theprice was 14 taka per ha/i of eggs in nearby
markets.I camebackand refusedto sell to him at the lower
prices. ... After a brief haggling, we agreed to buy and sell
ath13ntaka plteaeloa
_Halirna Khatuon, a poor, iliiterate woman
who sels eggs, Bangladesh
A subsidiaryof GrameenBank, GrameenTelecom operates a vil-
lage payphoneprogramthat leasescellulartelephonesto selected
bank members, mostly women in rural areas,who use the tele-
phone to provide services and earn money.Today around 2,000
village payphones are in place.Thetarget is to install40,000 tele-
phones by 2002, introducingtelefax and email servicesas well.
Thesephones havehelped lower the cost of informationgath-
ering. This can be seen in lower prices for poultry feed, more sta-
ble dieselprices,and lessspoilageof perishablegoodsdueto more
preciseshipmentdates.Women providingthe phoneserviceshave
gaii'redconfidence and new status as "phone ladies." Telephone
users include both rich and poor, but poor peoplemake more calls
for economic reasons.
seeBurr(2000).
Promoting core labor standards
Core labor standards have been set out in the Declaration
on the Fundamental Principles and Rights at Work
adopted by the members of the International Labour Or-
ganization in 1998. They include freedom of association
and the right to collective bargaining, elimination of
forced labor,effectiveabolition of child labor,and the elim-
46
inationofdiscrimination in employmentand occupation.
The goals underlying these core labor standardsare im-
portant, and it is widely agreed that the standards them-
selvesrepresent worthy targets for economicdevelopment.
This consensus is especially strong for the most ex-
ploitativeforms of child labor and forced labor. However,
there is no consensus regarding the best way to achieve
the labor conditions envisaged by these core standards.
How best to implement the objectivesset out in these stan-
dards is difficult to determine and depends a great deal
74 WORLD DEVELOPMENT REPORT 2000/200I
on the circumstances of individual countries. Some in-
dustrial countries take the position that the standards
should be enforced through trade agreements or devel-
opment cooperation.Many developing countries argue-
and rightly so-that applying trade sanctions in this way
can serve protectionist purposes for industrial countries
and that conditioning development cooperation will un-
fairly hamper development.
It is clear that simply adopting core labor standards
will not guarantee their realization. In developing coun-
tries problems meeting these standards may be a conse-
quence of poverty,
Consider child labor.Too often children'stime spent at
work comes at the expenseof their formalschooling-with
likelyadverse long-term consequences.But a child's earn-
ings may make the difference between survival and star-
vationfor the family,or they may help provide the resources
for a siblingto stay in school.4 7 In these circumstancessim-
ple bans on child labor can have adverse consequencesfor
poor families' incomes and can evenhave the unintended
effectof pushing children from work in the formal sector
to more exploitativework in firms outside the reach of for-
mal regulations.As a complement to standardsagainst the
most exploitativeforms of child labor, programs that pro-
videfinancialincentivesthat make it affordableto keep chil-
48
dren in school can be a very effectivestrategy.
Implementation of the standards on freedom of asso-
ciation and collectivebargainingalso raisescomplex issues
for economicdevelopment.Enshriningsuch rightscan help
eliminate abusive workplace practices and ensure fair
compensation, especiallyfor poor people, whose desper-
ate need for employment places them most at risk of un-
fair and exploitative employers. Unions also are an
important dimension of civilsociety,and consultationwith
unions can provide a valuable input into policy formula-
tion. However, empirical evidence on the economic ben-
efits of unionization and collectivebargainingis generally
quite mixed and suggeststhat both costs and benefits are
complex and context specific.49 Particularlyimportant are
the rules that govern collectivebargaining and resolution
of labor disputes.Some forms of collectivebargainingrules
may be better at producing efficient and equitable out-
comes than others.5 0 In any case,the exerciseof these rights
will best serve development objectives when unions and
employers are knowledgeable and independent and bar-
gain in good faith.
The core labor standards, then, set an important tar-
get, but a simple strategy of enforcing them through sanc-
tions is unlikely to produce the desired outcomes for
51
workers. Rather,promoting them aspart of a broad-based
development strategy through information, technical as-
sistance, capacity building, and complementary initia-
tives is likely to yield the greatest benefits. Using
incentives-such as programsto keepchildren in school-
to address the causesof suboptimal labor practices must
be a key part of this strategy.Along these lines, and also
deserving close attention, are interesting new ideas about
complementing public standardswith private (market-dri-
ven) standards that encourage employers to adopt desir-
able labor practices.52
Improving access tofinancial markets
for poor people
Accessto financial markets is important for poor people.
Likeall economic agents,low-incomehouseholdsand mi-
croenterprises can benefit from credit, savings, and in-
surance services.Such serviceshelp to manage risk and
to smooth consumption in the face of sharp fluctuations
in agricultural yields and prices, economic shocks, and
even natural disasters.Savingsand credit facilitiescan help
to make larger investments more affordable, and so allow
people to take advantage of profitable business oppor-
tunities and increase their earnings potential. For
economies as a whole, a large literature has documented
the importance of well-functioning financial markets for
growth.53
But financial markets, because of their special fea-
tures, often serve poor people badly.Asymmetric infor-
mation between lenders and borrowers creates problems
of adverse selection and moral hazard. The traditional so-
lution to these problems is for lendersto demand collateral
from borrowers. Since poor people have insufficient tra-
ditional forms of collateral(such as physicalassets)to offer,
they are often excluded from traditional financial mar-
kets. In addition, transactions costs are often high rela-
tiveto the small loans typicallydemanded by poor people.
And in areas where population density is low, physical ac-
cess to banking services can be very difficult: in the
mountains of Nepal people must walk six hours to and
from the nearest bank branch at an opportunity cost of
a day's wages.54 Facing such hurdles, poor people are
often discouragedand simply do not seek loans since they
believe that they will be denied credit or will not be able
to fulfill bank requirements. At the same time, conven-
tional banks often find it unprofitable to provide services
to poor people using traditional lending practices.
 MAKING
These failureshave been used to justify a high level of
government intervention in the form of targeted credit,
with government-owned financial institutions channel-
ing sizable resources at subsidized interest rates. Often,
this approach assumed that poor people required only
cheap credit, ignoring their demand for savings instru-
ments.55 Outcomes were disappointing. The lending in-
stitutions were not financiallyviable,and in countries from
Indonesia to Peru government-sponsored rural credit
programs collapsedunder the weight of their losses. Sub-
sidized interest rates distorted the financial markets. Tar-
get groups were not reached.5 6
So many lending institutions have emerged, but their
operations are hardly transparent. People do not know
how to access them. Those who have tried have been let
down by high levels of collateral demanded.
-From a discussiongroup. Malawi
Over the past two decades new approaches known
collectivelyas microfinancehaveemerged,applying sound
economic principles in the provision of financial services
to low-incomeclientsand using group as wellas individual
lending. Pioneers such as Grameen Bank in Bangladesh
and the village banks (unit desas) of Bank Rakyat In-
donesia captured attention worldwide by providing fi-
nancial products matching the needs of low-income
clients, using innovative collective monitoring through
group lending to strengthen repayment performance, and
charging interest rates that fully cover operational costs.5 '
In many cases these innovations led to much higher re-
payment rates than under previous schemes-and were
particularly effectivein reaching women.5 8
While such programs have become popular and rep-
resent a major step forward from previous public inter-
ventions, they are still no panacea for poverty. Not
surprising, simply providing access to credit does not
create investment opportunities: a study of rural house-
holds in Nicaragua and Romania found that removing
credit constraints would have only moderate impacts on
the number of households making investments and on
the amounts invested.59 In addition, small, locally based
microfinance organizations can be particularly vulnera-
ble to shocks such as natural disasters or fluctuations in
agricultural yields, which affect a large proportion of
their clientele at once. This can raise the riskiness of
their loan portfolios and make it more difficult for them
to provide more sophisticated financial products. Shar-
MARKETS WORK BETTER FOR POOR PEOPLE 75
ing these risksamong microfinanceorganizationsand pos-
sibly encouraging a greater role for larger and more ge-
ographically diversified and established financial
institutions can help in this respect.
Careful measurement of the economic impact of mi-
crofinance programs or institutions is fraught with
methodological difficulties,and the results of studies are
often contradictory.6 0 Nevertheless,evidence is gradually
emerging.For example,a recent reviewof 13 microfinance
institutions found that borrower households above or on
the poverty line experience a higher impact than house-
holds below the poverty line, suggesting that while ef-
fective, such institutions are not necessarilywell targeted
toward the pooresthouseholds. 6 1 Another studyfound
that the majority of microfinanceprograms reviewed
still requiredfinancialsubsidiesto be viable. 62 Increas-
ingly,the performanceof these institutionsis evaluated
by two primary criteria: their outreach to target clientele
and their dependence on subsidies.63 Although these cri-
teria do not provide a full assessmentof the economic im--
pact of microfinance institutions, they highlight the
socialcost at which microfinanceinstitutions havereached
their objectives.
These results on targeting and the prevalence of sub.-
sidy dependence point to the challenges faced by mi-
crofinanceprograms: continuing to move towardfinancial
viability while extending their outreach to their target
clientele. Best-practicedesign features of such institutions
as the village banks of Bank Rakyat Indonesia-interest
rates that fully cover costs, availability of well-rewarded
voluntary savings, performance-based compensation for
staff, intensive staff training, innovative low-cost distri-
bution networks, frequent loan collection, products
matching the demand of low-income groups, and effec-
tive management information systems-are all associated
with good financialperformance. Stronger capacitybuikl-
ing and better dissemination of these best practices can
help microfinance institutions wean themselves from
subsidies without compromising their ability to provicde
services to poor people.
Governments can improve financialintermediation for
the poor by providing complementary public goods and
improved regulation that recognize the special needs of
microfinance schemes. For example, better investment in
rural infrastructure and literacy promotion can help ex-
pand the reach of microfinance organizations, and creclit
information registries can lower informational costs atid
allow borrowers to build reputational collateral.
76 WORLD DEVELOPMENT REPORT 2000/2001
On the regulatory and supervisory fronts outdated
usury laws that prevent microfinance institutions from
establishingsufficientlyhigh spreadsbetween savingsand
lending rates to allow them to coverthe high transactions
costs on small loans should be eliminated. Improving the
legal framework for secured transactions, as is being
done in Argentina, Mexico, and Romania, can widen
credit opportunities for low-income people.
Well-functioning markets create opportunities for
poor people to escape poverty. But establishing such
markets where they are absent, making them work bet-
ter, and ensuring that poor people have free and fair ac-
cess are difficult and take time. At times, market reforms
fail entirely-or have unintended consequencesfor poor
people. The lessons of these failures point to the impor-
tance of designing and implementing reforms in a way
that is measured and tailored to the economic, social,and
political circumstances of a country. Market-friendly re-
forms create winners and losers.And when the losers in-
clude poor people, societies have an obligation to help
them manage the transition.
However,there is no presumption that making reforms
pro-poor means making reformsslowly.In some casespoor
people will benefit more from rapid market-oriented re-
forms,especiallyin areas that directlyaffecttheir economic
opportunities or that help break down entrenched mo-
nopoly privileges.In view of the urgent need to get coun-
tries onto dynamic, job-creating development paths, it
is critical that the difficulty of reform and the impossi-
bility of compensating every loser not lead to policy
paralysis.
Furthermore, to make markets work better for poor
people, macro reforms must be complemented by micro
reforms and improvements in poor people's access to
markets and information-through investment in in-
frastructure and modern technologies-as well as sources
of credit. Reducing labor market restrictions that limit
job creation and stifle competition while promoting core
labor standards remains a key challenge.

Expanding Poor People's
Assets and Tackling
Inequalities
L acking assets is both a cause and an outcome
of poverty. Poor health, deficient skils, scant access to
also lack assets because of stark inequalities in the dis-
basic services,and the humiliations of social exclusion
reflect deprivationsin personal, public, and socialassets.
Human, physical,and natural assets also lie at the core
of whether an individual, household, or group lives in
poverty-or escapes it. These assets interact with mar-
ket and social opportunities to generate income, a bet-
ter qualityof life,and a senseof psychologicalwell-being.
Assetsare alsocentralto coping with shodcks and reducing
the vulnerability that is a constant feature of poverty.
Assets and their synergies
If we get a road we would get everything else:community
center, employment, post office, water, telephone.
-Young woman in a discussiongroup,
Little'Bay,Jamaica
Poor people have few assetsin part because they live in
poor countries or in poor areas within countries. They
CHIAPTER5
tribution of wealth and the benefits of public action. In
Bolivia the under-five mortality rate of the poorest 20
percent of the population is more than four times that
of the richest 20 percent.' In West and Central Africa
the rich-poor gap in school enrollment ranges from 19
percentage points in Ghana to almost 52 percentage
points in Senegal.2 And in Ecuador 75 percent of house-
holds among the poorest fifth lack piped water, com-
pared with 12 percent amonigthe richest fifth.3 Poor
women and members of disadvantagedethnic or racial
groups may lack assetsbecauseof discrimination in the
lawor customarypractices.Lowassetsandlow income
are mutually reinforcing: low education translatesinto
low income, which translatesinto poor health and re-
duced educationalopportunities for the next generation.
There are powerfulcomplementaritiesacrossassets--
the benefits of one assetcan depend crucially on access
to another. The synergies between human capital as-
sets-such as a mother's education and her offspring's
nutrition levels-are well documented. In Vietnam
77
78 WORLD DEVELOPMENT REPORT 2000/2001

Box 5.1 Box 5.2

Interactions between human and physical Linksbetween the environment and health
capital
Therearemanycriticalinteractions betweentheenvironment
A studyof irrigation infrastructure
inVietnamuncovered im- andpoverty.Amongthe bestdocumented is thelinkbetween
portantcomplementarities betweeneducation andthegains the environment andthe healthof poorpeople.Pollution-
fromirrigation. Thestudytriedto explaindifferencesinfarm dirtywaterandair-s a majorcontributor to diarrhea andres-
profitsas a functionof irrigatedandnonirrigated landallo- piratoryinfections,the two most frequentcausesof death
cationswithcontrols fortheobservedfactorsthatdetermined for poorchildren.
the administrative landallocationsto households on decol- Research hasconsistently
shownthatimproving drinking
lectivization.Assumingthat placementof irrigationis not waterhaslesspronounced healthbenefits
thanimproving san-
basedon expectedratesof return,the resultssuggestthat itation.Butthebenefitsfromcleaner waterareenhanced when
householdswith high levelsof primaryschoolingbenefit sanitation is improvedandwaterquantityis optimal.When
mostfromirrigation. Thefigureshowshowthe marginal ben- hygiene isalsoimproved,increasingthequantity of waterdoes
efits fromirrigationwouldvaryacrosspercapitaconsump- moreto improvehealththanimprovingitsquality.Education
tionexpenditures if therewerenodifferences intheeducation onhygieneis oftennecessary, though,beforecommunities
levelsof adultsacrosshouseholds. Thebaselineshowsthe realizeits potentialhealthbenefits.
gainsat actuallevelsof education andcompares thosewith A studyof 144waterandsanitationprojectsfoundthat
the simulatedamountsthat would resultif eachheadof improvedwater andsanitationserviceswere associated
household had the maximum five years of primary withamedianreductionof 22percentintheincidenceof di-
education-orif alladultshadthe full five years.Moreedu- arrheaand65percentindeathsfromdiarrhea. But improved
cationraisesthe returnsto irrigation,andthe effect is par- excreta disposalandhandwashing canreduce under-fivemor-
ticularlystrongfor the poor,who tend to havethe least tality ratesby 60 percentandcasesof schistosomiasis by
education. 77 percent,of intestinalwormsby 29 percent,andof tra-
Lackof irrigationinfrastructure is onlyone of the con- chomaby 27-50percent.Otherworkhasfoundsignificant
straintsto reducingruralpovertyinVietnam.Butthe full re- relationships betweenair qualityandhealth.
turnsfromirrigation investmentswill notberealizedwithout Thesecriticalinteractions
betweentheenvironment and
concomitant investmentsin education. healthhighlightthe importance of workingacrosssectorsto
improvepovertyoutcomes.

Irrigationdeliversgreaterbenefitsto Source:
WorldBankforthcoming
b; Klees,Godinho,
andLawson-Doe
better-educatedhouseholdsin Vietnam 1999.

Marginalbenefitsfromirrigation,1993
Dongper square meter per year
400 All adultshave
full primary by 15 percent.5 In Morocco placeswith better rural roads
education also have much higher girls' primary school enrollment
rates and twice the use of health care facilities.6
Another important exampleof the interactionsbetween
300 assets lies in the influence of the environment on health
/ /Baseline / \ \ (box 5.2). Such interactions suggest that poor health in-
dicators in an urban slum, for example, may not be sig-
Headof household nificantly improved by a localhealth center without the
hasfullprimary
education benefits of an effectivesewage system. Increasing human
200 ) 0.5 1 3 6 well-being is thus likely to require action to simultane-
Annual
consumption
expenditure
percapita ously expand complementary assets.
Millionsof dong;log scale

Source:vandeWalle2000a. Publicaction to facilitatethe

accumulation of assets

research found that households with higher education Poor people are central agents in building their assets.Par-
levelshad higher returns to irrigation,with the largestben- ents nurture, care for, socialize, teach skills to, and help
efits goingto the poor (box 5.1).. In rural Philippineselec- finance the education of their children. Small farmers in-
trificationwasestimatedto increasethe returnsto education vest in their land and livestock, while the self-employed
 EXPANDING POOR PEOPLE S ASSETS AND TACKLING INEQUALITIES
invest in materials, equipment, and market contacts im-
portant to their business. Workers migrate to cities and
to other countries, and their remittances are often used
to invest in household assets. Poor women and men save
in housing, in rotating savings and credit societies, and
(whereavailable)in localbanks and credit associations.Still,
(whereavailable) in localbanks andcreditassociations.Still,
the scope for assetaccumulation by poor households is se-
verelyconstrainedby inadequacies
in the marketstheyface
and by weaknessesin public and private institutions for
service delivery.It is also constrained by lack of income,
with poor people suffering severe handicaps in financing
health, education, and other asset-relatedinvestments.
Why doesthe statehavea rolein expandingpoorpeo-
ple'sassets?For two basic reasons. First, markets do not
work well for poor people, because of their physical iso-
lation and because of market failures in the financial,
health, and insurance sectors, for example. Second, pub-
lic policy can reduce initial inequalities and increase the
opportunities for poor people to benefit from growth.
Equity and efficiencyconsiderationscan be largelyin-
dependent, but they generallyoverlap. Poor people, faced
with a failureof a private market, can be left with only the
poor will seek alternative solutions, the nonpoor have
more resourcesat their disposaland so will clearlyhavemore
options. Among the most effectiveantipoverty policiesare
.But ad
thosethat achieve moreequity through redistribution
thosethat achievemoreequitythroughredistributionand
simultaneously enhance the efficiencyof markets used by
poor people (box 5.3, seealso box 3.8 in chapter 3).
While there is a case for state involvement in ex-
panding
poor people's assets, there is no guarantee
panding poor people's assets, there is no guarantee that
that
the state will be effective. Ineffective state action and
unsatisfactory human outcomes partly reflect the fact
that government can influence only a few of the multi-
ple sources of well-being. But they also reflect the diffi-
culty many governments have in delivering goods and
services. Governments are constrained by the fiscal re-
sources at their disposal and their administrative capac-
ity to deliver services effectively.Also, even if they have
the resources and the capacity, state institutions may not
be responsive to the needs of poor people.
How can public action enable poor people to expand
their assets, and how can it tackle asset inequalities? In
three complementaryways.By using the power of the state
to redistribute resources, especially in services that cre-
ate assets,such as education, health, and infrastructure.
By implementing policy and institutional reforms to en-
sure effectivedelivery of services.And by engaging poor
79
Box 5.3
Win-win policiesin the health sector
Thecombinedobjectivesof greaterequityandefficiency are
easierto achieveinsome programsthan inothers,as illus-
tratedinthe following twoexamplesfromthe healthsector.
Some health services, suchas mosquitoandotherpest
controlandhealth educationon basic hygieneandnutrition,
arepurepublicgoods.Others, such as combatinginfectious
diseases,haveclearpositiveexternalities. Stillothers, such
as curativecare for noninfectious conditions,are private
goods.Governments are responsible for infectiousdisease
controlonefficiency grounds.Butsuch policieshaveimportant
equitybenefitsas well.Whilethe poor suffer more from al-
most all diseases than the nonpoor do, the difference is
greatestfor infectiousdiseases.In Indiathe pooresttenth
ofthe population is seventimesas likelyto sufferfromtu-
berculosisas the richesttenth.
Thegeneralinadequacy, if nottotalabsence,ofhealthin-
surancemarketsinmostdeveloping countriesexposesboth
the poorandthe nonpoorto substantial financialriskandin-
security(chapter8).(Thesystemicreasonsforthis failurewere
originally
discussedindetailbyArrow1963.)Whilepublicpro-
visionof insuranceis one policyoption,managing suchpro-
gramsis noteasy.Acommonwayofaddressing the insurance
problemis tosubsidizeexpensivecare,usuallythroughpub-
lichospitals.Butonequitygrounds,thevalueof subsidizing
hospitalsis notas clearas thatofattacking infectiousdisease.
Hospitals areusuallyinurbanareas,andthenonpoorendup
benefiting morefromthe subsidies.Subsidized provisionof
essentialhospitalcarecanenhanceequityaswellas efficiency.
itwillbe a win-win policy-enhancing bothequityandthe
efficiency ofmarketsusedbypoorpeople-onlyifpoorpeo-
plecaneffectively accesshospitalcare,andthatposesa sig-
nificantchallenge.
Source:Hammer1997; Devarajanand Hammer1998; WorldBank
1ss8t.
households and poor communities in decisions on the
choice, operation, monitoring, and evaluationof programs
and services that build their assets.
The rest of this chapter focuses on these three prin-
ciples, with examples from different asset categories. It
then discusses the important complementarities in set-
ting priorities across areas of action, using land reforn
and the provision of urban water and sanitation services
as illustrations.(Expanding accessto financialassetsis dealt
with in chapter 4. Mechanisms to improve the account-
ability and responsiveness of state institutions are the
focus of chapter 6. How to build social assets is covered
in chapter 7. Using assetsand protecting the assetsof poor
people during adverse shocks are discussed in chapters 8
8o WORLD DEVELOPMENT REPORT 2000/2001

and 9. And the role of the internationalcommunityin
assetbuildingis coveredin chapters10 and 11.)
Redistributing public spending
The coercivepowerof the statecanbe a potentforcesup-
portingassetformationbypoorpeople.But thereareclear
bounds to state action. In today'sgloballyintegrated
world intrusivestate action can undercut the function-
ing of marketsand the incentivesforprivateinvestment-
killingjob opportunities,not creatingthem.
In someinstancestherearesignificanttradeoffsbet veen
efficiencyand equity. States generallymirror the un-
equal politicalstructure they are founded on, and gov-
ernment actionoften reflectsthis. Coerciveland reform
underhighlyunequalland ownership,whilepotentially
good for poor peopleand good for efficiency,rarelyhas
enough politicalsupportto be effectivelyimplemented.
Nationalizingindustrial assetsis rarely good for poor
people (whodo not enjoythe benefits)and can damage
efficiency.But there are many other instancesin which
addressingassetinequalitycan enhanceefficiencyand be
goodforgrowth(seebox 3.8 in chapter3).The outcomes
of redistributivepoliciesfor equity and for efficiency
and growththus needto be carefullyevaluated.
The most importantdomainfor stateactionin build-
ing the assets of poor people is the budget. Evidence
suggests,for example,that publicspendingon education
and healthis not progressivebut is frequentlyregressive
(tables5.1 and 5.2). This is alsoan areawhereredistri-
bution is technicallyfeasibleand where tradeoffsbe-
tween redistributionand aggregategrowthmay be low
or negative(indeed,growth can be spurred).But trade-
offsalsoarisein choosingbetweencompetingredistrib-
utiveactionson whichpublic funds couldbe spent.The
choicehas to be guidedby an assessmentof the relative
effectiveness of differentinstrumentsin realizingthe ob-
jectivesof redistributionand povertyreduction.
Budgetaryactionin supportof assetredistributionre-
quirestwo things.The firstis a willingnessandcapacityto
raiseadequaterevenues anddevotea significant
shareofthose
revenuesto development(not to militaryspending,subsi-
diesto the nonpoorandto loss-making publicenterprises,
or illicittransfersto foreignbankaccounts).The secondis

Table 5.1
Publicspendingon education by income quintile in selecteddevelopingcountries,various years
Percent

Quintile
~~~~~~1 ~~~~~~~~5
Country Year (poorest) 2 3 4 (richestl
Armenia 1996 7 17 22 25 29
C6te d'lvoire 1995 14 17 17 17 35
Ecuador 1998 11 16 21 27 26
Ghana 1992 16 21 21 21 21
Guineaa 1994 9 13 21 30 27
Jamaica 1992 18 19 20 21 22
Kazakhstan 1996 8 16 23 27 26
Kenya 1992/93 17 20 21 22 21
Kyrgyz Republic 1993 14 17 18 24 27
Madagascar 1993/94 8 15 14 21 41
Malawi 1994/95 16 19 20 20 25
Morocco 1998/99 12 17 23 24 24
Nepal 1996 11 12 14 18 46
Nicaragua 1993 9 12 16 24 40
Pakistan 1991 14 17 19 21 29
Panama 1997 20 19 20 24 18
Peru 1994 15 19 22 23 22
Romania 1994 24 22 21 19 15
South Africa 1993 21 19 17 20 23
Tanzania 1993 13 16 16 16 38
Vietnam 1993 12 16 17 19 35

a. Includesonly primaryandsecondaryeducation.
Source:Li, Steel,and Glewwe1999;WorldBank1997i(forRomania),2000f(for Ecuador).
 EXPANDING POOR PEOPLE'S ASSETS AND TACKLING INEQUALITIES 8I

Table 5.2
Publicspendingon health by incomequintile in selected developingcountries,various years
Percent

Quintile
1 5
Country Year (poorest) 2 3 4 (richest)
Argentina 1991 33 6
Brazil 1990 8 18 30 25 20
Bulgaria 1995 13 16 21 26 25
Chile 1982 22 11
Ghana 1994 12 15 19 21 33
Indonesia 1987 12 14 19 27 29
Kenya 1992 14 17 22 22 24
Malaysia 1989 29 11
Mongolia 1995 18 20 19 19 24
South Africa 1993 16 17
Uruguayb 1989 37 21 17 14 11
Vietnam 1993 12 16 21 22 29

a. Distributionacrossthese quintilesnot distinguishedin originalsource.

b. Quintilesdefinedby householdsratherthan individuals.
Source:Filmerand Pritchett1999b.

the allocation and management of development spending
to support asset formation for poor people-especially
human and infrastructureassets.Increasingtransparencyin
budgets at both the national and the local level can reveal
the extent to which public resourcesare used in a pro-poor
manner and can improve local accountability (chapter6).
Public financing of servicesis a core element of poverty
reduction policy and practice. Experience suggests two
lessons. First, higher public spending on social services
and infrastructure may not translate into more or better
servicesfor poor people because programs for poor peo-
ple are too often of low quality and unresponsive to their
needs, and because the incidence of public expenditures
is often regressive.However,subsidiesto the nonpoor can-
not be fully avoided because gaining political support for
quality programs may sometimes require providing ser-
vices to a broader segment of the population than just the
poor alone. Second, it is important to use public re-
sources to relax demand-side constraints. Even when
health, education, and infrastructure servicesare publicly
financed, poor people faceconstraints that limit their abil-
ity to benefit from them (for example, complementary
costs, such as transportation to medical care).
Raising resources and making
public spending pro-poor
Asjust noted, effectivepublic redistributionrequiresa will-
ingness and capacity to raiserevenues,especiallyfrom the
nonpoor. In middle-income developing countries withr
high inequality, the nonpoor are often reluctant to con-
tribute their fair share. Unblocking this resistance re-
quires actions to build pro-poor coalitions (chapter 6).
Low-income countries have the added problem of low
public revenues: in 1997 government revenues in these
countries averagedabout 17.5 percent of GDP (exclud--
ing China and India).7 Compare that with around 29.6
percent for high-income countries.
One reason for the disparity is the high cost of rais--
ing revenues in poor countries, costs that sometimes
outweigh the benefits of public spending.8 This in turn
is explainedby a combination of factors:a narrow tax base,
high (and distorting) tax rates, weak tax administration,
and poor public sector governance. So the payoff from
tax reforms that seek to broaden the base,lower rates, and
strengthen revenue collection (often by contracting iv
out to the private sector) can be substantial, making ad-
ditional resourcesavailablefor effectiveredistribution. In
unequal societies,making tax collection as progressiveas
possible without seriously compromising efficiency is
also desirable.For example,inheritanceand realestatetaxes
could be very progressive.
Often, however, the real problem is that the limitecd
public resources are not spent on activities-such as ed-
ucation, health, slum upgrading, and rural development--
that help poor people accumulate assets. Part of the
reason is that many low-income countries are simply
82 WORLD DEVELOPMENT REPORT 2000/2001
spending too much on other areas, such as debt service,
subsidies to the nonpoor, loss-makingor inefficient pub-
lic enterprises, and the military. In the heavily indebted
poor countries more than a fifth of public funds goes to
debt repayments.9 Six heavily indebted poor countries in
Africa spend more than a third of their national budgets
on debt service and less than a tenth on basic social ser-
vices.1 0 Niger spends more than twice as much servicing
debt as it does providing primary health care. For several
other low-income countries, debt service is not the con-
straint becausethey are not meeting their debt obligations.
Still, debt levels and debt service obligations are unsus-
tainable for severalcountries and incompatible with help-
ing poor people accumulate assets.
Military spending in developingcountries fellfrom 4.9
percent of GDP in 1990 to 2.4 percent in 1995." In sev-
eral countries this lower military spending permitted
greater spending on health and higher education.12 But
in other countries-especially those experiencing armed
conflicts or facing unresolved tensions with neighbors-
military spending continues to cut into pro-poor spend-
ing. Many such countries have some of the worst health
and educationindicatorsin the world but spend more than
twice as much on the military as on education and health
combined. High military spending also has significant
costs in lost opportunities for asset building. Beyond
this crowding-out effect, the destruction of physical and
social infrastructure and the slowdown in growth often
associated with military conflicts further limit asset ac-
cumulation and poverty reduction (see box 3.2 in chap-
ter 3).
Spending on the military and on broader security
nevertheless has a role in development. The challenge is
to inform budget allocation by making an intelligent as-
sessment of the threats that a country faces. Better gov-
ernance and transparency in managing military spending
can help keep it in check.' 3 So can the peaceful resolu-
tion of ongoing or potential conflicts.
Betweenthe mid- 1980s and mid- I 990s public spend-
ing on education and health increased in a large num-
ber of low-income countries, though slowly. For 118
developing and transition economies, real per capita
spending increased on average by 0.7 percent a year for
education and 1.3 percent a year for health. Such spend-
ing also rose as a share of total spending and national in-
come. 14 But allocating more funds to these sectors is not
enough. To support asset accumulation by poor people,
the distribution within sectors must favor basic services
used more by the poor and with the greatest market
failures-typically not the case. For example, education
and health resourcesgo disproportionately to tertiary ed-
ucation and to hospital and curative care, used more by
better-off groups. 15
Several studies confirm that public resources favor
the better-off. In Nepal the richest quintile receivesfour
times as much public education spending as the poorest
quintile (see table 5.1). In Ghana the richest quintile re-
ceives nearly three times the public health spending re-
ceived by the poorest quintile (see table 5.2). 6
Infrastructure spending also tends to disproportionately
benefit wealthier groups.1 7 Subsidizing electricity in
Croatia and water in Russia helps the rich much more
than the poor. 18 In Bangladesh infrastructure subsidies
for the better-off are about six times those for the poor.'9
Governments face important political issuesin redis-
tributing public spending to support asset accumulation
by poor people. With finer targeting, public funds may
in principle reach more poor people. But such targeting
may lack political support from powerfulgroups that may
lose out. Hence the importance of building pro-poor
coalitions (chapter 6). This may require allocating some
of the resources to actions and programs that also bene-
fit the nonpoor.
Making public spending more pro-poor will involve
reducing military spending and subsidies to the non-
poor. Privatizing loss-making or inefficient public en-
terprises releasesresourcesthat can potentially be used to
address poor people's needs. Simplifying bureaucratic
procedures reduces not only wasteful spending but the
opportunities for corruption and diversion of resources
to illicit activities as well. Prudent macroeconomic man-
agement can lowerdebt payments and make space for pro-
poor spending. Periodic reviews of overall public
expenditure outcomes can shed light on how efficiently
public resourcesare used and how well they benefit poor
people (see box 9.2 in chapter 9).
For the poorest countries, domestic actions will not
suffice. These actions will have to be complemented by
efforts from the international community to bring about
debt reliefand expand government resourcebasesthrough
development cooperation (chapter 11).
Providing services and targeting subsidies
Public spending can provide services directly to poor
people-through the construction of roads, schools,
health clinics, or water supply schemes. But redistribu-
 EXPANDING POOR PEOPLE'S ASSETS AND TACKLING INEQUALITIES 83

tion can also be achieved by relaxing demand-side con-
straints for poor people by subsidizing the consumption
of privately provided services and covering complemen-
tary and opportunity costs. The Voicesof the Poorstudy
shows how the cost of services can prevent poor house-
holds from obtaining them (box 5.4).
Case studiesconfirm the cost constraintsthat poor peo-
ple face in accumulating a wide range of assets. In rural
areas in the KyrgyzRepublic 45 percent of patients sold
assets (produce or livestock)to pay for hospital care.20 In-
direct costs to households in forgone income from child
labor or in household chores no longer performed by
21
childrenwho are in school can alsobe significant. In rural
Madagascar,where access to water is poor, the high op-
portunity cost of girls' school attendance in time spent
fetching water significantlyreduces girls' education.22 In
Uganda primary enrollment nearly doubled in the
1997/98 school year when the requirement that parents
pay half the cost of school fees was lifted and parent-
teacher association levies were banned.23
Poor people often pay enormous amounts for infra-
structure services.In Nouakchott, Mauritania, most low-
income households spend 14-20 percent of their budgets
on water-costs reflecting the minimal water infrastruc-
ture and the higher cost of the small quantities poor peo-
ple buy.2 4 A cubic meter of water from private water
vendors in Port-au-Prince,Haiti, costs6-1 0 times as much
as a cubic meter from the public water service.25 Similarly,
high connection costs prevent poor households from en-
26
joying energy services. In rural areas, connecting to an
electricitygrid can cost $20-1 ,000.27 In too many cases
poor people simply do not have the choice of consuming
cheaper water and energy from a commercial network.,8
Redistribution, by providing servicesfor free or subsi-
dizingtheir demand, can help poor people expandtheir as-
sets. Freeprimary education for poor people is criticalfor
expanding their human assets,especiallyfor girls.Similarly,
subsidizingpreventionof infectiousdiseasesand helpingpoor
householdsfinance the costsofcatastrophic health episodes
need to be key elementsin strengtheningpoor people'shealth
assetsand reducing their vulnerabilityto health shocks (see
box 5.3; chapter 8).
In both education and health services-even when they
are provided for free-demand-side subsidies can help

Box 5.4
Lockedout by health and educationfees

Whethertoseekmedicaltreatmentoreducationfortheirchildrenpre- Difficultieswith payingschoolfees andother costsassoci-

sentsagonizing choices
for poorpeople.Amongparticipants inthe atedwith sendingchildrento schoolarealsowidelyreported.A
Voicesof thePoorstudy,illnesswasthe mostfrequenttriggerof a motherfromMillbank, Jamaica,explainedthatshecouldnotsend
slideintodeeperpoverty.
Nha,a26-year-dd fatherinVietnam, reported hersix-year-old daughterto schoolbecauseshecouldnot afford
thathehadhadto sellfourbuffalo,a horse,andtwo pigsto payfor the uniformandothercosts.Anotherdaughterhadto dropout
hisdaughter's operation.
Theoperation failedto cureher,andtheneed of schoolbecausethe familycouldnot affordthe $500for school
forfurthertreatmenttransformed
hisfamilyfromoneofthemostpros- fees.Thewomansaid,"My sonwill be readyfor schoolin Sep-
perousinthecommunity to oneof the poorest. temberbut I can't seehow I'll be ableto sendall threeof them
In Pakistan
manyhouseholds reportedthattheyhadborrowed to school."
largesumsof money,soldassets,or removeda childfromschool In somecountrieschildrenarepulledout of schoolbecause
at leastonceto covermedicalcosts.SaidanoldwomanfromGhana, feesareduewhenfamiliescanleastaffordthem.In Ethiopia pay-
"If you don't havemoneytoday,your diseasewill takeyou to mentsaredueat the startof the schoolyearin September, a time
your grave." of two importantfestivalsandthe harvest.Amadi,a 14-year-old
Althoughthe greatestfearfor poorpeopleis the riskof large boyin Nigeria,saidthathehadbeeninandout of schoolbecause
hospitalfees,illegalpaymentsfor primarycarecanalsobepainful. hisparentscouldnotpayhis schoolfeesregularlyandpromptly.
Corruptionin healthcare is widelyreported.Poorwomenfrom Hemissedhispromotionexamsseveraltimesandremainsin pri-
Madaripur, Bangladesh,saidthat the doctorin the government maryschoolwhile othershis agehavegone on to secondary
healthcarecenterignoredthem,givingpreferential treatmentto school.
patientswearinggoodclothesandto thosewho couldaffordside In formerlycentrallyplannedeconomiesthe costof school-
paymentsreferredto as "visit fees." A studyparticipantfrom ingisa seriousconcernfor poorfamiliesbecauseeducationwas
Vares,BosniaandHerzegovina, exclaimed,"Before,everyone free in the past.Peoplealsoreportedproblemswith teachers
couldget healthcare.But now everyonejust praysto Godthat solicitingbribesandspecial"tutoringfees" inexchange for pass-
theydon'tgetsickbecause everywhere theyjustaskfor money." ing gradesanddiplomas.

Source:
Narayan,
Chambers,
Shah,
andPetesch
2000;
Narayan,
Patel,
Schafft,
Rademacher,
andKoch-Schulte
2000.
84 WORLD DEVELOPMENT REPORT 20001200t
poor familiesinvest further in the human capital of their
children (to covertransport costs,for example).To increase
accessto educationfor girlsor minority,indigenous,or poor
children, public funds-in vouchers, stipends, scholar-
ships, grants, and so on-are paid directly to individuals,
institutions, and communities. In Bangladeshthe govern-
ment pays stipendscovering30-54 percent of directschool
expensesfor girlsin grades 6-10. In Colombia in the past
poor children receivedpublic vouchers to attend the sec-
ondaryschoolof their choice.In Balochistan,Pakistan,com-
munity grants are provided for girls to attend community
schools. These programs raise the demand for education
among poor households. Mexico'sProgresa,29for example,
has boosted enrollments among beneficiaries compared
with similar familiesnot in the program (box 5.5).
An alternative to transfers is subsidies on the price of
services. Few developing countries, however, have suc-
cessfullyimplemented price discrimination in health ser-
vicesthrough slidingscalefees.30 In most African countries
such exemptions tend to benefit wealthier groups (such
as civil servants).3 1 In Ghana'sVolta Region in 1995 less
than 1 percent of patients were exempt from health user
fees, and 71 percent of exemptions went to health service
staff.32 In Indonesia and Vietnam poor people can have
user feeswaivedthrough an affidavitof indigence, but few
people seem to take advantage of this-partly perhaps be-
Box 5.5
Mexico's Progresa:paying parentsto send
children~~~~~~~~~~~~
to
childrento school
Mexico's Progresa,an integrated povertyreductionprogram
initiated in 1997,subsidizeseducation,healthcare,and nutri-
tionfor poorruralhouseholds.
Itaimsto reducecurrentpoverty
andincrease investment
inhumancapital,breaking intergen-
erationalpoverty.Progresa
covers2.6millioniamilies-about
80percentof thepopulationinextremepovertyinruralareas.
Progresaprovides grants to poor families for each child
under18enrolledbetweenthe thirdgradeof primaryandthe
thirdgradeof secondary school.Thegrantsincrease for higher
gradesandareslightlyhigherfor girlsthanboys.Forachild
in the thirdyear of secondary school,grants are equalto 46
percent of the average earnings of an agriculturalworker.r
Familiesof childrenwho miss more than 15 percent of the
schooldays ina monthdo not receivethe grantthat month.
Progresahas pushed up enrollmentsat alllevels,withthe
largest effect (17percent) on the transitionfrom sixth grade
to the firstyear of secondaryschool(traditionally
when many
children drop out).
Source:
IDB2000.
cause of social stigma attached to declaring oneself indi-
gent.33 Sometimesprivateand nongovernmentalproviders
are in a better position to implement sliding scales,since
they frequentlyknow their patients' background and have
an incentive to charge what the market will bear.34
For water and energy many developing countries use
increasing block tariffs, charging a low tariff (often below
cost) for the first block of consumption and rising tar-
iffs for additional blocks. In Asia 20 of 32 urban water
utilities use this tariff structure.35 Such tariffs appear to
be more equitable (since they force firms and wealthier
consumers to subsidize consumption by poor house-
holds). They also discourage waste. But there are prob-
lems. In many developing countries few poor households
are served by network utilities, and governments choose
large initial consumption blocks, putting most of the fi-
nancial benefitin the hands of middle- and upper-income
consumers. Ironically, increasing block tariffs discrimi-
nate against poor households that share a water connec-
tion with several other households, because even if the
consumption of each household is low, total consump-
tion is high. In addition, tariffs charged to industries
have often been so high that they choose to self-provide,
undermining the financial viability of serviceproviders.36
Still, in countries where network accessis high, a well-
designed increasingblock tariffcan outperform cash trans-
fersadministeredby poorlyfunded socialprotection offices,
as in some countries in Eastern Europe and the former So-
viet Union.3 7 But in countries where poor households
Buschooliswer oo oseod
have limited access,the subsidy of choice should be sup-
port for a connection, not for consumption. One way to
mitigatehigh connectioncostsis to extendcreditto poor
38
users. Another IS to subsidizeall or part of the connec-
tion fee. Infrastructure subsidies can also be made more pro-
poor if financed through the general budget or through
industry levies in ways that are not discriminatory. 39 This
approach,compatiblewith free entry, provides strong i-
centives to serve the poorest if the subsidy is paid to the
provider only after service has been delivered.
How redistribution is best achieved through transfers
a p s i vi wn
identifv the poor and administer subsidies. If it is possi-
ble to identify poor people individually, any number of
policies can help in redistribution. Direct cash payments,
or the subsidy of any good at all, are fine if the benefits
can be restricted to poor people alone. But it usually is
not possible to tell precisely who is eligible, necessitating
more indirect means (box 5.6).
 EXPANDING POOR
Box 5.6
Some general principleson how to design
subsidies
Evenwhenpoorpeoplecannotbeidentified individually
by ad-
ministrative
means, subsidies
canbedesignedtoreachthepoor.
* Self-targeting. Programscanbedesigned toensure "self-
selection"-say,by payingwagesbelowprevailingmar-
ket rates. The Maharashtra Employment Guarantee
Schemein Indiarelieson providingworkthatonly poor
peoplewouldfindattractive(Ravallion
1999a;seebox8.9
in chapter8).
* Geographic cangoto specificloca-
targeting.Subsidies
tiQns,so that rural
and remoteareasreceivemostof the
benefits.Thisworksbestif the correlation between
povertyandlocationis high-lesswellif poorandnon-
poorliveclosetogether.It alsoworksbestifthesubsidy
is attached
to goodsthatarehardto transport, suchas
directservicesineducation andhealth.
targeting.
* Commodity Subsidies
shouldgo to commodi-
tiesthatpoorpeopleconsume proportionately
moreof
thanotherpeople,ensuringthat theywillreceivemost
of thesubsidy.
Foodandprimary educationusually
rank
highonthiscriterion.
Poolingrisk through insuranceis another way to address
cost constraints on demand for health care. Severalmiddle-
income countries are pursuing universal health insurance
(chapter 8). Chile managed to reach the 15 percent of its
popuLationnot coveredby socialinsuranceby creatinga na-
tional health fund (Fonasa) that collects both payroll de-
ductions for social insuranceand a generalrevenue subsidy
for health care. Still, public resourcesmay be better spent
and poor people may benefitmore if governmentsfocus on
insuringagainstcatastrophichealth incidents-which most
poor householdsare lessable to finance.4 0 Socialinsurance
schemes,evenwhen intendedto be universal,frequendyserve
the better-off first, with poor people receiving coverage
late. Indeed, before coveragebecomes universalpoor peo-
ple may suffer-since the demand and pricesfor privatecare
can increaseas a result of the insurance program, as was the
case in the Philippines.41
Institutionalreformsfor effective
delivery: governance, markets,
and competition
I heard rumorsabout assistancefor the poor, but no one
seems to know where it is.
-Fromn a discuission
gro1up.
Tanjugrejo,Indonesia
PEOPLE'S ASSETS AND TACKLING INEQUALITIES 85
Once countries have settled the political problem of howv
much should be redistributed and the more technicalques-
tion of what is to be redistributed, the next step is to make
surethat servicesdo in factreachthe poor. How can poor
peopleget effectivedeliveryof the servicesthey needto
form assets?The old model of universal state provisioll
too often fails because of lack of financial and adminis-
trativeresourcesor the failureto respondto poor peo-
ple'sneeds.
Part of the problem may be technical and logistica:l.
Poorpeopleoftenlivein remote,low-densityruralareas
that are expensiveand difficult to serve.Resourcesfor
povertyreductionmay simplynot stretch far enoughin
these environments.
But the problem is quite frequently management and
motivation, with inadequate incentives for conscientiouLs
servicedelivery(chapter6). For servicesthat requirethe
presenceof an educatedprofessional-education,health
care,judicial services-it is often difficult to induce skilled
civil servantsto live in remote or rural areas.4 2 With chil-
dren of their own, they often resist living where the edu-
cational and cultural opportunities are limited. Besides
geography there is another kind of "distance" between
providersof servicesand the poor. Since doctors, teachers,
and judges are highly educated, they are often from very
different social classesthan the communities they serve,
making interaction difficult and strained. If pay is deteir-
mined by civil service rules and differentials for difficult
postings do not fully compensate for living conditions, it
3
is verydifficultfor the public sectorto servepoor people.'L
Public provision is generally only part of the picture
for servicessupplied to poor people, however,and other
actors can often overcome the limitations on state pro-
vision. Religious groups often do much in providing ecd-
ucation. NGOs are also a major force in many countries:
in Bangladesh such agencies as the Bangladesh Rural
Advancement Committee play a substantial role in de-
livery,with better results than the government's. And the
private sectorhas alwaysdeliveredservices.More than half
the health servicesin developing countries are private. In
Boliviaalmost three-quarters of visits to health clinics for
treatmentof diarrheaor acute respiratoryinfectionsare
to privatefacilities.44
The privatesectoralsobeganplay-
ing a bigger role in infrastructure provision in the 1990s.
Soundgovernance,competition,and markets-and
free entry for multiple agents, whether government,
nongovernment, or private-are essentialfor effective
service delivery, especially to poor people. (Indeed,
86 WORLD DEVELOPMENT REPORT 2000/200I
nonpoor communities and people are more likely to
make effective use of state systems.) This is not an
issue of the state versus the market, but of the use of
different agents and mechanisms depending on the
type of activity. In education the national curriculum
and exams are a public function, but multiple agents
can provide schooling and communities can hold teach-
ers accountable.
The mix of state and market-and the mix of agents-
depends on the nature of the service and the institu-
tional context.4 5 Where governments are weak, there
might be a stronger case for open entry and reliance on
private and nongovernment agents. But this, too, may re-
quire more monitoring and regulatory capacity than the
government can muster. The importance of institutional
reforms, good governance, and markets in providing
quality servicesto poor people is illustrated here with ex-
amples from health and telecommunications.
Improving the delivery of health services
Sometimes I stayfor long hours until I can seeone of the
doctors, then afterwards the nurse comes and tells me
that he is not coming or he came but he will not be able
to see me.
-l'ooir womall,FlAlailail,
! t/
Despite impressive advances in health in recent decades,
and despite the potential effectiveness of policies and
programs, health services often fail to reach poor people.
A fundamental problem: it is difficult to maintain staff
in rural areas and to ensure conscientious care by those
who do show up. In countries as diverse as Brazil, India,
Indonesia, and Zambia, staffvacanciesin health posts are
much more numerous and last longer in poor and rural
areas than in richer and urban ones.
Improving health servicesin poor communities might
involve changing the incentive structure for public
providers, switching from public provision to public fi-
nancing of private or NGO providers, or changing the
type of servicesthe government is committed to offer,fa-
voring those whose delivery is easier to manage.
Changing incentives in the public sector is often dif-
ficult, with civil service rules often tightly constraining
hiring, firing, promotion, and pay (chapter 6). Different
methods, all with their own risks, have been used, such
as allowingprivate practice to supplement incomes,mak-
ing education subsidies contingent on public service, or
paying extra allowancesfor hardship posts. None of them
is problem free.
In recent years there have been more attempts to de-
centralize health services to subnational levels of gov-
ernment. This also changes the incentives for providers
because they have to satisfy a different set of employers.
Local governments may be more responsive to feedback
from clients. But the jury is still out on the benefits of
decentralizationin health. Sometimes decentralizationhas
simply shed responsibilities from the central govern-
ment-not an example of good decentralization. Suc-
cessfuldecentralization relies on increased participation
of people as monitors of quality (see next section on
participation).
Rather than provide services directly, governments
can make better use of the private sector and NGOs.
In many countries even very poor people prefer to
spend money on services from the private sector (or
from NGOs) if they perceive the quality to be higher
than that of public services.46 This preference can be
exploited by changing the role of government from
providerto financier.
Effectivepartnershipsdrawon the strengthsof each
sector-public, private,nonprofit-in improvingpro-
vision of health services to poor people. The World
Health Organization's immunization program has been
a remarkable success (box 5.7). In Brazil, by having
NGOs compete for funding, the government has har-
nessed the private sector's energy and expertisein the bat-
tle against AIDS and other sexuallytransmitted diseases.
NGOs can often reach segments of society that shy away
from official contact yet run the highest risk of HIV in-
fection. By the end of 1994 NGOs financed under this
system had distributed an estimated 2.6 million con-
doms and taken 11,000 calls to hotlines. Stringent gov-
ernment supervision has also been important in ensuring
that all but 4 of the 191 NGO-run projects financed have
gone forward without a hitch.47
Governments might rethink the types of servicesthey
choose to offer, based simply on what they can credibly
promise. Maintaining permanent staff in rural primary
health care clinics may be too difficult for some gov-
ernments to manage. Changing the mode of delivery or
the types of servicesmay be called for. India recently in-
troduced a campaign to combat polio, reducing reliance
on permanent clinics with short trips by medical per-
sonnel to rural areas-with good results.4 5 Similarly,
rather than relying on permanent health staff who are
 EXPANDING POOR
Box 5.7
Effective public-private partnership
in immunization
In1974the WorldHealthOrganization (WHO) launched the
Expanded Immunization Program, aimingfor80percentcov-
erageof childrenunderfiveby 1990. Although the program
startedas a WHOinitiative, it soonincludedmanymultina-
tionalorganizations,andvolunteers,privateentrepreneurs,
and governmentworkersdidthe actualworkwithincoun-
tries.TheRotarians, forexample,raisedmorethan$240mil-
lionto providepoliovaccineto some500millionchildrenin
103countries.Privatemanufacturers alsotookpart,provid-
inglargevolumes of vaccineat a lowpriceto the UnitedNa-
tionsChildren'sFund'sglobaldistribution networkand still
makinga profit.
Theresultshavebeenspectacular. By1990theprogram
hadachievedits goalof immunizing 80 percentof children
againstthe mostcommonchildhood diseases.Theinitiative
succeededbecauseit combinedpublic,private,and multi-
nationalefforts,witheachorganization usingitscomparative
strength tofit eachcountry's circumstances.
Source:vander Gaag1995.
difficult to monitor and motivate, governments might
find infrastructureprojects providing cleanwater and san-
itation both easier to manage and more likely to improve
health conditions.4 9 And rather than fighting doctors'
preference to work in hospitals, governments might
choose to focus more on gaining access to hospital ser-
vices for poor people.50
Providing telecommunications services
to poor people
Given the right policies and regulations, the private sec-
ror is well positioned to provide telecommunications
services to poor people. Better communications, bring-
ing new influences and broader views of the world, can
raisethe earnings of poor people.51 In Sri Lanka telephone
service in rural areas increased farmers' share of the price
receivedfor crops sold in the capital city from 50-60 per-
cent to 80-90 percent.52
Since the mid-i 980s developing countries have been
opening telecommunicationsto private participation
and competition.5 3 The arrangements range from private
investment in publicly owned companies (China) to
complete privatization and widespread competition, leav-
ing the state to focus on regulation (Chile). Private par-
ticipationhas generallyresulted in rapid growth in access,
lower prices, and better service. In Peru five years after
PEOPLE'S ASSETS AND TACKLING INEQUALITIES 87
reform, the number of fixedlines had increasedmore than
165 percent, the number of mobile lines had risen fromn
20,000 to nearly 500,000, and the number of localeswith
access to telephones had more than doubled. Between
1995 and 1996 the share of households in the poorest
quintile with telephonesincreasedfrom 1 percentto 7
percent.54
Large and small providers can offer services side-by-
side, facilitated by a wide range of innovative technolo-
telecommunicationsservicesin Senegal,costsdropped and
access more than doubled (box 5.8).55 Provision of pay
phones can greatly enhance poor people's access to
telecommunicationsservices, in countries
t elephon call rates particularnynecontrges
where telephone call rates are low but connection charges
are high.5 6 Advances in cellular technology have also
dramatically increased access to telecommunications in
countries where laws and regulations encourage geo-
graphically widespread coverage. Grameen Telecom, a
nonprofit in Bangladesh, uses cellular technology, coni-
bined with the entrepreneurial talents of rural women,
to provide servicesto villages(see box 4.6 in chapter 4).
India, Peru, South Africa, and Thailand have seen dra-
matic growth in privately owned and operated telecen-
ters, providing rural inhabitants with new sources of
information and new opportunities.5 7
Because private providers focus on the most prof-
itable marketsegments,some pockets of the population--
particularlypoorer groups-may not receiveaccessbecause
of the high cost and low revenue potential of extending
Box 5.8
Local entrepreneurs increase access
to telecommunications services
Allowinglocalentrepreneursto offertelecommunications
servicesis animportantfirststepinloweringthecostsofpub-
licaccess Payphonesinparticular benefitthosewho can-
morethan2,000privatetelecenters,eachwitha telephone
andmanywitha faxmachine-fourtimes the numberjust
two yearsbefore.By1998it had6,000.Sonatel(theSene-
galesepublictelecommunications company)franchises
phone
serviceto the telecenterowner,whomaychargea tariffup
to 140percentabovethe Sonatelpricepercallunit.Onav-
erage,telecenters havepaid$3,960 to Sonatelandkept
$1,584each year.Theresult: public
accesstotelephones has
morethandoubled.
Source:Ernberg1998;CSIR1998.
88 WORLD DEVELOPMENT REPORT 2000/2001
serviceto them. To avoid leavingout poor people, in-
novativepublic-privatepartnershipsand well-targeted
governmentsubsidiesmaybe needed. 58 Chileusedgov-
ernment resourcesto improve access for low-income
households,peoplewith disabilities,and public schools,
health centers,and librariesand auctionedsubsidiesto
privateprovidersto payfor ruraltelecommunications roll-
out. In about half the chosenlocations,bids to provide
servicedid not requiresubsidiesas initiallyexpected;the
demand analysisdone by the privatizationgroup per-
suadedprivateinvestorsof the profitabilityof providing
servicesin theseareas.Within twoyears90 percentof roll-
out objectiveshad been achievedfor about half the ini-
tial budget.
Despite successesin extendingtelecommunications
servicesto poor people,privatizationis unlikelyto sig-
nificantlyincreaseaccessin the absenceof greatercom-
petition and more effectiveregulationto preventabuse
of market power. To make private participation pro-
poor in telecommunications-and in infrastructure
more broadly-policymakers mayneed to refocusreg-
ulationsand transactionprocesses.A studyof telecom-
municationsprovisionin 30Africanand LatinAmerican
countriesfoundthat strongcompetitioniscorrelatedwith
per capita increasesin mainlines,payphones,and con-
nection capacity-and with decreasesin the price of
localcalls.It alsofound that well-designedregulationwas
importantin improvingconnectioncapacity. 6 0 A study
on infrastructure reform in Argentina suggests that
public-privatepartnershipscan, with the right policies
and regulations,alsoimproveaccessto infrastructurefor
poor people.6 1
Participation: choice, monitoring,
and accountability
The thirdprinciplefor publicactionto promoteassetac-
cumulation involvesengagingpoor communitiesand
poor people.Participationhas threemain objectives:
* To ensurethat the preferencesand valuesof commu-
nitiesare reflectedin the choiceand designof inter-
ventions.
* To use communityand participantmonitoringto im-
proveimplementation, transparency,
andaccountability,.
* To givepoor people more influenceover their lives.
Participation,while potent, is no panacea.Depend-
ing on localorganizationsand powerstructures,shifting
influenceto local communitiescan lead to greatercap-
ture of benefitsby localelites(chapter6). In Bangladesh
the extentto whichfood-for-education transfersgoto poor
or nonpoor householdsdependson the relativestrength
of organizationsfor the poor and the nonpoor.6 2 Simi-
larly,some localvaluesmay be inimicalto somegroups
of poor people-as with biasesagainstwomen in many
parts of theworld,againstlowercastesin India,or against
other disadvantagedethnic or socialgroups.Shame,de-
nial, and stigmaover HIV/AIDSare in somecountries
reasonsfor local inaction.
The complexitiesof participationimplythat it needs
to be fosteredbyactionsthat strengthenthe voiceof poor
groups in confronting social stratification or stigma.
This impliesthat participationneedsto be shaped in a
broaderinstitutionalcontext.Localgovernmentisin the
middleof the picture,with coreinteractionsbetweenmu-
nicipalitiesand communities-as in the designof the De-
centralizationand PopularParticipationLawsin Bolivia
in the pastdecade.But localgovernmentsoften needto
be strengthenedand made accountable:too often dis-
empoweredin the past, theyfacedifficultiesof weakca-
pacityand localcapture.Civilsocietyorganizationscan
also increase the influence of poor people and poor
communities.Thesebroaderissuesare takenup in chap-
ters 6 and 7; here the importanceof participationin en-
abling poor people to expand their assetsis illustrated
with examplesfrom education,localinfrastructure,and
forestmanagement.
Increasing local participation
and accountability in education
Parentsand localcommunitiesare demandingmore of
a say in children'seducation.As educationsystemshave
expandedin many developingcountries,concernshave
mounted about the qualityof instruction.Centralstruc-
tures are weakin dealingwith dailyadministrativetasks
and too distant to take effectiveaction againstteachers
who do not perform.
Community participationin primaryeducationfre-
quentlyfocuseson monitoringteacherperformanceand
ensuringthe availabilityof schoolsupplies.In the func-
tions most suitable for local management-in-service
training and pedagogicalsupervision-teachers unions
can complementlocal parent-teacherassociations.But
teachersunions often fiercelyopposedevolvingcontrol
of hiring and firingto locallevels,becausethat has often
resultedin delayedsalarypaymentsand at times abuse
by local officials,inciting teacherstrikesin Nigeriaand
 EXPANDING POOR
Zimbabwe. 63 Ongoing reform effortsin Nicaragua
are
seekingto overcomesuch problems(box 5.9).
The overalltrend in educationis to decentralize.In
1993 Sri Lankaestablishedschooldevelopmentboards
to promote communityparticipationin schoolmanage-
ment. In Bangladeshschool managementcommittees
have been reactivatedby the SocialMobilizationCam-
paign,to involvecommunitiesin education.ElSalvador
started involving rural communities in school manage-
ment in 1995.64 In several communities in developing
countries parent groups have responsibility for hiring
and firing teachers and for supplying and maintaining
equipment, under contract from education ministries. So
far, however,there has been little experiencewith full-scale
decentralization of teacher management to schools.
One of the few available quantitative impact evalu-
ations, for the Primary Education Project in Mexico,
shows that educational achievement improved signifi-
cantly during decentralization and that the lower the ini-
tial achievement level, the greater the scope for
Box 5.9
Localparticipation in Nicaragua's decentralized
education system
Since 1993Nicaragua'sMinistryof Educationhas been de-
centralizingpublicprimaryand secondary educationto local
managementboards, based on the followingmodel:
* Legal responsibilityfor public educationrests with the
ministry,but some teacher management is delegatedto
other levelsof the system.
* Thecentercontrolsteacherpreparation,establishesstaffing
levels,funds teacher salaries, and sets standards for
teacher qualificationsand pedagogicalperformance. It
alsodrafts regulationsand financialcontrols.
* Thedepartmentallevelis responsibleforsupervisoryfunc-
tions,providingpedagogicalsupporttoteachers,and mon-
itoringcompliancewith standards.
* Municipaleducationcouncilscomposed of localrepre-
sentatives discharge the administrativefunctions dele-
gated to the locallevel.These councilspayteachersalaries
(with central funds) and approveteacher appointments,
transfers, leavesof absence, and dismissals in accord
with central laws and regulations.They also oversee
teacherincentivesandissueincentivepayments to eligi-
bleteachers.
* Teachers areaccountable for whathappens intheirclass-
room.Parentsmonitortheir attendanceandreport to
the municipaleducation councils,whichinformteachers
weeklyof their statusin relationto incentives.
Source:Gaynor1998.
PEOPLE'S ASSETS AND TACKLING INEQUALITIES 89
6 5 There is also evidencethat supervision
improvement.
incentivesare the most cost-effective input for ruraland
indigenousschools.Nicaragua'sexperiencewith school
autonomy reformindicatesthat a higher level of deci-
sionmakingby schoolsis associatedwith higherstude.nt
test scores,particularlyin schoolsexertinggreater aa-
tonomy in teacher staffing, monitoring, and evalua-
tion.6 6 In Nigeriateachershaveshownup on time since
local supervisionwas introduced.6 7 And socialassessments
of Brazil'sMinas Gerais program and El Salvador'sCorn-
munity Participation in Education program (Educaci6n
con participaci6n de la comunidad, or Educo) indicate
that as a result of the programs, teachers meet more
often with parents and are regarded more highly by
them. Their attendance is also better, which lowers stu-
dent absences.68
Other evidence suggests that community manage-
ment of education can increaseefficiency.Preliminaryre-
sults for the Philippines show that primary schools thLat
rely more on local support have lower costs, holding
quality and enrollments constant.69 In Mauritius parent-
teacher associations have been so successful that govern-
ment funds are being used to further stimulate this
70
partnership.
Effective community management in education may,
however, be hard to achieve. Finding qualified people to
manage schools can be difficult, and the results are un-
even. Botswana has had trouble attracting talented peo-
1 71
pIe to school boards. In the Zambezia provice of
Mozambique parent management committees have led
to beneficial partnerships between communities and
schools in some villages, but in others they barely func-
. . .
tion. Many villagers are afraid to openly criticize school
staff, and committees have been co-opted by corrupt of-
ficials.7 2 There is also evidence that community in-
volvement may have little impact where adults are barely
literate. 73 Despite Educo's success in expanding access in
El Salao D itehas osdeliv her acheveent
El Salvador, it has not delivered higher achievement
scores than traditional schools in the poor rural coim-
munities that were the top priorities. 7 4
Overall, experience suggests that a strong regulatory
framework is needed and that training parents is vital to
make local monitoring of schools effective. Many other
concerns about greater household involvement in eclu-
cation can be addressed through public funding. 75 Con-
tinued monitoring and evaluation of local participation
in education can tell much about what works and whiat
does not.
90 WORLD DEVELOPMENT REPORT 2000/2001

Fosteringownershipthroughparticipation minority-mustbe ableto voicetheirdemands.Local

and choicein localinfrastructure communities reflectexistingsocial,ethnic,gender,and
economicdivisions, andunlessthequestionofwhocon-
Thepolicyof theparty is that thepeopleknow,thepeople stitutesthecommunityis understoodand addressed up
discuss,thepeopledo, but herepeopleonlyimplementthe front,menand localelitesmaydominatedecisionmak-
lastpart, whichis thepeopledo. ing and captureprojectbenefits.In ruralSub-Saharan
-- Fromzadiciissiowg1ro0p, -Fa 7T01b.V'iet;u,an Africamen often identifyroadsas high-priority inter-
ventions, whilewomen,whenconsulted, preferto improve
Communityinvolvementin planningand managing thefootbridges and pathsthat makeup thelocaltrans-
localinfrastructure servicescangreatlyincreaseowner- portsystemon whichtheyrely. 79 In a villagewatersup-
shipandsustainability-ifcommunities makeinformed plyprojectin ruralIndia,watersupplypointswereplaced
choices.Whilelocalinfrastructure is scarcein mostde- nearinfluentialhouseholds.S° Andin Hondurasbenefi-
velopingcountries,theinfrastructure thatexistsis often ciaryassessments showedthat in placeswherethesocial
poorlyconceived and maintained. Why? fundhad financedpipedsewerage, thechoiceof better-
Pasteffortsto providelocalinfrastructure haveoften offhouseholds withwaterconnections, mostcommunity
failedto involvecommunitiesin key decisions,with membershad wantedroads and bridges. 8 ' Provider
centralministriesdecidingon whatlocalcommunities agencies-local governments, NGOs,projectfacilitators-
neededand communitieslearningof a projectonly canreducetheriskof capturebyelites,but theseagen-
whenthe bulldozershowedup. Anotherproblemhas ciessometimes alsotry to capturebenefits.
beena lackofchoice.Facedwithtake-it-or-leave-it, few Requiringbeneficiaries to sharein the costof invest-
communities turn downa freeor heavilysubsidized in- mentcanalsoimproveownership."2 Contributions usu-
vestment. 76 Thiscankilllocalownership, andupkeepof allycomemorereadilywhenthecommunities and local
theinvestmentbecomessomebody else'sresponsibility. governments responsible foroperationandmaintenance
The incentivestructuresforagencystaffcanalsocom- aregivena voiceindesignandimplementation. Localcon-
promisecommunityparticipation-whenthe prefer- tributionsvary.In Ghanacommunities contribute5 per-
ence is speed in implementation.True community cent of the cost for improvedwatersystemsand 50
participationprocesses taketime-it tookyearsfor the percentfor sanitationsystems.Costsharingin menu-
Orangicommunityin Pakistanto agreeon the sewers drivenor socialfund-typeprojectsis typicallybetween
it wantedto install77 -but theygenerateownershipand 5 and20 percentofprojectcosts." 3 Significant financial
sustainability. contributions-between20 and 55 percentof project
Involving beneficiaries in decisionmaking isthestart- costs-have beensuggestedas importantfor sustain-
ing pointin creatinglocalownershipof infrastructure abilityin a studyof ruralwaterprojects." 4 Thereis also
assets,ownershipthat isimportantin threekeydimen- compelling evidencethat ownershipisa functionofthe
sions.First,ithelpsin choosing priorities,particularly in institutionalrelationship betweencommunities andser-
areasdifficultto tacklethrougheconomicanalysis, such viceproviders(chapter6).85
astherelativevalueofsocialandproductive investment, Thereisa trendtowardprovidinglocalinfrastructure
thecomplementarity betweeninvestments, andtargeting throughcommunity-based multisectoral approaches.Y6By
withincommunities. 78 Second, it isessentialforgoodop- givinggreaterchoice,suchapproacheshavethe poten-
erationandmaintenance, becausegovernments canrarely tial to respondbetterto theprioritiesof eachcommu-
be reliedon to performtimelyupkeepof localinfra- nity,contributingto ownershipandsustainability. 87 But
structure.Third,giventheprecarious budgetsituationof it maynot bedesirable to offeropenmenusin everypro-
mostdeveloping countrygovernments and thevastin- ject.If thereis a criticalneedforinstitutionalor policy
frastructure needs(andtotalcosts),localownershipisre- reformin wateror transport,forexample,multisectoral
quiredforcommunitycostsharingin investments and approaches areunlikelyto addressit (box5.10).Similarly,
operation. sometypesofinfrastructure involving morethanonecom-
Fortheparticipatory processto generateownership, munity,suchas roadslinkingmanycommunities, will
all groupsin the community-menand women,those rarelybe demandedby individualcommunities evenif
wellrepresentedin the communityand thosein the theyareneeded.Suchinfrastructure isthusbestsupplied
 EXPANDING POOR PEOPLE
Box5.10
Single-sector
and multisectorarrangements
for improvingruralroadsin Zambia
In Zambiain 1997therewas a criticalneedto clarifythe institu-
tionalarrangements for managing andfinancingruralroads.Many
communities hadconstructedroadswithoutthe involvement of
the localcouncil,motivatedby food aidfrom NGOsor by free-
standingprojects.But thesecommunities, whichhadbeenfully
compensated for theirwork,were unwillingto carryout mainte-
nanceonavoluntary basis.Localcouncils,strapped for resources,
were unableto assumeresponsibility for the roads.As a result,
scarceinfrastructure assetswere goingbackto bush,leaving
communitiesin isolation.
To improveruralaccessibility, the government of Zambiain-
cludeddistrictandcommunityroadsas partof a roadsectorin-
vestmentproject(Roadsip) in 1998.Recognizing the importance
of localownershipin infrastructure, Roadsipaddresses the insti-
tutionalarrangements for the entireroadsector-from the main
highwaysto the communityroads-and the governmentis ex-
ploringwaysto put communityownershipof roadsintolaw.
Experience showsthateffortsfor sustainable improvements in
ruralaccessibility
at the community levelalsohaveto addresssec-
tor policyandinstitutionalreformandmustinvolvethefutureown-
Source:World Bank1998u, 1999ee.
and managed by local governments, although in con-
sultation with communities,8 8 using a unified invest-
ment planning process. Such a process can be followed
by multisector or single-sector projects.
Among the attempts to introduce participatory
processes that allow choice, social funds have been the
most widely studied. Social htnds aim to empower com-
munities by promoting their participation in the selec-
tion, implementation, and operation and maintenance
of their development projects, usually for local infra-
structure.89 But merely making financing availablefor in-
vestments in a variety of sectors is not enough to ensure
that beneficiaries exercisetheir choice.9 0 In many social
fund projects community members are unaware of the
full range of options eligible for financing. In Peru only
16 percent of beneficiaries could cite more than 5 of 19
eligible project types.91 Furthermore, the mere fact that
communities have a choice does not necessarilymean that
it will be an informed one.
For local infrastructure investments to be effective
and sustainable, the demand-based approach gener-
ally has to be complemented by supply-side inputs
(capacity building, information, outreach). Balancing
a bottom-up identification of investments with care-
fully selected supply-side inputs will enhance the
S ASSETS AND TACKLING INEQUALITIES 9I
ersof eachroad.Improving the roadsownedby differentlevelsof
government tookaverticalsingle-sector
approach workingthrough
localgovernments andthe Ministriesof Transport andCommuni-
cations,Public Works,andLocalGovernment andHousing. Theap-
proachfor the communityroadswas a horizontalmultisectoral
approach workingwith the Zambian socialfund.
Only by workingthroughthe socialfund, whichhasa well-
establishedsystem for facilitating communityparticipation,
could Roadsipensurethat communitieswould choosetheir
priorityinvestment.Whencommunitiesrequestimprovements
to roads,footbridges,or paths,theypay 25 percentof project
costsandthe socialfundandRoadsippay75 percent.Oncom-
pletionof a roadproject,the communityforms a roadowners
associationandappliesto the nationalroadboardfor grantsup-
portfor maintenance (75percentcommunity,25 percentroad
fund). Providingcost-sharinggrantsto communitiesfor main-
tainingtheir roadsdoesnot haveto cost much.At $300a kilo-
meter, the cost of supporting the maintenanceof 5,000
kilometersof communityroadswould be less than 2 percent
of annualroadfund revenues.
prospects for equitable and sustainable infrastructure
services for poor people.9 2
Promoting local management offorests
Common property resources, because they possesschair-
acteristicsof both public and private goods, are subject to
free-rider problems that may lead to degradation or de-
pletion in a free market. Poor people suffer the most fromn
these problems when they depend heavily on natural re-
source assets.93 The common response has been state
management,widhregulationsto induce user behaviorcon-
sistent with resource conservation. But the deplorable
environmental outcomes under state-ledprograms, dwin-
dling public resources,and the generalshiftfrom top-down
to bottom-up partnership approaches have recendy in-
creased the emphasis on community-based natural re-
source management. This approach recognizes and
reinforces the role of communities living in and around
vulnerable natural resources, tapping their ideas, experi-
ence, values, and capabilities for preserving their natural
resources.94
Communities often manage natural resources in co-
operation with-and with support from-other coni-
munities and higher (or external) entities, such as local
or districtgovernments,governmentagencies,or NGOs.95
92 WORLD DEVELOPMENT REPORT 2000/2001
Often, such arrangementsapplyto forests.At the center
of joint forest management is an agreement between
governments and communities on the distribution of
use rights and the sharing of benefits,9 6 usually with
communitiesgettinga largershareof forestassetsifthey
communitiesgetting a larger share of.forest.assets.if.they.fringes
achieve agreed conservation and sustainability objec-
tives.97 The state, through the forestdepartment, is often
the owner of the forest and also regulates the system.9 t
The approach has been widely applied in South Asia-
for example, in the Indian state of Andhra Pradesh (box
5.11). In Africa community participation has helped re-
store forest resourcesin The Gambia and led to broader
participation in rural development in Zimbabwe.
But the mixed record in other cases signals the chal-
lengesin making joint management an effectivetool for
promoting poor people's access to key natural assets.
Overly centralizedadministrativestructures havebeen one
reason for failure.The experienceof many countries con-
firmsthat powerfulresistanceat the nationallevelto de-
volvingrightsto forestuserscan blunt effectivecommunity
participation. And forest users or communities, often
unorganizedand with diverginginterests,may lackthe
capacity, interest, and incentives to manage large forestate.
areas.99 The applicabilityand successof joint management
will in each case be determined by the institutional
context-including private interests, local norms, and
traditions-and by the quality of state and local organi-
zations and institutions. Part of the challengeof joint man-
agement is identifyingthe sociogeographic units that
can work together to manage and conserve natural
resources.l°
Severalapproaches have been used to overcome these
obstacles. Providing incentives for stakeholder partici-
pation is essential: granting secure tenure and rights to
forest users,lot more fully transferring management au-
thority over forests (rather than user rights alone) to
communities, sharing benefits, and using socially ac-
ceptable technologies that provide adequate revenue.t0 2
Effective mechanisms for resolvingconflicts are also crit-
ical, especiallywhere resource users' livelihoods compete
with other objectives, such as biodiversity protection or
103
sustainableforestry. Contracts betweenthe government,
villages, and fuelwood collectors in Burkina Faso and
Madagascarspecifywhich subgroups of users manage op-
tions in watershed and protected areas. Effective en-
forcement of these contracts is essential. In the Czech
Republic, Ecuador,the SlovakRepublic, and Ukraine sites
outside protected areas are rezoned to accommodate
Box 5.11
RejuvenatingIndia's forests through joint action
The stategovernment
ofAndhraPradesh
has introduced
joint
forest managementon a massivescale. Peopleon the
of forests are formingvillageorganizationsto pro-
tect forests-vana samrakashna samithi(VSSs).The orga-
nizations workwith thestate forestdepartment,sharingthe
responsibilitiesand benefitsofforestrestoration,protection,
andmanagement.
Theforestdepartmentis responsible for organizing and
providingtechnicaland administrativesupport to the VSSs.
Villages and VSSsare selectedcarefully, but peoplefrom
scheduledcastes and tribesare automatically eligiblefor
membership.TheVSSs protect the forest from encroach-
ment,grazing, theft,andfire,improvingit accordingto a joint
forest managementplan.As compensation,the VSSsare en-
titledto allthe forest's products(nontimberproductsas
well as allthe incomefromthe harvestof timberand bam-
boo)as longas they set asidehalfthe incomeforthe future
development of the forest.
The programgot off to a slow start in 1992because vil-
lagerswere hesitantto assume responsibilityfor forest man-
agement.Inaddition,
forestdepartmentstaffhadreservations
aboutjointforestmanagement.Butin1999 more than 5,000
VSSswere managing more than 1.2 millionhectares of de-
Resultsare impressive,and the programis expanding
rapidly.Thedegradedforestshavesprungbackto life,tim-
ber smugglinghas almoststopped,and cattlegrazingis
undercontrol.Therehas beennofurtherencroachment by
agricultureon lands managed by the VSSs. Manyvillagers
now work inthe forests, and outmigrationhas declined.Soil
conservation
hasresultedinhigherwatertablesinmanyareas,
increasing
agricultural
production. Andlocalplantsand ani-
malsareflourishing.
Source:Venkataraman
andFalconer
1999.
4
multiple land uses.10 Where local capabilities to control
and manage resource use have been eroded or have bro-
ken down, external assistance can help strengthen and
monitor resource sharing and management. °5
As in other examplesof local participation, social in-
equalities can reinforce the influence of politically pow-
erful and better-off groups, further reducing access for
other groups. 106 In particular, women and poor people
who depend on communal assets for their livelihoodscan
be pushed into deeper poverty if they are excluded. 107
While the forest management groups in Andhra Pradesh
successfullyinvolvedwomen, a few programs in other In-
dian states still allow only one household member to
participate-effectively excluding women. In several In-
 EXPANDING POOR PEOPLE'S ASSETS AND TACKLING INEQUALITIES 513

dianvillageswomenwerebarredfrom collectingany ershipremainsconcentrated, andeffortsto increase lan.d

forestproductson protectedlands.10 8 equalityhaveoftengenerated conflict. 1 13
Newapproaches
Evenwhenwomenarenot excluded,theirnumbers to landreformstresstheimportance ofbringingtogether
and their influencein managementcommitteesare variousstakeholders-thelandlessandtheirassociations,
low.'09 A policyand legalframeworkthat promotes theprivatesector(landowners), andgovernment institu-
participationby poorusersin the management of nat- tionsat thelocalandnationallevels.
uralresources 1 Greaterinclusion
canhelpchangethis."' Therearemanywaystogainsecureaccess toland-some
of womenwillalsorequireawareness-raising activities informaland othersformal,somespontaneous and oth-
114
to breakthroughsocietalnormsthatkeepwomenfrom ersrequiringextensive governmentintervention. ]n
playingan equalrolewith men. general,secureaccessto landcanbegainedthroughown-
11 5
ership,tenure,or customaryuserights. Ownership
Complementaritiesin publicaction rightsarethemostsecurebutalsotheleastlikelytobe en-
joyedby the poorand othersociallyexcludedgroups.
Recallthe threeprinciplesfor buildingassetsfor poor Poorpeopleoftengainaccessto landthroughtherental
people-redistribution bythestate,effective
governance and marketandcustomary userights.116 In India,oftheesri-
useof markets,andparticipation. Effectiveactiongener- mated19 percentof ruralhouseholdsthat leaseland,
allyinvolves applyingallthreeprinciples
in a particular
area morethan90percentarelandless orownverylittleland." 17
orsector.Theextenttowhicheachprincipleisappliedde- Whenlandis relatively abundantorpoorpeoplearewell
pendson thestructural conditions,thetypeofaction,the organized andinfluential,theycanhavesecurelandaccess
stateof governance, and the extentof participatory in- withoutformalpropertyrightsandregistration." 8 Thisis
volvement andsocialinequality.Takeeducation. Thereis stillthecasein manypartsofSub-Saharan Africa,though
a caseforpublicredistribution thatseeksto ensurefreeor in recentdecadespopulationgrowthandmarketintegra-
subsidized basiceducationforall.Effective delivery
often tionhaveaccelerated theshiftto individual landrights."1 9
involves multiple agentsprovidingschooling-public, pri- Publicactionis criticalto ensuringsecureaccessto
vate,andcivilsociety.
Andinpoorareasparticipation ofpar- landforpoorpeople.Landreformthatenhancesequity
ents is importantfor increasingcoverage, quality,and and productivitythroughgovernment-supported pro-
accountability. gramsis usuallywhatfirstcomesto mind(seebox3.8in
In thissectionthemixofactionsisillustrated fora nat- chapter3). Butmanyotheractionscanimproveaccessto
uralasset(land)andfortwophysicalassets(urbanwater land.Policies andlawsthatclearlydefinelandrightsand
andsanitation). Thecomplementarities implythataction protectpoorpeopleagainstlandgrabbingcangreatlyen-
is neededon severalfronts.Buttheprioritiesshouldde- hancetheirabilityto uselandascollateral andinvestinland
pendonwhatpoorpeoplelackmostrelativeto theirpo- theyalready"own."Well-functioning rentalmarketscan
tentialopportunities. Inpoorruralareasthismaybebasic raisetheefficiencyoflanduseandhelpthelandless clinib
economic infrastructure,
land-enhancinginvestment,water the"agricultural 20
ladder"to ownership.' Providing legal
andsanitation services,
andbasiceducation andhealthcare. assistanceto poorpeopleenablesthemto presstheirlegal
In urbanslumsitmaybeinfrastructure. Andsecureprop- claimto a plotofland.Similarly, publicsupportto ins-ti-
ertyrightsonlandareimportantforboth. tutionsthat protectwomen'srightscanbe instrumental
wheredeep-seated socialnormsandcustoms inhibitwomen
Enablinggoodgovernance,active markets, fromexercising effectivecontroloverlandevenwhenthere
and broadparticipation in land reform arelegalprovisions forthemto doso (chapter7).121
Landreformhasreturnedto thepolicyagendain thepast Attemptsatlandreformoftenfailbecausetheyrelyon
decade,as manydeveloping countriesmovebeyondim- government alone.Newapproaches emphasize continu-
plementingmacroeconomic reformsto addressingthe ousmechanisms ofadjustmentin landaccess,greaterre-
oftenweakmicro-level supplyresponses "'
in agriculture. lianceontraditionalformsofaccess, andgreateruseofland
It is easyto seewhyfarmlandis a keyassetforthe rural markets.'22 Brazil,
Colombia, thePhilippines, andSouth
poor.Butsecureaccessto landandforwhomandunder Africaareexperimenting withdecentralized, community-
whatconditions remainsa thornyissue.1
12 Poorpeople, based,demand-driven negotiationsbetweenstakeholders
especiallypoorwomen,oftenlacklandrights.Landown- to findlessantagonistic waysto improveaccessto lanl.
94 WORLD DEVELOPMENT REPORT 2000/2001
Such negotiated land reform relies to a great extent on
experiencegained in past attempts at land reform and on
the successful aspects of demand-driven social funds. It
addresses both supply and demand. Landlords are paid
in cash rather than with highly discounted government
bonds, as in the past. Beneficiaries receive grants for
.ictivirv.
prod enhancing investments, because experience
productivity-enhancing investments, because experience
shows that unless those who receiveland can make pro-
ductive use of it, reforms will be undermined.12 3
The negotiated approach has several innovative com-
ponents, including strong involvement of local governments
and communities in organizing land transfers and assist-
ing the beneficiaries-and a high degree of transparency
5.12). In Colombia municipalities must develop a
(box 5.12). In Colombia municipalities must develop a
comprehensive plan to identify potential sellers and ben-
eficiaries, and the plan is widely circulated to avoid cor-
ruption. Potential beneficiaries are offered training in farm
management and assistance in developing land use plans.
These plans must then be approved in public meetings of
municipal councils. 12 4 Beneficiaries are supported by na-
tional, regional, or state councils that provide technical guid-
ance and resolve administrative obstacles.
Theresults
ground from
on the
the community-based
approach are encouraging. Yet it Is still too early to fully
evaluate this new generation of reforms. While the ne-
gotiated land reform has been criticized for, among other
things, burdening beneficiaries with loans they cannot
repay and trying to replace expropriative land reform, re-
visions to the approach have sought to address these and
other concerns. Another issue is costs. Although in Brazil
and elsewhere this approach has achieved savings of as
much as 40 percent relative to expropriative reforms, it
still requires significant public outlays. 125 Its effectiveness
needs to be assessed relative to other instruments for re-
ducing povertv.126
1 2 6
ducingpoverty.
Providing water and sanitation services to
the urban poor
Thesewagerunsright in yourfront door, and when it
rains,the waterfloodsinto the houseand you needto lift
things. .. the wastebringssomebfugs.Herewe baverats,
cockroaches, spiders,and evensnakesand scorpions.
lYouiwa
o o/i ousWfl
,r-oup,
Xova(U ;nm, Br.zzil
To improvewater and sanitationservicesto urban resi-
dents,governmentsandmunicipalitiesthe worldoverare
Box 5.12
A new approach
to landreformin Brazil
Brazil
hasbeenexpropriating
andredistributing
landsincethe
mid-1960s, reachingan impressivenumberof beneficiaries,
butwithhighcostsand unevenquality.
Toincreasequality,lowercosts,and speedreform,five
stategovernmentsinthe Northeaststarteda pilotprogram
of negotiated,decentralized,
community-driven landreform
in1997. The program providesloans for land purchases to
landless rural dwellers (sharecroppers, renters, landless
workers, labor tenants) or to smallholders who organize
themselves in beneficiaryassociations. The associations
have to identifylandlordsinterested insellingthem land,an
approachthat fosters direct negotiationsbetween owners
and the associationsand reduces governmentintervention.
The pilot program also providesgrants for productivity-
enhancingcommunityprojects identifiedby the beneficiary
associations, drawing on well-establishedpoverty allevia-
tionprojects inthe fivestates. The philosophyof the projects
is that beneficiaryassociations are best placed to identify,
rank,and implementinvestments, drawingon technicalas-
sistance as necessary.Thesame participatoryphilosophyhas
been adopted in the land reform pilots.
Theprogramhas three grantelements constitutingabout
50 percent of the cost of the land:an inherentsubsidyin the
interest rate,a grantfor the complementaryinfrastructurein-
vestment,and an installationgrant.A creditwitha maximum
term of 20 years is providedfor the land.The land guaran-
tees the loanand has to meet a number of requirements,in-
cluding a price comparable to what prevails in the local
market.Sincethere is a maximumgrant per beneficiaryfam-
ilyandthe infrastructureinvestmentgrant is a residual,there
is an incentivefor the associationto buythe land at the low-
est possiblecost.
Althoughthe numberof beneficiaryfamiliesisstillmodest
(about10,000in330 projects),the pilotprogramhas pickedup
momentum,and its success has promptedthe central gov-
ernmentto expandthe programnationwide.Complementing
the larger,expropriativeprogram,the new programfocuseson
purchasingpropertiesthat cannotbe legallyexpropriated.
Source:Deininger
forthcoming.
exploring alternative approaches involving the private
sectorand localneighborhood
and civilsocietygroups.
Workingtogether,theseactorsseekto providequality
servicesto poor consumerswhileensuringfinancial
sustainability.
Waterandsanitation
services
havetraditionally
been
providedbypublic agencies.Whilethereare exceptions,
fewdevelopingcountrieshaveelicitedstrong,sustained
performancefrompublic waterand wastewaterutilities.
Weakperformanceincentivesand difficultiesin shelter-
 EXPANDING POOR PEOPLE
ing management decisions from political interference
havelocked many utilities into a cycleof poor service,low
user willingnessto pay,and insufficient maintenance.The
inability of supply to keep pace with increasing demand
from growing urban populations has forced poor house-
holds to find their own solutions. In many countries
small informal water vendors and sanitation providers
reach poor urban areas unserved by government utilities.
In West African cities independent entrepreneurs supply
most poor households (box 5.13). Similarly,in Guatemala
City and Lima, Peru, which both have major utility com-
panies, most families depend on private informal
providers.12 7 Although local suppliers can be more ex-
pensive than public providers, households would be
worse off without them.
Since the early 1990s there has been a marked increase
in large-scaleprivate participation in water and sanitation
in developing countries, reflecting a desire to deliverbet-
ter services at lower cost-including services to poor
urban neighborhoods. 128 Private participation can boost
service coverage and make utility operations more effi-
Box 5.13
West African businessespioneer water and
sanitation servicesfor the urban poor
Africa'sindependent
waterandsanitation
providerssuggest
that the market has found solutions that benefit everyone:
providers,utilities,and,aboveall, low-incomecustomers.
Recentstudiesin sevenWestAfricancitiesshowthathalf
the residentsrelyon privateindependent providersfor their
waterandat leastthree-quarters rely onindependent oper-
ators for sanitation. Depending on the city, independent
providers coverupto 85 percentof marginal andlow-income
neighborhoods,
andtheyservemanybetter-offfamilies
aswell.
Independentsanitationproviders,workingwithout an
officialmandateorarrangementwith localgovernments,in-
clude small informal operators as well as a few that have
grown and become "legitimate.' Providers rely on good
client relations, since their operationsare completely
demanddriven.Consumerabilityto pay andcompetition
amongprovidersdetermineprices.
Independentsanitationproviderschargehigherpricesthan
subsidized publiccompanies, but publiccompanies rarelyre-
couptheiroperatingcosts-let alonethe costsof installing
sewernetworks.Andthe independent providersaregener-
allyreliableandresponsive to their customers.
Theyextend
credit(for a few daysat least)andspreadcollectionsover
days and weeks, far easierfor poor clients to fathom than
the three-month
billsfrom publiccompanies.
Source:Solo1999.
S ASSETS AND TACKLING INEQUALITIES 95
cient, and the early results are encouraging. Coverage has
increased rapidly, and in some cases tariffshave fallen (as
in Manila, Philippines).The water supply systemin Cbte
d'Ivoire, which introduced the first private concession in
Sub-Saharan Africa, performs better than other urban
water systems in West Africa.129
But large-scaleprivate participation in water and sani-
tation does not automaticallymean better servicesfor poor
people. Unlesscarefullycrafted,contracts may precludethe
extension of servicesto low-income areas and create local
monopolies. Contracts often mandate tariffstructuresand
set connection fees that do not vary with the true cost of
connection.These featuresdiscourageconcessionairesfrom
delivering service to low-income areas.130 In Guayaquil,
Ecuador, residential water tariffs did not cover collection
costs. Everynew connection, even if fully grant financed,
31
wasa net revenuedrain on the utility.' To servelow-incomre
householdsbetter, concessionsin BuenosAires,Argentina,
and La Paz-ElAlto, Bolivia,have been restructured based
on negotiationsbetweengovernmentsand privateproviders
and input from local stakeholders.
Good pricing policy is a key element of pro-poor pol-
icy.'32 Whether water and sanitation utilities are publicly
or privatelymanaged, those most successfulin expanding
these servicescharge tariffsthat covercosts. Such tariffscan
increaseaccessfor low-incomehouseholdsby attractingpri-
vate investment to expandsupplyand enhancequality.They
can also end general government subsidies that go mostly
to the nonpoor, releasing public resources for more targeted
assistance to poor people. Notwithstanding the overall
trend in water and sanitation toward greater cost recovery,
governments can ensure greater accessfor poor people by
subsidizing connections or, where network access is higlh,
using well-designed block tariffs. 133
Additionalmeasures to benefit poor households and
attract private investors to water and sanitation include
simplifying contracts, contracting out some regulatory
functions, and increasing the predictability of regula-
tory discretion. 134 The designof regulation-particular]y
to reduce monopoly power-is also critical for pro-poor
outcme 135 Regulation can enhance competition by
outcomes.
permitting greater entry, including by nonconventional
suppliers, and by changing service standards to fit local
needs-for example, focusing on the potability of water
rather than on technical construction standards set at in-
dustrial country levels.' 3 6
Also important is involving users and local institutions
in designing private sector options with user preferences in
96 WORLD DEVELOPMENT REPORT 2000/2001
mind-and fullyleveragingthe presenceof alternativeser-
vice providers.This is the approach of the Water and San-
itation Program-a partnership of donors, governments,
and NGOs that focuseson poor rural and periurban areas.
In each context this demand-responsiveapproach must be
tailored to local conditions. Brazil'sProsanear (Waterand
Sanitation Program for Low-Income Urban Populations)
followssix principles to providesustainablewater and san-
itation servicesto poor households (box 5.14).
A similar approach aims to cover 35 towns in an
urban water and sanitation project in the Philippines.
Communities decide to participate (borrowing money
from the Development Bank of the Philippines) after ex-
tensiveconsultationsinvolvingconsumers, the mayor,and
the town council. Service charges are used to repay the
loan and cover operation and maintenance. The in-
volvement of users, together with local government de-
cisionmakingon participation, appears to ensure the
ownership needed to enhance prospects for long-term
sustainability.137
3* * political and social structures and the assets of poor peo-
The assets that poor people possess-or have access
to-directly contribute to their well-beingand havea po-
tent effecton their prospectsfor escapingpoverty.Human,
physical, natural, financial, and social assets can enable
poor people to take advantage of opportunities for eco-
nomic and social development (just as their lack can
prevent this). Expanding the assets of poor people can
strengthen their economic, political, and social position
and their control over their lives.Assetsempower the poor.
And assets help people manage risks (chapters 8 and 9).
Box 5.14
Sustainablewater and sanitation for Brazil's
urban poor
Thefirstphaseof Prosanear(1992-97)was a periodof learn-
ingthatledto sixguidingprinciples forsustainableprovision
ofwaterandsanitation inpoorurbanneighborhoods inBrazil:
* Startcommunity participationat theverybeginning of pro-
jectpreparation.
* Ensurethat cost-recoveryand subsidyrulesare clear
andtransparent.
* Makeformal,long-termarrangements foroperatingand
maintaining systemsan integralpartof the design.
* Discussallfeasibletechnicaloptionsandtheircostswith
communities.
* Coordinate projectswiththe localgovernment's urbande-
velopmentplanfromthe outsetof preparation.
* Confirmthat the localgovernmenthas a strongcom-
mitmenttothe projectandto povertyreduction.
Source:KatakuraandBakalian1998.
But because there is a two-waycausalrelationshipbetween
ple, it may be necessary to tackle exclusionary or weak
social structures in order to form assets (chapter 7).
Public action is essential to expand poor people's as-
sets and to tackle asset inequalities-particularly in the
distribution of human assets. Effectively using the re-
distributive power of the state and involving multiple
agents (civil society, markets, and the state) and stake-
holders in the provisionof servicesare crucial to this end.
Expanding poor people's assets is at the core of getting
the benefits of growth to reach poor people faster. And
it can be pro-growth (chapter 3).
 PART III

Emp owe rm eent

f . , 1^4
Making State Institutions
M'ore Responsive
to Poor People
iP overty is an outcome not only of economic
processes-it is an outcome of interacting economic,
social, and political forces. In particular, it is an out-
come of the accountability and responsivenessof state
institutions.1
As this chapter discusses,the state will delivermore
effectivelyto all its citizens, but to poor people in par-
ticular,if:
* Public administrations implement policies effi-
ciently and are accountableand responsiveto users,
corruption and harassment are curbed, and the
power ofthe state is used to redistribute resources
for actions benefiting poor people (chapter 5).
* Legal systems promote legal equity and are acces-
sible to poor people.
* Central and local governments create decentralized
mechanismsfor broadparticipation in the deliveryof
public servicesand minimizethe scopefor capture by
local elites.
* Governments generate political support for public
action against poverty by creating a climate favor-
CHAPTER 6
ableto pro-pooractionsand coalitions,facilitating
the growth of poor people's associations,and in-
creasing the political capacity of poor people.
* Politicalregimes honor the rule of law, allowthe ex-
pression of political voice, and encourage the par-
ticipation of poor people in political processes.
Publicadministrationand poverty
reduction
It is hard to get to the rghtperson in the munwapaity, and
whenyou do he says, "I'msorry, Lam not abl to he4pyou"
-From a discussiongroup, Zenica,
BosniaandHerzegovina
In most developing countries poor people have trou-
ble getting prompt, efficientservicefrom the public ad-
ministration (box 6.1). To change this, the first step is
building the capacityof public administration. Officials
also need tractable regulatory frameworks,with proper
performance incentives and mechanisms to ensure ac-
99
IOO WORLD DEVELOPMENT REPORT 2000/2001
Box 6.1
Poor people are often harassedby public officials
Poorwomenandmenin the Voicesof the Poorstudystressed
thatofficialsareoften unresponsive
to them.Theysharedcount-
lessexamplesof criminality,abuse,andcorruptionin their en-
counterswithpublicinstitutionsandsaidtheyhavelittlerecourse
to justice.Indescribing
theirencounters
withinstitutions,
poorpeo-
plealsodrewattentionto the shameandindignityof beingtreated
with arrogance, rudeness, anddisdain.
Whenthey assistyou theytreat you likea beggar.. . but we
aren't ... we pay taxes.... Theremust be transparencyin
government actions, tax money has to be well employed.
... They invent theseuselessconstructionsand grab our
money...
-Poor man,VilaJunqueira,
Brazil
Somereceiveus,othersdon't.It's awful.... Theyareabusive
... Theytreatonealmostlike a dog.... Themunicipalityonly Wewouldrathertreatourselvesthango to the hospital,where
servesthe high-societyones.... the angrynursemight inject us with the wrongdrug.
-From a discussiongroup, Esmeraldas,Ecuador
Source:Narayan,Chambers,Shah,and Petesch2000; Narayan,Patel,Schafft, Rademacher,and Koch-Schulte2000.
countability and responsivenessto clients, including poor
people.2 Poor organizationaldesignengenders inefficiency
and corruption, typically hurting poor people the most.
Focusing public action on social priorities
In nearlyeverycountry the public sector often pursues ac-
tivitiesthat are not sociallyjustifiedand, in somecases,that
generate rents for the elite. During the past two decades,
as societiesand their governments have become aware of
this problem, they have launched public sector reforms to
focus public action and programs on social priorities and
increasethe capacity of the state to reduce poverty.
Public sectorreform and modernization have great po-
tential to reduce poverty, if they are at the core of a de-
velopment strategy that establishes clear priorities for
public action. The functional and organizational struc-
ture of the public sector needs to be rationalized to im-
prove resource allocation for programs that are social
prioritiesand havegreatercapacityto reduce poverty.Most
important is to streamline and "rightsize" public ad-
ministrative entities and privatize public enterprises and
other operational public programs.
Beyond rationalizing the structure of the public sec-
tor there is a need to improve public management sys-
tems to make public programs more efficient and
accountable. Involving civil society in planning, moni-
toring, and evaluating public programs and policiesis also
Theofficialsof the socialassistancedepartmentareimpolite
andevencrudewithordinarypeoplefromthe village.Igo there
formy socialbenefitformychildren.I haveto waitfor twohours;
theytreatme verybadly.If I cry andshoutthat my child is ill,
they'll give me something.But it happensseldom.
-Woman, NovyGorodok, Russian Federation
We in the countryget up at 6 a.m. to takethe collectivebus.
We arrive.We go to the doctor at the hospital. Youarrive at
8 a.m.or sometimesnot until I p.m. Youarestuckthereuntil
the afternoon,without eating,without beingable to drink ...
you spendhoursand hourshungry. Youhave to go backbe-
fore the doctor hasseen you. Youmiss the bus. Youhaveto
go howeveryou can...
-Twenty-five-year-oldmother,LosJuries,Argentina
-Poor youth from Kitui,Tanzania
crucial to ensure steady progress toward a fully respon-
sive and accountable state.
Enabling and motivating public
administrations
Having the right performance incentives smooths the de-
livery of public services.Key incentives include merit-
based recruitment, clear specificationof tasks, rewards for
good performance,and insulation from excessivepolitical
pressure.3 Togetherwith skilled technocrats and close col-
laboration with the businesscommunity, these make up
what has been termed "the developmental state."4
Merit-based recruitment goes a long way toward im-
proving administrative performance. When nepotism or
cronyism exists, it is difficult to motivate staff to perform
well.5 Cross-country analyses indicate that merit-based
recruitment is associated with lesscorruption and fewer
delays (figure 6.1). Merit-based promotion is also essen-
tial for motivating staff. If there are few opportunities for
promotion, or if promotion is unrelated to performance,
staff have much less incentive to perform. What is im-
portant is to promote an evaluation culture, for staff and
for agencies. Also important to good performance are
clearly specified and tractable tasks and competitive
salaries.Compensation of public servantsthat is severely
out of line with that in the private sector affects perfor-
mance incentives and encourages corruption.6
 MAKING STATE INSTITUTIONS MORE RESPONSIVE TO POOR PEOPLE IOI

Figure 6.1 Legislativeoversight of the executive, carried out ac-

Merit-basedrecruitment in government is cording to transparent procedures, is an important part
associated
with lesscorruptionand of monitoring and improving performance. Public ad-
bureaucratic
delay
ministrations also need to be supported and actively
monitored by political leaders. Surveysin several devel-
oping countries show that public officials' performance
Corruption improveswhentheyknowthat electedrepresentatives
are
High overseeingtheir work. But sometimesthis processbecomes
subject to the personal goals or whims of elected repre-
sentatives, resulting in excessive political interference.
The quality of public service is reduced when public of-
ficials are held accountable more to their hierarchical
superiors than to the people they serve.7

Making the public sector more responsive

to client needs
Many different kinds of measures help improve public
sector service delivery. One important measure is sim-
plifying procedures and making them transparent to
LLow clients. In the Philippines several public agencies have
Low Merit-based recruitment High streamlined procedures to curb corruption. At the out-
set of a transaction clients receivea list of required doc-
High Bureaucratic delay uments along with a timetable showing how long the
process will take and a schedule of fees.8More generally,
simplifyingand improving regulatoryand tax systemsand
privatizing state-owned enterprises can reduce the op-
. . . . portunities and scope for corruption.
* . . Another important measure is disseminating informa-
. . . * tion to allow people to monitor public services. Using
newspapersand other popular information sourcesto dis-
seminate information on budget allocationsand spending
enablespeople to hold civilservantsaccountable,reducing
inefficiencyand corruption. In Uganda, when primary en-
rollments did not improve despite substantialincreasesin
budget alLocations,a survey of schools examined public
Low. spending on primary education. The study found that
Low High budgetallocationsmay not matter when institutionsor their
Merit-basedrecruitment popularcontrolis weak:in 1991-95 on averagelessthan
Note: The figure is basedon the responsesto a survey sent to 30 percentof the intendednonsalarypublicspendingonl
expertsin developing
countries,with 126completeresponsesfrom primaryeducationreachedschools.The governmenthas
35countries.Merit-based
recrurtmentcapturestheshareof higher
officials in core economic agencieswho enter the civil service since improved performance by increasingthe flow of in-
througha formalexamination systemor haveuniversity or formationwithinthe system.A major breakthroughwas
postgraduate degrees.A highscoreon corruption indicatesa strong
likelihoodthat high governmentofficials will demand special achieved
by makmg regularannouncements m local news-
payments and strong expectationsof illegalpayments throughout papersandon the radioof the publicfunds transferredto
low levels of govemment.A low score on bureaucratic delay districtsand postinginformationon transfersat each schoo:l.
indicatesgreaterspeedandefficiencyof the civilservice.
Source: Rauch and Evans 1999. A follow-up survey in 1999 showed dramatic improve-
ments since 1995, with schoolsreceivingcloseto 100 per-
cent of the nonwage public funding.9
102 WORLD DEVELOPMENT REPORT 2000/2001

Fostering communication between civil servants and Figure 6.2

their clients is also important. Many developing coun- Corruption is a regressivetax
try administrations have poor mechanisms for learning
about andrespondingto users'demands.In Indiathe "re- 10Percent
port card" on Bangalore'spublic servicesshowshow a pub- Bribe cost as a share Bribe cost as a share
lic feedback mechanism can make public agencies more of firms' revenue of households' income
accountable to their clients. Launched in 1993 by a 8
group of committed citizens,the report card provided cit- 7
izens' views of public service delivery in the city. Re- 6
spondents focused on agenciesthey dealt with to redress 5
a problem or to get a service-ranking their satisfaction 4
and indicating the time spent. The findings were dis- 3
seminated to public agencies,the media, and NGOs, trig-
gering some service providers to become more efficient 2
and accountable,The BangaloreCity Corporation helped I
set up an informalnetwork of city officialsand non- 0
, , ~~~~~Micro
Small Large Low Middle High
governmental groups to meet periodically and work out (<11) and (>99) income incomeincome
answers to prioritv problems.'l medium size (<$110) ($110- (>$329)
(11-99) 329)

t_urbingcorruption
Corruption takes a toll on economic performance, un-
dermines employmentopportunities, and cloudsprospects
for poverty reduction. Even petty corruption dramatically
raisesthe cost of engaging in productiveactivities.In West
Africa bribes in the transport industry are crippling. The
estimated cost of transporting goods from C6te d'Ivoire
to Niger includesbribes to customs, police, and transport
officialsthat represent three-quarters of payments to the
administration. 1" Similarly,a transport trip in Benin en-
countered 25 roadblocksover 753 kilometers-road-
blocks staffed by state agents who demanded bribes that
added up to 87 percent of the cost of the trip.' 2
The burden of petty corruption fallsdisproportionately
on poor people (figure6.2). For those without money and
connections, petty corruption in public health or police
servicescan have debilitating consequences. Corruption
affectsthe livesof poor people through many other chan-
nels as well.'3 It biases government spending away from
sociallyvaluable goods, such as education. It diverts pub-
lic resources from infrastructure investments that could
benefit poor people, such as health clinics, and tends to
increasepublic spending on capital-intensiveinvestments
that offer more opportunities for kickbacks, such as de-
fense contracts. "4It lowers the quality of infrastructure,
sincekickbacksare more lucrativeon equipment purchases.
Corruption also undermines public servicedelivery.
Streamliningbureaucraticprocedures,simplif,ringtax sys-
tems, eliminating excessiveregulations, and motivating
Note: Results are for Ecuador.For firms, the values in parentheses
refer to the numberof employees;for households,
the valuesrefer
to monthly household income. The figure is based on preliminary
data from a 1999 survey of 1,164 enterprises and another of 1,800
households.
Source: Kaufmann, Zoido-Lobaton, and Lee 2000.
public servantscan help reduce the opportunities for cor-
ruption. And community participationand monitoring can
keep it in check.
Poor people and the rule of law
There are four dragons: law court, prosecutor's office,
khokimiat, and head of police. Vobody can get anything
until they are satiated.
--iF ir li,nlo,)i)
cra gorp. ()itano,lali,I /.Zblkistan
The rule of law means that a country's formal rules are
made publicly known and enforced in a predictable way
through transparent mechanisms. Two conditions are es-
sential: the rules apply equally to all citizens,and the state
is subject to the rules. How state institutions comply with
the rule of law greatly affects the daily livesof poor peo-
ple, who are very vulnerable to abuses of their rights.
The rule of law is upheld through many channels, the
most formal being the legal and judicial system. The
legal and judicial system constrains and channels gov-
ernment action-and maintainsclear rules and procedures
for upholding an individual's constitutional rights. This
 MAKING STATE INSTITUTIONS MORE RESPONSIVE TO POOR PEOPLE 103

system is essential for guarding against abuse of power by tecting their interests. Legal systems, the product of
the state or other actors, and it requiresthat the judiciary power relations between different groups in society,typ-
be independent of the executiveand legislativebranches. ically focus on protecting the interests of those with pc-
The rule of law protects life and personal security and litical strength and representation. Making laws and their
guards against human rights abuses.Thus defined, the rule interpretation more sensitive to the needs of the disad-
of law is tremendously important for all citizens-but es- vantaged requires building coalitions to this end. This is
pecially for poor people, who have few private means of the goal, for example, of efforts to make laws more eq-
protecting their rights (box 6.2). uitable in their treatment of women and minorities
The rule of law is associated with better overall eco- (chapter 7).
nomic performance (figure 6.3), and in this sense it also Legal obstacles leave poor people vulnerable to ex-
promotes poverty reduction. It does this by creating a pre- ploitation by local bossesand the police,and arbitrary ha-
dictable and secure environment for economic agents to rassment, lawlessness,and violenceare constants in their
engage in production, trade, and investment, thereby lives. For poor people, a crucial aspect of the rule of law
expanding poor people's employment opportunities and is the ability to live without fear of lawlessnessand ha-
incomes.15 Market mechanisms depend on credible rassment. An effective modern police force is needed to
threats of punishment for breakingcontractualobligations, maintain order by enforcing the law, dealing with po-
backed by prompt methods for resolvingdisputes and en- tentially disorderly situations, and attending to citizens
forcing contracts. Without these deterrents, the transac- in distress.
tions costs of doing business can be very high.
Although the rule of law benefits poor people in many Making the legal system more responsive
ways, laws and statutes are not necessarilygeared to pro- to poor people
Even when the legal system is well run, poor people face
Box 6.2 constraints in tising it.16 Poor people typically have lil;-
Lawlessnesscontributes to poverty
Figure 6.3
Ata hospitalinBabatidistrictin Tanzaniaa newdeliveryofes- Better rule of law is associated with higher per
sentialmedicalsuppliespurchasedwithforeigncurrencydis- capita income
appearsfromthe publicdispensarywithinhoursbut is available
for purchaseat the doctor's home that evening.The poordo Per capita GDP
not receivethe freemedicalcarepromisedbythe government, u.s. dollars (PPP)
but those withthe rightconnectionsandthe abilityto paycan 25,000
secure pharmaceuticalsin abundance.
InJohannesburg,SouthAfrica,rates of theft and violent
crime are among the highest in the world. Wealthy resi- 20,000 *
dents can affordsophisticatedalarms,security guards, and
other forms of privatepolicingto protect their property and 0.
persons. Poorpeopleare stuck inpoorlybuilthomes, some- 15,000 *
times withouteven simplelocks,and are vulnerableto theft, ,
assault, murder,and other violentcrimes. I
In Pakistana man too oldto work is left without assets 10,000
°°e.
or income after his son is murdered.To gain access to his .
son's estate, he needs a succession certificate from the e J
civilcourt in Lahore-more than 160 kilometersaway. The 5 *t *
trainticketand the bribedemanded bythe clerkof the court . * _ * ,iA *
send the man deeperintodebt,yet after fivetripsto the court 0 _
inas manymonths,he stillhas not receivedthe stamped -2.5 -2.0 -1.5 -1.0 -0.5 0 0.5 1.0 1.5 2.0
piece of paper to which he is legallyentitled.The clerk re- Rule of law index
fuses to producethe certificate,whilethe authoritiesin the
man'shomevillagerefuseto give himaccessto his son's Note:Thefigureshows the relationship
betweenan indicatorof
assets untilthe certificateis produced. ruleof lawand per capitaGDPfor 166countriesin 1997-98.
Source:Kaufmann, 1999.
Kraay,and Zoido-Lobaton
Source:MichaelAnderson1999.
I04 WORLD DEVELOPMENT REPORT 2000/2001
tle knowledgeof their rights and may be deliberately
misinformed.Contemporarylegalsystemsarewrittenand
are conducted on the basis of written documents-
makingaccessinherentlydifficultfor poor people,who
usuallyhavelittle formal education.Language,ethnic,
caste,and genderbarriersand other exctusionaryprac-
ticesadd to theseproblems.
The intrinsiccomplexityof legalsystemsis exacerbated
in many developingcountriesby the superimpositionof
new lawsand constitutionalrightsover coloniallegisla-
tion and customarylaw. The resultingconfusionmakes
it difficultto know one'srights,introducesarbitrariness
in law enforcement,and enablesthe powerfulto choose
whichlegalsystemto apply.18 This reducespoor people's
confidencein the legalsystem.It alsogivesenormousdis-
cretion to authorities, often making connections and
bribescentralto negotiatingthe legalsystem.Makingthe
rulessimplerandclearerisespecially importantin theareas
of greatestconcernto poorpeople,suchaslabor disputes,
land titling, human rightsabuses,and policeviolence.
Althoughpoor peopleneedto accessthe legalsystem
for registrationand other administrativepurposes,they
use the judicial systemmuch less frequentlythan the
nonpoor.Courtsystemsin manydevelopingcountriesare
poorlyfunded and equipped,and mechanismsfor en-
forcingjudgmentsoftenweak.Theseaddto the otherprob-
lemspoor peoplefacein usingthe judicialsystem,such
as financialcosts.Waivingcourt feesforpeoplewith low
incomes could provide some relief.Ecuador and Peru
provideexemptionsforcourt feesin certaincases.To as-
sistpoor people,legalaid is providedin many develop-
ingcountries,but oftenmorein principlethan in practice.
To be effective,suchaid must be deliveredpromptly:in
TrinidadandTobagoit takesthe legalofficeabout three
months to processapplicationsforlegalassistance,in ef-
fect denyingaccessto thoseunableto wait that long.1 9
In addition to government-providedservices,legal
aid can be providedthrough alternativesources.Many
countriesrequirelawschoolgraduatesto providelegalaid
before becoming attorneys-others require practical
trainingof lawstudents.In Chileand Perulawyersmust
completeaspecifiedamountof practicaltrainingafterlaw
school,often in legalaid offices,therebyprovidingim-
portant resourcesforpoor people. 20
Streamliningthe operationof the judicialsystemto re-
ducecostsand delayswill addresssomeof the problems
poor peoplefacein the courts.21 Reformingcourt proce-
dures helps-simplifying rules (while respectingdue
process),shorteningproceedings,allowingpartiesto rep-
resentthemselves.Broaderreforms,suchaschangingthe
structureof the courts,alsohelpincreasepoorpeople'sac-
cessto justice.Smallclaimscourtsandother informalpro-
ceedingscan reducethe backlogandwidenaccess. 22 And
the teachingand practiceof lawcan be amendedto sen-
sitizethe legalprofessionto the needsof poor peopleand
to the use of the lawto furtherthe public interest.
Alternativedisputeresolutionmechanismshold con-
siderablepotentialforreducingthe delaysandcorruption
that characterize much disputesettlement.In ElSalvador
mediationprovidespartiesameansto settledisputeswith-
out a lawyerand withintwo months. 2 3 In SriLankathe
AsiaFoundationhas assistedthe Ministryof Justicesince
1990in establishinga nationalnetworkof communitv-
basedmediationboards.In 1998,100,000caseswerere-
ferredto the mediationboards,with two-thirdsof cases
resolvedto both parties'satisfaction.An independenteval-
uationfound that the boardsenjoyedan outstandingrep-
utation and successfullyprovided low-cost,accessible
justiceto poorpeoplein ruralareas. 24 In Bangladesh some
NGOs haveadoptedtheshalish(anindigenouspracticethat
usesoutsidepartiesto helpresolvedisputes)to aidwomen
andotherdisadvantaged groups,suchaslow-income farm-
erswithland-related disputes.A 1999studyin Dhakashows
that womenwhohavebeenthroughNGO-initiatedme-
diationexpressedsatisfactionwith the resultsbya four-to-
onemargin. 2 5 That the NGOs canbackup the mediation
processwith litigationis a factorin that success.
Thesealternativemechanismsmayprovidemore pre-
dictableoutcomesthan the formalsystem,becausecom-
munity mediatorsare typicallymore familiarwith the
detailsof casesthan are judges.2 6 The risk of such mech-
anismsis that theycan giveundue powerto conservative
forcesin a community(whichmight, forexample,be bi-
asedagainstgenderequity)and be subvertedto servethe
interestsof localelites.To minimizetheserisks,alternative
disputeresolutionmechanismsneedto be carefullyregu-
latedand supervisedby moreformallegalstructures.They
can alsobe introducedgradually-for example,through
pilotprogramssponsoredandsupervised by regularcourts.
Promoting legal service organizations
Civilsocietyorganizationssuch as legalserviceorgani-
zationsseek to help poor peoplegain accessto the ben-
efitsand protectionof the legalsysteminsideand outside
the court system(box6.3). Protectingindividualsagainst
unlawfuldiscriminationat work andevictionfrom their
 MAKING STATE INSTITUTIONS MORE RESPONSIVE TO POOR PEOPLE 105

homes, such organizationshelp peoplecollecttheir en- This goesfarbeyondthe conventionalideaof offeringfree

titlements,obtain basicservices,and get court ordersto
protectwomenfromdomesticviolence.Theycanalsopro-
tect communitiesfrom beingdispossessed.
Legalserviceorganizations
canhelppoorpeoplebytak-
ing legalactionon behalfof a group of plaintiffs.Often,
largenumbersof poor peoplesufferfromsimilarinjuries,
so seekingredressas a group providespoor people with oth-
erwise inaccessiblejudicial protection. Legaladvocacyor-
ganizations in Bangladesh helped avert the eviction of
urban slum dwellers.Evicted residentsbecame petitioners
in litigation, where the basic argument rested on funda-
mental constitutional guarantees:demolishing their home
deprives the poor of a livelihood, in violation of the con-
stitution.27 Public interest litigation can also benefit poor
people. In India it has improved the deliveryof some pub-
lic servicesand reduced environmental contamination.2 8
The most effectivelegalserviceorganizationswork out-
side the judicial system, protecting rights without re-
sorting to lawsuits-important, because the costs of
lawsuits can sometimes outweigh any resulting gains.
legalrepresentationto poor individualsand helpingpeo-
ple or communitiesasserttheirrightsthroughthe courts.
Moregenerally,the workof legalserviceorganizations
helpscreatea cultureof rightsthat changesthe waypeo-
ple think about themselvesrelativeto those who have
powerovertheirlives-spouses,landlords,employers, gov-
ernment agencies. This encourages poor people to avavil
themselves of the protection that the formal legal system
offers. These organizations also generate pressure for
changing the way the rules are applied by judges, bu-
reaucrats, and the police. Legal literacy and legal aid
have the maximum benefit if they help create a process
of self-empowerment and social empowerment that
moves citizens to activate their rights and to redefine
and reshape inequitable laws and practices.
Legalservice organizations help change the rules that
affect poor people, whether in constitutions, statutes,
regulations,municipal ordinances, or myriad other codes.
In Thailand the Women and the Constitution Netwo:rk
was very active in constitutional reform that led ito

Box 6.3
Legal service organizations help poor people gain access to the protections of the legal system

Almostanyform of legalassistance is of value.Butcomprehen- Cambodia

siveservicesfromindependent legalserviceorganizationsarees- Legalaid organizations in Cambodiaare strugglingto createa
peciallyvaluable
to poorpeople,andthedemandfor suchservices justicesystem-from almostnothing.TheCambodia Defenders
is high.Standardsshouldbedeveloped to makethemevenmore Project,establishedin 1994,focuseson criminaldefenseand
effective.Legalserviceorganizations alsoneedfinancialsupport communitylegaleducation. It collaborates with NGOsto provide
fromdonorsandcivilsociety,buttheyhaveto beallowedto func- servicesandrepresent womenincourt,especially in domesticvi-
tionautonomously, takingdirectionfrompoorpeoplethemselves. olencecases.Theorganization's lawyersruntrainingprograms,
commenton draftlaws,andworkwith civilsocietygroupsto ex-
Bangladesh plorelegaltoolsfor influencinggovernment. The LegalAid Soci-
TheAin-O-Salish Kendra(ASK),established in 1986,seeksto re- ety of Cambodia worksto increasepublicunderstanding of and
formthe lawthroughits representation of poorwomenandchil- respectfor the law,whileproviding free legalservicesincriminal
dren, organizedgroupsof workers,the ruralpoor, and slum andcivilcases.It is especially activein defending farmersbeing
dwellers.It provideslegalaidprimarilyonfamilymatters,includ- dislodgedfromtheirlandby powerfulbusinessinterests.
ingviolenceagainstwomen.Litigationon behalfof victimsis un-
dertaken in criminalcasesandwhenbasiclegalrightsareviolated. SouthAfrica
ASKinvestigates andmonitorsviolationsof lawandhumanrights, The LegalResourcesCenter,a nationalorganization foundedin
including policetorture,murder,rape,anddeathsingarmentfac- 1979to servepoorpeople,initiallyusedlegaladvocacy to exploit
tories.It alsomonitorspolicestationsto collectinformation onvi- contradictionsin the apartheidlegalsystem.Sincethe endof
olenceagainstwomenandchildrenandto trackcasesreported apartheid the centerhasusedlegaladvocacy to addresslandand
at the stations.The-work by ASKis significantbecause of thesub- housingissues.It successfully represented the Makulekecom-
stanceof whatit does-work onbasicissuesfordisenfranchised munityin its restitutionclaimto landin KrugerNationalPark.
people-andbecauseof the way it doesit-through mediation, Otheractivitieshaveincludedcasesto restorewaterserviceter-
discussion groups,legalawareness courtcases, minatedbecauseresidentsweretoo poorto payandto protect
training,individual
administrative andlegallobbying,grouprepresentation, andpub- the landrightsof anaboriginalcommunityin the privatization of
lic interestlitigation. a diamondmine.

Source:Manning1999.
io6 WORLD DEVELOPMENT REPORT 2000/2001
amendments recognizingequal rightsfor women.The net-
work followed up this successby launching a mass cam-
paign to educate Thai citizens-women and men-about
the new constitution and its implications.2 9
How can decentralization be made
pro-poor?
State institutions are often accused of being too remote
from the daily realities of poor people's lives, and de-
centralization is often recommended as a solution. De-
centralization can be powerful for achievingdevelopment
goals in ways that respond to the needs of local com-
munities, by assigning control rights to people who
have the information and incentives to make decisions
best suited to those needs, and who have the responsi-
bility for the political and economic consequences of
their decisions.3 0 It is not in itselfa goal of development,
but a means of improving public sector efficiency.And
there are important caveats.The most important is that
decentralization can bolster the power of elites in set-
tings with highly unequal power structures.3 1 To ben-
ctit poor people, it must have adequate support and
safeguards from the center and effective mechanisms of
participation.
Decentralization can mean different things. Here it
refers to the formal devolution of power to local deci-
sionmakers. Less extensive forms of decentralization in-
elude deconcentration(the central governmentposts
employees at the local level) and delegation (powers are
delegated to the local level).32 The size of decentralized
units of government can vary enormously: decentraliza-
tion to the state or province in Brazil, China, and India
merely breaks government into units the size of many
countries. Decentralization to smaller units increases the
scope for interaction with the citizenry served.
Decentralization can make state institutions more re-
sponsive to poor people, but only if it allows poor peo-
ple to hold public servants accountable and ensures their
participation in the development process. The pace and
design of decentralization affect its impact on efficiency,
accountability, participation, and ultimately poverty re-
duction. But only generalprinciples from successfulmod-
els can be transferred from one setting to another.3 3
Moving,programs closer to users
Localinformation has many advantages.It can help iden-
tifvmore cost-efficientways of building infrastructure,pro-
viding public services,and organizingtheir operation and
maintenance. A study in South Africa found that com-
munitv involvementreduced the cost of creating jobs and
improved the cost-effectivenessof transferring resourcesto
poor people (figure 6.4). Moreover, knowing what local
In In-
needsaremostpressingcanhelpthe disadvantaged.
donesiagreaterlocal control over funds led to more spend-
ing on health and education in prioritv areas for poor
people-and to more spending on small infrastructure,
boosting nonfarm emplovment and incomes.34
Local monitoring and supervision for many types of
projects and programs are more effective and less ex-
pensive because of proximity to the point of provision
and better interactions at the local level (box 6.4). In
Nicaragua students attending schools that were
"autonomous"-as measured by the share of decisions
on teacher staffing made by the school-achieved bet-
ter test scores than students in schools with limited or
no local autonomy.35
What is required to reach poor people?
Decentralization can greatlyenhance the state's capacity
to accelerate local development and reduce poverty, but
only if it is effectivelydesigned.Localauthoritiesand agen-
Figure6.4
Decentralizationlowers the cost of raisingthe
income of poor people in South Africa
Programcostoftransferring1 randto a poorperson,
1995-97
Rand
15
12
9
6
3
*
None Limited Partner Sole
implementer
Type of community involvement in program
Source:Adato and others 1999.
 MAKING STATE INSTITUTIONS MORE RESPONSIVE TO POOR PEOPLE 107

Box 6.4 While a certain degree of fiscaldevolution is needed for

Community monitoring can reduce effectivedecentralization,it carriesthe risk of exacerbating
environmental pollution inequalitiesbetween regions.37 In China, where provinces
and local bodiesare expectedto be self-financing,socialser-
Poorcommunities benefitdirectlyfrom regulations
thatre- vicesare greatlyunderfundedin poorerprovinces." 8 Mech-
ducepollution,butwhereenforcement isweak-as in many
developing countries-companies run littleriskof being anisms for redistribution from the central budget can
caught andpunished. Polluting
firmsthushavelittleincen- mitigatethese inequalities,but this is politicallycontentious.
tiveto cleanuptheiractivities,
andfirmsthatrespect legal The problem needs to be addressed through consensus
limitshavelessincentiveto cuttheirpollution. buildingand taxsharingsothat the centralgovernmenthas
A newapproach combines publicinformation
disclosure buidga to sarn tralsgovernmentshay
with market-basedincentivesto encouragefactoryman- resourcesto maketransferswherenecessary.
agersto improvetheirenvironmental
performance.In some Sitpportand safeguardsfrom the center. Central support
countries
localcommunity
representatives
negotiate
withgov- is required to ensure that national policies are adhered to
ernmentregulators
andfactorymanagers
to agreeonac- and to coordinate the interregional interests of different-
ceptable
pollution
levelsandsetpollution
charges
accordingly. administrative units-as with highway charges and ac-
Elsewhere,
publicinformationenables consumers,
bankers,
andstockholders
to evaluatea company's environmental cess to common water resources. Common macroeco-
record
beforedeciding
whether tobuya product,
lendmoney, nomic and redistributivegoals also need to be supported.
ortradethecompany'sshares. The danger of decentralization without safeguardsis il--
Becausepoorpeoplearelessabletoprotect
themselves lustrated by the situation of Brazil in January 1999,
fromindustrialpollution,theircommunities
particularly
value
public
information
about
whichcompanies
polluteandhow when one state's action threatened the macroeconomic
their dischargesaffect health.Where governmentshave stability of the entire country.39
providedlocalcommunitieswithreliable
pollution
data,poor Support for training is alsorequired.Studiesofsuccessful
peoplelivinginthevicinityof industrial
polluters
havene- decentralization indicate the importance of creating ad--
40
gotiatedbetterarrangements
forcompensation
andcleanup. Many local governments
Theresultsso fararepromising.
In Indonesia
thegov- ministrativecapacity.lack the
ernment hascutindustrial
waterpollutionsharply
bymoni- administrative capacity for large-scale decentralizatiori
toringfactorydischarges
andbringingpublicpressure
on and need training in accounting, public administration,
factories
bypublicizing
theiremissions
data. financialmanagement, public communications, and com--
Source:World Bank 1999k. munity relations. If subnational governments have strong
administrative capacity and accountability mechanisms,
decentralization can reduce the scope for corruption. IF
they do not, it can increasecorruption and reduce access
41
cies need considerable autonomy, including on fiscal to basicsocialservices, as in Central Asia,the South Cau--
42
matters, as well as considerable support and safeguards casus, and the Baltics.
from the center. Moreover, decentralized government Safeguards are also needed to monitor financial pro--
needs mechanisms to ensure high levelsof participation bity and discourage the capture of local bodies by pow--
in the design and monitoring of programs and policies erful elites. One of the most serious pitfalls o:F
by all sections of the population to be served. decentralization occurs when power imbalances are
Autonomy andfiscal decentralization.Local authorities large at the local level. In such a situation higher lev--
need to have enough fiscal control to plan their activities. els of government, less subject to local political pres-
But locallyraisedrevenuesare often onlya small part of the sures, may be more motivated than local bosses to help
budget of decentralizedunits, weakening ownership of lo- the disadvantaged. For example, the U.S. federal gov--
cally designedpoliciesand threatening their sustainability. ernment has a long history of doing more to proteci:
While decentralized units need an adequate budget base, minority civil rights than state governments, which
enforcing hard budget constraints is alsoessential,to make have greater representation of those interested in sub--
them accountable. If ad hoc funds from outside units are verting those rights.43 Studies of Argentina indicate that-
availableto meet budget shortfalls,localbodiescan losetheir subnational governments can sometimes be less effec--
incentive to function efficiently.Moreover, suchifunding tive than central governments at targeting poor areas.44
erodes the real power of the localbody and its ability to ef- Similar problems are noted elsewhere in Latin Amer-
fect change,as attention goesto extractingthese benefits.36 ica and South Asia.4 5
108 WORLD DEVELOPMENT REPORT 2000/2001
Participation. Widespread popular participation is
vital to successful decentralization-without it, the po-
tential benefits of local information cannot be realized.
Moreover, participation creates a virtuous circle. Partic-
ipating in local government helps build civil society and
ensure that majority needs are heard and goals are
achieved. It also helps increase the voice of poor people
in local affairs.
One direct way of ensuring participation is to hold
regular elections for local government. Electoral rules
can also foster broad participation by reserving seats for
marginal groups. In India a third of panchayat (local
council) presidents, vice presidents, and elected mem-
bers must be women. In addition, certain other disad-
vantaged groups must be allotted memberships and
executive positions proportional to their number in the
area. Such measures can transform power relations over
time.' 6
Participation on a more frequent basisthan just at elec-
tion time also needs to be fostered.In Bolivia,Brazil,and
the Philippines decentralization laws require local gov-
ernments to incorporate or formally associate grassroots
organizations with their deliberative procedures and to
goivesuch organizations a role in administering services
and projects.4 7 Successfulparticipatory budgetmaking in
Porto Alegre, Brazil, shows that having local communi-
ties decide on the use of municipal resources can be very
effective for local development.4 8 Good information
channels between governments and communities are
also necessary for good results. In Chile, where calls for
community fund proposalsare publiclybroadcastand for-
mats for project presentation are distributed through
municipalities, a survey of beneficiariesfound that fund
disbursementswere biased toward neighborhoods and so-
cial organizations well connected with municipal and
regionalgovernments.Those with weaker connectionsre-
ceived fewer funds.49
Decentralizing powers and resources to the
submunicipal level-such as neighborhoods or villages-
requires special effort, bur the benefits can be consider-
able.50 In South Africapartnershipsbetween communities
and local governments sharply increasethe probability of
long-term returns to the community.5 1In Guinea a pilot
project showed that communities are adept at designing
and managing such projects as building and maintain-
ing new infrastructure. Communities mobilized local
resources,used grant funds equitably and efficiently,and
targeted funds to help vulnerablewomen and children.5 2
The politics of poverty reduction:
pro-poor coalitions
If we aren't organized and we don't unite, we can't ask
for anything.
Poorwomnani,
FlorencioVarela,Argentina
Pro-poor coalitionsthat link the interests of the poor and
the nonpoor are important for poverty reduction. Im-
proving the capacity of poor people to participate pro-
ductivelyin economicactivityalsohelpslay the foundation
for fastergrowth. The state can support the growth of pro-
poor coalitions by:
* Fostering a political climate favorableto pro-poor ac-
tions and coalitions.
* Removing legal barriers to pro-poor associationsand
offering them technical and other support to scaleup
their activities.
* Fosteringstate-communitysynergiesand increasingthe
capacity of poor people to participate in development
and local governance.
Such transformations are essentiallypolitical and have to
be effected through political processesinvolving changes
in political configurations and power balances.5
Creating political supportfor pro-poor
actions and coalitions
The interestsof the poor and the nonpoor are intertwined
in many ways, making it beneficial for the nonpoor to
take an interest in redistributive measures and pro-poor
actions. This interest can be motivated by a recognition
that efforts to reduce poverrycan promote social and eco-
nomic development for the whole nation, thereby also
raising the living standards of the nonpoor. The indus-
trializing economies of East Asia, where the creation of
a skilled, healthy workforce was crucial to success, show
that investing in mass education and human capital for-
mation provides a significant boost to national eco-
nomic growth.
Control of communicable diseasesis another case in
which allcitizensbenefitfrom programsfocusedon the poor,
as it is almost impossiblefor any group to avoid these dis-
eases unless the sourcesof contagion are eradicated. Poor
people, undernourished and living in environments with
greater exposureto disease,are especiallyvulnerableto in-
fection.They are alsolesslikelyto receiveadequatepreventive
and curative health services.Thus poor people tend to
form pockets of contagion from which diseasescan spread
 MAKING STATE INSTITUTIONS
to other groups. This was one of the main driving forces
behind the public health movements in the West at the turn
of the 20th century (box 6.5). The spread of disease is in-
tensified today by the vastly increased volume of travel: if
health services are of poor quality or unaffordable for poor
people in one country, drug-resistant strains of malaria
and tuberculosis can spread around the globe. So both na-
tional and global efforts are needed to address some of the
health problems of poor people (chapter 10).
Another motivation for the nonpoor to support pro-poor
action is the specter of mass migration to urban areas, with
attendant problems of growing slums and rising demands
on already overburdened urban services. China and India
have reduced incentives for urban migration by providing
plying schools, health services, electricity, and other basic
amenities, but also ensuring that employment creation is
geographically dispersed and that transport networks allow
people to commute to work from their villages.
people to commute to work from their villages.
To build political support for publicaction against poverty,
governments have to enhance the perception of common in-
terests between the poor and the nonpoor."4 Key to this is
systematiayntroucingthe public debate the notion
that poverty reduction is a public good and can further the
well-being of the nonpoor. How these issues are framed in
the public debate can greatly influence the outcome. Povertys
character, causes,and solutions are malleable concepts, which
can be reinterpreted and represented m a vanretyof ways, many
of them conducive to public action against poverty. In the
early20th century state governments in the United Stateswere
persuaded, mainly by national middle-class women's orga-
nizations, to spend public money to support poor famlies-
on the grounds that this was the only way to protect the moral
and physical integrity of the nation.5 5
Understanding the benefits of helping the less for-
tunatecantusbeapowrful
stimulus for pu c astrong
.uate canthus beapowrfustmulmental
tion. Without such understanding, the living conditions
of the disadvantaged are sometimes used to justify their
further exclusion. Latin American elites have some-
times viewed poor people as a danger to public well-
being. This mindset makes it more difficult to eradicate
poverty and mitigate its negative impact on the econ-
omy and society.
Facilitating the growth ofpoor people's
associations
The state'smost important task in fostering poor people's
organizations is to remove legal and other barriers to
MORE RESPONSIVE TO POOR PEOPLE 109
Box 6.5
National coalitions against communicable
diseases in the West
Sanitary neglectis mistaken parsimony: thephysicalstrength
of a nationis amongthe chieffactorsof national prosperity.
-John Simon(1858)as citedinRosen(1993)
The publichealthmovementinEuropeandthe UnitedStates
broughtrapidimprovementsinthe healthconditionsof poor
and rich alikein the late 19th and early20th centuries,long
beforethe discoveryof antibiotics.The politicsof publicre-
sponsibilityfor reducingcommunicablediseases were mo-
tivated by a blend of economic, political,and humanitarian
interests. Industrialistswere concernedwith reducingthe
drain on labor force productivity.States were concerned
with havingenough fit young men to serve inthe army and
expandspheresof influence.Elitesfeltthat theirenvironment
was detrimentallyaffected bythe illhealthof poorpeople-
and that the dangers for the populationas a whole needed
to be reduced. Intellectualspointedout the connectionsbe-
tween illhealthand poverty,demandingradicalchange as a
solutionto the problemsof endemic andepidemicdiseases.
To reduce everyone's exposure to communicabledis-
eases, strenuouseffortshadto be made to improvethe health
of poorpeople.Measuresincludedcontroloffood and drugs,
smallpoxvaccinations,and quarantine. Central to the en-
deavor was securing a pure water supply,effective waste
disposal,cleanstreets, and reduced pollution.Housingreg-
ulationswere enforced to ensure adequate ventilation,toi-
let facilities,drainage,and sewerage in homes. Restrictions
on privatebehaviorincludedforbiddingspittingand urinat-
ing in public spaces and banning livestockfrom domestic
premises. Massive health education campaigns used ex-
tensive outreachto change personal healthbehaviorand in-
crease people's understandingof how to avoidillhealthand
care for the sick.These state interventions,combinedwith
risinglivingstandards,dramaticallyimprovedhealthand life
expectancybetween 1880 and 1920.
Paradoxically, improvementsincurativetechnologiesin
recent decades may have led to less vigilanceagainst com-
municablediseases in some developingcountries. These
powerfulcurativetechnologiesneed to be combinedwith
publichealthpoliciesaimedat improving environ-
sanitationand encouraging healthylifestyles.This
willhelp increaseeconomicgrowthand reduce poverty,
avertingnegativeconsequences fornational
andglobalhealth
as drug-resistant
strainsof diseasesmultiply.
Source:
Rosen1993;Preston
andHaines 1991;Schofield,
Reher,
and
Bideau1991;Caldwell
andothers1990.
forming associations and to provide an administrative
and judicial framework supportive of such associations.56
Without this, it is very difficult for poor people's associ-
ations to flourish and to influence public policy. Poor
110 WORI. ) DEVELOPMENT REPORT 2000/2001

people face enormous constraints in forming associations agencies.59 To counter these problems, some peasant or-
to increasetheir voiceand improvetheir circumstances. ganizationsin Boliviaand Ecuadorhaveworkedthrough
Thev usuallvtakelittlepart in politicsbecauseparticipa- NGOsto linkup with nationalagriculturalagencies, enor-
tion seemsirrelevantto theirprimaryconcerns,futile,or mouslyexpandingtheir rangeand effectiveness.60
both. They often have low expectations of their In mostdevelopingcountriesNGOs are centralactors
61
government-and mayevenfear reprisalsfrom state or in antipovertypoliciesand programs. The socialand ed-
localauthoritiesif theyorganize.Evenformatterswhere ucationalbackgroundof manyNGO staff enablesthem
the governmentis viewedas relevant,poor peopleseein- to interacteasilywith the staffofnationalinstitutions,and
dividualand collectiveeffortsto exertinfluenceas hav- theycanhelpcreatebridgesbetweentheseinstitutions, out-
ing little effect.When poor peopledo participate,their sideagencies,andgrassroots organizations.
NGOs canalso
classidentity is not the only influenceon their decision beveryeffectivein deliveringtechnicalassistanceto poor
to do so. As with other citizens,the forcesthat move people,62 as Mopawihas beenin Honduras(box6.6).
them to actionare often tangible,short term, and local. In Boliviaa Dutch NGO helpeda campesino federa-
Reducingasymmetriesin informationcan do much tionlinkwith researchinstitutionsinvolvedin the national
to changepoor people'shesitanceto participate-and to
empowerthem. Formaleducationenablespeopleto gain Box 6.6
accessto bettereconomicopportunities(chapter5) and NGOscanhelpmobilizeandempower
givesthem the means to articulatetheir needs and de- communities
mandsin publicforumsand in politicalprocesses.Allthis
is enhanicedby widespreaddisseminationof informa- Since1985theNGOMopawi(Moskitia
Pawisa,
or Devel-
*on. TodayIs information technology
II and lowrInor- opment of LaMosquitia)has been workingalongsidethe in-
digenouscommunities of La Mosquitia,a remotearea of
mation costs, combined with rising demand for greater westernHondurasandoneofthelastremaining areasoftrop-
access to public documents, can have powerful benefits icalforestinCentralAmerica.Overthe yearsMopawihas
for the poor. developed a large and complex development program. It
Major impediments to organization by poor people are has worked to change government policy for the region
throughcontinuouslobbyingand advocacy,helpingto form
lackof rimc,resources,information,andaccessto outside linksbetweengovernment,
international
NGOs,
research
sourcesof help. Added to that are phvsical constraints on organizations,
and indigenous organizations
to raiseaware-
collaboration, such as geographic dispersion and poor ness and inspireaction.Ithas workedwithlocalcommuni-
traiisport
con*tlLinications
aiid infrastrL CtLIre.EhIc ad Ities to find ways of improvinglivelihoodswithout harming
transport
communicationsinfrastructu.
and Ethnic the environment. Mopawi has also addressed deforestation
other socialdivisionsare another impediment (chapter7).9 inLaMosquitia,combining advocacywithpracticalpreven-
Despitethesedifficulties,manycountrieshaveseenan ex- tion.Mostof its staffarefromLaMosquitia, whichhas
plosionof community-based,participatorygrassrootsor- provedto bea major strength.
ganizationsin the past few decades.ThroughoutLatin Theorganization
hashelped
improve
people's
livelihoods
by identifyingalternative models of resource use and in-
Americapopularand indigenous organizations,
some- volvinglocalcommunities in decisionmakingand manage-
times based on traditional forms of association, now give ment.Forexample,smallbusinesses havebeenset up,and
voice to the underprivileged and deal with immediate experiments conductedwithagriculture
andagroforestry.
An
needs in health, schooling, and public infrastructure. agroforestry andpastureprojectbeingconductedwithset-
tlers and indigenouscommunitiesincludesexperimentswith
organzatonsrequiremanyformsof
Suchgrassroots sustainablemanagementof the forestand restorationof
support from the state and from civil society.They often degradedareas.Andeffortstargetingwomenareaimedat
need technicalassistanceand skill building to become sus- cultivating vegetableplotsto improvehealthand nutrition.
tainableand effective.Thev alsoneedhelp in scalingup Tostrengthen
localorganizations,
Mopawi
hasworked
ing up with Masta (MosquitiaAlsaTanka),the federation of repre-
their membership, range of functions, and political en- sentative indigenous organizationsin La Mosquitia.With
gagement. Many grassroots initiativesare limited in havetakenonlegalization
Mopawi'shelp,localorganizations
scopeand depth and never reachthe national political oflandownershipanduserightsanddevelopedtheircapacity
arena.Studiesin LatinAmericahavefound that someor- foradvocacy.
ganizations are effectivein addressingsomeof the im-
t~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Source:Brehm2000.
mediateconcernsof poor people,but their sustainability
is constrainedby difficultiesin linkingup with external
 MAKING STATE INSTIT'UTIONS
potato programbycontractingwith an internationalex-
pert to work with the federationat the onset of the pro-
ject. The consultant had no difficulty establishing
high-levelcontactswith researchinstitutions, and on
the basis of these meetingsthe federation established
strong links with the national potato program.The re-
sult washighercrop incomesfor federationmembers. 63
Sometimes NGOs reflect the political system in
whichtheythrive,or localinterestgroups,and thus may
not serve the interests of poor people as well as they
might.6 4 NGOs are no panacea-it is important that
they be accountablefor their actions,especiallyto the
poor groups that they seek to represent.
Fosteringstate-community synergies
in growth andpoverty reduction
The state canfacilitateinteractionsbetweenlocaladmin-
istrationsandcommunitiesto engenderdevelopmentand
reducepoverty.65 There are two main aspectsto thisrole:
reducing obstaclesto collectiveaction in communities
and encouraging greater collaboration between commu-
nities and local governments. To forge ties within com-
munities and facilitatelocal collectiveaction, the state can
initiate programs that build up the assets of poor people
and make public servicesmore accessible.Such programs
reduce the perception among poor people that their sur-
vival depends on avoiding risksand keeping their patrons
happy-releasing their energiesto pursue actions for up-
ward mobility and to collaborate with others on a more
equal footing.
The combination of a more egalitarian social organi-
zation at the community level and better local adminis-
tration enablesthe creation of powerfulcoalitionsfor rapid
development. Strong links between local administrations
and communities improve servicedeliveryand reduce the
potential for local capture of development programs.
This arrangement has been used successfullyin very dif-
ferent political and administrative settings: Brazil in the
1980s, the Republic of Korea in the 1960s and 1970s,
and Taiwan, China, in the 1950s.66
The example of Brazilshows that institutional change
is considerably more difficult in highly unequal settings.
Lacking the prior extensive land reform of East Asian
countries, the state in its efforts to reform local govern-
ment had to tackle problems of landlord interests and po-
litical connections with local government. This created
problems, because large landowners, private contractors,
and relief suppliers were accustomed to cornering re-
MORE RESPONSIVE TO POOR PEOPLE III
sources.In the 1987droughtthe stateused agricultural
extensionworkersto break the grip of patronagein the
distributionof drought relief.But sustainingsuch suc-
cessesrequiresmuch continuingeffort.
The Brazilianexperiencealsoshowsthat manyof these
obstaclescan be overcomeby bringinggrassroots electoral
pressureto bearon localgovernments.Politicalinterference
has beenkept at bayby the stategovernments'insistence
that municipalcouncilsfordisbursingdevelopmentfunds
haveat least80 percentrepresentation
fromend-usercom-
munities.Moreover, ifcommunitiesfeeltheyaretreatedun-
fairlyby the municipalcouncils,theycan applyfor funds
directlyfrom the stategovernment.67
The state can undertake severalkey actionsto foster
developmentalsynergiesbetweencommunitiesand local
governments (figure 6.5):
* Generatingcommunitydemandfor betterpublicad-
ministrationand servicedelivery,
throughintensivedis-
seminationof information.
* Formingdensenetworksbetweenthe stateand com-
munities and making availableto communities the in--
formation and technical, marketing, credit, and othe.r
support they need to implement programs.
* Changing localagencies'mode of operation by putting
pressure on them from aboveand below.In Brazil the
state used official job recognition to motivate staff.
a Motivating grassroots workers and leaders through
positive and negative sanctions, including respect:
from peers. Where the workers are also community
members, as in the Republic of Korea, the potential
sanctions are especially strong.
* Adjusting the roles of higher levels of government,
training and motivating their staff to focus on man-
aging overallstrategy,and providing technical support..
regulations, and facilitation.
These initiatives yield substantial political payoffsfor
the governments in legitimacy and in popular support.
In a municipal election in Brazil candidates said that if
they wanted to get elected, they had to support the new
arrangementsfor increasingpublic accountability of local
government and improving public service delivery.6 8
This helped to strengthen potentially weak governments
and motivate them to engage in these difficult tasks. At
the same time the conditions for a plural polity were
strengthened.
Collaboration between communities and local gov-
ernments can also promote many different forms of de-
velopment. In addition to local improvements in
112 WORLD DEVELOPMENT REPORT 2000/2001

Figure 6.5
State-community coalitionscan foster rapid development and better servicedelivery

Centralor state government a Ordevelrpliting

(incentivesfor action: a Politicalandeconomic

a Raiseaspirations,generate and grassroots

workers
demand (incentives
for action:
aOfferinformation
Onservices public
image,materialbenefits)

communities
(incentives
foraction: Pressure
forqualityservices
development and andpeermonitoring
empowerment)
political

Source:DasGupta,Grandvoinnet,
andRomaniforthcoming

infrastructure and living conditions, such collaborative Figure 6.6

efforts have delivered health and drought relief in Brazil Good political and administrativeinstitutions
and supported industrial production for export markets go hand in hand with economicgrowth
in Taiwan, China.
Changes are incremental and can often take time, Economicgrowth
index
but as successstories accumulate in a given setting, they 0.03
createa demonstration effectfor others.The examplessug- 0.02 *
gest that it is possible, in the space of decades, to reengi- .*
neer state institutions to quicken the pace of development, 0.01 * 0
growth, and poverty reduction. They alsoshow that with * *
creativepolitical thinking these changes are possibleeven 0 .- ---_--- -
in relatively weak institutional settings. S jS*

Political regimes and poverty -0.02

Voicelessnessand powerlessness are key dimensions of -0.03 *

poverty, and an important aspect of voice relates to po-04
litical rights and civil liberties.69 Democracy is intrinsi- -2.5 -2.0 -1.5 -1.0 -0.5 0 0.5 1.0 1.5 2.0
cally valuable for human well-being as a manifestation of Qualityof institutionsindex
human freedom. Political freedoms have enormous
70
im-
indexisestimatedon the basis of
Note: Thequalityof institutions
pact on the livcs and capabilities of citizens . rights,the qualityof
measuresof civillibertiesand political
Participatory political processes can also help build a governmentinstitutions,the ruleof law,andpublic spendingon
good institutional base for the polity, society,and econ- socialinsurance.
Source:Rodrik1999a.
omy, enabling all voices to be heard and to interact in de-
termining outcomes (figure 6.6).71 Civil and political
 MAKING STATE INSTITUTIONS
liberties,alongwith competitiveelections,are powerful
instrumentsfor holding governmentsaccountablefor
theiractions.To translatethispotentialinto reality,many
institutionsneedto be in placeto ensurethat democra-
tic processesfunction as they should-among them,
independentmediato monitorelectoraland administra-
tiveprocesses,an independentjudiciaryto upholdthecon-
stitution and rule of law, and strong parliamentary
institutionswith the capacityto monitor the executive
throughsuchmechanismsas publicaccountscommittees.
Buildingthese institutionstakestime, and constantvig-
ilance is required to ensure that democraticprocesses
functionastheyshould.Butit is worththe effort,forthese
processesofferthe most effectivemeansof guaranteeing
voiceand participation.
Promoting democratic politics to foster stable
environments for growth
Evidenceon the relationshipbetweenthe type of politi-
cal regimeand the rate of economicgrowthis mixed. 72
73
In part thisreflectsmeasurementproblems, and in part,
the experiencesof growth with significantpoverty re-
duction in a few notablydevelopment-orientedcoun-
tries,suchas the Republicof Korea,beforetheybecame
pluralistdemocracies. Thesecountriesdevelopedthe pre-
conditionsfor a developmentalstate-notably, political
elitescommittedto developmentandsupportedbyan ef-
ficientpublicadministrationthat wasinsulatedfrompo-
litical pressuresand had close links to the business
community. 74
Anothermajorfactorin the successoftheseeconomies
wastheir earlyemphasison equity-especiallyextensive
landreformanduniversaleducation,whichlaidthe foun-
dationforrapidandequitablegrowth. 75 Theseeffortswere
facilitatedin somecasesby the devastationcausedbywar
and the attendantdisempowerment of entrenchedelites.
Thesecircumstancesloweredthe resistanceof politically
powerfulvestedintereststo drasticland reform.
In most nondemocraticsettings,however,lackof in-
stitutionalizedaccountabilityhas resultedin poor per-
formance in growth and poverty reduction. Even
successfuldevelopmentalstatespoint to an important
lesson: undemocratic regimesface serious abuses of
state power,and they are prone to rapid policy rever-
sals that can make their development gains fragile.
These states are moving to resolvesomeof these prob-
lemsby changingtheir politicalinstitutionsto increase
officialaccountability.
MORE RESPONSIVE TO POOR PEOPLE 113
The checksand balancesof participatorydemocratic
regimes-and the proceduresfor consensusbuilding-
limitthe scopefor rent seekingand drasticpolicyrever-
sals,offeringa much more reliableand sustainablepath
to development. 76 Participatorypoliticalregimesareas-
sociatedwith more stablegrowth7 7 -very importantfor
povertyreduction,giventhe highly adverseeffectsthat
shockshaveon poor people(chapters8 and 9).There are
severalreasonsfor this association.
First,participatorypoliticalprocessesencouragethe
use of voicerather than violenceto negotiateconflict.
Combined with guaranteed political rights, these
processesreduce the potential for ethnic and other in-
tergroup conflict,avertingmajor sourcesof socialand
economicvulnerabilityfor poor people. 7 8 For example,
India'sstrong democraticpoliticalinstitutionshelp me-
diate the potentiallyconflictingdemandsof its highly
heterogeneouspopulation.
Second, political and civil rights and a free press
allowpeople to draw attention to their needsand de-
mand appropriate public action.79 This is especially
important for avertingor respondingquicklyto major
disasters.And third, democraticelectionsconferlegit-
imacyon governments,fosteringsociopoliticaland eco-
nomic stability.
How can democracy deliver more effectively
for poor people?
Democracy-both representativeand participatory-
is a good in itself.But democraticpolitical processes
aloneare not enough to ensurethat poverty reduction
is taken asa keypriorityin society'sefforts.Politicaland
social ideologiesshape the extent to which democra-
tic systemsactuallyreduce poverty.Different philoso-
phies underlying welfarepoliciesin OECD countries
produceverydifferentoutcomesin povertyreduction-
despite the fact that all these countries have a long
historyof democraticpoliticalinstitutionsand high per
capita income (box 6.7).
Representativepoliticsallowall citizens'intereststo
be expressed,but the outcomesdepend on how differ-
ent groupinterestsplayout.80 Groupsthat are politically
connectedor better educatedhavea natural advantage
overothersin influencingpublic policy.This is reflected
in the United Statesin the largediscrepanciesbetween
affluentandpoorcommunitiesin fundingappropriations
for lawenforcementand public schools.In developing
countries,wherethe distributionof educationand po-
114 WORLD DEVELOPMENT REPORT 2000/2001

Box 6.7 litical know-how is far more skewed than in the United
Politics and poverty in OECDcountries States, large segments of the population remain under-
served.81 In C6te d'Ivoire 35 percent of public educa-
Povertyis not restricted to developing
countries.Thereare tnon spending goes to the richest 20 percent of the
significantpocketsof povertyin someOECDcountries.Al- population (see table 5.1 in chapter 5), and 55 percent
82
thoughthesecountries areallaffluentmarketeconomies with m i
democraticsystems,ideologicaldifferencesand corre- of tertiarystudents comefrom this group. Democra
spondingdifferences in popularsupportfor povertyreduc- tic politics are also subject to manipulation by political
tionprograms resultinverydifferentlevelsof povertyfortheir leaders. Leaders may favor spending resources on im-
citizens. mediate consumption rather than investment, on pop-
A comparative studyof povertytrendsinGermany, the ulist rather than productive measures..3 In addition,
Netherlands, andthe UnitedStatesexaminedtheincidence
of povertyandhow it wasaffectedovera 10-yearperiodby interest groupscan be bought off or co-opted with fa-
government programs. Thestudyfoundwidedifferencesin vors from politicians. 8 '
povertyincidence amongthe threecountries, differences
that There are three main wavsto strengthen the institu-
were widenedby governmentprograms.Levelsof "pre-
government"poverty(basedon earnedandunearnedper-
sonalincomeexcluding taxesandtransfersfromgovernment) them more effective at reducing poverty. First, democ-
vary,largelya resultof markeddifferences in laborlawsand ratic processes must permeate all major levelsof deci-
othermarketfactors.Pensionpaymentsreducepovertyin sionmaking. Some regimes are more democratic in
allthreecountries,but the designandimpactof otherpub-
lic transfersandtaxesaimedat reducingpovertydiffer. principle than in practice. Others, like India, are gen-
Especiallystrikingarethe low levels of povertyin the umnely democratic at most levels but have historically
Netherlands,a resultof universalbenefits.Althoughthe found it difficult to ensure that political accountability
transfershavelargetargetingerrors-they goto the non- reaches all levels of decisionmaking, particularly for the
pooras well as the poor-they do not appearto havere- poor. Indias ongoing panchavati raj drive toward de-
sultedinslowereconomic growthcompared with theother
countries.I countries. ~~~~~~~~centralization
and communityempowermentis an ef-
fort to correct this by increasing the powers of local
Governmentprogramswiden differencesin elected councils.
poverty among OECD countries Second, citizens must be given svstematic access to
Share of population under age 60 living in poverty, information so that thev can hold their civilservants and
1985-94 politicians accountable. If information on budgets and
Percent
15 on the use of funds-from the federal to the local
level-is made available in newspapers and other in-
formation sources, people can hold their leaders ac-
countable for results. Such public accountability can help
10 ~ to reduce inefficiency and corruption. This dissemina-
tion of information needs to be legally mandated to en-
sure that it does not stop with a change of government.
Germany Progress in information technology and increasing ex-
posure to global currents help to create a new environ-
ment of public awareness that reinforces democratic
politics.
\Netherlands Third, strong civil societv organizations can promote
0 1 2 3 the political empowerment of poor people, pressuring the
Governmentpolicysteps state to better serve their interests and increasing the ef-
Note:Policysteps:0 = "pregovernment"
poverty
rate, fectiveness of antipovertv programs. 8 5 Case studies in
1 = socialinsurance
pensions,2 = 1 + otherpublictransfers, the Indian state of Kerala and elsewhere show that a
3 = 2 + taxes.Povertyisdefinedas havinglessthan halfthe highlv engagedcivilsocietvcontributes to better outcomes
median disposable household income of the country. Incomes . '6
(netof inflation)were cumulatedfor the 10-yearperiod. in health and education. What is needed is an enabling
Source:Goodinand others1999. institutional environment for civil society to develop
and thicken (box 6.8).8"
 MAKING STATE INSTITUTIONS MORE RESPONSIVE TO POOR PEOPLE II5

Box 6.8
The evolution of civil society and state reform in Mexico

Traditionally,Mexicohashadwell-institutionalized systemsforchan- mandedthatthe basicsocialcontractbetweenstateandsociety

nelingandcontrolling politicalactivities-andusingstateresources be renegotiated,politicalleadersand partiesattemptedto re-
to cementpoliticalsupportfor the regime.Thesesystemswere spondin waysthat wouldgivethem the leverage to determine
more concernedwith controllingsociety's demandsthan re- the scopeandnatureof that contract.Thisconflictremainedun-
sponding to them.Thestatehaddeveloped a highlyeffectiveand resolvedinthe early1990s,withthe potentialto developa more
sophisticated machineryfor co-optingandmanagingdemands openpoliticalsystemdependingonthecapacityof civilsocietyto
anddissent.Althoughthe capacityof civilsocietyto demandre- forcechange.
sponsiveness was limited,the statewassensitiveto the ongo- Inrecentyearsprospects of realchangehaveemerged in Mex-
ingneedto cementloyalties,garnersupport,andresolveconflict. ico.A muchmoreopenanddemocratic politicalprocesshasde-
In the 1980stheserelativelystrongpoliticalcapacitiescame veloped,andanindependent electoralcommission andcivilsociety
undersiege:responsiveness, representation, and participation organizations widelyencouraged citizensto voteaccording to their
becameissuesof greatcontention. Day-to-daymanagement of po- conscience, freeof coercionandinducements, intheJuly2000elec-
liticalandeconomicconflictbecamean increasingly difficulttask tion.Whilemuchremainsto bedoneto openthe doorfurtherto
for publicofficials.Financial resources reached historically
low lev- civilsocietyparticipation
andtheexpression of citizendemands, this
els,andgovernment legitimacyplummeted. Whilecivilsocietyde- is a cautious,lurching,butin the endirreversible, first step.

Source:Grindle
1996.

** * poor people are often voiceless-and their interests

Respect for the rule of law, an efficient public ad-
ministration, and high-quality political systems facil-
itate the emergence of state institutions inclusiveof poor
people. But the impact of these factors on poverty de-
pends on how effectively they are translated into em-
powerment at the community level. Even in states with
extensive political and civil liberties and with govern-
ments that are neither captured by elites nor corrupt,
figure little in public policy. Poor people need direct
voice in the interventions that affect their daily lives,
as well as the ability to organize and vote. Actions are
needed to bring down barriers-legal, political, ad-
ministrative, social-that work against particular groups
and to build up the assets of poor people to prevent their
exclusion from the market. Some major social barriers
to poverty reduction are discussed in the following
chapter.
 N~~~~~~~~~~
Removing Social Barriers
and Building . Social
0
SJ ocialinstitutions-kinshipsystems,community
organizations, and informal networks-greatly affect
poverty outcomes. They do so by affecting the pro-
ductivity of economic assets,the strategies for coping
with risk, the capacityto pursue new opportunities, and
the extent to which particular voices are heard when
important decisions are made. Social institutions can
help poor people get by and get ahead.' But they can
also place barriers between poor people or the socially
disadvantagedand the opportunityand resourcesthey
need to advance their interests. Discrimination on the
basis of gender, ethnicity, race, religion, or social sta-
tus can lead to socialexclusionand lock people in long-
term poverty traps.
Values,norms, and social institutions may reinforce
persistent inequalities between groups in society-as
with gender-based prejudice throughout much of the
world, the caste system in India, and race relations in
South Africa and the United States.2 In the extreme,
these socialdivisionscan become the basisof severede-
privation and conflict.Legaland other measuresto over-
CHAPTER 7
Institutions
comethese inequalitiesmust be accompaniedby ef-
forts to raise awareness about culturally based atti--
tudes such as those toward women and people of
different races, religions, or ethnic origin. Otherwise
these measures will be unable to produce real change.
Social barriers can take many forms. Here the focus is
on key barriers arising from gender inequality, social
stratification, and social fragmentation.3
Gender discrimination and poverty
Until we became organized as a SEWA cooperative, the
middlemen could cheat us. But now I can negotiate with
them as the representative of our cooperative and as an
elected member of our local council. One day near the bus
stop, I heard a couple of men saying, "There'sthe woman
who is giving us all this trouble. Shall we beat her up?" I
told them, "Go ahead and just try it. I have 40,000
women behind me. "
-Woman laborer,speakingat World Summitfor
Social Development and Beyond, Geneva,June 2000
117
II8 WORLD DEVELOPMENT REPORT 20002/001
The extentand manifestationsof genderinequalityvary
amongsocieties,shapedto a considerabledegreeby kin-
ship rules.4 Rulesof inheritancedetermineownershipof
productiveresources.Rulesof marriagedeterminewomen's
domesticautonomy:if theserulesrequirethat womenjoin
theirhusband'sfamily,womenhavefarlessautonomythan
if theyare ableto forma new householdor livewith their
own family(which is uncommon).The most pervasive
formsof genderinequalityappearwhereboth inheritance
and marriagerulesare heavilyweightedin favorof men.
Bycontrast,wheresuch rulesare more genderbalanced,
women havegreatervoicein the householdand in pub-
lic spacesand face fewerconstraintson becominginde-
pendent economicand socialactors.5
Norms for genderroles and rights form part of the
moral order of a communityand permeateother insti-
tutions, includingthose of the state.This further rein-
forcesgenderinequities,unlessconsciouseffortsare made
to avoidit. Legalsystemsplaya keypart,eitherreinforcing
customarygenderrightsand roles-or deliberatelyseek-
ing to alter them. Also important is the provisionof
publicgoodsandservices,whichoftenbypasswomenun-
less specificeffortsare made to reachthem.
. X
::,lii ,e. ; z rc
d access
ZC tv /resturces
Customarygendernorms and valuescan leadto politi-
cal, legal,economic,andeducationalinequalitiesthat per-
petuatewomen'slackof accessto resources,controlover
decisionmaking,and participationin public life.Greater
political representationcould help changethis-in no
country do womenhold more than a verysmallshareof
6
the seatsin parliament.
Legalsystemscan constrainwomen from becoming
independenteconomicactors.In manycountriesfamily
lawsare heavilystackedagainstwomen,restrictingtheir
rightsin divorceand in inheritanceof landand other pro-
ductiveresources.In most developingcountriestitlesto
land arenormallyvestedin men.7 Sincethe greatmajority
of the world'spoor peoplelive in agrariansettings,this
is a fundamentalsourceofvulnerabilityfor poorwomen.
Somecountriesuse the legalsystemto formalizecus-
tomaryrules that explicitlylimitwomen'srights.In the
Republic of Korea, for example, customary laws re-
stricting women's rights were formalized in the Civil
Code of 1962,and women'slegalrightshavebeen very
slowto improve.Afterdecadesof strugglebywomen'sor-
ganizations,keyamendmentsin 1990 gavewomen the
rightto inherittheirparents'andhusband'sproperty.8 Di-
vorcelawswerechangedto allowwomenequalrightsto
propertyacquiredduringmarriage,and child custodyis
no longergrantedautomaticallyto the father.Butthe law
continuesto insist on male householdheadship,which
women'sorganizationsseeas the main sourceof gender
inequalityin the familyand in other socialinstitutions.
So whilewomenin Koreahavebecomeeducatedand par-
ticipate activelyin the labor force,their unequalstatus
servesto maximizetheir economiccontributionwhile
minimizingadvancesin genderequity.9
In manycountrieswomencontinueto be deniedeven
basic legal rights.In Botswana,Lesotho,Namibia, and
Swazilandmarriedwomen,accordingto both customary
and commonlaw,areunderthe permanentguardianship
of their husbandand haveno independentrightto man-
age property (exceptunder prenuptial contract)." 0 In
Guatemala men can restrict the kind of employment
their wife can acceptoutsidethe home. In some coun-
tries women need their husband'spermissionto obtain
a passportand moveabout freely. 1l
Poorwomen facea doubledisadvantagein accessto
resourcesand voice-they arepoor,and theyarewomen.
Poor people have much less accessto education and
health care than the nonpoor, and the gender gap in
12 The same
theseservicesis largeramongpoorpeople.
is true forcredit and agriculturalextensionservices:un-
less strong countervailingmeasuresare taken, the poor
receivelessthanthe nonpoor,andwomenreceivethe least.
Studiesfrom many countriesshow that agriculturalex-
tensionagentsfocuson malefarmers,eventhoughwomen
are often the primarycultivatorbecausehusbandswork
offthe farm.13 So womenfacedisadvantagesnot only in
landownership,but in gainingaccessto the resourcesand
informationthat would improveyields.
7'hetoll of gender inequality on society
If the rights of men and women are flagrantlyunequal,
it is very difficult to establisha democraticand partici-
patorysociopoliticalorder and an environmentof equal
opportunity. Moreover,the more extreme manifesta-
tions of power inequality between men and women
constitutegrossviolationsof human rights.Domesticvi-
olencehas beenshownto be startlinglyprevalentaround
the world-among people at all income levels(seebox
8.1 in chapter 8).
In some societiesthe lowervalueassignedto women
and girlstranslatesinto excessmortality.Estimatesbased
on officialnational censusesfind that as a result of ex-
 REMOVING SOCIAL
cess female mortality, about 7 percent of girls under age
five are "missing"in China and Korea and more than 4
percent in India and other parts of South Asia.14 With-
out such discrimination there would be an estimated
60-100 million more
60-100 million women in
more women in the world.1515
the world.
Gender inequality also has strong repercussions for
human capital in the next generation, because the bur-
den of bearing and rearing children falls largely on
women. Women deprived of education and decision-
making power in the home face seriousconstraints in rear-
ing healthy, productive children. They also tend to have
more childrenthan they wish, compounding the pressures
on themselves and their family. Better-educated women
are able to communicate better with their spouse about
family size decisions, use contraception more effectively,
and have higher aspirations for their children.1 6
Low autonomy for women takes an independent toll.
Studiesin Chinaand India findthat evencontrollingfor
education, household income, and other socioeconomic
characteristics, low domestic autonomy is associatedwith
higher infant and child mortality rates.17 Studies con-
sistently show that women'seducation improveschild sur-
vival.18 And longitudinal studies in the United Kingdom
and the United States find that, controlling for other
household-level factors, mother's education is associated
with better child cognitive development.19
Among children of women who have greater financial
autonomy, either because they earn cash incomes of their
own or have a greater role in domestic decisionmaking,
nutrition and education are higher. Studies in Brazil
show that more income in the hands of mothers is asso-
ciated with better nutritional outcomes and physical de-
velopment.20 Microcredit programs in Bangladesh find
that giving income-generating loans to women improves
the nutritional status of their children, a result that does
not hold for men.21
Education and autonomy reinforceeach other.Women
with more education and greater domestic autonomy
are better able to nurture and protect their children.22 Low
education and low autonomy make it more difficult for
women to obtain medical care, comply with instruc-
tions, and follow up with the health care provider if the
instructions seem ineffective.They also make it more dif-
ficult for women to obtain healthcare information, pre-
vent illness, and care for the sick.
More equitable distribution of opportunities and re-
sources between men and women also leads more directly
to higher economic growth and productivity.2 3 Cross-
BARRIERS AND BUILDING SOCIAL INSTITUTIONS 119
Figure 7.1
Closingthe gender gap in schoolingmore
rapidlywould boost economicgrowth
Average
Percent annualgrowthin per capitaGNP,1960-92
3.5
3.0 Closingthe gap
at EastAsianpace
2.5
l
2.0 Actual
1.5
1.0
0.5
0
Middle East South Asia Sub-Saharan
and North Africa Africa
Source: World Bank (forthcoming a), estimated from Klasen (1999).
country analysisindicates that countries that investin girls'
education have higher rates of economic growth (figure
7.l).24 Country studies show the benefits of increasing
women farmers' accessto agriculturalextension,credit ser-
vices, and other productive inputs.2 5 Raising their edu-
cation increasestheir efficiencyas producers, by increasing
their adoption of new technologies and their efficiency
in using resources.Analysisfrom Kenya suggeststhat giv-
ing women farmers the same education and inputs as men
increasesyields by as much as 22 percent.2 6 For Burkina
Faso analysisof household panel data suggeststhat farm
output could be increased 6-20 percent through a more
equitable allocation of productive resources between
male and female farmers.27 Further analysis is needed to
determine the impact of such a reallocation on overall
household income and nutritional well-being.
Scope for change
While political and legal equality between men and
women have increased in most regions, it takes effort
and perseverance to change people's gender values and
beliefs.2 8 But much can be done-and much has been
done-to improve women's voice and access to re-
sources by increasing their political representation,
their legal rights, and their command over physical, fir-
nancial, and human capital (figure 7.2). Efforts are
120 WORLD DEVELOPMENT REPORT 2000/2001

Figure ,2
Trendsin female educationand life expectancyreflect increasingequality between women and men

Femalegrossprimaryenrollmentrate Femaleto malegrossprimaryenrollmentratio

Percent
120 1.1

10 1.0__

0.9

0.8

0.7

40 0.6
1970 1980 1990 1995 1970 1980 1990 1995

- East Asia - Europe and - Latin America - Middle East - South Sub- -- OECD
and Pacific Central Asia and the and North Africa Asia Saharan
Caribbean Africa

Femalelife expectancy Femaleto malelife expectancy

ratio
Years
85 1.15

80 _____--

75 ' 1.10

70

65 1.05

55 / 1.00

50

45 0.95
1970 1980 1990 1997 1970 1980 1990 1997

Source: World Bank forthcoming a.

under way in at least 32 countries to increase women's
political representation by reserving seats for them in
local and national assemblies.2 9 In India two amend-
ments to the constitution reserve a third of local coun-
cil seats for women, giving rise to a new class of women
(some 600,000 strong) with political influence; simi-
lar reservation is under consideration for higher polit-
ical levels.3 0 In Argentina at least a third of the
candidates on national election lists must be women.3 1
Women's legal rights have been broadened consider-
ably in many countries. In a growing number of countries
daughters and sons now have equal legal rights to inherit
from their parents. The existenceof such legal rights does
not mean that deeplv rooted cultural norms immediately
change, however. Moreover, the legal system often gives
people scope for implementing their own norms. For ex-
ample, the option ofwritingawill allows people to main-
tain cultural norms on inheritance favoringsons.3 2 When
 REMOVING SOCIAL
Box 7.1
Making land titling less gender biased
in Latin America
Thelandtitlingprocess,rifewithinequities,hasoften reduced
women's access to land.Statutorylaw inseveralLatinAmer-
icancountriesrequiredthat the beneficiariesof earlierland
reform programsbe heads of household.Sincecustom dic-
tated that men were the head of the household,itwas diffi-
cult for women to benefitfrom such programs. Duringthe
1980s and 1990s,however,reform measures changed, and
the more progressiveagrariancodes of the 1990s gave spe-
cialattentionto this problem.
A studybased on gender-disaggregateddatafor six coun-
tries (Chile,Colombia,Ecuador,Honduras,Mexico,and Peru)
shows that women make up a largershare of beneficiaries
under current land titlingprogramsthan under past agrarian
reforms.Still,severalobstaclesto improving women's formal
rghts to landremain:
* Women often are unaware of their rights or of the land
titlingprogram.
* Landtitlingprojectsare often arbitrary.The problemusu-
allystarts withlackof clarityaboutthe bundleof property
rights to land withinone household:those of the wife,
those of the husband,and those to jointlyacquiredprop-
erty.Atthe enforcementstage,this confusionoften works
to women's disadvantage.
* Some legaldispositionsare gender biased. Procedures
ceding rightsto land often aim at individualizing land
rights-one person per household. To work in favorof
women, landtitlingprogramsneed to givefemale heads
of householdpriority,as inChile.
Two sets of measures are particularlyimportantfor pre-
venting gender bias in land titlingand promotingthe rights
of women:
* Makingjoint titling of land to couples mandatory.Joint
titlingguarantees marriedwomen propertyrightsto land
that has beenjointlyacquired.InColombialandtitledjointly
to couples accountedfor 60 percentof landadjudications
in 1996, up from 18 percent in 1995. Land titled exclu-
sivelyto men declinedfrom63 percentto 24 percent over
the same period.
* Fostering partnerships between government depart-
ments and NGOsthat defend the rights of women-to
increase women's awareness of their rights and sup-
port them in claimingtitle to land in the face of a pos-
siblyhostile
bureaucracy
orfamily.InBolivia
andEcuador,
wherewomen'slandrightsfeaturedlittle in the nego-
tiationsleading
to newagrariancodesandwherethere
wasnomovement towardjointtitlingorspecial
rights
forwomen,thereforms didnotimprove women's land
rights.
WorldBank
Source: forthcoming
a{basedonDeere
andLeon
1997
Deere
1999); andLeon
forthcoming.
BARRIERS AND BUILDING SOCIAL INSTITUTIONS 121
legislation conflicts too sharply with customary law, prob-
lems can surface.33 Still, even if laws are not self-enforcing,
they are a necessary first step toward gender equity.
More direct efforts to ensure women s access to pro-
ductive resources include recent land titling programs to
ian Law gave top priority to redistributing land to house-
holds headed by women and to women who lacked
protection or had been displaced by war (including sin-
gle and childless women).34 The scheme-"a parcel of
one's own"-was the only guarantee of secure livelihood
for women and their children upon separation or divorce.
Several other Latin American countries have beenwork-
ing on this issue, with varying success (box 7. I).
ths ise, wh
Wn men ry su ccess(o 7.1).
Women also need more equitable access to credit and
associated productivity-enhancing services. Studies of'
the effect of networking schemes, such as group-based mi-
crocredit, suggest that these schemes have enormous po-
. r .
tential for reducig poverty Some of these credit programs,
such as Grameen Bank in Bangladesh, are targeted more
to women than to men.35 Using peer pressure and group
obligations rather than legal contracts, group-based
. *-
schemes rely on social collateral rather than traditonal
financial assets as security.3 6 The schemes have helped
women acquire nonland assets and have also been asso-
ciated with positive effects on girls' schooling. 37
Critical to these programs are services that comple-
ment credit and savings facilities, such as training in en-
trepreneurial skills-especially for women, who are
typically cut off from the normal paths for acquiring such.
s G t
cessful entrepreneurs. In southern Africa women own.
an impressive share of small, informal sector businesses:
67 percent in Zimbabwe, 73 percent in Lesotho, and
84 percent in Swaziland. 38 The next step is to ensure
greater access for women to business opportunities inL
the formal sector.
Recognizing the constraints women face in gaining ac-
cess to public services and other opportunities makes
antipoverty interventions more effective. In education,,
female teachers and separate sanitary facilities-or even
single-sex schools-can boost girls' enrollments in some
regions.3 9 Demand-side interventions can also be effec-
tive (box 7.2). In agriculturalextension,effortsto hire and
train female extension agents-and to focus extension ef-
forts on women farmers-help make new agricultural
methods and technologies more accessible to them and
increase productivity.
122 WORLD DEVELOPMENT REPORT 2000/2001

.iox 7.3
Usingsubsidies
to closegendergaps Towardstrongerfemalevoicein policymaking:
in education women'sbudgetinitiativesin southernAfrica

Evaluations of recentinitiativesthat subsidizethe costsof The SouthAfricanWomen'sBudgetInitiativebeganas an

schoolingindicatethat demand-side interventionscan in- innovative "joint venture"betweenseveralNGOsandnew
creasegirls'enrollments andclosegendergapsineducation. parliamentarians in thefirst post-apartheidgovernment. The
Aschoolstipendprogramestablished in Bangladesh in1982 parliamentarians were membersof the Genderand Eco-
subsidizesvariousschoolexpenses forgirlswhoenrollinsec- nomicPolicyGroupof the JointStandingCommitteeon Fi-
ondary school,Inthefirstprogram evaluation girls'enrollment nance,whilemanyof theNGOrepresentatives wereinvolved
rateinthe pilotareasrosefrom27percent,similarto the na- inbudget-related andmoregeneralpolicyresearch. Thepur-
tionalaverage, to 44percentoverfiveyears,morethantwice poseof the initiativehas beento highlightthe genderdi-
the nationalaverage(BellewandKing1993).After girls'tu- mensions of the government's budget-includingintaxation,
itionwaseliminated nationwidein 1992andthe stipendpro- expenditure, andthe budgetprocessitself-and to ensure
gramwasexpanded to all ruralareas,girls'enrollmentrate that genderequityis betterservedby the budgetprocess
climbedto 48percentat the nationallevel.Therehavealso andallocations.
beengainsin thenumberof girlsappearing for examsandin Theinitiativehasundertaken fourroundsof budgetanaly-
women'senrollments at intermediate collegesILiang1996). sis on a rangeof sectors.Whilethe earlyroundsfocused
Whileboys'enrollment ratesalsoroseduringthisperiod,they largelyonthe nationalbudgetprocess, the fourthhasbegun
did not riseasquicklyas girls'. to focusonlocalgovernment andondissemination of findings
Tworecentprogramsin Balochistan, Pakistan, illustrate andmessages to a broaderconstituency of SouthAfricans-
the potentialbenefitsof reducingcostsandimproving phys- to betterequipordinary citizensto engage inpolicydiscussions.
icalaccess.Beforethe projectstherewere questions about TheSouthAfricaninitiativehasinspired severalothers.A
whethergirls'low enrollments weredueto culturalbarriers three-yeargenderbudgetinitiativewasstartedin Ugandain
that causeparentsto holdtheir daughtersout of schoolor 1997,ledby the Parliamentary Women'sCaucus in cooper-
to inadequate supplyof appropriate schools.Programeval- ationwiththe ForumforWomenin Democracy, anNGO.Like
uationssuggestthat improvedphysicalaccess,subsidized the SouthAfricanprogram, the Ugandan initiative
involves the
costs,andculturally appropriate designcansharplyincrease coordinated effortsof parliamentarians andNGOresearchers.
girls'enrollments. Already apowerfulforceinUganda, theWomen'sCaucus has
Thefirst program,in Quetta,the capitalof Balochistan, pushedthroughseverallegislativechanges,includingthe
usesa subsidytied to girls'enrollmentto supportthe cre- clausein the localgovernmentlaw requiringthat women
ationof schoolsinpoorurbanneighborhoods by localNGOs. constituteat leasta thirdof executivecommitteemembers
Theschoolsadmitboysas longas they makeup lessthan at the parishandvillagelevels.Thegenderbudgetinitiative
half of total enrollments.In rural Balochistan the second hasfocusedonmacroeconomic policyandgender,including
programhasbeenexpanding the supplyof local,single-sex the effectsof structuraladjustment on poorwomen.
primaryschools for girlsbyencouraging parental involvement In Tanzania anotherthree-yearinitiative,alsostartedin
in establishingthe schools andbysubsidizing therecruitment 1997,is spearheaded bya coalitionof NGOsledbythe Tan-
of femaleteachersfrom the localcommunity.The results: zaniaGenderNetworkingProgram.It focuseson under-
girls'enrollments rose33 percentin Quettaand22 percent standingthe budget processesof the NationalPlanning
in ruralareas.Interestingly, both programsappearto have Commissionand the Ministry of Finance,how those
alsoexpandedboys'enrollments,suggesting that increas- processes affectgovernment spending onbasicservices, and
ing girls'educational opportunities mayhavespilloverben- how governmentspendingdecisionsaffectwomen'sand
efits for boys. men'saccessto healthandeducationservices. Theinitiative
hasbegundisseminating keyfindingsinsimplelanguage to
Source:
WorldBankforthcoming
a;Kim,Alderman,
andOrazem1998. makethembroadlyaccessible.

Source:WorldBankforthcoming
a; Budlender
1999;TGNP1999.

Increasing gender equity has enormous benefits in

establishinga culture of human rights as well as more im-
mediate material benefits through its effects on produc-
tivity and the human capital of the next generation.
Paths to gender equity include giving men and women
equal rights under the law,equal access to education and
health care, and equal accessto servicesrelated to income
generation. Gender budgeting and publication of gender-
disaggregated development indicators can help generate
public support for such efforts (box 7.3). All these in-
terventions need to be backed by efforts to increase the
political participation of women, so that they can con-
tribute more fully to society.
Antidiscriminatory legal, institutional, and policy
reforms for increasing gender equality have both in-
strumental value for development and poverty reduc-
 REMOVING SOCIAL BARRIERS AND BUILDING SOCIAL INSTITUTIONS I23

tion and intrinsic value for furthering human rights and Box 7.4
well-being. More equitable access to material resources Using development programsto break the
and to needed services increases economic productiv- power of agrarian elites: a casestudy from
iry and growth. More generally, increasing gender eq- eastern Uttar Pradesh,India
uity is an important component of efforts to encourage
greatercitizenparticipationin
greater citizen participation
publiclifeand in mon-
in public life and in mon-
The socioeconomichierarchyin the villageis apparent:there
isonelargewhite-washed brickmansionstanding outamong
itoring state institutions. a sea of mudhuts.Themansionis the homeof the talukdar,
the largelandowner whosedutyit was to collectlandrevenues
Social stratificationand poverty for thecolonialpower.The talukdar'sfamilylostsomeof its
holdingswhenlandceilingswere imposedin the 1950s,al-
thoughit heldontomuchof its landthrough fictitiousdivision.
Because we had no schooling we are almost illiterate. Forthe nextcoupleof decadesthe talukdarsfamilycon-
Sometimes we cannot even speak Spanish; we can't add. solidatedits relationships
withthe new powerstructures of
Store-owners cbeat uts,because we don't know how to thestate.Ina typicalpatternof diversifying familynetworks,
count or aniythingelse. They buy at the prices they want the fatherarrangedfor oneson to be in the policeservice,
whileanothermanaged the land.Theycontinuedto bethe
and less. Theycheat us because we are not educated,
pay mainsourceof creditandemployment for the villagers,who
-1iirka'enouswoman in Asocilicionde 10Ago:to,
fcuador acknowledged their socialsuperiorityby prefacingeveryin-
teractionwith the greeting"Touchingyourfeet, Lord."
Economicinequalitiesreinforced bysocial barriersmake Forthisfamily,welleducated
andwellconnected,
it was
easyto divertdevelopment fundsfor its own benefit.The
othervillagersgenerallyneverknewaboutthe entitlements
poverty.When social distinctionsbetween groups are theywerebeing deprivedof.Eveniftheydidknow, theycould
used to perpetuate inequalitiesin accessto material hardly protest,becausethe talukdar'sfamily hadgunsand
resources, they generate rigid sociopolitical hierarchies, was knownto rapeandmaimat will.
which constitute powerfuil social barriers explicitlyaimed Around1970agricultural extensionagenciesbroughtin-
formationon tubewellsto the village.Somemiddle-level
at preserving thestatus of the better-off. Theyplacecrip- peasantspooled theirresources to sinkatubewell to irrigate
plingconstraintson individuals.For poor people,naturally theircontiguousplots.Eager to maximize theirprofits, they
risk averse because they live closeto the margin of sur- began to sowcashcropsandraised thewagestheyoffered
vival,the prospect of incurring thewrath of powerfulelites agricultural
laborers.
Thetalukdars sonresponded bystrid-
ingaroundat the weeklymarketwith a gunslungoverhis
bychallenging these barriers is intimidating. Rigidstrat- shoulder,threateningto shootanyone whooffered laborers
ification also creates obstacles to collectiveaction: if the morethanthegoingrate.Thattemporarily thwartedthe
distribution of power in a community is too skewed, peasants' efforts.
prospects for trust and cooperation are low. Butnewopportunities offeredby tubewellsandthe
openingof agovernment milkcollectioncenterinthe village
Social inequality in villages undermines efforts to madeit moredifficult
forthetalukdars familytoretain itspo-
manage collectivegoods such as water.40 In the hands of sition.Overtimethemiddle-level peasantsincreased their
villageelites, control of these resourcescan be used to fur- incomes andoffered newsources ofemployment andcredit
ther discriminate against poor people. One of the most forpoorpeople. Thevillageshiftedawayfroma bipolar polity
glaring manifestations of inequality is in accessto land. towarda broaderdistribution
of power.
A studyofanother
UttarPradesh
villagenotedsimilar
ten-
In most developing countries large inequalities in land sions.Therethetalukdars
familyhadtriedsuchmethods as
ownership make it virtually impossible for poor people arsonandelectionriggingto maintainits power,butthe
to risefrom the bottom of the agrarianhierarchy.But land middle-level
peasants'
determineduseof newagrarian op-
reform and broader efforts to diversify economic op- portunitieseventually
weakenedthatpower.By the 1990s
the middle-level
peasantshadbecomeprosperous anded-
portunities can break some of these barriers and reduce ucated andformeda serious political
challenge
to thetaluk-
rural poverty (chapter 5; box 7.4). dar'sfamily,
defeatingtheminlocalelections.
Thehegemony
In many settings discrimination and social inequality of the oldcoloniallandowning elitehasbeeneffectively
are the outcome of entire social groups having little po- challengedthroughcontinuing development programs and
participatory
politicalinstitutions.
liticalvoice.These groups are discriminated againstor ne-
glectedin thedistribution ofpublicgoods, which translates Source:
DasGupta,
Grandvoinnet, forthcoming;
andRomani Dreze,
into lower accessto education and health-and lowerin- Laniouw,and Sharma 1998.
come. Most damaging are the poverty traps that arisefrom
124 WORLD DEVELOPMENT REPORT 2OOO/2001

Figure 7.3 42 Living in a better-off neighborhood

Minority groups in Vietnam have less access
to servicesthan nonminorities
Distanceto facility,1992-93
20 Kilometers
Minority
15 ethnicgroups
Majority e States, or even against majority groups, as during the
ethnicgroups ..
10 0 as in Latin America, blacks have lower education and
5 not explained by socioeconomic disparities alone.46
Up Ps Hi Dr M aropportunities
Upper Post Hospital District Market
secondary office center
school
Source:Adaptedfrom vandeWalleandGunewardena
2000.
active discrimination, which can inflict psychological
damage on those discriminated against.
Some poverty traps are created in part by geographic
isolation. Differential outcomes based on geographic iso-
lation are a form of stratification, even if not consciously
designed. For example, the disproportionately high
poverty among indigenous groups in Latin America
partly reflects their greater distance than others from
markets, schools, hospitals, and post offices.Similar con-
straints are documented for minority ethnic groups in
Vietnam (figure 7.3). Indigenous groups in Latin Amer-
ica also receive less education on average than non-
indigenous groups. Ethnic discrimination exacerbates
the effects: returns to schooling are lower among in-
digenous groups. Indigenous people are more likely than
others to be sickand less likelyto seekmedical treatment,
which may also help account for the difference in earn-
ings.41 This is a vicious circle, as low income reduces the
probability of improving one's health.
Isolation and lack of education can create poverty
traps that persist over generations, as children living in
different locations experience different types of human
capital accumulation. Even the neighborhood in which
one lives can have a powerful influence on income and
human capital.
to socialandculturalfactorsthatin-
exposesindividuals
43
h productivity.
creasetheir - Neghborhood
egbrod effects
fet can a
also reduce economic mobility and widen income dis-
parities across communities, as in Ethiopia.4 4 Similarre-
suits have been reported in industrial countries, where the
rich often live apart from the rest of the population.
Other poverty traps result directly from prolonged dis-
crimination against minority groups, as in the United
*'
apartheid regime in South Africa.45 In these countries,
income than whites. But their disadvantages run even
deeper: their life expectancy at birth is also lower, a gap
The cumulative effects of discrimination in education,
employment opportunities, and information weaken the
for members of these groups to find good
p by psy-
4
jobs. This dynamic is powerfullvboosted by the psv-
chological damage from discrimination-and the psy-
chological obstacles to upward mobility add to the
physicaland financialobstaclesto obtaining qualifications.
People cease to believe in their abilities and stop aspir-
ing to join the economic and social mainstream. This so-
cial dynamic emerges forcefully in the context of race
relations in the United States (box 7.5).
Mitigating the impact of social stratification requires
multifaceted approaches. Ensuring that public agencies
and other state institutions serve all sectors of the pop-
ulation equally can make a big difference. This practice
can be furthered bg mobilizingexcludedgroups to be more
assertiveof their needs and rights. In situations of active
discrimination, carefullydesigned affirmativeaction poli-
cies can help equalize access to opportunities.
Refor7ming institutions
In societies not deeply stratified, reform of state institu-
tions can increasesocial equity. A fairly simple reform is
to ensure that deliveryof public services does not neglect
disadvantaged groups. Broader reforms involve making
legal systems equitable and ensuring that administrative
and political institutions are accessibleand responsiveto
all. Rather than create barriers, these systems should fa-
cilitarethe full participation of the entire population. Cit-
izenship laws may also need reform-to reduce social
tensions and enable disadvantaged groups to participate
in political life, which is important to their ability to or-
ganize on their own behalf48 In some countries, having
 REMOVING SOCIAL BARRIERS AND BUILDING SOCIAL INSTITUTIONS I2, 5

Box 7.5 crease their access to health, education, and other pub-
Discriminationis psychologicallydevastating lic services, improving their living conditions and rais-
ing their incomes. Early results from innovative
In ananalysis
of socialexclusion
andtheneedfor affirma- "ethno-development"programs in Ecuador show the
tiveactioninthe UnitedStates,GlennLourydrawsattention importance of cultivating genuinedemand, enhancingself-
to the psychological
havocthatlong-standingdiscrimination
canwreak on black ghetto dwellers: management, and building local capacity-instructive49
lessonsfor development practitionersand policymakers.
Hereis a youngsterto whomonesays,"Why don't
youmarrythe girlyougot pregnant?Instead
of stand- Taking afl7rmative action
ing onthe streetcornerhustling,why don't yougo It
to the communitycollegeandlearnhow to runone Ii s
ofthesemachines inthehospital?
Youcouldlearnthat plemented by affirmativeaction programs-to counter thie
witha coupleof yearsat thecommunitycollegein- disabilities from long-standing discrimination. To com-
steadofbeing
a misfit,"andtheanswer
isnot,"I have pete in economic and political arenas, those discrimi-
donemy sumsandthe courseyousuggestsimply nated againstneedspecialassistance
in acquiring education,
doesnotpay."Instead, hisanswers, "Who,me?" natin and seLf-confidnce mative actionb
Hecannot seehimselfthus. information,and self-confidence. Affirmativeactionbe-
Blackghettodwellers in theUnitedStatesarea ginswith legislationagainstdiscriminationin accessto pub-
peopleapart,susceptible to stereotyping,
ridiculed lic and privategoods and services,such as housing, credit,
fortheircultural
styles,isolatedsocially,
experiencing transport public places and public office.
an internalized
senseof helplessness
and despair, ,p p ,
withlimitedaccesstocommunal networksofmutual Prominent in affirmative action are efforts to reduce
assistance.Inthefaceof theirdespair,
violence,
and the cumulative disadvantages of low access to education
self-destructive
behavior,
it ismorally
obtuse andsci- and employment. This typically involveshelping mem-
entifically
naivetoarguethatif "thosepeople"would bers of discriminated-against groups acquire skills and
just gettheir actstogetherwe wouldnot havesuch .. t . r
a horrificproblem.Socialprocesses
encourage
the de- accessto opportunites through fnancial support for
velopment ofself-destructive
behavior.
Thisisnotto education, preferentialadmissionto higher education, and
saythat individuals
haveno responsibility
for the job quotas.50 These policies, of two main types,make a
wrongchoicestheymaymake.Instead,it is to rec- big difference in outcomes:51
ognize
adeepdilemma,
onethatdoesnotleaveuswith policies seekto enhancethe perfor-
.Developmental
anygoodchoices.
Because thecreation
of a skilledworkforce
isa mance of members of disadvantagedgroups. Exam-
socialprocess,
themeritocratic
ideal-thatinafree ples are financial and other inputs to improve
societyindividuals
shouldbeallowedto riseto the educational qualifications,and management assistance
level of their competence-shouldbe tempered for those establishing their own business.
withanunderstanding
thatnoonetravelsthatroad
alone."Merit" is produced
throughsocialprocesses. * Preferential policies seek to reduce cumulative
Forthisreason, thereshouldbea collective
public disadvantages more rapidly by giving members of
effortto mitigate
theeconomicmarginality
of those disadvantaged groups opportunities even when they
blackswho languishin the ghettosof America.Pub-
lic goalsoughtnotto beformulated
in race-neutral maybe lessqualifiedthan others.ALthoughthe quick-
terms,evenif theinstrumentsadopted for thepur- est way to social and economic mobility, these policies
suitof thosegoalsare,in themselves,
color-blind. can backfireby reinforcing negative stereotypesabout
the lower abilities of the disadvantaged.5 2 Even qua.L-
Source:Passages
excerpted
fromvarioussectionsof Loury12000). ified membersof disadvantagedgroups cannot es-
cape this shadow.
A crucial role for affirmative action policies is to cre-
accountable judicial institutions would also help pro- ate role models who can alter the deep-rooted beliels
tect disadvantaged groups from discrimination. about different worth and abilities that permeate segre-
Poor, marginalized communities can be mobilized to gated societies (box 7.6). Such beliefs, psychologically
help reduce their poverty by drawing on and strength- devastating for the disadvantaged,are also shared by those
ening their social institutions. Groups with a strong col- who offer jobs and promotions, reducing the likelihood
lective identity-and a willingness to collaborate with that they will give equal consideration to minority can-
outside agents to forge new solutions-can work to in- didates, even when they have the necessaryqualifications.
1z6 WORLD DEVELOPMENT REPORT 2000/2001

1ox a negative effects are largelv associated with preferential

Using affirmative action against caste-based policies and can be averted through greater use of
discrimination
in India developmental policies. In the United States affirmative
action has redistributed income to women and minori-
inIndia
Thecastesystem separated
people
intoeconomic
and ties, with minimalloss of efficiency.5 3 Preferentialpoli-
socialstrataby birth,reinforcing
thesedivisions
throughdif-
ferences inritualstatus.Thisrigidhierarchy
remainedlargely ces maybe costlyi the long term, Jobquotas for
inplaceformanycenturies, despiteperiodicchallenges
from minorities may distort the allocationof labor,impede ef-
socialandreligious reformmovements. Butin 1950the ficiency,and create tensions between the "favored" and
government
newlyindependent ofIndiasetouttotransform theothers.54 Preferential policiescan alsohave negative
the system.Theconstitution
abolished inpri-
untouchability . .
vateor publicbehaviorandempowered thegovernment to polticalrepercussions.Polticalelites,seekig to bene-
takecorrectiveactionbyreducing
thesocialandeducational fit from political clientelism,can manipulatepolicies
disadvantages facedby lower-caste
people andintroducing aimed at reducing segregation or reserving employment
affirmativeactioninemployment.Seatsinthenational
par- for particular groups. Developmental policies, lesslikely
liamentandstateassemblieswerereserved formembersof to elicit resentment from other groups, are politically less
scheduled
(lower)castesandtribes,andanactwasadopted
makingthepractice
of untouchability
a criminal
offense. challengingthan preferentialpolicies,and have enor-
Theprocessofchange hasbeenfraughtwithdifficulties. mous potential for reducing the cumulative disadvantages
havebeenmounted
Legalchallenges againstthepolicies
on of longstanding discrimination.
grounds also reflected in the public debate-that lower-
castepeoplehavenomonopoly on povertyandthat the Social fragmentation and conflict
gainsofaffirmative
action
havebeencornered
bya subgroup
of the lower castes.And political resistancearosewhen the
scope ofpreferential
policieswasexpanded
inrecentdecades Group differentiation by such characteristics as eth-
sectorjobsforlower-
to reservelargersharesof government nicity, race, religion, and language can sometimes result
caste people.By contrast, the developmentalpoliciesaimed
at helping
lower-castepeople gainaccess to education
for in socialfragmentation, with groups perceivingthem-
upwardmobilityhavebeeneffective andlesscontentious. selves as having distinct interests even though they may
Despite
thesedifficulties,
theaffirmative
action
programs have similar socioeconomic status. Ethnicity-a multi-
havedonemuchto lowerthebarriers facedbylowercastes. dimensional phenomenon and a controversial notion-
Lower-castepeoplenow occupypositionsin the highest is basedon perceivedcultural differences between
walksof life,servingasrolemodels for others.Still,agreat groups
dealremains to bedone,aseconomic andeducational in- in a society, differences that form a powerful source of
equalities
persist.A surveyin1992-93found that57 percent identityand a baseforpoliticalmobilization. 55 Some
of heads of household wereilliterate
in scheduled castes, scholars have treated ethnicity as a form of capital-a
compared with 35 percent in other castes. And special ef-
fortsareneeded
inthefewremaining
regions
wherethe po- resourceor assetonwhch membersofa partculareth-
licearestilldominated byupper-casteinterests.Nevertheless, nic communitycallin their businessand politicaldeal-
the experience of affirmativeactionin Indiaillustrateshow, ings. 56 Common ethnic affiliations can be a basis for
wlthpoliticalwill,the effectsoflong-standingpatternsofdis- bonding social capital (see next section), providing
crimination canbeovercome, community members with a range of benefits (credit,
Source: Deshpande2000; Dushkin 1972; Galanter 1972; Srinivas employment,marital partners)while imposingsignif-
1987; Tumma a 1999. icant obligations and commitments (financial support,
conformity). Membership in an ethnic community can
also generate negative externalities, as with conflict be-
Affirmative action seeks to alter these perceptions of tween ethnic groups (box 7.7).7 Such divisions can be
different worth by bringing some members of obstacles to collectiveaction: in the United States greater
discriminated-against groups into the mainstream econ- ethnic fragmentation is associated with lower partici-
omy and society.This has an important demonstration ef- pation in civic activities. 5 8
fect: havingblack or low-castedoctors, for example, shows Ethnicity can become a basis for competition for po-
everyone,including their own group, that members of this litical power and for access to material resources.5 9 Un-
group can be good doctors. lessinstitutions of the state and civil societv offer forums
Do affirmative action programs reduce efficiency or for mediating intergroup rivalries and forging cross-
engender political strife? Evidence shows that these cutting ties among diverse ethnic groups, these ethnic
 REMOVING SOCIAL
cleavagescanleadto conflicts,
tearinga societyandecon-
omyapart,leavingeveryonevulnerable to poverty.
The extentto whichsocialfragmentation leadsto
conflictdependslargelyon administrative and political
institutions.To createa functioningsociety,a whole
rangeof socialand politicalinstitutionsmustworkto-
gether. By contrast, breakdowns in governanceand in the
deliveryof publicgoodsand relatedsocialservicescre-
ateconditions forsocialunrestandconflict-asdobreak-
downsin the institutionsof conflictmediation,suchas
representativepoliticsand theruleof law.
Ethniccleavages canaffectdevelopment outcomesin
manyways.Theycaninfluence theinternalorganization
ofgovernment andtheallocation ofpublicspending,lead-
ingto unequaldistributionofpublicgoodsandservices.
Theycanencourage rentseeking,reducingtheefficiency
60
ofpublicspending. Furthereconomicdistortions enter
whenpowerfulethnicgroupsusetheirpoliticalpowerto
increasetheirincomesrelativeto thoseof others.Recent
studiesin Ghanashowthatlocallydominantgroupsre-
ceivea 25 percentpremiumover the wagesof other
groupsin thepublicsector-a discrepancy thatleadsto
Box7.7
Ethnicdivisionsand civilconflict
Ethnic fragmentation,inits mostextremeformandundercon-
ditionsof economicdeprivation and nondemocratic govern-
ment,candescendintocivilconflict. Ethnicconflictintensified
inthesecondhalfof the 20thcentury,as the patternofcon-
flictshiftedfromwarsbetweennationsto conflictswithin
states.Civilconflict
isbotha causeanda consequence ofpoor
economic performance.Researchhasshownthatduring civilwars
percapitaoutputfallsbymorethan2 percenta yearonaverage,
Themostimportantcostofcivilconflict is lossof life-a
humanitarian tragedyandan obstacletoreconstruction. Other
costsincludedestruction of physical,
human,andsocialcap-
ital;lowerinvestmentinphysicalandhumancapital;disrup-
tionofmarketsandotherformsofeconomicandsocialorder;
diversionof humanresources andpublicexpenditureawayfrom
productive or productivity-enhancingactivities;migrationof
highlyskilledworkers;andtransfersoffinancialassets abroad.
These costs can trapcountriesin poverty-and in conflict.
Civilconflictcan also acceleratethe collapseof the state,
disproportionatelyhurtingpoor people.Andthe problemsof
civilconflictspillacross borders, increasingthe burdens of
neighboringcountries.In 1998there were an estimated 12.4
millioninternationalrefugeesand 18millioninternationallydis-
placedpeople,almosthalfof theminAfrica.
Source:Collier
and Hoeffler1998;Austin1999;Stewart, Humphreys,
andLea 1997;Collier1999c, Luckham1999.
BARRIERS AND BUILDING SOCIAL INSTITUTIONS 127
unrestand poorperformance 61 Suchdis-
in thesector.
tortionsin the distributionof resourcesand the effi-
ciencyof theiruseshowup in development outcomes.
In severalAfricancountries,forexample,childsurvival
is higherin dominantethnicgroups.62
Buildi7ng political allianices
Countrieswithhighethnicdiversity needto buildthepo-
liticalconditionsforintegrating
diversegroupssothatthey
63
can functioncollectively.With well-functioning ad-
ministrative and politicalinstitutions,
multiethnicsoci-
eties can be effectivelyshaped into an "imagined
community" ofnationandstate.64 Knittingdiversecom-
munitiestogetherthrougha multiplicity ofcivilandstaite
channels-to avertconflict-wasa majorgoalof the
earlydesignersof Europeanunity. 65 The communist
regimes oftheSovietUnionandYugoslavia, despitetheir
economic andpolitical
failures,
notonlyreducedeconomic
inequalities butalsomanagedethnicconflict.Withtheir
collapse, violentethnicconflictsbrokeout becausenoal-
ternativeideological and institutionalframeworkhad
evolvedto mediatethem.
In Sub-Saharan Africanation-states werefashioned
outofarbitrary divisions
ofterritorybycolonialpowers--
divisionsoften basedon convenientgeographicmarkers
such as linesof latitudeand longitude,with no consid-
eration of the social units of local populations. With dis-
parategroupsand fewsupraethnicinstitutionsto mediate
I v v c
among them, the creaton of nation and state has been
fraught with problems. Colonial rulers and local politri-
cians have often manipulated ethnic tensions for private
gain, sometimes leading to gruesome civil wars.66 In-
flamingethnictensionsandcivil unrestis a frequentstrat-
egy for gaining and keeping power in these circumstances,
since it justifies expanding brutal military forceswhile un-
dermining the capacity of opposition groups demanding
reform. Overtime ethnic minorities,especiallythosefac-
ing discrimination, inequality, or conflict, can become
ethno-classes,67 groups whose ethnicity-based sensibilities
and demands become independent causes of conflict.6 8
Constructing high-quality public institutions is essential
for ensuring that diverse identities become a develop-
mental asset, not a source of political division and violence
(figure 7.4) 69This is especially important in countries witi
abundant natural resources, such as oil, diamonds, and
minerals. 7 0 In environments with little institutional
128 WORLD DEVELOPMENT REPORT 2000/2001

Figure 7.4
Ethnic diversity is associated with violence In addition to removing socialbarriers,effectiveefforts
where institutional quality is low to reduce poverty require complementary initiativesto
build up and extend the social institutions of the poor.
Probability of state-sanctioned ethnic killings, 1960-90 Social institutions refer to the kinship systems, local or-
,r/ ------ -H-igh-~ - -~ ~ ~ ~~ ~~ ~5 -~~ ganizations, and netwvorksof the poor and can be usefully
0.5 / ', fragmentation discussed as different forms or dimensions of social
I - capital.
0.4
Bondin?g, bridging, and linikinigsocial capital
Low Distinguishing among differentdimensions of socialcap-
0.3 ethnolinguistic .dfeetosc

fragmentation ital within and between communities is useful for un-

0.2 derstanding the problems faced by poor people (box 7.8).
0 The strong ties connecting familymembers, neighbors,

close friends, and business associates can be called

0.1 ?-l l ____ _ _ _ ________ _ __ bonding socialcapital. These ties connect people who
share similar demographic characteristics.
0
Low High * The weak ties connecting individuals from different
Institutionalquality ethnic and occupational backgrounds can be referred
Source:Easterly2000a. to as bridging socialcapita /i2 Bridging social capital
implies horizontal connections to people with broadly
comparable economic status and political power. A
accountability and transparency,the exorbitant rents from theory of social capital that focuses only on rela-
these resourcesbecome a primary source of competition tions within and between communities, however,
among ruling factions. opens itself to the criticism that it ignores power.7 3
Civil society organizations and the state can do much * A third dimension, linkingsocialcapital,consists of the
to lay the institutional foundation for groups to cooper- vertical ties between poor people and people in posi-
ate for the common good. Institutions need to be par- tions of influence in formal organizations (banks,
ticipatory, credible, and accountable, so that people can agricultural extension offices, the police).74 This di-
see the benefits of cooperation. Underpinning these in- mension captures a vitally important additional fea-
stitutions need to be constitutional and legal systems ture of life in poor communities: that their members
and representative political systems,which allow groups are usuallyexcluded-by overt discrimination or lack
to work out their intereststhrough mechanismsother than of resources-from the places where major decisions
violence. Some social integration can be achieved by en- relating to their welfare are made.
couraging people to learn each other's languages.Another Research on the roles of different types of social net-
important requirement for effectivelyhelping excluded works in poor communities confirms their importance.
groups is to collect accurate data on them.7 t An analysis of poor villages in rural North India, for
example, shows that social groups play an important role
Bujidding social institutions and in protectingthe basicneedsof poorpeopleandmediating
social capital against risk (chapter 8). In contrast, the more extensive,
leveragednetworks of the nonpoor are used for strategic
d b?:Ul, e? we work togetber . womlen
re is ti
,ineral. advantage,such as procuring better jobs and higher wages
(ita iw
wari- . ewiood, and collect maizeflour firo77z
('cllecf tn and seizing new economic opportunities (in some cases
while the men dig thpegrae rUld bllry the by directly mobilizing to secure a disproportionate share
dei/> 1, . t uork together on colmmunityprojects like of public resources and services).75
?nottu jcs'
a 1ks
for a school . . .'onomen also work Strikingly similar results emerge from work on the
?jhtr/rn
cleaning around the boreboles. relationship between enterprise performance and the
structure of business networks in Africa. Poor entre-
 REMOVING SOCIAL BARRIERS AND BUILDING SOCIAL INSTITUTIONS IZ9

preneurs operating small local firms in traditional
dustries form "solidarity networks," sharing personal in-
formation about members' conduct and intentions in
order to reduce risk and uncertainty. In contrast, larger
regional firms form "innovation networks," which share
knowledge about technology and global markets in
order to increase productivity, profits, and market
share.7 6 Studies of agricultural traders in Madagascar
show that social relationships are more important to
traders than input prices. Close relationships with other
traders are used to lower the transactions costs of ex-
change, while ties to creditors and others who can help
out during times of financial hardship are vital sources
of security and insurance.7 7 In Bolivia, Burkina Faso,
and Indonesia field surveys attempting to measure
in- household social capital have found a positive associa-
tion with household consumption, asset accumulation,
and access to credit.7 8
Researchers and practitioners have long recognized
that the bonding and bridging social capital in local or-
ganizations is necessary but insufficient for long-term
development. In Kenya a participatory poverty assess-
ment found more than 200,000 community groups in
rural areas, but most were unconnected to outside re-
sources and unable to help poor people rise out of
poverty.7 9 The creation of linking social capital is es-
sential, and external support has often been important
in its emergence.8 0 External support-from NGOs anid
religiousorganizations, for example-can help create so-
cial capital that increases the voice and economic op-

Box 7.8
How does social capital affect development?

Thereareat leastfourviewson the relationship betweensocial complementone another.Macroinstitutionscan providean

capitalanddevelopment (Serageldin andGrootaert2000;Wool- enablingenvironmentin which microinstitutionsdevelopand
cockandNarayan 2000).Thenarrowestholdssocialcapitalto be flourish.In turn,localassociations helpsustainregionalandna-
the socialskillsof individuals-one'spropensityfor cooperative tional institutionsby giving them a measureof stabilityand
behavior, conflictresolution,tolerance, andthe like(Glaeser, Laib- legitimacy-andby holdingthem accountable for their actions
son,andSacerdote 2000). (Evans1996;Woolcock1998; Narayan1999; Serageldinand
A moreexpansive mesoviewassociates socialcapitalwithfam- Grootaert2000;Putnam1993).
iliesandlocalcommunityassociations andthe underlying norms Whilethe mechanisms by which socialcapitaloperateare
(trust,reciprocity)thatfacilitatecoordination andcooperation for generallywell understood, thereis lessconsensus onwhetherthey
mutualbenefit.Thisviewhighlightsthe positiveaspectsof social qualifysocialcapitalas "capital."In manycasesnormsandinsti-
capitalfor membersof theseassociations but remainslargely tutions havethe durabilityandlastingeffectsassociatedwith
silentonthe possibilitythatsocialcapitalmaynotimpartbenefits capital(Collier1998;Narayan andPritchett1999).Someargue, how-
to societyat largeandthatgroupmembership itselfmayentailsig- ever,that the sacrificeof a presentfor a futurebenefit,typicalof
nificantcosts. traditionalformsof capital,is not presentin socialnetworks-to
A morenuancedmesoviewof socialcapitalrecognizes that theextentthatthesenetworksarebuiltfor reasons otherthantheir
groupmembership canhavebothpositive andnegative effects(Cole- economic valueto participants (Arrow2000).Evenso,socialnet-
man1990;Burt 1992;Portes1995;MasseyandEspinoza 1997). worksandorganizations areclearlykeyassetsin the portfolioof
Thisapproach broadens the conceptof socialcapitalto includeas- resources drawnon by poorpeopleto manageriskandopportu-
sociationsinwhichrelationships amongmembersmaybehierar- nity.Theyarealsokeyassetsfor the rich,who advancetheir in-
chicalandpowersharingunequal. Theseformsof associations and tereststhrough suchorganizations ascountryclubsandprofessional
networksaddressawiderrangeof objectives: someof themserve associations,but their relativeimportanceis greaterfor poor
onlythe privateinterestsof members,whileothersaremotivated people.
by acommitment to servebroader publicobjectives. Thisviewem- Socialcapitalhasitsdarkside,however. Wheregroupsor net-
phasizes thatgroups,inaddition to providingbenefitsto members, worksareisolated,parochial, or workingat cross-purposes to so-
canmakesignificantnoneconomic claimsonthem. ciety'scollectiveinterests(gangs,drugcartels),the socialcapital
A macroviewof socialcapitalfocuseson the socialandpolit- withinthem servesperverseratherthan productivepurposes,
icalenvironment thatshapessocialstructuresandenablesnorms undermining development (Rubio1997;Levy1996;PortesandLan-
to develop.Thisenvironment includesformalized institutional
re- dolt1996). Organized crimesyndicates, suchasthoseinLatinAmer-
lationshipsandstructures,suchas government, politicalregime, icaandRussia,generatelargenegativeexternalities for the rest
ruleof law,the courtsystem,andcivilandpoliticalliberties.Insti- of society-lost lives,wastedresources,pervasiveuncertainty
tutionshaveanimportant effectonthe rateandpatternof economic (Rose1999).AndinIndia,for example, obligationsto familymem-
development (North1990;Fukuyama 1995;Olson1982). bersandpressures to fulfillcommunityexpectations leadmany
An integratingview of socialcapitalrecognizes that micro, younggirls to dropout of school(DrezeandSen1995;PROBE
meso,andmacroinstitutionscoexistandhavethe potentialto Team19991.
130 WORLD DEVELOPMENT REPORT 2000/2OOI
Box 7.9
The federation of comedoresin Peru:
the creation of linking socialcapital
Thecomedores(community
kitchens)movement,
oneof the
mostdynamic women'sgroupsin Peru,emerged
in the
mid-1980s. Participantssought to movebeyond theirtradi-
tionalsurvivalstrategy andmakedemands on thepolitical
system. Federations wereformed attheneighborhood level,
thenatthedistrictlevel,andfinallyatthemetropolitan and
nationallevels.Centralization
ofthemovement lowered the
costof inputs,suchasfoodandkitchen equipment, and in-
creased theavailability
of educationalworkshops.
Thehighest-level organization,theCNC(National Com-
missionof Comedores), became theofficiallyrecognized
representative of the comedores, Oneof its centralde-
mands wasto include allpoorwomeninwelfare programs,
notjustthosewith connections to the rulingparty.Besides
influencingpolicymaking, the comedores havehada signif-
icantimpacton localpowerrelationsintheshantytowns and,
by extension,onthe structureof the politicalsystem.
Althoughthe movement'sactionshavebeenlimitedby
thestructureof thePeruvian
state(withfewformalchannels
for political
action),
thenetworkof comedores represents
a Many casestudies show that socialcapital can improve
formofsocialcapitalthathasenhanced poorwomen'svalue
as anelectoralconstituency.
The comedoreshavealso in-
creasedwomen'snegotiatingpowerIntheirfamilies.
Source:
Houtzager
andPattenden
1999.
portunitiesof poor people(box7.9).This supportis most
effectivewhen it is sustainedover time, emphasizesca-
pacity building, and is based on a sensitive under-
standingof the localconditionsand a relationshipof trust
and partnership.
This approachcharacterizesthe work of Myrada,an
Indian NGO deliveringmicrofinanceservices.Myrada
actsas a medium-termintermediarybetweenpoor people
and commercialbanks.81Its initialtaskis to mobilizethe
bonding social capital within villagecommunitiesto
form credit managementgroups and then over time to
form regionalfederationsmade up of representatives
from eachcreditgroup (therebyenhancingeachgroup's
bridging social capital). From the outset credit man-
agementgroupshold accountswith commercialbanks,
progressively gainingthe confidenceand skillstheyneed
to participateindependentlyin formalinstitutions(link-
ing socialcapital).Afterfiveyearsof trainingand hard-
won experience,group members are able to manage
these accounts-and even arrangefor annual external
audits-without the involvementof Myradastaff,who
move on to start the processafresh.
U,sing social capital to improve program
effectiveness
The state plays a vital role in shaping the contextand cli-
mate in whichcivilsocietyorganizationsoperate(chap-
ter 6).82In somecasesthe statecan alsocreatesocialcapital.
In 1987theDepartmentof Health of the stateof Ceara,
Brazil,launched a rural health program-since then ex-
panded to most of the country-that increasedvaccina-
tion rates significantlyand reduced infant mortality. The
successof the programhas beenattributedlargelyto the
building of trust between government workers and poor
people. The program made building trust an explicit
part of the health workers' mandate by adopting a client-
centered,problem-solvingapproachto service delivery.
Workers were helped by government media campaigns
that publicized the program regularly and gave them a
senseof calling.The resultwasa total reversalof attitude:
motherswho oncehid their children from government
8
83
workerssawthe agentsastrue friendsof the community.
project design and sustainability (box 7. 10). Recent eval-
uationsofWorld Bank rural developmentprojectsshow that
outcomesturnheavilyon thenatureof the powerrelations
betweenkeystakeholdergroupsandon the fitbetweenex-
ternalinterventionsand localcapacities.Howrelationsbe-
tweenstakeholdersevolveover time has an important
bearingon the generationof trust.Projectandcommunity
leaderswho createconfidenceand goodwillare crucial,
suggestingthat high turnoveramongfield staff can un-
dermineprojecteffectiveness. The GalOya irrigationpro-
ject in SriLankahas succeededin a destituteregionwith
high levelsof ethnicviolencebecauseof the patienceand
long-standingcommitmentof fieldstaff(aptlycalledin-
stitutionalorganizers).
The project'skeycontributionhas
beenintegratinglocalknowledge withexternalexpertise
and
forgingcooperationbetweenNGOs and governmentof-
84 InAfricarecentinnovations
ficials. in community-driven
development programshaveshiftedresponsibility
formain-
taining hand pumps and latrinesdirectlyto communi-
ties.85 Wherepreviouslysuchitemsbrokedownquicklyand
took monthsto repair,theyare nowin goodcondition.
A keylessonforpractitionersand policymakersis the
importance of using existingforms of bridging social
capitalin poor communitiesas a basisfor scalingup the
effortsof local community-basedorganizations. 8 6 Cre-
ating more accessibleformal institutions helps poor
peoplearticulatetheir intereststo those in powermore
clearly,confidently,and persuasively.
 REMOVING SOCIAL BARRIERS AND BUILDING SOCIAL INSTITUTIONS 131

Box7.10
Mobilizingand creatingsocialcapitalin developmentprojects

Development programs havereliedonlocalgroupsof projectben- projects. InCoted'lvoireruralwatersupplyimprovedsignificantly

when
eficiaries or localassociationsto improvethe successof develop- responsibility for maintenance wasshiftedfromthe national water
mentprojectsfor morethantwo decades.'What is new is the distribution companytocommunitywater groups.Breakdown rates
umbrellalabelsocialcapitalto referto the underlying socialforce werereduced from50percentto 11percent, whilecostsfellnearly
or energy. 70percent.Theseresultsweresustained, however,onlyinvillages
In Bangladesh Grameen Bankreliesongroupsof poorwomen in whichwell-functioning community organizations
existedandde-
to implementprograms, andthe Bangladesh RuralAdvancement mandforwaterwashigh(Hino1993).
Committee ongroupsof villageworkerswithlittleornoland.In Pak- In manycases,challenging existingnormsandpracticesin-
istanthe AgaKhanRuralSupportProgram givesassistance to vil- creasesthe socialcapitalof previously
excludedgroupswhilede-
lageorganizations to supplement theirself-helpefforts.TheKenya creasingthe power of localelites,helpingreduceobstaclesto
TeaDevelopment Authorityworkedwith growercommitteesto povertyreduction.Development programssuchas women'smi-
promoteproduction, obtainingaone-third shareof thecountry'stea crofinancein Bangladesh changethe socialrelationships in a
exportswithin 15years.The6-Smovementin nineWestAfricancoun-village-indeed,theirsuccessdependsonit. Breaking the gripof
triesorganized peasant federations inmorethan2,000communities moneylenders,overcomingthe resistanceof certainreligious
to helpfarmersovercome the hardships of the dry season.The leaders, andgivingwomenmoredecisionmaking powerwithintheir
CenterforSocial andEconomic Development inBoliviahassupported household allrequireafundamental realignmentof traditionalso-
morethan250peasantorganizations thatpromoteprograms inagri- cialrelationships.
Manydevelopment programs areinherentlypo-
culture,livestock,forestry,artisanproduction, andcommunityin- litical(FoxandGershman 1999),andpowerfulvestedinterestscan
frastructure lUphoff1993;Krishna, Uphoff,andEsman1997). be expectedto mobilizeagainstreformsthat seekto erodetheir
Localgroupshavealsobeenusedfrequently in irrigation,
water positionin the nameof poorpeople.Development researchers,
supply, andsanitation programs.TheOrangiPilotProjectinPakistan policymakers, andpractitionersmustrecognize thesetensionsand
provided low-costself-helpsewerage facilities
andotherservices to respondappropriately.
poorsettlements andhelped autonomous localinstitutionsimplement
1. Amongthe firstsystematicevaluations
of communityparticipation
was EsmanandUphoff(1984).

practices and create visible role models for others to fol-

Many aspects of social norms and practices help gen-
erate and perpetuate poverty.Discriminatory practices as-
sociated with gender, ethnicity, race, religion, or social
status result in the social, political, and economic exclu-
sion of people. This creates barriers to upward mobility,
constraining people's ability to participate in economic
opportunities and to benefit from and contribute to eco-
nomic growth. It also constrains their effective partici-
pation in political processesand civilaction to ensure that
state institutions are accountable to citizens and respon-
sive to their needs.
Policies and programs for mitigating social exclusion
depend on the nature of the exclusion. In some casesex-
clusion can be addressedsimply by improving the outreach
of public servicesto neglected areas. Where more active
discrimination is involved, it is important to ensure eq-
uity in the law and in the functioning of state institutions.
In addition, affirmative action policies may be needed to
reduce the cumulative disadvantages of discriminatory
low. Where there is considerable ethnic heterogeneity
and social fragmentation, conflict can be averted through
efforts to increasethe civic interaction of different groups
and engage them in resolvingpotential conflicts through
political processes.Gender-based discrimination is qual-
itatively different from these other forms of discrimina-
tion because it involves intrahousehold distinctions in
assigning value to people and allocating resources ac-
cordingly.Reducing gender-based social barriers requires
changing deep-rooted beliefs about appropriate gender
roles, as well as taking action to ensure greater gender eq-
uity in the functioning of formal public institutions.
Increasing the participation of the poor in develop-
ment and reducing social barriers are important com-
plements to creating an environment in which they have
greater opportunity and security. This empowerment is
enhanced by scaling up social institutions, increasing the
capacity of poor people and the socially disadvantaged
to engage society's power structure and articulate their
interests and aspirations.
 PART IV

Securit
 f~~~~
Helping Poor People
Manage Risk
To be well is to know what will happen with me
tomorrow.
-,'1id/lc-agH 7ma, Razzad,Bulgaria
J overtymeansmorethaninadequateconsump-
tion,education,andhealth.Asthevoicesof thepoor
cryout, it alsomeansdreadingthefuture-knowing
that a crisismaydescendat any time,not knowing
whetherone willcope.Livingwith suchriskis part
of lifeforpoorpeople,and today'schangesin trade,
technology, andclimatemaywellbeincreasing therisk-
inessofeveryday life.Poorpeopleareoftenamongthe
mostvulnerablein societybecausetheyarethemost
exposedto a widearrayof risks.Their lowincome
meanstheyarelessableto saveandaccumulate assets.
That in turn restrictstheirabilityto dealwitha cri-
siswhenit strikes.
Economicgrowthisonewayof reducingthevul-
nerabilityof poorpeople.Astheirincomesrise,they
arebetterableto managerisks.However, at anypoint
in timethosewhoarepoorwillseetheirvulnerabil-
CHAPTER 8
ity lessenedif mechanisms to reduce, mitigate, and cope
with risks are available to them.
Poorpeoplehavedevelopedelaboratemechanisms
fordealingwithrisk.Butthemechanisms arefarfrom
capableof eliminatingvulnerability.Manyof the
mechanisms offershort-termprotectionat long-term
cost,preventinganyescapefrompoverty.
Thepolicyresponse to vulnerability
mustbeaimed
at helpingpoorpeoplemanageriskbetterby reduc-
ing and mitigatingriskand lesseningthe impactof
shocks.Suchpoliciesaddresstheimmediateproblems
ofshocksandtheinability to copewiththem.Butthey
alsolaythefoundationsforinvestmentbypoorpeo-
plethatcantakethemoutof poverty. Thisreportad-
vocatesa modularapproachto riskmanagement that
adaptssafetynets to the specificpatternof risk in
eachcountryandcomplements existingriskmanage-
mentarrangements. This chapterbrieflyreviewsex-
periencewithseventoolsespecially relevantforpoor
people:healthinsurance,oldageassistance and pen-
sions,unemployment insuranceandassistance,work-
fareprograms,
socialfunds,microfinanceprograms, and
cashtransfers.
135
136 WORLD DEVELOPMENT REPORT 2000/200I

Table 8.1
Main sourcesof risk

Idiosyncratic Covariant
Risksaffectingan Risksaffecting Risksaffecting
individualor groupsof households regionsor
Typeof risk household(micro) or communitiesImeso) nations(macro)

Natural Rainfall Earthquake

Landslide Flood
Volcaniceruption Drought
Highwinds
Health Illness Epidemic
Injury
Disability
Old age
Death
Social Crime Terrorism Civilstrife
Domesticviolence Gangactivity War
Socialupheaval
Economic Unemployment Changesin food prices
Resettlement Growthcollapse
Harvestfailure Hyperinflation
Balance of payments,
financial,
or currencycrisis
Technologyshock
Termsof tradeshock
Transitioncostsof
economicreforms
Political Riots Politicaldefaulton
socialprograms
Coupd'etat
Environmental Pollution
Deforestation
Nucleardisaster
Source:Adaptedfrom Sinhaand Lipton 11999)
andWorld Bank(2000q).

A typology of risks Thisdistinctionbylevelofriskis critical.A riskthat

One wayto understandrisksbetterand designappro-
priatepolicyresponsesis througha typologyof risksand
shocksto whichpeoplearevulnerable(table8.1).Risks
canbe classified
bythelevelatwhichtheyoccur(micro,
meso,and macro)and by thenatureof theevent(nat-
ural, economic,political,and so on). Microshocks,
oftenreferredto as idiosyncratic, affectspecificindi-
vidualsor households.Mesoshocksstrikegroupsof
householdsor an entirecommunityor village.These
shocksarecommon(orcovariant)to allhouseholdsin
the group.Shockscanalsooccurat the nationalor in-
ternationallevel.
affectsan entirevillage,for example,cannotbe insured
solelywithinthevillage.It requirespoolingwithareasnot
subjectto therisk.In practice,manyshockshaveboth
idiosyncraticand covariantparts,thoughmostempiri-
calstudiesfindthat theidiosyncratic partofincomerisk
is large.'Thischapterfocuseson risksthat usuallyhave
largeidiosyncratic
components: illnessandinjury,oldage,
violence,harvestfailure,unemployment, andfoodprice
risk(box8.1). Covariantrisksarediscussed in chapter3
(box3.2)and chapter7 (warand civilstrife)and chap-
ter 9 (macroeconomic shocksandnaturaldisasters).
Theextentto whicha riskiscovariantoridiosyncratic
dependsconsiderablyon the underlyingcauses.For
 HELPING POOR PEOPLE MANAGE RISK 137

Box 8.1
Poorpeople's exposureto risk

Poor people are exposed to a wide range of risks. der rate for adolescent males in poor neighborhoods was 11
times that in wealthier ones (Sinhaand Lipton 1999). Poor peo-
Illnessand injury ple frequently voice their fear of violence and the resulting pow-
Poor people often live and work in environments that expose erlessness, "I do not know whom to trust, the police or the
them to greater risk of illness or injury, and they have less access criminals."
to healthcare(Prasad,Belli,and DasGupta19991.Theirhealthrisks Crime also hurts poor people indirectly. Childrenexposed to
are strongly connected to the availability of food, which is af- violence may perform worse in school (Morrison and Orlando
fected by almost all the risks the poor face (natural disasters, 1999). A study of urban communities in Ecuador, Hungary,the
wars, harvestfailures, andfood price fluctuations; de Waal 1991). Philippines,and Zambiashowed that difficult economicconditions
Communicablediseasesare concentratedamong the poor, with leadto destruction of social capitalas involvement in community
respiratory infections the leadingcause of death (Gwatkin, Guil- organizationsdeclines,informalties among residentsweaken,and
lot, and Heuveline2000). A recent study of poverty in Indiafound gang violence, vandalism, and crime increase (Moser 1998). Vi-
that the poorare 4.5 times as likelyto contracttuberculosis as the olence and crime may thus deprive poor people of two of their
rich and twice as likely to lose a child before the age of two best means of reducing vulnerability: human and social capital.
(World Bank 1998t). Richand poor women alike are victims of domestic violence,
Illness and injury in the householdhave both direct costs (for but the incidenceis often higher in poor households. In Santiago,
prevention, care,and cure) and opportunity costs (lost income or Chile, 46 percent of poor women and 29 percent of wealthy
schoolingwhile ill; Sinha and Lipton 1999).The timing, duration, women suffer from domesticviolence;in Managua,Nicaragua,54
and frequency of illness also affect its impact.A study of South percent and 45 percent (Morrisonand Orlando 1999).
India found that householdscan compensatefor an illness during
the slack agriculturalseason,but illness during the peak season Unemployment and other labor market risks
leads to a heavy loss of income, especiallyon small farms, usu- Labor market risks include unemployment, falling wages, and
ally necessitating costly informal borrowing (Kochar1995). having to take up precariousand low-qualityjobs in the informal
sector as a result of macroeconomiccrises or policy reform. The
Old age first workers to be laid off during cutbacks in public sector jobs
Many risks are associatedwith aging: illness, social isolation, in- are usuallythose with low skills, who then join the ranks of the
ability to continue working, and uncertaintyabout whether trans- urban poor, a patternobservedin Africa and Latin America during
fers will provide an adequate living. The incidence of poverty the structural adjustment reforms of the 1980sand early 1990s
among the elderly varies significantly. In most Latin American (ECLAC1991;Sinha and Lipton 1999).The EastAsian crisis also
countries the proportion of people in poverty is lower for the el- hadpronouncedeffects on labormarkets,with realwages and non-
derlythan for the populationat large(IDB2000).In contrast,in many agricultural employment falling in all affected countries (World
countries of the former Soviet Unionthe incidenceof poverty is Bank 1999j).As state enterprisesin EasternEuropeand the coun-
above averageamong the elderly, particularly among people 75 tries of the former SovietUnionwere privatized,povertyincreased
and older (Grootaert and Braithwaite 1998; World Bank 20001). among displacedworkers with low educationand obsolete skills,
Women, because of their longer life expectancy, constitute the not qualifiedto work in emerging industries.Wagearrearsin Rus-
majority of the elderly,and they tend to be more prone to poverty sia intensifiedthe problem (Grootaertand Braithwaite 1998).
in old age than men (World Bankforthcoming a). The number of Fluctuations in demandfor labor often disproportionatelyaf-
elderly people in the developing world will increase significantly fect women and young workers. Most public sector retrench-
in coming decadeswith the rapiddemographictransition. ment programshave affected women's employment more than
Consultationswith poor peopleshow that income security is men's (WorldBankforthcominga),and womenare more likelythan
a prime concern of the elderly, followed closely by access to men to work for small firms, which tend to be more sensitive to
healthservices,suitable housing,and the qualityof familyand com- demand fluctuations (Horton and Mazumdar 1999).As incomes
munity life. Isolation,loneliness,andfearall too often markold peo- fall, poor householdstry to increasetheir labor market participa-
ple's lives (Narayanand others 1999). As an elderly woman in tion, especiallyfor women and children. This responsehas been
Ukraine put it, " If I lay down and died, it wouldn't matter, because documented in many countries (Horton and Mazumdar 1999;
nobody needs me. The feeling of being unnecessary, of being Grootaertand Patrinos 1999).
unprotected, is, for me, the worst of all."
Harvest failure and food price fluctuations
Crime and domesticviolence Weather-relateduncertainties (mainlyrainfall),plant disease,and
Crime and domestic violence reduce earningsand make it harder pests create harvestrisk for all farmers, but technologiesfor re-
to escapepoverty. While the rich can hire private security guards ducing such risks (irrigation, pesticides, disease-resistant
and fortify their homes, the poor have few means to protect
themselves against crime, In Sao Paulo, Brazil, in 1992the mur- (boxcontinues on nextpage)
138 WORLD DEVELOPMENT REPORT 2000/2001
Box 8.1
Poor people'sexposureto risk (continued)
varieties) are less availablein poor areas. In 1994-96 less than 20
percent of all cropland was irrigated in low- and middle-income
countries (only 4 percent in Sub-SaharanAfrica).
Fluctuationsin food pricesare a relatedrisk. Since poorhouse-
holds spenda largepartof their incomeon food, even small price
increasescan severelyaffect food intake. Householdsthat meet
their food needsthrough subsistenceagricultureare less vulner-
ablethan householdsthat have to buy all their food.
Liberalizationof markets often boosts the price of staples-
a benefit to small farmers if they are net sellers of food. Hurt are
the urban poor and the landless rural poor, as net food buyers,
and farmers who engage in seasonal switching, selling food
after the harvest when food is plentiful and cheap and buying it
when it is scarce and expensive (Sinhaand Lipton 1999).Where
transport facilitiesare good,traderscanstep in and equalizeprices
example,job losscan be an individualrisk, or it can be
common to most workers in a country if it is the result
of a macroeconomic crisis. The risk of becoming ill can
be idiosyncratic, or it can have a large common compo-
nent if there is an epidemic. The HIV/AIDS pandemic
is a health risk at the global level,with devastating effects
on poor people and poor countries (box 8.2).
Knowing the source of shocks is important for pre-
venting them, but identify-ingthe source is not always
straightforwvard.Many exogenous events can have simi-
lar effectson household income.A macroeconomicshock,
a hurricane, or a civil war can all lead to severe decline
in income and deplete a household's assets. But how a
shock is transmitted to households is greatly affected by
a countrys institutions. Not everydrought causesfamine,
illness,and death. The effectof a disaster depends on how
well the government functions, whether there is peace or
civilstrife, how well the safetynet and other institutions
include the poor, and so on.
The rvpology can be refined by distinguishing the
severity and frequency of shocks. Consumption smooth-
ing is more difficult with repeated shocks,because house-
holds may have depleted their assets in coping with the
initial shock, leaving them unable to absorb subsequent
shocks.2 And one shock might lead to another. A natural
disaster could wipe out poor people's food supply, leav-
ing them weak and susceptibleto illness.Severitycan range
from catastrophic (a natural disaster, death of the bread-
winner) to minor (aslight illness,a few dayswithout work
for casual laborers).
over the year through arbitrage, but such infrastructure is lack-
ing in many areas.In Madagascarthe meanprice of rice,the main
staple, rose 42 percent and the variance increased 52 percent
after the price liberalization of the 1980s. Two-thirds of rice
farmers were hurt because they consumed more rice than they
produced, and poverty deepened (Barrett 1996, 1998a).
Forthe rural poor, crop diversificationand income diversifica-
tion into nonfarm activities hold the greatest promise for reduc-
ing food price and harvest risks. Reducingconsumption as food
prices rise can have majorand lastingadverse health effects, es-
pecially for children. Successive harvest failures because of in-
sufficient monsoonsin Sri Lankain 1995and 1996led to increased
indebtedness in 80 percent of householdsin eight villages, and
30 percent of householdsreportedincreased incidenceof illness
(Sinhaand Lipton 1999).
The nature and magnitude
of vulnerability
Vulnerabilityaffectseveryone(box8.3). Evenwell-paidcivil
servantsare vulnerable to losing their jobs and sliding into
poverty.For the poor, and for peoplejust abovethe poverty
line, vulnerability is a graver concern because any drop in
income can push them into destitution. As a result, poor
people are highly risk averseand reluctant to engagein the
high-risk, high-return activitiesthat could lift them out of
poverty. One slip could send them deeper into poverty.
Largefluctuationsin income are common for poor peo-
ple.3 For South Indian villagesestimates of the coefficient
of variation of annual income from the main crops range
between 0.37 and 1.1 and are as high as 1.27 for total
farm profits.5 In rural Ethiopia three of four households
suffered a harvest failure over a 20-year period, resulting
in significant fluctuations in farm income.6
Furthermore, because poor people have fewer assets
and less diversified sources of income, these fluctua-
tions affect them more than other groups. In South In-
dian villages an increase in risk (from the monsoon
arriving too soon or too late) reduced farm profits for
the poorest quarter of households by 35 percent but left
the wealthiest farmers nearly unaffectedi. In Vietnam par-
ticipants in the Voicesof the Poor study said of harvest
losses due to floods:
The wealthy can recover losses in one year, but poor
people, who have no money, will never recover.
 HELPING POOR PEOPLE MANAGE RISK 139

Box 8.2
AIDS and poverty

More than 34 million peopleworldwide are infected with HIV, and
more than 18million peoplehavedied of AIDS.More than 90 per-
cent of peopleinfectedwith HIV/AIDSare in the developingworld.
Cross-countryevidenceindicatesthat both low incomeand unequal
distribution of income are strongly associatedwith HIV infection
rates. Countrieswith high gender inequalityalso have higher in-
fection rates.Sub-SaharanAfrica has more casesof existing and
new infectionsthan the rest of the world combined,though the
rateof increaseis now steepest in Asia and in the countries of the
former Soviet Union.
All 20 countries with the highest HIV prevalence are in Sub-
SaharanAfrica. In Botswana and Zimbabwe 1 in 4 adults is in-
fected. In 10 other African countries more than 1 in 10 adults
are infected. The effect on life expectancy will be devastating.
Had AIDS not affected these countries, life expectancy would
have reached 64 years by 2010-15. Instead, it will have re-
gressed to 47 years, reversing the gains of the past 30 years.
The impact on child mortality is also enormous. In Zambia and
Zimbabwe 25 percent more infants are dying than would have
without HIV.
Despite the strong correlation at the country level between
poverty and AIDS, the evidencefor individualsdoes not suggest
that poor people are most likely to be infected. Indeed, early on,
the disease struck mainlythe better-off groups. Evidencefor the
1980s and the first half of the I 990s indicates a positive corre-
lation between HIV infection and education, income, and socio-
economicstatus, probablybecausewealthierand better-educated
people were more likely to have multiple sexual partners. Non-
sexualmodesof transmission-intravenous drug use andmother-
to-child transmission-are associated more with poverty, In
recent years the profile of HIV-infected people has been chang-
ing rapidly, and AIDS is becoming a disease of poor people.
With the more educatedresponding to the information available
on AIDS and adopting protective sexualpractices (condoms),the
share of new infections is rising among low-income and less ed-
ucated people.
With 5 million people becoming infected annually,urgent ac-
tion is needed to stop the spreadof HIV/AIDS.Successfulinter-
vention programs require strong government commitment and
partnershipswith the privatesector, NGOs,and community lead-
ers. Interventionsshown to be effective include conducting pub-
lic informationcampaignsto changeindividualbehaviorand social
norms for sexualcontact; makingcondoms more availableand af-
fordable;providingvoluntarycounseling,testing, andtreatment of
sexuallytransmitted diseases;ensuring a safe blood supply; and
taking measuresto reduce mother-to-childtransmission. In addi-
tion, care activities needto be scaledup to support the vast num-
bers of people infected and affected.
AIDS hasa devastatingimpact on poor people. During the ill-
ness it leadsto loss of laborand causes poor householdsto dis-
pose of productive assets to payfor treatment. The impact of an
adult death from AIDS is more severe in poor households.The
recommended policy approach is to concentrateon poor house-
holds most in need of survivorassistance,focusing on the period
immediatelyafter a death, when food consumption hasfallen but
there has not yet been a permanentlydamaging impact.
The view that HIV/AIDSis a centraldevelopmentissue is em-
bodied in the InternationalPartnershipagainstHIV/AIDSin Africa,
launched in 1999by the cosponsors of the Joint United Nations
Programmeon HIV/AIDS(UNAIDS),including the World Bank.
In collaborationwith African governments, the program aims to
increase resources and technical support, establish targeted
preventionand treatmentefforts, andexpandthe knowledgebase
to assist countries.

Source:AinsworthandSemali1998;Basu1995;Over1998;Rugalema
1999;UNAIDS
2000;WorldBank1997d,1999m
.

Box 8.3
Some key terms: risk, risk exposure,and vulnerability

As traditionallydefinedand measured,povertyisa static concept- being-the risk of becoming ill, or the risk that a drought will
a snapshot in time. But insecurityand vulnerabilityare dynamic- occur,The uncertaintycanpertain to the timing or the magnitude
they describe the response to changes over time. Insecurity is of the event. Forexample,the seasonalfluctuation of farm income
exposureto risk; vulnerability,the resultingpossibility of a decline is an event known in advance,but the severity is not always pre-
in well-being. The event triggering the decline is often referred to dictable.Risk exposuremeasuresthe probabilitythat a certainrisk
as a shock, which canaffect an individual(illness, death),a com- will occur.Vulnerabilitymeasuresthe resilienceagainsta shock-
munity,a region,or evena nation(naturaldisaster,macroeconomic the likelihoodthat a shock will result in a decline in well-being.As
crisis). this chapter explores, vulnerability is primarily a function of a
Risk, risk exposure,and vulnerabilityare related but not syn- household'sasset endowment and insurancemechanisms-and
onymous. Risk refers to uncertainevents that candamage well- of the characteristics(severity, frequency)of the shock.
I40 WORLD DEVELOPMENT REPORT 2000/2001

In China 40 percentof an incomedeclineispassedon Table 8.2

aslowerconsumptionforthe pooresttenth ofhouseholds, Shocksfacedby ruralhouseholds
in Ethiopia
but onlv 10 percent for the richest third of households, Percentage
of households
because they have better access to insurance.8 reporting a hardship
One measure of the vulnerability of the poor and Event episodein past 20 years
near-poor is how often a household fallsbelow the poverty Harvestfailure
line. A studv of seven countries for which panel surveys (drought,flooding) 78
labl' .Policy shock (taxation,
are available found that in six of them the "sometimes forcedlabor) 42
poor group was significantly larger than the always Laborproblems
poor" group.9 A nine-year panel survey of households in (illness, death) 40
South Indian villagesfound that 20 percent of households Oxen problems
were poor in each of the nine years and that only 12per- Other livestockproblems
cent were never poor, with movement in and out of (illness,death) 35
poverty the norm for the vast majority of households.10 Landproblems(land
a) ~~~~~~~expropriation,
reform) 17
These findings show both the high vulnerability and the Asset losses 16
strong resilience of poor households-the ability to es- War 7
cape poverty again after suffering an income shock. Rel- Crime(theft,violence) 3
ative income mobility can be quite large. In South Africa Source:Dercon 1999.
29 percent of households in the poorest quintile moved
up two or more quintiles from 1993 to 1998, while in
Peru 37 percent of households did so between 1985 and taxation,land expropriation, and war (table 8.2). Rainfall-
1990.11 induced income shocks have idiosyncratic components
Another approach is to define long-term poverty as av- of 23 percent, but crop damage from other sources (pests,
erage long-term consumption below the poverty line animals, weeds) have idiosyncratic components of 65-87
and then to ask how much of measured povertv is tran- percent. Income shocks from illnesseshave an even larger
sitory.This approach implicitlyconsidersthe duration and idiosyncratic component.1 6 The cumulation of different
depth of transitions into and out of poverty. By this shocks is a source of significant stress for households:
method about half the estimated poverty in South Indian
villagest 2 and about half the severe poverty in China are As if land shortage is not bad enough, we live a life of
transitory.13 tension wor-rying about the rain: wvill it rain or not? We
Both methods suggest that transitory poverty is a live hour to hour.
large part of total poverty in many settings. Generally, -lt "oITUi kaliima, Fthiopia
households with the fewest assets are most likely to be
chronicallypoor.Education almost alwaysreduceschronic Responsesto risk by households
poverty, but its effects on transitory poverty differ. Bet- and communities
ter educated households in Cote d'Ivoire and Hungary
were found to recoverbetter from downward income fluc- For poor people, dealing successfullywith the range of
tuations, but in China education is not correlated with risks they are exposed to is often a matter of life or death.
transitory poverty. 4 The duration of transitory poverty To manage risks, households and communities rely on
also depends on the frequency of shocks: households are both formal and informal strategies (table 8.3). Informal
more likelyto bounce back from a singleshock than from strategies include arrangements that involve individuals
repeated income shocks.15 or households or such groups as communities or vil-
Vulnerabilityis multidimensional,and poor households lages. Formal arrangements include market-based activ-
face manifold risks, so variations in income and con- ities and publicly provided mechanisms. Informal and
sumption can occur for a varietyof reasons. Rural house- formal strategies are not independent: public policies
holds in Ethiopia, for example, face natural shocks such and the availabilityof formal mechanisms heavily influ-
as harvest failure, health-related shocks such as illness or ence how extensively informal arrangements are used
disability, and macro-level shocks such as the effects of and which kinds are used.
 HELPING POOR PEOPLE MANAGE RISK 141

Table 8.3
Mechanismsfor managing risk

Informalmechanisms Formalmechanisms
Individual Group Market Publicly
Objective andhousehold based based provided
Reducingrisk * Preventive health * Collectiveaction * Sound macroeconomic
practices for infrastructure, policy
* Migration dikes, terraces * Environmentalpolicy
* More secure income * Commonproperty * Educationand training
sources resource policy
management * Publichealth policy
* infrastructure(dams,
roads)
* Active labor market
policies
Mitigatingrisk
Diversification * Crop and plot * Occupational * Savings accounts in * Agriculturalextension
diversification associations financialinstitutions * Liberalizedtrade
* Income source * Rotatingsavings and * Microfinance * Protection of property
diversification credit associations rights
* Investment in physical
and human capital
Insurance * Marriageand * Investment in social Old age annuities * Pension systems
extended family capital Inetworks, * Accident, disability, * Mandated insurance for
* Sharecropper associations, rituals, and other insurance unemployment, illness,
tenancy reciprocalgift giving) disability,and other risks
* Bufferstocks
Copingwith * Sale of assets * Transfersfrom * Sale of financial * Social assistance
shocksa * Loans from money- networksof assets * Workfare
lenders mutual support * Loans from financial * Subsidies
* Childiabor institutions * Social funds
* Reduced food * Cash transfers
consumption
* Seasonal or temporary
migration
Note:The white shaded area shows household and community responses through informalmechanismsto improverisk mitigationand
coping. The dark shaded area shows the publiclyprovided mechanisms for insuring against risk and copingwith shocks-the social safety net.
a. Publiclyprovided copingmechanisms can also serve risk mitigatingpurposes if they are in place on a permanent basis.
Source:Adapted from Holzmannand Jorgensen (2000).

Risk management strategiescan be further classifiedas
risk reduction and mitigation measures (actions in antic-
ipation of a shock) and coping measures (actions in re-
sponse to a shock).t 7 Risk reduction aims at reducing the
probability of a shock or negative fluctuation. Individu-
als or households can sometimes take such action them-
selves (digging wells, getting vaccinated). But to reduce
most risks effectively,action is also needed at the meso or
macro level.The riskof floodingcan be reduced if the com-
munity builds a dike or the government builds a dam.
Sound economic and environmental policies, education
and training, and other measures can also reduce a wide
variety of risks (and are discussedelsewherein the report).
Risk mitigation aims at reducing the impact of shocks.
Householdsmitigaterisk through diversification(acquiring
assets whose returns are not perfectly correlated) and in-
surance.Common diversificationstrategiesareplanting dif-
ferentcropsand plots,combiningfarmand nonfarmincome
in ruralareas,and combiningwageincomeand income frorn
household enterprisesin urban areas.Households can take
most of these actions on their own-though group or gov-
ernmentaction (agriculturalextension,infrastructure)can
sometimes facilitatediversification.Households also miti.-
gate riskthroughinsurance,includingself-insurance,informal
insurance, and formal insurance-though market-based
formal insurance playsa minor role for poor people.
I42 WORLD DEVELOPMENT REPORT 2000/200I

Copingstrategiesaim to relievethe impactof a shock Evidencesuggests,however,that the net effectof these

after it occurs.Actionsby individualsinclude drawing
down savingsor sellingassets,borrowing,and callingon
support networks.Actionsbygovernmentinclude acti-
vating the transfersor workfaremechanismsthat con-
stitute the social safety net. If these measuresprove
insufficient,householdsmay need to reduceconsump-
effortsis limited and that the variabilityof farmers'in-
comeremainshigh. The incomeoptions typicallyopen
to farmerstend to movetogetherduringcrises.Drought,
for example,reducesnonfarm incomeas wellas harvest
incomebecausecrop failureleadsto a generalizeddrop
in incomethat reducesdemandfor nonfarm services. 20

tion or increaselabor supply.Many of these copingre-
sponsesforce a high long-term cost on householdsfor
a short-termbenefit.
This chapterfocusesprimarilyon howto improverisk
mitigation and coping by poor people. It examines
households'and communities'own responsesthrough
informalmechanisms. The chapterthenexploresthe con-
ditions for public action to supplementpoor people's
own risk managementefforts-and the forms this in-
tervention can take. In particular,it discussesthe range
of safety nets that can be used for risk mitigation and
The range of income options availableto farming
households is often quite restricted. Evidence from
Burkina Faso, Ethiopia, India, Kenya, and Tanzania
showsentryconstraints-including lackof workingcap-
ital,skills,andinputs-for manyactivitiesthat couldallow
farmersto diversifytheir incomes.Startup costsfor set-
ting up a shop or providingservicesare often 10-20
timesthe cost of other activitiesthat poor people typi-
callyundertake,suchas charcoalmaking,dung cakecol-
lection,or simplefood processing,activitiesthat provide
only weak income diversification. 2' As a result, poor

coping (seetable 8.3). farmersin Africatend to be lesseffectively

Mitigating risk through diversification
Many studiesdocumenthowhouseholdsthroughoutthe
developing world diversify their income sources to
smooth the flowof income over time." A reviewof 25
diversifiedthan
rich farmers(table8.4).22Poorfarmersin other parts of
the world havehad more successin diversifyingincome
sources.In Pakistan55 percent of farmers'income in
1986-89 came from nonfarm sources,and this share
was threetimesas high for poor as for rich farmers.2 3 In

studiesin Africashowsthat rural householdsreceivean Egypt as well,poor farmerswerefound to be more di-

averageof 45 percent of income from nonfarm activi- versifiedthan rich farmers.24

ties, with the share ranging from 15 to 93 percent.'9 Wherethe possibilitiesfor effectivediversificationare
Farmers also diversifyacross crops and plots and by limited, poor farmers will specializein low-risk,low-
working for other farmers. return activities,makingit hard to escapepoverty.Poor

Table 8.4
Income diversificationamong Africanfarmers

Averageshareof nonfarm Ratioof richfarmers'

incomein total income nonfarmshareto poor
Country Period (percent) farmers'
Botswana 1985-86 77 2.5
Burkina Faso 1981-84 37 2.5
Ethiopia 1989-90 36 1.2
Gambia 1985-86 23 1.3
Malawi 1990-91 34 1.0
Mozambique 1991 15 2.5
Niger 1989-90 52 2.0
Rwanda 1990 30 5.0
Senegal
North 1988-89 60 2.0
Central 1988-90 24 1.0
South 1988-90 41 2.6
Sudan 1988 38 1.0
Zimbabwe 1988-89 42 1.0

Source:Reardon1997.

Indian farmers devote a larger share of land to tradi-
tionalvarietiesof rice and castorthan to high-returnva-
25 Tanzanianfarmerswithout livestockgrowmore
rieties.
sweetpotatoes,a low-risk,low-returncrop,than do farm-
erswho ownlivestock.Asa result,returnsto farmingper
adult householdmemberare 25 percent higher for the
wealthiestgroup than for the poorest. 2 6 Poor farmers
are at a further disadvantagebecauseharvestshocksare
typicallycovariantovera fairlylargearea.This limitsthe
usefulnessof group-basedstrategiesandnetworksof mu-
tual support, becauseall or most group membersare
likelyto be affectedsimultaneously.27
Mitigating risk through insurance
In principle,anyshockwith a probabilitythat can be cal-
culatedfrom historicalrecordsis insurable.In practice,
there are almost no insurance markets in developing
countriesbecauseof problemsof contract enforcement
and asymmetricinformation. People, especiallypoor
people,haveto rely largelyon self-insuranceand infor-
mal insuranceinstead.These problemshavebeen over-
come in developedcountriesthrough strong legaland
other institutions.
Self-insurance.Householdsinsurethemselvesby ac-
cumulating assetsin good times and drawingon them
in bad. The strategyis effectiveif assetsare safeand have
a positiverate of return, especiallyif the rate of return
exceedsthe rate of time preference (of present con-
sumption over future consumption). In practice, re-
turns to assets may be negative, and many poor
householdshaveveryhigh ratesof time preference(they
are "impatient,"often out of necessity),which impedes
assetaccumulation. 28
Another problemis that assetvaluesand income are
often covariantfollowinga macro shock, so that the
valueof assetsis lowestjust when theyare neededmost.
A drought that destroysa harvestmay alsoweakenand
killcattle, whichfarmersin many poor countriesuse as
a bufferstock.The termsof tradeof assetsrelativeto con-
sumption goods may alsodeteriorateas a result of the
shock, as everyonetries to sellassetsand buy staplesat
the same time. Both supply and demand factorspush
downassetprices:the incomeshock induceseveryoneto
sellassets,and the declinein purchasingpowerreduces
demand(unlessbuyersfrom outsidethe shockzoneshow
up). In good timesthe processworksin reverse:every-
one wants to buy the buffer asset,pushingup its price
and making the strategy very costly.29
HELPING POOR PEOPLE MANAGE RISK I43
Simulationswith householdrisk modelssuggestthat
self-insurancequicklyloseseffectivenesswhen the co.r-
relationbetweenincomeand the termsof trade of assets
exceeds0.5. Householdsthen have to curtail the saleof
assetsduring crisesbecausethey gain so little extracon-
sumption in return. During the 1984-85 famine in
Ethiopiaassetterms of trade collapsed,and householcls
cut their consumption drasticallyrather than sell as-
sets.30 During the 1981-85 drought in Burkina Faso
livestocksalescompensatedfor only 15-30 percentof the
shortfallin crop income.31
Buyingand sellingcattle, though a commonstrategy
for copingwith incomefluctuations,is not a feasibleone
for manypoorhouseholds.Buyinga cowrequiresa large,
one-timeoutlay(andsignificantprior saving).In western
Tanzaniaa cowcostsabout a fifth of meanannual crop
income,explainingwhyonlyhalfof householdsowncat-
tle.32 Wherepossible,poorhouseholdsusesmalleranimals
(goats,sheep)or more divisibleitemsas bufferstocks.In
three SouthIndian villagesfarmersheld bufferstocksof
grainsand currencyas their mainriskmanagementstrat-
egy.33 In ruralChina, by contrast,householdsincreased
their holdingsof unproductiveliquidassetsonlyslightly
in responseto incomerisk.34
Becausethe indivisibilityand riskinessof manyassets
(pricerisk, survivalrisk for cattle) limit asset-basedrisk
managenment strategies,poor peopleneeda widerrange
of assetsand greaterstabilityof assetvalues.This would
allowthem to take better advantageof opportunitiesfor
incomegrowth(describedin part II of this report).Sav-
ings accountshold greatpromiseas a divisibleassetwith
a fixedvalueand positivereturn.Givensomeassurances
about the safetyof the financialinstitution holding the
accounts,the mainriskwouldbe inflation.Severalrecent
experienceshaveunderscoredthe greatdemand bypoor
householdsfor safesavingsaccounts.Bank RakyatIn-
donesiahas morethan 16millionlow-incomedepositors.
SafeSave,an NGO in Dhaka, Bangladesh,has adapted
the principlesof a traditionalrotatingsavingsand credit
association;its agentscollectsmallsums of moneydaily
for depositin members'accounts. 35
Informalinsurance.Householdsalsouse group-based
mechanismsof informalrisk sharingthat relyon the so-
cialcapitalof groupsof households.Typically,informal
insuranceinvolvesa mutual support networkof mem--
bersofa communityor extendedhousehold,oftenwithin
ethnicgroups;amongmembersof the sameoccupation;
or between migrants and their households of origin.
144 WORLD DEVELOPMENT REPORT 2000/2001

Like consumption smoothing, which aims to equal-
ize marginal utilitiesover time, group-basedinsuranceaims
to equalize marginal utilities across members of the
group.3 6 When one member's consumption falls, the
others transfer resources to rebalance marginal utilities.
These nerworks are effectiveonly against shocks common
to some members but not all. So the wider the group, the
less likely a shock is to affect all members, and the more
effective the}' all are at risk pooling.37
A network operates through transfers, gifts, or loans
between members, typically with expectations of reci-
procitv. Transfers respond to an emergency befalling a
member of the network, thus serving risk management
purposes, but they also fulfill a social function in forg-
ing community cohesion.3 8 The importance of gifts and
uations of high economic stress, norms and social pres-
sure may not be enough to ensure that members of the
group do in fact transfer resourcesto other members. In-
formal insuranceworks bestwhere people value future pro-
tection highly (rates of time preference are low) and fear
of future exclusionfrom the insurance scheme keepscom-
pliancehigh. But this works againstpoor people, who tend
to value current consumption highly relativeto future con-
sumption (usuallyout of necessity).For this reason, poor
people, even though they need insurance most, are more
likely to drop out of informal arrangements. Informal in-
surance alsoworks better when the rate of transfersis high
(becausefrequent interactions create trust in future com-
pliance) and shocks are idiosyncratic (because covariant
shocks can wipe out the entire network's resources). 39

transfers varies greatly. In Bulgaria fewer than a fifth of To determine the need for a formal safetv net, re-
households receive transfers; in Jamaica more than half searchershave tried to measure how well informal insur-
do (table8.5). In most countries the bulk of transfers goes ance works, but measurement has proved difficult. It is
to the poorest households, often representing a large hard to distinguish between the effects of informal in-
shareof income. Privatetransfersincreasethe poorest quin- surance and those of self-insurance. And because mea-
tile's share of aggregate income by about 50 percent in surement requires information about consumption and
Jamaica and Nepal and by almost 70 percent in Russia trends for all members (or a statistically valid sample of
(figure 8. 1). them), it is especiallydifficultwhen a networkextends past
The occurrence of transfers is not alwaysa sign of ad-
equateprotection againstcrises.The kev featureof informal Figure 8.1
insurance is reciprocity,self-enforcedby the group. In sit- Private transfersrepresent a large share of the
income of the poor
Table 8.5
Private cash and in-kind transfersfor poor Income share of poorest quintile, mid-1990s
households 8
Percent
After private transfers
Share
Shareof receivingtransfers 6
households All Poor Before private
Country giving house- house- 5 transfers -X
8
Iyear) transfers holds holds
4
Jamaica (1997) 13.1 53.0 65.0
Nepal (1996) 17.4 44.7 55.3 3
Peru (1994) 14.3 37.3 46.7
Panama (1997) 15.5 37.8 40.9 2
Kazakhstan
(1996) 20.2 27.5 33.8 1* *
Kyrgyz Republic 3 0
(19961 15.7 35.5 31.7 0
Russian 'A . ; ' Q
Federation P" 5' / '

(1997) 23.7 25.2 31.5 11,P

Bulgaria (1995) 15.0 17.0 21.4
Note: Quintiles are based on per capita income distribution.
a. Households in the lowest quintile of the per capita income Transfers are those accruing to the poor. Data are for most recent
distribution. year available.
Source: Cox, Galasso, and Jimenez 2000. Source: Cox, Galasso, and Jimenez 2000.

the boundariesof a villageor other geographicalentity.
EvidencefromC6ted'Ivoire,India,Thailand,and Uganda
suggeststhat informalinsuranceexists,but is farfromper-
fect.40 Evidencefrom China and India indicatesthat the
poorand the landlessaremuch lessprotectedfrom income
fluctuationsthan the rich and the largelandholders.41
Coping with shocks
When a shock hits, peoplecopeby cashingin their in-
surance:sellinglivestockor otherassets,or callingon sup-
port networksfor transfersor loans.If thesemechanisms
fail or fall short, householdsmay increasetheir labor
supply,workingmore hours, involvingmore household
members(womenor children),or migratingto unaffected
areas.If allelsefails,householdsreduceconsumptionand
go hungry. 42
The poorhavefeweroptionsthan the wealthyforcop-
ingwith shocks.Becausetheyown fewerphysicalassets,
poor people are more likelyto increasetheir labor sup-
ply. If the shock is covariantand the locallabor market
has collapsed,migrationis the only answer.And if the
men in the household migrate, women and children
may need to pick up the slack locally.4 3
Copingwith shocksoftenrequiresmorethantheseeco-
nomic responses.During a prolongedcrisispeoplemay
delaymarriageand childbearing,familiesmav movein
together(especiallyin urban areas),and peoplemay re-
sort to illegalactivities(theft, robbery,prostitution).Ul-
timately,the socialmechanismsmeantto helphouseholds
copewith shocksmay comeundone under the contin-
uing pressureof a prolongedshock.
Effectswithin households
So far,the discussionof risk managementhas viewedthe
householdasthe unit of impactand decision.Yetriskshar-
ing within the householdmay not be equal, and the
burden of the household'sresponsemay fall dispropor-
tionately on the weakestmembers, especiallywomen
andchildren.Twosituationsarepossible.A shockaffecting
the householdas a wholemay havedifferenteffectson
differenthouseholdmembers.Or an individualshock(ill-
ness, loss of job) may have different effects on con-
sumption dependingon whether the affectedperson is
a man or a woman.There is more evidenceon the first
situationthan the second.44
Becausepoor householdstend to have many chil-
dren, childrenare more exposedto povertyand vulner-
abilitythan other groups.Children in poor households
HELPING POOR PEOPLE MANAGE RISK I45
are especiallyvulnerableto fluctuationsin householdin-
comeandconsumption.They are more likelythan other
childrento be underweight, sothat furtherdeclinesin food
consumptioncan causeirreversibleharm. In Bangladesh
children'sgrowthsufferedduringmajorfloods. 4 5 In rurao
India childmortalityratesincreasedin timesof drought,
especiallyin landlesshouseholds. 46
The evidenceon genderbiasin the responsesto such1
shocks is mixed. No such bias was found following
floodsin Bangladesh. 47 Studiesin India, however,found
that girls' nutrition sufferedmore than boys'duringpe-
riods of low consumptionin the slack agriculturalsea-
son.48 Price changes also were found to affect girls'
consumptionmore than boys'.49 For childrenunderthe
age of two, rainfallshortageswere associatedwith more
deaths for girls than for boys.50
Some studies have found that women suffer more
than men from adverseshocks.Risingfoodpricesled to
largerreductionsin nutrient intakefor women than for
men in Ethiopiaand India.51 Culturalandtraditionalfac-
tors can increasewomen'sexposureto risk. Divorced
and widowedwomen in South Asia often face higher
health risks and are more likely than married women to
be poor becausethey loseaccessto theirhusband'sprop-
erty.52 In someAfricancountrieswomen mayloseaccess
to householdland when their husband dies.53 There is
alsoevidenceof a pro-malebiasin householdhealthand
nutritionexpenditures,but it is not clearwhetherthe bias
affectspoor householdsmore than others.A recentstudy
in Pakistanfound limited evidencethat genderbias in
health expendituresdecreaseswith risingincome. 54
On balance,the evidencepointsto importantdiffer-
encesin intrahouseholdeffectsfrom shocks.But the ev-
idencecomesmainly from SouthAsia.Whethersimilar
effectsoccur elsewhereis still unknown.
The poverty trap and the long-term
consequencesof inadequate risk management
As householdsmovecloserto extremepovertyand des-
titution,theybecomeveryriskaverse:anydropin income
could push them belowthe survivalpoint. The poorest
householdstry to avoidthis evenif it meansforgoinga
largefuture gain in income. Despite facingthe highest
risk,theyhavethe fewestresourcesfor dealingwith that
risk.Andforcedonto the mostmarginallands(floodplains,
hillsides)and into areaswith poor infrastructure,theyare
most at risk from natural disastersand usuallyfar frorn
health facilities.
I46 WORLD DEVELOPMENT REPORT 2000/200I
Extreme poverty deprives people of almost all means
of managing risk by themselves. With few or no assets,
self-insurance is impossible. With poor health and bad
nutrition, working more or sending more household
members to work is difficult. And with high default
risks, group insurance mechanisms are often closed off.
The poorest households thus face extremely unfavor-
able tradeoffs. When a shock occurs, they must obtain
immediate increases in income or cut spending, but in
so doing they incur a high long-term cost by jeopardiz-
ing their economic and human development prospects.
These are the situations that lead to child labor and mal-
nourishment, with lasting damage to children, and the
breakdown of families.
In C6te d'Ivoire severeeconomicrecessioncaused
households, especiallythe poorest, to sharply increasethe
labor supply of children.55 In rural India child labor was
found to play a significant role in households' response
to seasonalvariationsin household income.5 6 In everypart
of the world participantsin the Voicesofthe Poorstudy men-
tioned child labor as an undesirable coping mechanism.
In Egypt children were sent to work in a storehousepack-
ing vegetables.During periods of drought in Ethiopia chil-
dren were taken out of school and sent to towns to be
employed as servants,with their earningssent back to their
families. In the lean season in Bangladesh children work
on farms, tend cattle, or carry out household tasks in ex-
change for food. Parents are often aggrievedby the undue
physicallabor of their children and worry especiallyabout
the vulnerability of girls to beatings and sexual assaults.57
Inadequate risk management can also compromise
nutrition in poor households. After the devastatingfloods
of 1988 in Bangladesh, many households took out loans
to meet consumption needs, but landless households
were less able to do so and their children suffered more
severe malnutrition. 58 A study of rural Zimbabwe found
that the 1994-95 drought caused a 1.5-2-centimeter
decline in annual growth among children one to tvo years
old. Although this study found the reduction to be per-
manent, other studies have found evidence of catch-up
during subsequent good periods.5 9
What do households sufferingthese unfavorablelong-
term effects on the education and nutrition of their chil-
dren have in common? Low asset endowments (physical,
human, social) and little or no access to credit and in-
surance markets-a chronic trap for poor people, unable
to accumulate enough assets to escape poverty. When
households do not havesome threshold of assets,they are
forced to engage in defensive actions to protect the as-
sets they do have. One study estimated that poor house-
holds engaging in this strategy could have boosted their
incomes by 18 percent with a more entrepreneurial man-
agement strategy (but one that requiresaccessto credit).6 0
Dysfunctional factor markets can also create or ag-
gravatepoverty traps.Takechild labor.When a crisisstrikes
and households cannot borrow or when adult unem-
ployment is high or wages low, children are pulled out
of school and sent to work. The lost schooling leads to
a lifelong loss in earning ability for these children. Fail-
ures in the credit or labor markets thus transmit poverty
and vulnerability across generations.
Policy responses for improving
risk management
Since poor people cannot fully manage risk on their
own, any poverty reduction strategy needs to improve risk
management for the poor-reducing and mitigating risk
and coping with shocks. The strategy should include
formal and informal mechanisms, provided by both the
public and the private sector.
In principle and excludingcost considerations,the best
approach is to reduce the risk of harmful shocks.62 Next
would be risk mitigation to reduce the possible impact
of a shock. Coping would be a residual approach to ad-
dress the failures of the first two.63 In practice, different
direct and opportunity costs may well change the rank-
ing of options. Some risk reduction and mitigation strate-
gies are prohibitively expensive, especially those for
dealing with infrequent but catastrophic shocks.
Comparative cost data and cost-benefit analyses are
generallynot availableto help policymakerschoose from
different types of risk management interventions. Fur-
thermore, the distributional implications of different
strategies need to be considered. A comparative study in
India found that, at the margin, publicwork programsben-
efit the poorest quintile the most, while credit programs
benefit the second and third poorest quintiles the most.64
Most developing countries pay too little attention to
risk reduction and mitigation and rely too much on in-
terventions after disaster strikes. Efforts to cope with the
Mexican peso crisis of 1995 and the East Asian financial
crisisof 1997 have shown how difficultit is to put effective
safety nets in place after the fact (chapter 9).
The balance needs to shift from policies for coping
to those for reducing and mitigating risk. That means

ensuring that social safetynets such as workfareprograms,
targeted human development programs, and social funds
are in place on a permanent basis and can be scaled up
when a shock occurs (see table 8.3). Interventions fol-
lowing the 1998 floods in Bangladesh were effectivebe-
caue
N o Osnd etwrk f
th exstig thr
causeofnetwork
the existing
of NGOs and other
mechanisms ready to be activated to help poor peo-
ple.65 It also means providing better access to credit and
financial assets,facilitating income diversification, man-
aging labor market risk better (especiallychild labor), and
providinghealth insurance.Such actionswould allow
poor people to pursue higher-risk, higher-return activ-
itiesthat could pull them out of poverty.66 Social safety
nets can also serve as an automatic compensatory mech-
anism for the unwanteddistributionalefernment
anism for the unwanteddistributionaleffectsof policy
reforms (chapter 4). By doing so, they will help make
reform sociallyand politically feasible.While a new bal-
ance is needed,copingmechanisms willremainvital
for disngewit unfoe see enshc where
rand sinfeq
for dealingwithunforeseen
andinfrequent shockswhere
it is prohibitively expensive to put mechanisms in place
ahead of time.
Not every country needsto set up a comprehensive so-
cial safety net. But each does need to construct a mod-
r . r . .
ular system of programs based on its own patterns of risk
and to cultivate a suitable mix of providers (public and
private) and administrative arrangements (box 8.4). The
first step in selectingand designingprogramsis to un-
derstandthe generalprinciplesof how safetynets com-
plement existingriskmanagement arrangements.The next
is to identifv specific types of risk (illness, old age, un-
and the mechanisms for dealing with them.
employment)
Generalprinciples of safety nets
and risk management
Reducing risk is possible for some categories of risk but
not all. Forexample,buildinga dam can reducethe risk
of flooding. Immunizations and other public health cam-
paignscan reducethe risk of illness.Policiesundertaken
primarily for other purposes can also contribute to risk
reduction. Good education policies, including scholar-
ships for poor families, can reduce child labor. Environ-
mental policiescan limit deforestation, reducing damage
from hurricanes and deaths from mudslides. Sound
macroeconomic policies can reduce the risks of high in-
flation and unemployment.
But the focus in this chapter is primarily on mitigat-
ing risk (diversification and insurance) and on coping.
Making a wider variety of crops and extension services
HELPING POOR PEOPLE MANAGE RISK 147
Box 8.4
Managing risk:the modular approachto social
safety nets
Constructinga socialsafetynetis farfroman exactscience,
andthe processwillvary from countryto country depend-
ingonthe context,dataavailability,
andpolitical
urgency.But
theprocessshouldhavecertainanalyticelements,including
establishing
thecountrycontext,constraints,andchallenges;
identifying
sourcesof risk,vulnerable groups,andpotential
interventions;and identifyingthe optimalmixof programs.
Malawiillustratesthe mixof preferredprogramsthat can re-
sult, depending on prevailingconditions.
Malawiis a low-income country,withmorethanhalfits
population inseverepoverty.Thevastmajorityof the pop-
ulationdependsonsubsistenceagriculture.
Thereislittlegov-
revenue surplus to redistribute and limited
administrativecapacityto managecomplexprograms.There
is noformalsocialsafetynet.
Identifying of riskandvulnerablegroups
sources
VulnerablegroupsinMalawiwere identifiedonthe basisof
a povertyanalysisconductedintheearly1990s.Fourgroups
werefoundto be mostat risk:ruralhouseholdswith small
landholdings,female-headedhouseholds,AIDSorphansand
theirrelatives,
andthosewhocouldnotcareforthemselves.
In addition,four majorriskswere identified:seasonal price
icessadfo hrae,proi ruh,lrepn
~~~~~~~~~~~Increases
andfood shortages,periodicdrought,largeperi-
odicmacroeconomic shocks,andthethreatofHIV/AIDS. Po-
tentialinterventions
toaddressthese risksweredeveloped.
intervention$
Acost-effectivenessanalysisofexistingprogramswas con-
ductedbeforepotentialnew interventions were rankedby
priority.
The results,togetherwith considerationof the vul-
nerable groups,the risks, and the need to focus on
productivity-enhancinginterventions,led to the following
modularsystem of programs:
* Publicwork (riskmitigationand coping).
* Transfersfororphansinpoorcommunities(riskmitiga-
tionandcoping).
* Nationwidenutritionprogram(riskreductionand coping).
* Targeted cash transfers to the needy (coping).
Source:World Bankforthcoming b.
available to farmers can help rural residents to diversify.
Opening trading opportunities through investments itn
infrastructure and other means can also stimulate diver-
sification.But liberalizingmarkets (say,by privatizingstate
commodity boards) can have mixed effects and will nolt
alwaysbenefit poor people. Sometimes dealersstep in be-
tween farmers and export traders and capture most of the
gains from open trade.
I48 WORLD DEVELOPMENT REPORT 2000/2001
Policies should also make it easier for poor people to
build up assetswhile reducingthe covariancebetveen asset
values and income. Covariance is a big problem in rural
areas,where asset values (livestock)often move in tandem
with farm income. This could be addressed through bet-
ter integration of asset markets with the wider economv-
by investingin transport infrastructure,disseminatingprice
information, and removing structural and institutional
market barriers. Macroeconomic stability promotes more
stable asset prices, reducing inflation-driven deterioration
in the terms of trade of assets relative to consumption
goods. And easier access to credit would facilitate the ac-
quisition of costly indivisible assets, such as cattle.67
Another critical intervention is the provision of in-
surance, especially for covariant risk. Self-insurance has
limits, mainly because poor people cannot accumulate
enough assets,especiallyafter successiveshocks. And in-
formal insurance, which relies on risk sharing across a
community or network, is ineffective for covariant
shocks.
The first question with insurance is whether market
or government provision is more cost-effective than
informal mechanisms. Can the state provide less costly
insurance for risks that are self-insured by poor people
or insured through group-based risk sharing?Becausethe
Figure 8.2
Central government spendingon socialsecurity varied greatly in 1995
Percent
35
30
25
Share of government
20 i . expenditure
15
10 t t | I t V t I 0
5 10~~~ l |
Source:IMF 1995.
public sector can pool risks over a larger area, the possi-
biliry exists for providing insurance at a lower cost than
informal agents can (assumingthat information problems
can be dealt with; see below).Publicly provided insurance
could thus yield a net gain to society-if the state is per-
ceivedas credible and the insurance scheme is fiscallysus-
tainable.
But if trust in the state is low, few people will put their
faith in the government systemand giveup their personal
or group insurance. And even if credibility is not an
issue, fiscal constraints may prevent the state from mak-
ing payments during a crisis. People who had given up
their informal insurance mechanisms would then be left
worse off than before the state offered insurance. Rela-
tive cost-effectiveness, trust, and sustainabilitythus all need
to be consideredin deciding on government intervention.
Government spending on socialsafety nets variescon-
siderably.Figure 8.2 illustratesthis with one component:
spending on socialsecurity by the central government. But
costs are only part of the picture. These expenditures are
also investments in human capital formation. By provid-
ing poor people accessto basic servicesand allowing them
to undertake higher-risk, higher-return activities, the in-
vestments can have positive effects on poverty and eco-
nomic development. Costs are still likely to be an issue,
n Share
| * 7 l '_0_ of GNP]

but theycan often be reducedby more explicitlyconsid-
ering the roleof privateproviders. 68
Concernsthat formalsafetynetswilldisplaceself-in-
surance or group-based mechanismsalso need to be
considered.Empiricalestimatesof thiseffectvary,sothe
country context is important. A study of the urban
Philippines estimated that government-providedun-
employment insurance would displace 91 percent of
privatetransfersto the unemployed. 69 Anotherstudyes-
timated that providinga basicpension benefit to black
SouthAfricansdisplacedonly 20-40 percent of private
transfersto the elderly. 7 0 Studiesof other countriesalso
found displacement rateson the orderof 20-40 percent.71
Displacements of privatetransfersneednot implya so-
cialloss.Ifpovertyreductionobjectives areconsideredalong
with insuranceobjectives,theremaywellbe a net social
gain,despitethe displacements. 72 In South Africamany
of the displacedtransferswerefromyoung to oldhouse-
holds, both of them poor.The new pensionprogramleft
more moneyin the pocketsof pooryounghouseholdsand
alsocoveredmanyelderlyresidentswhohad not beenre-
ceivingprivatetransfers.Overall,then,the pensionscheme
significantly strengthenedSouthAfrica'ssocialsafetynet.73
When should the state step in and provide a social
safetynet forpoor people-and how?The generalanswer
is that it dependson the types of shockslikelyto occur
and the kindsof privateinsurancearrangementsin place.
* If informalarrangements insureadequatelyagainstidio-
syncraticrisk,the stateshouldstepin to insureagainst
covariantrisk. In most circumstancesprovidingthis
coverage willimproveoverallriskmanagementand in-
creasewelfare,without crowdingout informalinsur-
ance.74 But since households'overall risk exposure
will havedeclined,self-insurance(precautionarysav-
ings or other assetbuildups)maydecline. 75
* Whereinformalinsuranceis ineffective-because of en-
forcementproblemsor becauseshocksaretoo frequent
or too large-household welfarecould be increasedif
the socialsafetynet insuredagainstboth idiosyncratic
and covariantrisks.Whether coverageshould come
fromthe stateor privateinsurersdependslargelyon the
typeof risk.Thestateisoftenbestableto covercovariant
risks,but most idiosyncraticrisksmaybe betterhan-
dled by private providers(communities,insurance
firms).The government'sroleshouldthen be to facil-
itate and, if necessary, 76
regulateprivateprovision.
* Wheregroup-basedinformalinsuranceworkswell,the
stateshould avoidsafetynet programstargetedto in-
HELPING POOR PEOPLE MANAGE RISK 149
dividualsor households.Most safetynets target spe-
cific types of people or households:the ill, the el-
derly, the women heading householdswith many
children,and the like. The dangeris that improving
the riskposition of one personbelongingto a group-
basedinsuranceschemecreatesan incentiveto drop
out of the group. If this leads to the collapseof the
group scheme,members not coveredby the safety
net could end up worseoff.The solutionis to target
broad groups (say,a credit program for the entire
community or specificgroupswithin it), although
doingso canbe difficultbecauseinsurancegroupsdo
not alwayscoincidewith communitiesor other eas-
ily identifiabletargetgroups.Of course,if the safety
net protectsalmost everyone,the disappearanceof
informalinsurancearrangementsmay not matter,aLt
leastifthe formalsafetynet is more cost-effective and
7
sustainable. 7
In the end, decisionson safetynets needto weighthe
negativeeffectsof displacementagainstthe positiveef-
fectsof long-termimprovementsin the welfareof poor
households.Safetynets are not the only wayto improve
poor households'abilityto managerisk and to engagein
higher-risk,higher-returnactivities.Stable macroeco-
nomic policiesmaydo more to reduceemploymentrisk
than publicwork programs.But sound economicpoli-
ciesmay increasethe risk for somecategoriesof house-
holds.Tradeliberalizationmaylowerthe costof imported
clothesand utensils,reducingdemandfor weavingand
handicrafts-two activitieswith low entry costs fre-
quentlyusedbypoor peopleto diversifyincome. 7 8 Sothe
decisionabout providingsafetynets needsto be viewed
in the full contextof economicand socialpoliciesand of
the impactson householdrisk.
Wherethereis a strongconcernforthe poor,especially
theverypoor,the formal-informal, public-private balance
generallyshifts in favor of public provision of insur-
ance.79 Concernsfor sustainabilityand other incentives
in group-basedinsuranceand credit schemesgenerally
work againstinclusionof the poorest,who havea higher
perceivedrisk of default.80 Similarconcernstend to ex-
cludepoor people from market-basedinsurance.Thus
publicinsuranceprovisionis not likelyto undercut any
informalarrangementsinvolvingthe pooresthouseholds.
Furtherstrengtheningthe caseforpublicintervention
is the ineffectiveness of the insurancemechanismsused
bypoorpeopleagainstrepeatedshocks-mechanismsthat
alsotend to be costly. 81 A study of six South Indian vil-
150 WORLD DEVELOPMENT REPORT 2000/2001
lages found that farmers sacrifice as much as 25 percent
of average income to reduce exposure to harvest risk.82
Severalpractical issueshave to be considered in setting
up state insurance programs. These include obtaining in-
formation about the people to be insured and dealing with
the political economy issuesin providing insurance.
Obtaining information about people to be insured is
costly.That is why so many traditional credit and insur-
ance institutions are local. Moneylenders or members of
a rotating savings and credit association have a better
chance of knowing who is a bad risk than would an out-
side insuranceprogram.8" Asymmetricinformation creates
problems of moral hazard and adverse selection, leading
to the underprovision of insurance (relative to the social
optimum) by private providers.84 Because information
problemsare especiallyacutefor poor people,the socialgains
from government provision of insurance may be large.
Because the government has no comparative advan-
tage in obtaining local information on who should be in-
Box8.5
Istargetingby the communitya goodidea?
Inmost socialsafetynet programsthe centralgovernmentprovides
funds and sets the eligibility
criteria,ostensiblyguaranteeingequal
treatmentacross the country.Butlocalneeds may varyacross the
country,andbenefits mayleakto ineligible householdsinvaryingde-
grees. Inan effortto improvetargeting,an increasingnumberof pro-
grams relyoncommunitiesto determineeligibility rulesandidentify
beneficiaries.The success of this approachdepends in parton the
degree of socialcohesionin the communityand whetherthe com-
munitycan be effectivelymobilizedina consultativeprocess to al-
locate benefits.
Targetingefficiencyalso depends on the entitychargedwithal-
locatingbenefits.In Uzbekistanquasi-religious communitygroups
knownas maha/las targetchildbenefitsand othertypes of socialas-
sistance to low-incomefamilies.Theyhaveconsiderablediscretion
overamountsandcriteriaforassistance.Anextemalreviewconcluded
that benefits were targeted fairlywell. In Armeniasubsidies for
Advantagesanddisadvantages
of allowingcommunities
to allocatebenefits
Advantages
Better information is availableon needy households
Allocationcriteriaare adaptedto local needs
Decentralized administrationis more efficient
Communitymobilizationmay buildsocialcapital
Source:Conning
andKevane1999.
sured, coproduction is frequently recommended: the
government provides the financial and technical means,
and local institutions or peer groups take care of imple-
mentation and monitoring. Or the government provides
funds to communities, which are responsible for identi-
fying poor beneficiaries (box 8.5).
The political economy may strengthen or weaken the
case for publicly provided risk mitigation. The state may
well be the best agent to provide insurance, but lack the
necessary institutional strength, financial resources, or
management capacity.Capacitv building may then be re-
quired inside the government. The political support to
allocate resources may also be lacking, since it requires
getting the rich to support a program that does not ben-
efit them. If the insurance program is not self-support-
ing, it may have to be funded out of general tax revenue,
at the expense of other programs that benefit the rich.
(Chapter 6 discussesthe political economy of poverty re-
duction further.)
school textbooks are allocatedlocallyby parent-teacherassocia-
tionsor the schoolprincipal.Theprogramhas not been formallyeval-
uated,but informalappraisalssuggestthat the systemhas beenwell
accepted by parents,and itmay be expandedto other types of aid.
The KecamatanDevelopment Project in Indonesiaprovides
blockgrants to 10,000villages.Each communitydecides on the
use of the funds through an extensive process of information
dissemination,communityfacilitation,and proposalpreparationand
selection.Fieldassessments indicatethat the process worksbest
when bothtraditionalandofficialcommunityleadersare on board
from the start (KDPSecretariat 1999).
In Albaniathe EconomicSupport Programhelps poor rural
householdsand peoplewho lose their jobs inthe transition.Local
governments receive blockgrants to allocatewithintheir com-
munes. Localtargetingcompares favorablywith that of safety net
programs inother countries.
Disadvantages
Program may be used to serve interests of the elite
Participationof communityleadersmayhave opportunity
cost
Allocationrules maycauseincreaseddivisivenessin the
community
Externalitiesacrosscommunitiesmaynot be takeninto
account
 HELPING POOR PEOPLE MANAGE RISK I5I

Box8.6
Insuranceoptionsfor the informal sector

Thesimpleexpansion of statutorycoverage of formalsectorsocial deliveryof welfareservicesfromindividual employee-employer re-

insurance programs (pensions, unemployment insurance, disability lationships,removinga majorbottleneckto includinginformal
insurance) to smallenterpriseswill not meetthe riskmanagement sectorworkersin contributory schemes.
needsof the informalsector.Schemes needto accommodate the SeveralIndianstatesrecentlytrieda moreexperimental group
lowercontributory capacityandgreaterearnings volatilityof self- insurance scheme.In Gujaratabout20,000landlessagricultural
employed andinformalworkers. laborersreceivedlife and accidentinsurancecoverage.Most
Linesmayneedto beblurredbetweenstrictcontributory self- groupinsuranceschemesarenot fullyself-financed andrequire
financedinsuranceschemesandassistance paidout of general stategovernment contributions.
taxresources. Alsocalledfor areflexiblepartnershipsbetweendif- In surveysinformalsectorworkersregularly singleout health
ferentproviders:the state,privateinsurancecompanies,com- insurance astheirgreatestinsurance need.InChinaruralhealthin-
munities,NGOs,and organizations representingthe informal surancecovershospitalandprimaryhealthcarecoststhroughpri-
sectorworkers. vateandpubliccontributions. Premiums paidby beneficiaries
are
Manycontributory, andoften self-managed, schemesfor in- supplemented throughavillage publicwelfarefundandgovernment
formalsectorworkershaveemergedinrecentyears.Theyareei- subsidies.
theradaptedfromformalsystemsorbasedoncooperatives and InTanzania a pilotprojectin Dares Salaam provideshealthin-
mutual benefit societies.In some cases they complement surancethroughfivemutualassociations of informalsectorwork-
indigenous riskmanagement arrangements, suchasburialandro- ers.In Igunga,a townin the northwest,a communityhealthfund
tatingsavingsandcreditsocieties. coveringprimaryhealthcareserviceshasachieved 50percentpar-
In the Indianstateof AndhraPradeshpensioncoveragehas ticipation. Sincethe scheme reliesonmatching funds,sustainability
beengivento about425,000 home-based workersinthe beedi(leaf- remainsan issue.
rolledcigarettes)industry,underthe Employees'ProvidentFund Keyto thesuccessof contributory insurance schemesfor the
Act.A simpleprocedural mechanism-issuing identitycards-was informalsectoraretheirorganization aroundan association based
crucialto success.In addition,a welfarefund for beediworkers ontrustandmutualsupport(professional group,village)
andthead-
was set up by the centralgovernment,fundedthroughan ear- ministrative capacity to collectcontributions andprovidebenefits.
markedtaxcollectedfrom employersandmanufacturers in the Administrative capacitycanbefortifiedbyanumbrella organization
beediindustry.Thisdelinksthe collection of contributions andthe thatjoinslocalgroupsandprovidesthemwith technical support.

Source:van Ginneken1999.

Public risk mitigation may also reduce profit oppor-
tunities for the rich (from money lending) or undercut
patron-client relationships between rich and poor by
making the poor more independent. 8 5 Allowing the rich
(or at least the middle class) to participate in some insurance
programs and showing that insurance is less costly than
other poverty reduction efforts can boost political support
for publicly funded insurance. Above all, as chapters 5 and
6 argue, public risk mitigation will succeed only if poor
people have a channel for dialogue with government on
issues of risk and vulnerability.
Special considerations stem from the large (and grow-
ing) informal sectors in many developing countries (box
8.6). Employment in the informal sector in 12 Latin
American countries rose from 50 percent of the eco-
nomically activepopulation to 54 percent between 1990
and 1997.86A large "gray" economy has sprung up dur-
ing the economictransition in EasternEurope and the for-
mer Soviet Union. Unemployment insurance will not
reach workers in the informal sector, but social assistance
programs can. Community-based and integrated provi-
sion of insurance are two innovative approaches showirLg
promise. Such programs recognize the strong links between
labor market risks and other risks in the informal econ-
omy.87 Packages that combine different types of insurance
or assistance for the self-employed may be particularly at-
tractive. In Chile many self-employed people participate
in the pension system to get health insurance.8 8
Specific instruments and the lessons
of experience
While the general principles discussed here are useful in
framing choices for policymakers, it is their application
to specific casesand the lessons of experience that really
matter. Many tools are availablefor public interventions
to improve the ability of households to manage risk. Tlhe
rest of this chapter covers seven tools especiallyrelevant
for poor people: health insurance, old age assistance and
pensions, unemployment insurance and assistance,work-
fare programs, social funds, microfinance programs, and
152 WORLD DEVELOPMENT REPORT 2000/2001
cash transfers.Some of these instruments addressprimarily
one type of risk-others are usefulfor a wide range of risks.
Health insurance. Several studies have shown that
many households in developing countries cannot insure
against major illness or disability. Significant economic
costs are associatedwith these conditions, both in income
losses and in medical expenses. The policy response
should be to provide health insurance and to direct pub-
lic health spending to facilitiesservingprimarily poor peo-
ple (or poor areas).8'9
Some middle-income countries have set up universal
health insurance, as Costa Rica and the Republic of
Korea did in the 1980s. But most low-income countries
can offer only limited health insurance, usually provid-
ing minimum benefits for all illnesses ("first dollar cov-
erage") rather than full insurance for infrequent but very
costly illnesses.'0 This choice may look pro-poor (bene-
fits are provided regardless of income and there is no de-
ductible or copayment), but the evidence suggests that
catastrophic illnessesand disabling injuries create much
greater problems for poor people than frequent, minor
illnesses. Households in Indonesia were able to smooth
more than 70 percent of consumption fluctuations caused
by moderate health shocks, but only 40 percent of those
caused by large health shocks.9' An averagehospital stay
in Indonesia costs 131 percent of the annual income of
the poorest quintile of households, but only 24 percent
of the income of the richest quintile."2 In China house-
holds could smooth only 6 percent of consumption fluc-
tuations caused by overall medical care costs, but 100
percent of fluctuations involving health care expenses of
less than 50 yuan.93
Public provision of insurance against catastrophic
health risks could thus significantly improve the welfare
of poor people where households are unable to insure
against these risks themselves.The evidence further sug-
gests that premiums can be quite low (because major ill-
ness is rare) and well below households' willingness to
pay.94 Countries as diverse as Costa Rica and Singapore
have implemented health insurance schemes with near
universal coverage (box 8.7). Where administrative ca-
pacity or other constraints make catastrophic health in-
surance infeasible for poor people, subsidies for hospital
care can be used instead. For this to be pro-poor, how-
ever, there must be equity in referrals and access to hos-
pitals.95In both approachesthe objectiveis to avoid a need
for poor people to pay for medical emergencies through
debt, distress sales of assets, or cuts in consumption.
Injuries and chronic illnessesthat result in long-term
disability affect an estimated 5-10 percent of people in
developing countries.96 Disability is associated with low
education, poor nutrition, high unemployment and un-
deremployment, and low occupational mobility-all fac-
tors that increasethe likelihood of being poor. And being
poor adds to the risk of becoming disabled. Much dis-
ability in developing countries is caused by injuries or by
communicable, maternal, and prenatal diseases, some
of them preventable. Medical prevention of disease be-
comes easier with rising incomes, of course.
In the long run policy efforts need to focus on pre-
vention, especiallyon maternal and child health care. Pro-
grams to eradicate measles, to fight onchocerciasis (river
blindness), and to reduce micronutrient deficiencieshave
already greatly reduced disabilities.9 7 Preventive pro-
grams that keep simple diseasesfrom becoming chronic
disabilities are especially important for children.98 War
and civil conflict have also caused many disabilities.Land
mine accidents have increased sharply over the past 15
years: a study of four war-affected countries found that
6 percent of households had a member who had been
killed or permanently disabled by land mines.99
People with disabilities incur extra medical costs and
are often excluded from services and community activi-
ties.100 Most people with disabilities depend on their
families for support and cannot increase their labor sup-
ply in response to income crises. One study found that
61-87 percent of land mine victims went into debt to pay
101
their medical bills,and 12-60 percent had to sellassets.
Prevention and better health care hold the key to reduc-
ing disabilities in the future. Those who are already dis-
abled need community-based rehabilitation programs
and public transfers to the families that provide care.102
Old ageassistanceandpensions. The risksassociatedwith
old age have social as well as economic dimensions, and
policies need to address both. To reduce the social isola-
tion of many of the elderly, social policies should facili-
tate access to community groups or associations that
cater to the elderly. Proximity to health facilities is also
a major concern, since elderly people have difficulty
reaching faraway clinics.
On the economic side, many elderly are poor because
they have been poor all their lives.103 Poverty reduction
policies that increasepeople's income during their work-
ing liveswill also make them better off during retirement.
Well-functioning financial markets that facilitate saving
and investment will help workers accumulate financialas-

Box 8.7
Two universalhealth insurancesystems:CostaRica and Singapore
CostaRicaand Singaporehave vastly different income levels and
administrative capacity, but each has succeeded in establishing
universalhealthcarecoverage.Theyalso havesome commonchar-
acteristicsthat are helpful in targetingfee waiversto poorpeople,
such as almost universal literacy and a system of formal docu-
mentationof vital events (births,marriages)and transactions(em-
ployment contracts, utility bills).
CostaRica
In CostaRicathe public sectordesigns and carriesout healthcare
policies.Therole of the privatesectorin healthcare is very limited:
barely2 percent of hospital beds in the country are in privatefaci-
ities.TheCostaRicanSocialSecurityFundwas createdin 1943,and
coveragefor healthserviceswas extendedto the entirepopulation
in 1971.About85 percentof the populationactuallyparticipates. Fund-
ing comes throughpayrolldeductionsand voluntary,income-based
contributionsof the self-employed.Publicspendingon healthcare
hasremainedhigh, rangingfrom 4.7 to 6.8 percent of GDP during
1975-93. Universalhealthinsurancewent handin handwith health
carestrategiesaimedat preventingdisease,addressingspecificrisk
factors, and extending service coverageto ruraland urbanareas.
Healthindicatorsresponded.Between 1975and 1990infant mor-
tality declinedfrom 37.9 to 15.3 per 1,000live births,and medically
assistedbirths rose from 82.5 percent to 95.2 percent.
The 15 percent of the populationnot covered by the national
healthinsuranceprogramis concentratedat the lowest end of the
incomedistribution.A free healthinsuranceprogramcovers more
than three-fourthsof this group. Eligibilityis verified through sys-
tematic evaluationsby social workers, basedon documentation
provided by applicantson householdcomposition,earnings,and
housing conditions.The administrative reviews of applicantsare
methodicalandeffective: 55 percent of programbenefits goto the
poorest quintile.
Equityconcernsarefurther addressedin the primaryhealthcare
reform started in 1995.The country has 800 healthzones, each
servedby a comprehensivehealthcare teamthat ensuresuniversal
accessto primary care and suitable referral to higher-levelfacili-
Source:Grosh 1994; Prescott and Pradhan1999; Sauma1997.
sets over their lifetime. This is especiallyimportant for
informal sector workers and the self-employed, who
rarely participate in pension plans. Higher incomes and
better risk management for today's prime-ageworkerswill
also help them support their parents financially.
Formal pension systemsare limited in most developing
countries,coveringonly 16 percent of the labor force in the
developing world.'04 In the poorest countries in South
Asiaand Sub-SaharanAfricapensionscoverlessthan 10per-
cent of the labor force.1 5 Coveragecanbe increasedthrough
suitable reform,but thistakes time: coverageratesabove50
HELPING POOR PEOPLE MANAGE RISK 153
.
ties. Each healthcare team is supported by a health committee
set up by the community.
Still, all is not well in Costa Rican health care. Waiting times
are long, and there have been complaints of improper treatment
of users. As a result, many people entitled to public services go
to private providersfor low-cost procedures.And because eligi-
bility is not linked to a specific number of premiums, some peo-
ple pay premiums only when they need costly treatment. This
violates the solidarity principle of an insurance system. Costa
Ricahas introducedreforms to deal with these problems.
Singapore
Between 1984and 1993Singaporeset up a three-tieredsystem
of healthinsurance:Medisave,Medishield,and Medifund.Thepro-
gram insuresagainstintermediate-levelhealthrisks through indi-
vidualor householdMedisaveaccounts.Thesemandatorysavings
accounts, partof Singapore'scompulsorysocial security system,
are funded by a 40 percent payrolltax (sharedequallyby employers
and employees).Of this contribution, 6-8 percent is allocatedto
Medisaveaccounts,which can be used to payhospitalizationex-
pensesof up to about$170a day. Individualsare expectedto cover
minor health costs out of pocket or through private insurance.
Catastrophic healthrisksare coveredthroughMedishield,optional
backupinsurancefor expensesexceedingthe maximum coverage
providedby Medisave.Eighty-eight percentof Medisaveaccounthold-
ershave optedfor Medishieldcoverage.Thecoinsurancerate is 20
percent,andthe deductibleamountvarieswith the comfort classof
the medicalfacility.
Equitybackupis providedthroughsubsidiesfrom Medifund, to
remedy the nonprogressive nature of Medisave accounts and
Medishield.A catastrophichealthshockwould cost 55 percent of
annualpercapitaexpendituresfor thepoorestquintileof households,
and just 21 percentfor the richest quintile.TheMedifund subsidies
are differentiated
by classof facilityandthusare self-targeted
to poorer
users.As a last resort, patientswho are unableto payall their med-
icalbillscanapplyfor a means-testedgrant from their hospitalMed-
Ifundcommittee,financedfrom the government'sbudget surplus.
percent of the labor force are usuallyseenonly in countries
with annual per capita income exceeding$5,000.106
The general recommendation for pension reform is to
establisha multipillar system: combining a publicly man-
aged defined-benefit plan with a privately managed
defined-contribution plan, supplemented by voluntary
retirement savings. The publicly managed plan, funded
from general tax revenues, can address poverty and eq-
uity concerns. The privately managed plan, fully funded
by participant contributions, servesas wage replacement
after retirement.10 7 Several countries, mainly in Latin
154 WORLD DEVELOPMENT REPORT
Americaand EasternEurope, havemultipillar pension sys-
tems. But successfulmanagement of such systemsrequires
considerable administrative capacity.
Even a well-structured pension systemwill not initially
reach the poor. Coveragein formal pension systems tends
to
be much greater for high-income workers: in Chile
to be much greater for high-income workers:
more than 40 percent of workers in the poorest income
decile do not participate in the pension system, compared
with fewer than 20 percent of workers in the richest
108
decile. In general, coverage is lowest among
the uneducated, the self-employed, and women who
have worked in the household rather than in the labor
market for most of their lives.
Contributing to this lower coverage are market and in-
stitutional failures and incentives that discourage mdi-
viduals from seeking coverage. The profile of risks that
poor households face may mean that illness or harvest fail-
ure are of much more concern to them than old age in-
come security. In a credit-constrained environment
mandatory contributions to a pension system may be dif-
ficult for poor or self-employed households to meet. If
in addition the public pension svstem lacks credibility,
many households will continue to rely for old age income
security on traditional informal arrangements, often
based in the household, extended family, or tribe.10 9
Addressing the needs of the elderly poor thus requires
more than pensions. Preventive
moretan pesionsmeasures
Prevntive easurs include
incude facili-
fcili-AIDS
tating saving and investment and providing poverty re-
duction programs during people'sworking lives.Different
forms of direct and indirect support are needed for today's
elderly. Programs can provide assistanceto families that
care for live-in elderly."'' Retraining and workfare pro-
grams adapted to older workerscan make it easierfor them
to continue working. i And socialassistanceor socialpen-
sions should cover the poorest and the very old (categories
that frequently overlap) and those without family sup-
port (box 8.8). Widows will often make up a large part
of this group.112
Unemploymentinsuranceand assistance.Labor market
risk can be reduced significantlyby improving the func-
tioning of labor markets and by adopting sound macro-
economic policies. Many labor markets in developing
countries are segmented (effectivelybarring entry to some
groups) and excessivelyregulated. Reform of labor laws
and regulations needs to balance greater efficiency in the
labor market with promotion and enforcement of core
labor standards to protect vulnerable workers (chapter 4).
Eliminating the most exploitative forms of child labor
2000/2001
Box 8.8
Social pensionsin Chile and Namibia
Countriesas different
as ChileandNamibiahaveestablished
socialpensionschemesto coverthe mostvulnerable
elderly.
Chile has a multipillarpension system, with pensions that
in Chile depend on years of employment and contributions. Gender
differencesinearningsand inyearsin the labormarketleadto
widedifferencesinpensionpayments.Takea woman withan
incompleteprimaryeducationand withaveragetenure in the
the poor, labor market who retires at her statutory retirement age of 60:
she would receive only 29 percent of the pension of an equally
qualifiedman who retiresat his statutoryretirementage of 65.
Inadditionto thisformalpensionsystem,the government
financesa socialassistancepensionintendedforpoorwomen
and men over65 not coveredby the formalsystem. Because
the program is means tested rather than employment based,
benefitamounts are not differentiatedby gender. Sinceel-
derlywomen are generallypoorerthan elderlymen, the pro-
gram benefits women proportionately more than men,
especiallyin ruralareas.
Namibia administers a social pension program for indi-
vidualsoverage 60. UnlikeChile'sprogram,Namibia'sis uni-
versal, not means tested. In practice,88 percent of eligible
pensioners receive the pension. The social pension con-
tributes significantlyto poverty reduction. It is the main
source of income for 14 percent of rural households and 7
percent of urban households.
The social pension program also indirectlyhelps chil-
dren, because many elderly people in Namibialook after
grandchildrenandpaytheir schoolfees when the parents are
awayworkingor lookingfor workor because the childrenare
orphansordisabled.
CoxEdwards
Source: 2000;Subbarao
1998;WorldBankforthcoming
a.
should be a primary objective.'1 3 In the informal sector,
where laws and regulations are seldom applied, public ac-
tion can complement customary informal arrangements
to improve the environment in which workers operate.
Reform and enforcement need to be combined with
programs of skill enhancement, job search assistance,and
microenterprise development. Since experience with
government-run training programs is mixed, partner-
ships with the privatesector need to be explored. 14Labor
markets can also be made more effectiveby improving re-
lationships among labor market partners (employersor-
ganizations, trade unions, and government) and by
strengthening collective bargaining and contracting.' 5
Even a well-functioning labor market will not fully
eliminate the riskof unemployment or underemployment,
however. Displaced workers will need unemployment
benefits to protect them from large income losses and

poverty. In some countries the link between unemiploy-
ment and poverty is very strong. A study of poverty in
countries of Eastern Europe and the former Soviet Union
found a 40-80 percent higher incidence of poverty
among households that had one unemployed member
than among households that had no unemployed mem-
ber. Households with severalunemployed members had
poverty rates twice the national averageor more in some
countries.116 Typical unemployment programs in the re-
gion include retraining, wage subsidies, job counseling
and referral services,public work and community em-
ployment, and small business creation programs.1 1 7
Unemployment insurance, the traditional means of
mitigating the risk of job loss, is not appropriate for
most developing countries because of their low admin-
istrative capacity and large informal sectors. The irregu-
lar and unpredictable earnings typical of the informal
sector make it hard for workers to participate in a con-
tributory insurance program.1 8' Many of the market and
institutional failures discussed under pension systems
apply also to unemployment insurance.Better options for
assistingthe unemployedare means-testedsocialassistance
and public work programs (workfare)."9 Means testing
has proved difficult in most settings, but promising ap-
proaches that use easily observable indicators for target-
ing are being pilot tested.'20
Workfareprograms.Public work programs are a useful
countercvclicalinstrument for reachingpoor unemployed
workers.They can easilybe self-targetingby payingwages
belowmarket rates.A well-designedand well-fundedwork-
fare programis a mix of riskmitigationand coping.To mit-
igate risk, the programmust inspireconfidencethat it will
continue to be availableafter a crisis. Only if the govern-
ment is perceivedas credible will such programs induce
households to give up costly self-insuranceor group in-
surance,freeingresourcesfor other productivepurposes.12 '
The programfunctionsas a copingmechanismby providing
jobs when a crisisstrikes.Providinghouseholdswith income
followinga crisishelps them avoid costly and damaging
strategies(sellingassets,reducingfood intake).Somework-
fareprograms-such asTrabajarin Argentina,theTfemporary
Employment Programin Mexico,'22 and the Maharashtra
EmploymentGuaranteeScheme in India-have succeeded
in creating employment for poor people (box 8.9). Other
programs not originallydesigned as workfare programs
may actuallyperformverysimilarfunctions.This is the case
for Mexicolsself-targetedProbecat, which providestrain-
ing to the urban unemployed.'2 3
HELPING POOR PEOPLE MANAGE RISK 155
Workfareprograms are not necessarilyan inexpensive
wa) of delivering benefits to poor people. Their cost-
effectivenessneeds to be compared with that of alterna-
tive transfer programs. The cost per person-day of
employment createdvariesgreatlyacrosscountries, rang-
ing from as low as $1-2 in several South Asian programs
to $8 in Bolivia.The cost depends on the wage rate, type
of projects undertaken, costs of local private contrac-
tors, and administrative effectiveness. Wages typically
represent 30-60 percent of total costs.124
Socialfinds. Social funds help finance small projects
identifiedand implemented by poor communities, which
usuallyprovide cofinancing. Almost 50 countries, most
in Latin America and Sub-SaharanAfrica, operate social
funds or similar entities. The world's largest is in Egypt.
Recently,Eastern European and Central Asian countries
have begun to set up social funds, with 10 already in op-
eration or under preparation.
The first social fund was set up by the Bolivian gov-
ernment in 1987 as an emergency response to a general
economic dowvnturn.Generally,however,social funds are
not coping instruments. Instead, they address a wide
range of objectives, including infrastructure, commu-
nity development, social services,and support for de-
centralization. 2 5 But some have been used to respond to
emergencies-Hurricane Mitch in Central America
(chapter 9), civilwar in Cambodia, an earthquake in Ar-
menia, drought in Zambia. Social funds have also grad-
uallyassumeda greaterrole in risk mitigation-supporting
incomegenerationprojects,stimulatingschoolenrollment
and health center use, and strengthening the social cap-
ital of communities.They haveprovedto be flexible,quick
to respond, and cost-effective.But the record is mixed
whenit comes to sustainabilityand poverty reduction.16
Social funds use three targeting devicesto reach poor
people: investment selection (mainlv basic services),pro-
ject screening(to ensure that most beneficiariesare poor),
and geographical targeting (of poor areas). The poverty
targeting strategy and the demand-driven approach of
social funds are sometimes in conflict, To enhance their
effectiveness, many funds initially financed projects in
better-offcommunitieswithgood organizationalskills.The
poorest communities,which often havedifficultiesputting
investment proposals together, receivedfewer benefits.
To address this problem, some social funds (Ar-
gentina, Chile, Mali, Romania) have supported capac-
ity building in poor communities. Others have
temporarily assumed some implementation responsi-
I56 WORLD DEVELOPMENT REPORT 2000/2001
Box 8.9
Principlesof successfulworkfare programs
In many programsfor the poor a largeshareof benefits go to the
nonpoor. This problem has stimulated interest in self-targeting
schemes, such as public work programs(workfare), which have
beenespeciallyeffective.Two successfulworkfare programsare
the MaharashtraEmployment Guarantee Scheme in India and
Trabajarin Argentina.
Launched during the severe drought of 1970-73, the
Maharashtrascheme expanded rapidly to reach some 500,000
workers monthly. In a typical year the scheme provides 100 mil-
lion person-daysof employment.Argentinaset upTrabajarII in the
mid-1990s (asan expandedand reformedversionof anearlierpro-
gram)to cope with sharplyrising unemployment, which reached
18 percent in 1996-97and was concentratedamong poor people.
Projectselection
Both programsconcentrateon infrastructure projects (roads,irri-
gationschemes,embankments).Localauthorities,in collaboration
with communities and NGOs, proposeprojects, which must use
labor-intensivetechnologies, benefit the local community, and
target poor areas.
Wage rate and self-targeting
To ensurethat most participantsarepoor andto maintainincentives
for workersto moveon to regularwork when it becomesavailable,
programsshould payno more than the averagewage for unskilled
labor. Trabajarset the wage rate at about 75 percent of average
monthly earnings from the main job of the poorest 10 percent of
households in Greater Buenos Aires. The Maharashtrascheme
usesthe averagewage rateof ruralunskilledlabor.Both programs
havebeenhighlysuccessfulin reachingthe poorestof the poor.About
9 of 10Maharashtraschemeparticipantswere livingbelowthe local
povertyline;4 of 5 Trabajarparticipantswere poor byArgentinestan-
dards. Forthe poorest 5 percent of participants,programbenefits
were 74 percent of their pre-programincome.
Source:Jalan and Ravallion 1999c;Lipton 1998; RavaliBon1991, 1999a.
bilities while communities increased their capacity (box
8.10). Several funds are improving their poverty target-
ing. In Malawi and Zambia social funds are introduc-
ing poverty mapping to identify pockets of poverty 127
Beneficiary assessmentshave identified community ori-
entation (responsivenessto community priorities, help-
fulness in promoting social cohesion) as one of the
strengths of social funds.'28
Microfinance (credit, savings, and insurance). Micro-
finance programs can help poor households smooth con-
sumption during an adverse shock. Access to credit may
help them avoid distress sales of assets and replace pro-
ductive assets destroyed in a natural disaster. But micro-
finance programsdo more than help householdscopewith
Benefitsto the poor
Since poor people can rarelyafford to be totally idle, they often
give up some form of incometo join a workfarescheme. Estimates
suggest that forgone income could representas much as 50 per-
cent of the wages paid by workfare schemes. But because the
employment is guaranteed,it providesmajor insurance benefits
to poor people.Incomesin villageswhere the Maharashtrascheme
operates havejust half the variability of incomesin villages with-
out the scheme.Poorpeoplealsoderiveindirectgainsfrom a work-
fare programif the infrastructure created by the programbenefits
them. Experienceis mixed. In some cases better-off households
have appropriatedthe assetscreated (not an unqualifiedliability,
since it may increasethe political acceptanceof the scheme by
the rich, apparentlythe case in Maharashtra).
Principlesof success
Workfare programscan improve their effectiveness by adhering
to severalprinciples.
* Thewage rateshould be determinedby the local marketwage
for unskilledlabor, not by the program's budget. If resources
are insufficient to meet demand, the programshould target
areaswith a highconcentrationof poor people.Usingadditional
eligibility criteria should be avoided.
* Wageschedulesshould be genderneutral.Women can be en-
couragedto participatethroughsuitableprojectselection,decen-
tralizedwork sites, andthe provisionof child care.
* Labor intensity should be higher than the local norm for sim-
ilar projects.
* Communitiesshouldbe involvedin projectselectionto maximize
the captureby the poorof indirectbenefits of the infrastructure
created.
* To get the most risk mitigation, the program should be avail-
able at all times, expandingautomaticallyduring crises as de-
mand increases.
shocks-they can also provide capital to create or expand
microenterprises. Microfinance thus helps households
diversifv their sources of income and reduces their vul-
nerability to income shocks. Microfinanceprograms have
been especially important for women and households
headed by women, who often have difficulty getting
credit. However,microfinanceinstitutions, depending on
their size and diversification, are unlikely to be effective
against large covariant shocks (chapter 4).
Microfinance programs have been more successfulin
reaching moderately poor and vulnerable (not necessar-
ily poor) households than extremely poor households.
Most programs reach clients just above or just below the
poverty line. Efforts to direct microcredit programs ex-

plicitly to poor households often fail, although there is
evidence that some programs successfullyuse geographic
targeting to reach poor people.129 Having appropriate local
groups identify beneficiaries or targeting beneficiaries
by size of landholdings (as in the 0.5-acre limit used by
Grameen Bank) has proved more successful.130
Empirical studiesfind that clientsoften use loans to re-
duce risk rather than to cope with shocks, meaning that
loans are not usually"diverted"to consumption. Poor and
nonpoor clientsalike use loans to smooth consumption by
smoothing income flows, mainly by increasing diversifi-
cation. Loans help households accumulate a variety of as-
sets:physicaland productive(vehicles,equipment, housing,
livestock), financial (savings accounts), human (educa-
tion, health care),and social(contributions to funeralsand
weddings or to networks of mutual support).13 1
As a risk management tool, the key strength of micro-
finance programs is the knowledge that loans will be
available in time of need, making it possible for house-
holds to dispensewith lesseffectiveand lessdesirablestrate-
gies (child labor, money under the mattress). There is a
parallel here with employment guarantee schemes: the
confidence in future availability is the key to the success
of microfinance programs as a risk management tool.
The availability of microfinance servicesenables poor
householdsto move from reactiveto proactiveapproaches:
they can plan to mitigate risk. Most clients, well aware
of this benefit, go to great lengths to repay their loans so
Box 8.10
The EritreanCommunity Development Fund
Afterthe war of independence,the governmentof Eritrea
promisedto provideeach provincewithbasic economicand
socialinfrastructure.Butmanypoor communitieslackedthe
capacityto implementthe projectsthemselves. Eritrea'sin-
novativesolutionwas to combinesocialfund and publicwork
mechanismsin the EritreanCommunityDevelopmentFund.
Thefund combinesthe bottom-upselectionof projectswith
the top-down selection of interventionareas. Contracting
proceduresare keptflexibleto reacheven communitieswith-
out implementationcapacity.Ifa communitycannot form a
projectcommittee to supervisea project,the fund takesover
procurement,contracting, and technical supervision.If nec-
essary,thefundevenmanages
thecommunity's
contribution,
Thisflexibleapproachiscombinedwith anambitiouscapacity-
building program,which trains community and localgovern-
ment staff in projectdesign, maintenance,and operation.
Source:Frigenti,Harth,andHuque1998.
HELPING POOR PEOPLE MANAGE RISK 157
that they do not lose access to future loans. Clients con-
tinued to repay loans even during and after the floods in
Bangladesh.13 2 Evidencesuggeststhat microcredit has es-
pecially improved the lives of poor women, by strength-
ening their bargaining position with their husbands,
boosting their self-confidence, and increasing their par-
ticipation in public life.133
The success of microfinance in reducing vulnerabil-
ity through income diversification and asset accumula-
tion suggeststhat these programs should be a priority for
government and donor support.13 4 But expanding the
client base to poorer households remains a challenge.To
some degree, microfinance products could be redesigned
to reach poorer households. Loan size and repayments
could be made more flexible to better match the income
flows and repayment capacity of borrowers.13 5 There
probably is a practical limit to this accommodation, since
at some point the increasing costs of making such loans
will undermine the sustainability of microfinance insti-
tutions. The very poorest may well be more effectively
helped with targeted cash transfers.
Programeffectivenesswould be increasedby combining
microcreditwithsavingsand insuranceproductssothat cients
would not haveto take out loansto copewith illnessor deati
(box 8.1 1). Bank Rakyat Indonesia and SafeSave in
Bangladesh demonstrate the potential of combining
Box 8.11
The Self-EmployedWomen's Association
of India
Establishedin 1972,the Self-Employed
Women's Association
(SEWA)is a registeredtrade unionfor women in India'sinfor-
mal sector.SEWA's220,000
me-bsedtorkErs, membersare
andlbers adthawkers, vendors,
towits ven-
homebased workers,andiaborers Inadditionto its conven-
tionarsaborunion functions(ensuringminimumwages and
worksecurity)cuEtA providesTegal aidand operates a bank
and a socialsecurity scheme The bank offers savings ac-
counts and loans to members, The socialsecurityscheme,
whichinsuresao members, coers
which insuresabout 14 percent of SEWAmemberst covers
Slightly morethan halfthe cost
heaothlifeandasset insurance.
insurance programiscoveredby premiums.The rest is
of the
financed by SEWA and a public subsidy SEWA views this
arrangementas a first step towardincreasedcontributionsby
membersandself-sustainability. The combinationof banking,
nuac,aduinsrieshshle.nraeSW'
Insurance,and union services has helped increaseSEWA's
planshto andheath bnefits a d a pensionc on
plansto expandhealthbenefitsandadda pensioncomponent.
Source:
LundandSrinivas1999b;MiraiChatterjee,
generalsecretary,
SEWA,emalcommunication,3 May2000.
158 WORLD DEVELOPMENT REPORT 2000/2001
microcredit with savings. Other microfinance programs
havesuccessfullyintroduced lifeinsurance,at low ratesand
136
with limitedbenefits(burialcostsand repaymentof debts).
Cash transfers. Cash transfers (excluding transfers
through such contributory systemsas regularpensions and
unemployment insurance) include social assistancepay-
ments for the elderly, child allowances, targeted human
development programs, and fee waivers for basic ser-
vices. In countries with large informal sectors, where
formal unemployment insurance is not feasible, means-
tested social assistance is an important way of assisting
the unemploved and underemployed.
The role of cash transfers in a social risk management
strategy depends on a country's income. In high-income
countries cash transfers are part of social insurance, of-
fering a broad guarantee of minimum income. In tran-
sition economies family assistance payments represented
0.4-5.1 percent of GDP in 1992-93. Cash social assis-
tance programs operate in only a few Asian countries,
where they account for less than 1 percent of GDP, and
are negligible in Africa and Latin America.
Cross-countryexperiencesuggeststhat familyassistance
and targeted social assistance are effective for reducing
poverty in the short term, especiallyin countries with rel-
atively little poverty. The difficulty is finding an appro-
priate targeting mechanism compatible with the country's
administrative capacity.Decentralized solutions may be
preferableif communities havebetter information on who
is needy (see box 8.5).13
Targetedhumandevelopmentprogramsforpoorhouse-
holds with childrentransferincomein cashor in kind on the
basisof such observablecriteriaas children'sage,attendance
in school,or participationin a healthcareprogram.They thus
servethe dual objectivesof povertyreductionand human de-
velopment.When effective,they preventthe long-termdam-
age to children that occurs when households, unable to
adequatelymanagerisk, respond to shocksby underfeeding
their childrenor pulling them out of schoolto work.
In the Bangladesh Food-for-Education program the
transfer to a household of 100 kilogramsof rice increased
the probabilityof boys' schooling by 17 percent and girls'
schoolingby 160 percent.138 The BrazilianBolsaEscolapro-
gram targetsscholarshipsto regionsand communitieswhere
child labor is greatest,seekingto keepchildren in school by
compensatingparents for the income children would have
earned.The Mexicanscheme Progresaprovideshealth and
educationbenefitsfor 2.6 millionhouseholdsin 2000. Eval-
uation resultssuggestthat the program is able to target ben-
efits to the poorest householdsand that it has raisedthe en-
rollmentsof childrenin beneficiaryhouseholds(seebox 5.5
in chapter 5).139
Fee waivers can be effective in counteracting falling
school enrollment in the aftermath of a crisis or shock.
Following the crisis in Indonesia, primarv school enroll-
ment of boys in the poorer areas of Jakarta fell 8.3 per-
cent and junior secondary enrollment fell countrywide,
with the greatest drops in poorer areas. In 1998 the In-
donesian government abolished entrance fees for public
schools and lowered monthly fees and exam fees at the
primary level, providing relief for many parents who had
fallenbehind on fee payments as a result of the crisis. An
individual scholarship program and block grants to
schools, both targeted to poorer areas, supplemented the
fee waivers to restore school enrollment rates.1 40
Poor people are exposed to a wide array of risks that
make them vulnerableto income shocks and lossesof well-
being. This chapter argues that helping poor people man-
age risk is thus an essential part of poverty reduction
programs-and should complement effortsto increaseav-
erage income and improve the distribution of income,
which are discussed elsewhere in this report. The focus
has been on risks occurring primarily at the individual,
household, and community (micro and meso) levels,
such as illness and injury, crime and domestic violence,
old age, harvest failure, and fluctuations in food prices
and demand for labor. (Chapter 9 discusses macro-level
riskssuch as macroeconomiccrisesand natural disasters.)
Poor people respond to their risk exposurethrough di-
versification of assets and sources of income and various
typesof self-insurance(bufferstocks, savings)and informal
insurance (networks of mutual support)-all means to
reduce the risk or soften its impact. Where these pre-
emptive mechanisms prove inadequate, households cope
with shock by increasing or diversifying labor supply
(child labor, migration), selling assets, or reducing
consumption.
These mechanismswork, but not well enough. Volatil-
ity in household income remains high in many areas,
and many households suffer episodic declines in well-
being. Some recover,but not all do. Shocks common to
a large area, which can wipe out an entire network's re-

sources,aremostlikelyto overwhelmthe riskmanagement
toolsof poor households.Andbecauseshocksdo not af-
fectallmembersof poorhouseholdsequally,with women
and childrenfrequentlythe most at risk,inadequaterisk
management can cause long-term harm to children
throughmalnourishment, childlabor,andlossof schooling.
In most developingcountriestoday,riskmanagement
emphasizesinterventionsaftera disasterstrikes.The bal-
ance needsto shift to favorpoliciesto reduceand miti-
gate risk. Health, environmental, labor market, and
macroeconomicpoliciescan all reducerisk. And safety
HELPING POOR PEOPLE MANAGE RISK 159
netsput in placebeforeadverseshockshit can serveboth
risk mitigationand copingpurposes.
Tocounterthe incentiveandinformationproblemsthair
excludepoor peoplefrom many market-basedinsurance
mechanisms, the statehasa specialrolein providingor reg--
ulatinginsuranceand settingup safetynets. This reporir
advocatesa modular approachthat adaptsthe safetynet
to the specificpattern of risk in eachcountryor areaand
complementsexistingrisk managementarrangements.
Many solutionswill involvepartnershipsamong poor
communities,the privatesector,and the state.

Mlanagcing Econom'ic
Cises and N~*atural
Disasters
There is nowhere to work. We get sick and we don't
have the money to get cured, we don't have medicines
because they are expensive. The government makes
everythingexpensiv... We don't have moneyto buy
fertiizers, seeds, everything is in dollrs. We don't
have anything to eat. Everythingis so expensive,
-From a discussiongroup of aduiltwomen at
the rime of a banking crisis,Juncal, Ecuador
[Security is]. . the ability of persons to cope with
disasters.
-From a discussiongroup, Little Bay, Jamaica
cnomic crises and natural disasterscan bring
deep and suddencollapsesin national output--and sharp
increasesin income poverty;Together with violent con-
flicts(seebox 3.2 in chapter 3), they are great sourcesof
vulnerability
and inseurity.
Worse,becauseofdthe
collateral
damagetheycause,suchasirreversible
lossofhumancap-
ital, they affect not only the current living standardsof
poor peoplebut theirabilityto escapefrom povertyaswell.
Malnutrition and dropout rates among poor chil-
CHAPTER 9
dren may rise during economic crises and natural
disasters. Poor households are often forced to sell
their meager assets at depressed prices. These re-
sponses perpetuate chronic poverty, possibly reducing
future economic growth because of the irreversible
losses in human and physical capital. That is why
preventing economic crisesand natural disasters is so
crucial. And that is why, when they occur, among the
top priorities should be to protect poor people. Re-
quired for that protection are not only resources but
also the instruments (safety net programs) to channel
those resources to poor households. While develop-
ing countries and transition economies in general are
vulnerable to crises and natural disasters, SMAl states
are especially vulnerable to adverse external events
because of their remoteness and isolation, high degree
of openness, susceptibilityto natural disasters,and lim-
ited diversification.1I
Preventingand copingwith
economic crises
Even our limited accessto schoolsand health is nou'
beginning to disappear. We fear for our childrens'
161
i6z WORLD DEVELOPMENT REPORT
fiiture.. What is tbe justice ini sending our children
the garbagesite every daeyto suipport thefiunilvy?
.! si.l
i>/s/ ljirts'cf
l',l -;; li.' li.
Economywide crises entail sharply falling output, de-
clining incomes, and rising unemployment. Pervasivein
the 1990s, they came in different forms: fiscal crises,
balance of payments crises,terms of trade shocks,currency
crises,banking crises,hyperinflation.The economic crises
in Mexico in 1995, in East Asia in 1997, and in Brazil
and Russia in 1998 received wide media coverage. But
they were not the only episodes of economic distress.2
Most crises have been brought on by varying combina-
tions of policy mismanagement and such external factors
as terms of trade shocks, volatile capital flows, and con-
tagion in international capital markets.
Economic criseshurt both the poor and the nonpoor,
but they are far more devastating for those already in
poverty or nearly poor. even if they are not hurt dis-
proportionately. The welfare losses are larger for poor
households and those who fall into poverty than for the
rest of the population. Poor people are unlikely to have
enough savings or self-insurance to see them through
bad times, and they have little or no access to insur-
ance schemes, whether social or market based (chap-
ter 8).
An economic crisisaffects the living standards of poor
people and those living close to poverty through differ-
ent channels:
* Typically, real wages fall and unemployment rises,
driving down labor earnings.
* Nonlabor incomes fall as economic activity slows,
and the prices of the goods and services produced by
poor people may fall relative to other prices.
* Privatetransfers,particularlyfrom family members, are
likelyto shrink as living standardsfall acrossthe nation.
* The meager assets of poor people are exposed to in-
flation or a collapse in prices.
* Macroeconomic crises slow the accumulation of
human, financial, and physicalcapital, weakening the
ability of poor people to escape poverty.
Is the observed fall in incomes during crises made
worse by the policies to respond to the crises?The debate
on this is long-standing.That risingpovertycoincideswith
the policy responses does not mean that the policies
caused the rise. Crises can occur because of past unsus-
tainable macroeconomic policies or inability to adjust to
2000/200I
tu
externalshocks(termsof tradeshocks,higherinternational
'' . interestrates,suddenmovements in capital flows as a re-
sult of contagion). In such circumstances restrictive fis-
cal and monetarypoliciesare inevitableand lesscostlythan
the alternative of delaving such measures, which could
lead to a larger crash.
Once adjustment policies are accepted as inevitable,
the way governments introduce fiscal austerity can
worsen the adverse effects on the living standards of the
poor and near-poor. For example, removing food or
fuel subsidies would exacerbate the effects on poor peo-
ple-unless compensatory measures are taken (chapter
4). So would increasing the rates and sometimes the cov-
erage of indirect taxes on food and other products that
figure large in the consumption basket of poor people.
Net government transfers may decline as governments
cut social assistance as part of a fiscal austerity program.
Reducing the quantity and quality of public services
used by the poor and near-poor would also worsen their
situation.
But government actions can also mitigate the impact
of crises on poor people. The task of the policymaker
is to implement the combination of macroeconomic
measures that results in the lowest cost in forgone out-
put and affords the greatest protection to the living
standards of poor people. A key element of a poverty-
sensitive response is the right composition of revenue-
raising measures and fiscal cuts. A poverty-sensitive
response should also allow for the expansion of safety
nets targeted to poor people (the "social insurance"
component of social spending) during periods of macro-
economic adjustment.
Social i7npact of cises
There is a strong link between macroeconomic down-
turns and rising income poverty (table 9. 1; see also fig-
ure 2. 1).3 During crisesmany people become temporarily
poor, and social indicators tend to worsen or to im-
prove more slowly. Data suggest that the human capi-
tal of poor people, particularly poor children, can
deteriorate. The damage can be irreversible,affecting the
ability of these children to escape povertywhen they reach
adulthood.
In most countries in East Asia poverty rose as a result
of the financial crisesof the late 1990s: it is estimated that
it rose almost 50 percent in Indonesia and that urban
poverty doubled in the Republic of Korea.4 In both
 MANAGING ECONOMIC CRISES AND NATURAL DISASTERS I63

Table 9.1
Effect of economiccriseson incidenceof poverty in selectedcountries
Percent
.

Before Year After

Countryandtypeof crisis crisis ofcrisis crisis
Argentina, hyperinflation and currency 25.2 47.3 33.7
(1987) (1989) (1990)
Argentina,contagion 16.8 24.8 26.0
(1993) (1995) (1997)
Indonesia, contagion and financial 11.3 18.9 11.7
(1996) (1998) (1999)
Jordan, currency and terms of trade 3.0 .. 14.9
(1986-87) (1989) (1992)
Mexico, currency and financial 36.0 .. 43.0
(1994) (1995) (1996)
RussianFederation,financial 21.9 32.7
(1996) (1998)
Thailand, currency and financial 11.4 12.9a
(1996) (1998)
Not available.
Note:Basedon nationalpovertylinesand percapitahouseholdincomeexceptfor Indonesia(percapitaexpenditure),Mexico(household
income),and Russia(householdexpenditureper equivalentadult).Datafor Argentinareferto GreaterBuenosAires.ForIndonesiapoverty
estimatesbefore andduringthe crisisare basedonthe full SUSENAS (thenationalsocioeconomicsurvey)conductedin February1996and
1999;estimatesafterthe crisisare basedon a smallersample.Figuresare not comparableacrosscountriesbecausepovertylinesdiffer.
a. Basedon the socioeconomic surveyconductedbetweenFebruary1998andJanuary1999,which doesnot fully reflectthe impactof the
crisis.Estimatesfrom a smallersurveyconductedduringJune-September1999put the povertyincidenceat 15.9percent.
Source:Ministeriode Economiade Argentina1998;WorldBank1994c,1999dd;ECLAC1999b;Lokshinand Ravallion2000b.

countries, however, poverty fell as the economies recov-
ered. In Russia the incidence of poverty rose from 21.9
percent to 32.7 percent between 1996 and 1998. In
every crisis in Latin America and the Caribbean the in-
cidence of poverty increased and several years later re-
mained higher than it had been before the crisis.
Inequality may rise, fall, or remain unchanged during
a crisis. In Latin America inequality (as measured by the
Gini coefficient)rose in 15 of 20 crisisepisodes for which
there are data. In East Asia during the recent crisis, how-
ever, inequality remained practically unchanged, and in
Mexico following the peso crisis in 1995 it fell. When
crisesare accompaniedby increasesin inequality,economic
contractions can more than reverse previous gains in
poverty reduction. In Latin America the poverty reduc-
tion from a 3.7 percent increase in per capita income for
urban areas and a 2 percent increasefor rural areas in the
1970s was reversedby just a 1 percent declinein per capita
income in the 1980s.5 Even if inequality increases, the
poorest fifth of the population is not always hurt dis-
proportionately. In Latin America the income share of the
middle fifths of the population often fell most during the
1980s debt crisis, but the share of the top tenth always
rose, sometimes substantially.6
The impact of economic crises on living standards is
not fully captured by measures of inequality and income
poverty. Economic crises are characterized by extensive
mobility: previouslynonpoor people may fall into poverty,
and previously poor people may escape it. Evidence of
sharp downward and upward mobility was found after
the 1998 crisis in Russia, for example.7 Mean expendi-
tures of people classified as poor in 1996 actually rose,
and 42 percent of them escaped poverty after the crisis.
By contrast, 61 percent of those who were poor after the
crisis had not been poor in 1996. Put another way, 20
percent of the population fell into poverty as a result of
the economic downturn. Even though overall inequalitv
fell and a large share of the poor escaped poverty after the
crisis, there was substantial downward mobility for many
who were not previously poor and for some who were al-
ready poor. Those who become poor during economic
crisesoften have different characteristicsthan the chron-
ically poor. For example, they may be better educated. A
study in the Philippines found that households with
more education are more vulnerable to wage and emr-
ployment shocks.8
I64 WORLD DEVELOPMENT REPORT 2000/2001

Table 9.2
Social impacts of economic crisesin selectedcountries

Maincrisisindicators Healthindicators Educationindicators

Argentina * Per capitaGDPfell4.1%. a Per capitadailyproteinintake * Growthingross primary
1995 * Per capitaprivateconsumption fell3.8%in 1995,but enrollmentdeclinedfrom
fell 5.6%. increased 1.9% in 1996. 2.2% in 1993to 0.8% in
1996.

Mexico * Per capitaGDPfell7.8%. * Amongchildrenunderage * Grossprimaryenrollment

1995 * Per capitaprivateconsumption 1, mortalityfromanemia increased0.44%in 1994,
fell 11.1%. increasedfrom6.3deaths but fell0.09%in 1995.
per 100,000livebirthsin
1993to 7.9 in 1995.
* Amongchildrenages 1-4,
the mortalityrate fromanemia
rose from1.7to 2.2per 100,000.

Indonesia * Per capitaGDPfell 14.6%. a The shareof womenwhose a The dropoutratefor children
1998 * Per capitaprivateconsumption bodymass indexis belowthe in the poorestfourthof the
fell5.1%. levelat whichrisksofillness populationrosefrom1.3%in1997
and deathincreaserose 25%. to 7.5%in 1998for thoseages
* Most indicatorsof child 7-12and from14.2%to 25.5%
nutritionalstatus remained for those ages 13-19.In both
stable.Theexceptionmay cohortsthe poorestfifth
be the weight (conditionalon experiencedthe largest
height)of childrenunderage increase.
3, suggestingthat familiesmay m Theshare of children
be investingin some members inthe poorestfourthof the
at the expense of others. populationnot enrolledin
schoolrose from4.9%in 1997
to 10.7%in 1998for those
ages 7-12 and from 42.5% to
58.4% for those ages 13-19. In
bothcohortsthe poorestfourth
had the largest increase.
Note:Gross enrollment ratios are used because net ratios were not available. These data should be used with caution.
Source: World Bank 1999cc; IDB Statistical and Social Database; PAHO 1998; Thomas 1999; Frankenberg, Thomas, and Beegle 1999.

Most social indicators either deteriorate or improve at
a slower pace during a macroeconomic crisis (table 9.2).
Social indicators such as infant mortality rates continued
to improve in Latin America in the 1980s, though more
slowlv than in the previous decade. But health indicators
more sensitive to consumption or income downturns
worsened. In Chile the share of low-birthweight infants
and undernourished children rose as the economy de-
clined. In Mexico infant and preschool mortality caused
by nutritional deficiency rose in the 1980s, reversing the
trend of the previous decade, and rose again with the eco-
nomic crisisof 1995. In Argentina and Venezuelathe daily
per capita intake of protein declined as per capita GDP
fell. In Indonesia the share of women whose body mass
index is below the level at which risksof illness and death
increaserose by a quarter in 1998, and the averageweight
of children under age three declined.
School attendance and literacy also take hits during
crises. In the Philippines secondary school enrollments
increased only 0.9 percent between the 1997/98 and
1998/99 academic years, after growing at an averagean-
nual rate of 2.6 percent in the previous five years. In Mex-
ico the proportion of each graduating class that enrolled
in the next education leveldeclined during the 1980s debt
crisis, particularly among high school and university stu-
dents. The percentageof age-appropriatechildren entering
primary school also declined. In rural areas the dropout
rate rose by 40 percent. In Argentina and Mexico growth
in gross primary enrollment slowed in 1995. A study for
South India found that children are often taken out of
 MANAGING
school in response to adverse shocks.9
There is no question that economiccrisesincreasetran-
sitorypoverty " They can alsoincreasepersistentor chronic
poverty because of hard-to-reverseeffects on the human
capital of poor people. While the trends cited for malnu-
trition, infant mortality, and enrollment are national av-
erages, they most likely reflect a deterioration in these
indicatorsamong poor people. For Indonesia, information
by income group shows that the dropout rate in the low-
est fourth of the income distribution rose from 1.3 per-
cent in 1997 to 7.5 percent in 1998 among children ages
7-12 and from 14.2 percent to 25.5 percent among those
ages13-19. The proportion of poor children not enrolled
in school increasedfrom 4.9 percent to 10.7 percent.
Recent research shows a link between macroeco-
nomic downturns and education indicators. The aver-
age annual increase in years of schooling in 18 Latin
American countries fell from 1.9 years in the 1950s and
1960s to 1.2 in the 1970s and 1980s. Worsening macro-
economic conditions (short-term GDP shocks, volatil-
ity, and adverse trade shocks) explain 80 percent of the
decline, according to one study." 1 As evidence from
Mexico shows, the negative "income effect" of falling in-
come tends to outweigh the positive "price effect" of the
lower opportunity cost of attending school.12 Simulation
results suggest that the gross secondary enrollment rate
in Mexico would have been 11 percentage points higher
in 1991 if the economy had grown during the 1980s at
half the rate of the 1970s.
Avoiding crises
Clearly,avoidingcrisesshould be a top priority in any anti-
poverty strategy.There is wide agreement on the kind of
macroeconomic and financial policies governments need
to reduce vulnerability to policy-inducedcrisesor adverse
external shocks.13 They should avoid profligate fiscal
and monetary policies,overvaluedexchangerates, and un-
sustainable current account deficits-all problems in the
1970s and 1980s.
Many parts of the world have made great progress in
steering awayfrom irresponsiblefiscal policy. Leadingex-
amples are the largeeconomiesin Latin Americaand some
of the transition economies, where the ensuing fall in in-
flation rates has helped build investor confidence and re-
duced, if not eliminated, the potential long-term effects
of inflation on efficiencyand growth.1 4 Lower inflation
has also helped reduce poverty, since high inflation often
hurts the poor more than the nonpoor. In Argentina, for
ECONOMIC CRISES AND NATURAL DISASTERS I65
example, ending hyperinflation brought about a signif-
icant one-time drop in the incidence of poverty: ini
Greater Buenos Aires the incidence of poverty dropped
from 34.6 percent in 1989 to 22.6 percent in 1991.15
The 1990s saw various types of crises, triggered by
weak banking systems and weak financial regulation in
a world of large and volatile international capital flows.
Liberalizing the financial sector was expected to put
economies on a more stable footing. But the transition
from more repressed to more open financial systems in
the developing world has been difficult to manage.
Banking criseshave been more numerous in the past two
decades, when stroke-of-the-pen financial liberalization
became popular. 16
Some of the reforms introduced in the financial sec-
tor backfired because the institutional rules allowed ex-
cessivelyrisky behavior while the costs of that behavior
had to be paid by society as a whole. A vivid example is
the Mexican financial crisis of 1995.17At the root of the
crisis was a weak banking system, its fragility traced to
the privatization processused for the banks, some aspects
of the financial liberalization program, and weak regu-
latory institutions. Rescuing the banking sector will cost
Mexican taxpayers an amount equal to about 20 percent
of GDP (in present value terms).
To prevent financial crises, governments need to imt-
provethe prudential regulationand supervisionof financial
intermediaries, introduce new standards for data dis-
semination, and implement corporate bankruptcy re-
form."8 These measures are already under way in many
developing countries, but there is still a long way to go.
At the same time, a cautious approach should be taken
to capital account liberalization. Controls on capital
inflows-such as those Chile used until recently-can he
an appropriate instrument for tempering the volatility of
capital flows.There is evidence that capital controls can
shift the composition of capital flows toward longer-
maturing investments.19
Other initiatives and measures are also important for
avoiding crises, such as mechanisms to diversify and in-
sure against risk. Some governments, such as Chile, self-
insure using fiscal stabilization funds. Others, such aLs
Argentina, negotiate contingent credit lines between the
central bank and private international financial institu.-
tions to ensure accessto foreign currency in the event of
a sudden slowdown in capital inflows.20
However, actions at the national level may not be
enough to prevent economywidecrises. Domestic actiorns
i66 WORLD DEVELOPMENT REPORT 2000/2001
will have to be complemented by actions at the interna-
tional level to fosterglobal financial stability (chapter 10)
and help countries, particularlythe poorest and the small-
est, manage commodity price shocks.
Formulating a crisis response that protects
poor people
No matter how skillful the economic management, crises
are likely to affect the developing world and transition
economies for some time to come. That is why articu-
lating a response to crises must take into consideration
its impact on poor people. A poverty-sensitive response
to crisis should steer toward:
* Helping poor householdsmaintain their consumption.
* Ensuring that poor people do not lose whatever ac-
cess they have to basic social services.
* Preventing permanent reversals in the accumulation
of human and physical capital.
* Averting self-defeating behavior, such as criminal ac-
tivity, prostitution, and exploitative forms of child
labor.
A poverty-sensitive response should also provide mech-
anisms for those at risk of becoming poor as a result of
the crisis.
What does it take to protect those who are alreadypoor
and those at risk of becoming so from sharp declines in
short-term income? Appropriate macroeconomic re-
sponses and well-functioning safety nets can enhance
equity and result in better growth outcomes. Some of the
recommendations here are already being incorporated in
the standard approach for dealing with crises. The Re-
public of Korea, for example, introduced or expanded
safety nets relativelyquickly in the wake of the 1997 fi-
nancial crisis (box 9.1). But in general the response con-
tinues to be ad hoc-with measures thrown together in
the heat of a crisis.
Adopting the right macroeconomicpolicy mix. Re-
sponding with the right macroeconomic policy mix after
an adverse shock is one of the biggest challenges policy-
makers face. Driven by political considerations, policy-
makersmay postpone needed adjustment and stabilization
measures becausethey are painful-making the situation
far worse. Peru was an extreme casein the 1980s.The gov-
ernment refused to implement an adjustment program
and in July 1985 announced a cap on external debt pay-
ments (a de facto unilateral moratorium) equal to 10 per-
cent of exports. Peru did well for a while, but the
disequilibria continued to mount and in 1988 the econ-
omy crashed, with per capita GNP falling by 13.4 per-
cent and real wages by 40.6 percent. Altogether, real
wages fell by 67 percent between 1988 and 1990.21
The 1997 crisisin Thailand showswhat happens when
there are no correctivemeasuresto address the buildup of
vulnerability.22 True, the financial panic of domestic and
international investors suddenly concerned about the fate
of their portfolios lit the fuse for the explosion. But the
buildup of structural vulnerabilities provided the
dynamite-sharply rising short-term debt that far ex-
ceededinternationalreserves,a financialsectorthat had done
a poor job of intermediating capital inflows and found it-
self saddled with hugely mismatched assetsand liabilities,
and corporationsthat were massivelyoverleveragedand ex-
posed to interest and exchangerate fluctuations.
Not all problems arise from a failure to adjust to an
adverseshock or from unsound macroeconomic policies.
In some casesthe policy response errs in the direction of
too much adjustment, with fiscal and monetary policy
more restrictive than necessary to restore equilibrium in
the currency market, the current account, or the capital
account. Overreaction can cause more pain than neces-
sary and in some circumstances can be self-defeating.An
initial overreactionon the fiscalfront can lead to a higher
fiscal deficit down the road because the larger-than-
expected recession lowers government revenues, defeat-
ing the purpose of the initial austerity measures. The
reason for overshooting often is that cautious policy-
makers prefer to err on the side of excessiveadjustment,
since timid adjustment can be far more devastating.
Although it may be hard to tell whether a policy pack-
age is excessivelyrestrictive, there are some indications
that those in place in EastAsia during the recent crisiswere
just that. In Thailand the tax increasein September 1997
made the ensuing recessionworse. In Korea the restrictive
fiscalpolicy initially made room for the expected costs of
bank restructuring. But the fiscal target was subsequently
relaxedas both the authoritiesand the internationalfinancial
institutions recognizedthat it was unrealisticin light of the
larger-than-expectedslowdown in growth. Aiming toward
the original target in the face of worsening economic con-
ditions would have been self-defeating.And for Malaysia
and the Philippines the trend of cyclicallyadjusted deficits
(for both revenuesand expenditures)suggeststhat they did
not relax their fiscalpolicy, even though the actual deficit
made it look as though they had.
Even if excessivelyrestrictive policies are later cor-
rected, the short-term costs can be significant, particu-
 MANAGING ECONOMIC
Box 9.1
Providingsocial protection in responseto crisisin the Republicof Korea
The Republic of Koreawasmakingsustainedprogressin reduc-
ing povertyin the 1990s:the urbanpovertyratefell an average
20 percenta yearduring1990-97,andtherewere no increases
in inequality.Butthen economiccrisisstruck,sharplyincreasing
unemployment andpoverty.Theincidence of poverty
in urbanareas
doubledfrom9 percentin 1997to 19.2percentin 1998.1 Unem-
ploymentrosefrom2.6 percentin the secondquarterof 1997to
apeakof 8.7 percentinearly1999.Realwagesdeclined 20.7per-
cent.Most of the newlyunemployed were low-paidworkers:in
December1998three-quarterswere temporary,daily, self-
employed, or unpaidfamilyworkers,andabout20 percentwere
the headof a household with no otherincomeearners.
Expansionary fiscalpoliciesin 1998and1999were criticalin
stemmingthe economicdownturn.Socialprotectionspending
was increasedthreefold-from 0.6 percentin 1997to 2.0 per-
cent in 1999.The governmentusedthreemaininstrumentsof
socialprotectionto help the unemployed,the poor,and the
elderly:
* Unemployment insurance.Koreaexpandedits nascentun-
employmentinsuranceprogram-the only such program
amongthe EastAsiancrisiscountries-fromfirmswith more
than30 employees to allfirms.It alsoincludedtemporary and
dailyworkers,shortenedthe contribution periodrequiredfor
eligibility,andextendedthe durationof unemployment bene-
fits. Thisexpandedthe eligibleworkforcefrom 5.7 million
workersat the beginningof 1998to 8.7 millionat the endof
theyear.Beneficiaries increased tenfold-fromaround18,000
1.Thepoverty
rateswerecalculated
usingseasonally
adjusted
expenditure
dataanda national
Source:WorldBank1999w,2000d.
larly for poor people. If there are viciouscyclesof poverty,
low education, and poor health, a recessioncan causeper-
manent damage for the poor.
Do the macroeconomicresponsesto crisesthat are best
for the overall output levelsof the economy differ from
those that would be best for the incomes of poor people?
Perhaps. Different policy combinations imply different
costs for the poor than for the nonpoor because of the
way the reduction in per capita output is distributed.2 3
But even if distributive outcomes were the same, the
poor and the nonpoor could well prefer different policy
packages.24
Poor people are more likelyto prefer an adjustment that
leads to the smallest drop in GDP at any point in time
even if it implies a slower recovery.Nonpoor people are
more likely to prefera program that reduces income more
severelyin the short run but yields higher growth in the
medium run. This difference results simply from the fact
that the welfare losses from an economic downturn are
CRISES AND NATURAL DISASTERS I67
in January1998to 174,000in March1999,still only 10per-
centof the unemployed workforce.
* Publicwork.Sincemostof Korea'sjoblessdidnotbenefitfrom
the expansion of unemployment insurance,thegovernment in-
troduceda temporarypublicworkprogramin May1998,en-
rolling76,000workers.By January1999the programwas
providing437,000jobs,thoughthe numberof applicants was
higherstill,at 650,000.By the first quarterof 1999the public
workprogramwas benefitingaround2.5 timesas manypeo-
pleasthe unemployment insurance program.
* Livelihoodprotection. In May1998thegovernment introduced
a temporarylivelihoodprotectionprogram,with fundingto
cover750,000 beneficiaries.It alsointroduced a means-tested
noncontributory socialpensionfor 600,000elderlypeople.
Althoughthegovernment's socialprotectionresponse wasquite
exemplary, publicspending onhealthandeducation didnotincrease
in linewith the overallbudget,andrealspending eitherfell or re-
mainedconstant.Butevenwithinthesmallerenvelope for health,
spending on primarycarewasprotected.
Thegovernment is nowfocusing onconsolidatingsocialsafety
nets,reducing incomedisparities, andcreatingthe basisfora com-
petitiveandknowledge-based economy. Policiesto achievethese
objectivesincludea law guaranteeing a minimumstandardof liv-
ing,to takeeffectinOctober2000.Thelawwill entitleallKoreans
livingunderthepovertylinetoreceiveincomesupportforliving,ed-
ucation, andhousing expenses. Nearly2 millionpoorpeopleareex-
pectedto benefit,fourtimesthe currentnumber.
povertylineequivalent
to about$8a day(in1993PPPdollars).
higher for poor people. Moreover,becausepoor people live
close to the subsistence level, their preference may reflect
application of the safety principle (minimizing the prob-
ability that their income will fall below a certain level).Or
the poor and nonpoor may discount future consumption
differently,with the poor putting a largerpremium on pre-
sent consumption than the nonpoor.
The distributive and intertemporal implications of al-
ternative adjustment policies are important, but policy-
makers rarelyhavethe luxury ofchoosing among different
adjustment paths. In general, the optimal combination of
policies-to achievethe necessarybalance of payments acl-
justment with the smallest decline in output-depends
on initial conditions.2 5 When a currency is under speCL-
lative attack, a spike in interest rates will in most casesbe
needed to stop the attack. But when a country intro-
duces adjustment measuresearlyon, the government may
have more freedom to choose among different policy
combinations and thus be more likely to manage a so.ft
168 WORLD DEVELOPMENT REPORT 2000/2001

landing. Unfortunately, macroeconomic analysis in its to poor people should be protected if not expanded.
current state can offer little guidance in assessingthe dis- It may seem obvious that governments should protect
tributive and intertemporal implications of alternative spending that benefits poor people and expand the safety
policy packages, clearly an area in need of far more ana- net programs targeted to them. But this does not neces-
lyrical and applied research. sarily happen in practice. Recent research in some coun-
Protectingspendingthat benefitspoorpeople. How gov- tries in Latin America has found that a 1 percent decline
ernments raiserevenues and cut public (nondebt) spend- in per capita GDP leads to an estimated 2-3 percent de-
ing has important policy implications for who bears the cline in targeted public spending per poor person.27 And
burden of adjustment and whether poor people are pro- a study on the Argentine employment program Trabajar
tected.26 To design a poverty-sensitivefiscal adjustment found that its performance in reaching poor people de-
to avoid or respond to a crisis, policymakers need to as- teriorated sharply with cuts to its budget.28
sessthe distributional effectsof spending programs.A use- There may be severalreasons for such "antipoor" pat-
ful tool for this is the public expenditure review(box 9.2). terns in fiscal adjustment. Without budgetary guidelines
As a general rule, areas important for poor people- to direct fiscal austerity, governments may go for pro-
basic education, preventive health care, water and sani- portional cuts to minimizebureaucraticinfightingand ease
tation, rural infrastructure-should be protected from acceptanceby the legislature.Another reason may be that
budget cuts to ensure that services are adequate. That governments lack the instruments to target resources to
means ensuring that schoolsand health posts in poor areas the poor-instruments that are difficult to put in place
have at least the basic minimum of supplies. General sub- in the heat of a crisis. Even if the instruments exist, po-
sidieson food staples might need to be maintained in the titicalforces may be such that the resourcesgoing to poor
short run-even if the benefits leak to the nonpoor- people are cut more than proportionately. In some coun-
unless they can be effectively replaced by targeted pro- tries information can be the major constraint:governments
grams. Safety nets and social assistanceprograms targeted may lack reliable records of their budget or programs.
What can be done to counter these factors? One way
Box 9.2 to protect spending that benefits poor households is for
Publicexpenditure reviews to assessthe impact the government and legislature to rank current programs
of fiscalretrenchmenton poorpeople by their importance as part of the budget approvalprocess.
When spending cuts are needed, the order of the cuts
Publicexpenditurereviews-assessments of publicsector would be determined bvthe priority assignedto each pro-
issuesthatfocuson theefficiencyandrationale
of the public
budget-could beusefultoolsfor evaluating
theimpactof gram. Government agencies could be required to evalu-
fiscaladjustment
programs andpublicsectorreformson ate social programs to helppolicymakers identify those
socialprogramsandsafetynets.In economywide crises thatare most cost-effectivein reducing poverty and there-
thatleadto spending
cuts,thesereviewscouldhelpestab- fore should be protected during a crisis.
lish a transparent
budgetmechanism
for rationalizing,
allo- Peru has introduced guidelines for protecting pro-
cating,
executing,
andmanaging public
spending to protect
poorpeopleandensureprivate
sectorefficiency. grams that benefit poor people as part of its public finance
Public
expenditure
reviewstypically
analyzeandproject reform law (box 9.3).The guidelines combine fiscal rules
public
revenues
anddeterminethelevelandcomposition of with measures to increase fiscal transparency and ac-
publicspending,
assessing
the allocation
of resources
among countabilitv. T
andwithinsectors. Whenplanningfiscalretrenchment,
a he programcreatesa stablizaton fund with
shortreviewshould bedone,focusing
onthesectorsthatac- the proviso that programs benefiting poor people should
countforthebulkof thepublic
budget
(agriculture,
education, be protected. Although such budget protocols may not
health,
infrastructure).
Thereviewshouldrankexpenditures be classified as antipoverty programs, they can have an
onsocialprograms, considering
the tradeoffbetweenthese impor rant
effect on overt b rotecting o-poor
programsandothernonessential spending(suchas military e p yu g prp
spending)that couldbe minimizedduringa crisis.Thistype spending during fiscal retrenchment.
of adjustmentisclearlymoreefficientinprotectingvulnera- If benefits targeted to poor people are cut for politi-
blegroups andmaintaining
private
sectorefficiency
thanthe cal economy reasons, a third party-such as the multi-
typicaluniform
spending
cut. lateral lending organizations-could advocate for the
Source:World Bank 1999v. poor and help governments implementing austerity mea-
sures design a viable way to protect programs and spend-
 MANAGING ECONOMIC CRISES AND NATURAL DISASTERS i6-9

ing that benefitthe poor.This happenedto somedegree One wayto avoid thispoliticaleconomyconstraintis to
in severalcountriesin the 1990s. designprogramswith lowmarginalbenefitsor high mar-
Changesin the incentivesystemembeddedin targeted ginal costsfor the nonpoor. 29

programscould alsofacilitatecuts for nonpoor benefi- Evaluatingdifferenttypesof spendingcan be difficult

ciariesduringperiodsof austerity.The argumentis this: when data are poor, the casein most developingcoun-
it is often saidthat for politicaleconomyreasonssome tries.Efficiencyindicatorsarealmostnonexistent,anddata
of the benefitsof targetedprogramshaveto goto the non- on actualspending,asopposedto budgetedamounts,are
poor-through "leakage"-to ensure continuing sup- availableonly afterlonglags.Usuallyan evaluationshould
port for programs.The same forceswill presumablyact take the availableintermediateinformation and com-
to limit the welfarelossesto the nonpoor from cuts. plementit to determinewhetherpublic resourcesreach
the intended beneficiarieseffectively.
A socialmonitor-
Box 9.3 ing and earlyresponseunit, such asthe one set up in In-
Protectingpoorpeopleduringfiscaladjustment: donesiaduringits recentcrisis,can helpensurequickand
Peru'sFiscalPrudenceandTransparency Law reliableinformationfor evaluatingspendingin specific
30 Wherefieldsurveysare infeasible(be-
socialprograms.
Prudence
Peru'sFiscal andTransparency
Law,overwhelm- causeof budget or time constraints), recent household
sur-
inglyapproved by the national
congress in 1999,doesmuch b
to ensurethat socialprotectionis maintained
duringafiscal veys can e used to try to determie an efficent and
adjustment. rational allocation of government resourcesamong social
First,the lawestablished fiscalrulesonthe maximuman- programs and safety nets.
nualdeficitof the consolidated publicsector,cappingit at 2 Putttingsafetynetsinplace beforea crisis.If the problern
percentof GDPin 2000,1.5percentin 2001,and1 percent is a lack of instruments for protecting poor people, the so-
thereafter.(Theconsolidated publicsectorincludesthe cen-
tral andregionalgovernments, decentralized agencies,and lution is to introduce, during normal times, safety net pro-
national publicenterprises; it excludes localgovernments and grams that can operate as insurance in times of economic
their agenciesandenterprises.) In the eventof a national distress. Safety nets are important for several reasons.
emergency,internationalcrisis,or fall in GDP,the fiscal
deficitcan increaseto 2 percentof GDP.The lawalso set Theycanplayacrucal rolei mitigatigtheeffectsofcrses
limitson increasesin publicspendinganddebt.The maxi- on the poor and protecting the near-poor from fallinginto
mumannualgrowthof nonfinancial publicspending is equiv- poverty. A study estimated that if the targeted programn
alentto the inflationrateplus2 percentage points,implying Progresa (see box 5.5 in chapter 5) had existed when the
a futurereductioninthe relativesizeof the publicsector. 1995 crisis hit Mexico,the poverty gap index in rural areas
Second,the lawcreateda fiscalstabilization fund,to be and he
s ed povery gap
index m
gives greas
fundedfromthreesources: the revenuesabovethe average and the squared povery gap index (whichgivesgreater
collectedduringthe previousthree years,three-fourthsof weight to the poverty of the poorest)31 would have declined
futureprivatization proceeds, andhalfof allrevenues fromfu- by 17 percent and 23 percent in the year after the crisis.-52
ture concessions. (Savingsaccumulated inthefundinexcess Safety nets can also help prevent irreversible damage to tlhe
of 3 percentof GDPwill betransferred to the publicpension . .
fundor usedto reducepublicdebt.)Upto 40percentof the human capital of poor people. And theycan aid poltical
fund'sresources canbe usedin a givenyearif currentrev- acceptance of stabilization and reforms, preventing con-
enuesfall below the averagecollectedover the previous flicts over resource distribution that can create stalemates,
threeyears.Fundresources canalsobeusedinemergencies, deepen economic crises, even cause governments to fall.
suchasaneconomiccrisisora naturaldisaster.
Third,the law mandatesthat the fund's spendingon tar- Recent work has shown that istitutional weaknesses, i-
geted poverty reduction programs be given priorityover cluding lack of safety nets, have been responsible for
spendingon other programs. many crises over the past 25 years.33
To enhance fiscal transparency, the law introduced a Most developing countries lack effective safety nets that
three-yearfiscalframeworkto be developed,approved,and
published by the government. And to improve fiscal ac- protect poor people from the output, employment, and
countability,it requiresthat the finance minister submit to price risks associated with systemic adverse shocks. When
congressand publishannualreportsassessing the execution these mechanisms are not in place before a crisis occurs,
of the fiscal goals in the multiyearframework. policymakers are often forced to improvise or to use pro-

Source:Ruprah1999. grams designed for other purposes and other beneficia-

ries. Emergency responses to emergency situations are
often prepared without technical analysis to identify the
170 WORLD DEVELOPMENT REPORT 2000/2001
groupsmost vulnerableto the shocksand to evaluatethe
cost-effectivenessof differentsocialprotection options.
Programsput in placeand operating-even on a small
scale-before criseshit do betterat protectingpoor peo-
ple than ad hoc emergencymeasures.
To be effective,safetynetsshouldincludea widerange
ofprograms-publicworkprograms,scholarships forpoor
children,cash transfers,food-relatedtransfers,foodsubsi-
dies, socialfunds, and fee waiversfor essentialservices
(chapter8). Socialprogramsthat focuson long-termde-
velopment(forexample,suchtargetedhumandevelopment
programsasMexico's Progresa)canalsoperforma safetynet
function during economicdownturns.The appropriate
mixof safetynet programswilldependon the characteris-
ticsof the poor and vulnerable,the type of crisis,and the
government'sinstitutionaland administrative capacity.
The internationalcommunitycan playan important
partbyprovidingpolicyadvice,contributingfinancialsup-
port, and helpingpolicymakersdesignand fund safety
nets. Internationalfinancialinstitutionscan help coun-
tries design pro-poor fiscaladjustment programs and
safetynets and, for countriestoo poor to fund a safety
net duringa crisis,can providefinancing. 34
Reducing vulnerability to natural
disasters
The biggest shoek we ever had was Hurricane Gilbert; . ..
all what we found after Gilbert was one wooden chair.
- ll 'ThYbw,R
Knnau jbJunZ..;
Economic development is repeatedly interrupted by nat-
ural disasters-by earthquakes, droughts, floods, land-
slides, volcanic eruptions, windstorms, forest fires. Like
economic crises, natural disasters can cause sharp in-
creasesin poverty and slow the pace of human develop-
ment. And like economic crises, they hurt poor people
in the short run and diminish their chances of escaping
poverty in the longer run.
The damage to agricultureand infrastructure variesby
type and intensity of natural disasters, as do the impli-
cations for their indirect and secondaryimpacts. Droughts,
for example, can result in heavycrop and livestock losses
while leaving infrastructure and productive capacity
largely unaffected.
Between 1988 and 1997 natural disasters claimed an
estimated 50,000 lives a year and caused damage valued
at more than $60 billion a year.35 Dramatic as these fig-
ures are, the fiill human and economiccosts are even
greater.Human costsincludeinjuriesand temporaryand
permanent disabilities,temporary and permanent dis-
placementof people,the breakupof familiesand social
networks,increasedpovertyand disease,and psycholog-
icalscars.Economiccosts,basedlargelyon directphysi-
calimpactsor lossesof fixedcapitaland inventory,arealso
underestimated.Many indirectand secondaryeffectson
economicactivity-such as changesin fiscalpolicies,the
long-termconsequencesofthe reallocationof investment
resources,or the lossesin humancapital-go unrecorded.
Overthe past 10yearsthe incidenceof naturaldisas-
tershasincreased. 36 This couldbe due in part to socialfac-
tors,assettlementshavesprungup in hazardousareas.The
urbanpoor in megacities-for example,in Riode Janeiro
and itsfavelas-are often forcedto build on steep,mar-
ginalland prone to landslidesthat killor leavehomeless
thousandsof peopleeveryyear.But thereare alsonatural
factors.The El Nifio events,associatedwith anomalous
floods,droughts,and storms,are gettinglargerand more
frequent.37 And warmingof the surfaceof theAtlanticis
increasingthe frequencyand severityof hurricanes. 38
Still,it is often askedwhether it would be more correct
to labelmanyof thesedisastersas "human-made"rather
than "natural."They areprobablyboth.
Inpact of natural disasters on poor countries
and poor people
Z AUnfortunatelyfir mne, the land on which I made myyfarm
was a swampy area and when it rained the whole farm
submerged with water and was destroyed.
-F---Haol'ywa
una, At1/nsuBokr, (7/7an/a
Developing countries, especiallytheir most densely pop-
ulated regions, suffer the brunt of natural disasters. Be-
tween 1990 and 1998, 94 percent of the world's 568 major
natural disasters and more than 97 percent of all natural
disaster-related deaths were in developing countries (fig-
ure 9.1). In Bangladesh alone three storms, four floods,
one tsunami, and two cycloneskilled more than 400,000
people and affected another 42 million. In southern
Africa in 1991-92, Malawi, South Africa, Zambia, and
Zimbabwe experiencedseveredroughts.39 In Latin Amer-
ica and the Caribbean major natural disasters associated
with El Nifio, Hurricane Mitch, Hurricane Georges, and
the Quindio earthquake in Colombia claimed thousands
of livesand caused billions of dollars of damage between
 MANAGING
Figure 9.1
Developingcountriesbore the brunt of natural disastersin 1990-98
Percent
100
80
80
40
20 Major disasters
EastAsia Europeand LatinAmerica MiddleEast
and Pacific CentralAsia andthe andNorthAfrica
Caribbean
Note:A disasteris classifiedas majorif it causedmorethan 50deathsor affectedmorethan 100,000people.
Source:USAID,OFDA1999.
1995 and 1998.4°In 1998 severe flooding of the Yangtze
River caused devastation in China, and a largeearthquake
occurred in Armenia. Another long series of disasters
struck in 1999-a major earthquake in Turkey, a cy-
clone in Orissa, India, floodsin central Vietnam, torrential
rains and catastrophic mudslides in parts of Venezuela,
floods in Mozambique. The list goes on.
Poverty and lagging development amplify the ad-
verse effects of natural disasters. Developing countries
are particularly vulnerable, because they have limited ca-
pacity to prevent and absorb these effects. People in
low-income countries are four times as likely as people
in high-income countries to die in a natural disaster.41
Despite similar patterns of natural disasters in Peru and
Japan, fatalities average 2,900 a year in Peru but just 63
in Japan.4 2 Averagecosts as a proportion of GDP are 20
percent higher in developing countries than in industrial
economies.4 3
Poor people and poor communities are frequently
the primary victims of natural disasters, in part because
they are priced out of the more disaster-proof areas and
livein crowded, makeshift houses.4 4 The incidence of dis-
asterstends to be higher in poor communities, which are
more likely to be in areas vulnerable to bad weather or
seismic activity. And there is evidence that the low qual-
ECONOMIC CRISES AND NATURAL DISASTERS 171
U
Disaster-related
deaths
SouthAsia Sub-Saharan Developing High-income
Africa countries countries
ity of infrastructure in poor communities increases their
vulnerability.
While natural disasters hurt everyone affected by
them, poor families are hit particularly hard because in-
jury, disability, and loss of life directly affect their main
asset, their labor. Disasters also destroy poor households'
natural, physical, and social assets, and disrupt social as-
sistance programs.4 5 Long-term disabilities and the de-
struction of assets can trap families in chronic poverty.
Malnutrition impairs children's ability to learn.
The few studies that have analyzed the impact of nait-
ural disasters on poverty show that the harm to current
and future living standards can be significant. In Ecuador
El Nifio may have increased the incidence of poverty in
affected areas by more than 10 percentage points.46 In
Honduras Hurricane Mitch caused an estimated 7 per-
cent declinein agriculturaloutput in 1998.47Lossof crops
was extensive, affecting a quarter to a half of households.
Rural households, most dependent on agriculture,lost the
most.48
In the 1984 drought in Burkina Faso the income of
the poorest third of the rural population fell 50 percent
in the Sahelian zone, the poorest agroclimate, and 7 per-
cent in the Sudanian zone.4 9 There was also evidence that
poor people sold livestockout of desperation. Becausethey
I72 WORLD DEVELOPMENT REPORT 200012001
had verysmallstocksof animalsto beginwith, thesedis-
tress salesmay have dangerouslydepleted their buffer
stocks, leaving them extremelyvulnerable to future
droughtand othershocksand possiblytrappingthem per-
manentlyin dire poverty. 50
Studiesof the impactof the 1994-95droughtin Zim-
babwefound that women and young childrenwere the
mostaffected.For women,the drought'seffecton health
(as measuredby body mass)was temporary.With good
rains the followingyear,they regainedmuch of the lost
body mass. But for children ages 12-24 months the
drought will probablyhave a permanent effect.These
young children lost an average1.5-2.0 centimetersof
linear growthin the aftermathof the drought.The im-
pact wasmostsevereamongchildrenin householdswith
little livestock,the principalassetof thesehouseholdsfor
smoothingconsumption. 51 The drought had no impact
on men'shealth.
On balance, female-headedhouseholds fare worse
than male-headedhouseholdsfollowinga natural disas-
ter, in part because of their smaller averageresource
base.52 Customaryor formallawscan makethis worse.
Amongthe Tongaof Zambia,for example,a widowhas
no entitlementto any of the household'spossessions. 53
The effectof a natural disasteron povertycan go well
beyond the householdsdirectly affected.Researchon
Sub-SaharanAfricasuggeststhat both agriculturaland
overallGDP are sensitiveto downwardfluctuationsin
rainfall.The 1991-92 drought in southernAfricaslowed
growthin agriculturaland total output in Malawi,South
Africa,Zambia,and Zimbabwe. 54 The impactof drought
shockson GDP andthe recoverytime dependin part on
the economicimportanceof the agriculturalsectorand
its integrationand linkswith industry.The second-round
and subsequenteffectsare more pronouncedin more in-
tegratedeconomies.In Senegaland Zimbabwethe effect
of droughts spilled over from agricultureto manufac-
turing.5 5 The value of Zimbabwe'smanufacturingout-
put declined9.5 percentin 1992,largelyasa resultof the
1991-92 drought,andexportreceiptsfrom manufactures
declined6 percent.5 6
The destructionof infrastructureby catastrophicnat-
ural disastersalsohas both immediateimplicationsand
longer-term,second-roundpovertyeffects.In Asia,for
example,where 70 percent of the world'sfloodsoccur,
the averageannualcostof floodsoverthe pastdecadewas
estimatedat $15 billion, with infrastructurelossesac-
countingfor 65 percent. 57
The needto replacedamagedinfrastructurein disaster-
stricken countries diverts governmentresourcesfrom
longer-termdevelopmentobjectivesand consumesa sig-
nificant share of multinational lending resources.In
Mexicoas much as 30 percentof the funds approvedby
the World Bank for improvingrural water supply over
the past decade have been diverted to postdisaster
rehabilitation.58
Risk reduction aizd mitigation:
lessening vulnerability to disasters
Cumulativeexperience with naturaldisasterspointsto an
urgentneed to movefrom fatalismto prevention,from
responseto preparation,from mobilizingresourcesafter
the fact to reducingand transferringrisk beforethe fact.
There is a distinct differencein approachto emergency
managementbetweenmany developingand developed
countries.Developingcountriesemphasizepreparedness
and response-makingsurethat the resourcesto respond
to emergenciesare availableand readyfor dispatchand
then that theyaredispatchedquicklyand usedefficiently
afteran emergencyhas occurred.Developedcountriesin-
creasinglyemphasizereducingor mitigatingthe impacts
of disasters(box9.4).
Disaster reduction and mitigation can lessen the
disruption causedby natural disasters,save lives, and
protect property. From a purely economic point of
view,investingin risk reduction paysoff. For example,
a cost-benefitanalysisfor eight cities in the Argentina
FloodRehabilitationProjectfound an internaleconomic
rate of return of 35 percent. The estimated $187 mil-
lion (1993 dollars)in avoideddamagesfrom the 1997
floodmorethan coveredthe $153 millionin investment.
By installing flood control dams and improving
drainage,the Rio FloodReconstructionand Prevention
Project reduced total floodable areas by 40 percent,
achievingan estimated6.5 benefit-costratio for seven
subbasinsof the Iguacu and Sarapui Rivers.59 Com-
prehensivedisasterrisk managementcan be integrated
into developmentinvestmentdecisions.In Turkey in-
ternational lendersand donors worked with the gov-
ernment to develop a new disaster management
frameworkin the aftermath of the 1999 earthquakes
(box 9.5).
Resettlement-tailoredto the needsofpoor people-
is often the appropriaterisk reductionstrategyin flood-
proneor volcanicareas.Whereresettlementis not feasible
or desirable,neighborhoodimprovementprogramsarean
 MANAGING
Box 9.4
Mitigation is the cornerstoneof emergency
management in the United States
Mitigation-the ongoingeffort to lessenthe impact disasters
haveonpeopleandproperty-is thecornerstoneof emergency
managementin the UnitedStates.It involveskeepinghomes
away from floodplains, engineering bridges to withstand
earthquakes,creating and enforcingeffective building codes
to protect propertyfrom hurricanes,and more.
Overthe past 10years the U.S. FederalEmergencyMan-
agement Agency (FEMA) has spent $25 billion to help peo-
ple repair and rebuild communities after natural disasters.
Other governmentagenciesand insurancecompanies have
respondedwith billionsof dollarsmore. Beyondthis, the costs
of emergenciesalso includelost lives,jobs,and businessop-
portunities. A big emergency can reduce local GDP by as
much as 10 percent.
In 1995the highand escalatingcosts of emergenciesled
FEMAto adopt a nationalmitigationstrategy,with two goals:
to protect peopleand structuresfrom disastersand to min-
imize the costs of disasterresponseand recovery.FEMA es-
timates that every dollar spent on mitigation saves two in
responseand recovery.
The strategy promotes a community-basedapproachto
reducingvulnerabilityto natural hazards:
I Altering the hazard(seedingclouds during a drought).
I Avertingthe hazard(buildingdams to controlfloodwaters).
I Avoidingthe hazard(movingpartsof communitiesout of
floodplains).
3 Adapting to the hazard (constructing earthquake-proof
buildings).
In February2000 FEMAannouncedProject Impact: Buil-
ding Disaster-Resistant Communities,a projectto provideex-
pertise and technical assistanceto about 200 communities
striving to become disasterresistant. Three principlesdrive
the project: Preventiveactions must be decided at the local
level. Private sector participation is vital. Long-term efforts
and investments in prevention are essential.
Source:Olsson 2000.
alternative. In these programs residents of low-income
urban areas improve their houses themselvesor with com-
munity help. The programs reduce building code viola-
tions by training informal sector construction workers in
mitigation techniques and by providing finance for low-
cost improvements that bring housing to stipulated stan-
dards. International assistance, channeled through local
NGOs, has often helped turn housing reconstruction ef-
forts into low-cost opportunities for mitigating risksin fu-
ture disasters (box 9.6). Other important neighborhood
upgrading activities include constructing drainage works
and reducing the risk of flooding and mudslides.
ECONOMIC CRISES AND NATURAL DISASTERS 173
Box 9.5
Mitigating the risks of natural catastrophes:
lessonsfrom the 1999 earthquakesin Turkey
A powerful earthquake shook northwestern Turkey in the
summer of 1999, killing more than 17,000 people, injuring
tens of thousands, and razingseveral population centers.
Three months later a second quake hit, raising the number
of victims and the social and economic losses.Industry and
businessesin the areas hit by the quakes had contributed
more than 35 percent of the country's GDP. Their destruc-
tion is likelyto affect growth in Turkeyfor manyyearsto come.
The international community assisted Turkey in relief
andimmediaterecoveryefforts. In partnershipwith the Turk-
ish government, the EuropeanInvestment Bank,the Coun-
cil of Europe's SocialDevelopmentFund, andother donors,
the World Bank coordinatedthe preparationof a framework
for a $1.7 billion reconstructionprogram.A crucialpartof the
framework is a disastermanagement and responsesystem
to prevent similar losses in the future.
Disasterand landdevelopmentlaws will be reviewedand
modified, and the capacity of municipalities to regulate,
plan,and implement disaster-resistantdevelopment will be
strengthened. Pilot projects in selected municipalities will
help planning and building departments develop risk-based
municipalmaster plans,meansfor effectiveimplementation
of building codes, municipal regulations to ensure that
builders follow appropriate licensing procedures,and pro-
grams for evaluating existing buildings.
The government's earthquake insurance program will
expand its catastrophic risk management and risk transfer
capabilities.The program will create an insurance mecha-
nism to make funds readily availableto owners (those pay-
ing real estate taxes) who need to repair or replace a
dwellingdestroyedor damagedby an earthquake.It will also
ensure the financial solvency of the insurancepool after all
but the most catastrophic events and reduce the govern-
ment's financial dependence on donors following major
earthquakes.
Source: Kreimer 1999.
Low-cost local initiatives can also reduce the vulner-
ability of communities' income to natural disasters. In
rural areas such initiatives might focus on environmen-
tal conservation and reforestation. For places prone to
droughts and floods, community food banks can help.
In Burkina Faso local cereal banks were introduced to
improve storage, lower food prices, and stabilize themn
over the year, including during the drought season.60
Community agricultural cooperatives can help sma'll
farmers obtain credit or crop insurance. And various
strategies can help diversify the economic activities
within a community.
I74 WORLD DEVELOPMENT REPORT 2000/2001
Box 9.6
Turning reconstructioninto risk mitigation with
the help of a local NGO
In a poor area of Peru partlydestroyed by an earthquake in
1990,Caritas,a localNGO,initiateda reconstructionprogram
that was also designedto mitigateearthquake-relatedrisks.
After consulting with the community, Caritas decided to
construct housing from quincha,a local materialcapableof
withstandingearthquakes.Todirectlyassistthe neediestfam-
ilies,such as householdsheaded bywomen,Caritasprovided
materialsinexchangefor participationincommunalwork.An
earthquakein 1991showed the advantageof usingquincha:
most houses resisted the earthquake,which registered6.2
on the Richterscale.
Source:
Schilderman
1993.
Reducing economic vulnerability also involves
encouraging-or mandating-the purchase of private
insurance for those who can afford it and identifying
mechanisms for transferring risk, such as catastrophic rein-
surance and catastrophe bonds (box 9.7). While risk
transfer mechanisms can efficiently cover much of the cost
of repairing and rebuilding infrastructure, freeing up
scarce government resources, they may not be easy to apply
in poor countries. For one thing, they require systems for
verifving damage that cannot be easily manipulated by
those (governments, for example) who would collect the
insurance benefits. To deal with this problem in floods,
for example, a country could establish a high-quality
measuring and ieporting system. This would facilitate in-
surance contracts that link payment schedules to a rain-
fall index.61
Coping with natural disasters
In the emergency phase following a disaster, efforts should
focus on providing food, water, shelter, and medicine. That
makes temporary repair of such infrastructure as roads and
water supply critical. Priorities need to be based on the
magnitude ofdamages and level ofvulnerability. The most
vulnerable groups-women, children, and the elderly-
need special attention. Involving women in the man-
agement of shelters, establishing workfare programs
adapted to women's needs, and ensuring gender neu-
trality in housing acquisition can improve recovery for
women and households headed by women. Expanding
early childhood development programs, particularly
mother and child feedingprograms, is also important. Re-
building schools should be a top priority-to avoid loss
Box 9.7
Mitigating risk with catastrophe bonds
Catastrophebonds-or cat bonds-offer an alternativeto in-
surance incountrieslackingactiveprivateinsurancemarkets.
A before-the-factrisktransfer mechanism,cat bonds provide
financialprotectionagainst disaster losses.
Consider a government that wants protection against
the riskof flooddamageto one of its water treatment plants
inthe next year. Experts estimate the chance of a flood at
1 in 100, a risk low enough to induce an institutionalin-
vestor to purchase a cat bond whose payoffis tied to flood
damage to the treatment plant.The investorbuys the bond
at the beginningof the risk periodat par. At the end of the
risk periodthe investorloses the entire principalif the water
treatment plant is damaged. But if no damage occurs, the
investorrecoversthe principalplus interest,normallyabove
the market rate to reflectthe riskof losingthe principal.
The government invests the funds, which willbe used
onlyif a catastropheoccurs, in risk-freesecurities.The cost
to the government is equal to the difference between the
interest rate it receives from the risk-free securities and the
interest rate it paysthe bondholder-acost analogousto pay-
ing an insurancepremium.The value of the bond-and the
government's interest payments-would be lower if the
governmentflood-proofs the treatmentplant.So,besidesper-
forming an insurance function, the cat bond gives the gov-
ernment an incentiveto invest in mitigationefforts.
A potential problemwith catastrophebonds is the diffi-
cultyof verifyingdamage. The publicagency operatingthe
water treatment plant mightexaggerate damage to ensure
that the bondholderpays. One way to deal with this moral
hazardis to tie payoutsto an objectiveindex(such as flood
height)rather than to actual damage.
Source:
Kunreuther 1999.
of human capital and perhaps to provide shelter for dis-
placed people. Cash transfers to poor families reduce the
likelihood that they will need to pull their children out
of school. Where children need to participate in recov-
ery efforts, schools can adopt flexible schedules.
Following a widespread natural disaster, national
and local governments need to establish a macroeco-
nomic management scheme to tackle fiscal and current
account effects-lower tax revenues and higher public
spending, lower exports and higher imports. A calamity
fund like that in Mexico can improve governments'
ability to cover the costs of coping with natural disas-
ters (box 9.8). Calamity funds should focus on ab-
sorbing the catastrophic risks that cannot be absorbed
by third parties, such as disaster-related damage affect-
ing farmers and urban dwellers unable to afford private
 MANAGING ECONOMIC CRISES AND NATURAL DISASTERS I75

Box 9.8 insurance, and providing social assistance to poor dis-

the Mexican
Sharingthe costsof catastrophes: astervictims.
fundfor naturaldisasters In the aftermath of a natural disaster targeted inter-
national assistancecan help maintain macroeconomicsta-
Withtremendous diversityin geography andciimate,Mex- bility,accelerate recovery, and protect poor people. But
ico is susceptibleto a wide rangeof naturaldisasters- e .
floods,droughts,earthquakes, wildfires,tropicalcyclones, to enable countries to accept financal assistance, inter-
volcaniceruptions.Since1980directdamagefromnatural national financial institutions may have to relax some ad-
disastershas totaledsome$6.5 billion,and about7,000 justment targets during crises. In Zambia in 1992-93 the
peoplehavelosttheirlives. tight public spending policy being implemented as part
In 1996,to help reducethe country'svulnerabilityto tf pdjubi
spendin poic beinghimplemented raisp
natural disasters,the governmentestablishedFonden o adjustment restricted the governmentsablity to raise
(Fondoparadesastresnaturales,or Fundfor NaturalDis- external financing because of the lack of counterpart re-
asters).Thisfederalfund wasto befinancierof lastresort sources in the local currency. As a result, international
for emergencyresponseequipment,disasterrelief activ- funds for drought relief were not fully used.6 2
ities, and reconstructionof publicinfrastructureandpro- The reconstruction period provides an opportunity to
tected areas.
In 1998,following
a period ofparticularly
heavylossesfrom reducevulnerabilityto naturalevents(seebox 9.6).Targeted
naturaldisasters,
thegovernmentdecided to useFonden assistancelocallyfor the most affectedpopulations (withthe
toprovide
morestrategically, incentives
forinsurance
useand poor the top priority) and consultation with affectedcom-
disastermitigation.After broadconsultationwith stake- munities and householdsshould be a key strategy Includ-
holders,in March1999the government changedFonden's ing local people in reconstruction activities can foster
guidelines
operating to:
* Increaseclarityandtransparency
inthedecision
rulesfor leadership and promote solidarity,helping reduce the psy-
grantingaccessto the fundandin lossassessment chological trauma caused by natural disasters (box 9.9).63
processes. Countries that have social or rural infrastructure in-
* Limitmoralhazard by encouraging greateruseof private
insuranceby Fonden'sbeneficiariesand establishing vestment funds can use them to channel resources effi-
clearcost-sharingformulasfor financingdisasterlosses ciently.6 4 Given their experience in building infrastructure
falling underthe responsibilityof state andmunicipal and providing social services to communities, these funds
governments. can quickly identify local spending priorities and help min-
* Encourage mitigationin the reconstruction
programsfi- prioriIesthelp mu-
nanced byFonden andinbeneficiaries'regularinvestment Imlze corruption. The HondurasSocialInvestmentFund
programs. performed this role in the aftermath of Hurricane Mitch.
* Refinance
disaster
response activities
initiallyfinanced It financed the opening of several important secondary
throughemergencyliquidity
facilities
to speeddisaster roads, began the rehabilitation of some water systems in
recovery. secondary towns and undertook cleanup activities.The
Thesechangesarebeingformalizedthroughvoluntary s
agreementsbetweenthefederal
government andthestate fund rapidly put in place a decentralized operating struc-
governmentsthatsetouttheparties'rightsandresponsi- ture and responded to the pressing needs of local munic-
bilities,
Fonden's
rules,andagreed
cost-sharing for
formulas ipalities and communities. Procedures established by the
disasterreliefandreconstruction
activities.Theagreements fund immediately after the disaster simplified project
willalsoleadtotheestablishment
oftrustsbetween thefed-
eralgovernment andeachstate.Under thetermsof each preparation and authorization and expedited contracting
trust,spending decisions
andcontractingof eligible
emer- and disbursement of funds. Closemonitoring by local res-
gencyactivitieswillbecarried
outbyatechnical committee idents of private contractor equipment and employmenr
actingon
consistingof stateandmunicipalrepresentatives, of local residentsin cleanup activities helped ensure that
advicefrom federalentities.
If successful,
thesemeasureswill increasetrans- funds were well accounted for.65
parency, accountability,and efficiency in the use of The experienceproduced valuable lessons on how to
Fonden's andredistribute
resources thecostsof natural improve the capacity of such funds to cope with disaster:
betweengovernment
disasters andthe privatesector. providing adequate financial resources to meet post-
Overtime they will alsoreducethe shareof costsborne g q
bythefederalgovernment
for mitigating
andcopingwith disasterneeds,ensuringthat ivestment projectsgo be-
disasters. yond rehabilitation and cleanup to include expansion of
existing facilities and new construction, and expanding
Source:Barham
2000. the fund's mandate to allowdirect social assistanceto vul--
nerable people.66
I76 WORLD DEVELOPMENT REPORT 2000/2001

Box 9.9
Involvingcommunities in postdisasterreconstruction:lessonsfrom the Maharashtra Emergency
EarthquakeRehabilitationProgram

On 30 September 1993an earthquake struck the Indianstate of
Maharashtra,killing some 8,000 people and damaging 230,000
housesin Latur,Osmanabad,and 11other districts. With the help
of the World Bank, the government of Maharashtracreated the
MaharashtraEmergencyEarthquakeRehabilitationProgram.The
programinstitutionalizedcommunity participationand formal con-
sultation with beneficiariesat all stages.
The programdivided communities into two categories:those
that needed to be relocated-the 52 villages that sustainedthe
worst damage-and those that needed to be reconstructed, re-
paired, or strengthened. The Tata Institute of Social Sciences
worked in the 52 relocationvillages,which had some 28,000fam-
ilies. The Societyfor Promotion of Area ResourceCenters orga-
nized community participationin the 1,500 villages-with some
190,000families-in which rebuilding or repair was to take place.
Over time the program became a people's project. As re-
sults materialized,community participation received greater ac-
ceptance. Initially skeptical, officials in the project management
unit later came to acknowledge community participation as an
effective tool for dealing with problems that arise during
implementation.
Participationalso had a positive psychologicaleffect on com-
munities. Involvinglocalpeople inthe reconstructionhelpedthem
overcomethe traumacausedby the earthquake.Recognizingthis,
the governmentbeganreconstructionin small villagesevenbefore
the rehabilitationprogrambegan,appealingto donors,corporations,
NGOs, and religiousorganizationsto "adopt" villages for recon-
struction. Some organizationsalso worked on social issues,such
as schoolingfor children.
Information on the program, its processes, and the mecha-
nismsfor redresswas accessible-and awarenesswas high. The
participatoryprocess opened many informal channelsof commu-
nication between the people and the government,helpingto nar-
row the gap between them. Beneficiaries learned of their
entitlementsand workedhardto securethem. Peoplewho felt that
their grievanceswere not addressedappropriatelyin the village or
taluka(anadministrativeunitthat includesseveralvillages)couldtake
them to the districtauthoritiesand the governmentin Mumbai.

Source:Vatsa1999.

Workfare programs can usefully be introduced or ex- * *

panded in disaster areas in conjunction with reconstruc-
tion operations, providing a livelihood to people who Large adverse shocks-economic crises and natural
can no longer support themselves (chapter 8). They can disasters-cause poor people to suffer not only in the short
also help people affected by the less visible impacts of a run. They undercut the ability of the poor to move out of
disaster, such as the poor fishers in Ecuador and Peru who poverty in the long run as well, by depleting their human
fell deeper into poverty as fish fled the waters warmed by and physical assets. Particularly harmful are the effects on
El Nifio. In Northeast Brazil the program Frente de Tra- poor children, who may suffer irreversible damage if a cri-
balho (Work Front) provided similar employment op- sis or natural disaster increases malnutrition or forces them
portunities in periods of drought. During the 1979-84 to drop out ofschool. Integral to any poverty reduction strat-
drought it employed up to 3 million workers in con- egy should thus be measures to prevent and manage economic
struction and drought-related jobs. 6 7 Public work programs crises and natural disasters-and to establish safety nets, with
that build social or community infrastructure or help in ensured financing, to help poor people cope when these ad-
cleanup and reconstruction can also be a good option. verse shocks do occur.
 PART V

International
0
Actions
 t ~ .- ~~ 4
Ft~~~~~~~~~
Harnessing Global Forces
for Poor People
T hroughout this report wehaveseen that poli-
ciesand institutionsat the country and locallevelare
the keysto enhancingthe opportunity,empowerment,
and securityof poor people.But the livesof poor peo-
ple are alsoaffectedby forcesoriginatingoutsidetheir
countries' borders-global trade, capital flows,offi-
cialdevelopmentassistance, technological
advance,dis-
eases,and conflicts,to namejust a few.Actionsat the
global level are therefore crucial complements to
country-levelactions.They can acceleratepovertyre-
ductionand helpnarrowthe gaps-in income,health,
and other dimensions-between rich countries and
poor.
This chapterdiscusses fourkeyareasof international
action for povertyreduction:
* Expandingmarket accessin rich countriesfor de-
velopingcountries'goodsand services.
* Reducingthe risk of economiccrises.
* Encouragingthe productionof internationalpub-
lic goods that benefit poor people.
* Ensuringa voicefor poor countriesand poor peo-
ple in globalforums.
Alsoimportantforpovertyreductionisdevelopment
cooperation-foreign aid and debt relief-discussed
CHAPTER 10
in chapter 11. Other globalforcesthat affectthe poor
include international labor migration, commodity
price volatility,global warming and environmental
degradation,promotionof politicalandhuman rights,
and the internationalarmssalesand tradein illicitgems
that spur or prolong conflictin countries.Severalof
these werediscussedin last year'sWorldDevelopment
Report.
Expanding market access
in high-income countries
At first glance, it seemsthat rich countries benefit
more from the opportunitiesof the globaleconomy.
After all,they haveaveragedfastergrowththan poor
countriesoverthe past40 years.But it is alsotrue that
poor countriesthat are more integratedwith interna-
tional marketshave grown as fast as or faster than
rich countries.' As chapter3 detailed,trade can pro-
vide a powerfulengine for growth and poverty re-
duction. It has alsobeen arguedthat tradewith richer
countriescan speedthe processof "catch-up." 2
Expandingaccessto rich countrymarketscan thus
do much to help poor countries grow faster and to
179
I80 WORLD DEVELOPMENT REPORT
Figure10.1
High-incomecountriesprotectmanufacturing
andagriculture
Taffs on mohancr ifnported by high-income countries,19In
Percent
20
15
Developing
country
10 exports
_ _
income
country World
5 exports exports
o _ ' * * 1
Manufactures Agricultural products
Source: Hertel and Martin 1999a.
reducepovertyin the developingworld.This is particu-
larlt so for agriculturalproducts,since more than two-
thirds of the developingworld'spoor peoplelivein rural
areas.Not only do foreignmarketsrepresentimportant
sourcesof demandfor developingcountries'agricultural
goods-because the demandfor basic food productsis
inelastic-but exportingcan expandnonfarm employ-
ment andstimulatethe entireruraleconomy.Agricultural
exportshavebeen shownto be a strong determinantof
overallagriculturalgrowth.3 So it is disturbingthat while
world trade in manufacturedproductsexpandedat 5.8
percent a year from 1985 to 1994, agriculturaltrade
grewat only 1.8percent.
One reasonfor thisslowgrowthis the continuingpro-
tectionof agriculturalproductsby developedcountries-
protection not only through tariffsand quotasbut also
through exportsubsidies. 4 The tariffsthat high-income
countriesimposeon agriculturalgoodsfrom developing
countries, especiallysuch staples as meat, sugar, and
dairy products, are almost five times those on manu-
factures(figure10.1). The European Union'stariffson
meat productspeakat 826 percent.5 These barriersare
hugeobstaclesfordevelopingcountriesstrivingto break
into exportmarkets.High-incomecountries'agricultural
tariffsand other distortions,such assubsidies,havebeen
estimatedto causeannual welfarelossesof $19.8 billion
for developingcountries-equivalent to about 40 per-
2000/2001
cent of the officialdevelopmentassistancegivento de-
velopingcountriesin 1998.6 This is a serioussetbackto
developmenteffortsin poor countries.
general, trade reforms in poor countries have
failedto delivertheir fullbenefitsbecausethey havenot
been matched by reformsin rich countries. For man-
ufactured goods(includingfood products),which now
account for almost three-quartersof developingcoun-
try exports, tariffs facing developingcountry exports
to high-income countries are, on average,four times
those facing industrial countrv exports to the same
maret
~~~~~market.
High-incomecountries'tariffsarenot only higherfor
manufacturesfrom developingcountries,they alsoes-
calate with the level of processing.For example, in
Japan and the European Union fully processedfood
F * products facetariffs twiceashigh as those on products
in the first stage of processing.In Canada the ratio is
even higher,with tariffs on fullyprocessedfood prod-
ucts 12 times thoseon products in the first stage.This
escalation can discourage industrialization efforts in
developingcountries.
Developedcountries'trade barrierscan place signif-
icant constraints on poor countries' efforts to grow.
Finding ways to unblock the political obstaclesto re-
movingsuch barrierswould do much to aid poverty re-
duction in the developingworld. Bysomeestimatesthe
welfarelossesfor high-incomecountriesfrom their own
distortionarytradepoliciesare large-$63 billiona year
for agriculturaldistortionsalone. It should be feasible
to put in placecompensatorymechanismsfor the rela-
tivelysmall-but politicallypowerful-groups of pro-
ducers as part of an agreementto lower trade barriers.
But more than anything,reducingtrade barrierswill re-
quire real politicalwill on the part of the leadersof de-
velopedcountries.Specialpriorityshouldgo to reducing
the scopeand scaleof protectionon agriculturalgoods,
labor-intensivemanufactures,and services.
Reducing the risk of economic crises
Aschapter9 details,economiccrisesin developingcoun-
triescan be devastatingfor poor people.So creatingthe
conditionsformacroeconomicstabilityis essentialforen-
hancing the securityof the poor and avoidingreversals
in povertyreduction.
Countriescan take measureson their own to reduce
the risk of macroeconomiccrises(chapter9). Amongthe
 HARNESSING
most importantare sound macroeconomicpoliciesand
adequateprudentialregulationand supervisionof finan-
cialinstitutions.But evenif a countryfollowssuch poli-
cies,it can stillbe hit by contagionandby wavesof panic
or herdbehaviorin worldcapitalmarkets.A premiummust
thereforebe placedon ensuringstabilityin the interna-
tionaleconomy,particularlyin the financialsector.
Internationaleffortsto achievestability,intensedur-
ing the Asiancrisis,havetaperedoff as the crisiseased.
One focushas been to create and enforceinternational
standardsfor financialdata disseminationand financial
practices.The goalis to ensurethat financialmarketsand
the public have timely and reliable data for making
decisions-and to ensurethat financialinstitutionsrun
effectively.Towardthis end, the InternationalMonetary
Fund (IMF) has developedstandardson financialdata
dissemination,financial sector soundness, and fiscal,
monetary,and financial transparency.Other standard-
settingbodiesareworkingon bankruptcy,corporategov-
ernance, securitiesmarket regulation,and accounting
and auditing.
But effortshavestalledin other areas.For example,
there has been littleprogressin settingup earlywarning
devicesthat could alert the internationalcommunityto
danger.8 Effortshavebeen similarlyunproductivein de-
signingclearguidelinesforprivatesectorinvolvementin
crisispreventionand resolution,which can limit moral
hazard,strengthenmarket disciplineby fosteringbetter
risk assessment,and improve the prospects for both
debtors and creditorsin debt workouts.There is a risk
that an apparentlackof urgencyin the aftermathof the
Asianrecoverycouldleadto inaction-but historyteaches
that more crisesare a realpossibility.
Recognizingthis, developingcountriesmay wish to
implement short-term safeguardsto limit their expo-
sure.9 These safeguardsare of two types: controls on
capitalflowsand measuresto enhanceliquidity.Controls
on capital-including Chilean-typetaxes on inflows,
quantitativecontrols on the banking sector's interna-
tional short-termliabilities,and restrictionson capital
outflows-have their problems,rangingfromevasionto
implementationdifficulties and opportunisticimposition.
They can alsorestricta country'saccessto much-needed
capital.But eachtypeof control can be effectivein some
situationsin dampeningthevolatilityof capitalflows,thus
helpingto preventcrises.
One wayof enhancinga country'sliquidityis to main-
tain higher reserves.But besidesbeing expensivefor the
GLOBAL FORCES FOR POOR PEOPLE I81
governmentand perhapscreatinga significantfiscalbur-
den, evenlargereservesarelikelyto be inadequatein some
situations.An alternativeis to imposehigherliquidityre-
quirements on the banking sector,effectivelyshifting
the burden of holding reservesto the privatesector (and
possiblymakingbankssafer,with beneficiallong-termef-
fects).Anotheris to contractwith an institutionfora con-
tingentcreditline. Bothprivatebanksand the IMF offer
such arrangements,which provide varying degrees of
automaticaccessto creditat predeterminedinterestrates.
Even if these short-term safeguardsare put in place
countrieswill often be unable to withstandseriousin-
ternationalvolatility.That is whyprioritymust be placecd
on increasingthe momentumfor internationalsystemic
financialreformsthat promote stabilityand ensurethe
availabilityofliquidityfor countriesfacingsevereadverse
shocksor hit by economywidecrises.
Producing pro-poor international
public goods
Many of the challengesfacingpoor countrieshavesolu-
tions that involvethe production of internationalpub-
lic goods.One importantcharacteristicof public goods
is the difficultyof restrictingpeoplefromconsumingthem
without paying-free riding-once they are produced.
This characteristicmeans that if production of public
goodswereleft to the market, therewould be an under--
supplyunlessthe governmentsteppedin to producethe
goods or to provide incentives(such as subsidies)for
their production.Governmentshavelong intervenedin
thisway,providingsuchnationalpublicgoodsasdefense,
infrastructure,lawand order,and rulesand standards.
The problem is more complexforinternationalpub-
lic goods,such as control of communicablediseasesor
researchto raiseyieldsin agriculture.Just as for national
publicgoods,the incentives-for countriesor for the pri-
vate sector-to produce internationalpublic goods are
weakor absent.Butthereis no worldgovernmentto help
spurthe productionof these goods-countries must de-
cideto cooperateto producethem.Today,asinternational.
problemsgrow more pressing,attention is focusingon.
how this cooperationcan be achieved. 10
Indeed, internationalcooperationhas had some re-
markablesuccessesin producing and spreadingpublic
goods.The green revolution-one of the 20th century's
most important developmentadvances-was an out-
come of internationalresearchon high-yieldingplant
I82 WORLD DEVELOPMENT REPORT 2000/200I
Box 10.1
A successstory: the fight against river blindness
in Africa
The internationaleffort to controlriver blindness(oncho-
cerciasis) is oneof the mostsuccessful programsinthe his-
tory of development cooperation. A painfulanddebilitating
diseasecausedby aparasitic worm,riverblindness hasbeen
virtuallyeliminated in the 11WestAfricancountries included
in the Onchocerciasis ControlProgram.Beforethe program
beganin 1974,morethana millionpeoplewereinfectedwith
the disease,sufferingfrom itching,disfigurement, eye le-
sions,and,for 100,000 of them,blindness. Whenthe program
windsdownin2002,aftera28-year effortto eliminate theblack
flies that carrythe parasite,34 millionpeoplewill be pro-
tected,600,000 casesof blindness will havebeenprevented,
and5 millionyearsof productive laborwill havebeensaved.
Partnersin the programhaveincludedAfricangovern-
ments,localcommunities, internationalorganizations, bilat-
eral donors,corporations, foundations,andNGOs.A key
contributorhasbeenthe MerckCorporation, whichhasdis-
tributedthe drugivermectinfree of charge.
While the programhas been highly successful,on-
chocerciasis remainsa problemincountriesoutsidethe pro-
gramarea.Soin1996the AfricanProgram for Onchocerciasis
Controlwascreated, extending the effortto controlriverblind-
nessto the 19 remainingAfricancountrieswhereit is en-
demic. Seventydevelopmentpartnersparticipatein this
project.
Source:WorldBank
lwww.worldbank.org/gper).
varietiesat institutesestablishedaround the world expressly
to develop technologies to relievethe world's food prob-
lem. More recently, international cooperation in the
campaign againstriverblindnessin Africabrought tremen-
dous benefits to 11 poor countries (box 10.1). Another
success story is the Montreal Protocol on ozone deple-
tion: 165 parties to the protocol agreed to full phaseout
of 94 ozone-depleting substances.
Still, international public goods have receivedrelatively
little attention in international cooperation."l And there
have been failures-the Kyoto Protocol on greenhouse
gasesthat contribute to globalwarming, for example, has
languished. 2 Given the potential that some public goods
hold for poverty reduction, more attention to ensuring
their provision is warranted. The benefits of such goods,
and the difficulty of creating the right incentives for
their production, are well illustrated by the attempts to
control infectious diseasesand boost agricultural yields-
two international public goods that would do much to
help poor people. There are many others, as well.
Figure 10.2
The burden of HIV/AIDS is heavilyconcentrated
in Sub-SaharanAfrica
Shareof adultsinfectedwith HIV,1997
Percent
8
7
6
5
4
3
2
1
*_*_ _ _ _
0 Low-incomeMiddle-incoeHigh-income
Sub-Saharan
countries countries countries Africa
Source:
WorldBank2000s.
Controlling infectious diseases
The potential benefits of international cooperation to con-
trol infectious diseases are exemplified in the AIDS pan-
demic. More than 34 million people worldwide are
infected with HIV, and more than 18 million have died
of AIDS.' 3 The epidemic continues largelyunabated: 5.4
million people were infected with HIV in 1999, and
some 15,000 are infected every day. AIDS has no cure-
nor is there yet a preventive vaccine. More than 90 per-
cent of the infections are in the developing world, nearly
70 percent in Sub-Saharan Africa (figure 10.2). Despite
its concentration in developing countries, AIDS poses a
risk to all countries-not only through its health effects
but also through its destabilizingeconomic and social ef-
fects. 4 The United States this year classifiedAIDS as a
national security risk.
While preventivebehavior is the key to controlling the
epidemic, an effectivevaccine could help dramatically.15
But progressin developing a vaccine has been slow.More
than 25 candidate vaccineshave been tested, but only one
is in large-scaleefficacytrialsin humans.There are two main
reasonsfor the slowprogress.The first is scientific:the cor-
relatesof HIV immunity are unknown, and many differ-
ent approaches will probably have to be tested in parallel,
with little certaintyabout their effectiveness.This raisesthe
up-front costs and risks for investors in AIDS vaccine de-
 HARNESSING
velopment.The second reason is economic:investors
would likelytake the risksassociatedwith researchif de-
mandweresufficient, but therearetoofewmarketincentives
to investin an AIDS vaccinethat would be effectiveand
affordablein developingcountries.Africa,forexample,ac-
countsforonly 1 percentof world drug sales.
The resultis that internationalinvestmentin research
and developmentfor an AIDS vaccine is quite low-
$300-350 milliona year." 6 Of this, $50-120 millionis
estimatedto comefrom the private sector,which has a
crucialrolein convertingresearchto productdevelopment
and distribution.And most of the researchfocuseson a
vaccinethat could be marketedin North Americaand
WesternEurope.Only about $ 10-25 millionis spentan-
nuallyon developmentof a vaccineforthe virussubtypes
andhealthsystemsofdevelopingcountries.17 In contrast,
more than $2 billionis spent eachyear on researchand
developmentforAIDS treatment,much of it in the pri-
vate sector,driven primarilyby the market represented
by the 3 million people with HIV/AIDS in industrial
countries.
Whatis true forAIDS istrue for otherdiseasesaswell.
TheWorldHealthOrganization estimatesthat only 10per-
cent of the $50-60 billionin healthresearchworldwide
eachyeargoesforthe diseasesthat afflict90percentofthe
world'speople.18 Developingcountriesaccountfor only
about 8 percent of world spendingon researchand de-
velopment,mainlybecausetheylack resources. 1 9 Of the
1,233new medicinespatentedbetween1975and 1997,
only 13 (1 percent)were fortropicaldiseases.The effect
of the researchand spendinggapsis devastating:malaria,
tuberculosis,and AIDS cause5 milliondeathsa year-
about 9 percentof alldeathsin the world-most of them
in developingcountries.Even when medicalremedies
exist,countriesmaynot be ableto affordthem. Despite
an effectivevaccine,hepatitis B still kills some 92,000
peoplea year,and chronichepatitisB contributesto an-
other 700,000deathsthroughcirrhosisandlivercancer. 20
About350millionpeoplearechronically infectedhepatitis-
B carriers,ableto transmitthe diseasefor manyyears.
The internationalcommunitycouldaccelerate progress
on vaccinesin twoways.First,internationalorganizations
and nationalgovernmentscould "push"researchand de-
velopment by subsidizingor reducingthe costsof vac-
cine developmentand strengthening the capacity of
developingcountrieswith a strong scientificbase to be
partners in vaccineresearch.For example,in 1996 the
RockefellerFoundationlaunchedthe InternationalAIDS
GLOBAL FORCES FOR POOR PEOPLE I8,3
VaccineInitiative,an internationalnonprofitthat stim-
ulatesinvestmentin and demandforAIDS vaccinesfor
globaluse.The initiativeworkswith the publicand pri-
vate sectorson targetedsupport to researchand devel-
opment for novelvaccineapproachesand on measures
to reduce obstaclesto private investment.Donor gov-
ernments,fortheirpart, could providetax breaksor sub-
sidiesforproductdevelopmentrelevantto poorcountries.
Second,the internationalcommunitycould demon-
strateor ensurea substantialfuturemarketin developing
countriesforvaccines.It couldpledgeto fullyimplement
programsfor the childhoodvaccinesalreadyon the mar-
ket (immunizationratesin many countrieshaveslipped
in the pastdecade).To ensurea largemarket forvaccines
in poor countries,it couldcreatea fund or other credible
precommitmentmechanismforpurchasing,for the poor-
est countries,many dosesof vaccinesshownto be both
21 Pricesshouldcovernot justpro-
effectiveand affordable.
ductioncostsbut someof the researchcostsaswell.Mul-
tilateraldevelopmentbanks mightalsoissuecontingent
loansfor vaccinepurchaseto developingcountries,to be
releasedonce a vaccineis developed.Similar arrange-
ments couldbe put in placefor other medicaladvances.
Boosting agricultural yields
Likeadvancesin medicalresearch,advancesin agricul-
tural technologycan haveprofound effectson the lives
of poorpeople(box 10.2).The greenrevolutionisamong
the most famous examplesof an international public
goodusedfor development.The revolutionbeganwhen
foundations,governments,and NGOs took the lead in
tryingto transferto farmersin developingcountrieswhat
scientistsalreadyknew about plant geneticsand new
high-yieldingvarietiesof grains.Privatecompanieshad
shownlittle interestbecauseof the difficultyof making
an adequate return on investmentsin new varieties--
farmers could simply collect seeds from the original
plants. Complementarypublic effortsat the national
levelwereessential.Many developingcountries(suchas
Braziland India) establishednational agriculturalre-
searchorganizationsto developsecond-generation mod-
ern varietiesbetter suitedto local conditions.They also
set up agriculturalextensionservicesto disseminatethe
knowledgeto farmersand get feedbackon the new va-
rietiesand cultivationtechniques.
Theseeffortshad a dramaticeffecton the livesof the
ruralpoor.In Africathe adoptionofimprovedmaizeraised
yieldsan estimated12-14 percent,with gainsas high as
I84 WORLD DEVELOPMENT REPORT 2000/2OOI
Box 10.2
Research,maize, and pigs in rural Guizhou
Anyonewho doubtsthe impactof agricultural
researchon
farmincomeandhouseholdfoodsecurity(andthuspoverty)
shouldvisitruralareas
in Guizhou,
thepoorest
province
of
China.In remotevillages,on smallfarmsset in the moun-
tainous
countryside, therehasbeenanalmostmiraculous turn-
aroundin the livesof poorpeoplethanksto the introduction
of qualityproteinmaize.
Until recentlyannualincomeswere lessthan $50 per
capita,andfor upto threemonths a yearfamilieshadvirtu-
allynofood.ThenhybridswereintroducedinGuizhou in1994.
Qualityproteinmaizeishigheryielding
thanconventionalva-
rieties,butmoreimportant,it hashigherlevelsof two es-
sentialaminoacidsvitalforthe growthof children.
Todaythe
local peopleare better fed, and surplusmaizehas been
usedto producepork,increasing
foodsecurity
anddispos-
ableincomes.Theextraincomehasbeenusedfor yield-
enhancinginvestmentssuchasirrigation.
Havingtransformed the livesof 25,000familiesinGuizhou,
cultivationof the hybridvarietyof maizeis beingadaptedto
neighboring
provinces.
Source:
Bale1999.
40 percent reported in areas with favorableconditions.2 2
A survey in southern India concluded that the averagereal
income of small farmers rose 90 percent in 1973-94
and that of the landless-among the poorest in farming
communities-125 percent.23 Higher productivity also
brought lower prices. It has been estimated that wheat
priceswould have risen 34 percent more in 1970-95 with-
out the international agricultural research efforts-rice
prices, 41 percent more. And because of the lower prices,
1.5-2 percent fewer children in developing countries
are malnourished.2 4
Despite these advances, the growth rate of cereal
yieldsin developing countries has been declining steadily,
from 2.9 percent a year in 1967-82 to 1.8 percent in
1982-94. With demand for foodgrains in developing
countries predicted to increase 59 percent in the next
25 years, the challenge for agriculture remains signifi-
cant, particularly if yield growth is to be environmen-
tally sustainable.2 5
One tvpe of technology that might make a significant
difference is biotechnology-using living organisms to
make or modify products to improve plants and ani-
mals. With far greater speed and accuracy than conven-
tional technology, biotechnology can identify desirable
traits and introduce them into plant and animal strains
(an example of such traits is increased nutritional qual-
ity, as in vitamin A rice). More research is needed on the
potential benefits and risks of specific uses of biotech-
nology in developing countries. But it is likely that
biotechnology,if steered by the rightpolicies, including
biosafetymeasures could be a key part of the solution
be
to the problemsof food security and poverty.
So far, however, biotechnology has had littleimpact
in most developingcountries. Unlike the advances of the
greenrevoluton, much of the progressi biotechnology
has been concentrated in the private sector. Govern-
ment funding of agricultural research, so crucial in the
green revolution, has stagnated or even declined, a ca-
sualty of general fiscal restraint and a more skeptical view
of the social benefits of investingin science (despite the
high returns on agricultural research).2 7 Private institu-
tions now hold a majority of the patents in biotechnol-
10.3). Because the knowledge is private, the cost of ac-
quiring it is much greater. Figuring out how to allow de-
veloping countries to capitalize on advances in
biotechnology research remains a key challenge for pol-
icymakersconcerned with food security and poverty.Part
of the answer may lie in how intellectual property rights
are used.
Safeguarding the interests ofpoor people
in the intellectualpropeerty rights regime
Intellectualproperty rights are important for encouraging
innovation, particularly in such areas as medicine and
agriculture.When creatorsof knowledgedo not retain ex-
clusiverights of ownership for a period of time, there is far
lessincentive to produce new knowledge.This was one of
the arguments for layingdown standards under the Agree-
ment on Trade-Related Aspects of Intellectual Property
Rights (TRIPS), negotiated in the Uruguay Round of
trade negotiations in 1986-94. But intellectual property
rights can sometimesprevent the distribution of potential
international publicgoods helpful to poor countries,which
can seldom afford the prices charged by patent owners.2 8
Three trends in intellectual property rights are partic-
ularly worrying to developing countries. The first is that
basic researchand knowledgeare increasinglybeing gen-
erated by private companies alone. The second is that in-
dustrialcountries continue to account for the vast majority
of patents worldwide-97 percent.2 9 Only 31 of the
26,088 applications for patents filed in 1997 under the
auspicesof the African Intellectual Property Organization
 HARNESSING GLOBAL FORCES FOR POOR PEOPLE I85

Box 10.3 Developingcountrieshaverespondedto thesetrends

Most biotechnology
patentsare private by proposing safeguardsfor the intellectualproperty
rights regime.Amongthem:
Thepublicsectorisofteninstrumental inpioneering biotech- * Recognizing the rightsof farmerscultivatingtraditional
nologyresearch,
latertransferring
it to privatefirms.Thatpat- varieties.
ternisevidentintheutilityandplantpatents
directly
involving
insecttoxicityof theBacillus (Bt)microorgan-
thuringiensis * Prohibitingthe patentng of lfe formsor biological
ism. Until 1987the publicsectorheldthe majorityof the processes.
patents.Sincethentheownershipof patentsinforce(whose * ReconcilingWorldTradeOrganization(WTO) provi-
overall
number hasincreased)hasshifted
dramatically
toward sions on intellectual property rights with the Interna-
theprivatesector(seefigure).Patents
arenowparticularly
concentrated
inthe"big6," thesixlarge
corporations
actively tionalConventonon Biodiversiyand theInternational
consolidating
theirglobal
positions
inagricultural
biotechnology Undertaking on Plant GeneticResources.
research,
intellectual
property,
andmarkets (Dow,Novartis, * Ensuringaccessto essential medicinesatreasonablecost.
Aventis,
Monsanto,
AstraZeneca,
andDuPont). Negotiatinga new intellectualpropertyrightsregime
that encouragesprivateinnovationwhilesafeguarding the
Holdingsofbiotechnology
patentshaveshifted interestsof poorcountriesand poor peoplein the benefits
sharplytowardtheprivatesector of that innovationwill taketime and much debate.Like
Share
ofBacillus patents
thuringiensis by the productionof allinternationalpublicgoods,it will
typeofholder cetn atcpto
Percent requirecreatingincentivesfor participationby all those
100 * Universities with an interest in the outcome, includingthe private
andpublic sector.
institutionssetr
80 El Independent
biotechnology Ensuringa voicefor poor people
60 firms
and
individuals iin global
lblfrmforums

40
nOthercorporate
entities Actionswitha globalreacharegenerallydiscussedin global
_ Big6 and internationalforums,such as nation groups,inter-
20 * d d nationalorganizations,and UnitedNationsconferences
r0 * [ [ and other gatherings.Ensuringthat poor countries,and
1987 1994 1999 especially
poorpeoplein thesecountries,havea strong
voicein theseforumswillhelp ensurethat these institu-
tions respondto the needs of poor people. Productive
partnerships-whether to agreeon standards,produce
public goods, or work toward other common goals--
werefrom residentsof Africa.And only 7 of 25,731 ap- requirethat all partnershavean effectivevoice.
plicationsregisteredthat year by the AfricanRegional
IndustrialPropertyOrganizationwerefiledbyresidents.30 Strengtheningthe capacity ofpoor countries
The third trend is that genetic science-enabling to represent their interests
companiesto patent such innovationsas recombinant Not all partnershipsshould be global-because not all
DNA techniques,monoclonalantibodies,and new cell internationalproblems are global. Solutionsto an in-
and tissuetechnologies-is gainingprimacy.This raises ternationalproblem-such as river blindnessor pollu-
a concernthat a systemof propertyrightsdesignedto pro- tion in a lakeborderingtwo countries-should be guided
tect industrialmachinerymay not be ableto copefairly primarilyby the countriesaffected.3 2 If those countries

and effectivelywith the complexitiesof geneticallyma-
nipulatedorganisms. 31 In somecasesbreedersof plantva-
rietiesprotected by patents can prevent farmers from
reusingharvestedseed.And if broadlywritten, patents
on biotechnology processessuchasresearchtoolscandeter
inventionin other fieldsusingthe sameprocesses.
need assistance,financial or otherwise, the assistance
should go to the smallestrelevantgroup-for example,
the Economic Community of West African States for
cross-borderproblemsinvolvingonly its memberstates.
This principleof subsidiaritycan be applied all the way
up the geographicscaleof internationalpublicgoods,but
i86 WORLD DEVELOPMENT REPORT 2000/2001
it must also be reconciled with economies of scale and
scope.33
Subsidiarity implies that regional institutions should
be significantlystrengthened to handle cross-borderprob-
lems. Given the importance of ownership, such institu-
tions would in many cases be a better choice for solving
local problems than such global institutions as the World
Bank and the United Nations. And because most re-
gional institutions lack wide-ranging expertise, sector-
specific organizations should also be strengthened to
assist when needed.
But many problems are global, and participation by
developing countries in finding solutions is just as im-
portant as for regional problems. Since international in-
stitutions will generally facilitate the discussions of
global problems, these institutions need to take the lead
in making information available, ensuring all parties a
seat at the table, and strengthening countries' capacity
to analyze issues and effectively communicate their
interests.
Because knowledge is essential to decisionmaking,
international organizations must place a premium on
transparency in information and in their operations. In
addition to publishing as much information as possible,
they need to ensure independent evaluation of their ac-
tions-to make themselves more accountable and more
effective. This is the direction in which international or-
ganizations have been moving in the past few years.
Even with all the right information, developing coun-
tries cannot represent their interests without a seat at the
table. Many global decisions continue to be made mainly
by the group of seven largest industrial democracies (the
G-7). Mechanisms are needed to ensure that developing
countries contribute effectivelyto those decisions.3 4 Bet-
ter progress has been made in discussions about the in-
ternational financial architecture. In 1999 the Group of
20 was establishedto conduct ongoing discussionson pre-
venting and managing systemicfinancial crises.Seven de-
veloping countries (Argentina, Brazil, China, India, the
Republic of Korea, Mexico, and South Africa) are part
of this group. Still, the arrangement lacks formal provi-
sions for including any of the poorest or smallest coun-
tries, which, though not yet integrated enough into the
global economy to present a risk of starting systemic
crises, can certainly be affected by them. A better model
for integrating developing countries into global problem
solving is the Global Environment Facility,which works
to foster international cooperation to protect the envi-
ronment. Half the representativeson its council are from
developing countries (box 10.4).
In addition to participating in discussions and solu-
tions, developing countries must be able to represent
their own interests well-and this requires capacity
building. For example, poor countries are at a signifi-
cant disadvantage in WTO negotiations on such issues
as labor, the environment, and intellectual property
rights. Why? Negotiating in the WTO is a continuous
process, involvingas many as 45 meetings or more a week
by one estimate. Yet only two-thirds of developing
countries even have offices in Geneva, including only
12 of the 29 least developed WTO members, and these
offices frequently must represent the country at other
international organizations as well. Moreover, devel-
oping country officials often lack the expertise to par-
ticipate in the increasinglytechnical trade debates. It has
been estimated that almost 60 percent of the develop-
ing country members of the WTO are handicapped in
their participation.35
One attempt to addresssuch problems is the Integrated
Framework for Trade-Related Assistanceto Least Devel-
oped Countries, which seeksto enhance the trade-related
assistanceprovided by the six participating international
agencies and other development partners.36 Despite
"needs assessments" submitted by 40 poor countries,
progress has been slow, with new donor projects in just
one country (Uganda). Developing countries have ex-
pressed disappointment with the limited financial
pledges.37 At the request of donors, an independent re-
view is being conducted with the hope that the pro-
gram's weaknesses can be corrected. If the problems can
be resolved, the program could be a model for capacity
building in' other areas to help developing countries rep-
resent their interests.
Building global networks ofpoor people's
organizations
At last those above will hear us. Beforenow, no one ever
asked us what we think.
-Poor man, Guatemala
Like the voices of poor countries, the voices of organi-
zations of poor people are essential in ensuring that
global actions are targeted toward poverty reduction.
Such organizations, particularlywhen linked up in global
coalitions amassing strength and capacity, can have a
 HARNESSING
Box 10.4
The Global Environment Facility:a model for developingcountry participation
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
TheGlobalEnvironment Facility(GEF)is afinancialmechanism for
fosteringinternationalcooperation andactionto protectthe global
environment. Through grantsandconcessional financing,
it funds
the additionalcostsincurredwhen a national,regional, or global
development projectalsoaddresses environmental concernsre-
latedto biological
diversity,climatechange,international waters,
anddepletionof the earth'sozonelayer.Effortsto stem land
degradation arealsoeligiblefor funding.
TheGEFwasstartedin 1991,andaftera trialperiodwascap-
italizedby34 nations(including 13developing countries)
at $2bil-
lionfor fouryears.In 1998,36 countriesdonateda totalof $2.75
billionto keepthefacilityrunninguntil2002.Its governingstruc-
ture ensuresrepresentation by all stakeholders. The GEFas-
sembly,with representatives fromall 165participatingcountries,
meets everythree yearsto reviewgeneralpolicfes.The GEF
council,with representativesfrom32countries(16developing, 14
developed, and2 transitioneconomies), meetseverysixmonths
onoperational policiesandprograms. TheGEFSecretariat trans-
latesthe decisions of the assemblyandcouncilintoaction.
TheGEF'sthreeimplementing agencies-theUnitedNations
Development Programme, the UnitedNationsEnvironment Pro-
gramme,andthe WorldBank-developprojectsfor GEFfunding
Source:Porterandothers1998.
major influence on international debates. For example,
a coalition of the Jubilee 2000 movement and other
groups concerned with debt reduction worked closely with
international financial institutions and industrial coun-
try governments to forge a consensus for deeper, faster,
and broader debt relief for heavily indebted poor coun-
tries (chapter 1 1).
Innovative solutions are needed to increase poor peo-
ple'sconnections to each other and to global decision-
makers. The most important shift needed is in the
mind-set of global actors-to be directly informed by the
experiences of poor men and women who will be affected
by or are expected to benefit from global actions. Also crit-
ical is information technology, which can help build net-
works to channel the voices of the poor to global
decisionmakers. With the right tools and organization,
these networks can be powerful in spurring the integra-
tion of poor people's priorities and analyses into global
discussions.
One such global network of poor people is HomeNet.
It was created in the mid- I 990s by unions, grassroots or-
ganizations,and NGOs working with home-based work-
ers and street vendors in developing and developed
GLOBAL FORCES FOR POOR PEOPLE 187
andimplement themthroughexecuting agencies.
Theypartner with
a widevarietyof organizations to executethe projects,including
government agencies,otherinternationalorganizations,
privatein-
stitutions,andinternational,
national,andlocalnongovernmental
andcivilsocietyorganizations.
Eachparticipatingcountry has a politicalfocal point-the
contactpoint with the GEFSecretariatand other participating
countries-andanoperational focalpoint,whichidentifiesproject
ideasthat meetcountryprioritiesandensuresthatGEFpropos-
als areconsistentwiththem.Theseorganizations helpto ensure
countryownership,asdothe 16regionalNGOsthatdisseminate
information andprovidecoordination betweennationalandlocal
NGOsandthe GEF.
A recentindependent evaluationof theGEFfoundthatinashort
timeandwith few resources, it hadperformedeffectivelyincre-
atingnewinstitutional arrangements andapproaches andin lever-
agingcofinancingfor GEFprojects.It has also hada positive
impacton policiesandprogramsin recipientcountries,Although
there is room for improvement,particularly in effortsto main-
streamattention to theenvironment, theevaluatorsconcluded that
the GEFhadpotentialfor muchgreatersuccessandthat donors
shouldstrengthenit.
countries and concerned about the adverse impact of
globalization on the livelihoods of poor women in the in-
formal economy. HomeNet's objective was international
recognition of the rights of home-based workers, em-
bodied in an International Labour Organization (ILO)
convention. That convention was ratified by the ILO in
1996, thanks in part to an alliance of researchers at Har-
vard University and the United Nations Development
Fund for Women (UNIFEM), who compiled statistics
for HomeNet to make the informal economy visible. In
1997 the alliance of grassroots organizations, researchers,
and international organizations gave birth to WIEGO
(Women in Informal Employment: Globalizing and Or-
ganizing), a global network to promote better statistics,
research,and policy in support of poor women in the in-
formal economy. HomeNet, with active member orga-
nizationsin more than 25 countries,publishesa newsletter
that reaches organizations in more than 130 countries.
Strengthening such networks will fortify a much-
needed voice in international cooperation: the voice of
the poor themselves. Just as for national policies, their
voice is essentialin ensuring that globalpoliciesmeet their
needs.
I88 WORLD DEVELOPMENT REPORT 2000/200I
* * * for poverty reduction depend on policy changes in high-
The four areas of action highlighted in this chapter il-
lustrate the importance of international cooperation in
the fight against poverty. Many of the most pressing
problems in developing countries-from trade barriers
to financial crises to infectious diseases-can be solved
only with cooperation from high-income countries. Yet
in the past, international cooperation has consisted pri-
marily of financial transfers from rich countries to poor
countries, notably aid. But aid is not enough-prospects
income countries and cooperative actions at the global
level.These include lowering trade barriers, increasingfi-
nancialstability,producing international public goods that
particularly benefit poor people, and ensuring a voice for
poor countries and poor people in global forums.
The need for these international actions should re-
define the role of international cooperation in poverty
reduction. Even with more effective aid, the subject of
the next chapter, progress against poverty will be slower
without the international actions recommended here.

Reforming Development
Cooperation
to Attack Poverty
D evelopment cooperation is being reformed.
Fromthe relationshipbetweendonor and recipientto
the way in whichaid is deliveredand the framework
for debt relieffor the poorestcountries,many of the
old waysof assistingdevelopmentarebeginningto be
replacedby new forms.
Much of this is due to a reaffirmedcommitment
by the internationalcommunityto fightpoverty.The
World Summit for Social Developmentin Copen-
hagenin 1995set forth the goalof eradicatingpoverty
in the world through decisivenational actions and
internationalcooperation.Donorshaveincludedhalv-
ing povertybetween1990and 2015 and other targets
amongtheir internationaldevelopmentgoals(seebox
2 in the overview).'In the 12th replenishmentof the
InternationalDevelopmentAssociation (IDA)in 1998,
donors reaffirmedtheir missionto supportprograms
to reducepoverty and improvethe quality of life in
IDAs poorest member countries. 2 The Jubilee 2000
movementhelped put deeper debt reliefat the heart
of developmentcooperationstrategiesfor povertyre-
CHAPTER 11
duction. And donors are working to resolvedifferences
in approaches to poverty reduction through the
OECD'sDevelopmentAssistanceCommittee(DAC),
whichexpectsby mid-2001to agreeon guidelinesfor
povertyreduction to help donor agenciesmaketheir
programsmore effective. 3
But while the international community's com-
mitment to attack povertywas strengthenedin the
1990s,officialdevelopmentassistanceshrank.This,
despite the optimism at the start of the 1990s that
development cooperation would reap a post-cold
war "peacedividend"from cutbacksin militaryspend-
ing.4 Indeed, after peakingin 1992 (in real terms),
officialdevelopmentassistancefell consistentlyover
the decade-despite the robust economicgrowthof
DAC countries-rebounding only slightlyin 1998
during the global financial crisis (figure 11.1). Six-
teen of the 21 DAC countries spent a smallershare
of their GNP on developmentassistancein 1997-98
than in 1988-92.5The regionaldistribution of this
aid remained roughly constant between 1987 and
189
I90 WORLD DEVELOPMENT REPORT 2000/200I
Figure 11.1
While donor countries'economiesgrew after
1992, their development assistanceshrank
GNPand officialdevelopmentassistance
per capitain
DACcountries
U.S. dollars
30,000
29,000
28,000
27,000
26,000
25,000
10 Officialdevelopmentassistance
75
50
25
2519932 1993 1994 1995 1996 1997
Note:GNPis in 1995U.S.dollars,officialdevelopmentassistance
in 1998U.S. dollars.
Source:WorldBankstaff estimatesbasedon OECDand World
Bankdata.
1998, apart from an increase in the share going to Eu-
rope and Central Asia (figure 11.2). But total develop-
ment assistancefell in every region except Latin America
and the Caribbean after 1992-93 (figure 11.3). Pre-
liminarv estimates show that official development as-
sistance rose again in 1999, by about 5 percent, though
it is too soon to know whether this reflects more than
the response to the Asian crisis and indicates a much-
needed real and sustained reversal of the downward
trend in the 1990s.
The decline has been costly for many countries. Al-
though it has coincided with massive inflows of private
capital to developing countries, very little of that capi-
tal goes to the poorest countries. Net private capital
flows to low- and middle-income countries reached
$268 billion in 1998 and now dwarf aid flows in some
countries. Overall, private flows to developing coun-
tries surged during the 1990s, from 43 percent of total
resource flows in 1990 to 88 percent in 1997, just be-
fore the East Asian financial crisis. However, inflows of
private capital have been concentrated in relatively few
countries; a large number of countries receive little or
nothing. In 1997, before the financial crisis, the top 15
developing country recipients received 83 percent of
Figure 11.2
With the exception of Europe and Central Asia,
the regional distributionof officialdevelopment
assistanceremained roughly constant ...
Averageannualshareof net receiptsof official
developmentassistance
Percent
35
30
25
20
20 1987-88
15 1997-98
.992-93
~~~~~~~~~~~~~~~~~10.-
5
1998 0 Eas Asia Europe Lati
n Middle South Sub-
and and AmericaEastand Asia Saharan
Pacific Central andthe North Africa
Asia CaribbeanAfrica
Note:Sharesdo not sumto 100percent,as someaid is not
allocatedby countryor region.
Source:OECD,DACdata.
private capital flows to developing countries, leaving
some 140 developing countries and territories (with
about 1.7 billion people) to share the remainder. Almost
entirely left out were the 61 low-income countries be-
sides China and India.6 For example, all of Sub-Saharan
Africa received only 1.2 percent of flows to developing
countries in 1998. These are the countries that need aid
most, and they are hit hard by its decline.
There is no single reason for the decline. Donors ini-
tially cited their fiscal deficits as a large part of the prob-
lem. Yet even as these deficits declined (from 4.3 percent
of GDP in 1993 to 1.3 percent in 1997), official devel-
opment assistancecontinued to shrink, dropping 14 per-
cent from 1996 to 1997.' A more likely explanation is
that donors continue to view development cooperation
through a strategic lens rather than a poverty lens, see-
ing other uses for their money as strategically more im-
portant. Historically, aid flows have been determined
more by political and strategic interests than by poverty
reduction goals.'
Perhaps more noteworthy is the decline in support
from the traditional proponents of official development
 REFORMING DEVELOPMENT COOPERATION TO ATTACK POVERTY
Figure 11.3
... while receiptsfell after 1992-93 in all
regionsbut LatinAmericaandthe Caribbean
Billions of U.S. dollars
20
15
10 1992-93
|
1997-98
| * i
*997-98
EastAsia Europe Latin Middle South Sub-
and and AmericaEastand Asia Saharan
Pacific Central andthe North Africa
Asia Caribbean Africa
Source: OECD, DAC data.
assistance. The preeminence of geopolitical interests is
not new.9 But what is new is the falloff in countervail-
ing support from advocates for development assistance
on humanitarian grounds. Many fell victim to "aid fa-
tigue" and were far lessvocal supporters in the 1990s than
before.
Not everycountry was affectedby aid fatigue-indeed,
aid flows increased from some countries-but its symp-
toms were clearly evident. For example, in the United
States a comprehensive poll found that an overwhelm-
ing majority of the population favored foreign aid in
principle-and that only 35 percent thought it should
be cut from current levels.10 Yet more than 80 percent
of respondents believed that waste and corruption kept
foreign aid from reaching the people who need it. This
kind of public disillusionment may have made it harder
for donor governments to maintain foreign aid, let alone
increase it. If aid is not working, the sentiment goes, the
money could be better spent elsewhere.
In contrast to the rise in aid fatigue in some placeswas
a major upsurge in support and activism around debt re-
duction, most notably under the auspices of the Jubilee
2000 movement of religious organizations and other
civil society groups. They rallied around the cause of
191
cutting debt for poor countries to support poverty re-
duction and human development. So there is clearly
continuing support for the principle of providing re-
sources for improving the lives of poor people in the de-
veloping world, but widespread questioning of the
traditional mechanisms for providing such resources.
Is aid working? Can it work better? What is the role
of debt reduction in concessional support? Developing
countries themselveswill largelydetermine through their
own policies whether they achieve the international de-
velopment goals. But aid and debt relief can provicle
crucial support. So finding out how to make these moire
effective-and then doing what it takes-remains vital.
In answering these questions, this chapter outlines a
vision for a better system of development cooperation,
one based on new thinking and new practices. This vi-
sion includes a reformed framework for country-focused
aid and debt relief for the poorest countries-under-
pinned by a renewed emphasis on the policy and insti-
uional envirent and the fundamental prioit of
tutional environment and the fundamental priority of
poverty reduction. Donors would work in partnership
with countries, directing aid and debt reliefalong the lines
of a broad-based poverty reduction framework (as ad-
vanced by many donors and laid out in this report), sup-
porting countries that can put these resourcesto good use
for poor people.
Supporting good policiesand institutions is important,
but it is not enough. We learned in the 1990s that process
is as important as policy in foreign aid and the manage-
ment of unsustainable debt burdens. The way donors and
recipients interact strongly influencesthe effectivenessof
development cooperation. Relationships have tended to
follow the preferences of donor countries, leaving recip-
ient countries with little sense of ownership of the aidl-
financed activities. Along with advancing a broad-based
poverty reduction framework, this report emphasizes
how much local realities matter in development.That aid
relationships have too often failed to take local realities
into account, undermining ownership, is an important
flaw.
If development cooperation is to attack poverty ef-
fectivelyand efficiently, donors will need to:
* Pay more attention to local conditions and country
ownership.
* Deliver aid in ways that intrude less on government
functions, including greater use of sectorwide ap-
proaches and a movement away from old forms of aid
conditionality.
192 WORLD DEVELOPMENT REPORT 2000/200I
* Provide sustained support for policy and institutional
environments that are strongly conducive to poverty
reduction, in preference to ones that are not.
The chapter begins by exploring how these new ap-
proaches can make aid more effective. It then examines
the issues associated with relieving the debt problem of
poor countries.
Making aid more effective in reducing
poverty
Recent studies confirm what anecdotal evidence has long
hinted: the experienceof aid has been mixed.1 Early pre-
dictions that aid would close the financing gap that pre-
vented developing countries from moving ahead have not
come to pass. If all the aid that went to Zambia between
1961 and 1994 had gone into productive investment, and
if investment had been as important to growth as initially
predicted, the country's per capita income would have
been more than $20,000 in 1994, not $600.12
And yet there have been many aid successes.The On-
chocerciasis Control Program is but one example (see
box 10.1). Aid was important, in different periods, in East
Asias extraordinary successin poverty reduction over the
past few decades. The rapid progress in Vietnam in the
1990s is another example. So aid can work. The challenge
for the international community is to understand how to
make it work consistently-and then to do what it takes.
The key problems with aid
Aid's difficulties in reducing poverty go deeper than the
sway of geopolitical interests over development inter-
ests, which has often directed aid to countries whose
policies were not focused on reducing poverty. Aid has
been hindered by the frequent differences in donors'
perspectiveson development policies,even though the past
50 years have been punctuated by times of relatively
wide consensus on the best way to pursue development.i3
Donor differences have played a key role in preventing
aid from achieving full effectiveness.Donors have often
failedto coordinate their efforts, countries have not taken
ownership, and there has been heavy use of condition-
ality both at the project level and economrywide.
In the first two decadesafter World War II state-led in-
dustrialization was generally seen as the best way to pur-
sue development, a consensus undone in the 1970s by
world events, including the demise of the fixed exchange
rate systemand two oil shocks,which had devastating im-
pacts on developing countries. It was widely believedthat
government interference in the economy had prevented
developing countries from adjusting to these shocks.Sub-
sequently, a new consensus began to form, eventually to
be known as the "Washington consensus" (seebox 4.1 in
chapter 4) .4 To many, including staff at the World Bank
and other multilateralfinancialinstitutions,fiscalprudence,
free markets, and outward orientation had clearlydemon-
strated their superiority as the most efficient way for
15
countriesto growand develop.
But it has become clear that simple strategies for de-
velopment and poverty reduction are elusive.While mar-
kets are a powerful forcefor poverty reduction, institutions
that ensure that they operate smoothly and that their ben-
efits reach poor people are important as well. As the 21st
century begins, donors are coalescing around a develop-
ment strategy that includes investing in people through
health and education services,promoting inclusive and
equitable growth, supporting good governance, and pro-
tecting the environment. IG This strategy also recognizes
the centrality of local conditions: that the most effective
development policies will vary by situation.
Despite this growing consensus on the broad devel-
opment framework,agreementon the right policiesin par-
ticular conditions has tended to elude donors and
recipients. Donors come to development problems with
their own mandates, histories, ideologies,and politicalre-
alities and often do not see situations in the same way as
other donors or the recipient countries. Even in health
and education, which all donors agree are essential, the
right reforms are open to debate. As an analyst com-
mented, there is "a bewilderingmultitude of national sys-
tems and experiences, with varied (and hotly debated)
advantagesand disadvantagesassociatedwith each."17 So
while the days of adhering strictly to either state-led or
market-led solutions are over, between these extremes lie
a host of options, and the debate on them is far from over.
The lack of consensus on the broad outlines and the
details of national and local policies and projects has re-
duced the effectivenessof development assistance.1 This
effect is especially evident in problems of ownership,
donor coordination, fungibility, and conditionality-
the four main issues affecting aid in the 1990s.
Ownership.Because donors and recipients often dis-
agree, donors have looked for ways to ensure that their
money is spent as they intend. They have run their own
projects, required detailed reports from countries on pro-
jects, and attached conditions-usually policy oriented-
 REFORMING DEVELOPMENT
to the use of funds. A major study on relations between
donors and African recipients found that "in spite of
some improvements, donors still tend to dominate the
project cycle and pay inadequate attention to the pref-
erences of the government or project beneficiaries."1 9
These efforts to ensure that aid is spent effectively,evi-
dence now shows, have often had the opposite effect by
diminishing ownership by the recipient country.
Analysesshow that ownership is a key ingredient of aid
20
effectiveness. How stronglya country believesthat a pro-
ject or reform will bring benefits affects the effort put into
the activity,the domesticresourcescontributed,and the com-
mitment to the activity after the donor has left-all sub-
stantial determinants of success.To succeed, reforms and
projects must fosterownership by the people for whom the
policy or project is ostensibly being implemented.
Donor coordination.When different donor priorities
and project-relatedconditions (including donor-specific
reporting and procurement requirements) are multi-
plied many times over, they can create an unworkable en-
vironmentfor a recipientgovernment.Just the sheernumber
of donors and donor projects can be challenging.At one
point therewere 405 donor-funded projectsin the Mozam-
bican Ministry of Health alone. In the early 1990s in Tan-
zania there were 40 donors and more than 2,000 projects.
In Ghana during the same period 64 different government
or quasi-governmentinstitutions were receivingaid.2 1 Co-
ordinating these effortsto support a coherent development
strategy-even at the sector level-is nearly impossible.
Fungibility. Studies show that aid funds allocated to
a particular sector tend to free up for other purposes
money that the government would otherwise have spent
in that sector.22This means that in funding specific pro-
jects or sectors,donors may actuallybe helping to increase
spending on sectors they do not want to finance, such as
the military. This has profound implications for devel-
opment cooperation. Project-levelevaluations will not re-
flect the true impact of aid, since aid is likelyto be freeing
up resources for other activities.23
Even where resourcesare fungible, donor support can
still have some impact, from the design of certain poli-
cies to institutional development. Moreover,in countries
highly dependent on aid, donors as a group could lead
to shifts in government resource allocations, because of
the sheersize of flows.A potentiallyimportant part of this
is the preference of donors to support development bud-
gets, which can lead to a net shift in resources out of the
recurrent budget-not alwaysa good thing for develop-
COOPERATION TO ATTACK POVERTY 193
ment because of the importance of recurrent spending in
maintaining basic social and economic services.2 4
Conditionality. Donors know that even properly im-
plemented projects will have limited impact in poor pol-
icy environments.2 5 Awell-built schoolwill be useful only
if money is budgeted annually for teachers, books, and
supplies-and if the economic environment enableschil-
dren to go to school. The role of good policies and in-
stitutions in ensuring sustainable results suggests that
aid should flow more to countries with a good overallpol-
icy environment and good policies for poverty reduction.
But the relationship between good policies and aid flows
has not been strong.2 6
This finding would be understandable if aid were
spurring policy reform by influencing countries to change
their policiesor by helping them do so. This has been the
intention of many donors, and it is one reason (fungibility
is another) that many of them have reduced the share of
their portfolio allocated to projects and increased the
share allocated to program and policy-based aid.2 7 Most
program and policy-based aid has been tied to the en-
actment of certain policy reforms. But studies in the
1990s showed little systematic relationship between con-
ditionality and policy changes,though casestudies do find
positive effects under some conditions, especiaLlywhere
conditionality supports the hand of reforming groups.283
The dynamics between aid donors and recipients ex-
plain why conditionaLityfails.Recipientsdo not seethe con-
ditions as binding, and most donors are reluctant to stop
giving aid when conditions are not met.29 As a result,
compliance with conditions tends to be low, while the re--
leaserate of loan tranchesremainshigh.30 Thus aid has often
continued to flow despite the continuation of bad policies.
In addition to performing poorly in influencing
policy reform, policy-relatedconditions, often combined
with project-related conditions, severely burden de-
veloping country administrators-a problem that has
become more pronounced as conditionality has ex-
panded. Conditions on World Bank adjustment loans,,
having mushroomed in the 1980s, continued to grow
in the 1990s along with the expanding development:
agenda.3 1 As one recent assessment put it: "Although
much has been added to the conditionality menu since
1981, nothing has been taken off."32 The time gov-
ernment officials spent negotiating and monitoring
these conditions is time they could better have spent
analyzing development problems and designing de-
velopment strategies. Ownership has been shown to be
I94 WORLD DEVELOPMENT REPORT 2000/2001
central to the sustainabilityof both projectsand pol-
icy reform,and the fact that the deliveryof aid weak-
ens it is a fundamental flaw of current development
cooperation mechanisms.
Solutions that accommodate different
perspectives
While the dominant formsof donor-recipientrelations
havealloweddonors to pursue their own priorities,the
result has generallybeen a fragmentedsystemthat un-
dermines their efforts.The challengein reformingin-
ternationaldevelopmentcooperationis to accommodate
differentperspectiveson developmentwithout overbur-
deningthe recipientor underminingownership.
Achievingglobaluniformityin developmentstrategies
mightbe one solution,but historyshowsthat uniformity
is undesirable.Developmentis determinedto a greatex-
tent bylocalconditions,includingsocialinstitutions,so-
cial capability,ethnic fragmentation, inequality,and
geography.3 3 In studiesthesevariablessignificantly
explain
the variationin growthratesoverthe past30 years.34 Stud-
iesalsoshowthat externalshocks-and the abilityto re-
spond to them-can haveas much effecton growthas
policiesdo.35 The approachto designingdevelopment
strategiesshould thereforebe flexibleenough to adjust
to both internaland externalconditions.
This perspectivebeganto takehold in thedevelopment
communityin the late1990s.Combinedwith newthink-
ing on aid effectiveness, it has promptedproposalsto ad-
dress the problemsof aid. Three prominent themesare
ownershipand partnership, less intrusive aid delivery
mechanismsthat focuson the overallpolicyand expen-
ditureframework,and selectivityTogether,theyform the
agendafor the internationalcommunityto improvede-
velopmentcooperationin the comingdecade.
Ownership andpartnership. Recognizingthe importance
of ownershipandthe problemof donorcoordination,most
donorshaveembracedpartnershipasa guidingprinciple
in interactionsamongdonors,governments,and citizens
in developingcountries. 3 6 Most partnershipframeworks
havetwo parts.The firstis a partnershipbetweenthe re-
cipientgovernmentand its citizens,whoshareresponsi-
bilityfordevelopingtheir nationaldevelopmentstrategy.
Thisstrategycan takeshapethrougha consultationprocess
involvinggovernment,civilsociety,and the privatesec-
tor.The secondis a partnershipbetweenthe government
and donors,with donorsdesigningtheirassistancestrate-
gies to support the government'sstrategy.In the new
thinkingthe focusis on how to shapethis externalpart-
nership,or contract,in a waythat providesthe incentives
forcountry-driven, long-termpovertyreductionstrategies
whilealsostrengtheningthe internalpartnershipsneces-
saryfor socialstabilityand economicdevelopment.
Consultationsbetweengovernmentsand civilsociety
and betweengovernmentsand donorshavebeencarried
out in a number of countriespiloting the World Bank's
ComprehensiveDevelopmentFramework,the European
Union'spartnershipapproach,and othersuchapproaches.
The consultationsunder the ComprehensiveDevelop-
ment Frameworkhaveprovedfruitfulin severalcountries
-such as Bolivia,the DominicanRepublic,andGhana-
but havealsohighlightedthe needforgovernmentcom-
mitmentand forcapacityas keyingredientsof successful
consultations(box 11.1).
This emergingapproachto developmentcooperation
has been incorporated into the new initiative by the
WorldBankand InternationalMonetaryFund (IMF)to
link their supportof low-incomecountriesto nationally
designedpovertyreductionstrategies,workingwithinthe
principlesof the ComprehensiveDevelopmentFrame-
work (box 11.2). Concessionalfunds and debt relief
fromthe WorldBankand IMF willbe linkedto the goals
of povertyreductionstrategiespreparedby governments
in consultationwith civilsocietyorganizations,the pri-
vate sector,and donors.Basedon a good understanding
of the povertysituationin the country,the strategieswill
identifyactionswith the greatestexpectedimpactand set
up monitoringand evaluationprocesses.The goalis for
these strategies,describedin povertyreduction strategy
papers,to form the basisforassistancenot only from the
WorldBankand IMF,but from other assistanceagencies
as well.37 Similarinitiativesare underwayin the regional
developmentbanks.
Lessintrusiveaid deliverymechanisms focusingon the
overallpolicy and expenditureenvironmentDonors have
usedmanymeansto influencerecipientcountrypolicies.
Old forms of policy conditionalityhaveoften had dis-
appointingresults,dependingon countrycircumstances
and how the conditionality was used. Policy review
processesalsohavehad limited success.Publicexpendi-
ture reviews,for example,have evaluatedthe leveland
compositionof countries' expendituresand identified
waysto improveexpenditurepolicyand use donor funds
moreefficiently (seebox 9.2in chapter9). Butseveralstud-
ieshavefound this type of interventionto be ineffective
in manycases,largelybecauserecipientcountrieshavenot
beencloselyinvolvedin the reviews-and sohavefelt lit-
tle inclinationto complywith the findings. 38
 REFORMING DEVELOPMENT COOPERATION TO ATTACK POVERTY 195

Box 11.1
Learningabout the consultativeprocessthrough the ComprehensiveDevelopment Framework

In 1999the WorldBankannounced itsComprehensive Develop- All discussions

with donorsnow takeplaceinthe contextof
mentFramework, atoolfor improving countryownership anddonor the nationalactionplan.At a consultative groupmeetinginApril
coordination indevelopment cooperation.Theframeworkis based 1998donorspledged45 percentmorethan they hadin 1997.
on four principles: countryownershipof the policyagenda,part- Donorshavealsobeenencouraged to formulatetheirstrategies
nershipwithallstakeholders, attentionto socialandstructuralcon- in supportof the nationalactionplan.TheWorldBankrecently
cernsas well as macroeconomic andfinancialissues,and a redesigned itscountryassistancestrategyto alignit with theplan,
long-term,holisticapproach builton nationalconsultations. choosingto supportthreeof the plan'sfour pillars.Thegovern-
Thecountrydevelopsits national strategyin consultation
with mentcontinuesto leaddonorcoordination, chairingthe consul-
civilsocietyandtheprivatesector-andthen,withdonors,designs tative groupmeetingin Parisin 1999,where it presentedits
a matrixlinkingdevelopment goalsanddevelopment actors.The versionof the Comprehensive DevelopmentFramework. It has
activitiesof actorsinsupportof eachgoalarelistedin the matrix, alsoagreedwith donorson intermediate indicators
for monitoring
revealinganygapsor overlaps. outcomes.
Theframeworkis beingimplemented in13 countries,encour- Othercountrieshavenot progressed asfar.Thedifficultiesof
agingwideconsultation betweengovernments andtheircitizens
and somehighlight potentialproblemareas.Forexample, it is clearthat
enhancing partnerships with donorsin the designof comprehen- countryownership dependslargely on national capacity. Thecoun-
sivenational development strategies.But progresshasbeenvar- try mustbeableto holdbroadconsultations withall elementsof
ied,reflectingdifferentstartingdatesandcountrycircumstances. societyandto conductthe complexanalysis necessary to design
Boliviais anearlycase.In late1997the newgovernment em- nationalstrategiesthatbalancemacroeconomic andfinancialis-
barkedonananalysis of thecountry'sdevelopment challengesand sueswith social,structural,and institutionalconcerns.And,of
the preparation of anational actionplanto addressthem.A keypart course,the countrymust beableto implementthe strategy.
wasa national consultation with a widerangeof representatives Without this ownership-andthe countryleadershipfrom
of civilsociety-NGOs,unions,religiousorganizations, opposition it-donor coordination will remaindifficult.Whilethere is some
parties,andacademics-andthe privatesectorto discussdevel- evidencethatsomedonorcountriesaregraduallyaligningtheir
opmentconstraints andproposesolutions.Theresultsof thisna- strategieswiththoseof recipientcountries,strongerleadership
tionaldialoguewerepresentedto the government asinputto the by the recipientgovernmentwill be requiredto accelerate
nationalactionplan. progress.
Source:Wolfensohn1999;WorldBank1999d,
1999u.

Box 11.2
The new poverty reduction strategy initiative

Thepovertyreductionstrategyinitiativeof theWorldBankandIn- years,all embeddedin a long-termframeworkfor povertyre-

ternationalMonetaryFundseeksto link externalsupportto do- duction.While the actualform of the strategywill be decided
mestically developed,results-based povertystrategies.It is also by the country-there is no single blueprint-most strategies
intendedto improvetheeffectiveness of WorldBankandIMF re- would likelyinclude:
lations(andthoseof otherdonorsaswell)withrecipientcountries. * Long-termgoalsfor key povertyreductiontargets,andthe
As importantasthe recipientcountrystrategyis theprocesslead- macroeconomic, structural,andinstitutionalframeworkfor
ingup to it. A broad,participatory
dialoguewith representatives achievingthem(see,for example,Uganda's goalsin box1.7).
of civilsocietyandthe privatesectoris expectedto: * Mechanisms for monitoringandevaluatingprogresstoward
* Helpnational authorities
developa betterunderstanding of the the povertyreductiontargets,linkedto publicactions.
obstaclesto povertyreductionandgrowth-and devisegood * A consistent policyandinstitutional
f ramework thatincludes the
indicatorsof progressin povertyreduction. underpinningsfor rapid,sustainedgrowthandpovertyreduc-
* Deepena sharedvisionof desiredpovertyreductiongoals tion (includingmacroeconomic policies,institutionalreforms,
acrosssociety. sector strategies,andassociateddomesticandexternalfi-
* Leadto formulationof prioritiesfor publicactionsto achieve nancing needs).
the desiredpovertyreductionoutcomes. Donorscan helpby providingtechnicalassistancein some
* Encourage the developmentof participatoryprocessesfor areas.Initialexperience inAfricaandLatinAmericaindicates that
settingpovertyreductiongoalsandmonitoringimplementa- countriesarestrongin layingout a povertyprofileandageneral
tion andprogress. povertyreductionstrategybutweakerin preparing quantifiedtar-
Theresultswill beperiodicallyreportedin povertyreduction gets,costingthe strategy,andevaluatingtradeoffsunderlim-
strategypapersexpectedto reflect a broadlyowneddevelop- itedresources. As in otheraspectsof development cooperation,
ment strategy.The strategieswill generallyfocus on three- the countryshoulddetermineits own needfor assistance-to
year cycles, with annualprogressreports in the intervening maintainownershipof this importantprocess.
Source:
IMFandIDA1999;
WorldBankandIMF2000a.
196 WORLD DEVELOPMENT REPORT 2000/2001

Perhapsmore surprising,donor compliancehas been Figure11.4

weakas well.A recentevaluationfound publicexpendi- Aiddoesnot go onlyto poorcountries
ture reviews to have had little effect on either recipient
country policies or donor
, t lending0 practices."
1 So donors Averageannual netrecipts
of totalof offcialdevelopment
~~~~~~~~~~~~~~~assistance
as a share
are searchingfor newmechanismsforstrengtheningpol- Percent
icy environments that encourage country ownership
rather than undermine it. They have begun, for exam- 35 1997-98
ple, to encouragecountries to participate fullyin the
30
publicexpenditurereviewprocess,and theyareexperl- 1987-88 I
mentingwith new instruments. 25
One new instrument that has receivedmuch atten- 2 l
tion is the sectorwideapproach:the governmentdesigns 20
an overall sector strategy,and donors sign on to fund 15
the sector, not individual projects. This resolvesthe 10
problemof donor coordinationby eliminatingthe need
for it: all activityin the sector is conducted by the re- 5
cipient country, using its own funds in addition to n
thoseof donors.This instrument respondsto a broader Least Other Middle- High-income
developed low-income income countries
policy environmentwhile alsoensuringownership.Al- countries countries countries
though the approach is too new to have a track record, Note:Leastdeveloped
countries
arethe48countriesso defined
by
some early experiences are promising (box 11.3). the UnitedNationsonthe basisof certainthresholdsfor income,
Some proponentshave suggested applying the prin- economicdiversification,
and socialdevelopment.
p ~~~~~~~~~~~~~Source:
OECD,DAC2000.
ciples of the sectorwide approach to all development
cooperation (box 1 1.4). Others consider project-based
lending to be desirable and consistent with the new dle-income countries (most donors target aid to the
thinking on development cooperation for poverty re- poorer countries), that still means that global aid is not
duction. Project support can be effective for results- heavily targeted to areas where the incidence of poverty
based sector development-if it falls within a sector is greatest. Aid, and especiallynonconcessional develop-
framework that systematicallylinks investments and pol- ment flows,still has a role in reducing poverty in middle-
icy and institutional development to poverty outcomes income countries, when the policy environment is sound
(and to intermediate indicators for tracking and in- and the resources are well targeted.
terpreting progress). The choice of instrument will de- In addition to targeting poverty, donors should allo-
pend on the policy and institutional conditions of cate aid on the basis of the policy environment. Aid has
particular countries (or sectors within countries) and been shown to be effective in promoting growth and
the preferences of individual donors. But a premium poverty reduction in poor countries with sound eco-
should be placed on putting the country in charge and nomic policiesand sound institutions-ineffective where
ensuring that the mechanisms of aid delivery do nor these are lacking.4 ' Aid driven by political and strategic
compromise its ownership. interests rather than by the recipient country's develop-
Selectivity. For aid to be most effective at reducing ment policy environment is largelywasted from a poverty
poverty, it must be well targeted. If all aid money were reduction perspective. Severalinstruments have been de-
allocated on the basis of high poverty rates and reasen- veloped to assessthe policyand institutional environment
ably effectivepolicies and institutions, a recent study es- in recipient countries, generallycoveringmacroeconomic
timates, even today's small aid flows could lift 19 million management, structural policies, policies for social in-
people out of poverty each year-almost twice the esti- clusion (poverty,gender), and public sector management
mated 10 million now being helped.40 (box 1 1.5).
Currently, about a third of aid goes to middle-income Factoring in the level of poverty and the quality of
countries, whose average GNP per capita is roughly six policiesshould make aid much more efficient in reducing
times that of low-income countries (figure 11.4). While poverty, and there is evidence that donors began to do
only a few major donors target more of their aid to mid- this in the 1990s.42 In replenishing IDA in 1998, for
 REFORMING DEVELOPMENT COOPERATION TO ATTACK POVERTY 19 7

Box 11.3
Sectorwide development cooperation

To addressproblemsof ownership,donorcoordination, andfun- Theapproach ensuresfultownership by the countryandelim-

gibility,donorsareexperimenting with poolingtheirresources to inatesproblems of donorcoordination. Withthecountry havingmore
supportsectorwidestrategiesdesigned andimplemented by the ownership andcontroloverwhathappens, theuseof resources can
recipientgovernment. Thecountry,inconsultation withkeystake- bemuchmoreefficient.Butit alsomeansgreatchangesindonor-
holders, designsasectorstrategyandabudget framework extending recipientrelationsandperhapsgreaterdifficultiesin implementa-
severalyearsforward,anddonorsput their moneyintothe cen- tion. Severalsectorwideprograms havestumbledbecauseof the
tral expenditurepoolfor the sector.The approachencourages recipientcountry'sinadequate institutional
capacity.Lackof con-
countryownershipof sectorstrategiesandprograms. It alsolinks sistencywiththe macroeconomic programhasbeenanother prob-
sectorexpenditure withtheoverallmacroeconomic framework. And lem.Anddonorsoftenhavetoo manyrequirements andthustoo
it ensurescoordination of donorandrecipientactivities. muchof aproblem(ortoo littleinterest)in harmonizing them(Har-
Somebenefitsof asectorwide program areevidentintheZam- roldandassociates 1995).Furthermore, thesearrangements greatly
bianhealthsector.In 1994the government presenteditsnational diminishdonorcontrolandmonitoring of exactly
howmoneyisspent.
healthpolicyandstrategyto donorsand-to ensureequitable dis- Thechangesrequiredimplythat gainingsupportfor the ap-
tributionof services
andcoherent implementation of thestrategy- proach will bedifficult.Therecipient
government hasto beverycon-
askedthemnot to fund specificprovinces or projectsbut to fund fident,because strictadherence to a sectorwide approachmeans
the Ministryof Healthcentrally.Hesitantat first,donorsbeganto donorsthatdonot participate in commonimplementation arrange-
comply.An independent evaluationin 1997foundthat "health mentsarenotallowedto actinthe sector(thatis,theydo nothave
workersare bettermotivated;clinicsare functioning;fundsare theirownprojects). Theresultmaybelessdonorfundingforasec-
flowingto thedistricts;
somemodicum of decentralizationis inplace; tor.Governments mighttherefore optfor lessstrictsectorwidepro-
land)animportantpartof the privatesectorhasbecomeformally grams,choosinginsteadto allowdonorsto implement projectsas
involved." longastheyfit intotheoverallsectorstrategy.

Box 11.4
The common pool for development cooperation

Seeingthe potentialof the sectorwideapproach, somepropose andthe confidence to followthroughevenif donorsdo not sup-
extendingthe ideato the countrylevel(Kanbur,Sandier, andMor- port it. In addition,donorsmightresistcommonpoolsat the na-
rison1999).Donorswouldcedecompletecontrolto the recipient tionallevelbecausetheywouldlikelymeana reductionin donor
countrygovernment-advancing their own perspectiveon de- staff,sincedonoragencies wouldnolongerbedeveloping andmon-
velopmentstrategythroughdialoguewith the countryandwith itoringprojectsor negotiating andmonitoringconditions.
oneanotherratherthan throughspecificprogramsandprojects. However,like the sectorwideapproach,the commonpool
Ratherthanfundtheirownprojects, donorswouldgivecentralbud- approach wouldensurefull ownershipby the countryandelimi-
get supportto countrieswithgooddevelopment strategies(and natedonorcoordination problems. It wouldalsopreservetwo im-
the capacityto implementthem). portantbenefitsof thecurrentdevelopment cooperationapproach:
A countrywouldfirst developitsown strategy,programs, and * Theknowledgetransferred in donor-implemented projects,an
projectsin consultation
with its peopleandwithdonors.It would important sideeffectof aid.A roadbuilding project,forexample,
then presentits plansto donors,whichwould put unrestricted mighttransferknowledgeof engineering or evenprojectac-
financinginto acommonpoolof development assistance, to be counting to localworkers. Thistransfer wouldnotbelostinacom-
usedalongwiththe government's own resources to financethe mon poolarrangement. Recipient countriescouldstill ensure
development strategy.Earmarking woulddisappear. Donormon- knowledge transferthroughtheirchoiceof companies andthe
itoringandcontrolof specificprojectsandprogramswouldnot termsof contracts.
be permitted.Andnoconditionswouldbe placedon donoraid. * Thesupportthatconditionality givesto reformfactionsin gov-
How muchdonorsgivewoulddependon theirassessments ernments. Supportfor reformelements ina countryis perhaps
of the country'spolicyenvironment,includinghow the country the onlyeffectivepartof the presentsystemof conditionality.
cameto agreementonthe strategyandhow capableit is of im- Donor-imposed conditionscanstrengthenthe positionof re-
plementingthe strategyandmonitoringprogress. In thiswaythe formersin national debatesorserveas a "self-imposed" con-
commonpoolapproach wouldbea morerigorous formof condi- strainton government officials.Theapproach to conditionality
tionality,becausedonorswouldneedto evaluatethe overallpol- in a commonpoolarrangement wouldbe far different,but it
icy environment,direction,andcapacityof countries.These wouldnotsacrficethisbenefit. Donorscouldstrengthen thehand
assessments wouldbemadeknownto the countryandto other of reformers by publicizingthe criteriausedto assesscountry
donorsduringthedialogueleadingupto the financingdecision, strategies andadjusting thevolumeof theirassistance. Thiswould
Thisapproach wouldentailmanyof the samechallenges fac- formthe basisfor amoreopenandhonestrelationship between
ing the sectorwideapproach,includingthe needfor recipient donorsandrecipients andpreservethebenefitsof thecurrent
countriesto haveboththe capacityto implementtheirstrategy conditionality whileeliminating its problems.
198 WORLD DEVELOPMENT REPORT 2000/2001
Box 11.5
Assessing country policies and institutions
TheWorldBankhas designeda measureof policyand institutional
soundness-the CountryPolicyand Institutional Assessment,which
givesequalweightto 20 componentsthat haveevolvedas the mea-
sure hasbeen refined.Eachcomponentis ratedbycountryspecialists
on a scale of 1-6 usingstandard criteria.Althoughcare is taken to
ensure that the ratings are comparablewithinand between re-
gions,the scores includean irreducibleelement of judgment.But
when the measure has been includedin regression analyses of
growthalongwithother commonlyused policyvariables,it has had
statisticalsignificance,while other policymeasures have not. It
thus appears to be a good summaryindicatorof the overallpolicy
environmentfor economicdevelopment.The 20 components:
Economicmanagement
Managementof inflationand the current account
Fiscalpolicy
Managementof externaldebt
Managementand sustainabilityof the development program
Structuralpolicies
Tradepolicyand foreignexchangeregime
Financialstabilityand depth
Bankingsector efficiencyand resource mobilization
Source:ColtlerandDollar2000;WorldBank1999h.
example, donors called for allocating funds on the basis
of each country's policy performance.4 3
How selectivityis applied will likely evolve as the in-
ternational community continues to learn about the en-
44
vironments in which aid is most effective. Some analysts
stress that the level of poverty in a country is more im-
portant to aid effectivenessthan the policy environment,
though both are crucial.45 Others show that external
shocks-such as declining terms of trade, volatility in
export prices, and even climate change-can impede
countries' efforts in growth and poverty reduction (chap-
ter 9).46 It has been argued that aid can make a largerdif-
ference in these countries (and thereforebe more effective)
than in countries not experiencing shocks.47 Refining the
criteriafor selectivityshould continue.Adheringto the basic
principle that aid should go where it is most effectivein
reducingpovertywill be key if the internationalcommunity
is to achieve the international development goals.
Implementation di]ficulties
and practical steps
These three components-ownership and partnership, aid
deliverymechanismsthat are lessintrusive,and selectivitv-
Competitiveenvironmentfor the privatesector
Factorand product markets
Policiesand institutionsfor environmentalsustainability
Policiesfor social inclusion and equity
Equalityof economicopportunity
Equityof publicresource use
Buildingof human resources
Safetynets
Povertymonitoringand analysis
Publicsector management and institutions
Propertyrightsand rule-basedgovernance
Qualityof budgetaryand financialmanagement
Efficiencyof revenue mobilization
Efficiencyof publicexpenditures
Transparency,accountability,and corruptionin publicservices
Developinga consistent basis for ratingeconomicand struc-
tural policies has been relativelystraightforward,but doing so
for social inclusionand public sector management has proved
more challenging.Work to refine the indicators and reference
points continues.
provide the framework for substantially improved inter-
national development cooperation. But progress toward
that vision will not be easy.Each component of improved
development cooperation brings great challenges in
implementation.
For example, while almost everyone agrees that part-
nership is a good idea, there is no consensus on how to
implement it.48Some analysts note that ownership is rel-
ative and that reaching consensus on strategies is essen-
tially a political process, involving the same power
relations that exclude poor people from discussions or
discriminate against them (as seen in chapter 6).9 Oth-
ers voice doubts that donors will really come to terms
with the implications of ownership and partnership for
their actions: that donors should interfere less in recip-
ient country policymaking.50 Many donor practices-
such as maintaining control overresourcemonitoring and
tying aid to specificprocurement requirements-run con-
trary to the idea of partnership.5i The recipient coun-
try's capacity to design and implement development
strategies and its ability (and willingness) to hold broad
consultations with all elements of society also pose sig-
nificant challenges.
 REFORMING DEVELOPMENT
The combinationof greaterselectivityand a broader,
lessintrusiveapproachto delivering development assistance
presentsits ownchallenges.Determininghowmuchsup-
port to giveto a sectoror nationalbudgetis difficult-and
likelyto provecontentious. Somecountryexpenditures may
notsmfio"torrcmovement
not seemtofinta"es"poe
bt
donors will have to evaluatethe poverty reduction impact
of the overallprogram, not the individual expenditures.
A more fundamental problem ariseswhen a country
does not have an overall policy environment worth sup-
porting, so that aid is largely ineffective. How should
donors proceed?
Most important, they must understand that policies are
driven primarily by the domestic political economy-and
that donorsare simply not very effectivein influencing
them.52 But donors can have some influence by tailoring their
involvementto a countrys commitment to reform.Until a
country commits seriouslyto reform, the best that donors
can do is to providetechnicalassistanceand policydia-
logue,without largebudget or balance of paymentssupport
(box 11.6). If donors pour large amounts of aid into poor
policyenvironments,they arelikelyto sustainpoor policies
longer. When the country finally commits to reform, evi-
dence shows that finance should be increased as policies im-
prove.53
In addition to this more nuanced approach to influ-
encing policy reform, donors can address the challenges
enc,ngpir r d ohelp
of the new development cooperation framework by tak-
ing several other steps:
* Molvethedonor-recipientdialogueto the countryand turn
itsleadershipoveraswelLDonor-recipient consultations-
consultativegroups or roundtables-have traditionally
taken place in donor countries, chaired by the World
Bank, the United Nations DevelopmentProgramme,or
another donor institution. Meetingsare now beginning
to be held in recipientcountriesand chaired by their gov-
ernments, to fosterownership.
• Continue to experiment with sectorwide approaches.
National capacity-and donor-recipient partner-
ships-can be built up sector by sector. While many
countries will for some time not have the overall tech-
nical capacity, accountability, and transparency to
monitor funds to the satisfaction of donors, these
may be more advanced in some sectors than in oth-
ers. The advanced sectors could be funded through the
sectorwide approach as soon as possible, taking into
account the lessons from experience with this ap-
proach.54 And donors should continue to improve their
COOPERATION TO ATTACK POVERTY I99)
Box 11.6
Howaidcanhelpin countrieswith a weak
policy environment
Whena countryhas poor policiesand no coherent political
tochangethesituation,aidcanhavea limitedbut
t, effective role, as Ghana,Uganda,and Vietnamall illustrate.
In their prereform periods (before 1983 for Ghana, 1986 for
Uganda,and 1991forVietnam),these countriesreceivedvery
littleaid,probablyreflecting
theirgovernments'
politicales-
trangement from the West. But the aidwas instrumentalin
layingthe foundationfor policyreform.
For example, when Ghana was dealing with a macro-
economic crisis in the early 1980s,its well-trainedecono-
mists found the policydialoguewith internationalfinancial
institutionsto be helpfulin workingout plans. A few years
later,when Uganda'sleaderswere tryingto designnew poli-
cies,donorsfinancedhelpfulstudytoursto Ghana.In1991
the United Nations Development Programme and World
Bankorganized
a meetingforVietnamese
leaderswitheco-
nomicministersfromIndonesia,
the Republic
of Korea,and
Malaysia,who laid out some key policiesthat had worked
for them and also some of the detailed issues in stabiliza-
foreigninvestment,andothereco-
tion,tradeliberalization,
nomicpolicies.
In successfulcases politicalleaderslearnfromother
countriesandfrom their own mistakes.Low-keyassistance
can help withthis policylearning,whichgenerallyhas to take
placeat a country's own pace. Evenin countriesthat do not
reformfor a longtime, technicalassistance can laythe foun-
dationforpolicylearning.InKenya,forexample,donorsare
supportingthe Instituteof PolicyAnalysisand Research to
develop localcapacityin research and policyanalysis.
is notgoingto havelargeben-
Thiskindof capacitybuilding
efits as long as vested interests resist serious reform. But
it is an essential foundation ifa politicalmovementfor change
develops.
Source:Devarajan,Dollar,and Holmgren2000; WorldBank1998b.
own practices-for example, by harmonizing proce-
dures and reporting requirements among themselves--
so that they can contribute effectively to these new aid
relationships.
* Strengthenmonitoringand evaluationpractices. Donors'
systems of monitoring and assessing the impacts of
their own projectshave failedto focus on how poor peo-
ple benefit.5 5 But doing this will be even more impor-
tant (and challenging)when looking at a sectorwide or
nationwide program. Donors should encourage local
monitoring by participants, to ensure ownership of the
results. Furthermore, donors tend to be weak in dis-
seminating information and incorporating knowledge
200 WORLD DEVELOPMENT REPORT 2000/200I

56
from evaluations. Feedbackand learning are essential Figure 11.5
to successful aid practices, and donors must ensure NGOs channeledsome $10 billionto
that theyIhtte
happen effectively Asspr
apnercley part offtl,unr
this, donors and
n developing countries
of it from the in 1998, about half
official sector
recipients should continue to strengthen their efforts
against corruption, a major obstacleto economic per- Billionsof U.S.dollars
formance that occasionallyalso affects donor agencies. 6
* End tied aid. In 1998 almost a quarter of officialdevel-
opment assistancewas tied, meaning that the procure- 5
ment contracts were limited to the donor country or a 4
group of countries. Driven by domestic politicalinter-
ests,this practice goesagainstthe very free-marketprin- *
ciples that most donors are trying to encourage in
developingcountriesand resultsin inefficientuse of aid. 2
It has been estimated that tying aid reduces its value by
15-30 percent.57 The practiceshould be ended as quickly 1
as possible,and contracts should go to the best bids.5 8 *
* Make technicalassistance demand driven. Turning more 0
responsibilityover to recipient countries for designing NGOs' own DAC Official
national strategies
development and leading consulta- grants to contributions development
nationaldevelopmentstrategiesand leading consulta- developing to NGOs assistance
tion meetingswillrequire rapid capacity development. countries administered
Recipient countries will also need strong auditing and by NGOs
accounting skills if donors are to relinquish monitor- Note:These data covermoney used for developmentor relief
andcontrol
* ~~~~~~~~~~~purposes,
Theamountof officialdevelopmentassistance
ingand controlof projects. But technicalassistance,the administered by NGOsis a rough estimate only,because many
obvious choice for building capacity, has a spottv donors do not report this information.
Souirce:OECD,DACdata.
record at best, particularly in countries where capac-
iry is already weak. The main reasonis that it has often
not been demand driven-it has often been tied aid gone into funding their evaluation and monitoring ef-
and designed to develop capacity only in donor-sup- forts.6 2 Even with better monitoring, though, NGO
ported activities.59 Instead, technical assistanceshould projects face the same problems of fungibility as donor
be incorporated into a national strategy and expendi- projects, and policy environments strongly influence
ture plan, with the recipient government deciding their effectiveness.Donors and NGOs should continue
what assistance it needs and who should provide it. This to improve their working relationships, sharing best
is likely to require initial support to countries on how practicesfor makingaid more effectivein the long term.
to use the market for technical assistance. * Relievemore debt. Debt relieffor the poorest countries
* Continue to learn about how to work effectivelywith is essentialfor effectiveaid. Heavy debt burdens reduce
NGOs. Relationships between donors and NGOs are incentivesfor policyreform,while debt negotiationsand
complex, with much room for improvement. 6 0 Good the constant circulation of new aid money to service old
data on the extent and effectiveness of donor-NGO debt distract government officials from the needs of their
relationships are scarce, but an estimated $5 billion in citizens. The next section turns to this issue.
aid is now channeled through NGOs, either in sub-
sidiesto their activitiesor in contractsto implement Relieving the debt burden
donor activities(figure11.5). NGOs appearto be an of poor countries
effectivechannel for aid when they are involved early
in projects (at the design phase), when they are cho- The most prominent issuein development cooperation at
sen for their proven capacity and experience, and the end of the 20th century and the beginning of the 21st
when they are treated as partners rather than con- has been debt relieffor the poorestcountries.There has been
tractors.6 1 The long-term impact of NGO projects re- a steady increaseover two decades in the indebtedness of
mains unknown, perhaps because so little money has a group of poor countries now referred to as the heavily
 REFORMING DEVELOPMENT COOPERATION TO ATTACK POVERTY ZOI

Figure11.6 Box11.7
As percapitaincomein the heavilyindebted The Enhanced HeavilyIndebtedPoorCountries
poorcountrieshasgonedown,debthasgone DebtReliefInitiative
up-and viceversa
The HeavilyIndebted PoorCountries (HIPC)Debt Relief Ini-
Medianpercapita Medianpresentvalueof external tiative was announcedin late 1996.Realizingthat the initia-

997 U.S. dolars debtasa shareof GDP tive did not go far enough,leadersof the Groupof Seven(G-7)
350 80 countriesendorsedan EnhancedHIPCInitiativeat a summit
in Cologne,Germany,in July 1999. The enhancedinitiative
Externaldebt was approvedby the full membershipof the World Bankand
InternationalMonetary Fund in September 1999as an inte-
70 gralpart of the new poverty reductionstrategy initiative (see
box 11.2). The EnhancedHIPC Initiative changed the eligi-
Percapita / ~ \ | \ bility requirements for debt relief and the timing of relief.
30030income rN \ 6
6
Eligibility
To be eligible, a country must be very poor, have an unsus-
tainabledebt burden, and pursue good policies.
50 * Pooris definedas both eligiblefor support underthe IMF's
Poverty Reduction and Growth Facility (the reformed
and renamedEnhancedStructuralAdjustment Facility,or
250 ESAF)and eligible only for concessionalfinancingfrom
1979 1985 1990 1995 1997 the World Bank,through the InternationalDevelopment
Association.
Note:Theobservedassociation betweendecliningincomeand * An unsustainabledebt burdenis definedas a stock of debt
risingdebt shouldnot beviewedas implyingthat debt reduction that is morethan 150 percent of exportsin present value
will automaticallyresultin higherincomes.Governmentpoliciesare terms after the full use of traditional debt relief mecha-
the keyto growthandpovertyreduction,and badpoliciescanlead nisms or (for countries with certain structuralcharacter-
to both higherdebt and lowerincomes. istics)a ratioof debt to governmentrevenueof more than
Source:Easterly1999c 250 percent.

I Good policies are interpreted to mean macroeconomic,

structural,and social policies consistentwith poverty re-
duction and sustained growth.
indebted poor countries (figure 11.6). Public attention has These new eligibilitycriteria increasefrom 26to 33 the num-
been drawn to their plight in large part through the tire- ber of countries likely to qualify for relief.
less efforts of NGOs in developed and developing coun-
tries, whose campaign for
for debt
tries, cancellation
cancellation by
whosecampaign
debt by 2000 has
2000hasThe frle HIPCInitiative provides for the possibility of
TmnEnhanced
captured the world's interest. 6 3 At the 1999 annual meet- interim relief for countriesafter they passthe decisionpoint,
ings of the World Bank and IMF, member countries agreed when the World Bank and IMF determine a country's eligi-
on an enhanced plan for debt relief, an acknowledgment bility. A reductionin debt service payments is thereforepos-
of the detrimental effects of debt on country policy envi- sible even before a country reaches the completion point,
when the stock of debt is reduced. Underthe earlier HIPC
ronments and overall expenditure frameworks (box 11.7). agreement the debt stock was reducedonly after comple-
tionof two full ESAFprograms-a minimum of sixyears.Now
The effects of heavy debt burdens the completionpointcan bemovedupif the country's per-
Many heavily indebted poor countries spent as much as formance is particularlygood. Relief is intended to be front-
loaded as much as possible.
a fifth of their annual budgets on debt service in the 1 990s, Combinedwith traditional debt relief arrangements,the
and some spent much more. 6 4 Because this is often more EnhancedHIPCInitiative is likely to cut by half the net pre-
than the amount spent on social programs, debt servic- sent value of public debt for the 33 countries likely to qual-
ing is viewed by many as a severe impediment to im- ify. As many as 20 countries may reach a decision point on
proving the lives of the world's poor. debt relief by the end of 2000, depending on progress in de-
velopingtheir povertyreductionstrategiesand on how much
It has been argued, however, that debt servicing is financing is availablefrom donors.
not really a problem because heavily indebted poor
countries receive more money from donor countries Source:WorldBanklwww.worldbank.org/hipc).
than they pay back. Actual debt service payments are
202 WORLD DEVELOPMENT REPORT 2000/200I

Figure 11.7
Concessional
transferslargelycompensate
for negativenet transfersof nonconcessional
resources

Net transfersto heavilyindebtedpoorcountriesby creditor,1988-97

Billions of U.S. dollars
25

20

15

10

- ---

-5

-10
International Bilateral Other Other International Private Bilateral International
Bank
Development concessional multilateral multilateral MonetaryFund nonconcessional
for Reconstruction
Association concessionalnonconcessional andDevelopment

Source:Easterly
1999c.

almost alwaysfar less than scheduled payments, because
the countries cannot make the full payments. The
debts are serviced by rescheduling some loans and fi-
nancing the rest through a combination of new loans
and grants.6 5 Overall, while net transfers of noncon-
cessional resourcestend to be negativebecause new non-
concessionalborrowing is stronglydiscouraged, transfers
of concessional resources tend to more than compen-
sate (figure 11.7).
However, heavy debt burdens bring additional prob-
lems that can affect a country's growth performance and
ability to focus government action on social priorities.
Debt service is financed largely by scarce domestic bud-
getary resources and thus competes with domestic re-
current spending, while concessional assistance goes to
new investment projects. This mix can mean resources
for new health centers and roads but not for nurses or
maintenance. In addition, many grants go to donor-
managed activities that are not included in the budget.
These are subject to all the problems of ownership and
donor coordination discussed above and can contribute
to the further institutional weakening of an already weak-
ened, insolvent state.66 And debt negotiations and mon-
itoring take up much of the already stretched time and
capacity of government officials.
These resource inflows can also be unstable, making
it difficult for governments to manage their spending and
maintain sound fiscal policies.67 Furthermore, if resource
flows are positive because countries have to rely on con-
tinuous recheduling and on grants and concessionallend-
ing, their access to private capital flowswill remain very
low. And where debts are not serviced in full, countries'
debt stocks continue to grow, creating a potential disin-
centive to investment, since investors may fear that fu-
ture profitswill be affectedby debt-relatedmacroeconomic
problems or higher taxes to service debt.6 8
Debt is thereforeas much a problem of how grossflows
and debt management affect ownership, policy, and ca-
pacity as it is a problem of net flows. In this, it shares
many of the problems that have diminished the effec-
tivenessof aid. Debt reliefcan play an important role here
by reducing the burden on recurrent budgets and allowing
government officialsto focus on sound spending strate-
gies rather than continual renegotiation of debt. And it
can be particularly crucial for countries emerging from
civil conflict and war.
 REFORMING DEVELOPMENT
There is also some evidencethat high debt service
obligations(includingthose to international financial
institutions) tend to weaken the link between conces-
sional flows and the quality of the policy and institu-
tional framework-and so the effectiveness in reducing
poverty. This could be because donors try to avoid de-
faults on loans, and as countries become more in-
debted, donors give new loans to cover the old ones.
(Between 1989 and 1997 debt relief for the 41 heav-
ily indebted poor countries totaled $33 billion and
new borrowing $41 billion.)6 9 Not only does this com-
promise the ability of donors to target aid to where it
will be most effective, but it may also deter reform in
countries with poor policies, because they have less
incentive to reform if they can expect relief and resources
anyway.70
Debt relief can ease all these problems by reducing
the gross flows and, if structured correctly, encouraging
a structure of new inflows that is more effective for
poverty reduction.
An improved initiative for debt reliejf
To be effective, debt relief needs to be delivered in ways
that encourage country ownership, using instruments
that provide incentives to use the resources for poverty
reduction. This is the same issue as for traditional aid flows,
but in the context of a one-time decision to reduce debt.
How much impact debt relief has on net transfers to a
country depends, of course, on what happens to gross aid
flows-on whether the resources for debt relief are ad-
ditional or not. But even if the resourcesare not entirely
additional, debt relief can ease policy and budgetary con-
straints for the recipientcountry, since it freesup resources
from the recurrent budget. What will guarantee that
these resourcesare used for poverty reduction? There are
two related challenges:
* Linking resources from debt relief to results in
poverty reduction.
* Strengthening accountability in the use of public re-
sources,to minimize diversion to other uses (especially
through corruption).
The lessons from the past-including those from the
experience with aid outlined above-indicate that both
are best tackled through their links to the overall policy
and institutional environment, especially for public re-
source use. Experience also shows that debt relief alone
will not improve policies. Twenty years of gradually in-
creasing debt relief have not improved policies in heav-
COOPERATION TO ATTACK POVERTY 2C3
ily indebtedpoor countries. 71 That is why the principile
is to grant debt reliefon the basisof reputation-an es-
tablished track record in using resources effectivelyfor
poverty reduction.
The design of the Enhanced Heavily Indebted Poor
Countries (HIPC) Debt Relief Initiative incorporates
these lessons. Debt relief will be granted to eligiblecoun-
tries with a viable and comprehensive poverty reduction
strategy and a framework for linking public actions to
monitorable results in poverty reduction. The strategy is
to be defined through a participatory process involving
government, the private sector, and civil society. The
participatory process is important for the design of the
strategy-and to help ensure good use of external (and
internal) resources.Debt reliefwill be integratedwith other
sources of external finance in the country's overall bucd-
getary framework for poverty reduction, rather than
being earmarked for certain expenditures. The goal of the
Enhanced HIPC Initiative is to contribute directly to
poverty reduction and to ensure that countries that re-
ceive debt relief do not have policies that will lead themn
deeplyinto debt again.
In May 2000 Uganda became the first country to re-
ceive debt relief under the Enhanced HIPC Initiative
(box 11.8). The relief was based on several years of
progress in the participatory formulation of its poverty
reduction strategy, results in key areas (getting children
into school, reducing income poverty through agricul-
tural and aggregate growth), and mechanisms to help
increase accountability for public funds and reduce
leakages.
The cost of the Enhanced HIPC Initiative is esti-
mated at $28 billion. If the debt relief is to be add.i-
tional, financing must come from outside the normal
aid and concessional lending budgets of donor insti-
tutions. Under current plans the cost will be financed
roughly equally by bilateral and multilateral credi-
tors. Although many donors have endorsed the En-
hanced HIPC Initiative and made political
commitments for funding, the mobilization of re-
sources has been slow, and some donors have not yet
committed to the initiative. Because a key principle un-
derlying the initiative is that debt relief should be co-
ordinated among all creditors, with broad and equitable
participation, this lagging of resources and commit-
ments seriously endangers the initiative. Donors need
to give high priority to securing sufficient funding
for the Enhanced HIPC Initiative.
204 WORLD DEVELOPMENT REPORT 2000/200I
Box 11.8
How debt relief fits into a poverty reduction strategy: Uganda's Poverty Action Fund
Fundamentalin the fightagainst povertyis improvingthe overall
allocationof resources, includingthose from debt relief,through
more poverty-orientedand transparent budgets. There are many
ways of achievingthis end, and in Ugandaa specialfund to use
the savingsfrom debt reliefis provinguseful.
The governmentchose to create the PovertyActionFund as
a conduitfor the savings from debt relief under the HIPCInitia-
tive (about$37 milliona year; the EnhancedHIPCInitiativeis ex-
pected to doublethis amount).Thefund has been earmarkedfor
prioritiesof the povertyeradicationaction plan adopted in 1997
to addresspovertyandsocialconditions. The planemphasizesmain-
tainingmacroeconomicstabilitywhileincreasingthe incomesand
the qualityof lifeof poor peopleby developingruralinfrastructure,
promotingsmallbusinesses and microenterprises,creatingjobs,
and improvinghealthservicesand education.The PovertyAction
Fundfocuseson schools, ruralfeeder roads,agriculturalextension,
Source:UNICEF
andOxfamInternational
1999.
Many questions remain about the implementation
of debt relief and of the new development cooperation
framework advanced in this chapter. Despite the fi-
nancing difficultiesof the Enhanced HIPC Initiative,some
observers call for even deeper and faster debt relief, ar-
guing that the debt deemed "sustainable" under the En-
hanced HIPC Initiative is still too burdensome.7 2 How
to move quicklyto relievedebt while still allowingenough
time to build country ownership of the poverty reduc-
tion strategy is another concern. Some countries wonder
about their capacity to prepare their own poverty assess-
ments and poverty reduction strategies. Others question
whether donors can support the formulation and im-
plementation of poverty reduction strategies without
undermining country ownership. Questions also remain
about the participatory process-how best to consult
with poor people, how to fit consultative processes into
the context of national political processes,and how to de-
velop effective feedback and monitoring systems. And
countries wonder how well donors will be able to realign
and district-levelwater and sanitation.Specificoutcome targets
have been identified,such as the constructionof 1,000additional
classroomsto support the primaryeducationprogram.
Twocrucialfeaturesofthe PovertyActionFundare its integration
intothe overallbudget andthe Ugandangovernment'seffortto cre-
ate a transparentand accountablestructure of management. Re-
ports on financialallocationsare released at quarterly meetings
attended by donors and NGOs.The Inspector General's office
monitorsthe use of funds at the districtand nationallevels.This
self-imposedconditionalityreflectsthe government'sstrong com-
mitmentto tacklingcorruption.Butit is also an attempt to address
creditorconcernsaboutthe capacityof a debtorcountryto linkdebt
reliefto povertyreduction.Severalmeasures havebeen proposed
for improvingmonitoring,rangingfrom includingdistrict-leveloffi-
cialsinthe quarterlymeetingsto havinglocalNGOsdo community-
based monitoringof the povertyfund's spending.
their procedures and interventions along the lineslaid out
in their poverty reduction strategies.7 3 All these issuesre-
flect the state of international development cooperation
at the turn of the 21st century. There is profound, on-
going change in the way developing and developedcoun-
tries work together to fight poverty.
While many issuesremain, the right direction for the
international community is clear. Country-focused as-
sistance should incorporate a greater emphasis on part-
nership between donors and developing countries. It
should apply lessintrusive mechanismsof aid deliverythat
focus on the overall policy and expenditure environ-
ment. And it should exercisegreaterselectivityin allocating
aid where it will be most effective.More aid and debt re-
lief need to be availableto countries with effectivepoverty
reduction programs. Donor evaluations of these pro-
grams must be informed by an awareness of the condi-
tions each country faces and by the new approach to
poverty reduction presented in this report. And to relieve
the burden of the heavilyindebted poor countries, donor
countries should finance the Enhanced HIPC Initiative
with money additional to their aid budgets.
 Bibliographic Note
T hisreportdrawson awide rangeof WorldBank
documentsand on numerous outsidesources.Back-
ground papers and notes were prepared by Daron
Acemoglu,MichelleAdato,MaryB.Anderson,Michael
R.Anderson,SimonAppleton,GarethAustin,Michael
Banton,PranabBardhan,PaoloBelli,TimothyBesley,
PilwhaChang, MoniqueCohen, MichelleConnolly,
RichardC. Crook, RobertA. Dahl, ParthaDasgupta,
SheltonDavis, Alainde janvry,Stefan Dercon,Ann
Elwan, Gary S. Fields,Gary Gereffi,GregoryGraff,
GeorgeGray-Molina,LawrenceHaddad, John Har-
riss,RonaldJ. Herring,John Hoddinott,NaomiHos-
sain,Peter P.Houtzager,RajshriJayaraman,Noushin
Kalati, Marcus Kurtz, Edward E. Leamer,Jennifer
Leavy,DavidLindauer,MichaelLipton,FrancesLund,
Daniel S. Manning, James Manor, Martha Argelia
Martinez,Jacob Meerman,Mick Moore, SamuelA.
Morley, KimberlyJ. Niles, Anthony Oliver-Smith,
Jonathan Pattenden, Anan Pawasuthipaisit,Louis
Pouliquen,KameshwarPrasad,JamesPutzel,Danny
Quah, ElisaReis,JamesA. Robinson,FranciscoRo-
driguez,ElisabethSadoulet,SombatSakuntasathien,
PeterK. Schott,JenneferSebstad,SaurabhSinha,Lina
Song, SmitaSrinivas,Alan Sturla Sverrisson,Robert
M. Townsend,BenTurok,AshutoshVarshney, Howard
White, LaurenceWhitehead,L. AlanWinters,Quentin
Wodon, ShahinYaqub,and DavidZilberman.
Backgroundpapersfor the report are availableei-
ther on the World Wide Web (wwwworldbank.org/
poverty/wdrpoverty) or through the World Develop-
ment Report office.The viewsexpressedin these pa-
persare not necessarilythoseof the WorldBankor of
this report.
Many people,both insideand outside the World
Bank,gaveadviceand guidanceto the team.Valuable
commentsand contributionswereprovidedbyTaoufik
BenAbdallah,RichardAdams,Nisha Agrawal,Sadiq
Ahmed,Martha Ainsworth,GeorgeAkerlof,Harold
Alderman,Titus Alexander,Jock Anderson, Hutton
Archer,Anthony Atkinson, Gareth Austin, Robert
Ayres,MalcolmBale, Namrata Bali, Andrew Balls,
AbhijitBanerjee,PranabBardhan,ChristopherBarham,
DouglasBarnes,TamsynBarton,AnanyaBasu,Kaushik
Basu,AmieBatson,AnthonyBebbington,Alan Berg,
TimothyBesley,GordonBetcherman,AndreBeteille,
Surjit Bhalla,Vinay Bhargava,Ela Bhatt, Mihir R.
Bhatt,Hans Binswanger, NancyBirdsall,YonasBiryu,
Mark Blackden,RebeccaBlank,David Bloom,Zelco
Bogeic:,Jan Bojo,ReneBonnel,Ed Bos,CesarBouil-
l6n, FrancoisBourguignon,SamuelBowles,CarlosA.
205
206 WORLD DEVELOPMENT REPORT 2000/2001
Primo Braga, John Briscoe, Penelope Brooks, Stephen
Brushett, Robin Burgess,Sara Calvo, Sarah Cambridge,
Roy Canagarajah, Gerard Caprio, Teresa Carbo, Guy
Carrin, Soniya Carvalho, Robert Chambers, Jacques
Charmes, Celine Charveriat, Mirai Chatterjee, Mrinal
Datta Chaudhuri, Rodrigo Chaves, Sandeep Chawla,
Shaohua Chen, Susan Chen, Kenneth Chomitz, Alberto
Chong, Ralph Christy, Mariam Claeson, John Clark,
Monique Cohen, PaulCollier,Tim Conway,GiovanniAn-
drea Cornia, Uri Dadush, Dana Dalrymple, Amit Dar,
Koen M. Davidse, Adrian Davis, Gloria Davis, Alain de
Janvry,Samantha De Silva,Naa dei Nikoi, Angus Deaton,
Klaus Deininger, Lionel Demery, Stephen Denning, Ste-
fan Dercon, Mahendra Dev, Shantayanan Devarajan,
Ishac Diwan, David Dollar, Philippe Dongier, Donna
Dowsett-Coirolo, Jean Dreze, Jean-Luc Dubois, Steven
Durlauf, Chris Dye, Tim Dyson, William Easterly,Judith
Edstrom, Dag Ehrenpreis, Lars Ekengren, Ibrahim El-
badawi, David P. Ellerman, Diane Elson, Gunnar Eske-
land, WolfgangFengler,Marco Ferroni, Deon Filmer,Ben
Fine, Ariel Fiszbein, Ann Florini, Emmanuel Forestier,
Justin Forsyth, Paul Freeman, Jose Furtado, Andreas
Galanakis, Emanuela Galasso, Joaquin Garcia, Michel
Garenne, Roberta Gatti, Guido Geissler,Alan Gelb, Paul
J. Gertler, Coralie Gevers, Ashraf Ghani, Maitreesh
Ghatak, Alan Gilbert, Michael Goldberg, Jeff Goldstein,
Fr. Xabier Gorostiaga (and his colleaguesfrom Asociaci6n
de Universidades Confiadas a la Compaifa de Jesus en
America Latina), Vincent Gouarne, Heather Grady, Peter
Grant, StefanieGrant, Cheryl Gray, Duncan Green, Mar-
garet Grosh, Sumit Guha, Patrick Guillaumont, Sanjeew
Gupta, Davidson R. Gwatkin, Lawrence Haddad, Peter
Hakim, Gillette Hall, Kristin Hallberg, Jeffrey Hammer,
Lucia Hanmer, Nancy Happe, Caroline Harper, Ricardo
Hausmann, YujiroHayami, John Healey,Gerry Helleiner,
Jesko Hentschel, Alicia Herbert, Norman L. Hicks, John
Hoddinott, Robert Holzmann, Peter P. Houtzager,Albert
D. Howlett, Chia-Hsin Hu, Gregory Ingram, Keiko Itoh,
VijayJagannathan,SelimJahan, K. Jankovsky,Mahieujar-
ret, Renana Jhabvala, Emmanuel Jimenez, Ian Johnson,
Gerd Johnsson, Ben Jones, Christine Jones, Steen Jor-
gensen, Sonia Kapoor, Dani Kaufmann, Masahiro Kawai,
Allen Kelley,Charles Kenny, Michel Kerf, Christine Kes-
sides, Roger V. Key, Anupam Khanna, Stuti Khemani,
Tony Killick, Ronald Kim, Elizabeth King, Stephan
Klasen,Jeni G. Klugman, SteveKnack, GrzegorzKolodko,
ValerieKozel,Annette Krauss,Alcira Kreimer,Jean-Louis
Lamboray,Jack Langenbrunner, Patricia Laverley,Richard
Leete, Arianna Legovini, Danny Leipziger, Brian Levy,
Maureen Lewis, Michael Lipton, Jennie Litvack, Laszlo
Lovei, James Christopher Lovelace, Landis Mackellar,
Fran,ois RegisMahieu, Nick Manning, Tamar Manuelyan
Atinc, Timothy Marchant, Rachel Marcus, Tiffany Mar-
lowe, Ricardo Martin, Will Martin, Antonio Martin del
Campo, Keith Maskus, Andrew Mason, Simon Maxwell,
Bill Mayville, Elizabeth McAllister, Milla McLachlan,
John Mellor, Jean-Roger Mercier, Tom Merrick, Rick
Messick, Dilip Mookherjee, William Moomaw, Michael
Moore, Mick Moore, Jonathan Morduch, Daniel Mor-
row, Robert Moulie, Peter Mousley, Ranjana Mukherjee,
Joseph Mullen, Rinku Murgai, Edmundo Murrugara,
Philip Musgrove,David Nabarro, Mustapha Nabli, Reena
Nanavaty, Deepa Narayan, Richard Newfarmer, Juan
Pablo Nicolini, Michel Noel, Barbara Nunberg, Veronic
Nyhan, Abena D. Oduro, Marcelo Olarreaga, Jonathan
Olsson, Azedine Ouerghi, Mead Over, Margaret Owen,
Howard Pack,Truman Packard,SheilaPage, Robert Pala-
cios, Ok Pannenborg, Sulekha Patel, Harry Anthony Pa-
trinos, Guillermo Perry,Jean Pesme, Patti Petesch, Guy
Pfeffermann, Claire Pierangelo, Jean-Philippe Platteau,
Boris Pleskovic,Louis Pouliquen, Alexander Preker, Gio-
vanna Prennushi, William C. Prince, Lant Pritchett, Fe-
licity Proctor, James Putzel, Dagmar Raczynski, Atiqur
Rahman, Mamphela Ramphele, James Rauch, Martin
Ravallion,SusanRazzaz,Thomas Reardon, Ritva Reinikka,
Ana L. Revenga,Carolyn Reynolds,Helena Ribe, Michelle
Riboud, Peter Roberts, Richard D. Robinson,Alberto Ro-
driguez,John Roemer, Halsey Rogers,Andrew Rogerson,
Jaime Ros, Jaime Saavedra,Elisabeth Sadoulet, David E.
Sahn, Joanne Salop, Susana Sanchez,Todd Sandler, Sven
Sandstrom, Filomeno Santa Ana, Justine Sass,David Sat-
terthwaite, Dieter Schelling,Anita Schwarz, Christopher
Scott, Jennefer Sebstad, Marcelo Selowsky,Amartya Sen,
Elena Serrano, Nemat Shafik, Shekhar Shah, Jim Shea,
Geoffrey Shepherd, Lynne D. Sherburne-Benz, John D.
Shilling, Paul Bennett Siegel, Hilary Silver,William Sil-
verman, Marcia Simoes, John Sinclair, Saurabh Sinha,
Richard Skolnick, Tova Maria Solo, Paul Spray, Lyn
Squire, T. N. Srinivasan, Nicholas Stern, David Stiedl,
David Stifel, Joseph E. Stiglitz, Kalanidhi Subbarao,
Parita Videt Suebsaeng,Eric Valdeman Swanson, Vinava
Swaroop,Simon Szreter,CeciliaTacoli,Kazuo Takahashi,
Vito Tanzi (and the team at the FiscalAffairs Department
of the International Monetary Fund), David Tarr, Ju-
dith Tendler, Sumeet Thakur, Duncan Thomas, Kirsten
Thompson, Robert Thompson, Erik Thorbecke, Mari-
 BIBLIOGRAPHIC NOTE 207

anoTommasi,LeeTravers,KerstinTrone,CarrieTudor, Chapter 1
WendyTyndale,ZafirisTzannatos,ChristopherUdry,Al- Unlessotherwisenoted,allquotations
inthechapter
aredrawnfrom
bertoValdes,Dominiquevan de Walle,JulievanDome- VoicesofthePoor(Narayan, Chambers,Shah,andPetesch2000;
len, M. Willem van Eeghien,Wouter van Ginneken, Narayan, Patel,Schaft,Rademacher, andKoch-Schulte2000).
1.Sen1999,p. 87.
WarrenVanWicklin,Jan Vandemoortele, KrishnaVatsa, 2. Szekely
andothersforthcoming.
Anthony Venables,MathewA. Verghis,Louis-Charles 3. HaddadandKanbur1990.
Viossat,Tara Vishwanath,Milan Vodopivec,Joachim 4. RavallionandvandeWalle1991.
von Amsberg, Jayshree Vyas, Robert Wade, Mike 5. Thepovertygapindexisthe sumoftheincomeshortfalls
Waghorne, Adam Wagstaff, Michael Walton, Kevin ofallpoorpeople-theamount bywhichtheirincomes fallshort
Watkins (and an Oxfam team), Catherine Wart, Richard of the poverty line-divided by the total population.The squared
poverty
gapindexisthesumofthesquared
shortfalls.
Thepoverty
Webb,L.AlanWinters,Quentin Wodon,AdrianWood, gapindexis
John Worley,Gustavo Yamada,Jacob Yaron, Shahid Q
Yusuf,Roberto Zagha,and ElaineZuckerman. N (y Y ),
The team wasassistedby studentsfrom-n rhe In- whereN = the total population,y7 = the povertyline,
ternship Programof the WashingtonCenter: Anju Ag- y, =incomeofindividual i, Q=the totalpopulationbelowthe
garwal,Waldo Aleriano,Juan CarlosArandia, Hector povertyline,and a = 1. For the squaredpovertygapindex,
Cabrera, Mario de la Cruz, Celeste de la Huerta, x=2.Whena =0, themeasure isthewell-known
headcount ratio.
Joaquin de la Torre, Alison Drury, Nilima Gulrajani, SeeFoster,Greer,andThorbecke (1984)andFosterandShorrocks
Tomoko Hagimoto, Daniel Hernandez Ruiz, Vir- (1988).
ginia Iglesias, Mika Iwasaki, Alejandra Lua, Felix 6. Rowntree 1901.
Marklein, Nadia Montiel, Mark SchlueterMaklnNdi
and Neil 7. WorldBank 1999j.
a8. onearShuete, Formoredetails seetheinformationonpovertyreduction
Thompson. strategypapersat www.worldbank.org/poverty/strategies/
Despite effortsto be as comprehensiveas possiblein index.htm.
compiling the list above,some who contributed may 9. Mecoviis Programa paraelMejoramientode lasEncucs-
havebeeninadvertentlyomitted.The teamapologizesto tasy laMedici6n delasCondicionesdeVidaenAmerica Latina
aywweenreeas ra y elCaribe(Program fortheImprovement ofSurveysandtheMea-
anywho wereand reiteratesits gratefulnessto all who surementof LivingConditionsin LatinAmericaand the
contributed. Caribbean). Seewww.iadb.org/sds/pov.
10.Townsend1985.
Endnotes 11.Atkinson andBourguignon forthcoming;
ChenandRaval-
lion 2000.
Overview 12.World Bank1999t.
Unlessotherwise noted,allquotations
in theoverviewaredrawn 13.WorldBank20001.
fromVoices ofthePoor(Narayan, Chambers, Shah,andPetesch 14.Demery1999.
2000;Narayan,Patel,Schafft,Rademacher, andKoch-Schulte 15.Wodon,Ayres,Barenstein, Lee,Peeters,Siaens,arLd
2000). Yitzhaki2000.
1.Sen1999. 16.Theincidence ofurbanpovertyin Colombia hasincreased
2. Thisaccountis of a visitbyRaviKanbur,directorof the steadily
since1996,however.
reportuntilMay2000. 17.WorldBank2000e.
3. Notethat 1998is themostrecentyearforwhichdataare 18.Suryahadi andothers1999.Povertyappears to havefallen
available.
However, figuresfor1998arepreliminary, fromFebruary 1996to aroundthethirdquarterof 1997.There
4. Whilethesenumbersprovide asenseofbroadtrends,they areseveral alternative
estimates
ofIndonesian poverty,discussed
shouldbetreatedwithcautionin lightof thedatashortcomings in detailin Suryahadi andothers(1999),buttheytella reason-
mentioned in chapter1andthefactthatfiguresfor1998areten- ablyconsistent story.The increasein thepovertyratefromthe
tativebecauseofthelimitednumberofsurveys available
(seebox lowestpoint (August-October 1997)to the highest(Septem-
1.2). ber-December 1998)isaround164percent,whichcanberegarded
5. HanmerandNaschold 1999. asthemaximum impactofthecrisisonpoverty.Thepoverty rate
6. HanmerandNaschold1999;McGee1999. peakedaroundthemiddleofthesecondhalfof 1998,following
7. Fora discussion
of therelationship
betweenenvironment thelargesurgein thepriceofriceandbeforethebeginning ofthe
andgrowth,seechapter4 ofWorldBank(200 0p). stabilizationofgeneralinflation.
8. ForfurtherinformationseetheUNICEFstatistical
data- 19.Dataonthedistribution ofconsumptionarenotavailable
baseatwww.unicef.org/statis. forboth1996and1998.Thusan estimate ofconsumption was
208 WORLD DEVELOPMENT REPORT 2000/2001
obtainedbymultiplyinlgall incomesby the share of aggregatepri-
vateconsumptionin nationalincomebasedon nationalaccounts
data. Actual consumption data, available for 1998, suggest that
the estimation procedures yield broadly accurate headcount ratios.
20. World Bank 1999bb.
21. World Bank 1999t.
22. Demery 1999.
23. iLUstig 1998.
24. Yao 1999.
25. Minot 1998; Jalan and Ravallion 1999b.
26. Lopez and della Maggiora forthcoming.
27. Bonilla-Chacin and Hammer 1999.
28. Wagstaff forthcoming b.
29. Wagstaff forthcoming b.
30. World Bank 1998t.
31. World Bank 1999t.
32. Filmer and Pritchett 1998.
33. Filmer 1999b; Ghana Statistical Service 1'999.
34. Liu, Hsiao, and Eggteston 1999.
35. Vclla 1997.
36. World Bank forthcoming a.
37 . Filmer 1999b.
38. World Bank 1998t.
39. In China ethnic minority groups make up less than 9 per-
cent of the population, but are believed to account for about 40
percent of the remaining absolute poor (World Bank 2000e).
40. Psacharopoulos and Patrinos 1994.
41. Psacharopoulos and Patrinos 1994.
42. World Bank 1999s.
43. Gragnolati 1999.
44. U.S. Census Bureau 1999.
45. World Bank 19 9 7g. A 1994 survey by the National Coun-
cil of Applied Econonmic Research shows that in rural India roughly
halfthe members ofscheduled castes and scheduled tribes lived below
the poverty line. and that while these groups represent about a third
of the population, they make up 43 percent of poor people.
46. Kozel and Parker 2000; PROBE Team 1999; World Bank
I 998r, table 2: World Bank 1999z, subnational database of India.
47. Dreze and Sen 1995.
48. Deaton 2000.
49. Baulch and Hoddinott forthcoming.
5)l. In addition, one caveat in interpreting these results is that
the pcople who movefrequently in and out of poverty may be pri-
marily those whose income is permanently close to the poverty
line. lfthat is the case, mansvwho are considered transitorily poor
could actuallv be chronically poor.
51. Jalani and Ravallion 1998b, 1999b.
52. Fields 1999.
53. Falkingham 1999.
54. Okrasa 1999.
55. Braithwaite 1997.
56. Klugmani and Braithwaitc 1998.
Chapter 2
Unless othervise noted, all quotations in the chapter are drawn from
Voicesof tbe Poor (Narayan, Chambers, Shah, and Petesch 2000:
Naravan, Patel, Schafft, Rademacher,and Koch-Schulte 2000).
Referencematerialfor this chapteris cited in the relevant
chapters of the report rather than repeating it here.
Chapter 3
Unless othervise noted, all quotations in the chapter are drawn
from Voicesof tbe Poor (Narayan, Chambers, Shah, and Petesch
2000; Naravan, I'atel, Schafft, Rademacher, and Koch-Schulte
2000).
1. Lipton and Ravallion 1995.
2. Figures are in constant 1990 U.S. dollars adjusted for dif-
ferences in purchasing power parity, as reported in Maddison
(1995).
3. Bourguignon and Morrisson 1999.
4. Sen 1999.
5. On the effects of human capital on growth, see Barro
(1997) and Bhargava and others (2000).
6. Summers 1993. T'here is debate on this. See Benhabib and
Spiegel (1994). who find no effect of human capital growth on out-
put growth, and Krueger and Lindahl (1999). who argue that the
lack of association is due to measurement error.
7. See, for example, Mankiw, Romer, and Weil (1992), who
find a consistently negative and sometimes significant effect of pop-
ulation growth on output growth consistent with the predictions
of the Solow model. Levine and Renelt (1992) find more mixed
evidence of the robustness of this result, while Kelley and Schmidt
(1994) argue more strongly for a negative effect. Young (1995)
and Bloom and Williamson (1997) provide evidence on the im-
portance of demographic change for growth in East Asia.
8. On openness and growth, see Sachs and Warner (1995) and
Frankel and Romer (1999). A critical review of the evidence is pro-
vided by Rodriguez and Rodrik (1999), and a rebuttal of this re-
view can be found in Bhagwati and Srinivasani (1999). On fiscal
policy and growth, see Easter-ly and Rebelo (1993). On inflation
and growth, see Bruno and Easterly (1998) and Barro (1997). And
on finanicial development and growth, see Levine (1997).
9. See Burnside and Dollar (forthcoming).
10. On civil uLnrest,terms of trade shocks, and growth, see East-
erly and others (1993). On volatility and growth, see Ramey and
Ramey (1995). And on slow trading partner growth, see Easterly
(2000b).
II. On corruption and growth, see Mauro (1995). On rule
of law and growth, see Kaufmann, Kraav, and Zoido-Lobaton
(1999).
12. Rodrik 1998.
13. On ethnic fragmentation and growth, see Easterly and
Levine (1997). On the role of institutions, see Easterly (2000a).
14. See Gallup, Sachs, and Mellinger (1999).
15. Ros 2000.
16. World Bank 2 0 0 0p.
17. World Bank 1997c.
18. World Bank 1997a.
19. Rovaume du Maroc, Ministere de la Prevision Economique
et du Plan, Direction de la Statistique 1999; World Bank 2000j.
20. Bruno, Ravallion, and Squire 1998; Deininger and Squire
1996b; Ravallion and Chen 1997; Dollar and Kraay 2000.

21. Dollar and Kraay2000.
22. Li, Squire, and Zou 1998.
23. World Bank 1997b.
24. Thorbecke and Jung 1996.
25. Bourguignon and Morrisson 1998.
26. Psacharopoulosand Patrinos 1994.
27. Banerjeeand others 2000.
28. Birdsalland Londofio 1997; Deininger 1999a.
29. Ravallion 1998.
30. Thomas and Wang 1998; Klasen 1999; Dollar and Gatti
1999.
31. Alesinaand Rodrik 1994; Perssonand Tabellini 1994; Per-
otti 1996a; Clarke 1995; Deininger and Squire 1998 (with land
inequaliy) .Rodriguez(1999) providesa surveyof the cross-coun-
try evidence.
32. Li and Zou 1998; Forbes forthcoming.
33. Banerjeeand Duflo 2000.
34. World Bank 1993a; Pritchett and Summers 1996; Filmer
and Pritchett 1999a; WHO 1999b; van Doorslaer and Wagstaff
1997; Gwatkin and others 2000; Schalick and others 2000;
Wagstaff 2000; Wagstaff and Watanabe 2000; Kakwani 1993;
Jamisonandothers 1996;KimandMoody 1992;Anandand Raval-
lion 1993.
35. Behrman and Knowles1997; Oxfam International 1999;
Filmerand Pritchett 1999a;Behrman 1987; Hanushekand Kimko
forthcoming; PROBE Team 1999.
36. Straussand Thomas 1998.
37. Lavv and others 1996; Thomas, Lavy,and Strauss 1996.
38. Pritchett and Summers 1996.
39. Barro 1997;Bhargavaand others2000; Hamoudi and Sachs
1999.
40. Pritchett and Summers 1996.
41. On income inequaliry and average health, see Rodgers
(1979), Flegg (1982), Waldmann (1992), and Filmerand Pritch-
ett (1999a).
42. Anand and Ravallion 1993; Bidani and Ravallion 1996.
43. Dreze and Sen 1995.
44. Filmer and Prichett 1999a.
Chapter 4
Unless otherwise noted, all quotations in the chapter are drawn
from Voicesof the Poor(Narayan, Chambers, Shah, and Petesch
2000; Narayan, Patel, Schafft, Rademacher, and Koch-Schulte
2000).
1. Rodrik (2000) presentsevidencethat democraticallyelected
governments are more likely to undertake reforms.
2. SeeCollier,Dollar,and Stern (2000)for a discussionof these
shifting perspectives.
3. Eichengreen (1999), IDB (1997), and Tommasi and Ve-
lasco (1996) all provide carefuldiscussionsof the experiencewith
reforms.
4. Tommasi and Velasco 1996.
5. Kornai (2000, p. 24-25) emphasizesthis point and cautions
againstevaluationsbasedon short-termresults."The transformation
of societyis not a horse race. The main indicator of successis not
who passesthe winning post first.Excessiveemphasison speedleads
BIBLIOGRAPHIC NOTE 209
to impatience, aggressiveness,and arrogance.... The transition
from socialismto capitalism .. . is a trial and error process.
Each element in the process might be very rapid, fairly rapid or
slow. Each has its own appropriate speed."
6. This point is emphasizedin World Bank and IMF (2000b).
7. ForMexico, seeLustig (1998); for transition economies,see
EDRB (1999).
8. Birdsall,Graham, and Sabot 1998.
9. IDB 1997.
10. Lindauer 1999.
11. World Bank 1996f.
12. Bajpai and Sachs 1999.
13. Easterly 2000b.
14. Berg and Taylor 2000.
15. Tommasi and Velasco(1996) suggestthis classificationof
reforms and provide a thorough review of the political economy
of reform.
16. Dollar and Kraay 2000; Gallup, Radelet, and Warner
1998. There is some debate,however, over the cross-countryev-
idence on the distributionalimpact of tradeliberalization. See,for
example,Lundberg and Squire (2000).
17. See, for example, Morley (1999) and Dollar and Kraay
(2000).
18. World Bank 20001.
19. Haltiwanger and Singh 1999.
20. de Ferranti and others 2000.
21. This discussiondraws heavily from Meerman (1997).
22. Sarris 1994; Alderman 1994.
23. Winters 1999.
24. This discussion draws heavily on Akiyama and others
(forthcoming).
25. Sahn, Dorosh, and Younger 1997.
26. Sahn, Dorosh, and Younger 1997.
27. World Bank 20001.
28. World Bank 20001.
29. Kemal 1994.
30. Hanson and Harrison 1999.
31. Freeman 1995.
32. The country studiesareV&ez,Kugler,and Bouill6n (1999)
for Colombia; Contreras and others (2000) and Legovini,Bouil-
16n, and Lustig (1999) for Mexico; and Ruprah and Marcano
(1999) for Venezuela.Alsosee Revenga (1997).
33. Berman and Machin 2000.
34. Pessino(1997) providesevidenceon how labor marketreg-
ulations hindered labor market adjustment in Argentina in dte
1990s.
35. Papageorgiou,Choksi, and Michaely 1995.
36. Haltiwanger and Vodopivec 1999.
37. CUTS 1999.
38. Parker, Riopelle,and Steel 1995.
39. Birdsalland de la Torre (2000) provide a valuable discus-
sion of regulatoryburdens in Latin America.
40. Hallberg 1999.
41. Hallberg 1999.
42. Asia Foundation and AKATIGA 1999.
43. SEWA 1997.
44. Haan, Coad, and Lardinois 1998.
210 WORLD DEVELOPMENT REPORT 2000/200I
45. Chisari, Estache, and Romero 1999.
46. ILO 2000.
47. Basu 1999a.
48. Examplesof such programs include Progresa (Programa
de Educaci6n, Salud y Alimentaci6n, or Program of Education,
Health, and Nutrition) in Mexicoand BolsaEscolain Brazil.ITzan-
natos (1998) shows that different incentives may be required to
keep children of different ages in school.
49. Martin and Maskus(2000)reviewthe economicarguments
for core labor standardswith an emphasison implicationsfor trade.
Aidt, Schlemmer-Schulte,and Tzannatos(2000) providea lengthy
review of the empirical evidence on the benefits of the right to
unionize and bargain collectively.
50. Pencavel(1997) reviewsarguments that the gains won by
unionized workers come at the expenseof nonunionized work-
ers rather than from profits.
51. It is increasinglyclear that it is counterproductive to use
tradesanctionsas an internationalenforcementmechanismfor core
labor standards, as the costs of such sanctions are too often borne
by the poor. See Marcus (2000) for a reviewof arguments.
52. Sable, O'Rourke, and Fung 2000.
53. Levine 1997.
54. Bennett, Goldberg, and Hunte 1996.
55. Adams 1984; Paxton and Cuevas 1996.
56. Seevon Pischke,Adams, and Donald 1984; Yaron, Ben-
jamin, and Piprek (I1997);Bravermanand Guasch (1993).
57. SeeChavesand Gonzalez-Vega(1996) and Yaron (1992),
for example.
58. Morduch 1999c.
59. Chaves and Sanchez2000; Sanchez2000.
60. See,for example,Khandker(1998) and Morduch (I 999b).
61. Hulme and Mosley 1996.
62. Morduch 1999c.
63. For example,the subsidydependence index, developedin
Yaron (1992), measuresthe extent to which the lending interest
rate would have to be raised to coverall operating costs in the ab-
sence of subsidies.
C,hapter 5
Unless othenvise noted, all quotations in the chapter are drawn
from Voicesof'thePoor (Naravan,Chambers, Shah, and Petesch
2000; Naravan, Patel, Schafft, Rademacher, and Koch-Schulte
2000).
1. Wagstaff 2000.
2. Ghana Statistical Service1999; Filmer 1999b.
3. Instituto Nacional de Estadisticav Censos 1998.
4. van de Walle 2000a.
5. Barnes, van der Plas, and Floor 1997.
6. Levy1996.Strongassociationsareoften found betweenroads
and social indicators,though the direction of causalityis unclear
(van de Walle 2000b).
7. China raised about 5.8 percent of GDP in central govern-
ment revenues, while India raised 12.2 percent (World Bank
2000s).
8. Ahmad and Stern (1987) calculated the additional cost of
raisinga rupee through the salestax in India as 60 paise.Devarajan,
Suthiwart-Narueput, and Thierfelder (2000) calculated the cost
of raising Cameroon's export tax as 1.7 times the revenue raised.
9. The debt levelsof the 41 heavilyindebted poor countries
on averagefar exceed their GDP and are about twice the levels
considered to be sustainable (van Trotsenburg and MacArthur
1999).
10. UNICEF and Oxfam International 1999.
11. Gupta, Schiff, and Clements 1996.
12. On average in 18 heavilv indebted poor countries the
share of military spending in GDP fell by the same amount (0.6
percent) as the increase in the share of education and health
spending (IMF 1999).
13. Costa Rica showsthat low militaryspendingneednot com-
promise externalor internal security.
14. Experience varied across regions, however. There were
largeincreasesin realper capita spending on educationand health
in Asiaand LatinAmericaand the Caribbean.In Sub-SaharanAfrica
per capita spending fell in education and roseonly moderatelyin
health. In transition economiesboth education and health spend-
ing declined sharply (Gupta, Clements, and Tiongson 1998).
15. Patrinos and Ariasingam1997.
16. Benefitincidenceanalvsisis a well-establishedtool for un-
derstanding who benefits from public spending, but it has limi-
tations. Average benefits, even when correctly measured, may
not be a reliableguide to the change in aggregatespending on a
givenprogram or to the distributionalimpact of a reallocationbe-
tween programs. Program participation may be nonhomoge-
neous, causing the marginal impacts of program expansion or
contraction to differ greatly from the average impacts (Lanjouw
and Ravallion 1999).
17. van de Walle 1996.
18. In Croatia in 1998 more than 90 percent of the energysub-
sidywent to nonpoor households.In Russiain 1997 about 74 per-
cent of the water subsidy went to middle- and high-income
consumers.In both casessubsidiesincreased rather than reduced
inequality (Loveiand others 2000).
19. World Bank 1994e.
20. Lewis2000.
21. In somecaseschildlabor may be a consequencerather than
a causeof childrenbeingout of school(OxfamInternational1999).
22. Bredie and Beehary 1998.
23. Oxfam International 1999.
24. Azandossessi2000; Pouliquen 1999b.
25. Constance 1999.
26. Barnesand Halpern 2000.
27. Barnes,van der Plas, and Floor 1997.
28. In energy,for example, the prospects of better servicesfor
many poor people in the foreseeablefuture will depend on find-
ing cheaper off-grid sources (Villagran2000).
29. Programade Educaci6n,Saludy Alimentaci6n,or Program
of Education, Health, and Nutrition.
30. Nolan and Turbut 1995.
31. Gilson 1998.
32. Nyonator and Kutzin 1999.
33. Gertler and Hammer 1997.
34. World Bank 1996a.
35. Boland and Whittington 2000.
 BIBLIOGRAPHIC NOTE 2I.[

36. This is the case for water supply in Jakarta. Indonesia nity organizerAkhterHameed Khan (WorldBank 1992b).Know-
(Bolandand Whittington 2000). ing the tradeoffberweenownershipand speed of implementation,
37. Loveiand others 2000. he allowed the participatory decisionmakingprocessesto take all
38. Barnes,van der Plas, and Floor 1997. the time needed.
39. Irwin 1997. 78. Pouliquen 1999a.
40. World Bank 1998t. 79. Barwell 1996.
41. Gertler and Solon 1998. 80. White 1997.
42. Filmer, Hammer, and Pritchett 1998. 81. Walker and others 1999.
43. Chomitz and others (1998) show that the paydifferential 82. Pouliquen 1999a.
needed to get medical personnelto outer areasin Indonesiais sev- 83. Rawlings.Sherhurne-Benz,and van Domelen forthcoming.
eral times any realisticamount. 84. Sara and Katz 1997.
44. See www.worldbank.org/html/extdr/hnp/health/ppi/ 85. Tendler and Freedheim 1994.
contents.htm. 86. Since 1987the World Bankhasapprovedfundingfor about
45. Girishankar 1999a. 100 multisectoralsocial fund-type projectsworth a total of $3.4
46. World Bank 1998t; Filmer, Hammer, and Pritchett 1998. billion in more than 60 countries (Parker and Serranoforthcom-
47. van der Gaag 1995. ing). Many other donors are also supporting socialfunds.
48. World Bank 1998t. 87. While choice among investmentscan help ensure sustain-
49. World Bank 1998t; Bonilla-Chacinand Hammer 1999. ability,unrestrictedchoiceisrarein multisectoralprojects.Yeta truly
50. Filmer, Hammer, and Pritchett 1998. demand-drivenapproachrequiresan open menu of investments.
51. Bettercommunications also havea positive impact on so- 88. Community ownership is difficult to instill for infra-
cial capital (Pouliquen 1999a). structurethat servesmany communities.Such infrastructureis usu-
52. World Bank 1998aa. allybetter managedby localor higher-levelgovernment(Malmberg
53. Izaguirre 1999.More than 90 developingcountriesopened Calvo 1998).
theirtelecommunicationssectorto privateparticipationin 1990-98. 89. Social funds are cross-sectoralfinancing mechanismsthat
54. Osiptel 1996. focus on poor communities but also channel grants to local gov-
55. For more on telecenters in Senegal, see www.idrc.ca/ ernmentsand NGOs. Originallyemergencyfinancingmechanisms,
acacia/engine/eng_6.htm and www.sonatel.sn/c-telece.htm. they have evolvedinto long-term mechanismsfor developmentfi-
56. World Bank 1994 e. nancing (chapter8).A common pitfallis to assumethat socialfund
57. Ernberg 1998. projects provide open choice. Some studies of social funds indi-
58. Subsidies may be required in telecommunications and cate that the menus of eligible projects may not be inclusive
such sectors as water and energy (Wellenius 1997; Barnes and enough and that projects maybe too narrowly defined(Owen and
Halpern 2000). van Domelen 1998; Walker and others 1999).
59. Wellenius 1997. 90. Pouliquen 1999a.
60. Wallsten 1999. 91. Owen and van Domelen 1998.
61. Chisari, Estache, and Romero 1999. 92. Carvalho 1999a.
62. Ravallionand Wodon forthcoming. 93. In Zimbabwe a study showed that roughly 35 percent of
63. Gaynor 1996. income in rural households comes from freelyprovided environ-
64. World Bank 1995b. mental goods and that the share increases as income declines
65. Gaynor 1998. (Cavendish 1999).
66. King and Ozler 1998. 94. Uphoff 1998.
67. Gaynor 1996. 95. In 1990 the government of India issued a circular to state
68. Gaynor 1998. governments recommending the adoption of joint forest man-
69. Jimenez and Paqueo 1996. agement on areas of state forestland. By 1995, 15 states had
70. World Bank 1995b. adopted such programs, involving local communities in manag-
71. World Bank 1995b. ing and protecting forestlandsin return for rights to use specified
72. Oxfam International 1999. forest products (Arnold forthcoming).
73. Jimenez and Sawada 1998. 96. For example,allowing villagecommunities to own neigh-
74. Educo's failure to raisestandardized test scores in mathe- borhood forests and grasslandswhile the government owns far-
matics and languagesmay reflectthe absenceof direct incentives off forestlands (Murty 1994).
for teachers, parents, and parent-teacher associations to do so 97. Arnold forthcoming.
Jimenez and Sawada 1998). 98. Arnold forthcoming.
75. These concerns include increased difficulties in imple- 99. In addition, veryhigh population growth rates may so re-
menting systemwideeducationpolicies,enforcingbroadernational duce the benefits to members that their incentive to participate
objectives,avoiding social segregation,and promoting equity if is eroded.
schoolsacceptstudentsbasedon abilityto pay(WorldBank 1995b). 100. Uphoff 1998.
76. Pouliquen 1999b. 101. A forestryproject in Nepal alloweduser communities to
77. The Orangi project wasfounded in 1980 by the commu- take over forest management, followingapproval of forest man-
ZI2 WORLD DEVELOPMENT REPORT 2000/200I
agement plans, and issued certificatesensuring long-term rights
to forest benefits (World Bank 1989).
102. Experiencealso shows that community participation is
higher when appropriate technologies allow for an adequate re-
turn from activities.For example, fuelwoodplantations that use
closelyspacedplanting,whichresultsin poor survivalratesfor trees,
elicit low levelsof community involvement. By contrast, com-
munity participation is high when treesare planted farther apart
and the plantations allow for an annual income flow from non-
timber products (agricultural intercrops, fodder or thatch grass,
commerciallyvaluableseedsor leaves;Banarjeeand others 1997).
103. Devisingeffective,autonomous,and honest court systems
is as important as designingappropriate property rights systems
(Ostrom forthcoming).
104. Banarjeeand others 1997.
105. Carnev 1998.
106. Banarjeeand others 1997.
107.Women and childrenareoften the primatycollectorsof fuel
and fodder for home consumption and saleto urban markets and
so act as de factomanagersof forests.Forestproducts are especially
importantwherewomen and childrenareunableto obtain sufficient
incomefrom agricultureor wageemploymentand fewother options
exist(Arnoldforthcoming).In such casesforest-relatedactivitiesare
likelyto be labor intensiveand householdbased(Arnold 1998).
108. Agarwal1997.
109. Agarwal1997.
110. Carney 1998.
111. de Janvry, Gordillo,Platteau,and Sadoulet forthcoming.
112.de Janvry, Gordillo, Platteau,and Sadouletforthcoming.
A secure land title facilitatesaccess to credit and induces more
environment-friendly farming practices, particularly in forest
areas. It also provides an incentive to invest in productivity-
enhancing techniques and improve local infrastructure (Hoff,
Braverman, and Stiglitz 1993; Schneider 1995).
113. Deininger and Binswanger1999. Secureproperty rights
are also a major issuefor the urban poor.
114.de Janvry, Platteau, Gordillo,and Sadoulet forthcoming.
115. Mearns (1999) distinguishesindividuals' rights, claims,
or interestsin land according to whether they may be legallyup-
held under prevailinglaw (strictlegality),are sociallyperceivedto
be legitimate regardlessof their strict legality(sociallegitimacy),
or are exercisedin practice and therefore translate into effective
control over land (effectivecontrol).
116.Mearns 1999. Sharmaand Dreze (1996), however,found
that in Palanpur, India, tenancy has lost much of its equalizing
influence.As agriculturehas becomemore capital intensive,there
has been growingequalitybetween landlords and tenants, in part
reflectingthe exclusionofthe landlessfrom tenancycontracts and
the increasedleasing-in by large landowners.
117. Parthasarthy as cited in Mearns (1999).
118.de Janvry,Gordillo, Platteau,and Sadoulet forthcoming.
119. Platteau and Baland forthcoming.
120. Sadoulet, Murgai, and de Janvry forthcoming. See box
4.4 in chapter 4.
121. Mearns 1999.
122. de Janvrv, Gordillo, Platteau,and Sadouletforthcoming.
123. Deininger 1999b.
124.de Janvry, Gordillo, Platteau,and Sadoulet forthcoming.
125. Deininger forthcoming.
126.de janvry, Gordillo, Platteau,and Sadoulet forthcoming.
127. Manv local suppliers have network servicesthat under-
sell even subsidizedpublic companies (Solo 1998a).
128. The number of large-scalewater and sanitation man-
agement contracts, leases,concessions,and divestituresin devel-
oping countries increasedfrom 4 in 1993 to 29 in 2000. Private
investment in the sector peaked in 1997 at $8.4 billion (Roger
1999).
129.The main reason for this performance is the good fit be-
tween the institutional endowment, with the long-standing ac-
ceptance of private participation, and the guarantees it provides
to investors (Menard and Clarke 2000).
130. Brook Cowen and Tynan 1999.
131.The systemsoughtto covercoststhrough industrialcross-
subsidies,but the resultingtariffswere so high that most industries
opted out. While servicedeclined to all users,poor households-
both servedand unserved-suffered the most (Yepes1999).
132. Price reforms should be implemented before switching
to private providers or introducing programs to increase public
utility efficiency(World Bank forthcoming b).
133. If needed in countries where connections are few,sub-
sidiesshould be used for new connections, not for consumption.
Where connections are widespread,such as in many countries in
EasternEurope and the former SovietUnion, well-designedblock
tariffs can be more pro-poor (Loveiand others 2000).
134.BrookCowen 1997.Regulatorystructuresneedto be both
sufficientlyrigid to provide investorsand managerswith the cer-
tainty they need to judge long-term profitability and sufficiently
flexible to adjust to changing conditions (Savedoffand Spiller
1999).
135. While much efforrhas gone into regulation to stop util-
itiesfrom abusingtheir monopoly power,relativelylittle has been
done to reducethat monopolypower (Webb and Ehrhardt 1998).
136. Brook Cowen and Tynan 1999.
137. World Bank 1998o.
ChJapter6
Unless otherwise noted, all quotations in the chapter are drawn
from Voicesof the Poor(Narayan, Chambers, Shah, and Petesch
2000; Narayan, Patel, Schafft, Rademacher, and Koch-Schulte
2000).
1. Skocpol 1992; Bates 1989; North 1990.
2. World Bank 19971.
3. Evansand Rauch 1999.
4. Woo-Cumings 1999; Evans 1999; Cheng, Haggard, and
Kang 1999.
5. Collier forthcoming.
6. van Rijckeghem and Weder 1997; Adams 1998.
7. Wade 1985.
8. Grandvoinner and Tamesis 2000.
9. Ablo and Reinikka 1998.
10. Paul 1998.
11. Marches Tropicauxet Mediterraneens1995.
12. IRIS-USAID 1996.
 BIBLIOGRAPHIC NOTE 2I3

13. Wei 1999a, 1999b. 66. For Brazil,see Tendler (1997); for Korea,Whang (1981);
14. Tanzi and Davoodi 1997; Gupta. Davoodi, and Alonso- and for Taiwan, China, Fei, Ranis, and Kuo (1979).
Terme 1998; Gray and Kaufmann 1998. 67. Coirolo 2000.
15. Norton 1998. 68. Coirolo 2000.
16. Grandvoinnet 2000. 69. Sen 1997b.
17. Berkovitz,Pistor, and Rischard 1999; Posner 1997. 70. Sen 1999.
18. Bouderbala and Pascon 1970. 71. The empiricalresult shown in figure 6.6, however,has to
19. Dakolias 1996. be taken with caution. It could also be the case that higher growth
20. Dakolias 1996. leads to institutions of better quality. Nevertheless, historical
21. Garro 1999. analysisrevealsthat the creationof some foundationalinstitutions
22. World Bank 1999a. (propertyrights,for example)waskeyin the processof development.
23. Dakolias 1996. 72. For example,Sah (1991), Bardhan(1997b),Harriss(1999),
24. Asia Foundation [www.asiafoundation.org]. Varshney (1999a), Niles (1999), and Moore and Putzel (1999).
25. Mc Clymont and Golub 2000. 73. Alesina and Rodrik 1994; Barro 1996b; Brunetti 1997;
26. Dakolias 1996. Campos 1994; Bardhan 1999a.
27. Manning 1999. 74.Johnson 1982;Wade 1991; H. Stein 1999;Woo-Cumings
28. Michael Anderson 1999; Manning 1999. 1999.
29. Yost 1999. 75. Acemoglu and Robinson 1999.
30. Bardhan 1997b. 76. Dethier 1999.
31. Bardhan 1997b; R. Adams 1986. 77. Rodrik 1999a.
32. Ranis and Stewart 1994. 78. Collier 1999c; Collier and Gunning 1999.
33. Tommasi and Weischelbaum 1999. 79. Sen 1997b; Besleyand Burgess2000.
34. Ranis and Stewart 1994. 80. Diamond, Linz, and Lipset 1988; Joseph 1999.
35. King and Ozler 1998. 81. Bardhan and Mookherjee 1999.
36. Bardhan 1997b. 82. Li, Steele,and Glewwe1999.Seealso Castro-Lealand others
37. Brown and Oates 1987. (1999).
38. Hsiao 1995. 83. Przeworskiand Limongi 1993; Bratton and van de Walle
39. EIU 1999a. 1997; Sah 1991; Varshney 1999a.
40. Crook and Sverisson1999. 84. Burki and Perry 1998.
41. Fisman and Gatti 1999; Treisman 1998. 85. de Waal 1999.
42. World Bank 20001. 86. Dreze and Sen 1995.
43. Foner 1989. 87. Joshi and Moore forthcoming.
44. Ravallion1999b,forthcoming;Litvackand Rondinelli1999.
45. Bardhan 1997b; Bardhan and Mookherjee 1999; Harriss
1999; Dreze and Sen 1995; Burki, Perry, and Dillinger 1999. Chapter 7
46. Blair 2000. Unless otherwise noted, all quotations in the chapter are drawn
47. Manor 1999. from Voicesofthe Poor(Narayan, Chambers, Shah, and Petesch
48. World Bank 1996b. 2000; Narayan, Patel, Schafft, Rademacher, and Koch-Schulte
49. Owen and van Domelen 1998. 2000).
50. Fox 1995. 1. Briggs1998; Hirschman 1984.
51. Adato and others 1999. 2. Banton 1999.
52. World Bank 1999b. 3. World Faiths Development Dialogue2000.
53. Bhatt 2000. 4. This section draws heavilyon World Bank (forthcoming a)
54. Moore and Putzel 1999. for examplesand other materials.
55. Skocpol 1992. 5. Das Gupta 1995, 1999; Goody 1976, 1990.
56. Moore and Putzel 1999. 6. World Bank forthcoming a.
57. Huntington and Nelson 1976. 7. Grayand Kevane1996;Deereand Leon 1997;Agarwal1994;
58. Uvin 1995. Sairo, Mekonen, and Spurling 1994.
59. Paerregaard1998. 8. KIHASAand UNDP 1998; Kim 1991.
60. Bebbington 1996. 9. Das Gupta and others 2000.
61. Riddell and Robinson 1995. 10. World Bank forthcoming a.
62. Brown and Ashman 1996. 11. UNDP 1995.
63. Bebbington 1996. 12. Filmer 1999b.
64. van de Walle forthcoming;Narayan,Chambers,Shah, and 13. Quisumbing 1994; Chi and others 1998; Saito and Spurl-
Petesch 1999; Narayan, Patel, Schafft, Rademacher, and Koch- ing 1992.
Schulte 2000. 14. Sen and Dreze 1989; Coale 1991; Klasen 1994.
65. Das Gupta, Grandvoinnet, and Romani forthcoming. 15. Das Gupta and others 1997; Zeng and others 1993.
214 WORLD DEVELOPMENT REPORT 2000/2001

16. Jejeebhoy 1995. 64. Anderson 1983.

17. Ren 1996; Das Gupta 1995. 65. Monnet 1988.
18. Hobcraft 1993; Hill and King 1995; Klasen 1999. 66. Austin 1999; Reno 1998.
19. Joshi and others 1999. 67. Gurr and others 1993.
20. Thomas 1990, 1997. 68. SeeChua (1998) on "ethno-nationalism" in Kazakhstan,
21. Khandker 1998. the Philippines,South Africa,Thailand, and Vietnam.
22. Jejeebhoy 1995. 69. Easterly 2 000a.
23. Dollar and Gatti 1999. 70. Collier and Hoeffler 2000; Woolcock, Pritchett, and
24. Klasen 1999. Isham 1999.
25. Saito,Mekonen, and Spurling 1994; Quisumbing 1994; 71. IDB 2000, p.2 7. Of 26 countries in the region, only four
Due and Gladwin 1991. have adequate data on Afro-Latin populations: Bolivia, Brazil,
26. Quisumbing 1996. Guatemala, and Peru.
27. Udry and others 1995; Udrv 1996. 72. Girtell and Vidal 1998; Narayan 1999.
28. Gopal 1999. 73. Fine 1999.
29. World Bank forrhcominga. 74. Woolcock 2000.
30. Sen 1999; Singh 1999. 75. Kozel and Parker 2000.
31. Jones 1998. 76. Barr 1998, 1999. Seealso van Dijk and Rabellotti (1997).
32. Das Gupta and others 2000; Agarwal1994. 77. Fafchamps and Minten 1999.
33. Agarwal1994; Andors 1983; Das Gupta and others 2000; 78. Grootaert 1999b;Grootaertand Narayan2000; Grootaert,
Davin 1976, 1995; Uberoi 1999; Wolf 1985. Oh, and Swamy 1999. For an overview of these studies, see
34. Deere and Leon 1999. Grootaert (1999a).
35. Pitt and Khandker 1998. 79. Narayan and Nvamwaya 1996.
36. Besleyand Coate 1995. 80. Bebbington and Carroll 2000.
37. Khandker, Khalily,and Khan 1996. 81. Fernandez 1994.
38. Rhyne and Holt 1994. 82. Singerman 1995.
39. SeeWorld Bank (forthcoming a). 83. Tendler 1997.
40. Wade 1987. 84. Uphoff 1992.
41. Psacharopoulosand Patrinos 1994. 85. World Bank forthcoming c.
42. However,spatialpoverty traps can (and do) emergepurely 86. Fox 1992; Bebbington 1999.
out of the interaction of factor market imperfections with exter-
nalities. SeeJalan and Ravallion (1999a).
43. Borjas 1992. C1hapter8
44. Jones 1998. Unless otherwisenoted, all quotations in the chapter are drawn
45. Borjas 1997; Borjas and Sucyoshi 1997; Turok 1999. from Voicesof the Poor(Narayan, Chambers, Shah, and Petesch
46. Davey Smith and others 1998; Lowelland Wood 1998; 2000; Narayan, Patel, Schafft, Rademacher, and Koch-Schulte
Meerman 1999. 2000).
47. The strongest empirical evidence for this comes from 1. Deaton 1997; Townsend 1995; Udry 1991.
poor urban communities in the United States. See, among 2. The recurringnature of shocks is also referredto as the de-
others, Wilson (1996) and Sampson, Morenhoff, and Earls greeofautocorrelation.Deaton(1991)developsatheoreticalmodel
(1999). showingthe effectof repeated shockson buffer stock behavior.
48. Herbst 1999. 3. Some downwardincome fluctuations are predictable, such
49. van Nieuwkoop and Uquillas 2000. as preharvestfood shortagesin rural areas. Whilethis food short-
50. Bardhan 1997a; Coate and Loury 1993. fallis strictly speakingnot a risk (although its magnitude may be
51. The distinction between the two types of affirmativeac- uncertain), householdswill still try to smooth consumption, and
tion policiesis made powerfullyby Loury (2000). thiscarriescosts.They may needto borrow,to build and fill a grain
52. Coate and Loury 1993; Steele 1999. store,or to callon a socialnetrvorkfor support, activitiesthat may
53. Holzer and Neumark 1999. be harder for the poor (Sinhaand Lipton 1999).Poor farmersmay
54. Bardhan 1997a. be forced to "presell" their crops to wealthier farmers or middle-
55. Horowitz 1999. men at below-market prices (Search 1999). Similarly, income
56. Borjas 1997; Casella and Rauch 1997. declinesassociatedwith old age areoften predictable,but their mag-
57. Bates 1999. nitude may not be becauseof uncertaintiesof health and life ex-
58. Alesina and La Ferrara 1999. pectancy. The chapter's discussion of risk management includes
59. Turton 1997. management of predictable fluctuations because many of the
60. SeeAlesina,Baqir, and Easterly (1998) on U.S. cities. tools availablefor dealing with risk also help households smooth
61. Collier and Garg 1999. predictablefluctuations.
62. Brockerhoffand Hewett 1998. 4. Townsend 1994.
63. Varshney 1999b. 5. Rosenzweigand Binswanger1993.
 BIBLIOGRAPHIC NOTE 215

6. Dercon 1999;Kinsey,Burger,and Gunning1998. 40. Ravallionand Chaudhuri1997;Deaton 1997;Town-

7. Rosenzweig and Binswanger1993. send 1995;McDonald,Schiller,and Ueda1999.
8. Jalanand Ravallion1997a. 41. Jalanand Ravallion1997a;Townsend1994.
9. Baulchand Hoddinottforthcoming. 42. See,forexample,Kochar(1999)andJacobyand Skoufias
10. Gaihaand Deolalikar1993. (1997)forIndia,Moser(1998)for EcuadorandZambia,andDer-
11. Maluccio,Haddad,and May 1999;Glewweand Hall con (1999)for Ethiopia.
1998. 43. Migrationis a copingstrategybutalsoservesinsuranceand
12. Ravallion1988. incomediversification purposes.Of course,muchmigrationtakes
13. Severepovertywasmeasuredbythe squaredpovertygap placefor reasonsotherthan riskmanagement. A detaileddiscus-
indexUa-lan and Ravallion1997a). sion of the roleof migration(domesticand international)is be-
14. Grootaert,Kanbur,and Oh 1997;Jalan and Ravallion yondthe scopeof this chapter.
1999c;Lokshinand Ravallion2000a. 44. Udry 1999.
15. Lokshinand Ravallion2000a. 45. Foster1995.
16. Derconand Krishnan2000a. 46. E. Rose1999.
17. The three-tieredframeworkof riskreduction,risk miti- 47. Foster1995.
gation, and copingis developedin Holzmannand Jorgensen 48. Behrman1988.
(2000)and WorldBank(2 000 q). Someof the literatureon risk 49. Behrmanand Deolalikar1990.
usesa differentclassificationthat labelsonlyex ante strategiesas 50. E. Rose1999.
"riskmanagement" and expost strategiesas "coping."Riskman- 51. Derconand Krishnan2000b;Behrmanand Deolalikar
agementis seenas all effortsat reducingincomevariability(di- 1990.
versification,migration). Coping consists of consumption 52. Sinhaand Lipton 1999;Lanjouwand Stern1991.
smoothing,eitheracrosstime(borrowing,self-insurance, saving) 53.FAO1995.The effectsofAIDSmaychangethecustomary
or acrosshouseholds(risksharing,throughformalor informalin- treatmentofwidows.In a TanzanianvillageAIDSwidowsareno
surance).SeeAldermanand Paxson(1992). longerdispossessed, sothat theycantakecareof the manyAIDS
18.Households diversifytheirincomesourcesnotonlyto deal orphansand compensateforthe reductionin themalelaborsup-
withriskbut alsoto increaseincome.Throughsuchactivitiesas ply(Rugalema1999).
intercropping and seasonally
varyingtheirlaborsupply,households 54. Aldermanand Gertler1997.
increasetheir meanincomeand in somecasesmayreducetheir 55. Grootaert1998.
riskaswell(Dercon1999). 56.Jacobyand Skoufias1997.
19. Reardon1997. 57. Narayan,Chambers,Shah,and Petesch2000.
20. Fafchamps,Udry,and Czukas1998. 58. Foster1995.
21. Dercon1999. 59. Hoddinottand Kinsey1998.
22. Reardon1997;Derconand Krishnan1996. 60. Zimmermanand Carter1999.
23. Adams1995. 61. Barrettand Carter1999;Grootaertand Kanbur1995a.
24. Adams1999. 62. Thispriorityrankingreflectsa concernfor preventingthe
25. Morduch1990. hardshipand deprivationthat frequentlyresultfrom shocks,es-
26. Dercon1996. peciallyfor thepoor.It doesnotimplythatallriskshouldbe elim-
27. Sinhaand Lipton 1999. inated,becausethis wouldhavestrong, undesirableincentive
28. The theoreticalmodelof precautionarysavingand asset effects.In the extreme,the absenceof allincomeriskwouldlead
buildingis discussedin Deaton(1991)and Dercon(1999). to an incentiveto undertakeno activityat all-the essenceof tht
29. The covariancebetweenassetvaluesandincomeis notjust moralhazardproblem(Devarajanand Hammer1997).
a problemfor the poor.Asthe recentEastAsianexperiencehas 63. Thisthree-pronged policyresponsetoriskis elaborated
fur-
shown,an economiccrisisusuallycoincideswitha collapseof the ther in Holzmannand Jorgensen(2000).
securitiesmarket,undercuttingtheuseof financialassetsforpre- 64. WorldBank1998t.
cautionarypurposes. 65. Zaman1999.
30. Dercon 1999. 66. WorldBank2 000 q.
31. Fafchamps,Udry,and Czukas1998. 67. Dercon1999.
32. Dercon1998. 68. Holzmannand Jorgensen2000.
33. Lim and Townsend1998. 69. Coxand Jimenez1998.
34. Jalanand Ravallion1998a. 70. Jensen1998.
35. Morduch1999a. 71. CoxandJimenez1998.
36. Dercon 1999.The extentto whichmarginalutilitiesare 72. Morduch1999a.
equalizeddependson the internalhomogeneityof the network. 73. Morduch1999a;Jensen1998.
37. Widernetworksmayfind it moredifficultto enforcethe 74. Dercon1999.
reciprocalobligationsamongmembers,however. 75. Deaton1991.
38. Morduch1999a. 76. Morduch1999a;Baulchand Hoddinottforthcoming.
39. Dercon 1999;Coateand Ravallion1993. 77. Dercon 1999.
216 WORLD DEVELOPMENT REPORT 2000/2001
78. Dercon 1999.
79. Sinha and Lipton 1999.
80. Besley1995a; Sebstadand Cohen 1999.
81. Sinha and Lipton 1999.
82. Walker and Ryan 1990.
83. Sinha and Lipton 1999.
84. Moral hazard, in this context, means the possibility that
people will engage in riskier behavior once they know the risk is
insured. Adverseselectionrefers to the tendency of people most
exposed to a certain risk to apply for insurance.
85. Sinha and Lipton 1999.
86. Holzmann, Packard, and Cuesta 1999.
87. Lund and Srinivas 1999b.
88. Holzmann, Packard, and Cuesta 1999.
89. Birdsalland Hecht 1995.
90. Gertler 2000.
91. Gertler and Gruber 1997.
92. Prescott and Pradhan 1999.
93. Gertler 2000.
94. Gertler 2000.
95. World Bank 1998t.
96. Part of the difficultv in measuring the incidence of dis-
ability stems from definitional confusion.The World Health Or-
ganization (WHO 1980) defines disability as any restriction or
lack (resulting from an impairment) of rhe ability to perform an
activity considerednormal for a human being. What constitutes
"normal" activity is clearly open to debate. In addition, many
surveys measure impairment rather than disability.And there is
a long-standing debate on whether a corrected impairment con-
stitutes a disability. See the discussion in Elwan (1999).
97. llwan 1999.
98. Khan and Durkin 1995.
99. Andersson,Palha da Sousa,and Paredes 1995.
100.Abu-Habib 1997; Harris-White 1996.
101. Andersson,Palha da Sousa,and Paredes 1995.
102. Elwan 1999.
103. Flanagan 1999.
104. By comparison, in OECD countriespension coverageis
85 percent of the labor force (Palaciosand Pallares-Miralles1999).
105. Palaciosand Pallares-Miralles1999.
106. Holzmann, Packard, and Cuesta 1999.
107. World Bank 1994b.In somecircumstancesthe defined-
contribution pillar could also be managed publicly.
108. Holzmann, Packard, and Cuesta 1999.
109. Holzmann, Packard, and Cuesta 1999; James 2000.
110.James 2000.
111. Flanagan 1999.
112. Flanagan1999; Grootaert and Braithwaite1998; World
Bank forthcoming a.
113.Extensivediscussionsof child labor issuesand policiescan
be found in GrooLaertand Kanbur (1995a), Grootaert and Pa-
trinos (1999), and Fallon and Tzannatos (1998). Also see chap-
ter 4.
114.World Bank 1995c, 2 000 q; Dar and Tzannatos 1999.
115. World Bank 1995c.
116. Grootaert and Braithwaite 1998.
117.World Bank 20 00 q.
118. van Ginneken 1999.
119. World Bank 20 00 q.
120. Grootaert and Braithwaite 1998.
121. Ravallion1991; Baulch and Hoddinott forthcoming.
122. Wodon 2000b.
123. Wodon and Minowa 2000.
124. Subbarao 1997.
125. This diversity of objectivesmeans that socialfunds can
be a deliveryvehicle for a wide range of programs. In that sense
they differ from specificinterventions, such as pensions or health
insurance,discussedelsewherein the chapter.
126. Jorgensenand van Domelen 1999; Frigenti, Harth, and
Huque 1998; Bigio 1998.
127.Jorgensenand van Domelen 1999; Frigenti, Harth, and
Huque 1998.
128.van Domelen and Owen 1998.
129. Ravallionand Wodon forthcoming.
130. Sebstadand Cohen 1999; Lipton 1998.
131. Sebstadand Cohen (1999) provide many sources docu-
menting these different uses of loans.
132. Zaman 1999.
133. Sebstadand Cohen 1999.
134.The specificrolesgovernment,donors, and NGOs should
play in the deliveryof microfinanceremainsan open question.Ex-
periencesacrosscountries varywidely.But programsoperated di-
rectlyby governmentstend to havehigher defaultrates (Morduch
1999a).
135. Sebstad and Cohen 1999.
136. Morduch 1999a, 1999c; Zeller 1999.
137. Subbarao and others 1997.
138. World Bank 1999n.
139. IDB 2000. Progresa is Programa de Educaci6n, Salud y
Alimentaci6n, or Program of Education, Health, and Nutrition.
140. Filmer and others 1998.
Chapter 9
Unlessotherwisenoted,allquotationsin the chapterare drawn from
Voicesof the Poor(Narayan, Chambers, Shah, and Petesch 2000;
Narayan, Patel, Schafft,Rademacher,and Koch-Schulte2000).
1.Joint Commonwealth Secretariat-World Bank Task Force
on the Small States 2000.
2. World Bank 20001.
3. See,for example,Lustig(forthcoming),World Bank(1999j),
Poppele, Sumarto, and Pritchett (1999), Reyes and Mandap
(1999), and Lokshin and Ravallion(2000b).
4. The estimates for Indonesia should be treated cautiously,
as the largedevaluationsof the rupiah made it difficultto estimate
price levels,particularlyin rural areas.
5. de Janvry and Sadoulet 2 000a.
6. See Lustig (1995).
7. Lokshin and Ravallion2000b.
8. Datt and Hoogeveen 2000.
9. Jacoby and Skoufias 1997.
10. See, for example,Jalan and Ravallion (1997a) for China
and Gaiha and Deolalikar (1993) for India.
11. Behrman, Duryea, and Szekely 1999.
 BIBLIOGRAPHIC NOTE 217

12. Binder 1999. 46. Vos, Velasco,and de Labastida 1999.

13. See Eichengreen (1999) and IDB (1995). 47. IDB 2000.
14. Tommasi 1999. 48. Morris and others 2000.
15. Morley 1994. These numbers differ from those in table 49. Reardon and Taylor 1996.
9.1 because they are from a different source. 50. Asset reduction was also the coping strategy of many
16. Caprio and Klingebiel1996. householdsin Zimbabwein the afternath of the 1991-92 drought.
17. Lustig 1998. Householdswere forcedto selltheir goats-a form of savingsused
18. These are not matters for emergingmarkets only. Banks to pay for secondary education-to sustain short-term con-
and other financialactors in developedcountries haveacted irre- sumption (Hicks 1993).
sponsibly. In 1998 the U.S. Federal Reservewas forced to or- 51. Hoddinott and Kinsey 1998, 2000. Owens 2000.
chestrate a bailout of Long-Term Capital Management, a hedge 52. Bonitatibus and Cook 1996.
fund based in Connecticut. 53. Ainsworth and Over 1994.
19. De Gregorio, Edwards,and Valdes forthcoming. 54. Benson and Clay 1998.
20. Kiguel 1999. 55. Benson and Clay 1998.
21. Dornbusch and Edwards 1991. 56. Robinson 1993.
22. World Bank 1998f, 2000e. 57. Freeman and MacKellar 1999b. Major destruction of in-
23. See, for example, Bourguignon and Morrisson (1992), frastructure can have an asymmetriceffect on output: the lossof
World Bank (1991b), Thorbecke (1991), deJanvry, Fargeix,and output caused by the destruction of infrastructure can exceedthe
Sadoulet (1991), and Bruno, Ravallion, and Squire (1998). increase in output caused by the eventual expansion of infra-
24. See the example in Lustig (2000). structure (World Bank 1994e).
25. See, for example, the discussion by Perry and Lederman 58. Kreimer 1999.
(1999). 59. World Bank 1999c, 1999e.
26. Gupta and others 1998. 60. Anderson and Woodrow 1989.
27. Wodon, Hicks, Ryan, and Gonzalez 2000; Hicks and 61. The government of Nicaragua, for example,is consider-
Wodon 2000. ing establishingsuch a system (World Bank 2000m).
28. Ravallion 1999c. 62. Benson and Clay 1998.
29. Ravallion1999b, pp. 13-14. 63. Levav 1999.
30. SocialMonitoringand EarlyResponseUnit (www.smeru.or. 64. For a discussion on social funds, see chapter 8.
id/about.htm). 65. Davis and Oliver-Smith 1999.
31. See note 5 for chapter 1. 66. Davis and Oliver-Smith 1999.
32. Davis, Handa, and Soto 1999. 67. Magalhaesand Glantz 1992.
33. Rodrik 1998.
34. See, for example,World Bank (forthcoming b).
35. International Federation of Red Cross and Red Crescent Chapter 10
Societies 1999, tables 5 and 12. Unless otherwise noted, all quotations in the chapter are drawn
36. Freeman 1999. from Voicesof the Poor(Narayan, Chambers, Shah, and Petesch
37. Patterns vary greatly, however, even within regions. The 2000; Narayan, Patel, Schafft, Rademacher, and Koch-Schulte
Sahel, for example,has experienceda significant decline in aver- 2000).
age rainfallbetween 1960 and 1990 compared with the previous 1. Ades and Glaeser 1999.
three decades.In contrast, southeasternAfricahas experiencedan 2. Quah 1999.
18- to 20-year cycleof relativelydrier and wetter periods,and East 3. Scandizzo1998.
Africahas not shown a change in trends or any cyclicalpatterns 4. Another reasonisthat developingcountriesmay not respond
(Benson and Clay 1998). to trade opportunities. This underlines the importance of policy
38. Gray and others 1999. and institutionalreformsin developingcountries to create a more
39. Benson and Clay 1998. favorableincentive framework.
40. USAID, OFDA 1999. The database maintained by 5. World Bank and IMF 2000b.
USAID's Office of U.S. Foreign Disaster Assistanceincludes all 6. Anderson, Hoekman, and Strutt 1999.
natural hazardsdeclareda disaster by the U.S. governmentas well 7. Anderson, Hoekman, and Strutt 1999.
as major undeclared disasters causing a substantial number of 8. Goldstein, Kaminsky,and Reinhart 2000.
deaths and injuriesand substantial damageto infrastructure,agri- 9. World Bank 20 00 g.
cultural production, and housing. 10. Kaul,Grunberg, and Stern 1999a; Kanbur, Sandler,and
41. International Federation of Red Cross and Red Crescent Morrison 1999.
Societies 1993, p. 34. 11. Cook and Sachs(1999) estimatethat lessthan 10 percent
42. Anderson and Woodrow 1989. of aid has gone to international public goods. Although it is dif-
43. Funaro-Curtis 1982. ficult to disaggregateaid flowsin this way, the 10 percent figure
44. Bhatt 1999b. is calculatedon the basisof how much aid goes to regional rather
45. Benson 1997a, 1997b, 1997c. than national projects.
2I8 WORLD DEVELOPMENT REPORT 2000/2001
12. See World Bank (1999dd) for a comparison and analysis
of the Kyoto and Montreal Protocols.
13. UNAIDS 2000.
14. National IntelligenceCouncil 2000.
15. Pilot prevention projectsand national programsin Brazil,
India, Senegal,and Thailand show that greater condom use and
safer injecting and sexualbehavior can have substantial impact.
t6. World Bank 2000a.
17. Estimate by the International AIDS Vaccine Initiative.
18. WHO 1999b.
19. These expendituresinclude but are not limited to health
research.
20. WHO 1999a.
21. For a discussionof how such a fund might operate, see
Sachs, Kremer,and Hamoudi (1999).
22. Byerleeand Heisey 1996; Smale and Heisey 1994.
23. Rosenzweig1998.
24. Evenson2000.
25. PinistrLp-Andersen,Pandya-Lorch,and Rosengrant1997;
de Janvry and others 2000.
26. Persley and Doyle 1999.
27. Alston, Pardey, and Smith 1998.
28. However,intellectualproperty rights can also facilitatethe
diffusion of knowledge.becauseparent claims are published and
sometimes contain valuable information for other potential in-
ventors (David 1993). Furthermore, patents play a rolein the cre-
ation of marketsfor informationand knowledgeby providingmore
informationto buyersand sellersof technology(PrimoBraga,Fink,
and Sepulveda2000).
29. UNDP 1999a.
30. World Bank 2 000s.
31. UNDP 1999a.
32. Breton 1965; Olson 1969.
33. Kanbur, Sandler, and Morrison 1999.
34. See Woods (1999) and Helleiner (2000a).
35. Blackhurst 1997a; Ohiorhenuan 1998; Michalopoulos
1999.
36. This initiative was instituted by the High-Level Meet-
ing on Integrated Initiatives for LDCs' Trade Development in
October 1997. The participating agenciesare the International
Monetary Fund, International Trade Center, United Nations
Conference on Trade and Development, United Nations
Development Programme, World Bank, and World Trade
Organization.
37. World Bank and IMF 2000b.
C'hapter I1l
1. OECD, DAC 1996.
2. IDA 1998.
3. Ehrenpreis 1999.
4. World Bank 1990, p. 4 .
5. OECD, DAC 2000.
6. Countries classifiedby the World Bank as low income in
1997 were those with a GNP per capita of $785 or less.
7. Germati and Randel 1998.
8. Alesina and Dollar 1998; Alesina and Weder 1999.
9. See,for example,Schraeder,Hook, and Taylor (1998) and
Maizels and Nissanke (1984).
10. Kull, Destler, and Ramsay1997.
11. See Boone (1994) and World Bank (1998b).
12. Easterly 1999a.
13. For a discussionof the changingconsensuson development
strategiesover time, see Thorbecke (2000).
14. Williamson 1990.
15. While much structural adjustment lending, especiallyin
the 1980s, emphasizedthese principles,most aid, including most
World Bank lending, continued to be in traditional projects, es-
peciallyin infrastructure,ruraldevelopment,and, increasingly,the
socialsectors and the environment.
16. See, for example,IDA (1998).
17. Nelson 1999, p. 22. Alsosee Colclough (1996).
18. Kanbur, Sandler,and Morrison 1999.
19. van de Walle and Johnston 1996, p. 55.
20. van de Walle andJohnston 1996. Seealso Collier (1997).
21. Wuyts 1996; van de Walle and Johnston 1996; Aryeetey
1996.
22. Fevzioglu,Swaroop,and Zhu 1998.
23. Devarajan, Squire, and Surhiwart-Narueput 1997.
24. Devarajanand Swaroop 1998.
25. World Bank 1992a; Isham and Kaufmann 1999.
26. Burnsideand Dollar forthcoming;Devarajan,Dollar, and
Holnigren 2000.
27. Mosleyand Eeckhout 2000.
28. Alesina and Dollar 1998. Although no systematic rela-
tionship can be found between conditionality and reform, a re-
cent 10-countrycasestudy found that conditionalitycan be useful
at the beginningof reform processesif the government is fullybe-
hind the reforms(Devarajan,Dollar,and Holmgren 2000).In these
casesthe governmentusesthe conditionsas a kind of "self-restraint"
mechanism. The problem, however, is that donors have not
adapted conditionality to different scenarios.This is likelyto be
the reason that no systematicrelationship appears between con-
ditionality and reform.
29. SeeMosley,Harrigan,andToye (1995)and Collier(1997).
30. SeeWorld Bank (1992a) and Mosley, Harrigan, and Toye
(1995).
31. Killick, Gunatilaka, and Marr 1998.
32. Mosleyand Eeckhout 2000.
33. On social institutions, see Collier and Gunning (1999).
On socialcapability,see Temple andJohnson (1998). On ethnic
fragmentation,see Easterlyand Levine(1997). On inequality,see
Benabou (1996) and Perotti (1996 a). And on geography, see
Bloom and Sachs (1998).
34. Temple 1999. I'hesefactors are also likely to have had an
effect on aid effectiveness(see Hansen and Tarp 2000).
35. Easterlyand others 1993; Rodrik 1998.
36. See,for example,OECD, DAC (1996), United Kingdom,
Secretaryof State for International Development (1997), World
Bank Partnerships Group (1998b), and UNDP (1999b).
37. IMF and IDA 1999.
38. Berg 2000. Also see World Bank (1994a) and Datta-
Mitra (1997).
39. World Bank 1998k.
 BIBLIOGRAPHIC NOTE 219

40. Collier and Dollar 2000.
41. World Bank 1998b.
42. Dollar 2000.
43. IDA 1998. IDA funds are allocated using a country per-
formance rating that combines the Country Policy and Institu-
tional Assessment (CPIA-given a weight of 80 percent in the
rating) with a measure of the IDA project portfolio performance
(weightedat 20 percent). In casesof severelvweak governancea
downward adjustment may be made to the rating to adjust for the
fact that governance criteria make up only a fourth of the CPIA
(World Bank 1999h).
44. For example,Llavadorand Roemer (2000) haveproposed
a new analyticalframeworkfor the allocation of aid, based on the
notion of equalizingopportunities among recipient countries for
achievinggrowth.
45. Benyon 1999.
46. See, for example,Guillaumont, Guillaumont Jeanneney,
and Brun (1999), Rodrik (1998), and Collier and Gunning
(1997).
47. Guillaumont and Chauvet 1999.
48. OECD 1999.
49. See, for example,Seymour and others (2000) and Buse
(1999).
50. Helleiner2000b; World Bank PartnershipsGroup 1998a.
51. Sweden,Ministry for ForeignAffairs 1999.
52. Killick 1997.
53. Devarajan, Dollar, and Holmgren 2000.
54. Many such lessonsare outlined in Walt and others (1999)
and Foster (1999).
55. OECD 1999.
56. OECD 1999.
57. Jepma 1991.
58. Chinnock and Collinson 1999.
59. Berg 1993; World Bank 1996 e.
60. van Rooy 1998.
61. Gibbs, Fumo, and Kuby 1999.Also see Kruse and others
(1997).
62. SeeDenmark, Ministry of ForeignAffairs (2000).
63. Jubilee 2000 Coalition (www.jubilee2000uk.org).
64. Oxfam International 1999; Sachs and others 1999.
65. Killick,Gunatilaka, and Marr 1998.
66. Sachsand others 1999.
67. See, for example,Claessens,Oks, and Wijnbergen (1993).
68. Claessensand others 1997.
69. Easterly1999c.Newborrowingdoesnot includerescheduling.
70. Bruno and Easterly 1996.
71. Easterly 1999c.
72. Sachs2000.
73. The issuesin this paragraphwere all raisedin wide-ranging
consultations held by the World Bank and International Mone-
tary Fund to inform and consult with partners about the poverty
reduction strategy initiative. SeeWorld Bank and IMF (2000a:2.
220 WORLD DEVELOPMENT REPORT 2000/2001
Background papers and notes
Acemoglu, Daron, and James A. Robinson. "On the Political
Economy of Institutions and Development."
Adato, Michelle,Timothy Besley,LawrenceHaddad, and John
Hoddinott. "Participation and Poverty Reduction: Issues,
Theory and New Evidencefrom South Africa."
Andcrson,MVlary B. "The Impactsof Natural Disasterson the Poor:
A Background Note."
Anderson, Michael R. "Accessto Justice and LegalProcess:Mak-
ing Legal Institutions Responsive to Poor People in LDCs."
Appleton, Simon, and Lina Song. "Income and Human Devel-
opment at the HouseholdLevel:Evidencefrom SixCountries."
Austin, Gareth. 'BackgroundNote on Civil Strifeand Povertyin
Sub-SaharanAfrica."
Banton, Michael. "Discrimination and Poverty."
Bardhan, Pranab. "PoliticalEconomy, Governance, and Poverty
Reduction."
Centre L.-J. Lebret."Attacking Poverty."
Chang, Pilwha. "Gender and Violence."
Connolly, Michelle."The Impact of Removing Licensesand Re-
strictions to Import Technology on Technological Change."
Crook,RichardC., andAlanSturlaSverrisson. "To What ExtentCan
DecentralizedFormsof GovernmentEnhancethe Development
of Pro-poorPoliciesand ImprovePoverty-AlleviationOutcomes?"
Dahl, Robert A."A Note on Politics,Institutions, Democracyand
Equality."
Dasgupta, Partha. "Valuation and Evaluation: Measuring the
Quality of Life and Evaluating Policy."
Davis, Shelton, and Anthony Oliver-Smith. "Post-Hurricane
Mitch Rehabilitation and Reconstruction Mission."
de Janvrv, Alain, Gregory' Graff, Elisabeth Sadoulet, and David
Zilberman."TechnologicalChange in Agricultureand Poverty
Reduction."
Dercon,Stefan."Income Risk,Coping Strategies,and SafetyNets."
Elwan, Ann. "Poverty and Disability."
Fields, Gary S. "Distribution and Development: A Summary of
the Evidencefor the DevelopingWorld."
Gereffi, Gary, and Martha Argetia Martinez. "Blue Jeans and
Local Linkages:The BlueJeans Boom in Torreon, Mexico."
Harriss,John."How Much DifferenceDoesPoliticsMake?Regime
DifferencesacrossIndian Statesand RuralPovertyReduction."
Herring, RonaldJ. "Political Conditions for AgrarianReformand
Poverty Alleviation."
Hossain, Naomi, and Mick Moore, with Noushin Kalati,James
Manor, and Elisa Reis. "Elites, Povertyand Development."
Houtzager,Peter P., and Jonathan Partenden. "Finding the Shape
of the Mountain: When 'the Poor' Set the Agenda."
INADES Formation Tchad. "Synthse de la consultation de la
Banque mondiale sur la pauvreteau Tchad."
Jayaraman,Rajshri. "Kerala and Uttar Pradesh:A Case Study."
Kurtz, Marcus. "The Political Economy of Pro-poor Policies in
Chile and Mexico."
Leamer, Edward E., and Peter K. Schott. "Natural Resourcesas
a Source of Latin American Income Inequality."
Lindauer, David. "Labor Market Reform and the Poor."
Lund, Frances, and Smita Srinivas."Learning from Experience:
A Framework for Social Protection for Workers in the Infor-
mal Economy."
Lustig, Nora. "Crisesand the Poor: SociallyResponsibleMacro-
economics."
Manning, Daniel S. "The Role of Legal ServicesOrganizations
in Attacking Poverty."
Meerman, Jacob. "Slow Roads to Equality: A Study of Four
Hard-Core Minorities-Issues from the Literature Review."
Moore, Mick, and James Putzel. "Politics and Poverty."
Moore, Mick, Jennifer Leavy,Peter P. Houtzager, and Howard
White. "Polity Qualities: How GovernanceAffects Poverty."
Morley, Samuel A. "The Impact of Reforms on Equity in Latin
America."
Niles, Kimberly J. "Economic Adjustment and Targeted Social
Spending:The Roleof PoliticalInstitutions (Indonesia,Mex-
ico, and Ghana)."
Pawasuthipaisit,Anan, Sombat Sakuntasathien, and Robert M.
Townsend."Reportto the Ford Foundation:Impactof the Thai
Crisis."
Pouliquen, Louis."Infrastructure and Poverty."
Prasad, Kameshwar,Paolo Belli,and Monica Das Gupta. "Links
between Poverty, Exclusion, and Health."
Quah, Danny. "6x 109: SomeDynamicsof Global Inequalityand
Growth."
Rodriguez, Francisco. "Inequality, Economic Growth and Eco-
nomic Performance."
Sebstad, Jennefer, and Monique Cohen. "Microfinance, Risk
Management and Poverty."
Sinha, Saurabh, and Michael Lipton. "Damaging Fluctuations,
Risk and Poverty: A Review."
Tlurok,Ben. "South Africa: From Exclusionto Integration."
Varshney,Ashutosh. "Democracy and Poverty."
Whitehead, Laurence, and George Gray-Molina. "The Long-
Term Politicsof Pro-poor Policies."
Winters, L. Alan. "Trade, Trade Policyand Poverty: What Are
the Links?"
Wodon, Quentin. "ExtremePoverty and Human Rights:Essays
on Joseph Wresinski."
Yaqub, Shahin. "How Equitable Is Public Spending on Health
and Education?"

Bibliography
Ablo, Emmanuel, and Ritva Reinikka.1998. "Do Budgets Really
Matter?" PolicyResearchWorking Paper 1926. World Bank,
Washington, D.C.
Abu-Habib, Lina. 1997. Genderand Disability:Womens Experi-
encesin the Middle East. Oxford: Oxfam International.
Acemoglu,Daron, and James A. Robinson. 1999. "On the Po-
litical Economy of Institutions and Development." Back-
ground paper for World Development Report 2000/2001.
Massachusetts Institute of Technology, Cambridge, Mass.;
Universityof California at Berkeley;and World Bank, Wash-
ington, D.C.
Adams, Dale. 1984. "Mobilizing Household Savings through
Rural Financial Markets." In J.D. von Pischke, Dale W.
Adams,and Gordon Donald, eds., RuralFinancialMarketsin
DevelopingCountries:Their Useand Abuse. Baltimore, Md.:
Johns Hopkins University Press.
Adams,Dale W., DouglasGraham, and J.D. von Pischke. 1992.
"Microenterprise Credit Programmes:Deja Vu." World De-
velopment20(10): 1463-70.
, eds. 1983. Limitations of Cheap Credit in Promoting
Rural Development.Economic Development Institute. Wash-
ington, D.C.: World Bank.
Adams,Richard.1986. "Bureaucrats,Peasants,and the Dominant
Coalition: An Egyptian Case Study."Journal ofD)evelopment
Studies22: 336-54.
1995."Sourcesof IncomeInequalityand Povertyin Rural
Pakistan."ResearchReport 102.InternationalFood PolicyRe-
searchInstitute, Washington, D.C.
- 1998. "The PoliticalEconomy of the Food SubsidySys-
tem in Bangladesh."Journal of DevelopmentStudies35(10):
66-88.
- 1999. "Nonfarm Income, Inequality,and Land in Rural
Egypt." PolicyResearchWorking Paper 2178. World Bank,
Washington, D.C.
Adato, Michelle, Timothy Besley,LawrenceHaddad, and John
Hoddinott. 1999. "Participation and Poverty Reduction: Is-
sues, Theory and New Evidence from South Africa." Back-
ground paper for World Development Report 2000/2001.
InternationalFood PolicyResearchInstitute and World Bank,
Washington, D.C.
Ades,Alberto,and EdwardGlaeser. 1999. "Evidenceon Growth,
IncreasingReturnsand the Sizeof the Market." QuarterlyJour-
nalofEconomics 114(3): 1025-46.
Agarwal, Bina. 1994. "Gender and LegalRights in Agricultural
Land in India." Economicand PoliticalWeekly30: A39-56.
1997. "Environmental Action, Gender Equity and
Women's Participation." Development and Change 28(1):
1-44.
Aghion, Philippe. Eve Caroli, and Cecilia Garcia-Penialosa.
1999. "Inequality and Income Growth: The Perspective of
New Growth Theories." Journal ofEconomicLiterature 27:
1615-60.
Ahmad,Ehtisham,and NicholasStern. 1987."AlternativeSources
of Government Revenue:Illustrationsfrom India, 1979-80."
BIBLIOGRAPHIC NOTE 221
In DavidNewberyand NicholasStern,eds., The Theoryof Tax-
ationfor DevelopingCountries.New York:Oxford University
Press.
Aho, Gilbert SylvainLariviere, and Frederic Martin, eds. 1997.
Manuel d'analysede lapauvrete:Applicationsau Benin. Que-
bec City: UniversiteLaval,UniversiteNationaledu Benin,and
United Nations Development Programme.
Aidt, Toke, Sabine Schlemmer-Schulte,and Zafiris Tzannatos.
2000. "Core Labor Standardsand the Freedomof Association."
World Bank, Washington, D.C.
Ainsworth, Martha, and Mead Over. 1994. "AIDS and African
Development." WorldBank ResearchObserver9(2): 203-40.
Ainsworth, Martha, and Innocent Semali. 1998. "Who Is Most
Likely to Die of AIDS? Socioeconomic Correlates of Adult
Deaths in Kigera Region, Tanzania." In Martha Ainsworth,
Lieve Fransen, and Mead Over, eds., ConfrontingAIDS:Ev-
idencefrom the Developing World. Luxembourg: European
Commission.
Ainsworth, Martha, Kathleen Beegle, and Andrew Nyamete.
1996. "The Impact of Women's Schooling on Fertility and
Contraceptive Use:A Study of Fourteen Sub-SaharanAfrican
Countries." World Bank EconomicReview 10: 85-122.
Akin, John, David K. Guilkey, and Hazel E. Denton. 1995.
"Quality of Servicesand Demand for Health Care in Nig,--
ria: A Multinomial Probit Estimation." Social Scienceand
M1edicine40(11): 1527-37.
Akiyama,Takamasa,John Baffes,Jonathan Coulter, Donald F.
Larson,and Panos Varangis.Forthcoming."Commodity Mar-
ket Reform in Africa:Why, How and What." World Bankand
Natural ResourceInstitute, Washington, D.C.
Alderman, Harold. 1994. "Ghana: Adjustment's Star Pupil?" In
DavidSahn, ed.,AdjtustingtoPolicyFailureinAfricanEconomies.
Ithaca, N.Y.: Cornell UniversityPress.
Alderman,Harold, and Paul Gertler. 1997."FamilyResourcesand
Gender Differencesin Human Capital Investments:The De-
mand for Children's Medical Care in Pakistan." In Lawrence
Haddad, John Hoddinott, and Harold Alderman, eds., In-
trahouseholdResourceAllocationin DevelopingCountries:Meth-
ods, Models, and Policy. Baltimore, Md.: Johns Hopkins
UniversityPress.
Alderman, Harold, and Victor Lavv. 1996. "Household Re-
sponses to Public Health Services:Cost and Quality Trade-
offs." World Bank ResearchObserver11(2): 3-22.
Alderman,Harold, and Christina Paxson. 1992. "Do the Poor In-
sure?A Synthesisof the Literature on Risk and Consumption
in Developing Countries." Policy Research Working Paper
1008. World Bank, Washington, D.C.
Alderman, Harold, Peter Orazem, and Elizabeth M. Paterrno.
1996. "School Quality, School Cost, and the Public/Private
SchoolChoicesof Low-IncomeHouseholdsin Pakistan.'"Im-
pact Evaluation of Education ReformsPaper 2. World Batik,
Washington, D.C.
Alderman, Harold, Simon Appleton, Lawrence Haddad, Lina
Song, and YisehacYohannes. 2000. ReducingChild Malrnu-
trition: How Far Does Income Growth Take Us?Washington,
D.C.: World Bank.
Alesina,Alberto, and David Dollar. 1998. "Who Gives Foreign
222 WORLD DEVELOPMENT REPORT 2000/2001
Aid to Whom and Why?" NBER Working Paper 6612. Na-
tional Bureauof Economic Research,Cambridge, Mass.
Alesina,Alberto, and Eliana La Ferrara. 1999. "Participation in
HeterogeneousCommunities." NBER Working Paper 7155.
National Bureau of Economic Research,Cambridge, Mass.
Alesina,Alberto, and Dani Rodrik. 1994. "Distributive Politics
and Economic Growth." QuarterlyJournalof Economics109:
465-90.
Alesina,Alberto,and BeatriceWeder. 1999."Do Corrupt Govern-
ments Receive Less Foreign Aid?" NBER Working Paper
7108. National Bureau of Economic Research,Cambridge,
Mass.
Alesina,Alberto, RezaBaqir, and William Easterly.1998. "Pub-
lic Goods and Ethnic Divisions." Policy ResearchWorking
Paper 2108. World Bank, Washington, D.C.
Alston, JulianI M., Philip G. Pardey, and Vincent H. Smith.
1998."FinancingAgriculturalR&D in Rich Countries:W hat's
Happening and Why." A ustralianjournalofAgriculturaland
ResourceEconomics42(1): 51-82.
Alston,LeeJ., Gary D. Libecap,and Robert Schneider.1996."The
Determinants and Impact of Property Rights: Land Titles on
the Brazilian Frontier." NBER Working Paper 5405. Na-
tional Bureau of Economic Research,Cambridge, Mass.
Altieri, Miguel A., and Peter Rosett. 1999. "Ten ReasonsWhy
BiotechnologyWill Not EnsureFood Security,Protect the En-
vironment and Reduce Poverty in the DevelopingWorld."
Paper presented at International Conference on Ensuring
Food Security, Protecting the Environment, and Reducing
Poverty in DevelopingCountries: Can BiotechnologyHelp?
ConsultativeGroup on InternationalAgriculturalResearchand
U.S. Academy of Sciences,October, Washington, D.C.
Altimir,Oscar, and Luis Becaria.1998. "Poliricamacroeconomica
y pobreza en AmericaLatina y et Caribe." In Enrique Ganuza,
LanceT'aylor,and SamuelA. Morley, eds., Efectosdeloscam-
biosmacroeconomicosy de as reformassobrelapobrezaurbana
en laArgentina. Mexico City: Mundi-Prensa.
Amis, Philip. 1994. "Indian Urban Poverty:LaborMarkets, Gen-
der, and Shocks."Journal of internationalDevelopment6(5):
635-43.
Anand, Sudhir, and RaviKanbur. 1993. "Inequality and Devel-
opment: A Critique."Journalof DevelopmentEconomics41:
19-43.
Anand, Sudhir, and Martin RavalLion.1993. "Human Develop-
ment in Poor Countries: On the Roleof Private Incomes and
Public Services."Journal ofEconornicPerspectives7: 133-50.
Anderson, Benedict. 1983. Imagined(Communities: Reflectionson
the Originand Spread ofA'ationalism.London: Verso.
Anderson,Jock R. 1998. "SelectedPolicyIssues in International
Agricultural Research: On Striving for International Public
Goods in an Era of Donor Fatigue." WorldDevelopment26(6):
1149-62.
Anderson,Jock R., and Dana G. Dalrymple.1999. 'World Bank,
the Grant Program, and the CGIAR: A Retrospective Re-
view."OED WorkingPaper Series,no. 1. World Bank,Wash-
ington, D.C.
Anderson, Kym. 1999. "Agriculture,DevelopingCountries, and
the NWTOMillennium Round." Centre for InternationalEco-
nomic StudiesDiscussionPaper 99/28. UniversityofAdelaide,
Australia.
Anderson, Kym, Bernard Hoekman, and Anna Strurt. 1999.
"Agricultureand the WTO: Next Steps." Paper presented at
SecondAnnual Conferenceon Global EconomicAnalysis,Av-
ernaes Conference Centre, 20-22 June, Helnaes, Denmark.
Anderson, Marn B. 1999. "The Impacts of Natural Disaster on
the Poor:A BackgroundNote." Backgroundpaper for World
DevelopmentReport2000/2001. Cornell University, Ithaca,
N.Y.; and NVorldBank, Washington, D.C.
Anderson, Mary B., and Peter Woodrow. 1989. Risingfrom the
Ashes:DevelopmentStrategies in TimesofDisaster.Boulder,Colo.:
Westview.
Anderson, Michael R. 1999. "AccesstoJustice and LegalProcess:
Making Legal Institutions Responsive to Poor People in
LDCs." Background paper for World DevelopmentReport
2000/2001. World Bank, Washington, D.C.
Andersson,Neil, Cesar Palhada Sousa,and SergioParedes. 1995.
"Social Cost of Land Mines in Four Countries: Afghanistan,
Bosnia,Cambodia, and Mozambique." BritishMedicalJour-
nal3l 1: 718-2 1.
Andors, Phyllis. 1983. The Unfinished Liberation of Chinese
Women, 1949-1980. Bloomington:Indiana UniversityPress.
Aoyama,Atsuko. 1999. Towarda VirtuousCircle:ANutritionRe-
view of the Middle Eastand North Africa.Washington, D.C.:
World Bank.
Appleton,Simon,and Lina Song. 1999."Income and Human De-
velopment at the Household Level:Evidencefrom Six Coun-
tries." Background paper for World Development Report
2000/2001. Universityof Bath, Department of Economicsand
International Development; and World Bank, Washington,
D.C.
Appleton, Simon, with Tom Emwanu, Johnson Kagugube,and
James Muwonge. 1999. "Changes in Poverty in Uganda,
1992-1997." Working Paper 99.22. Oxford University,Cen-
tre for the Study of African Economies.
Aristy,Jaime, and Andres Dauhajre. 1998. "Efectosde las polit-
icasmacroeconomicas y socialessobrela pobrezaen la Republica
Dominicana." Inter-American Development Bank, Sustain-
able Development Department, Washington, D.C.
Arnold,MichaelJ.E. 1998. "Forestrvand SustainableRuralLiveli-
hoods." In Diana Carney, ed., SustainableRural Livelihoods:
What ContributionCan We Make?Implementing theSustain-
able Rural LiveliboodsApproach.London. U.K. Department
for International Development.
. Forthcoming. "Devolution of Control of Common-
Pool Resourcesto LocalCommunities:Experiencesin Forestry."
In ALainde Janvry, Gustavo Gordillo,Jean-Philippe Platteau,
and ElisabethSadoulet,eds.,AccesstoLand, RuralPovertyand
PublicAction. New York: Oxford University Press.
Arrow, Kenneth. 1963. "Uncertainty and the WelfareEconom-
ics of Medical Care."AmericanEconomnic Review53: 941-73.
. 2000. "Observationson SocialCapital." In Partha Das-
gupta and Ismail Serageldin,eds., Social Capital:A Multifac-
etedPerspective.Washington, D.C.: World Bank.
Arulpragasam,Jehan, and Carlo delNinno. 1996. "Do Cheap Im-
ports Harm the Poor?Rural-Urban Tradeoffs in Guinea." In

DavidE. Sahn,ed., Economic ReformandthePoorin Africa.
Oxford:Clarendon.
Aryeetey,Ernest.1996."AidEffectiveness in Ghana."Overseas
DevelopmentInstitute,London.
AsiaFoundationand AKATIGA.1999. TheImpactofEconomic
CrisisonIndonesian
SmallandMedium Enterprises.
[akarta:Asia
Foundation.
Asiimwe,Delius.1997."InformalHealth Marketsand Formal
HealthFinancingPolicyin Uganda."LondonSchoolof Hy-
giene,London;and MakerereInstituteof SocialResearch,
Kampala.
Atkinson,AnthonyBarnes.1981."On Intergenerational Income
Mobilityin Britain."Journalof Post-Keynesian
Economics 3:
194-218.
1998. "EquityIssuesin a GlobalizingWorld:The Ex-
perienceof OECDCountries."Paperpresentedat conference
oneconomicpolicyand equity,International MonetaryFund,
June,Washington,D.C.
Atkinson,AnthonyBarnes,and Fran,ois Bourguignon.Forth-
coming."Povertyand Inclusionfroma World Perspective."
In Pierre-Alain
MuetandJosephE. Stiglitz,eds..GlobalGoy-
ernance,Marketsand Equity.NewYork:OxfordUniversity
Press.
Austin, Gareth. 1999. "BackgroundNote on Civil Strifeand
Povertyin Sub-Saharan Africa."Backgroundpaperfor World
Development Report2000/2001. LondonSchoolof Econom-
icsandPoliticalScience;andWorldBank.Washington,D.C.
Azandossessi,Arsene. 2000. "The Struggle for Water in Urban
Areasof Nouakchott, Mauritania." WATERfront 13. United
Nations Children's Fund, New York.
Bajpai,Nirupam, and JeffreySachs. 1999. "The Progressof Pol-
icy Reformand Variationsin Performanceat the Sub-National
Levelin India." DevelopmentDiscussionPaper 730. Harvard
Institute for International Development, Cambridge, Mass.
Bale,Malcolm. 1999. "The Rural Poor: A Thematic Approach."
World Bank,East Asiaand PacificRegion,Washington.D.C.
Bale,Malcolm, and Tony Dale. 1998. "Public SectorReform in
New Zealand and Its Relevanceto Developing Countries."
WorldBank ResearchObserver13(1): 103-21.
Banarjee, Ajit, Gabriel Campbell, Maria Concepcion J. Cruz,
Shelton H. Davis, and AugustaMolnar. 1997. "Iarticipation
in Forest Management and Conservation."Environment De-
partment Paper 49. World Bank, Washington, D.C.
Banerjee, Abhijit V., and Esther Duflo. 2000. "Inequality and
Growth: What Can the Data Say?"MassachusettsInstitute of
Technology, Department of Economics,Cambridge. Mass.
Banerjee.Abhijit V., Paul Gertler, and Maireesh Ghatak. 1998.
"Empowerment and Efficiency:The Economics of Agrar-
ian Reform." Massachusetts Institute of Technology Work-
ing Paper 98-22. Department of Economics, Cambridge.
Mass.
Banerjee, Abhijit V., Dilip Mookherjee, Kaivan Munshi, and
Debraj Ray.2000. "Inequality,Control Rights and Rent Seek-
ing: SugarCooperatives in Maharashtra."Boston University;
Massachusetts Institute of Technology, Cambridge. Mass.;
and Instituto de AnalisisEcon6mico, Barcelona.
Banerji,Arup, and Hafez Ghanem. 1997. "Does the Type of Po-
BIBLIOGRAPHIC NOTE 223
liticalRegimeMatterfor Tradeand LaborMarketPolicies"'
Economic Review11(1):171-94.
Banton,Michael.1999. "Discrimination and Poverty."Back-
groundpaperfor WorldDevelopmentReport2000/2001. Cam-
bridgeUniversity;and World Bank,Washington.D.C.
Bardhan,Pranab.1997a."Methodin the Madness?A Political-
EconomyAnalysisof the EthnicConflictsin LessDeveloped
Countries."WorldDevelopment25(9): 1381-98.
.1997b. TheRoleof Governancein EconomicDevelopmer,t:
A PoliticalEconomyApproach. OECD DevelopmentCentre
Study.Washington,D.C.: WashingtonCenterPress.
. 1999a."Democracy and Development: A ComplexRe-
lationship."In IanShapiroand CasianoHacker-Cord6n, eds..,
Democracy's Values.NewYork:CambridgeUniversityPress.
Economy,Governance
.1999b. "Political and PovertyRe-
duction."Backgroundpaperfor WorldDevelopment Report
2000/2001.Universityof Californiaat Berkeley,Instituteof
InternationalStudies;and WorldBank,Washington,D.C.
Bardhan,Pranab,andDilipMookherjee. 1999."Capmure andGov-
ernanceat Localand NationalLevels."Universityof Califcr-
niaat Berkeley,Instituteof InternationalStudies;andBoston
University,Departmentof Economics.
Bardhan,Pranab, and ChristopherUdry. 1999. Development
Microeconomics. NewYork:OxfordUniversityPress.
Barham.Christopher.2000. "The MexicanFund for Natural
Disasters(Fonden)." Box for WorldDevelopmentReport
2000/2001.World Bank,Washington,D.C.
Barnes. Douglas F., and Jonathan Halpern. 2000. "The Role of
EnergySubsidies."In Energy SectorManagement Assistance
Programme, Energyand DevelopmentReport 2000: Energv
Servicesforthe WorldsPoor.Washington, D.C.: World Bank.
Barnes.DouglasF., Robert van der Plas,and Willem Floor. 19S97.
"Tacklingthe RuralEnergyProblemin DevelopingCountries."
Financeand Development34: 11-15.
Barr. Abigail.1998. "Enterprise Performanceand the Functional
Diversity of SocialCapital." Working Paper Series,no. 98-1.
Oxford University,Institute of Economics and Statistics.
. 1999. "Collective Action and Bilateral Interaction in
Ghanaian Entrepreneurial Networks." Oxford Universir.y,
Centre for the Study of African Economies.
Barraclough, Solon. and Daniel Moss. 1999. Toward Greater
FoodSecurityin CentralAmericafollowingHurricaneMitch.
Boston: Oxfam America.
Barrett. Christopher. 1996. "On Price Risk and the InverseFarm-
Size Productivitv Relationship."Journal ofDevelopmentEco-
nomics51(2): 193-215.
. 1998a. "Food Aid: Is It DevelopmentAssistance,Trade
Promotion, Both or Neither?"AmericanJournalofAgricultural
Economics80(3): 566-71.
. 1998b."ImmiserizedGrowth in LiberalizedAgriculture."
WorldDevelopment26: 743-53.
. 1999. "On Vulnerability,Asset Povertyand Subsidiar-
ity." Comments to Ford-Rockefeller Foundation Seminar
Series Session,Managing Vulnerability and Shocks within
the Agro-Food System,20 May, New York.
Barrett, Christopher, and Michael Carter. 1999. "Can't Get
Ahead for Falling Behind: New Directions for Development
224 WORLD DEVELOPMENT REPORT 2000/2001
Policy to Escape Poverty and ReliefTraps." U.S. Agencyfor
International Development,Washington, D.C.
Barrett, Christopher, and Paul Dorosh. 1996. "Farmers'Welfare
and ChangingFood Prices:NonparametricEvidencefrom Rice
in Madagascar."AmericanJournalofAgriculturalEconomics 78:
656-69.
Barrett, Scott. 1999. "Montreal versus Kvoto: International Co-
operationand the Global Environment."In Inge Kaul,Isabelle
Grunberg, and Marc Stern, eds., GlobalPublic Goods:Inter-
national Cooperationin the 21st Century.New York: Oxford
University Press.
Forthcoming."A Theory of Full International Cooper-
ation." Journalof TheoreticalPolitics.
Barro, Robert J. 1996a. Health and EconomicGrowth.Washing-
ton, D.C.: Pan American Health Organization.
1996b. "Institutions and Growth: An Introductory
Essay."Journalof EconomicGrowth 1(1): 145-48.
.199)7. DeterminantsofEconomicGrowth:A Cross-Country
EmpiricalStudy. Cambridge, Mass.:MIT Press.
1999. "Inequality, Growth, and Investment." Harvard
Universitv, Department of Economics, Cambridge, Mass.
Barro, Robert J., and Xavier Sala-i-Martin. 1995. Economic
Growth.New York:McGraw-Hill.
Barros, Ricardo, Rosane Mlendonca,and Sonia Rocha. 1995.
"Brazil:Welfare,Inequality,Poverty,SocialIndicators,and So-
cial Programs in the 1980's." In Nora Lustig, ed., Coping
withAusterity. Washington, D.C.: Brookings Institution.
Barwell,lan. 1996. "Transport and the Village:Findings from
African Village-LevelTravel and Transport Surveysand Re-
lared StLdies." Africa Region Series Discussion Paper 344.
World Bank, Washington, D.C.
Basta, Samir S., Alan Berg, and Susan Brems. 1986. Guidelines
for Workin Autrition. Population, Health, and Nutrition De-
partment. Washington, D.C.: World Bank.
Basu,AlakaMalwade.1995. "PovertyandAIDS: The ViciousCir-
dce."Division of Nutritional SciencesWorking Paper 95.02.
Cornell University,Ithaca, N.Y.
Basu, Kaushik. 1998. "Child Labor: Cause, Consequence, and
C(ure,with Remarks on International Labor Standards." Pol-
icy ResearchWorking Paper 2027. World Bank, Washington,
D.C.
. 1999a. "Child Labor: Cause, Consequence, and Cure,
with Remarkson International Labor Standards."Journalof
FconomicLiterature38: 1083-119.
1999b. "InternationalLaborStandardsand Child Labor."
CJlalleuge42(5): 80-93.
Basu, Kaushik,and Pham Hoang Van. 1998. "The Economicsof
Child Labor." AmericanEconomicReview88: 412-27.
Bates,Robert H. 1989. BeyondtheMiracle of theMarket: ThePo-
liticalFconomyofAgrarianDevelopmentinKenya.New York:
Cambridge University Press.
1997. "Political Institutions and Economic Growth in
Africa."Development Discussion Paper 583. Harvard Insti-
tute for International Development, Cambridge, Mass.
- 1999. "Ethnicity, Capital Formation and Conflict."
Center for International Development Working Paper 27.
Harvard University, Cambridge, Mass.
Batson, Amie, and Piers Whitehead. 1999. "HIVIAIDS Vac-
cines: What Motivates Private Investment in R&D?" AIDS
Vaccine Task Force Commissioned Paper. Mercer Manage-
ment and World Bank, Washington, D.C.
BaudotSundberg,Barbara,and William Moomaw,eds. 1999. Peo-
ple and Their Planet.New York:St. Martin's.
Baulch, Bob, and John Hoddinott. Forthcoming. "Economic
Mobility and Poverty Dynamics in Developing Countries."
Journal of DevelopmentStudies.
Baum, Warren. 1986. Partners against Hunger. Washington,
D.C.: World Bank.
Bebbington,Anthony. 1996. "Organizationsand Intensification:
Campesino Federations, Rural Livelihoodsand Agricultural
Technology in the Andes and Amazonia." World Develop-
ment 24(7): 1161-78.
. 1997. "SocialCapital and Rural Intensification: Local
Organizationsand Islandsof Sustainabilityin the RuralAndes."
GeographicalJournal163(2): 189-97.
. 1999. "Capitals and Capabilities:A Frameworkfor An-
alyzing Peasant Viability, Rural Livelihoods and Poverty."
WorldDevelopment27(12): 2021-44.
Bebbington, Anthony, and Thomas Carroll. 2000. "Induced So-
cial Capital and Federations of the Rural Poor." Social Capi-
tal Initiative Working Paper 19. World Bank, Social
Development Department, Washington, D.C.
Bebbington,Anthony,and Thomas Perreault.1999. "SocialCap-
ital, Development and Access to Resources in Highland
Ecuador." EconomicGeography75(4): 395-418.
Becker,Gary. 1960. "An EconomicAnalysisof Fertiliuy."In Uni-
versities-National Bureau Committee for Economic Re-
search, Demographic and Economic Change in Developed
Countries.National Bureau of Economic Research Confer-
ence Series, no. I1. Princeton, N.J.: Princeton University
Press.
Behrman, Jere R. 1987. "Schooling in Developing Countries:
Which Countries Are the Over- and Underachievers and
What Is the Schooling Impact?" Economicsof EducationRe-
view6(2): 111-27.
. 1988. "Intra-HouseholdAllocationof Nutrients in Rural
India:Are BoysFavored?Do Parents Exhibit Inequality Aver-
sion?" OxfordEconomicPapers40(1): 32-54.
. 1996. "The Impact of Health and Nutrition on Educa-
tion." WorldBank ResearchObserver11(1): 23-37.
Behrman,Jere R., and Anil Deolalikar.1990. "The Intra-House-
hold Demand for Nutrients in Rural South India: Individual
Estimates, Fixed Effects and Permanent Income."Journalof
Human Resources24(4): 655-96.
Behrman, Jere R., and James C. Knowles. 1997. "How Strongly
Is Child SchoolingAssociatedwith HouseholdIncome?"Uni-
versity of Pennsylvania, Philadelphia; and Abt Associates,
Cambridge, Mass.
Behrman, Jere R., Suzanne Duryea, and Miguel Szekely.1999.
"Agingand EconomicOptions: LatinAmericain a World Per-
spective." Inter-American Development Bank, Office of the
Chief Economist, Washington, D.C.
Bellew,Rosemary T., and ElizabethM. King. 1993. "Educating
Women: Lessons from Experience." In Elizabeth M. King

and Anne M. Hill, eds., Barriers,Benefits,and Policies.Balti-
more, Md.: Johns Hopkins UniversityPress.
Benabou, Roland. 1996. "Inequality and Growtlh." In Ben
Bernanke and Julio Rotemberg, eds., National Bureau ofEco-
nomicResearchMacroeconomicsAnnual Cambridge,Mass.:MIT
Press.
Benhabib,Jess,and Mark M. Spiegel.1994. "The Roleof Human
Capital in Economic Development: Evidence from Aggre-
gate Cross-Country Data." Journal of Monetary Economics
34(2): 143-73.
Bennett, Lynn, Mike Goldberg,and PamelaHunte. 1996. "Own-
ership and Sustainability:Lessonson Group-BasedFinancial
Servicesfrom South Asia."Journal of InternationalDevelop-
ment 8(2): 271-88.
Benson,Charlotte. 1997a."The EconomicImpact ofNatural Dis-
astersin Fiji." Working Paper 97. OverseasDevelopment In-
stitute, London.
- 1997b. "The Economic Impact of Natural Disastersin
the Philippines." Working Paper 99. OverseasDevelopment
Institute, London.
- 1997c. "The Economic Impact of Natural Disastersin
Vietnam." Working Paper 98. OverseasDevelopment Insti-
tute, London.
Benson,Charlotte,and EdwardClay. 1998. TheImpactofDrought
onSub-SaharanAfricanEconomies: A PreliminaryExamination.
World Bank Technical Paper 401. Washington, D.C.
Benyon, Jonathan. 1999. "Assessing Aid and the Collier/Dollar
Poverty EfficientAid Allocations:A Critique." U.K. Depart-
ment for International Development,London.
Berg, Elliott J. 1993. Rethinking TechnicalCooperation:Reforms
for CapacityBuildingin Africa.New York:United Nations De-
velopment Programme.
- 2000. "Aid and Failed Reforms: The Case of Public
Sector Management."In Finn Tarp, ed., ForeignAid andDe-
velopment:LessonsLearnt and Directionsfor the Future.Lon-
don: Routledge.
Berg,Janine, and Lance Taylor. 2000. "External Liberalization,
Economic Performance, and Social Policy." Center for Eco-
nomic PolicyAnalysisWorking Paper SeriesI, Working Paper
12. New School University, New York.
Berkovitz,Daniel, Katherina Pistor, andJean-Francois Rischard.
1999. "Economic Development, Legalityand the Transplant
Effect." University of Pittsburgh; and Max Planck Institute,
Rostock, Germany.
Berman, Eli, and Stephen Machin. 2000. "Skilled-BiasedTech-
nology Transfer: Evidence of Factor-Biased Technological
Change in Developing Countries." Boston University, De-
partment of Economics.
Berry,Sara. 1989. "SocialInstitutions and Accessto Resources."
Africa 59(1): 41-55.
Bertrand,Marianne, Douglas Miller,and Sendhil Mullainathan.
1999. "Public Policy and Extended Families:Evidence from
South Africa." Princeton University,Princeton, N.J.
Besley,Timothy. 1995a. "Property Rights and Investment In-
centives:Theory and Evidence from Ghana." Journal of Po-
liticalEconomy103(5): 903-37.
- 1995b. "Savings, Credit and Insurance." In Jere R.
BIBLIOGRAPHIC NOTE 225
Behrmanand T.N. Srinivasan,eds.,HandbookofDevelopment
Economics.Amsterdam: ElsevierScience.
Besley, Timothy, and Robin Burgess. 1998. "Land Reform.,
PovertyReduction and Growth: Evidencefrom India." New
Series,DERP-13. London Schoolof Economicsand Political
Science.
. 2000. "Does Media Make Government More Respon-
sive?Theory and Evidencefrom Indian FamineReliefPolicy."
Suntory and Toyota International Centres for Economicsand
Related Disciplines (STICERD), London School of Eco-
nomics and Political Science.
Besley,Timothy, and StephenCoate. 1995. "Group Lending, Re-
payment Incentivesand Social Collateral."Journal of Devel-
opmentEconomics46: 1-18.
Bhagwati,Jagdish, and T.N. Srinivasan. 1999. "Outward Ori-
entation and Development:Are the RevisionistsRight?"Paper
contributed to festschrift in honor of Anne Krueger. [wwwv.
columbia.edu/cu/economics/].
Bhalla,Surjit. 2000. "Growth and Povertyin India:Myth and Re-
ality." Oxus Researchand Investments,New Delhi.
Bhargava,Alok. 1997."ALongitudinalAnalysisofInfant andChild
Mortality Rates in Developing Countries." Indian Economic
Review32: 141-53.
Bhargava,Alok, Dean Jamison, LawrenceLau, and Christopher
Murray. 2000. "Modellingthe Effectsof Health on Economic
Growth." World Health Organization, Geneva.
Bhatt, Ela. 1989. "Toward Empowerment." World Development
17(7): 1059-65.
. 2000. "Notes on the Politicsof PovertyReduction." In
Gudrun Kochendorfer-Luciusand Boris Pleskovic,eds., In-
clusion,Justice,and PovertyReduction.Berlin:German Foui-
dation for International Development.
Bhatt,Mihir. 1999a."Natural Disasteras National Shocksto Poor
and Development."Paper presentedat WorldDevelopmentRe-
port 2000/200] Summer Workshop.World Bank,July, Waslh-
ington, D.C.
. 1999b. "Vulnerability Matters." Paper presented at In-
ternationalDecadefor Natural DisasterReductionProgramme
Forum, United Nations Development Programme, July,
Geneva.
Bidani, Benu, and Hyeok Jong. 1999. "Explainingthe Dynanr-
icsof Inequality in Thailand: 1981-1996." World Bank, East
Asia and PacificRegion. Washington. D.C.
Bidani, Benu, and Martin Ravallion. 1996. "Decomposing So-
cial Indicators Using Distributional Data."Journal ofEcono-
metrics77(1): 125-40.
Bigio.Anthony,ed. 1998. SocialFundsand ReachingthePoor:Ex-
periencesand Future Directions.Washington, D.C.: Wodd
Bank, Economic Development Institute.
Binder,Melissa.1999. "SchoolingIndicatorsduring Mexico'sLost
Decade." Economicsof EducationReview 18(2): 183-99.
Binswanger,Hans, and Shahidur Khandker. 1995. "The Impact
of Formal Finance on the Rural Economy of India." Journal
of DevelopmentStudies32(2): 234-62.
BiodiversitySupport Program. 1993. AfricanBiodiversity:Foun-
dationfor the Future.Washington, D.C.
Birdsall,Nancy. Forthcoming. "New Findings in Economicsand
226 WORLD DEVELOPMENT REPORT 2000/2001
Demography: Policy Implications." In Nancy Birdsall,Allen
C. Kelley,and Steven Sinding, eds., PopulationDoesMatter:
Demogrdphy,Growth,and Povertyin theDevelopingIYorld.New
York:Oxford University Press.
Birdsall,Nancy, and Augusto de la Torre. 2000. "EconomicRe-
form in Unequal Latin American Societies." Carnegie En-
dowment for InternationalPeaceand Inter-AmericanDialogue,
Washington, D.C.
Birdsall,Nancv,and Robert Hecht. 1995. "Swimmingagainstthe
Tide: Strategiesfor ImprovingEquity in Health." Human Re-
sources Development and Operations PolicyWorking Paper
55. World Bank, Washington, D.C.
Birdsall,Nancy, and Robert Z. Lawrence. 1999. "Deep Integra-
tion and Trade Agreements:Good for Developing Coun-
tries?" In Inge Kaul, IsabelleGrunberg, and Marc Stern, eds.,
Global Public Goods:International Cooperationin the 21st
Century.New York: Oxford UniversityPress.
Birdsall,Nancy, and Juan Luis Londofio. 1997. "Asset Inequal-
invMatters:An Assessmentof rhe World Bank'sApproach to
PovertyReduction."AmericanEconomicReview87(2): 32-37.
Birdsall,Nancy, Carol Graham, and Richard H. Sabot. 1998. Be-
yond Tradeoff:Market Reformand EquitableGrowth in Latin
America.Washington, D.C.: Inter-American Development
Bank and BrookingsInstitution.
Birdsall,Nancy,Allen C. Kelley,and Steven Sinding, eds. Forth-
coming. PopulationDoesMatter: Demography,Growth, and
Povertyin the DevelopingWorld New York: Oxford Univer-
sitv Press.
Bisat, Amer, R. BarryJohnston, and Vasudevan Sundararajan.
1999. "SequencingFinancialReformand Liberalizationin Five
DevelopingCountries." In Barry R. Johnston and Vasudevan
Sundararajan,eds., SequencingFinancialSectorReforms:Coun-
try Experiencesand Issues.Washington, D.C.: International
Monetarv Fund.
Blackden, Mark, and Chitra Bhanu. 1999. Gender, Growthand
Poverty Reduction:Special Program of Assistancefor Africa.
World Bank Technical Paper 428. Washington, D.C.
Blackhurst, Richard. 1997a. "The Capacity of the WTO to Fill
Its Mandate." In Anne Krueger, ed., The WTO as an Inter-
nationalOrganization.Chicago: Universityof ChicagoPress.
.997b. "The WTO and the Global Economyv."World
Economy20: 527-44.
Blair. Harry. 2000. "Participation and Accountability at the Pe-
riphery: Democratic Local Governance in Six Countries."
WVorld Development28(1): 21-39.
Blair.Harrv, and Gary Hansen. 1994. "Weighingin on the Scales
of Justice: StrategicApproachesfor Donor-Supported Ruleof
LawPrograms."AssessmentReport 7. U.S.Agencyfor Inter-
nationalDevelopment,Programand Operations,Washington,
D.C.
Blanchard, Olivier,and Andrei Shleifer.2000. "Federalismwith
and without Political Centralization: China versus Russia."
NBER Working Paper 7616. National Bureauof Economic
Research,Cambridge, Mass.
Bloom, David E., and JeffreyD. Sachs. 1998. "Geography, De-
mography, and Economic Growth in Africa." BrookingsPa-
person EconomicActivity 2: 207-95.
Bloom, David E., and Jeffrey G. Williamson. 1997. "Demo-
graphic Transitions and Economic Miracles in Emerging
Asia." NBER Working Paper 6268. National Bureau of Eco-
nomic Research,Cambridge, Mass.
Bodart,Claude,andJennie L. Litvack.1993."UserFees PlusQual-
it EqualsImproved Accessto Health Care: Resultsof a Field
Experimentin Cameroon." SocialScienceandMedicine37(3):
369-83.
Boland, John, and Dale Whittington. 2000. "Water Tariff De-
sign in Developing Countries: Disadvantages of Increasing
Block Tariffs (IBTs) and Advantages of Uniform Price with
Rebate(UPR) Designs."World Bank, DevelopmentResearch
Group, Washington, D.C.
Bonilla-Chacin,Maria, and JeffreyS. Hammer. 1999. "Life and
Death among the Poorest." World Bank, Development Re-
searchGroup, Washington, D.C.
Bonitatibus, Ester, and Jonathan Cook. 1996. "Incorporating
Gender in Food Security Policies:A Handbook for Policy-
makers in Commonwealth Africa." Commonwealth Secre-
tariat, London.
Boone, Peter. 1994. "The Impact of ForeignAid on Savingsand
Growth." London Schoolof Economicsand PoliticalScience.
Booth, David,John Milimo, Ginny Bond, and SilverioChimuka.
1995. "Copingwith Cost Recovery:A Study ofthe Social Im-
pact and Responsesto Cost Recoveryin BasicServices(Health
and Education) in Poor Communities in Zambia." Task Force
on PovertyReductionWorkingPaper 3. SwedishInternational
Development Cooperation Agencv, Stockholm.
Borjas,GeorgeJ. 1992."Ethnic Capitaland IntergenerationalMo-
bility." QuarterlyJournal of Economics107(1): 123-50.
.1997. "To Ghetto or Not to Ghetto: Ethnicity and Res-
idential Segregation."NBER Working Paper 6176. National
Bureau of Economic Research,Cambridge, Mass.
Borjas,George J., and Glenn T. Sucyoshi. 1997. "Ethnicity and
the IntergenerationalTransmissionof WelfareDependency."
NBER Working Paper 6175. National Bureau of Economic
Research,Cambridge, Mass.
Bouderbala,Negib, and Paul Pascon.1970. "Ledroit et le faitdans
la soci6t6 composite." Essai d'introduction au systeme ju-
ridique marocain BESM 117. Rabat.
Bourguignon, Francois. 1999. "Inclusion, Structural Inequality
and Poverty:Interplay of Economic and SocialForces."World
Bank, ResearchAdvisorvStaff,Washington, D.C.; and Delta,
Paris.
Bourguignon, Franc,ois,and SatyaChakravarty. 1998. "Multidi-
mensionalMeasuresof Poverty."Delta Working Paper 98-12.
Paris.
Bourguignon,Francois,and Pierre-AndreChiappori. 1992. "Col-
lective Models of Household Behavior."EuropeanEconomic
Review36(2-3): 355-64.
Bourguignon,Francois,and Christian Morrisson,eds. 1992. Ad-
justmentandEquity in DevelopingCountries:ANewApproach.
Paris: Organisation for Economic Co-operation and
Development.
. 1998. "Inequality and Development: The Role of Du-
alism."Journal ofDevelopmentEconomics57: 233-57.
1999. "The Size Distribution of Income among World

Citizens." Delta and University of Paris, Departement et lab-
oratoire d'economie theorique et appliquee, Paris.
Bourguignon, Francois,Albert Berry, and Christian Morrisson.
1981. The World Distributionof Incomesbetween1950 and
1977. Paris: Ecole Normale Superieure.
Bourguignon, Fran,ois, Francisco Ferreira, and Nora Lustig.
1997. "The Microeconomics of Income Distribution Dy-
namics in East Asia and Latin America." Researchproposal.
Inter-American Development Bank and World Bank, Wash-
ington, D.C.
Bourguignon, Fran,ois, Myra Fournier, and Marc Gurgand.
1998. "Distribution, Development and Education: Taiwan,
1979-1994." World Bank, ResearchAdvisory Staff,Wash-
ington, D.C..
Bourguignon, Fran,ois, Sylvie Lambert, and Akiko Suwa-
Eisenmann. 1996. "Distribution of Export Price Risk in a De-
veloping Country." CEPR Discussion Paper 1482. Centre
for Economic PolicyResearch, London.
Boyce,James K. 1987. AgrarianImpassein Bengal.Agricultural
Growth in Bangladeshand West Bengal1949-1980. Oxford:
Oxford UniversityPress.
Braithwaite,Jeanine D. 1997. "The Old and New Poor in Rus-
sia."In Jeni Klugman,ed., Povertyin Russia:PublicPolicyand
Private Responses.Economic Development Institute. Wash-
ington, D.C.: World Bank.
Brandsma,Judith, and Rafika Chaouali. 1998. "Making Micro-
finance Work in the Middle East and North Africa."World
Bank, PrivateSectorDevelopmentand FinanceGroup, Wash-
ington, D.C.
Bratton, Michael, and Nicolas van de Walle. 1997. Democratic
Experimentsin Africa:RegimeTransitionsin ComparativePer-
spective.New York: Cambridge UniversityPress.
Brautigam, Deborah. 1997. "Substituting for the State: Institu-
tions and Industrial Development in EasternNigeria." World
Development25(7): 1063-80.
Braverman,Avishay,and Jose-Luis Guasch. 1993. "Administra-
tive Failuresin Government Credit Programs."In Karla Hoff,
Avishay Braverman, and Joseph E. Stiglitz, eds., The Eco-
nomics of Rural Organization: Theory,Practice,and Policy.
New York: Oxford UniversityPress.
Bredie,Joseph W.B., and Girindre K. Beehary. 1998. SchoolEn-
rollmentDeclinein Sub-SaharanAfrica.World Bank Discus-
sion Paper 395. Washington, D.C.
Brehm, Vicky Mancuso. 2000. "Environment, Advocacy, and
Community Participation: MOPAWI in Honduras." Devel-
opment in Practice10(1): 94-98.
Breton,Albert. 1965. "ATheory of Government Grants." Cana-
dianJournal of Economicsand PoliticalScience31: 147-57.
Briggs,Xavier. 1998. "Brown Kids in White Suburbs: Housing
Mobility and the Multiple Faces of SocialCapital." Housing
PolicyDebate9(1): 177-221.
Brockerhoff,Martin, and Ellen Brennan. 1998. "The Povertyof
Cities in Developing Regions." Populationand D)evelopment
24(1): 75-114.
Brockerhoff, Martin, and Paul Hewett. 1998. "Ethnicity and
Child Mortality in Sub-SaharanAfrica."Population Council
Working Paper 107. New York.
BIBLIOGRAPHIC NOTE 227
Brook Cowen, Penelope. 1997. "Getting the Private Sector Inr-
volved in Water: What to Do in the Poorest of Countries."
Viewpoint 103.World Bank, Finance,Private Sector,and In-
frastructure Network, Washington, D.C.
Brook Cowen, Penelope,and Nicola Tynan. 1999. "Reachingthe
Urban Poor with Private Infrastructure." Viewpoint 188.
World Bank, Finance,Private Sector,and InfrastructureNer.-
work, Washington, D.C.
Brown, Charles, and Wallace Oates. 1987. "Assistanceto th,e
Poor in a Federal System."Journal of Public Economics32:
307-30.
Brown, David, and Darcy Ashman. 1996. "Participation, Social
Capital,and IntersectoralProblem Solving:Africanand Asian
Cases." WorldDevelopment24(9):1467-79.
Brunetti, Aymo. 1997. "Political Variables in Cross-Countiy
Analysis."Journal of EconomicSurveys11(2): 163-90.
Bruno, Michael, and William Easterly. 1996. "Inflation's Chil-
dren: Tales of Crises That Beget Reforms."NBER Working
Paper 5452. National Bureau of Economic Research, Cani-
bridge, Mass.
. 1998. "Inflation Crises and Long-Run Growth." Jour-
nal ofMonetary Economics41: 3-26.
Bruno, Michael, Martin Ravallion,and Lyn Squire. 1998. "Eq-
uity and Growth in Developing Countries: Old and New
Perspectiveson the Policy Issues." In Vito Tanzi and Ke-
young Chu, eds., IncomeDistributionand High-QualityGrowth.
Cambridge, Mass.:MIT Press.
Budlender,Debbie. 1999. "The South AfricanWomen's Budget
Initiative."Paper presentedat workshop on pro-poor, gender-,
and environment-sensitivebudgets, United Nations Devel-
opment Programme in partnership with United Nations D,-
velopment Fund for Women, Community Agencyfor Social
Enquiry, 28-30 June, New York.
Bunce,Valerie.1999. "The PoliticalEconomyof Post-Socialism."
Slavic Review58(4): 756-93.
Burki, Shahid Javed, and Guillermo Perry. 1998. Beyond the
WashingtonConsensus: InstitutionsMatter.Washington, D.C.:
World Bank.
Burki,ShahidJaved,GuillermoPerry,andWilliamDillinger.1999.
BeyondtheCenter:Decentralizingthe State.Washington, D.C.:
World Bank.
Burnside,Craig, and David Dollar. Forthcoming. "Aid, Policies,
and Growth." AmericanEconomicReview.
Burr, Chandler. 2000. "Grameen VillagePhone: Its Current Sta-
tus and Future Prospects."InternationalLabourOrganization,
Geneva. [www.ilo.org/public/english/employment/ent/
papers/grameen.htm].
Burt, Ronald. 1992. StructuralHoles:The SocialStructureof Corn-
petition. Cambridge, Mass.:Harvard University Press.
Buse, Kent. 1999. "Keeping a Tight Grip on the Reins: Donor
Control over Aid Coordination and Management in
Bangladesh."Health Policyand Planning14(3): 219-28.
Byerlee,Derek, and Paul Heisey. 1996. "Past and Potential Ira-
pacts of Maize Researchin Sub-SaharanAfrica:A Critical As-
sessment." FoodPolicy21(3): 255-77.
Caldwell,John. 1986. "Routes to Law Morality in Poor Coun-
tries." Populationand DevelopmentReview12: 171-220.
228 WORLD DEVELOPMENT REPORT 2000/2001
Caldwell,John, SallyFindley,Pat Caldwell,Gigi Santow,Wendy
Cosford, Jennifer Braid, and Daphne Broers-Freeman,eds.
1990. "What We Know about Health Transition: The Cul-
tural, Socialand BehaviouralDeterminants of Health." Health
Transition Series2(1). Australian National University, Na-
tional Centre for Epidemiology and Population Health,
Canberra.
Campos,Jose Edgardo. 1996. TheKeyto theAsian Miracle:Mak-
ing Shared Growth Credible.Washington, D.C.: Brookings
Institution.
Campos, Nauro. 1994. "Why Does DemocracyFoster Economic
Development?An Assessmentof the EmpiricalLiterature."Uni-
versity of Southern California, Los Angeles.
Cao, Yuanzheng,YingyiQian, and BarryWeingast. 1998. "From
Federalism, Chinese Style, to Privatization, Chinese Style."
CEPR DiscussionPaper 1838. Centrefor EconomicPolicyRe-
search, London.
Caprio, Gerard, and Daniela Klingebiel.1996. "Bank Insolven-
cies: Cross-Countrv Experience." Policy ResearchWorking
Paper 1620. World Bank, Washington, D.C.
Carbonnier,Gilles.1998. Conflict,PostwarRebuildingandtheEcon-
omy:A CriticalReviewof the Literature.War-Torn Societies
Project OccasionalPaper 2. Geneva:United Nations Research
Institute for Social Development.
Cardoso, Eliana. 1992. "Inflation and Poverty."NBER Working
Paper 4006. National Bureau of Economic Research, Cam-
bridge, Mass.
Carney,Diana, ed. 1998.SustainableRuralLivelihoods: What Con-
tribution Can We Make? Implementingthe SustainableRural
LivelihoodsApproach.London: U.K. Department for Inter-
national Development.
Carter, MichaelR., and Rene Salgado.Forthcoming."Land Mar-
ket Liberalizationand the Agrarian Question in Latin Amer-
ica." In Alain de Janvry, Gustavo Gordillo, Jean-Philippe
Platteau, and Elisabeth Sadoulet, eds., Accessto Land, Rural
Povertyand PublicAction.New York:OxfordUniversityPress.
Carvalho, Soniya. 1999a. "Social Funds." Paper presented at
Center for Urban Development Studies, Harvard University,
12 October, Cambridge, Mass.
- 1999b. "What Are Social Funds Really Telling Us?"
Paper presented at International Conference on Poverty Re-
duction and Evaluation,World Bank, 5-7 June, Washington,
D.C.
Case, Anne, and Angus Deaton. 1998. "Large Cash Transfers to
the Elderly in South Africa." EconomicJournal 108(450):
1330-61.
Casella, Alessandra, and James E. Rauch. 1997. "Anonymous
Market and Group Ties in InternationalTrade."NBER Work-
ing Paper 6186. National Bureau of Economic Research,
Cambridge, Mass.
Castro-Leal,Florencia,Julia Dayton,LionelDemery, and Kalpana
Mehra. 1999. "Public SocialSpending in Africa: Do the Poor
Benefit?" WorldBank ResearchObserver14(1): 49-72.
Cavendish,William. 1999. "EmpiricalRegularitiesin the Poverty-
Environment Relationship of African Rural Households."
Working Paper Series, no. WPS/99.21. Oxford University,
Centre for the Study of African Economies.
Centre L-J.Lebret. 1999. "AttackingPoverty."Backgroundpaper
for WorldDevelopmentReport2000/2001. World Bank,Wash-
ington, D.C.
CGIAR (Consultative Group on International Agricultural Re-
search),SystemReview Secretariat.1998. "The International
ResearchPartnership for Food Securityand SustainableAgri-
culture."Third System Reviewof the Consultative Group on
International AgriculturalResearch.Washington, D.C.
Chambers,Robert. 1983.RuralDevelopment: PuttingtheLastFirst.
Essex:Longman Scientificand Technical.
.1995. "Povertvand Livelihoods:Whose RealityCounts?"
Institute for Development StudiesDiscussionPaper 347. Sus-
sex University, Brighton.
Chang, Pilwha. 1999a. "Gender and Violence." Background
paper for WorldDevelopmentReport2000/2001. Asian Cen-
ter for Women's Studies, Seoul; and World Bank, Washing-
ton, D.C.
. 1999b. "Impact of Economic and Financial Crisis on
Women in South Korea." Asian Center for Women's Stud-
ies, Seoul.
Chaves,Rodrigo, and Claudio Gonzalez-Vega.1996. "The De-
signof SuccessftlRuralFinancialIntermediaries:Evidencefrom
Indonesia." World Development24(1): 65-78.
Chaves,Rodrigo, and Susana Sinchez. 2000. "Romania: Finan-
cial Markets, Credit Constraints, and Investment in Rural
Areas." Report 19412-RO. World Bank, Romania Country
Unit, Washington, D.C.
Chen, Shaohua, and Martin Ravallion. 2000. "How Did the
World's Poorest Fare in the 1990s?"Policy ResearchWork-
ing Paper. World Bank, Washington, D.C.
Cheng, Tun-jeng, Stephan Haggard, and David Kang. 1999.
"Institutions and Growth in Korea and Taiwan: The Bu-
reaucracy."In YilmazAkyuz,ed., EastAsianDevelopment:New
Perspectives. London: Frank Cass.
Chi, Truong, Thi Ngoc, LisaLeimarPrice,and Mahabub M. Hos-
sain. 1998. "Impact of IPM Training on the Male and Female
Farmers'Knowledgeand Pest Control Behavior:A Case Study
of Vietnam." Working Paper. International Rice Research
Institute, Manila.
Chin, Christine B.N. 1997. "Wallsof Silenceand LateTwentieth-
Century Representations of the Foreign Female Domestic
Worker: The Case of Filipinaand Indonesian Female Servants
in Malaysia."InternationalMigrationReview31: 353-85.
Ching, Panfila. 1996. "User Fees,Demand for Children's Health
Care andAccessacrossIncomeGroups:The PhilippineCase."
SocialScienceand Medicine 41(1): 37-46.
Chinnock, Jeffrey,and Sarah Collinson. 1999. PurchasingPower:
Aid Untying, TargetedProcurementand Poverty Reduction.
London: ActionAid. [www.actionaid.org].
Chisari, Omar, Antonio Estache, and Carlos Romero. 1999.
"Winners and Losersfrom the Privatizationand Regulation
of Utilities:Lessonsfrom a General Equilibrium Model of Ar-
gentina." WorldBank EconomicReview13: 3 57-78.
Chomitz, Kenneth. 1999. "Environment-PovertyConnections in
Tropical Deforestation." Paper presented at World Develop-
ment Report 2000/2001 Summer Workshop, World Bank,
July, Washington, D.C.

Chomitz, Kenneth, and David Gray. 1996. "Roads, Land Use,
and Deforestation:A SpatialModel Appliedto Belize."World
Bank EconomicReview 10(3):487-512.
Chomitz, Kenneth, with Gunawan Setiadi, Azrul Azwar,Nusye
Ismail, and Wadiyarti. 1998. "What Do Doctors Want? De-
veloping Incentivesfor Doctors to Servein Indonesia's Rural
and Remote Areas." Policy ResearchWorking P'aper 1888.
World Bank, Washington, D.C.
Chong, Alberto, and Jesko Hentschel. 1999. "The Benefits of
Bundling."World Bank,PovertyDivision,Washington,D.C.
Chua, Amy L. 1998. "Markets, Democracy and Ethnicity: To-
ward a New Paradigm for Law and Development." YaleLaw
Journal 108(1): 1-107.
Claessens,Stijn, Daniel Oks, and Swedervan Wijnbergen. 1993.
"Interest Rates, Growth, and External Debt: The Macroeco-
nomic Impact of Mexico's Brady Deal." Policy Research
Working Paper 1147. World Bank, Washington, D.C.
Claessens, Stijn, Enrica Detragiache, Ravi Kanbur, and Peter
Wickham. 1997. "Highly Indebted Poor Countries' Debt: A
Review of the Issues." Journal of African Economies6(2):
231-54.
Clark, Diana, and Chang-Tai Hsieh. 1999. "Schoolingand Labor
Market Impact of the 1968 Nine-YearEducation Program in
Taiwan." University of California at Berkeley.
Clarke, George R.G. 1995. "More Evidenceon Income Distrib-
ution and Growth." Journal of DevelopmentEconomics47:
403-27.
Coale, Ansley.1991. "ExcessFemale Mortality and the Balance
of the Sexes: An Estimate of the Number of 'Missing Fe-
males'." Populationand DevelopmentReview 17: 517-23.
Coate, Stephen, and Glenn Loury. 1993. "Will AffirmativeAc-
tion PoliciesEliminate Negative Stereotypes?"AmericanEco-
nomicReview 83: 1220-40.
Coate, Stephen, and Martin Ravallion. 1993. "Reciprocitywith-
out Commitment: Characterization and Performance of In-
formal Insurance Arrangements." Journal of D)evelopment
Economics40: 1-24.
Cohen, John M. 1995. "Ethnicity, Foreign Aid and Economic
Growth in Sub-SaharanAfrica: The Case of Kenya." Discus-
sion Paper 520. Harvard Institute for International Develop-
ment, Cambridge, Mass.
Coirolo, Luis. 2000. Personal communication. World Bank,
Latin America and the Caribbean Region, Washington,
D.C.
Colclough, Christopher. 1996. "Education and the Market:
Which Parts of the Neoliberal Solution Are Correct?" World
Development24(4): 589-610.
Coleman, Brett E. 1999. "The Impact of Group Lending in
Northeast Thailand." Journal of DevelopmentEconomics60:
105-41.
Coleman,James. 1988. "SocialCapital in the Creation of Human
Capital." AmericanJournalof Sociology94: S95-120.
- 1990. Foundationsof Social Theory.Cambridge, Mass.:
Harvard University Press.
Colletta,Nat, MarkusKostner,and Ingo Wiederhofer.1996a. Case
Studiesin War-to-PeaceTransition: The Demobilizationand
Reintegrationof Ex-Combatantsin Ethiopia, Namibia, and
BIBLIOGRAPHIC NOTE 229
Uganda.World Bank Discussion Paper 331. AfricaTechni-
cal Department Series.Washington, D.C.
. 1996b. The TransitionfromWarto Peacein Sub-Saharan
Africa.Washington, D.C.: World Bank.
Collier, Paul. 1997. "The Failure of Conditionality." In Cather-
ine Gwin and Joan Nelson, eds., Perspectiveson Aid and De-
velopment. Policy Essay 22. Washington, D.C.: Overseas
Development Council.
. 1998. "SocialCapital and Poverty." Social Capital Ir[i-
tiativeWorking Paper4. World Bank, SocialDevelopmentDe-
partment, Washington, D.C.
. 1999a."Ethnicity,Politics,and EconomicPerformance."
World Bank, Development Research Group, Washington,
D.C.
- . 1999b. "On the EconomicConsequencesof CivilWar."
Oxford EconomicPapers51(1): 168-83.
. 1999c. "The Political Economy of Ethnicity." In Boris
PleskovicandJoseph E. Stiglitz,eds.,Annual WorldBankCon-
ference on DevelopmentEconomics1998. Washington, D.C(.:
World Bank.
- . 2000. "Implications of Ethnic Diversity."World Bank,
Development Research Group, Washington, D.C.
. Forthcoming. "Consensus-Building, Knowledge, and
Conditionality." In Boris Pleskovic and Joseph E. Stiglitz,
eds., Annual World Bank Conferenceon Development Eco-
nomics2000. Washington, D.C.: World Bank.
Collier, Paul, and David Dollar. 1999. "Can the World Cut
Povertyin Half? How Policy Reform and EffectiveAid Can
Meet the DAC Targets." World Bank, Development Re-
search Group, Washington, D.C.
- . 2000. "Aid Allocationand Poverty Reduction." World
Bank, Development ResearchGroup, Washington, D.C.
Collier, Paul, and Ashish Garg. 1999. "On Kin Groups and
Wages in the Ghanaian Labour Market." Oxford Bulletin of
Economicsand Statistics61: 133-52.
Collier, Paul, and Jan Willem Gunning. 1997. TradeShocksin
DevelopingCountries:Theoryand Evidence.Oxford:Clarendon.
. 1999. "Explaining African Economic Performance."
Journal of EconomicLiterature37: 64-111.
Collier, Paul, and Anke Hoeffler. 1998. "On the Economic
Causes of Civil War." Oxfrrd EconomicPapers50: 563-73.
.2000. "Greed and Grievance in Civil War." Policy Re-
searchWorking Paper 2355. World Bank, Washington, D.C.
Collier,Paul, DavidDollar, and NicholasStern.2000. "FiftyYears
of Development."PaperpresentedatAnnualWorldBankCc,n-
ferenceon Development Economicsin Europe, World Batik,
26-28 June, Paris.
Colson, Elizabeth. 1999. "Gendering Those Uprooted by D)e-
velopment." In Doreen Marie Indra, ed., EngenderingForced
Migration: Theoryand Practice.Oxford: Berghahn Books.
Conning, Jonathan, and Michael Kevane. 1999. "Communiry-
BasedTargetingMechanismsfor SocialSafetyNets." Williams
College,Department of Economics,Williamstown, Mass.
Connolly,Michelle.1999. "The Impactof RemovingLicensesand
Restrictionsto Import Technologyon TechnologicalChange."
Backgroundpaper for WorldDevelopmentReport2000/2001.
World Bank, Washington, D.C.
230 WORLD DEVELOPMENT REPORT 2000/2001
Constance, Paul. 1999. "What Price Water?" IDB Anmerica
(July-August): 3-5.
Contreras, Dante, David Bravo,Tomas Rau, and Sergio Urzua.
2000. "Income Distribution in Chile, 1990-1998: Learning
from Microeconomic Simulations." Universidad de Chile,
Department of Economics, Santiago.
Cook, Lisa D., and JeffreySachs. 1999. "Regional Public Goods
in International Assistance."In Inge Kaul, IsabelleGrunberg,
and Marc Stern, eds., GlobalPublic Goods:InternationalCo-
operationin the21stCentury.NewYork:OxfordUniversityPress.
Coppin, Addington, and Reed Neil Olsen. 1998. 'Earnings and
Ethnicity in Trinidad and Tobago." Journal of Development
Studies34(3): 116-34.
Cornia,GiovanniAndrea.1999. "Liberalization,Globalizationand
Income Distribution." WIDER Working Paper 157. United
Nations University, Helsinki.
Cornia, Giovanni Andrea, Richard Jolly, and Frances Stewart.
1987. Adjustment with a Human Face. New York: Oxford
Universitv Press.
Cowell, Frank. 1995. AMeasuring Inequality.New York: Prentice
Hall/Harvester Wheatsheaf.
Cox, Donald, and EmmanuelJimenez. 1998. "Risk Sharing and
PrivateTransfers:What about Urban Households?"Economic
Developmentand Cultural Change46(3): 621-37.
Cox, Donald, EmanuelaGalasso,and EmmanuelJimenez. 2000.
"Inter-Country Comparisons of Private Transfers." World
Bank, Development ResearchGroup, Washington, D.C.
Cox Edwards,Alejandra.2000. "Pension Projections for Chilean
Men and Women: Estimatesfrom Social SecurityContribu-
tions." Background paper for forthcoming World Bank Pol-
icy ResearchReport En GenderingDevelopment.World Bank,
Washington, D.C.
Crook, Richard C., and Alan Sturla Sverrisson.1999. "To What
Extent Can DecentralizedFormsof Government Enhancethe
Development of Pro-poor Policies and Improve Poverty-
AlleviationOutcomes?" Backgroundpaper for WorldDevel-
opmentReport2000/2001. U.K. Department for International
Development,London; and World Bank, Washington, D.C.
CSIR (Council of Scientific and Industrial Research). 1998.
"Knowledge in Development: Multi-Media, Multi-Purpose
Community Information Centers as Catalystsfor BuildingIn-
novative Knowledge-BasedSocieties."Background paper for
World DevelopmentReport1998. World Bank, Washington,
D. C.
Currie, Janet, and Ann Harrison. 1997. "Sharing the Costs: The
Impact of Trade Reform on Capital and Labor in Morocco."
Journal of LaborEconomics15: 44-71.
CUTS (ConsumerUnitv and Trust Societv). 1999. "Conditions
Necessaryfor the Liberalizationof Trade and Investment to
ReducePoverty." Final report to U.K. Department for Inter-
national Development, London.
Dahl, Robert A. 1999. "A Note on Politics,Institutions, Democ-
racy and Equality." Background paper for World Develop-
mentReport2000/2001. World Bank, Washington, D.C.
Dakolias, Maria. 1996. TheJudicial.Sectorin Latin America and
the Caribbean:Elementsof Reform.World Bank Technical
Paper 319. Washington, D.C.
Daly, Mary, Greg Duncan, George Kaplan, and John Lynch.
1998. "Macro-to-MicroLinks in the RelationbetweenIncome
Inequalityand Mortality."Milbank Quarterly76(3): 315-39.
Dar, Amit, and Zafiris Tzannatos. 1999. "Active Labor Market
Policies." Social Protection Discussion Paper 9901. World
Bank, Human Development Network, Washington, D.C.
Das Gupta, Monica. 1995. "LifecoursePerspectiveson Women's
Autonomy and Health Outcomes." AmericanAnthropologist
97(3): 481-91.
. 1999. "LifeboatversusCorporate Ethic: Social and De-
mographic Implications of Stem and Joint Families."Social
Scienceand Medicine49(2): 173-84.
. 2000. "Social Exclusion and Poverty: Preliminary
Thoughts for World DevelopmentReport2001." In Gudrun
Kochendorfer-Luciusand BorisPleskovic,eds., Inclusion,Jus-
tice,and PovertyReduction.Berlin:German Foundation for In-
ternational Development.
Das Gupta, Monica, and Li Shuzhuo. 1999. "Gender Biasand
the Marriage Squeeze in China, South Korea and India,
1920-1990:The Effectsof War, Famineand FertilityDecline."
Developmentand Change30(3): 619-52.
Das Gupta, Monica, Helene Grandvoinnet,and Mattia Romani.
Forthcoming."State-Communitv Synergiesin Development:
Layingthe Basisfor CollectiveAction." PolicyResearchWork-
ing Paper. World Bank, Washington, D.C.
Das Gupta, Monica, Jiang Zhenghua, Xie Zhenming, and Li
Bohua. 1997. "The Status of Girls in China." Proceedingsof
the Symposium of Chinese Scholars on the Demography of
China. International Union for the Scientific Study of Pop-
ulation, October, Beijing.
Das Gupta, Monica, Sunhwa Lee, Patricia Uberoi, Danning
Wang, Lihong Wang, and Xiaodan Zhang. 2000. "State Pol-
icy and Women's Autonomy in China, South Korea and
India, 1950-2000: Lessons from Contrasting Experience."
BackgroundpaperforforthcomingWorldBankPolicyResearch
Report EnGenderingDevelopment.Also presented at Popula-
tion Associationof Americaannual meeting, March, Los An-
geles.
Dasgupta, Partha. 1999. "Valuation and Evaluation: Measuring
the Quality of Lifeand EvaluatingPolicy."Backgroundpaper
for World DevelopmentReport 2000/2001. Cambridge Uni-
versity;and World Bank, Washington, D.C.
Datt, Gaurav. 1998. "Poverty in India and Indian States: An
Update." Discussion Paper 47. International Food PolicyRe-
searchInstitute, Food Consumption and Nutrition Division,
Washington, D.C.
- 1999. "Has Povertyin India Declinedsincethe Economic
Reforms?"Economicand PoliticalWeekly34(50).
Datt, Gaurav,and Hans Hoogeveen.2000. "El Nifo or El Peso?
Crisis, Povertyand Income Distribution in the Philippines."
World Bank,East Asia and PacificRegion, Washington,D.C.
Datt, Gaurav, and Martin Ravallion. 1998. "Farm Productivity
and Rural Poverty in India." Journal of DevelopmentStudies
34: 62-85.
Datta-Mitra, Jayati. 1997. Fiscal Management in Adjustment
Lending.Washington, D.C.: World Bank.
Davey Smith, George,James D. Neaton, Deborah Wentworth,

Rose Stamler, and Jeremiah Stamler. 1998. 'Mortality Dif-
ferencesbetween Blackand White Men in the USA: Contri-
bution of Incomeand Other RiskFactorsamongMen Screened
for the MRFIT." Lancet351(9107): 934-36.
David, Paul. 1993. "Intellectual Property Institutions and the
Panda'sThumb: Patents,Copyrights,and TradeSecretsin Eco-
nomic Theory and History." In Mitchel B. Wallerstein,
Roberta A. Schoen, and Mary E. Mogee, eds., GlobalDi-
mensionsoflntellectualProperty Rightsin Scienceand Technology.
Washington, D.C.: National Academy Press.
Davin, Delia. 1976. Womenwork:Womenand the Party in Revo-
lutionaryChina. Oxford: Clarendon.
- 1995. "Women, Work and Property in the Chinese
Peasant Household of the 1980s." In Diane Elson,ed., Male
Bias in theDevelopmentProcess.New York:Manchester Uni-
versity Press.
Davis, Benjamin,Sudhanshu Handa, and Humberto Soto. 1999.
"Crisis,Poverty,and Long-TermDevelopment:Examiningthe
Mexican Case." International Food PolicyResearchInstitute,
Washington, D.C.
Davis,Shelton,and AnthonyOliver-Smith.1999."Post-Hurricane
Mitch Rehabilitation and Reconstruction Mission." Back-
ground note for WorldDevelopmentReport2000/2001. World
Bank, Washington, D.C.
de Ferranti, David, Guillermo Perry, Indermit Gill, and Luis
Serven.2000. SecuringOur Future in a GlobalEconomy.Latin
American and Caribbean Studies Series.Washington, D.C.:
World Bank.
De Gregorio, Jose, Sebastian Edwards, and Rodrigo Valdes.
Forthcoming."Controls on Capital Inflows:Do They Work?"
Journal ofDevelopmentEconomics.
de Haan, Arjan. 1997. "Poverty and Social Exclusion:A Com-
parison of Debates on Deprivation." Poverty Research Unit
Working Paper 2. SussexUniversity, Brighton.
deJanvry, Alain, and Elisabeth Sadoulet. 1999a. "AssetPositions
and Income Strategiesamong Rural Households in Mexico:
The Role of Off-farm Activitiesin PovertyReduction." Uni-
versity of California at Berkeley,Department of Agricultural
and Resource Economics; and World Bank, Development
Research Group, Washington. D.C.
- 1999b. "Growth, Povertyand Inequality in Latin Amer-
ica:A Causal Analysis,1970-94." Paper presented at confer-
ence on social protection and poverty, Inter-American
Development Bank, February,Washington, D.C.
1999c. "Rural Povertyand the Design of EffectiveRural
Development Strategies."University of California at Berke-
ley, Department of Agricultural and Resource Economics;
and World Bank, Development ResearchGroup, Washing-
ton, D.C.
- 2 000a. "Growth, Povertyand Inequality in Latin Amer-
ica: A Causal Analysis, 1970-1994." In Nora Lustig, ed.,
Shieldingthe Poor:SocialProtectionin the DevelopingWorld.
Washington, D.C.: Brookings Institution.
2000b. "Making Investment in the Rural Poor into
Good Business:New Perspectivesfor Rural Development in
Latin America."Paper presented at Social Equity Forum, an-
nual meeting of boards of governorsof Inter-American De-
BIBLIOGRAPHIC NOTE 231
velopment Bank and Inter-American Investment Corpora-
tion, 24-25 March, New Orleans.
de Janvry,Alain,Andre Fargeix,and ElisabethSadoulet.1991. "Po-
liticalEconomyofStabilizationPrograms:Feasibility,Growth
and Welfare."Journalof PolicyModeling 13: 317-45.
de Janvry, Alain, Gustavo Gordillo,Jean-Philippe Platteau, and
Elisabeth Sadoulet, eds. Forthcoming. Accessto Land, Rural
Povertyand PublicAction.New York:Oxford UniversityPress.
de Janvry, Alain, Gregory Graff, Elisabeth Sadoulet, and David
Zilberman. 2000. "TechnologicalChange in Agricultureand
PovertyReduction."Backgroundpaper for WorldDevelopment
Report2000/2001. University of California at Berkeley,De-
partment ofAgriculturaland ResourceEconomics;and World
Bank, Development ResearchGroup, Washington, D.C.
de Janvry, Alain, Jean-Philippe Platteau, Gustavo Gordillo, and
Elisabeth Sadoulet. Forthcoming. "Accessto Land and Land
PolicyReforms."In Alain de Janvry, Gustavo Gordillo,Jean-
PhilippePlatteau, and ElisabethSadoulet,eds.,AccesstoLand,
Rural Povertyand PublicAction. New York: Oxford Univer-
sity Press.
de Janvry, Alain, Elisabeth Sadoulet, Benjamin Davis, and Gus-
tavo Gordillo de Anda. 1996. "Ejido Sector Reforms: From
Land Reform to Rural Development." In Laura Randall, ed.,
ReformingMexico'sAgrarian Reform.Armonk, N.Y.: M.E.
Sharpe.
de la Rocha, Mercedes. 1995. "The Urban Family and Poverty
in Latin America."Latin AmericanPerspectives 22(2): 12-31.
De Silva, Samantha. 1999. "Community Contracting in Bank-
Funded Projects:A ReviewofStakeholder Experience."World
Bank, Human Development Network, Washington, D.C.
De Soto, Hernando. 1989. The OtherPath: The InvisibleRevo-
lution in the Third World.New York: Harper and Row.
. Forthcoming. The Mysteryof Capital.
de Waal, Alex. 1991. "Emergency Food Security in Western
Sudan: What Is It For?" In S. Maxwell, ed., To Cure,All
Hunger:FoodPolicyand FoodSecurityin Sudan. London: In-
termediate Technology Development Group.
. 1999. "Democratic Political Processesand the Fight
againstFamine."Universityof London, Schoolof Oriental aLnd
African Studies,International AfricaInstitute.
Deaton, Angus.1991. "Savingsand LiquidityConstraints."Econo-
metrica59(5): 1221-48.
. 1997. TheAnalysisofHouseholdSurveys:AMicroeconomnic
Approach.Baltimore, Md.: Johns Hopkins University Press.
. 2000. "Enrollment of Children in School in the 42nd
(1986-87) and 52nd (1995-96) Rounds of the NSS." World
Bank, Washington, D.C.
Deaton, Angus,andAlessandroTarozzi. 1999."Pricesand Poverty
in India." Princeton University, Princeton, N.J.
Deere,Carmen Diana, and MagdalenaLeon. 1997. "Women and
Land Rights in the Latin American Neoliberal Counter Re-
forms."Women in InternationalDevelopmentWorking Paper
264. Michigan State University, East Lansing.
. 1999. "Institutional Reform of Agriculture under Neo-
liberalism:The Impact of the Women's and IndigenousMcve-
ments." Keynote address at the conference Land in Latin
America:New Context, New Claims,New Concepts, Centre
232 WORLD DEVELOPMENT REPORT 2000/2001
for Latin AmericanResearchand Documentation, Centre for
Resource Studies, and Wageningen Agricultural University,
26-27 May, Amsterdam.
- Forthcoming.Gender,Propertyand Empowerment:Land,
State and Market in Latin America.Pittsburgh: Universityof
Pittsburgh Press.
Deininger, Klaus. 1999a. "Asset Distribution, Inequality, and
Growth." World Bank, DevelopmentResearchGroup,Wash-
ington, D.C.
- . 1999b. "Making Negotiated Land Reform Work: Ini-
tial Experiencesfrom Brazil,Colombia,and South Africa."Pol-
icy ResearchWorking Paper 2040. World Bank, Washington,
D.C.
- 1999c. "Negotiated Land Reform: Brazil, Colombia,
South Africa." Draft. World Bank, Development Research
Group, Washington, D.C.
Forthcoming."Negotiated Land Reformas One Way of
Land Access:Initial Experiencesfrom Colombia, Brazil and
South Africa." In Alain de Janvry, Gustavo Gordillo, Jean-
PhilippePlatteau, and ElisabethSadoulet,eds.,Accessto Land,
Rural Povertyand Public Action. New York: Oxford Univer-
sity Press.
Deininger, Klaus, and Hans Binswanger. 1999. "The Evolution
of the World Bank's Land Policy." WorldBank ResearchOb-
server14(2): 247-76.
Deininger, Klaus,and Gershon Feder. 1998. "Land Institutions
and Land Markets." Policy Research Working Paper 2014.
World Bank, Washington, D.C.
Deininger, Klaus, and Bart Minten. 1996."Poverty, Policies and
Deforestation: The Case of Mexico." EconomicDevelopment
and Cultural Change47: 313-44.
Deininger, Klaus, and Pedro Olinto. 2000. "Asset Distribution,
Inequality, and Growth." Policy Research Working Paper
2375. World Bank, Washington, D.C.
Deininger,Klaus,and Lyn Squire.1996a. "MeasuringIncome In-
equality: A New Database." Development Discussion Paper
537. Harvard Institute for International Development,Cam-
bridge, Mass.
- 1996b. "A New Data Set Measuring Income Inequal-
itv." WorldBank EconomicReview10(3): 565-91.
1998. "New Ways of Looking at Old Issues:Inequality
and Growth."Journal ofDevelopmentEconomics57: 259-87.
Delion,Jean. 1999. "ProducerOrganizations:Donor Partnerships
in Project Implementation." World Bank, Africa Region,
Washington, D.C.
Demery,Lionel.1999. "PovertyDynamicsin Africa:An Update."
World Bank,AfricaRegion,PovertyReductionand SocialDe-
velopment Unit, Washington, D.C.
Demery, Lionel, and Lyn Squire. 1996. "Macroeconomic Ad-
justment and Povertyin Africa:An EmergingPicture." World
Bank ResearchObserver11: 39-59.
Demirguic-Kunt,Asli, and Enrica Detragiache. 1998. "Financial
Liberalizationand FinancialFragility."PolicyResearchWork-
ing Paper 1917. World Bank, Washington, D.C.
Denmark, Ministry of Foreign Affairs. 2000. Impact Study of
Danish NGOs. Copenhagen.
Dercon, Stefan. 1996. "Risk, Crop Choice and Savings: Evi-
dence from Tanzania." EconomicDevelopmentand Cultural
Change44(3): 485-514.
. 1998. "Wealth, Risk, and Activity Choice: Cattle in
Western Tanzania." Journal of DevelopmentEconomics55:
1-42.
. 1999. "Income Risk, Coping Strategiesand SafetyNets."
Backgroundpaper for WorldlDevelopmentReport2000/2001.
KatholiekeUniversiteitLeuven;Oxford University,Centre for
the Study of African Economies;and World Bank, Washing-
ton, D.C.
Dercon, Stefan,and Pramila Krishnan. 1996. "Income Portfolios
in Rural Ethiopia and Tanzania: Choices and Constraints."
Journal of DevelopmentStudies32(6): 850-75.
. 1998. "The Urban Labour Marketduring StructuralAd-
justment: Ethiopia 1990-1997." Centre for the Study of
African Economies Working Paper Series, no. 98-9. Oxford
University.
---. 1999. "Vulnerability, Seasonality and Poverty in
Ethiopia." Oxford University,Centre for the Study of African
Economies.
. 2000a. "In Sicknessand in Health: Risk Sharingwithin
Households in Rural Ethiopia." Journal of PoliticalEconomy
108(4).
. 2000b. "Vulnerability, Seasonality and Poverty in
Ethiopia." Journal of DevelopmentStudies.
Dervis,Kemal,Jaimede Melo, and ShermanRobinson.1982. Gen-
eral Equilibrium Modelsfor DevelopmentPolicy.New York:
Cambridge UniversityPress.
Desai, Meghnad. 1991. "Human Development: Concepts and
Measurement." EuropeanEconomicReview 35: 350-57.
Deshpande, Ashwini.2000. "Does Caste Still Define Disparity?
A Lookat Inequality in Kerala,India." AmericanEconomicRe-
view 90: 322-25.
Dethier, Jean-Jacques. 1999. "Governance and Economic Per-
formance:A Survey."Discussion Paper on Development Pol-
icy 5. UniversitatBonn, Zentrum fur Entwicklungsforschung.
Devarajan, Shantayanan, and Jeffrey Hammer. 1997. "Public
Expendituresand Risk Reduction."World Bank,Development
ResearchGroup, Washington, D.C.
. 1998."Risk Reductionand Public Spending."PolicyRe-
searchWorking Paper 1869. World Bank, Development Re-
search Group, Washington, D.C.
Devarajan,Shantayanan, and Vinaya Swaroop. 1998. "The Im-
plications of ForeignAid Fungibility for Development Assis-
tance." Policy ResearchWorking Paper 2022. World Bank,
Washington, D.C.
Devarajan,Shantayanan,David Dollar,and Torgny E. Holmgren,
eds. 2000. Aid and Reform in Africa. Washington, D.C.:
World Bank.
Devarajan, Shantayanan, Lyn Squire, and Sethaput Suthiwart-
Narueput. 1997. "Beyond Rate of Return: Reorienting Pro-
ject Analysis." World Bank ResearchObserver12(1): 35-46.
Devarajan, Shantayanan, Sethaput Suthiwart-Narueput, and
KarenThierfelder.2000. "The Marginal Cost of Taxation in
DevelopingCountries." World Bank, Development Research
Group, Washington, D.C.
Devarajan,Shantayanan,DanyangXie, and Heng-fu Zou. 1998.
 BIBLIOGRAPHIC NOTE 233

"ShouldPublicCapitalBe Subsidizedor Provided?" Journal Developmentin Palanpur,1957-93."In PeterLanjouwarid

ofMonetaryEconomics 41(2):319-31. NicholasStern,eds.,Economic Development in Palanpurover
DFID (U.K.DepartmentforInternationalDevelopment).1999. FiveDecades. Oxford:Clarendon.
SustainableLivelihoods GuidanceSheets.London. Dube, ShyamaCharan. 1997."Ethnicity:Myth, Historyand
Diamond,Larry.1996. "Is the Third WaveOver?"Journalof Politics."In ArvindM. Shah,BaburaoS. Baviskar,and E.A.
Democracy 7(3). Ramaswamy, eds.,SocialStructureand Change.Vol.4. New
Diamond,Larry,JuanJ. Linz,and SeymourMartinLipset,eds. Delhi:Sage.
1988.Democracy in DevelopingCountries.
London:Adaman- Due,Jean,and ChristinaGladwin.1991."Impactsof Structural
tine. AdjustmentPrograms onAfricanWomenFarmersandFemale-
Diop, Francois,AbdoYazbeck,and RicardoBitran.1995."The HeadedHouseholds."AmericanJournalofAgricultural Eco-
Impactof AlternativeCostRecoverySchemeson Accessand nomics73(5):1431-39.
Equiryin Niger."HealthPolicyandPlanning10(3):223-40. Duflo,Esther.2000a."Child Healthand HouseholdResources
Dollar,David.2000."HasAidEfficiency Improvedinthe 1990s?" in SouthAfrica:Evidence fromtheOldAgePensionProgram."
World Bank, DevelopmentResearchGroup,Washington, AmericanEconomicReview90(2).
D.C. - . 2000b."Schooling and LaborMarketConsequences of
Dollar,David,and RobertaGatti. 1999."GenderInequality,In- SchoolConstructionin Indonesia:Evidencefrom an Un-
comeand Growth:AreGoodTimesGoodfor Women?"Pol- usualPolicyExperiment."Massachusetts Instituteof Tech-
icy ResearchReporton Genderand DevelopmentWorking nology,Departmentof Economics,Cambridge,Mass.
PaperSeries,no. 1. World Bank,Washington,D.C. Durlauf,Steven.1996."ATheoryof PersistentIncomeInequal-
Dollar,David,and AartKraay.2000. "GrowthIs Good for the iry."JournalofEconomic Growth1(1):75-93.
Poor."World Bank,DevelopmentResearchGroup,Wash- Duryea,Suzanne.1998."Children'sAdvancement throughSchool
ington,D.C. in Brazil:The Roleof TransitoryShocksto HouseholdIn-
Dollar,David,RaymondFisman,and RobertaGatti. 1999."Are come."Inter-AmericanDevelopmentBank, ResearchDe-
WomenReallythe 'Fairer'Sex?Corruptionand Womenin partment,Washington,D.C.
Government."PolicyResearchReporton Genderand De- Dushkin,Lelah.1972. "ScheduledCaste Politics."In Michael
velopmentWorkingPaperSeries,no. 4. WorldBank,Wash- Mahar,ed., TheUntouchables in Contemporary India.Tucson:
ington,D.C. UniversityofArizonaPress.
Dornbusch,Rudiger,and SebastianEdwards.1991.Macroeco- Dyson,Tim. 1996.Population andFood:GlobalTrendsandEu-
nomicsofPopulism.Chicago:Universityof ChicagoPress. tureProspects. NewYork:Routledge.
Dow, William,PaulGertler,RobertShoeni,John Strauss,and Easterly,William.1999a."The Ghost of the FinancingGap:
DuncanThomas. 1997. "Health Care Prices,Health and Testingthe GrowthModelusedin the InternationalFinan-
LaborOutcomes:Experimental Evidence."Rand,SantaMon- cial Institutions."Journalof Development Economics 60(2):
ica,Calif. 423-38.
Downing,Thomas.1996. "MitigatingSocialImpoverishment . 1999b."HappySocieties: The Middle-Class Consensus
WhenPeopleAre Involuntarily
Displaced."In Christopher
Mc- and EconomicDevelopment." PolicyResearch WorkingPaper
Dowell,ed., UnderstandingImpoverishment: TheConsequences 2346.World Bank,Washington.D.C.
of Development-Induced Displacement.Oxford: Berghahn . 1999c."HowDid HighlyIndebtedPoorCountriesBe-
Books. comeHighlyIndebted?Reviewing TwoDecadesof Debt Re-
Dreze,Jean. 1999."Militarism,Developmentand Democracy." lief." PolicyResearchWorking Paper2225. World Bank,
Lecturegivenat MaharajaSayajiraoUniversityof Baroda, Washington,D.C.
October,Baroda,India. . 2000a."CanInstitutions Resolve EthnicConflict?"
World
Dreze,Jean, and BhatiaBela.1998. "ForDemocracyand De- Bank,Development ResearchGroup,Washington,D.C.
velopment."DelhiSchoolof Economics,Centerof Develop- . 2000b."TheLostDecades... and the ComingBoom?
mentEconomics; andCambridgeUniversity, CentreofSouth Policies,Shocks,and DevelopingCountries' Stagnation,
AsianStudies. 1980-98."WorldBank,Development Research Group,Wash-
Dreze,Jean,and Haris Gazdar.1997."UttarPradesh:The Bur- ington,D.C.
den of Inertia."In JeanDrezeand AmartyaSen,eds.,Indian Easterly,William,and StanleyFischer.1999."Inflationand the
Development: SelectedRegional
Perspectives.
Oxford:OxfordUni- Poor."Paperpresentedat AnnualWorldBankConference on
versityPress. DevelopmentEconomics,28-30 April,Washington,D.C.
Dreze,Jean, and AmartyaSen. 1995.India:EconomicDevelop- . 2000. "Inflationand the Poor."PolicyResearch Work-
mentandSocialOpportunity. NewDelhi:OxfordUniversity ing Paper2335.World Bank,Washington,D.C.
Press. Easterly,William,andAartKraay.1999."SmallStates,SmallProb-
-. 1999."PublicActionand SocialSecurity:Foundations lems?"PolicyResearchWorkingPaper2139. World Bank,
and Strategy."In EhtishamAhmad,Jean Dreze,John Hills, Washington,D.C.
and AmartyaSen,eds.,SocialSecurityin DevelopingCountries. Easterly,William, and RossLevine. 1997. "Africa'sGrowth
NewYork:OxfordUniversityPress. Tragedy:Policiesand EthnicDivisions."Quarterly Journalof
Dreze,Jean,PeterLanjouw,andNareshSharma.1998."Economic Economics 62(11):1203-50.
234 WORLD DEVELOPMENT REPORT 2000/2001
- 2000. "It's Not Factor Accumulation: Stylized Facts
and Growth Models." World Bank, Development Research
Group, Washington, D.C.
Easterly,William,and SergioRebelo.1993. "FiscalPolicyand Eco-
nomic Growth: An Empirical Investigation."JournalofMIon-
eratyEconomics32: 417-58.
Easterly,William, Roumeen Islam, and Joseph E. Stiglitz. 1999.
"Shaken and Stirred: Explaining Growth Volatility." World
Bank, Development ResearchGroup, Washington, D.C.
Easterly, William, Norman Loayza, and Peter Montiel. 1997.
"Has Latin America's Post-Reform Growth Been Disap-
pointing?"Journal of InternationalEconomics43: 287-311.
Easterly, William, Carlos A. Rodriguez, and Klaus Schmidt-
Hebbel. 1994. Public SectorDeficitsand MacroeconomicPer-
formance. New York: Oxford UniversityPress.
Easterly,William, Michael Kremer,Lant Pritchett, and Lawrence
Summers. 1993. "Good Policy or Good Luck? Country
Growth Performance and Temporary Shocks." Journal of
MonetaryEconomics32: 459-83.
Eastwood, Robert, and Michael Lipton. 1998. "Demographic
Transition and Poverty: Effectsvia Economic Growth, Dis-
tribution, Conversion." Sussex University, Department of
Economics,Brighton.
1999. "The Impacts of Changes in Human Fertility on
Poverty."JournalofDevelopmentStudies36(1): 1-30.
EBRD (European Bank for Reconstruction and Development).
1998. TransitionReport 1998: The FinancialSectorin Tran-
sition. London.
- 1999. TransitionReport1999: Ten Yearsof Transition.
London.
Eccles,Stephen, and Catherine Gwin. 1999. SupportingEfective
Aid: A Frameworkfor Future Concessional Funding of Multi-
lateralDevelopmentBanks. PolicyEssay23. Washington,D.C.:
Overseas Development Council.
Echeverri-Gent, John. 1988. "Guaranteed Employment in an
Indian State: The Maharashtra Experience." Asian Survey
(28): 1294-310.
ECLAC (Economic Commission for Lacin America and the
Caribbean). 1989. "Antecedentes estadisticosde la distribu-
cion del ingreso en el Peru, 1961-1982." Serie distribucion
del ingreso 8. Santiago.
- 1991. "El perfil de la probeza en America Latina a
comienzos de los afios 90." Santiago,Chile.
1993. "Antecedentesestadisticosde Ladistribucionde in-
greso en los aios 80." Serie distribucion del ingreso 13.
Santiago.
- 1995. Human Settlements:The Shelter of Development.
Santiago.
1997a.SocialPanoramaofLatin America1996. Santiago.
1997b. StatisticalYearbookfor Latin America and the
Caribbean.Santiago.
1999a. EconomicIndicators.Santiago.
1999b.SocialPanoramaofLatin America1998. Santiago.
Economic and Political Weekly Research Foundation. 1998.
"Marathwada Earthquake: Efforts at Participatory Rehabili-
tation and the Role of Community Participation Consul-
tants." Mumbai.
Edwards, Sebastian. 1994. "Trade and Industrial Policy Reform
in Latin America." NBER Working Paper 4772. National
Bureau of Economic Research,Cambridge, Mass.
. 1999. "How EffectiveAre Capital Controls?" NBER
Working Paper 7413. National Bureau of Economic Re-
search, Cambridge, Mass.
Ehrenpreis, Dag. 1999. "Development Cooperation in Support
of PoverryReduction." Paper presented at researchworkshop
on poverty, World Bank, 6-8 July, Washington, D.C.
Eichengreen,Barry. 1999. TowardaNew InternationalFinancial
Architecture:A PracticalPost-Asia Agenda.Washington, D.C.:
Institute for International Economics.
EIU (Economist IntelligenceUnit). 1999a.BusinessLatinAmer-
ica. London.
. 1999b. Country Reporton Indonesia.London.
Ekbom, Anders, and Jan Bojo. 1999. "Poverty and Environment:
Evidenceof Linksand Integration into the CountryAssistance
Strategy Process." Environment Group Discussion Paper 4.
World Bank, AfricaRegion, Washington, D.C.
Ellis, Frank. 1999. "Rural Livelihood Diversity in Developing
Countries: Analysis,Policy,Methods." U.K. Department for
International Development, London.
Elwan, Ann. 1999. "Poverty and Disability." Background paper
for WorldDevelopment Report2000/2001.World Bank, Wash-
ington, D.C.
English, Philip. 1998. "Mauritius: Reigniting the Engines of
Growth-A Teaching Case Study." World Bank, Economic
Development Institute, Washington, D.C.
Ernberg,Johan. 1998. "UniversalAccessfor Rural Development:
From Action to Strategies."Paper presented at International
Telecommunication Union Seminaron Multipurpose Com-
munitv Telecentres,7-9 December, Budapest.
Escobal,Javier,Jaime Saavedra,and Maximo Torero. 1998. "Los
activos de los pobres en el Peru." Documento de trabajo 26.
Grupo de Analisispara el Desarrollo, Lima.
Esman, Milton, and Norman Uphoff. 1984. Local Organiza-
tions:Intermediariesin RuralDevelopment.Ithaca, N.Y.: Cor-
nell University Press.
Esteban,Joan-Maria, and Debraj Ray. 1994. "On the Measure-
ment of Polarization." Econometrica62(4): 819-51.
Estes,Richard. 1988. Trendsin WorldSocialDevelopment:The So-
cial ProgressofNations, 1970-1987. New York: Praeger.
. 1996. "SocialDevelopment Trends in Asia, 1970-1994:
The Challenge of a New Century." SocialIndicatorsResearch
37(2).
. 1998. "Trends in World Social Development,
1970-1995: Development Challenges for a New Century."
Journalof DevelopingSocieties14(1).
Evans,Peter. 1996. "Government Action, SocialCapital and De-
velopment: Reviewingthe Evidence on Synergy." WorldDe-
velopment24(6): 1119-32.
. 1999. "Transferable Lessons?Reexaminingthe Institu-
tional Prerequisitesof East Asian Economic Policies."In Yil-
maz Akyuz, ed., East Asian Development:New Perspectives.
London: Frank Cass.
Evans,Peter,and James Rauch. 1999. "Bureaucracyand Growth:
A Cross-National Analysisof the Effects of 'Weberian' State

Structureson EconomicGrowth."AmericanSociological
Re-
view64(5):748-65.
Evenson,Robert.2000. "CropGeneticImprovementand Agri-
cultural Development." ConsultativeGroup on International
AgriculturalResearch,Washington, D.C.
Fafchamps,Marcel, and Bart Minten. 1999. "SocialCapital and
the Firm: Evidencefrom Agricultural Trade." SocialCapital
InitiativeWorkingPaper 17.World Bank, SocialDevelopment
Department, Washington, D.C.
Fafchamps, Marcel, Christopher Udry, and Katherine Czukas.
1998. "Drought and Saving in West Africa: Are Livestock
a Buffer Stock?" Journal of Development Economics 55:
273-305.
Fajnzylber,Pablo, and Daniel Lederman. 1999. "Economic Re-
forms and Total Factor Productivity Growth in Latin Amer-
ica and the Caribbean (1950-1995): An Empirical Note."
Policy ResearchWorking Paper 2114. World Bank, Wash-
ington, D.C.
Fajnzylber,Pablo,Daniel Lederman,and Norman Loayza.1998.
Determinantsof Crime Ratesin Latin America and the World:
An EmpiricalAssessment.Washington, D.C.: World Bank.
- 1999. IncomeInequalityand ViolentCrime.Washington,
D.C.: World Bank.
Falkingham, Jane. 1999. "Welfare in Transition: Trends in
Poverty and Well-Being in Central Asia." Centre for the
Analysisof Social ExclusionPaper 20. London School of Eco-
nomics and Political Science.
Fallon,Peter,and ZafirisTzannatos. 1998. ChildLabor:Issuesand
Directionsfor the World Bank. Washington, D.C.: World
Bank.
FAO (Food and Agriculture Organization of the United Na-
tions). 1995. The Effectsof HIV/AIDS on FarmingSystemsin
EasternAfrica. Rome.
Feder, Gershon, Tongroj Onchan, Yangyuth Chalamwong, and
Chica Hongladarom. 1988. Land Policiesand FarmProduc-
tivity in Thailand.Baltimore,Md.: Johns Hopkins University
Press.
Fei, John C.H., Gustav Ranis, and Shirley W.Y. Kuo. 1979.
Growthwith Equity:The TaiwanCase.NewYork:Oxford Uni-
versity Press.
Fernandez, Aloysius. 1994. The Myrada Experience:Alternative
ManagementSystemsJbrSavingsand Creditfbr theRuralPoor.
Bangalore:Myrada.
Fernandez-Arias,Eduardo, and Peter Montiel. 1997. Reformand
GrowthinLatinAmerica:ALlPain, No Gain?Washington,D.C.:
Inter-American Development Bank.
Ferreira, Francisco,and Julia A. Litchfield.1998. "Education or
Inflation? The Roles of Structural Factors and Macroeco-
nomic Instability in Explaining Brazilian Inequality in the
1980s." STICERD Discussion Paper 41. London School of
Economics and Political Science, Suntory and Toyota Inter-
national Centres for Economics and Related Disciplines.
Ferreira, Francisco,and Ricardo Paes de Barros. 1999a. "Climb-
ing a Moving Mountain: Explainingthe DecliningIncome In-
equality in Brazil from 1976 to 1996." Inter-American
Development Bank, Washington, D.C.
- 1999b. "The Slippery Slope: Explainingthe Increasein
BIBLIOGRAPHIC NOTE 235
ExtremePovertyin UrbanBrazil,1976-96."PolicyResearch
WorkingPaper2210. WorldBank,Washington,D.C.
Ferreira,Francisco,PeterLanjouw,and MarceloNeri.2000. "A
New PovertyProfile for BrazilUsing PPV, PNAD and Cen-
sus Data." Texto para discussao418. PontificiaUniversidade
Cat6lica do Rio de Janeiro, Departamento de Economia.
Feyisetan,Bamikale,and Martha Ainsworth. 1996. "Contracep-
tive Useand the Quality, Price.and Availabilityof FamilyPlan-
ning in Nigeria." World Bank EconomicReview 10: 159-87.
Feyzioglu,Tarhan,VinayaSwaroop,and Min Zhu. 1998. "APanel
Data Analysisofthe FungibilityofForeign Aid." WorldBank
EconomicReview12(1): 29-58.
Fields,Gary S. 1987. "Measuring Inequality Change in an Econ-
omy with Income Growth." Journalof DevelopmentEconom-
ics26: 357-74.
. 1991. "Growth and Income Distribution." In Geoige
Psacharopoulos,ed.,EssaysonPoverty,Equity,and Growth.New
York: Pergamon.
. 1999. "Distribution and Development: A Summary of
the Evidencefor the DevelopingWorld." Background paper
for WorldDevelopmentReport2000/2001.CornellUniversity,
Ithaca, N.Y.; and World Bank, Washington, D.C.
Filmer, Deon. 1999a. EducationalAttainment and Enrollment
Profiles:AResourceBook Basedon an AnalysisofDemographic
and Health SurveysData. Washington, D.C.: World Bank.
. 1999b. "The Structure of Social Disadvantagein Eclu-
cation: Gender and Wealth." PolicyResearchReport on Gen-
der and Development Working Paper Series, no. 5. World
Bank, Washington, D.C.
Filmer, Deon, and Lant Pritchett. 1998. EducationalEnrollment
andAttainmentin India:HouseholdWealth,Gender,Villageand
StateEffects.Washington, D.C.: World Bank.
. 1999a."The Effectof HouseholdWealth on Educational
Attainment:Evidencefrom 35 Countries."Populationand De-
velopmentReview25(1): 85-120.
. 1999b. "The Impact of Public Spending on Health:
Does Money Matter?"SocialScienceand Medicine49: 1309-23.
Filmer,Deon,JefferyHammer,andLantPritchett. 1998."Health
Policyin PoorCountries:WeakLinks in the Chain." PolicyRe-
searchWorking Paper 1874. World Bank,Washington, D.C.
Filmer, Deon, ElizabethKing, and Lant Pritchett. 1999. Gender
Disparityin SouthAsia:Comparisons betweenand within Coufn-
tries.Washington, D.C.: World Bank.
Filmer,Deon, Haneen Sayed,Boediono,Jiyono,Nanik Suwaryani,
and Bambang Indriyanto. 1998. "The Impact of Indonesia's
Economic Crisis on BasicEducation: Findings from a Survey
of Schools." World Bank, Development Research Group,
Washington, D.C.
Fine, Ben. 1999. "The DevelopmentalState Is Dead: Long Live
Social Capital?" Developmentand Change30: 1-19.
Fiscella,Kevin, and Peter Franks. 1997. "Poverty or Income In-
equalityas Predictorof Mortality:LongitudinalCohort Study."
British MedicalJournal314(7096): 1724-27.
Fisman, Raymond, and Roberta Gatti. 1999. "Decentralization
and Corruption: Cross-Country and Cross-State Evidence."
World Bank, Development Research Group, Washington,
D.C.; and Columbia BusinessSchool, New York.
236 WORLD DEVELOPMENT REPORT 2000/200I
Fiszbein,
Ariel,GeorgePsacharopoulos, SamuelMorley,Haeduck
Lee,and BillWood. 1993."La pobrezay la distribucionde
losingresosen AmericaLatina:Historiadeldeceniode 1980."
LatinAmericaand theCaribbeanTechnicalDepartmentRe-
port 27. WorldBank,Washington,D.C.
Flanagan,Kerry.1999."Agingand Poverty:A Policyfor the El-
derlv Poor.' World Bank, SocialProtectionDepartment,
Washington,D.C.
Flegg,A. 1982."InequalityofIncome,Illiteracy,
andMedicalCare
asDeterminants of InfantMortalityin Developing Countries."
PopulationStudies36(3):441-58.
Foley,Gerald.1997."RuralElectrification in CostaRica:A Case
Study."Joint UnitedNationsDevelopmentProgrammeand
World Bank EnergySector ManagementAssistancePro-
gramme,Washington,D.C.
Foley,Michael,and Bob Edwards.1999."Is It Time to Disin-
vestin SocialCapital?"JournalofPublicPolicy19: 141-73.
Foner,Eric.1989.Reconstruction:
Americas Unfinished Revolution,
1863-1877.NewYork:Harperand Row.
Forbes,Kristin.Forthcoming."AReassessment of the Relation-
ship betweenInequalityand Growth."AmericanEconomic
Review.
Foster, Andrew. 1995. "Prices, Credit Markets and Child
Growth in Low-Income Rural Areas." EconomicJournal
105: 551-70.
Foster, Andrew, and Mark Rosenzweig. 1995. "Learning by
Doing and Learningfrom Others: Human Capital and Tech-
nical Change in Agriculture." Journal of Political Economy
103(6): 1176-209.
Foster, James E., and Anthony Shorrocks. 1988. "Poverty Or-
derings." Econometrica56(1): 173-77.
Foster,James E., Joel Greer,and Erik Thorbecke. 1984. "A Class
of Decomposable Poverty Measures." Econometrica52(3):
761-66.
Foster, Mick. 1999. Lessonsof Experiencefrom Sector-Wide Ap-
proachesin Health. Geneva:World Health Organization.
Fox, Jonathan. 1992. "Democratic Rural Development:Leader-
ship Accountabilityin Regional Peasant Organizations." De-
velopmentand Change23(2): 1-36.
- 1995. "Governanceand RuralDevelopment in Mexico:
StateIntervention and PublicAccountability."JournalofDe-
velopmentStudies32: 1-30.
1996. "How Does Civil SocietyThicken? The Political
Construction of SocialCapital in Rural Mexico." World De-
velopment24(6): 1089-103.
Fox, Jonathan, and John Gershman. 1999. "Investing in Social
Capital?Comparative Lessons from Ten World Bank Rural
Development Projects in Mexico and the Philippines." Uni-
versity of California at Santa Cruz, Department of Latin
American and Latino Studies.
Frankel, Jeffrey, and David Romer. 1999. "Does Trade Cause
Growth?"AmericanEconomicReview 89: 379-99.
Frankenberg,Elizabeth, Duncan Thomas, and Kathleen Beegle.
1999. "The Real Costs of Indonesia's Economic Crisis: Pre-
liminary Findings from the Indonesia Family Life Surveys."
Labor and PopulationWorking Paper Series,no. 99-04. Rand,
Santa Monica, Calif.
Freeman,Paul. 1999. "The IndivisibleFaceof Disaster."In
World Bank,Investingin Prevention:A SpecialReportonDis-
asterRiskManagement. Washington,D.C.
Freeman,Paul,andLandisMacKellar. 1999a."EconomicImpacts
of Natural Disasters."Box for WorldDevelopment Report
2000/2001.World Bank,Washington,D.C.
. 1999b."Impactof NaturalDisasterson Infrastructure."
Boxfor WorldDevelopment Report2000/2001.WorldBank,
Washington,D.C.
Freeman,Richard.1995."AreYourWagesSetin Beijing?"Jour-
nalofEconomic Perspectives
9(3): 15-32.
Frigenti,Laura,and AlbertoHarth,withRumanaHuque.1998.
"LocalSolutionsto RegionalProblems:The Growthof So-
cialFundsand PublicWorksandEmployment Projectsin Sub-
SaharanAfrica."World Bank, AfricaRegion,Water and
Urban2 DivisionandInstitutionalandSocialPolicyDivision,
Washington,D.C.
Fukuyama,Francis.1993. TheEnd ofHistoryandtheLastMan.
NewYork:AvonBooks.
. 1995.Trust:TheSocial1'irtues
andtheCreationofPros-
perity.NewYork:FreePress.
Funaro-Curtis,Rita. 1982. NaturalDisastersand the Development
Process:A Discussion of Issues.Arlington, Va.: Evaluation
Technologies.
Gaiha, Raghav,and Anil Deolalikar. 1993. "Persistent, Expected
and Innate Poverty: Estimates for Semi-arid Rural South
India, 1975-1984." CambridgeJournal of Economics17(4):
409-21.
Galanter, Marc. 1972. "The Abolition of Disabilities:Untouch-
ability and the Law." In Michael Mahar, ed., The Untouch-
ables in ContemporaryIndia. Tucson: Universitv of Arizona
Press.
Galasso,Emanuela,and Martin Ravallion.2000. "Distributional
Outcomes of a DecentralizedWelfare Program." Policy Re-
searchWorking Paper 2316. World Bank, Washington, D.C.
Galeotti,Marzio,and AlessandroLanza. 1999. "DesperatelySeek-
ing (Environmental) Kuznets." Working paper. Fondazione
Eni Enrico Mattei, Milan.
Gallup, John Luke. 1997. "Ethnicity and Earnings in Malaysia."
Discussion Paper 593. Harvard Institute for International
Development, Cambridge, Mass.
Gallup,John Luke, Steven Radelet, and Andrew Warner. 1998.
"Economic Growth and the Incomeof the Poor." Harvard In-
stitute for International Development,Cambridge, Mass.
Gallup, John Luke,Jeffrey Sachs,and Andrew Mellinger. 1999.
"Geography and Economic Development." InternationalRe-
gional ScienceReview22(2): 179-232.
Gardner, L. Bruce. 1995. "Policy Reform in Agriculture:An As-
sessmentofthe ResultsofEight Countries."UniversityofMary-
land, Department of Agricultural and Resource Economics,
College Park.
Garro, Alejandro. 1999. "Accessto Justice for the Poor in Latin
America." In Juan E. Mendez, Guillermo O'Donnell, and
Paulo SergioPinheiro, eds., The (Un)Rule ofLaw and the Un-
derprivilegedin Latin America.Notre Dame, Ind.: University
of Notre Dame.
Gatti, Roberta. 1999. "Corruption and Trade Tariffs, or a Case

for Uniform Tariffs." PolicyResearchWorking Paper 2216.
World Bank, Washington, D.C.
Gaynor, Cathy. 1996. "Decentralizationof Primary Education:
Implicationsat School and Community Level-The Case of
Nigeria and Tanzania." World Bank, Economic Develop-
ment Institute, Washington, D.C.
.1998. DecentralizationofEducation:TeacherManagement.
Directions in Development Series.Washington, D.C.: World
Bank.
Gereffi, Gary, and Martha ArgeliaMartinez. 1999. "Blue Jeans
and Local Linkages:The BlueJeans Boom in Torre6n, Mex-
ico." Background paper for World Development Report
2000/2001. World Bank, Washington, D.C.
German, Tony, and Judith Randel. 1998. "Targeting the End of
Absolute Poverty:Trends in Development Cooperation." In
Judith Randel and Tony German, eds., The Reality of Aid,
1998/1999: An IndependentReviewof PovertyReductionand
DevelopmentAssistance.London: Earthscan.
Gertler, Paul. 2000. "Insuring the Economic Costsof Illness." In
Nora Lustig,ed., Shieldingthe Poor:SocialProtectionin theDe-
velopingWorld.Washington, D.C.: BrookingsInstitution and
Inter-AmericanDevelopment Bank.
Gertler, Paul, and Paul Glewwe. 1989. The Willingnessto Payfor
Educationin DevelopingCountries:Evidencefrom Rural Peru.
Living Standards Measurement Study Working Paper 54.
Washington, D.C.: World Bank.
Gertler, Paul, and Jonathan Gruber. 1997. "Insuring Consump-
nion against Illness." NBER Working Paper 6035. National
Bureau of Economic Research,Cambridge, Mass.
Gerder,Paul, andJeffreyHammer. 1997. "Policiesfor PricingPub-
licly Provided Health Services." Policy Research Working
Paper 1762. World Bank, Washington, D.C.
Gertler, Paul, and Orville Solon. 1998. "Who Benefitsfrom So-
cialHealth Insurance in Low-IncomeCountries?" University
of California at Berkeley,Haas School of Business.
Gertler, Paul, and Jacques van der Gaag. 1990. The Willingness
to Payfor MedicalCare:Evidencefrom Two DevelopingCoun-
tries.Baltimore, Md.: Johns Hopkins University Press.
Ghana StatisticalService.1999. "PovertyTrends in Ghana in the
1990s."Report preparedby governmentof Ghana for the tenth
consultativegroup meeting, 23-24 November, Accra.
Ghatak, Maitreesh.1999."Group Lending,LocalInformationand
Peer Selection." Journal of Development Economics60(1):
27-50.
Gibbs, Christopher, Claudia Fumo, and Thomas Kuby. 1999.
Nongovernmental Organizationsin World Bank-Supported
Projects.A Review.Washington, D.C.: World Bank.
Gilbert, Alan. 1998a. "A Home Is for Ever?Residential Mobil-
ity and Home Ownership in Self-Help Settlements." Envi-
ronmentand Planning31(6): 1073-91.
- . 1998b. "Market Efficiencyand the SecondaryHousing
Market in Third World Cities."Report for U.K. Department
for International Development. London.
Gilbert, Alan, Oscar 0. Camacho, Rene Coulomb, and Andres
Necochea. 1993. In Searchof a Home: Rental and Shared
Housing in Latin America. Tucson: University of Arizona
Press.
BIBLIOGRAPHIC NOTE 237
Gilson, Lucy. 1998. "The Lessons of User Fee Experience in
Africa."In AlisonBeattie,Jan Doherty, Lucy Gilson, Eyitayo
Lambo, and Paul Shaw, eds., SustainableHealth Care Fi-
nancingin SouthernAfrica: Papersfrom an EDI Health Polity
Seminarheldin Johannesburg,SouthAfrica,June 1996 Wash-
ington, D.C.: World Bank.
Girishankar, Navin. 1999a. "Reforming Institutions for Service
Delivery."PolicyResearchWorking Paper 2039. World Bank,
Washington,D.C.
. 1999b. "Securingthe PublicInterestunder PluralisticIn-
stitutional Design."World Bank,Operations EvaluationDe-
partment, Washington, D.C.
Girishankar, Navin, and Nicholas P. Manning. 1999. "Toolkit
for AssessingConstraints on Frontline Delivery in Decen-
tralized Settings."World Bank, Poverty Reduction and Eco-
nomic Management Network, Washington, D.C.
Gittell, Ross. and Avis Vidal. 1998. Community Organizing:
BuildingSocialCapitalasa DevelopmentStrategv. NewburyPark,
Calif.: Sage.
Glaeser, Edward, David Laibson, and Bruce Sacerdote. 2000.
"The Economic Approach to Social Capital." NBER Work-
ing Paper 7728. National Bureau of Economic Research,
Cambridge, Mass.
Glewwe,Paul. 1991. "Investigating the Determinants of House-
hold Welfare in C6te d'Ivoire." Journalof DevelopmentEco-
nomics35: 307-37.
Glewwe,Paul,and Dennis de Tray. 1989. The Poorin Latin Amer-
icaduringAdjustment:A CaseStudy of Peru.LivingStandards
Measurement Study Working Paper 56. Washington, D.C(.:
World Bank.
Glewvwe, Paul, and Gilette Hall. 1998. "Are Some Groups Mere
Vulnerable to Macroeconomic Shocks than Others? Hy-
pothesisTests Basedon Panel Data from Peru."JournalofDe-
velopmentEconomics56: 181-206.
Glewwe, Paul, Michele Gragnolati, and Hassan Zaman. 20C00.
"Who Gained from Vietnam's Boom in the 1990s?An Analy-
sis of Povertyand Inequality Trends." PolicyResearchWork-
ing Paper 2275. World Bank. Washington, D.C.
Gokcekus.Omer, RanjanaMukherjee,and Nick Manning. 2000.
"Institutional Arrangements Affect Performance." World
Bank, Poverty Reduction and Economic Management Net-
work, Washington, D.C.
Goldstein,Anne. 1999."Thinking Outside Pandora's Box."Back-
ground paper for forthcomingWorld Bank PolicyResearchRe-
port EnGenderingDevelopment. World Bank,Washington,D.C.
Goldstein, Morris, Graciela Kaminsky, and Carmen Reinhart.
2000. AssessingFinancial Vulnerability:An EarlyWarningS5ys-
temforEmergingMarkets.Washington,D.C.: Institute for In-
ternational Economics.
Gonzalesde Olarte, Efrain,and PilarGavilanoLlosa.1999. "Does
PovertyCauseDomesticViolence?SomeAnswersfrom Lima."
In Andrew Morrison and Maria Loreto Biehl, eds., Too Close
to Home: Domestic Violence in the Americas. Washington,
D.C.: Inter-American Development Bank.
Goodin, Robert E., Bruce Headey,Ruud Muffels,and Henk-Jan
Dirven. 1999. The Real Worldsof WelfareCapitalism.Cam-
bridge: Cambridge University Press.
238 WORLD DEVELOPMENT REPORT 20002/OOI
Goody,Jack. 1976. Productionand Reproduction:A Comparative
Study of the DomesticDomain. Cambridge: Cambridge Uni-
versity Press.
- 1990. The Oriental,the Ancient and the Primitive: Sys-
temsofMarriage and the Family in the Pre-industrialSocieties
of Eurasia.Cambridge: Cambridge University Press.
Gopal, Gita. 1999. Gender-RelatedLegal Reform and Accessto
EconomicResources in EasternAfrica.World Bank Discussion
Paper 405. Washington, D.C.
Gragnolati, Michele. 1999. "Children's Growth and Poverty in
Rural Guatemala." Policy Research Working Paper 2193.
World Bank, Washington, D.C.
Graham, Carol. 1992. "Politicsof Protectingthe Poor during Ad-
justment: Bolivia'sEmergencySocial Fund." WorldDevelop-
ment20: 1233-51.
.1994. SafetyNets,Politics,
and thePoor:Transitionsto Mar-
ketEconomies.Washington, D.C.: Brookings Institution.
1996. GenderIssuesin PovertyAlleviation:RecentExpe-
rienceswith Demand-BasedProgramsin LatinAmerica,Africa
and EasternEurope.Developmentand Technical Cooperation
Department. Geneva:International Labour Office.
Grandin, Barbara, and P. Lembuya. 1992. "AmboseliNational
Park, the SurroundingGroup Ranchesand Community Con-
servation."African Wildlife Foundation, Nairobi.
Grandvoinnet, Helene. 2000. "Rule of Lawand Povertv Reduc-
tion: SomeIssues."In AsbjornKjonstadand PeterRobson,eds.,
Povertyand Law. London: Hart.
Grandvoinnet,Helene,and PaulineTamesis.2000. "FightingCor-
ruption: The Case of the Philippines." In Helene Grand-
voinnet, Irene Hors, and Pauline Tamesis, eds., Fighting
Corruption:ComparativeCountry CaseStudiesin Five Devel-
opingCountries.Paris:Organisationfor EconomicCo-operation
and Development, Development Center; and United Na-
tions Development Programme.
Granovetter,Mark. 1994."BusinessGroups." In Neil Smelserand
Richard Swedberg,eds., TheHandbookofEconomicSociology.
Princeton, N.J.: Princeton UniversityPress.
Gray, Cheryl,and Daniel Kaufmann. 1998. "Corruption and De-
velopment." Financeand Development35(3): 7-10.
Gray, Leslie,and Michael Kevane. 1996. Land TenureStatus of
African Women.Washington, D.C.: World Bank.
Gray, William, Christopher Landsea,Paul Mielke,and Kenneth
Berry.1999."Summaryof 1999AtlanticTropical CycloneAc-
tivity and Verification of Author's SeasonalActivity Predic-
tion." ColoradoStateUniversity,AtmosphericScienceFaculty,
Fort Collins.
Greenland, David J. 1997. "International Agricultural Research
and the CGIAR System: Past, Present, and Future."Journal
of InternationalDevelopment9: 449-58.
Greenwald,Bruce C., and Joseph E. Stiglitz. 1990. "Asymmet-
ric Information and the New Theory of the Firm: Financial
Constraintsand Risk Behavior."NBER Working Paper 3359.
National Bureau of Economic Research,Cambridge, Mass.
Greif, Avner. 1994. "Cultural Beliefsand the Organization of So-
ciety:A Historical and Theoretical Reflection on Collectivist
and Individualist Societies."Journalof PoliticalEconomy102:
912-50.
Grimard, Franque. 1997. "Household Consumption Smoothing
through Ethnic Ties: Evidencefrom Cote d'Ivoire."Journal
ofDevelopmentEconomics53: 391-422.
Grindle, Merilee. 1996. ChallengingtheStare:Crisisand Innova-
tion in Latin AmericaandAfrica. New York:Cambridge Uni-
versit' Press.
. 1997. GettingGoodGovernment:CapacityBuildingin the
PublicSectorsofDevelopingCountries.Cambridge,Mass.:Har-
vard University Press.
Grootaert, Christiaan. 1997. "SocialCapital: The MissingLink?"
In World Bank, Expandingthe Measureof Wealth:Indicators
of EnvironmentallySustainahLeDevelopment.Washington,
D.C.
. 1998. "Child Labor in Cote d'lvoire: Incidence and
Determinants." PolicyResearchWorking Paper 1905. World
Bank,Washington, D.C.
. 1999a. "Does Social Capital Help the Poor? A Synthe-
sis of Findingsfrom the LocalLevelInstitutions Studies in Bo-
livia, Burkina Faso and Indonesia." World Bank, Social
Development Department, Washington, D.C.
. 1999b. "SocialCapital, Household Welfareand Poverty
in Indonesia." Local Level Institutions Working Paper 6.
World Bank, SocialDevelopment Department, Washington,
D.C.
Grootaert, Christiaan, and Jeanine Braithwaite. 1998. "Poverty
Correlates and Indicator-Based Targeting in Eastern Europe
and the Former Sovier Union." Policy Research Working
Paper 1942. World Bank,Washington,D.C.
Grootaert, Christiaan, and Ravi Kanbur. 1995a. "Child Labor:A
Review."PolicyResearchWorking Paper 1454. World Bank,
Washington, D.C.
. 1995b. "The Lucky Few amidst Economic Decline:
Distributional Change in C6te d'lvoire as Seen through
Panel Data Sets." Journal of Development Studies 31 (4):
603-19.
Grootaert, Christiaan, and Deepa Narayan. 2000. "Local Insti-
tutions, Povertyand HouseholdWelfarein Bolivia."LocalLevel
Institutions Working Paper 9. World Bank, Social Develop-
ment Department, Washington, D.C.
Grootaert, Christiaan, and Harry Anthony Patrinos, eds. 1999.
ThePolicyAnalysisofChildLa6bor:AComparativeStudy. New
York: St. Martin's.
Grootaert, Christiaan, Ravi Kanbur, and Gi-Taik Oh. 1997.
"The Dynamics of Welfare Gains and Losses:An African
Case Study."Journal of DevelopmentStudies33(5): 635-57.
Grootaert,Christiaan,Gi-Taik Oh, and Anand Swamy.1999."So-
cial Capital and Development Outcomes in Burkina Faso."
Local LevelInstitutions Working Paper 7. World Bank, So-
cial Development Department, Washington, D.C.
Grosh, Margaret. 1994. AdministeringTargetedSocial Programs
in LatinAmerica:FromPlatitudestoPractice.Regionaland Sec-
toral Studies Series.Washington, D.C.: World Bank.
Grosh, Margaret, and Paul Glewwe,eds. 2000. DesigningHouse-
hold Survey Questionnairesfor DevelopingCountries:Lessons
from 15 Yearsof the Living Standards Measurement Study.
Washington, D.C.: World Bank.
Guillaumont, Patrick, and Lisa Chauvet. 1999. Aid and Perfor-

mance:A Reassessment.Clermant-Ferrand: Centre d'Etudes
Recherches
le DevelopmentInternational.
Guillaumont,Patrick,SylvianeGuillaumontJeanneney,and
Jean-FrancoisBrun. 1999."HowInstabilityLowersAfrican
Growth."JournalofAfrican Economies 8(1): 87-107.
Gupta,Dipankar,ed. 1991.SocialStratification.
NewDelhi:Ox-
ford UniversityPress.
Gupta,Sanjeev,HamidDavoodi,andRosaAlonso-Terme. 1998.
"DoesCorruptionAffectIncomeInequalityandPoverty?" IMF
WorkingPaper98/76. InternationalMonetaryFund,Wash-
ington,D.C.
Gupta, Sanjeev,Jerald Schiff,and BenedictClements.1996.
"WorldwideMilitarySpending,1990-95."IMF Working
Paper96164.InternationalMonetaryFund, Washington,
D.C.
.1998. "Public Spending on Human Development."Fi-
nanceandDevelopment35(3): 10-13.
Gupta, Sanjeev,CalvinMcDonald,ChristianSchiler,MarijinVer-
hoeven, Zelco Bogetic,and G. Schwartz. 1998. "Mitigating
the Social Cost of the Economic Crisis and the Reform Pro-
grams in Asia."Paper on PolicyAnalysisand Assessment98/7.
International Monetary Fund, Washington, D.C.
Gurr, Ted Robert, BarbaraHarff, Monty G. Marshall, and James
R. Scarritt. 1993. Minorities at Risk:A Global View of Eth-
nopoliticalConflicts.Washington, D.C.: Institute of Peace
Press.
Gwatkin, Davidson R., Michel Guillot, and Patrick Heuveline.
2000. The Burden of Diseaseamong the GlobalPoor:Current
Situation,Future Trends,and ImplicationsforStrategy.Wash-
ington, D.C.: World Bank.
Gwatkin, Davidson R., Shea Rutstein, KierstenJohnson, Rohini
Pande, and Adam Wagstaff. 2000. SocioeconomicDiferences
in Health,NutritionandPopulation.Washington,D.C.: World
Bank. [www.worldbank.org/poverty/health/data/index.htm].
Haan, Hans Christian, Adrian Coad, and Inge Lardinois. 1998.
MunicipalSolid WasteManagement:InvolvingMicroand Small
Enterprises.Turin: International Labour Office.
Haddad, Lawrence, and Ravi Kanbur. 1990. "Are Better-off
Households More Unequal or Less Unequal?" Policy Re-
searchand ExternalAffairsWorking Paper 373. World Bank,
Washington, D.C.
Haddad, Lawrence, and Thomas Reardon. 1993. "Gender Bias
in the Allocationof Resourceswithin Households in Burkina
Faso: A DisaggregatedOutlay Equivalent Analysis."Journal
ofDevelopmentStudies29(2): 260-76.
Haddad, Lawrence, Marie J. Ruel, and James L. Garrett. 1999.
"Are Urban Poverty and Undernutrition Growing? Some
Newly AssembledEvidence." Food Consumption and Nu-
trition Division DiscussionPaper 63. International Food Pol-
icy ResearchInstitute, Washington, D.C.
Haggard, Stephan. 1998. "Institutions and Growth in Koreaand
Taiwan: The Bureaucracy."JournalofDevelopmentStudies34:
87-111.
Haggard, Stephan, and Steven B. Webb. 1993. "What Do We
Know about the Political Economy of Economic Policy Re-
form?" WorldBank ResearchObserver8(2): 143-68.
Haggard,Stephan,ElizaWillis,and Christopherda C.B. Garman.
BIBLIOGRAPHIC NOTE 239
1999."Politicsof Decentralization
in LatinAmerica."Latin
AmericanResearch Review34(1):7-56.
Hallberg,Kristin.1999.Smalland Medium-Scale Enterprises:
A
Frameworkfor Intervention.
Washington,D.C.:World Bank.
Haltiwanger,John, and ManishaSingh.1999."Cross-Countxy
Evidenceon PublicSectorRetrenchment."WorldBankEco-
nomicReview13(1):23-66.
Haltiwanger,John,and MilanVodopivec.1999."GrossWorker
and Job Flowsin a TransitionEconomy:An Analysisof Es-
tonia." PolicyResearchWorkingPaper2082.World Bank,
Washington,D.C.
Hammer,Jeffrey.1997. "EconomicAnalysisfor Health Pro-
jects."WorldBankResearch Observer12(1):47-72.
Hamoudi,Amar,and JeffreySachs.1999. "EconomicConse-
quencesof HealthStatus:A Reviewof Evidence."HarvardUni-
versity, Center for International Development, Cambridge,
Mass.
Hanmer, Lucia.1994."What Happensto WelfareWhen UserFees
Finance Health Care?The Impact of Gender on PolicyOut-
comes:Theoryand Evidencefrom Zimbabwe."Institute of So-
cial Studies Working Paper Series, no. 180. The Hague,
Netherlands.
Hanmer, Lucia, and Felix Naschold. 1999. "Are the Interna-
tional Development Targets Attainable?"OverseasDevelop-
ment Institute, Portland House. London.
Hansen, Henrik, and Finn Tarp. 2000. "Aid EffectivenessDis-
puted." In Finn Tarp, ed., ForeignAid and Development:
LessonsLearntandDirectionsfortheFuture. London: Roudledge.
Hanson, Gordon, and Ann Harrison. 1999. "Trade and Wage In-
equalityin Mexico."IndustrialandLaborRelationsReview52(2):
271-88.
Hanushek, Eric, and Dennis D. Kimko. Forthcoming. "School-
ing, Labor Force Quality and the Growth of Nations."Amer-
ican EconomicReview.
Haque,NadeemUl, and JahangirAziz.1997."The Qualityof Gov-
ernance: 'Second Generation' Civil ServiceReform in Africa."
JournalofAfrican Economics8(1): 68-106.
Haque, Nadeem Ul, and Ali Khan. 1997. "Institutional Devel-
opment: SkillTransferencethrougha Reversalof 'Human Cap-
ital Flight' or Technical Assistance." IMF Working Paper
97/89. International Monetary Fund, Washington, D.C.
Harberger, Arnold. 1998. "A Vision of the Growth Process."
AmericanEconomicReview 88: 1-32.
Harrison, Ann, and Gordon Hanson. 1999. "Who Gains from
Trade Reform? Some Remaining Puzzles." NBER Working
Paper 6915. National Bureau of Economic Research, Carn-
bridge, Mass.
Harriss,John. 1999. "How Much DifferenceDoes PoliticsMake?
RegimeDifferencesacrossIndian Statesand Rural PovertyRe-
duction." London Schoolof Economicsand PoliticalScience.
Harriss,John, Janet Hunter, and Colin M. Lewis.1995. TheNew
InstitutionalEconomicsand Third WorldDevelopmentLondon:
Routledge.
Harris-White,Barbara.1996. "The PoliticalEconomyofDisability
and Development,with Special Referenceto India." United
Nations ResearchInstitute for SocialDevelopmentDiscussion
Paper 73. Geneva.
240 WORLD DEVELOPMENT REPORT 2000/2001
Harrold, Peter, and associates.1995. The Broad SectorApproach
to InvestmentLending:SectorInvestmentPrograms. World Bank
Discussion Paper 302. Washington, D.C.
Harvey,PharisJ., Terry Collingsworth,and BamaAthreya.1998.
"Developing EffectiveMechanisms for Implementing Labor
Rights in the Global Economy." International Labor Rights
Fund, Washington, D.C.
Hashemi, Syed M., and Sidney R. Schuler. 1997. "Sustainable
Banking with the Poor: A Case Study of the Grameen Bank."
John Snow Inc. Research and Training Institute Working
Paper 10. Arlington, Va.
Hausmann, Ricardo, and Michael Gavin. 1995. "Overcoming
Volatility in L.atinAmerica."Seminar Series,95-34. Interna-
tional Monetary Fund, Washington, D.C.
Heath, Julia A. 1998. "The Financing and Provisioning of Edu-
cation and Health Servicesin DevelopingCountries:A Review
Article."EconomicsofEducation Review 17(3): 359-62.
Heggie,Ian G., and PiersVickers.1998. CommercialManagement
and Financingof Roads.World Bank Technical Paper 409.
Washington, D.C.
Helleiner, Gerry K. 2000a. "Developing Countries in Global
Economic Governance and Negotiation Processes." Paper
prepared for World Institute for Development and Economic
ResearchprojectNew Rolesand Functions for the United Na-
tions and the Bretton Woods Institutions. University of
Toronto, Department of Economics.
.2000b. "ExternalConditionality, Local Ownership,and
Development." In Jim Freedman, ed., TransformingDevel-
opment.Toronto: University of Toronto Press.
Heller, Patrick. 1996. "Social Capital as a Product of Class Mo-
bilization and State Intervention: Industrial Workers in Ker-
ala, India." World Development24(6): 1055-71.
Hellman, Joel, Geraint Jones, Daniel Kaufmann, and Mark
Schankerman.2000. "MeasuringGovernance,Corruption,and
State Capture: How Firms and BureaucratsShape the Busi-
ness Environment in Transition Economies."Policy Research
Working Paper 2312. World Bank Institute, Governance,
Regulation, and Finance; and European Bank for Recon-
strucrion and Development, the Chief Economist's Office,
Washington, D.C.
Hentschel,Jesko, andJesse Bump. 2000. "Urban Income Poverty:
Some Cross-Country Comparisons." World Bank, Poverty
Group, Washington, D.C.
Herbst, Jeffrey. 1999. "The Role of Citizenship Lawsin Multi-
ethnic Societies:Evidencefrom Africa."In RichardJoseph,ed.,
State, Conflictand Democracyin Africa.Boulder, Colo.: Lynne
Rienner.
Herrera, Javier. 1999. Dynamique de la pauvrete et de linegalite
au Perou,1997-1998. Paris: Institute for Resource Develop-
ment/DIAL.
Herring,RonaldJ. 1999."PoliticalConditionsforAgrarianReform
and PoverryAlleviation."Backgroundpaper for WorldDevel-
opmentReport2000/2001. World Bank, Washington, D.C.
Herring, RonaldJ., and Rex Edwards. 1983. "Guaranteeing Em-
ployment to the Rural Poor: Social Functions and Class In-
terestsin EmploymentGuarantee Schemesin Western India."
WorldDevelopment11: 575-92.
Hertel, Thomas, and Will Martin. 1999a. "Developing Country
Interests in LiberalizingManufactures Trade." Purdue Uni-
versity,AgriculturalEconomicsDepartment, West Lafayette,
Ind.; and World Bank, Washington, D.C.
. 1999b. "Would Developing Countries Gain from In-
clusionof Manufacturesin the WTO Negotiations?"Paperpre-
senredat conferenceon the WTO and the Millennium Round,
20-21 September,Geneva.
Herzog, Henry W. 1997. "Ethniciuvand Job Tenure in a Seg-
mented LabourMarket:The Casefor New Zealand."Australian
EconomicReview30(2): 167-84.
Hicks, David. 1993.An Evaluationofthe Zimbabwe DroughtRe-
liefProgram,1992-1993: TheRolesofHousehold-Level Response
and DecentralizedDecisionMaking. Harare: World Food Pro-
gramme.
Hicks, Norman, and Quentin Wodon. 2000. "EconomicShocks,
SafetyNets, and Fiscal Constraints: Social Protection for the
Poor in Latin America."XII Seminario Regional de Politica
Fiscal:Compendio de Documentos 2000, Economic Com-
missionfor LatinAmericaand the Caribbean, United Nations,
Santiago.
Hill, Anne, and ElizabethKing. 1995. "Women's Education and
Economic Well-Being."FeministEconomics1(2): 1-26.
Hino, Toshiko. 1993. "Community Participationin 'Programme
de restructurationde I'hydrauliquevillageoise'in C6te d'lvoire."
World Bank, Washington, D.C.
Hirsch, Aaron. 1999. "Report on the Internet Center in Wa,
Northern Ghana." World Bank, Washington, D.C.
Hirschman, Albert. 1968. Journeys toward Progress:Studies of
EconomicPolicy-Makingin Latin America.New York:Green-
wood.
. 1984. GettingAheadCollectively:Grass-RootsOrganiza-
tions in Latin America.New York: Pergamon.
Hobcraft, John. 1993. "Women's Education, Child Welfare and
Child Survival:A Reviewof the Evidence." Health Transition
Re'ieus3: 159-75.
Hoddinort,John, and LawrenceHaddad. 1995."Does FemaleIn-
come Share Influence Household Expenditures? Evidence
from C6te d'Ivoire." OxfordBulletin of Economicsand Statis-
tics 57(1): 77-96.
Hoddinott, John, and Bill Kinsey. 1998. "Child Growth in the
Time of Drought." International Food PolicyResearchInsti-
tute, Washington, D.C.
. 2000. "Adult Health in Time of Drought." Food Con-
sumption and Nutrition Division Discussion Paper 79. In-
ternational Food PolicyResearchInstitute, Washington,D.C.
Hoff, Karla, Avishay Braverman, and Joseph E. Stiglitz, eds.
1993. TheEconomicsof'RuralOrganization:Theory,Practice,
and Policy.New York: Oxford University Press.
Holzer, Harry, and David Neumark. 1999. "AssessingAffirma-
tive Action." NBER Working Paper 7323. National Bureau
of Economic Research,Cambridge, Mass.
Holzmann, Robert, and SteenLau Jorgensen. 1999. "SocialPro-
tection as Social Risk Management: Conceptual Under-
standings for the Social Protection Sector Strategy Paper."
SocialProtection DiscussionPaper 9904. World Bank,Wash-
ington, D.C.

2000. "Social Risk Management: A New Conceptual
Framework for Social Protection and Beyond." Social Pro-
tection Discussion Paper 0006. World Bank, Human Devel-
opment Network, Washington, D.C.
Holzmann,Robert,Truman Packard,and Jose Cuesta.1999. "Ex-
tending Coveragein MultipillarPension Systems:Constraints
and Hypotheses, PreliminaryEvidence and Future Research
Agenda." Paper presented at World Bank conferenceon new
ideas about old age security, 14-15 September,Washington,
D.C.
Hopenhayn, Hugo, and Juan Pablo Nicolini. 1999. "Hetero-
geneity and Optimal Unemployment Insurance." Paper pre-
sented at conference on social protection and poverty,
Inter-American Development Bank, February,Washington,
D.C.
Horowitz, Donald. 1999. "Structure and Strategyin IEthnicCon-
flict: A Few Steps toward Synthesis."In Boris Pleskovicand
Joseph E. Stiglitz, eds., Annual World Bank Conferenceon
Development Economics 1998. Washington, D.C.: World
Bank.
Horton, Sue, and Dipak Mazumdar. 1999. "Vulnerable Groups
and Labor: The Aftermath of the Asian Financial Crisis."
Paper presented at World Bank-International Labour Orga-
nization seminar on economic crisis, employment, and labor
markets in East and SoutheastAsia.
Hossain, Mahabub. 1988. "Credit for Alleviation of Rural
Poverty: The Grameen Bank of Bangladesh."Research Re-
port 65. International Food PolicyResearchInstitute, Wash-
ington, D.C.
Hossain, Naomi, and Mick Moore, with Noushin Kalati,James
Manor, and Elisa Reis. 1999. "Elites, Poverty and Develop-
ment." Background paper for World Development Report
2000/2001. World Bank, Washington, D.C.
Hotchkiss, David. 1993. "The Role of Quality in the Demand
for Health Care in Cebu, Philippines." University of North
Carolina, Department of Economics,Chapel Hill.
Houtzager,Peter P., and Jonathan Pattenden. 1999. "Finding the
Shape of the Mountain: When 'the Poor' Set the Agenda."
Backgroundpaper for WorldDevelopmentReport2000/2001.
U.K. Department for InternationalDevelopment,London;and
World Bank, Washington, D.C.
Hoy, M., and Emmanuel Jimenez. 1997. "The Impact on the
Urban Environment of Incomplete Property Rights." World
Bank, Development ResearchGroup, Washington, D.C.
Hsiao, William.1995. "The ChineseHealth Care System:Lessons
for Other Nations."SocialScienceand Medicine41(8): 1047-55.
Hulme, David, and Paul Mosley. 1996. FinanceagainstPoverty.
New York: Routledge.
Humana, Charles. 1992. WorldHuman RightsGuide.New York:
Oxford University Press.
Humplick, Frannie, and Azadeh Moini-Araghi. 1996a. "Decen-
tralized Structures for Providing Roads." Policy Research
Working Paper 1658. World Bank, Washington, D.C.
- 1996b. "Is There an Optimal Structure for Decentral-
ized Provision of Roads?" Policy Research Working Paper
1657. World Bank, Washington, D.C.
Huntington, Samuel,and Joan Nelson. 1976. No EasvChoice:Po-
BIBLIOGRAPHIC NOTE 241
liticalParticipationin DevelopingCountries.Cambridge,Mass.:
Harvard UniversityPress.
Hutchinson, Paul. Forthcoming. "Combating Illness." In Ritva
Reinikka and Paul Collier, eds., Ugandas Recovery:The Role
of Farms,Firms and Government.Washington, D.C.: World
Bank.
IDA (International Development Association). 1998. Additions
to IDA Resources:Twelfth Replenishment-A Partnershipfor
PovertyReduction.Washington, D.C.: World Bank.
. 2000. "Uganda: Poverty Reduction Strategy Paper and
Joint World Bank-IMF StaffAssessmentof the PRSP."World
Bank, Washington, D.C.
IDB (Inter-American Development Bank). 1995. Overcoming
Volatility:EconomicandSocialProgress in LatinAmerica,1995
Report.Washington, D.C.
. 1997. Latin Americaafter a Decade of Reforms.Wash-
ington, D.C.
. 2000. SocialProtection for Equity and Growth.Sustaini-
able Development Department, Poverty and Inequality Ad-
visoryUnit, Washington, D.C.
. Variousyears.StatisticalandSocialDatabase.
Washington,
D.C. [www.iadb.org/int/sta/ENGLISH/staweb/statshp.htm].
IDRC (InternationalDevelopment ResearchCentre). 1998. Aca-
cia Page. [www.idrc.ca/acacia/engine/eng-6.htm].
IFC (International Finance Corporation) and FIAS (Foreign In-
vestment AdvisoryService). 1997. ForeignDirectInvestment.
Lessonsof Experience5. Washington, D.C.
ILO (International Labour Organization). 2000. "International
Labour Standards and Human Rights." [www.ilo.org/public/
english/standards/norm/index.htm].
Imber, Mark. 1996. "The Environment and the United Na-
tions." In John Vogler and Mark Imber, eds., The Environ-
ment and InternationalRelations.London: Routledge.
IMF (International Monetary Fund). 1995. International Fi-
nancialStatistics.Washington, D.C.
. 1999. "Military SpendingContinues to Stabilize:Some
Countries IncreaseSocial Spending."International Monetary
Fund Survey.Washington, D.C.
. 2000. Progress
in StrengtheningtheArchitectureof the In-
ternationalFinancialSystem.Washington, D.C.
IMF (InternationalMonetary Fund) and IDA (InternationalDe-
velopment Association).1999. PovertyReductionStrategyPa-
pers: OperationalIssues. Washington, D.C.: International
Development Association.
IMF (International Monetary Fund), OECD (Organisation for
EconomicCo-operation and Development),UN (United Na-
tions), and World Bank. 2000. 2000: A BetterWorldforAII--
Progress towards the International Development Goals.
Washington, D.C. [www.paris2l.org/betterworld/].
INADES Formation Tchad. 1999. "Synthse de la consultation
de la Banque Mondiale sur la pauvrete au Tchad." Back-
ground paper for WorldDevelopmentReport2000/2001.World
Bank, Washington, D.C.
Independent Review Team of Zambia's Health Sector. 1997.
ComprehensiveReviewoftheZambian HealthReforms.Vol. 1,
Main Report.Geneva:World Health Organization.
Independent Task Force.1999. SafeguardingProsperity in a Global
242 WORLD DEVELOPMENT REPORT 2000/2001
FinancialSystem: The Future InternationalFinancialArchitec-
ture.Sponsored by the Council on Foreign Relations. Wash-
ington, D.C.: Institute for International Economics.
INEGI (Instituto Nacional de Estadistlca Geografla e Infor-
mitica). 1992. Encuesta nacional de ingresos y gastos de los
hogares1992. Mexico City.
- 1994. Encuesta nacional de ingresosy gastosde los hogares
1994. Mexico City.
- 1996. Encuesta nacional de ingresosygastos de los hogares
1996 Mexico City.
lnstituto Nacionalde Estadisticay Censos. 1998.Encuestade condi-
cionesde vida. Quito, Ecuador.
Instituto Nacional de Estadisticay CensosArgentina. 1991."Cen-
sus." [www.indec.mecon.ar/default.htm].
InternationalFederationof Red Cross and Red Crescent Societies.
1993. World DisasterReport1993. Geneva.
- 1999. WorldDisasterReport1999. Geneva.
International Institute for Environmentand Development. 1992.
"SpecialIssueonApplicationsof WealthRanking."RapidRural
AppraisalNote 15. London.
InternationalTask Forceon CommodityRisk Managementin De-
veloping Countries. 1999. "Dealing with Commodity Price
Volatility in DevelopingCountries: A Proposal for a Market-
BasedApproach." World Bank, Washington, D.C.
IRIS (Center for Institutional Reform and the Informal Sector)
and USAID (U.S. Agency for International Development).
1996. Governanceand theEconomyin Africa: ToolsforAnaly-
sis and Reform of Corruption.Washington, D.C.: USAID.
Irwin, Timothy. 1997. "Price Structures, Cross-Subsidiesand
Competition in Infrastructure."Viewpoint108. World Bank,
Finance, Private Sector, and Infrastructure Network, Wash-
ington, D.C.
Isham, Jonathan. 1999. "The Effect of Social Capital on Tech-
nologyAdoption: Evidencefrom Rural Tanzania." Paper pre-
sentedatAnnual MeetingofAmerican Economic Association,
January, New York.
Isham, Jonathan, and Daniel Kaufmann. 1999. "The Forgot-
ten Rationale for Policy Reform: The Productivity of In-
vestment Projects." Quarterly Journal of Economics114(1):
149-84.
Isham, Jonathan, Deepa Narayan, and Lant Pritchett. 1995.
"DoesParticipationImprovePerformance?EstablishingCausal-
ity with SubjectiveData." WorldBank EconomicReview9(2):
175-200.
Izaguirre,Ada Karina. 1999. "Private Participation in Telecom-
munications: Recent Trends." Viewpoint 204. World Bank,
Finance, Private Sector, and Infrastructure Network, Wash-
ington, D.C.
Jacoby,Hanan G., and EmmanuelSkoufias.1997. "Risk,Financial
Markets and Human Capital in a DevelopingCountry." Re-
view of Economic Studies 64(3): 311-35.
Jalan,Jyotsna, and Martin Ravallion. 1997a. "Are the Poor Less
Well Insured? Evidence on Vulnerability to Income Risk in
Rural China." Journal ofDevelopmentEconomics58: 61-81.
- 1997b. "Spatial PovertyTraps?"PolicyResearchWork-
ing Paper 1862. World Bank, Washington, D.C.
- 1998a. "BehavioralResponsesto Risk in Rural China."
Policy Research Working Paper 1978. World Bank, Wash-
ington, D.C.
. 1998b. "DeterminantsofTransient and Chronic Poverty:
Evidencefrom Rural China." PolicyResearchWorking Paper
1936. World Bank, Washington, D.C.
. 1999a."China's LaggingPoorAreas."AmericanEconomic
Review 89: 301-05.
. 1999b. Do Transient and Chronic Poverty in Rural China
Share Common Causes?Washington, D.C.: World Bank.
. 1999c. "Income Gains to the Poor from Workfare: Es-
timates for Argentina's Trabajar Program." Policy Research
VWorkingPaper 2149. World Bank, Washington, D.C.
James,Estelle.2000. "Old-AgeProtectionfor the Uninsured:What
Are the Issues?"In Nora Lustig, ed., ShieldingthePoor:Social
Protectionin theDevelopingWorld.Washington, D.C.: Brook-
ings Institution and Inter-AmericanDevelopment Bank.
Jamison, Dean T., Julio Frenk, and Felicia Knaul. 1998. "Inter-
national CollectiveAction in Health: Objectives, Functions,
and Rationale." Lancet 351: 514-17.
Jamison, Dean T., Jia Wang, Kenneth Hill, and Juan-Luis Lon-
dofno.1996. "Income,Mortality and FertilityControl in Latin
America:Country-LevelPerformance,1960-90." World Bank,
Latin Americaand the Caribbean Region, Washington, D.C.
Javaraman, Rajshri. 1999. "Kerala and Uttar Pradesh: A Case
Study." Background paper for World Development Report
2000/2001. World Bank, Washington, D.C.
Jejeebhoy,Shireen. 1995. Women'sEducation,Autonomy,and Re-
productive Behavior: Experience from Developing Countries.
New York: Oxford UniversityPress.
Jensen, Robert. 1998. "Public Transfers, Private Transfers, and
the 'Crowding Out' Hypothesis: Theory and Evidence from
South Africa."John F. Kennedy School of Government Fac-
ulty Working Paper R98-08. Harvard University,Cambridge,
Mass.
Jepma, Catrinus J. 1991. The Tying ofAid. Paris: Organisation
for Economic Co-operation and Development.
Jha, Prabhat, and David Nalor. 1999.A FineBalance:Privateand
PublicHealth Carein UrbanIndia. Washington, D.C.: World
Bank.
Jimenez,Emmanuel.1986. "The PublicSubsidizationof Education
and Health in Developing Countries: A Review of Equity
and Efficiency." WorldBank ResearchObserverI (1): II 1-29.
.1987. PricingPolicyin the SocialSectors.Baltimore,Md.:
Johns Hopkins UniversityPress.
Jimenez, Emmanuel, and Marlaine Lockheed. 1995. Publicand
PrivateSecondaryEducationin DevelopingCountries:A Com-
parativeStudy. World Bank Discussion Paper 309. Washing-
ton, D.C.
Jimenez,Emmanuel,and Vicente Paqueo. 1996. "Do Local Con-
tributions Affect the Efficiencyof Public Primary Schools?"
Economics of Education Review 15(4): 377-86.
Jimenez, Emmanuel, and Yasuyuki Sawada. 1998. "Do
Community-ManagedSchoolsWork? An Evaluationof El Sal-
vador's Educo Program." Impact Evaluation of Education
ReformsPaper8. World Bank, DevelopmentResearchGroup,
Washington, D.C.
Jodha, Nerpat S. 1986. "Common Property Resourcesand the

Rural Poor in Dry Regionsof India." Economicand Political
Weekly21(27): 1169-8 1.
Johansen, Frida. 1993. PovertyReductionin EastAsia: The Silent
Revolution.World Bank Discussion Paper 203. Washington,
D.C.
Johnson, Chalmers A. 1982. MITIand theJapaneseMiracle:The
GrowthofIndustrialPolicy,1925-1975. Stanford,Calif.:Stan-
ford University Press.
Johnson, Simon, Daniel Kaufmann, and Pablo Zoido-Lobaton.
1998. "Regulatory Discretion and the UnofficialEconomy."
AmericanEconomicReview88(2): 387-92.
Joint Commonwealth Secretariat-World Bank Task Force on
the Small States.2000. "Small States:Meeting Challengesin
the Global Economy." Washington, D.C.
Jones, Patricia. 1998. "Skill Formation and Inequality in Poor
Countries: How Much Do Ethnic Neighborhoods Matter?"
Journal ofAfrican Economies7(1): 62-90.
Jorgensen,SteenLau, and Julievan Domelen. 1999. "Helping the
Poor Manage Risk Better: The Role of Social Funds." Paper
presented at Inter-AmericanDevelopment Bank Conference
on Social Protectionand Poverty,4-5 February,Washington,
D.C.
Joseph, Richard. 1999. State, Conflictand Democracyin Africa.
Boulder, Colo.: Lynne Rienner.
Joshi, A., and Mick Moore. Forthcoming. "Rights. Institutions
and Poverty."In Alison L. Booth and Paul Mosley,eds., The
New PovertyStrategies:What Have They Achieved?London:
Macmillan.
Joshi, Heather, Elizabeth Cooksey, Richard Wiggins, Andrew
McCulloch, GeorgiaVerropoulou, and Lynda Clarke. 1999.
"Diverse Family Living Situations and Child Development:
MultilevelAnalysisComparing Longitudinal Evidence from
Britain and the United States." Internationaljournal of Law,
Policyand the Family 13(3):292-314.
Kabeer,Naila,and SubrahmanianRamya.1996. "Institutions,Re-
lations and Outcomes: Framework and Tools for Gender-
Aware Planning." Discussion Paper 357. Institute of
Development Studies, SussexUniversitv, Brighton.
Kahk6nen, Satu, and Louis Pouliquen. 1999. "Institutions and
Rural Infrastructure Delivery."World Bank, Department of
Regional Economic and Social Policy,Washington, D.C.
Kakwani, Nanak. 1993. "Performance in Living Standards: An
International Comparison."Journal of DevelopmentEconom-
ics41: 307-36.
1999. "Poverty and Inequality during the Economic
Crisis in Thailand." National Economic and Social Develop-
ment Board/AsianDevelopmentBank.Indicatorso/Well-Being
and PolicyAnalysis3(1).
Kakwani, Nanak, and Jaroenjit Pothong. 1998. "Impact of
Economic Crisison the Standard of Livingin Thailand." Na-
tional Economic and Social Development Board /Asian De-
velopment Bank. Indicatorsof Well-Beingand PolicyAnalysis
2(4).
Kakwani,Nanak, and NicholasM. Prescott.1999."Impact of Eco-
nomic Crisis on Poverty and Inequality in Korea." World
Bank, EastAsiaand PacificRegion,Human DevelopmentSec-
tor Unit, Washington, D.C.
BIBLIOGRAPHIC NOTE 243
Kalton,Graham. 1983. CompensatingforMissing SurveyData.Ann
Arbor: Universityof Michigan Press.
Kanbur, Ravi. 1986."StructuralAdjustment,MacroeconomicAd-
justment and Poverty: A Methodology for Analysis."CEPR
Discussion Paper 132. Centre for Economic PolicyResearch,
London.
. Forthcoming. "Income Distribution and Development."
In Anthony B. Atkinson and Fran,ois Bourguignon, eds.,
HandbookofIncomeDistribution.New York:North Holland-
Elsevier.
Kanbur, Ravi, and Nora Lustig. 1999. "Why Is Inequality Back
on the Agenda?"Department of Agricultural, Resource,and
Managerial EconomicsWorking Paper 99-14. Cornell Urii-
versity,Ithaca, N.Y.
Kanbur, Ravi, and Lyn Squire. 1999. "The Evolution of Think-
ing about Poverty: Exploring the Interactions." Department
of Agricultural, Resource,and Managerial EconomicsPaper
99-24. Cornell University, Ithaca, N.Y.
Kanbur,Ravi, and Todd Sandler,with KevinM. Morrison. 1999.
The Future of DevelopmentAssistance:CommonPoolsand Li-
ternationalPublic Goods.PolicyEssay25. Washington, D.C.:
OverseasDevelopment Council.
Katakura,Yoko, and AlexanderBakalian. 1998. "PROSANEAR:
People, Poverty and Pipes." United Nations Development
Programme-World Bank Water and Sanitation Program,
Washington. D.C.
Katz, Travis, and Jennifer Sara. 1998. "Rural Water Supply:
Global Study." United Nations Development
Programme-World Bank Water and Sanitation Program,
Washington, D.C.
Kaufmann, Daniel, and Shang-jin Wei. 1999. "Does 'GreaLse
Money' SpeedUp the Wheels of Commerce?"NBER Work-
ing Paper 7093. National Bureau of Economic Research,
Cambridge, Mass.
Kaufmann,Daniel, Aart Kraay,and Pablo Zoido-Lobaton. 1999.
"GovernanceMatters." PolicyResearchWorking Paper 2196.
World Bank, Washington, D.C.
Kaufmann, Daniel, Pablo Zoido-Lobaton,and YoungLee.2000.
"Governanceand Anticorruption:EmpiricalDiagnosticStudy
for Ecuador." World Bank, Washington, D.C.
Kaul, Inge, Isabelle Grunberg, and Marc Stern. 1999a. "Defin-
ing Global Public Goods." In Inge Kaul, IsabelleGrunberg,
and Marc Stern, eds., GlobalPublic Goods:InternationalCo-
operationin the 21st Century.New York: Oxford University
Press.
, eds. 1999b. GlobalPublic Goods:International Cooper-
ation in the21st Century.New York:Oxford UniversitvPress.
Kawachi,Ichiro, and LisaBerkman. 2000. "SocialCohesion, So-
cialCapitaland Health."In LisaBerkmanand Ichiro Kawachi,
eds., SocialEpidemiology.New York:Oxford UniversityPress.
Kayani, Rogati, and Andrew Dymond. 1997. OptionsforRural
TelecommunicationsDevelopment.World Bank Technical
Paper 359. Washington, D.C.
KDP (Kecamatan Development Project) Secretariat.1999. "Ke-
camatan Development Project: First Annual Report." Gov-
ernment of Indonesia, National Coordination Team, Jakarta.
Kelley,AllenC., and Robert M. Schmidt. 1994. Populationand
244 WORLD DEVELOPMENT REPORT 2000/2001
IncomeChange:RecentEvidence.World Bank DiscussionPaper
249. Washington, D.C.
Kemal, A.R. 1994. "Structural Adjustment, Employment, In-
come Distribution and Poverty."PakistanDevelopmentReview
33: 901-11.
Kennedy, Bruce, Ichiro Kawachi, Roberta Glass, and Deborah
Prothrow-Stith. 1998. "Income Distribution, Socioeconomic
Status, and Self-RatedHealth in the United States:Multilevel
Analysis."BritishMedicalJournal 317: 917-21.
Kenny, Charles. 1999. "Accessto Telecommunications and In-
formaticsIs of Great Benefitto the Poor, and This Has to Mean
PublicAccess."Boxfor WorldDevelopment Report2000/2001.
World Bank, Washington, D.C.
Kessides, Christine. 1997. "World Bank Experience with the
Provision of Infrastructure Servicesfor the Urban Poor: Pre-
liminarv Identification and Reviewof BestPractices."World
Bank, Urban Development Division, Washington, D.C.
Khan, Naila, and Maureen Durkin. 1995. "Framework: Preva-
lence." In P. Zinkin and H. McConachie, eds., Disabled
Childrenand DevelopingCountries.Cambridge: Cambridge
UniversityPress.
Khandker, Shahidur R. 1998. FightingPovertywith Microcredit:
Experiencein Bangladesh.New York:Oxford UniversityPress.
Khandker,ShahidurR., BaquiKhalily,andZahed Khan,eds. 1996.
CreditProgramsforthePoor:Householdand Intrahousehold Im-
pacts and ProgramSustainability.Washington, D.C.: World
Bank and BangladeshInstitute of Development Studies.
Kiguel,Miguel. 1999. "The Argentine Currency Board." Docu-
mento de Trabajo 152. Universidaddel CEMA, BuenosAires.
KIHASA (Korea Institute for Health and Social Affairs) and
UNDP (United Nations Development Programme). 1998.
Korea:Human DevelopmentReport1998. Seoul.
Killick,Tony. 1997."Principals,Agents,and the Failingsof Con-
ditionality."JournalofInternationalDevelopment9(4): 483-95.
Killick,Tony, with Rumani Gunatilakaand Ann Marr. 1998.Aid
and the PoliticalEconomyofPolicyChange.London: Overseas
Development Institute.
Kim, Elim. 1991. A Studyof theFamilyLaw and lts ReformMove-
ment. Seoul: Korean Women's Development Institute.
Kim, Jooseop, Harold Alderman, and Peter Orazem. 1998. "Can
Private SchoolsSubsidiesIncreaseSchoolingfor the Poor?The
Quetta Urban FellowshipProgram."Impact Evaluationof Ed-
ucation ReformsPaper 11.World Bank,DevelopmentResearch
Group, Washington, D.C.
Kim, Kwangkee,and Philip Moody. 1992. "More Resources,
Better Health? A Cross-National Perspective."SocialScience
and Medicine34(8): 837-42.
King, Elizabeth,and Peter Orazem. 1999. "EvaluatingEducation
Reforms:Four Casesin DevelopingCountries." World Bank,
Development ResearchGroup, Washington, D.C.
King, Elizabeth,and Berk Ozler. 1998. "'XWhat's Decentralization
Got to Do with Learning?The Caseof Nicaragua'sSchoolAu-
tonomy Reform." Impact Evaluationof Education Reforms
Paper 9. World Bank, Development ResearchGroup, Wash-
ington, D.C.
Kinsey,Bill, KeesBurger, and Jan W. Gunning. 1998. "Coping
with Drought in Zimbabwe: Survey Evidence on Responses
of Rural Households to Risk." World Development 26(1):
89-110.
Klasen,Stephan. 1994."'MissingWomen'Reconsidered." World
Development22(7): 1061-71.
. 1999. "DoesGender Inequality ReduceGrowth and De-
velopment?Evidencefrom Cross-Country Regressions."Pol-
icy ResearchReport on Gender and Development Working
Paper Series,no. 7. World Bank, Washington, D.C.
Klees, Rita, Joana Godinho, and Mercy Lawson Doe. 1999.
"Health, Sanitation,and Hygiene in Rural Water Supply and
SanitationProjectsand Other World Bank-FinancedProjects."
World Bank, Europe and Central Asia Regional Studies Pro-
gram, Washington, D.C.
Klitgaard, Robert. 1990. TropicalGangsters.New York: Basic
Books.
. 1997. "Cleaning Up and Invigorating the Civil Ser-
vice." Report prepared for Operations Evaluation Depart-
ment. World Bank, Washington, D.C.
Klugman,Jeni, and Jeanine Braithwaite. 1998. "Poverty in Rus-
sia during the Transition: An Overview." World Bank Re-
searchObserver3(1): 37-58.
Knack,Stephen, and Philip Keefer. 1995. "Institutions and Eco-
nomic Performance: Cross-Country Tests Using Alternative
Institutional Measures."Economicsand Politics7: 207-27.
. 1997. "Does Social Capital Have an Economic Payoff?
A Cross-Country Investigation." QuarterlyJournal of Eco-
nomics 112: 1251-88.
Knack,Stephen, and Nick Manning.2000. "PolicyVolatility and
Poverty." World Bank, Poverty Reduction and Economic
Management Network, Washington, D.C.
Kochar, Anjini. 1995. "Explaining Household Vulnerability to
Idiosyncratic Income Shocks." American EconomicAssocia-
tion Papersand Proceedings85(2): 159-64.
. 1999. "Smoothing Consumption by Smoothing In-
come: Hours-of-Work Responses to ldiosyncratic Agricul-
tural Shocksin RuralIndia." ReviewofEconomicsand Statistics
81(1): 50-61.
Kohli, Atul. 1987. The State and Povertyin India: The Politicsof
Reform.New York: Cambridge Universitv Press.
Kormendi, Roger, and Philip McGuire. 1985. "Macroeconomic
Determinants of Growth: Cross-Country Evidence."Journal
ofMonetary Economics16(2): 141-63.
Kornai, Janos. 2000. "Ten Yearsafter 'the Road to a Free Econ-
omy": The Author's Self-Evaluation." Paper presented at
World BankAnnual Conference on Development Econom-
ics, 18-20 April, Washington, D.C.
Kozel,Valerie,and BarbaraParker.2000. "IntegratedApproaches
to PovertyAssessmentin India." In Michael Bamberger,ed.,
IntegratingQuantitativeand QualitativeResearchin Develop-
mentProjects.Washington, D.C.: World Bank.
Kreimer, Alcira. 1999. "Learning Lessons from the Earthquake
in Turkey." Box for World DevelopmentReport2000/2001.
World Bank, Washington, D.C.
Kreimer, Alcira, and Edward Echeverria. 1991. "Case Study:
Housing Reconstruction in Mexico City." In Alcira Kreimer
and Mohan Munasinghe, eds., Managing Natural Disaster
and the Environment:SelectedMaterialsfrom the Colloquium

ontheEnvironment andNaturalDisasterManagement.Wash-
ington,D.C.:World Bank.
Kreimer,Alcira,EdwardEcheverria,and MarthaPreece.1991.
"CaseStudy:ReconstructionafterNorthChina'sEarthquake."
In AlciraKreimerand MohanMunasinghe,eds.,Managing
NaturalDisasterandtheEnvironment: Selected
Materials
from
theColloquium ontheEnvironmentand NaturalDisasterMan-
agement. Washington,D.C.: World Bank.
Kreimer,Alcira,MargaretArnold,ChristopherBarham,Paul
Freeman,RoyGilbert,FrederickKrimgold,RodneyLester,
JohnD. Pollner,andTomVogt.1999.Managing Disaster
Risk
inMexico:MarketIncentives
forMitigationInvestment.Wash-
ington,D.C.:World Bank.
Kremer,Michael.1999."Purchase Pre-commitmentsforNewVac-
cines:Rationaleanda ProposedDesign."HarvardUniversity,
Departmentof Economics, Cambridge,Mass.;and Brookings
Institution,Washington,D.C.
Krishna,Anirudh,Norman Uphoff,and Milton Esman,eds.
1997.Reasons for Hope:InstructiveExperiences
in RuralDe-
velopment.West Hartford,Conn.:Kumarian.
Krueger,Alan B., and MichaelLindahl. 1999."Educationfor
Growthin Swedenand theWorld."SwedishEconomic Policy
Review6(2):289-339.
Krueger,Anne0. 1993."EastAsia:LessonsforGrowthTheory."
Paper presented at Fourth Annual East Asian Seminar on
Economics,NationalBureauof EconomicResearch,17-19
June, SanFrancisco.
Krugman,Paul. 1998. "SavingAsia:It's Time to Get Radical."
Fortune138:75-80.
Kruse,Sten-Erik,Timo Kyllonen,SatuOjanpera,RogerC. Rid-
dell, and Jean Vielajus.1997. "Searchingfor Impactand
Methods:NGO EvaluationSynthesisStudy."Reportpre-
paredfor Organisationfor EconomicCo-operationand De-
velopment,Development AssistanceCommitteeExpertGroup
on Evaluation,Paris.
Kull,Steven,I.M.Destler,and ClayRamsay.1997."TheForeign
PolicyGap:HowPolicymakers Misreadthe Public."Univer-
sityof Maryland,Centerfor Internationaland SecurityStud-
ies,CollegePark.
Kunreuther,Howard.1999."Incentives forMitigationInvestments
andMore EffectiveRiskManagement:The Needfor Public-
PrivatePartnerships."University
of Pennsylvania.
Operations
and InformationDepartment,Philadelphia.
Kurtz,Marcus.1999."The PoliticalEconomyof Pro-poorPoli-
ciesin Chileand Mexico."Backgroundpaperfor WorldDe-
velopmentReport2000/2001. WorldBank,Washington,D.C.
Lachaud,Jean-Pierre.1999."Pauvretemenageet genreen Afrique
sub-Saharienne: Nouvellesdimensionsanalytiques."Univer-
site Montesquieu-BordeauxIV, Centre d'Economie du
Developpement, Bordeaux.
Lago,Ricardo.1991. "The Illusionof PursuingRedistribution
through Macropolicy: Peru's Heterodox Experience,
1985-1990."In RudigerDornbuschand SebastianEdwards,
eds., Macroeconomics of Populism.Chicago:Universityof
ChicagoPress.
Lal, Deepak,I. Natarajan,and RakeshMohan. 2000. "Eco-
nomicReformsand PovertyAlleviation: A TaleofTwo Sur-
BIBLIOGRAPHIC NOTE 24 5
veys."Universityof Californiaat LosAngeles,Department
of Economics.
Lamontagne,Jacques.1999. "NationalMinorityEducationin
China:A NationwideSurveyacrossCounties."In GerardA.
Postiglione,ed., ChinasNationalMinority
Education:Culture,
Schooling and Development.London:Falmer.
Lampietti,Julian,AnthonyKolb,SumilaGulyani,and Vahram
Avenessian.2000. "PricingServicesto Protect the Poor:
Lessons fromArmenia."WorldBank,EuropeandCentralAsia
Region,Washington,D.C.
Lanjouw,Jean Olson.1997. "Demvstifying PovertyLines."Se-
ries on PovertyReduction.United Nations Development
Programme,NewYork.
Lanjouw,Peter,andMartinRavallion.1999."BenefitIncidence
and the Timingof ProgramCapture."WorldBankEconomic
Review13(2):257-74.
Lanjouw,Peter,and NicholasStern.1991."PovertyinPalanpur."
WorldBankEconomic Review5(1):23-56.
Larrain,Felipe,and AndresVelasco.1999."ExchangeRatePol-
icyfor EmergingMarkets:One SizeDoesNot FitAll."Har-
vardUniversity,DepartmentofEconomics, Cambridge,Mass.
Lavy,Victor.andJean-MarcGermain.1994.QualityandCost.in
HealthCareChoicein Developing Countries.
LivingStandards
Measurement StudyWorkingPaper105.Washington,D.C.:
World Bank.
Lavy,Victor,John Strauss,DuncanThomas,and Philippede
Vreyer.1996."QualityofCare,SurvivalandHealthOutcomes
in Ghana."Journalof HealthEconomics 15:333-57.
Leamer,EdwardE.,andPeterK.Schott.1999."NaturalResources
as a Sourceof Latin AmericanIncomeInequality."Back-
groundpaperfor WorldDevelopmentReport2000/2001. World
Bank,Washington,D.C.
Legovini,Arianna,CesarBouill6n,andNoraLustig.1999."Can
EducationExplainIncomeInequalityChangesin Mexico?"
Inter-American DevelopmentBank,Washington,D.C.
Lembuya,Peter.1992.AmboseliNationalPark,theSurrounding
GroupRanches and Community Conservation.Nairobi:African
WildlifeFoundation.
Lensink,Robert,and HowardWhite. 1998."Doesthe Revival
of InternationalPrivateCapitalFlowsMeanthe End of Aid?
An Analysisof DevelopingCountries'Accessto PrivateCap-
ital." WorldDevelopment 26: 1221-34.
Leonard,David.1987."PoliticalRealitiesofAfricanManagement."
WorldDevelopment 15:899-9 10.
Lepkowski, James,GrahamKalton,and DanielKazprsyk.1989.
"Weighting Adjustmentsfor PartialNon-responseinthe 1984
SIPPPanel."Paperpresentedat sectionon surveymethodsre-
search,meetingof AmericanStatisticalAssociation.
Levav,Itzhak.1999."TheIndivisibleFaceof Disaster."In World
Bank, Investingin Prevention:A SpecialReporton Disaster
RiskManagement. Washington,D.C.
Levine,Ross.1997. "FinancialDevelopmentand Economric
Growth:Viewsand an Agenda."JournalofEconomicLitera-
ture35(2):688-726.
Levine,Ross,and DavidRenelt.1992."A Sensitivity Analysisof
Cross-Country Regressions."American EconomicReview 84(4):
942-63.
246 WORLD DEVELOPMENT REPORT 2000/2001
Levison, Deborah, Richard Anker, Shahid Ashraf, and Sandya
Barge. Forthcoming. "Is Child Labor Really Necessary in
India's Carpet Industry?" In Richard Anker and others, eds.,
Economicsof Child Laborin SelectedIndustriesof India. New
Delhi: Hindustan.
Levy,Hernan. 1996. "MoroccoImpactEvaluationReport: Socio-
Economic Influenceof RuralRoads-A Studyof Rural Roads
Financed under the World Bank's Fourth Highwav Project."
World Bank, Operations Evaluation Department, Washing-
ton, D.C.
Levy,Margaret.1996."Socialand UnsocialCapital:A ReviewEssay
oniRobert Putnam's A,akingDemocracyWork." Politicsand
Society24(1): 45-55.
Lewis,Maureen. 2000. "Informal Health Paymentsin EasternEu-
rope and Central Asia:Issues,Trends, and PolicvImplications."
World Bank, Europe and Central Asia Region, Washington,
D.C.
Lewis,Maureen, Gunnar S. Eskeland,and Ximena Traa-Valerezo.
Forthcoming. "Challenging El Salvador's Rural Health Care
Strategy."PolicyResearchWorking Paper 2164. World Bank,
Washington, D.C.
Li, Guo, Diane Steele,and Paul Glevwe. 1999. "Distribution of
Government Education Expenditures in Developing Coun-
tries." World Bank, Development Research Group, Wash-
ington, D.C.
Li, Hongyi, and Heng-fu Zou. 1998. "Income Inequality Is Not
Harmful for Growth: Theory and Evidence." Review of De-
velopmentEconomics2(3): 318-34.
Li, Hongyi, Lyn Squire,and Heng-fu Zou. 1998. "ExplainingIn-
ternational Inequality and Intertemporal Variations in In-
come Inequality." EconomicJournal 108: 26-43.
Liang, Xiaoyan. 1996. "Bangladesh Female Secondarv School
AssistanceProject." World Bank, Human Development De-
partment, Washington, D.C.
Lim, Young-Jae,and Robert Townsend. 1998. "General Equi-
librium Models of Financial Systems:Theory and Measure-
ment in Village Economies."ReviewofEconomicDynamics1:
59-118.
Lin, Justin Yifu. 1995. "Commentary: Food Policy in China-
In Retrospect and Prospect." IFPRI Report 17(3). Interna-
tional Food Policy ResearchInstitute, Washington, D.C.
Lindauer, David. 1999. "Labor Market Reform and the Poor."
Backgroundpaper for WorldDevelopmentReport2000/2001.
World Bank, Washington, D.C.; WellesleyCollege, Welles-
ley, Mass.; and Harvard Institute for International Develop-
ment, Cambridge, Mass.
Lindauer, David, and Barbara Nunberg, eds. 1996. Rehabilitat-
ing Government:Payand EmploymentReformin Africa.Alder-
shot: Avebury.
Lipton, Michael. 1977. WhyPoorPeopleStay Poor:UrbanBiasin
WorldDevelopment.Canberra: AustralianNational Universitr
Press.
1996. DefiningandMeasuringPoverty:Conceptuallssues.
New York: United Nations Development Programme.
1998. Successesin Anti-Poverty.Geneva: International
Labour Office.
Lipton, Michael,and Martin Ravallion. 1995. "Poverty and Pol-
icy." In Jere R. Behrmanand T.N. Srinivasan,eds., Handbook
ofDevelopmentEconomics. Vol. 3. Amsterdam:North Holland.
1Little,Roderick. 1988. "Missing-DataAdjustments in LargeSur-
veys."JournalofBusinessand EconomicStatistics6(3):287-96.
Litvack,Jennie, and Carol Bodart. 1993. "User Fees Plus Qual-
ity Equals Improved Accessto Health Care: Resultsof a Field
Experiment in Cameroon." SocialScienceand Medicine37(3):
369-83.
Litvack,Jennie, and Dennis A. Rondinelli,eds. 1999. Market Re-
form in Vietnam:BuildingInstitutionsfor Development.West-
port, Conn.: Quorum Books.
Litvack,Jennie, and JessicaSeddon, eds. 1999. "Decentralization
Briefing Notes." World Bank, Poverty Reduction and Eco-
nomic Management Network, Washington, D.C.
Liu, Yuanli,William C. Hsiao, and Karen Eggleston. 1999. "Eq-
uirv in Health and Health Care: The Chinese Experience."So-
cial Scienceand Medicine49: 1349-56.
Livi-Bacci,Massimo. 1997. "Population, Constraint, and Adap-
tation: A Historical Outlook." In Robert Dorfman and Peter
Rogers, eds., Sciencewith a Human Face:Festschriftin Hon-
our ofRogerRandallRevelle.Cambridge, Mass.:Harvard Uni-
versity Press.
Llavador,Humberto G., and John E. Roemer.2000. "An Equal-
Opportunity Approach to the Allocation of International
Aid." YaleUniversity, Department of Political Science,New
Haven, Conn.
Lockheed, Marlaine, and Qinghua Zhao. 1993. "The Empty
Opportunity: LocalControl of SecondarySchoolsand Student
Achievement in the Philippines." Policy ResearchWorking
Paper 825. World Bank, Washington, D.C.
Lockheed,Marlaine,Dean T. Jamison,and LawrenceLau. 1980.
"Farmer Education and Farm Efficiency: A Survev." Eco-
nomicDevelopmentand Cultural Change29: 37-134.
Lokshin, Michael, and Martin Ravallion. 2000a. "Short-Lived
Shocks with Long-LivedImpacts? Household Income Dy-
namicsin a Transition Economy."World Bank,Development
ResearchGroup, Washington, D.C.
. 2000b. "WelfareImpact of Russia's 1998 Financial Cri-
sis and the Responseof the Public SafetvNet." World Bank,
Development ResearchGroup, Washington, D.C.
Londofno,Juan Luis,and MiguelSzekely.1997. PersistentlPoverty
and ExcessInequality:Latin America, 1970-1995. Washing-
ton, D.C.: Inter-American Development Bank.
Lopez, Ramon, and Carla della Maggiora. Forthcoming. "Rural
Poverty in Peru: StylizedFacts and Analytics for Policy." In
Ramon Lopezand Alberto Valdes,eds., RuralPovertyin Latin
America:Analytics,New Empinrical Evidenceand Policy.Wash-
ington, D.C.: World Bank.
Lopez, Ramon, and Alberto Valdes. Forthcoming. "Fighting
Rural Poverty in Latin America:New Evidence and Policy."
In Ramon Lopez and Alberto Valdes, eds., Rural Povertyin
Latin America:Analytics,New EmpiricalEvidenceand Policy.
Washington, D.C.: World Bank.
Lopez,Ramon,T. Thomas, and Vinod Thomas. 1998."Economic
Growth and the Sustainabiliryof NaturalResources."University
of Maryland, Department of Agriculturaland Resource Eco-
nomics, College Park.

Lora,Eduardo,and FelipeBarrera.1997.Unadecadade reformas
estructurales
enAmericaLatina:elcrecimiento, laproductividad
y lainversidn,ya nosoncomoantes.Inter-American Develop-
ment Bank, Office of the Chief Economist,Washington,
D.C.
Loury,GlennC. 2000."SocialExclusion and EthnicGroups:The
Challengeto Economics." DiscussionPaperSeries,no. 106.
BostonUniversity,Institutefor EconomicDevelopment.
Lovei,Laszlo,EugeneGurenko,MichaelHaney,PhilipO'Keefe,
and Maria Shkaratan.2000. "MaintainingUtility Services
for the Poor:PoliciesandPracticesin Centraland EasternEu-
ropeand theFormerSovietUnion."WorldBank,Europeand
CentralAsiaRegion,Washington,D.C.
Lowell,PeggyA.,and CharlesH.Wood.1998."SkinColor,Racial
Identityand LifeChancesin Brazil."LatinAmericanPerspec-
tives25(3):90-109.
Luckham,Robin. 1999."ComplexPoliticalEmergenciesand
theState:Failureandthe Fateofthe State."ThirdWorldQuar-
terly20(1):27-50.
Lund, Frances,and SmitaSrinivas.1999a."Learningfrom Ex-
perience:A Frameworkfor SocialProtectionforWorkersin
the InformalEconomy."Backgroundpaperfor WorldDe-
velopmentReport2000/2001.World Bank,Washington. D.C.
.1999b. "Learningfrom Experience:A GenderedApproach
to Social Protection for Workers in the Informal Economy."
Background paper for workshop on social protection for
women in the informal sector, International Labour Organi-
zation, 6-8 December, Geneva.
Lundberg,Mattias,and Lyn Squire.2000. "Inequalityand Growth:
Lessonsfor Policy."World Bank, Global Development Net-
work, Washington, D.C.
Lundberg,Shelley,RobertA. Pollak, and TerenceJ. Wales. 1997.
"Do Husbandsand WivesPoolTheir Resources?Evidencefrom
the United Kingdom Child Benefit."Journal of 'Human Re-
sources32(3): 463-80.
Lustig,Nora. 1988."From Structuralismto Neostructuralism:The
Searchfor a HeterodoxParadigm."In PatricioMeller,ed., The
LatinAmericanDevelopmentDebate.Boulder,Colo.:Westview.
. 1997. "The SafetyNets Which Are Not SafetyNets: So-
cial Investment Funds in Latin America." Paper presented at
Harvard Institute for International Development and United
Nations Development Programmeconferenceon governance,
povertyeradication,and socialpolicy,November,Cambridge,
Mass.
- 1998. Mexico: The Remaking of an Economy. 2d ed.
Washington, D.C.: Brookings Institution.
Forthcoming. "Crises and the Poor: SociallyResponsi-
ble Macroeconomics."Economia,Journal of theLatin Ameri-
canand CaribbeanEconomicAssociation1(1).
, ed. 1995. CopingwtithA usterity:Povertyand Inequality
in Latin America.Washington, D.C.: Brookings Institution.
,ed. Forthcoming.ShieldingthePoor:SocialProtectionin
the DevelopingWorld.Washington, D.C.: BrookingsInstitu-
tion.
Lustig, Nora, and Darryl McLeod. 1997. "Minimum Wages and
Povertyin DevelopingCountries: Some Empirical Evidence."
In SebastianEdwards and Nora Lustig,eds., LaborMarketsin
BIBLIOGRAPHIC NOTE 247
LatinAmerica.Washington.D.C.: BrookingsInstitution.
Lustig,Nora,and MiguelSzekely.
1998.Economic Trends,
Poverty
andInequalityin Mexico.Washington,D.C.: Inter-American
DevelopmentBank.
Maddison,Angus. 1987. "Growth and the Slowdownin Ad-
vancedCapitalistEconomies: Techniquesof QuantitativeAs-
sessment." JournalofEconomicLiterature
25(2):649-98.
. 1995. "Monitoringthe World Economy,1820-92."
Organisationfor EconomicCo-operationand Development,
DevelopmentCentre,Paris.
Magalhaes,Antonio,and MichaelH. Glantz. 1992. Socioecs-
nomicImpactsof ClimateVariations and PolicyResponses in
Brazil.Brasilia:FundacaoGrupoEsquelBrasil.
Mahar,J. Michael,ed. 1972.TheUntouchables in Contemporary
India.Tucson:Universityof ArizonaPress.
Maizels,Alfred,and MachikoK. Nissanke.1984."Motivations
for Aid to DevelopingCountries."WorldDevelopment 12:
879-900.
Mallick,Ross.1998.Development, EthnicityandHumanRights
in SouthAsia.New Delhi:Sage.
MalmbergCalvo,Christina.1998. Optionsfor Managingand
Financing RuralTransportInfrastructure.
World BankTech-
nical Paper 411. Washington, D.C.
Malthus, Thomas. 1985 [1798].An Essayon thePrincipleof Pop-
ulation. London: Penguin Classics.
Maluccio, John, LawrenceHaddad, and Julian May. 1999. "So-
cialCapitaland IncomeGenerationin SouthAfrica,1993-98."
International Food Policy Research Institute, Washington,
D.C.
Mankiw, N. Gregory, David Romer, and David N. Weil. 1992.
"A Contribution to the Empirics of Economic Growth."
QuarterlyJournal of Economics107(2): 407-37.
Manning, Daniel S. 1999. "The Role of Legal ServicesOrgani-
zations in Attacking Poverty." Backgroundpaper for World
DevelopmentReport 2000/2001. World Bank, Washingtcn,
D.C.
Manor, James. 1999. The PoliticalEconomyof DemocraticDe-
centralization.Directionsin DevelopmentSeries.Washingtcn,
D.C.: World Bank.
Marches Tropicauxet Mediterraneens.1995. "Une evaluation de
la crise des transports en Afrique sub-saharienne." 2612:
2629-31.
Marr, Ana. 1999. "The Poor and Their Money: What Have We
Learned?"ODI PovertyBriefing4. OverseasDevelopment In-
stitute, London.
Marshall,Adriana. 1998. "State Intervention, the Labor Market,
and Inequalityin Argentina."In AlbertBerry,ed., Poverty,E-o-
nomicReformandIncomeDistributionin LatinAmerica.Lon-
don: LynneRienner.
Martin, Will, and Keith E. Maskus.2000. "Core Labor Standards
and Competitiveness:Implicationsfor Global Trade Policy."
World Bank, Development Research Group, Washington,
D.C.; and Universityof Colorado, Boulder.
Maskus, Keith E. 1997. "Should Core Labor Standards Be Imn-
posed through International Trade Policy?" PolicyResearch
Working Paper 1817. World Bank, Washington, D.C.
. 1999. "Comments on Core Labor Standards and Inter-
248 WORLD DEVELOPMENT REPORT 2000/2001
national Trade Policy."Paper presented at World Bank sym-
posium, 20 January, Washington, D.C.
-2000. "Regulatory Standards in the WTO: Comparing
Intellectual Property Rights with Competition Policy, Envi-
ronmental Protection, and Core Labor Standards." Working
Paper 00-1. Institute for International Economics,Washing-
ton, D.C.
Massey, Douglas, and Karin Espinosa. 1997. "What's Driving
Mexico-U.S. Migration? A Theoretical, Empirical, and Pol-
icy Analysis."Americanjournal of Sociology102(4): 939-99.
Matin, Imran, David Hulme, and Stuart Rutherford. 1999.
"Financial Services for the Poor and Poorest: Deepening
Understanding to Improve Provision." Finance and Devel-
opment Research Program Working Paper 9. University of
Manchester, Institute for Development Policy and
Management.
Mauro, Paolo. 1995. "Corruption and Growth." QuarterlyJour-
nal of Economics110(3): 681-712.
.1998. "Corruptionand the Composition of Government
Expenditure." Journal of PublicEconomics69: 263-79.
Maveneke,Taparandava N. 1998. "LocalParticipation as an In-
strument for Natural ResourcesManagementunder the Com-
munal Areas Management Programme for Indigenous
Resources(CAMPFIRE) in Zimbabwe." Paper presented at
internationalworkshop on community-basednaturalresources
management, World Bank, 10-14 Ma', Washington, D.C.
Mc Clymont, Mary, and Stephen Golub, eds. 2000. Many Roads
toJustice:The Law-RelatedWork of FordFoundationGrantees
around the World.Washington, D.C.: Ford Foundation.
McCulloch, Andrew, and HeatherJoshi. 1999. "Child Develop-
ment and Family Resources:An Explorationof Evidencefrom
the Second Generation of the 1958 British Birth Cohort."
Working Paper 99-115. Universityof Essex,Institute for So-
cial and Economic Research.
McDonald, Calvin, Christian Schiller,and Kenichi Ueda. 1999.
"Income Distribution, Informal Safety Nets, and Social Ex-
penditures in Uganda." IMF Working Paper 99/163. Inter-
national Monetary Fund, Washington, D.C.
McGee, Rosemary. 1999. "Meeting the International Develop-
ment Targets: What Are the Prospects and Key Challenges?
Uganda Country Study." Christian Aid, London.
McGuire, Paul B., John Conroy, and Thapa Ganesh. 1998. Get-
ting theFrameworkRight:Policyand Regulation for Microfinance
in Asia.Brisbane:Foundation for Development Cooperation.
McKernan, Signe-Mary. 1996. "The Impact of Micro-Credit
Programs on Self-EmploymentProfits: Do Non-credit Pro-
gram Aspects Matter?" Brown University,Providence, R.I.
Meade, Donald C., and Carl Liedholm. 1998. "The Dynamics
of Micro and Small Enterprises in Developing Countries."
World Development26(1):61-74.
Mearns, Robin. 1999. "Accessto Land in Rural India." PolicyRe-
searchWorking Paper 2123. World Bank,Washington, D.C.
Meerman, Jacob. 1997. ReformingAgriculture:The World Bank
GoestoMarket. Washington, D.C.: World Bank, Operations
Evaluation Department.
1999. "SlowRoads to Equality:A Study of Four Hard-
Core Minorities-Issues from the Literature Review."Back-
ground paperfor WorldDevelopmentReport2000/2001.World
Bank, Washington, D.C.
Meier, Gerald M., and Dudley Seers,eds. 1984. Pioneersin De-
velopment.New York: Oxford UniversityPress.
Mejia, J. Antonio, and Rob Vos. 1997. "Poverty in Latin Amer-
ica and the Caribbean: An Inventory, 1980-95." Working
Paper Series, no. 1-4. Inter-American Development Bank,
Washington, D.C.
Mellor, John W. 1999. "Pro-poor Growth: The Relation be-
twveenGrowth in Agricultureand PovertyReduction."Abt As-
sociares,Bethesda,Md.
Menard, Claude, and George Clarke. 2000. "Reforming Water
Supply in Abidjan, Cote d'Ivoire: A Mild Reform in a Tur-
bulent Environment." World Bank, Development Research
Group, Washington, D.C.
Michalopoulos,Constantine. 1999. "DevelopingCountries' Par-
ticipation in the World Trade Organization." WorldEconomy
22(1): 117-44.
Milanovic, Branko. 1998. Income,Inequality,and Povertyduring
the Transition from PlannedtoMarketEconomy.New York:Ox-
ford University Press.
. 1999. "True World Income Distribution, 1988 and
1993." Policy ResearchWorking Paper 2244. World Bank,
Washington, D.C.
MIMAP (MicroImpactsof MacroeconomicAdjustmentPolicies).
1999a. "After the Crisis:A Look at the Elementary and Sec-
ondary Enrolment in the Philippines." MIMAP Indicators
6(2). [wwvw.pins.ph.net/mimap/v620699c.htm].
. 1999b. "An Analysisof the Social Impact of the Finan-
cial Crisis in the Philippines." MIMAP ResearchResults6(1).
[www.pins.ph.net/mimap/v610399a.htm].
. 1999c. "National Statistics Office (NSO) in MIMAP
Database." [www.pins.ph.net/mimap/eindune.htm].
Mingat, Alain, and Jee-Peng Tan. 1998. "The Mechanics of
Progressin EdLication:Evidence from Cross-Country Data."
Policy Research Working Paper 2015. World Bank, Wash-
ington, D.C.
Ministerio de Economfa de Argentina. 1998. Informeeconomico
28. BuenosAires.
Minot, Nicholas W. 1998. "Generating DisaggregatedPoverty
Maps: An Application to Vietnam." MSSD DiscussionPaper
25. InternationalFood PolicyResearchInstitute,Washington,
D.C.
MNA Live Database. "Human Development Indicators: Coun-
try at a Glance." World Bank, Washington, D.C.
Moctezuma, Esteban. 1999. Keynote address by Mexican Min-
ister of Social Development, Inter-American Development
Bank Conferenceon SocialProtection and Poverty,February,
Washington, D.C.
Molinas,Jose R. 1998. "The Impact of Inequality,Gender, Ex-
ternal Assistanceand Social Capital on Local-LevelCoopera-
tion." WorldDevelopment26(3): 413-31.
Monnet, Jean. 1988. Memoires.Paris: Fayard.
Moock, Peter. 1994. "Education and AgriculturalProductivity."
InternationalEncyclopediaof Education1. Oxford: Pergamon.
Moomaw, William, and Mark Tullis. 1999. "Population, Af-
fluence or Technology?An Empirical Look at National Car-

bon DioxideProduction."In BarbaraSundbergBaudotand
WilliamMoomaw,eds.,Peopleand TheirPlanet.NewYork:
St. Martin's.
Moore,Mick, and JamesPutzel. 1999."Politicsand Poverty."
Backgroundpaperfor WorldDevelopment Report2000/2001.
SussexUniversity,Brighton;LondonSchoolof Economics and
PoliticalScience,DevelopmentStudiesInstitute; andWorld
Bank,Washington,D.C.
Moore,Mick,JenniferLeavy,PeterP. Houtzager,and Howard
White. 1999. "PolityQualities:How GovernanceAffects
Poverty."Backgroundpaperfor WorldDevelopment Report
2000/2001.World Bank,Washington,D.C.
Morduch,Jonathan.1990."Risk,Production,and Saving:The-
ory and Evidencefrom Indian Households."HarvardUni-
versity,Departmentof Economics,Cambridge,Mass.
- 1999a."Betweenthe Marketand State:Can Informal
InsurancePatchtheSafetyNet?"WorldBankResearch Observer
14(2):187-207.
- 1999b."DoesMicrofinanceReallyHelp the Poor?Ev-
idencefrom FlagshipProgramsin Bangladesh."Princeton
University,Departmentof Economics,Princeton,N.J.
1999c."TheMicrofinance Promise."JournalofEconomic
Literature37(4): 1569-614.
-. 1999d."TheRoleof Subsidies in Microfinance: Evidence
fromthe GrameenBank."JournalofDevelopment Economics
60(1):229-48.
Morley,SamuelA. 1994.PovertyandInequalityin LatinAmer-
ica:PastEvidence andFutureProspects.
PolicyEssay13.Wash-
ington,D.C.: OverseasDevelopmentCouncil.
- 1999."ImpactofReformson Equityin LatinAmerica."
Backgroundpaperfor WorldDevelopmentReport2000/2001.
World Bank,Washington,D.C.
Morley,Samuel,and CarolaAlvarez.1992."Recessionand the
Growth of Poverty in Argentina." Working Paper 92-W02.
Vanderbilt University,Nashville,Tenn.
Morris, David. 1979. Measuring the Conditions of the World's
Poor:The PhysicalQualityof Life Index. Oxford: Pergamon.
Morris,Saul, OscarNeidecker-Gonzales,CalogeroCarletto,Mar-
cial Munguia, and Quentin Wodon. 2000. "Impact of Hur-
ricane Mitch in the Poorest Communities of Honduras."
International Food Policy Research Institute, Washington,
D.C.
Morrison,Andrew,and Maria BeatrizOrlando. 1999. "Socialand
EconomicCosts of DomesticViolence:Chile and Nicaragua."
In Andrew Morrison and Maria Loreto Biehl, eds., TooClose
to Home: Domestic Violencein the Americas.Washington,
D.C.: Inter-American Development Bank.
Moser, Caroline. 1998. "The Asset Vulnerability Framework:
ReassessingUrban Poverty Reduction Strategies." WorldDe-
velopment26(1): 1-19.
Mosley,Paul, and Marion J. Eeckhout. 2000. "From Project Aid
to ProgrammeAssistance."In Finn Tarp, ed., ForeignAid and
Development:LessonsLearntand Directionsfor theFuture.Lon-
don: Routledge.
Mosley, Paul, Jane Harrigan, and John Toye. 1995. Aid and
Power.London: Routledge.
Mukherjee, M., and G.S. Chatterjee. 1974. "On the Validity of
BIBLIOGRAPHIC NOTE 249
NSSEstimatesofConsumptionExpenditure." In T.N. Srini-
vasanand P.K.Bardhan,eds., PovertyandIncomeDistriba-
tionin India.Calcutta:IndianStatisticalInstitute.
Murthi, Mamta,Anne-CatherineGuio, and Jean Dreze.1995.
"Mortality,FertilityandGenderBiasin India:A District-Level
Analysis."Development EconomicResearchDiscussion Paper
61. LondonSchoolof Economicsand PoliticalScience.
Murty, Mamta.1994."Managementof CommonProperryRe-
sources:Limits to VoluntaryCollectiveAction."Environ-
mentalandResource Economics4(6).
Musgrove, Philip. 1996.PublicandPrivateRolesin Health:Tlhe-
oryand FinancingPatterns.World Bank DiscussionPaper
339.Washington,D.C.
Mwabu,Germano,MarthaAinsworth,and AndrewNyamere.
1993."Qualityof MedicalCareand Choiceof MedicalTreat-
mentin Kenya:AnEmpiricalAnalysis." AfricaTechnicalDe-
partmentTechnicalPaper9. WorldBank,Washington,D.C.
Narayan,Deepa.1998. "ParticipatoryRuralDevelopment."In
ErnstLutz,ed.,Agriculture andtheEnvironment. Washington,
D.C.: WorldBank.
. 1999."Bondsand Bridges: SocialCapitaland Poverty."
PolicyResearchWorkingPaper2167. World Bank,Wash-
ington,D.C.
Narayan,Deepa,and DavidNyamwaya.1996."Learningfrom
the Poor:A ParticipatoryPovertyAssessment in Kenya."En-
vironmentDepartmentPaper034.Participation Series.World
Bank,Washington,D.C.
Narayan,Deepa,and LantPritchett.1999."Centsand Sociabil-
ity:HouseholdIncomeandSocialCapitalin RuralTanzania."
Economic Development and CulturalChange 47(4):871-97.
Narayan,Deepa,andTalatShah.1999."GenderInequity,Poverty
and SocialCapital."World Bank, PovertyReductionand
EconomicManagementNetwork,Washington,D.C.
Narayan,Deepa, Robert Chambers,Meera K. Shah, and Patti Pe-
tesch. 1999. "Global Synthesis:Consultations with the Poor."
Paper presented at Global SynthesisWorkshop,World Bank,
Poverty Group, 22-23 September,Washington, D.C.
-. 2000. VoicesofthePoor:CryingOutfor Change.NewYork:
Oxford UniversityPress.
Narayan, Deepa, with Raj Patel, Kai Schafft,Anne Rademacher,
and Sarah Koch-Schulte.2000. Voicesof the Poor:Can Any-
one Hear Us?New York: Oxford University Press.
National Intelligence Council. 2000. "The Global Infectious
DiseaseThreat and Its Implicationsfor the United States."Na-
tional IntelligenceEstimate 99-17D. Washington, D.C.
Nelson,Joan. 1999. ReformingHealthandEducation: The World
Bank, the Inter-AmericanDevelopmentBank,and ComplexIn-
stitutionalChange.PolicyEssay26. Washington, D.C.: Oser-
seas Development Council.
Nepal,Department of Roads.1997. "EnvironmentalManagement
Guidelines." Geo-Environment Unit, Kathmandu.
Newman, John, Steen Jorgensen, and Menno Pradhan. 1991.
Workers'Benefits from Bolivias EmergencySocialFund. Living
Standards Measurement Study Working Paper 77. Washing-
ton, D.C.: World Bank.
Ng, Francis,and SandyYeats.1996. "Open EconomiesWork Bet-
ter! Did Africa'sProtectionist PoliciesCause Its Marginaliza-
250 WORLD DEVELOPMENT REPORT 200012001
tion in World Trade?" PolicyResearchWorking Paper 1636.
World Bank, Washington, D.C.
Nigerian Economic Society.1997. PovertyAlleviationin Vigeria.
SelectedPapersfor the 1997 Annual Conference.University of
lbadan.
Niles, KimberlyJ. 1999. "Economic Adjustment and Targeted
Social Spending: The Role of Political Institutions (In-
donesia,Mexico, and Ghana)." Backgroundpaper for World
Development Report 2000/2001. U.K. Department for In-
ternational Development, L.ondon;and World Bank, Wash-
ington, D.C.
Nolan, Brian, and Vincent Turbat. 1995. Cost Recoveryin Pub-
lic Health Services in Sub-Saharan Africa. Economic Devel-
opment Institute Technical Materials. Washingron, D.C.:
World Bank.
North, Douglass. 1990. Institutions,Institutional Change,and
Economic Perfbrmance.New York: Cambridge University
Press.
1994. "Economic Performancethrough Time." Amer-
ican EconomicReview84(3): 359-68.
Norton, Seth. 1998. "Poverty,PropertyRights and Human Well-
Being:A Cross-NationalStudy." Catojournal1 8(2):233-450.
Nyonator, Frank, and Joseph Kutzin. 1999. "Health for Some?
The Effects of User Fees in the Volta Region of Ghana."
Health Policy and Planning 14(4): 329-41.
Oates, Wallace E. 1999. "An Essayon FiscalFederalism."Jour-
nal ofEconomicLiterature37: 1120-49.
Ocampo, Jose Antonio, and LanceTayLor.1998. "Trade Liber-
alization in Developing Economies: ModestBenefits but Prob-
lems with Productivity Growth, Macro Prices, and Income
Distribution." Center for Economic PolicyAnalysisWorking
Paper 8. New School for SocialResearch, New York.
Odaka. Konosuke, and Juro Teranishi. 1998. Markets and Gop-
ernment:In SearchofBetter Coordination.Tokyo: Maruzen.
OECD (Organisation for Economic Co-operation and Devel-
opment). 1996. Trade, Employment, and Labor Standards: A
Study ofCore Workers'Rights and International Trade. Paris.
1999. DAC ScopingStudy of Donor PovertyReduction Poli-
cies and Practices. London: Overseas Development Institute.
OECD (Organisation for Economic Co-operation and Devel-
opment), DAC (Development Assistance Committee). 1996.
Shapingthe21st Century: The Contribution ofDevelopment Co-
operation. Paris.
- 1997. DAC Guidelines on Conflict, Peace, and Develop-
ment Cooperation. Paris.
- 1999. Development Cooperation:Efforts and Policiesofthe
Members ofthe DevelopmentAssistance Committee-1998 Re-
port. Paris.
- 2000. "Development Cooperation: 1999 Report." DAC
Journal 2000 1(1).
Ohiorhenuan, John. 1998. "Capacity Building Implications ofEn-
hanced African Participation in Global Rules-Making and
Arrangements." Paper prepared for the collaborative research
project Africa and the World Trading System, African Eco-
nomic Research Consortium, Nairobi.
Okrasa, Wlodzimierz. 1999. "Who Avoids and Who Escapes
from Poverty during the Transition? Evidence from Polish Panel
Data, 1993-96." PolicyResearchWorking Paper 2218. World
Bank, Washington, D.C.
Olinto, Pedro, Benjamin Davis, and Klaus Deininger. 1999.
"Did the PoorBenefitfrom Land MarketLiberalizationin Mex-
ico?Panel-DataEvidenceof the Impact of the EjidoReforms."
World Bank, Latin America and the Caribbean Region and
Development ResearchGroup; and International Food Pol-
icy ResearchInstitute, Washington, D.C.
Olson, Mancur. 1969. "The Principle of 'Fiscal Equivalence':
The Division of Responsibilitiesamong Different Levelsof
Government." American Economic Review Papers and Pro-
ceedings59: 479-87.
. 1982. TheRiseandDeclineofNations:EconomicGrowth,
Stag/lation,and Social Rigidities.New Haven, Conn.: Yale
UniversityPress.
Olsson,Jonathan. 2000. "Mitigation Is the CornerstoneofEmer-
gencyManagement in the United States." Boxfor WorldDe-
velopmentReport2000/2001. World Bank, Washington, D.C.
"Openness,Macroeconomic Crisis and Poverty." 1999. Rappor-
teur's report of a dialogue and consultation on World Devel-
opmentReport2000/2001,World Bank Institute and Institute
of Strategic and International Studies Malaysia, May, Kuala
Lumpur.
Osiptel (Organismo Supervisorde Inversi6n Privada en Teleco-
municaciones). 1996. La Apertura del mercado de las teleco-
municacionesen el Peru.Lima.
Ostrom, Elinor. 1997. "Investingin Capital, Institutions,and In-
centives."In ChristopherCtague,ed., Institutionsand Economic
Development: Growth and Governance in Less-Developed and
Post-SocialistCountries.Baltimore, Md.: Johns Hopkins Uni-
versity Press.
. Forthcoming. "The Puzzleof Counterproductive Prop-
erty Rights Reforms: A Conceptual Analysis." In Alain de
Janvry, Gustavo Gordillo, Jean-Philippe Platteau, and Elisa-
beth Sadoulet, eds., Accessto Land, Rural Povertyand Puhlic
Action. New York: Oxford University Press.
Ostrom, Elinor, L. Schroeder, and S. Wy'nne. 1993. Institutional
Incentives and Sustainable Development. Infrastructure Policies
in Perspective. Boulder, Colo.: Westview.
Over, Mead. 1998. "The Effects of Societal Variables on Urban
Rates of HIV Infection in Developing Countries: An Ex-
ploratory Analysis." In Martha Ainsworth, Lieve Fransen, and
Mead Over, eds., Confronting AIDS: Evidence from the De-
veloping World. Luxembourg: European Commission.
Owen, Daniel, and Julie van Domelen. 1998. "Getting an Ear-
ful: A Review of Beneficiary Assessments of Social Funds." So-
cial Protection Discussion Paper 9816. World Bank,
Washington, D.C.
Owens, Trudy. 2000. "The Determinants of Income Growth in
Rural Households and the Role of Aid: A Case Study of Zim-
babwe." Ph.D. diss. Oxford University.
Oxfam International. 1999. Education Now: Break the Cycle of
Poverty. Oxford.
Oxford Analytica. 1998a. "Central America: Hurricane Impact."
10 November. [ww-w.oxfam.com/index.html].
. 1998b. "Indonesia: Ethnic-Chinese Plight." 15 July.
[www.oxfam. com/index.html].

1999. "Indonesia: New Management." 25 October.
[www.oxfam.com/index.html].
Pack, Howard. 1999. "Poverty-ReducingPolicyReform." Paper
presented at WorldDevelopmentReport2000/2001 Summer
Workshop, World Bank, July, Washington, D.C.
Paerregaard,Karsten. 1998. "AlleviatingPoverty in Latin Amer-
ica: Can Local Organizations Be of Any Help?" Working
Paper 9810. Center for Development Research,Copenhagen.
PAHO (Pan AmericanHealth Organization). 1998. Healthin the
Americas1998. Washington, D.C.
Palacios, Robert, and Montserrat Pallares-Miralles.1999. "In-
ternational Patterns of Pension Provision."Discussion paper.
World Bank, Social Protection Department, Washington,
D.C.
Palley,Thomas. 1999."The EconomicCasefor InternationalLabor
Standards:Theoryand Some Evidence."AmericanFederation
of Labor and Congress of Industrial Organizations (AFL-
CIO), Washington, D.C.
Pande, Rohini. 1999. "Minority Representation and Policy
Choices: The Significanceof Legislator Identity." London
Schoolof Economicsand PoliticalScience.
Papageorgiou, Demetris, Armeane Choksi, and Michael
Michaely. 1995. "Liberalizing Foreign Trade in D)eveloping
Countries: The Lessonsof Experience."World Bank, Wash-
ington, D.C.
Park, Kyung Ae. 1990. "Women and Revolution in China: The
Sources of Constraints on Women's Liberation." Koreaand
WorldAffairs 14(4).
Parker, Andrew, and Rodrigo Serrano. Forthcoming. Promoting
GoodLocal Governancethrough SocialFunds and Decentral-
ization.Washington, D.C.: World Bank.
Parker, Edith A., Amy J. Schulz,BarbaraA. Israel,andlRose Hol-
lis. 1998. "Detroit's East Side Village Health Worker Part-
nership: Community-BasedLay Health AdvisorIntervention
in an Urban Area." Health Education and Behaviour 25(1):
24-45.
Parker,Roland, Randall Riopelle,and William Steel. 1995. Small
EnterprisesAdjustingto Liberalizationin FiveAfricanCountries.
World Bank Discussion Paper 271. Washington. D.C.
Parthasarthy,Gupta. 1991. "LeaseMarket, PovertyAlleviationand
PolicyOptions."EconomicandPolitical Weekly26(13):A31-38.
Patrinos, Harry Anthony, and David Ariasingam. 1997. Decen-
tralization of Education:Demand-SideFinancing.Directions
in Development Series.Washington, D.C.: World Bank.
Patten, Richard, and Jay Rosengard. 1991. ProgresswvithProfits:
TheDevelopmentofRural Bankingin Indonesia.San Francisco:
ICS Press.
Paugam, Serge,ed. 1996. L Exclusionletatdessavoirs.Paris: Edi-
tions La Decouverte.
Paul, Samuel. 1998. "Making Voice Work: The Report Card on
Bangalore'sPublic Services."Policy ResearchWorking Paper
1921. World Bank, Washington, D.C.
Paul, Samuel,and M. Shan. 1997. "Corruprion in Public Service
Delivery." In Sanjivi Guhan and S. Paul, eds., Corruptionin
India: AgendaforAction. New Delhi: New Delhi Press.
Pawasuthipaisit,Anan, Sombat Sakuntasathien,and Robert M.
Townsend. 1999. "Report to the Ford Foundation: Impact of
BIBLIOGRAPHIC NOTE 25[
the Thai Crisis."Backgroundpaper for WorldDevelopment Re-
port 2000/2001. World Bank, Washington, D.C.
Paxton, Julia, and Carlos Cuevas. 1996. A WorldwideInventory
of Microfinance Institutions. Sustainable Banking with the
Poor Project. Washington, D.C.: World Bank.
Pearce, David, and others. 1996. "The Social Costs of Climate
Change." In James Bruce, Hoesung Lee, and Erik Haites,
eds., Climate Change1995: Economicand SocialDimensiorns
of Climate Change-Contribution of WorkingGroupIII to the
SecondAssessmentReport of the IntergovernmentalPanel on
Climate Change.New York:Cambridge UniversityPress.
Pencavel,John. 1997. "The LegalFramework for CollectiveBat-
gaining in DevelopingCountries." In SebastianEdwards and
Nora Lustig, eds., LaborMarkets in LatinAmerica.Washing-
ton, D.C.: Brookings Institution.
Perotti, Roberto. 1992. "Income Distribution: Politics and
Growth." AmericanEconomicReview82: 311-16.
. 1993. "Political Equilibrium, Income Distribution, and
Growth." ReviewofEconomicStudies60: 755-76.
- 1996a. "Growth, Income Distribution, and Democ-
racy: What the Data Say." Journal of EconomicGrowth ]:
149-87.
. 1996b. "Redistributionand Non-consumption Smooth-
ing in an Open Economy." Review of EconomicStudies 63:
411-33.
Perry,Guillermo, and Daniel Lederman. 1999.Adjustmentsafer
SpeculativeAttacks in Latin Americaand Asia:A Tale of Two
Regions?Washington,D.C.: World Bank.
Persley,GabrielleJ., andJohnJ. Doyle. 1999. "Biotechnologyfor
Developing-Country Agriculture:Problems and Opportuni-
ties-Overview." Focus 2, Brief 1. International Food Policy
ResearchInstitute, Washington, D.C.
Persson, Torsten, and Guido Tabellini. 1994. "Is Inequality
Harmful for Growth?"AmericanEconomicReview84: 600-2 L.
Pessino, Carola. 1997. "Argentina: The Labor Market during
the Economic Transition." In Sebastian Edwards and Nora
Lustig,eds.,LaborMarketsin LatinAmerica.Washingcon,D.C.:
Brookings Institution.
Pfaff, AlexanderS.P. 1996. "What Drives Deforestation in the
Brazilian Amazon? Evidence from Satellite and Socioeco-
nomic Data." Columbia University, Department of Eco-
nomics, New York.
Phipps, ShelleyA., and Peter S. Burton. 1998. "What's Mine Is
Yours?The Influenceof Male and FemaleIncomes on Patterns
of Household Expenditure." Economica65: 599-613.
Piketty, Thomas. 1998. "Self-FulfillingBeliefsabout Social Sta-
tus."Journal of PublicEconomics70: 115-32.
. Forthcoming. "Theories of Persistent Inequality and
Intergenerational Mobility." In Anthony B. Atkinson and
FrancoisBourguignon,eds., HandbookofIncomeDistribution.
New York: North Holland-Elsevier.
Pinstrup-Andersen,Per, Rajul Pandya-Lorch, and Mark Rosen-
grant. 1997. "The World Food Situation: Recent Develop-
ments, Emerging Issues, and Long-Term Prospects." Food
Policy Report. International Food Policy ResearchInstitute,
Washington, D.C.
Pitt, Mark, and Shahidur Khandker. 1998. "The Impact of
252 WORLD DEVELOPMENT REPORT 2000/2001
Group-Based Credit Programs on Poor Households in
Bangladesh:Does the Gender of Participants Matter?" Jour-
nal of PoliticalEconomy106: 958-96.
Platteau,Jean-Philippe,and Jean MarieBaland.Forthcoming."Im-
partible InheritanceversusEqual Division:A ComparativePer-
spective Centered on Europe and Sub-Saharan Africa." In
Alainde Janvry, GustavoGordillo,Jean-PhilippePlatreau,and
ElisabethSadoulet,eds.,AccesstoLand,RuralPovertyand Pub-
licAction.New York: Oxford UniversityPress.
Polanyi, Karl. 1957. The Great Transformation.Boston: Beacon.
Poppele, Jessica, Sudarno Sumarto, and Lant Pritchett. 1999.
"SocialImpacts of the Indonesian Crisis: New Data and Pol-
icy Implications." SMERU Report. Social Monitoring and
Early ResponseUnit, Jakarta.
Porter, Gareth, Raymond Clemen,on, Waafas Ofosu-Amaah,
and Michael Philips. 1998. Study of GEF's Overall Perfor-
mance.Washington, D.C.: Global Environment Facility.
Portes, Alejandro. 1995. The EconomicSociologyoflmmigration:
EssaysonNetworks,Ethnicityand Entrepreneurship. New York:
Russell Sage Foundation.
- 1998. "Social Capital: Its Origins and Applications in
Contemporary Sociology." Annual Review of Sociology24:
1-24.
Portes,Alejandro,and Patricia Landolt. 1996. "The Downside of
Social Capital." American Prospect26: 18-21, 94.
Portes, Alejandro, and Julia Sensenbrenner. 1993. "Embedded-
ness and Immigration: Notes on the Social Determinants of
Economic Action." American Journal of Sociology98(6):
1320-50.
Posner, Daniel. 1999. "Ethnic Fractionalization:How (Not) to
MeasureIt? What Does (and Doesn't) It Explain?"Paper pre-
sented at annual meeting of American Political ScienceAsso-
ciation, Atlanta.
Posner,Richard. 1997."SocialNorms and the Law:An Economic
Approach." AmericanEconomicReview87(2): 365-69.
Postel, Sandra. 1999. "Water, Food and Population." In Barbara
Sundberg Baudot and William R. Moomaw, eds., Peopleand
TheirPlanet.New York:St. Martin's.
Pottebaum, David A. 1999. Economicand SocialImplicationsof
War and Conflict.Cornell University,Agricultural Econom-
ics Department, Ithaca, N.Y.
Pouliquen, Louis Y. 1999a. "Infrastructure and Poverty." Back-
ground paperfor WorldDevelopmentReport200012001.World
Bank, Washington, D.C.
- 1999b. "Rural Infrastructure from a World Bank Per-
spective: A Knowledge Management Framework." World
Bank, Environmentally and Socially Sustainable Develop-
ment Network, Washington, D.C.
Pradhan, Sanjay. 1996. "Public-PrivatePartnerships for Service
Provision." Backgroundpaper for WorldDevelopmentReport
1997. World Bank, Washington, D.C.
Prakash,S. 1997. "Povertyand Environment Linkagesin Moun-
tains and Uplands: Reflectionson the 'Poverty Trap'." Col-
laborative Research in the Economics of the Environment
and Development Working Paper Series, no. 12. Interna-
tional Instutute for Environment and Development, Envi-
ronmental Economics Programme, London; Institute for
Environmental Studies, Amsterdam; and Vrije Universiteit,
Amsterdam.
Prasad, Kameshwar,Paolo Belli,and Monica Das Gupta. 1999.
"LinksbetweenPoverty, Exclusion,and Health." Background
paper for WorldDevelopmentReport2000/2001. World Bank,
Washington, D.C.
Preker, Alexander. 2000. "Partnership: Private Sector Role in
Poverty and Health." Box for World Development Report
2000/2001. World Bank, Washington, D.C.
Preker,Alexander,and Richard Feachem. 1996. Market Mecha-
nismsand theHealthSectorin CentralandEasternEurope.World
Bank Technical Paper 293. Washington, D.C.
Prennushi, Giovanna. 1999. "Nepal: Povertyat the Turn of the
Twenty-First Century." South Asia Region Internal Discus-
sion Paper 174.World Bank, Washington, D.C.
Prescott,Nicholas,and Menno Pradhan. 1999."Coping with Cat-
astrophic Health Shocks."Paper presented at Inter-American
Development Bank conference on social protection and
poverty, 4-5 February,Washington, D.C.
Preston, Samuel H., and Michael R. Haines. 1991. Fatal Years:
ChildMortality in Late-Nineteenth-CenturyAmerica.Prince-
ton, N.J.: Princeton University Press.
Primo Braga,CarlosA., Carsten Fink, and ClaudiaPaz Sepulveda.
2000. IntellectualPropertyRightsand EconomicDevelopment.
World Bank Discussion Paper 412. Washington, D.C.
Prince of Wales BusinessLeaders Forum. 1998. Building Com-
petitivenessand Communities.London.
Pritchett,Lant. 1996 a. "MindYour P's and Q's: The CostofPub-
lic Investment Is Not the Value of Public Capital." PolicyRe-
searchWorking Paper 1660. World Bank, Washington, D.C.
-. 1996b."Where HasAll the Education Gone?"PolicyRe-
searchWorking Paper 1581. World Bank,Washington, D.C.
. 1997. "Divergence,BigTime."Journal ofEconomic Per-
spectives11: 3-17.
Pritchett, Lant, and LawrenceSummers. 1994. "Desired Fertil-
ity and the Impact of Population Policies."Populationand De-
velopmentReview20(1): 1-55.
. 1996. "Wealthier Is Healthier." Journal ofHuman Re-
sources31(4): 841-68.
PROBE Team. 1999. PublicReporton BasicEducationin India.
New Delhi: Oxford University Press.
Prunier, Gerard. 1997. The Rwanda Crisis:Historyofa Genocide.
New York:Columbia UniversityPress.
Przeworski, Adam, and Fernando Limongi. 1993. "Political
Regimes and Economic Growth." Journal of EconomicPer-
spectives7:51-69.
Psacharopoulos,George,and Harry Anthony Patrinos. 1994. In-
digenousPeopleand Povertyin Latin America:An Empirical
Analysis.Washington, D.C.: World Bank.
Pulley,RobertV. 1989. MakingthePoorCreditworthy: A CaseStudy
of the IntegratedRural DevelopmentProgramof India. World
Bank Discussion Paper 58. Washington, D.C.
Putnam, Robert. 2000. BowlingAlone: The Collapseand Revival
ofAmerican Community.New York: Simon and Schuster.
Putnam,Robert,with RobertLeonardiandRaffaellaNanetti.1993.
Making DemocracyWork: Civic Traditionsin Modern Italy.
Princeton, N.J.: Princeton Universiry Press.

Putzel,James. 1997."Accountingfor the 'Dark Side' of SocialCap-
ital: Reading Robert Putnam on Democracy."Journalof In-
ternationalDevelopment9(7): 939-49.
Qian, Yingyi,and Barry R. Weingast. 1996. "China's Transition
to Markets:Market-PreservingFederalism,ChineseStyle."Jour-
nal of PolicyReform2: 149-85.
Quah, Danny. 1997. "Empirics of Growth and Distribution:
Stratification, Polarization,and ConvergenceClubs." Centre
for Economic Performance Discussion Paper 324. London
School of Economics and Political Science.
.1999. " 6 x 109:Some Dynamicsof Global Inequalityand
Growth." Background paper for World DevelopmentReport
2000/2001. London School of Economics and Political Sci-
ence, Suntory and Toyota International Centres for Eco-
nomicsand RelatedDisciplines;and World Bank,Washington,
D.C.
Quisumbing, AgnesR. 1994. "Improving Women's Agricultural
Productivity as Farmers and Workers." Education and Social
PolicyDepartment Discussion Paper 37. World Bank, Wash-
ington, D.C.
- . 1996. "Male-Female Differences in Agricultural Pro-
ductivity: Methodological Issues and Empirical Evidence."
World Development24: 1579-95.
Quisumbing, Agnes R., and John A. Maluccio. 1999. "Intra-
household Allocation and Gender Relations:New Empirical
Evidence." Background paper for forthcoming World Bank
Policy Research Report EnGendering Development.World
Bank, Washington, D.C.
Ramamurti, R. 1999. "Why Haven't DevelopingCountries Pri-
vatized Deeper and Faster?" WorldDevelopment27: 137-56.
Ramey, Garey, and Valerie A. Ramey. 1995. "Cross-Country
Evidenceon the Link betweenVolatilityand Growth." Amer-
icanEconomicReview85(5): 1138-51.
Ranis, Gustav. 1979. "AppropriateTechnology in a Dual Econ-
omy: Reflectionon Philippineand Taiwanese Experience."In
EdwardA. Robinson, ed., AppropriateTechnologies for Third
World Development.New York: St. Martin's.
Ranis, Gustav, and FrancesStewart. 1994. "Decentralization in
Indonesia." Bulletin of Indonesian EconomicStudies 30(3):
41-72.
Rao,J. Mohan. 1999. "Equity in a Global Public Goods Frame-
work." In Inge Kaul, IsabelleGrunberg, and Marc Stern, eds.,
Global Public Goods:International Cooperationin the 21st
Century.New York:Oxford University Press.
Rao,Vijayendra.1998a."DomesticViolenceand Intra-household
ResourceAllocationin Rural India: An Exercisein Participa-
tory Econometrics." In Maithreyi Krishnaraj, Ratna M. Su-
darshan, and Abusaleh Shariff, eds., Gender,Populationand
Development.New York: Oxford UniversityPress.
- 1998b. "Wife-Abuse, Its Causes and Its Impact on
Intra-household Resource Allocation in Rural Karnataka:A
Participatory Econometric Analysis." In Maithreyi Krish-
naraj, Ratna M. Sudarshan, and AbusalehShariff, eds., Gen-
der,PopulationandDevelopment.New York:Oxford University
Press.
Rauch,James, and Peter Evans. 1999. "BureaucraticStructure and
BureaucraticPerformance in Developing Countries." Work-
BIBLIOGRAPHIC NOTE 253
ing Paper 99-06. Universityof California at San Diego, De-
partment of Economics.
Rauch, James, and Vitor Trindade. 1999. "Ethnic Chinese Ner-
works in International Trade." NBER Working Paper 7189.
National Bureau of Economic Research,Cambridge, Mass.
Ravallion,Martin. 1988. "Expected Povertyunder Risk-Induced
WelfareVariability."EconomicJournal 98: 1171-82.
. 1991. "Reaching the Rural Poor through Public Em-
ployment: Arguments, Evidence and Lessons from South
Asia." World Bank ResearchObserver6(2): 153-76.
. 1992. Poverty Comparisons: A Guide to Conceptsand
Methods.LivingStandardsMeasurementStudyWorking Paper
88. Washington, D.C.: World Bank.
. 1993.PovertyComparisons: FundamentalsofPureandAp-
plied Economics.Chur, Switzerland: Harwood Academic.
. 1997a. "Can High-Inequality Developing Countries
EscapeAbsolute Poverty?"EconomicsLetters56(1): 51-57.
. 1997b. "Good and Bad Growth: The Human Devel-
opment Reports." WorldDevelopment25: 631-38.
. 1998. "Does AggregationHide the Harmful Effects of
Inequality on Growth?" EconomicsLetters61(1): 73-77.
. 1999a."AppraisingWorkfare." WorldBankResearch Ob-
server14(1): 31-48.
. 1999b. " Is More TargetingConsistentwith LessSpend-
ing?" International Tax and Public Finance6(3): 411-19.
. 1999c. "On Protecting the Poor from Fiscal Contrac-
tions." World Bank, Development Research Group, Wash-
ington, D.C.; and Universitedes SciencesSociales,ARQADE,
Toulouse.
. 2000a."How LongBeforeMost of the PoorLivein Urban
Areas?"World Bank, Development ResearchGroup, Wash-
ington, D.C.
. 2000b. "Inequality Convergence."World Bank, Devel-
opment ResearchGroup, Washington, D.C.
. Forthcoming."MonitoringTargetingPerformanceWhen
DecentralizedAllocationsto the PoorAre Unobserved."World
Bank EconomicReview.
Ravallion,Martin, and Shubham Chaudhuri. 1997. "Riskand In-
surance in Village India: A Comment." Econometrica65:
171-84.
Ravallion,Martin, and Shaohua Chen. 1997. "What Can New
Survey Data Tell Us about Recent Changes in Distribution
and Poverty?" WorldBank EconomicReview 11(2): 357-82.
Ravallion,Martin, and Gaurav Datt. 1996. "How Important t-o
India'sPoor Is the SectoralCompositionof EconomicGrowth?'
WorldBank EconomicReview 10(1): 1-25.
. 1999. "When Is Growth Pro-Poor? Evidence from the
DiverseExperiencesof India's States."PolicyResearchWork-
ing Paper 2263. World Bank, Washington, D.C.
Ravallion, Martin, and Binayak Sen. 1994. "The Impacts en
Rural Poverty of Land-Based Targeting: Further Results for
Bangladesh."WorldDevelopment22(6):823-38.
Ravallion,Martin, and Dominique van de Walle. 1991. "Urban-
Rural Cost-of-Living Differentials in a Developing Econ-
omy."Journalof UrbanEconomics29(1): 113-27.
Ravallion,Martin, and Quentin Wodon. 1999. "Poor Areas, or
Only Poor People?"Journal ofRegionalScience39: 689-711.
254 WORLD DEVELOPMENT REPORT 2000/2001
- 2000. "Does Child Labor Displace Schooling? Evidence
on Behavioral Responses to an Enrollment Subsidy." Eco-
nomicJournal I 10: 158-75.
. Forthcoming. "Banking on the Poor? Branch Placement
and Nonfarm Rural Development in Bangladesh." Review of
Development Economics.
Rawlings, Laura, Lynne Sherburne-Benz, and Julie van Domelen.
Forthcoming. "Evaluating Social Fund Performance across
Countries." World Bank, Human Development Network,
Washington, D.C.
Reardon, Thomas. 1997. "Using Evidence ofHousehold Income
Diversification to Inform Study of the Rural Nonfarm Labor
Market in Africa." WorldDevelopment25(5): 735-47.
Reardon, Thomas, and J. Edward Taylor. 1996. "Agroclimatic
Shock, Income Inequality, and Poverty: Evidence from Burk-
ina Faso." World Development24(5): 901-14.
Reardon, I'homas, Christopher Delgado, and Peter Matlon.
1992. "Determinants and Effects of Income Diversification
amongst Farm Households in Burkina Faso." Journal of 'De-
velopment Studies 28(2): 264-96.
Reardon, 'Thomas, J. Edward Taylor, Kostas Stamoulis, Peter
Lanjouw, and Arsenio Balisacan. Forthcoming. "Effects of
Non-farm Employment on Rural Income Inequality in De-
veloping Countries: An Investment Perspective." Journal of
Agricultural Economics.
Reinikka, Ritva, andJakob Svensson. 1999. "Confronting Com-
petition: Investment Response and Constraints in Uganda."
Policy Research Working Paper 2242. World Bank, Wash-
ington, D.C.
Remenyi, Joc. 1991. Where Credit Is Due: Income-Generating
Programmes for the Poor in Developing Countries. London: In-
termediate Technologies.
Ren, X.S. 1996. "Regional Variation in Infant Survival in China."
Social Biologyl43(1-2): 1-19.
Renner, Michael. 1994. "Budgeting for Disarmament: The Costs
of War and Peace." Worldwatch Paper 122. Worldwatch In-
stitute, Washington, D.C.
Reno, William. 1998. Warlord Politics and African States. Boul-
der, Colo.: Lynne Rienner.
Revenga, Ana. 1997. "Employment and Wage Effects of Trade
Liberalization: The Case of Mexican Manufacturing." Jour-
nal oflLahor Economics 15(3): 20-43.
Re'es, Cecilia, and Anne Bernadette E. Mandap. 1999. "The So-
cial Impact of the Regional Financial Crisis in the Philip-
pines." Micro Impact of Macroeconomic and Adjustment
Policies Research Paper Series, no. 41. Policy and Development
Foundation, Manila.
Rhyne, Elisabeth, and Sharon Holt. 1994. Women in Finance and
Enterprise Development. Washington, D.C.: World Bank.
Riddell, Roger C., and Mark Robinson, with John de Coninck,
Ann Muir, and Sarah White. 1995. Non-governmental Orga-
nizations and Rural PovertyAlleviation. NewYork: Oxford Uni-
versity Press.
Rios-Rull, Jose Victor. 1994. "Population Changes and Capital
Accumulation: The Aging of the Baby Boom." Universiry of
Pennsylvania, Department of Economics, Philadelphia.
Robbins, Donald J. 1995. "Should Educational Spending Be Re-
distributed from Higher to Primary Education in LDCs? A
Note with Application to Chile." Revista deAnalisis Economico
10(1): 37-51.
. 1996. "Evidence on Trade and Wages in the Develop-
ing World." Technical Paper 119. Organisation for Economic
Co-operation and Development, Paris.
Robinson, James A. 1996. "When Is a State Predatory?" Univer-
siry of Southern California, Department of Economics, Los
Angeles.
Robinson, Peter. 1993. "Economic Effects ofthe 1992 Drought
on the Manufacturing Sector in Zimbabwe." Overseas De-
velopment Institute, London.
Robinson, Richard, and David Stiedl. 2000. "Decentralisation of
Road Administration: Review of Experience." U.K. Depart-
ment for International Development, London.
Rodgers, Gerry. 1979. "Income and Inequalitv as Determinants
of Mortality: An International Cross-Section Analysis." Pop-
ulation Studies 33: 343-51.
Rodgers, Gerry, Charles G. Gore, and Jose B. Figueiredo. 1995.
SocialExclusion: Rhetoric, Reality, Responses.Geneva: Inrerna-
tional Institute for Labor Studies.
Rodriguez, Francisco. 1999. "Inequality, Economic Growth and
Economic Performance." Background paper for World De-
velopmentReport2000/2001. World Bank, Washington, D.C.
Rodriguez, Francisco, and Dani Rodrik. 1999. "Trade Policy
and Economic Growth: A Skeptic's Guide to the Cross-
National Evidence." Universiry of Maryland, Department of
Economics, College Park.
Rodrik, Dani. 1996. "Labor Standards in International Trade: Do
They Matter and What Do We Do about Them?" In Robert
Z. Lawrence, Dani Rodrik, and John Whalley, eds., Emerg-
ingAgenda for Global Trade: High Stakes for Developing Coun-
tries. Policy Essay 20. Washington, D.C.: Overseas
Development Council.
. 1997. "Trade Policv and Economic Performance in
Sub-Saharan Africa." Harvard University, John F. Kennedy
School of Government, Cambridge, Mass.
. 1998. "W here Did All the Growth Go? External Shocks,
Social Conflicts, and Growth Collapses." NBER Working
Paper 6350. National Bureau of Economic Research, Cam-
bridge, Mass.
. 1999a. "Institutions for High-Quality Growth: What
Thev Are and How to Acquire Them." Paper presented at In-
ternational Monetary Fund conference on second-generation
reforms, 8-9 November, Washington, D.C.
. 1999b. The New Global Economy and the Developing
Countries:Making OpennessWork. Policy Essay 24. Washington,
D.C.: Overseas Development Council.
. 1999c. "Why Is There So Much Economic Insecurity
in Latin America?" NBER Working Paper 6350. National Bu-
reau of Economic Research, Cambridge, Mass.
. 2000. "Institutions for High-Qualitv Growth: What
They Are and How to Acquire Them." NBER Working Paper
7540. National Bureau of Economic Research, Cambridge,
Mass.
Roemer, John E. 1996. Theories of Distrihutive Justice. Cam-
bridge, Mass.: Harvard University Press.

1998. EqualityofOpportunity. Cambridge, Mass.:Har-
vard University Press.
Roger, Neil. 1999. "RecentTrends in Private Participationin In-
frastructure." Viewpoint 196. World Bank, Finance, Private
Sector, and Infrastructure Network, Washington, D.C.
Ros, Jaime. 2000. Development Theory and the Economicsof
Growth.Ann Arbor: University of Michigan Press.
Rose,Elaina. 1999. "ConsumptionSmoothing and ExcessFemale
Mortality in Rural India." Reviewof Economicsand Statistics
81(1): 41-49.
Rose, Richard. 1995. "Russia as an Hour-Glass Society:A Con-
stitutionwithout Citizens." EuropeanConstitutionalReview4(3):
34-42.
- 1999. "Getting Things Done in Anti-modern Society:
Social Capital Networks in Russia."Social Capital Initiative
Working Paper 6. World Bank, SocialDevelopment Depart-
ment, Washington, D.C.
Rose-Ackerman,Susan. 1997. "Corruption and Development."
In Boris Pleskovicand Joseph E. Stiglitz,eds., Annual World
Bank Conferenceon DevelopmentEconomics1992. Washing-
ton, D.C.: World Bank.
Rosen, George.1993. A HistoryofPublic Health.Baltimore,Md.:
Johns Hopkins UniversityPress.
Rosenzweig,Mark. 1998. "SocialLearningand EconomicGrowth:
Empirical Evidence." Background paper for World Develop-
ment Report 1998/99. World Bank, Washington, D.C.
Rosenzweig,Mark,and Hans Binswanger.1993. "Wealth,Weather
Risk and the Composition and Profitabilityof AgriculturalIn-
vestment." EconomicJournal103: 56-78.
Rowntree, Benjamin Seebohm. 1901. Poverty:A Study of Town
Life. London: Macmillan.
Royaume du Maroc, Ministere de la PrevisionEconorniqueet du
Plan, Direction de la Statistique.1999. "Enquete nationalesur
les niveaux de vie des menages 1998/99: Rapport de syn-
these."
Rozelle, Scott. 1996. "Stagnation without Equity: Patterns of
Growth and Inequality in China's Rural Economy." China
Journal 3: 63-91.
Rubin, Donald B. 1987. Multiple Imputationfor Nonresponsein
Surveys.New York:John Wiley & Sons.
Rubio, Mauricio. 1997. "PerverseSocial Capital: Some Evidence
from Colombia." Journal of EconomicIssues31(3): 805-16.
Rugalema,Gabriel. 1999. "Adult Mortality as Entitlement Fail-
ure: AIDS and the Crisis of Rural Livelihoodsin a Tanzan-
ian Village." Ph.D. diss. Institute of Social Studies, The
Hague.
Ruprah, Inder. 1999. "Towards Fiscal Prudence, Transparency
and Accountabilityin Peru: A Proposal."Inter-AmericanDe-
velopment Bank, Division 5 Regional Operations Depart-
ment, Washington, D.C.
Ruprah, Inder, and Luis Marcano. 1998. "Poverty Alleviationin
Venezuela:Who to Target and How Not to Adjust in a Cri-
sis." Inter-AmericanDevelopment Bank, Division 5 Regional
Operations Department, Washington, D.C.
- 1999. "Digginga Hole: IncomeInequalityin Venezuela."
Inter-AmericanDevelopment Bank, Division 5 RegionalOp-
erations Department, Washington, D.C.
BIBLIOGRAPHIC NOTE 255
Rutkowski, Michal. 1999. "Russia's Social Protection Malaise."
SocialProtection DiscussionPaper 9909. World Bank,Wash-
ington, D.C.
Sable,Charles,Dara O'Rourke, and Archon Fung. 2000. "Ratch-
eting Labor Standards: Regulation for Continuous Improve-
ment in the Global Workplace." World Bank, Human
Development Network, Washington, D.C.
Sachs,Jeffrey.1999. "Sachson Development:Helping the World's
Poorest." Economist,14-20 August, pp. 17-20.
. 2000. "Science, the Global Division of Labor, and In-
ternational Public Policy." Keynoteaddress at Annual World
Bank Conferenceon Development Economics, 18-20 April,
Washington, D.C.
Sachs,Jeffrey,andAndrewWarner. 1995. "EconomicReformand
the Process of Global Integration." BrookingsPaperson Eco-
nomicActivity 1: 1-117.
Sachs,Jeffrey,Michael Kremer,and Amar Hamoudi. 1999. "The
Case for a Vaccine Purchase Fund." Center for International
DevelopmentPolicyPaper 1. Harvard University,Cambridge,
Mass.
Sachs,Jeffrey,KwesiBotchwey,MaciejCuchra, and SaraSievers.
1999."ImplementingDebt Relieffor the HighlyIndebtedPoor
Countries." Center for International Development Policy
Paper 2. Harvard University, Cambridge, Mass.
Sadoulet, Elisabeth, Rinku Murgai. and Alain de Janvry. Forth-
coming. "Accessto Land via Land Rental Markets." In Alain
de Janvry, GustavoGordillo,Jean-PhilippePlatteau, and Elis-
abeth Sadoulet, eds., Accessto Land, Rural Povertyand Public
Action. New York: Oxford University Press.
Sah, Raaj K. 1991. "Fallibilityin Human Organizations and Po-
litical Systems."Journal ofEconomicPerspectives 5(2): 67-88.
Sahn, David E., and David C. Stifel.1999."Poverty Comparisons
overTime and acrossCountries in Africa."Working Paper 95.
Cornell University, Food and Nutrition Policy Program,
Ithaca, N.Y.
Sahn, David E., Paul A. Dorosh, and Stephen D. Younger.1997.
StructuralAdjustment Reconsidered.New York: Cambridge
UniversityPress.
Sahn, David E., David C. Stifel,and Stephen D. Younger. 1999.
"Inter-temporalChangesin Welfare:PreliminaryResultsfrom
Nine AfricanCountries." Cornell University,Departmentsof
Economicsand Nutritional Science,Ithaca, N.Y.
Saito, Katrine, and Daphne Spurling. 1992. DevelopingAgricul-
tural Extensionfor Women Farmers.World Bank Discussion
Paper 156. Washington, D.C.
Saito, Katrine, Hailu Mekonen, and Daphne Spurling. 1994.
Raising the Productivity of Women Farmersin Sub-Saharan
Africa.World Bank DiscussionPaper 230. Washington, D.C.
Sakurai,Takeshi. 1997. CropProductionunderDroughtRisk and
EstimationofDemandforFormalDroughtInsurancein theSahel.
Tokyo: Ministry of Agriculture, National ResearchInstitute
of AgriculturalEconomics, Forestry and Fisheries.
Salmen,LawrenceF. 1995. "ParticipatoryPovertyAssessment:In-
corporatingPoor People'sPerspectivesinto PovertyAssessment
Work." World Bank, EnvironmentDepartment,Washington,
D.C.
Sampson, Robert, Jeffrey Morenhoff; and Felton Earls. 1999. "Be-
256 WORLD DEVELOPMENT REPORT 2000/2001
yondSocialCapital:SpatialDynamicsof CollectiveEfficacy
for Children."AmericanSociological Review64(5):633-60.
Sanchez,Susana.2000. "NicaraguaFinancialMarkets."Back-
groundpaperfor NicaraguaPoverty Assessment
2000.World
Bank,LatinAmericaandthe CaribbeanRegion,Washington,
D.C.
SandIer,Todd. 1997. GlobalChallenges: An Approachto Envi-
ronmental, Political,
andEconomic Problems.
NewYork:Cam-
bridgeUniversityPress.
- 1998."Globaland RegionalPublicGoods:A Progno-
sisfor CollectiveAction."FiscalStudies19(3):221-47.
- 1999. "Intergenerational PublicGoods:Strategies,Ef-
ficiency,and Institutions."In IngeKaul,IsabelleGrunberg,
and MarcStern,eds.,GlobalPublicGoodsInternationalCo-
operationin the21st Century.NewYork:OxfordUniversity
Press.
Sara,Jennifer,andTravisKatz.1997."MakingRuralWaterSus-
tainable:Reporton the Impactof ProjectRules."UnitedNa-
tionsDevelopmentProgrammeand WorldBankWater and
SanitationProgram,Washington,D.C.
Sarris,Alexander.1994."HouseholdWelfareduringCrisisand
Adjustmentin Ghana."Journalof AfricanEconomies 2(2):
195-237.
Satterthwaite,David.2000. "UrbanUpgradingand Water and
SanitationServicesin Low-IncomeAreas."Box for World
DevelopmentReport2000/2001. World Bank,Washington,
D.C.
Sauma,Pablo.1997."CostaRica:A PublicApproach."In Elaine
Zuckerman andEmanuelde Kadt,eds.,ThePublic-Private Mix
in SocialServices:Health Careand Education in Chile, Costa
Rica and Venezuela.Washington, D.C.: Inter-American De-
velopment Bank.
Saunders,RobertJ.,JeremyJ. Warford,and BjornWellenius.1983.
Telecommunicationsand EconomicDevelopment.Baltimore,
Md.: Johns Hopkins University Press.
Savedoff,William, and Pablo Spiller. 1999. "SpilledWater: In-
stitutional Commitment in the Provision of Water Services."
Inter-AmericanDevelopment Bank, Washington, D.C.
Sawadogo,Kimseyinga. 1997. "La pauvrete au Burkina Faso:
Une analysecritique des politiques et des strategies d'inter-
vention locales."Working Paper 57. European Centre for
Development Policy Management, Maastricht.
Scandizzo,PasqualeL. 1998. "Growth, Trade, and Agriculture:
An InvestigativeSurvey." FAO Economic and Social Devel-
opment Paper 143. Food and Agriculture Organizationof the
United Nations, Rome.
Schalick, Lisa,Wilbur Hadden, Elsie Pamuk, Vicente Navarro,
and GregoryPappas.2000. "The Widening Gap in Death Rates
among Income Groups in the United States from 1967 to
1986." InternationalJournalof Health Services30(1): 13-26.
Schiavo-Campo,Salvatore, Giulio de Tommaso, and Amitabha
Mukherjee. 1997a. "An International Statistical Survey of
GovernmentEmploymentand Wages."PolicyResearchWork-
ing Paper 1806. World Bank, Washington, D.C.
1997b. "Government Employment and Pay:A Global
and Regional Perspective." Policy ResearchWorking Paper
1771. World Bank, Washington, D.C.
Schick,Allen.1998."WhyMost DevelopingCountriesShould
Not Try NewZealand'sReforms."WorldBankResearch Ob-
server13(1):123-31.
Schieber,George,ed. 1997. Innovationsin HealthCareFinanc-
ing:Proceedings of a WorldBank Conference,
March10- 1,
1997.WorldBankDiscussionPaper365.Washington,D.C.
Schieber,George,and AkikoMaeda.1997. "A Curmudgeon's
Guideto FinancingHealth Carein DevelopingCountries."
In GeorgeSchieber,ed., Innovationsin HealthCareFinanc-
ing:Proceedings of a WorldBank Conference,
March10-11,
1997.WorldBankDiscussionPaper365. Washington,D.C.
Schiff,Maurice.1998."EthnicDiversityand EconomicReform
in Sub-SaharanAfrica."Journalof AfricanEconomies 7(2):
348-62.
Schilderman, Theo. 1993."Disastersand Development:A Case
Study from Peru."Journalof InternationalDevelopment 5:
415-23.
Schneider,RobertR. 1995. Government andtheEconomy on the
AmazonFrontier. WorldBankEnvironmentPaper11.Wash-
ington,D.C.
Schofield,Roger,David Reher, and AlainBideau.1991. The
DeclineofMortalityin Europe.Oxford:Clarendon.
Schraeder,PeterJ., StevenHook,and BruceTaylor.1998."Clar-
ifyingthe ForeignAid Puzzle:A Comparisonof American,
Japanese,French,and SwedishAid Flows."WorldPolitics
50(2): 294-320.
Schuler,Margaret,and SakuntalaKadirgamar-Rajasingham,eds.
1992. LegalLiteracy:A Toolfor Women'sEmpowerment.
Washington,D.C.: OEF International,Women, Law,and
Development.
Schultz, T. Paul. 1994. "Human Capital, Family Planning, and
Their Effects on Population Growth." American Economic
Review84(2): 255-60.
. 1999. "PreliminaryEvidence of the Impact of Progresa
on School Enrollment from 1997 and 1998." International
Food Policy ResearchInstitute, Washington, D.C.
Scoones, Ian. 1995. "Investigating Difference: Applications of
Wealth Ranking and Household SurveyApproaches among
Farming Households in Southern Zimbabwe." Development
and Change26(1): 67-88.
Search,Leila.1999. "Tanzania:Report on a Village Immersion."
Box for World DevelopmentReport2000/2001. World Bank,
Washington, D.C.
SebelloMendoza, Meyra, and Mark W. Rosegrant. 1995. Pric-
ing Behaviorin PhilippineCornMarkets:ImplicationsforMar-
ket Efficiency.Washington, D.C.: International Food Policy
ResearchInstitute.
Sebstad,Jennefer,and Monique Cohen. 1999."Microfinance,Risk
Managementand Poverty." Backgroundpaper for WorldDe-
velopmentReport2000/2001.U.S.Agencyfor InternationalDe-
velopment and World Bank, Washington, D.C.
Secretariade Programaci6n Econ6mica y Regional. 1998. Car-
acterizaci6ny evoluciondelgastopuiblicosociaLBuenos Aires:
Ministerio de Economia y Obras y Servicios.
Sell, Susan. 1996. "North-South Environmental Bargaining:
Ozone, Climate Change, and Biodiversity." Global Gover-
nance2: 97-118.

Sen,Amartya. 1984. "Poor,RelativelySpeaking."InAmartyaSen,
ed., Resources,Values and Development. New York: Basil
Blackwell.
- 1992. "Missing Women." British Medical Journal
304(6827): 587-88.
- . 1993. "Capability and Well-Being." In Martha Nuss-
baum and Amartya Sen, eds., The Quality of Life. Oxford:
Clarendon.
- 1997a. "Development Thinking at the Beginning of
the 21st Century." New Series,DERP-2. London School of
Economicsand PoliticalScience.
1997b. InequalityReexamined.Cambridge, Mass.:Har-
vard University Press.
1997c. On EconomicInequality.Oxford: Clarendon.
1999. DevelopmentasFreedom.New York: Knopf.
Sen, Amartya,and Jean Dreze. 1989. Hunger and PuhlicAction.
Oxford: Clarendon.
Serageldin,Ismail,and ChristiaanGrootaert. 2000. "Defining So-
cial Capital: An Integrating View." In Partha Dasgupta and
Ismail Serageldin, eds., Social Capital:A Multifaceted Per-
spective.Washington, D.C.: World Bank.
Serieux,John E. 1999. "Reducingthe Debt of the Poorest: Chal-
lengesand Opportunities." North-South Institute, Ottawa.
SEWA (Self-EmployedWomen's Association).1997. Liberaliz-
ingfor the Poor.Ahmedabad.
Seymour, FrancesJ., and Navroz K. Dubash, withJake Brunner,
FrancoisEkoko,Colin Filer,Hariadi Kartodihardjo,and John
Mugabe. 2000. The Right Conditions:The WorldBank, Struc-
turalAdjustment,andForestPolicyReform.Washington,D.C.:
World ResourcesInstitute.
Shah, Anwar.1999. "Balance,Accountabilityand Responsiveness:
Lessons about Decentralization." Policy ResearchWorking
Paper 2021. World Bank, Washington, D.C.
Shah, Shekhar. 1999. "Coping with Natural Disasters:The 1998
Floods in Bangladesh."Paper presented at WorldDevelopment
Report 2000/2001 Summer Workshop, World Bank, July,
Washington, D.C.
Sharma, Naresh, and Jean Dreze. 1996. "Sharecropping in a
North Indian Village."Journalof DevelopmentStudies27(2):
277-92.
Shaw, R. Paul, and Martha Ainsworth, eds. 1995. Financing
Health Servicesthrough UserFeesand Insurance:CaseStudies
from Sub-SaharanAfrica.World Bank Discussion Paper 294.
Washington, D.C.
Shkolnikov,Vladimir, Giovanni Andrea Cornia, David A. Leon,
and France Mesle. 1998."Causesof the RussianMortality Cri-
sis:Evidenceand Interpretations." WorldDevelopment26(11):
1995-2011.
Siegel, Paul, and Jeffrey Alwang. 1999. "An Asset-Based Ap-
proachto SocialRisk Management:A ConceptualFrameworkL"
SocialProtectionDiscussionPaper 9926. World Bank, Wash-
ington, D.C.
Silver, Hilary. 1994. "Social Exclusion and Social Solidarity:
Three Paradigms." International Labour Review 133(5-6):
531-78.
Sindzingre,Alice.2000. "A ComparativeAnalysisof African and
East Asian Corruption." In Arnold J. Heidenheimer and
BIBLIOGRAPHIC NOTE 257
Michael Johnston, eds., Political Corruption. London:
Transaction.
. Forthcoming. "Dimensions economiques des reformes
de l'etat en Afriquesub-saharienne."In Comi Toulabor and
Dominique Darbon, eds., Reformede l'etat:Reconstructionin-
stitutionnelleet modesde regulation.Paris: Karthala.
Singerman, Diane. 1995. Avenuesof Participation:Family,Poli-
tics,and Networksin Urban Quartersof Cairo.Princeton,N.J.:
Princeton UniversityPress.
Singh, Balmiki Prasad. 1999. "Democracy, Ecology and Cul-
ture: The Indian Experience."Sardar PatelMemorial Lecture,
31 October, New Delhi.
Sinha, Saurabh, and Michael Lipton. 1999. "Damaging Fluctu-
ations, Risk and Poverty: A Review." Backgroundpaper for
World Development Report 2000/2001. Sussex University,
Poverty Research Unit, Brighton; and World Bank, Wash-
ington, D.C.
Skocpol,Theda. 1992. ProtectingSoldiersand Mothers:The Po-
liticalOriginsof SocialPolicyin the UnitedStates.Cambridge,
Mass.:Harvard University Press.
Skoufias, Emmanuel, Benjamin Davis, and Jere R. Behrman.
1999. "An Evaluation of the Selection of BeneficiaryHouse-
holds in the Education, Health, and Nutrition Program (Pro-
gresa)of Mexico."InternationalFood PolicyResearchInstitute,
Washington, D.C.
Smale, Melinda, and Paul Heisey. 1994. "Maize Research in
MalawiRevisited:An EmergingSuccessStory?"Journalofln-
ternationalDevelopment6(6): 689-706.
Snodgrass,Donald R., and Richard H. Patten. 1989. "Reform of
Rural Credit in Indonesia: Inducing Bureaucraciesto Behave
Competitively."DevelopmentDiscussionPaper 315. Harvard
Institute for International Development, Cambridge, Mass.
Solo,Tova Maria. 1998a. "Competition in Water and Sanitation:
The Role of Small-Scale Entrepreneurs." Viewpoint 165.
World Bank, Finance,Private Sector,and Infrastructure Net-
work, Washington, D.C.
. 1998b. "KeepingParaguay'sAguateroson Stream." Wall
StreetJournal.27 November.
. 1999. "West African BusinessesPioneerWater and San-
itation Servicesto the Poor." Boxfor World DevelopmentRe-
port 2000/2001. World Bank, Washington, D.C.
Solow,Robert. 1956. "A Contribution to the Theoryof Economic
Growth." QuarterlyJournal of Economics70: 65-94.
Soobader,Mah-Jabeen, and FeliciaLeClere. 1999. "Aggregation
and the Measurement of Income Inequality:Effects on Mor-
bidity." SocialScienceand Medicine 48: 733-44.
Spohr, Chris. 2000. "Essayson Household and the Workforcein
Taiwan." MassachusettsInstitute of Technology,Department
of Economics, Cambridge, Mass.
Srinivas,Mysore. 1987. "The Dominant Caste in Rampura." In
Mysore Srinivas,ed., The Dominant Casteand OtherEssays.
New Delhi: Oxford University Press.
Srinivasan, T.N. 2000. "Growth, Poverty Reduction, and In-
equality."Paper presentedat Annual World Bank Conference
on Development Economicsin Europe, World Bank, 26-28
June, Paris.
Srinivasan,T.N., and Pranab K. Bardhan. 1974. PovertyandI n-
258 WORLD DEVELOPMENT REPORT 2000/2001

comeDistributionin India. Calcutta:StatisticalPublishing SoniyaCarvalho,KeneEzemenari,CarolGraham,and Alan

Society.
Steele,ClaudeM. 1999."Thin Ice:'StereotypeThreat'and Black
CollegeStudents."AtlanticMonthly284(2):44-54.
Stein,Ernesto.1999. 'FiscalDiscipline andSocialProtection:Are
TheyCompatible?" Inter-AmericanDevelopmentBank,Re-
searchDepartment,Washington,D.C.
Stein,Howard.1999."The Developmentof the Developmental
Statein Africa:A TheoreticalEnquiry."Paperpresentedat
African StudiesAssociationmeeting, 11-14November, Uni-
versicyof Pennsylvania, Philadelphia.
Steward,Fred. 1978. Technology and Underdevelopment. Lon-
don: Macmillan.
Stewart,Frances.1995. "The SocialImpactsof Globalization
and Marketization."In CUner Kirdarand LeonardSilk,eds.,
People:FromImpoverishment to Empowerment. NewYork:
NewYorkUniversitvPress.
Stewart,Frances,Frank P. Humphreys,and Nick Lea. 1997.
"CivilConflictin DevelopingCountriesoverthe LastQuar-
ter of a Century:AnEmpiricalOverviewofEconomicandSo-
cial Consequences."OxfordDevelopmentStudies25(1):11-41.
Stiglitz, Joseph E. 1998a. "Genderand Development:The Role
of the State." Paperpresentedat genderand development
workshop,2 April,WorldBank,Washington,D.C.
.1998b."The Roleof International FinancialInstitutions
in the CurrentGlobalEconomy."Addressto ChicagoCoun-
cil on ForeignRelations,Februarv,Chicago.
- 1998c."Towardsa NewParadigmfor Development:
Strategies,Policies,and Processes."
PrebischLectureat United
NationsConferenceon Trade and Development,October,
Geneva.
.1999. "QuisCustodietIpsosCustodes? CorporateGov-
ernanceFailurein theTransition."Keynoteaddressat Annual
WorldBankConferenceon DevelopmentEconomicsinEu-
rope,World Bank,21-23 June, Paris.
Stiglitz,JosephE., and HirofumiUzawa.1969.Readingsin the
ModernTheoryofEconomic Growth.Cambridge,Mass.:MIT
Press.
Stock,Elisabeth,andJan deVeen.1996.Expanding Labor-Based
Methodsfor RoadWorksin Africa.World BankTechnical
Paper347. Washington,D.C.
Strauss,John, and DuncanThomas.1998."Health,Nutrition,
and EconomicDevelopment." JournalofEconomic Literature
36: 766-817.
Streeten,Paul.1984."BasicNeeds:SomeUnsettledQuestions."
WorldDevelopment 12:973-82.
Streeten,Paul,ShahidJ.Burki,Mahbubul Haq,NormanHicks,
and FrancesStewart.1981.FirstThingsFirst:MeetingBasic
HumanNeedsintheDeveloping Countries.
Oxford:OxfordUni-
versityPress.
Subbarao, Kalanidhi.1997."PublicWorksasan Anti-poverty' Pro-
gram:An Overview of Cross-CountryExperience." American
JournalofAgricultural Economics
79: 678-83.
1998."Namibia'sSocialSafetyNet:Issuesand Options
for Reform."PolicyResearchWorkingPaper1996.World
Bank,Washington,D.C.
Subbarao,Kalanidhi,AniruddhaBonnerjee, JeanineBraithwaite,
Thompson.1997.SafetyNetPrograms andPovertyReduction:
Lessonsfrom Cross-Country Experience.
Directionsin Devel-
opmentSeries.Washington,D.C.: World Bank.
Summers,Lawrence H. 1992."Investingin Allthe People."Pak-
istanDevelopment Review31:367-93.
. 1993."Foreword."In ElizabethM. KingandM. Anne
Hill,eds., Women'sEducation in Developing
Countries:
Barri-
ers,Benefits,and Policies.Baltimore,Md.: Johns Hopkins
University Press.
Summers,Robert,and Alan Heston. 1991."The Penn World
Table(Mark5):An ExpandedSetof InternationalCompar-
isons, 1950-1988."Quarterly Journalof Economics106(2):
327-68.
Suryahadi,Asep,Sudarno Sumarto,YusufSuharso,and Lant
Pritchett.1999."TheEvolutionof Povertyduringthe Crisis
in Indonesia,1996to 1999."WorldBank,Washington,D.C.;
and SocialMonitoringand EarlyResponseUnit,Jakarta.
Swamy,AnandV., SteveKnack,YoungLee,and OmarAzfar.
1999. "Genderand Corruption."Universityof Maryland,
Center on InstitutionalReformand the InformalSector,Col-
lege Park.
Sweden,MinistryforForeignAffairs.1999."MakingPartnerships
Workon the Ground."Reportof a workshop,30-31 August,
UlvsundaCastle,Stockholm.
Szekely,Miguel,Nora Lustig,JoseAntonioMejia,and Martin
Cumpa.Forthcoming.HowMVany PoorPeople inLatinAmer-
icaAre ThereReally?Washington, D.C.: Inter-AmericanDe-
velopmentBank.
Szreter,Simon. 1997. "EconomicGrowth,Disruption,Depri-
vation,Diseaseand Death:On the Importanceof the Politics
of PublicHealth."Population andDevelopmentReview23(4):
693-728.
. Forthcoming."SocialCapital,the Economyand Edu-
cation in HistoricalPerspective."In StephenBaron,John
Field,and Tom Schuller,eds., SocialCapital:CriticalPer-
spectives.
Oxford:OxfordUniversitvPress.
Tacoli,Cecilia.1998. "Rural-UrbanInteractions:A Guide to
the Literature."EnvironmentandUrbanization10(1):147-66.
Tanzi, Vito. 1998. "FundamentalDeterminantsof Inequality
and the Roleof Government."IMFWorkingPaper98/178.
InternationalMonetaryFund,Washington,D.C.
. 1999."TheQualityof thePublicSector."
Paperpresented
at International Monetary Fund conferenceon second-
generationreforms,8-9 November,Washington,D.C.
Tanzi,Vito,and HamidDavoodi.1997."Corruption,PublicIn-
vestment,and Growth."IMFWorkingPaper97/139.Inter-
nationalMonetaryFund,Washington,D.C.
Temple,Jonathan.1998."InitialConditions,SocialCapital,and
GrowthinAfrica."Journal ofAfricanEconomics73(3):309-47.
. 1999."TheNewGrowthEvidence." JournalofEconomic
Literature 37: 112-56.
Temple,Jonathan,and PaulA.Johnson.1998."SocialCapabil-
ity and EconomicGrowth."Quarterly Journalof Economics
113(3):965-90.
Tendler,Judith. 1993.NewLessonsfrom OldProjects:
TheWork-
ings of RuralDevelopment in NortheastBrazil.Operations

EvaluationStudy.Washington,D.C.: WorldBank.
- 1997. GoodGovernance in the Tropics.Baltimore,Md.:
JohnsHopkinsUniversityPress.
Tendler,Judith,and SaraFreedheim. 1994."Trustina Rent-Seek-
ingWorld:Health and GovernmentTransformedin North-
eastBrazil."WorldDevelopment 22(12):1771-91.
Teranish,Juro. 1996."SectoralResourceTransfer,Conflictand
Macro-Stability in EconomicDevelopment:A Comparative
Analysis."In MasahikoAoki,Hyung-KiKim,and Masahiro
Okuno-Fujiwara,eds., TheRoleofGovernment in EastAsian
EconomicDevelopment: Comparative InstitutionalAnalysis.
NewYork:OxfordUniversityPress.
Ter-Minassian,Teresa. 1997.FiscalFederalism in Theoryand
Practice.Washington,D.C.: InternationalMonetaryFund.
TGNP (TanzaniaGenderNetworkingProgramme).1999.Bud-
getingwitha GenderFocus.Dar es Salaam.
Thomas,Duncan.1990."Intrahousehold Resource Allocation:
An
InferentialApproach."JournalofHuman Resources25:635-64.
- 1997."Incomes,Expenditures, and HealthOutcomes:
Evidence on IntrahouseholdResource Allocation."In Lawrence
Haddad,John Hoddinott,and HaroldAlderman,eds.,In-
trahousehold ResourceAllocation
in Developing Countries:Mod-
els,Methods,and Policy.Baltimore,Md.: Johns Hopkins
UniversityPress.
- . 1999."EconomicCrisisandPoverty:EvidencefromIn-
donesia."Universityof Californiaat LosAngeles,Depart-
mentof Economics.
Thomas,Duncan,VictorLavy,andJohn Strauss.1996."Public
PolicyandAnthropometricOutcomesin the C6ted'Ivoire."
JournalofPublicEconomics 61: 155-92.
Thomas,Vinod,and YanWang.1998."MissingLessonsof East
Asia:Openness,Education,and the Environment."Paper
presented at AnnualBankConference on Development in Latin
Americaandthe Caribbean,WorldBank,June,Montevideo.
Thorbecke,Erik.1985."TheSocialAccountingMatrixandCon-
sistency-Type PlanningModels."In GrahamPyattandJeffrey
I. Round,eds.,SocialAccounting Matrices:A Basisfor Planning.
Washington,D.C.: World Bank.
- 1991. "Adjustment,Growthand IncomeDistribution
in Indonesia."WorldDevelopment 19(11):1595-614.
.2000."The Evolutionof theDevelopmentDoctrineand
the Role of ForeignAid, 1950-2000."In Finn Tarp, ed.,
ForeignAidandDevelopment: Lessons LearntandDirectionsfor
theFuture.London:Routledge.
Thorbecke,Erik,andHong-SangJung. 1996."AMultiplierDe-
compositionMethodto AnalyzePovertyAlleviation." Jour-
nalofDevelopment Economics 48:279-300.
Timmer,C. Peter.1997."HowWellDo the PoorConnectto the
GrowthProcess?" ConsultingAssistance on EconomicReform
DiscussionPaper 178. HarvardInstitutefor International
Development,Cambridge,Mass.
Timmer,MarcelP., and AdamSzirmai.1997."Growthand Di-
vergencein ManufacturingPerformancein Southand East
Asia."ResearchMemorandumGd-37.Universityof Gronin-
gen,GroningenGrowthand DevelopmentCentre.
Tomich,ThomasP., and Meinevan Noodwijk,eds. 1998.Al-
ternativesto Slash-and-Burn in Indonesia:SummaryReport
BIBLIOGRAPHIC NOTE 259
andSynthesis ofPhaseH. Bogor:InternationalCentrefor Re-
searchin Agroforestry.
Tommasi,Mariano.1999."On High Inflationand the Alloca-
tionof Resources."JournalofMonetaryEconomics 44:401-21.
Tommasi,Mariano,and AndresVelasco.1996."WhereAreWe
inthe PoliticalEconomyof Reform?"Journal ofPolicyReform
1(2):187-238.
Tommasi,Mariano,and FedericoWeischelbaum. 1999."APrin-
cipal-AgentBuildingBlockfor the Studyof Decentralization
and Integration."HarvardUniversity,Cambridge,Mass.
Townsend,Peter. 1979.Povertyin the UnitedKingdom:A Sur-
veyof Household Resources
andStandardsofLiving.London:
AllenLane.
. 1985. "SociologicalApproachto the Measurementof
Poverty:A RejoindertoProfessorAmartyaSen."OxfordEco-
nomicPapers37:659-68.
Townsend,Robert.1994."Riskand Insurancein VillageIndia."
Econometrica 62(3):539-91.
. 1995."Consumption Insurance:An Evaluationof Risk-
BearingSystemsin Low-Income Economies." JournalofEco-
nomicPerspectives 9: 83-102.
Treisman,Daniel.1998."The Causesof Corruption:A Cross-
NationalStudy."UniversityofCaliforniaat LosAngeles,De-
partmentof PoliticalScience.
Tsui,KayYuen.1995."Multidimensional Generalizations
ofthe
Relativeand AbsoluteIndices:The Atkinson-Kolm-Sen Ap-
proach."JournalofEconomicTheory67:251-65.
. 1997. "Multidimensional PovertyIndices."Chinese
Universityof Hong Kong,Departmentof Economics.
Tuck, Laura,and Kathy Lindert.1996. From Universal Food
Subsidiestoa Self-TargetedProgram:
A CaseStudyin Tunisian
Reform.World Bank DiscussionPaper 351. Washington,
D.C.
Tummala,KrishnaK. 1999."Policyof Preference: Lessonsfrom
India,the UnitedStates,and SouthAfrica."PublicAdminis-
trationReview59(6):495-509.
Turok,Ben. 1999."SouthAfrica:FromExclusionto Integration."
Backgroundpaperfor WorldDevelopmentReport2000/2001.
World Bank,Washington,D.C.
Turton,David,ed. 1997. WarandEthnicity:GlobalConnections
andLocalViolence. Rochester,N.Y.:Universityof Rochester
Press.
Tzannatos,Zafiris.1998."ChildLaborand SchoolEnrollment
in Thailand in the 1990s."SocialProtectionPaper9818.
WorldBank,Washington,D.C.
Uberoi, Patricia.1999. "Gender and State Policiesin India,
1950-2000."World Bank,Washington,D.C.
Uddin,Sharif.1999."The'Village Phone':BringingUniversalSer-
viceto RuralAreasin Bangladesh."Paperpresentedat National
TelephoneCooperative Association's
FirstInternationalCon-
ferenceon RuralTelecommunications, November,Wash-
ington,D.C.
Udry,Christopher.1991."CreditMarketsin NorthernNigeria:
Creditas Insurancein a RuralEconomy."WorldBankEco-
nomicReview4(3):251-71.
. 1996."Gender,AgriculturalProduction,and theTheory
oftheHousehold."JournalofPoliticalEconomy104(5):1010-46.
260 WORLD DEVELOPMENT REPORT 2000/200I

. 1999."Povertv,Riskand Households."Paperpresented ficeofU.S.ForeignDisasterAssistance). 1999.International

at WorldDevelopmentlReport2000/2001 SummerWorkshop, Disaster Database.
Washington,D.C.
WorldBank,July,Washington,D.C. U.S.CensusBureau.1999.Poverty in theUnitedStates1998.Wash-
Udry, Christopher,John Hoddinott, Harold Alderman,and ington,D.C.: U.S.Departmentof Commerce.
Lawrence Haddad. 1995."GenderDifferentialsin FarmPro- Uvin,Peter.1995."FightingHungerat the Grassroots:Pathsto
ductivity:Implicationsfor HouseholdEfficiencyand Agri- ScalingUp." WorldDevelopment 23: 927-39.
culturalPolicy."FoodPolicy20(5):407-23. Valdes,Alberto.1994."AgriculturalReformsin Chileand New
UN (UnitedNations).1998.PovertyReduction Strategies:
A Re- Zealand:A Review." JournalofAgriculturalEconomics45(2):
view.NewYork. 189-201.
UNAIDSUoint United NationsProgrammeon HIV/AIDS). . 1999."ARuralPovertyProfileof theRegion."In Ramon
1998.AIDS EpidemicUpdate.Geneva. Lopezand AlbertoValdes,eds.,RuralPovertyin LatinAmer-
. 2000.ReportontheGlobalHIV/AIDSEpidemic. Geneva. ica:Analytics,NewEmpirical Evidence andPolicy.Washington,
UNDP(UnitedNationsDevelopment Programme). 1995.Human D.C.: World Bank.
Development Report1995.NewYork:OxfordUniversity Press. van de Walle,Dominique.1995."Towarda Synthesis."In Do-
1996. Human Development Report1996 New York: miniquevande WalleandKimberlyNead,eds.,PublicSpend-
OxfordUniversityPress. ingandthePoor:TheoryandEvidence. Baltimore,Md.:Johns
1997. Human Development Report1997 New York: HopkinsUniversitvPress.
OxfordUniversityPress. .1996. Infrastructure
andPoverty in Vietnam.LivingStan-
1998. Overcoming HumanPoverty.NewYork:United dardsMeasurement StudyWorkingPaper121.Washington,
Nations. D.C.: World Bank.
.1999a.HumanDevelopment Report1999.NewYork:Ox- . 2000a."AreReturnsto InvestmentLowerforthe Poor?
ford UniversityPress. Humanand PhysicalCapitalInteractionsin RuralVietnam."
.1999b.UnitedNationsDevelopmentProjectsProgramming World Bank, DevelopmentResearchGroup,Washington,
Manual.NewYork:UnitedNations. D.C.; and Universityof Toulouse,Departmentof Econom-
UNESCO (United NationsEducational,Scientific,and Cul- ics.
tural Organization).Database.[unescostat.unesco.org/enl . 2000b."ChoosingPro-poorRuralRoadInvestments."
statslstatsO.htm]. World Bank, DevelopmentResearchGroup,Washington,
UNFPA(UnitedNationsPopulationFund)and AustralianNa- D.C.
rionalUniversity.1998. "South East AsianPopulationsin . 2000c."HumanCapitalandLaborMarketConstraints
Crisis:Challengesto the Implementationof the ICPD Pro- in DevelopingCountries:A CaseStudyof Irrigationin Viet-
grammeof Action."UNFPA,NewYork. nam."World Bank,DevelopmentResearchGroup, Wash-
Unger,Danny.1998.BuildingSocialCapitalin Thailand:Fibers, ington,D.C.
Finance,andInfrastructure.
NewYork:CambridgeUniversity van de Walle, Dominique,and Dileni Gunewardena.2000.
Press. "Sourcesof Ethnic Inequality in Viet Nam." PolicyResearch
UNICEF (United Nations Children's Fund). 1994. Childrenat Working Paper 2297. World Bank, Washington, D.C.
Work.East Asia and PacificRegional Office. Bangkok. van de Walle, Dominique, and Kimberly Nead, eds. 1995. Pub-
UNICEF (United Nations Children's Fund) and Oxfam Inter- licSpendingand thePoor:Theoryand Evidence.Baltimore,Md.:
national. 1999. "Debt Relief and Poverty Reduction: Meet- Johns Hopkins UniversityPress.
ing the Challenge." Paper presented at HeavilyIndebted Poor vande Walle,Nicolas.Forthcoming. ThePoliticsofPermanentCri-
Countries ReviewSeminar, United Nations Commission for sis: Managing African Economies,1979-1999. New York:
Africa,July, Addis Ababa. [www.worldbank.orglhipc]. Cambridge University Press.
United Kingdom, Secretaryof State for International Develop- van de Walle, Nicolas,and Timothy Johnston. 1996. Improving
ment. 1997. EliminatingWorldPoverty:AChallengefjrthe21st AidtoAfrica. PolicyEssay21. Washington,D.C.: OverseasDe-
Centuy.White Paper on InternationalDevelopment.London: velopment Council.
Her Majesty'sStationery Office. van der Gaag, Jacques. 1995. PrivateandPublicInitiativesWork-
Uphoff, Norman. 1992. Learningfrom Gal Oya:Possibilitiesfor ing Togetherfor Health and Education.Directions in Devel-
ParticipatoryDevelopmentand Post-NewtonianSocialScience. opment Series.Washington, D.C.: World Bank.
Ithaca, N.Y.: Cornell UniversityPress. van Dijk, Meine Pieter, and Roberta Rabellotti. 1997. Enterprise
.1993. "GrassrootsOrganizationsand NGOs in RuralDe- Clustersand Networksin DevelopingCountries.London: Frank
velopment: Opportunities with Diminishing States and Ex- Cass.
panding Markets." WorldDevelopment21(4): 607-22. van Domelen, Julie, and Daniel Owen. 1998. "Getting an Ear-
- 1998. "Community-BasedNatural ResourcesManage- ful:A Reviewof BeneficiaryAssessmentsof SocialFunds." So-
ment: ConnectingMicroand Macro Processes,and Peoplewith cial Protection Discussion Paper 9816. World Bank,
Their Environments." Plenary presentation at International Washington, D.C.
Workshop on Community-Based Natural Resources Man- van Doorslaer,Eddy, and Adam Wagstaff.1997. "Income-Related
agement, World Bank, 10-14 May, Washington, D.C. Inequalities in Health: Some International Comparisons."
USAID (U.S.Agencyfor InternationalDevelopment),OFDA (Of- Journal ofHealth Economics16: 93-112.

vanGinneken,Wouter,ed. 1999.SocialSecurityfor theExcluded
Majority.Geneva:InternationalLabourOffice.
vanNieuwkoop,Martien,and JorgeE. Uquillas.2000. "Defin-
ingEthnodevelopment in OperationalTerms:Lessonsfrom
the EcuadorIndigenousand Afro-Ecuadoran PeoplesDevel-
opmentProject."Sustainable Development WorkingPaper6.
WorldBank,LatinAmericaandthe CaribbeanRegion,Wash-
ington,D.C.
van Rijckeghem,Caroline,and BeatriceWeder.1997."Corrup-
tionand the RateofTemptation:Do LowWagesin CivilSer-
vice Cause Corruption?" IMF Working Paper 97173.
InternationalMonetaryFund,Washington,D.C.
vanRooy,Alison,ed.1998.CivilSociety andtheAidIndustry.Lon-
don:Earthscan.
van Trotsenburg,Axel,and AlanMacArthur.1999."The HIPC
Initiative: Delivering Debt Relief to Poor Countries."
[www.worldbank.org/hipc/related-papers/related-papers.htmll. Press.
Varian,HallR. 1990."MonitoringAgentswith OtherAgents."
JournalofInstitutionaland Theoretical
Economics 146:153-74.
Varma,Keshav.1999."CityofAhmedabad SlumNetworkingPro-
ject."Self-Employed Women'sAssociation, Ahmedabad.
Varshney,Ashutosh.1999a."Democracyand Poverty."Back-
groundpaperfor WorldDevelopmentReport2000/2001. World
Bank,Washington,D.C.
- 1999b. "MassPoliticsor Elite Politics?India's Eco-
nomicReformsin Comparative Perspective."
In JeffreySachs,
AshutoshVarshney,and NirupamBajpai,eds.,Indiain the
EraofEconomic Reforms.NewDelhi:OxfordUniversity Press.
-2000. EthnicConflictand CivicLife:HindusandMus-
limsin India.NewHaven,Conn.:YaleUniversityPress.
Vatsa,Krishna.1999."CommunityParticipationin Postdisaster
Reconstruction: Lessons
LearnedfromtheMaharashtraEmer-
gencyEarthquake Rehabilitation
Prograrn."Boxfor WorldDe-
velopmentReport2000/2001. WorldBank,Washington,D.C.
Velez,CarlosEduardo,AdrianaKugler,and CesarBouill6n.
1999. "A Microeconomic Decompositionof the Inequality
U-turninUrbanColombia: LaborMarketForcesandBeyond."
Inter-American Development Bank,Poverty and InequalityAd-
visoryUnit,Washington,D.C.
Vella,Venanzio.1997."Healthand NutritionalAspectsof Well-
Being."In JeniKlugman,ed.,Povertyin Russia:PublicPolicy
andPrivateResponses. Washington,D.C.:World Bank.
Venkataraman, Arjunamurthy,andJuliaFalconer.1999."Reju-
venatingIndia's DecimatedForeststhroughJoint Action:
Lessonsfrom AndhraPradesh."ProjectBrief.World Bank,
RuralDevelopmentDepartment,Washington,D.C.
Vietnam,GeneralStatisticsOffice.1998. VietnamLivingStan-
dardsSurvey1998.Hanoi.
Villagran,Eduardo.2000. "KeyDriversof ImprovedAccess:
Off-GridService."In EnergySectorManagementAssistance
Programme,Energyand Development Report2000. Energy
ServicesjortheWorld'sPoor.Washington,D.C.:WorldBank.
Visaria,Pravin.2000. "Povertyin India during1994-98:Alter-
nativeEstimates."Instituteof EconomicGrowth,Dehli.
von Pischke,J.D., DaleW. Adams,and Gordon Donald,eds.
1984.RuralFinancial MarketsinDeveloping Countries:Their
UseandAbuse. Baltimore,
Md.:JohnsHopkinsUniversity Press. Wallsten,ScottJ. 1999."AnEmpiricalAnalysisof Competition,
BIBLIOGRAPHIC NOTE 26I
Vos, Rob, MargaritaVelasco,and Edgarde Labastida.1999.
"Economic and Social Effects of El Nifio in Ecuador,
1997-1998."Inter-American DevelopmentBank,Washing-
ton, D.C.
Waddington,Catriona,and K.A.Enyimayew. 1989."Priceto Pay:
The Impactof UserChargesinAsante-Akim District,Ghana."
International Journalof HealthPlanningand Management
4(1): 17-47.
. 1990. "A Price to Pay, Part 2: The Impact of User
Chargesin the VoltaRegionof Ghana."International Jour-
nalofHealthPlanningandManagement 5(4):287-312.
Wade,Robert.1985."The Marketfor PublicOffice:WhytheIn-
dianStateIs Not Betterat Development." WorldDevelopment
13(April):467-97.
- . 1987. Village Republics: EconomicConditionsfor Collec-
tiveActionin SouthIndia.NewYork:CambridgeUniversity
. 1991.GoverningtheMarket: Economic TheoryandtheRole
ofGovernment inEastAsianIndustrialization.Princeton,N.J.:
PrincetonUniversityPress.
. 1992."Howto Make Street-Level BureaucraciesWork
Better:IndiaandKorea."International DevelopmentStudiesBul-
letin23: 51-54.
Wagstaff,Adam.1999.Inequalities in ChildMortalityin theDe-
velopingWorld:HowLargeAre They?How Can TheyBe Re-
duced?Washington, D.C.:WorldBank,HumanDevelopment
Network.
. 2000. "Socioeconomic Inequalitiesin ChildMortality:
ComparisonsacrossNine DevelopingCountries."Bulletin
of the WorldHealthOrganization 78(1).
. Forthcoming a. If theHealthof thePoorMattersMore:
Child SurvivalInequalitiesin Nine DevelopingCountries.
Washington,D.C.: World Bank, Human Development
Network.
. Forthcomingb. Inequalities in ChildHealth.Washing-
ton, D.C.: WorldBank,HumanDevelopmentNetwork.
Wagstaff,Adam,and Eddyvan Doorslaer.2000. "IncomeIn-
equalityandHealth:What DoestheLiteratureTellUs?"An-
nualReviewofPublicHealth21: 543-67.
Wagstaff,Adam,and NaokoWatanabe.2000. "Socioeconomic
InequalitiesinChildMalnutritioninthe Developing World."
WorldBank,Health,Nutrition,and PopulationTeam,Wash-
ington,D.C.
Waldmann,Robert. 1992. "IncomeDistributionand Infant
Mortality."QuarterlyJournal ofEconomics 107:1283-302.
Walker,Ian, Rafaeldel Cid, Fidel Ordonez,and FlorenciaRo-
driguez.1999. "Ex-postEvaluationof the HonduranSocial
InvestmentFund."World Bank,Washington,D.C.
Walker,Ian, MargaritaVelasquez,FranciscoOrdonez,and Flo-
renciaRodriguez.1997."Regulation,Organizationand In-
centives:The PoliticalEconomyof PotableWaterServicesin
Honduras."WorkingPaperR-314. Inter-AmericanDevel-
opmentBank,Washington,D.C.
Walker,Thomas,and JamesRyan.1990.VillageandHousehold
Economies inIndia'sSemi-aridTropics. Baltimore,Md.:Johns
HopkinsUniversityPress.
262 WORLD DEVELOPMENT REPORT 2000/200I
Privatization,and Regulationin Africaand LatinAmerica."
PolicyResearchWorkingPaper2136. World Bank,Wash-
ington,D.C.
Walt,Gill,EnricoPavignani,LucyGilson,and KentBuse.1999.
"ManagingExternalResourcesin the Health Sector:Are
ThereLessonsfor SWAps?"HealthPolicyandPlanning14(3):
273-84.
Ward-Batts,Jennifer.1998."ModelingFamilyExpendituresto
TestIncomePooling."Paperpresented at 1998Population As-
sociationof AmericaMeetings,2-4 April,Chicago.
Waterbury,John. 1973."Endemicand PlannedCorruptionin a
MonarchicalRegime."WorldPolitics 25(4):533-55.
Webb,Michael,and DavidEhrhardt.1998."ImprovingWater
ServicesthroughCompetition."WorldBank,Finance,Private
Sector,and Infrastructure
Network,Washington,D.C.
Weber,EugeneJ. 1976. Peasants into Frenchmen: TheModern-
izationofRuralFrance,1870-1914Stanford,Calif.:Stanford
UniversityPress.
Wei, Shang-Jin.1999a.Corruption and Poverty:
New Evidence.
Washington,D.C.: WorldBank.
- 1999b."Corruptionin EconomicDevelopment:Ben-
eficialGrease,MinorAnnoyance,or MajorObstacle?"Policy
ResearchWorkingPaper2048. World Bank,Washington,
D.C.
Weidenbaum,Murray,and SamuelHughes.1996. TheBam-
booNetwork:HowExpatriateChinese Entrepreneurs
Are Cre-
atinga New EconomicSuperpower in Asia.NewYork:Free
Press.
Weiner,David.1999."AtSeattle,StarttheTradeand LaborDi-
alogue."ODC Viewpoint.OverseasDevelopmentCouncil,
Washington,D.C.
Weiner,Myron. 1991.TheChildandtheStateinlndia:.ChildLabor duction."World Bank, PovertyReductionand Economic
andEducation Policyin Comparative
Perspective.
Princeton,N.J.:
PrincetonUniversityPress.
Wellenius,Bjorn.1997."ExtendingTelecommunications Service
to Rural Areas:The ChileanExperience."Viewpoint105.
WorldBank,Finance,PrivateSector,and Infrastructure Net-
work,Washington,D.C.
Whang, In-Joun. 1981. Managementof Rural Changein Korea.
Seoul:National UniversityPress.
White, Judy. 1997."EvaluationSynthesisof RuralWater and San-
itation Projects."EvaluationReport EV596.U.K. Department
for International Development, London.
Whitehead, Laurence, and George Gray-Molina. 1999. "The
Long-Term Politicsof Pro-poor Policies."Backgroundpaper
for WorldDevelopmentReport2000/2001.World Bank,Wash-
ington, D.C.
Whittington, Dale, Donald T. Lauria,and Xinming Mu. 1989.
"Paying for Urban Services:A Study of Water Vending and
Willingness to Pay for Water in Onitsha, Nigeria." Report
INU40. World Bank, Infrastructure and Urban Develop-
ment Department, Washington, D.C.
WHO (World Health Organization). 1980. International
Classificationof Impairments, Disabilities and Handicaps.
Geneva.
- 1998. TB: A Crossroad-WHO Report on the Global
Tuberculosis Epidemic 1998. Geneva.
. 1999a. RemovingObstacles to HealthyDevelopment:
WHOReportonInfectious Diseases.Geneva.
.1999b. WorldHealthReport1999:MakingaDifference.
Geneva.
Wilkinson,Richard.1996. Unhealthy Societies:
TheAfflictionsof
Inequality. London:Routledge.
Williamson, John. 1990.LatinAmericanAdjustment: HowMuch
HasHappened?Washington, D.C.:InstituteforInternational
Economics.
. 1993. "Democracy and the 'WashingtonConsensus'."
WorldDevelopment21: 1329-36.
Wilson,WilliamJulius.1996.WhenWorkDisappears: TheWorld
of theNew UrbanPoor.NewYork:Knopf.
Winters,L. Alan.1999."Trade,TradePolicy,and Poverty:What
AretheLinks?"Background paperfor WorldDevelopment Re-
port 2000/2001. SussexUniversity,EconomicsDepartment,
Brighton;and World Bank,Washington,D.C.
Wodon,Quentin.1997."FoodEnergyIntakeand Cost of Basic
Needs:MeasuringPovertyin Bangladesh." Journalof Devel-
opmentStudies34: 66- 101.
. 1999."Growth,Inequality,
andPoverty:ARegional Panel
forBangladesh." PolicyResearchWorkingPaper2072.World
Bank,Washington,D.C.
. 2000a."ExtremePovertyand HumanRights:Essays on
JosephWresinski." Backgroundpaperfor WorldDevelopment
Report2000/2001.World Bank,Washington,D.C.
. 2000b."IncomeMobilityand Riskduringthe Business
Cycle."WorldBank,PovertyReductionand EconomicMan-
agementNetwork,Washington,D.C.
. 2000c."PublicWorksEmployment andWorkfarePro-
grams:Optimizingthe Timing of Benefitsfor PovertyRe-
ManagementNetwork,Washington,D.C.
. Forthcoming. "MicroDeterminantsof Consumption,
Poverty,Growth, and Inequalityin Bangladesh."Applied
Economics.
Wodon, Quentin, and NormanHicks. 1999. "Protectingthe
PoorduringCrisisthroughPublicSpending? Framework and
Application to Argentinaand Mexico." World Bank, Poverty
Reduction and Economic Management Network, Washing-
ton, D.C.
Wodon, Quentin, and Mari Minowa. 2000. "Training for the
Urban Unemployed:A Reevaluationof Mexico's Probecat."
World Bank, PovertyReduction and Economic Management
Network, Washington, D.C.
Wodon, Quentin, Norman Hicks, BernadetteRyan, and Gabriel
Gonzalez. 2000. "Are Governments Pro-poor but Short-
Sighted?Targeted and Social Spending for the Poor during
Boomsand Busts."World Bank, PovertyReduction and Eco-
nomic Management Network and Latin America and the
Caribbean Region, Washington, D.C.
Wodon, Quentin, Robert Ayres,Matias Barenstein,Kihoon Lee,
Pia Peeters, Corinne Siaens, and Shlomo Yitzhaki. 2000.
Povertyand Policyin Latin Americaand the Caribbean.World
Bank Technical Paper 467. Washington, D.C.
Wolf, Margery. 1985. RevolutionPostponed:Womenin Contem-
porary China. Stanford, Calif.: Stanford UniversityPress.

Wolfensohn,JamesD. 1999."A Proposalfor a Comprehensive
Development Framework: A DiscussionDraft."WorldBank,
Officeof the President,Washington,D.C.
Woo-Cumings,Meredith,ed. 1999. TheDevelopmental State.
Ithaca,N.Y.:CornellUniversityPress.
Woods,Ngaire.1999."GoodGovernancein InternationalOr-
ganizations."GlobalGovernance 5:39-61.
Woolcock,Michael.1998. "SocialCapitaland EconomicDe-
velopment:Towarda TheoreticalSynthesis and PolicyFrame-
work."Theoryand Society27(2): 151-208.
.1999."LearningfromFailuresin Microfinance." Amer-
icanJournalofEconomics andSociology 58(1):17-42.
.2000."Managing Risk,Shocks,andOpportunities in De-
velopingEconomies:The Roleof SocialCapital."In Gustav
Ranis,ed., Dimensions of Development. New Haven,Conn.:
YaleCenterPress.
-Forthcoming.UsingSocialCapital.GettingtheSocialRe-
lationsRightin the Theoryand PracticeofEconomicDevelop-
ment.Princeton,N.J.: PrincetonUniversityPress.
Woolcock,Michael,and DeepaNarayan.Forthcoming."Social
Capital:Implicationsfor Development Theory,Researchand
Policy."WorldBankResearch Observer.
Woolcock,Michael,Lant Pritchett,andJonathanIsham.1999.
"The SocialFoundations of Poor Economic Growth in
Resource-Rich Countries."Paperpreparedfor UnitedNations
University-World Institutefor DevelopmentEconomicsRe-
searchprojecton naturalresourcesand economicgrowth,
Helsinki.
WorldBank.1980.WorldDevelopment Report1980.NewYork:
OxfordUniversityPress.
-. 1989."NepalHillCommunityForestry."SouthAsiaRe-
gion,Washington,D.C.
1990.WorldDevelopment Report1990:Poverty.
NewYork:
OxfordUniversityPress.
-1991a.Jordan: PublicExpenditureReview.Washington,
D.C.
199 lb. "A SymposiumIssue on the Analysisof Poverty
andAdjustment."WorldBankEconomicReview5(2): 177-393.
- . 1991c. World DevelopmentReport 1991: The Challenge
of Development.New York: Oxford UniversityPress.
- 1992a. "Effective Implementation: Key to Develop-
ment Impact." Report of Portfolio Management Task Force.
Washington, D.C.
.1992b. WorldDevelopmentReport 1992:Developmentand
the Environment.New York: Oxford University Press.
.1993a. "Area DevelopmentProjects."Lessonsand Prac-
rices 3. Operations Evaluation Department, Washington,
D.C.
- 1993b. The EastAsian Miracle:EconomicGrowthand
Public Policy. Policy Research Report. New York: Oxford
UniversityPress.
- 1993c. Uganda:SocialSectors.Country Study. Wash-
ington, D.C.
1993d. World DevelopmentReport 1993: Investing in
Health.New York: Oxford UniversityPress.
.1994a. AdjustmentinAfrica:
Reforms,Results,andtheRoad
Ahead. New York: Oxford University Press.
BIBLIOGRAPHIC NOTE 263
. 1994b.AvertingtheOldAgeCrisis.PolicyResearchRe-
port. NewYork:OxfordUniversityPress.
. 1994c. "HashemiteKingdomof Jordan: PovertyAs-
sessment."Vol. 1. Report 12675-JO.Washington,D.C.
. 1994d."Republicof Turkey-Second Health Project:
EssentialHealthServicesand ManagementDevelopmentin
Easternand SoutheasternAnatolia."StaffAppraisalReport
12765-TU.Washington,D.C.
. 1994e.WorldDevelopment Report1994:Infrastructure
for Development. NewYork:OxfordUniversityPress.
. 1995a."Argentina:Argentina'sPoor-A Profile."Sec-
tor Report13318.Washington,D.C.
. 1995b.Priorities
andStrategies for Education:
A World
BankReview.Washington,D.C.
. 1995c.WorldDevelopment Report1995:Workersin an
IntegratingWorld.NewYork:OxfordUniversityPress.
. 1996a."AccesstoEducationandHealthCarein Uganda."
EasternAfricaDepartmentandPovertyand SocialPolicyDe-
partment,Washington,D.C.
. 1996b."DemocraticandPopularParticipationinthePub-
licField:The Experience of the Participative
Budgetin Porto
Alegre(1989-1995)." EconomicDevelopmentInstitute,
Washington,D.C.
. 1996c.Mexico:RuralPoverty. Washington,D.C.
. 1996d.Poverty Reductionandthe WorldBank:Progress
and Challenges in the1990s.Washington,D.C.
. 1996e."TechnicalAssistance." Lessonsand Practices7.
OperationsEvaluationDepartment,Washington,D.C.
. 1996f.WorldDevelopment Report1996:FromPlanto
Market.NewYork:OxfordUniversityPress.
. 1997a.CantheEnvironment Wait?Priorities
for EastAsia.
Washington,D.C.
. 1997b.China2020:Sharing RisingIncomes.Washing-
ton, D.C.
. 1997c.ClearWater,BlueSkies:China's Environmentin
theNew Century.Washington,D.C.
. 1997d. ConfrontingAIDS: PublicPrioritiesin a Global
Epidemic.Policy ResearchReport. New York: Oxford Uni-
versity Press.
. 1997e.ExpandingtheMeasureof Wealth:IndicatorsofEn-
vironmentallySustainableDevelopment.Environmentally and
SociallySustainableDevelopmentStudiesand MonographsSe-
ries, no. 17. Washington, D.C.
. 1997f. "Health, Nutrition and Population." Sector
StrategyPaper.Human Development Network,Washington,
D.C.
. 1997g. India:Achievementsand Challengesin Reducing
Poverty.Country Study. Washington, D.C.
. 1997h. "India: Andhra Pradesh Hazard Mitigation and
EmergencyCycloneRecoveryProject."Memorandumand Rec-
ommendation of the President P7100. Washington, D.C.
. 1997i. "Romania Povertyand SocialPolicyReport."Eu-
rope and Central Asia Region,Washington, D.C.
. 1997j. Rural Development.From Vision to Action-A
SectorStrategy.Environmentallyand SociallySustainableDe-
velopment Studies and Monographs Series, no. 12. Wash-
ington, D.C.
264 WORLD DEVELOPMENT REPORT 2000/200I
- . 1997k."SocialFunds PortfolioReview."SocialFunds
ThematicGroup,Washington,D.C.
- . 19971.WorldDevelopment Report1997:TheStatein a
ChangingWorld.NewYork:OxfordUniversityPress.
.1998a."AfricaRegionFindings:Listeningto Farmers-
ParticipatoryAssessment of PolicyReformin Zambia'sAgri-
culturalSector."Report105.AfricaRegion,Washington,D.C.
-.1998b.AssessingAidWhatWorks,WhatDoesn't,andWhy.
PolicyResearchReport.NewYork:OxfordUniversityPress.
- 1998c."Bangladesh Emergency FloodRecovery Project."
ReportP7264.SouthAsiaRegion,Washington,D.C.
- 1998d.Bangladesh:FromCountingthePoortoMaking
thePoorCount.CountryStudy.Washington,D.C.
- 1998e."CambodiaPovertyAssessment." EastAsiaand
PacificRegion,Poverty ReductionandEconomicManagement
UnitandHumanDevelopment SectorUnit,Washington, D.C.
1998f.EastAsia:TheRoadtoRecovery. Washington, D.C.
1998g. El SalvadorRuralDevelopment Study.Country
Study.Washington,D.C.
- 1998h.Ethiopia:SocialSectorReport.CountryStudy.
Washington,D.C.
1998i.GlobalEconomicProspects and the Developing
Countries1998/1999.Washington,D.C.
- 1998j. "Honduras:HurricaneEmergencyProject."
MemorandumandRecommendation of the PresidentP7280.
Washington,D.C.
.1998k.ThelmpactofPublicExpenditureReviews:AnEval-
uation.Washington,D.C.
- 19981."Indonesia:Educationin Indonesia,fromCrisis
to Recovery."EducationSector Unit Report 18651-IND.
EastAsiaand PacificRegion,Washington,D.C.
- 1998m."NicaraguaBasicEducationProject."StaffAp-
praisal Report.Latin Americaand the CaribbeanRegion,
Humanand SocialDevelopmentGroup,Washington,D.C.
- 1998n."PakistanPublicExpenditureReview:Reform
Issuesand Options."PovertyReductionand EconomicMan-
agementReport18432.EastAsiaand PacificRegion,Wash-
ington,D.C.
- 1998o."PhilippinesLocalGovernmentUnits Urban
Waterand SanitationProject."ProjectAppraisaland Super-
visionDocument.EastAsiaand PacificRegion,Washington,
D.C.
. 1998p. "Philippines
SocialExpenditurePriorities."Sec-
tor Report 18562-PH.EastAsiaand PacificRegion,Wash-
ington,D.C.
- . 1998q."Post-conflict Reconstruction:The Roleof the
WorldBank."Environmentally and SociallySustainableDe-
velopmentNetwork,Washington,D.C.
.1998r. "RapidSocialAssessments,
January-April1998."
In SocialImpactof theEastAsianFinancialCrisis.Washing-
ton, D.C.
-1 1998s."RecentExperiencewithInvoluntaryResettlement:
Overview."Report 17538. OperationsEvaluationDepart-
ment,Washington,D.C.
- 1998t."ReducingPovertyin India:Optionsfor More
EffectivePublicServices."Report 17881-IN.Washington,
D.C.
. 1998u."RoadSectorInvestmentProgram-Zambia."
AfricaTechnicalDepartment,Washington,D.C.
. 1998v."Rwanda PovertyNote:Rebuildingan Equitable
Society-Povertyand PovertyReductionafterthe Genocide."
World DevelopmentSourcesReport17792-RW.Washing-
ton, D.C.
.1998w. "Security,
PovertyReduction andSustainableDe-
velopmentChallengesfor the NewMillennium."SocialDe-
velopmentDepartment,Washington,D.C.
. 1998x.SocialImpactof the EastAsianFinancialCrisis.
Washington,D.C.
. 1998y."ThailandEconomicMonitor."Departmental
WorkingPaper18936.ThailandResidentUnit, Bangkok.
. 1998z."The World Bankand ClimateChange:East
Asia."GlobalEnvironmentUnit,Washington,D.C.
. 1998aa.WorldDevelopment Report1998/99:Knowledge
for Development. NewYork:OxfordUniversityPress.
. 1999a."Accessto Justice:The EnglishExperiencewith
SmallClaims."PREMNote 40. PovertyReductionand Eco-
nomicManagementNetwork,LegalInstitutionsThematic
Group,Washington,D.C.
. 1999b."AfricaRegion."CommunityActionProgram
WorkingGroup Notes.AfricaRegion,Washington,D.C.
. 1999c."Argentina:The FloodRehabilitationProject."
ImplementationCompletionReport 18769.Washington,
D.C.
. 1999d."Bolivia:
Implementing the Comprehensive De-
velopmentFramework." Report19326-BO.Bolivia,Paraguay,
and PeruCountryManagementUnit, Washington,D.C.
. 1999e."Brazil:Rio FloodReconstruction and Preven-
tion Project."Performance AuditReport19497.Washington,
D.C.
. 1999f."CitieswithoutSlums:ActionPlan for Moving
Slum Upgradingto Scale."Urban DevelopmentDivision,
Washington,D.C.
-. 1999g. "CopingwiththeCrisisinEducationandHealth."
In ThailandSocialMonitor2. Washington,D.C.
. 1999h. "CountryAssessments and IDA Allocations."
Washington,D.C. [www.worldbank.org/ida/idalloc.htm].
. 1999i.Global Development Finance1999.Washington,
D.C.
. 1999j.GlobalEconomicProspects and the Developing
Countries 2000.Washington,D.C.
. 1999k.Greening Industry:New Rolesfor Communities,
Markets, andGovernments.PolicyResearchReport.NewYork:
OxfordUniversityPress.
. 19991. "Honduras:CountryAssistance Strategy."Report
19893-HO.LatinAmericaandthe CaribbeanRegion,Wash-
ington,D.C.
. 1999m.IntensijyingActionagainstHIVIAIDSinAfrica:
Responding toa Development Crisis.Washington,D.C.
. 1999n. "Managingthe SocialDimensionsof Crises:
GoodPracticesin SocialPolicy."HumanDevelopmentNet-
work,Washington,D.C.
-. 1999o."Moldova:PovertyAssessment TechnicalPapers."
Report19846.Europeand CentralAsiaRegion,Washington,
D.C.

1
1999p. Peru:ImprovingHealthCareforthePoor.Coun-
try Study.Washington,D.C.
- 1999q. "PovertyAlleviationin Jordan in the 1990s:
Lessonsfor the Future."Report19869-JO.MiddleEastand
North AfricaRegion,Washington,D.C.
- 1999r."Povertyand Policyin LatinAmericaand the
Caribbean."LatinAmericaand the CaribbeanRegion,Office
of the ChiefEconomist,RegionalStudiesProgram,Wash-
ington,D.C.
- 1999s. Povertyand Social Developmentin Peru,
1994-1997.WashingtonD.C.
- 1999t. "PovertyTrendsand the Voicesof the Poor."
PovertyReductionand EconomicManagementNetwork,
Washington,D.C.
1999u."ProgressReportto theWorldBank'sExecutive
Board."Washington,D.C.
-1999v."PublicExpenditureReviews:Progressand Po-
tential."PREMNote 20. PovertyReductionand Economic
Management Network,PublicExpenditureThematicGroup,
Washington,D.C.
-1999w."Republicof Korea:Establishing a NewFoun-
dationfor SustainedGrowth."Washington,D.C.
- 1999x."AReviewof WorldBankParticipatoryPoverty
Assessments."PovertyReductionand EconomicManage-
ment Network,Washington,D.C.
- 1999y."AStrategicViewof Urbanand LocalGovern-
ment Issues:Implicationsfor the Bank." Transportation,
Water,and UrbanDevelopmentDepartment,Washington,
D.C.
- 1999z."Sub-nationalDatabaseof the World'sLargest
Countries." Sub-nationalRegionalEconomicsThematic
Group,Washington,D.C.
.1999aa."Turkey:EconomicReforms,LivingStandards
and SocialWelfareStudy."Europeand CentralAsiaRegion,
Washington,D.C.
- 1999bb."Vietnam-DevelopmentReport2000:At-
tackingPoverty."CountryEconomicMemorandum.Wash-
ington,D.C.
- . 1999cc.WorldDevelopment Indicators1999.Washing-
ton, D.C.
- 1999dd.WorldDevelopment Report1999/2000:Enter-
ingthe21st Century.NewYork:OxfordUniversityPress.
.1999ee. "Zambia: Road SectorInvestment PlrogramSu-
pervision Reports."Africa Region, Washington, D.C.
- 2000a. "Acceleratingan AIDS Vaccine for Developing
Countries:Recommendationsfor the World Bank."AIDSVac-
cine Task Force, Washington, D.C.
- 2000b. CanAfrica Claimthe21st Century?Washington,
D.C.
- 2000c. "China: Overcoming Rural Poverty." East Asia
and PacificRegion, Washington, D.C.
- 2000d. "East Asia and the Pacific Quarterly Brief."
Washington, D.C.
- 2000e. East Asia: Recoveryand Beyond.Washington,
D.C.
- . 2000f. "Ecuador: Crisis, Poverty and Social Services."
Washington, D.C.
BIBLIOGRAPHIC NOTE z65
2000
. g. GlobalDevelopment
Finance2000.Washington,
D.C.
. 2000h. "Health,Nutrition, Population,and Poverty
CountryInformationSheets."Health,Nutrition,and Popu-
lationDepartment,Washington,D.C.
. 2000i."India-Policies to ReducePovertyand Accel-
erateSustainable Development." SectorReport.SouthAsiaRe-
gion,Washington,D.C.
. 2000j."Kingdom of Morocco:PovertyUpdate."Human
DevelopmentGroup,Washington,D.C. Draft.28 June.
- . 2000k. "MaintainingUtility Servicesfor the Poor."
Washington,D.C.
. 20001.MakingTransition WorkforEveryone: Poverty
and
InequalityinEuropeandCentralAsia. EuropeandCentralAsia
Region,Washington,D.C.
. 2000m."Nicaragua:RainfallRiskManagement." Pro-
jectAppraisalDocument.Washington,D.C.
. 2000n. "A Note on EconomicReformsand Perfor-
mancein Sub-Saharan Africa."AfricaRegion,Officeof the
ChiefEconomist,Washington,D.C.
. 2000o."Philippines PovertyAssessment." EastAsiaand
PacificRegion,Washington,D.C.
00
. 20 p. TheQualityof Growth. NewYork:OxfordUni-
versityPress.
. 20 00 q. "SocialProtectionSectorStrategy:From Safety
Net to Trampoline."Human DevelopmentNetwork,Social
ProtectionTeam,Washington,D.C.
. 2000r. "Water,Sanitationand Poverty."PovertyRe-
ductionStrategyPaper.Washington,D.C.
. 2000s.World Development Indicators
2000.Washington,
D.C.
. Forthcominga. EnGendering Development. PolicyRe-
searchReport.NewYork:OxfordUniversityPress.
. Forthcoming b. PovertyReductionStrategySourcebook.
Washington,D.C.
. Forthcomingc. Sourcebook on Community-Driven De-
velopment in theAfricaRegion.Washington,D.C.
World Bankand IMF (InternationalMonetaryFund). 2000a.
"ProgressReport on PovertyReductionStrategyPapers."
Paperpreparedfor meetingofJointMinisterialCommitteeof
the Boardsof Governorsof the Bankand the Fund on the
Transferof RealResources to DevelopingCountries(Devel-
opment Committee), World Bank and International Mone-
tary Fund, 31 March, Washington, D.C.
. 2000b. "Trade, Development, and PovertyReduction."
Issues paper prepared for meeting ofjoint Ministerial Com-
mittee of the Boards of Governors of the Bank and the Fund
on the Transfer of Real Resourcesto Developing Countries
(Development Committee), World Bank and International
Monetary Fund, 31 March, Washington, D.C.
World Bank Partnerships Group. 1998a. "Partnership for De-
velopment: From Vision to Action." Briefingto Board of Ex-
ecutive Directors, September,Washington, D.C.
. 1998b. "Partnership for Development: Proposed Ac-
tions for the World Bank." Discussion paper. Washington,
D.C.
World Faiths Development Dialogue. 1999. "A Different Per-
266 WORLD DEVELOPMENT REPORT 2000/2001
spectiveon Povertyand Development." Commenton World
Development Report2000/2001.London.
.2000. "ANewDirectionforWorldDevelopment? Com-
ment on the FirstFullVersionof the WorldDevelopment Re-
port2000/2001." OccasionalPaper3. London.
Wuyts,Marc.1996."Foreign Aid,Structural
Adjustment, and Pub-
lic Management:The MozambicanExperience."Develop-
mentand Change27(4): 717-49.
Yao,Shujie.1999."EconomicGrowth,IncomeInequalityand
Povertyin ChinaunderEconomicReforms." Journalof De-
velopment Studies35(6): 104-30.
Yaqub,Shahin. 1999. "How EquitableIs Public Spendingon
Healthand Education?" Backgroundpaperfor WorldDevel-
opmentReport2000/2001. SussexUniversity,PovertyRe-
searchUnit,Brighton;andWorldBank,Development Research
Group,Washington,D.C.
Yaron,Jacob.1992.Successful RuralFinanceInstitutions.World
BankDiscussionPaper150.Washington,D.C.
Yaron, Jacob, McDonald Benjamin Jr., and Stephanie
Charitonenko.1998."PromotingEfficientRuralFinancialIn-
termediation."WorldBankResearch Observer13(2):147-70.
Yaron,Jacob,McDonaldBenjaminJr., and GerdaPiprek.1997.
RuralFinance: Issues,DesignandBestPractices.
Environmen-
tallyand Socially
Sustainable
Development StudiesandMono-
graphsSeries,no. 14.Washington,D.C.: World Bank.
Yepes,Guillermo.1999."Do Cross-Subsidies Help the Poorto
BenefitfromWater and WastewaterServices? Lessonsfrom
Guavaquil."WaterSupplyand SanitationProgramWorking
Paper.UnitedNationsDevelopmentProgrammeand World
Bank,Washington,D.C.
Yost,Carol.1999."Genderand Law:Challengesand Opportu-
nitiesfor Development." Remarksat WorldBankseminaron
Genderand Law,June,Washington,D.C.
Young,Alwyn.1995."The Tyrannyof Numbers:Confronting
the Statistical
Realitiesof EastAsianGrowth."QuarterlyJour-
nalofEconomics 110(3):641-80.
Younger,Stephen.1996. "LaborMarketConsequencesof Re-
trenchmentfor CivilServantsin Ghana."In DavidE. Sahn,
ed., Economic ReformandthePoorinAfrica.Oxford:Claren-
don.
Zaman,Hassan.1999."Assessing the Impactof Microcrediton
Povertyand Vulnerabilityin Bangladesh."PolicyResearch
WorkingPaper2145.WorldBank,Washington,D.C.
Zeller,Manfred.1999."TheRoleofMicrofinance for Incomeand
ConsumptionSmoothing."Paperpresentedat Inter-Ameri-
can DevelopmentBankconferenceon socialprotectionand
poverty,4-5 February,Washington,D.C.
ZengYi,Tu Ping,Gu Baochang,XuYi,LiBohua,and LiYong-
ping. 1993."Causesand Implicationsof the RecentIncrease
in the ReportedSexRatioat Birthin China."Population and
Development Review19(2):283-302.
Zimmerman,Frederic,and MichaelR. Carter. 1999. "Asset
Smoothing,ConsumptionSmoothingand theReproduction
of InequalityunderRiskand Subsistence Constraints."De-
partmentof AgriculturalandAppliedEconomicsStaffPaper
402. Universityof Wisconsin,Madison.
/~~~~~~~~~~~~~~~~~

Selected World
D)evelopment
Indicators
 to the SelectedWorldDevelopment
Introduction Indicators 270

Map TheWorldbyIncome 273

Tables

WorldView
Table1 Sizeofthe economy 274
Table2 Qualityof life 276

People
Table3 Population
and laborforce 278
Table 4 Poverty 280
Table5 Distributionof incomeor consumption 282
Table 6 Education 284
Table7 Health 286

Environment
Table8 Landuseand agricultural
productivity 288
Table9 Wateruse,deforestation,
andprotectedareas 290
Table10 Energyuseandemissions 292

Economy
Table11 Growthof theeconomy 294
Table12 Structureofoutput 296
Table13 Structureofdemand 298
Table14 Centralgovernment finances 300
Table15 Balanceofpaymentscurrentaccountand international
reserves 302

StatesandMarkets
Table16 Privatesectorfinance 304
Table17 Roleof government in theeconomy 306
Table18 Powerand transportation 308
Table19 Communications, information,andscienceandtechnology 310

GlobalLinks
Table20 Globaltrade 312
Table21 Aidand financialflows 314

Table1a Keyindicatorsfor othereconomies 316

Technical
Notes 317

DataSources 333

Classification
of Economies
by IncomeandRegion,2000 334

269
Introduction
to the Selected World
Development Indicators
T he SelectedWorldDevelopmentIndicators
providesa coreset of standardindicatorsdrawn from
the WorldBank'sdevelopmentdatabases.The layout
of the 21 tables retains the tradition of presenting
comparativesocioeconomicdata for more than 130
economiesfor the most recentyearforwhichdata are
available
and foran earlieryear.An additionaltablepre-
sentsbasicindicatorsfor 74 economieswithsparsedata
or with populations of less than 1.5 million,
The indicatorspresentedherearea selectionfrom more
than 500 included in the World Bank's World Devel-
opmentlndicators2000. Publishedannually,the World
Development Indicators reflectsa comprehensiveview
of the development process.Its opening chapter reports
on the recordof and the prospectsfor socialand economic
progressin developingcountries,measuredagainstseven
internationaldevelopmentgoals.Its fivemain sectionsrec-
ognizethe contributionof a wide rangeof factors:human
capitaldevelopment,environmentalsustainabiity,macro-
economicperformance,privatesectordevelopment,and
the globallinks that influencethe external environment
for development.AseparatelypublishedCD-ROM data-
basegivesaccessto more than 1,000 data tables and 500
time-seriesindicators for 223 countries and regions.
270
Organization of the Selected World
Development Indicators
Tables1-2, WorldView,offeran overviewof keyde-
velopmentissues:How rich or poor are the people in
eacheconomy?What is theirreallevelof welfareas re-
flectedin childmalnutritionandmortalityrates?What
is the life expectancyof newborns?What percentage
of adults are illiterate?
Tables 3-7, People, show the rate of progress in so-
cial development during the past decade. They in-
clude data on population growth, labor force
participation, and income distribution. They also pro-
vide measures of well-being such as health status,
poverty rates, school enrollment and achievement,
and gender differences in educational attainment.
Tables 8-10, Environment, bring together key in-
dicators on land use and agricultural output, defor-
estation and protected areas, water resources, energy
consumption, and carbon dioxide emissions.
Tables 11-15, Economy,present information on the
structure and growth of the world's economies, in-
cluding government finance statistics and a summary
of the balance of payments.
 SELECTED
Tables16-19, StatesandMarkets,focuso0ithe roles
of the publicand privatesectorsin creatingthe necessary
infrastructure
foreconomicgrowth.Thesetablespresentin-
formationon privateinvestment,stockmarkets,the eco-
nonic activities
of thestate(includingmilitaryexpenditure),
informationtechnology, and researchand development.
Tables20-2 1, GlobalLinks,contain informationon
trade and financialflows,includingaid and lendingto
developingcountries.
Becausethe World Bank'sprimarybusinessis provid-
inglendingandpolicyadviceto its low-andmiddle-income
members,the issuescoveredin thesetablesfocusmainlyon
these economies.Where available,informationon the
high-incomeeconomiesis alsoprovidedfor comparison.
Readersmaywishto referto nationalstatisticalpublications
and publicationsof the Organisationfor EconomicCo-
operationand Development(OECD) and the European
Unionformoreinformation on thehigh-income economies.
Classification of economies
As in the rest of the report, the main criterion used in
the SelectedWorld DevelopmentIndicatorsto classify
economiesand broadlydistinguishstagesof economic
developmentis GNP per capita. Economiesare classi-
fiedinto threecategoriesaccordingto income.The clas-
sificationusedin thisedition has beenupdatedto reflect
the World Bank'scurrent operational guidelines.The
GNP per capitacutofflevelsare as follows:low income,
$755 or lessin 1999;middle income,$756-9,265; and
high income, $9,266 or more. A further division at
$2,995 is made between lower-middle-incomeand
upper-middle-income economies.Economiesarefurther
classifiedby region. See the table on classificationof
economiesat the endof thisvolumefora listof economies
in eachgroup (includingthose with populationsof less
than 1.5 million).
From time to time an economy'sclassificationis re-
visedbecauseof changesin the abovecutoffvaluesor in
the economys measuredGNP per capita. When such
changesoccur,aggregatesbasedon those classifications
are recalculatedfor the past period so that a consistent
time seriesis maintained.Between1999 and 2000 sev-
erallargecountrieschangedclassification, resultingin sig-
nificantchangesin the income and regionalaggregates.
For example,revisionsto estimatesof China'sGNP per
capitahavecausedthat economyto be reclassifiedfrom
low to lowermiddle income.The followingchangesare
WORLD DEVELOPMENT INDICATORS 271
alsoreflected:Turkeymovedfromupper middleincome
to lowermiddle income;Georgia,Ukraine,and Uzbek-
istan from lower middle income to low income; Do-
minica and SouthAfricafrom lowermiddle income to
upper middleincome; and Hondurasfrom low income
to lowermiddle income.
Data sources and methodology
Thesocioeconomic and environmental datapresentedhere
aredrawnfrom severalsources:primarydata collectedby
the World Bank, member country statisticalpublica-
tions,researchinstitutes,and suchinternationalorgani-
zationsas the OECD, the InternationalMonetaryFund
(IMF), and the UnitedNationsand its specializedagen-
cies(seethe Data Sourcesfollowingthe TechnicalNotes
fora completelisting).Althoughinternationalstandards
of coverage,definition,and classificationapply to most
statisticsreportedby countriesand internationalagencies,
thereareinevitablydifferences in timelinessandreliability
arisingfrom differencesin the capabilitiesand resources
devotedto basicdatacollectionand compilation.Forsome
topics,competingsourcesof datarequirereviewby World
Bankstaff to ensurethat the most reliabledata available
are presented.In someinstances,whereavailabledata are
deemed too weakto provide reliablemeasuresof levels
and trendsor do not adequatelyadhereto international
standards,the data are not shown.
The data presentedare generallyconsistentwith those
in WorldDevelopmentIndicators2000, though data
have been revisedand updated wherevernew informa-
tion has becomeavailable.Differencesmayalsoreflectre-
visionsto historicalseriesand changesin methodology.
Thus data of differentvintagesmaybe publishedin dif-
ferent editionsof WorldBankpublications.Readersare
advisednot to compiledata seriesfrom differentpubli-
cations or differenteditions of the same publication.
Consistenttime-seriesdataareavailableon the WorldDe-
velopmentIndicators2000 CD-ROM.
Alldollarfiguresarein currentU.S.dollarsunlessoth-
erwisestated.The variousmethodsusedto convertfrom
nationalcurrencyfiguresare describedin the Technical
Notes.
Summary measures
The summarymeasuresat the bottom of eachtable are
totals (indicatedby t if the aggregatesincludeestimates
272 WORLD DEVELOPMENT REPORT 2000/200I
for missingdata and nonreportingcountries,or by an s
for simplesums of the data available),weightedaverages
(w), or median values (m) calculated for groups of
economies.Data for the countries excludedfrom the
maintables(thosepresentedin tablela) havebeenincluded
in the summarymeasures,wheredata are available,or by
assumingthat theyfollowthe trendof reportingcountries.
This givesa more consistentaggregatedmeasurebystan-
dardizingcountrycoverageforeachperiodshown.Where
missinginformationaccountsfor a third or more of the
overallestimate,however,the group measureis reported
as not available.The sectionon statisticalmethodsin the
TechnicalNotes providesfurther informationon aggre-
gation methods. Weightsused to construct the aggre-
gatesare listedin the technicalnotes for eachtable.
Terminology and country coverage
The term countrydoesnot implypoliticalindependence
but may referto any territory for which authoritiesre-
port separatesocialor economicstatistics.Dataare shown
for economiesastheywereconstitutedin 1999,and his-
toricaldata arerevisedto reflectcurrentpoliticalarrange-
ments.Throughoutthe tables,exceptionsare noted.
On I July 1997 China resumedits exerciseof sover-
eigntyover Hong Kong.On 20 December1999 China
resumedits exercise of sovereignty
overMacao.Unlessoth-
erwisenoted, data for China do not include data for
Hong Kong, China; Taiwan,China; or Macao,China.
Datafor the DemocraticRepublicof the Congo (Congo,
Dem.Rep.,in the tablelistings)referto the formerZaire.
For clarity,this edition alsousesthe formalname of the
Republicof Congo (Congo,Rep.,in the tablelistings).
Data are shown wheneverpossiblefor the individual
countriesformedfrom the formerCzechoslovakia-the
CzechRepublicand the SlovakRepublic.On 25 Octo-
ber 1999the UnitedNationsTransitionalAdministration
for EastTimor (UNTAET)assumedresponsibility for the
administrationof East Timor. Data for Indonesiain-
cludeEastTimor. Data are shownfor Eritreawhenever
possible,but in most casesbefore 1992 Eritrea is in-
cludedin the data for Ethiopia.Data for Germanyrefer
to the unifiedGermanyunlessotherwisenoted.Data for
Jordanreferto the EastBankonlyunlessotherwisenoted.
In 1991the Union of SovietSocialistRepublicswasdis-
solvedinto 15 countries(Armenia,Azerbaijan,Belarus,
Estonia, Georgia, Kazakhstan,the KyrgyzRepublic,
Latvia, Lithuania, Moldova, the Russian Federation,
Tajikistan,Turkmenistan, Ukraine, and Uzbekistan).
Wheneverpossible,data are shownfor the individual
countries.Data for the Republicof Yemenreferto that
country from 1990onward;data for previousyearsrefer
to aggregateddataforthe formerPeople'sDemocraticRe-
public of Yemenand the formerYemenArab Republic
unlessotherwisenoted. In December1999 the official
name of Venezuelawas changedto RepublicaBolivari-
ana de Venezuela(Venezuela,RB, in the table listings).
Wheneverpossible,data are shownfor the individual
countriesformedfrom the former SocialistFederalRe-
public ofYugoslavia-Bosniaand Herzegovina,Croatia,
the formerYugoslavRepublicof Macedonia,Slovenia,and
the FederalRepublicof Yugoslavia. All referencesto the
FederalRepublicofYugoslavia in the tablesareto the Fed-
eral Republicof Yugoslavia(Serbia/Montenegro)unless
otherwisenoted.
Technical notes
Becausedata quality and intercountrycomparisonsare
oftenproblematic,readersare encouragedto consultthe
TechnicalNotes,the tableon classification
of economies,
and the footnotesto the tables.For more extensivedoc-
umentation,see World DevelopmentIndicators2000.
The Data Sourcessectionfollowingthe TechnicalNotes
lists sourcesthat contain more comprehensivedefini-
tions and descriptionsof the conceptsused.
For more informationabout the SelectedWorld De-
velopmentIndicatorsand the WorldBank'sother statis-
ticalpublications,pleasecontact:
InformationCenter,DevelopmentData Group
The World Bank
1818 H StreetN.W
Washington,D.C. 20433
Hotline: 800-590-1906or 202-473-7824
Fax:202-522-1498
Email:info@worldbank.org
Website: www.worldbank.org/wdi
To order World Bank publications,email your re-
quest to books@worldbank.org, write to World Bank
Publicationsat the addressabove,or call202-473-1155.
The World by Income Low $755 or less
Lower middle $756-2,995
This map presents economies classified according to World Bank estimates of 1999 GNP per capita. Upper middle $2,996-9,265 fl
Not shown on the map because of space constraints are French Polynesia (high income); High $9,266 or more
American Samoa (upper middle income); Fiji, Kiribati, Samoa, and Tonga (lower middle income);
and Tuvalu (no data). No data K]
30988
_IHRD
JULY20

R-~i.n Fed-rtio

Cd,-

w _ ASt,_, .m.X
t-ze
,_C,~A ~ '~'~

t/- -11l! j i Dem Rep turundi; .ro Repo- f9 t eA !

Hdo. E-Mo.6F _ed So r
-zca
;1, g.,7

Tt. -- k T-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~lr
El F..~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~e~ ono
SO . . . .d _.M...-
Fd-tV

* / s.vnumearrd v ~~~~~~~~~~~~~~,4.
V_s sid. T-6-d P,,-p-eC ~ ~ Mo g~ro
G 4g. jK-y.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~F

Zooood
GV
 __ 7C ~~~CZC ~ 0 -
C-,C E; 74O~ 0--C-Cm-O-OC-OC- -
O -00 -000
00 0-0 -000
C -0 - C->>> >>~C CI

= moC' o
0
, - CC' CC> -,C- ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~-C--C--'--c--c

- CCC C' C' C' -' C - - - CC-

-. - -< CC C--C
9 0

3
cF C CC - ---- CCC'~~--- --- - -- - - - - -C -C-0-C-C-C- ->- - C'
- - --- --- - --- ---- -- - - -- --

- CC ~ ~ ~ ~ ~ ~~~ -----

- -- -- - - -- - - -- 3--C -- -- -- -- -- -- - - ---
O - -0

CO-'CCCD- 3C-C C
C - 00
-Y
-
-
YC-CC-
-
CC CC
3-- CC0 - ~3C0C
C3CC3CC3-'
-- C-
-'-----'
- --- CC
CC
-
CCC C
-0C 3C-C- ---
CC -
-33-3CC--
----- -
-CCO-CCC-
-0 - C
-3OCD
-C -
C.
-
C- -
---
-----
- CC
- - -
$CD-CCCC-CCCC
-----
- -
-C0--C-
-I--
-C
-
-'-
C- -C
--.C0-
-
- Cr

-H-
CCC -'---0 - -- CCC ~~~~~~~-- -- -C C-
- CC---C -. -CO ~ 0
C C

0C-CCCC CC'CC'-CC'C'CC-C
C - -. COCCCCD-O 0 C OC OCD-O -C' -OC- - -' ---CC C- - -- C CC-'--C3-C - 0 -C-CCC - - - C--- 00-

03 ::~~~~ E 2- Eg,~~~~~-u
- - - - - - - -v

- -'-C'CC CDC-C -0 ~ ~ ~ ~ ~ ~ '-CO- -- ~ - CO

-- - -----
- -- '-C-- C C - - -- C-C-
 a- 0 0 00 t 000000
_Em0 r 0 -- 00- 0- ° 0 - 0 - -00 0-0 - ------- 0 - E

5- c coo, oco ooc cr, o0 -

-~~~~~~~~~~~~~~~~~~~~
0.COC 0 - - -o000-00
-- 0-- - 0- 0 m- -00- - - - -0-- 0 00--
---- 0 0- ° -c~~~~~

a =
0NmN voe.
C- >N0N
ss.rcrmsoi
v¢o¢0o... oc

5 & X
X~~~'a C0.NOO 0. 00
-
00c0 0 0.00
o 0 0000 0 ..
-0 0
- - S-
_~.
-_
0
m -m00 - - C -o . -
-.
m00m
E_ - -
00
- om°
.- R-
0 0 000
Cr
00
re o 000000000000
° meo
00rX
o -

-T -- T- - -7 - -
Z .' 2 coo, o 00- 0 m0.- cfoO 0 -cO-00 com00 Cm co- 0.-0. -rm00 00 0 00 0 0 0 0000000 u000 0m 3- O 0r O Csr. o°e-Co

000 0-w0 -OCcc

0 OcOO o-- - r- - --oo 00 -oc
C-- - -o- --- -00C
- -o cc

z _- Oc -o
0 oco 00C D000 wr oc
l 000.Z
CD
0- D c r Cr4 -ZC c 0. - 0 D.°0m I COI o :0 0 0000 0 0-0 0 0 ,-C r0 00 o 0 ' -0 00
- - 0 C00 0 0On .oi c0 10 C-cC
SO

Q~~~ 0 *J0JC0mCr 0 CkOroCDcrmnvD_ __sO. 0oC00m .00b .0N 000 00 0 -00 t00000 f 0.-Cr 00o..0-0 00 000. cC -O

W ~~~~~.2 .- , I I I - wr

-0 ~ ~ ~ - - -0 - - 0-0- - 0

10 - - - 0

0, G0. CO 0. 0 0,_ ,,0 0 ,_0_ r 0S>. v0G0 ' ) 00 0 00000.- 00 00 0 000 00 0 0 i Nv~GA rS

A~~~ 3 ~0 - -n Cr 0 0w | . C C 0 Cr Cr 00 Cr 00oc nos 0 0I 0 < 0i

nnnv r o* u0-0 .-. t- 7| S |.D0
mrooo
|m u c 10
0.0 E

0 ~ ~ ~ ~ ~~~E ZL0<~c:. :0
Elw
= ic <
2 f 0zF:E
0 0 Z0 Cr000000000- -00C -0 .-- 00 00 00 0 0.Cr 000 001 0C 00 N 00 00 0 0 59

E S 0 - o 0_ - 00 _I-
m

U g c G 2~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ E E _ t~ 9o
276 WORLD DEVELOPMENT REPORT -2000/2001

Table2. Qualityof life

Growth of Accessto
privateconsumption Adult sanitation
percapita ofealchide dr Lifeexpectancy illiteracy
rate i ra
Avg.annalgrowt
of chil Under-5 at birth %ofpeople Urban areas
ratg.annul0gowt malnutritionrmortality
rate Years 15andabove population urba
rate
19811-98
l%l, %of cnhildren Per1,000 1998 1998 %oftotal
I ubn pop.
Diatribution-under age5 withaccess
Economy corrected 1992-98a 1980 1998 Males Females Males Females 1980 1999 1990-968
Albania ... 8 57 31 69 75 9 24 34 41 97
Algeria -2.3 -1.5 13 139 40 69 72 24 46 44 60
Angoa -9.5 . . 261 204 45 48 -. . 21 34 34
Argentina . 2 38 22 70 77 3 3 83 90 80
Armenia .. 3 . 18 71 78 1 3 66 70-
Australia 1.7 1.1 0 13 6 76 82 . .. 86 85
Austria 2.0 1.5 .1 7 6 75 81 . . 65 85 100
Azerba jan .. 10 . 21 68 75 . .. 53 57 67
Bangladesh 2.1 1.4 56 211 96 58 59 49 71 14 24 77
Belarus -2.7 -2.1 . . 14 63 74 0 1 57 71
Belgium 1.6 1.2 .1 5 6 75 81 . 9.5 97 100
Benin -0.4 .. 29 214 140 52 55 46 77 27 42 54
Bolivia 0.1 0.1 8 170 78 60 64 9 22 46 62 77
Botsvvana 3.0 . .94 105 45 47 27 22 15 50 91
Brazil 0.7 0.3 6 80 40 63 71 16 16 66 81 74
Bugaria -0.8 -0 5 .. 25 15 67 75 1 2 61 69 100
Burkina Fato 0.4 0.2 33 .. 210 43 45 68 87 9 18 78
Burundi -0 9 -0 6 .. 193 196 41 44 45 63 4 9 60
Cambodia ... 330 143 52 55 43 80 12 16
Cameroon -2.0 .22 13 10 5 56 0 33 1 4873
Canada 1.4 09 13 7 76 82 . .. 76 77
Central AfricanHepublic -0.8 -0.3 23 .. 162 43 46 43 68 35 41
Chad ... 39 235 172 47 50 51 69 19 23 74
Chile 4.0 17 1 35 12 72 78 4 5 81 85 82
China 7.2 4.3 18 65 36 68 72 9 25 20 32 58
HonngKong,Chna 4.8 ... . . 76 82 4 11 92 100
Colombia 12 0.5 8 58 28 67 73 9 9 64 73 76
Congo, Dem.Rep. -45 .. 34 210 141 49 52 29 53 29 30 23
Congno Rep. 0.2 ... 125 143 46 51 14 26 41 62 15
CostaRica 0.8 0.4 5 29 15 74 79 5 5 43 48 100
C8tedIlvoire -22 -1.4 24 170 143 46 47 47 64 35 46 59
Croatia . .1 23 10 69 77 1 3 50 57 71
CzechRepublic I.. 19 6 71 78 . .. 75 75
Denmark 1.8 1.4 .. 10 . 73 78 . .. 84 85 100
Donminican Republic 0.0 0.0 6 92 47 69 73 17 17 51 64 76
Ecuador -0.2 -0.1 .. 101 37 68 73 8 11 47 64 87
Egypt,ArabRep. 2.0 1.4 12 175 59 65 68 35 58 44 45 20
ElSalvador 3.0 1.4 11 120 36 67 72 19 25 42 46 78
Eritrea ... 44 -. 90 48 52 34 62 14 18 12
Estonia -1.0 -0.7 .. 25 12 64 75 .-. 70 69
Ethiopia -0.4 -0.3 48 213 173 42 44 58 70 11 17
Finland 1.4 1.1 .. 9 5 74 81 . .. 60 67 100
France 1.6 1.1 .. 13 5 75 82 . .. 73 75 100
Georga ... . 20 69 77 . .. 52 60
Germany ... .16 6 74 80 . .. 83 87
Ghana 0.2 0.2 27 157 96 58 62 22 40 31 38 53
Greece 1.9 1.3 .. 23 8 75 61 2 5 58 60 100
Guatema a 02 0.1 27 .. 52 61 67 25 40 37 39 91
Guinea 1.0 0.6 .. 299 184 48 47 . .. 19 32
Haiti ... 28 280 116 51 56 50 54 24 35 42
HondLuras -0.1 -0.1 25 183 46 67 72 27 27 35 52 81I
Hungarv -0 1 -0.1 .. 26 12 66 75 1 1 57 64 100
India 27 1.7 177 83 62 64 33 57 23 28 46
Indonesiu 4.0 2.9 34 125 52 64 67 9 20 22 40 73
Iran,IluamicRep. 0.5 .. 16 126 33 70 72 18 33 50 61 89
Ireland 2.8 1.9 .. 14 7 73 79 -. - 55 59 100
Israel 3.3 2.1 .. 19 8 76 80 2 6 89 91 100
Italy 2.1 1.6 17 6 75 82 1 2 67 67 100
Jamaica 1.3 0.6 10 39 24 73 77 16 10 47 56 89
Japan 2.8 2.1 .. 11 5 77 84 ... 76 70
Jordan -1.5 -1.0 5 31 66 73 6 17 60 74
Kazakhstan.. . 8 20 59 70 ... 54 56
Kenya 0.4 8.2 23 115 124 50 52 12 27 16 32 69
Korea.Rep. 6.5 4.4 .27 11 69 76 1 4 57 81 100
Kuwait -.. 2 35 13 74 80 17 22 90 97 100
KyrgyzRepublic .. 11 . 41 63 71 -. . 38 34 87
LaoPIR .. 40 200 .. 52 55 38 70 13 23 70
Latvia -.. 26 19 64 76 0 0 66 69 90
Lebanonn . 3 30 68 72 9 21 74 89 100
Lesotho 0.8 0.4 16 168 144 54 57 29 7 13 27
Lithuania ... 24 12 67 77 0 1 61 68
Macedonia, FYR ... .69 16 70 75 ... 54 62 68
Madagascar -2.2 -1.2 40 216 146 56 59 28 42 16 29 50
Malaw 0.8 .. 30 265 229 42 42 27 56 9 24 70
Malaysia 2.9 1.5 20 42 12 70 75 9 18 42 57 100
 SELECTED WORLD DEVELOPMENT INDICATORS 2 77

WORLDVIEW

Growth of Accessto
privatecnupinAdult sinturban
consumpitio Prevalence Lifeexpectancy illiteracy
rate i ra
per.cnnalproth ofchild Under-5 atbirth %of people Urbani areas
annulDgowt
ravg. malnutritionmortality
rate Years 15and above population
%of
rate
1980-98
l%l, % ofchildren Pe ,0 9819 fttl
urbanpop.
Distribation-
underage5 Per 1,000 1998 998 __of_tota_ with access
Economy corrected 1992-980 1980 1998 Males Females Males Females 1980 1999 1990-98e
Mali -1.0 -0.5 27 218 49 52 54 69 19 29 58
Mauritania 0.8 0.5 23 175 140 52 55 48 89 27 56 44
Mexico 0.2 0.1 .74 35 69 75 7 11 66 74 81
Moldova .... 22 63 70 1 2 40 46 96
Mongolia ... 9 60 65 68 28 49 52 63
Moroccs 1.9 1.2 10 152 61 65 69 40 66 41 55 69
Mozambique -1.0 -0.8 26 213 44 47 42 73 13 39 53
Myanmar . .43 134 118 58 62 11 21 24 27 42
Namibia -1 4 .. 26 114 112 54 55 18 20 23 30 77
Nepal 2.0 1.3 57 180 107 58 58 43 78 7 12 34
Netherlands 16 1.1 .1 1 7 75 81 . .. 88 89 100
NewZealand 08 0.4 .. 16 7 75 80 . .. 83 86
Nicaragua -2.2 -1.1 12 143 42 66 71 34 31 50 56 34
Niger -2.2 -1.1 50 317 250 44 48 78 93 13 20 71
Nigeria -4.2 -2.1 39 198 119 52 55 30 48 27 43 61
Norway 1.6 1.2 .1 1 6 76 81 . .. 71 75 100
Pakistan 20 1.4 38 161 129 61 63 42 71 28 36 53
Panama 2.4 1.2 8 38 25 72 76 8 9 50 56 99
PapuaNewGuinea -0.6 -03 30 . 76 57 59 29 45 13 17 82
Paraguay 1.7 0.7 .61 27 68 72 6 9 42 55 20
Peru -04 -0.2 8 126 47 66 71 6 16 65 72 82
Philippines 08 04 30 81 40 87 71 5 5 38 58 88
Poland .. .1 1 89 77 0 0 58 65 100
Portugal 3.1 20 .. 31 0 72 79 6 11 29 83 100
Romania 0.4 03 6 36 25 66 73 1 3 49 56 81
RussianFederation .3 20 61 73 0 1 70 77
Rwanda -1.0 -0.7 29 . 205 40 42 29 43 5 6
SaudiArabia .. 85 26 70 74 17 36 66 85 100
Senegal -0.6 -0.4 22 . 121 51 54 55 74 36 47 83
SierraLeone -3.1 -1.2 336 283 36 39 .. 24 36 17
Singapore 4.8 13 6 75 79 4 12 100 100 100
SlovakRepublic -2.1 -1.7 .. 23 10 69 77 .. . 52 57
Slovenia ... .18 7 71 79 0 0 48 50 100
SouthAfrica -0.1 00 9 91 83 61 66 15 16 48 52 79
Spain 2.2 1.5 .16 7 75 82 2 4 73 77 100
Sri Lanka 2.9 1.9 38 48 18 71 76 6 12 22 23 33
Sweden 0.7 0.5 .. 9 5 77 82 -. 83 83 100
Switzerland 0.5 0.3 11 5 76 82 .. . 57 68 100
SyrianArabRepublic 0.9 13 73 32 67 72 13 42 47 54 77
Tajikistan .... 33 66 71 1 1 34 28 83
Tanzania 0.0 0.0 31 176 136 46 48 17 36 15 32 97
Thailand 5.1 3.0 .. 58 33 70 75 3 7 17 21 98
Togo -01 .. 25 188 144 47 50 28 62 23 33 57
Tunisia 1.1 0.7 9 100 32 70 74 21 42 52 65 100
Turkey 2.6 1.5 10 133 42 67 72 7 25 44 74 99
Turkmenistan ... . 44 63 70 .. . 47 45 70
Uganda 1.9 1.2 26 180 170 42 41 24 46 9 14 75
Ukraine ..... 17 62 73 0 1 62 68 70
UnitedKingdom 26 1.6 .. 14 7 75 80 .. 89 89 100
UnitedStates 1.0 1.1 1 15 . 74 80 ... 74 77
Uruguay 2.6 1.5 4 42 19 70 78 3 2 85 01 56
Uzbekistan 5.5 3.7 19 .. 29 66 73 7 17 41 37 46
Venezuela, RB -0 8 -0.4 5 42 25 70 76 7 9 79 87 64
Vietnam . 40 105 42 66 71 5 9 19 20 43
Yemen,Rep. .. 46 198 96 55 56 34 77 19 24 40
Zambia -3.6 -1 8 24 149 192 43 43 16 31 40 40 40
Zimbabwe 0.4 0.2 16 108 125 50 52 8 17 22 35 99
World 1.3w 30w 123w 75w 65w 69w law 32w 40w 46w .. w
Lowincome 1.4 .. 177107 59 61 30 49 24 31 56
Middleincome 2.2 14 79 38 67 72 10 20 38 50
Lowermiddleincome 36 15 83 39 67 72 10 23 31 43 59
Uppermiddleincome 1.5.. 6 35 7 74 9 1 64 6
Lowandmiddle income 1.9 .. 135 79 63 67 18 33 32 41
EastAsia& Pacific 5.6 22 82 43 67 71 9 22 22 34 61
Europe& CentralAsia 8 . 26 65 74 2 5 59 67
LatinAmerica& Caribbean 0.6 8 78 38 67 73 11 13 85 75
MiddleEast&NorthAfrica .. 15 136 55 66 69 26 48 48 58
SouthAsia 2.6 51 180 89 62 63 35 59 22 28 46
Sub-SaharanAfrica -1.2 33 188 151 49 52 32 49 23 34
Highincome 2.2 15 6 75 81 .. 75 77
Note:Fordatacomparability
andcoverage,seethe Technical
Noses.
Figuresin itaicsarefor yuarsotherthunnoose
specified
a.Dataarefor mostrecentyearanailable
 - CD ~~~~~
~~~ n n n p n ~C~- -- CO -
00CgO 0 ' B0CC

C- o&Ycr
00 COF F -octB(fl C
0
~~~~~~~~-
- 0
---- -

aa - a~~- -- -- I-- ----- ----. =- - - - -<I -~C - - - P-

- -

Zn Zn~~~~~~~~~~~~~~~~~~~~~~

- - --
-<
--- - - -
0 -c -, Zn~~~~ -F - ~~~~~~~~~~~~
o

- ---- -- -- - ---- - --
--- - - -- -- - - -- --- -I - -- -- 0 z
- 0 ----- c- no-C -- --- - --- - - - - - - - --

-C-n- -o- o-C

-- - - 4-Z
-n - - -Zn -CC -C - - - -Zn - -F-C- - -0-t-
0-0- -NQo-O-
- - -1-0-0- 0-aot-t
- to-
-Ž - 0- - -

Ni-"' CO-no C' t-o-C 0 -- -- -- -- -l -o - - -- -- C--C - 0- - . -.- n-o -

nononco n-ono -0---- no-nt--0 --- -on - -no n- -n--ono

- - -Z
- - - -- - -
ton--o --
- - - - - - - -
-- o o-t-n -ooto-0o-- noon n- - ---- nono--- o-n-noI---
- -- -oooot
n-c--- - - o-o -
-u a

Znnon-toCCtoC
CnoCo-CZ Ott no--oonoo-notocrClC.ro--- Z Z- -ttC-
C,nt- - n-t-
- - - - - -- - - - - - -
no-n-
~ ~CC-Cotto-
~ --
~ n -CFo co--
~ ~ ~ C
~
1Co- ]
Zn
00
C
ET '

- - - - -- - - - - - - - - - - - - - - - -C,

- - - - - - - - - - - - 5~~~~~~~~~~~~~~~~~~n
E
 N.-t )0R-N 10oo -'-N-t - '-o -v
N .. J -o C -ON (0t omeomN'1 '-10N '-oR 10o

o e
LU
CL s !' - - - - - -r - - -D - - cn - - - - - :Ios rote
- -t -rw n<moooo

o = -~~~~~~~~~~~~~~~~-----

; WS oO n m oCCC (00004 1
mmS<0 10 o- 10 B
oSmm|omom>mN4
O 10o

- (C c-
C' -s -_- -- --- - o -
------- 0 0=me)c0m
O-0O(-0C-0(-1101-1-0 c0
-01-CO-0- 0- N
N-- ro 00 0N (
m -O11-11-10 1-11--
- 00.N- (000(1w
0)-10- - -01 - (010m
- --r- -0010-1-
- - l00 - f--O -- 0-JC-1- -
- 00 -w---00O0N11 -

H *VLJ. ~ (0140(00(0 10CO 1000) N-100- CO-N-N- - - - - 0000

- - - - N- 0- - - -O---N --
--- 1 -00( -0N- -- -- -- --

0 N - C O 0N C 10 0 10) unv0
Now C O 1.01 0 C Cr N -1 C N - 10 0 1 0c OCO c Oe 10 1m0) 0 Co o -
= 0 Cr 1 0 0 01 1 r ,1 1C O ( 0 J O C 1 0 0
- - - - - - - 0 - - - - - - - - - - -01-0 -10-0 mmm - - - - --- - -C ° e CO 1010c ( O 0 0CO10C

A N-CON-OCOCOCO-COCO 0.--1-0C-O-C 0OC - 0- -00

- C-- - -- -- -CO--- --- - (O-
-000-1-N -. -0 -------- -O - - - - - - - - -

- 010-- -- ---- ( 00 0 -001 -) 0

- - - 10( -C CO N -N-O-CO-CN-N- -NCOC
- (
-CO-O-,-r-

W.o -£tms .no,< . mm.S0 - mm-w .m0-00 - .o m-on

- -- -- t o. -- .T - 9 > (0 0LC1- L 0

CO-- 0 10010)0)0e C_ Ds o
0)e 01m00)0(0 l 0
0~~~~~~~~ o (000)00 N-CC1 m e OCO 00
_rCo1 N-(O fl (0m0(erm
00 xt me c

-4
0 CCCOCC
--- - - --- N-C - N-C - 10--- 000-- ---CC C O - -cO- - C-
- 00)1)00-0 -N-O-CO-C -- --CO-C.- - - - O0~O

C.)~~~~~~~~~~~~~~~~~~~~~~~g1

| a( 00r1 0 10 c n0 Cs. i0n 0r mCO L m

r--
u: - - - - i- - -m --m_010

CC C z- CO -0 C --N - -O -N C OCO1--(00 1 10 -N-0 0

) N (0 ' '0 o-N0 E

.' N-. cC.0
[E0)000)10010~
.010.N-COCO
E 5 E E E E ° D = E o z z z z z c X =a=°c====
c'=
O~~~~E 2
280 WORLD DEVELOPMENT REPORT 2000/2001

Table 4. Poverty

National poverty
lines International
poverty
lines
Population
belowthe Population
belowthe Population
PovertyPopulation
Poverty
povertyline (%) povertv
line(%)below gapat below gapat
Survey Survey
_oetlie% poetln() Survey $1aday $1aday $2aday $2aday
Economy year Rural UrbanNational year Rural Urban National year % %%
Albania
Algeria 1988 16.6 7.3 12.2 1995 30.3 14.7 22.6 1995 <2 <0.5 15.1 3.6
Angola
Argentna 1991 25.5 1993 . 17.6
Armenia
Auistrali
I.
Austra
Azerbaijan 1995 . 68.1
Bangladesh 1991-92 46.0 23.3 42~7 1995-96 39.8 14.3 35.6 1996 29.1 5.9 77,8 31.8
Belarus 1995 . .. 22.5 . 1998 <2 <0.5 <2 0.1
Belgium . .
Benin 1995 . . 330 . .
Bolivia 1993 . 29.3 . 1995 79.1 .. 1990 11.3 2.2 36,6 13.5
Botswana . -. . 1965-86 333 12.5 61.4 39.7
Brazil 1990 32.6 13.1 17.4 . 1997 51 1.3 17.4 6.3
Bulgaria .-. -.. 1995 <2 <0.5 7.6 1.6
BurkinaFanso. . . . . 1994 61.~2 25.5 95.6 59.9
Burundi 19990 . . 36.2
Cambodia 1993-94 43.1 24.9 39.0 1997 49.1 211 36.1 .
Camefoon 1964 32.4 44.4 40.0 . . .

Cnanada . ... .
Central
AfricanRepublic. . . . . 1993 96.6 38.1 84.0 56,4
Chad 1995-96 67.0 630 640
Chile 1992 . .. 216 1994 - 29.5 1994 4.2 0.7 20.3 5.9
China 1966 7.9 <2 6.0 1996 4.6 <2 4.6 1968 16.5 4.2 53.7 21.0
HonigKong, China.. . ... ...

Colomnbia 1991 290 7.6 16.9 1992 71.2 8.0 17.7 1996 11.0 3.2 28.7 11.6
Congo,Dum.Rep.
Congo.Rep.-
CostaRica .. 1996 96 3.2 26.3 16.1
COtedIlvoiro . 1995 12.3 2.4 49.4 16.8
Croatia
CzechRepublic .. 1993 <2 <9.5 <2 <05
Denmark.. .
DominicanRepublic 1969 27.4 23.3 24.5 1992 29.8 10.9 26.6 1996 3.2 0.7 16.0 5,0
Ecuador 1994 47.0 25.0 35.90 1995 20.2 5.6 52.3 21.2
Egypt,ArabRep. 1995-96 23.3 22.5 22.9 . 1995 3.1 0.3 52.7 11.4
ElSalvador 1992 55.7 43.1 46.3 . 1996 25.3 10.4 51.9 24.7
Eritea
Estonia 1995 14.7 6.8 6.9 1995 4.9 1.2 17.7 6.6
Ethiopia . 1995 31.3 8.0 76.4 32.9
Finland
France
Georga 1997 9.9 12.1 111I
Germany
Ghana 1992 34.3 26.7 31.4
Greecoe . ...
Guatemala 1969 71.0 33.7 57.69. 1669 39.8 1668 64,3 36.6
Guinea 1994 - 40.0 -.
Haitni 1067 - 65.9 1995 66.0
Honiduras 1992 46.6 56.0 50.9 1993 51.0 57.0 53.6 1996 48.5 17.5 68.8 36.9
Hungary 1969 . .. 1.6 1993 . .. 8.6 1993 <2 <0.5 4.0 0.9
India 1992 43.5 33.7 40.9 1994 36.7 30.5 35.6 1997 44.2 12.0 86.2 41.4
Indonesia 1996 12.3 9.7 11.3 1998 22.0 17.83 203 1999 15.2 2.5 66.1 22.6
Iran,IslamicRep. .

Irelarnd . . ..

Italy .
Jamaca 1092 . .. 34.2 .. 1996 3,2 0.7 25.2 69

Jordan 1991 . .. 15.0 1997 . .. 11.7 1097 <2 <0.5 7.4 14

Kazaknstan 1996 39.0 36.0 34 6 .. . . 1996 1.5 0.3 15.3 39
Kenya 1992 46.4 293 420 .. . . 1994 26.5 9.6 62.3 27.5
Korea,Rep. . -.
.. . 1993 <2 <0.5 <2 <03.5

KyrgyzRepublic 1063 48.1 26.7 4660 1997 64.5 28.5 51.0

LaoPDA 1983 53.0 24.0 46.1...
Latvia .. . . . 1998 <2 <0.5 8.3 2.0
Lebanon . . ......
Lesotho 1993 53.9 27.8 49.2 .. 1963 43.1 20.3 65.7 36.1
lithuania .. 1996 <2 <0.5 7.8 2.0
Macedo0nia,
FYR . . .. ..
Madagascar 1993-94 77.0 47.9 70.9. 1993 60.2 24.5 89.6 51.3
Malawi 1990-91 - 5490
Malaysia 1989 .. . 15.5
 SELECTED WORLD DEVELOPMENT INDICATORS 281

PEOPLE

National
poverty
lines poverty
International lines
Population
belowthe Population
belowthe PovertyPopulation
Population Poverty
(%)
Survey l ~ ~~ Survey poverty
pvryline line(do) below gapat below gapat
Survty $1a day $1a day $2aday $2aday
Economy year Rural UrbanNational year Rural Urban National year % %%
Mali .. . .. . 1894 72.8 37.4 906 60.5
Mauritania 1989-90 . . 57.0 . . . 1995 3.9 1.0 221 6.6
Mexico 1999B . 10.1 .. . . 1995 17.9 6.1 42.5 18.1
Moidona 1997 26.7 .. 23.3 . 1992 7.3 1.3 31.9 10.2
Mongolia 1995 33.1 36.5 36.3 .. 1995 13.9 3.1 50.0 17.5
Morocco 1990-91 16.0 7.6 13.1 1996-99 27.2 12.0 19.0 1990-91 <2 <03.5 7.5 13
Mozambique . . ... . .. 1996 379 12.0 78.4 36.6
Myanmar. .. ... . ..
Namibia .. . .. . . 1993 34.9 14.0 5568 30.4
Nepal 1995-96 44.0 23.0 _42.0 .. 1995 37.7 97 62.5 37.5
Netherlands ..
NewZealand. . ...
Nicaragua 1993 76.1 31.9 50.3 .
Niger 1969-93 660 52.0 63.0 ... 1995 61.4 33.9 65.3 54.8
Nigeria 1965 495 31.7 43.0 1992-93 36.4 30.4 34.1 1997 70.2 34.9 90.6 59.0
Norway. ... . ..
Pakistan 1991 36.9 26.0 34.0 .. 1996 31.0 6.2 647 35.0
Panama 1997 64.9 15.3 37.3 . . . 1997 10.3 3.2 251 10.2
PapuaNew Guinea . . . . .
Paraguay 1991 26.5 19.7 21 6 . 1995 19.4 8.3 36.5 16.6
Peru 1994 670 461 53.5 1997 64.7 40.4 490 1996 15.5 5.4 41.4 17.1
Philippines 1994 53.1 26.0 406 1997 51 2 225 40.6.
Poland 1993 . .. 23.8 . .. . 1993 5.4 43 10.5 6.0
Portugal . .. .. . .. 1994 <2 <05 <2 <0.5
Rumania 1994 27.9 20.4 21.5 .. 1994 2.8 0.8 27.5 6.9
Ruosian Federation 1994 .. . 30.9 .. 1996 7.1 1.4 25.1 8.7
Rwanda 1993 . .. 512 .. 1983-85 35 7 7.7 64.6 36.7
Saudi Arabia.. . ....
Senegal . .. ... 1995 26.3 7.0 678 26.2
SierraLeone 1969 76.0 53.0 66.0 .. 1969 57.0 39.5 74.5 51.8
Singapore . . . ..
SlovakRepublic . .. ... 1992 <2 <0.5 <2 <05
Slovenia .. 1993 <2 <0.5 <2 <0.5
SouthAfrica . 1993 115 1.6 35.6 13A4
Spain . . .. . .

Sri Lanka 1985-86 45.5 26.6 40.6 1990-91 36.1 28.4 35.3 1995 6.6 10 45.4 13.5
Sweden . . . . ..
Switzerland.. . .. .. ..
SyrianArabRepublic. . ... . ...

Tajikistan.. . .. .. .

Tanzania 1991 .51.1 .. . . 1993 19.9 46 59.7 230

Thailand 1990 . . ta 1992 15.5 10.2 13.1 1998 <2 <0.5 28.2 71
Togo 1967-89 . .. 32.3 . . ..
Tunisia 1965 29.2 12.0 19.9 1990 21.6 6.9 14.1 1990 <2 <05 11.0 2.9
Turkey . . ... . .. 1994 2.4 05 16.0 S.0
Turkmenistan .. . .. 1993 20.9 5.7 59.0 23.3
Uganda 1993 . .. 55.0 . . . 1992 36.7 114 77.2 35.6
Ukraine 1995 . .31 7 .. 1996 <2 <03.5 23.7 4.4
UnitedKingdom
UnitedStateo
Uruguay .. . .1969 <2 <0.5 6.6 1.9
Uzbekiotan . .. 1993 3.3 0.5 26.5 7.3
Venezuela,RB 1969 . .. 31.3 .. 1996 14.7 5.6 364 15.7
Vietnam 1993 57.2 25.9 50.9. . ..
Yemen,Rep. 1992 19.2 16.6 19.1 . . 1998 5.1 0.9 35.5 10.1
Zambia 1991 66.0 46.0 660 1993 86.0 1996 72.6 37.7 91.7 61.2
Zimbabwe 19919-91 31.0 10.0 25.5 .. . . 1990-91 36.0 96 642 29.4
Note:Fordatacomparability
andcoverage,
tee the Technical Notes.Figuresin italicsareforyearsotherthanthoseupecified.
28z WORLD DEVELOPMENT REPORT 20002OOI

Table 5. Distributionof income or consumption

Percentage
shareof incomeorconsumption
Economy Surveyyear Giniindex Lowest10% Lowest20% Second20% Third20% Fourth20% Highest20% Highest10%
Albania
Ageria '995a' 35.3 28 7.0 11.6 161 22.7 426 268
Angna
Argentina
Armemia
Australa 19941' 35.2 20 59 120 172 236 413 254
Austria 1987'' 231 44 104 14.8 185 22.9 333 193
Azerbai an
Bangacesh 1995-96'b 33.6 3.9 87 12.0 15.7 208 42.8 28.6
Belarus 199B'b 21.7 5.1 114 15.2 18.2 219 33.3 200
Belgum 1992''- 25.0 3.7 95 14.6 184 23,0 34.5 20 2
Benin
Bolivia 1990'- 4220 2.3 506 9.7 14 5 22 0 48.2 31 7
Botswana
Brazil 1996' 600 0.9 25 5.5 100 183 63.8 476
Bulgaria 1995'- 283 34 85 13.8 179 22.7 37.0 22.5
BurkinaFasc 19943 482 22 55 8.7 12 0 1807 55.0 395
Burund 1992'3 333 3.4 79 12.1 16.3 221 41.6 266
Cambodia 1997' 404 2.9 6.9 10.7 14.7 201 47.6 33.8
Carmerton
Canada 1994-d 31.5 28 75 12.9 17.2 230 39.3 23.8
CentralAfricaer
Republic 1993'b 613 07 2.0 49 96 10.5 65 0 47.7
Chad
Chle 1994-' 565 14 35 66 10.9 18.1 610 46.1
China 1998-: 40.3 2.4 5.9 102 15.1 22.2 46.6 30.4
HongKong,ChFna
Coomba 10996' 571 1 3.9 66 11.1 18.4 6069 461
Congo, Dem.Rep
Congo, Rep
CostaRica 1996' _ 47.0 13 4,0 808 '3.7 21.7 518 34.7
Crted'lvoire 1995'' 36.7 31 71 112 15.6 21.9 44.3 28.8
Croatia 19980b 26.8 4.0 93 138 178 22.9 362 21.6
Czech Republic 1996'' 25.4 4.3 103 145 177 217 359 224
Denmark 1992'1 24.7 3.6 98 14.9 183 22.7 345 20.5
Dominican Repubt
c 1996'' 48.7 1.7 43 0.3 131 20.6 537 378
Ecuador 1995a' 43.7 22 54 9.4 14.2 21.3 49.7 33 8
Egypt,ArabRep 1995a,' 2809 4.4 98 13.2 16.6 214 39.0 25.0
ElSavador 19960 52.3 1.2 3.4 7.5 12.5 20.2 56.5 40.5
Ertrea
Estonia 1995': 35.4 2.2 602 120 170 231 41.8 262
Ethiopia 1995< 400 30 71 109 14 5 19.8 47.7 33.7
Fnland '991'' 256 42 100 14.2 17.6 223 35.8 21.6
France '995'' 327 2.8 72 12.6 172 22.8 4092 25.1
Georgia
Cerranv 1994d 300 3.3 02 13.2 175 227 3805 23.7
Ghana 1997'' 327 36 8.4 '2 2 15.8 21.9 41.7 26.1
Greece 1993'' 327 30 75 124 169 22.8 40.3 25.3
Guatemala 1989'' 596 06 2.1 50 10.5 186 63.0 46.6
Guinea 1994'h 4093 26 6.4 10.4 14.8 212 47.2 32.0
Haiti
Honduras 1996'' 53.7 12 3.4 71 11.7 19.7 580 421
Hungary 1996-' 308 39 808 125 16.6 22.3 399 24.8
India 19973b 370 35 01 116 15.0 19.3 461 33.5
Indones a 1996-' 36.5 36 800 113 15.1 208 44.9 30.3
Iran,IslamicRep : ,_ _ -
Ireand 1987'1 35.9 25 6.7 it 0 '64 22.4 429 27.4
Israe 1992'' 35.5 28 6.9 114 '6.3 22.9 42 5 26.9
Italy 1995'' 27.3 35 897 149 '.81 22.9 36.3 21.8
Jamaca 19963. 36.4 2.9 7.0 115 '5.0 21.8 43.9 28.9
Japan 1993'' 24.9 4.8 '0.6 142 17.6 22.0 357 21.7
Jerdan 199730 3694 3.3 76 11 4 155 21.1 444 2958
Kazakhstan '8996' 35.4 2.7 67 11.5 164 23.1 423 263
Kenya 1994'') 44.5 108 50 97 142 20.9 502 349
Korea, Rep '593,'" 316 2.9 75 12.9 174 22.9 39.3 243
Kuwait - -
Kyrgyzieoublic 1997'- 40.5 2.7 63 102 147 21.4 47.4 317
LaoPDR 19923- 30.4 42 96 129 163 21.0 40.2 264
Latvia 1998&- 32.4 2.9 76 12.9 17.1 221 40.3 259
Lebanor
Lesotho 1986-87" 56., 099 26 6.5 112 194 60.1 43.4
Lithuana 1996',' 324 3.1 78 12.6 168 224 40.3 256
Macedonia, FYR
Madagascar 1993'', 46.0 1.9 51 9.4 133 20.1 52.1 36.7
Malawi
Malaysia 1995A 485 18 4.5 0.3 13.0 20.4 53.8 37.9
CA'

c-X -X O s0 6040-40 10-- 4raSc- - 01- 04S64G

°an< - -- - - - --- - - - - - -- - -
A,< 0 0 4
044,601 010 o-o 01610 6 DNO n NN04
04- 04- -000 4400004

0 04 co 0. 0r 0- 04 0-4 m4 -0 07 04 04 G4 04 0- 04 C 4 0 _4 - :4 04I 04 4 04 0-I 04 0O CO 0r 04 04 0 04 04 04 04 04) 004

EM E a
Oo e a ]SOO
z 0

E ._004~ 04- 0404-0-040-0404040- - -o -0---0 0 .E

._ -0- . 400 -- - - o-- - - - - - - -

04 t-440004- 400044-44004 -- 0-40 - 40-4000-0

H 40
~ 040
0 0-040 04 4040 04 4040, 0--04004044040 040,04 4040 040 040
0440 -4
040404040404040404
-4' - 0
~ ~~
400-
- -
1
0

ae

v: ~ 4
~ ~ (D CN040440
400404-40o
C4 0 4r 4004v
S>q>:C
4DCO
04
r4
,-440
40400404
044-440
04,04400
04044O44
,0404040 e040
404
040
040044400
,-0040 C0400
: :040:
COCD
040
01oZ
04000
caoDs
_44440
604C044060
aDC nw

e4 04 0 0 0 to 04 0 04 04 LO 04 04 04 04 04 40 04 04 to 04 04 04 04 04 40 040 40 04 04 04 : 04 04 04 04 04 04 10 04 04 04 m 04 040_44

2* 0-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~2

~~~~~~~Xt0
o
D4 bsNta 0,
76
0

0 --- -- - --- …4,04~~~~-

E2 ~~ ~ ~ ~ ~ b
oo~~~~~~~
444 40 I

E2 -rE 4
E~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
40004004004: 04004ME040-04-04- 404-0 044.0 -
:4404444 .040040PE !4> 40FE04 -- 0440040
284 WORLD DEVELOPMENT RFPORT 2,000/2001

Table 6. Education
Public expenditure Net enrollmentratioO Percentage of cohort
%of relevant age groLip reaching grade 5 Expected years of schooling
on education
of GNP Primary Secondary Males Females Males Females
Economy 1980 1997 1980 1997 1980 1997 1980 1996 1980 1996 1980 1997 1980 1997
Albania 37 81 83
Algera 7.8 51 82 96 43 69 90 94 85 95 10 12 7 10
Angola - 83 35 81 31 - - 8 9 7 7
Argentna 2.7 15 97 100 59 77
Armenia 20
Australia 5.5 5.4 100 100 81 96 12 17 12 17
Austria 55 54 100 I 00 91 97 11 15 11 14
Azerbaijan 30
Bangadesh II 22 60 75 18 22 18 26 5 3
Belarus 59 85
Beiglum 6.0 3.1 100 I 00 96 100 - 14 77 13 17
Benin 3,2 53 68 25 28 59 64 62 57
Bolivia 44 4.9 79 97 34 40 -
Botswana 6.0 8,13 76 80 40 89 80 87 34 93 7 12 8 12
Brazil 3.6 5. 7 80 97 46 66
Bulgar a 4.5 3.2 98 98 75 78 93 90 11 12 11 12
Burk na Fasc 2.2 1.5 15 32 5 13 77 74 74 77 2 3 1 2
Burunc 3,4 4 0 20 36 B 17 100 96 3 5 2 4
Carnbodia 2.9 1 00 100 15 39 51 46
Cameroon 3.8 71 62 40 40 70 70 8 6
Canada 6.9 6 9 100 100 84 95 15 17 15 17
Centra Africar Republic 57 46 27 19 63 50
Chad 17 26 4B 13 18 62 53
Ch e 4.6 3.6 93 90 70 B5 94 100 97 100 73 13
China 2.5 2.3 84 100 63 70 93 94
Hcrig Kong, Ch na 2 4 2.9 98 91 67 69 98 - 99 - 12 12 11 12
Colorfib a 19 4.7 73 89 60 76 36 70 39 76
Congo, Dem. Rep. 2 6 71 58 44 37 56 59 7 4
Congo, Rep, 7 0 6.1 97 78 98 B4 81 40 83 78
Costa Rica 78 5.4 89 89 39 40 77 86 82 89 10 10
Cbte d'lvo re 7 2 50 55 58 39 34 86 77 79 77
Croatia 5,3 100 100 80 72 98 9,q II 12
Czech Republic 5 1 95 100 93 100 - 13 13
Berimark 6.7 8.1 96 100 89 95 99 loo 99 99 14 15 13 75
Dom nican Reputi 2 2 23 99 91 50 79 - - II - II
Ecuador 5.6 35 92 100 66 51 84 86
Egypt, Arab Rep 5.7 4.B 72 95 43 75 92 88 12 10
E Salvador 3,9 2.5 70 89 23 36 46 76 48 77 70 10
Eritrea 1.8 29 38 73 67 5 4
Estop a 72 100 100 100 86 96 97 12 13
Ethlomia 3. 7 4.0 28 35 19 25 50 57 51 50
Finland 5 3 7,5 100 100 87 95 IDO loo 15 17
France 5.0 6.0 100 100 94 99 15 15
Georg a 5,2 93 89 97 76 71 11
Germariy 4,8 100 100 82 95 16 16
Ghana 31 42
Greece 2.0 3,1 100 100 75 91 99 98 13 14 12 14
Guatemala 18 11 59 74 28 35 .. 52 47
Gu nea I 9 30 46 20 15 59 85 41 68
Ha t 1,5 - 33 34
Honduras 12 3 6 79 88 44 36
Hungary 4.7 4 6 95 98 71 97 96 97 13 13
Ind a 10 32 65 77 41 60 - 62 55
Indonesia 1.7 1,4 89 99 42 56 88 88 10 10
Iran, Is amic Rep 7.5 40 72 90 50 81 92 89 12 71
Ireland 6.3 6 0 100 100 go 100 99 100 11 14 12 14
Israel 8.2 7,6 - -
Italy 4 9 100 100 70 95 99 98 99 99
Jama ca 7.0 7.4 98 96 71 70 91 97 it - 11
Japar, 5.8 3.6 100 100 93 100 100 100 100 100 14 13
Jordan 6.6 6 8 73 68 53 41 100 98 12 12
Kazakhslan 4 4
Kenya 6.8 6.5 91 65 55 61 60 62 -
Korea, Rep 3.7 3.7 100 100 76 100 94 98 94 99 12 75 11 14
Kuwait 2.4 5 0 85 65 81 63 12 9 12 9
Kyrgyz Republic 5.3 100 100 100 78
Lao PDR 2 1 72 73 53 63 57 54
Latvia 3.3 6.3 IDO 100 90 81 12 13
Le ormon - 2.5 76 -
Lesotro 5.1 8.4 67 69 69 73 50 72 68 87 7 9 10 70
Lithuania 5.4 81
Macedonia, FYR 5.1 95 55 95 95 11 11
Madagascar 4A 1.9 61 49 33 -
Malavvi 3A 5,4 43 99 39 73 48 40 -
Malaysia 6.0 4.9 92 100 48 64 97 98 97 100
 SELECTED WORLD DEVELOPMENT INDICATORS 285

PEOPLE
Publicexpenditure Net enrollmentratioa Percentage of cohort
on educatiion %of relevant age group reachinggrade 5 Expectedyears of schooling
%of GNp Primary Secondary Males Females Males Females
Economy 1980 1997 1980 1997 1980 1997 1980 1996 1980 1996 1980 1997 1980 1997
Mali 37 22 20 38 10 18 48 92 42 70
Mauritania 5.1 57 61 68
Mexico 47 4.9 98 100 67 66 85 86
Moldova 34 106
Mongolia 57 100 85 89 56 - 7 9
Morocco 6.1 5.0 62 77 36 38 79 76 78 74 8 5
Mozambique 31 35 40 40 22 52 39 5 4 4 3
Myanmar 1.7 1.2 71 99 38 54
Namibia 1.5 9.1 86 91 67 81 76 82
Nepal 18 3.2 66 78 26 55
Netherlands 7.7 5.1 100 100 93 100 94 98 13 16 13 16
NewZealand 58 7.3 100 100 85 93 97 97 14 16 13 17
Nicaragua 3.4 39 71 79 51 51 40 52 47 57 9 9
Niger 32 2.3 22 24 7 9 74 72 72 74 3 2
Nigeria 6.4 0.7
Norway 6.5 74 99 100 B4 98 100 100 100 loo 13 15 13 16
Pakistan 21 2.7
Panama 49 5.1 89 90 65 71 74 79 11 12
PapuaNew Guinea
Paraguay 15 4.0 91 96 37 61 59 77 58 80 10 10
Peru 3.1 2.9 87 94 BO 84 78 74 11 10
Philippines 17 34 95 100 72 78 68 73 11 11
Poland 7.5 99 99 73 87 12 13 12 13
Portugal 3.8 5.8 99 100 45 90 14 15
Romania 3.3 3.6 91 100 ]DO 76 12 12
RussianFederation 3.5 3.5 92 100 98 88
Rwanda 27 59 69 74 -
SaudiArabia 41 7.5 49 60 37 59 82 87 86 91? 7 10 5 9
Senegal 37 37 60 19 20 89 89 82 85
Sierra Leone 3.5
Singapore 2.8 3.0 100 91 66 76 100 100
SlovakRepublic 50
Slovema 57 95
South Africa 79 68 100 62 95 72 79 74 74
Spain 23 50 100 100 79 92 95 94 13 13
Sri Lanka 27 34 96 100 59 76 92 91
Sweden 90 83 100 100 83 100 98 97 99 97 12 14 13 15
Switzerland 47 5.4 100 100 80 84 13 15 12 74
SyrianArabRepublic 4.6 3.1 90 95 48 42 93 93 88 94 11 10 8 9
Tajikistan 2.2
Tanzania 68 48 89 78 90 84 10 7
Thailand 3.4 4.8 92 88 25 48
Togo 56 45 79 82 65 58 59 79 45 60
Tunisia 54 7.7 83 100 40 74 89 90 84 92 10 7
Turkey 2.2 22 81 100 42 58 93 96 II 9
Turkmenistan
Uganda 13 2.6 82 73
Ukraine 56 7.3
UnitedKingdom 56 5.3 100 100 88 92 13 16 13 17
UnitedStates 6.7 5.4 90 100 94 96 14 15 15 16
Uruguay 2.3 J3 87 94 70 84 97 99
Uzbekistan 7.7
Venezuela,RB 4.4 5.2 83 83 24 49 86 92 10 11
Vietnam 30 96 100 47 55
Yernen,Rep. 70
Zambia 4.5 22 77 72 35 42 88 82 7
Zimhabwe 53 72 93 20 59 82 78 76 79
World 3.9 m 4.8 m 81 w 9OW 60 w Sow w w w w
Low income 3.4 3.3 66 76 38 51
Middle income 18 48 86 97 63 71
Lowermiddleincome 3.5 4.8 85 98 64 70 92 92
Uppermiddleincome 4.0 5.0 B8 96 59 75
Low and middle income 3.5 41 78 89 53 63
East Asia& Pacific 2.5 2.9 86 99 59 67 93 93
Europe& CentralAsia 51 92 100 84 81
LatinAmerica& Caribbean 3.8 36 85 94 55 66
MiddleEast &NorthAfrica 5.0 52 74 87 46 66 88 84
South Asia 2.0 31 64 77 38 55 62 55
Sub-SaharanAfrica 3.8 41
High income 5.6 5.4 97 100 87 96
Note:Fordata comparabilityand coverage,see the TechnicalNotes. Figuresin italicsare for years other than those specified.
a. UNESCO enrollmentestimates and projectionsas assessed in 1999
 --- >O >OC 0 oo
a ~-.o
0.-...~ 070mGCoo
38~j~B-o~>Po
0 0 0V'Oin>
or~~
0
m
m3
m orom
2,-.~=-~2
,2,00,u
0m~80C
000 C
0000700-0-00
000 0000-0-)0-0-)0-0-)0->>->->>
00~,3~~3DCr- 000DO
> I>

o---- )----- -- 0----0 ----- 3-0----,--- 3- -~--. - - - - __ - - -- - - - - - - xL'

S~3 T0 00E 0 3E - - --- -0- 0 0- --C

010) ---'-- 0) 0) 2, ~ ~ ~~~~~~~~~- -0- - --- -- - 0- -0

- ~
-, - 0- -- ----

3*M=
- --- - - - -- --- - --- - - - - - - - - -- -- -- -- - -- -- -- -- -* -- 0
-- - - - - -- --- - - - - -- - - - -- - - - - - - -- - - - -- - - - -- - -0
- - - - - -- -- - - -- -- - -- -- -- - -- -- -- -- - - - - - -- - - - - -- - - - - -m
- - - - - -- - - - - - - - - - - - - - - - - - -

OL~~~~~~~~t
00 IL g
~~EX
IJ,5XI c-c cc -c :sX C cc
Dsr2 cc,. Q cccccc -- inOccoc

0.
0. O..fo .- nzU7c,-c crNJcc .- L)it COSns.- A in nSobSC 3-
III o.-cCCn
C)B

U U

ccc r-ca cccc.3-cccc- or--c-,ccLfl

ccccc ---
cc',-cccc - --
cc- - - - - -0 ------ - -- -c- --- N- - a5 c .
z 2~1

, _cc cc c-;c-a ro cc cc c, cc> cc cccc .- cccccc ccc,cccc cciA Ec

c.- c cccccc -- cc ----- ca. c,ccc.cc-cc.-

------ - c -ca-, -ncccc
a,-c - cc---> -ccc--.-cc-cc,- -c-cc
cc-ccccc c

>~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:>w
-c-c-ccc-c- ----
- w -0, --- cc-c- - - - - - cc~cccc
-c- -ccccccc-c-a-.- c - -cccc -cccca
-. cc- 4 -cccc--a-,c-c---cc

-cLo_ _ o °d =E ~~~~~~~~~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~c

Xw";;_ cL
gS. En - -c

- - - . ~ ~~~~~~~~
- - - - -E
.4~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~E EE E ~ ~~~ Wc

~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~'~~~~~~~
- cc-
-cc cc. cccccc
ccc - :. - mccc, cc a, ccccc - - cc :cccccc a, ccc, cc ccccc :.- - cc cc -'
cc-cc cc.-cc cc cccc
ccc, cc cccccc cc .3-.--cc - cc cc cc .3-cc cc cc~~~~~~~~~~~~~~~~~~~~~~~~~~~r
0~~~~~~~~~~~~~~~~~i, R7
cc 0~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~F F
o~~~~~~~~~~~~E a cc
CWc,
'ccc,c ccccc3 c -ccc cc.-cc cO ccc a,c3-c cc- c.ciccccc.cc cO-cE cccccccc
cc c c 0 c a,cc
c cccc , c cccc c , c cccc c c c a,~5 ccc c ccc,c cccc c ca c c c cccca,a,ccc cc _c
_ 03 A 1czV V D;>5> a',~~~~~-
 00 :000--00---100-- r-.- - ,---1-
--- -- --- - -- 0-10 -000-1- 1 1-0- --- 0 01-00
-1-0--- 10-0-0
-1-0-000-O---O- C-I 0 -1
-10 0- - -1-

U- i -

1010 10s _S0 osv000 0 1 1 0 1 0 00 0 111 0 sO R 0)10 n coCwo 01- C 000101 0'°0°° 1=0 1011 0 S0°0 0-1

o_ 10 0im 0° 100 303 0000.100 10

0010C'J0-010-00 t 1Ot° 0 Ct 0001 0101
0001 - - - 110-
0-0 010-
-0 00 1-C-
10 0 n00 :0 - -100CO0
00 -00 0-01
-C
00 0-1
0ESn0 -1
000 010-0 0oo LI 00 0 10 0 10I 01

0-100100000010m0
.- ---- 000 --- -0
00>
-
:0.
1
10--CoO
1-
:0101
0
:01010
C ~ '10
- 02.
:0 100 :0
C
:,-00
1CIN
o0 :0-0
1
oo0
0 0
e -0
'2100
.0 1 : 110 0

-t 0CC I 00 0 0 10 0 101 00 0 01 . 00 00=0 05 =-01 00 1 er r r N. m0 1R01001- _ 00.0 1-00o011 00010 0mL - oR

0 ° 00

CC 10- ~ CO- ----- 1010----- IN - IN- C-O IN

, 03 o .C7 0_

- 00- -------- 1101-…… ----- --- -----0- ----- - C-0 0- -N - - -CO1 ---- -

4S Q1~-1 - I .- o 0 :0- ro:o:1m0

00ooo 0 Ln :010001010
-0-00CC -00
0010 -:-:1---
0101 010100 -0 0 0
_ooo 1010101 :-: :-:-
01 oo1 -00 --
oO0001000111 0 - 00 - 0-
0oooo010. o : 00 : : 100 : : 001 mo

--- - - - - -- 0 - ----- ----- ----- ----- --- -0 1-0 ------ 1 ----- 010
--- - 010-
-- 101

O tEs 10. tlso

W
$~~
-
CL~~~~C,1
.300=o
aSi 4To10101
.e 10
: -10
00
:0 :
,-
10000-0000000
10 0 10
10100CC
0
1010
10
: :1
10
0 000 : : 1-10
101
:0-010e00
10 10 o 100101010ces
-
00
001
00
0 oP
0 :.-1-1 0 :01
10 i=os
=°Sm -00s
co :_reNC

O t C,'- - ~~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ -----

Z 0
2 0 @ * (qn 4 Oo .~ 2oe 4 110100
O
010 CIO
10 I O.~
1010101010-O
Cr031010 O
0101000001101- 000-- 00-1-100- u 0101 10 m0 100 - .- 00-
0.-1-1 101 - -C r - -m11---11.
-- 0 0010-Q
10.CIN00 J
CO 1 00 -00O 10-0 0I 1-0.-0011 1 00 0o

Cl CC0 00 010 0 00000 o1 00 -oooo_ ooo0 6oooo 666o 66 6 6666 6666 6666 ro c-oo00 00 0 o m oO

N a 03G0G G 0

0
I.. 0 0 - 0 01 1 1F1 ~ mfC 2

***1
0 1 0100 0010 0001 00 1 0 - 10 <1 0 1 0- 000 1000110 01100-01 1 - 00 -0 0--
 SELECTED WORLD DEVELOPMENT INDICATORS 289

ENVIRONMENT
Agricultural
Agricultural productivity
Landunder machinery Agr.valueadded
permanent Tractors peragricultural
crops Irrigatedland Amablelend pertehousand worker Foodproductionindex
%of landarea %of cropland Hectarespercapita agricultural
workers 1995dollars 1989-91= 100
Economy 1980 1997 1979-81 1995-97 1979-81 1995-97 1979-81 1995-97 1979-81 1996-98 1970-81 1996-98
Mali 00 0.0 2.9 2.1 0.31 042 0 1 241 271 17.9 1145
Mauritania 0.0 00 22.8 9.7 0.14 0.21 1 1 285 433 06.5 1047
Mexico 08 11 20.3 22.0 034 0.27 106 20 1,082 2,164 83.8 1202
Moldova 121 .. 14.1 0.41 .. 85 .. 1,474 532
Mongolia 0.0 6..4 071 053 32 21 932 1,151 08.1 805
Morocco 11 19 15.2 13.1 0.38 033 7 10 1,146 1,036 55.8 107.2
Mozambique 0.3 03 2.1 3.4 0.24 0.10 1 1 127 100.1 130.9
Myanmar 07 0.9 10.4 154 0.28 022 1 0 ... 877 1381
Namibia 0.0 00 006 0.8 0.04 052 180 11 802 1,233 107.4 123.5
Nepal 0.2 0.5 22.5 302 0.16 0.13 0 0 162 189 65.9 1172
NetherJands 0.0 1.0 505 610 8.06 0.06 561 631 21,663 . 86.5 99.0
NewZealand 3.7 0.4 5.2 8.7 0.80 0.42 619 437 ... 90.7 124.8
Nicaragua 15 2.4 6.0 32 039 0.54 6 7 1,620 1,821 117.0 1227
Niger 00 0.0 0.7 1.3 0.62 0.53 0 0 222 195 101.4 127.8
Nigeria 20 2 07 0 03 0.512 414 624 57.9 142.5
Norway .. . . 0.20 0.22 824 1,270 17,044 32,600 921 1009
Pakistan 0.4 07 72.7 80.8 0.24 017 5 13 394 626 86.4 136.2
Panama 16 2.1 5,0 4.91 0.22 019 27 20 2,122 2.512 05.6 9900
PapuaNewvGuinea 1.1 13 .0.01 001 1 1 716 700 86.2 1078
Paraguay 03 02 3.4 2.0 0.52 0.44 14 25 2,018 3,448 6006 1200
Peru 0.3 04 32.8 42.0 0.18 0 15 5 3 1,340 1,063 773 140.5
Philippines 14.8 14.8 14.0 162 0.00 0.07 1 1 1,347 1,352 060 125.8
Poland 11 1.2 0.7 07 0.41 037 112 281 .. 1,751 879 062
Portugal 7.8 8.2 20.1 2108 0.25 022 72 208 ... 719 970
Rumania 2.9 26 2189 31.3 0.44 0.41 39 84 .. 3.101 1128 859
RussianFederation . 01 . 4.0C .. 0.86 . 106 . 2,476 64.4
Rwanda 10.3 122 04 63 0.15 0.12 0 6 371 249 88.7 79.1
SaudiArabia 0.0 0.1 289 423 0.20 019 2 12 3.167 10,742 29.7 78.0
Senegal 0.0 0.2 26 31 042 0.26 0 0 341 320 74.2 100.4
SierraLeone 8.7 0.8 41 53 014 0.1I8 0 368 411 84.5 99.5
Singapore 9.8 .. 00 0.08 3 10 13,937 41,673 154.3 31.8
SlovakRepublic .. 2.6 125 E 0.28 .92 .. 3,378 -. 74.7
Slovenia .. 2.7 07 . 0.12 . 3.082 .. 26,517 .. 100.3
SouthAfrica 0.7 08 0.4 79 0 45 8.38 94 68 2,899 3,958 92 6 1008
Spain 9.9 07 14.8 18.1 6.42 0.38 200 540 -. 13,499 021 1181I
Sri Lanka 159 150 28.3 30.7 000 005 4 2 649 726 983 109.1
Sweden . . 036 032 715 958 .1001 1008
Switzerland 85 0.6 6.2 5.6 006 006 494 627 95 8 95 8
SyrianArabRepublic 25 41 96 20.5 0.60 0.33 29 68 942 1407
Tajikistan 69 .. 79.7 . 0.13 .. 38 . 396 590
Tanzania 1.8 180 38 38 0.12 010 1 1 .. 174 769 10080
Thailand 35 66 16.4 23.9 0.35 028 1 7 634 932 804 1126
Togo 66 6.6 0.3 0.3 0.76 0.49 0 0 345 539 770 1359
Tunisia 9.7 129 49 76 0.51 032 30 39 1,743 2,959 67.6 1214
Turkey 4.1 3.4 9.6 147 0.57 042 38 58 1,860 1.858 75.0 1113
Turkmenistan . 0.1 .. 035 82 .101.7
Uganda 8.0 8.8 0.1 0.1 832 026 8 1 . 345 704 1071
Ukraine . 1.7 7.4 .. 0.65 .89 . 1,454 .. 523
UnitedKingdom 83 0.2 2.0 1.7 812 0.18 726 883 . . 920 99.7
UnitedStates 82 0.2 108 12.8 0.83 0.67 1,230 1,484 . 36,001 945 117.9
Uruguay 03 8.3 5.4 10.7 0.48 0.39 171 173 9,021 9,862 069 13080
Uzbekistan 0..9 803 . 0.19 .59 2,113 . 1090B
Venezuela, 80 09 1.0 3.6 5.7 019 812 50 59 4,041 5,036 803 1144
Vietnam 19 4.7 24.1 31.0 011 808 1 4 230 638 1405
Yemen, Rep. 02 02 19.0 31.3 0.10 009 3 2 .. 338 750 1207
Zambia 00 0.0 0.4 0.9 089 0.57 3 2 328 209 740 1045
Zimbabwe 03 63 3.1 4.7 0.36 0.27 7 7 307 347 81.9 101.9
World 0.9 w 1.0w 17.8w 19.2w 0.24w 0.24w 18 w 20 w .w .w 75.7w 130.3w
Lowincome 1.8 1.4 19.9 255 0.22 019 .71 3 1243
Middle income 10 1.8 23.9 19.4 0.18 023 .... 71.0 143.8
Lowermiddleincome 1.1 09 321 223 0.14 021 6908 151 1
Uppermiddleincome 1.6 12 101 11.4 0.32 0.29 79.1 118.5
Lowandmiddleincome 1.0 12 21 9 219 0.20 021 4 8 .. 566 71.6 134.7
EastAsia&Pacific 1.5 2.6 37.0 363 011 011 2 2 ... 7.0 152.1
Europe & CentralAoia 3.2 04 1106 104 814 0.6 101 . 2,180
LatinAmerica& Caribbean 1.1 1.3 116 135 8.32 0.28 25 35 . 8084 1239
MiddleEast& NorthAfrica 04 8.7 25.8 35.5 029 0.21 12 25 ... 70.1 138.0
SouthAsia 1.5 2.1 28.7 397 6 23 0.16 2 5 265 356 70.4 122.1
Sub-Saharan Africa 0.7 0.9 4.0 42 032 0.25 3 2 418 379 78.8 124.3
Highincome 0.5 8.5 9.8 112 046 8.41 520 900 . 93.1 107.5
Nore:Fordatacomparability andcoverage,
seesthe Technical Notes Figures.nitalicsarefor yoursother thanthose specified
a Includes Luxembourg. b.IncludesTaiwan,China
 - - - - -- -5-0 - - - : -- - - - 5-5-0-50
-0 -- - -5-0-00 -5-0-0--0- - -0-0 - - -or- -c -0-00-0 -0-0-- -- -0-5000- -0- 00- 50000-0 - - -O
-0 ---- - - - - C'- 0- ---

- 00 0 5 0
-0 -00- -- -- 0 05 --- 0 0 0-
-00 ------- 0--00-
------ -0-0-5- 00- OO-O-
-o-Obs-o -- -0-0-0-0-

~~ 0500 00 05005 0 0.0 005o0-0-5-- -0-- - -- 05-½ - 05 ------- - - -0 - - 0

3 I- o 2

a L00, 0 00 0 :0 0 05 0 0 05 '00 , : 00 0 00 -L-o 0 0 0s--O oo- r o o- 0 05 00L,00 0 5

= o--£005 o 0000- '-- ------ 00 0 -o.o - - -6 6 H o - -oo-. --o-c-A-b- - -o6- - ---- -c -000 -o -oo-

Cw~

0)
0) 050 :0 cs : -- - - :-:0-0- -- :-5-05-:-
-- -- - :00 00- :0 50- - -00 [--:-Os- - -0 - -0 - :.:00 . o :0 00

o~~ ,-0 050 05 '-0 0 0 00 050500 N~~~~~~~-

0 0- 0050 us
-- 00 0-0000- 005-- -00

~~~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
C0 ---

(00 E s 7

co ~ ~ -0 : :5:: : - :.5 0 0 : 0 0 ~ ::0 0 :- - 00 : : s::: 0 'O so :: 0

-C 0 0500 0 0 00 0 0
C~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~1
05~~~~05 ~~~~
00 0 0 0 0 0 0 005 00 0 O 0 0 0 0 0
9
55
Z
Z ~-Q) c h-o ooo ocoa, - r-OOocn - C -0-L cc)m -
-O -aco rh- - O--
Oc--- -- oooco
ccoccaco
0 c - oc
-Oc-ch-C
- -o -r o o-coc
fLrOo-c c a co o 00
in
-oc-
a
-O'- -,
0
-
-
c-I h
~~ ~~~-
-a,-
-----
-
0-0
a,
-Oh--a
h--c0-I ----
- - - - -- - -C- -

0; :E
Sr C
~~
Zcnsccc no car-co.- 0000Cc-I- c-O-O-a- -c-- O- - - O.O-I(
--- c-- -ccO h-c-o c- i-c--cch -- C0-
-OCc -E o o hc -r-a

:'-h- o c W B0Ra, c-<

[- Cv ° G a,'-- a, _ - G c a, Oh- G n _ - - -B - - B_S_--
--
Ca , - - - - -

03oEa - - - - - - - - - - - 0
H B O c 0 0t B c (. >

- - - - - -- - - - - - -
E~~~~~~~~~~~ - - - - - - - - - - - - - - - - - - - - °-|- - - - - - - - - -

Ew oc
-cocae cc-c- E cca,cd 6a, - jc-i c co cocoa, coco , -coca -a,-cc cocoa ca, co6 E oco c - c- - c - - -'
-- - - - o-c- - n cc--- oo - -a c o-c- - --- -I--- --
a,--a,--0 a -
- -Cco a--L--l--a -- -0 - 0 E-0 h 0 , c h-a-

- ,
C --- 0c'fl
-a,-----
- -a- - 0a-,a,
-&,5-$c$
~-0.-
-00 ------ c-c-- - a -c c-Ic--c
0
c
0
no--aa
o o
c -

E2 - - - - - - - - - - - -~~~~~~~~~~~- -ccC-

- -- -- - -- ~ o ; -'

E -~~~~~~~~~~~~~~~~~~ 5-c~~~~~~~~~~~~~~D

C -M 2Coo
k R: z 9 4 -0 f E-6~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C~C
c

a~ ~ ~~~ R DZ :
 m,C070- ~ Ei -y 5r =~ 000 0- ~ m-0 0 -lTWll-rrr
000 0000 - -- m-- m amm a m a»>>> r mN

BaoO~ DC
PI
CDCCCDC
M
00
WctD) Ce0),
- -
P.
-D-a -- -Bo - 0--
a
- - - 0
-1 0
-Bo- 0
at
-- --
-D-D -C0- -10-
~ C
-'0O -D
- ia~

zi 0~~~~~~~~~~~~

DC ~ ~ ~ ~ ~ ~
- --- - -P --

bi ~ M ~ ~ ~ ~ ~

C -. -- - ----- ----- ----- -0.--'-C - --- - -N - a--- - -- - -- --

-- - 'a-.--c -C - - - -NCm N--C

- -- C -C-NC -C - ' 7- -- - - -CNNC - - N--C

NC
NCNCNC-N -- aac O-N-N-0 - -- - -C ONC -- - -- NC- -- - -NC-m N-'-0a-N- -4 C CIN-aN - -CN
N - - - - - - .
10 NC NCNC a -'-NO -aC - -CN
m, - N- - - -C- - -N-N-a-0- - -N- m -mC - - - - . m NCa- . m. Ce C- a C ca C

NC C0~-J F-Cc-C- -C- - -C0N

-NC - m ' C-C C C-- 0(---J-C a-CN- -c.-a
C0N -- C-al -- .- '- -' -

-0 - NCCD
C( NCNCCC cii NCNCONCNCNC-a-C-N- a- a-aNC-
-C-C -C- -N- :-m,,-aNC -C: - -m -C - a - - - C ERoCm Em'a-.
C 0 o b
- - -- - - -- - -- -- - -- - - -

a- -'---'C- -- NC - N-0 NC- C Z---

CC ---- - -- --- - -
N ---- NC- - -N- -C--NC - N- -C---'-C N NC - - CC'Ca a N
aa aC~ NC ma -am - -C- - N a a- aNC
---- t ---- -.
CCCNN C 0m- -NC-a N- - ' --- NN -NeN 0N C N -ci - 0- C-'-'-' - maW- - -- C
 SELECTED WORLD DEVELOPMENT INDICATORS 293

ENVIRONMENT
Commercial
energyuse GDP perunit Cro ixd msin
Thousandmetric PercaPita ofenergy use Netenergy
imports' abndoid msin
tonsof Kgof oil Av.nul PPP$ perkg %of commercial Total Percapita
oilequivalent equivalent Aeg.rannuaof oilequivalent energyuse Millionmetrictons Metrictons
Economy 1990 1997 1990 1997 1990-97 1990 1997 1990 1997 1990 1996 1990 1996
Muli .. . . .- 0.4 0.5 0.0 00-
Mauritania. . 2.0 2.9 13 12
Mexico 124,187 141.520 1,492 1,501 -0.3 4.2 51 -57 -58 295.0 348.1 3.5 3.8
Moldova 9,959 4,430 2,283 1,029 -11.2 2.0 2.1 100 96 218 12.1 5.0 28
Mongolia .-.. -. . . . 100 8.9 4.5 3.6
Morocco 6,745 9,275 281 340 2.9 99 95 89 88 23.5 27.9 1.0 1.0
Mouambique 7,318 7,664 517 461 -1.7 1.0 1.0 6 0 1.0 1.03 0.1 01
Myanmar 10,787 13,009 266 296 16 . .. -1 6 41 7.3 0.1 0.2
Namibia
Nepal 5,834 7,160 311 321 0.6 2.8 3.7 3 8 -0.6 16 0.0 0.1
Netherlando 66,593 74,910 4,454 4,800 11 38 4.6 10 13 138.9 1552 9.3 10.0
NewZealand 14,157 16,679 4,120 4,435 0.9 3.4 4.0 14 15 23.8 298 69 8.0
Nicaragua 2,174 2,573 568 551 -0.8 2.8 39 31 41 2.6 2.9 0.7 06

Nigeria 70,905 88,852 737 753 0.2 1.0 1.1 -112 -115 68.7 83.3 0.9 0.7
Norway 21,456 24,226 5,050 5.501 10 36 48 -460 -778 47.7 67.0 11.2 153
Pukistan 43,238 56,818 400 442 17 33 39 21 28 67.9 94.3 0.6 0.8
Panama 1.535 2,328 640 856 4.0 56 6.1 61 65 3.1 6.7 1.3 25
PapuaNewGuinea .. .. .. .. 2.4 2.4 0.6 05
Paraguay 3,097 4,191 734 824 21 5.3 5.5 -48 -661 2.3 37 9.5 0.7
Peru 11,549 15,127 535 621 24 5.3 7.3 -6 19 22.2 26.2 1.0 1.1
Philippines 28,294 36,251 452 520 2.3 08 7.2 4.4 57 44.3 63.2 0.7 0.9
Poland 100,114 105,155 2.626 2,721 0.0 2.1 2.7 1 4 347.6 356.8 91 9.2
Portugal 16.419 20,400 1,659 2,051 2.9 6.6 7.1 87 89 42.3 47.9 4.3 4.8
Romania 61,117 44,135 2,634 1,957 -2.6 2.3 32 35 30 1551 119.3 6.7 5.3
RussianFederation 908,433 591,982 6,112 4,019 -6.2 1.6 17 -40 -57 1,954.4 1,570.5 13. 103.7
Rwanda . .. . . . . 05 0.5 01 01
SaudiArabia 03,275 98,449 4,004 4,906 1.7 25 21 -483 -395 177.1 267.8 11.2 13.8
Senegal 2,213 2,770 302 315 05 368 41 38 40 2.9 3.1 0.4 0.4
SierraLeone . . 3
D.. 04 0.1 0.1
Singapore 13,357 26,878 4,938 8.661 84 2.8 2.9 .. 100 41.9 65.8 155 21.6
SlovakRepublic 21,363 17,216 4,044 3,198 -2.7 2.1 3.0 75 73 43.0 396 81 74
Slovenia 5,250 . 2,627 .. 3.6 4.4 55 .. 13.0 .. 6.5
SouthAfrica 91,229 107,220 2,592 2,636 0.3 31 3.3 -26 -33 291.1 292.7 6.3 7.3
Spain 90,552 107.328 2,332 2,729 2. 5.3 5.9 62 71 211.7 232.5 55 5.9
Sri Lanka 5,476 7.159 322 386 2.4 6.2 7.6 23 39 3.9 7.1 0.2 0.4
Sweden 47,747 51,934 5,579 5,869 1.0 3.1 35 38 36 485 54.1 5.7 6.1
Switzerland 24,998 26,218 3,724 3,699 -0.3 6.2 6.0 61 56 42.7 44.2 0.4 6.3
SyrianArabRepublic 11,028 14,642 984 983 -0 3 24 30 -89 -124 35.8 44.3 3.0 3.1
Tajikistan 3,266 3,384 616 562 -820 4.0 16 43 63 21.3 5.6 3.8 1.0
Tanzania 12,529 14.258 492 455 -1 1 09 10 6 5 2.3 2.4 0.1 0.1
Thailand 43.706 79,063 786 1,310 8.5 49 4.7 39 42 95.7 205.4 1.7 3.4
Togo .. . .. .0.7 068 02 0.2
Turisia 5,683 6,805 697 738 11 5.5 7.2 -7 2 13.3 162 1.6 16
Turkey 52,408 71,273 935 1,142 29 50 5.7 51 61 143.8 178.3 2.6 2.9
Turrnmenistan 18,923 12,181 5,159 2,615 -6.7 11 1.0 -293 -54 34.2 34.2 8.5 7.4
Uganda . ... . . . 06 1.0 0.1 0.1
Ukraine 252,631 150,059 4,868 2,060 -7.5 1.3 1.1 46 46 631.1 397.3 12.1 7.6
UnitedKingdom 213,090 227,077 3,702 3,663 0.7 4.4 53 2 -16 563.3 557.0 9.8 9.5
UnitedStates 1,025,680 2,162,190 7,720 6,076 0.8 29 3.6 14 22 4,824.0 5,301.0 19.3 200a
Uruguay 2,233 2.883 719 883 2.2 82 97 46 62 39 5.6 1.3 1.7
Uzbekistan 43,607 42,553 2,130 1,798 -3.1 11 1.1 12 -15 106.5 95.0 5.0 41
Veeezuela,
8B 40,851 57,530 2.095 2,526 13 24 2.4 -221 -255 1136 1445 5.8 6.5
Vietnam 24.451 39,306 360 521 4.0 2.7 3.2 -1 -11 22.5 376 0.3 05
Yemen,Rep. 2,665 3,355 224 208 -1.8 3.0 35 -267 -469 .
Zambia 5,220 5,087 671 634 -1 0 1.1 1.2 8 7 2.4 2.4 0.3 0.3
Zimbabwe 8,934 9,926 917 866 -1 2 26 3.1 9 18 16.6 184 17 1.6
World 8,6008414
t 9,431,190
t 1.105
w 1.692
w 0.0w ..w .w ..w ..w 16,183.16t
22,690.1
t 3.3w 4.0w
Lowincome 1.122,683 1,194,696 607 563 -1.2.. . -17 -9 1,376
8 2,433.8 0.7 1.1
Middleincome 3,297,t30 3,523.253 1,397 1,368 -0.2 -28 -33 5,772
8 9,524.1 27 37
Lowermiddleincome 2,426,017 2,384,656 1,302 1,178 -1.2 -16 -20 3,721.6 6,734.6 2.2 33
Uppermiddleincome 870,913 1,138,397 1,753 2,068 2.2 -63 -65 2,051.2 2,789.6 4.3 51
Lowandmiddle income 4.420,513 4,717,949 1.049 1,005 -0.H -26 -28 7,150.8 11,959.5 1.8 2.5
E-ast
Asia&Pacific 1,188.126 1,647,162 743 042 38 . .3,289.6 4,717.5 2.0 2.7
Europe &CentralAsia 1,799,836 1,240,586 3,966 2,690 -56 18 22 -5 -13 924.8 3,448 9 9.1 7.3
LatinAmerica& Caribbean 457,439 575,389 1,057 1,161 1.4 -31 -35 966.4 1,207.5 22 2.5
Middle East& North
Africa 266,687 374,375 1,134 1,354 2.1 3.3 3.3 -266 -225 737.6 967.2 3.3 3.9
SoothAsia 435,330 556,496 394 443 19 . .. 9 15 765.9 1,125 1 07 0.9
Sub-Saharan Africa 273,093 323,921 705 695 -0.2 .465.3 471.7 0.9 08
Highincome 4,187,901 4,713,241 4,996 5,369 1.1 . . 23 24 9,033.5 10,732.1 11.9 12.3
Ndote:Fordatacomparability
andcoverage, NotesFiguresinitalicsareforyears
nonthe Technicol otherthan those,specified
a.A negative
valueindicates
thatacountry
isa netexporter
 01 Dg C [OODa.Ž-Dn n 03 0 2~ -z = n DnD -Do 'C&C BZs C D B1DCDCD
WWWDCDCD0 BB-D m

o ~~~~~~~~~'cc
CDS'D
am z "wo BcDCD 3D1-C
F o
° r O--<=C Q1 DCC .
=

DCD
0
~~~
~ CD
0
-'CDD)
CD
0
~~Ž- CDCDCDCCDDC
- a a~~ a - - ---- - Bm C-''' -) -P-

•i~~~~~gCD--IS--CD-- - - -- - -- -- - - - - - -D -C -D -Ž--.-- C-

rno eow w e- c -C -D -D - - - - - -- -' - -C- a- , C- - '- C- - D- -C 1- C-- -D - - t-

01 CD~~~~~~~~~~~~
CD - CD--- --- ---

'DC -t\)0)
CID
I
-:
O ~ CS cn: e-0 : CD
b
o-1 en
e c
CD oX
co
a
: (D
CD
nWO
CD
CD
o
_-
C-
-H - -* -,, - -
n
--
-D - - - - - - - - -
acn
- -
_ ^>>__s I _;, e,<,, _=

2.
CD] CC CD C D- a CD - - - - - -- - -

0
h-c r4m 4011 Ž-4ŽI
mt 2CO-
c.) CD-$e
Ža
- ID C 01010101N-Ž- ' nCDDD ND NCC ON .cnDDC- mD h-Df 1-C4 cn -n 01-DC ].' -'N -0h- .DC1.ŽD, N4DYDCDOM 4D.- 1ŽMm-~ o1 oI

I CD CD a- -0- 1:mmeI- Oc1uo1s°I

011- -'0: 4Žh.Ž-1- 01ND01 010'1 Ž-J h--D-Ž-C01-'h- 0011-C F 4 :':C >D01DCDOC cn O Oh-i 01Ž- CD-'CDNOnJ 001011--- 01- OhDOC m4
-- IND a

o
Mr cs9:o x 4 > cov - cer1cov_v4N
Nco_
{s
>o
4ore
;m
A>
xco
vm
MceM
^torsoa so°eD
tauM
oX
0 0
scDh-01.a-4Ž--.JNDO'DDCD
N,D01NJCCDCCD -DŽ-----Nv]---'CCD
.D.0-'N4¢ Že-
Dh
-e NCOCDGD
CD
1 C]Ž-iC
4101 h-D
CDOCD-IŽ-h-D
-DDOOCD¾CD __JF-4h-JDDO4
-'OCDŽ-'DCDOCD (-CODOC oCDDCD01O01
01Ž-i DCDh-Dh-DŽ-DŽ-4 -00CD01 CD

LNo
M _ =
4_nm
[c-o11 M v | e ooTz
 SELECTED WORLD DEVELOPMENT INDICATORS 29 5

ECONOMY
Averageannual% growth
Gross GOP Agriculture Industry Services Exportsof goodsGross
domestic
domestic
product implicit
deflator valueadded valueadded valueadded andservices investment
Economy 1980-90 199G-99 1980-90 1990-99 1980-90 1990-99 1980-90 199D-99 1980-90 1990-99 1980-90 1990-99 1990-99
Mali H. 3 4.5 8.5 3.3 2.8 4.3 94 1.9 2.7 4.8 99 -0.8
Mauritania 1.8 4.1 8.4 8.1 1.7 5.2 4.9 28 0.4 48 21 1.6 6.8
Mexico 1.1 27 715 18.3 08 13 11 3.8 1.4 2.4 720 143 3.9
Moldova 3.8 -11 5 .. 142.5 .. -62 .. -118 .. -14 8 .. 48 -20 0
Mongolia 54 07 -1.6 08.5 1.4 2.9 6.6 -1.8 5.9 0.8-
Morocco 4.2 2.3 7.1 3.2 07 00 30 317 4.2 25 57 3.0 15
Mozambique -071 6.3 30.3 364 6.6 52 -4.5 99 9.1 55 -8.8 134 131
Myanmar 0.1 63 12.2 25.9 0.5 49 1.5 101I 0.8 66 19 75 147
Namibia 13 34 13.7 98 1.9 3.8 -0.6 25 2.3 34 0.7 4.3 2.5
Nepal 4.0 48 11.1 806 4.0 23 0.7 70 39 60 3.9 14.3 5.7
Netherlands 23 2.7 18 21 3.4 37 1.6 12 26 23 45 4.8 15
NewZealand 17 209 10.8 15 3.8 26 1.1 3.5 1.8 34 4.0 54 81
Nicaragua -109 3.2 4223 38.8 -2.2 54 -2.3 4.1 -15 1.1 -3 9 10.3 126
Niger -0 1 2.5 19 6.4 1.7 33 -1.7 1.9 -07 1.9 -29 1.7 54
Nigeria 1.6 24 16.7 348 3.3 29 -1.1 17 3.7 31 -0.3 2.5 5.8
Norway 28 3.7 5.6 1.8 -0.2 417 33 5.5 27 32 52 61 51I
Pakistan 6.3 40 6.7 10.7 4.3 43 7.3 49 6.8 46 8.4 2.7 2.1
Panama 0.5 42 1.9 20 2.5 17 -1.3 55 0.7 42 -0.9 0.0 12.1
PapsuaNew Guinea 1.9 4.0 5.3 78 108 2.3 19 6.9 20 28 33 9.5 68
Paraguay 25 2.4 24.4 137 306 2.8 03 2.8 31 2.0 122 5.1 1.5
Peru -0 3 54 231.3 287 2.7 58 -0.9 07 -0 7 40 -1.6 90 9.0
Philippines 10 32 14.9 84 1.0 15 -009 34 28 39 3.5 9.8 4.1
Poland 2.2 47 53.5 24 5 -0.4 00 0.3 63 28 38 6.6 106 11.9
Portugal 3.1 25 180 53 . -0 4 . 07 . 22 8.7 56 3.5
Romania 0.5 -1 2 25 105.5 . -0 5 -1.6 -1 0 61 -11 8
Russian Federation . -6.1 .. 189.6 . -6.3 -9.8 -1.8 . 23 -13 3
Rwanda 2.2 -1 5 4.0 163 05 -3.9 25 2.0 55 -1 2 34 -6 0 21
SaudiArabia 0.0 16 -40 1.4 13.4 07 -2.3 15 1.3 2.0
Senegal 3.1 32 6.5 52 2.8 1.6 4.3 4.6 208 34 3.7 2.6 31
SierraLeone 1.2 -408 62.8 312 3.1 1.63 17 -7.1 -2 7 -5.3 02 -122 -10.3
Singapore 6.7 60 1.9 16 -6 2 0.4 53 7.9 76 8.0 .. 5
SlovakRepublic 20 19 18 10.0 1.6 -0.2 20 -4.9 0.8 8.0 120 46
Sloveniia 2.4 .. 23.5 . 0.1 2.0 3.8 .. -0.5 10.2
SouthAfrica 1.0 1.0 15.5 10.2 209 1.0 07 0.9 24 2.4 189 5.3 30
Spain 3.0 22 0.3 40 25 .. 57 10.9 -0 5
SriLanka 4.0 5.3 110 9.7 22 1.5 46 7.4 47 5.6 4.9 8.4 6.2
Sweden 2.3 15 7.4 2.1 15 . 2.8 . 2.2 4.3 83 -2.2
Switzerland 2.0 0.5 34 1.5 .... 35 2.2 -0 4
SyrianArabRepublic 1.5 5.7 153 8.7 -0 6 . 06 .. a01 . 73 4.7 79
Tajikistan . -9.8 300.0 -12.2 -17.2 -10.7
Tanzania' 3.1 . 23.2 3.0 2.0 2.5 . 9.5 -1.7
Thailand 7.6 47 3.9 4.6 3.9 27 9.8 6.7 7.3 55 141 94 -2.9
Togo 1.7 2.5 48 8.3 56 4.5 11 2.6 -0.3 0.5 01 1.5 11.6
TLunisia 3.3 46 74 4.7 28 21 31 4.5 3.5 53 56 51 3.4
Turkey 54 4.1 452 77.9 13 1.0 7.8 4.8 4.4 4.3 . 11.9 4.6
Turkmenistan .. -3.5 . 622.8 .
Uganda 2.9 72 113.8 137 2.1 3.7 5.0 127 2.8 81 1.6 163 9.0
Ukraine . -10.8 . 440.0 -5 8 .. -15.5 -317 .. -3.6 -24.8
UnitedKingdom 32 2.2 57 29 . 39 6.0 1.8
UnitedSnates 3.0 3.4 4.2 18 25 . 49 . 21 47 93 70
UJruguay 04 37 61.3 380D 0.0 43 -0.2 17 0.8 4.5 4.3 7.0 89
Uzbekistan -2 0 3567 -1 0 . -517 -11I
Venezuela,RB 1.1 17 19.3 476 3.0 07 1.6 26 0.5 08 2.8 5.6 2.0
Vietnam 4.6 8.1 210.8 16.8 43 4.9 . 130 . 86 . 277 25.5
Yemen, Rep. . 30 . 260 . 50 . 7.8 . -1 8 . 10.2 77
Zambia 1.0 1.0 42.2 5689 3.6 -4 4 08B -4.3 -1 5 10.6 -3 4 1.8 11.3
Zimbabwe 3.6 24 11.6 238 3.1 43 3.2 -1 2 31 36 4.3 11.0 -0 7
World 3.2w 2.5w 2.7w 1.6w ..w 3.0w -.w 2.5w 5,2w 6.9w 2.9w
Lowincome 4.4 24 30 2.5 54 1.1 57 4.7 3.3 53 -1.4
Middle
income 3.2 35 3.5 20 3.6 4.4 35 36 7.3 88 40
Lowermiddleincome 40 34 4.0 20 6.1 52 5.3 37 7.6 67 35
Uppermiddleincome 2.5 36 28 1.9 2.3 3.9 28 3.6 7.0 108 4.4
Lowandmiddle income 3.4 33 3.4 22 38 3.9 38 3.7 6.6 8.2 30
EastAsia & Pacific 8.0 7.4 44 3.4 95 9.8 87 86 11.1 126 70
Europe& CentralAsia 2.4 -2.7. -3.0 . 3.5 0.5 . 4.4 -79
LatinAmerica& Caribbean 1.7 34 2.2 2.0 13 3.6 18 3.4 54 87 49
MiddleEast& NorthAfrica 2.0 3.0 55 2.5 03 2.1 22 3.5
SouthAsia 5.7 57 32 3.7 68 6.3 6.6 7.0 6.5 96 608
Sub-Saharan Africa 1.7 2.4 25 2.5 12 15 24 2.5 24 44 36
Highincome 3.1 24 0.8 2.6 2.2 50 6.5 2.6
Note.Fordatacomparabiiity
andcoverage, seethe TechnicalNotes.Figuresin italics arefor yearsothernoonthosespecified
a.Datapriorno1992includeEnitrea.
b. Datapriorto 1990refernothe Fuderal
Republic of Germany beforeunificationc Datacovermuainland
Tanzana only
296 WORLD DEVELOPMENT REPORT 2000/2001

Table12. Structureof output

Gross
domestic product Valueaddedas%of GDP
Millionsofdollars Agriculture Industry Manufacturing Services
Economy 1990 1999 1990 1999 1990 1999 1990 1999 1990 1999
Albania 2,102 3,058 37 54 47 25 .. 16 21
Algeria 61,902 47,015 14 13 45 54 12 11 41 33
Angola 10,260 5,861 18 7 41 70 5 63 41 23
Argentina 141,352 281,942 8 6 36 32 27 22 56 61
Armenia 4,124 1,911 17 33 52 32 33 22 31 35
Australia 297,204 389,691 3 29 15 . 67
Austria 159,499 208,949 3 32 . 23 - 65
Azerbaijan 9,837 4,457 19 43 . 6 38
Bangladesh 29,855 45,779 28 21 24 27 15 17 48 52
Belarus 34,911 25,693 24 13 47 46 39 39 29 40
Belgium 196,134 245,706 2 1 30 28 21 18 68 71
Benin 1,045 2,402 36 38 13 14 8 a 51 48
Bolivia 4,868 8,516 15 16 30 31 17 17 54 54
Botswana 3,766 5,996 5 4 56 45 5 5 39 51
Brazil 464,989 760,345 8 9 39 29 25 23 53 62
Bulgaria 20,726 12,103 18 18 51 27 20 31 55
BurkinaFaso 2,765 2,643 32 32 22 27 16 21 45 41
Burandi 1,132 701 56 52 19 17 13 9 25 30
Cambodia 1,115 3,117 56 51 11 15 5 6 33 35
Cameroon 11,152 8,781 25 44 29 20 15 11 46 36
Canada 572,673 612,049 3 33 18 - 64
CentralAfricanRepublic 1,488 1,053 48 55 20 20 11 9 33 25
Chad 1,739 1,574 29 38 18 14 14 11 53 48
Chile 30,307 71,092 8 8 39 33 19 16 53 59
China 354,644 991,203 27 17 42 50 33 24 31 33
HongKonc,China 74,784 158,611 0 0 25 15 18 6 74 85
Colombia 46,907 88,596 19 14 31 24 15 12 51 61
Congo, Dem.Rep. 9,348 6,964 30 58 28 17 11 . 42 25
Congo,Rep. 2,799 2,273 13 10 41 48 8 6 46 42
CostaRica 5,713 11,076 16 14 24 22 19 17 60 64
Coted'lvoire 10,796 11,223 32 24 23 24 21 20 44 52
Croatia 13,370 21,752 10 9 34 32 28 21 56 59
CzechRepub ic 34,880 56,379 8 4 45 39 48 57
Denmark 133,361 174,363 4 27 18 69
DlominicanRepublic 7,074 17,125 13 11 31 35 18 16 55 54
Ecuador 10,686 18,712 13 12 38 33 19 22 49 55
Egypt,ArabRep. 43,130 92,413 19 17 29 33 24 27 52 50
l Salvador 4,807 12,229 17 10 26 28 22 22 57 61
Eritrea 437 670 29 16 19 27 13 14 52 57
Estonia 6,760 5,101 17 6 50 27 42 16 34 66
Ethiopiaa 6,842 6,534 49 49 13 7 8 38 44
Finland 134,806 126,130 6 35 23 58
France 1,195,438 1,410,262 3 2 29 26 21 19 67 72
Georgia 12 4,192 32 22 33 13 24 12 35 65
Germany 1,719,510 2,081,202 1 1 , 29 24 34 36
Ghana 5,886 7,606 45 36 17 25 10 9 38 39
Greece 82,914 123,934 11 . 22 13 67
Guatemala 7,650 18,016 26 23 20 19 15 13 54 58
Guinea 2,818 3,693 24 23 33 36 5 4 43 41
Haiti 2,981 3,871 33 30 22 20 16 7 45 50
Honduras 3,049 5,342 22 18 26 30 16 18 51 52
Hungary 33,056 46,355 15 6 39 34 23 25 46 60
India 322,737 459,765 31 28 27 25 17 16 42 46
Indonesia 114,426 140,964 19 20 39 45 21 25 41 35
Iran, Islamic Rep, 120,404 101,073 24 . 29 . 12 . 48
Ireland 45,527 84,861 8 . . .. . . 53
Israel 52,490 99,068 . . ..
Italy 1,093,947 1,149,958 3 3 33 31 22 20 63 67
Jamaica 4,239 6,134 6 8 43 33 20 15 50 59
Japan 2,970,043 4,395,083 3 2 41 37 28 24 56 61
Jordan 4,020 7,616 8 2 28 27 15 15 64 71
Kazakhstan 40,304 15,594 27 10 45 30 9 23 29 60
Kenya 8,533 10,603 29 27 19 17 12 11 52 56
Korea,Rep. 252,622 406,940 9 5 43 44 29 32 48 51
Kuwait 18,428 29,572 1 52 .. 12 .. 47
KyrgyzRepublic 1,629 35 44 36 22 28 19 29 35
LaoPDR 865 1,373 61 53 15 22 10 17 24 25
Latvia 12,490 6,664 22 5 46 33 34 22 32 63
Lebanon 2,B38 17,229 . 12 .. 27 17 .. 61
Lesotho 622 874 23 18 34 38 .. 43 44
Lithuania 13,264 10,454 27 10 31 33 21 19 42 57
Macedonia,FYR 2,635 3,445 14 11 32 28 .. .. 54 60
Madagascar 3,081 3,733 32 30 14 14 12 11 53 56
Ma awi 1,803 1,820 45 38 29 18 19 14 26 45
Malaysia 42,775 74,634 19 14 40 44 26 35 41 43
 SELECTED WORLD DEVELOPMENT INDICATORS 297

ECONOMY
Gross
domestic
product Value
added
as%ofGDP
Millions
ofdollars Agriculture Industry Manufacturing Services
Economy 1990 1999 1990 1999 1990 1999 1990 1999 1990 1999
Mali 2,421 2,714 46 47 16 17 9 4 39 37
Mauritania 1,020 959 30 25 29 29 10 10 42 46
Mexico) 262,710 474,951 7 5 26 27 19 21 67 68
Moldova 10,583 1,092 43 21 33 24 .. 18 24 55
Mongolia .. 905 15 33 41 28 . . 44 40
Morocco 25,821 35.238 18 17 32 32 18 17 50 51
Mozambique 2,512 4,169 37 32 18 24 10 13 44 44
Myanmar .. 57 53 11 9 8 8 32 38
Narmibia 2,340 3,075 12 13 38 33 14 15 50 55
Nepal 3,628 4,904 52 41 16 22 6 9 32 37
Netherlands 283,872 384,786 4 29 -. 19 .. 67
NewZealand 43,103 53,622 7 26 .1 8 .. 67
Nicaragua 1,009 2,302 31 26 21 21 17 14 48 53
Niger 2,481 2,967 35 40 16 17 7 6 49 43
Nigeria 28,472 43,286 33 41 41 62 6 5 26 -3
Norway 115,453 145,449 3 2 31 32 12 11 66 88
Pakistan 40,010 59,880 26 26 25 25 17 17 49 49
Panama 5,313 9,606 160 8 16 18 10 9 73 74
PapuaNewGuinea 3,221 3,571 29 21 30 15 9 8 41 64
Paraguay 5.265 8.065 28 26 25 22 17 16 47 52
Peru 32,802 57.318 7 8 38 39 27 24 55 54
Philippines 44,331 75.350 22 17 34 31 25 21 44 52
Poland 61,197 154,146 8 4 48 33 20 44 63
Portugal 69,132 107,716 6 .. 37 .. 27 57-
Romania 38,299 33,750 20 16 50 40 .. 30 30 44
RussianFederation 579,068 375,345 17 7 48 34 35 58
Rwanda 2,584 1,956 33 46 25 20 19 12 42 34
SaudiArabia 104,870 128,89q2 6 7 50 48 8 10 43 45
Senegal 5,698 4,791 20 18 19 25 13 17 61 57
SierraLeone 897 669 47 44 20 24 4 4 33 32
Singapore 36,638 84,945 0 0 35 36 27 26 65 64
SlovakRepublic 15,485 19,307 7 4 59 32 .. 23 33 64
Slovenia 1Z673 20,653 6 4 46 39 35 28 49 57
SouthAfrica 111,997 131,127 5 4 40 32 24 19 55 64
Spain 491,938 562,245 5 .. 35 .. 23 60-
Sri Lanka 8,032 15,707 26 21 26 28 15 17 48 51
Sweden 229,756 226.388 3 34 22 63
Switzerland 228,415 260,299 . ..
SyrianArabRepublic 12,309 19.380 29 .. 24 48
Tajikistan 4.857 1,778 27 6 34 30 39 65
Tanzaniab 4,220 8.777 48 48 16 14 9 7 36 38
Thail-and 85,345 123.807 12 13 37 40 27 32 50 49
Togo 1,628 1,506 34 43 23 21 10 9 44 36
Tunisia 12.291 21,188 10 13 30 28 17 18 54 59
Turkey 150,721 188,374 18 1 02 0 1 256
Turkmenistan 6,333 2,708 32 25 30 42 .. 29 38 34
Uganda 4,304 6,349 57 44 11 18 6 9 32 38
Ukraine 91,327 42,415 26 14 45 34 36 29 30 51
UnitedKingdom 975,512 1,373,612 2 .. 35 .. 23 63
UnitedStaten 5,554,100 8,708,870 2 2 28 26 19 18 70 72
Uruguay 8,355 20,211 11 9 32 29 26 19 57 63
Uzbekistan 23,673 16,844 33 31 33 27 13 34 42
Venezuela, RB 48,593 103,918 5 5 50 24 20 12 44 71
Vietnam 6,472 28,567 37 26 23 33 19 .. 40 42
Yemen, Rep 4,660 6,769 27 17 30 49 10 11 43 34
Zambia 3,288 3,325 18 17 45 26 32 11 37 57
Zimbabwe 8,784 5,710 16 19 33 24 23 17 50 56
World 21,39064t 30,211,993
t 6w 4w 34w 32w 22w 21w 60w 61w
Lowincome 889.723 1.067,242 29 27 31 30 18 18 41 43
Middleincome 3,525,445 5,488,604 13 10 39 36 25 23 47 55
Lowermiddleincome 1,820,097 2,575,942 21 15 39 40 26 23 40 46
Uppermiddleincome 1,722,041 2,918,403 8 7 39 32 24 24 53 61
Lowendmiddle income 4,413,061 6,557,913 16 12 38 35 23 22 46 54
EastAsia& Pacific 925,765 1,888,729 20 13 40 46 29 28 40 41
Eurspe& Central Asia 1,240,214 1,093,237 17 10 43 32 . .40 58
LatinAmerica& Caribbean 1,146,895 2,055,025 9 8 36 29 23 21 56 63
MiddleEast& NorthAfrica 402,799 596,253 15 .. 38 .13 .. 47
SouthAsia 410,341 595,915 30 28 26 25 17 16 44 47
Sub-Saharan Africa 297,397 332,744 18 18 34 32 17 17 48 50
Highincome 16,967,888 23,662,676 3 2 33 30 22 21 64 84
Note:Fordatacomparability
andcoverage, seetbe Technical
Notes,Figuresin italicsarefor yearsotherthanthosespecified.
a.Datapriorto 1992includeEritrea.b.DatacovermainlandTanzania
only.
 .. s00co ,-r00 000-C-- 00-CC N 7r cr-c 0=0' OC-C-"-O rC-C - OCCOf COPn_ 00i OCCOC NO rcO rc> - O'--'-
00 = Y7ciu== 1[ =9 ,- x ,7 e
l74c>] 7 7 sr r
rN
f IX I} ~ 00000
I= r r -5)
r
T = ? [= n rxr7

'|>eZ
0i '
Yr R 0 0CC- P~ - x9Cn0oC-o 0COC)r-C C c -;- -) ' 0 c = 0fi00CflYCg
¢>-N\ -C- - -P'=-0-- Cc-- -c -= - - 0- 0--- - *- - -0-0-

I . ?I
.0 y - -
>-m- ---- = r -= ' ; P
,,, - *_ , -- 0

.N, rCC0 .. 0r:0r rN,.A- C-k>F. 0; aY0 ccrNFC 000 ONCc r00c 00-,RY;>Yb:. 00cc- Yrcc NC a3-rNCr.-r 000 r00 Yr N-0Ot<

'0E

;YT <-- < - N 0- - O O-N - N N0 -0 0IO-0- c

:~~~~~~~~~ R 9II PR T _2 ~~~~~

CDre : a, c=

8
- E oOOccccDc ° o0ccN NO-ccO -00000
°
CEOC-CO -0-0(00--C 00000 OctN(-cc 0°000 cc0n-0C0 cc'ONN cccco0z0 000>0CC- cccc0cN N0-E
 SELECTED WORLD DEVELOPMENT INDICATORS 299

ECONOMY
Percentageof GDP
Private Generalgovernment Grossdomestic Grossdomestic Exportsof goods Resource
consumption consumption investment savings andservices balance
Economy 1990 1999 1990 1999 1990 1999 1990 1999 1990 1999 1990 1999
Mali 80 80 14 12 23 20 6 8 17 22 -17 -12
Mauritania 69 73 26 15 20 22 5 12 46 39 -15 -1O
Mexico 70 70 8 7 23 24 22 23 19 31 -1 -1
Moldova 63 92 14 12 25 18 23 -4 49 49 -2 -21
Mongolia 57 63 30 18 34 26 13 20 21 50 -21 -6
Morocco 69 67 15 15 25 23 16 18 19 29 -10 -5
Mozambique 101 79 12 10 16 35 -12 11 8 13 -28 -24
Myanmar 89 89 b b 13 12 11 11 3 1 -2 -1
Namibia 51 64 31 26 34 20 18 9 52 53 -16 -11
Nepal 83 80 9 10 18 19 8 11 11 22 -11 -9
Netherlands 59 59 15 14 22 20 27 27 54 56 5 7
NewZealand 63 63 17 15 19 21 20 21 28 29 1 1
Nicaragua 59 65 43 14 19 37 -2 1 25 37 -21 -36
Niger 84 83 15 13 6 10 1 4 15 15 -7 -6
Nigeria 56 88 15 12 15 11 29 0 43 17 15 -11
Norway 49 48 21 20 23 25 30 32 41 41 7 7
Pakistan 74 78 15 11 19 15 11 11 16 15 -8 -4
Panama 60 59 18 16 17 34 21 25 38 34 5 -9
PapuaNewGuinea 59 48 25 15 24 36 16 37 41 66 -8 1
Paraguay 78 73 5 10 22 19 10 17 23 41 -6 -2
Peru 70 65 8 15 21 22 22 20 12 14 0 -2
Philippines 72 68 10 16 24 21 18 16 28 56 -6 -5
Poland 50 74 19 9 25 28 32 18 28 20 7 -10
Portugal 63 64 16 19 29 26 21 17 34 31 -7 -9
Romania 66 81 13 9 30 15 21 10 17 29 -9 -5
Russian Federation 49 57 21 14 30 14 30 29 18 48 0 15
Rwanda 84 89 10 13 15 14 6 -1 6 6 -8 -16
SaudiArabia 40 41 31 32 20 21 30 26 46 36 10 5
Senegal 76 76 15 10 14 21 9 14 25 34 -5 -7
SierraLeone 82 93 10 13 9 5 8 2 24 14 -1 8
Singapore 46 39 10 10 37 33 44 52 202 7 19
SlovakRepublic 54 50 22 22 33 30 24 28 27 64 -9 -11
Slovenia 55 56 19 21 17 25 26 24 84 57 9 -1
SouthAfrica 63 63 20 19 12 16 18 18 24 25 6 3
Spain 62 62 16 16 25 21 22 22 17 28 -3 1
SriLanka 76 71 10 10 22 25 14 19 30 36 -8 -6
Sweden 51 53 27 26 21 14 22 21 30 44 0 7
Switzerland 57 61 14 14 28 20 29 25 36 40 1 4
SyrianArabRepublic 70 70 14 11 15 29 16 18 28 29 0 -11
Talikistan 65 21 23 .. 14 .. -10
Tanzania' 84 72 17 13 23 18 -1 14 12 20 -23 -4
Thai;and 57 57 9 11 41 21 34 32 34 57 -8 12
Togo 71 83 14 12 27 14 15 6 33 32 -12 -8
Tunisia 58 63 16 12 32 28 25 24 44 42 -7 -3
Turkey 69 68 11 11 24 24 20 21 13 26 -4 -3
Ttrkmenistan 49 23 40 28 .. -13
Uganda 92 84 8 10 13 17 1 6 7 11 -12 -12
Ukraine 57 56 17 26 27 21 26 18 28 40 -1 -3
UnitedKingdom 63 64 21 20 19 16 17 15 24 29 -3 0
UnitedStates 67 68 18 15 17 10 15 17 10 12 -1 -1
Uruguay 69 78 14 9 11 14 17 13 26 19 6 -1
Uzbekistan 61 59 25 22 32 19 13 19 29 22 -19 0
Venezuela,
RB 62 77 8 6 10 15 29 17 39 21 19 3
Vietnam 86 71 8 8 13 29 6 21 26 44 7 -7
Yemen,Rep 73 72 18 15 15 21 9 13 15 37 -6 -B
Zambia 64 85 19 10 17 17 17 6 36 29 -1 -11
Zimbabwe 63 69 19 16 17 18 17 15 23 46 0 -2
World S1w 62w 15w 15w 24w 22w 23w 23w 19w 22w Ow Ow
Lowincome 66 70 12 11 24 20 21 19 17 27 -3 -1
Middleincome 59 62 14 12 26 24 27 26 21 28 1 1
Lowermiddleincome 57 59 13 11 31 27 30 30 21 32 -1 3
Uppermiddleincome 60 65 15 12 23 22 25 23 21 25 2 0
Lowandmiddleincome 60 63 14 12 26 24 26 25 21 26 0 1
EastAsia& Pacific 54 53 11 10 35 33 35 37 26 39 0 5
Europe& CentralAsia 55 64 18 12 28 20 26 23 23 38 -1 4
LatinAmerica& Caribbean 65 68 13 13 19 21 22 20 14 16 2 -1
MiddleEast& NorthAfrica 58 60 20 21 24 22 22 19 33 25 -2 -3
SouthAsia 69 71 11 10 23 22 19 19 9 12 -4 -4
Sub-SaharanAfrica 66 69 18 16 15 17 10 14 27 27 2 -2
Highincome 62 62 16 15 23 21 23 22 19 22 0 1
Note:`ondatacomparability
andcDverage, seetheTechnicalNotes.Figuresin italicsarefor yearsotherthanthosespecified
a Datapriorto1992includeEritrea.b.General government
consumpton figuresarenotavailableseparately; theyareincludedin privateconsumption.
c. Datacovermainland
Tanzaniaonly.
300 WORLD DEVELOPMENT REPORT 2000/2001

Table14. Centralgovernmentfinances
Percentage
ofGDP Percentage
oftotalexpenditureb
Current Current Current Capital Overall Goods Social
taxrevenue nontaxrevenue expenditure expenditure dleficit/surplus' andservices servicesc
Economy 1990 1998 1990 1998 1990 1998 1990 1998 1998 1998 1990 1998 1990 1998
Albania 14.8 4.5 251 . 4.7 -4.5 185 28.8
Algera 30.7 1.5 . 21 5 77 2,9
Angala- -
Argenitna 94 1214 10 11 19.1 141 0.5 12 -04 -7,5 29.7 214 571 6365
Armenia. .. ....

Ausnrala 238 229 21 1.0 22.2 233 2.1 1.3 2.1 2.9 27.4 30.6 50.7 66.3
Austia 314 348 30 2,5 34.0 37 9 3.3 2.6 -4.5 -27 24.5 22,6 68A4 682
Azercaijan - 192 . 1.1 1908 5.3 .. -3.9 34. 37.6
Bangladlesh
Belarus - 30.8 28.7 0.5 2.1 31.7 27.5 59 47 -5.1 -0.9 36.2 254 572 43.7
Belgium 41.5 43.3 1.5 0.7 45.0 444 2.3 22 -5 6 -2 0 18.4 19.1

Boaivia 0.4 15.1 5.2 24 139 18.5 2.5 34 -1.7 -2.3 53.2 40.0 381 51.2
Botsawana 277 14.7 234 29.5 26.5 285 7.3 68 113 0.4 41.4 4648 33.9 42.7
Brazil 19,0 37 . 46.1 . 0.7 -. -5.8 .. 147 . 33.0
Bmigara 34.5 27.0 12.9 609 53.5 30.5 1. 3.8 -9.2 2.83 33.5 32.9 30.3 450
Burkena Fasa 101 08 -. 11.0 3A
4 -1.3 . 60.3 20.5
Burundi 183 17.7 1.9 1.0 14.2 17,3 11.8 37 -33 -5 5 32.0 55.2 22.3 2320
Cambondia . .. .
Cameracn 108 .. 43 .. 14.7 . 5,5 . -5.9 . 51.0 . 30.1
Canada 108.3 . 2.7 . 25 3 05 -4.9 21.3 . 477
CentreAfricanRepablic
Chad 96 0.5 . 9.5 123 -47 41.4
ChIn 18.3 18.4 4.3 3.6 181 14.0 23 36 9.8 04 28,5 2808 83.9 71.3
China 4.0 57 2.4 02 -..
. -109 -1.5 . .. 2.5 25
Hang Konag. Caina . . . . . .
Colombai 008 101 21 1lA 7.7 12.8 22 32 34 -4 7 25.1 29.4 321 45,2
Cango,GemRep 94 43 0.7 1.0 15.7 99 3.1 05 -6.5 -08 72.9 94.5 6.17 0.5
Canga, Rep. 14.7 6,6 70 228 34.8 32 9 0.8 4.1 -14.1 -8.6 54.4 49.7
CasteRica 19.7 2321 3.3 322 22.7 272 2.9 2.9 -3.1 -2.8 55.4 4717 58.7 59.6
Cotea caoirn 100 21.0 2.1 0.9 24.5 17.1 0.9 7.1 -2.9 -1.3 68.8 44.0 30.0-
Croaaaia 31.9 43.3 11l 2.1 364 40.8 J.2 4.93 -4.6 0.6 53,9 48.7 6328 63.5
CzechRepublic .. 31.9 .. 1. 330 - 3.0 -. -1.6 . 14.3 . 67.1
CernmarK 32.3 .. 5.5 . 377 . 1.3 . -0.7 . 20.3 .. 50.5-
Dominicanr_Republic
_ __ 10.8 155 1.2 1.4 8.5 11.6 5.1 43 061 04 3809 43,3 44.0 44.2
Ecuador 17.0 0.4 7119 . 206 . 37 .. 41.5 .. 321
Egypt,Awab Rep 10.7 16.6 5.2 9.7 230 2323 4.0 74 -5.7 -2 0 37,2 40-9 32.1 23,6
El Salnadar
Ertrea
Entonia 256 29.0 0.6 '.0 21.9 33.1 1.8 2.8 04 -0 1 22,3 42.9 48q,7 55.5
tEhiappem 12.2 . 5.1 . 24.6 . 4.4 -0.0 . 75.3 .. 21.6
Finlana 283 24.1 208 3.9 207 34 2 2.1 11 02 -25 10.6 19,0 60.8 55.3
Franca 3756 39.2 2.8 285 40.0 44.6 2.5 2,0 -2.1 -3 5 25.7 22,6 68.2
Georgia . 4.06 . 1.03 . 02 . 0.4 . -2.5 .. 45.5 . 26.1
Germany____27.0 26.9 1.4 5.1 28.2 3196 19 1.4 -2.2 -0.9 32.0 318 95.0 69.8
Ghanna 11.4 .. 1.9 107 .. 2.5 0.2 .. 48.2 . 46.4
Greece 28.0 2085 2.3 2~4 4809 284 41 4.3 -23.2 -4.4 31.5 203 32.1 35,0
Guatranala.. . .10
Guinea 11.5 1010 4.6 94 1008 1. 2 -3.3 41 36.7 3490
Hla ri _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Honduras . . . . . . .
Hungary 44 7 31.4 0.2 42 50.2 30.4 1.9 4.9 0.0 -6.1 36.5 16.0 46.5 46.85
Inea 9.0 896 24 3.0 14.2 120 1.8 1.6 -7.5 -5.2 10.2 20.1 0.1 9.2
manoner a 17.0 159 1.0 1.2 10.4 12.2 0.0 5.7 0.4 -2.4 22.9 18.9 13.2 26.2
IraniIslamicRep. 72 11.2 10.0 15.3 15.0 177 4.0 0.0 -1.8 0.3 53.6 54.5 40.3 41.8
Ireanc 32.8 31.6 2.4 1.8 393 322 20 322 -2.1 -0.4 18.08 18 1 553 548.
lsame 33.7 35.8 5.7 0.4 47.3 45.1 3.0 3.0 -5.3 -1.2 37.4 33.2 30.2 57.6
ta,y 37~3 3086 12 20 4308 4272 4,0 2A -19.3 -33 16.5 10.2
Jamaica
J.apan 13.7 00 137 2,0 .. -19 1386 52.0
Jordan 10.8 19.8 7,5 67 301 24.3 5.0 57 -3.5 -23 54.5 62.6 36.6 44.6
Knazkl,stan.. ..... ....
Kenya 202z 2325 2.2 2.7 22.0 2556 5.5 24 -3.8 -0 9 40.7 44.5 20.5 29q6
KareaRep. 1509 173 17 27 13.8 7137 24 328 -0.7 -173 30.1 21,6 27.8 278
Kuvvait 15 1,5 . 45 1 440 10.2 8.6 ..

KyrgyzRepualic-
LasPOP
Latvai 20.0 .. 38 . 30 . 2.5 0.1 .. 29.0 50.4
Lebanona. 12.7 . 4.3 26.0 . 6.1 -15 1 . 20.7 .. 10.4
Leacaha 34.7 34.9 4.3 0.8 283 40.9 220 07 -1 0 -3 7 39 2 75.6 34.3 35.7
Liahnranin 293 25.4 2,6 1.3 231 27.2 5.9 3.1 1.4 -04 11.7 54.5 . 86.1
Meacedonia, FYR. . . . ... ...... .
Madagascar 0.4 4.5 21 02 01 10.5 6. 4
68 -0.0 -1.3 35.3 24.6 22.4 16.5
Manac 10,0 27 20,2 . 6.4 . -1.7 -. 54.6 .. 22.4
M-anans 10.6 18,9 75 41 23.3 15.2 7.3 45 -2.1 2.9 42.6 40.5 35.6 42.5
 SELECTED WORLD DEVELOPMENT INDICATORS 30I

ECONOMY
0
Percentage
of C3DP Percentage
oftotalexpenditure
Current Current CurrenFt Capital Overall Goods Social
taxrevenue nontaxrevenue expenditure expenditure deficit/surplusOandservices servicesc
Economy 1990 1998 1990 1998 1990 1998 1990 1998 1990 1998 1990 1998 1990 1998
Mali
Mauritania - . . .
Mexico 13.7 130 1.6 1.7 15.5 14.3 2.5 1.9 -25 -1.1 247 23.5 30.6 48 1
Moldova . . . . .
Mongolia 15.6 13.5 2.7 6.0 187 19.8 2.9 32 -6 0 -10.8 242 23.6 247 254
Morocco 22.9 . 3.5 .. 26.8 .. 8.0 -2.2 -. 47.69 . 27.0
Mozambique - . -
Myanmar 6.2 4.5 43 3.3 11.4 4.3 4.6 4.5 -5.1 -0.9 . .. 35.7 15.5
Namibia 27.4 .. 4.0 - 28.4 -. 5.1 . -1.2 -- 71.7
Nepal 7.0 8.8 1.4 1.8 . .. . .. -6.8 -4.7 . .. 24.4 28.9
Netherlands 428 42.7 4.2 3.0 48.6 46.0 3.6 1.7 -4.5 -1.7 15.2 15.4 64.6 63.9q
NewZealand 36.3 32.1 82 2.1 43.1 32.5 0.6 09 4.0 0.5 22.1 50.9 69.2 71.0
Nicaragua 29.3 .. 4.3 . 688
9 3.1 .. -35.6 .. 42.6 . 36.2

Norway 32.3 34.1 102 91 39.2 34.7 2.1 1.6 0.5 0.7 18.8 17.3 440 42.1
Pakistan 133 126 58 33 19.8 188 2.8 25 -5.4 -6.3 38.7 48.5
P-anama 17.7 18.4 709 70 23.3 25 0 0.4 2.0 3.6 02 67.3 54.9 667 65.2
PapaaNewGuinea 20.1 . 5.2 . 31.0 3.7 . -3.5 . 58.7 .. 30.4
Paraguay 9.2 . 3.1 .. 7.8 .. 1.6 . 2.9 . 54.2 .. 31.5
Peru 8.4 13.7 0.6 2.3 151 13.8 1.3 2.6 -6.4 -0.2 30.0 41.8
Philippines 14.1 17.0D 2.1 2.0 165 163 31 2.2 -3.5 01 42.4 51.1 22.5 26.5
Poland . 32.8 . 29 I.. 35.7 . 2.0 .. -1.0 ..I 25.2 6.9.6
Portugal 28.4 32.1 4.0 3.6 34.3 356 4.5 5.2 -4.6 -2.1 38.0 39.5
Romania 36.9 24.4 3.5 21 27.9 29.1 5.9 2.9 69 -39 25.8 30.I 42 9 490
RussianFederatior . . . . . . . . .
Rwanda 9.5 1.3 .. 12.7 . 63 . -5.3 .. 52.9
SauidiArabia
Senegal.. . . ... . ...
SierraLeone 3.9 9.9 0.2 0.3 58 13.0 05 4.2 -1.8_ -5 8 76.2 39.0 25_.8
Singapore 15.4 16.2 11.5 8.4 16.4 11.8 5.1 5.1 10.8 11.8 49.8 36.7 32.2 23.2
Slonak Republic.. . .. .
Slovenia.. . ... .. .
SouthAfrica 24.3 24.5 2.0 1.7 27.0 28.4 3.1 1.2 -4 1 -2 9 51.9 26 6
Spain 28.8 28.1 1.6 2.0 30.8 34.2 32 1.9 -3.2 -5.5 18.7 15.9 50.6 48.3
Sri Lanka 19.0 14.5 2.0 2.7 22.3 19.7 61 5.3 -7.8 -8.8 30.2 37 1 27.5 30.0
Sweden 38.4 358 5.8 4.5 39 8 417 1.0 1.1 1.0 -1.6 13.7 14.5 61.8 53.2
Switzerland 19.4 22.0 14 1.7 22.1 27 0 1.2 1.0 -0. -1.3 29.9 29.1 66.4 729
SyrianArabRepublic 16.7 16.4 5.1 7.a 16.0 15.5 5.6 9.1I 03 -02 . .. 12.6 164
Tajikistan . . . . . ..

Tanzania ... . . ...

Thailand 17.1 144 15 108 115 117 2.6 67 4.6 -3.4 592 51.2 32.2 38.3
Togo . . . ..
Tunisia 24.0 24.8 6.7 4.8 27.0 25.9 7.6 6.7 -5.4 -3 1 31.7 37.9 36.9 46.6
Turkey 11.63 1917 2.1 28B 15.1 26.5 2.3 3.4 -3 0 -8.4 52.0) 32.8 26.3 25.7
Turkmenistan .. . . .. . . .. .
Uganda .. . .. . ... .
Ukraine.. . .. ... ... .
UnitedKingdom 33.3 36.3 3.2 2.0 34 2 36.3 3.8 1.5 0.6 0.6 31.1 28.3 52.8 57.5
UnitedSnates 18.0 204 1.5 14 21.7 20.4 1.8 06 -3.9 0.9 27.4 21.9 43.4 538
Uruguay 252 300 1.3 2.3 24.0 31.6 19 17 0.4 -0.8 346 31.7 61.6 75.8
Uzbekistan . . . . . . . . .
Venezuela.
RB 184 128 5.3 4.6 17.4 16.0 3.3 3.8 00 -2.8 27.2 23.9
Vietnam 15.8 . 2t .. 14.3 .. 5.8 . -1 1 . .. . 30.5
Yemen,
Rep. 11.2 13.7 8.3 23.1 228 31.2 9.6 6.8 -9.1 -2.3 61.9 43.5 219 22.4
Zambi-a .
Zimbabwe 21.7 26.4 2.4 3.0 24.5 _336 2.8 2.1 -5.3 -50 51.2 48.5 55.3
Note:Fordatacomparabilityandcoverage, oeetbe TechnicalNotes.Figuresin italicsareforyearnotherthanthosespecified.
a. Includes
grants.b.Totalexpenditure
includeslendingminusrepayments. c. Refers to education,
health,socialsecurity,
welfare,housing,andcommunity
amenities.
d. Datapriorto
1992 includeEritrea.
 CO m-c-m -- --- 0)0 - -0(0 c20Coo L., Z 5-> = - -(( C40c

cm mm m ETC ~~~~ ~--- ---- - - c-, - - c 0'-C

cC' cm 0 cm -~~~~~~~~~~~~~~~~~
0,0~C I

Ci- 01 CO~~~~~2-
- -0 - - 7cm

- - ----
0 '< CC
-c -0- - - - - -
CCi~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~0
CO 0~ ~ ~ ~ ~ ~ ~ ~~~~~~~~2
-- -- ---- - --- - --- -- - - - - -

- -- -- -- -- --- -- - - -- - -- - - -- -- - -- 0

---- - - --- -- -- - CD~

CD

0' -' CrNN)QN)NN)-

- - - -N- -C -. - - - - - - -00O-)0-00-C -o'-tC- C- --- 'r .0 -0

-~ ,tCO
-'C0'0Toro CO)Coo- 00 0 )mOC N C'O 0C CC'N)COCr 0-00CO)NN))0N)N) too-c co Ootc,Co mC-o o-

CO ~~~~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~ - - - - - - - - - - - - - - - - - -O
 SELECTED WORLD DEVELOPMENT INDICATORS 303

ECONOMY
Millionsof dollars
Goodsand services Netcurrent Currentaccount Grossinternational
Exports Imports Net income transfers balance reserves
Economy 1990 1998 1990 1998 1990 1998 1990 1998 1990 1998 1990 1999
Mali 420 637 830 899 -37 -45 225 126 -221 -178 198 350
Mauritania 471 393 520 471 -46 -32 86 187 -10 77 59 224
Mexico 48,805 129,523 51,915 138,441 -8,316 -13,056 3,975 6,014 -7,451 -15,960 10,217 31,782
Moldova .. 765 1,228 33 83 .. -347 0 186
Mongolia 493 540 1.096 671 -44 0 7 56 -640 -75 23 136
Morocco 6,239 9,970 7,783 11,358 -988 -1,101 2,336 2,345 -196 -144 2,338 5,689
Mozambique 229 531 996 1,132 -97 -141 448 313 -415 -429 232 654
Myanmar 641 1,634 1,1B2 2,789 -61 38 77 515 -526 -602 410 265
Namibia 1,220 1,605 1,584 1,908 37 61 354 403 28 162 50 305
Nepal 379 1,108 761 1,646 71 13 60 103 251 -421 354 843
Netherlands 160,447 224,762 147,652 200,897 -631 8,905 -2,943 -7,185 9,221 25,585 34,401 10,098
NewZealand 11,683 16,017 11,699 15,859 -1,576 -3,693 138 338 -1,453 -2,596 4,129 4,455
Nicaragua 392 761 682 1,656 -217 -151 262 . -305 . 166 510
Niger 533 332 728 479 -54 -24 14 -22 -236 -192 226 39
Nigeria 14,556 9,855 6,969 13,377 -2,738 -2,291 85 1,570 4,988 -4,244 4,129 6,485
Nomray 47,07B 54,768 38,911 54,440 -2,700 -898 -1,476 -1,591 3,992 -2,161 15,788 20,400
Pakistan 6,217 16,017 9,351 12,819 -966 -2,330 2,748 3,430 -1,352 -1,702 1,046 1,511
Panama 4,438 8,023 4,193 8,869 -255 -525 219 159 289 -1,212 344 823
PapuaNewGuinea 1,381 2,091 1,509 1,872 -103 -259 156 87 -76 47 427 205
Paraguay 1,609 3,893 2,094 4,277 260 61 55 58 -171 -265 675 987
Peru 4,128 7,488 4,686 10,494 -1,733 -1,484 316 . -1,384 .. 1,891 8,730
Philippines 11,430 36,973 13,967 39,631 -872 3,510 714 435 -2,695 1,287 2,036 13,230
Poland 19,037 43,387 15,095 52,007 -3,386 -1,178 2,511 2,897 3,067 6,961 4,674 24,535
Portugal 21,554 34,621 27,146 45,323 -96 -579 5,507 4,031 -181 -7,250 20,579 8,427
Romania 6,390 9,519 9,901 12,790 161 -392 106 753 -3,254 -2,918 1,374 2,690
Russian Federation 53,883 87,734 48915 74,078 -4,500 -12,000 . -415 468 1,241 . 8,457
Rwanda 145 112 359 482 -17 -8 145 236 -86 -143 44 174
SaudiArabia 47,445 43,551 43,939 44,417 7,979 2,768 -15,637 -15,053 -4,152 -13,150 13,437 16,997
Senegal 1,453 1,319 1,840 1,627 -129 -37 153 264 -363 -81 22 404
SierraLeone 210 75 215 161 -71 -15 7 . -69 . 5 39
Singapore 67,489 128,766 64,953 113,698 1,606 3,783 -421 -1,177 3,122 17,614 27,748 76,843
SlovakRepublic .. 13,012 15,346 . -158 366 .. -2,126 .. 3,371
Slovenia 7,900 11,143 6,930 11,405 38 146 46 112 978 -4 112 3,168
SouthAfrica 27,119 34,526 21,017 32,687 -4,096 -3,029 60 -746 2,065 -1,936 2,583 6,353
Spain 83,595 161,294 100,870 160,165 -3,533 -7,513 2,799 3,249 -18,009 -3,135 57,238 33,115
SriLanka 2,293 5,648 2 965 6,661 -167 -178 541 903 -298 288 447 1,636
Sweden 70,560 103,130 70490 89,268 -4,473 -5,785 -1,936 -3,438 -6,339 4,639 20,324 15,019
Switzerland 96,926 120,542 96402 106,277 9,746 16,018 2,329 -3,736 6,941 24,547 61,284 36,321
SyrianArabRepublic 5,030 4,930 2955 4,788 -401 -606 88 523 1,762 59
Tajikistan . 604 .. 731 .. -38 57 . -107
Tanzania 538 1,144 1,474 2,353 -185 -139 562 560 -559 788 193 775
Thailand 29,229 65,903 35,870 48,704 -853 -3,566 213 414 7,281 14,048 14,258 34,063
logo 663 693 847 823 -32 7 132 .. -84 358 122
Tunisia 5,203 8,482 6,039 9,131 -455 -857 828 831 463 -675 867 2,262
Turkey 21,042 54,541 25,652 55,412 -2,508 -2,985 4,493 5,727 -2,625 1,871 7,626 23,340
Turkmenistan 1,238 614 857 1,608 0 33 66 27 447 -934 1,513
Uganda 246 634 676 1,871 -77 -9 78 539 -429 -706 44 763
Ukraine .. 17,621 18,828 . -871 .. 782 . -1,296 469 1,046
UnitedKingdom 238,568 372,594 263,985 386,529 -818 23,589 -7,624 -10,754 -33,859 -1,100 43,146 29,834
UnitedStates 536,058 933,906 615,992 1,098,181 28,431 -12,209 -27,821 -44,075 -79,324 220,559 173,094 60,500
Uruguay 2,158 4,225 1,659 4,507 -321 -185 8 67 186 -406 1,446 2,587
Uzbekistan 3,148 .. 3,182 -61 43 -52
Venezuela,RB 18,806 19,021 9,451 19,870 -774 -1,559 -302 -154 8,279 -2,562 12,733 12,277
Vietnam 1,913 11,974 1,901 13,507 412 -689 49 951 -351 -1,271 429 2,002
Yemen,Rep. 1,490 1,708 2,170 2,771 -454 -422 1,872 1,256 739 -228 441 1,010
Zambia 1,360 1,057 1,897 1,140 -437 -485 380 -594 , 201 45
Zimbabwe 2,012 2,535 2,001 2,742 -263 -346 112 . -140 . 295 268
World 4,251,942t 6,766,816t 4,257,615t 6,696,346t
Lowincome 130,884 209,252 148,102 243,846
Middleincome 699,711 1,374.233 666,224 1,356,768
Lowermiddleincome 289,307 634,614 302,617 610,185
Uppermiddleincome 409,317 739,691 365,852 746,047
Lowandmiddleincome 829,625 1,583,740 814,842 1,603,017
EastAsia& Pacific 239,776 614,457 240,892 497,263
Europe& CentralAsia 188,731 340,843 187,584 363,280
LatinAmerica& Caribbean 169,084 335,772 146,919 393,251
MiddleEast& NorthAfrtca 134,093 131,866 134,828 154,797
SouthAsia 34,113 70,684 49,041 89,001
Sub-Saharan Africa 80,330 89,935 74,324 104,277
Highincome 3,418,264 5,183,326 3,436,033 5,096,364
Note:Fotdatacompatability
andcoverage, seethe Technical
NotesFiguresin italicsareforyearsotherthauthosespecified
a.IncludesLuxembourg.
b. Datapriorto 1992includeEritrea.
304 WORLD DEVELOPMENT REPORT 2000/200I

Table16. Privatesectorfinance

Interest
ratespread Domesticcredit
Private
investment Stockmnarket Listed (lendingminus provided
by
%ofgrossdomestic capitalization domestic depositrate) thebanking
sector
fixedinvestment Millionsof dollars companies Percentagepoints %ofGOP
Economy 1990 1993 1990 1999 1990 1999 1990 1999 1990 1999
Albafnia .. . .. .2.1 8.7 .. 47.2
Algera . ...... 2.5 74.7 45.8
Angola . . ..... 7.5 .. -10.4
Argentina 67.4 92.6 3,268 83,867 179 128 . 3.0 32.4 34.8
Armenia . .25 - 95 .. 11.5 58.7 1,08
Australia 742 80.1 107,611 427,683 1.069 1.217 4.5 29 103.5 93.7
Austria .. .11,476 33,025 97 97 .. 3.4 123.0 132.1
Azerbaijar . ... 4 .2 .- 57.2 10.6
Bangladesh 50.1 67.2 321 865 134 211 4.0 5.4 24.1 33.4
Belarus .. . -.. .. 27.2 . 37.4
Belgium . .. 65,449 164,942 182 172 6. 4.3 70.9 147.9
Benin 44.8 59.5 ... .9.0 . 22.4 9.6
Bolivia 39.3 59.7 .116 .. 16 16.9 23.1 30.7 66.2
Botswana .- 261 1,052 9 15 1.6 5.2 -46.4 -69.7
Breeil 767 80.0 16,354 227,962 561 478 .-. 89.8 50.5
Bulgaria 3.6 42.2 .. 706 . 860 8.9 9.6 0.1 0.0
BurkinaFasoc. ....... 9.0 .. 13.7 13.2
Burundi . . . 23.2 30.7
Carmbodia 86.0 73.9 ... . 10.2 .7 4
Camereoon. . . . 11.0 17.0 31.~2 17.7
Canada 86.3 68.5 241,929 800,914 1,144 3,767 1.3 1.5 858 100.9
CentralAfricarn
Repubiic . .. .11.9 17.0 12 9 11.7
Chad . 54.1 . .. .11.0 179 117.5 12.0
Chico 7 5 80.8 13.645 68,228 215 285 8.6 4.1 73.0 98.8
COno 33.8 47.5 2,028 330.703 14 950 0.7 3.6 98.0 130.4
HongKong.China -. . 83,397 689,090 284 695 3.3 4. 156.3 1411
Colombia 61.5 55.9 1.416 11,590 80 145 8.8 9.1 388 40.5
Conga, Dorn. Rep -. -.. .. . 25.3
Congo,Rep. . .. ... 11.0 17.0 29.1 19.7
CostaRica 76.9 73.2 475 2,303 82 22 11.4 11.4 29.9 38.3
COteolcOoire 57.9 70.2 549 1,514 23 38 9.0 -. 44.5 25.9
Croatia .. . . 2,584 2 59 499.3 10.6 . 48.4
CzechRepu,blic . .. 11,796 164 .. 4.2 6.5.1I
Denmark .. .39,063 105.293 258 233 9.2 4.7 13.~0 57.1
DominicanRepublic 73.1 82.9 -. 141 .. 6 15.2 9.0 31.5 37.0
Ecuador 78.3 82.9 69 415 85 28 -0.0 15.1 17.2 870
Egypt,ArabBop. 62.3 68.9 1,765 32,638 573 1.032 7.0 3.7 106. 86.3
ElSalvador 81.5 77.3 .. 2,141 .. 40 32 4.7 32.0 4.4.4
Eritreae
Estonia . 1,780 . 25 .. 4.5 68.0 34.0
Ethiopia . .. 3.6 4.2 50.4 51.3
Finland .. 22,721 349,409 73 147 4.1 3.5 84.3 57.5
France . .. 314,384 1.475,457 578 968 6.1 3.7 106.1 107.1
Georga . ...... 18.8 13.1
Germany . . 355,073 1,432,190 413 933 4.5 6.4 109.5 146.9
Ghava . .78 916 13 22 .13.2 30.8
Greece 49.8 .. 15,228 204,213 145 281 81 63 733 65.4
Guatemala 79.9 79.8 215 5 5.1 11.9 17.4 17.1
Gajivea . . .. 0.2 .. 5.4 8.8
Haiti .. .. ... . 15.5 32.9 24. 1
Hondairas . 40 458 28 71 9.3 18.2 40.9 28.5
Hurngary . .505 16.317 21 66 41 3.1 105.5 52.2
India ,56.7 70.1 36.567 184,605 2,435 5.863 50,6 44.8
Indonesia 67.5 77.2 6,081 64,087 125 277 3.3 1.9 45.5 61.1
Iran,lvlam,cRep. 53.6 55.5 34,282 21,830 97 295 .-. 62.1 49.4
Ireland .. . . 42.458 .. 84 5.0 3.2 57.3 100.0
lorae . 3,324 63,820 216 044 12.0 5.0 106.2 85.3
Italy - 148,766 728.273 220 241 7.3 4.0 90.1 93.5
Jamaica .. .911 2,530 44 46 6.0 16.0 34.8 49.1
Japan 79.4 70.9 2,617,679 4,546,937 2.071 2,470 34 2.0 266.8 142.4
Jordan -. 2,001 5,827 105 152 2,2 3.2 110.0 801
Kazakhrstan .. 2,260 17 .. .. 11.1
Kenya 41,8 56.0 453 1,409 54 57 5.1 128 52.9 50.1
Korea,Rep. 79.9 73.4 110,594 308.534 669 725 0.0 1.4 57.2 85.2
Kawait ... 18,614 .76 0.4 2.8 217.6 11886
KyrgyzRepublic . 25.3 . 20.0
LaoPDR .... 2,5 18.6 51 10.7
Latvia 301 70 9.2 - 18.8
Lebanon 1.921 .. 13 23.1 7,0 132,6 134.9
Lesotlro . 7.4 11.6 27.4 -0.2
Lithuania . .. 1.138 .. 54 .. 8.1 -. 13.1
Macedonia, FYR . .. 8 .. 2 . 9.1 . 9.0
Madagascar 48.5 46.9 ... 5.3 19.0 26.2 15.8
Macawi 51.8 27.7. ... 8.9 20.4 17.8 7.3
Molaysia 84.5 72.0 48,611 145,445 282 757 1.3 3.2 77.9 160.5
 SELECTED WORLD DEVELOPMENT INDICATORS 305

STATESANDMARKETS
Interest
ratespread Domesticcredit
Privateinvestment Stockmarket Listed (lending
minus provided
by
%ofgross domnestic capitalization domestic depositrate) thebanking
sector
fixedinvestment Millions
ofdollars companies Percentagepoints %ofGDP
Economy 1990 1997 1990 1999 1990 1999 1990 1999 1990 1999
Mali . .... 9.0 .. 13.7 16.1
Mauritania 46.2 49,4 .. 51 .. 54.7 03
Mexico 760 81.4 :32,725 154,044 199 186 . 16.3 366 29 1
Moldova 38 .58 80 62.8 30 8
Mongolia ... 32 .. 418 17.9 68.5 11.6
Morocco 683 67.9 966 13,695 71 55 0.5 62 430 838
Mozarmbique ... . 1168 15.6 6.2
Myanmar 2.1 51 3268 29.1
Namibia 61.7 62.2 21 691 3 14 108 7.7 20.4 53.7
Nepal . .418 .. 100 2.5 40 2859 41 4
Netherlands 87.3 86.9 119,825 605,209 260 344 0.4 0.7 107.4 131.5
NewZealand 74.7 86.8 0,835 26,352 171 114 4.4 3.9 81.6 121.3
Nicaragua 54.7 62.0 .. .12 5 1189 2066 1448
Niger -. 90 .. 16.2 9.2
Nigeria . .. 1,372 2,940 131 194 5.5 131 23.7 154
Norway 65.8 .. 26,136 63,696 112 195 46 28 674 608
Pakistan 51.7 58.3 2,850 6,965 487 765 50.9 47.0
Panamna 86.9 640 226 3,584 13 31 3.6 31 52.7 93.6
PapoaNewGuinea 79.7 79.9 ...... 9. 34 3568 310
Paraguay 8667 67.0 423 55 8.1 10.5 14.9 271
Peru 831 84.5 012 13,392 294 242 2,3350 145 162 249
Philippines 81 8 801 5,927 48,105 153 226 406 36 23.2 64.1
Poland 41.2 534 144 29,577 9 221 462.5 6.3 16.8 36.5
Portugal . .. 9,201 66,408 181 125 7.8 2.8 71.8 1079
Romania 9.8 35.4 .. 873 5,825 . 79.7 18.8
RussianFederation 244 72,205 13 207 . 26 0 1* 350
Rwanda ... 63 11 132
SaudiArabia 48,213 60,440 59 73 144 468
Senegal ... 90 338 228
SierraLeone 120 17.3 26.3 50.1
Singapore 34,3063 1~98,407 150 355 2.7 4.1 62.2 83.7
SlovakRepublic ... 723 845 . 6.7 . 67,5
Slovenia 2,180 24 28 142.0 5.1 36.8 40 1
SouthAfrica 65.6 72.7 137,540 262,47t 732 608 21 58 978 73.4
Spain 11V,404 431.068 427 718 54 21 1108 114.4
Sri Lanka .917 1,564 175 239 -6 4 -4.8 43.1 31 9
Sweden 78.9 797 97,929 373,278 258 277 6.8 3.9 145.5 122.1
Switzerland 160,044 693,127 182 239 -089 27 1790 18417
SyrianArabRepublic .. ... 566 28 9
Tajikistan
Tanzania 181 .4 22.1 39.2 139
Thailand 84 8 65 9 23,896 58,365 214 392 2.2 4.3 911 1260
Togo ... 90 21.3 222
Tunosia 50.5 508 533 2,706 13 44 .. . 62.5 69.2
Tarkey 68.2 77.6 19,065 112,716 110 285 .25.9 366
Turkmenistan ... . .. 26 6
Uganda . ... 74 128 178 7.6
Ukraine 1,121 . 125 34.3 83.2 24.3
UnitedKingdom 83.6 87.0 848,866 2.933.280 1,701 1,945 272 2.7 123.0 1291
UnitedStates 04.0 85.8 3,058,434 16,935,114 6,599 7,651 . 114.7 1701
Uruguay 71.5 720 168 36 17 76.6 39.0 60.1 43.2
Uzbekistan .. .119 4 .
Venezuela, RB 34.8 43.6 8361 7,471 76 07 77 10.8 374 1695
Vietnam 53 159 219
Yemen,top. 62 8 322
Zambia 291 8 9.5 20.1 67.8 63 5
Zimbabwe 2,395 2,514 57 70 2.9 16.9 41.7 32.8
World 78.1w 76.0w 9,3981,391
s 36,M030=0
s 25,424
s 49,640
s 125.1
w 1364w
Lowincome 48 1 56.5 54,588 268.082 3,446 8,332 43 7 42 7
Middleincome 72.2 74.6 430,570 2,159.249 4.914 16,560 021 80.2
Lowermiddleincome .. 58,226 751.775 1,833 11,451 63 4 02 5
Uppermniddleincomie 73.8 _77 9 372,344 1.407.474 3,081 5,100 60.2 603
Lowandmiddle income 64.5 9669 485,198 2,427,331 8,360 24,092 58.4 72.2
EastAsia & Pacific 633 56.9 197,109 955,379 1,443 3,754 71.0 112.5
Europe& CentralAsia .19,065 269,207 110 9.000 . 33 5
LatinAmerica& Caribbean 743 79.8 78,470 584,985 1,748 1,938 58.7 27.6
MiddleEast& NorthAfrica -. . 5,265 151,562 817 1,863 54 3 72 2
SouthAsia 55.9 6889 42,655 194.475 3,231 7,198 46.3 441
Sub-SaharanAfrica . .. 142,594 275,723 1,011 1,138 55.6 43 2
Highintcome 81 9 792 8,913,233 33,603,476 17,064 24,749 140.1 139.0
Nore.Fordatacomparability
andcoverage,
usee theTechnical
Nores.Figuresin italicsarofor yearsotherthantnose specified
 : :00o [0 °mU :0 :0 0 0 .0 m00 :0 00 :0 0 000-0 : 0 00 _ o, :00 N<>nN
._ _ *0 [E (00

x~~~~~~~~~c -1-i
~~~~~~~~~~c
4,~

mo : : : :--
. - : :n0a,- m
4,- - - - - -,-a - a, - o-,o a, 0
o--- a, -0 - C-- e D 0 - - 0-- aa ,0 00
-- LOLOa-
a, r0, 00 -e 0a 0 00 m
4, : a--00 : ma, : - -,- - --
- - - - :-
- M
0.0 ? w=-
'- 0 .- 0 - - - - -
- ' - - - - --00?-0
C0 - - -, a, _ .? Ea,-aa- -
w-a,>-; 00--

- - - 0 - - - - - - - - - - - a-
-: - o0 m-, - - -
- Omm -
-- -O -
o- ra,a,ra
: - - r0 o oom
or;rm
0
o~~~~~a 00
00 0 - r- t-2- a -, -
m - - -
000 - a- aL, ri0r - - .- rm-
a,rw
-
ma, .- a, r i a 0 a,0ai,r ci oa,, 000000 a,a,o 0a a,a

° t ,~= a . *..,w0?

0 ~ ~ ~ = ~~~~m-o
l a,0 000 0,c 000 o c7 r00 0,0,a 0 ,m ro ,-a r0 m0- 0 0O , a,;,aa 00 0 m a, 0 0r :00a 0z :n :0,o

-- -- 0 0 000 0 a, a, . 0
- - - _ w . a, - 0 -0 -
= 0, w : :0 - - - -- 0- ?- *- 0? -w- - - - -
0~~~o =

- F ,|Aa ,rA
X- a - - - -a A , -- - -G - -0- --- -a, 0 - 00 ' 00000 ,-- -_- a- - 00 -i 0 -

cr, 4 0)00-00000000000 00- 000000- C0-0-,0000000 :0 - 0 0-0-0 0 0 0 0,-0 0 00 0 - ,0-0.,,0

 9 o0~ CO o 0 E,C
-m ccncc
ZnZn as c-- CZO00 -- tEcBBa c sm WDC~ COOC 0)T X 2 ~~Zn
ZZO Zfl03< 0
aa,noam

t z o co eD~ _ ; ,n D ~ ZD O snn , O - . 3 oc

v:~~~~~~~~~~~~~~~~~~~~~~~~~~
X_ < R E: E- cO*°

J
_. *PV
z j : D O. _ D cn0 O O M aaswc 3C1 -r - Zg cn cn
Z*n . c n c D Ž0tDC). : J -c UC
ACO_C
:ZC

S2, V Er
' C a-
bDQ4>- a;[

~~~~~~~~~~~a
~~~~~ ~ ~ ~
a ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

0,0 ---- - ---- -- ~~~~~~~~ -.

| c c) c t ( p 0 2 I _ 3, 2 t~~~~~~~~~~~~

o>o
Ia
0
o)-.N) ~~~~inzcnc0 cor= tJN) CJ1gS
-- --
N)c> ciC4_14
---
C-.)
----
- - - C-
- - -
- c-,N, - - - -- -'-0D-100 - N
- - )OC -- -1 -' - CS) -
Zn~~~wziZ
Zn -.----0

| p~~~~~~
N, t- =:e
00~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
a<)|WwDmocC> ,tct'>8o
- - -CD N S' Ni ' - - -'' -0 - . - 0- - 0-5 C - :
_- g a - - - - - )-,- i C-A- -, - -C C O N) - - -CD

nc o> So c :t : :C c c0:0 )\ c n c: D1 DO ocno .iW C c C n::r -. X>°3

a a

| -< °-t > O . o o A c W . D n n: D D S > D n OO:E z ' s C C ° : D °o 8 oj-SXL

308 WORLD DEVELOPMENT REPORT 2000/200I

Electric
power Goods
transported Goods
Transmission
and byroad transported
byrail Arpsegr
Consumption
parcapita distribution
losses Paved
roads Millions
of Ton-km
per$million Arpsegr
carried
Kilowatt-hours %of output % of totalI ton-kmhauled -of GOP(PPPI_ Thousands
Economy 1990 1997 1990 1997 1990 1998 1990 1998 1990 1998 1998
Albania 810 851 18 53 .. 30.0 7,195 1,830 67,204 2,029 21
Algeria 449 566 14 15 67.0 68.9 14,000 . 23,449 .3,382
Angola 60 64 25 28 25.0 250 .... 553
Argentina 1,123 1,634 17 17 28.5 29.5 . 31,570 8.,447
Armenia 2,545 1,141 16 21 992 100.0 . 213 419,134 53,150 365
Austra IuI 7,572 8,307 7 6 35.0 38,7 ... 76,786 .. 30,186
Austria 5,587 6,051 6 6 100.0 100.0 13,300 15,700 86,340 79,889 5,872
Azerbaijan 2,584 1,831 13 23 .. 92.3 .. 706 ... 669
Bangladesh 43 76 34 15 7.2 9.5 . 6,543 5,148 1,153
Belarus 3,700 2,607 11 15 95.8 95.6 . 9,747 1,094,182 4869,369 226
Belgium 5.817 7,055 5 5 812 80.7 25,000 36,000 45,300 32,071 8,748
Benin 37 43 214 786 20.0 20.0 . 1
9....
Boinvia 271 391 15 11 4.3 5.5 ... 47,053 .. 2,118
Botswana . . . 32.0 23.5 .... .124
Brazil 1,425 1,743 14 17 9.7 9.3 .. 51.486 31,663 28,091
Bulgaria 4,046 3,203 11 14 91.6 92,0 13,823 307 303,350 154,935 828
BurkinaFauo -. 16,6 16.0 . ~ . .. 102
Burundi . .. . .. .12
Cambodia . . 75 7,5 . 1,200 .. 78,146
Cameroon 204 161 13 20 1105 12.5 .. 37,699 37,719 -278
Canada 14,972 15,829 7 4 35.0 35.3 54,700 72,240 411,103 440,137 24,653
Central AfricanRepublic .. . . . 2.7 144 60 ... 91
Chad 1 0.8 0.8 . .. .98
Chile 1,78 2,011 9 1398 13.8 . 23,140 7,959 5,150
China 471 714 7 8 . . .. 687,164 304.775 53,234
HongKong,Clra 4,178 4,959 11 14 100.0 100.0 . ... 12,254
Colomnia 751 865 21 22 11.9 12.0 6,227 . 2,113 1,945 8,290
Congo, Dem. Rep. 122 120 13 3 . .. . 32,460
Congo,Rep 254 197 0 1 9.7 9.7 ... 189,871 .241
CostaRica 1,111 1,353 8 9 15.3 21.0 2,243 3,070 ... 1,170
Cotedlvoire 158 181 18 16 8.7 9.7 - 15,674 19,827 162
Croatia 2,765 2,429 18 19 .. . . . 201,699 60,241 828
Couch Republi 4,649 4,817 7 6 100.0 100.0 .. 33,912 .. 143,684 1,601
Denmark 5,650 6,027 6 5 100.0 100.0 13,700 14,700 18,759 12,268 5,947
Dominican Repdublic 437 620 25 29 44.7 49.4 ... .. 34
Ecuador 467 611 23 23 13.4 16.8 2,638 3,753 -- 1,919
Egypt,ArabRap. 697 803 12 12 72,0 78.1 31,400 31,500 21,444 20,062 3,895
ElSalvaoor 358 537 16 13 144 19.8 ... .1,694
Eritroa .194 21.8 .
Estonia 4,332 3,466 7 16 51.8 22.1 .. 3,791 354,547 519,696 297
Ethiopia 21 21 1 1 15.0 15,0 - 2,120 790
Finland 11,922 13,689 5 4 61.0 64,0 26,300 25,400 97,605 92,017 6,771
France 5,321 6,060 7 6 . 100.0 180,500 237,200 51,687 43,309 42,232
Georgia 2,711 1,142 18 10 93.8 93.5 .. 98 272.479 146,315 110
Germany 5,729 5,626 4 4 960 99.1 182,800 3017,800 .- 49.280
Ghana 301 276 1 0 19.0 24.1 ... 6,191 .. 210
Greece 2,802 3,493 8 9 91.7 91.8 78,900 96,200 5,763 2,196 6,403
Guateomala 242 404 16 13 24.9 276 -. 506
Guinea . . 15.2 16.5 . .36
Haiti 61 42 1 4 1 4
Honduras 365 411 20 24 211 20.3-
Hungary 3,048 2,840 14 13 50.4 43.4 1.836 14 177,696 74,713 1.749
India 254 363 18 19 47.3 45.7 ... 199,742 137,082 16,521
Indonesia 156 329 15 12 46.0 46.3 . 9,570 9,125 12,614
Iran, IslamicRep. 829 1,163 19 22 50.0 ... 44,931 46.269 9,200
Ireland 3,385 4,559 9 8 94.0 94.1 5,100 5,500 14,784 4,875 10,401
Israel 3.902 5.069 5 9 100.0 100.0 -. 16,931 9.605 3,699
Italy 3,784 4,315 6 7 100.0 100.0 177,900 207,200 20,143 18,885 27,463
Jamaica 686 2,170 18 11 64.0 70.7 . ... 1,454
Japan 6.125 7.241 4 4 69.2 74.9 274.444 306,263 11,356 7,854 101.701
Jordan 959 1,196 8 10 100.0 100.0 ... 102,326 40,974 1,187
Kazakhsaun 5,905 2,595 9 15 55.1 86.5 . 4,637 3,964,8051,498,375 566
Kenya 115 127 10 17 129 13.9 -. 80.740 41,917 1,138
Korea, Rep. 2,202 4,847 5 4 71.5 74,5 31.841 74,504 37,095 20,362 27,109
Kuwait 6,675 12.886 9 29 86. .. 2,241
KyrgyzRepublic 1,903 1.372 8 34 90.0 91,1 . 350 620
LaoPDR.. . . . 24.0 13.8 120 ... 124
Latvia 3.281 1,758 19 29 134 38.6 .. 4.108 854,603 788,435 229
Lebanon 369 1,930 7 13 95.0 95,0 .... 716
Lesotno .. . 180 17.9 . .. 28
Litnuania 3,228 1,B81 5 11 81.83 91.0 .. 5,611 .. 346,800 259
Macedonia, FYR - 59.9 6386 1.708 1,210 62,004 47,137 489
Madagascar .. 15.4 11.6 . .. 601
Malawi -.... 22.0 190 . . 15,207 11,185 158
Malaysia 1,096 2.352 10 9 70.0 75.1 .. 15.555 7,339 13,654
 SELECTED WORLD DEVELOPMENT INDICATORS 309

STATESANDMARKETS
Electric
power Goods
transported Goods
Transmission
and byroad transported
byrail
Consumption
percapita distribution
losses Pavedroads Millions
of Ton-kmper$ million Airpassengers
Kilowatt-hours %ofoutput %of total ton-kmhauled ofGDP(PPP) Thousands
Economy 1990 1997 1990 1997 1990 199t 1990 1998 1990 1998 1998
Mali .. 10.9 121 . 53,612 34,053 91
Mauritania ... . .. 11.0 11.3 .... 250
Mexico 1,204 1,459 12 14 35.1 29,7 108,684 154,083 68,768 62,182 17,717
Moldova 2,279 1,217 8 27 87.1 87.3 .. 780 . 118
Mongolia .. .102 3.4 1,871 723 1,351,705 653,947 240
Morocco 340 423 9 4 49.1 52.3 2,638 2,086 67,356 49,613 3,012
Mozambique 35 47 16 31 168 187 . 10 ... 201
Myanmar 43 57 26 35 10.9 12.2 . .. .333
Numibia .... 108 83 . . 285,327 129,941 214
Nepal 28 39 29 28 37.5 41.5 ... .. 754
Netherlands 4,917 5,736 4 4 88.0 98.0 31,800 45,000 12,187 9,938 18,676
NewZealand 8,887 8,380 11 11 57.0 58.1 ... 49.742 51,977 8,655
Nicaragua 284 286 18 26 10.5 10.1 . .... 51
Nigerf . 290 7.9 . .91
Nigeria 77 84 38 32 380 30.9 .. 3,359 4,834 313
Norway 22,824 23,409 6 8 80.0 74.5 7,940 11,838 .- 14,292
Pakistan 267 333 21 24 540 57,0 352 90,268 41,763 26,278 5,414
Panama 883 1,152 24 22 32 0 28.1 . .. 860
PapuaNewGuinea .. .3.2 3.5 ... 1,110
Paraguay 470 759 0 2 8.5 9.5 .... .222

Peru 491 607 18 18 9.9 12.9 ... 8.138 4,757 2.775

Philippines 336 432 13 17 . 19.8 6..5 4 6,732
Poland 2,525 2,451 8 12 61.6 65.6 49.800 69,543 395,542 209,664 2,213
Portugal 2,379 3,206 11 10 . .. 12,200 13,500 13,550 13.975 7.023
Romania 2,337 1,704 9 12 510 67.6 13,800 15,785 345,140 146,252 908
Russian Federation 5,821 3,981 8 10 74.2 . ... 1,726,768 1,04Z132 15,224
Rwanda . .9.0 9.1 .
SaudiArabia 3,181 4,085 9 8 40.6 301 .. 4,324 3,843 11,816
Senegal 94 107 14 17 27.2 29.3 .. 61,227 35,183 121
SierraLeone ... .10.6 80 ... 0
Singapore 4,702 7,944 3 4 97 1 97.3 . ... 13,331
SlovakRepublic 4,432 4,243 8 9 987 99.0 4,180 4,750 .. 224,788 107
Slovenia 4,875 4,955 6 5 72.0 906 3,440 325 120,357 92,945 460
SouthAfrica 3,676 3,800 6 8 29.8 11.8 .... 362,402 283,262 6,480
Spain 3,239 3,899 9 9 740 99.0 10,900 16,500 22,227 17,569 31,594
Sri Lanka 153 227 17 17 32.0 95.0 19 30 5,077 2,035 1,213
Sweden 14,061 14.042 6 7 71.0 77.5 26,500 33,100 122,858 99,690 11,878
Switzerland 6,997 6.885 6 6 .. 10,400 13,250 ... 14,299
SyrianArabRepublic 683 776 265 . 720 23.1 .. 41,508 26,484 685
Tajikistan 3,346 2,177 9 12 716 827 . .. 592
Tanzania 51 54 20 14 370 42 . 77,249 71,671 220
Thailand 690 1,360 11 9 55 3 975 .. 15,903 8,835 15,015
Tago . ... 21 2 31.6 . ... 91
Tunisia 532 709 11 11 76.1 78.9 . . 52,684 42,976 1,859
Turkey 801 1,275 12 18 .. 28.0 .. 152,210 30,183 20,310 9,949
Turkmnenistan 2,293 934 8 I11 735 81.2 .... .521
Uganda .. . . . . .. 13,470 4,990 100
Ukraine 4,308 2,449 7 16 93.7 96.5 .. 18,266 1,476,624 987,824 1.066
UnitedKingdom 4,769 5,241 8 7 100.0 100.0 133,000 152,500 17,203 . 61.940
UnitedStanes 10,558 11,822 0 6 58.2 58.8 1,073,100 1,534,430 356,829 213,751 588,171
Uruguay 1,220 1,710 14 19 74.0 90.0 ... 11,124 6,290o 557
UzbekistaTn 2,383 1,645 9 9 79.0 87.3 .... 317,391 1,560
Venezuela,
RB 2,307 2,488 18 21 35.6 33 6 .... 354 3,737
Vietnam 04 203 24 18 23.5 25.1 1, 124
I... 11.367 2,304
Yemen,Rep 108 93 12 26 91 8.I .... 765
Zambia 503 563 11 11 16. . 81,810 57,858 49
Zimbabwe 933 919 6 13 140 47.4 ... 215,462 140,231 7899
World 1,928
w 2.053w 8w 8w 39.0m 43.1m 1,466,869
s
Lowincome 373 357 13 17 174 188 53,5866
Middle
income 1,243 1,340 9 11 50.5 49.7 292,223
Lowermiddleincome 1,061 1,042 9 11 50.5 43.7 153,612
Uppermiddleincome 1,926 2.434 tO 12 50.4 47.1 138,611
Lowandmniddle income 860 896 10 12 288 29.5 345,809
East Asia& Pacific 465 771 8 8 235 174 133,490
Europe & CentralAoia 3,853 2,603 8 12 742 86.5 41,165
LatinAmerica& Caribbean 1,131 1,402 14 16 21.9 26.0 89,378
MiddleEast&NorthAfrica 920 1,159 13 13 67.0 50.2 40,144
SouthAsia 228 324 19 18 37.5 570 25,390
Sub-Saharan Africa 446 446 9 10 16.8 150 16,242
Highincome 7.294 8.238 7 6 86.3 93.9 1,121,061-
Note Fordatacomparability
andcoverage,
seethe Technical
Noses.
FigurDsin italicsareforyearsotherthanthosespecified.
 (033GO0 TO0CW) GD ONOCiCt Cflrflmmrfl GOCOOC 0000 00000
cxcxcx-a OOLaLaNIO a LaLaLaLaLa
-a0CN LaLa»>
CD CD CD C »»> rn
C
CDCD< CDCDCDCDO CDDOa-,a CDcxaa CDam m CDCNCD D-LaC aam-.c, cOCOa -aa CD aacx w4
CDcxCDCDCcC3 Pm 9
mmBa a-'am- a lBSaaaaa-a 0 NOE C(NNJ
aOOCDa'm cx-a aa a c,a amoam- g 0 m 9 5 0 CDm maaa: Bcxcx CDaaB CD CD Dr CNCD
0
Ca oa in-in cx aC aa aDa CDCjC a aa - - -9- mcxin
a a m DC 933 m ma a 0O>m CC-' -. CD a2-a B 9:aa aaaB
ama a a a cx a-, aCta mLaGOao
0 (NO a
0 0
a
cx CD ma
-cx m aCD - cx 0 m
-rn cx mu 20 CD DC CD
-< -. CD -cx '- a 0 a
a C a a a cx
GO a
- a CD Q

'C '. m
- -(DO (4Cc, (01 -'NO--' -. - Go COO (4CC- cx - NO cx NONO-' -. C
cm
aC-
COLa
cm-coacxcxacmOŽ-
c0Ž -. jC-a
'-acm
cm
a
cxaaccm
maCDec
-ONO
a-a-
acm
acm Go.
Got-N
aoGOso-
-'-'
-
-'cm
acm--'
cxLa-N.cxcmccm---'-'
-a- aca r-oa-N.cmcx
a
LaaoaN0
-aa La
mONO
La
acxr-oao----i
CCC .NOND
aNOLaCom
a - "Oca
LaNOCCGO
CoNOaoao
GOC-o---aa
La 14
La
- m
GO
0

.14 0

-NO--'COcm LaD)--' -' aLa-GOOC< a.acm aNO--CDCGO 0-DO LaLaCCDT- cx -aGO-a aGONOcxGOLaaGOCOCO C LaLaNODOGONOLaCOCO -. 14
(0=* ' 91
NOaLaC-'
OCCDCOOGO
ac-'-a----
caocGoCo
CDCGOOaa
OGŽO.N.-O
cma-ONOa
creacca
maNOLa
C-OCCC-DCLa
acmoo-cx-OGO
COC-.OcacxaO
GOO1GO LaLa
a (01 -Omen
ccx
Ca-.
CDCCC-
.N-CCCD
cx-CDOGOLaNOCOcx
a-wa, a
am GocmLaacx
-'to.COC---aGO.aCCGOcxGO--J---GOGO
aNOOGO .cmcxca
amCncrnao
C-CDLaNOLaDONOaCCOŽ.-
COCGOGOCI -N.--.a---'-o
'0--'
-'a
Z
4. H
-4

CtCc,-mCa
Co
Na-N.
NOOCOD La
COCCONO-NCOC
0-Go
coaCoNacto
-m-------'r-o
N N.-.N.Na-N
coca-C,
N.C-OaaCO
--- La
cmGoOcGoC-a
-omcamO
-- to-
-toCoola
NN--CCcCCa
cm-NGOO,a
a-mCooo-
aCa-c(O,
OG
Ou---o
Cm "
CoCa
Lao-CO1CoC-o
Laa-DN.
'"-
Lac,1-JNoc
-0
GO
C0GOGo
-- 'Coo
DNoCO,O
NoCco
COLa
NOC-O-'C-
N.
-'NaN,
CN
cx Go Go
La-'-
C-OC--o.N.C.aC-o
cx
Co---'-
c,
aCCCCO-,O.ThNO
GO NOena
cm-Go
-N.Laac,a
CoN 0
a,aaCCCNa
=-
(0
mt
CD
EGO
01
o -.
H

cx -'GO NON. - cm-Oo-C -'w cmcm-cmcm 'GO aGoGO---'c,1NO a GO-' (02 NO aLa -NO toE..: a
cx
: N
0
0
aGOGOaa Sm5 La N-ŽAC-O---CoCDCOO Lacm-NOa cxcmLao.GO NON-Cac aaCOl-'NOcac,CNOG (ONOO-GONOGoCDCGO-- 0
14=

La-'cxCLaNOLCCcxa.C
-- '-0 coCa aNOCaNONO
cxm -. cxCocmLac,1
COcmLacx (CLa LaLa
0) COOGOLa
-'La C-C cx-'acx-aNONO aca-'NOC
DO-NO cxc
-a-La---a
a 0-NO-aDOa-'CCo acm Lac,ICNOLa
-'a - NOLaaLacmIcx cx -CD-'NO DO aLa cx La _ $ 2

I4.
La La GO aLa GOGODC-DO Go La La cm GOOD GO -' -0(0-' GO NOLa -C GO GONO a GO- C
a cx N. a >-. Cata 90 DO LaGODocx-'CNOm--.o NOacx- -a -oLa N. aDD -0-0--GO C-a cxO aa C 0 0 oLa-OCCa GO-' CD-N.m-N.
a: GO: La NO cx LaaCC1 N. GO aNODO NONOcxCOCDC OGOLaCC cx NONO -CD: -'cm ODGOCCO - LaNO (ONO a a: N.. -LaLaLaO : : -IC-a NOGOCCDGO 00

NO GO -cm GOal a -NOcm DO NO GO-' a GOOGO NO a NO -C -'NO--'GOLaGO CC

Laa a-ON. actcxc OLaCNO-' a-'-CLac C-'OGONO a c-GOa cx-'-' COLa a a-a-' cx-'CD)- La OCLaLaNO OmNOCC cacao- CCGO-'NOGO NOOLaCOD) NOaNOaa CD 'a;
CNONO-C NOLaGOGOGO c,1CCNOoNO a-a-a-'- aLaNOGO-' Lamcmaa a--t-N.aa mccmcxcm aLaNONNa 'CNOcma LaO-aC--' CC-"NO-'La NOc-C-a a NO-'DOGO - aaNO '*14
GOONOONO GOGOGOONO CGOCLaDO LaaNOCcx CDOGOLaaO OLaCIN. 0N0N.NO' -'-'-N.GO NO WLaCCDLaNO aO-'La LaNO cxNO-'NOcx aGOaaa NOLaO(C NOCCCGOCO -'aC-'-C

-'NO NO La NO NO NONO NO GO-'- NO cx NOCNO-GO-- (N '

La
GO La
--'GONO
CC
La
La
a
NO
GOa
-'
ca
LaGOGOLa----
La - -' LaaOŽ-a
a - DO a a cx
ODOGOLaCCD-'NOLaLaNO-GO
La La-NONOCOLa C .N.cx
La.C La-'p GO-.CD
cx-'
a
-- 'NONO
cxN.DO
cx La GO
-N.LaNCNOLa
La C
aLa
i a
cx
B
aQO14
=
0 0
GO NOLam - La La OGO : -a Ca: CO CO ONOaNO: - ODGO -': COLD. cx COcOC - CONOaC NO - - acm aLa - GOD(NO a. NOOCCO LaNO-DODOLa NO: - j 0

0.

CD 9114 n
La D NO NO NOOD -' NO - NONO - cx NO-' -' .
-- 'GO. O( Ocx-N-a aCamacm aLa--' La cxcx: La: GONO La: ac-N. -'aaa GO - a- aLa ocm Co. : : La Ca. La a-am NO-' LaLa '00 00(0

GO La
La cm -' Dl - I,
a NO NCOCD NO NO NOCO
-' NO -' cmDOa -'-0 CONONOLa La LaCCNO DO cx DC cx COLO La La
DO CCNO CO La NONODO CODOLaNCDD- La acCaLa cx C, LaNODO COCO LaLa (CD a GO cx LaN. COGO aNO Go (N
CON) acm GO: NO acm-. cx -aDICC La: cmaCu GO: a.Go NOLaa-'.CD a- GO-N.a ON-GO: : : C-ON.D( aCLa : Na : cx a acxN.: -a cmC,. -'NO GO-a: = -

a, 0
z
LaNONONO NO NO GO-aDO La CONOLa NO CCGO -DONDa NO LaNO a cm NO GO La NO NOOGO NO NO -'
9' 919'?'?' 91 9' 9' >91o 9' S&LaN NO 9191 GOCCGoCo1 9' DCCC -'tO LaO-' 91 91 -' La La LaLa LaLa La CD La
NOLa--'LaCOO ND CO LaLa 00cm-CD La GO-' 'GOLaLaCCDGO DC DOGO -ac-a -'Go COiN. NO La CO-La NO-GO cmNO Ca NOC -
DOGODOCCNO La La La LaGOc La GO-CLa -' GODO GOca -aaa-o NO OCOC ccxGO GONOCO CON. cm a -aaa cm LaC'CaaDO LaGo Ca &
NONO-aDOGO GO a - - LaLa DO: aLaLa - DO -'OCOGO-aGOGO- NODC. La: ODOONO - GOCCNO. - DC-NO - - - La -ONOONO Ca LaCICOGON. COLa: 00cm '
 SELECTED WORLD DEVELOPMENT INDICATORS 3I1

STATESANDMAR(
Scientistsand High-
Per1,000people Internet engineers technology Patent
hoStSb in R&D exports applications tiled'
Daily TelevisionTelephone Mobile Personal Per10,000 Permillion % ofmint. 1997
newspapersRadios setse mainlinesatelephonesa computersa people people exports
Economy 1996 1997 1888 1998 1998 1998 January 2000 1987-97d 1998 ResidentsNonresidents
Mali 1 54 12 3 0 0.7 001
Mauritania 0 151 91 6 0 5.5 022
Mexico 97 325 261 164 35 470 40.90 214 19 429 35,503
Moldova 60 740 297 150 2 64 2.97 330 7 295 25,030
Mongolia 27 151 63 37 1 54 019 910 1 186 261197
Morocco 26 241 160 54 4 25 0.33 .. 0 90 237
Mozambique 3 40 5 4 0 1.6 0.09 6
Myanmar 10 95 7 5 0 . 0.00
Namibia 19 144 37 89 12 186 1174
Nepal 11 30 6 8 0 .. 0.12
Netherlands 306 979 543 593 213 317.6 51703 2,219 30 5.227 85,402
NewZealand 219 990 508 479 203 282.1 703.33 1,663 -. 1,735 33,402
Nicaragua 30 385 190 31 4 78 204 4
Niger 0 69 27 2 0 02 003
Nigeria 24 223 65 4 0 5.7 0.01 15
Norway 588 915 579 660 474 373.4 09940 3,664 19 1,518 30,489
Pakistan 23 98 08 19 1 3.9 0.34 72 0 16 782
Panama 82 299 187 151 29 2771 4.33 .. 0 31 142
PapuaNewGuinea 15 97 24 11 1 0.70
Paraguay 43 192 101 55 41 9.6 3.02 2
Peru 0 273 144 87 30 18.1 360 233 3 40 756
Philippines 79 159 108 37 22 151 158 157 71 125 3,440
Poland 113 523 413 220 50 43.9 47.26 1,350 3 2,401 30,137
Portugal 75 304 542 413 309 01.3 9067 1,182 4 92 106,595
Romania 300 319 233 162 29 102 1102 1,307 2 1,709 27,346
RussianFederation 105 410 420 175 46 14.69 3,8 2 15,277 32.043
Rwanda 0 102 0 2 1 0.30 35
SaudiArabia 57 321 262 143 31 49.6 1208 1 57 1,001
Senegal 5 142 41 16 2 11.4 032 3
SierraLeune 4 253 13 4 0 915 .. 9,506
Singapore 360 022 348 562 346 4584 452.25 2,318 59 8.188 29,467
SlovakRepublic 185 580 402 286 87 65.1 47.96 1,606 3 234 27,973
Slovnei-a 199 436 356 375 04 250.9 103.71 2,251 4 205 27,162
SouthAfrica 32 317 125 115 56 47.4 39.17 1,031 9
Spain 100 333 5138 414 179 144.0 105.36 1,305 7 2,656 110,911
Sri Lanka 29 209 92 20 0 41 0.03 191 .. 81 26,322
Sweden 445 932 531 674 464 3614 870.83 3,826 23 7,8393 107,107
Switzerland 337 1,000 535 675 235 4218 429.01 3,006 16 5,014 107,038
SyrianArabRepublic 20 279 70 95 0 1.7 0.00 30
Talikistna 20 142 285 37 0 . 035 666 23 24,742
Tanzania 4 279 21 4 1 186 008 0
Thailand 63 232 236 84 32 21 6 6.48 103 31 238 5,205
Togo 4 210 18 7 2 6.8 026 98
Tunisia 31 223 198 01 4 14.7 0.10 125 2 46 128
Turkey II1 180 286 24 5 232 1.291233 27,985
Turkmenistan .. 276 201 92 1 0.62 52 24,584
Uganda 2 120 27 3 1 15 0.06 21 ... 49,769
Ukraine 54 884 490 191 2 13.8 539 2,171 . 4,692 28,036
UnitedKingdom 329 1,436 045 557 252 263.0 32139 2,448 26 26,591 121,619
UnitedStatus 215 2,146 847 661 256 4586 1,939.97 3,676 33 125,8308 110,884
Uruguay 293 607 241 258 60 912 76.09 2 32 320
Uzbekistan 3 465 275 65 1 006 1,763 817 26,490
Venezuela, RB 206 468 185 117 67 43.0 5.01 209 3 201 2,323
Vietnam 4 107 47 26 2 64 0.02 .30 27,410
Yemen, Rep. 15 64 29 13 1 12 602
Zambia 12 121 137 9 1 . 0.53 . 96
Zimbabwe 10 93 30 17 4 90 1.71 .2 3 21,9666
World .w 418w 247 w 146w 55w 70.6w 120.02w .w 22w 798.007 s 3,602,785 s
Lowincome .. 157 76 2F3 ~ 2 3.2 0.37 . 16,764 689,407
Middleincome .. 359 257 109 31 229 996 668 16 133,150 784,961
Lamermiddleincome 322 250 90 10 136 2.83 763 15 34,272 445,265
Uppermiddleincomre 89 493 285 176 76 531 35688 660 20 98,678 330,696
Lowandmiddleincome .. 263 172 69 17 156 540 .. 17 149,914 1,465,458
EustAsia& Pacific . 302 228 73 25 141 2 90 492 26 106,342 184,266
Europe & Central As:a 102 442 353 209 23 346 1687 2,534 8 31,061 605,716
LatinAmerica & Caribbean 71 420 255 123 45 33 9 22.33 .. 12 1,708 175,004
MiddleEast& NsrtnAfrica 33 274 135 81 6 99 0.55 .. 509 1,297
SouthAsia .. 112 81 19 1 2.9 0.22 137 4 10,236 26,322
Seb-Saharan Africa 12 198 52 14 5 75 2.73 38 392,621
Highincome 286 1,286 861 567 265 311.2 777.22 3,166 33 646,093 2,137,327
Ndote:Fordatacomnparability
andcoverage, onetheTechnical Notes.Figuresinitalicsareforyoursotherthanthosespeclfien.
a.Dataarefromthe International
Telecommunication Union's
IITU)WorldTelecnmmunication DevelopmentReport 1999.Pleasecitethe ITUforthird-party
usaofthesedata b. Dataarefrom
theIntemetSsftwareConsnortiummwww isc orgi c.Otherpatentapplications filedin 1997includethosefiledunderthe auspicesofthe AfricanIntellectual
Property Organizationl131by
residents,26,057by nonresidents),
AfricanReginnal Industrial
Property Organzation(7byresidents, 25,724bynonresidents),European PateutOffice144,664 byresidents,
53,339by
nonresidents),
andEurasianPatentOrganizatisn (250byresidents, 26,207bynonresidents) Thuoriginalinformationwasprovided oythe WorldIntellectualProperty Organizainon(WIPO). The
lnternational
Bureau ofWIPO assumes no liabilityorresponsibility with regardtothe transfsrmation
ofthesedtat. d. Dataareforthe mostretoolyearavailablee.Totalforresidents and
nonresidents.
312 WORLD DEVELOPMENT REPORT 2,000/2001

Table20. lbl rd

Merchandise
exports Exports
of
commercial
services MrhniemptsImports
Mrhniempts commercial of
services
Millionsof dollars Mfg.%oftotal Millionsof dollars Millionsof dollars Mfg.O/ oftotal Millionsof dollars
Economy 1990 1998 1990 1998 1990 1998 1990 1998 1990 1998 1990 1998
Albania 230 210 .. 68 32 83 380 840 67 29 119
A geria 11,330 10,300 3 3 479 . 9,715 9,320 68 82 1,155
Angoa 3,910 2,880 0 .. 5 226 1,578 2,120 .. . 1,2198 1,738
Argentina 12,353 26,441 29 35 2.264 4,494 4,076 31,404 78 89 2,876 8,714
Armenia .. 225 . 54 . 118 -. 895 . 39 .. 175
Australia 39,752 55,900 18 29 9,833 15,809 42.032 84,630 80 87 13,388 16,928
Austria 41.265 62,584 88 83 22,755 31,817 48,146 68,028 81 84 14,104 30,035
Azerbaijan .. 605 13 320 . 1,075 - 75 .692
Bangladesh 1,671 3,831 77 91 208 252 3,508 6,874 56 69 554 1.1780
Belarus . 7,015 . 78 935 . 8,510 57 .. 444
Belgium' 117,703 178.811 77 70 24,090 36,076 119,702 166,790 68 78 24,298 34,10
4
Benin 288 420 .. 109 102 265 671 . .. 113 170
Bolivia 926 1,103 5 30 133 238 687 1,983 85 86 201 423
Botswana 1,784 1,948 -. 183 241 1,946 2,387 . . 371 517
Brazil 31,414 51,120 52 55 3,708 7,083 22.524 60,730 56 76 6,733 15,743
Bulgaria 5,030 4.300 - 61 837 1,786 5,100 4,980 49 50 800 1,398
BurkinaFaso 152 315 .. 34 - 530 750 . . 196
Burundi 75 65 .. .7 3 231 150
8 59 33
Cambodia 88 320 .. 50 09 164 880 - 64 185
Cameroon 2,002 1,870 9 8 369 1,400 1,520 78 67 1,018
Canlada 127,628 214,710 59 66 18,350 30,281 123,244 206,066 81 85 27,470 35,240
CentraAfricanRepublic 170 160 . 43 17 154 280 - 61 166
Chad 188 270 .. 23 . 285 255 .223
Chile 8,372 14,830 11 17 1706 4,030 7,678 18,779 75 81 1,883 4.077
China* 82,001 183,809 72 87 5,748 24,040 53,345 140,237 80 81 4,113 28,775
HongKong,China 82,39000 174,863 b 95 95 18,128 34,523 84,725b 186,759 b 85 89 11,018 22,786
Colombia 6.766 10,852 25 32 1,548 1,990 5,590 14,635 77 79 1,683 3.462
Congo,DermRep 989 600 .. 127 B.87 322 .689
Congo,Rep. 981 1,250 .65 45 621 470 . . 748 553
CostaRica 1,448 5,511 27 56 563 1,315 1,990 6,230 66 86 540 1,166
COtee'lvoire 3,072 4,575 . 385 451 2,097 2,991 .. 56 1,375 1,314
Croatia 4,541 .. 76 - 3,964 6,384 .. 75 1,880
CzechRepuoblic I2,170 26,350 .. 88 -. 7,396 12,880 28,700 02 5.665
Denmark 38,870 48,173 60 65 12.731 14,830 33,333 46,086 73 77 10,106 15,460
OominicanRepublic 735 795 78 8 1,086 2,421 2,062 4,716 .435 1,300
Ecoador 2.714 4,203 2 10 508 757 1,861 5,576 84 76 601 1,178
Egypt,ArabRep. 2,585 3,130 42 44 4,813 7,832 9,216 16,166 56 59 3,327 5,886
El Salvador 982 1,263 38 47 301 277 1,263 3,112 63 68 296 538
Eritrea ..
Estonia 3,240 66 1,476 -4,785 71 .. 814
Ethiopia 561 .. 261 348 1081 1,450 , . 348 405
Finland 26,571 43.145 03 86 4,562 6,703 27,001 32,338 76 77 7,432 70679
Franca 216,588 305,362 77 80 66,274 84,627 234,436 289,421 74 80 50,455 65,420
Georgia 1909 I . 278 . 1,055 . .335
Germany 421,100 542,812 88 86 51,605 78,903 355,688 470,656 74 73 79,214 125,039
Ghana 897 1,700 8 79 162 1,205 1,850 70 .. 226 433
Greece 8,105 10,765 54 54 6.514 9,224 18,777 28,754 70 73 2,756 4,196
Guatemala 1,163 2,582 24 33 313 581 1.649 4,651 69 77 363 759
Guinea 671 800 .. 91 66 723 1,090 .243 274
Haiti 160 175 85 84 43 178 332 797 . . 71 370
Honduras 831 1,575 6 17 121 361 035 2,500 71 74 213 396
Hungary 10,000 22,995 63 82 2977 4,870 19,340 25.705 70 64 2,264 3,941
India 17,075 33,626 71 74 46090 11,067 23,642 42,742 51 55 5,943 14,182
Indonesia 25,675 48,847 35 45 2,488 4,340 21,837 27,337 77 69 5,898 11,744
Iran,IsamicRep. 16,870 13,100 .. 343 902 15,716 12,500 . . 3,703 2,302
Ireland 23,743 64,380 70 84 3,286 6,586 20,669 44.526 76 81 5,145 20,005
Israel 12,080 22,993 87 82 4,546 8,980 16,793 29,342 77 82 4,825 9,626
Italy 170,364 242,348 88 80 48,570 66,621 181,668 215,576 64 72 46,602 62,887
Jamaica 1,135 1,312 69 70 976 1,727 1,859 2,997 61 65 667 1,233
Japan 287,581 387,027 98 04 41,384 61.795 235,368 280,494 44 56 84,281 110,705
Jordan 1,664 1,800 51 42 1.430 1,819 2,600 3,836 51 58 1,118 1,588
Kazakhstan .. 5,340 .. 23 . 887 . 4.240 . 70 1,128
Kenya 1,031 2,008 29 24 774 638 2,125 3,197 66 64 598 603
Korea, Rep. 65.016 132,313 94 91 96155 23,843 69,844 93,282 63 61 10,050 23,523
Kuwai 7,042 9.554 6 14 1,054 1,496 30972 8,619 79 81 2,805 4,243
KyrgyzRepublic 515 . 38 . 58 .. 840 - 48 .. 177
LaoPDR 78 370 -. 11 116 291 553 . . 25 02
Latvia , 1,810 58 290 1,035 . 3,190 . 74 120 717
Lebanon 404 662 . .. 2,529 7,070 ,
Lesotho 59 194 . . 34_ 46 672 863 . .. 48 50
Lithuania 3,710 61 .. 1,096 . 5,795 .. 69 .. 816
Macedonia. FYR ,. .. 72 .. 130 . ., 47 .. 297
Madagascar 319 243 14 28 129 264 571 514 69 61 172 326
Maiavi 417 528 5 .. 37 .. 581 637 78 .. 268
Malaysia 29,416 73,305 54 79 3,769 11,296 29,258 58,326 82 85 5,394 13,230
*Taiwan,China 67,142 110,518 93 93 6.937 16,660 54,831 104,946 69 73 13,923 23,240
 SELECTED WORLD DEVELOPMENT INDICATORS 3I3

GLOBALLINKS

Merchandise
exports ~~Exports
of
commercial
services Mecadseiprs
Imports
commercial
of
services
Millions
ofdollars Mfg.%oftotal Millionsof dollars Millions of dollarsMfg.%oftotal Millionsof dollars
Economy 1990 1998 1990 1998 1990 1998 1990 1998 1990 1998 1990 1998
Mali 359 556 2 .. 71 62 619 750 53 352 324
Mauritania 469 448 . .. 14 24 387 319 . . 126 130
Mexico 40,711 117,500 43 85 7,222 11.937 43,548 130,811 75 85 10,063 12,621
Moldova .. 635 .. 25 . 117 . 1,025 . 57 .. 191
Mongolia 660 370 .. 10 48 75 924 472 . 65 155 142
Morocco 4,265 7,266 52 49 1,671 2,558 6,800 10,276 61 58 940 1,414
Mazarmbique 126 210 .. 17 103 286 878 910 62 206 396
Myanmar 325 1,065 10 .. 93 529 270 2.666 81 72 429
Namibia 1,085 1.460 . . 106 315 1,163 1,680 . . 341 449
Nepal 210 474 83 90 166 433 686 1,245 67 42 159 188
Netherlands 131,775 201,001 59 70 29,621 51,706 126,098 187,357 71 77 28,995 46,506
NewZealand 9,488 12,071 23 32 2,415 3,651 9,501 12,495 81 61 3,251 4,5038
Nicaragua 330 573 6 8 34 151 638 1,492 59 09 73 264
Niger 282 298 . . 22 . 308 377 .. 209
Nigeria 13,670 9,729 I 2 965 884 5,627 10,002 67 75 1,901 4,054
Norway 34,047 39,645 33 30 12,452 13,953 27,231 36,103 82 83 12,247 15,211
Pakistan 5,589 8,594 79 84 1,240 1.473 7,546 9,415 54 55 1,897 2,468
Panama 340 784 21 17 907 1,563 1,539 3,350 70 78 666 1,129
PapuaNewGuinea 1.144 1,772 10 . 198 318 1,193 1,232 73 .. 393 794
Paraguay 959 1,021 10 15 404 469 1.352 3,200 77 69 361 535
Peru 3.230 5,735 18 24 715 1,653 3.470 9,840 61 73 1,071 2,191
Philippines 8,068 29,414 38 90 2,697 7,465 13.041 31,496 53 80 1,721 10,087
Polapd 14,320 28,230 59 77 3,200 10,890 11,570 47,055 63 80 2,847 6,559
Portugal 16,417 24,177 80 87 5,054 8,512 25,263 36,912 71 77 3,772 6,708
Romania 4,960 8,300 73 81 610 1,192 7,600 11,835 39 73 787 1,838
Russian Federation .. 74,200 28 12,937 .. 59,100 .. 44 .. 16,127
Rwanda 110 60 . . 31 31 288 285 . . 96 115
SaudiArabia 44,417 39.775 7 88 3,031 4,421 24,069 30,013 81 78 12,694 8,678
Senegal 761 965 23 .. 356 329 1,219 1.407 51 .. 368 389
SierraLeone 138 7 .. 45 149 95 . . 67
Singapore 52,752h1 109,895b 72 86 12,719 18,243 60,899b1 101,496 b 73 84 8,575 17,884
SlonakRepublic .. 10,775 84 . 2,275 .. 13.005 . 77 .. 2.272
Slovenia .. 9,048 90 .. 2,045 10,110 .. 80 .. 1,520
South Africa 23,549 26,362 22 54 3,442 5,109 18,399 29,242 77 70 4.096 5.278
Spain 55,642 109,037 75 78 27,649 48,729 87,715 132,789 71 76 15,197 27,495
SriLanka 1.983 4,735 54 425 888 2,685 5,917 65 . 620 1,325
Sweden 57,540 84,705 83 82 13,453 17,675 54,264 68,177 79 81 16,959 21,620
Switnerland 63,784 78,876 94 93 161,232 25,795 69,681 80,094 84 85 11,086 15,273
SyrianArabRepublic 4.212 2,890 36 10 740 1,551 2,400 3,895 62 69 702 1,297
Tajikistan .. 600 .... . 770 -
Tanzania 415 676 ., 10 131 534 1,027 1,453 .. 66 288 885
Thailand 23.070 54,456 63 74 6.292 13,074 33,379 42,971 75 78 6,160 11,874
Togo 268 230 9 114 .. 581 623 67 .. 217
Tunisia 3,526 5,750 69 82 1.575 2,662 5,542 8,338 72 79 682 1,153
Turkey 12.959 26,974 68 77 7,882 23,161 22,302 45,921 61 76 2,794 9,441
Turkmenlistan .. 920 . . 269 . 980 81 .. 669q
Uganda 147 501 . .2 1 165 213 1,414 . . 195 693
Ukraine .. 12,635 . ... 3,922 .. 14,675 . . .. 2,545
UnitedKingdom 185,172 272,832 79 85 53,172 99,097 222,077 315.145 75 82 44.608 78,219
UnitedStates 393,592 682,497 74 82 132,184 239,957 516,987 942,645 73 81 97.940 165,827
Uruguay 1,693 2.769 39 39 460 1,382 1,343 3.808 69 79 363 866
Uzbekistan .. 2,390 . .. .2.750 .
Venezuela,RB 17,497 17,193 10 19 1,121 1,297 7,335 15,727 77 82 2,390 4,824
Vietnam 2,404 9,361 -. 1832 2,530 2,752 11,494 .. 126 3.153
Yemen,Rep. 692 1,496 1 . 2 166 1,571 2,167 31 .. 493 510
Zambia 1,309 740 .. . 94 1,220 700 .. . 370
Zimbabwe 1,726 2,111 31 32 253 . 1,847 2,772 73 77 460
World 3,328.357
t 5,253.926
t 72w 79w 754.507
a 1.279,291
s 3,408.529
t 5,3836450t71w 77w 783.538
s 1,266.754
s
Lowincome 110,592 165,177 48 52 16,955 31,453 118,035 177,252 64 63 30,934 57,435
Middleincome 613,527 1,124,846 54 71 104,602 236,330 572,952 1,147,658 70 74 131,347 243,647
Lowermiddleincome 281,205 499,085 59 66 57.873 122,664 297.148 508,531 70 73 63,395 116,077
Uppermiddleincome 332,321 625,765 51 74 46,729 113,666 275,891 639,025 70 78 67,952 127,570
Lowandmiddle income 724,243 1,290,207 54 69 121,557 267,783 692,669 1,327,01670 74 162,281 301,082
EastAsia& Pacific 220,817 537,290 68 81 31,420 88,106 230,492 413,466 73 75 34,539 104,151
Europe & CentralAsia 140,625 248,210 .. 57 31,468 79,758 143,083 300,176 .. 67 37,865 59,087
LatinAmerica&Caribbean 141,932 272,768 34 49 26,660 49,114 120,241 336,683 69 80 33.386 64,007
Middle East& North Africa 126,139 109,391 17 19 15,626 24,438 101,726 125,552 69 69 27,525 28,711
SouthAsia 27,790 51,606 71 77 6,8338 12,970 39,339 67,217 54 56 9,317 19,272
Sub-SaharanAfrica 66,763 70,732 20 36 9,547 13,397 56,115 81,867 .. 71 19,649 25,874
Highincome 2,604,220 3,963,915 77 82 632,950 1,011,508 2,717,343 4,058,694 71 78 621,257 965,672
Note:Fordatacomparabilityandconerage,seethe Technical
Notes.Figuresin italicsareforyearsotherthanthosespecified.
a.IncludesLuxembourg.b.Includesreexports.
 oECa00 °'
- °r°WWcae-
co rCOeCe0
- ao
a-a-.-
- -- 6 =Ln= oi
e004)00)
-- d -
o [---
-
CoOa vco-o ooaCocoa, aeC-C C
eO CO
a- C
C
eO
aC) Co aO
- o 000
(0
c
- -.
o a,
0-a-CO
aa-oO
- a-
,a, - o- -a - - --
COO oa,o eeeOa-a,ee .Lr

C
, 2
ta
|, n
c o a,rs7o
~
c o6
m
>
.
a--a-
:- o C mo -: - Om
o
:- -
:00
m0 -
_ _oaiaia~ lfr
-
m
- -
C o
-.-
0?
a-- -
a- a,om
0
*- r
C :
La ma-a-
- -C-
oa-
-a
c oo--o--
-o
-- -
-
-:-
c
- ma-
a-
m-
a--
:
-D
, aoO :(0 a0
r<a-ka,
C
a, o

- 0--- -a- ---- > o- _ - - : ° ° m° <: m D f °:D

S
OD
fm >'o
mn
b-m
m - mD D
-m-0 - - - -

CD 4

o a-

a, r.- - - a-- - - - a--- ,n -. _ -=,, - , a,- a - -, - - --- a- - - -, ,-,

r-I - 0) I j- I

0 -- 0a a,a a L aE-A, ana o - - - - a- - -

~~ X u w w w w w w z w~~~~~~~~~~~~~~~~~~, E cw
O _0 c a, s aCO
ww aSwa°
a-CO a,a--:
Oa-aa-O 0 000 O 00 a-CCOa-a--U CO oa--, L - : 0 .,COC a-acoo - - : C LaO,:a,-a-.a- aCOCO- C
C4a -- - an--aa,-OO EO-a P aO ,O Oaaaa -aaa- , , 0 aC- OaC.
a- -a< :--C -- aa OOaa CCaC
 SELECTED WORLD DEVELOPMENT INDICATORS 315

GLOBALLINKS
External debt
Netprivate
capital
flows Foreign
directinvestment Total Official
development
assistance
Millionsof dollars Millionsof dollars Millionsof dollars PresentvalueDolrpecaia %fGN
______________
_____________
_____________% of GNP ol r e a ia %o N
Economy 1990 1998 1990 1998 1990 1998 1998 1990 1998 1990 1998
Mali -8 17 -7 17 2,407 3,202 84 a 57 33 20.0 13.5
Mauritania 0 3 7 5 2,000 2,500 148 117 68 220 17.8
Mexico 8,253 23,188 2,634 10,230 104,431 150,959 39 2 0 0.1 0.0
Moldova 62 85 .39 1,035 58 2 8 0.3 20
Mongolia 28 7 27 19 350 739 49 6 79 . 20 6
Morocco 341 905 165 322 24,450 20,687 54 44 19 4.2 15
Mozambique 35 209 0 213 4,653 8,208 74a 71 61 42.4 282
Myanmar 153 153 101 70 4,005 5,6800 4 1
Namibia . .90 168 5.1 5.8
Nepal -8 -1 3 12 1,640 2,046 -31 23 18 11.5 83
Netherlands .. 12,352 33.346
NewZealand 1,735 2,657
Nicaragua 21 171 1 184 10,708 5,068 262a 87 117 33.7 201
Niger 9 -23 -1 1 1,726 1,659 55C 51 29 164 144
Nigeria 467 1,026 580 1,051 33,440 36,315 74 3 2 10 05
Norway ... 1,003 3,507
Pakistan 182 806 244 500 20,663 32,220 41 10 0 29 1.6
Panama 127 1,450 132 1,206 0,679 6,8689 78 41 8 20 0.3
PapuaNewGuinea 204 230 155 110 2,594 2,692 69 107 70 13.3 104
Paraguay 07 236 76 256 2,104 2.305 25 13 15 11 0.9
Peru 50 2,724 41 1,030 20,067 32,397 55 19 20 13 0.8
Philppines 630 2,587 530 1,713 30,500 47,817 66 20 0 29 0.9
Poland 71 9,653 80 6,365 49,366 47,700 28 35 23 24 06
Portuqa~ 2,011) 1,703
Rormania 4 1,826 13 2,031 1,140 9,513 23 10 16 0.6 09
RussianFederation 5,562 19,346 2,764 50,707 183,601 82 2 7 0.4
Rwanda 6 7 8 7 712 1.226 34 42 43 113 173
SaLzd1Arabia . .3 1 00 0.6
Senegal 42 24 57 40 3,732 3,861 50 112 56 1409 10.8
SierraLeonie 36 5 32 5 1,151 1,243 126 15 22 7.9 16.2
Singapore . 5,575 7,210 .. -1 1 a06 0.0
SlovakRepublc 278 1,400 6) 562 2,008 0,803 45 1 29 0.10 0.0
Slovenia 165 ... 20 0.2
SeuthAfrica 783 .. 550 24,712 108 12 0.4
Spain . 13,084 11,392-
Sri Lanka 54 325 43 193 5.863 0,526 41 43 20 0.3 3.2
Swooens 1,083 19,413
Switzerland .. 4,061 5,488 .
SyrianArabRepteblic 18 76 71 00 17.060 22,435 136 56 10 5.9 1.0
Tajikistan . -3 . 18 10 1,070 49 2 17 04 509
Tanzania 4 157 II 172 6,438 7,8603 71 b 46 31 293 125
Thailand 4,309 7,825 2.444 6,941 20,165 66,172 70 14 11 09 06
Togs 0 0 1 0 1,275 1,448 68 74 29 103 856
Tunisia -122 604 76 656 7,601 11,070 56 48 16 3.3 08
Turkey 1,782 1,641 684 040 49.424 102,074 40 22 6 0a 60
Turkmenistan , 473 . 130 ., 2,266 78 I 4 01 66
Uganda 16 198 0 200 2,583 3,835 35 a 41 23 15.8 70
Ukraine 369 2,087 743 551 12,718 30 6 8 03 009
UnitedKingdem ... 32,518 67,481
UnitedStates ,. 48,954 193.373
Uruguay -102 406 It 164 4,415 7,606 36 17 7 07 01
Uzbekistan 40 592 401 200 60 3,162 17 0 6 0.0 008
Venezuela, RB -126 6,066 457 4,435 33,170 37,003 46 4 2 02 00
Vietnam 16 032 ifi 1,200 23,270 22,359 76 3 15 2.4 4.3
Yomen,Rep. 30 -210 -13' -210 0,345 4,138 56 34 19 08 5.5
Zombie 194 40 203 72 0,910 0,865 181 62 30 10.0 114
Zimhabwe 85 -217 -12 76 3,247 4,716 00 35 24 40 47
Werld .. s .. s 193,382 s 619,258 s .. s ..s 13w 9w ..w 0,6w
Lowincome 6,048 12,231 2,20' 10,674 418,922 570,545 9 7 26 13
Middleincome 35,059 255,469 21,029 160,267 1,041,421 C 1,956,501
C18 12 07 0.4
Lowermiddlencome .. ,24 16 12 1.0
Uppermiddleincome .,18 6 04
Lowandmiddleincome 42,606 287,706 24,130 170,042 1,4606,343
C 2,536,046C 11 8 1.2
0.7
EastAsia & Pacific 18,720 67,249 11.135 64,162 274,071 667,522 4 4 0.5
Europe& CentralAsia 7,048 53,342 1,051 24,350 220,428 400,539 13 0.6 14 06
LatinAmerica& Caribbean 12,412 126,854 8,180 69,323 475,867 786,010 11 0.4 9 0.2
MiddleEast& NorthAfrica 369 9,223 2,458 5,054 183,205 208,059 42 2.1 18 10
SouthAsia 2,174 7,581 464 3,659 129,809 163,775 5 15 4 0.0
Sub-SeahaanAfrica 1,283 3,452 830 4,304 176,873 230,132 38 9.9 21 4.1
Highincome 169,252 448,316
Note Fordatacomparability
asscoverage, seethe TecnnicalNoten.Figuresin italicsarefor yearsotherthanthose specifed
a Dataarn fromdebt sustainability
analyses
undertakenas partofthe HeavilyIndebted PesrCountriesIHIPCIn tiat ye Present
valueestmnate
coverspubiicandpubliclyguaranteed
debtonly b. DatarefersomainlandTanzania
only.c. Includesdatafor Gibraltarnot includedinothertables
316 WORLD DEVELOPMENT REPORT 2000/2001

Tablela. Key indicators for other oeconomies

Gross
national
product
IGNP)
GNPpercapita GNPmeasured Adult Carbon
Surface Population ~~at ppp Life illiteracy dioxide
area density Avg.annual Avg.annual expectancy rate emissions
Population
ThousandsPeopleper Millions growth
0
growth MillionsPercapita atbirth %ofpeopleThousands
Thousandsof sq.km sq.km of dollars rateI% Dollars rate(%) of dollars (dollars) Years 15andabove oftons
Economy 1999 1999 1999 1999' 1998-99 l 999b 1998-99 1999 1999 1998 199t 1996
Afghanistan 25,869 6521 40 . . . 46 65 1,176
AmericanSamoa 63 02 317 . 282
Andorra 65 05 144
AntiguaandBarbuda 67 04 153 . 75 322
Aruba 94 02 495 e . 1,517
Bahamas,The 298 139 30 . . 74 5 1,707
Bahrain 665 07 963 . .. . . 73 14 10,578
Barbados 267 04 620 d 76 8.35
Belize 247 23.0 11 673 4.7 2,730 1.1 1,109 4,492 75 7 355
Bermuda 64 01 1,280 e . 462
Bhutan 782 47.0 17 399 690 510 2.9 1,169 1,496f 61 . 260
Bosniaard Herzegovina 3,001 51.1 76 6.1 9 3.0 73 .. 3,111
Brunei 322 58 61 e 76 9 5,071
CapeVerde 429 4.0 106 569 803 1,330 5.1 1,499' 3,497 1 68 27 121
Cayman Islands 36 0.3 133 e .. 282
Channels ands 149 0.3 480 e 79
Comoros 544 22 244 189 -1.4 350 308 740' 1,360f 60 42 55
Cuba 11,150 110.9 102 .9 76 4 31,170
Cyprus 760 9.3 82 9,086 4.2 11,960 33 13,977 18,395f 78 3 5,379
Djibouti 648 23.2 28 511 790 50 38 366
Dominca 73 0.8 97 231 -0.1 3,170 01 352 4,825 76 . 81
Equatorial Guinea 443 28.1 16 516 85 1,170 57 50 19 143
Faeroesarands 44 1.4 31 , e 630
Fji 801 18.3 44 1,771 7.8 2,210 64 3,634 4,536 73 8 762
French Polynesia 231 4.0 63 , e 72 . 561
Gabon 1,208 267.7 5 4,043 -3.9 3,350 -6 2 6,435 5,325 53 . 3,690
Gambia, The 1,251 11.3 125 430 5.2 340 22 18671 1,4921 53 65 216
Greenland 56 341.7 0 , e 68 . 509
Grenada 97 0.3 285 335 7.5 3,450 66 567 5,047 72 . 161
Guam 151 0.6 275 , e 77 4,078
Guinea-Bissau 1,185 36.1 42 195 4.9 160 28 705f 595f 44 63 231
Guyana 856 215.0 4 653 30 760 2.2 2,7741 3,2421 64 2 953
Iceland 277 '03.0 3 8,109 6.0 29,280 4.9 7,280 26,283 79 2,195
Iraq 22,797 438.3 52 9 . 59 46 91,387
Isleof Man 76 0.6 129 d
Kribati 88 0.7 121 81 165 910 -187 2821 3,186f 61 22
Korea,Oem. REp 23,414 120.5 194 c , ,, 63 254,326
Liberia 3,044 1114 32 , .. 47 49 326
Libya 5,419 1,759-5 3 da . 70 22 40,579
Liechtenste a 32 02 200
Luxemnourg 432 26 167 19,285 51 44,640 38 16,523 38,247 77 8,281
Mlacao, Chnia 469 00 23,450 , e 78 1,407
Mvtaldives 278 03 925 322 72 1,10 33 984f 3.545f 67 4 297
Malta 379 03 1,184 3,492 35 9,210 29 5,7101 15,066f 77 9 1,751
Marshalls ands 64 02 353 100 05 1,560 2.6
Mauscritics 1,170 20 576 4,203 49 3,590 40 10,123 8,652 71 16 1,744
Mavotte 120 04 341 d
Micronesia, Fed Sts. 116 0.7 165 210 03 1,810 10 . 67
Monaco 32 00 16,410 e
NetherandsAntilles 215 0.8 268 e . 76 4 6,430
NewCaedonia 213 186 12 , e 73 1,751
NorthernMarianas ands 68 0,5 143
Oman 2,348 2125 11 , 73 31 15,143
Palau f9 05 40 c 71 245
PuertoRico 3,890 9'0 439 . . 76 7 15,806
Oatar 757 110 69 e 74 20 29,121
Samoe 170 28 60 181 1.0 1,060 0.5 6660 3,915f 69 20 132
SaGTomeardPrincipe 145 10 151 40 52 270 2.9 1931 1,335 64 . 77
Sevchelles 80 0.5 177 520 -3.0 6,540 4.1 826' 10,381f 72 . 169
Solomon Islands 429 28.9 15 320 -0.4 750 -3.4 836 1,949f 71 . 161
Somaa 9,388 6377 15 ' 48 .. 15
St K tts andNevis 41 0.4 114 262 2.1 6,420 2.0 401 9,801 70 .. 103
St. Luca 154 06 253 501 3' 3,770 1.5 775 5,022 72 191
St Vincent andthe Grenadnes 114 0.4 293 307 4.0 2,700 32 532 4,667 73 125
Sudan 28,993 2,505.8 12 9,435 6.0 330 3.6 37,641f 1,2981 55 44 3,473
Suriname 413 163.3 3 .9 .. 7
70. 2,099
Swazisand 1,018 17.4 59 1,379 2.0 1,360 -0,9 4,274 4,200 56 22 341
Tonga 100 0.8 138 172 2.2 1,720 12 4271 4,281 71 . 117
Tr nidadansTobagc 1,293 5.1 252 5,661 4.5 4,390 3.9 9,368 7,262 73 7 22,237
UnitedArabnEmirates 2,815 83.6 34 e 75 25 61,843
Vanuatu 189 12.2 16 221 2,0 1,170 -5.2 5251 2,771f 65 . 62
V rginIslands(U.S.) 120 0.3 352 eI 77 12,912
WestBankandGaza 2,839 6.2 446 4,559 5.2 1,610 13 71
Yugoslavia, FR(Serb/Mont I 10,616 1022 104 .. 9 72 36,197
Note Figuresin ta icsarefor yearsotherthanthosespecified
a Purchas ngpowerparityuseethe TechnicalNotesbsCalculated usingtheWorldBankAtlasmethodc Estimated to be owincomeI$755orless),d. Estimated to beuppermiddle
income 1$2,9969,265). e. Estimated
to be high ncome($9,266or more).f Theestimateis basedonregresson,othersareextrapolated fromthe latestInternationalComparison
Programme benchmarkestimatin8 Estimated to be lowermiddleincome 1$756-2,995).
 Technical Notes
T hesetechnicalnotesdiscussthe sourcesand methods
usedto compilethe 149indicatorsincludedin thisyear'sSe-
lectedWorldDevelopmentIndicators.The notesfollowthe
orderin whichthe indicatorsappearin the tables.
Sources
The data publishedin the SelectedWorldDevelopmentIn-
dicatorsare takenfrom WorldDevelopment Indicators2000.
Wherepossible,however,revisionsreportedsincetheclosing
date of that editionhave been incorporated.In addition,
newlyreleasedestimates of populationandGNP percapitafor
1999areincludedin table 1.
The WorldBankdrawson a varietyof sourcesfor thesta-
tisticspublishedin the WorldDevelopment Indicators.Dataon
externaldebt arereporteddirectlyto the World Bankby de-
velopingmembercountriesthroughtheDebtorReportingSys-
tem. Otherdata are drawnmainlyfrom the UnitedNations
and its specialized
agencies,fromthe InternationalMonetary
Fund (IMF),and from countryreportsto the World Bank.
Bankstaffestimatesarealsoused to improvecurrentnessor
consistency.Formost countries,nationalaccountsestimates
areobtainedfrommembergovernments throughWorldBank
economicmissions. In someinstancestheseareadjustedbystaff
to ensureconformitywithinternationaldefinitionsand con-
cepts.Most socialdatafromnationalsourcesare drawnfrom
regularadministrative files,specialsurveys,or periodiccensus
inquiries.The Data Sourcessectionfollowingthe Technical
Notesliststhe principalinternationalsourcesused(aswellas
the sourcescitedin the TechnicalNotes).
Data consistency
and reliability
Considerable efforthasbeenmadeto standardize the data,but
fullcomparability cannotbe assured,and caremust be taken
in interpretingthe indicators.Manyfactorsaffectdata avail-
ability,comparability,and reliability:statisticalsystemsin
manydeveloping economies arestillweak;statisticalmethods,
coverage,practices,and definitionsdifferwidely;and cross-
countryandintertemporalcomparisons involvecomplextech-
nicaland conceptualproblemsthat cannotbe unequivocally
resolved.For thesereasons,althoughthe data aredrawnfrom
the sourcesthoughtto be mostauthoritative,theyshouldbe
construedonlyas indicatingtrendsand characterizing major
differencesamongeconomiesrather than offeringprecise
quantitativemeasuresof thosedifferences. Also,nationalsta-
tisticalagenciestendto revisetheirhistoricaldata,particularly
for recentyears.Thus data of differentvintagesmaybe pub-
lishedin differenteditionsofWorldBankpublications. Read-
ersareadvisednotto compilesuchdatafromdifferenteditions.
Consistenttime-seriesdataare available on the WorldDevel-
opmentIndicators 2000 CD-ROM.
Ratios and growth rates
Foreaseof reference,the tablesusuallyshowratiosand rates
ofgrowthratherthan the simpleunderlyingvalues.Valuesin
theiroriginalformareavailable on the WorldDevelopment In-
dicators2000 CD-ROM. Unlessotherwisenoted, growth
ratesarecomputedusingthe least-squares regressionmethod
(seethe sectionon statisticalmethodsbelow).Becausethis
methodtakesinto accountallavailableobservations duringa
period,the resultinggrowthratesreflectgeneraltrendsthat
are not undulyinfluencedby exceptional values.To exclude
theeffectsof inflation,constantpriceeconomicindicatorsare
usedin calculatinggrowthrates.Datain italicsare for a year
or periodotherthan that specifiedin the columnheading-
up to twoyearsbeforeor afterfor economicindicators,and
up to threeyearsfor socialindicators,whichtend to be col-
lectedlessregularlyand to changelessdramaticallyovershort
periods.
Constantpriceseries
An economy'sgrowthis measuredby the increasein value
addedproducedbythe individualsand enterprisesoperating
in that economy.Thus measuringrealgrowthrequiresesti-
matesof GDP and its componentsvaluedin constantprices.
The World Bankcollectsconstantpricenationalaccountsse-
riesin nationalcurrencies
and recordedin thecountry'sorig-
inalbaseyear.To obtaincomparableseriesof constantprice
data,it rescalesGDP and valueaddedby industrialoriginto
317
318 WORLD DEVELOPMENT REPORT 2000/2001
a common referenceyear, currently 1995. This processgivesrise
to a discrepancy between the rescaled GDP and the sum of the
rescaledcomponents. Becauseallocating the discrepancywould
give rise to distortions in the growth rate, it is left unallocated.
Summary measures
The surmimary measuresfor regionsand income groups,presented
at the end of most tables,are calculatedby simple addition wvhen
they areexpressedin levels.Aggregategrowth ratesand ratios are
usuallycomputed as weighted averages.The summary measures
for socialindicators are wveightedby population or subgroups of
population, except for infant mortality, which isweighted bv the
numcer of births. Seethe notes on specificindicatorsfor more in-
formation.
For summary measuresthat covermany years,calculationsare
based on a uniform group of economiesso that the composition
of the aggregatedoes nor change over time. Group measuresare
compiled only if the data availablefor a givenyear account for at
least two-thirdsof the full group, as defined for the 1995 bench-
mark vear.As long as this criterion is met, economies for which
data are missingare assumed to behavelike those for which there
are estimates. Readers should keep in mind that the summary
measuresare estimatesof representativeaggregatesfor each topic
and that nothing meaningfulcan be deduced about behaviorat the
country levelbyworking back from group indicators.In addition,
the weighting processmay rsult in discrepanciesbetweensubgroup
and overalltotals.
Table 1. Size of the economy
Population is based on the de facto definition, which counts all
residents, regardless of legal starus or citizenship, except for
refugeesnot permanently settled in their country of asylum,who
are generally considered part of the population of their country
of origin.The indicatorsshown aremidyearestimates(seethe tech-
nical note for table 3).
Surface area is a country's total area, including areas under
inland bodies of water and coastalwateerwvays.
Populationdensityis midyearpopulationdividedby land area.
Land area is a country's total area excluding areas under inland
bodiesof water and coastalwaterways.Density is calculatedusing
the most recentlyavailabledata on land area.
Gross national product (GNP), the broadest measureof na-
rionalincome,measurestotal value added from domestic and for-
eign sourcesclaimed by residents.GNP comprisesgrossdomestic
product (GDP) plus net receipts of primary income from non-
residentsources.Data are convertedfrom nationalcurrencyto cur-
rent U.S. dollars using the World Bank Atlas method. This
involvesusing a three-year averageof exchange rates to smooth
the effectsof transitory exchangerate fluctuations. (See the sec-
tion on statiscicalmethodsbelowfor further discussionof the Atlas
method.) Average annual growth rate of GNP is calculated
from constant price GNP in national currency units.
GNP per capita is GNP divided by midyear population. It is
convertedinto current U.S.dollarsby the Atlasmethod. The World
Bank uses GNP per capita in U.S. dollars to classifyeconomies
for analyticalpurposesand to determineborrowingeligibility.Av-
erageannual growth rate of GNP per capita is calculated from
constant price GNP per capita in national currency units.
Becausenominal exchange rates do not alwaysreflect inter-
nationaldifferencesin relativeprices,table I also showsGNP con-
verted into international dollars using purchasing power parity
(PPP)exchangerates.GNP measuredat PPP is GNP converted
to internationaldollars by the PPP exchangerate.At the PPP rate,
one international dollar has the same purchasing power over do-
mestic GNP that the U.S. dollar has over U.S. GNP. PPP rates
allowa standard comparisonof real price levelsbetweencountries,
just as conventional price indexes allow comparison of real val-
ues over time. The PPP conversion factors used here are derived
from the most recent round of price surveysconducted by the In-
ternational Comparison Programme, a joint project of the World
Bank and the regionaleconomic commissionsof the United Na-
uions.
T'his round of surveys,completed in 1996 and covering 118
countries, is based on a 1993 referenceyear. Estimatesfor coun-
tries not included in the survey are derivedfrom statisticalmod-
els using availabledata.
Rankings are based on 206 economies and include the 74
economies with sparse data or populations of less than 1.5 mil-
lion from table I a. Rangeestimatesfor GNP and GNP per capita
have been used to rank many of these 74 economies-such as
Liechtenstein,which ranks second in GNP per capita.
Table 2. Quality of life
Growth of private consumption per capita is the average an-
nual rate of changein privateconsumption dividedby the midyear
population. (Seethe definitionof privateconsumption in the tech-
nical note to table 13.) The distribution-corrected growth rate is
I minus the Gini index (see the technical note to table 5) multi-
plied by rhe annualrate of growth of privateconsumption.Growth
in private consumption per capita is generallyassociatedwith a
reductionin poverty,but wherethe distribution of incomeor con-
sumption is highly unequal, the poor may not share in that
growth. The relationship between the rate of poverty reduction
and the distribution of income or consumption, as measured by
an indexsuch as the Gini index, is complicated.But Ravallionand
Chen (1997)have found that the rate of poverty reduction is, on
average,proportional to the distribution-correctedrate of growth
of private consumption.
Prevalenceof child malnutritionis the percentageof children
under age 5 wvhose weight for age is less than minus 2 standard de-
viationsfrom the median of the referencepopulation,whichis based
on children from the United States,assumedto be well nourished.
Weight for age is a compositeindicator of both weight for height
(wasting)and height for age (stunting). Estimates of child mal-
nutrition are from the WVorldHealth Organization.
Under-5 mortality rate is the probability that a child born
in the indicatedyear *willdie before reachingage 5, if the child is
subject to current age-specificmortality rates. The probability is
expressedas a rate per 1,000.
Life expectancy at birth is the number of years a newborn
infant would liveif patterns of mortalityprevailingat its birth were
to stay the same throughout its life.
 SELECTED
Age-specificmortality data such as infant and child mortality
rates,alongwith lifeexpectancyat birth, are probablythe best gen-
eral indicatorsof a community'scurrent health statusand areoften
cited as overallmeasuresof a population's welfareor qualityof life.
The main sourcesof mortality data are vital registration systems
and direct or indirect estimates based on sample surveysor cen-
suses. Becausecivil registerswith relativelycomplete vital regis-
tration systemsare fairly uncommon, estimatesmust be obtained
from sample surveysor derived by applying indirect estimation
techniquesto registration,census,or surveydata.Indirect estimates
rely on estimated actuarial ("life")tables, which may be inappro-
priate for the population concerned. Lifeexpectancyat birth and
age-specificmortalityratesaregenerallyestimatesbasedon the most
recently available census or survey; see the primary data docu-
mentation table in W"orldDevelopmentIndicators2000.
Adult illiteracy rate is the percentageof people age 15 and
above who cannot, with understanding, read and write a short,
simplestatement about their everydaylife. The definition here is
based on the concept of functional literacy: a person's ability to
use reading and writing skills effectivelyin the context of his or
her society.Measuringliteracyusingsuch a definitionrequirescen-
sus or sample surveymeasurementsunder controlled conditions.
In practice, many countries estimate the number of illiterate
adults from self-reporteddata or from estimates of school com-
pletionrates. Becauseof these differencesin method, comparisons
acrosscountries-and even over time within countries-should
be made with caution.
Urbanpopulationis the shareof the population livingin areas
defined as urban by each country.
Access to sanitation in urban areasis the percentageof the
urban populationservedby connectionsto public sewersor house-
hold systemssuch as pit privies,pour flush latrines, septic tanks,
communal toilets, or other such facilities.
Table 3. Population and labor force
Total population includesall residentsregardlessof legal status
or citizenship,exceptfor refugeesnot permanentlysettled in their
country of asylum,who are generallyconsideredpart of the pop-
ulationof their countryof origin.The indicatorsshownare midyear
estimates.Population estimatesare usuallybased on national cen-
suses. Intercensal estimates are interpolations or extrapolations
based on demographic models. Errors and undercounting occur
even in high-income economies;in developingcountries such er-
rors may be substantial becauseof limits on transportation, com-
munication, and the resourcesrequired to conduct a full census.
Moreover,the international comparabilityof population indica-
tors is limited by differencesin the concepts,definitions,data col-
lection procedures, and estimation methods used by national
statisticalagenciesand other organizationsthat collectpopulation
data. The data in table 3 are providedby national statisticaloffices
or by the United Nations Population Division.
Average annual population growth rate is the exponential
rate of change for the period (see the section on statisticalmeth-
ods below).
Population ages 15-64 is a commonly accepted measureof
the number of peoplewho are potentially economicallyactive.In
WORLD DEVELOPMENT INDICATORS 3I9
many developingcountries,however,children under age 15work
full or part time, and in some high-income economies manv
workers postpone retirement past age 65.
Total labor force comprisespeople who meet the definition
establishedby the International Labour Organization (ILO) for
the economicallyactive population: all people who supply labor
for the production of goodsand servicesduring a specifiedperiod.
It includesboth the employed and the unemployed.Althoughna-
tional practicesvary, in generalthe labor force includesthe armed
forcesand first-timejob seekersbut excludeshomemakersand other
unpaid caregiversand workersin the informalsector.Data on the
labor force are compiled by the ILO from census or labor force
surveys.Despite the ILO's efforts to encourage the use of inter-
national standards, labor force data are not fully comparable be-
cause of differences among countries, and sometimes within
countries, in definitions and in methods of collection,classifica-
tion, and tabulation. The labor force estimates reported in table
3 were calculated by applying activity rates from the ILO data-
base to the World Bank's population estimates to create a labor
force seriesconsistentwith those estimates.This proceduresome-
times results in estimatesthat differslightlyfrom those published
in the ILO's Yearbookof labour Statistics.
Averageannual labor force growth rate is calculated using
the exponential endpoint method (see the section on statistical
methods below).
Females as a percentageof the labor force show the extent
to which women are activein the labor force.Estimatesare from
the ILO database. These estimates are not comparable interna-
tionally because in many countries large numbers of women as-
sist on farms or in other family enterprises without pay, and
countries use different criteria to determine the extent to which
such workers are to be counted in the labor force.
Childrenages 10-14 in the labor force are the shareof that
age group that is active in the labor force. Reliable estimates of
childlaboraredifficultto obtain.In manycountrieschildlabor
is illegal or officiallypresumed not to exist and is therefore not
reportedor included in surveysnor recorded in officialdata. Data
are alsosubject to underreportingbecausethey do not include chil-
dren engagedin agriculturalor householdactivitieswith their fam-
ilies.
Table 4. Poverty
Surveyyearisthe yearin which the underlvingdatawere collected.
Ruralpopulation below the nationalpovertyline is the per-
centageof the rural population living belowthe rural poverty line
determinedby national authorities.Urbanpopulationbelow the
national poverty line is the percentageof the urban population
living below the urban poverty line determined by national au-
thorities. Total population below the national poverty line is
the percentage of the total population living below the national
povertyline.National estimatesare based on population-weighted
subgroup estimates from household surveys.
Population below $1 a day and population below $2 a day
are the percentagesof the population living below those levelsof
consumption or income at 1993 prices, adjusted for purchasing
power parity.
320 WORLD DEVELOPMENT REPORT 2000/2001
Poverty gap at $1 a day and poverty gap at $2 a day are cal-
culated as the mean shortfall below the poverty line (counting the
nonpoor as having zero shortfall), expressed as a percentage of the
poverty line. This measure reflects the depth of poverty as well as
its prevalence.
International comparisons of poverty data entail both con-
ceptual and practical problems. Different countries have differ-
ent definitions of poverty, and consistent comparisons between
countries based on the same definition can be difficult. National
poverty lines tend to have greater purchasing power in rich
countries, where more generous standards are used than in poor
countries.
International poverty lines attempt to hold the real value of
the poverty line constant between countries. The standard of $1
a day, measured in 1985 international prices and adjusted to
local currency using PPP conversion factors, was chosen for World
Development Report 1990: Poverty because it is typical of poverty
lines in low-income economies. For this year's report, the stan-
dard has been updated to $1.08 a day, measured in 1993 inter-
national prices. PPP conversion factors are used because they
take into account the local prices of goods and services that are
not traded internationally. However, these factors were designed
not for making international poverty comparisons but for com-
paring aggregates in the national accounts. As a result, there is no
certainty that an international poverty line measures the same de-
gree of need or deprivation across countries.
Problems can arise in comparing poverty measures within
countries as well as between them. For example,the cost of food
staples-and the cost of living generally-are typically higher in
urban than in rural areas. So the nominal value of the urban
poverty line should be higher than the rural poverty line. But the
difference berween urban and rural poverty lines found in prac-
tice does not always reflect the difference in the cost of living. For
some countries the urban poverty line in common use has a
higher real value-meaning that it allows poor people to buy
more commodities for consumption-than does the rural poverty
line. Sometimes the difference has been so large as to imply that
the incidence of poverty is greater in urban than in rural areas, even
though the reverse is found when adjustments are made only for
differences in the cost of living.
Other issues arise in measuring household living standards. The
choice between income and consumption as a welfare indicator
is one. Incomes are generally more difficult to measure accu-
rately, and consumption accords better with the idea of a standard
of living than does income, which can vary over time even if the
standard of living does not. But consumption data are not always
available, and when thev are not, there is little choice but to use
income. There are still other problems. Household survey ques-
tionnaires can differ widely, for example, in the number of dis-
tinct categories of consumer goods they identify. Survey quality
varies, and even similar surveys may not be strictly comparable.
Comparisons across countries at different levels of development
also pose a potential problem because of differences in the rela-
tive importance of consumption of nonmarket goods. The local
market value of all consumption in kind (including consumption
from a household's own production, particularly important in un-
derdeveloped rural economies) should be included in the measure
of total consumption expenditure. Similarly, the imputed profit
from production of nonmarket goods should be included in in-
come. This is not always done, although such omissions were a
far bigger problem in surveys before the 1980s than today. Most
survey data now include valuations for consumption or income
from own production. Nonetheless, valuation methods vary: for
example, some surveys use the price at the nearest market, whereas
others use the average farm-gate selling price.
Whenever possible, consumption has been used as the welfare
indicator for deciding who is poor. The international poverty mea-
sures in table 4 are based on the most recent PPP estimates of con-
sumption in 1993 prices from the World Bank. When only
household income is available, average income has been adjusted
to accord with either a survey-based estimate of mean consump-
tion (when available) or an estimate based on consumption data
from national accounts. This procedure adjusts only the mean, how-
ever; nothing can be done to correct for the difference between
the Lorenz (income distribution) curves for consumption and in-
come.
Empirical Lorenz curves were weighted byvhousehold size, so
they are based on percentiles of population, not of households.
In all cases the measures of poverty have been calculated from pri-
mary data sources (tabulations or household data) rather than ex-
isting estimates. Estimates from tabulations require an interpolation
method; the method chosen is Lorenz curves with flexible func-
tional forms, which have proved reliable in past work.
Table 5. Distribution of income or
consumption
Survey year is the year in which the underlying data were collected.
Gini index measures the extent to which the distribution of
income (or, in some cases, consumption expenditure) among in-
dividuals or households within an economy deviates from a per-
fectly equal distribution. The Gini index measures the area between
the Lorenz curve (described in the technical note to table 4) and
a hypothetical line of absolute equality, expressed as a percentage
of the maximum area under the line. As defined here, a Gini index
of zero would represent perfect equality, and an index of 100 would
imply perfect inequality.
Percentage share of income or consumption is the share that
accrues to deciles or quintiles of the population ranked by income
or consumption. Percentage shares by quintiles may not add up
to 100 because of rounding.
Data on personal or household income or consumption come
from nationally representative household sur-veys.The data in the
table refer to different vears between 1985 and 1999. Footnotes
to the survey year indicate whether the rankings are based on in-
come or consumption. Distributions are based on percentiles of
population, not of households. Where the original data from the
household survey were available, they have been used to directly
calculate the income or consumption shares by quintile. Otherwise,
shares have been estimated from the best available grouped data.
The distribution indicators have been adjusted for household
size, providing a more consistent measure of income or con-
sumption per capita. No adjustment has been made for differences
in the cost of living in different parts of the same country because
 SELECTED
the necessarydata are generallyunavailable.For further details on
the estimation method for low- and middle-income economies,
see Ravallionand Chen (1996).
Becausethe underlyinghouseholdsurveysdifferin method and
in the type of data collected, the distribution indicators are not
strictly comparable acrosscountries. These problems are dimin-
ishingas surveymethods improveand becomemore standardized,
but strict comparabilityis still impossible.The income distribu-
tions and Gini indexes for the high-income economies are calcu-
lated directly from the LuxembourgIncome Study databaseusing
an estimation method consistent with that applied for develop-
ing countries.
The followingsourcesof noncomparability should be noted.
First, the surveyscan differ in many respects,including whether
they use income or consumption expenditure as the living stan-
dard indicator.Incomeis typicallymore unequallydistributedthan
consumption. In addition, the definitionsof income used in sur-
veys areusuallyverydifferentfrom the economic definitionof in-
come (themaximum levelof consumption consistentwith keeping
productive capacityunchanged). Consumption is usuallya much
better welfareindicator, particularlyin developingcountries. Sec-
ond, households differ in size (number of members) and in the
extent of income sharing among members. Individuals differ in
age and in consumption needs. Differencesbetween countries in
these respectsmay bias distribution comparisons.
Table 6. Education
Public expenditure on education is the percentageof GNP ac-
counted for by public spending on public education plus subsi-
dies to private education at the primary, secondary,and tertiary
levels.It may excludespending on religiousschools, which play
a significantrolein many developingcountries.Data for somecoun-
tries and for some years refer to spending by the ministry of ed-
ucation of the central governmentonlyand thus excludeeducation
expendituresby other central government ministriesand depart-
ments, local authorities, and others.
Net enrollment ratio is the ratio of the number of children
of official school age (as defined by the education system) who
are enrolled in school to the population of the corresponding of-
ficial school age. Enrollment data are based on annual enroll-
ment surveys,typicallyconducted at the beginning of the school
year. They do not reflectactual attendance or dropout rates dur-
ing the school year. Problems affecting cross-countrv compar-
isons of enrollment data stem from inadvertent or deliberate
misreporting of age and from errors in estimates of school-age
populations. Age structures from censuses or vital registration
systems, the primary sourcesof data on school-agepopulations,
are commonly subject to underenumeration, especiallyof young
children.
Percentageof cohort reachinggrade5 isthe shareof students
enrolled in the first gradeof primary school who eventuallyreach
fifth grade. Becausetracking data for individual students are not
available,aggregatestudent flowsfrom one grade to the next are
estimatedusingdata on enrollmentand repetitionby gradefor two
consecutiveyears.This procedure,calledthe reconstructedcohort
method, makesthreesimplifyingassumptions:that dropouts never
WORLD DEVELOPMENT INDICATORS 32I
return to school;that promotion, repetition,and dropout rates re-
mainconstantoverthe entireperiodin whichthe cohortis enrolled;
and that the same rates apply to all pupils enrolled in a given
grade, regardlessof whether they previouslyrepeated a grade.
Expected years of schooling are the averageyears of formal
schoolingthat a child isexpectedto receive,includinguniversityed-
ucationand yearsspentin repetition.This indicatormay alsobe in-
terpretedas an indicatorof the total educationresources,measured
in schoolyears,that a childwill requireoverthe courseof his or her
"lifetime"in school.
Data on education are compiled by the United Nations Ed-
ucational, Scientific, and Cultural Organization (UNESCO)
from official responsesto surveys and from reports provided by
education authorities in each country. Becausecoverage,defini-
tions, and data collection methods vary acrosscountries and over
time within countries, data on education should be interpreted
with caution.
Table 7. Health
Public expenditure on health consistsof recurrent and capital
spending from government (central and local) budgets, external
borrowings and grants (including donations from international
agenciesand nongovernmentalorganizations),and social(or com-
pulsory)health insurancefunds. Becausefew developingcountries
have national health accounts, compiling estimates of public
healthexpenditure iscomplicatedin countrieswherestate,provin-
cial, and localgovernments are involved in health care financing.
Such data are not regularly reported and, when reported, are
often of poor quality. The data on health expenditure in table 7
are the product of an effort to collectall availableinformation on
health expendituresfrom national and localgovernment budgets,
national accounts,household surveys,insurancepublications, in-
ternational donors, and existing tabulations.
Accessto improved watersource refers to the percentage of
the population with reasonableaccessto an adequate amount of
water (including treated surface water and untreated but un-
contaminatedwater, such as from springs,sanitarywells,and pro-
tected boreholes). In urban areas the source may be a public
fountain or standpipe located not more than 200 metersfrom the
residence. In rural areas the definition implies that household
members do not have to spend a disproportionate part of the day
fetching water. An "adequate" amount of water is that needed to
satisfy metabolic, hygienic, and domestic requirements, usually
about 20 liters per person per day.
Accessto sanitation is the percentageof the population with
disposalfacilitiesthat can effectivelyprevent human, animal, and
insect contact with excreta.Suitable facilitiesrange from simple
but protected pit latrines to flush toiletswith sewerage.To be ef-
fective,all facilitiesmust be correctly constructed and properly
maintained.
Infant mortalityrateis the number of infants who die before
reaching 1 year of age, expressedper 1,000 live births in a given
year (seethe discussionof age-specificmortality rates in the tech-
nical note to table 2).
Contraceptiveprevalence rate is the percentage of women
who are practicing,or whosesexualpartnersare practicing,any form
322 WORLD DEVELOPMENT REPORT 2000/2001
of contraception.It is usuallymeasured for marriedwomen aged
15-49 only. Contraceptiveprevalenceincludes all methods: inef-
fectivetraditionalmethodsas well as highlyeffectivemodern meth-
ods. Unmarriedwomen are often excluded from the surveys,and
this may bias the estimate.The rates are obtainedmainly from de-
mographicand healthsurveysand contraceptiveprevalencesurveys.
Total fertility rate is the number of children who would be
born to a woman if she were to live to the end of her childbear-
ing yearsand bear children in accordancewith current age-specific
fertilityrates. Data are from vital registrationsystemsor, in their
absence,from censusesor sample surveys.Provided that the cen-
suses or surveysare fairly recent, the estimated rates are consid-
ered reliable. As with other demographic data, international
comparability is affected by differencesin definitions, data col-
lection, and estimation methods.
Maternalmortalityratio isthe number of women who die dur-
ing pregnancyor childbirth,per 100,000livebirths.Householdsur-
veyssuchas the demographicand healthsurveysartemprto measure
maternalmortalitybyasking respondentsabout survivorshipof sis-
ters.The main disadvantageof this method is that the estimatesof
maternal mortalitythat it producespertain to 12yearsor so before
the survey,makingthem unsuitablefor monitoringrecentchanges
or observingthe impact of interventions.In addition,measurement
of maternalmortalityissubjectto manvtypesof errors.Evenin high-
income countrieswith vital registrationsystems,misclassification
of maternaldeaths has been found to leadto seriousunderestima-
tion. The data in the table are officialestimatesbased on national
surveysor derivedfrom officialcommunity and hospital records.
Some reflectonly births in hospitalsand other medicalinstitutions.
In somecasessmallerprivateand rural hospitalsare excluded,and
sometimes even primitive local facilitiesare included. Thus the
coverageis not alwayscomprehensive,and cross-countrycompar-
isons should be made with extremecaution.
Table 8. Land use and agricultural
productivity
Land underpermanentcrops is land cultivated with crops that
occupy the land for long periods and do nor need to be replanted
after each harvest,excluding trees grown for wood or timber. Ir-
rigated land refers to areas purposely provided with water, in-
cludingland irrigatedbycontrolledflooding.Arablelandincludes
land defined by the Food and Agriculture Organization (FAO)
as land under temporarycrops (double-croppedareasare counted
once), temporary meadowsfor mowingor for pasture, land under
marker or kitchengardens,and land remporari]yfallow.Land aban-
doned as a result of shifting cultivation is not included.
The comparability of land use data from different countries
islimited by variations in definitions,statisticalmethods, and the
quality of data collection. For example, countries may define
land risedifferently.The FAO, the primary compilerof these data,
occasionally adjusts its definitions of land use categories and
sometimesrevisesearlierdata. Becausethe data thus reflectchanges
in data reporting proceduresas well as actual changesin land use,
apparent trends should be interpreted with caution.
Agriculturalmachinery refers to wheel and crawler tractors
(excludinggarden tractors) in use in agriculture at the end of the
calendar year specified or during the first quarter of the follow-
ing year. Agriculturalproductivity refers to agricultural value
added per agricultural worker, measured in constant 1995 U.S.
dollars.Agriculturalvalue added includesthat from forestryand
fishing.Thus interpretationsof land productivityshould be made
with caution. To smooth annual fluctuations in agriculturalac-
tivirv, the indicators have been averagedover three years.
Food productionindex coversfood crops that are considered
edibleand that contain nutrients. Coffee and tea are excludedbe-
cause,although edible,they haveno nutritive value.The food pro-
duction index is prepared by the FAO, which obtains data from
official and semiofficialreports of crop yields, area under pro-
duction, and livestocknumbers. Where data are not available,the
FAO makesestimates.The index is calculatedusing the Laspeyres
formula:the productionquantitiesofeach commodityare weighted
by average international commodity prices in the base period
and summed for each year.The FAO's indexmay differfrom those
of other sourcesbecause of differencesin coverage,weights, con-
cepts,time periods,calculationmethods, and use of international
prices.
Table 9. Water use, deforestation, and
protected areas
Freshwaterresourcesreferto total renewableresources,which in-
clude flowsof rivers and groundwater from rainfall in the coun-
trv and river flowsfrom other countries. Freshwaterresourcesper
capita arecalculatedusing the World Bank's population estimates.
Data on freshwaterresourcesare based on estimatesof runoff
into riversand rechargeof groundwater.These estimatesare based
on different sourcesand referto differentyears, so cross-country
comparisonsshould be made with caution. Becausethey are col-
lected intermittently, the data may hide significantvariations in
total renewablewater resourcesfrom one year to the next. These
annual averagesalso obscurelarge seasonaland interannual varia-
tions in water availabilitywithin countries. Data for small coun-
tries and countriesin arid and semiaridzones are less reliablethan
those for larger countriesand countries with more rainfall.
Annual freshwater withdrawals refer to total water with-
drawals, nor counting evaporation losses from storage basins.
They also include water from desalination plants in countries
where these are a significantsource of water. Withdrawal data are
for singleyearsbetween 1980and 1998unlessotherwiseindicated.
Caution is advisedin comparing data on annual freshwaterwith-
drawals,which are subject to variations in collection and estima-
tion methods. Withdrawals can exceed 100 percent of renewable
supplies when extraction from nonrenewable aquifers or desali-
nation plants is considerableor when there is significantreuse of
water. Withdrawals for agriculture and industrv are total with-
drawalsfor irrigation and livestockproduction and for direct in-
dustrial use (including withdrawals for cooling thermoelectric
plants).Withdrawalsfordomesticuse includedrinkingwater, mu-
nicipal use or supply, and use for public services,commercial es-
tablishments,and homes.For most countries,sectoralwithdrawal
data are estimated for 1987-95.
Access to improvedwater source refers to the percentageof
the population with reasonableaccessto an adequate amount of
 SELECTED
water (including treated surfacewater and untreated but uncon-
taminated water,such as from springs,sanitarywells, and protected
boreholes; see table 7). While information on accessto water is
widelyused,it is extremelysubjective,and such termsas "adequate"
may have very different meanings in different countries. Even in
high-income countries treated water may not alwaysbe safe to
drink. Althoughaccessto an improvedsourceis often equatedwith
connection to a public supplysystem, this does not take account
of variations in the quality and cost (broadly defined) of the ser-
vice once connected. Thus cross-country comparisons must be
made cautiously. Changesover time within countries may result
from changes in definitions or measurements.
Annual deforestation refers to the permanent conversion of
forest area (land under natural or planted stands of trees)to other
uses,includingshifting cultivation,permanent agriculture,ranch-
ing, settlements,and infrastructuredevelopment.Defotesredareas
do not include areas logged bur intended for regeneration or
areas degraded by fuelwood gathering,acid precipitation, or for-
est fires. Negative numbers indicate an increasein forest area.
Estimatesof forestarea are from the FAO's Stateof the Worlds
Forests 1999,whichprovidesinformationon forestcoveras of 1995
and revisedestimates of forest cover in 1990. Forest coverdata
for developingcountriesarebasedon country assessmentsthat were
prepared at different times and that, for reporting purposes, had
to be adapted to the standard referenceyears of 1990 and 1995.
This adjustment was made with a deforestation model designed
to correlate forest cover change over time with certain ancillarv
variables,including population change and density. initial forest
cover, and ecologicalzone of the forest area under consideration.
Nationallyprotectedareasreferto totallyor partiallyprotected
areasof at least 1,000 hectaresthat are designatedas nationalparks,
natural monuments, nature reserves,wildlife sanctuaries, pro-
tected landscapesand seascapes,or scientificreserveswith limited
public access.The indicator is calculatedas a percentageof total
area. For smallcountrieswhoseprotectedareasmay be smallerthan
1,000 hectares,this limit will resultin an underestimate of the ex-
tent and number of procectedareas.The data do not include sites
protected under local or provincial law.
Data on protected areas arecompiled from a varietyof sources
by the World Conservation Monitoring Centre, a joint venture
of the United Nations Environment Programme. the World
Wide Fund for Nature, and the World Conservation Union. Be-
causeof differencesin definitions and reporting practices, cross-
country comparabilityislimited. Compounding these problems,
the data availablecover different periods. Designating land as a
protected areadoes not necessarilymean, moreover,that protec-
tion is in force.
Table 10. Energy use and emissions
Commercial energy use is measured as apparent consumption,
which is equal to indigenous production plus imports and stock
changes,minus exportsand fuelssupplied to ships and aircraften-
gaged in international transportation. The InternatiornalEnergy
Agency(IEA)and the United Nations StatisticsDivisioni(UNSD)
compileenergy data. IEA data for countriesthat are not members
of the Organisationfor EconomicCo-operationand Development
WORLD DEVELOPMENT INDICATORS 323
(OECD) arebased on nationalenergydata that havebeen adjusted
to conform with annual questionnaires completed by OECD
member governments. UNSD data are compiled primarily from
responses to questionnaires sent to national governments, sup-
plemented by officialnational statisticalpublicationsand by data
from intergovernmental organizations. When official data are
not available,the UNSD bases its estimates on the professional
and comrmercialliterature. The variety of the sourcesaffectsthe
cross-countrycomparability of data.
Commercialenergy use refers to domestic primary energyuse
beforetransformationto other end-useenergysources(suchas elec-
tricity and refined petroleum products). It includesenergy from
combustible renewables and waste. All forms of commercial
energy-primary energy and primary electricitv-are converted
into oil equivalents.To convert nuclear electricityinto oil equiv-
alents.a notional thermal efficiencyof 33 percent is assumed;for
hydroelectricpower, 100 percent efficiencyis assumed.
GDP per unit of energy use is the P'I'PGDP per kilogram
of oil equivalentof commercialenergy use. PPP GDP is grossdo-
mestic product converted to international dollars using purchas-
ing power parity rates. An international dollar has the same
putrchasingpower as a U.S. dollar has in the United States.
Net energy imports are calculated as energy use less produc-
tion. both measured in oil equivalents.A negativevalue indicates
that the country is a net exporter of energy.
Carbon dioxide emissions referto those stemming from the
burning of fossil fuels and the manufactLre of cement. These
emissionsinclude carbon dioxide produced during consurnption
of solid, liquid, and gas fucls and from gas flaring.
The Carbon Dioxide Information AnalysisCenter (CDIAC),
sponsored by the U.S. Department of Energy, calculatesannual
anthropogenicemissionsof carbon dioxide.These calculationsare
based on data on fossilfuel consumption from the World Energy
Data Set maintained by the UNSD. and data on world cement
manufacturing from the Cement Manufacturing Data Set main-
tained by the U.S. Bureau of Mines. Each vear the CDIAC re-
calculates the entire time series from 1950 to the present,
incorporating its most recent findings and the latest corrections
to its database. Estimatesexclude fuels supplied to ships and air-
craft engaged in international transportation because of the dif-
ficulty of apportioning these futelsamong the countriesbenefiting
from that transport.
Table 11. Growth of the economy
Gross domesticproduct isgrossvalue added,at purchaserprices,
by all resident producers in the economy plus any taxes and
niinus any subsidiesnot included in the value of the products. It
is calculatedwithout deducting for depreciation of fabricatedas-
sets or for depletion or degradation of natural resources. Value
added is the net output of a sector after adding up all outputs and
subtracting intermediate inputs. The industrial origin of value
added is determined by the International Standard Industrial
Classification (ISIC) revision 3.
GDP implicit deflator refiects changesin prices for all final
demand categories, such as government consumption, capital
formation, and international trade, as well as the main compo-
324 WORLD DEVELOPMENT REPORT 2000/2001
nent, privatefinal consumption. It is derived as the ratio of cur-
rent to constant price GDP. The GDP deflator may also be cal-
culated explicitlyas a Paascheprice index in which the weights
are the current period quantities of output.
Agriculturevalue added corresponds to ISIC divisions 1-5
and includesforestryand fishing.Industryvalue added comprises
the followingsectors: mining (ISIC divisions 10-14), manufac-
turing (15-37), construction (45), and electricity,gas, and water
supply (40 and 41). Servicesvalue added corresponds to ISIC
divisions 50-99.
Exportsof goods and servicesrepresentthe value of all goods
and servicesprovidedto the restof the world. Included is the value
of merchandise,freight,insurance,travel,transport, and other ser-
vices such as communications and financial services. Factor and
property income (formerly called factor services), such as invest-
ment income, interest, and labor income, is excluded, as are trans-
fer payments.
Gross domestic investment consists of outlays on additions
to the fixed assets of the economy plus net changes in the level of
inventories. Additions to fixed assets include land improvements
(fences,ditches, drains, and so on); plant. machinery, and equip-
ment purchases;and the constructionof buildings,roads,railways,
and the like, including commercial and industrial buildings, of-
flces, schools, hospitals, and private dwellings. Inventories are
stocks of goods held by firms to meet temporary or unexpected
fluctuations in production or sales.
Growth rates are annual averages calculated using constant price
data in local currency. Growth rates for regional and income
groups are calculated after converting local currencies to U.S.
dollars at the averageofficial exchangerate reported by the In-
cernational Monetary Fund (IMF) for the year shown or, occa-
sionallv, using an alternative conversion factor determined bv
the World Bank's Development Data Group. Alternative con-
version factors and methods of computing growth rates are de-
scribedin the section on statisticalmethods below.For additional
information on the calculation of GDP and its sectoral compo-
nents, see the technical note to table 12.
Table 12. Structure of output
Gross domestic product represents the sum of value added by
all producers in the economy (see the technical note to table 1I1
for a more detailed definition and for definitions of agriculture,
industry, manufacturing, and services value added). Since
1968 the United Nations' System of National Accounts (SNA)
has called for estimates of GDP by industrial origin to be valued
at producer prices (including taxes on factors of production, but
excluding indirect taxes on final output). Some countries, how-
ever,report such data at basicprices-the pricesat whichfinal sales
are made (including indirect taxes)-and this may affect esti-
mates of the distribution of output. Total GDP as showniin this
table is measuredat purchaserprices.GDP components are mea-
sured at either basic or producer prices.
Amongthe difficultiesfaced by compilersof nationalaccounts
is the extentof unreportedeconomicactivityin the informalor sec-
ondary economy. In developing countries a large share of agricul-
tural output is either not exchanged(becauseit isconsumedwithin
the household)or not exchangedfor money.Financialtransactions
also may go unrecorded.As a result, agriculturalproduction often
must be estimated indirectly,using a combinationof methods in-
volvingestimatesof inputs, yields,and areaunder cultivation.
The output of industry ideally should be measured through
regular censuses and surveys of firms. But in most developing
countries such surveys are infrequent and quickly go out of
date, so many results must be extrapolated. The choice of sam-
pling unit, which may be the enterprise (where responses may
be based on financial records) or the establishment (where pro-
duction units may be recorded separately),also affectsthe qual-
ity of the data. Moreover,much industrialproduction is organized
not in firms but in unincorporated or owner-operated ventures
not captured by surveys aimed at the formal sector. Even in large
industries, where regular surveys are more likely, evasion of ex-
cise and other taxes lowers the estimates of value added. Such
problems become more acute as countries move from state con-
trol of industry to private enterprise because new firms go into
business and growing numbers of established firms fail to report.
In accordance with the SNA, output should include all such un-
reported activitv as well as the value of illegal activities and
other unrecorded, informal, or small-scaleoperations. Data on
these activities need to be collected using techniques other than
conventional surveys.
In industries dominated bv large organizations and enter-
prises, data on output, employment, and wages are usually read-
ily available and reasonably reliable. But in the service industry
the many self-employedworkers and one-person businessesare
sometimes difficultto locate, and their ownershave little incen-
tiveto respond to surveys,let alone report their full earnings.Com-
pounding these problemsare the many formsof economicactivity
that go unrecorded, includingthe work that women and children
do for little or no pay. For further discussionof the problems en-
countered in using national accounts data, see Srinivasan(1994)
and Heston (1994).
Table 13. Structure of demand
Private consumption is the market value of all goods and ser-
vices, including durable products (such as cars, washing ma-
chines, and home computers), purchased or received as income
in kind by households and nonprofit institutions. It excludes pur-
chases of dwellings but includes imputed rent for owner-occu-
pied dwellings. In practice, it may include any statistical
discrepancy in the use of resources relative to the supply of
resources.
Private consumption is often estimated as a residual,by sub-
tracting from GDP all other known expenditures. The resulting
aggregatemay incorporate fairly large discrepancies.When pri-
vateconsumption is calculatedseparately,the household surveys
providingthe basisfor a largepart of the estimatestend to be one-
yearstudies with limited coverage.Thus the estimatesquicklybe-
come outdated and must be supplemented by price- and
quantiry-based statistical estimating procedures. Complicating
the issue, in many developingcountries the distinction between
cash outlays for personal business and those for household use may
be blurred.
 SELECTED
Generalgovernment consumption includes allcurrentspend-
ing for purchasesof goodsand services(includingwagesand
salaries)by alllevelsof government,excludingmostgovernment
enterprises.It alsoincludesmostexpenditureon nationaldefense
and security,someof whichis nowconsideredpartof investment
in the 1993SNA.
Grossdomesticinvestmentconsistsof outlayson additions
to the fixedassetsof the economyplusnet changesin the levelof
inventories.Fordefinitionsof fixedassetsand inventories,seethe
technicalnote to table11.Underthe 1993SNAguidelines, gross
domesticinvestmentalsoincludescapitaloutlaysonidefensees-
tablishmentsthat may be used by the generalpublic, such as
schools,hospitals,and certaintypesofprivatehousiisgfor family
use.Allotherdefenseexpenditures aretreatedascurrentspending.
Investmentdata maybe estimatedfrom directsurveysof en-
terprisesand administrative recordsor basedon the commodity
flowmethod,usingdata fromtradeand constructionactivities.
The qualityofpublicfixedinvestmentdata dependson the qual-
ity of governmentaccountingsystems,whichtend to be weakin
developingcountries.Measuresof privatefixed investment-
particularlycapitaloutlaysbysmall,unincorporated enterprises-
areusuallyveryunreliable.
Estimatesofchangesininventories arerarelycompletebutusu-
ally includethe most importantactivitiesor commodities.In
somecountriestheseestimatesarederivedas a compositeresid-
ual alongwithaggregateprivateconsumption.Accordingto na-
tional accountsconventions,adjustmentsshould be made for
appreciationof inventoryvaluedue to pricechanges,but this is
not alwaysdone.In economieswhereinflationis high,this ele-
ment canbe substantial.
GrossdomesticsavingsarethedifferencebetweenGDPand
totalconsumption.
Exportsof goodsandservicesrepresentthevalueof allgoods
and services (includingfreight,transport,travel,and otherservices
suchascommunications, insurance,and financialservices)
provided
to the restof theworld.Dataonexportsand importsarecompiled
from customsreturnsand from balanceof paymentsdata ob-
tainedfromcentralbanks.Althoughdata on exportsand imports
from the paymentsside providereasonablyreliablerecordsof
cross-bordertransactions, theymaynot adherestrictlyto the ap-
propriatevaluationand timingdefinitionsofbalanceof payments
accountingor, more important,correspondto the change-of-
ownershipcriterion.(In conventionalbalanceof paymentsac-
counting,a transactionis recordedasoccurringwhenownership
changeshands.)This issuehasassumedgreatersignificance with
the increasingglobalization ofbusiness.Neithercustomsnor bal-
anceof paymentsdata capturethe illegaltransactionsthat occur
inmanycountries. Goodscarriedbytravelers acrossbordersinlegal
but unreportedshuttletrademayfurtherdistorttradestatistics.
Resourcebalanceis the differencebetweenexportsof goods
and servicesand importsof goodsand services.
Table 14. Central government finances
Currenttaxrevenuecomprisescompulsory,unrequited,nonre-
payablereceiptscollectedby centralgovernments for publicpur-
poses.It includesinterestcollectedon tax arrearsand penalties
WORLD DEVELOPMENT INDICATORS 325
collectedon nonpaymentor latepaymentof taxes.It is shownnet
of refundsand othercorrectivetransactions.
Currentnontaxrevenueincludesrequited,nonrepayable re-
ceiptsfor publicpurposes,suchasfines,administrative fees,or en-
trepreneurialincomefromgovernmentownership ofproperty,and
voluntary,unrequited,nonrepayable currentgovernmentreceipts
other than fromgovernmentsources.Thiscategorydoesnot in-
cludegrants,borrowing,repaymentof previouslending,or sales
of fixedcapitalassetsor of stocks,land,or intangibleassets,nor
doesit includegiftsfromnongovernment sourcesforcapitalpur-
poses.Together,taxandnontaxrevenuemakeupthe currentrev-
enueof the government.
Currentexpenditure includesrequitedpaymentsotherthan
forcapitalassetsor forgoodsor services tobe usedin theproduction
of capitalassets.It alsoincludesunrequitedpaymentsfor purposes
other than permittingthe recipientsto acquirecapital assets,
compensatingthe recipientsfor damageto or destructionof cap-
ital assets,or increasingthe financialcapitalof the recipients.
Currentexpendituredoesnot includegovernmentlendingor re-
paymentsto the government,or governmentacquisitionof eq-
uiryfor publicpolicypurposes.
Capitalexpenditure is nonmilitaryspendingto acquirefixed
capitalassets,land,intangibleassets,and strategicand emergency
stocks.Alsoincludedareunrequitedcapitaltransfersfor acquir-
ingcapitalassetsor to increasethe financialcapitalof recipients.
Overalldeficit/surplus is currentand capitalrevenueandof-
ficialgrantsreceived,lesstotalexpenditureand lendingminusre-
payment.
Goodsand servicesexpenditurecomprisesallgovernment
paymentsin exchangeforgoodsandservices,includingwagesand
salaries.
Socialservicesexpenditure comprisesexpenditure on health,
education,housing,welfare,socialsecurity,and community
amenities.It alsocoverscompensationfor lossof incometo the
sickand temporarilydisabled;paymentsto the elderly,the per-
manentlydisabled,and the unemployed;family,maternity,and
childallowances; and the costof welfareservicessuchas care of
the aged,thedisabled,and children.Manyexpenditures relevant
to environmental protection,suchas pollutionabatement,water
supply,sanitation,and refusecollection,are includedindistin-
guishablyin this category.
Dataon governmentrevenuesand expenditures arecollected
by the IMF,throughquestionnaires distributedto membergov-
ernments,and bythe OECD.In general,the definitionof gov-
ernment excludesnonfinancialpublic enterprisesand public
financialinstitutions(suchas thecentralbank).Despitethe IMF's
effortsto systematize and standardizethe collectionof publicfi-
nancedata,statisticson publicfinanceareoftenincomplete,un-
timely, and noncomparable.Inadequatestatisticalcoverage
precludesthe presentationof subnationaldata, makingcross-
countrycomparisonspotentiallymisleading.
Total centralgovernmentexpenditureas presentedin the
IMF's Government FinanceStatisticsYearbook is a more limited
measureof generalgovernmentconsumptionthan thatshownin
the nationalaccountsbecauseit excludesconsumptionexpendi-
ture bystateandlocalgovernments. At thesametime,theIMF's
conceptof centralgovernment expenditure is broaderthanthe na-
326 WORLD DEVELOPMENT REPORT 2000/2001
rional accounts definition because it includes government gross
domestic investment and transfer payments.
Central government finances can refer to one of two ac-
counting concepts: consolidated or budgetary. For most countries
central government finance data have been consolidated into one
account, but for others only budgetary central government accounts
are available. Countries reporting budgetary data are noted in the
primary data documentation table in World Development Indica-
tors 2000. Because budgetary accounts do not necessarily include
all central government units, the picture they provide of central
government activities is ustially incomplete. A key issue is the fail-
ure to include the quasi-fiscal operations of the central bank.
Central bank losses arising from monetary operations and subsi-
dized financing can result in sizable quasi-fiscal deficits. Such
deficits may also result from the operations of other financial in-
termediaries, such as public development finance institutions.
Also missing from the data are governments' contingent liabili-
ties for unfunded pension and insurance plans.
Table 15. Balance of payments current
account and international reserves
Goods and services exports and goods and services imports to-
gether comprise all transactions between residents of a coLntry and
the rest of the world involving a change in ownership of general
merchandise, goods sent for processing and repairs, nonmonerary
gold, and services.
Net income refers to compensation earned by workers in an
economy other than the one in which they are resident, for work
performed for and paid for by a resident of that economy, and in-
vestment income (receipts and payments on direct investment, port-
folio investment, and other investment, and receipts on reserve
assets). Income derived from the use of intangible assets is recorded
under business services.
Net current transfers consist of transactions in which residents
of an economy provide or receivegoods, services,income, or fi-
nancial irems without a quid pro quo. All transfers not consid-
ered to be capital transfers are current transfers.
Current account balance is the sum of net exports of goods
and services, inconme, and current transfers.
Gross international reserves comprise holdings of monetary
gold, special drawing rights, reserves of IMF members held by the
IMF, and holdings of foreign exchange under the control of mon-
etarv authorities. 'I'he gold component of these reserves is valued
at year-end london prices ($385.00 an ounce in 1990 and
$290.25 an ounce in 1999).
'['he balance of payments is divided into two groups of accounts.
The current account records transactions in goods and services,
income, and current transfers. The capital and financial account
records capital transfers, the acquisition or disposal of nonproduced,
nonfinancial assets (such as patents), and transactions in financial
assets and liabilities. Gross international reserves are recorded in
a third set of accounts, the international investment position,
wvhich records the stocks of assets and liabilities.
The balance of payments is a double-entry accounting system
that shows all flows of goods and services into and out of an
economy; all transfers that are the counterpart of real resources
or financial claims provided to or by the rest of the world with-
out a quid pro quo. such as donations and grants; and all changes
in residents' claims on, and liabilities to, nonresidents that arise
from economic transactions. All transactions are recorded twice:
once as a credit and once as a debit. In principle the net balance
should be zero, but in practice the accounts often do not balance.
In these cases a balancing item, called net errors and omissions,
is included in the capital and financial account.
Discrepancies may arise in the balance of payments because
there is no single source for balance of payments data and no way
to ensure that data from different sources are fully consistent.
Sources include customs data, monetary accounts ofthe banking
system, external debt records, information provided by enter-
prises, surveys to estimate service transactions, and foreign exchange
records. Differences in recording methods-for example, in the
timing of transactions, in definitions of residence and owner-
ship, and in the exchange rate used to value transactions
-contribute to net errors and omissions. In addition, smuggling
and other illegal or quasi-legal transactions may be unrecorded or
misrecorded.
The concepts and definitionsunderlying the data in table 15
are based on the fifth edition of the IMF's BalanceofPayments Man-
ual That edition redefined as capital transfers some transactions
previously included in the current account, such as debt forgive-
ness, migrants' capital transfers, and foreign aid to acquire capital
goods. Thus the current account balance now more accurately
reflects net current transfer receipts in addition to transactions in
goods, services (previously nonfactor services), and income (pre-
viously factor income). Many countries still maintain their data col-
lection systems according to the concepts and definitions in the
fourth edition. WXherenecessary, the IMF converts data reported
in earlier systems to conform with the fifth edition (see the primary
data documentation table in World Development indicators2000).
Values are in U.S. dollars converted at market exchange rates.
Table 16. Private sector finance
Private investment covers gross outlays by the private sector (in-
cluding private nonprofit agencies) on additions to its fixed do-
mestic assets. When direct estimates of private gross domestic fixed
investment are nor available, such investment is estimated as the
difference between total gross domestic investment and consoli-
dated public investment. No allowance is made for the depreci-
ation of assets. Because private investment is often estimated as
the difference between two estimated quantities-domestic fixed
investment and consolidated public investment-private invest-
ment ma) be undervalued or overvalued and subject to errors over
time.
Stock market capitalization (also called market value) is the
sum of the market capitalizations of all firms listed on domestic
stock exchanges, where each firm's market capitalization is its share
price at the end of the year times the number of shares outstand-
ing. Market capitalization, presented as one measure for gauging
a country's level of stock market development, may not be strictly
comparable across countries as a result of conceptual and statis-
tical weaknesses such as inaccurate reporting and differences in ac-
counting standards.
 SELECTED
Listed domestic companies refer to the number of domesti-
callyincorporatedcompanieslistedon the countrys stockexchanges
at the end of the year. Excludedare investment companies,mu-
tual funds, and other collective investment vehicles. Data on
stockmarket capitalizationand listed domesticcompaniesarefrom
Standard & Poor's EmergingStockMarketsFactbook2000.
Interest rate spread (alsoknown as the intermediation mar-
gin) is the difference between the interest rate charged by banks
on short- and medium-term loans to the private sector and the
interest rate offered by banks to resident customers for demand,
time, or savingsdeposits. Interest ratesshould reflectthe respon-
sivenessof financial institutions to competition and price incen-
tives. However, the interest rate spread may not be a reliable
measure of a banking system's efficiency,to the extent that in-
formationabout interestratesis inaccurate,that banksdo not mon-
itor all bank managers,or that the government sets deposit and
lending rates.
Domestic credit provided by the banking sector includes
all credit to various sectorson a grossbasis,with the exceptionof
credit to the central government,which is net. The banking sec-
tor includes monetary authorities, deposit money banks, and
other banking institutions for vhich data are available(including
institutions that do not accept transferabledeposits but do incur
such liabilities as time and savingsdeposits). Examplesof other
banking institutions include savingsand mortgageloan institu-
tions and building and loan associations.
In general,the indicators reported in table 16 do not capture
the activitiesof the informal sector,which remains an important
source of finance in developingeconomies.
Table 17. Role of government in the
economy
Subsidies and other current transfers include all unrequited, non-
repayable transfers on current account to private and public en-
terprises and the cost to the public of covering the cash operating
deficits on sales to the public by departmental enterprises.
Value added by state-owned enterprises is estimated as sales
revenue minus the cost of intermediate inputs, or as the sum of these
enterprises' operating surplus (balance) and their wage payments.
State-owned enterprises are government-owned or -controlled eco-
nomic entities that generate most of their revenue by selling goods
and services. This definition encompasses commercial enterprises
direcdy operated by a government department and those in which
the government holds a majoriry of shares directly or indirectly
through other state enterprises. It also includes enterprises in which
the state holds a minority ofshares, if the distribution ofthe remaining
shares leaves the government with effective control. It excludes
public sector activity-such as education, health services, and road
construction and maintenance-that is financed in other ways,
usuallyfrom the government'sgeneral revenue.Becausefinancial
enterprises are of a different nature, they have generallybeen excluded
from the data.
Military expenditures for members of the North Atlantic
Treaty Organization (NATO) are based on the NATO definition,
which covers military-related expenditures of the defense ministry
(including recruiting, training, construction, and the purchase of
WORLD DEVELOPMENT INDICATORS 327
military supplies and equipment) and other ministries.Civilian-
type expendituresof the defense ministry are excluded. Military
assistanceis included in the expendituresof the donor country.
Purchasesof military equipmenton credit are includedat the time
the debt is incurred, not at the time of payment. Data for coun-
tries that are not membersof NATO generallycoverexpenditures
of the ministry of defense;excluded are expenditures on public
order and safety,which are classifiedseparately.
Definitions of military spending differdepending on whether
they include civil defense, reserves and auxiliary forces, police
and paramilitaryforces,dual-purpose forcessuch as military and
civilianpolice, military grants in kind, pensions for military per-
sonnel, and socialsecurity contributions paid by one part ofgov-
ernment to another. Official government data may omit some
military spending, disguisefinancingthrough extrabudgetaryac-
counts or unrecorded use of foreign exchangereceipts, or fail to
include militaryassistanceor secretimports of militaryequipment.
Current spending is more likelyto be reported than capitalspend-
ing. In some casesa more accurate estimate of military spending
can be obtained by adding the value of estimated arms imports
and nominal military expenditures.This method may understate
or overstate spending in a particular year, however,because pay-
ments for arms may not coincide with deliveries.
The data on militaryexpendituresin table 17 are from the U.S.
Department of State'sBureau ofArms Control. The IMF's Goy-
ernmentFinianceStatisticsYearbookis a primary source for data on
defensespending. It uses a consistentdefinition of defense spend-
ing based on the NATO definition and the United Nations' clas-
sification of the functions of government. The IMF checks data
on defense spending for broad consistencywith other macroeco-
nomic data reported to it, but it is not alwaysable to verify the
accuracy and completeness of such data. Moreover, country cov-
erage is affected by delays or failure to report data. Thus most re-
searchers supplement the IMF's data with assessments by other
organizations such as the Bureau of Arms Control, the Stockholm
International Peace Research Institute, and the International In-
stitute for Strategic Studies. However, these agencies rely heavily
on reporting by governments, on confidential intelligence estimates
of varying quality, on sources that they do not or cannot reveal,
and on one another's publications.
Composite ICRG risk rating is an overall index of investment
risk in a country, taken from the PRS Group's International
CouintryRisk Guide. The index is based on 22 components of risk.
The PRS Group collects information on each component, groups
these components into three major categories (political, financial,
and economic), and converts the information into a single nu-
merical risk assessment ranging from 0 to 100. Ratings below 50
indicate very high risk and those above 80 very low risk. Ratings
are updated every month.
Institutional Investor credit rating ranks, from 0 to 100, the
probability of a country's default. A high number indicatesa low
probability of default. Institutional Investor country credit ratings
arebased on information providedby leadinginternationalbanks.
Responses are weighted using a formula that gives more impor-
tance to responses from banks with greater worldwide exposure
and more sophisticated country analysis systems. Ratings are up-
dated even, six months.
328 WORLD DEVELOPMENT REPORT 2000/2001
Risk ratings may be highly subjective, reflecting external per-
ceptions that do not always capture a country's actual situation.
But these subjective perceptions are the reality that policvmakers
face in the climate they create for foreign private inflows. Coun-
tries not rated favorably by credit risk rating agencies typically do
not attract registered flows of private capital. The risk ratings
presented here are not endorsed by the World Bank but are in-
cluded for their analytical usefulness.
Highest marginal tax rate is the highest rate shown on the
schedule of tax rates applied to the taxable income of individuals
and corporations. The table also presents the income threshold
above which the highest marginal tax rate applies for individuals.
Tax collection systems are often complex, containing many ex-
ceptions, exemptions, penalties, and other inducements that af-
fect the incidence of taxation and thus influence the decisions of
workers, managers, entrepreneurs, investors, and consumers. A po-
tentially important influence on both domestic and international
investors is the tax system's progressivity, as reflected in the high-
est marginal tax rate on individual and corporate income. Mar-
ginal tax rates for individuals generally refer to employment
income. For some countries the highest marginal tax rate is also
the basic or flat rate, and other surtaxes, deductions, and the like
may apply.
Table 18. Power and transportation
Electric power consnmption per capita measures the produc-
tion of power plants and combined heat and power plants less trans-
mission, distribution, and transformation losses and own use by
heat and power plants. Electric power transmission and distri-
bution losses are losses in transmission between sources of sup-
ply and points of distribution, and in distribution to consumers,
including pilferage.
The International Energy Agency collects data on eLectric
power production and consumption from national energy agen-
cies and adjusts those data to meet international definitions-for
example, to account for self-production by establishments that, in
addition to their main activities, generate electricity wholly or
partly for their own use. In some countries self-production by
households and small entrepreneurs is substantial because of their
remoteness or because public power sources are unreliable, and in
these cases adjustments may not adequately reflect actual output.
Although power plants' own consumption and transmission
losses are netted out, electric power consumption includes con-
sumption by auxiliary stations, losses in transformers that are
considered integral parts of those stations, and electricity produced
by pumping installations. Where data are available, consumption
covers electricity generated by all primary sources of energy: coal,
oil, gas, nuclear, hydroelectric, geothermal, wind, tide and wave,
and combustible renewables. Neither production nor consump-
tion data capture the reliability of supplies, including the frequency
of outages, breakdowns, and load factors.
Paved roads are those surfaced with crushed stone (macadam)
and hydrocarbon binder or bituminized agents, with concrete, or
with cobblestones, as a percentage of all the country's roads, mea-
sured in length. Goods transported by road are the volume of
goods hauled by road vehicles, measured in millions of metric tons
times kilometers traveled. Goods transported by rail are the
tonnage of goods transported times kilometers traveled per mil-
lion dollars of GDP measured in PPP terms. Air passengers car-
ried include those on both domestic and international passenger
routes.
Data for most transport sectors are not internationally com-
parable because-unlike for demographic statistics, national in-
come accounts, and international trade data-the collection of
infrastructure data has not been "internationalized." Data on
roads are collected by the International Road Federation (IRF),
and data on air transportation by the International Civil Aviation
Organization. National road associations are the primary source
of IRF data; in countries where such an association is lacking or
does not respond, other agencies are contacted, such as road di-
rectorates, ministries of transportation or public works, or cen-
tral statistical offices. As a result, the compiled data are of uneven
quality.
Table 19. Communications,information,
and science and technology
Daily newspapers are the number of copies distributed of news-
papers published at least four times a week, per 1,000 people. Ra-
dios are the estimated number of radio receivers in use for
broadcaststo the general public, per 1,000 people. Data on these
two indicators are obtained from statistical surveys carried out by
UNESCO. In some countries definitions, classifications, and
methods of enumeration do not entirely conform to UNESCO
standards. For example, some countries report newspaper circu-
lation as the number of copies printed rather than the number dis-
tributed. In addition, many countries impose radio license fees to
help pay for public broadcasting, discouraging radio owners from
declaring ownership. Because of these and other data collection
problems, estimates ofthe number ofnewspapers and radios vary
widely in reliability and should be interpreted with caution.
Television sets are the estimated number of sets in use, per
1,000 people. Data on television sets are supplied to the Inter-
national Telecommunication Union (ITU) through annual ques-
tionnaires sent to national broadcasting authorities and industry
associations. Some countries require that television sets be regis-
tered. To the extent that households do not register some or all
of their sets, the number of licensed sets may understate the true
number of sets in use.
Telephone mainlines are all telephone lines that connect a cus-
tomer's equipment to the public switched telephone network, per
1,000 people. Mobile telephones refer to users of portable tele-
phones subscribing to an automatic public mobile telephone ser-
vice using cellular technology that provides access to the public
switched telephone network, per 1,000 people. The ITU compiles
data on telephone mainlines and mobile phones through annual
questionnaires sent to telecommunications authorities and oper-
ating companies. The data are supplemented by annual reports
and statistical yearbooks of telecommunications ministries, reg-
ulators, operators, and industry associations.
Personal computers are the estimated number of self-contained
computers designed to be used by a single person, per 1,000 peo-
ple. Estimates by the ITU of the number of personal computers
 SELECTED
are derivedfrom an annual questionnaire,supplemented by other
sources. In many countries mainframe computers are used ex-
tensively,and thousandsof usersmaybe connectedto a singlemain-
framecomputer; in such casesthe number of personal computers
understates the total use of computers.
Internet hosts are computersconnected directlyto the world-
wide network; many computer users can access the Internet
through a single host. Hosts are assignedto countrieson the basis
of their country code, though this code does not necessarilyin-
dicate that the host is physicallylocated in that countrv. All hosts
lacking a country code identification are assignedto the United
States. The Internet Software Consortium changed the methods
used in its Internet domain surveybeginning in July 1998. The
new surveyis believed to be more reliableand to avoid the prob-
lem of undercounting that occurs when organizations restrict
download to their domain data. Nevertheless,some measure-
ment problems remain, so the number of Internet hosts shown
for each country should be consideredan approximation.
Scientists and engineers in research and development
(R&D) are the number of people trained to work in any field of
sciencewho areengagedin professionalresearchand development
activity(includingadministrators),per million people. Most such
jobs require completion of tertiary education.
UNESCO collectsdata on scientists,engineers,and R&D ex-
penditure from its member states, mainly from officialrepliesto
UNESCO questionnairesand specialsurveys,as well as from of-
ficial reports and publications, supplemented by information
from other national and international sources.UNESCO reports
either the stock of scientistsand engineersor the number of eco-
nomicallyactive persons qualifiedto be scientistsand engineers.
Stock data generallycome from censusesand are less timely than
measuresof the economicallyactive population. UNESCO sup-
plements these data with estimatesof the number of qualifiedsci-
entists and engineers by counting the number of people who
have completed education at ISCED (International Standard
Classificationof Education) levels6 and 7. The data on scientists
and engineers,normallycalculatedin termsof full-time-equivalent
staff, cannot take into account the considerable variations in the
quality of training and education.
High-technology exports are products with high R&D in-
tensity. They include high-technology products in aerospace,
computers,pharmaceuticals,scientificinstruments, and electrical
machinery.
The methodology used for determining a country's high-
technologyexportswas developedby the OECD in collaboration
with Eurostat (the StatisticalOffice of the European Communi-
ties).Termed the "productapproach" to distinguishit from a "sec-
toral approach," the method is based on the calculation of R&D
intensity (R&D expenditure divided by total sales)for groups of
products from six countries (Germany,Italy, Japan, the Nether-
lands, Sweden,and the United States).Becauseindustrial sectors
characterizedby a few high-technologyproductsmay alsoproduce
many low-technologyproducts, the product approach is more ap-
propriate for analyzinginternational trade than is the sectoralap-
proach. To construct a list of high-technology manufactured
products (servicesare excluded),the R&D intensitywascalculated
for products classifiedat the three-digit level of the Standard In-
WORLD DEVELOPMENT INDICATORS 329
ternational Trade Classification (SITC) revision3. The final list
wasdetermined at the four- and five-digitlevel.At this level,be-
cause no R&D data were available,final selectionwas based on
patent dataand expert opinion.This methodologytakesonlyR&D
intensity into account. Other characteristicsof high technology
arealso important, such as know-how,scientificand technicalper-
sonnel,and technologyembodiedin patents;consideringthesechar-
acteristicswould result in a different list (see Hatzichronoglou
1997).
Patent applications filed are the number of documents, is-
sued by a government office,that describean invention and cre-
ate a legalsituation in wvhichthe patented invention can normally
be exploited (made, used, sold, imported) only by, or with the
authorizationof, the patentee.The protection of inventionsislim-
ited in time (generally 20 years from the filing date of the ap-
plication for the grant of a patent). Information on patent
applications filed is shown separately for residents and nonresi-
dents of the country. Data on patents are from the World In-
tellectual Property Organization, which estimates that at the
end of 1996 about 3.8 millionpatents were in force in the world.
Table 20. Global trade
Merchandiseexportsshowthe fo.b. (freeon board) value,in U.S.
dollars,of goods provided to the rest of the world. Merchandise
imports show the c.i.f. (cost plus insurance and freight)value, in
U.S. dollars,of goods purchased from the rest of the world. Man-
ufacturedexportsand importsreferto commoditiesin Standard
InternationalTrade Classification(SITC) sections 5 (chemicals),
6 (basicmanufactures),7 (machinery),and 8 (miscellaneousman-
ufactured goods),excluding division68 (nonferrousmetals) and
group 891 (arms and ammunition). Commercialservices com-
prise all trade in services,including transportation, communica-
tion, and businessservices,excludinggovernmentservices,which
compriseservicesassociatedwith governmentsectors(such as ex-
penditureson embassiesand consulates)and with regionaland in-
ternational organizations.
Data on merchandiseexportsand imports arederivedfrom cus-
toms records and may not fully conform to the concepts and de-
finitions in the fifth edition of the IMF's Balanceof Payments
Manual. The value of exports is recorded as the cost of the goods
deliveredto the frontier of the exporting country for shipment-
the fo.b. value.Manv countries collectand report trade data in
U.S. dollars.When countries report in local currency, the value
is converted at the averageofficialexchangerate for the period.
The value of imports is generallyrecorded as the cost of the goods
when purchased by the importer plus the cost of transport and
insurance to the frontier of the importing country-the c.i.f.
value.Data on imports of goods arederivedfrom the same sources
as data on exports.In principle,world exports and imports should
be identical.Similarly,exportsfrom an economyshouldequal the
sum of imports by the rest of the world from that economy. But
differencesin timing and definition result in discrepanciesin re-
ported valuesat all levels.
The data in table 20 were compiled by the World Trade Or-
ganization (WTO). Data on merchandise trade come from the
IMF's InternationalFinancialStatistics Yearbook,supplemented
330 WORLD DtVEIOPMENT REPORT OOO,i200t
be data from the Cormmodity'rade (COMTRADE) databasc
maintained by the United Nations Statistics Divisionand fiom
national publications for countries that do not report to the
lNIM'. Data on trade in manu.facturcs coome from the
COMTRADE database. Where data were not available from
the WITD,World Bank staff estimited shares of matanuflctures
using the rtost re_ent intformationavailable front the
COMTlRADEdatabase.Whcreveravailablc,VFFA/ireportsmer-
chandise [rade dataon tlie hasis of thc generalsystem of trade.
whichincludesgoodsimportedfor reexport.T'woeconomicswsith
substantial reexports. Hong Koong(Chinal and Singapore. arc
noted in the table. (oods transported through a country en
route to another are not included. Data on trade in commercial
servicesarc drawn from the IMPF'sBalanceof Payvmerntrsdatabase,
supplementedby nationalpublicacionsfrom countriesthat do not
report to the MF.
Table21. Aid and financialflows
Net privatecapitalflowsconsistosfprivatedebt and nondebtflows.
Privatedebt flowsincludecommercialbarik lendingg. bonds,and
other privarccredits~privatenondeht flowsare foreigndirect in-
vestmeint and portfolio equity investnenrt. Foreign direct in-
vestment is nlet itftlows of investmeit to acquire a lasting
manrra..mentinterest (10 percent or more of votingstock) in an
enterpriseoptrating in at economyothierthati lhataofritc investor.
It isthe suImoffeILlitn
capitalfloWs,rtiusvcatmertofearrings, other
long-telrt capital flows,and short-term capitalflowsas showvitin
the balanceof pavyeirits.
T'hedat oi foreigndircct investmeritare based oil balanceof
paymentsdata reportedby the I1P. suppIcmentedby data on net
foreiin direct investmentreportedby the OECD and of'ficialna-
tional sources. he internationallyaccepteddefinitionof foreign
direcrinvestrmrntisthat providedin the fifth edition0oftheIMF's
Baslance ofPf'/Trmershtoudl, The EC1D)hasalso ptblished a de-
rinitioni,in conwiltaCtioti
with the IMF, Etirostat.andrileUnited
Natiort.. Becauseof the mtultiplicityof sourcesandtldifferencesin
definitionsand reportingmethods,more than one estiatet otffor-
eign direct insestment may existfor a countrv,and data m11aV not
be comparableacrosscountries.
Foreigndirect investmentdata do inotgivea corimplete picture
of international investmentin an econome.Balanceofpayrnents
data on1foreigndirect investmentdornot includecapital raisedin
the host economies,whiclhhasbecomean inmportantsource of fi-
isancingfor investmnenit projects in some developingcoutntries.
There is also increasisnawarenessthat foreigndirect investmern
data are limitedbecase they captire onilycross-borderinvestmern
flows involving cqiri partiipationand omit noneqttis-CroSS-
bordcr transactionssuch as itraflrmn flowsofgoods and services.
For a detailed disctissioiiof the data issues,see the World Bank's
I;rlrd Debtrflabes 1993-394(voltime 1. chapter 3).
Total externaldebt is debt owed to nonresidents repayable
in foreignlcurrency goods.or services.It is the scimiof ptblic. pub-
liclygiuaranteed,and prisvarenonguaranteedlong-tereidebt, rise
of NMbF credit,andshort-term debt. Short-term debt includesall
debt having an original maturity of one year or lessand interest
,rears Otl long-term debt. Presentvalue of externaldebt is
the sum of short-term externaldebt plusthe discounted sum of
total debt servicepayrnentsdue on public, publiclvguaranteed,
and privare nonguaranteedlong-term externaldebt over the life
of existingloans.
Data on theexterna clebtof low-and middle-incomeeconomies
aregatheredby the World Bankthrough its[ebtor ReportinigSys-
tem. World Bankstaff calculatethe indebtednessof developing
countriesusingloant-by-loanreportssubmitted by thesecountries
ott long-termpublicand publiclyguaranteed borrowing alongwith
information on short-term debt collected by the countries or
fromii creditors through the reportingsystemsofQtheBankfor In-
ternationl Sertlemenitsand the OECD. These data are supple-
mentedby informationon loansandcreditsfrom majormultilateral
baniksand loanstarementsfrom officiallendingagenciesin nmajor
creditor coulntrics,and by estimates from NWorldBankcountry
economistsaJidIMFdesk officers.In addition,somecountriespro-
vide data on private nonguaranteeddebt. In 1996, 34 countries
reportedtheirprivatenourguaranteed debt to theWorldBank;es-
timates evcre made for 28 additionalcountriesknownto havesig-
inificantprivatedebt.
The presentvalueof externaldebt providesa measureof fu-
ture debt serviceobligationsthat can be comparedwith-such in-
dicators as CNP. It is calculated bv discountingdebt service
huterest plusamortization)due on long-term externaldebt over
the life of existing loans. Short-term debt is includedat its face
valule.Data oti debt are in U.S. dollars converted at officialex-
charigerates.The discount rate applied to long-term debt is de-
tertuined bv the currency of repavmentof the loan and is based
on the OECD's commnercial interest referencerates. Loanisfrom
the InternationalBankfor Reconstructionand Developmentand
credits from the InternationalDevelopmentAssociationare dis-
counitedusing a referencerare for specialdrawing rights, as are
obligationsto the IMbF.Wihenthe discount rate is greaterthan the
interest rate of the loai, the present value is less than the nomi-
nal sLium of futurc debt serviceobligations.
Official development assistance (ODA) consists of dis-
btirsementsof loans (net of repaymentsof principal) and grants
madicon concessionalterms by officialagenciesof the members
of the DevelopmentAssistanceConmmittee(DAC), by multilat-
efal instittitions,and by certain Arab countriesto promote eco-
nornic developmnent and welfarein recipienteconomieslisted by
DAC as developing.Loanswith a grant element ofmore than 25
percenltare included in ODA, as are technicalcooperation and
assistance.Alsoincluded are aid flows(net of repayments)from
officialdonors to the transition economies of Eastern Eiarope
and the forisierSoviet Unioniand to certain higher-incomede-
velopingcountriesand territoriesas deternminedby DAC. These
flowsare sonetimes referred to as official aid aid are provided under
terms and conditions sinmilar to those for ODA. Data for aid itsa
shareofGNIParecalculatedusingvaluesin U.S. dollarsconverted
at officialexchangerates.
The data cover bilateral loans and grants fromnDAC coutn-
tries. multilateral organizations. and certaini Arab countries.
They do not reflectaid given by recipient countries to other de-
veloping counrries. As a result, some countries that are net
donors (such as Saudi Arabia) are shown in the table as aid
recipienits.
 SELECTED
The data do not distinguishamong differenttypesof aid (pro-
gram,project,or foodaid;emergencyassistance; peacekeepingas-
sistance;or technicalcooperation),each of which may havea
verydifferenteffecton the economy.Technicalcooperationex-
pendituresdo not alwaysdirectlvbenefitthe recipienteconomy,
to the extentthat they defraycostsincurredoutsidethe country
for salariesand benefitsof technicalexpertsand for overheadof
firmssupplyingtechnicalservices.
Because the aiddata intable21 arebasedon informationfrom
donors,theyarenot consistentwithinformationrecordedbyre-
cipientsin the balanceof payments,whichoftenexcludesallor
sometechnicalassistance-particularly paymentsto expatriates
madedirectlybythe donor.Similarly,grantcommodityaid may
notalwaysberecordedin tradedataor in the balanceof payments.
Althoughestimatesof ODAin balanceof paymentsstatisticsare
meantto excludepurelymilitaryaid,the distinctionis sometimes
blurred.The definitionused by the country of origin usually
prevails.
Statistical methods
Thissectiondescribesthe calculationof the least-squares
growth
rate,the exponential(endpoint)growthrate,the Giniindex,and
theWorldBank'sAdasmethodfor calculatingtheconversion fac-
tor usedto estimateGNP and GNP per capitain U.S.dollars.
Least-squares growth rate
Least-squaresgrowthratesareusedwhereverthereis a sufficiently
longtimeseriesto permita reliablecalculation.No growthrate
is calculatedif more than half the observationsin a periodare
missing.
The least-squaresgrowth rate, r, is estimated by fitting a lin-
ear regressiontrend line to the logarithmic annual values of the
variablein the relevant period. The regressionequation takes the
form
In X, = a + bt.
which is equivalentto the logarithmictransformationof the com-
pound growth equation,
X, = Xo (1 + r)t.
In this equation Xis the variable, t is time, and a = In XOand b =
In (1 + r ) are the parameters to be estimated. If by is the least-
squaresestimateof b, the averageannual growth rate,r, is obtained
as [exp(b ) - 1] and is multiplied by 100 to express it as a
percentage.
The calculated growth rate is an averagerate that is represen-
tative of the availableobservationsoverthe entire period. It does
not necessarilymatch the actual growth rate between any two
periods.
Exponential growth rate
The growth rate between two points in time for certain demo-
graphicdata, notablylabor force and population,iscalculatedfrom
the equation
WORLD DEVELOPMENT INDICATORS 33I
r = In (p,, 1p1 )!n,
wherep, andp1 are the lastand firstobservations in the period.
n is the numberof yearsin the period,and In is the naturallog-
arithm operator.This growthrateis basedon a modelof con-
tinuous,exponentialgrowthbetveentvo pointsin time.It does
not take into accountthe intermediatevaluesof the series.Nor
doesit correspondto theannualrateof changemeasuredat a one-
yearinterval,whichis givenby (pn- l )-PsIp,-
The Gini index
The Giniindexmeasuresthe extentto whichthe distributionof
income(or,in somecases,consumptionexpenditure) amongin-
dividualsor householdswithinan economydeviatesfroma per-
fectlyequal distribution.A Lorenzcurveplots the cumulative
percentages of totalincomereceivedagainstthe cumulativeper-
centageof recipients,startingwith the poorest individualor
household.The GiniindexmeasurestheareabetweentheLorenz
curveand a hypotheticallineof absolureequality,expressedas a
percentageof themaximumareaunderthe line.Thusa Giniindex
of zerorepresentsperfectequality,and an indexof 1100perfect
inequality.
TheWorldBankusesa numericalanalysis program,POVCAL.
to estimatevaluesof the Giniindex,seeChen, Datt,and Raval-
lion (1993).
World Bank Atlas method
In calculatingGNP and GNP per capitain U.S.dollarsfor cer-
taimoperationalpurposes,the World Banklusesa syntheticex-
changerate commonlycalledthe Atlasconversionfactor.The
purposeof the Atlasconversionfactoris to reducethe impactof
exchangerate fluctuations in the crrss-coiintrv comparisonofna-
tional incomes.
The Atlas conversion factor for any vear is the averageOf a
country's exchange rate (or alternative conversion factor) for
that year and its exchangerates for the two preceding years, ad-
justed for the difference between rhe rute of inflation in the
country and that in the Group of Five (G-5) countries (France,
Germany, Japan, the United Kingdom. and rhe United States).
A country's inflation rate is measured by the change in its GNP
deflator.
The inflation rate for the G-S countries, representing inter-
national inflation, is measured by the change in the SDR defla-
tor. (Special drawing rights, or SDRs. are the IMF's unit of
account.) The SDR deflatoris calculated.asa weightedaverageof
the G-5 countries' GDP deflators in SDR terms, the weights
being the amount of each countrv's currency in one SDR unit.
Weights vary over time because both the composition of the
SDR and the relativeexchangerates for each currencvchange.The
SDR deflator is calculatedin SDR terms first and then converted
to U.S. dollars using the SDR to dollar Atlas conversion factor.
The Atlas conversion factor is then applied to a country's GNP.
The resulting GNP in U.S. dollars is divided by the midyear
populationfor the latestof the three yearsto deriveGNP per capita.
When officialexchangerates are deemed to be unreliableor
unrepresentativeduring a period, an alternativeestimateof the ex-
change rate is used in the Atlasformula (see below).
332 WORLD DEVELOPMENT REPORT 2000/2OOI

The followingformulas describe the calculation of the Atlas t,p'S$is the SDR deflator in U.S. dollar terms for year t, Yt$ is the
conversion factor for year t: Atlas GNP in U.S. dollars in year t, Yris current GNP (local cur-
rency)for year t, and A' is the midyear population for year t.

e1e
3t
3j =
r
2
tPt
, ppss
_$t pss
1e
1
Pt / Pt
s j'
Alternative conversion factors
The World Bank systematicallyassessesthe appropriatenessof of-
ficialexchangerates as conversionfactors.An alternativeconver-
sionfactorisusedwhen theofficialexchangerate is judgedto diverge
and the calculation of GNP per capita in U.S. dollars for year t by an exceptionallylargemargin from the rate effectivelyapplied
to domestictransactionsof foreigncurrenciesand traded products.
g =
(Yr/Jt)IeY! This is the case for only a small number of countries(see the pri-
mary data documentation table in WorldDevelopmentIndicators
whereet isthe Atlasconversionfactor(nationalcurrencyto theU.S. 2000). Alternativeconversionfactors areused in the Atlasmethod
dollar)for yeart, eristhe averageannualexchangerate (nationalcur- and elsewherein the SelectedWorld Development Indicators as
rencyto the U.S. dollar) for year t, pt is the GNP deflator for year single-yearconversionfactors.

Ahmad,Sultan. 1992."RegressionEstimatesof Per CapitaGDP Basedon
Purchasing Power Parities." Policy ResearchWorking Paper 956.
World Bank, International EconomicsDepartment, Washington,
D.C.
- 1994. "ImprovingInter-Spatialand Inter-TemporalCompara-
bility of National Accounts."Journalof DevelopmentEconomics4:
53-75.
Ball,Nicole. 1984."MeasuringThird World SecurityExpenditure:A Re-
searchNote." WorldDevelopment12(2): 157-64.
Chen,Shaohua,GauravDatt,and Martin Ravallion.1993."IsPovertyIn-
creasingin the DevelopingWorld?"PolicyResearchWorking Paper.
World Bank,Washington,D.C.
Councilof Europe.Variousyears.RecentDemographic Developments
in Eu-
ropeand NorthAmerica.Strasbourg:Council of Europe Press.
Eurostat(Statistical
Officeof the EuropeanCommunities).1999.EU Trans-
port in Figures.Luxembourg.
-Variousyears.Demographic Statistics.Luxembourg.
-Variousyears.StatisticalYearbook.Luxembourg.
FAO (Foodand AgricultureOrganizationof the United Nations). 1999.
Stateof the World'sForests1999. Rome.
-Variousyears.ProductionYearbook.FAOStatisticsSeries.Rome.
Hatzichronoglou,Thomas.1997."Revisionof the High-TechnologySec-
tor and Product Classification." STI Working Paper 1997/2.
OrganisationforEconomicCo-operationand Development(OECD)
Directoratefor Science,Technology,and Industry. Paris.
Heston,Alan. 1994."ABriefReviewof SomeProblemsin UsingNational
AccountsData in Levelof Output Comparisonsand GrowthStudies."
Journalof DevelopmentEconomics 44: 29-52.
ICAO (InternationalCivilAviationOrganization).1999.CivilAviationSta-
tisticsoftbeWorld1998.ICAOStatisticalYearbook.23rded. Montreal.
IEA (International Energy Agency).Various years.EnergyBalancesof
OECD Countries.Paris.
-Variousyears.EnergyStatisticsandBalances ofNon-OECDCoun-
tries.Paris.
-Variousyears.EnergyStatisticsof OECDCountries.Paris.
IFC (InternationalFinanceCorporation).1999. Trendsin PrivateInvest-
ment in DevelopingCountries1999. Washington,D.C.
ILO (International Labour Organization). 1999. Key Indicatorsof the
LabourMarket Geneva.
-F Variousyears.Yearbookof LabourStatistics.Geneva.
IMF (InternationalMonetaryFund). 1986.A Manualon GovernmentFi-
nanceStatistics.Washington,D.C.
- 1993. BalanceofPaymentsManual.5th ed. Washington,D.C.
-Variousyears.Directionof TradeStatisticsYearbook.Washington,
D.C.
- . Variousyears.GovernmentFinanceStatisticsYearbook.Washing-
ton, D.C.
-Variousyears.InternationalFinancialStatisticsYearbook.
Wash-
ington, D.C.
InstitutionalInvestor.2000. March.New York.
Internet SoftwareConsortium.2000. InternetDomain Survey.January.
[www.isc.org].
IRF (InternationalRoad Federation).1999. WorldRoadStatistics1999.
Geneva.
ITU (InternationalTelecommunicationUnion). 1999. WorldTelecom-
municationDevelopmentReport1999. Geneva.
OECD (Organisationfor Economic Co-operationand Development).
Variousyears.Geographical DistributionofFinancialFlowsto Aid Re-
cipiens:Disbursements, Commitments,CountryIndicators.Paris.
Variousyears.NationalAccounts.Vol. 1,Main Aggregates. Paris.
Variousyears.NationalAccounts.Vol. 2, DetailedTables.Paris.
Data Sources
OECD (Organisationfor Economic Co-operationand Development),
DAC (DevelopmentAssistanceCommittee).Variousyears.Develop-
ment Co-operation. Paris.
PricewaterhouseCoopers. 1999a. CorporateTaxes:WorldwideSummaries
1999-2000. New York.
. 1999b.IndividualTaxes:Worldwide Summaries1999-2000.New
York.
PRSGroup. 2000.InternationalCountryRiskGuide.March.EastSyracuse,
N.Y.
Ravallion, Martin,and ShaohuaChen. 1996.'What CanNew SurveyData
Tell Usabout the RecentChangesin LivingStandardsin Developing
and TransitionalEconomies?'PolicyResearchWorking Paper1694.
World Bank,Washington,D.C.
- . 1997."Can High-InequalityDevelopingCountriesEscapeAb-
solutePoverty?"EconomicLetters56: 51-57.
Srinivasan,T.N. 1994."DatabaseforDevelopmentAnalysis: AnOverview."
JournalofDevelopment Economics44(1): 3-28.
Standard& Poor's.2000.Emerging StockMarkets Factbook 2000.NewYork
UN (United Nations). 1968.A Systemof NationalAccounts:Studiesand
Methods.SeriesF, no. 2, rev. 3. New York.
. 1999a. WorldPopulation The 1998Revision.NewYork.
Prospects:
. 1999b.WorldUrbanization Prospects:
The1998Revision.NewYork.
- Variousissues.MonthlyBulletinof Statistics.StatisticsDivision.
New York.
. Variousyears.EnergyStatisticsYearbook. StatisticsDivision.New
York.
- Variousyears.NationalIncomeAccounts. StatisticsDivision.New
York.
. Variousyears.Populationand VitalStatistics Report.StatisticsDi-
vision.New York.
. Variousyears.StatisticalYearbook.StatisticsDivision.NewYork.
. Variousyears.Updateon the NutritionSituation.Administrative
Committeeon Coordination,Subcommitteeon Nutrition. Geneva.
UNCTAD(UnitedNationsConferenceon Trade and Development).Var-
ious years.Handbookof InternationalTradeand Development Statis-
tics.Geneva.
UNESCO (United Nations Educational,Scientific,and Cultural Orga-
nization).1999a. StatisticalYearbook.Paris.
. 1999b. WorldEdurationReport.Paris.
UNICEF (UnitedNations Children's Fund). Variousyears.TheStateof
the WorldsChildren.New York:OxfordUniversityPress.
UNIDO (United NationsIndustrialDevelopmentOrganization).Various
years.InternationalYearbookofIndustrialStatistics.Vienna.
U.S. Departmentof State,Bureauof ArmsControl. 1999. WorldMili-
taryExpenditures andArms Transfers1998.Washington,D.C.
WHO (WorldHealth Organization). 1997. Coverage of MaternityCare.
Geneva.
. Variousyears. WorldHealthReport.Geneva.
- . Variousyears.WorldHealthStatistics Annual Geneva.
World Bank.1993a.PurchasingPower of Currencies:ComparingNational
IncomesUsingICPData.Washington,D.C.
. 1993b. WorldDebt Tables1993-94. Vol. 1. Washington,D.C.
. 2000a. GlobalDevelopmentFinance2000.Washington, D.C.
. 2000b. WorldDevelopment Indicators2000. Washington,D.C.
World IntellectualPropertyOrganization.1999.industrialPropertySta-
tistics.PublicationA. Geneva.
World ResourcesInstitute, in collaborationwith UNEP (United Nations
EnvironmentProgramme)and UNDP (United Nations Develop-
ment Programme).Various years. WorldResources: A Guide to the
GlobalEnvironment.NewYork Oxford UniversityPress.
WTO (WorldTrade Organization).Variousyears.AnnualReport.Geneva.
333
Classification of Economies
by Income and Region, 2000
Sub-Saharan
Africa Asia Europe
andCentral
Asia Middle
East
andNorth
Africa
Eastand Eastern
Income Southern West EastAsia South Europe
and Restof Middle North
group Subgroup Africa Africa andPacific Asia Central
Asia Europe East Africa Americas
Angola Benin Cambodia Afghanstan Armenia Yemen,
Rep. Haiti
Burundi BurkinaFaso |Indonesia Bangladesh Azerbaijan Nicaragua
Comoros Cameroon Korea,Dem. Bhutan Georgia
Congo, Dem. CentralAfrican Rep, India Kyrgyz
Rep. Repubic LaoPDR Nepal Republc
Eritrea Chad Mongolia Pakistan Moldova
Ethiopia Congo, Rep. Myanmar Tajkistan
Kenya COted'lvoire Solomon Turkmenistan
Low Lesotho Gambia, The Islands Ukraine
ncome Madagascar Ghana Vietnam Uzbekistan
Maoaw Guinea
Mozambique Guinea-Bissau
Rwanda Liberia
Somalia Mali
Sudan Mauritania
Tanzana Niger
Uganda Nigeria
Zamba SaoTome
Zmbabwe andPrincipe
Senegal
SierraLeone
Togo

Namibia
Swaziland
SIazilano
{
CapeVerde
Equateria
EqGuinea
China
FirioSri
FKiriati
Maldives

L
i Lanka
Albania
Belarns
Bosniaand
Turkey ran,Islamic
Repa
Iraq
Algeria
Djibouti
Egypt,Arab
Be ize
Bolivia
Colombia
Marshall Herzegovina Jordan Rep. CostaRica
Islands Bulgaria SyrianArab Morocco Cuba
Micronesa, Kazakhstan Republic Tunisia Dominican
Fed.Sts. Latvia WestBank Republic
PapuaNew Lithuania andGaza Ecuador
Lower Guinea Macedonia, ElSalvador
Philippines FYR' Guatemala
Samoa Romania Guyana
Thailand Russian Honduras
Tonga Federation Jamaica
Vanuatu Yugoslavia, Paraguay
Fed.Rep.' Peru
St. Vincentand
theGrenadines
Middle Suriname
income
Gabon American Croatia Isleof Man Bahrain Libya Antiguaand
Botswana Samoa Czech Lebanon Malta Barbuda
Mauritius Korea,Rep. Republic Oman Argentina
Mayortte Malaysia Estonia Saudi Barbados
Seychelles Palau Hungary Arabia Brazil
South Africa Poland Chile
Slovak Dominica
Republic Grenada
Upper Mexico
Panama
PuertoRico
St. Kittsand
Nevis
St. Lucia
Trinidad
andTobago
Uruguay
Venezuela, RB

|Subtotal125 2323 26 2 10 7 33

334
 SELECTED WORLD DEVELOPMENT INDICATORS 335

Sub-Saharan
Africa Asia Europeand CentralAsia MiddleEastandNorthAfrica

Eastand Eastern
Income Southern West EastAsia South Europeand Restof Middle North
group Subgroup Africa Africa andPacific Asia CentralAsia Europe East Africa Americas

Australia Austria Canada

Japan Belgium UnitedStates
NuwZealand Denmark
Finland
France'
Germany
Greece
Iceland
Ireland
OECD Italy
Lasembourg
Netherlands
Norway
Portugal
Spain
Sweden
Switzerland
United
High Kingdom
income Brunei Slovenia Andorra Israel Aruba
French Channel Kuwait Bahamas, The
Polynesia Islands Datar Bermuda
Guam Cyprus UnitedArab Cayman Islands
HongKong, Faeroe Emirates Netherlands
China' Islands Antilles
Macao, Greenland Virgin
Comna' Liechtenstein Islands(U.S.)
Non-DECO New Monaco
Caledona
N Mariana
Islands
Singapore
Taiwan,
China

Total 157 25 23 35 8 27 27 14 7 41

a.Former
YugoslavR
epublic
ofMacedenia
b FederaRepublic
ofYugoslsa c TheFrench
aSerbha/Msntenegro) overseas
departments
French
Guana,Guadeupe,
Martinique,
andHuniDn
arencluded
inFrance.
d.Ont Julyt997China
resumed
itsexercise
ofsovereigntv
overHong
Kong. e On20December
t999Chnaresumed
itsexerc
seofsovereignty
overMacao
SourceWorldBankdata

The World Bank classifieseconomies for operational and analy-
ticalpurposesprimarilyby GNP per capita.Everyeconomyisclas-
sifiedas low income,middle income(subdividedinto lower middle
and upper middle),or high income.Other analyticalgroups,based
on geographicregionsand levelsof externaldebt, are also used.
Low-incomeand middle-incomeeconomiesare sometimesre-
ferred to as developingeconomies.The term is used for conve-
nience;it does not implythat all such economiesare experiencing
similar development or that other economies have reacheda pre-
ferred or final stageof development.Nor does classificationby in-
come necessarilyreflectdevelopment status.
This table classifiesall World Bank member economies as
well as allother economieswith populationsof more than 30,000.
Economiesare divided among income groups according to 1999
GNP per capita,calculated using the World Bank Atlasmethod.
The groups are as follows: low income, $755 or less;lower mid-
dle income, $756-2,995; upper middle income, $2,996-9,265;
and high income, $9,266 or more.
DISTRIBUTORSAND
BOOKSELLERS
OF WORLD
BANK GROUP
PUBLICATIONS
Prices and credit termsvary
from country to country. Please
consultyour local distributor or
booksellerbeforeplacing an
order
DISTRIBUTORS
ARGENTINA
World PublicationsSA
Av. Cordoba1877
1120 Buenos Aires
Tel: (5411) 4815 8156
Fax: (54 11)481581t56
Email.wpbooks@irfovia.com.ar
AUSTRALIA, PAPUA NEW
GUINEA, FIJI, SOLOMON
ISLANDS, VANUATU, AND
SAMOA
D.A. Informaton Services
648 Whitehorse Road
Mitcham 3132
Victoria, Australia
Tel: (61 3) 9210 7777
Fax (61 3) 9210 7788
Ema1:service@dadirectrcomau
URL. www dadirect.com.au
AUSTRIA
Gerold andCo.
Weihburggasse26
A-1010 Wien
Tel' (43 1)51247310
Fax (43 1) 512 473129
Email. buch@geroldtelecomat
BANGLADESH
Micro IndustriesDevelopment
AssistanceSociety (MIDAS)
House5, Road 16
Dhanmondi R/Area
Dhaka 1209
Tel: (880 2) 326427
Fax: (8802) 8111188
Email:midas@fsbdnet
BELGIUM
Jean de Lannoy
Av du Roi 202
1060 Brussels
Te. (32 2) 538 5169
Fax I32 215380841
Emainljean.de larnoy@
infoboard be
URL www.lean-de-lannoy.be
BOSNIAAND HERZEGOVINA Tel:(4202)24462232,24462254 D-53113Bonn
Book Trading Company
Sahinpasic
MarsalaTita 29/Il
71000 Sarajevo
Tel (387 33) 21 05 20
Fax (38733) 66 88 56
Email:talib@btcsahinpasic.com Havelkova22
URL wvw.btcsahipasic com
BRAZIL
PublicacoesTecnicas nterna-
cioraisLtda
RuaPeixoto Gomide, 209
BelaVista
01409-901Sac Paso. SP
Tel: (55 11) 259 6644
Fax 15511)2586990
Emal. webmaster@pt com.br
URL: www.pti.com.br
CANADA ECUADOR Emai pap4gioa forthnet gr JAMAICA MOROCCO
Renouf PublishingCo Ltd Libr Mundi- Lbreria nternacional URL www papasotiriougr Ian Randle PublisnersLtd LibrarieInternatonale
5369 Canotek Road Juan Leon Mera 851 206 Old Hope Road 70, RueT ssoule
Ottawa, Ontario KIJ 9J3 PG Box 17-01-3029 HAITI Kingston6 P0 Box 302
Tel:1613)745-2665 Quito Culture Dffusion Tel (876)927 2085 Rabat(SoUiSSi)MA 10001
Fax (613)745-7660 Tel: (5932)521606 Mr. Yves C rmentJumele Fax: (8761977 0243 Tel (2127)7501 83
Email. orderdept@ Fax: 159321504 209 5, Rue Capons Email irpl@cols com Fax (2127) 75 86 61
renoufbooks.com Ernaillibrmu@llbrimundicomec C.P.257
URL. www renoufbookscom Port-au-Prince JAPAN NEPAL
CODEU Te (509)23 9260 Eastern Book Service(EBS) Everest Media nternat onal
CHINA Ruizde Castilla 763, Edif. Fax. (509)23 4858 3-13Hongo3-chome,Bunkyo-ku Services(P) Ltd
ChineseCorporaton for Expocoor Tokyo 113 GPOBox 5443
Promoton and Human ies Primer p so Of #2 HONGKONG,CHINA;MACAU Tea (81 3) 3818 0861 Kathrnandu
15, Ding Hul Dong Li, Quito Asia 2000 Ltd. Fax (81 3) 3818 0864 Tel: j9771) 416 026
Kun Lan Hotal Te: (5932) 507-383 Sales& C rculationDepartment Email orders@svt-ebsco.jp Fax: (977 1) 250 176
HaidianDistrict 100036 Fax: (5932)507-383 302 SeabirdHouse URL. www.svt-ebs colp Email: emisptd@w ink com.np
Beij ng Email codeu@impsatnet ec 22-28 WyndhamStreet,Central
Tel (86 10) 88117711 Hong Kong, China KENYA NETHERLANDS
Fax: (86 10) 88129871 EGYPT,ARAB REPUBLICOF Tel 1852)2530 1409 LegacyBooks De Lindeboom/lnternationale
Email:wangleng99@vahoo.oonn Al Anram Distribution Agency Fax (852)2526 1107 LoetaHouse Publikatiesb v
Al GalaaStreet Email sales@asia2000 com.hk PO Bax 68077 MA de Ruyterstraat20A
China Book Import Centre Cairo URL www.asia2000 com.hk Nairob 7482 BZ Haaksbergen
P.O. Box 2825 Tel: 1202) 578 60 83 Te (254 2, 330 853 Tel: (31 531574 0004
Beijing Fax: (20 2)5759388 HUNGARY Fax (2542)330854 Fax (31 53) 5729296
Euro Info Service Emai legacy@form-netcom Email booksOdelndeboorncom
Ch na Financa & Economic MERIC(Midd e EastReaders Hungexpo EuropaHaz (Pf. 441 URL www,delindeboom com
Pubishing House information Center) H-1441 Budapest Afr ca Book Servce lE.A.) Ltd
8, Da FoSi DongJ e 2 Bahrat Alv St Tel. (361()2648270; 264 8271 Quaran House,Mfangano NEW ZEALAND
Beijing Building 'D" 1st Floor,Apt.24 Fax (36 1) 264 8275 Street EBSCONZ Ltd
Tel:(86 10) 6401 7365 Cairo Email euroinfo@euroinfo.hu P.O Box 45245 Private Mail Bag 99914
Fax. (86 10) 6401 7365 Tel: (20 2) 341 3824 JRL. wwweuronfo.hu Nairobi New Market
Fax:(20 2) :3419355 Tel: 254 2) 223 641 Auckland
COLOMBIA Email:order@meric-cocom INDIA Fax (2542) 330 272 Ten (64 9) 524 8119
nfoenlace Ltda /An IHSGroup URL.www meabservercom eg Allied PublishersLtd. Fax (64 9) 524 8067
Company 751 Mount Road KOREA,REPUBLICOF Ema1:WGent%ess-nz.ebsco@
Calle 72 No. 13-23 - Piso 3 For subscriptionorders and Madras600 002 Euyoo Publishing Co, Ltd. iss ebsco,com
Editicio NuevaGranada publications in French only. Tel: (91 44) 852 3938 46-1, Susong-Dong
Santatb de Bogota, D.C. M ddle EastObserver Fax (91 44) 852 0649 Jongro-Gu Oasis Official
Te (57 1)2609474,2609480 41, Sherif Street Email: allied mds@smb Seoa P.O. Box 3627
Fax (57 1)2480808 11111 Cairo spr ntpg.ems vsn net in Tel (82 2) 7343515 We lington
Email: nfoenlace@andinet.com Tel: (20213926919 Fax (822)7329154 Tel 1644)4991551
Fax.
N20 2) 393 9732 INDONESIA Ema .eulyoo@chal an net Fax 1644) 4991972
COTE DIVOIRE Emal: Pt ndra Limited Emai oasis@actixgennz
Centre dEditon etde Diffu- inquiry@meobserver,com Jalan Borobudur20 DayangBooks Trading Co. URL wwwoasisbooks co.nz
sion Africaines (CEDA) URL: wwvv.mebsenrvercom PO Box 181 InternationalDivision
04 B.P541 Jakarta 10320 954-22, Bangbae-Dong, NIGERIA
Abidjan04 FINLAND Te 1622113904290 Socho-ku University Press Pic
Tel: (225)246510 Akateeminen Kirakauppa Fax (62 21) 3904289 Seanl Three Crowns Buiidinguercho
Fax: 4225)25 0567 PL 128 (Keskuskatu14 Tel:(82 21582 3588 Private Mad Bag 5095
Email: info@ceda-cicom FIN-00101Helsinke PF Book Fax 4822) 521 8827 sbadan
URL. wwwu.ceda-ci.com Tel (35891121 4385 JI dr.SetiaBudhi No 274 Email: dico3@chollan.net Te. (23422)411356
Fax (35891 1214450 Bandung40143 Fax (234 22) 412056
CYPRUS Emaeakatlaus@akateeminen.com Tel: (62 22) 2011 149 LEBANON Em'a unipress@skannet com
Center for Appiied Research URL. wwv .akateeminencom Fax.(62 22) 2012 840 Librairiedu Leban
6, DiogenesStreet, Engomi Email:pfbao@kbandung P.O. Bax 11-9232 PAKISTAN
P.O Box2006 FRANCE wasantara net id Be rut OxfordUniversity Press
Nicosia EdbonriscEs ka DJB/Offil b Te (9619)217944 5 BangaloreTown, Sharae
Tel 13572) 59 0730 12, rue duQuatre-Septembre IRAN Faxn(9619)217434 Fasal
Fax (357 2166 2051 75002 Paris KetabSara Co Publishers Email hsageh@cyberianet.lb PG Box 13033
Einail.nt tzim@sting. Tel: (33 1) 42 86 58 88 P O. Box 15745-733 URL www.hibraire-du- Karachi75350
cycollegeac cy Fax: (33 1) 42 60 45 35 Tehran 15117 liban com.lb Tel (92 21) 446307, 449032;
Erial: offi b@offilibfr Te (98 21) 871 6104 440532
CZECH REPUBLIC URL: wwwoffilibfr Fax (98 21) 871 2479 MALAYSIA Fax: (92 21) 4547640,449032
ManagementPress,NT Emael ketab-sara@neda.net.ir University of Malaya Coopera- ESma l: ouppak@theoffice.net
Publishing,s.r o GERMANY five Bookshop,Limited URL: wwvv oup.com pk
Nam.W. Churchi a 2 UNO-VERL.AG KowkabPublishers P.O.Box 1127,JalanPantaiBaru
130 59 Prague3 Am Hofgarten 10 P.O BOX 19575-511 59700 KualaLumpur PakBook Corporation
Tehran Tel: (603)7565000 Azz Chambers21
Fax (4202) 2446 2242 Tel: (49 228) 949 020 Tel: (98 21) 258 3723 Fax: (60 3) 755 4424 Queen's Road
Emai mgmtpress@ Fax (49 228)949 0222 Fax: (98 21) 258 3723 Email. umkoop@rm.net.my Lahore
mgmtpresscz Emai unoverlag@aol.com Email.kowkabpub@tavana.net Tel (92 42) 636 3222,636 0885
URL www mgmtpress.cz URL www.uno-verlagde MEXICO Fax 19242i 636 2328
ISRAEL INFOTEC Email pbc@brainset.pk
USIS,NIS Prodeona GHANA Yozmot Literature Ltd Av. San FernandoNo 37
Epp Books Services P.O.Box 56055 Col Toriello Guerra M,rzaBook Agency
130 00 Prague3 Post Office Box 44 3 YohananHasandlarSt. 14050 Mexico D.F 65, Shahrah-e-Quaid-e-Azam
Tel: (42 2)2423 1486 TUC Tel Aviv 61560 Tel 4525) 624 2800 Lahore 54000
Fax:1422i 2423 1114 Accra Ten 1972345285397 Fax 1525) 624 2822 Tel 19242) 7353601
Email posta@usiscrcz Te: (23321) 778 843 Fax (97234 5285 397 Email infotec@rtn.netmx Fax. (92 42) 576 3714
URL:mww usscr.cz Fax (233211779099 URL www.rtn net mx Emal) merchant@brainnet.pk
Email eppaafrrcaorisne.com.gh ITALY
DENMARK L.cosaLibreriaCommissionaria Munds-PrensaMexico, S.A PERU
Samfundsitteratur GREECE Sansoni SPA. de CV. Editoral Desarrollo SA
SolblergPlads3 Papasotirou SA.. Via Ducadi Ca abr a t/1 c/Rio Panuco,141- Colon a Apartado3824
DK-2000Frederiksberg Internationa Technical 50125 Firenze Cuauhtemoc Ica 242. OF 106
Tel (45 381 153870 Bookstore Tel: (39 55) 648 31 06500 Mexico DF Lma 1
Fax (45 381153856 35, StournaraStr Fax 13955) 641 257 Te (52 5) 533 56 58 Tel 451 14)285 380
Email ck@sl.cbsdk 10682 Athens Email. hicosa@ftbcc.r Fax (52 5) 514 67 99 Fax 15114)286 628
URL. www.sl cbs dk Tel: (301)3641826 URL: www ftoc.i/licosa Ema 10152452361@
Fax 1301) 364 8254 compuserve com
PHILIPPINES
Internationa Booksouice
Center,Inc,
1127-A Antipolo St.
Barangay,Venezuela
Makati City
Te: (63218966601
Fax:(63 21 8966497
POLAND
InternationalPublishingService
Ul Piekna31t37
00677 Warsaw
Tel 1482r 628 6089
Fax (482) 8217255
Cmailbooks%.pskp.atrr.om.pl
JRL www ips com pl
PORTUGAL
Livraria Portugal
Apartado2661
RuaDo Carmo70-74
1200 Lisbon
Tea (351 1)3474982
Fax (351 1 3470264
ROMANIA
CompanDe LbrariiBucuresosa
Str Lipscaninr.26, sector 3
Bucharest
Tel (401)3139645
Fax (401)3124000
RUSSIAN FEDERATION
lzdatelstvo << Ves Mir >
Moscow 101831
Tel 1795) 917 8749
FdX (79519179259
Email ozimarin@glasnet ru
UHL:wwvw.vesmir.tsxorg
SENEGAL
LibrairieClairatriq,e
2, RueEl Hadl Mbaye Gueye
Placede l'Independance
B P. 2005
Dakar
Tel (221)822 21 69
uax(221t 821 84 09
SINGAPORE;TAIWAN, CHINA, Cotes-de-Montbenon30
MYANMAR; BRUNEI
HemispherePublishingServices Tel 14121 34' 3229
GoldenWheel Building
41 KallaugPuddingRoad,#04-03 Email institutionnel@
Sirgapore 349316
Tel: (65) 741 5166
Fax.(65) 742 9356
Email:info@hem sphere.com.sg Techniques
SLOVAKREPUBLIC
Slovart G.T.G.Ltd
Krupinskl 4
P0. Box 152
852 99 Bratislava5
Tel. (42 7) 839 471, 472: 473
Fax (427)18 485
Email:gtg@internet.sk
SLOVENIA
Gospoduiskivestiik Publish-
ingGroup
Dunajskacesta 5
1000 Llubljana
Te: (386 61)133 83 47
Fax:138661113380 30
Email repansekg8gvesxnik si ATTN Centra Books Distribu- Academicand BaobabBocks
UHL www gvestniksx
EUROPANindex htm
SOUTHAFRICA,BOTSWANA Kiong-Teoy
Foesingle titles
Oxford UniversityPress
SouthernAfr ca
P.O Box 12119
NI C0ty7463
CaneTown, South Africa
Te (27 21) 595 4400
Fax (27 21(65954430
Email oxford@oupco.za
For subsciption orders
Inteinanoriai Suoscription
Service
P.O. Box41095
Craighall
Johanneslbw42024,South
Africa
Tel: (27 11)880 1448
Fax: (27 11)880 6248
Emal: iss@is.co,za
SPAIN
Murdi-Piensa Lbros, s a.
Castello37
28001 Madrid
Tel (34 911436 37 00
Fea (34 9115753998
Ema1.libreiagmrmundiprensa
es
URL www.rmundiprensaes
Mundi-PrensaBarcelona
Consellde Cent No 391
08009 Barcelona
Tel 134314883492
Fax (34 3i 487 7659
Emailbarcelna@nmundiprensa.es P.O.Box 3, Omega Park
SRI LANKA,THE MALDIVES
LakeHouse Bookshop
PO Boo 244
100, Sir ChittampalamGar-
diner Mawatra
Ccolmbo 2, Sri Lanka
Te 1941t 32 104
Fax (941) 432104
Ema, LHL@nsr lankanet
SWEDEN
For periodicalsandserialsonly: Email:chris.allen@theso.cn.uk Tel: (962 6) 465 6123
Wennergren-Wiiliams
InformationsServiceAB
PC Box 1305
S-17125 Solna
Tel i46817059750
Fau (46 8) 27 0071
Email mel(Bmwvse
SWITZERLAND
LibreriePayotS A
Serv,Le rlit turionne
1002 Lausanne
Fae (41 2113413235
payct-bbrairech
ADECOVarnDermen Edibons 246 Le ThanhTon Street
Ch. de Lacuez41
CH-1807Bionay
Tel: 41 21) 943 2613
Fax:(41 21) 943 3605
TANZANIA
TEPUSA
The Network of Technical
Pubelcationsin Africa
P 0 Boe 22638
Dar es Sa 0am
Tel 125551) 114876
Fax (25551)112434
Email tepusacgintaftica.com
THAILAND
CentracInternationalLtd.
tlion Co , Ltd
Sinnrat Bldg 13th Floor
3388/42-06Rams 4 Rd.
BangkokIo1 0 0
Tel: (66 21367-5030-41X178
Fax (66 213675049
TRINIDAD & TOBAGOAND
THE CARIBBEAN
SystematicsStudies Ltd.
St Aagusnn ShoppingCenter
EasternMain Road
St Augustine BOOKSELLERS Fax: 48 2216822233-6321724
Tel: (868)645 8466 Email:abe@ikp.atm.com.pl
Fax: 1868)645 8467 CHINA
Email: tobeAtrindad.net China NationalPublications TAIWAN, CHINA
Imoort Tycoon Information,Inc.
UGANDA & ExportCorporaton Ms. Eileen Chen
GustronLimted 160ongti EastRoad 5 Floor, No. 500
P.O. Box 9997 PostCode 100020 Chang-ChunRoad
MadhvaniBuilding Beijing Taipei 105, Taiwan
Plot 16/4, Jinja Road Tel: !866 2) 8712 8886
Kampala HUNGARY Fax:(886 2) 8712 4747;8712
Tel: (25641) 251467 Foundationfor Market Econ- 4777
Fax: 125641)251468 omy Email:eiutpe@ms21.hinet.net
Email:gus@swiftuganda.com 112 Pf 249
1519 Budapest THAILAND
UKRAINE Tel: 1361) 2042951; 2948 Chulalongkomn UniversityBook
LIBRA PublishingHouse Fax.(361)2042953 Center
Ms. SophiaGhemborovskaya Email: ipargazd@hungaiy.net PhyathaiRoad
53/80SaksahanskohoStr. Bangkok10330
252033, Kiev 33 INDIA Tel: (66 2) 235 5400
Tel: 7 44) 227 62 77 Bookwell Fax:(66 2)255 4441
Fax P744) 227 62 77 HeadOffice. 2/72, Nirankari
Colony Book Link Co. Ltd,
UNITEDKINGDOM Delhi- 110009 118/4EkamaiSoi28
Micro nfo Ltd. Tel.(91 11) 7251283 Vadhana
SalesO'fice: 24/4800,Ansani Bangkok10110
Alton Road,DaryaGanj Tel: 662)711 4392
HampshireGU34 2 PG New Delhi-110002 Fax:(662)711 4103
Tel: (441420)86 848 Tel: (91 11) 326 8786,325 Email:bbatpt@glau.ac.th
Fax: 44 1420) 89 889 7264
Email: vbank@microinfo.co.uk Fax: (9I 11)3281315 TURKEY
LRL: wwwmicroinfoco.uk Email: bkwell@nde.vsnl.net.in Donya lnfotelA.S
"Globus' DonyaBasinevi
The Slationery Office JORDAN 100, Yil Mahallesi
51 Nine Elms Lane GlobalDevelopmentForum 34440 Bagcilar-lstanbul
LondonSW85DO (GDF) Tel: (0 212)6290808
Tel. (44 20)7 873 8372 P.O.Box 941488 Fax:(90 212)629 46 89; 629
Fax:(4420)78738242 Amman 11194 4627
Enar:dunya@dunya-azete.Domtr
URL www.the-statonery- Fax: (9626) 465 6123 URL: http. www.duiya.com
office.co.ukai/ Email:gdf@index.com.jo
UNITED ARABEMIRATES
VENEZUELA,R.B.de KOREA,REPUBLICOF Al RawdhaBookshcp
Tecni-CienciaLibros,S.A. Seiong Books,Inc. P.O.Box 5027
Sr. LuisFernandoRamirez, 81-4Neung-dong Shar)ah
Director Kwangjin-ku Tel: 19716) 734 687
CentroCuidadComercial Seoul 143-180 Fax: (9716) 384 473
Tamanaco Tel (822)4980300 Email:alrawdha@hctmail.com
Nivel C-2 Fax: (82 2) 3409 0321
Caracas Email: danielchoi@sejong URUGUAY
Tel. (58 2) 959 5547 bookscom LibreriaTe,nca Uruguaya
Fax. (58 2) 9595636 UFL: htbp:www.sejong Colonria1543, Piso7, Of. 702
Email: IfrgOOl@ibm.net bookscom Casila deCorreo 1518
Montevideo11000, Uruguay
VIETNAM MALAYSIA Tel: 159521490072
FAHASA(TheBook Distribu- MDC PublishersPrintersSDN Fax: (59821413448
tion Co of HochiminhCity) BHD Email:ItuOcs comry
MDC Building
District 1 2718, Jalan Permata Empat
HochiminhCity Taman Permara,Ulu Kelang
Tel: (84 8) 829 7638, 822 5446 53300 KualaLumpur
Fax:(848) 822 5795 Te: (60 3) 408 6600
Email:fahasa-sg@hcm.vnn.vn Fax: (60 3) 408 1506
URL:www tinet.com.vn/fahasa Email. mdcpp@2mwsmcom.my
UFL wvw2mws.com.my/mdc
ZAMBIA
UniversityBookshop.Univer- NEPAL
sity of ZamDia BazaarInternational
GreatEPstRoadCampus 228 SanchayaKosh Building
P.O.Box 32379 GFOBcx 2480, TddeviMarg
Lusaka Kathmaridu
Tel: (260 1)252576 Te: (9771)255125
Fax:(2601 253952 Fax: (977 1) 229437
Email:hunene@admin.unza.zm Email. bazaar@mos.com.np
ZIMBABWE NIGERIA
Mosuro Booksellers
(Pvt.)Ltd 5 OluwareObasa Street(Near
4 ConaldRoad Awolowo Ave.)
Graniteside P.O. Box 30201
P.O. Box 567 Ibadan
Harare Te: (2342) 810-2560
Tel: (2634) 755 035 Fax: (2342) 810-2042
Fax:(263 4) 755052 Email:KmosuroMinkservaom.ng
Email:Academic@Africaon-
line.Co.Zw POLAND
A.B E. Marketing
Ul. Grzybowska37A
00-855 Warsaw
Te: (48 22) 654 06 75
 A t the start of a new century, poverty r emains a global prohlem ofl'huge
proportions. 01' the world's 6 billion people, 2.8 billion live on less than $2 a
day and 1.2billion on less than Si a day. Eight ofevery 100 inf'ants do not live
to see their fifth birthday. Nine oflevery 100 boys and 1 oftevery 100 girls
w.vhoreach school age do not attend school. Poverty is also evident in poorl
people's lack of political power and voice and in their extreme vulnerability to
ill health, economic dislocation, personal violence, and natural disasters. And
the scourge of' HI/AIDS, the frequency and brutality of civil conflicts, and
rising disparities between rich countries and the developing wsorld have
increased the sense of deprivation and injustice for many.

1II,/), ,A'. ' b u j (wrhichfollows tlo

other W\7orld l)evelopment Reports on poverty, in 1980 and 1990) argues
nevertheless that major r eductions in all these dimensions of'poverty ar-e indeed
possible-that the interaction of'markets, state institutions, and civil society can
harness the forces of' economic integration and technological change to ser ve the
interests of poor people and increase their share of' society's prosperity.

Actions are needed in three complementary areas: promoting economic

opportunities for pool' people through equitable growth, better access to
markets, and expanded assets; facilitating empowerment by making state
institutions more responsive to poor people and removing social barriers
that exclude women, ethnic and racial groups, and the socially
disadvantaged; and enhancing security by preventing and managing
economyvvide shocks and providing mechanisms to i'educe the sources of'
vulnerability that poor people face. But actions by countries and
communities vill not be enough. Global actions need to complement
national and local initiatives to achieve maximum benefit l'or poor people
throughout the wvorld.

The World Bank

1818 H Street, N.W.
Washington, D.C. 20433, U.S.A.

Telephone: 202 477 1234

Facsimile: 202 477 6391
Telex: MCI 64145 WORLDBANK
MCI 248423 WORLDBANK
Internet: www.worldbank.org
'19W [I E-mail: books@worldbank.org

Oxford University Press

0-19-521129-4