1999 ANNUAL REVIEW OF DEVELOPMENT EFFECTIVENESS · William Battaile, Ruchira Corcoran, Deepa...

W O R L D B A N K O P E R A T I O N S E V A L U A T I O N D E P A R T M E N T

1999 ANNUALREVIEW OFDEVELOPMENTEFFECTIVENESS

Nagy Hannawith

Ramgopal AgarwalaWilliam Battaile

Deepa ChakrapaniRuchira Corcoran

Federico Mini

1999

The World Bank

http://www.worldbank.org/html/oed Washington, D.C.

Copyright © 2000

The International Bank for Reconstruction

and Development/THE WORLD BANK

1818 H Street, N.W.

Washington, D.C. 20433, U.S.A.

All rights reserved

Manufactured in the United States of America

First edition November 1999

The opinions expressed in this report do not necessarily represent the views of the World Bank or itsmember governments. The World Bank does not guarantee the accuracy of the data included in thispublication and accepts no responsibility whatsoever for any consequence of their use. The boundaries,colors, denominations, and other information shown on any map in this volume do not imply on thepart of the World Bank Group any judgment on the legal status of any territory or the endorsement oracceptance of such boundaries.

The material in this publication is copyrighted. The World Bank encourages dissemination of its workand will normally grant permission promptly. Permission to photocopy items for internal or personaluse, for the internal or personal use of specific clients, or for educational classroom use is granted by theWorld Bank, provided that the appropriate fee is paid directly to the Copyright Clearance Center, Inc.,222 Rosewood Drive, Danvers, MA 01923, U.S.A., telephone 978-750-8400, fax 978-750-4470.Please contact the Copyright Clearance Center before photocopying items. For permission to reprintindividual articles or chapters, please fax your request with complete information to the RepublicationDepartment, Copyright Clearance Center, fax 978-750-4470.

All other queries on rights and licenses should be addressed to the Office of the Publisher, World Bank,at the address above, or faxed to 202-522-2422.

Photographer: Still Pictures (Father carrying daughter - Nepal)

ISSN 1019-4363

ISBN 0-8213-4706-3

Printed on recycled paper.

iii

vii Acknowledgmentsix Foreword, Prefacio, Préfacexi Executive Summary, Resumen, Résumé Analytiquexxiii Abbreviations and Acronyms

1 1. Toward a Comprehensive Development Strategy2 Long-Term, Holistic, Balanced3 Ownership3 Partnership and Development Cooperation3 Results, Learning, Accountability4 Challenges and Tensions

5 2. Challenges at the Project Level5 Keeping up with Complexity6 Varied Performance among Regions, Sectors, and Types of Lending7 Institutional Development and Sustainability Flagging8 Bank and Borrower Performance Must Improve9 Project Performance and CDF Principles

11 3. Challenges in Sectoral and Thematic Programs11 Short versus Long Term12 Comprehensiveness versus Selectivity13 Speed versus Broad-Based Ownership13 Partnership versus Country Capacity and Transaction Costs14 Accountability for Results versus Local Capacity

17 4. Challenges at the Country Level17 Addressing Social Issues18 Addressing Institutional Capacity and Structural Issues19 A Long Way to Go20 Managing the Tensions in the CDF21 Country Cases

23 5. Promising Approaches23 Learning Process, not Blueprints24 Managing Complexity by Sequencing24 Comprehensive Analysis and Selective Actions25 Sectorwide Approaches25 Adaptable Conditionality26 Time to Build Consensus26 Broadening Participation27 Information for Accountability and Learning27 Capacity Building to Manage for Results28 From Aid Coordination to Development Partnership28 Linking Global and Country Strategies

31 6. Implications for Development Effectiveness32 Implications for the Bank

C o n t e n t s

1 9 9 9 A n n u a l R e v i e w o f D e v e l o p m e n t E f f e c t i v e n e s s

iv

Annexes35 Annex 1: Project Performance Trends and Outstanding Projects47 Annex 2: Challenges in Implementing the CDF Principles at the Sector Level49 Annex 3: Development Performance in Selected Countries55 Annex 4: Emphasizing People-Centered Development in a Holistic Framework57 Annex 5: Ownership, Participation, and Conditionality61 Annex 6: Managing for Results65 Annex 7: Partnership and Strategic Selectivity71 Annex 8: International Public Goods and Aid Effectiveness75 Annex 9: Results-Based Management Glossary77 Annex 10: Managing Development Effectiveness:

An Overview from the CODE Chairperson

79 Endnotes

81 Bibliography81 Background Papers81 Selected Bibliography

Figures4 1.1 Interaction of CDF Principles6 2.1 Demandingness, Complexity, Riskiness, and Outcome, by Exit Fiscal Year

18 4.1 The Bank’s Contribution to Implementing CDF Principles36 A1.1 Aggregate Project Performance Index36 A1.2 Satisfactory Project Outcomes37 A1.3 Outcome, Sustainability, and ID Impact, by Region and Exit Fiscal Year Group37 A1.4 Satisfactory Operations by Sector and Exit Fiscal Year Group38 A1.5 ID Impact38 A1.6 Sustainability39 A1.7 Borrower Performance39 A1.8 Bank Performance47 A2.1 Sector Scores on CDF48 A2.2 Sector Scores on Individual CDF Principles49 A3.1 Disappointing Performance in Reducing Poverty…49 A3.2 … And in Promoting Economic Growth55 A4.1 Sustainable Livelihood Framework

Boxes8 2.1 Updated Determinants Analysis Puts Spotlight on Bank Performance

24 5.1 Challenges and Promising Approaches58 A5.1 Conditionality Scenarios61 A6.1 Barriers to Performance Orientation62 A6.2 Pooling and Joint Reviews in Ghana’s Health Sector Reinforce Accountability63 A6.3 Malaysia: Sustainable Penang Initiative67 A7.1 Learning from Business Alliances68 A7.2 Capacity Building Assistance to Kazakhstan and Vietnam for Aid Coordination69 A7.3 Advantages and Disadvantages of In-Country Aid Coordination Meetings71 A8.1 Public Goods Primer72 A8.2 Progress in Controlling River Blindness

v

Tables2 1.1 Three Development Paradigms

41 A1.1 Outstanding and Poor Performers among Recently Evaluated Projects42 A1.2 Outcome, Sustainability, ID Impact, and Aggregate by Sector, Network, Lending

Type/Source, Region, and WDI Income Group for Exit Fiscal Years 1990–93,1994–97, and 1998–99 (by projects)

44 A1.3 Outcome, Sustainability, ID Impact, and Aggregate by Sector, Network, LendingType/Source, Region, and WDI Income Group for Exit Fiscal Years 1990–93,1994–97, and 1998–99 (by real disbursements, FY96US$)

50 A3.1 Relevance of Sample of Evaluated Countries51 A3.2 Country Performance: Per Capita Growth, Poverty, and Life Expectancy52 A3.3 Country Performance: Savings53 A3.4 Country Performance: Investment Efficiency65 A7.1 Country Characteristics and Aid Coordination66 A7.2 Donor Characteristics and the Aid Coordination Environment

C o n t e n t s

vii

ACKNOWLEDGMENTS

his report was prepared by a team led byNagy Hanna, including Ramgopal Agarwala,William Battaile, Ruchira Corcoran, Deepa

Shafik, William E. Stevenson, and Michael Walton.Pablo Guerrero also provided valuable comments.

The report benefited from comments of the DGOand the OED management team: Robert Picciotto,Wendy Jarvie, Elizabeth McAllister, Gregory Ingram,Ruben Lamdany, and Osvaldo Feinstein. Bruce Ross-Larson and Meta de Coquereaumont provided editorialassistance. Parveen Moses provided excellent support.The report was published by the Partnerships andKnowledge group by the Dissemination and Outreachunit, directed by Elizabeth Campbell-Pagé (task teamleader), including Caroline McEuen (editor), KathyStrauss and Lunn Lestina (graphics and layout), Bar-bara Yale (editorial assistant), and Juicy Qureishi-Huq(administrative assistance).

TChakrapani, and Federico Mini. The report benefitedfrom the contributions of OED staff, as well asbackground papers from the Institute of DevelopmentStudies, the Overseas Development Institute, and Ox-ford Policy Management. These contributors and par-ticipants are listed in the Bibliography. Special thanksto Keith Bezanson and Robert Chamber of IDS, toSimon Maxwell and Mick Foster of ODI, and toWilliam Branson of Princeton University.

Valuable guidance was provided by an AdvisoryCommittee including Robert M. Buckley, Carl Dahlman,Christian Delvoie, Shantayanan Devarajan, IshacDiwan, David R. Dollar, Marco Ferroni, Alan HaroldGelb, Ian Johnson, Geoffrey B. Lamb, Pierre Landell-Mills, Ashoka Mody, John Page, Anthony J. Pellegrini,Guillermo Perry, Guy Pfeffermann, Ian C. Porter,Wolfgang H. Reinicke, Neil D. Roger, Nemat Talaat

Director-General, Operations Evaluation: Robert Picciotto

Director, Operations Evaluation Department: Elizabeth McAllister

Manager, Corporate Evaluations and Methods: Wendy Jarvie

Task Manager: Nagy Hanna

ix

FOREWORD PREFACIO PRÉFACE

F o r e w o r d

EN

GL

IS

H

ES

PA

NO

L

FR

AN

CA

IS

Despite the potential benefitsof globalization and technologicalchange, world poverty has in-creased and growth prospects havedimmed for developing countries.Inequality and instability have in-creased. Aid flows have stagnated.Public dissatisfaction with the effi-cacy of development assistance hasgrown.

The Comprehensive Develop-ment Framework (CDF), launchedby the President of the World Bankin January 1999, is a response tothese alarming trends. It is far tooearly to evaluate the initiative. In-stead, this year’s Annual Review ofDevelopment Effectiveness exam-ines the lessons of development ex-perience through the lens of the CDFprinciples.

As in past years, the Reviewtracks the Bank’s operational perfor-mance based on the findings ofrecent OED evaluations. The over-all performance trends are positive,despite a major increase in therelevance, demandingness, andcomplexity of Bank operations—asubstantial achievement. But futuregains will be more difficult toachieve, and quality at exit appearsto have plateaued below the Strate-gic Compact target of 75 percentsatisfactory. A growing gap betweenoperational complexity and localcapacity confirms the need for theenhanced capacity building effortshighlighted in Mr. Wolfensohn’sAnnual Meetings speech.

A pesar de los posibles beneficiosde la globalización y del cambiotecnológico, la pobreza mundial ha

aumentado y las perspectivas decrecimiento se han oscurecido en lospaíses en desarrollo. La desigualdad y lainestabilidad han aumentado. Los flujosde ayuda se han estancado. La opiniónpública cada vez desconfía más de la

eficacia de la asistencia para eldesarrollo.

El Marco Integral de Desarrollo(MID), puesto en marcha por elPresidente del Banco Mundial en enerode 1999, es la respuesta a esas

tendencias alarmantes. Es todavíademasiado pronto para evaluar esainiciativa. Por ello, en el Examen anualde la eficacia en términos de desarrollode este año se analizan las enseñanzasobtenidas de las actividades de

desarrollo desde la perspectiva de losprincipios del MID.

Como en años anteriores, en elExamen se consideran los resultadosoperacionales del Banco teniendo encuenta las comprobaciones de las

evaluaciones recientes delDepartamento de Evaluación deOperaciones (DEO). Las tendenciasgenerales observadas son positivas, loque representa un logro notable si setiene en cuenta que al mismo tiempo

ha aumentado notablemente lapertinencia, dificultad y complejidadde las operaciones del Banco. Pero losprogresos futuros serán más difícilesde conseguir y la calidad final parecehaberse estabilizado por debajo del

objetivo del Pacto Estratégico, es decir,

Malgré les avantages potentiels dela mondialisation et du progrèstechnologique, la pauvreté dans le

monde avance et les perspectives decroissance s’assombrissent pour les paysen développement. Les inégalités secreusent, l’instabilité s’accroît et le vo-lume d’aide se tasse. Le manqued’efficacité de l’aide au développement

suscite un mécontentement grandissant.Le cadre de développement intégré

(CDI), lancé par le président de laBanque mondiale en janvier 1999, viseà corriger cette évolution inquiétante.S’il est encore bien trop tôt pour

évaluer les résultats de cette initiative,la présente édition de l’Examen annuelde l’efficacité du développement dressele bilan de l’année écoulée enappliquant les principes du CDI.

Comme par le passé, nous

mettons en lumière le résultat desopérations de la Banque à partir desconclusions d’évaluations récemmentréalisées par l’OED. L’évolution de laperformance globale est positive, alorsmême que les opérations de la Banque

sont d’une portée beaucoup plus largeet qu’elles sont devenues plusexigeantes et plus complexes. C’est làun résultat non négligeable. Lesprogrès seront toutefois désormaisplus difficiles à réaliser, comme le

montre l’évaluation de la qualité àl’achèvement des projets, qui sembleplafonner au-dessous du niveau fixépar le Pacte stratégique (75 %d’opérations jugées satisfaisantes). Lefossé qui se creuse entre la complexité

des opérations et les moyens locaux


x

Robert PicciottoDirector-General, Operations Evaluation

EN

GL

IS

H

ES

PA

NO

L

FR

AN

CA

ISImplementing the CDF

principles, both within andoutside the 13 pilot countries,raises tough challenges for theBank and its partners. TheReview draws on extensiveevaluation evidence, litera-

ture reviews and research findings,country assistance evaluations,background papers, and workshopsto identify the tensions and dilem-mas likely to be involved in CDFimplementation. It identifies prom-ising approaches to meet thechallenges implicit in CDF imple-mentation. Finally, it draws broadimplications for the Bank’s futuredevelopment effectiveness.

un 75% de proyectos

calificados como satisfactorios.La diferencia creciente entre lacomplejidad operacional y lacapacidad local confirma lanecesidad de impulsar lasiniciativas de fortalecimiento de

la capacidad destacada por el Sr.Wolfensohn en su discurso conocasión de las Reuniones Anuales.

La aplicación de los principios delMID, dentro y fuera de las iniciativasexperimentales, plantea difíciles retos

para el Banco Mundial y susasociados. En el Examen se evalúanatentamente los testimoniosdisponibles, se examinan laspublicaciones y los resultados de lainvestigación, las evaluaciones de la

asistencia a los países, los documentosde antecedentes y los seminariosrealizados con el fin de descubrir lastensiones y dilemas que acompañarán,probablemente, a la aplicación delMID. Se localizan planteamientosprometedores para responder a los

desafíos implícitos en la aplicación delMID. Finalmente, se extraen algunasconclusiones generales sobre la eficaciafutura del Banco en términos dedesarrollo.

montre bien qu’il est nécessaire

de redoubler d’efforts pourrenforcer les capacités, ainsi quel’a rappelé M. Wolfensohn dansson allocution devantl’Assemblée annuelle.

L’application des principes

du CDI dans le cadre des opérationspilotes, et au-delà de celles-ci, est unetâche ardue pour la Banque et sespartenaires. Pour mettre en évidenceles tensions et les dilemmes que la miseen œuvre du CDI risque de faire

surgir, le présent examen s’appuie surde nombreux faits observés dans lesévaluations, sur des étudesdocumentaires et des travaux de re-cherche, sur des évaluations de l’aideapportée aux pays, sur des documents

de référence et sur des ateliers. Ilprésente des formules prometteusespour surmonter les difficultésinhérentes à cet exercice. Enfin, il tireles grandes conséquences de la situa-tion du point de vue de l’efficacité queles opérations de la Banque pourront

avoir sur le développement des pays.

xi

EXECUTIVESUMMARY

RESUMEN RÉSUMÉANALYTIQUE

E x e c u t i v e S u m m a r y

EN

GL

IS

H

ES

PA

NO

L

FR

AN

CA

IS

Changes in the global envi-ronment over the past decade haveprofoundly altered the context ofBank operations. Many developingcountries have undertaken majorpolitical and economic reforms,opening their economies and politi-cal processes. At the same time,

globalization has brought new op-portunities for gain—and new risksof instability and rising inequalitywithin and among states. Aid flowshave stagnated, even as new de-mands for official resources haveescalated. Despite worldwide im-provements in social indicators,development progress has been dis-appointing, poverty trends haveworsened, and most low-incomecountries remain heavily dependenton aid. Per capita incomes actuallyfell during 1985–95—by 1.4 percenta year in low-income countries (ex-

Los cambios registrados en elentorno mundial durante el pasadodecenio han modificado profundamenteel contexto de las actividades del BancoMundial. Muchos países en desarrollohan emprendido grandes reformas

políticas y económicas, que hanrepresentado una apertura de sus

economías y de los procesos políticos.Al mismo tiempo, la globalización hatraído consigo nuevas oportunidades, ynuevos riesgos de inestabilidad yaumento de la desigualdad dentro de

cada país y entre unas naciones y otras.Los flujos de la ayuda se han estancado,aun cuando se han multiplicado lasnuevas demandas de recursos oficiales.A pesar de los progresos mundiales delos indicadores sociales, el avance del

desarrollo ha sido decepcionante, lastendencias de la pobreza se hanagravado y la mayoría de los países deingreso más bajo continúan

Les changements survenus sur lascène internationale au cours des dixdernières années ont profondémentmodifié le contexte dans lequel laBanque opère. De nombreux pays ontentrepris d’importantes réformes qui se

sont traduites par une ouverture sur leplan économique et politique. Au mêmemoment, la mondialisation ouvrait desperspectives de gains mais faisait aussinaître des risques d’instabilité et creusaitles inégalités sur le plan tant national

qu’international. Malgré la montée enflèche de la demande d’aide publique, lesflux d’assistance ont fléchi. En dépitd’une amélioration des indicateurssociaux partout dans le monde, la pro-gression du développement a été

décevante, les statistiques sur la pauvretése sont aggravées et la plupart des paysà faible revenu sont restés trèsdépendants de l’aide. Pendant la périodecomprise entre 1985 et 1995, le revenupar habitant a en fait baissé, de 1,4 %

par an dans les pays à faible revenu(sans la Chine, ni l’Inde) et de 0,7 %dans les pays à revenu intermédiaire.

Dans ce contexte, le manqued’efficacité du développement est devenuun sujet de préoccupation croissante

tant pour les autorités des paysdéveloppés et des pays endéveloppement que pour le public engénéral. Face à cette situation, on arecherché différents moyens d’améliorerla façon dont la coopération pour le

développement est abordée. Le Cadre dedéveloppement intégré (CDI) reprend lesgrands thèmes autour desquelss’organise le consensus de lacommunauté du développement. Lecadre stratégique de lutte contre la


xii

ES

PA

NO

L

FR

AN

CA

IS

EN

GL

IS

H

cluding China and India) andby 0.7 percent a year inmiddle-income countries.

Inevitably, governments ofdeveloped and developingcountries, as well as the publicat large, have become more

concerned about development effec-tiveness. This has led to a wide-ranging search for improvedapproaches to development coopera-tion. The Comprehensive Develop-ment Framework (CDF) encapsulatesmajor themes around which a consen-sus of the development community iscrystallizing. The Poverty ReductionStrategy Papers—to be preparedjointly by the World Bank, the Inter-national Monetary Fund, and bor-rowing countries—are a closelyrelated response.

Since its unveiling a year ago,the CDF has evoked considerableinterest. The basic elements of theCDF are not new. What is new is theirjoint articulation as a guide to devel-opment assistance. First, develop-ment constraints are structural andsocial and cannot be overcomethrough economic stabilization andpolicy adjustment alone—they re-quire a holistic, broad-based ap-proach. Second, policy reform andinstitutional development cannot beimported or imposed: without domes-tic ownership, reforms and invest-ments are not sustainable. Third,successful development requires part-nerships among government, localcommunities, the private sector, civilsociety, and development agencies.Fourth, development activities mustbe guided and judged by results. CDFpilot activity is beginning in 13 coun-tries, and its dimensions are stillevolving. Although it is too early toassess the CDF as such, this year’sAnnual Review of Development Ef-fectiveness examines development

dependiendo fuertemente de laayuda. Los ingresos per cápitaretrocedieron de hecho en 1985-95, un 1,4% al año en los paísesde ingreso bajo (con exclusión deChina y la India) y un 0,7% al

año en los países de ingresomediano.

Inevitablemente, los gobiernos delos países desarrollados y endesarrollo, así como la sociedad engeneral, se muestran ahora más

preocupados por la eficacia entérminos de desarrollo. Ello ha dadolugar a una búsqueda generalizada denuevos enfoques que permitanorientar mejor la cooperación en mate-ria de desarrollo. El MID compendia

los grandes temas en torno a loscuales se está cristalizando el consensode todos los interesados en eldesarrollo. En ese mismo contexto seestán preparando los documentos deestrategia para la reducción de la

pobreza, obra conjunta del BancoMundial, el Fondo MonetarioInternacional y los países prestatarios.

Desde que se propuso, hace unaño, el MID ha suscitado considerableinterés. Sus elementos básicos no son

nuevos. Lo novedoso es suarticulación conjunta como guía de laasistencia para el desarrollo. Primero,los obstáculos al desarrollo sonestructurales y sociales y no puedensuperarse únicamente con políticas de

ajuste y de estabilización económica,sino que requieren un planteamientoglobal, de base amplia. Segundo, lareforma de las políticas y el desarrolloinstitucional no pueden importarse niimponerse: si en los propios países no

hay una identificación con las nuevaspropuestas, ni las reformas ni lasinversiones serán sostenibles. Tercero,un desarrollo eficaz requiere elestablecimiento de asociaciones entrelos gobiernos, las comunidades lo-

cales, el sector privado, la sociedad

pauvreté, qui sera conjointementélaboré par la Banque mondiale,le Fonds monétaire internationalet les pays emprunteurs, est uneintervention étroitement liée àcette question.

Depuis son apparition, il ya un an, le CDI suscite un intérêtconsidérable. Les éléments sur lesquelsil se fonde ne sont pas nouveaux. Enrevanche, la façon dont ils s’articulentles uns avec les autres pour servir de fil

conducteur à l’aide au développementest une donnée nouvelle.Premièrement, les problèmes auxquelsse heurte le développement étant denature structurelle et sociale, lesmesures de stabilisation économique

et le recadrage des politiques nesuffiront pas pour les surmonter ;l’approche doit être globale etmultisectorielle. Deuxièmement, ni lesréformes, ni le développementinstitutionnel ne peuvent être importés

ou imposés : la pérennité des réformeset des investissements passe par uneadhésion sans réserve au niveau local.Troisièmement, le développement nepeut réussir que s’il fait appel à despartenariats entre les pouvoirs publics,

les populations locales, le secteurprivé, la société civile et les organismesd’aide au développement. Enfin, les ac-tions de développement doivent avoirle souci du résultat. Le CDI est lancé àtitre pilote dans 13 pays et il n’a pas

encore revêtu sa forme définitive. Bienqu’il soit encore trop tôt pour dresserun premier bilan, la présente éditionde l’Examen annuel de l’efficacité dudéveloppement s’appuie sur lesprincipes du CDI pour apprécier le

chemin parcouru et tirer desenseignements qui faciliterontl’application des dispositions prévues.

Évolution générale des projetsLes faits observés dans les évaluationsmontrent que le résultat des projets est

xiii

EN

GL

IS

H

ES

PA

NO

L

FR

AN

CA

IS


experience through the lens ofCDF principles, with a view todrawing lessons for CDF imple-mentation.

Project TrendsEvaluation evidence confirms

the importance of country policyand institutional factors addressedby the CDF to project performance.Projects are important vehicles ofBank assistance, and most perfor-mance at the project level willcontinue to be a critical element ofdevelopment effectiveness. Projectevaluations have shown improve-ments over the past decade, alongwith a considerable increase inproject demandingness and com-plexity—a considerable achieve-ment. Adjustment lending, inparticular, has delivered a rela-tively high share of satisfactoryoutcomes. But there has been arecent plateauing in performance:72 percent of evaluated projectsexiting in FY98–99 show satisfac-tory outcomes, while the share ofexiting projects likely to sustaintheir benefits remains below half.

As the complexity of exitingprojects has risen, borrower imple-mentation performance has beenstagnant, which has contributed tothe plateauing of performance.Analysis shows borrower imple-mentation performance to be a keydeterminant of project success, sec-ond only to Bank supervision. Thewidening gap between complexityand local capacity highlights theimportance of these inputs, as wellas the need for intensified attentionto institutional constraints. Despitesubstantial improvements in recentyears, only 40 percent of exitingprojects currently have substantialinstitutional development impact.

civil y los organismos dedesarrollo. Cuarto, lasactividades de desarrollo debenorientarse y juzgarse teniendoen cuenta los resultados. Laaplicación experimental del

MID ha comenzado en 13países, y sus dimensiones están aún enevolución. Aunque es todavíademasiado pronto para evaluar elMID en cuanto tal, en el Examenanual de la eficacia en términos de

desarrollo de este año se examina larealidad del desarrollo desde laperspectiva de los principios del MID,con el fin de extraer enseñanzas parasu aplicación.

Tendencias de los proyectosLos datos relativos a las evaluacionesconfirman la importancia de los factoresnormativos e institucionales de lospaíses contemplados en el MID para

determinar los resultados de losproyectos. Los proyectos son vehículosimportantes de asistencia del Banco, porlo que la multiplicación de esosresultados continuará siendo unelemento crítico de la eficacia del

desarrollo. Las evaluaciones de losproyectos han demostrado losprogresos conseguidos en el pasadodecenio, a pesar de que ha aumentadotambién notablemente la dificultad ycomplejidad de los proyectos, lo que

representa un importante logro. En par-ticular, el financiamiento para fines deajuste ha conseguido una proporciónrelativamente elevada de calificacionessatisfactorias. Pero se ha observadotambién una estabilización de los

resultados: el 72% de los proyectosevaluados finalizados en los ejercicios de1998-99 presentan resultadossatisfactorios, mientras que la parte delos proyectos concluidos conprobabilidades de continuar

produciendo beneficios se mantiene pordebajo de la mitad.

étroitement lié aux politiquesnationales et aux facteursinstitutionnels, aspects au cœurdu CDI. Les projets étant l’un desgrands moyens d’assistance de laBanque, la performance à ce

niveau continuera à influer defaçon déterminante sur l’efficacité dudéveloppement. Sur les dix annéesécoulées, le résultat des évaluations deprojet est en amélioration ce qui, comptetenu de l’exigence et de la complexité

toujours croissantes des opérations,constitue un bilan remarquable. Lesprêts à l’ajustement, en particulier, ontdonné lieu, pour une assez large part, àdes résultats jugés satisfaisants. On atoutefois constaté récemment un certain

tassement : 72 % des projets évalués quisont sortis du portefeuille pendant lesexercices 98 et 99 ont enregistré desrésultats satisfaisants, mais moins de lamoitié sont susceptibles de confirmerdurablement cette performance.

Les projets étant devenus pluscomplexes, la performance desemprunteurs au niveau de l’exécutiondes opérations est restée stationnaire,ce qui contribue à la stagnation desrésultats. Les analyses montrent que

cette performance est un élément quivient immédiatement après la supervi-sion de la Banque dans les facteurs quidéterminent le succès d’un projet. Lecreusement de l’écart entre lacomplexité des opérations et les

capacités locales souligne l’importancede ces relais et la nécessité d’unemeilleure prise en compte descontraintes institutionnelles. Malgréune nette amélioration au cours desdernières années, seuls 40 % des

projets sortant du portefeuille ontactuellement un impact important surle développement institutionnel.

Une tâche ardueLe CDF repose sur des principes quitirent parti des synergies. Mais celles-ci


xiv

EN

GL

IS

H

ES

PA

NO

L

FR

AN

CA

ISTough Challenges Ahead

The CDF principles are syner-gistic. But their combinationcarries with it critical tensionsthat will require careful man-agement.

Short term versus longterm. Social transformation and in-stitutional development requirelong-term and sustained efforts. Yetthe incentives and processes in de-veloping countries and aid agencieshave tended to focus on the shortterm. Blueprint planning and theproject approach embody this short-term focus. The absence of a long-term perspective has been especiallydetrimental in dealing with thestructural dimensions of reform,such as privatization, civil servicereform, and deregulation of thefinancial sector.

Comprehensiveness versus selec-tivity. The Bank’s expanding agendahas increased the complexity, de-mandingness, and risk of projects andprograms. In the face of these pres-sures, selectivity has not always beenexercised. Yet evaluation findingshighlight the risk of excessive com-plexity, especially when it is built intothe design without regard to domesticcapacity or ownership. Projects thatinvolve several implementing agen-cies or cross-sectoral assets tend toperform poorly.

Speed versus broad-based own-ership. Sustainable reform requiresbroad-based ownership. Yet part-ners may disagree about the distinc-tive roles of the state, the privatesector, and civil society. Achievingagreement among partners may de-lay reform. Governance reformsmay be required to institutionalizeparticipation—a long-term process.

Ownership versus conditionality.How should the apparent tensionbetween country ownership and do-

A medida que haaumentado la complejidad delos proyectos concluidos, eldesempeño de los prestatariosen la ejecución de los proyectosse ha estancado, lo que ha

contribuido a la estabilizaciónde los resultados. Los análisis

efectuados revelan que los resultadosde la ejecución de los proyectos son undeterminante fundamental del éxito deéstos, cuya importancia sólo es infe-

rior a la supervisión del Banco. Ladiferencia creciente entre lacomplejidad y la capacidad local ponede manifiesto la importancia de esosaportes, así como la necesidad deintensa atención a los obstáculos

institucionales. A pesar de lasconsiderables mejoras de los últimosaños, sólo el 40% de los proyectosconcluidos tienen actualmenteimportantes repercusiones en eldesarrollo institucional.

Arduos desafíos para el futuroLos principios en que se fundamenta elMID son sinérgicos, pero lacombinación de éstos conlleva tensiones

críticas que se deberán manejar consumo cuidado.

Corto plazo versus largo plazo.La transformación social y eldesarrollo institucional exigenesfuerzos sostenidos y a largo plazo.

Sin embargo, los incentivos y procesosen los países en desarrollo y en losorganismos de ayuda se han centradoen el corto plazo, tendencia que se hacaracterizado por una planificacióndetallada y un enfoque en los

proyectos. La falta de una perspectivaa largo plazo ha sido especialmenteperjudicial para abordar lasdimensiones estructurales de lareforma, como la privatización, lareforma de la administración pública y

la desreglamentación del sectorfinanciero.

génèrent aussi des antagonismesqu’il faudra gérer prudemment.

Le court terme et le longterme. La transformationsociale et le développementinstitutionnel supposent un ef-

fort de longue haleine.Pourtant, les mécanismes d’incitationet le mode de fonctionnement quiprévalent dans les pays endéveloppement et dans les organismesd’aide ont tendance à privilégier le

court terme. La planification à l’aidede modèles et l’approche-projets’inscrivent dans cette perspective. Lesdimensions structurelles des réformes(privatisation, réforme de la fonctionpublique et déréglementation du

secteur financier, par exemple) ontparticulièrement souffert de cette ab-sence d’optique à long terme.

Exhaustivité et sélectivité. Du faitde l’expansion du programme de tra-vail de la Banque, les projets et les

programmes se sont fait plus com-plexes, risqués et exigeants. Face àcette situation difficile, on ne procèdepas toujours de façon sélective.Pourtant, les évaluations montrentqu’une complexité excessive comporte

des risques, surtout lorsqu’il n’a pasété tenu compte des capacités ou de lavolonté d’appropriation au niveau lo-cal. Les projets faisant intervenirplusieurs organismes d’exécution oudes actifs dans plusieurs secteurs

obtiennent généralement des résultatsmédiocres.

Rapidité et large adhésion. Pourqu’une réforme soit durable, il fautque l’adhésion à l’action menée soittotale. Pourtant, les partenaires ne

sont pas toujours d’accord sur lesrôles respectifs de l’État, du secteurprivé et de la société civile. La réformepourra prendre du retard du fait dutemps nécessaire aux différents acteurspour parvenir à un accord. Il faudra

parfois réformer le mode de gestion

xv

EN

GL

IS

H

ES

PA

NO

L

FR

AN

CA

IS


nor interest in performance (of-ten enforced through condition-ality) be resolved? Conditional-ity is widely viewed as anattempt to impose reforms andblueprint solutions on clients.Yet ownership is essential to

sustainable reform. New forms ofconditionality—reflecting the CDFprinciples—should be instituted.

Partnership versus country ca-pacity and ownership. Countries inneed of partnership often lack thecapacity to coordinate aid. Govern-ments, civil society, the private sector,and external donors may have differ-ent agendas, requiring a gradualapproach to consensus building. Thelarger and more diverse the partners,the higher the transaction costs andthe greater the difficulty of combiningownership and partnership. Wherecountry commitment is lacking, co-herent views among external partnersmay be perceived as “ganging up.”

Country-led partnership versusdonor accountability. All develop-ment assistance agencies are account-able to their authorities, so placing anunresponsive government “in thedriver’s seat” involves risks. Thechallenge is to find common groundand achieve results over the long termthrough patient nurturing of reformand capacity building.

Results-orientation versus localcapacity. The performance of theBank and most developing countriesin monitoring and evaluation hasbeen weak. Yet the internationaldevelopment goals, the recent atten-tion to governance, and the move toprogrammatic lending reinforce theneed for results-based managementand stronger evaluation capacitiesand local accountability systems.

Country-based programs versusglobal public goods. Developmentassistance efforts and approaches

Integridad versusselectividad. El programa deactividades del Banco, que va enaumento, presenta una mayorcomplejidad, dificultades yriesgos de los proyectos y

programas. Ante talespresiones, no siempre se ha

aplicado el criterio de la selectividad.Con todo, en las conclusiones de lasevaluaciones practicadas se destaca elriesgo de una excesiva complejidad,

sobre todo cuando ésta se incluye en eldiseño sin tener en cuenta la capacidadlocal o la identificación del país con laspropuestas. Los proyectos queentrañan la participación de variosorganismos de ejecución o activos

intersectoriales tienden a arrojarresultados deficientes.

Velocidad versus ampliaidentificación. Para que las reformassean sostenibles es preciso que existauna amplia identificación con ellas. No

obstante, los asociados puedendiscrepar con respecto a las diferentesfunciones del Estado, el sector privadoy la sociedad civil. El logro de unconsenso entre los asociados puededemorar las reformas. Puede ser

necesario introducir reformas de lafunción de gobierno parainstitucionalizar la participación, unproceso a más largo plazo.

Identificación versuscondicionalidad. ¿Cómo debe

resolverse la aparente tensión queexiste entre la identificación del país yel interés de los donantes en eldesempeño (el que a menudo se exige através de la condicionalidad)? Existe laimpresión generalizada de que la

condicionalidad es un burdo intentode imponer reformas y solucionesplanificadas a los clientes. Sin em-bargo, la identificación con lasreformas es fundamental para queéstas sean duraderas. Deben instituirse

nuevas formas de condicionalidad,

des affaires publiques pourinstitutionnaliser la participa-tion, une entreprise de longuehaleine.

Adhésion au projet etconditionnalité. Comment

l’apparente contradiction entrel’adhésion du pays à l’action menée etles résultats demandés par les bailleursde fonds (souvent au moyen de laconditionnalité) peut-elle être résolue ?La conditionnalité est très souvent

perçue comme une tentative nondéguisée d’imposer des réformes ou unmodèle aux pays clients. Et pourtant,il ne peut y avoir de réformes durablessi le pays ne s’identifie pas à l’actionmenée. Il conviendrait de repenser la

conditionnalité, sous une forme plusrespectueuse des principes du CDI.

Partenariat, capacité locale etadhésion du pays. Les pays devantfaire appel à des partenaires multiplessont rarement à même de coordonner

l’aide reçue. Les objectifs desgouvernements, de la société civile, dusecteur privé et des bailleurs de fondspeuvent en effet être différents, ce quiobligera les partenaires à rechercherprogressivement le consensus

nécessaire. Plus les partenaires sontnombreux et différents, plus les coûtsde transaction sont importants et plusil est difficile de concilier adhésion etpartenariat. En l’absence, de volontépolitique de la part du pays

bénéficiaire, une communauté de vuesdes bailleurs de fonds peut être perçuecomme une « coalition ».

Partenariat piloté par le pays etresponsabilité des bailleurs de fondsenvers leur autorité de tutelle. Tous les

organismes d’aide au développementayant à répondre de leur gestiondevant leur autorité de tutelle, « passerles rênes » à un gouvernement peumotivé comporte des risques. Il s’agirade trouver un terrain d’entente et

d’obtenir des résultats à long terme en


xvi

EN

GL

IS

H

ES

PA

NO

L

FR

AN

CA

ISare focused at the project and

country levels. With global-ization, development prob-lems require multilateralsolutions and stronger linksbetween national strategiesand international policies.

Promising ApproachesIn order to resolve the above dilem-mas and implementation chal-lenges, it will be necessary to em-ploy creativity and innovation andto heed the lessons of experience.Several promising practices emergefrom this review. All are predicatedon country commitment to povertyreduction and sustainable growth,and a few are quick to take to scale.

• An adaptive learning processstarts small and combines top-down direction with bottom-upexperimentation and learning.This shifts the emphasis fromup-front analysis and detaileddesign toward developing flex-ible solutions, building localcapacity, and relying on socialprocesses and monitoring sys-tems for adaptation and learn-ing during implementation.

• One way out of the excessivecomplexity of projects andprograms is to sequence inter-ventions within a long-termstrategy that builds on pastlearning. Such sequencing canstart by piloting comprehensiveapproaches at the local level,then scaling them up as part ofa long-term process of capacitybuilding and decentralization.

• Tensions between comprehen-siveness and selectivity can beeased by combining compre-hensive analysis with strategicactions. Participatory ap-proaches can enable poorpeople to analyze their realities

que reflejen los principios delMID.

Asociaciones versuscapacidad e identificación de los

países. Los países que necesitanestablecer asociaciones amenudo carecen de la capacidad

para coordinar la ayuda. Losgobiernos, la sociedad civil, el sectorprivado y los donantes externos

pueden tener distintos programas, loque hace necesario adoptar unplanteamiento gradual para laformación de consenso. Mientras másnumerosos y diversos sean losasociados, mayores serán los costos

de transacción y las dificultades paracombinar la identificación y lasasociaciones. Cuando en un país faltael sentido de compromiso, lacoherencia de opiniones entre losasociados externos puede percibirse

como una “alianza”.Asociación de esfuerzos

encabezada por el país versusresponsabilidad de los donantes.Todos los organismos de asistenciapara el desarrollo son responsables

ante sus autoridades, de manera queponer a un gobierno que no reaccionaen el “ asiento del conductor” planteaciertos riesgos. El desafío consiste enencontrar un terreno común y lograrresultados a largo plazo fomentando

con paciencia la reforma y eldesarrollo de la capacidad.

Orientación a los resultados ver-sus capacidad a nivel local. Laactuación del Banco y de la mayoría delos países en desarrollo en lo que

respecta a las actividades deseguimiento y evaluación ha sidodeficiente. Con todo, las metasinternacionales de desarrollo, lareciente atención a los asuntosrelativos a la función de gobierno y el

avance hacia el financiamiento paraprogramas refuerzan la necesidad deuna gestión basada en los resultados,

renforçant patiemment lescapacités et en laissant mûrir lesréformes.

Obligation de résultat etcapacités locales. Le travail de

suivi et d’évaluation de laBanque et de la plupart des pays

en développement laisse à désirer. Etpourtant l’existence d’objectifsinternationaux de développement,l’accent récemment mis sur le mode

d’administration des affairespubliques et le passage à des prêtsprogrammatiques rendent d’autantplus nécessaires des mesures visant àaxer la gestion sur les résultats, àaméliorer les capacités d’évaluation et

à renforcer les systèmes de contrôle.Programmes au niveau des pays

et biens publics de caractère mondial.Les actions et stratégies d’aide audéveloppement sont conçues au niveaudes projets et des pays. Du fait de lamondialisation, la solution aux

problèmes de développement passepar l’adoption de solutionsmultilatérales et par un renforcementdes liens existant entre les stratégiesnationales et les politiquesinternationales.

Des possibilités intéressantesPour sortir des dilemmes mentionnésplus haut et résoudre les problèmes de

mise en œuvre qui en découlent, ilfaudra faire preuve de créativité etd’innovation et savoir tirer les leçons del’action menée. Plusieurs possibilitésintéressantes se dégagent du présentexamen. Toutes reposent sur la volonté

du pays de faire reculer la pauvreté et depromouvoir une croissante durable,mais rares sont celles qui peuvent êtrerapidement portées à une plus grandeéchelle.

• Tout processus d’apprentissage

évolutif doit démarrer à petiteéchelle et allier une démarchevenant d’en haut à une acquisi-

xvii

EN

GL

IS

H

ES

PA

NO

L

FR

AN

CA

IS


in a comprehensive way, andthen express their prioritiesand choose interventions se-lectively. An example of aholistic yet selective approachis the sustainable livelihoodsapproach of the U.K. Depart-

ment for International Develop-ment, CARE, and the UnitedNations Development Program(UNDP).

• Sectorwide and programmaticapproaches can help reduce therigidities and burdens associ-ated with the proliferation ofdonor-financed projects. Learn-ing and innovation pilots fol-lowed by adaptable programloans can support prioritygoals, particularly in sectorswhere flexibility is critical. Butmoving prematurely fromprojects to full-scale sector ap-proaches is risky. Capacitybuilding must accompany theprocess of scaling up. More-over, sectorwide approachesshould be tailored to countriesand sectors.

• Conditionality should be recon-ciled with country ownership.Empirical analysis of pastcountry reforms over a longhorizon and of the recenthigher-impact adjustment lend-ing in Africa confirms that con-ditionality is best managed as aflexible, noncoercive policycompact adapted to differentstages of reform.

• Broadening ownership acrossmany stakeholders with diverseinterests and capabilities requirestime and early mobilization, par-ticularly of communities and thepoor. It implies the systematic useof participatory processes; in-forming and giving voice to theweak partners, particularly

así como de una mayorcapacidad de evaluación y desistemas para asegurar laresponsabilidad a nivel local.

Programas basados en los

países versus bienes públicosmundiales. Las iniciativas y

enfoques relativos a la asistencia parael desarrollo se centran al nivel de losproyectos y de los países. Con elproceso de globalización, los

problemas del desarrollo exigensoluciones multilaterales y elestablecimiento de vínculos másestrechos entre las estrategiasnacionales y las políticasinternacionales.

Planteamientos prometedoresPara resolver los dilemas y retos en ma-teria de aplicación señaladosanteriormente, será necesario recurrir a

la creatividad y la innovación, y poneratención a las enseñanzas de laexperiencia. Del presente examen surgenvarias prácticas prometedoras. Todas sefundamentan en la determinación de lospaíses para reducir la pobreza y lograr el

desarrollo sostenible, y unas pocas sepueden aplicar de inmediato en mayorescala.

• Un proceso de aprendizajeadaptativo comienza en formamodesta y en él se combina una

dirección de arriba hacia abajocon experimentación yaprendizaje de abajo hacia arriba.Con esto se cambia de énfasis alpasar de un análisis inicial y undiseño detallado a la formulación

de soluciones flexibles, la creaciónde capacidad a nivel local y elapoyo en los procesos sociales yen sistemas de seguimiento quepermitan la adaptación y elaprendizaje durante el proceso de

aplicación.• Una manera de evitar proyectos y

programas excesivamente

tion des connaissances et à uneexpérimentation partant dubas. Cela permettra de moinsmettre l’accent sur les analyses

initiales et les plans détaillés etde privilégier la recherche de so-lutions souples, le renforcement

des capacités locales et le recoursaux systèmes de suivi et auxmécanismes encadrant les

conduites au sein des collectivitéspour favoriser l’adaptation etl’apprentissage pendantl’exécution des activités.

• L’un des moyens de remédier à lacomplexité des projets et

programmes est d’échelonner lesinterventions en appliquant unestratégie qui intègre les acquis surle long terme. Pour échelonner lesinterventions, on peut commencerpar expérimenter des méthodes

globales au niveau local, pour lesappliquer ensuite à plus grandeéchelle dans le cadre d’un proces-sus de renforcement des capacitéset de décentralisation qui s’inscritdans la durée.

• En complétant l’analyse globaled’actions stratégiques, on peutatténuer l’antinomie existant entreexhaustivité et sélectivité. Demême, en associant les pauvres àl’action menée, on peut leur

permettre d’analyser la réalité defaçon globale, puis de fixer leurspriorités et de sélectionner les in-terventions en conséquence. Lafaçon dont le Département pourle développement international

(Royaume-Uni), CARE et lePNUD abordent la question desmoyens d’existence viables est unbon exemple d’une approche à lafois globale et sélective.

• Les approches par secteur et par

programme peuvent aider àréduire les rigidités et lescontraintes liées à la prolifération


xviii

EN

GL

IS

H

ES

PA

NO

L

FR

AN

CA

IS

women and marginalizedpopulations; and creating anenabling environment for scal-ing up and institutionalizingparticipation.• Partnership requires twocomplementary roles for the

complejos es establecer unorden de sucesión de lasintervenciones dentro de unaestrategia a largo plazo que sefundamente en las leccionesextraídas del pasado. Tal

secuenciamiento puedecomenzar con la aplicación ex-

perimental de planteamientosglobales al nivel local, que luego sepueden ir incrementando comoparte de un proceso a largo plazo

de desarrollo de capacidades ydescentralización.

• Las tensiones entre la integridad yla selectividad pueden suavizarsemediante la combinación de unanálisis cabal y acciones

estratégicas. Los enfoquesparticipatorios pueden permitirque la gente pobre analice surealidad en forma integral, y luegoexpresar sus prioridades yescoger selectivamente las

intervenciones. Un ejemplo de unplanteamiento integral peroselectivo es el enfoque sobremedios de vida sostenibles delDepartamento de DesarrolloInternacional del Reino Unido,

CARE y el PNUD.• Los enfoques sectoriales y

programáticos pueden ayudar areducir las rigideces y la cargaasociadas a la proliferación deproyectos financiados por

donantes. Los préstamosexperimentales para el aprendizajey la innovación seguidos depréstamos adaptables paraprogramas pueden apoyar laconsecución de metas prioritarias,

sobre todo en sectores en los quela flexibilidad es fundamental.Pero pasar prematuramente delos proyectos a enfoquessectoriales integrales planteariesgos. El proceso gradual debe ir

acompañado de la creación de

de projets financés par desbailleurs de fonds. Desopérations pilotes pour ledéveloppement desconnaissances et l’innovation,complétées de prêts à des

programmes évolutifs, peuventaider à la réalisation d’objectifs

prioritaires, surtout là où la lati-tude d’action joue un rôledéterminant. Il est toutefois risquéde transposer prématurément au

niveau de l’ensemble d’un secteurles activités réalisées à l’échelond’un projet. Le passage à l’échellesupérieure doit s’accompagnerd’un travail de renforcement descapacités. L’approche devra en

outre être adaptée au pays et ausecteur considérés.

• Il faut concilier conditionnalité etadhésion du pays à l’actionmenée. L’analyse sur une longuepériode des réformes entreprises

dans les pays et l’examen desrécentes opérations de prêt àl’ajustement, qui ont eu desconséquences importantes enAfrique, montrent que laconditionnalité gagne à prendre la

forme d’un programme souple,non imposé, qui évolue au fil deson application.

• Pour amener un grand nombre departies prenantes sanscommunauté de moyens ni

d’intérêts à souscrire pleinement àl’action menée, il faut compter avecle temps et mobiliser rapidement lesintéressés, tout particulièrement lapopulation et les pauvres. À cettefin, il faudra systématiquement faire

appel à la participation, s’adresseret donner la parole aux partenairesles plus faibles, notamment auxfemmes et aux exclus, et créer lesconditions qui permettrontd’élargir et d’institutionnaliser cette

participation.

Bank: supporting country lead-ership and building the capac-ity to exercise that leadership,and engaging its developmentassistance partners to promoteselectivity, coordinate interven-tions, and harmonize proce-dures. Partnerships may implyup-front investments and mustbe effectively monitored andmanaged if transaction costs todonors and countries are todecline over time.

Evaluating DevelopmentEffectivenessThe changing global economy andthe CDF suggest several principlesfor evaluating development effec-tiveness.

Evaluation should focus on re-sults, and this depends on accuratetracking of progress toward devel-opment goals, with a clear focus onpoverty reduction and growth.Tracking development outcomesshould comply with the comprehen-sive development agenda agreed bythe government and its partners.

As the focus of the developmenteffort moves from projects to thehigher plane of country programs,so must the evaluation process. Re-sources and skills should be investedin developing appropriate indica-tors and information systems. Thecurrent preoccupation with projectperformance and evaluation shouldbe complemented by a sectoral andcountrywide focus. Public sector re-form needs to include building theevaluation capacity of countries.

xix

Development effective-ness should be evaluated interms of shared objectives,joint responsibility for out-comes, reciprocal obligationsto achieve results, and distinctaccountability for perfor-mance. In particular, donors

and governments should team up toinvolve civil society and the privatesector in monitoring and evaluation,and to help all stakeholders acquirethe needed skills and attitudes. Par-ticipatory monitoring and evaluationhold significant promise for sociallearning and managing for results.

Evaluation should be informedby the global perspective of theInternational Development Goalsendorsed by the development com-munity. Far from implying rigid,top-down global planning, this two-way link means adapting the inter-national goals to country conditionsand priorities and enhancing part-nerships at all levels.

Implications for the BankThe CDF is not a blueprint—thehallmark of the planning era andthe adjustment era. Its aim is tolaunch a process that adapts con-tinually to changing circumstances.Customization should thus replacethe “one-size-fits-all” mindset.

To this end, the Bank must domore to learn and listen. The CDFpoints to the pivotal role of knowl-edge and learning in development.Too often, it is assumed that theBank has the answers, and that theonly problem is to sell the preferredsolution to clients. But the Bank isnot a storehouse of universally ap-plicable knowledge just waiting tobe transferred. That is why it shouldempower its clients to tap globalknowledge, connect clients to oneanother and to other sources of

EN

GL

IS

H

ES

PA

NO

L

FR

AN

CA

IS


capacidad. Además, losenfoques sectoriales deberíanadaptarse a las circunstanciasde los países y los sectores.• La condicionalidad deberíaconciliarse con la identificación

de los países. El análisisempírico de las reformas

introducidas en el pasado duranteun prolongado horizonte tempo-ral y el financiamiento para finesde ajuste otorgado recientemente

a África confirman que la mejormanera de gestionar lacondicionalidad es mediante unpacto de políticas flexible, nocoercitivo, adaptado a lasdiferentes etapas de la reforma.

• Para que la identificación demuchas partes interesadas condiversos intereses y capacidadespueda extenderse se requieretiempo y un proceso temprano demovilización, especialmente de las

comunidades y de los pobres.Para ello es preciso recurrirsistemáticamente a procesosparticipatorios; informar y daroportunidades de expresión a losasociados más débiles, sobre todo

a las mujeres y a las poblacionesmarginadas, y crear condicionespropicias para incrementar einstitucionalizar la participación.

• Para las asociaciones se requierendos funciones complementarias

de parte del Banco: el respaldo alos dirigentes nacionales y eldesarrollo de la capacidad paraejercer ese liderazgo, y laparticipación de sus organismosasociados de asistencia para el

desarrollo a fin de promover laselectividad, coordinar lasintervenciones y armonizar losprocedimientos. Las asociacionespueden entrañar inversionesdesde un comienzo y deben

vigilarse y gestionarse de manera

• Pour promouvoir lepartenariat, l’action de laBanque doit être double. Elledoit aider les pays à jouer un

rôle moteur en contribuant aurenforcement de leurs capacitéset elle doit inciter ses partenaires

bailleurs de fonds à encourager lasélectivité, à coordonner les inter-ventions et à harmoniser les

procédures. Les partenariatsimpliquent parfois uninvestissement initial et doiventêtre efficacement suivis et réguléspour que les coûts de transactionà la charge des bailleurs de fonds

et des pays diminuent au fil dutemps.

Évaluation de l’efficacité dudéveloppementLa mondialisation de l’économie et lecontenu du CDI donnent à penser queplusieurs principes doivent s’appliquer àl’évaluation de l’efficacité dudéveloppement.

L’évaluation doit être axée sur les

résultats. Pour cela, il faut pouvoirsuivre avec précision les progrèsréalisés par rapport aux objectifs dedéveloppement, en s’intéressant toutparticulièrement à la croissance et à lalutte contre la pauvreté. La mise en

lumière des fruits du développementdoit correspondre au programmed’ensemble arrêté d’un commun ac-cord par le pays et ses partenaires.

Le recentrage de l’effort dedéveloppement, qui est passé des projets

aux programmes à l’échelle du pays,doit s’accompagner d’un recadrage ana-logue du processus d’évaluation. Ilconvient de mobiliser les ressources et lescompétences nécessaires pour mettre aupoint les indicateurs et les systèmes

d’information voulus. Outre l’attentionactuellement portée à la performance età l’évaluation des projets, il faut mettrel’accent sur les résultats au niveau de


xx

EN

GL

IS

H

ES

PA

NO

L

FR

AN

CA

ISexperience, and learn with

them about what works.Practicing selectivity

through partnership willrequire continuous change.Bank management has initi-ated many changes under the

Strategic Compact. More changewill be needed to sharpen the Bank’sfocus and improve the developmenteffectiveness of its operations. Thatwill involve new skills, new ap-proaches, new incentives, and newattitudes—to make room for genu-ine partnerships. Above all, theBank will have to be realistic andstrategic about its evolving com-parative advantage in a fast-changing environment.

eficaz a fin de que los costos detransacción para los donantes ylos países puedan disminuir a lolargo del tiempo.

Evaluación de la eficacia entérminos de desarrollo

La cambiante economía mundial y elMID plantean varios principios para

evaluar la eficacia en términos dedesarrollo.

La evaluación debe centrarse en losresultados, y para ello es necesario unseguimiento exacto de los progresoshacia la consecución de las metas en ma-

teria de desarrollo, con un claro enfoqueen la reducción de la pobreza y en elcrecimiento. El seguimiento de losresultados en términos de desarrollodebe ajustarse al programa integral dedesarrollo convenido con el gobierno y

sus asociados.A medida que el énfasis de las

iniciativas en pro del desarrollo avanzandel nivel de los proyectos al plano máselevado de los programas nacionales, lopropio debe ocurrir con el proceso de

evaluación. Se deben invertir recursos yconocimientos para formularindicadores y sistemas de informaciónadecuados. La actual preocupación porlos resultados de los proyectos y laevaluación debe complementarse con un

enfoque sectorial y nacional. La reformadel sector público debe incluir eldesarrollo de la capacidad de evaluaciónpor parte de los países.

La eficacia del desarrollo debeevaluarse en términos de los objetivos

compartidos, la responsabilidadconjunta por los resultadosobtenidos, las obligaciones recíprocaspara alcanzar resultados y ladistinción de responsabilidades conrespecto al desempeño. En particular,

los donantes y los gobiernos debenunirse para hacer participar a lasociedad civil y al sector privado en lasactividades de seguimiento y

l’ensemble du secteur et du pays.À cette fin, les réformes dusecteur public doiventcomprendre le renforcement de lacapacité d’évaluation des pays.

Des objectifs communs, uneresponsabilité conjointe des

résultats à obtenir, une obligation derésultat réciproque et des impératifsdistincts en ce qui concerne lescomptes à rendre sont les critères à

utiliser pour évaluer l’efficacité dudéveloppement. Il faut, en particulier,que les bailleurs de fonds et lespouvoirs publics s’emploient de con-cert à associer la société civile et lesecteur privé au travail de suivi et

d’évaluation, et à aider toutes les par-ties prenantes à acquérir lescompétences et l’état d’espritnécessaires. Ici, l’approche participa-tive laisse bien augurer de l’avenir, carelle est instructive pour la société etpermet une gestion soumise à des

critères de performance.Le travail d’évaluation devra

s’inscrive dans la perspective mondialedes Objectifs internationaux dedéveloppement approuvés par lacommunauté du développement. Loin

de déboucher sur une planificationimposée d’en haut et rigide au niveaumondial, ce double lien doit permettred’adapter les objectifs internationauxà la situation et aux priorités du payset de renforcer les partenariats à tous

les échelons.

Répercussions pour la BanqueLe CDI n’est pas un modèle et ne vise

pas à marquer de son estampille l’âge dela planification et de l’ajustement. Il apour objectif de lancer un processus quine cessera de s’adapter à l’évolution de lasituation. Le « sur mesure » est doncappelé à remplacer la « taille unique ».

Pour cela, il faut que la Banquecherche davantage à apprendre et àécouter. Le CDI insiste sur le rôle

xxi

ES

PA

NO

L

FR

AN

CA

ISevaluación, y para ayudar a

todas las partes interesadas aadquirir y desarrollar losconocimientos y las actitudesnecesarios. El seguimiento y la

evaluación con participaciónofrecen grandes esperanzas

para el aprendizaje social y la gestiónde los resultados.

La evaluación debe fundamentarseen la perspectiva global de las metas

internacionales en materia de desarrolloaprobadas por la comunidad deldesarrollo. Muy lejos de implicar unaplanificación global rígida, de arribahacia abajo, este vínculo bidireccionalsignifica adaptar las metas

internacionales a las condiciones yprioridades de cada país y mejorar lasasociaciones en todos los niveles.

Repercusiones para el BancoEl MID no es un plan detallado, lo cualcaracterizó a la era de la planificación yla era del ajuste. Su objetivo es poner enmarcha un proceso que se adaptecontinuamente al cambio decircunstancias. En consecuencia, la

adaptación debe reemplazar a lamentalidad de “modelo único”.

Para ello, el Banco debe esforzarsemás para aprender y escuchar. El MIDdestaca el papel fundamental quedesempeñan el conocimiento y el

aprendizaje en el desarrollo. Condemasiada frecuencia se supone que elBanco tiene todas las respuestas, y queel único problema es vender lasolución preferida a los clientes. Peroel Banco no es un almacén de

conocimientos de aplicación universalen espera de ser transferidos. Por esarazón, el Banco debe dar a sus clienteslos medios para aprovechar losconocimientos, poner en contacto asus clientes entre sí y con otras fuentes

de experiencia, y sacar conclusionescon ellos acerca de lo que daresultado.

déterminant que la diffusion dusavoir et le développement desconnaissances jouent dans ledéveloppement. Trop souvent,

on part du principe que laBanque détient la réponse etqu’elle n’a plus qu’à vendre la

meilleure solution aux clients. Mais laBanque ne stocke pas desconnaissances d’application générale

qui attendent d’être transférées. C’estpourquoi elle doit donner aux pays lapossibilité d’exploiter lesconnaissances existant à travers lemonde, mettre ses clients en rapportles uns avec les autres, leur fournir

d’autres sources d’information etapprendre, avec eux, les formules quidonnent des résultats concluants.

Pour agir de façon sélective dansle partenariat, il faut être constammentprêt à s’adapter. La direction de la

Banque a inscrit de nombreuxchangements dans le Pacte stratégique.Mais l’institution devra continuer àévoluer pour mieux circonscrire sonaction et améliorer l’impact de sesopérations sur le développement. Cela

passe par de nouvelles compétences,de nouvelles approches, de nouvellesincitations et un nouvel état d’esprit,autant d’éléments qui permettront lanaissance de véritables partenariats. Etsurtout, la Banque devra faire preuve

de réalisme et faire des choixstratégiques pour tenir compte del’évolution de son avantagecomparatif dans un environnement enpleine mutation.



xxii

Para poner en práctica la

selectividad mediante lasasociaciones se requeriráncambios continuos. Laadministración del Banco haemprendido muchos cambiosen el marco del Pacto

Estratégico. Serán necesariosotros cambios para definir mejor elenfoque del Banco y para aumentar laeficacia de sus operaciones entérminos de desarrollo. Para ello serequerirán nuevas aptitudes, nuevos

planteamientos, nuevos incentivos ynuevas actitudes, para dar cabida aauténticas asociaciones. Por encima detodo, el Banco deberá ser realista yestratégico con respecto a sucambiante ventaja comparativa en un

entorno en rápida evolución.

ES

PA

NO

L

xxiii

AFR - Africa RegionAPL - Adaptable Program LoanAPPI - Aggregate Project Performance IndexARDE - Annual Review of Development EffectivenessBP - Background PaperCAE - Country Assistance EvaluationCARE - Cooperative for Assistance and Relief Everywhere, Inc.CAS - Country Assistance StrategyCDF - Comprehensive Development FrameworkCG - Consultative GroupCGIAR - Consultative Group on International Agricultural ResearchDAC - Development Assistance CommitteeDANIDA - Danish Cooperation AgencyDFID - Department for International DevelopmentDGO - Director-General, Operations EvaluationEAP - East Asia and Pacific RegionECA - Europe and Central Asia RegionESSD - Environmentally and Socially Sustainable Development NetworkFERD - Foreign Economic Relations Department (Vietnam)GDI - Gross Domestic InvestmentGDP - Gross Domestic ProductGDS - Gross Domestic SavingsHIAL - Higher-Impact Adjustment LendingICR - Implementation Completion ReportID - Institutional Development (Impact)IDA - International Development AssociationIDS - Institute of Development Studies at Sussex UniversityLCR - Latin America and Caribbean RegionLIL - Learning and Innovation LoanM&E - Monitoring and EvaluationMENA - Middle East and North Africa RegionNAFI - National Agency for Foreign Investment (Kazakhstan)NEP - New Economic Program (Bolivia)NGO - Nongovernmental OrganizationODI - Overseas Development InstituteOECD - Organization for Economic Cooperation and DevelopmentOED - Operations Evaluation DepartmentPREM - Poverty Reduction and Economic ManagementQAG - Quality Assistance GroupRBM - Results-Based ManagementSA - South Asia RegionSDC - Swiss Agency for DevelopmentSPA - Special Program of Assistance to AfricaUN - United NationsUNDP - United Nations Development ProgramUNICEF - United Nations Children’s FundUSAID - United States Agency for International DevelopmentWBI - World Bank InstituteWDR - World Development Report

abbreviations and acronyms

1

11TOWARD ACOMPREHENSIVEDEVELOPMENT STRATEGY

Global development during the 1990s presents a paradoxical picture. Markets and

democratic principles continue to extend their reach, and global markets for trade,

finance, and information are increasingly integrated. These trends should have helpeddeveloping countries to accelerate their growth and reduce poverty. Yet the fight against poverty

is being lost, and the efficacy of the development assistance system is being questioned.

A necessary condition for poverty reduction—growth in per capita incomes—is not being achieved bya majority of developing countries:

• During 1985–95, the aggregate of low-incomecountries (excluding China and India) experi-enced a decline in per capita income of 1.4percent a year—and middle income countries, adecline of 0.7 percent.1

• The number of poor people living on less thanUS$1 a day rose from 1,197 million in 1987 to1,214 million in 1998. Excluding China, thereare 100 million more poor people in developingcountries than a decade ago.

These numbers have led to much soul-searching inthe development community. The United Nations De-velopment Programme (UNDP) has adopted humandevelopment as its focus. The Organization for Eco-nomic Cooperation and Development (OECD) Devel-opment Assistance Committee (DAC) has formulated a

set of international goals for the first decades of thenext century that stress poverty reduction, education,health, and the environment. Most bilateral agencieshave realigned their policies to emphasize these priori-ties. OED’s evaluations (ARDE 1998) have stressed theimportance of looking beyond projects, getting clientsto own and broaden their programs of reform, coordi-nating the Bank’s work with that of other donors, andinjecting more accountability into the developmenteffort—both for countries and for donors.

The Comprehensive Development Framework(CDF), proposed by President Wolfensohn in January1999, pulls together many strands of thought withinfour overarching themes:

• Holistic. A development strategy has to gobeyond macroeconomic management and incor-porate governance, human, and social develop-ment objectives.

• Ownership. The country has to be in the driver’sseat in formulating and implementing develop-


2

ment strategy, and it must involve the privatesector and civil society.

• Partnership. Donors and multilateral institutionsshould harmonize theirprograms and prac-tices, concentrate onareas of comparativeadvantage, and workwith country partnersin a framework of mu-tual accountability.• Results-orientation.

Success has to be judged by progress on theground toward equitable and sustainablegrowth.

The principles of the CDF differ significantly fromthose of the planning and adjustment eras (table. 1.1).The planning era focused on investment, and neglectedpolicies. The adjustment era focused on policies, andneglected public investment. Both neglected institu-tions. By contrast, the CDF emphasizes institutions,including governance, the judicial framework, finan-cial intermediation, and social capital creation. Itdraws on both old and recent lessons of development.

Long-Term, Holistic, BalancedA recurrent theme of OED reports has been thatinstitutions matter (Picciotto and Wiesner 1998). Evalu-ation evidence confirms the need for a broad-basedconcept of development, including effective public

TABLE 1.1 THREE DEVELOPMENT PARADIGMS

Comprehensive DevelopmentPlanning Adjustment Framework

• Pervasive market failures• Government-led development

• Centrally driven; detailedblueprints

• Investment-led development• Resource allocation by

administrative fiat

• Planners and engineers dominant

• Industrialization with importsubstitution

• Donors fill resource gap

• Donors place foreign experts

• Marginal role for monitoringand evaluation

• Pervasive government failures• Market-led development

• Short-term adjustments

• Incentive-led development• Investments and institutions

follow it

• Economists and financial expertsdominant

• Liberalization and privatization

• Donors determine resourceenvelope

• Donors impose policies

• Donor-driven monitoring of policyimplementation

• Situation-dependent failures• Country-led development through

partnerships

• Long-term vision, socialtransformation, adaptive learningprocess

• Investment, incentives, andinstitutions considered jointly

• Multidisciplinary approach

• Liberalization, regulation, andindustrial policy to match statecapability

• Country drives aid coordinationbased on comparative advantages

• Donors provide advisoryassistance to empowerstakeholders with options

• Participatory monitoring andevaluation to enhance learning andadaptation

The principles of theCDF differ significantly

from those of theplanning and

adjustment eras.

3

To w a r d a C o m p r e h e n s i v e D e v e l o p m e n t S t r a t e g y

expenditure management, suitable institutional capa-bilities, and safety nets. Recent research also highlightsthe value of a comprehensive approach to createvirtuous circles (Dollar 1999; Thomas and others1999). Governance and institutional quality are key togrowth, poverty reduction, and project success (Evansand Battaile 1998; Buckley 1999).

Poverty is multidimensional and location-specific.It can be reduced only through programs tailored to thelocal context (Narayan 1999; World Bank 1999l). AnOED case study of health determinants suggests theneed to go beyond the confines of the health sector—transport, pollution, communications, the environ-ment, and education must all be considered to getresults (World Bank 1999b). Similarly, getting girlsinto schools and keeping them there transcends theeducation sector—concurrent investments in transport,water, sanitation, and domestic labor-saving technolo-gies may also be necessary.

OwnershipLack of ownership has compromised the developmenteffectiveness of many Bank-supported operations (WorldBank 1998e, 1998f, 1998m, 1999c). Projects have tendedto perform poorly when they were prepared by outsiders,did not engage stakeholders and beneficiaries, exceededlocal implementation capacity, and did not engenderborrower commitment. Success in both investment andadjustment operations has been associated with owner-ship and beneficiary participation.2 It is important tonurture ownership among senior government officialsinvolved in negotiations, civil servants concerned withimplementation, and those affected by the intervention, aswell as among Bank staff (World Bank 1998m).

Decentralization involves risks, but can help im-prove ownership (World Bank 1999o). Over time,decentralized systems create commitment to reform aslower-level officials, accountable to elected officials,move to higher posts, bringing with them enthusiasmfor pluralistic, consensual, and responsive modes ofgovernance (Crook and Manor 1999 BP).3 Decentrali-zation also broadens the scope for partnerships amonglocal government institutions, civil society, andgrassroots communities.

Partnership and Development CooperationSpurred by global economic and political change,development cooperation is undergoing fundamental

changes in its rationale, strategy, and mode of opera-tion. The development assistance system is too frag-mented and onerous, particularly for poor and weakcountries. Recent studies have concluded that there isan urgent need for a country-led partnership approachto development assistance (World Bank 1998i, 1999m).An OED study recommends that the Bank “let go” inareas where it does not have a comparative advantage,and concentrate instead on mainstreaming promisingprograms, relying on other agencies to pilot newapproaches (World Bank 1999c). Similarly, formulticountry, grant-based collaborative programs, anOED study concludesthat the Bank shouldteam up with privatefoundations and develop-ment assistance agencieswith a long tradition ofawarding grants.

Conflicting interestsamong donors and barriers to progress within countriesmust be addressed in order to achieve reforms. TheBank can help forge strategic alliances with otherlenders and donors to overcome these constraints andnurture consensus on policy changes and capacitybuilding requirements (World Bank 1995b). Activeinvolvement by senior management and well-staffedResident Missions facilitate cooperation (Kreimer andothers 1998). The OED aid coordination study alsohighlights the need to build local capacity for effectiveaid management (World Bank 1999m).

Results, Learning, AccountabilityThe aid business has been overly focused on inputs,financial commitments and disbursements, and supply-driven technical assistance. An OED study on resettle-ment suggests that disappointing outcomes were largelythe product of reliance on “plans” rather than “results” asthe touchstone of quality management (World Bank1998k). Better performance indicators and learning pro-cesses can help manage projects and strategies towardsuccessful outcomes (World Bank 1999d).

An extensive evaluation literature provides evi-dence of the crucial role of monitoring for developmenteffectiveness (World Bank 1999d). Results orientationconnected to a learning process requires mechanisms togenerate and share knowledge, both within a countryand among partners. OED’s assessment of public

Lack of ownershiphas compromiseddevelopmenteffectiveness.


4

expenditure reviews recommends a focus on the goalsof expenditure reform from the beginning, both forproper selection of expenditure instruments and forbetter implementation (Datta-Mitra 1997).

Challenges and TensionsThe CDF principles, a regular staple of research andevaluation findings, are rarely practiced. We need tounderstand why and find ways to overcome theimplementation challenges. There is tension as well assynergy among the CDF principles. The distinctivecontribution of the CDF is to have brought these well-established principles together. Each is valid on itsown, but it is their combination that promises highrewards, as well as significant risks (figure 1.1).

The challenges arise from current incentives andcapacity constraints within developing countries and

FIGURE 1.1 INTERACTION OF CDF PRINCIPLES

Holistic,Long-term

Partnership Ownership

Results

the development assistance system. Leadership, a newauthorizing environment, and broad-based ownershipof the CDF principles must be combined with system-atic learning.

5

22CHALLENGES AT THEPROJECT LEVEL

Trend analysis shows that the quality at exit of Bank-supported development projects

has improved over the 1990s, but the most recent evaluations suggest a stalling of gains

below the Strategic Compact goal of 75 percent satisfactory.1 The quality at exit has

taken a slight downturn from a peak of 74 percent satisfactory for projects exiting the portfolio in

FY97 to 72 percent for FY98–99 exits. Moreover, institutional development impact and sustain-

ability of project benefits continue to lag.

The demandingness, complexity, and risk ofprojects have also climbed over the past decade,making the modest performance gains more impres-sive.2 But stagnant performance in borrower imple-mentation has led to a growing gap between complex-ity and capacity, contributing to the recent downturn inproject outcomes. Improving project performance willrequire greater emphasis on developing borrower ca-pacity. Lessons from project experience also point tothe continuing importance of the Bank’s appraisal andsupervisory roles.

Keeping up with ComplexityThe changing nature of Bank-financed projects must bekept in mind when considering results trends. Theexiting portfolio has become more ambitious relativeto the implementation environment (figure 2.1). Devel-opment work in traditional sectors has become morecomplex, and the Bank has expanded its activities inmore challenging areas (environment, human develop-

ment, governance). The increased complexity and themore adverse operating environment suggest that themodest improvement in results during the 1990s repre-sent a substantial improvement. But the gains in projectperformance appear stalled, highlighting the need forintensified attention to institutional constraints.

OED evaluators assess project results along threerelated dimensions—outcome at the time of evaluation,sustainability of ben-efits, and institutionaldevelopment impact.Last year’s ARDE intro-duced a new perfor-mance indicator thatcombines these three di-mensions into an aggregate project performance index(APPI). The latest evaluation results show a broad-based improvement in the aggregate performance ofexiting projects.

Across exit years, the APPI shows a leveling off in

Demandingness,complexity, and risk ofprojects have climbedover the past decade.


6

performance since the improvement from the decade’snadir in FY94, both by projects and by disbursements (seefigure A1.1). Since FY97 the APPI by projects hasremained at 6.7, and by disbursements at roughly 7.1.

The recent results reflectstability in overall per-formance, but also hidethe dynamics among thethree dimensions of theindex. A modest down-ward shift in the out-comes of recently evalu-ated projects has beenoffset by an increase ininstitutional develop-ment impact (althoughfrom a low base).

The percentage of projects with satisfactory outcomesat exit—projects that have efficiently met or exceededtheir major relevant objectives—has increased over the1990s, rising above the 70 percent threshold of the early1990s (see figure A1.2).3 The share of satisfactory projectsexiting in FY98–99 is 72 percent, confirming a movebeyond the trend for FY90–96, when outcome perfor-mance averaged 66 percent satisfactory. But while FY98is the second consecutive year in which more than 70percent of exiting projects had satisfactory outcomes, theshare is down from 74 percent in FY97, and preliminaryresults for FY99 (which may not be statistically signifi-cant) show a further decline to 70 percent. The recent

downturn in project outcomes combined with the steadyincrease in complexity suggests a possible tradeoff be-tween the two under current conditions. To ensure futureperformance gains, greater project complexity must bemet with improvements in borrower capacity and in-creased vigilance in quality assurance.

Over the short term, outcome performance isexpected to remain below Strategic Compact levels.The Quality Assurance Group (QAG) reports an in-creased share of projects exiting the portfolio withunsatisfactory ratings, including most of the projects tobe evaluated by OED over the next six months. Butamong the remaining active projects, a declining shareare at risk of failing to achieve their developmentobjectives, which is a promising signal for improve-ments over the long term.

Varied Performance among Regions, Sectors, andTypes of LendingRegions fall into three groups according to the perfor-mance of projects exiting in FY98–99 (see figure A1.3).The top performers are Latin America and the Caribbean,Europe and Central Asia, and East Asia and the Pacific—all with more than 80 percent of projects rated satisfactoryin outcome, and more than half as likely to be sustain-able. The Middle East and North Africa and South Asiarank in the middle, with roughly two-thirds of projectshaving satisfactory outcomes. Sub-Saharan Africa rankslowest on all three dimensions, despite solid improvementin project outcomes and sustainability. The Region had

FIGURE 2.1 DEMANDINGNESS, COMPLEXITY, RISKINESS, AND OUTCOME, BY EXIT FISCAL YEAR

Note: Results for FY99 are preliminary (43 percent coverage).Source: World Bank, Operations Evaluation Department data.

0

10

20

30

40

50

60

70

80

90Percent

FY90 FY91 FY92 FY93 FY94 FY95 FY96 FY97 FY98 FY99

Demandingness (% substantial) Complexity (% substantial) Riskiness (% substantial) Outcome (% satisfactory)

To ensure futureperformance gains,

greater projectcomplexity must be met

with improvements inborrower capacity andincreased vigilance in

quality assurance.

7

C h a l l e n g e s a t t h e P r o j e c t L e v e l

the second-largest gain in the share of projects withsatisfactory outcomes between FY94–97 and 1998–99,from 55 to 61 percent.

Performance has improved dramatically in LatinAmerica and the Caribbean over the past decade, andmore recently in Europe and Central Asia. The gains haveput projects in these regions on par with those in East Asiaand the Pacific, where the financial crisis has led to adeterioration in outcomes and sustainability. The LatinAmerica and the Caribbean Region has seen improve-ment in all three dimensions. The gains in institutionaldevelopment impact are particularly promising for theRegion’s future performance; 52 percent of projectsexiting in FY98–99 were rated as having substantialimpact, the highest share among all the Regions for thisperiod. In Europe and Central Asia, recently exitingprojects show solid gains in outcomes and sustainability(83 percent with satisfactory outcomes and 69 percentwith likely sustainability in FY98–99). But performancein institutional development impact has improved onlyslightly, and self-evaluations of the active portfolio in theRegion suggest a diminished development impact in thefuture. Active commitments at risk rose from 25 percent inFY98 to 47 percent in FY99, making the Europe andCentral Asia portfolio the riskiest in the Bank. Theincreased riskiness reflects a rapid deterioration in theRussia portfolio.

Outcome performance varied significantly amongsectors between FY94–97 and 1998–99, deteriorating insix sectors while improving in five (see figure A1.4).4 Butseveral of the improving sectors made tremendous gains,lifting Bankwide performance. Gains were particularlystrong in three sectors: the share of projects with satisfac-tory outcomes rose by 41 percentage points in publicsector management (to 93 percent), by 25 points in urban(85 percent), and by 17 points in transportation (93percent). The latest self-evaluation data show that publicsector management is a key problem area for the Bank,however, suggesting much lower performance in thefuture. The two best-performing sectors in the FY94–97exit cohort—social sector and multisector—both experi-enced deterioration in project performance. Performancedeclined most for multisectoral projects, with the shareshowing satisfactory outcomes falling from 83 percent forprojects exiting in FY94–97 to 65 percent for those exitingin FY98–99. The decline reflected lower than averageperformance for structural adjustment and technicalassistance loans.

Overall, adjustment loans have shown strong im-

provement in outcomes since the 1980s and early 1990s—satisfactory outcomes rose from 69 percent in FY90–93 to85 percent in FY98–99.5 Performance has improved inSub-Saharan Africa (79 percent satisfactory, up from 69percent in FY94–97) and in the Finance, Private Sector,and Infrastructure Net-work (94 percent, upfrom 74 percent). Butsectoral adjustmentlending has accountedfor the bulk of the gains;structural adjustmentlending has shown stag-nating performance.

Despite the overallimprovement in theperformance of adjust-ment lending, concerns have been expressed about itssocial and environmental content. A 1996 evaluation ofcompleted adjustment operations by OED has identi-fied a need for greater emphasis on cost-effectivemanagement of public social expenditures, along witheffective safety nets to protect the extremely poor andthose vulnerable to the effects of adjustment(Jayarajah, Branson, and Sen 1996). Managementreviews of the extent of recent progress are under way,following QAG and Environmentally and SociallySustainable Development Network (ESSD) assess-ments. Better dissemination of relevant Bank policydirectives to increase the congruence between policyand practice should be considered.

Institutional Development and Sustainability FlaggingSustainability of project benefits and institutional de-velopment impact remain major concerns (see figuresA1.5 and A1.6).6 Fewer than half the evaluated projectsexiting in FY98–99 are likely to have sustained results,a performance similar to that throughout the 1990s. Avolatile external environment has suppressed improve-ments in sustainability, particularly in East Asia andthe Pacific, because of the recent financial crisis, and inEurope and Central Asia. Operations with large dis-bursements perform somewhat better on sustainabil-ity—two-thirds of evaluated disbursements for fiscal1998–99 are judged likely to have sustained results.

An emphasis on institutional development is criti-cal if the Bank is to move beyond simple projectfinancing to long-lasting improvements in developingcountries and full ownership by borrowers. At the

Fewer than half theevaluated projectsexiting in FY98–99 arelikely to have sustainedresults, a performancesimilar to thatthroughout the 1990s.


8

project level, the Bank has been making slow progressin institutional impact. Still, only 39 percent of theexiting projects in FY98–99 show substantial institu-tional development impact. Better ways to achieve thiscrucial objective must be found.

Recent evaluations show marked improvement in theinstitutional impact of projects directed primarily toinstitutional development. Half deliver substantial im-pact, a significantly higher share than the Bankwideaverage. These focused institutional development inter-ventions, which make up only a quarter of the exitingportfolio, thus contribute disproportionately to the gainsin the institutional impact of Bank-supported projects.

Promising evidence suggests that projects can

produce an institutional development impact at thenational level.7 Among countries in which more thanhalf the evaluated projects achieve substantial institu-tional development impact, three-quarters show corre-sponding improvements in institutional quality at thenational level. This impact is strongest for projects thatadvance the regulatory framework, streamline publicsector operations, and improve the enabling environ-ment for private sector activity.

Bank and Borrower Performance Must ImproveFor this Review, OED conducted econometric analysesof the key determinants of success for investment andadjustment lending (box 2.1). The results for invest-

TBOX 2.1 UPDATED DETERMINANTS ANALYSIS PUTS SPOTLIGHT ON BANK PERFORMANCE

his Reviewrevisitedthe economet-

tory—as the dependentvariable.

Investment projects.The results for the mostrecently exited investmentprojects (FY98–99)broadly confirm theconclusion of the earlieranalyses: borrower andBank performance are themost important determi-nants of projectsuccess, with Bank super-vision and borrowerimplementation perfor-mance both crucial indetermining the finaloutcome. Other thingsbeing equal, improvedBank supervision increasesthe likelihood that aproject will be successfulby 51 percentage points—and satisfactory borrowerperformance by 43. Qual-ity at entry increases aproject’s likelihood ofsuccess by 28 percentagepoints, and borrowercompliance by 20. Bor-

rower preparation turnedout to be insignificant.

Compared with theresults for projects exitingin FY95–97, the moststriking difference is thegreater importance ofBank supervision; itscontribution to the likeli-hood of success increasedfourfold. This finding isintuitive given theincreased complexity ofBank projects and thestagnant performance inborrower implementation.

Adjustment operations.Bank and borrowerperformance also turnedout to be key determi-nants of the success ofadjustment projects,along with politicaleconomy factors.Borrower performanceduring implementation iscritical: when it is satis-factory, a project’s likeli-hood of success increasesby 50 percentage points.

Satisfactory quality atentry increases the esti-mated likelihood ofsuccess by 12 percent-age points, Bank super-vision by 13. Politicaleconomy factors alsohelp predict whether aproject will succeed: aone standard deviationincrease in theincumbent’s time inpower at the start ofthe operation reducesthe likelihood ofsuccess by 13 percent-age points, and a onestandard deviationincrease in the numberof average governmentcrises reduce it by 5percentage points. Theother politicaleconomy variable, adummy capturingwhether the incumbentwas democraticallyelected, turned out tobe insignificant.

ric analysis of thedeterminants of perfor-mance for investmentprojects using a meth-odology employed inthe 1994 and 1997ARDEs. And for thefirst time, a similartreatment was given toadjustment operations,using a methodologydeveloped in the Devel-opment ResearchGroup. The main goalwas to assess the rela-tive impact of Bankand borrower inputs onthe success of projects,while controlling forcountry factors such asthe macroeconomicenvironment and qual-ity of governance.Single-equation probitanalysis was applied,with outcome—satis-factory or unsatisfac-

9

ment projects show that Bank supervision has becomethe most important factor in determining outcomes.This makes intuitive sense, given the increased com-plexity of Bank projects and the volatility of theoperational environment. Bank and borrower inputsare both important for adjustment operations, butborrower performance is of greatest significance. Thisfinding confirms the prime importance of countryownership to successful policy reform.

For projects exiting in FY98–99, overall borrowerperformance—based on the average of preparation,implementation, and compliance with legal covenants—rose to 73 percent satisfactory, up from 68 percentsatisfactory in FY94–97 (see figure A1.7). This improve-ment in performance, consistent with last year’s findings,keeps borrower inputs on par with Bank inputs. Three offour evaluated projects now show overall satisfactoryBank and borrower inputs.

Yet while overall borrower performance is up,implementation performance continues to hover around60 percent satisfactory, and remains the weakest aspect ofborrower project processing. This stagnant performancereflects the difficulty of improving borrower capacity andunderscores the need for sustained emphasis on capacitybuilding. The determinants analysis suggests that im-proved borrower implementation would lead to majorgains in the development impact of projects.

Bank performance shows mixed results for evaluatedprojects exiting in FY98–99 (see figure A1.8). Comparedwith results for FY94–97 projects, identification perfor-mance has declined from 82 percent satisfactory to 78percent, appraisal performance has stagnated at 62percent satisfactory, and Bank supervision has improvedto 76 percent satisfactory. Disaggregating the Banksupervision measure turns even this positive result into amixed signal, however: while the share of projectsreporting satisfactory supervision was 80 percent forFY98, it dropped to 69 percent for the partial FY99sample. As the econometric determinants analysis con-firms, this drop is closely associated with the downturn inoutcomes for evaluated projects exiting in FY99.

CDF would have significant implications for Bankappraisal and supervision. Upstream activities are theentry point for several key principles of the framework.Project appraisal would include consideration of ben-eficiary participation and local capacity to ensureownership. And it must build in a focus on results fromthe beginning. Equally important, if not more so, isBank supervision, which would assist the borrower in

ensuring the flexibility to adapt to changes in theimplementation environment and in working with andbenefiting from partners. This process role is critical indeveloping a learning environment while maintaininga focus on development impact.

Project Performance and CDF PrinciplesA sample of recent project evaluations was reviewed toassess the importance of CDF principles in projectperformance.8 For investment projects, the review cov-ered all four pillars of the framework. A smaller numberof adjustment project evaluations was reviewed to assessthe balance between attention to social issues and themacroeconomic components of the reform program.

India’s Agricultural Development Project in TamilNadu provides a good example of the way synergiesamong policy reforms, institutional development, andpublic investment can help achieve broad-based, sus-tainable rural development. The project shows how acommitted borrower, with a coherent policy frameworkand decentralized project execution, can implement asuccessful, long-term, holistic program. The TamilNadu Project now serves as a model for rural develop-ment projects in other Indian states.

The Ghana Feeder Roads Project shows the benefitsof ownership and strong partnership among donors.The project aimed to help the rural poor by improvingaccess to markets and agricultural areas and bypromoting labor-inten-sive construction meth-ods to provide employ-ment. It also sought tointegrate the transportand agricultural sectorsmore closely. The De-partment of FeederRoads, the implement-ing agency, exhibitedstrong ownership andcommitment during allstages of the project. The project’s institutional devel-opment activities improved the department’s workprogramming and resulted in a thriving constructionand consultancy industry in Ghana. Partnership withthe Danish Cooperation Agency (DANIDA) and theU.K. Department for International Development wasexceptional, with periodic formal meetings enhancingcoordination.

Bank loans often supply the institutional frame-

Determinants analysissuggests that improvedborrowerimplementation wouldlead to major gains inthe developmentimpact of projects.

C h a l l e n g e s a t t h e P r o j e c t L e v e l


1 0

work for expanded activities that attract other multilat-eral and bilateral donors. This additional funding canhelp multiply project benefits and impacts, as inPoland’s Environmental Management Project.Bangladesh’s Agricultural Support Services Project andPoland’s Energy Resource Development Project, bothcarried out in partnership with another donor, illustrate

the benefits of donorcooperation. The suc-cess of these projectscan be attributed inpart to joint missions,the smooth exchangeof information, and thecoordination of techni-cal assistance.

Where coordina-tion is lacking, projectssuffer. This was the case

for Kenya’s Forestry Development Project. Originallydesigned, in collaboration with five donors, to fosterdonor coordination and promote balanced developmentof the forestry sector, the project was drastically pareddown after donors withdrew. But even the reduced projectsuffered from lack of cooperation, this time between theBank and the borrower.

Chile’s Primary Education Improvement Projectand India’s Industrial Technology Development Projecthighlight the benefits of building flexibility into projectimplementation. An adaptable design gave India’sproject the flexibility to respond to a changing eco-nomic environment, contributing strongly to its suc-cess. And in Chile’s education project, adapting instruc-tional processes in response to the results of monitoringby the Bank and the borrower ensured highly satisfac-tory outcomes. But adequate resources for effective andpermanent monitoring are seldom provided. Manyprojects, such as Mauritius’s Industrial and VocationalTraining Project, fail to design clear performancetargets for project activities—and even when targetsare identified, supervision missions often ignore them.

Adjustment operations face the challenge of bal-ancing a focus on macroeconomic issues with an equalfocus on their social implications. Panama’s EconomicRecovery Project shows how this can be done. Theproject included a poverty reduction program to miti-gate the employment effects of labor market reformsand to reduce the cost of basic consumption needs,including public transportation. The government alsoset up a social emergency fund to protect those mostsusceptible to malnutrition and disease.

Chile’s PrimaryEducation Improvement

Project and India’sIndustrial TechnologyDevelopment Project

highlight the benefits ofbuilding flexibility into

project implementation.

1 1

33CHALLENGES INSECTORAL ANDTHEMATIC PROGRAMS

T he difficulties involved in practicing the CDF principles vary across sectors (see

Annex 2). CDF principles have been easier to practice in technology-driven sectors,

such as infrastructure, than in the softer sectors, such as human development, or thesocial and structural dimensions of hard sectors. The challenges involved within and across CDF

principles reflect dilemmas the Bank has been grappling with for decades.

Five key tensions are at play: short versus longterm, comprehensiveness versus selectivity, speed ver-sus broad-based ownership, partnership versus countrycapacity, and accountability for results versus localcapacity. Resolving such tensions lies at the core ofquality management in development assistance.

Short versus Long TermMany factors contribute to the short-term orientation ofdevelopment efforts: the project approach, financialcrises, political instability, the election cycle, and theincentive systems of the civil service in developingcountries, as well as the incentives and planningprocesses of donors and the Bank. Yet the mostfundamental problems of development, such as institu-tional development and governance, require long-termstrategies and sustained efforts.

A long-term perspective is especially important indealing with the structural dimensions of reform. Amongrecent examples of failures to take the long view areprivatization in transition economies, civil service reform,

and deregulation of the financial sector (Mathieu 1998;Stiglitz 1999b; World Bank 1999c). Privatization in-creases inequality if the appropriate regulatory frame-work and environment for private sector development aremissing. In transitioneconomies, the rush tomass privatization,without establishing theunderpinnings of capi-talism, led to corruptsales, lack of restructur-ing, insider-dominatedtransactions, and un-regulated actions by in-vestment funds.

The more ambi-tious the reform, themore time and re-sources are needed to prepare the way. A long-termcommitment is essential to success. OED’s study offinancial sector reform found that countries with

Many factors contributeto the short-termorientation ofdevelopment efforts,yet the mostfundamental problemsof development requirelong-term strategiesand sustained efforts.


1 2

unsatisfactory outcomes averaged only 1.5 reformoperations dealing with the financial sector over a 12-year period; countries with satisfactory outcomes aver-aged 2.3 operations. Equally, resettlement operationscall for involvement of affected communities manyyears before the infrastructure investments take place(World Bank 1998j).

Comprehensiveness versus SelectivityMore comprehensive approaches frequently implygreater complexity and implementation difficulties forsectorwide and multisectoral programs. Thematic,cross-sectoral, and structural dimensions are particu-larly challenging for sector-bound aid agencies andgovernment ministries. For example, it took decades ofexternal pressure and top management leadership tobring environmental considerations into the Bank’soperations. Gender, public sector management, andprivate sector development issues raise similar chal-lenges (Murphy 1997). Cross-sector themes are oftenignored or resisted by sector-bound organizations, bothwithin the Bank and in developing countries.

Following the debt crisis, the Bank broadened itsview to address systemic problems. Although now morerelevant, the success rate of the more demandingfinancial sector operations dropped to 50 percent, evenprior to the 1997 financial crisis.1 An OED review of

recently closed opera-tions supporting finan-cial sector reformsshows that success isoften attributable togovernment ownershipand commitment, con-sensus-building, a fa-vorable political cli-mate, and good policydialogue (Mathieu1998).

A focus on discreteinvestments rather thanintegrated packages ofinvestments has been a

frequent feature of multisectoral programs, as noted in arecent review of social protection activities based onbeneficiary assessments. Going around ministries hasadvantages, but these are often enjoyed at the expense ofsustainability. Services provided through social fundagencies are particularly vulnerable because of the

programs’ weak links to existing government structures.The Bolivia Social Fund’s emphasis on speed and au-tonomy from line ministries worked against fittingprojects into sectoral plans. The assessments for socialfunds in Armenia, Ecuador, and Peru highlighted the needfor complementary actions, such as funding educationalmaterial, equipment, and other inputs along with infra-structure in school projects or including training for waterand sanitation projects. According to the Ecuador benefi-ciary assessment, if the social fund does not permit thefinancing of complementary works, serious considerationshould be given to not financing the project. “Inattentionto complementary requirements can . . . put at risk theimpact of the project, not to mention the satisfaction with,use, and maintenance of projects.”

Multisectoral program services have been difficult tosustain without good coordination with ongoing pro-grams of sectoral ministries. In Mali’s integrated Health,Population, and Rural Water Supply Program, imple-mented by different line agencies, the population andwater supply components were not always well coordi-nated with the health component. With the involvement ofmany agencies, programs proved difficult to monitor—and implement. The challenges of managing multisec-toral, multiagency programs are compounded by weakincentives and mechanisms for intersectoral coordination,both in countries and within the Bank.

Integrated programs may also generate tensionsbetween line agencies and oversight bodies, such asministries of finance. Activities requiring recurrentfunding can create ownership conflicts between centraland local governments, especially in non-revenue-earning operations such as highways. Implementationof a Thailand highways project, an integrated interven-tion designed to address cross-sectoral issues, wasmarred by conflicts among agencies. Implementingagencies were strongly committed to the physicalworks, but less committed to policy reforms, in partbecause policymaking rested elsewhere in government.Similarly, urban development projects involving mul-tiple sectors have provoked detrimental competitionamong oversight agencies, making them unmanage-able. While the Bank has approved no new IntegratedUrban Development Projects since 1986, the perfor-mance of some recently completed operations, such asthe Brazil Salvador Metropolitan Development, suf-fered from needlessly complex designs involving dis-parate activities implemented by too many agencies.

Sectorwide approaches are necessarily ambitious,

The challenges ofmanaging

multisectoral,multiagency programs

are compounded byweak incentives and

mechanisms forintersectoral

coordination, both incountries and within

the Bank.

1 3

C h a l l e n g e s i n S e c t o r a l a n d T h e m a t i c P r o g r a m s

complex, and demanding of Bank involvement, espe-cially supervision. In the energy sector, for example,both the sector development model and the Bank’sagenda have grown in complexity. The global movefrom public monopolies toward privatization andderegulation has required tackling a much broaderrange of issues: sector unbundling, private participa-tion, regulation, competition, interregional trade, re-settlement, environment, access by the poor, andrenewable energy sources, among others. Operational-izing a sectorwide approach through programmaticlending has had mixed results. Energy sector loans toPakistan, the Philippines, and Turkey fell short of theirobjectives because they were too complex—the FY89Pakistan energy sector adjustment loan had more than40 conditions. Phased or incremental approaches havebeen more successful. In China a succession of powersector operations of incrementally increasing policyintensity succeeded because of a realistic assessment ofinstitutional capacity, judicious use of technical assis-tance loans to build greater capacity, and effective useof economic and sector work. But a gradualist ap-proach may be difficult for the Bank to replicate incountries where the assistance strategy can accommo-date only one energy operation every few years, orwhere strong government ownership may develop onlyduring brief periods of political opportunity, as inArgentina and Bolivia.

Speed versus Broad-Based OwnershipThe lack of government ownership of reforms or commu-nity ownership of local projects has undermined develop-ment efforts. OED evaluations show that ownership isdifficult to achieve in sectors that have a broad array ofstakeholders with different interests, such as health andeducation; in thematic and structural areas, such asenvironment, rural development, and civil service reform;and in sectors such as agriculture, where resourcesprovided through state channels are under pressure.Partners may have different views of the roles of the state,the private sector, and civil society. Coalition buildingand media campaigns to overcome vested interests orhold the bureaucracy to account may not be feasible.Governance reforms may be required to institutionalizeparticipation, and this may take decades to accomplish.

Speed often compromises ownership. In a spirit ofpriority setting and capacity building, the Bank hashelped a number of governments address environmen-tal issues through National Environmental Action

Plans and programs to strengthen national and localenvironmental institutions. But a 1996 OED reviewrevealed that the environmental plans had generallynot elicited local ownership. Many of the plans wereprepared in haste and driven by deadlines that left littletime for participation.2 Making the plans a requirementfor lending further eroded country support.

The interests of different ministries (and the priori-ties of the center and the districts) can vary, and evenconflict. Uneven stakeholder commitment and weakcapacity can pose risks. OED evaluators found that thesuccess of health sector projects was significantlycorrelated with how well program designers hadassessed ownershipby key stakeholders,including concreteevidence of commit-ment. In education,too, the number of stakeholders is very large, withmany agencies and institutions involved in executingpolicies. Responsibility for selecting policy reformmeasures and deciding on mechanisms to encouragesupport must come from within the country and begrounded in broad-based support for reform. The clearimplication is that borrowers should be encouraged totake a leadership role in the preparation of projects andthat all stakeholders, including women and the poor,should be fully engaged.

Thematic strategies have a better chance of beingsuccessfully implemented when a range of public andprivate stakeholders participate. Lessons from TheGambia, Ghana, Madagascar, and Mauritius suggestthat this holds true for the environment. In Madagascarand Mauritius, national environmental action planswere able to increase local environmental capacitymore than in some other countries because of substan-tial local ownership. The more participatory planswere also successful in information gathering andpublic education.

Partnership versus Country Capacity andTransaction CostsPartnerships of external partners may be essential tocoordinate sectorwide programs and reduce demandson government capacity. OED evaluators have spot-lighted examples of effective coordination with donorsin such sectors as transport, telecommunications, andenergy. In other areas, however, partnerships have along way to go, especially in rural development,

Speed often compromisesownership.


1 4

education, and private sector development.While aid coordination can reduce demands on

government capacity, a lack of implementation capacitycan still undermine reform. An evaluation of the GhanaPrivate Sector Adjustment concluded that “required exper-tise should be on board before the process begins, andpotential legal issues, like land transfers, which provedproblematic in Ghana, should be carefully reviewed inadvance.” An evaluation of Jamaica’s Private SectorDevelopment Adjustment highlighted the need to assessborrower capacity to implement reforms, as well asborrower commitment. The borrower’s own evaluationwas highly critical of unrealistic demands by the Bank onJamaica’s weak implementation capacity.

Partnership and coordination imply high transac-tion costs. In a Ghana health operation, the Bankadopted a sectorwide approach but provided onlymarginally higher supervision resources than for aconventional investment project. This limited theBank’s ability to establish local presence, includeappropriate technical expertise in supervision missions,

or adequately partici-pate in coordinationmeetings.

A better alterna-tive may be to pro-mote strategic selectiv-ity—that is, to shareresponsibility amongdonors, to rely more

on pooled technical assistance support under govern-ment control, and to resist the inclination to take on alltasks and cover all bases.3 More intense supervision ofincreasingly complex Bank-assisted projects will un-doubtedly raise their effectiveness, but complexprojects also involve opportunity and transaction costsfor governments: the attention of senior officials isdirected to Bank missions at the cost of neglecting otherissues that may be more pressing. Donor pressure alsodiverts a disproportionate share of scarce local budgetsand staff to service a bewildering number of donorprojects.

An OED review of completed infrastructureprojects raises a number of issues surrounding the roleof external financiers:

• Too little involvement of partners and cofinan-ciers in project design and supervision

• Divergence, sometimes serious, between externalfinanciers and the Bank on stated project objectives

• Discrepancies in sector policies and proceduresfor program implementation

• Little interest in contributing to the Implementa-tion Completion Reports (ICRs), depriving thereports of the views of cofinanciers

• Problems with cofinanced technical assistance,such as reluctance of cofinanciers to replaceinappropriate technical assistance personnel

• Dropping out of the project at appraisal afterexpressing support during preparation.

Accountability for Results versus Local CapacityIn tracking the development impact of projects, the Bankhas been weak, almost across the board. Monitoring andevaluation (M&E) for results has been easier to achieve ininfrastructure lending than in structural lending (institu-tions, public and private sector development), sociallending (health, education), or thematic lending (environ-ment, rural development). OED evaluators have consis-tently identified weak M&E capacity (including trackinginputs and outputs) and the need for greater attention tosectorwide and thematic efforts. Decentralization,privatization, and weak regulation have all compromisedmonitoring and evaluation. Chile’s power sector istypical: the combination of unbundling, privatization,and weak regulation led to a deterioration in monitoringand evaluation. A clear lesson is that the Bank shouldfocus on sectorwide M&E as an integral part of itsassistance to setting up regulatory frameworks. A broaderM&E issue is that of cross-sectoral links and the dearth ofdata on the social impact of energy policies and sectorreform.

Identification of relevant indicators has been achallenge, both at the technical level (balancing processwith outcome indicators, and ensuring coverage of keyissues) and in negotiating a limited list among keystakeholders. A concern expressed in Ghana’s healthsector program was that the indicators placed greateremphasis on donor priorities than on the priorities ofnational stakeholders or consumers. In addition, institu-tional mechanisms have rarely been crafted in order tocreate incentives for increased performance at the districtand facility levels. Excessive emphasis on process indica-tors (decentralization, budget allocations) may compro-mise achievement of outcomes (as in Zambia) (WorldBank 1999d).

M&E in thematic programs has also been weak,particularly in newer dimensions such as gender,informatics, and the cross-cutting areas of environment

In tracking thedevelopment impact ofprojects, the Bank has

been weak, almostacross the board.

1 5

and poverty. The Bank has rarely used its lendingportfolio to systematically collect evidence on what worksin reducing poverty, what does not, and why. Few ruralprojects have supported those who work with the poor or

C h a l l e n g e s i n S e c t o r a l a n d T h e m a t i c P r o g r a m s

enhanced monitoring of resource allocations to the poor.Indicators used were generally input measures, such as thenumber of personnel trained or wage expenditures, ratherthan outcomes. Seldom were these gender-disaggregated.

1 7

44

T

CHALLENGES AT THECOUNTRY LEVEL

his chapter assesses Bank Country Assistance Strategies (CASs) from the perspective of

the CDF. It focuses on results—growth and poverty reduction over the long term. It also

highlights the tensions in implementing principles of the CDF at the country level andidentifies examples of good practice. The analysis covers 28 countries that have received OED

Country Assistance Evaluations (CAEs). These evaluations were prepared to fit the Bank’s CAS

schedule, but relevance, regional coverage, and importance to the Bank’s portfolio were

additional selection criteria. The sample thus includesdeveloping countries that range widely in size andincome, and achieves regional balance.1

The facts of growth and poverty in the 28 countriesbetween 1981 and 1997 are sobering (see Annex 3):

• In 40 percent of the countries, per capita incomeeither failed to grow or shrank.

• In 25 percent, the share of the population inabsolute poverty increased.

• In 23 percent, life expectancy declined.• In 54 percent, the people experienced stagnating

per capita income, rising poverty, declining lifeexpectancy, or a combination of these events.

• In 85 percent, per capita income grew 1 percent ayear or less in the 1990s.

• In 59 percent, gross savings as a percentage of GDPwere low (less than 10 percent) or declining.

• In 67 percent, investment efficiency was less than10 percent or declining.2

These findings confirm the view underlying theCDF: that the battle against poverty is being lost andthat business as usual will not accomplish the objec-tives of the development community.

Addressing Social IssuesThe CDF conceives of development as a social transfor-mation process. Increased emphasis on social servicesforms part of the new development consensus. TheUnited Nations Children’s Fund (UNICEF) and theInternational Labor Organization, along with otherobservers, have argued that the inadequate progress inreducing poverty can be traced to neglect of socialpriorities under adjustment programs (World Bank1999l). Yet the country evaluations show no significantdecline in the share of social spending in governmentbudgets during the period under study; countries suchas Côte d’Ivoire, Ghana, Morocco, Poland, and Togofully sustained their social spending.

OED’s forthcoming study on poverty shows that


1 8

many countries launched social protection and devel-opment programs during the 1990s, some with finan-cial support from the Bank. An earlier OED report onthe social dimensions of adjustment concluded that in

most countries thatimplemented adjust-ment policies, “realspending per head onhealth, education, andsocial security andwelfare programs ei-ther rose during the ad-justment period or re-bounded soon after . . .few of the Bank’s early

adjustment programs provided for safety net programs,but such programs have become more common”(Jayarajah, Branson, and Sen 1996: xi).

Social problems stem more from the inefficiencyand anti-poor bias of social programs than from lack ofbudgetary allocations. The CAS for Côte d’Ivoire notesthat the unfavorable cost-benefit ratio in education wasproduced by four factors—abnormally high wages forteachers, low internal efficiency caused by high drop-out rates, low employment potential, and inequitableaccess.3 Public spending on education typically benefitsthe wealthy at the expense of the poor, and boys at the

expense of girls. Health sector spending reflects similarbiases: more is spent on hospitals than on basic healthcenters, on curative than on preventive care, in capitalcities than in rural areas, and on wages than on basicsupplies (particularly generic drugs).

Addressing Institutional Capacity and Structural IssuesDisappointing growth is associated with neglect ofstructural issues. Many OED reports emphasize a gapbetween institutional capacity and policy reform. A1996 review highlighted the importance of institutionaland capacity development issues over and abovemacroeconomic stability, price liberalization, tradereform, and wage and interest rate liberalization.4 Thereport emphasized the need for prudential regulationand banking supervision, laws for entry and exit offirms (bankruptcy), technology development, technol-ogy transfers and licensing, arbitration mechanisms tosettle labor market disputes, labor force training incooperation with the private sector, and improvedinformation on market opportunities, particularly forexports.5

Sequencing is an important part of the CDF.Problems arise when financial sector liberalizationprecedes the development of a regulatory and supervi-sory system, or when privatization comes before asound framework for regulation and competition.

FIGURE 4.1 THE BANK’S CONTRIBUTION TO IMPLEMENTING CDF PRINCIPLES

Results Focus

Partnership

Ownership

Holistic

0 10 20 30 40 50 60 70 80 90 100

Outcomes, impact, and monitorable indicators

Development of local capacity for M&E

Tracking poverty

Tracking sustainable growth

Harmonization of donor procedures

Encouraging strategic selectivity

Encouraging stakeholder participation

Capacity building for country-driven aid coordination

Country-driven strategy formulation/implementation

Client commitment to reforms

Long-term view of client potential and constraints

Inter-sector coordination

Building capacity for governance

Balance in lending operations

Balance in policy dialogue and sector work

Percent of countries rated modest-negligible

Social problems stemmore from the

inefficiency and anti-poor bias of social

programs than fromlack of budgetary

allocations.

1 9

C h a l l e n g e s a t t h e C o u n t r y L e v e l

These problems were highlighted in several CAEs. Inaddition, OED’s forthcoming study on poverty high-lights the role of the composition of growth in povertyalleviation—rural growth, in particular, was found tobe critical in reducing poverty.

A Long Way to GoThe CDF principles imply a different approach tomanaging country assistance programs. For this report,a new calibration was tested to determine what theperformance rating of the Bank would have been if theCDF principles had been the agreed benchmark. To thisend, the principal authors of CAEs were asked to ratethe Bank’s contribution in helping client countriesimplement CDF principles on a scale of 1 (poor) to 4(excellent) (figure 4.1). These experienced evaluatorsrated the Bank’s contribution to implementing theprinciples of the CDF as modest or poor in nearly two-thirds of cases. This suggests that country assistancepractices will have to change substantially when theCDF is mainstreamed. While the principles are notnew, the Bank’s commitment to giving them priority is.

An upcoming OED report on poverty has reviewedcountry assistance strategies, and preliminary findingsindicate that (World Bank 1999l):

• More than 70 percent focus on three main policyareas—trade and tariff reform, macroeconomicbalance, and liberalization and deregulation.

• Fewer than 50 percent address other aspects ofthe structural agenda, such as regulatory poli-cies, support for the informal sector, or wage andemployment policies.

• Roughly one in five contains a good discussion ofdistribution or equity issues; one in three coversissues relating to the sectoral composition ofgrowth.

• Fewer than 40 percent emphasize policy supportfor the informal sector or wage and employmentpolices directed at removing labor market con-straints for the poor.

There has been a significant shift in Bank lending,away from traditional investments in agriculture, andtoward investments in human capital development.Agricultural lending declined from 27 percent in the1980s to 16 percent in the 1990s, while lending forhuman capital development increased from under 7percent to more than 21 percent. Lending to industry

fell significantly in the 1990s, to less than 2.5 percent ofthe portfolio in FY99. Still, few CASs explicitly addressissues of income distribution, or propose interventionsthat directly address inequality (as distinct from inequi-ties in access to services). Despite the stated goal of pro-poor growth, policy prescriptions have deviated littlefrom those of the 1980s.

OED’s 1999 review of poverty assessments findsthat nearly half do not adequately assess individualelements of the poverty reduction strategy.6 They failto address the links between poverty and such macro-economic policies as trade and exchange rate policy, orsuch sectoral issues as food and agricultural policy andrural development. They give only limited attention tolocal dimensions of poverty reduction, including in-vestment decisions. And while all assessments recog-nize the importance of labor-intensive growth to pov-erty reduction, few have analyzed this issue.

Slightly more than a third of the sampled povertyassessments achieved a high country-level impact—asmeasured by the influence on the poverty debate,policy design andimplementation, andinstitutional develop-ment in the country—while slightly under athird had a low impact.The impact of povertyassessments is stronglycorrelated with theirquality, with the ad-equacy of time for their preparation, and with thedegree of partnership, consultation, information shar-ing, and knowledge transfer at the country level. Asurvey of stakeholders shows that while six of sevenwere satisfied overall with the poverty assessment, theyviewed knowledge transfer and local partnership andconsultation as the least satisfactory aspects.

An OED study of public expenditure reviews foundthat, except in the cases of Ghana and Zimbabwe, thereviews remain primarily internal Bank documents thatevince little sense of ownership from the government.Client governments do not believe that they caninfluence the process or the output. Some perceive thereviews as an attempt by the Bank to micromanageexpenditure policies. Even where the government isheavily involved in the process, the impact on policy isoften weak because of inappropriate sequencing orgovernance constraints. In brief, implementation of the

Despite the stated goalof pro-poor growth,policy prescriptionshave deviated littlefrom those of the1980s.


2 0

CDF principles will call for continuing improvementsin country assistance practices.

Managing the Tensions in the CDFThe principles of the CDF pose important challenges.Four are highlighted here. (Chapter 5 proposes somepromising approaches to address them.)

• Ownership versus conditionality. How shouldthe apparent conflict between country ownershipand donor interest in performance (often en-forced through conditionality) be resolved? Howshould the need for ownership be reconciled withthe need for policy reform and sound develop-ment priorities when country commitment islacking?

Conditionality is widely viewed as a crude attemptto generate policy reform in exchange for grants orloans. OED studies have shown that when applied as aone-sided, coercive instrument, conditionality can becounterproductive and incompatible with ownership.Some observers have rushed to declare all forms ofconditionality a failure. This conclusion is not sup-ported by the evidence. But the principles of ownershipand partnership clearly call for reconciling the ac-countability of donors and countries through reciprocalconditionality.

• Country-led partnership versus country capacityand commitment. How should donor demands forfinancial accountability be reconciled with client-led partnership, particularly when countries lackcapacity? How can donors play to their compara-tive advantage, and thus maximize the effective-ness of the development assistance system?

Two enabling conditions for country-led partner-ship are often missing: country commitment to soundpolicies and development priorities, and institutionalcapacity to manage and coordinate aid (World Bank1999m). Most donors and international agencies, in-cluding the Bank, continue to impose barriers tocountry-led coordination. These aid-delivery transac-tions costs include: divergent and complex administra-tive procedural requirements and the numerousmissions countries must accommodate. For highly aid-

dependent countries, these burdens can add up tothousands of reports and missions each year.

• Poor accountability record versus scaling up.Despite continued emphasis by OED on theimportance of M&E for learning and account-ability, the record remains far from satisfactory.The international development goals and recentattention to governance reinforce the need forenhanced accountability for results.

Past activities to build M&E capacity have focusedat the project level to satisfy donor requirements. Theresulting lack of domestic ownership of M&E hasundermined the use of its findings and the acceptance ofperformance measurement for sound governance. Thelimited capacity created through donor-driven, project-based M&E has been dissipated at project completion.The CDF raises the bar by emphasizing learning andaccountability for results at all levels—well beyondprojects.

• Country focus versus global public goods. Withglobalization, development problems increasinglyrequire multi-country efforts and strengthened linksbetween national strategies and international poli-cies and programs. How can international effortsaimed at global public goods complement nationalefforts through the CDF?

As global integration deepens, the number of devel-opment problems calling for supranational policy re-sponses grows. These cross-border challenges arise from acombination of market failure, government failure, andsystemic failure. The challenge of overcoming suchfailures creates a new role for development assistance.The country focus remains critical, but official develop-ment finance also must help meet the growing deficit inthe supply of international public goods.

Global forces (including technological change) arecreating not only far-reaching growth opportunities,but also a host of potential problems—including capi-tal flight, financial contagion, illicit drug trade, cross-border environmental problems, the spread of disease,waves of migrants and refugees, and loss of bio-diversity and cultural heritage. Building purposefuland inclusive partnerships is as important in addressing

2 1

these transboundary development challenges as it is intackling national challenges under the CDF.

Country CasesDevelopment assistance strategies in several countrieshave gone some way toward implementing the prin-ciples of the CDF and resolving some of the inherenttensions. Six such cases offer lessons of good (andsometimes bad) practice.

AlbaniaThe Albania review highlights some important CDFlessons. While it stresses the need to be holistic, client-focused, and results-oriented, it reveals that, with aCDF approach (like most things), the devil is in thedetails. Three lessons emerge.

First, a holistic approach requires selectivity. The22 projects supported by the Bank between 1992 and1996 covered the social sectors, infrastructure, power,agriculture, the enterprise and financial sectors, andthe environment. Ostensibly the Bank was beingholistic. Despite—or perhaps because of—its breadth,the strategy lacked depth, as well as a focus on thefinancial sector and governance, which would haveproduced the greatest impact on long-term develop-ment prospects.

Second, it may not be easy to identify who is in thedriver’s seat. During 1992–96, there was strong owner-ship by the government of many key components of theBank’s program, including the initial stabilizationprogram, enterprise reform, and agricultural adjust-ment. Yet in the year preceding the crisis, the govern-ment exhibited arbitrary behavior that contributed topolarization. By the time the pyramid schemes col-lapsed in 1997, the country was poised for a conflictthat pitted the president and his predominantly north-ern supporters against groups from the south. Owner-ship by government alone is not enough; it must bebroad-based.

Third, keeping track of results without assessingsustainability can be misleading. In the early years ofBank involvement, Albania registered high growthrates (nearly 9 percent annually). Agricultural produc-tion boomed. The government successfully completed amass privatization program. By all accounts, Albaniawas the hope of southeastern Europe. By 1997, it wasapparent that the good news on growth and agricul-tural production obscured serious problems with theinstitutions governing financial intermediation and

state-society relations (particularly service delivery).Albania’s experience highlights the need to go beyondtraditional monitoring of macro and fiscal trends.

BangladeshBangladesh was thought to have little chance of survivalafter independence in 1971. The Bank, in partnership withother stakeholders, did much to help the new countrygrapple with its myriad challenges by following aholistic, long-term approach in providing support. Sincethen the country has made remarkable gains in bothmacroeconomic and so-cial development. TheBank ensured that thecountry was not just theowner of the develop-ment program, but alsoits author. It played asupportive role in keysector programs—foodproduction, family plan-ning, and education—within the framework of a comprehensive developmentprogram prepared by the country. The Bank encouragedbroad participation and active NGO involvement informulating and implementing programs. Finally, itallocated adequate resources to the country assistanceprogram and contributed to improved aid coordination.

BoliviaThe case of Bolivia demonstrates the need to go beyondimproved macroeconomic management to achievepoverty reduction. Although the New Economic Pro-gram (NEP) launched in 1985 with strong Bank anddonor support was remarkable in achieving stabiliza-tion of the economy, it did little to achieve povertyreduction. Social and structural reforms lagged until1993–97, when a “made in Bolivia” program tochange the role of the State was launched; it includedhealth, education, and income-generating programs. InAugust 1997 the new government initiated a participa-tory process known as the national dialogue, andcrafted a comprehensive framework for development,presaging many features of the CDF.

The Bolivian experience offers three lessons withbroad application. First, ownership of the reformprogram evolved slowly, finally extending to all majorsegments of civil society after 14 years—after thebenefits of the early reforms had become apparent.

Despite—or perhapsbecause of—itsbreadth, the strategylacked depth, as wellas a focus on thefinancial sector andgovernance.

C h a l l e n g e s a t t h e C o u n t r y L e v e l


2 2

Second, establishing partnerships among aid agencieswas a huge challenge. Many donor-financed consult-ants took line management rather than advisory posi-tions, which reduced, rather than reinforced, ownershipof reforms. Finally, vested interests often stalled thereforms. The struggle against corruption and towardreform of the civil service, Customs, and the judiciaryfaced many difficulties that could be resolved onlythrough the continued involvement of all developmentparticipants.

EthiopiaWith a GNP per capita of US$110 and the vastmajority of its 60 million people living on less thanUS$1 a day, Ethiopia is one of the world’s poorestcountries. Yet since 1991, when the country opted for areformist approach to development, it has made im-pressive economic gains. The Ethiopian experiencehighlights the importance of the commitment of keystakeholders, and their ownership of development

programs, as well asthe need for relation-ships built on trust. Be-cause of the strong re-lationship between thecountry’s leadershipand the Bank team, theBank sometimes ac-cepted government po-sitions (on the size ofsector programs, forexample), although

not entirely convinced they were the correct choice. Animportant lesson: policy advice is more likely to beeffective if the client feels that its views are considered,and the tone is neither hectoring nor patronizing.

IndonesiaIn Indonesia, the Bank’s long-standing strategy success-fully supported many of the development outcomessummarized under the CDF. Growth was substantial,comprehensive, and balanced. Economic managementkept the economy on track through boom and bust.Dutch disease was avoided. The use of oil windfallrevenues was spread over time, and they were plowed

into agriculture, education, and social and transportinfrastructure. When oil revenues dropped, demandrestraint combined with a flexible exchange rate policystrengthened incentives for non-oil exports. Over time,the dependence on oil declined, poverty was reduced,and social services improved. There was also goodpartnership and strong government ownership; loanconditionality was no longer essential.

In the years leading to the crisis, complacency setin, and both the Bank and government ignored gover-nance weaknesses, financial sector dysfunctions, and afragile social situation. The eventual crisis, rooted inthe regional recession, declining export markets, andexcessive foreign borrowing, exposed cracks in theBank’s assistance strategy.

UgandaThe CDF emphasis on client ownership, a holistic andlong-term approach, and stakeholder partnership is notnew to Uganda. The country team focused on many keyconcerns of the CDF as early as 1987. Havingidentified a group of reformers in government, theInternational Development Association (IDA) sup-ported the government’s rehabilitation program. Inaddition to financial support, IDA assisted with policydialogue, training, and technical assistance, not only toidentify reforms, but also to set out the analyticalrationale in favor of market reforms and establishbenchmarks for the measurement of progress.

The Bank adopted a holistic approach. Well awarethat the success of the macroeconomic reform programhinged on the adoption of far-reaching reforms in boththe financial sector and the key agricultural sector, theBank reinforced the government’s reform program witha battery of macroeconomic and sector adjustmentoperations and technical assistance credits. The Bankalso made extraordinary efforts to keep donors in-volved in reform through co-financing and technicalassistance.

Since FY92, the country’s reform program has beentransiting from macroeconomic adjustment programsinto a plethora of microeconomic, sector, and thematicstrategies. While holistic, these strategies are difficultto pursue simultaneously and pose a challenge toconsensus-building.

The Ethiopianexperience highlightsthe importance of the

commitment of keystakeholders,

as well as the need forrelationships built

on trust.

2 3

55PROMISINGAPPROACHES

This chapter outlines some promising approaches to address the challenges and tensions

at the project, sector, and country levels (box 5.1). These approaches are synthesized

from the lessons of experience and evaluative evidence presented in the backgroundpapers prepared for this report. More systematic learning is needed to enrich the tool kit.

Learning Process, not BlueprintsThe blueprint approach has been a common pitfall. Itseems to simplify decisionmaking and reduce uncer-tainties, while it imposes standard solutions in poorlyunderstood sociopolitical contexts and varied localrealities and ignores social capital and local institu-tions, sustainability and learning, and the capacity toadapt during implementation.

OED lessons indicate that adaptive experimenta-tion and sustainable learning through multiple initia-tives yield better results than a one-size-fits-all bestpractice blueprint.1 A learning process means startingsmall, building in a bias for action, avoiding newbureaucracies, and supporting multidisciplinary teams.It emphasizes flexible, evolutionary, participatory,goals-oriented, and client-driven processes. It calls forthinking thematically and managing across sectors,without undercutting professional rigor and account-ability. Building capacity to learn and act strategicallyis as important as preparing plans.

Inevitably a tradeoff between detailed analysis andup-front design and adaptation during implementation isinvolved. The new approach implied by the CDF requires

a significant shift of resources from program design tosupervision, implementation support, and participatoryM&E. A fundamental lesson from both developmentplanning and corporateexperience is that strate-gies emerge from con-tinuous interactions oftop-down and bottom-up learning processes.Top-down approachesare best combined withbottom-up approachesto enable local initiatives, identify and scale up successes,promote stakeholder commitment, and ensure learningand adaptation over time.

Adaptable Program Loans (APLs) embody someaspects of the learning process approach—agreementon long-term objectives and broad directions, withdesign limited to start-up processes and institutionbuilding. A learning process was embedded in Brazil’sWater and Sanitation for low-income settlements (thePROSANEAR Program). Design was demand-basedand iterative, shaped during implementation by benefi-

Building capacity tolearn and actstrategically is asimportant as preparingplans.


2 4

ciary participation, feedback, and learning. The pro-gram developed partnerships among residents for theselection and management of water and sanitationsystems. Community mobilization and group decision-

making were carriedout differently in eachcommunity, dependingon factors such as lev-els of social cohesionand organization. In

an adaptive learning approach, the program encour-aged ongoing evaluation of each community’s experi-ence for rapid feedback to the next subproject.

Managing Complexity by SequencingProject complexity has been rising since the early1990s (Chapter 2). Managing the tradeoffs betweencomprehensiveness and selectivity requires an under-standing of the sources of complexity. Bank thinkingabout development has become more complex. Insuffi-cient “upstream” sector work before project prepara-tion and appraisal leads to inadequate selectivity.Fiduciary requirements are becoming more demand-ing. There is pressure to add components to deal withthe Bank’s expanding agenda and survive the internal

approval process. Career and budgetary incentivesencourage the design of large projects and discouragepriority setting (World Bank 1998a). At the countrylevel there is often a desire for large resource transfers,full-scale national coverage, and “empire building” bythe implementing agency.

One way out of crippling complexity is to take along-term view by sequencing a series of projectswithin a long-term strategy that builds on past learn-ing. The real issue is often premature complexity.Projects that build on past learning and are integratedinto existing practices can be complex and successful,as the Bangladesh Population Project and the BrazilHealth Program show.2 “Repeater” projects havehigher rates of success, especially when carried outwithin a strategic and long-term framework.

Another promising approach is to phase in cover-age, and to expand geographical scope in line withgovernment capacity to manage policy reform andimplement the program. Accordingly, growth andpoverty alleviation interventions should be pilotedregionally, and progressively tailored and broadened.This regional-focus approach is enhanced by explicitattention to capacity building and decentralization, asimplied in the Bank’s Rural Vision to Action Strategy.

BOX 5.1 CHALLENGES AND PROMISING APPROACHES

Challenges and tensions Approaches

• Learning process, not blueprints• Managing complexity by sequencing• Comprehensive analysis and selective actions• Sectorwide approaches tailored to countries and

sectors

• Adaptable conditionality for a mutualcommitment process

• Building consensus• Broadening participation

• Information for accountability and learning• Capacity building to manage for results

• From aid coordination to developmentpartnership and capacity building

• Links between global and country strategies

• Short versus long term• Comprehensiveness versus selectivity

• Ownership versus conditionality

• Speed versus broad-based ownership

• Accountability for results versus local capacity• Poor accountability record versus scaling up

• Partnership versus country capacity andtransaction costs

• Country focus versus global public goods

One way out of cripplingcomplexity is to take a

long-term view.

2 5

P r o m i s i n g A p p r o a c h e s

Learning and Innovation Loans (LILs) offer a suitabletool for managing complexity by starting small. Ad-equate supervision resources should be available toensure learning during implementation.

Comprehensive Analysis and Selective ActionsComprehensive analysis can be combined with strate-gic investments. It is best conducted with key partnerswith a view to exercising selectivity in line with apartner’s comparative advantage and country capacity.

Participatory approaches can be used to enablepoor people to analyze their realities, express theirpriorities, and make explicit their choices and tradeoffsto improve selectivity and results. The ParticipatoryPoverty Assessment process holds promise, as evi-denced by its recent application to the Ghana andUganda programs, where it brought the realities andpriorities of the poor to the attention of nationalpolicymakers. For participatory approaches to lead tosuperior selectivity and results, participants must haveaccess to information, options, and learning experi-ence, as social fund programs have shown.3

An example of this approach is the sustainablelivelihood approach, which emphasizes people-cen-tered development in a holistic framework. Sustainablelivelihoods provides an analytical structure for under-standing the factors that influence poverty and toidentify where interventions might best be made. Thisapproach has been adopted by the U.K. Department forInternational Development, the United Nations Devel-opment Program (UNDP), and CARE, among othergroups. It proposes an integrated and dynamic way ofunderstanding poverty and thinking about poorpeople’s livelihoods—the capabilities, assets (human,natural, financial, social, and physical), and activitiesrequired to earn a living (see Annex 4). The approachbuilds on what people have and how they live theirlives to add to their accumulation of assets and removebarriers to the realization of their own livelihoodchoices.

Central to this approach is a recognition of people’sdiverse livelihood goals—such as health, income, orreduced vulnerability—and the complex householdstrategies adopted to meet them. Strategies are drivenby preferences and priorities shaped by vulnerability toshocks and seasonal variations. Options are alsoinfluenced by structures (the form and organization ofgovernment and the private sector) and processes(policies, laws, institutions).

Sectorwide ApproachesA sectorwide approach provides an opportunity to shiftattention from inputs to monitoring against agreedintermediate and outcome indicators. The Ghana andBangladesh sectorwide approaches are good examples.They have been supported by two programmaticinvestment instruments: sector investment and mainte-nance loans and APLs. The sector investment andmaintenance loan is intended to bring sectoral invest-ments, policies, and performance in line with economicpriorities and to ensure efficient operation and mainte-nance of investments. The focus is on institutionalcapacity to plan, implement, and monitor investments(Jones and Lawson 1999 BP).

APLs are especially well suited to the support ofsectorwide approaches, particularly in sectors wherefinancial flexibility is needed. A recent review con-cluded that several operations would probably not havebeen brought to the approval stage without this newinstrument, because of difficulties in accurately predict-ing activities, costs, implementation arrangements,and results beyond three or four years.

Moving from projects to a full-scale sectorwideapproach (with pooling of donor finances) is risky if doneprematurely (Jones andLawson 1999 BP). Sucha change takes timeand systematic capac-ity building. Its paceshould vary accordingto the quality of macro-economic and publicexpenditure manage-ment, sector-level poli-cies and resources,quality of sector-levelmanagement, degree ofaid dependence, andother country- and sec-tor-specific factors. Because sectorwide approaches addto program complexity for donors, more resources arerequired for supervision and implementation assistance.Risks should be managed by supporting capacity build-ing, setting clear performance targets and safeguards,strengthening financial accountability, and emphasizingM&E.

The fiduciary risks are higher for sectorwide orprogrammatic lending than for project lending, sincemore fungible forms of financing are included (SIDA

The fiduciary risks arehigher for sectorwide orprogrammatic lendingthan for project lending,since more fungibleforms of financing areincluded. But these risksshould be balancedagainst the costs ofbusiness as usual.


2 6

1999). But these risks should be balanced against thecosts of business as usual. The proliferation of projectsputs an enormous burden on weak administrations,often undermining local capacity building andsustainability.

Adaptable ConditionalityConditionality should be understood as a credibleindicator of commitment by the Bank and its partners,not as an attempt to force externally designed policychanges on unwilling governments (see Annex 5). TheBank and external partners act as enabling agencies tosupport the country’s motivation for a reform processthat is guided by genuine learning from successes andfailures (Branson and Hanna 1999 BP). This kind ofconditionality is agreed and consensual. It represents apolicy compact based on mutual commitment.

Support for conditionality as a commitment pro-cess comes from a reevaluation of the Dollar-Svenssondata (in World Bank 1998o), using the country as the

unit of observation,and a recent study onhigher-impact adjust-ment lending (HIAL)(Dollar and Svensson1998; World Bank1999i). When the dataare re-analyzed usingthe country as the unitof analysis, past suc-cess becomes a highlysignificant predictor of

future success. This result supports the view of condi-tionality as a process of mutual commitment, since theBank can refer to lending history in formulating futureconditionality. Such conditionality is adapted as acountry increases its ownership of reform, assisted bycapacity building to achieve parity in the relationship.Ongoing reformers can be offered the option of ex postconditionality, while credible new reformers mightchoose to adopt floating tranche loans, as in the HIALapproach in Africa.

In this context, conditionality is part of a policyreform compact: the Bank and the borrower firstdevelop and then nurture mutual trust and commitmentas reform proceeds. Conditionality is the Bank’s side ofa continuing relationship, while ownership is thecountry’s side. A model for this approach is therelationship between a commercial bank and its cus-

tomers: as long as the customer projects a credible pathof earnings, lending continues.

Assessing ownership should lead to selectivity.Energy lending to reforming states in India in the 1990sis a promising example of selectivity used to reinforcereform. Taking advantage of India’s decentralizeddecisionmaking to demonstrate the benefits of reform,the Bank shifted its policy dialogue from the federal tothe state level and engaged only reforming states.Ownership and partnership were strengthened by wait-ing until the climate was right. The Bank haltedlending to the power sector in India for three years,until it found evidence of real ownership of policyreform in selected states. With other partners, it thenengaged in capacity building to solidify local owner-ship. Subsequent lending produced more sustainableresults than the earlier approach.

Time to Build ConsensusMobilizing the support of beneficiaries cannot startearly enough. Pilot projects do not always proceedsmoothly, even when rooted in strong communitysupport. Flexibility is essential, along with a willing-ness to listen and develop a program incrementally,informed by lessons learned.

Recent irrigation operations in India, the Philip-pines, and Turkey show the importance of allowingtime for interventions to take effect on a sociallyappropriate scale:

• In Andhra Pradesh, India, in the early 1990s,irrigator groups were formed around pipe com-mittees of 20–100 farmers. This group sizeallowed the local elite to continue to dominateand led to water allocation disputes among pipecommittees. The democratic election of muchlarger groups in the late 1990s overcame thisproblem.

• In the Philippines, large national irrigationschemes were effectively no more than fee-collection groups for the government agency andhad limited responsibility for operations andmaintenance. Water-user groups were more suc-cessful in small communal irrigation projectsthat had more autonomy.

• In Turkey these lessons were taken into account.Efforts were made to build a consensus amongstakeholders—a process greatly facilitated by theWorld Bank Institute—before irrigation systems

Conditionality is part ofa policy reform

compact: the Bank andthe borrower firstdevelop and then

nurture mutualtrust and commitment

as reform proceeds.

2 7

were turned over to water-user groups. Largergroups were also more likely to be financiallyviable and could be built around existing institu-tions, such as municipalities.

Broadening ParticipationHow should participation be broadened across interestgroups and scaled-up to the national level? Extensiveevaluation and research findings point to severallessons (see Annex 5):

• Integrate a learning process. A well-knownsuccess in broadening participation in a govern-ment bureaucracy is the Philippines NationalIrrigation Administration, which adopted a step-by-step approach to building the capacity of userassociations to manage local irrigation systems.

• Beware of procedural inflexibility. Changingorganizational systems and procedures to facili-tate participatory development is a complexundertaking. But the costs of not doing so can beheavy, as in the government of Uttar Pradesh’sDoon Valley Project in the Himalayas (Shepherd1999).

• Avoid rushing. Participation cannot be rushed—and the larger the scale, the greater the risk.When the Indonesian government tried to insti-tute nationwide village development planning inless than a year in 1995–96, there was little or noownership of the process at the village level. Toolittle time was spent building partnerships withNGOs, whose skills might have enabled thegovernment to do a better job.

• Persevere. The district-level Rural IntegratedProject Support Program in the Lindi andMtwara regions of Tanzania has evolved into aholistic program involving local government,agriculture, natural resource management,transport and marketing, education, savings andcredit, health, and water. This broad participa-tion emerged from the accumulated experienceof repeated mistakes and learning over 25 years.

• Identify a champion. A pronounced shift towardparticipation by government or NGOs can oftenbe traced to a single leader or strong alliance ofsupporters. Taking the lead, these championshave often battled institutional inertia and politi-cal pressures from wealthy elites.

• Change attitudes and behavior. For participatory

approaches to work, attitudes and behavior mustchange among all actors and at all levels(Blackburn and Holland 1998). The bottom line inparticipatory change is individual and personal.

The CDF expands the ability of the Bank to betterdeal with the difficult governance issues that ofteninvolve stakeholders other than government. For ex-ample, in the area of clean government and improvedpublic financial accountability, the Bank’s concernsextend well beyondgovernment, towardother elements of thestate, such as civil so-ciety, the media, andthe watchdog agenciescharged with monitor-ing and protecting thepublic interest underthe laws of the country. The judiciary and legislaturealso take on new emphasis in capacity building.

A recent OED study finds mixed results for efforts tobroaden NGO interventions (Gibbs, Fumo, and Kuby1998). Many NGOs are reluctant to increase the scope oftheir operations or to enter large-scale partnerships.Scaling-up can pose a challenge to an NGO’s manage-ment capacity and create obligations to members that aredifficult to sustain. Any scaling-up initiative must bepreceded by an assessment of capacities and commitment.A critical step is to involve all stakeholders in developingperformance indicators, a process that clarifies expecta-tions and priorities.

Information for Accountability and LearningTargets for monitoring progress are effective tools forguiding decisionmakers during implementation. Recenteducation projects have invested heavily in setting andmonitoring targets. The Mozambique Second Educa-tion Program used indicators to help sustain theoperation after the credit closed. It used targets set atappraisal and added new ones to establish a five-yearplan agreed with the borrower. It also identifiedoutstanding policy and implementation issues, leadingto agreement on remedial actions.

Experience in health, nutrition, and populationprojects also shows the importance of effective M&Edesign. Selecting a limited number of well-chosenindicators and attending to capacity for data collectionand analysis increase the focus on results and the

For participatoryapproaches to work,attitudes and behaviormust change among allactors and at all levels.



2 8

likelihood of achieving development impact. Strength-ening borrower systems for the collection, analysis,and use of health information in policymaking takestime. It requires attention and resources during pro-gram design and implementation and strengthenedincentives to achieve results and use the information.

Capacity Building to Manage for ResultsDeveloping a results-oriented public sector is a keychallenge in many developing countries (Annex 6). Aninitial focus on performance monitoring in selectedsectors or ministries can create a demonstration effect,easing the way to broader application to programevaluation as opportunities arise (Marcel 1999).

The emphasis on managing performance forprojects and programs is being complemented by awider focus on governance and intersectoral synergies.Dissemination of M&E information on governmentperformance can support the active involvement ofcivil society in the assessment of that performance.Civil society also has a role to play in influencing theevaluation agenda.

Key stakeholders such as NGOs, the media, andparliaments—particularly those representing and em-powering the poor—gradually learn how to use perfor-mance concepts and tools and to understand theirlimitations and weaknesses. Donors and governments

can help build such ca-pacities among thesestakeholders.

P a r t i c i p a t o r yM&E is one step to-ward building capacityto learn and managefor results and apply-ing the CDF principles

(Estrella and Gaventa 1998). As multiple stakeholderswork together to develop indicators of success, theirdiffering expectations and priorities are brought intothe open. Stakeholders must then negotiate to develop amore generally applicable framework, thereby build-ing ownership in the outcomes and reflecting partner-ship in the assessment itself. Participatory M&E alsoallows the tracking of holistic goals, both tangible andintangible. Developing large-scale participatory M&Erequires skills unlike those needed for traditionalevaluation. Stakeholders need help to acquire skills inindicator development, monitoring, facilitation, andconflict resolution.

From Aid Coordination to Development PartnershipEffective aid coordination guides countries and donorstoward agreement that all partners will accept mutualresponsibility and distinct accountability for develop-ment outcomes. For countries, this requires commit-ment to developing sound policies and effective institu-tions. For donors, it calls for adoption of a developmentorientation, the exercise of selectivity that reflectscomparative advantage, accommodation of country-led efforts to achieve coherence and selectivity, and theprovision of effective capacity-building assistance tocreate a level playing field among partners.

Although many countries have expressed a strongdesire to take the lead in aid coordination, only a few,such as the Republic of Korea, Malaysia, and Thai-land, can be said to have fully assumed the role. Manycountries lack the capacity to take the lead, and somestill lack the commitment and resolve to do so. Thetime has come for a real change—rapid, but deliberateand finely executed. A logical and critical step is for theBank—in consultation with other donors and coun-tries—to make country responsibility a reality wher-ever this is feasible. The Bank’s chief role would be tosupport country leadership and to help build thecapacity needed to exercise that leadership effectively.By giving the country a chance to exert leadership, theBank would be better positioned to assist in buildinglong-term capacity, in concert with the principles of theCDF and the OECD/DAC (see Annex 7).

The costs of poor partnership and inefficient aidcoordination—in decisionmakers’ time, gaps in assis-tance, and distortions in country priorities—are borneprimarily by developing countries. For donors and theBank, the tensions between practicing partnership andreducing the transaction costs involved may be high in theshort term. There are up-front costs, but they should beviewed as long-term investments in building the infra-structure of partnership and creating skills, trust, andlearning. And the costs are likely to decline sharply ifpartnerships are pursued selectively and strategically. TheM&E of partnership and coordination can be strength-ened to assess the cost-effectiveness of different ap-proaches. Much can be learned from the experience ofprivate firms in selecting and building strategic alliances(Doz and Hamel 1998; Picciotto 1998).

Linking Global and Country StrategiesMission-oriented transnational networks addressinghighly visible and urgent human priorities—such as the

Participatory M&E isone step toward

building capacity tolearn and manage for

results.

2 9

Onchocerciasis Control Program and the ConsultativeGroup on International Agricultural Research—canserve as a rallying device for contributing partners.Selectivity is ensured up-front by the choice of publicgoods to be created, shared learning occurs as a matterof course, and motivation and coordination amongdonors and partners may be achieved more easily thanin multisectoral national programs (see Annex 8).

Attention needs to be given to the interface betweeninternational and national public goods—a new dimen-sion of aid coordination. The implication is not thatinvestments in international public goods should waituntil conditions are right in most countries, but thatconditions need to be nurtured through transnationalcollaboration programs.

The CDF provides guidance. Efforts to build

Setting priorities at thenational level underthe CDF can helpidentify areas whereinternational programsare needed tosupplement nationalefforts.


regional and global partnerships should follow itskey tenets of inclusiveness and comprehensiveness.Setting priorities at the national level under the CDFcan help identify ar-eas where interna-tional programs areneeded to supple-ment national ef-forts. And buildingcapacity in nationaland local institutions(state and nonstate)will be critical inachieving the devel-opment goals of glo-bal programs.

3 1

66IMPLICATIONSFOR DEVELOPMENTEFFECTIVENESS

Increased relevance implies greater complexity. Managing complexity effectively calls for new

forms of partnership that facilitate strategic selectivity in line with comparative advantage. Sus-

tainability implies ownership of reforms, empowerment of domestic actors, and adequate domes-tic capacity. Thus institutional development—which provides the foundations for balanced develop-

ment, domestic ownership, participation, and partnership—emerges as the linchpin of the CDF.

Development effectiveness also depends on aresults orientation, a critical aspect of the CDF. Theframework is a compass, not a blueprint. Managing iteffectively will therefore require accurate monitoringof progress toward development goals.

The results orientation of the CDF reflects hard-won lessons of Bank lending. OED’s independentassessment of operational performance trends showsthat quality management must be strengthened further:

• A plateau seems to have been reached in theshare of projects with satisfactory outcomes.

• Institutional development continues to improve,but from a very low base.

• Sustainability remains low, and the operatingenvironment is becoming riskier.

• The Bank is being called on to deal with morecomplex and demanding development problems.

The changing global economy and the CDF sug-gest four principles for evaluating development effec-

tiveness. First, a results-based approach should bedesigned up front. That is, programs should be“evaluable” and the tracking of development outcomesand leading indicators should be carried out jointlywith the governmentand its partners, with aclear focus on povertyreduction and growth.Process indicators forkey principles, such asownership and part-nership, should also bemonitored.

Second, as the de-velopment effort shiftsits focus to the higherplane of country pro-grams, so must theevaluation process.Resources and skills need to be invested in developingappropriate indicators and information systems. The

Programs should be“evaluable” and thetracking of developmentoutcomes and leadingindicators should becarried out jointly withthe government and itspartners, with a clearfocus on povertyreduction and growth.


3 2

current preoccupation with project performance andevaluation should be complemented by a sectoral andcountrywide focus.

Third, evaluation must become participatory. Devel-opment effectiveness must be considered in the context ofshared objectives, joint responsibility for outcomes, recip-rocal obligations to achieve results, and distinct account-

ability for performance.Donors and govern-ments should team up tolink M&E systems withcivil society. Buildingcountry capacity forM&E is imperative.

Fourth, the growinginterdependence of theworld’s economies callsfor evaluating country-

based development programs from a global perspective,taking account of international development goals. Coun-try assistance strategies should link country goals,progress indicators, and national strategies with thesegoals. But these links should not imply top-down, target-based planning. Instead, the international goals must beadapted to country circumstances and priorities (Max-well, Foster, Naschold, and Conway 1999 BP). A globalperspective also calls for greater harmonization of evalu-ation across development agencies.

Finally, given the emphasis on the learning andprocess aspects of the development programs, Bankperformance indicators should emphasize professional-ism and partnership, creativity and innovation, pru-dence and probity, and flexibility and responsiveness.Sharing the knowledge emerging from country-basedprograms and pilots should leverage their learningcosts and accelerate the diffusion of lessons andpromising approaches within and among countries.

Implications for the BankWhile the Bank has experience in implementing indi-vidual CDF principles, it has yet to mainstream their jointapplication across the board. In effect, the Bank is movinginto uncharted territory. To equip itself to implement theframework effectively, the Bank will have to continuouslyexamine the results of ongoing experiments.

The CDF thus implies continued change in the Bank.A “one-size-fits-all” mentality will be replaced by a“customization” mindset. Every effort will be made toensure that the CDF does not become another blueprint,

repeating the pattern of the planning and adjustment eras.And the framework will have to be continually adapted.Rigidity is a real implementation risk.1

Learning will complement dissemination. In theplanning and adjustment eras, donors and centralgovernments too often adopted a hegemonic planningmentality that excluded vital local knowledge andpractical know-how.2 Similarly, in the adjustment erait was assumed that the Bank had all the answers—theonly problem was to sell them to its clients throughconditionality. Under the CDF, Bank staff can no longerpretend that the Bank is a storehouse of universallyapplicable knowledge. Instead, they will be induced toempower their country counterparts and learn withthem how to achieve development results effectively.

The CDF is certain to increase the demand fornonlending tools and advisory services, to engenderownership, partnership, and long-term holistic think-ing. The use of these tools will either empower clientsor lead to cognitive dependency—the end result willdepend on the skills and attitudes of users. Tendenciestoward bureaucratization and excessive documenta-tion—the pitfalls of the planning era—will need to beovercome by participatory and creative approaches tostrategy development. Donor-led economic work andpolicy prescriptions—the hallmarks of the adjustmentera—should give way to country-led approaches thatwould build on local processes and experiences anddevelop a strong commitment to policy reforms.

Bank processes have been attuned to a differentparadigm, and will have to change for implementationof the CDF. Bank management has taken steps to adaptthe organization under the Strategic Compact, but asurvey of Bank staff (and local donor representatives)found that more than half did not consider an activegovernment role a prerequisite for in-country aidcoordination (World Bank 1999m). Further internalchanges, both subtle and demanding, will be necessaryto fulfill the potential of the CDF.3

Finally, a new development architecture is needed toaddress the crisis of global poverty and mainstreampractice of the CDF principles throughout the develop-ment system. The Bank should promote the developmentof this architecture, in partnership with other multilateralinstitutions and developing countries. Among the firststeps is to harmonize donor procedures, build an informa-tion-sharing network, and innovate ways to strengthenclient voice and build local capacity for partnership,broad-based participation, and learning from results.

Every effort will bemade to ensure that the

CDF does not becomeanother blueprint,

repeating the pattern ofthe planning andadjustment eras.

ANNEXES

3 5

ANNEX 1. PROJECT PERFORMANCE TRENDS AND OUTSTANDING PROJECTS

OED evaluates all closed projects. These evalua-tions are the basis for the estimated trends in projectperformance and lessons from outstanding projectspresented here. The number of projects exiting theportfolio has risen significantly over the past two fiscalyears, and the number of independent project evalua-tions has increased nearly 20 percent since the lastARDE, greatly adding to the evaluation knowledgebase.

The newly evaluated cohort consists mainly ofinvestment projects approved in the early 1990s andfast-disbursing adjustment operations approved inFY96–98. This yields a complete picture of the FY98cohort of exiting projects and a partial look at the FY99cohort.1 The composition of the evaluated portfoliohas evolved over time, with the share of exitingprojects in Europe and Central Asia and in the humandevelopment sectors more than doubling during the1990s. The strong performance of adjustment opera-tions detailed in tables A1.2 and A1.3 is especiallynotable and will require further analysis.

Performance TrendsOED evaluators assess project results along threerelated dimensions—outcome (which combines rel-evance, efficacy, and efficiency), sustainability ofproject benefits, and institutional development (ID)impact. Last year’s ARDE introduced another perfor-mance indicator for quality at exit that combined thesekey dimensions: the Aggregate Project PerformanceIndex (APPI). In parallel, OED assesses Bank andborrower inputs during key stages of the project cycle.Figures A1.1 and A1.2 present the aggregate andoutcome trends by year of exit from the portfolio.

Aggregate Project Performance IndexThe APPI is defined by assigning cardinal weights tothe ratings of each of the three results-oriented dimen-sions (outcome, sustainability, and ID impact), thensumming them in a simple way. Thus, the APPI rangesfrom 2 (for projects with a highly unsatisfactoryoutcome, benefits that are unlikely to be sustainable,and negligible institutional development impact) to 10(for projects that performed strongly on all threemeasures). It is noteworthy that of the three dimen-sions, outcome is the key force behind the index. Theindex draws a line between satisfactory and unsatisfac-

tory outcomes: an index measure of 6 represents thedivide. A project with an unsatisfactory outcome willnever score higher than 6, no matter what ratings itreceives along the other two dimensions.

OutcomeOutcome assessments are based on a determination ofwhether the project achieved most of its major relevantgoals efficiently and with few shortcomings. Anevaluator’s judgement about outcome essentially boilsdown to answering the question: Did the projectachieve satisfactory development results, consideringthe importance and relevance of its major statedobjectives and the associated costs and benefits? Theoutcome rating takes into account relevance (to checkwhether the project’s objectives were consistent withthe country’s development strategy), efficacy (to exam-ine whether the operation achieved its stated goals),and efficiency (to assess results relative to inputs bycosts, implementation times, and economic and finan-cial returns). Outcome is rated on a six-point ordinalscale: highly satisfactory, satisfactory, marginally sat-isfactory, marginally unsatisfactory, unsatisfactory,and highly unsatisfactory.

Significant Variation among Regions and SectorsFigures A1.3 and A1.4 show the wide variation inproject results by region and sector. The three mainOED dimensions of outcome, sustainability, and IDimpact are presented in three separate graphs, with theregional information segregated into the three exitfiscal year periods of 1990–93, 1994–97, and 1998–99.Of the 16 sectors in the Bank, the sectoral figurehighlights 13—these are the sectors with at least 10projects exiting in FY98–99. Like the regional repre-sentation, each sectoral performance is segregated intothe three exit fiscal year periods of 1990–93, 1994–97,and 1998–99.

Institutional DevelopmentEmphasis on institutional development is critical toenabling the Bank to move beyond project financingand to achieve long-lasting improvements in develop-ing countries, with ownership by borrowers. OEDmeasures ID impact as the extent of the improvement inthe ability of an agency or a country to use its humanand financial resources effectively and to efficiently


3 6

organize economic and social activities prompted bythe project. ID impact is rated on a three-point scale:substantial, modest, and negligible. The most recenttrend in ID impact is presented in figure A1.5.

SustainabilityConceptually, sustainability differs from outcome, be-cause it focuses on the features that determine whetherthe operation will last over its intended useful life.

Thus, sustainability is defined as the likelihood, at thetime of evaluation, that the project will maintain itsresults in the future. Assessments of sustainability takeinto account a wide variety of factors, includingcountry conditions (for example, the government’scommitment to the future of the project, as documentedin the operational plan); government economic andfinancial policies, including policies on cost recovery;availability of funds for operation and maintenance;the political situation; sector conditions; and conditionsspecific to the operation (particularly the quality ofproject management and the capacity of project institu-tions). Sustainability is rated on a three-point scale:likely, uncertain, and unlikely. Aggregate results areportrayed in figure A1.6.

Borrower PerformanceBorrower performance is defined as the assumption ofownership rights and responsibilities and delivery ofthe inputs needed to prepare and implement the project.OED assesses borrower performance along three di-mensions—preparation, implementation, and compli-ance with legal covenants. Of particular importanceare the assignment of qualified personnel, the provisionof the required financial resources, and compliancewith the conditions of the loan agreement, includingreporting and auditing requirements and specific pro-curement rules. The performance of implementingagencies is rated on such elements as the quality of

FIGURE A1.2 SATISFACTORY PROJECT OUTCOMES

Note: By exit and approval fiscal year. Results are preliminary for FY99 exits (43 percent coverage) and FY92 approvals(57 percent coverage).Source: World Bank, Operations Evaluation Department data.

50 50

60 60

70 70

80 80

90 90

9290888684FY829897969594939291FY90

By Exit FY (percent satisfactory)

99

By projects

By projects

By disbursements By disbursements

By Approval FY (percent satisfactory)

FIGURE A1.1 AGGREGATE PROJECT PERFORMANCEINDEX

Note: By exit fiscal year. Results for FY99 are preliminary (43percent coverage).Source: World Bank, Operations Evaluation Department data.

5

6

7

8

9APPI average

FY90 91 92 93 94 95 96 97 98 99

By projects

By disbursements

3 7

FIGURE A1.3 OUTCOME, SUSTAINABILITY, AND IDIMPACT, BY REGION AND EXIT FISCAL YEAR GROUP

Note: The figures for FY99 exits are preliminary(43 percent coverage).Source: World Bank, Operations Evaluation Department data.

FIGURE A1.4 SATISFACTORY OPERATIONS BY SECTORAND EXIT FISCAL YEAR GROUP

Note: Includes only sectors with at least 10 projects exiting inFY98–99.Source: World Bank, Operations Evaluation Department data.

A n n e x 1 . P r o j e c t P e r f o r m a n c e T r e n d s a n d O u t s t a n d i n g P r o j e c t s

Industry

Industry

Water supply& sanitation


Electric power& other energy

Electric power& other energy

Agriculture

Agriculture

Population,health & nutrition

Population,health & nutrition

Multisector

Multisector

Oil & gas

Oil & gas

Finance

Finance

Education

Education

Social sector

Social sector

Urbandevelopment

Urbandevelopment

Transportation

Transportation

Public sectormanagement

Public sectormanagement

By Projects (percent satisfactory)

Exit FY90–93 Exit FY94–97 Exit FY98–99

0

0

20

20

40

40

60

60

80

80

100

100

By Disbursements (percent satisfactory)

Percent

Percent

0

0

0

10

10

10

20

20

20

30

30

30

40

40

40

50

50

50

60

60

60

70

70

70

80

80

80

90

Percent

Percent

Percent

90

90

AFR

AFR

AFR

EAP

EAP

EAP

ECA

ECA

ECA

LCRRegion

Region

Region

LCR

LCR

MNA

MNA

MNA

SAR

SAR

SAR

Outcome (percent satisfactory)

Sustainability (percent likely)

Institutional Development Impact(percent substantial)



3 8

management and staff associated with the operation;the use and effectiveness of technical assistance, includ-ing training, advisers, and contractual services; theadequacy of M&E systems; and the extent and qualityof participation by the intended beneficiaries, includingtheir contribution to the project’s outcome. Perfor-mance by fiscal year groups is shown in figure A1.7.

Bank PerformanceBank performance is defined as the quality of servicedelivered by the Bank, especially in those tasks for

FIGURE A1.5 ID IMPACT

Note: By exit fiscal year. Results for FY99 exits are preliminary (43 percent coverage).Source: World Bank, Operations Evaluation Department data.

FIGURE A1.6 SUSTAINABILITY

Note: By exit fiscal year. Results for FY99 exits are preliminary (43 percent coverage).Source: World Bank, Operations Evaluation Department data.

98 9897 97996 9695 9594 9493 9392 9291 91FY90 FY90

Percent Percent

99 99

By Projects By Disbursements

0 0

10 10

20 20

30 30

40 40

50 50

60 60

70 70

Likely

Uncertain

Unlikely

Likely

Uncertain

Unlikely

98 9897 9796 9695 9594 9493 9392 9291 91FY90 FY90

Percent Percent

99 99

By Projects By Disbursements

0 0

10 10

20 20

30 30

40 40

50

60

50

60

70 70

Substantial Substantial

Modest Modest

Negligible Negligible

which it has primary responsibility, such as appraisaland supervision. OED assesses Bank performancealong three dimensions—identification, appraisal, andsupervision. These three important considerations injudging Bank performance encompass the provision ofappropriate advice as an integral part of the appraisaland supervision process, the ability to adapt to chang-ing circumstances (if necessary, by restructuring aproject), and the oversight of loan/credit conditionality.Performance by fiscal year group is shown in figureA1.8.

3 9

FIGURE A1.7 BORROWER PERFORMANCE

Note: By exit fiscal year. The figures for FY99 exits are preliminary (43 percent coverage).Source: World Bank, Operations Evaluation Department data.

FIGURE A1.8 BANK PERFORMANCE

Note: By exit fiscal year. The figures for FY99 exits are preliminary (43 percent coverage).Source: World Bank, Operations Evaluation Department data.



00

2020

4040

6060

8080

100Percent

100Percent

PreparationPreparation ImplementationImplementation ComplianceCompliance




00

2020

4040

6060

8080

100100


Identification IdentificationAppraisal AppraisalSupervision Supervision


PercentPercent

Lessons from Outstanding ProjectsOf the 289 operations evaluated since last year’sARDE, OED assessed 21 (7 percent) as outstanding and8 (3 percent) as having particularly poor performance(table A1.1). The selection criteria for outstandingprojects include a highly satisfactory outcome, likelysustainability, and substantial ID impact. Theseprojects are judged to have met or exceeded all theirmajor goals, and many had innovative designs or are

judged to be potentially replicable in other countries orsectors. Their success can be traced to borrowerownership, consistent monitoring, good supervision, aninnovative design allowing flexibility in responding tochanging conditions, and continuity in the Bank staffassigned to them.

Several of these characteristics are amply demon-strated by Argentina’s Provincial Reform Loan, the firstBank adjustment operation targeting subnational enti-


4 0

ties. The project’s main objective was to support thefederal government’s effort to promote the reform ofprovincial finances. The loan’s contributions wentbeyond the conditions for each tranche by providing aforum for the provinces to agree on a reform strategy.The operation also set in place support mechanismsthat not only advanced reform among the provincialgovernments, but also helped to curb the contagioneffects of the Mexican crisis at the provincial level. Theoperation’s success is directly related to the consistencyand commitment of both the national government andthe Bank. The government’s strategy was transparentand uniform across provinces. And the Bank’s lendingportfolio was fully consistent with the government’sstrategy, with common rules for all provincial lending.

This project offers lessons directly applicable to theincreasingly important area of subnational lending andserves as a model for operations in Latin America andelsewhere. A telecommunications project in Sri Lankademonstrates the importance of staff continuity andstrong borrower commitment in the difficult area ofinstitutional development. The project sought publicreforms and restructuring in telecommunications, andregulation and promotion of private investments in thesector. Initially classified as a problem project becauseof slow disbursement, the project eventually exceededall its stated objectives, thanks to a strong commitmentto dialogue on both sides. The Bank team responsiblefor the project remained unchanged from identificationto completion.

4 1


TABLE A1.1 OUTSTANDING AND POOR PERFORMERS AMONG RECENTLY EVALUATED PROJECTS

Country Project Loan or credit

Outstanding performersArgentina Provincial Reform L3836Argentina Provincial Pension Reform Adjustment L4116Bolivia Capitalization Program C2761Bolivia Hydrocarbon Sector Reform C2762China Shuikou Hydroelectric 2 L3515Georgia Transport Rehabilitation C2809Ghana Feeder Roads C2319Hungary Enterprise and Financial Sector Adjustment L4141India Industrial Technology Development C2064India Technician Education C2130India Maharashtra Emergency Earthquake C2594Indonesia Village Infrastructure L3888Korea, Rep. Environmental Research and Education L3612Latvia Structural Adjustment L4126Macedonia, FYR Transit Facilitation L3868Morocco Emergency Drought Recovery L3935Mozambique Economic Recovery 3 CN010Panama Economic Recovery L3438Poland Environmental Management L3190Romania Transportation L3593Sri Lanka Telecommunications 2 C2249Poor performersBulgaria Agricultural Development L3771Bulgaria Private Investment and Export Finance L3631Indonesia Financial Sector Development L3526Morocco National Agricultural Credit L3088Morocco National Rural Finance L3662Nigeria National Population C2238Philippines Rural Electrification L3439

Vanuatu Housing C2262

Note: Covers only the 289 projects evaluated since last year’s ARDE.Source: World Bank, Operations Evaluation Department data.

19

99

An

nu

al

Re

vi

ew

of

De

ve

lo

pm

en

t E

ff

ec

ti

ve

ne

ss

42 TABLE A1.2 OUTCOME, SUSTAINABILITY, ID IMPACT, AND AGGREGATE BY SECTOR, NETWORK, LENDING TYPE/SOURCE, REGION, AND WDI INCOME

GROUP FOR EXIT FISCAL YEAR GROUPS 1990–93, 1994–97, AND 1998–99(BY PROJECTS)


Outcome Sust. ID impact Outcome Sust. ID impact Outcome Sust. ID impact

Projects Share % % % Aggregate Projects Share % % % Aggregate Projects Share % % % Aggregate# % sat. likely sub. index # % sat. likely sub. index # % sat. likely sub. index

Sector

Agriculture 252 30 59 34 27 6.1 235 24 69 42 35 6.3 71 19 62 42 39 6.4

Education 67 8 82 64 38 7.0 79 8 72 44 27 6.4 39 10 79 49 33 6.8

Electric Power & Other Energy 79 9 69 59 34 6.5 84 9 60 54 28 6.1 24 6 61 39 30 6.0

Environment 0 — — — — — 9 1 56 44 22 6.5 5 1 80 40 20 6.5

Finance 53 6 57 43 26 6.2 51 5 59 49 38 6.3 27 7 74 58 48 7.0

Industry 50 6 58 46 24 6.2 41 4 51 39 32 5.9 10 3 40 33 30 5.4

Mining 12 1 55 50 50 6.2 9 1 78 78 67 7.1 3 1 67 67 100 7.6

Multisector 62 7 71 52 35 6.7 100 10 83 54 27 6.8 31 8 65 48 13 6.2

Oil & Gas 27 3 78 52 67 7.3 24 2 79 58 42 7.0 13 3 69 69 31 6.8

Population, Health & Nutrition 21 2 67 43 14 6.2 42 4 67 55 26 6.4 28 7 64 50 25 6.3

Public Sector Management 35 4 56 40 31 6.0 57 6 56 44 23 5.8 29 8 97 72 52 7.6

Social Sector 1 0 100 100 100 8.3 16 2 88 25 56 6.9 20 5 80 20 45 6.7

Telecommunications 11 1 64 73 36 6.7 22 2 82 73 36 7.0 4 1 100 100 100 9.6

Transportation 95 11 74 55 34 6.6 108 11 76 50 33 6.7 29 8 93 72 69 7.9

Urban Development 44 5 77 41 27 6.5 50 5 60 36 16 5.9 21 6 85 45 45 6.9

Water Supply & Sanitation 37 4 69 40 43 6.7 38 4 55 32 18 5.7 23 6 48 26 30 5.7

Network

Environmentally & Socially

Sustainable Development 252 30 59 34 27 6.1 244 25 68 42 35 6.3 76 20 63 42 37 6.4

Finance, Private

Sector & Infrastructure 408 48 68 50 35 6.5 427 44 65 48 30 6.3 154 41 71 52 46 6.8

Human Development 89 11 78 59 33 6.8 137 14 72 45 30 6.5 87 23 75 43 33 6.6

Poverty Reduction

& Economic Management 97 11 66 48 34 6.4 157 16 73 50 25 6.5 60 16 80 60 32 6.9

43

TABLE A1.2 OUTCOME, SUSTAINABILITY, ID IMPACT, AND AGGREGATE BY SECTOR, NETWORK, LENDING TYPE/SOURCE, REGION, AND WDI INCOMEGROUP FOR EXIT FISCAL YEAR GROUPS 1990–93, 1994–97, AND 1998–99 (CONTINUED)(BY PROJECTS)


Outcome Sust. ID impact Outcome Sust. ID impact Outcome Sust. ID impact

Projects Share % % % Aggregate Projects Share % % % Aggregate Projects Share % % % Aggregate# % sat. likely sub. index # % sat. likely sub. index # % sat. likely sub. index

Lending type

Adjustment 113 13 69 51 35 6.6 152 16 79 59 33 6.9 53 14 85 66 45 7.2

Investment 733 87 66 45 32 6.4 813 84 66 44 30 6.3 324 86 70 46 38 6.6

Lending source

IBRD only 448 53 69 58 36 6.7 489 51 71 55 33 6.5 170 45 76 59 45 7.0

IDA/blend 398 47 63 33 28 6.1 476 49 65 38 29 6.2 207 55 69 41 34 6.4

Region

Africa 278 33 54 28 25 5.8 320 33 55 28 24 5.8 122 32 61 34 26 6.0

East Asia and Pacific 159 19 80 70 45 7.2 158 16 84 69 41 7.1 58 15 81 54 50 7.1

Europe and Central Asia 50 6 73 56 38 6.9 87 9 72 62 40 6.6 47 12 83 69 41 7.2

Latin America

and Caribbean 158 19 64 55 33 6.4 179 19 78 54 38 6.9 74 20 81 66 52 7.3

Middle East

and North Africa 92 11 70 51 31 6.5 91 9 66 44 21 6.1 26 7 69 38 44 6.6

South Asia 109 13 73 38 30 6.4 130 13 67 47 27 6.2 50 13 66 42 34 6.5

WDI 1999 Income Group

Lower 390 46 59 30 26 6.0 477 49 60 33 26 6.0 183 49 66 36 28 6.2

Lower-middle 291 34 73 58 38 6.8 314 33 74 54 30 6.6 129 34 76 57 50 7.0

Upper-middle 142 17 69 62 35 6.7 150 16 81 69 45 7.0 57 15 81 72 46 7.2

High 23 3 82 82 45 7.5 24 2 79 75 33 7.0 8 2 88 75 57 8.1

Total/average 846 100 66 46 32 6.4 965 100 68 47 31 6.4 377 100 72 49 39 6.7

Note: Percentages exclude projects not rated. The data for FY99 exits represents a partial sample (118 out of 277) and reflects the processing of all ICRS received throughJuly 1999. The processing of the remainder of the FY99 sample is ongoing; completion is expected by spring 2000. Access the table on-line athttp://wbln0018.worldbank.org/oed/oeddoclib.nsf/htmlmedia/pubparade.html

An

ne

x 1

. Pr

oje

ct P

er

for

ma

nc

e T

re

nd

s an

d O

utsta

nd

ing

Pr

oje

cts

19

99

An

nu

al

Re

vi

ew

of

De

ve

lo

pm

en

t E

ff

ec

ti

ve

ne

ss

44 TABLE A1.3 OUTCOME, SUSTAINABILITY, ID IMPACT, AND AGGREGATE BY SECTOR, NETWORK, LENDING TYPE/SOURCE, REGION,

AND WDI INCOME GROUP FOR EXIT FISCAL YEAR GROUPS 1990–93, 1994–97, AND 1998–99(BY REAL DISBURSEMENTS, FY96US$)


Disburse Outcome Sust. ID impact Disburse Outcome Sust. ID impact Disburse Outcome Sust. ID impact

$ Share % % % Aggregate $ Share % % % Aggregate $ Share % % % Aggregatemillions % sat. likely sub. index millions % sat. likely sub. index millions % sat. likely sub. index

Sector

Agriculture 14,860 21 69 42 31 6.4 13,136 17 79 52 42 6.7 3,742 12 64 54 50 6.9

Education 2,921 4 89 76 59 7.4 3,956 5 73 49 33 6.7 2,686 8 87 51 36 7.1

Electric Power & Other Energy 9,692 14 66 75 40 6.5 9,618 13 69 61 31 6.5 1,803 6 59 34 28 5.9

Environment 0 - - - - - 285 0 75 58 30 7.0 107 0 82 44 17 6.5

Finance 4,785 7 48 42 28 6.1 7,704 10 64 60 49 6.9 7,843 24 95 92 47 7.6

Industry 7,210 10 71 64 26 6.7 5,053 7 57 47 43 6.3 810 2 50 47 46 6.1

Mining 671 1 46 65 63 6.3 606 1 95 95 28 7.0 431 1 98 98 100 8.2

Multisector 11,490 17 82 66 53 7.5 10,716 14 90 61 32 7.1 2,146 7 75 54 20 6.6

Oil & Gas 1,995 3 74 91 44 7.6 1,650 2 96 87 40 7.8 1,998 6 70 84 6 6.5

Population, Health & Nutrition 716 1 71 55 33 7.0 1,626 2 85 68 35 6.9 1,830 6 81 71 37 6.9

Public Sector Management 1,658 2 79 45 70 7.4 2,530 3 79 65 34 6.6 2,126 7 99 73 70 8.2

Social Sector 29 0 100 100 100 8.2 1,022 1 99 55 41 6.8 882 3 94 63 76 7.2

Telecommunications 360 1 40 47 19 5.7 1,202 2 93 95 64 8.3 233 1 100 100 100 9.3

Transportation 7,431 11 80 62 37 7.0 9,453 12 86 54 33 6.9 2,770 9 87 69 67 7.6

Urban Development 3,671 5 79 57 33 6.9 3,737 5 74 51 14 6.3 1,514 5 97 58 53 7.6

Water Supply & Sanitation 1,907 3 65 38 40 6.5 3,361 4 55 23 13 5.3 1,565 5 44 18 23 5.7

Network

Environmentally & Socially

Sustainable Development 14,860 21 69 42 31 6.4 13,421 18 79 52 41 6.8 3,849 12 64 53 49 6.9

Finance, Private

Sector & Infrastructure 37,722 54 68 63 35 6.7 42,385 56 72 56 35 6.6 18,966 58 82 72 43 7.1

Human Development 3,666 5 86 72 54 7.4 6,604 9 80 55 35 6.8 5,398 17 86 60 43 7.0

Poverty Reduction

& Economic Management 13,148 19 82 63 56 7.5 13,246 18 88 61 32 7.0 4,272 13 87 63 45 7.4

45

TABLE A1.3 OUTCOME, SUSTAINABILITY, ID IMPACT, AND AGGREGATE BY SECTOR, NETWORK, LENDING TYPE/SOURCE, REGION,AND WDI INCOME GROUP FOR EXIT FISCAL YEAR GROUPS 1990–93, 1994–97, AND 1998–99 (CONTINUED)(BY REAL DISBURSEMENTS, FY96US$)


Disburse Outcome Sust. ID impact Disburse Outcome Sust. ID impact Disburse Outcome Sust. ID impact$ Share % % % Aggregate $ Share % % % Aggregate $ Share % % % Aggregate

millions % sat. likely sub. index millions % sat. likely sub. index millions % sat. likely sub. index

Lending type

Adjustment 21,745 31 74 62 44 7.1 22,155 32 80 64 41 7.0 11,910 17 93 84 51 7.6

Investment 47,651 69 71 58 37 6.7 53,500 77 75 53 33 6.6 20,574 30 74 56 41 6.8

Lending source

IBRD only 51,719 75 72 65 42 6.9 52,136 75 77 61 38 6.8 23,371 34 83 71 48 7.2

IDA/blend 17,677 25 71 42 31 6.4 23,519 34 76 47 30 6.6 9,113 13 76 54 37 6.9

Region

Africa 10,706 15 59 23 25 5.9 12,570 18 68 33 28 6.2 4,275 6 69 40 23 6.3

East Asia and Pacific 14,720 21 89 80 53 7.6 15,843 23 86 75 43 7.3 10,378 15 90 80 43 7.2

Europe and Central Asia 5,812 8 67 54 41 6.8 9,378 14 83 71 46 7.1 3,719 5 73 71 36 7.0

Latin America and Caribbean 21,712 31 66 68 34 6.8 18,103 26 82 58 45 7.1 7,056 10 90 78 71 7.8

Middle East and North Africa 4,900 7 73 57 45 6.8 6,490 9 64 34 14 5.8 1,731 2 83 47 56 7.1

South Asia 11,546 17 76 52 39 6.7 13,271 19 68 55 24 6.2 5,326 8 69 50 31 6.7

WDI 1999 Income Group

Lower 22,508 32 68 38 32 6.3 27,169 39 69 44 26 6.3 9,759 14 69 45 28 6.5

Lower-middle 22,016 32 80 71 48 7.2 28,368 41 81 61 35 6.9 9,928 14 79 60 59 6.9

Upper-middle 23,215 33 68 66 36 6.9 18,835 27 81 66 50 7.1 7,482 11 87 81 59 7.7

High 1,658 2 84 85 53 7.6 1,283 2 78 78 32 7.0 5,316 8 98 99 7 7.3

Total/average 69,396 100 72 59 39 6.8 75,655 100 77 56 35 6.7 32,485 100 81 66 44 7.0

Note: Percentages exclude projects not rated. The data for FY99 exits represents a partial sample (118 out of 277) and reflects the processing of all ICRS received throughJuly 1999. The processing of the remainder of the FY99 sample is ongoing; completion is expected by spring 2000. Access the table on-line athttp://wbln0018.worldbank.org/oed/oeddoclib.nsf/htmlmedia/pubparade.html

An

ne

x 1

. Pr

oje

ct P

er

for

ma

nc

e T

re

nd

s an

d O

utsta

nd

ing

Pr

oje

cts

4 7

ANNEX 2. CHALLENGES IN IMPLEMENTING THE CDF PRINCIPLES AT THE SECTOR LEVEL

Sectors have varied in their success in delivering on theprinciples of the CDF. The traditional infrastructuresectors have performed better than others (figure A2.1).Background papers commissioned for this review andan informal survey of OED evaluators turned up thesame results. The principal evaluators for some of thesectors were asked to rate the Bank’s contribution tohelping client countries implement the principles of theCDF on a scale of 1 (poor) to 4 (excellent). The resultsare detailed below.

Long-Term Perspective and a Holistic ApproachA key impediment to development has been the lack ofa long-term perspective. Its absence has been especiallydetrimental in trying to deal with the structuraldimensions of reform. OED evaluators also highlightthe importance of comprehensive analyses in elucidat-ing cross-sectoral dependencies. One outcome of thismore comprehensive approach has been greater com-plexity, which often creates difficulties for sectorwideand multisectoral programs and projects. There is acomplementary need to simplify program design fol-lowing comprehensive analyses to avoid overloadinglimited government capacity.

In a holistic approach, thematic, cross-sectoral, andstructural dimensions can provide crucial integratingthemes. Yet integrating such dimensions has provedparticularly challenging and has often been resisted byaid agencies and government ministries with a sectoralorientation. It took decades of external pressure andleadership by top management, for example, to integrateenvironmental considerations into Bank operations.

A notable example concerns information technologyand knowledge management. As the importance of theinformation technology revolution to development be-came clear, the Bank established a Knowledge Manage-ment Unit. Education and other sectors began to explorethe use of information technology, as in distance educa-tion. Over the past decade, Bank lending for knowledgeapplications has spread, figuring as a component in 60–90 percent of projects. Yet this dimension of developmenthas yet to be integrated into country assistance and sectorstrategies. Moreover, Bank studies and Internal AuditDepartment reviews have concluded that lending andpolicy assistance in this area lack standards, a coherentstrategy, quality assurance, and proper integration withother sectors (Hanna 1991; Hanna and Boyson 1993).

Ownership and PartnershipOwnership has been difficult to achieve in sectors witha broad array of stakeholders, such as health andeducation. Yet lack of ownership—by governments forreforms and by communities for local projects—hasundermined development efforts across the board. Ingeneral, ownership by government, the private sector,and civil society has been easier to achieve in technol-ogy-driven sectors that are open to private participa-tion (energy, telecommunications) than in softer sectors(education, environment, and rural development). Pri-vate sector development nevertheless scored well onownership, presumably because of the openness toprivate participation (figure A2.2).

The picture is similar for partnership. Partnershipsbetween external stakeholders are essential in coordinat-ing sectorwide programs to avoid fragmentation oradding unnecessarily to the demands on government.OED evaluators have found cases of solid coordinationwith other donors in technology-driven sectors such astransport, telecommunications, and energy, and weakcoordination in the social sectors. Partnerships with other

FIGURE A2.1 SECTOR SCORES ON CDF

Telecommunications


Energy

Private sectordevelopment

Education

Urban

Environment

Transportation

Rural

Sector

0.0 0.5 1.0 1.5

Score

2.0 2.5 3.0 3.5 4.0


4 8

external stakeholders clearly have a long way to go in allareas, especially in rural development, education, andprivate sector development.

Managing for ResultsManaging for results has been easier to achieve intechnology-driven sectors than in structural dimensions(institutions, public and private sector management),

FIGURE A2.2 SECTOR SCORES ON INDIVIDUAL CDF PRINCIPLES

social sectors (education), or thematic areas (environ-ment, rural development). OED evaluators have con-sistently found poor M&E systems for results (as wellas for tracking inputs and outputs) and a need forgreater attention to this area in sectorwide and the-matic efforts.

Results focus

Partnership

Ownership

Holistic

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Rural

Education


Private sectordevelopment

Environment

Urban

Transportation

Telecommunications

Energy

Sector

4 9

ANNEX 3. DEVELOPMENT PERFORMANCE IN SELECTED COUNTRIES

OED has prepared CAEs for 28 countries. Whilenot chosen with a view toward forming a representativesample of developing countries, they were selected byOED in the context of the Board calendar and where anevaluation was judged useful for the design of futurestrategies or lending programs. Other considerationsincluded relevance, regional coverage, and importanceto the Bank’s portfolio. The 28 countries capture a widerange of country size, income, and regional representa-tion (table A3.1).

The facts on growth and poverty in the 28 countriesbetween 1981 and 1997 are sobering. They suggest thatthe battle against poverty is being lost, not won (tableA3.2 and figure A3.1):

• In 40 percent of the countries, per capita incomeeither failed to grow or shrank.

• In 23 percent, the share of the population belowthe poverty line increased.

• In 25 percent, life expectancy declined.• In 54 percent, the people experienced at least one of

these three events: per capita income stagnated, thepoverty rate increased, or life expectancy declined.

• In 85 percent, per capita income grew 1 percent ayear or less in the 1990s.

FIGURE A3.1 DISAPPOINTING PERFORMANCE INREDUCING POVERTY . . .

23

0 20 40 60 80 100

Percent of countries in sample

85

40

25

54

Poor–modest growthin per capita incomes

Declining percapita income (1)

Increasing poverty (2)

Declining lifeexpectancy (3)

(1) And/or (2) and/or (3)

FIGURE A3.2 . . . AND IN PROMOTING ECONOMIC GROWTH

Percent of countries in sample

Savings

0

20

40

60

80

59

26

48

Low ordeclining

savings rate

Significantdecline in

savings rates

Low rate ofsavings (<10%)

0

20

40

60

80Percent of countries in sample

Investment

67

46

25

Low ordeclining efficiency

of investment

Decliningefficiency ofinvestment

Low efficiency ofinvestment (<10%)


5 0

Some observers have argued that the inadequateprogress in reducing poverty may be the legacy ofneglect of social spending during periods of austerityrequired by adjustment programs. But the CAEs do notsuggest any significant decline in social expenditures asa share of the total. Many countries made a deliberateeffort to maintain social expenditures—including Côted’Ivoire, Ghana, Morocco, Poland, and Togo.

Growth performance can be broken down into twoparts: domestic resource mobilization, indicated by theratio of savings to GNP, and efficiency of investment.1

The savings rate declined in 26 percent of thecountries in the group during 1981–97, and in 48percent it was less than 10 percent. A majority of thecountries (59 percent) experienced a low or decliningsavings rate—or both (table A3.3 and figure A3.2).

In 42 percent of the countries, investment efficiencywas lower in 1981–97 than in the 1970s. And in aquarter it was less than 10 percent. Two-thirds of thecountries had either low or declining investment effi-ciency (table A3.4).

TABLE A3.1 RELEVANCE OF SAMPLE OF EVALUATED COUNTRIES (PERCENT)

Classification In total In sample

Income category

Low 39 57

Lower-middle 39 36

Upper-middle 24 7

Total (number) 157 28

Region

Latin America and the Caribbean 22 14

Sub-Saharan Africa 31 32

Middle East and North Africa 10 7

South Asia 5 14

Europe and Central Asia 17 14

East Asia and the Pacific 14 18

Total (number) 157 28

Size

High population 33 57

Medium population 34 36

Low population 33 7

Indicator Low-income a Middle-income a

Average annual growth of per capita income

1985–95 -1.4 -0.7

1997–98 -5.9 -1.5

Negative per capita growth rate

(percent of countries) 53.7 41.5

a. According to the World Development Report 1997 (page 265): Economies are divided according to 1995 GNP per capita, calculatedusing the World Bank Atlas method. The groups are: low-income, $765 or less, and middle-income, $766–$9,385.Source: World Development Report 1999, 1997, and 1990.

51

An

ne

x 3

. De

ve

lop

me

nt P

er

for

ma

nc

e in

Se

lec

ted

Co

un

trie

s

a. Urban poverty only; rural rate much higher.b. Poverty is estimated to increase due to East Asia crises.c. Only rural poverty.d. Considered qualitative assessment that people are worse-off in the 1990s than in the 1980s.

TABLE A3.2 COUNTRY PERFORMANCE: PER CAPITA GROWTH, POVERTY, AND LIFE EXPECTANCY

GNP per capita growth rate (%) Poverty: % of population (year) Life expectancy (years)

Country 1980s 1990s 1981–97 Datapoint 1 Datapoint 2 Datapoint 3 Datapoint 4 Datapoint 5 Early 1980s 1990s

Albania 2.58 35.5 (96) 70 72Azerbaijan -7.22 61.0 (95) 69 71Bangladesh 0.68 1.47 0.96 40.9 (83) 33.7 (85) 41.3 (88) 42.6 (91) 35.5 (95) 50 58Bolivia a 0.84 60.1 (89) 61.6 (93) 54 61Burkina Faso 0.44 0.21 0.35 45 44Cambodia 1.08 45 54Côte d’Ivoire -1.47 0.70 -0.89 11.1 (85) 17.8 (88) 32.8 (93) 36.8 (95) 50 47Croatia 1.14 71 72Ecuador -0.23 0.22 0.19 35.0 (94) 65 70Ethiopia -0.40 1.62 -0.22 52.0 (94) 42 43Ghana 0.05 0.57 0.57 36.9 (88) 41.8 (89) 31.4 (92) 54 60Indonesia b 1.80 2.44 2.25 21.6 (84) 17.4 (87) 15.1 (90) b 56 65Jamaica 0.17 -1.73 0.58 29.8 (88) 34.2 (92) 71 75Kenya c 0.33 0.14 0.03 47.9 (82) 46.4 (92) 56 52Malawi c 0.00 0.49 0.10 42 (91) 45 43Maldives 1.78 57 67Morocco 0.93 0.10 0.60 16.5 (85) 7.0 (91) 60 67Mozambique 0.67 1.56 0.19 44 45Nepal d 0.82 0.91 0.91 36.2 (77) 41.4 (86) 44.6 (96) 49 57Papua New Guinea -0.04 0.56 0.54 52 58Philippines -0.71 0.94 0.01 39 (91) 36 (94) 62 68Poland 0.91 2.37 0.54 6.0 (88) 20.0 (94) 71 73Sri Lanka 1.09 1.83 1.24 27.3 (86) 22.4 (91) 69 73Thailand 2.54 2.39 2.82 22.2 (88) 17.9 (90) 13.1 (92) b 65 69Togo d -0.20 0.05 -0.62 32.3 (89) 50 49Ukraine -12.18 30.0 (95) 70 67Yemen -0.25 19.1 (92) 50 54Zambia -1.20 -0.17 -0.73 69 (91) 51

Percent of countries with: GNPpc =< 0%50% (10/20) 39% (11/28) 40%(8/20) (not significantly different from 0)

Life expectancy declines 25% (7/28)Increasing poverty 23% (5/22)Growth <0% and/or increasing poverty and/or declining life expectancy 54%


5 2

TABLE A3.3 COUNTRY PERFORMANCE: SAVINGS

Savings rate (% GDP) Linear trend savings

Country 1980s 1990s 1981–97 a T-Stat

Albania 30.33 -22.31 -4.28 -3.89

Azerbaijanb 5.84

Bangladesh 11.69 13.84 0.16 2.78

Bolivia 13.39 9.46 -0.50 -3.94

Burkina Faso 0.40 6.88 0.89 8.68

Cambodiac 5.41 0.20 0.91

Côte d’Ivoire 18.72 16.74 -0.22 -0.78

Croatia 8.66

Ecuador 21.51 22.55 -0.03 -0.26

Ethiopia 4.81 5.30 0.23 2.03

Ghana 4.81 8.65 0.46 3.88

Indonesia 30.11 31.40 0.16 1.87

Jamaica 17.65 25.12 0.84 4.56

Kenya 19.94 17.99 -0.27 -1.95

Malawi 11.68 4.75 -0.82 -4.45

Maldivesd

Morocco 16.55 16.19 0.12 1.11

Mozambique -8.56 0.40 1.01 3.11

Nepal 10.38 11.51 0.08 0.93

Papua New Guinea 12.20 28.95 1.81 8.37

Philippines 20.38 15.90 -0.52 -8.09

Polanda 30.28 17.57 -1.15 -1.49

Sri Lanka 13.09 15.28 0.22 1.99

Thailand 27.56 35.89 0.95 10.63

Togo 11.46 7.93 -0.41 -1.95

Ukrainea 27.59 -1.84 -1.94

Yemena -1.87 0.70 0.54

Zambia 13.78 7.62 -0.38 -1.37

Percent of countries with:

Significantly declining rate of savings 26% (7/27)

Savings rate < 10% in 1990s 48% (13/27)

Declining or/and low savings rate 59% (16/27)

a. For the period 1990–97 for which data are available.b. Data available only since 1991 and 1993 for Azerbaijan and Croatia.c. For period 1988–97.d. No data available.Note: Savings for 1980s and 1990s are calculated as a simple average for this period (GDS as % of GDP). Rate of savings is calcu-lated as a linear trend—Savings (% GDP) = a + (Rate * Time) for the period 1981–97.

5 3

TABLE A3.4 COUNTRY PERFORMANCE: INVESTMENT EFFICIENCY

GDP growth (%) GDI/GDP (%) Investment efficiency (%)

Country 1971–80 1981–97 1971–80 1981–97 1971–80 1981–97

Albania 0.51 24.67 2.07

Azerbaijana

Bangladesh 1.79 4.75 13.92 19.66 12.86 24.18

Bolivia 1.78 15.55 11.45

Burkina Faso 3.34 3.67 21.85 20.82 15.30 17.64

Cambodia 5.49 13.83 39.70

Côte d’Ivoire 5.48 1.70 24.49 13.11 22.37 12.96

Croatiaa

Ecuador 9.05 2.61 24.21 20.25 37.39 12.90

Ethiopia 3.02 13.34 22.62

Ghana 0.52 3.12 9.35 13.37 5.57 23.32

Indonesia 7.87 6.82 22.45 28.57 35.08 23.87

Jamaica -0.60 1.55 22.16 27.31 5.68

Kenya 8.18 3.24 24.11 22.00 33.94 14.73

Malawi 6.25 3.16 27.14 18.04 23.02 17.53

Maldivesa

Morocco 5.56 3.25 23.23 23.26 23.91 13.97

Mozambique 2.46 19.40 12.66

Nepal 2.11 4.90 13.11 20.55 16.10 23.84

Papua New Guinea 2.57 2.98 23.51 24.95 10.94 11.93

Philippines 5.93 2.35 27.77 22.40 21.35 10.49

Poland 1.48 23.93 6.19

Sri Lanka 4.43 4.64 18.97 24.90 23.34 18.64

Thailand 6.89 7.46 26.17 34.59 26.31 21.57

Togo 4.41 1.63 29.45 17.18 14.97 9.47

Ukraine -9.30 28.69 -32.39

Yemena

Zambia 1.45 1.19 29.71 14.72 4.89 8.06

Percent of countries with:

Declining efficiency of investment 42% (11/24)

Low efficiency of investment (<10%) 25% (6/24)

Declining/low efficiency of investment 67% (16/24)

a. No data available.Note: The Investment Efficiency Ratio is defined as the GDP growth rate divided by the investment GDP ratio. This is the inverse ofthe conventional Incremental Capital Output Ratio (ICOR), but is adjusted for changes in terms of trade between investment goodsand overall GDP. Simple averages are used.

A n n e x 3 . D e v e l o p m e n t P e r f o r m a n c e i n S e l e c t e d C o u n t r i e s

5 5

The sustainable livelihood approach to develop-ment adopted by the U.K. Department for InternationalDevelopment, CARE, and the UNDP proposes a holis-tic understanding of poverty and an integrated anddynamic way of thinking about poor people’s liveli-hoods. This approach defines a livelihood as “thecapabilities, assets (including both material and socialresources), and activities required for a means ofliving. A livelihood is sustainable when it can copewith and recover from stresses and shocks, and main-tain or enhance its capabilities and assets both now andin the future, while not undermining the naturalresource base” (Chambers and Conway 1992).

The sustainable livelihood approach provides ananalytical structure to guide the understanding offactors that influence poverty and to identify whereinterventions might best be made (figure A4.1). Itrecognizes people’s diverse livelihood goals—betterhealth, more income, reduced vulnerability—and thecomplex strategies that households adopt to meet thesegoals. These strategies are shaped by preferences andpriorities that arise in a context of vulnerabilityresulting from shocks, changing trends, and seasonalvariations. They are also affected by structures (such asthe form and organization of government and theprivate sector) and processes (policies, laws, andinstitutions).

The approach identifies five kinds of capital assetson which people can build or draw: human, natural,financial, social, and physical. It aims to build on whatpeople have and how they live their lives, to add to theaccumulation of assets, and to remove the barriers topeople’s ability to advance their livelihood choices.

Early experience with the sustainable livelihoodapproach has shown that its common frameworkfacilitates cross-sectoral and multidisciplinary think-ing. It makes explicit the choices and tradeoffs inplanning and execution. It emphasizes a small numberof entry points, with multisectoral links evolving overtime. The iterative approach demands learning andsequencing, and thus requires that funding partners beflexible and responsive.

But the experience has also raised new challenges.The approach can result in a huge agenda, and thusrisks becoming too multifaceted and time-intensive.Institutional issues—such as developing closer opera-tional links with the social sectors—need more atten-tion. And while the approach may improve the under-standing of the problems faced by the poor, effectingchange in a world of entrenched power structuresremains a huge challenge. Another challenge is todevelop effective tools for managing tradeoffs andsequencing—and for measuring change in people’slivelihoods and security.

ANNEX 4. EMPHASIZING PEOPLE-CENTERED DEVELOPMENT IN A HOLISTIC FRAMEWORK

FIGURE A4.1 SUSTAINABLE LIVELIHOOD FRAMEWORK

• Privatesector

• SHOCKS

• TRENDS

• SEASONALITY

Influence& access

Inorder

toachieve

• More income

• Increasedwell-being

• Reducedvulnerability

• Improved foodsecurity

• More sustainableuse of naturalresources

VULNERABILITYCONTEXT

TRANSFORMINGSTRUCTURES &PROCESSES

STRUCTURES

PROCESSES

• Laws

H

S

P F

N

KeyH = Human capitalN = Natural capitalF = Financial capital

S = Social capitalP = Physical capital

LIVELIHOODSTRATEGIES

• Levels ofgovernment

• Policies

• Culture

• Institutions

LIVELIHOOD ASSETS LIVELIHOODOUTCOMES

5 7

ANNEX 5. OWNERSHIP, PARTICIPATION, AND CONDITIONALITY

Ownership at its broadest involves commitment bythe public sector, the private sector, and civil society toa national development effort. At its narrowest itinvolves beneficiaries, local government, and commu-nity organizations in project design and implementa-tion. Ownership can apply to a national developmentstrategy, country assistance strategy, sectoral reformprograms, or a project.

Ownership and ParticipationParticipation is important because it increases thelikelihood of a project’s relevance and effectiveness.OED findings suggest that participation promotesownership and broad, long-term attention to humanissues. It improves sustainability of development pro-grams and promotes learning and results-based orienta-tion—key goals of the CDF.

Benefits of ParticipationAt the project level there are clear benefits to ownershipand participation. Participation promotes a more holis-tic approach. For instance, in Bangladesh and India,participation quickly uncovered the necessary adapta-tions that turned projects in roads, lighting, reliablepower, and better services into good investments. Thisin turn facilitated a longer-term view. And the moreprojects that are controlled by community clientsrather than by the Bank or government, the more likelythe projects are to have a long-term perspective.

Experience also shows that participation increasestransparency and accountability—and reduces corrup-tion. When civil works belong to the people who usethem, there is no tolerance for siphoning off projectfunds in bribes. Contractors are grateful for communitysupport and protection from extortion. While closecommunity supervision might lead to more complaints,contractors were able to fix problems more quickly andcheaply, reducing complaints on completion and im-proving relationships with communities.

Scaling-up ParticipationIt takes time and effort to foster participation, particu-larly to scale up participation from a project toprograms, sectorwide reform, multisectoral efforts, ora national strategy. There have been few successes inscaling up NGO involvement from projects to widerprograms. Bank evaluations and research name the

difficulties that the Bank and its clients are likely toface in trying to scale up participation (and promoteconsultation) in national development and countryassistance strategies, or even in advisory and analyti-cal activities. Not all NGOs want to increase the scaleof their operations to work on Bank-supported projects.Scaling up can also strain an NGOs managementcapacity and create financial obligations to staff andclients that are difficult to sustain.

Any scaling up initiative must be preceded by acareful assessment of the capacities and commitment ofall parties. A critical step is to involve all stakeholdersin developing success indicators, a process that canexpose differences in expectations and priorities. Nego-tiating to develop a common framework helps to buildownership of outcomes.

OED findings suggest several lessons for success inscaling up participation:

• Integrate a learning process. Perhaps the bestknown success in scaling up participation is thePhilippines National Irrigation Administration,which adopted a step-by-step approach to build-ing user associations’ capacity to manage localirrigation systems. The use of a learning processapproach and the introduction of a number ofmanagerial innovations allowed local prioritiesand processes to guide the activities of thebureaucracy—rather than the other way around(Korten 1980, 1988). A working group wasformed to act as a catalyst for change. It broughttogether frontline workers and higher-level staffto define a common vision of change, and thenworked through a learning cycle to achieve thatchange over a number of years.

• Beware of procedural inflexibility. Changingorganizational systems and procedures to facili-tate participatory development is a complexundertaking. But the costs of not doing so can beheavy, as in the Government of Uttar Pradesh’sDoon Valley Project in the Himalayas (Shepherd1999). Examples across the world show thatsimply adding participatory rural assessmentsonto existing programs will not lead to partici-pation.

• Avoid rushing to scale up. Participation cannotbe rushed, and it is especially dangerous to try to


5 8

do so on a large scale. The Indonesiangovernment’s attempt to implement nationwidevillage development planning in less than a yearin 1995–96 illustrates the pitfalls of forcedparticipation. There was little or no ownership ofthe process at the village level, and too little timewas devoted to building partnerships withNGOs, whose skills might have enabled thegovernment to do a better job.

• Persevere. The Rural Integrated Project SupportProgram in the Lindi and Mtwara regions ofTanzania is one of the best documented cases of agovernment scaling-up program. This district-level holistic program involved participation bystakeholders in local government, agriculture,natural resource management, transport andmarketing, education, savings and credit, health,and water. The successful collaboration andresults visible today are an outgrowth of theaccumulated experience of repeated mistakesand learning over some 25 years.

• Identify a champion. A strong shift towardparticipation in government or NGO practicecan often be traced to an alliance of individuals

or a single person. These champions have oftenled the battle against institutional inertia, indi-vidual hostility, and political pressures from thewealthy elite. For example, the regional commis-sioners of the Mtwara and Lindi regions inTanzania allied themselves with the Rural Inte-grated Project Support Program to introduceparticipatory approaches.

• Change attitudes and behavior. For participatoryapproaches to work, actors at all levels will needto change attitudes and behavior (Blackburn andHolland 1998). Otherwise a new authoritarianmanager can come into an organization and setthe clock back, ruining a participatory cultureand the practices patiently nurtured by a prede-cessor. The bottom line in participatory changeis individual and personal.

Ensuring attitude and behavior change also poses amajor challenge for the Bank. The Bank may be askedto help the government develop participation—forexample, through World Bank Institute seminars for awide range of participants, including members ofparliament and the press. Cooperation with local

At least fourconditionalityscenarios canA

BOX A5.1 CONDITIONALITY SCENARIOS

be imagined, rangingfrom the most flexibleto the most restrictive.

• Ex post conditionsfor lending to anongoing reformerwith a track recordof success. Here theprocess of mutualcommitment isfully under way.The country ismaking progress inreforms and theBank is lending.An example maybe Indonesia during1985–95.

• A higher-impact struc-tural adjustment lend-ing model of floatingtranches. This modelmight apply to a coun-try seen to be a cred-ible reformer withstrong ownership, butno real track record.The governmentchooses the sequenceand timing of theconditions and trancherelease. Four coun-tries with floatingtranches—Cameroon,Côte d’Ivoire,Malawi, and Tanza-nia—outperformedothers.

• Regular multiple

tranches that could beconverted to floatingtranches as ownershipdevelops or experi-mental single-trancheoperations. Thismodel could apply tocountries withnascent or developingownership, but withrecent changes inregime or othersources of credibilityin reform. The single-tranche operationcould apply to apotential reformer,with the promise ofadditional lending tofollow if reform actu-ally comes about.

• Advice, continuingcontact, and readi-ness to commit sup-port if circum-stances change. Thismodel could applyto a country with agovernment not yetready for reform.

All the scenarios areconsistent with adjust-able conditionality andtranches. As countriesprogress, they wouldmove up the condition-ality ladder. The possi-bility of this kind ofsequencing shouldstrengthen themovement towardmutual commitment.

5 9

research organizations can also help develop policyknowledge and contribute to ownership through theparticipation of local researchers and officials.1

Broad participation in program development, withcooperation between the Bank and the government,should strengthen ownership on both sides. Broadsupport can lend greater credibility to governmentownership, which in turn can result in greater Bankflexibility regarding conditionality.

Ownership and ConditionalityConditionality, through its relationship to ownership,can foster mutual commitment to ongoing reform.Traditionally the Bank has attached conditionality totranche releases on specific, single loans. The CDFsuggests a broader view of conditionality as theoutcome of an ongoing process that allows the bor-rower and the Bank to develop and nurture mutual trustand commitment, a process that adjusts as countryownership grows. Empirical evidence on successfuladjustment lending suggests different forms of condi-tionality for countries at varying phases of the reformprocess. It proposes a new view of flexible conditional-ity, combined with capacity building, to level theplaying field for all parties in the relationship.

Reconciling Conditionality and OwnershipUnder the CDF, conditionality can play a key role inthe Bank’s dialogue with reforming borrowers. Ratherthan an attempt to force externally designed policychanges on unwilling governments, conditionalityshould be used as a credible indicator of commitmentby the Bank and its partners. Under the CDF, condition-ality should represent a transparent and explicit under-standing of sustained Bank commitment in support of anew or ongoing policy reform program formulated bythe country, with wide participation by the privatesector and civil society in cooperation with externalpartners. The program would be owned by the country,and conditionality would define the parameters ofexternal support.

Viewed in this way, conditionality could be appliedto actions already taken—with further support expectedon the assumption that the reform process will continue(box A5.1). As the reform process and ownershipdeepen, conditions would become more flexible. A newreformer might begin with a single up-front experimen-tal tranche, or with normal conditions that could beconverted to floating tranches as the process develops.

The government would choose the sequence and timingof sector reforms as external support is calibrated to thequality of the program. A model might be the relation-ship between a commercial bank and its customers. Aslong as the customer can project a credible path ofearnings, lending continues. Thus conditionality is theBank’s side of a continuing relationship, and ownershipis the country’s side.

It is difficult to establish criteria for ownership.2

The frameworks commonly used to assess ownership—leadership analysis and stakeholder analysis—are help-ful, but insufficient. Reform-readiness analysis is oftenproposed as an alternative, but it does not eliminateuncertainty. Broad participation in program develop-ment, with cooperation between the Bank and thegovernment, strengthens ownership on both sides.Broad support can lend credibility to governmentownership, leading, in turn, to greater flexibility onconditionality.

Conditionality as a Commitment ProcessTwo recent studies support the view of conditionality as acommitment process: a study of HIAL and a reevaluationof the 1998 Dollar-Svensson data using the country as theunit of observation (Dollar and Svensson 1998; WorldBank 1999i). Both studies support the view of condition-ality as a process of mutual commitment.

The HIAL study found better policy outcomes andeconomic impacts for countries receiving HIAL thanfor other groups of IDA countries. This performancegain was attributed to greater selectivity in lending andmore flexibility in disbursement arrangements. HIALintroduced floating tranches, with the governmentdeciding when to meet conditions, which were gener-ally sectoral, and when to draw on the correspondingtranche. Several loans had an initial tranche based onselection criteria (generally macroeconomic stabiliza-tion), followed by floating tranches. Analysis for thisreport of the relationship between conditionality andtranche arrangements and outcomes found that coun-tries with higher-impact adjustment loans performedbetter during the loan period than before, with thelargest average increase in growth of 3.5 percent.While the samples are small, this provides encouragingsupport for flexibility in conditionality and for condi-tionality as a process of mutual commitment.

The 1998 study by Dollar and Svensson of theimportance of country political economy variablescompared with variables under Bank control in predict-

A n n e x 5 . O w n e r s h i p , P a r t i c i p a t i o n , a n d C o n d i t i o n a l i t y


6 0

ing success or failure (as measured by OED ratings)concluded that “the results have clear implications forthe management of policy-based lending. They suggestthat the role of adjustment lending is to identifyreformers, not to create them.” The study uses the loanas the unit of analysis. But structural adjustmentlending tends to be repetitive: 27 percent of countriesreceiving adjustment loans since 1979 received morethan five loans. Past successes could be predictors offuture successes. When the data are re-analyzed usingthe country as the unit of analysis, past successbecomes a highly significant predictor of future suc-cess. This result supports the view of conditionality as a

process of mutual commitment, since the Bank can uselending history to formulate future conditionality.

This view of flexible conditionality differs from theview of conditionality as coercion—an attempt toinduce the borrower to take action against its will.Participation, by involving other stakeholders, reducesthe risks of coercion implicit in a secret negotiationbetween a hard-pressed borrower and a powerfullender. Equally, transparency and participation shouldminimize the risks of shirking or policy reversals.Under the CDF, this form of conditionality is comple-mentary to ownership.

6 1

BOX A6.1 BARRIERS TO PERFORMANCEORIENTATION

ANNEX 6. MANAGING FOR RESULTS

There are growing pressures on both governmentsand donors to make and track progress in achievingdevelopment objectives. These pressures include:

• Citizen demands for more accountability andgreater value for the money

• Scarce donor funds and growing competitionamong countries

• Market perceptions of national economic com-petitiveness

• Greater willingness of civil servants to take asystematic approach to learning from experience.

Developing countries are placing more emphasis ongood governance and are adapting approaches andmodels from developed countries. One popular model isresults-based management (RBM), which focuses ongoals, objectives, and the results chain—the linkage ofresources, processes, outputs, client reach, outcomes, andimpact. It emphasizes the perspectives of clients and otherstakeholders, and can help build a client-oriented serviceculture. At the core of RBM is a performance measure-ment system focusing on results, not just inputs, processes,or narrowly defined outputs. A number of governmentshave implemented elements of RBM, and their experi-ence—particularly in measuring government perfor-mance—provides key lessons for a more performance-oriented approach to government (Mackay 1998b;OECD 1995, 1997a).

RBM and the CDFThe CDF implies an RBM approach: keep the focus onachieving development effectiveness. RBM is a means toevaluate CDF processes, and thus to help governmentsand donors learn what works, what does not, and why,from the CDF pilots and their approaches. CDF countries,which will adopt different approaches built around thefour core CDF themes, can share lessons about thestrengths and weaknesses of these approaches.

Experience shows that countries must demand afocus on results (UNDP 1997); donor-driven efforts arenot enough. Donors can “sell” a greater performanceorientation to countries, but an infrastructure thatmeasures performance—through systematic M&E—will founder unless the government accepts the donor’sarguments or reaches its own conclusions about themerits of such an approach (box A6.1).

Countries and donors often disagree on ways totrack implementation and impact, and few localinstitutions can bridge the gap between the technicaltask of research and evaluation and the political task ofpolicymaking. The experience of the U.S. Agency forInternational Development (USAID) suggests that coun-tries are more likely to monitor policy implementationand systematically use the findings in decisionmakingwhen:

• Think tanks and other credible, nongovernmentparties help monitor.

• Explicit responsibility is assigned for translatingresearch and evaluation findings into policyimplications.

• Evaluation findings are discussed in public forums.

It may take a decade or more to build a strong,systematic approach to measuring government perfor-mance. The challenge goes beyond building skills andcivil service infrastructure. It also encompasses build-ing sustained government and civil service commit-ment to performance orientation.

Incentives are crucial to ensuring a performanceorientation (Picciotto and Weisner 1998). To under-stand a country’s incentive frameworks, one mustdiagnose its institutions and consult closely with itsgovernment (as in Zimbabwe) (Mackay 1998a, 1998b)and, where agreeable to government, with civil societyand the private sector. M&E findings can enhanceaccountability, improve manager performance, andhelp guide budget allocations (including budget cuts).

Where government performance is poor, M&E can

• Poor demand and ownership in countries• Lack of a culture of accountability (often re-

lated to ethics or corruption)• Absence of evaluation, accounting, or audit-

ing skills• Poor quality of performance information, and of

accounting and auditing standards and systems• Little integration of M&E findings and

mechanisms into decisionmaking.


6 2

BOX A6.2 POOLING AND JOINT REVIEWS IN GHANA’S HEALTH SECTOR REINFORCE ACCOUNTABILITY

hana’s healthsector demon-strates an

evaluated and allows anemphasis on joint account-ability for the donors andthe government.

The government andthe donors conduct ajoint review of the healthsector every year byassessing 20 performanceindicators. They alsoidentify priorities for thecoming year. Two suchjoint reviews have beenconducted since 1997.

Although close govern-ment-donor collaborationhas been effective, thereis still work to be done.The joint review facedproblems such as:

• Incomplete evalua-tions. In-depth evalu-ations of selectedproblem areas—pub-lic health, institu-tional care, andsystems of manage-ment and support—had been planned aspart of the 1998review, but were notcompleted.

• Deficiencies in thehealth ministry’scapacity to collect,analyze, use, andreport performancemonitoring data,relating not only tothe ministry but also

to the sector.• Lack of evaluations

on the outcomes orimpact of healthministry programs.

• Inability of thereview to take fullaccount of theactivities andservices providedby bodies externalto the health minis-try—such as otherministries, NGOs,and the for-profitprivate sector.

• A need for greaterdialogue with civilsociety, includingthe district level.

Ginnovative approach todonor funding andreviews of sectorperformance. Donorshave agreed to pool asignificant proportion oftheir funds for this sec-tor, giving up the rightto select which projectsto finance in exchangefor a voice in develop-ing sectoral strategy andallocating resources.This pooling approachmeans that the cost-ef-fectiveness of specificdonor funds cannot be

help to improve accountability. Measuring govern-ment performance increases the pressure for a moreresponsive public sector. In Bangalore, India, forexample, an NGO surveys citizens on the quality ofmunicipal government services and the extent ofcorruption in delivery (Paul 1998).

RBM has synergies with many types of publicsector reform. These include public expenditure man-agement (medium-term expenditure frameworks andperformance budgeting), civil service reform (perfor-mance contracts for senior civil servants), intergovern-mental fiscal decentralization, commercialization andprivate sector delivery of public services, servicedelivery or beneficiary surveys, participation and the“voice” of civil society, and anticorruption efforts thatstrengthen financial management and oversight bodies.Efforts to foster RBM are thus likely to benefit fromclose links with related public sector reforms.

A holistic approach to governance encouragesgreater performance orientation. A cross-sectoral per-spective puts pressure on lagging ministries to keep upwith leading ministries. But experience also suggeststhat a holistic approach to RBM will not be easy totrigger across all ministries. It might be more realistic

to start with performance monitoring in selected sectorsor ministries, and then extend the approach to others,and add measurement tools such as project andprogram evaluation as opportunities allow.

Chile took such an incremental approach (Marcel1999: 265–325). Initial monitoring identified someproblem programs, and this led to questions aboutprogram outcomes and impact that have since beenanswered by in-depth evaluations.

An emphasis on performance management forprojects and programs should be complemented by afocus on governance and intersectoral synergies. Do-nors have attempted to foster good governance and aperformance culture at the national and ministrylevels. But attempts to strengthen M&E have focusedon the project level, typically only to satisfy donorrequirements. The lack of domestic ownership of M&Ehas undermined acceptance of performance measure-ment for sound governance. Moreover, the limitedM&E capacity created through donor-driven, project-based activities was often lost once the project wascompleted. Ghana is starting to work with donors,including the Bank, to build more robust national andsectoral approaches to performance management and

6 3

measurement. Donor adoption of a sectorwide ap-proach helps this process along (box A6.2).

The CDF provides an opportunity to involve civilsociety in assessing government performance and indiscussions about national and sector priorities andprogress in achieving development goals. Civil societyhas a number of potentially important roles to play. Itcan provide a deeper understanding of the constraintsand limitations on government activities. It can pres-sure governments and civil servants to improve man-agement of inefficient or ineffective government activi-ties.1 It can use and learn from information onperformance indicators and the findings of evaluationsand reviews. It can voice views of ordinary citizens,NGOs, and the private sector—the consumers ofgovernment goods and services. The beneficiary assess-

ments now conducted in many countries provide suchopportunities. Civil society can also influence theevaluation agenda by identifying government activitiesin need of assessment. Box A6.3 demonstrates anothertype of civil society involvement.

Donors and governments both have a role to play inbuilding the capacities of key stakeholders such as NGOs,the media, and parliaments. These stakeholders need tohave some familiarity with performance concepts andM&E tools to make full use of information on govern-ment performance; they also need to understand itslimitations and weaknesses. In South Africa, for example,an NGO works directly with parliament on a fee-for-service basis to help it understand and digest informationon the performance of government ministries.

n late 1997, theMalaysian State ofPenang initiated the

the extent to which citizensfeel they have a voice inthe changes affecting theirneighborhoods.

The initiative wasdesigned as a pilotproject for communityindicators. State assemblymembers, state govern-ment officials, businessrepresentatives, NGOs,community-based organi-zations, academics, andconcerned individualsorganized into fiveroundtable groups of 50people each. The groupsfocused on the economy,environment, socialjustice, culture, and popu-lar participation.

During one- and two-day facilitated discus-sions, each groupprepared vision state-

ments and made a list ofkey performance indica-tors for its topic. Thegroups chose indicatorsbased on their rigor,relevance, and availabil-ity. In community healthcare, for example, indi-cators included cancerrates, infant mortality,quality of health care,levels of patient satisfac-tion, and the ratio ofhealth care facilities andprofessionals to the popu-lation. The groups useddata to identify trendsand implications for thesustainability of develop-ment, and their relation-ship to other indicatorsor issues.

In late 1998, theroundtable findings werepublished in the first

People’s Report onPenang, and reviewedby the state assembly.The intention is that theroundtable groups willreconvene each year toreview improvementsand identify problemareas.

The initiativeappears to have beensuccessful in generat-ing stakeholderownership of evalua-tion, showing thatRBM is possible atthe community level.The initiative alsodemonstrates the po-tential for achievinga collaborative part-nership approachamong diverse groupsand interests.

IBOX A6.3 MALAYSIA: SUSTAINABLE PENANG INITIATIVE

Sustainable PenangInitiative, with supportfrom the CanadianInternational Develop-ment Agency, UNDP,and the UN Economicand Social Commissionfor Asia and the Pacific.The initiative is a long-term pilot response tocommunity concernsabout the costs ofPenang’s rapid develop-ment: the sustainabilityof growth; its environ-mental impact and theconsequences for health;the distribution of gainsfrom development; theimpact of growth onPenang’s values, tradi-tions, and heritage; and

A n n e x 6 . M a n a g i n g f o r R e s u l t s

6 5

Since its emergence some three decades ago, theconcept of partnership for development has won broadacceptance. Donors and beneficiary countries are ex-pected to participate jointly in analyzing macroeco-nomic issues, developing action plans for nationalpriorities, and strengthening or building local institu-tions. Ultimately, partnership is expected to transformbeneficiary countries from aid recipients to aid manag-ers. Some governments—Bangladesh, Bolivia, Ethio-pia, Ghana, Tanzania, Uganda, and Vietnam—havetaken the initiative in setting the agenda for local aidmeetings and sharing a lead role in annual or biannualapex Consultative Group (CG) meetings. They are alsoparticipating more intensively in preparatory activitiesand analytical work, such as country economic memo-randa and sector reports, that affect aid coordinationexercises.

The idea of partnership was first articulated in the1969 report of the Commission on International Devel-opment, Partners in Development, commissioned byWorld Bank President Robert MacNamara. The reportcalled for a “new partnership” based on reciprocalrights and obligations. Not until 1995, however, waspartnership officially made one the of Bank’s six“guiding principles.”2

Partnership gained additional prominence inDAC’s seminal document, Shaping the 21st Century:The Contribution of Development Co-operation(OECD and DAC 1996). It put forward internationallyagreed goals as part of a new partnership agenda forpoverty reduction that calls on developing countries to

ANNEX 7. PARTNERSHIP AND STRATEGIC SELECTIVITY1

society and with external partners.” In a furtherrefinement, the World Bank’s Partnership and Develop-ment (World Bank 1998i) defines partnership as acollaborative relationship in which parties jointlydevelop and agree on objectives and expected out-comes, division of labor and responsibilities, rights andaccountabilities, and capacity development to ensurethat weaker members are able to participate fully.3

These notions are also embodied in the CDF.

Effective Country-Led Aid CoordinationThis notion of partnership implies a matrix such as thatshown in table A7.1, with country-driven arrangementsexpected to make the greatest contribution to develop-ment effectiveness.4 High country commitment andinstitutional capacity are likely to be associated withcountry-driven aid coordination arrangements, and lesscountry commitment and institutional capacity withdonor-driven arrangements. Where capacity is high butcommitment is low, the matrix implies joint sponsor-ship, or intensive donor involvement in tandem withthe government to minimize resource misallocationunder flagging or missing policy commitment. Wherecommitment is high but capacity is low, country-drivencoordination is possible if institutional capacity isstrengthened. Institutional capacity includes organiza-tional capability and the rules and incentive structuresgoverning policy formation and resource allocation.Incentives include both the internal incentive (anddisincentive) structures of a bureaucracy and theexternal incentives resulting from market competitionand from the various mechanisms for citizen voice orparticipation (World Bank 1998i: 9–10).5

Donor CharacteristicsHow can a country move to the country-driven quad-rant? Depending on where a country stands, the answeris to strengthen country commitment, strengthen capac-ity, or do both. But the characteristics of donors andinternational agencies also affect the ease of moving toeffective country-led aid coordination arrangements. Themost advantageous case for country-led aid coordinationis that of a small number of donors, all with a highdevelopment orientation (table A7.2). As the number ofdonors increases, the environment for aid coordinationbecomes more challenging, and as nondevelopmentmotives for providing aid (such as commercial interests)

TABLE A7.1 COUNTRY CHARACTERISTICS ANDAID COORDINATION

Institutional Country commitment capacity High Low

High Country-driven Joint-sponsorship

Low Country-driven Donor-driven (with institutional strengthening)

commit to an effective policy environment for pro-poorgrowth, and for donors to increase financial support forsuch policies. The report gave highest priority todevelopment strategies grounded in an “open andcollaborative dialogue by local authorities with civil


6 6

increase, effective country-led coordination becomesmore difficult. While the end of the Cold War has reducedthe dominance of the ideological motive for providingaid, geopolitical considerations are still at work.6

Effective aid coordination involves a process lead-ing to an agreement by the recipient country, donors,and international agencies that all partners will acceptmutual responsibility for achieving development out-comes, along with distinct accountabilities. For therecipient country, accountabilities include making acommitment to sound policies and effective institu-tions. For the donors, accountabilities include adoptinga development orientation, achieving coherence withcountry development priorities and selectivity alongthe lines of comparative advantage (thus addressing the“numbers” issue), providing effective capacity buildingassistance to level the playing field among partners,and supporting country-led efforts to achieve theseobjectives.

The Limits of PartnershipThere are limits to the extent to which the realities ofpartnership match the vision. These include tensionsbetween harmonization and accountability and be-tween conditionality and partnership, as well as issuesof governance constraints and strategic selectivity.

Harmonization and AccountabilityHarmonization of donor policies and procedures does notremove the need for recipient accountability for their useof funds to meet the donors’ fiduciary responsibilitytoward their governing bodies or parliaments. Account-ability, however, has been one-sided. Donors have notbeen held accountable to recipients for the methods orrationale of the allocation of aid resources or thepredictability of resource availability. The reluctance ofsome donors to untie aid illustrates this point.

Conditionality and PartnershipIn normal business partnerships, the partners mutuallyagree to abide by certain conditions and obligations. In

development cooperation, the donor normally imposesconditions on the recipient, which must abide by them.The relationship between the two is far from symmetrical.It is embedded in an institutional culture of donorinstitutions that views those who offer aid as belonging toa fundamentally different league from those who receiveit. Changes in rhetoric have not altered the aid culture,even under the pressure of determined leadership.

Governance ConstraintsGuided by its Articles of Agreement, the Bank dealsprimarily with governments. Partnership has meant, forall intents and purposes, an effort to improve relationshipswith government borrowers. This has constrained theability to form ties with the private sector and civilsociety. In recent years the Bank has expanded themeaning of partnership to include NGOs, civil societyorganizations, and the private sector. But governanceconstraints impose limits on such partnerships, particu-larly where NGOs and other civil society bodies operateunder the government’s wings or on the edge of legiti-macy. True partnership, based on open and free dialogue,is not possible in these conditions. As the Bank’s GeneralCounsel pointed out, “the Bank does not have a legalright to oversee the governance of its borrowing membersor to participate in such governance. It is neither a worldpoliceman, nor a world government.” Bilateral donorsare far less restricted in their ability to offer advice onpolitical issues.

Strategic SelectivityPartnership and selectivity are inseparable, both con-ceptually and operationally, which makes it difficult todistinguish selectivity from partnership without seri-ously compromising partnership.7 Strategic selectivitymay be exercised on the basis of several characteristics,each with distinct implications:

• Need. Donors may disagree with the governmentor each other about which problems, sectors,regions, or groups are priorities. Donors mayfocus exclusively on “fashionable sectors,” suchas health and education, at the expense ofcomplementary investments, such as rural roads.

• Comparative advantage of donors. In an idealworld, donors would divide their labor accord-ing to comparative advantage. In practice thereis no ready agreement on what this means.Consider the often difficult relationships between

TABLE A7.2 DONOR CHARACTERISTICS AND THEAID COORDINATION ENVIRONMENT

Number of Development orientationdonors and agencies High Low

Low Favorable Relatively Difficult

High Challenging Very Difficult

6 7

uch can belearnedfrom the

ances to increase reach,mobilize resources,source new knowledge,lower overhead costs,increase their responsive-ness to clients, and focuson core businesses andcompetencies. They areusing the Internet forcollaborative planningand facilitating learningand action on projects

across geographical loca-tions and teams. Corpo-rations are becoming, ineffect, multiorganiza-tional or networked orga-nizations. Successful stra-tegic alliances depend ona clear strategic fit andmutual agenda, incre-mental processes ofincreasing involvementand sharing information,

cultural fit and part-nership orientation,collaborative plan-ning and learning,dedicated inter-organizational com-munication, invest-ment in mutuallybeneficial goals, rein-forcement of partner-ship successes, and asense of “co-destiny.”

Source: Buono 1997; Hamel and Prahalad 1994; Johnston and Lawrence 1988; Kanter 1994.

MBOX A7.1 LEARNING FROM BUSINESS ALLIANCES

corporate world aboutpartnerships and strate-gic alliances, particu-larly about definingand exploitingcomparative advan-tages among partners.Corporations are enter-ing into strategic alli-

the Bank and the specialized agencies of the UN,or among UN agencies themselves. The com-parative advantage of donor institutions canconflict with commercial interests.

• Country capacity. Resources should be allocatedto the most efficient institutions in the recipientcountry. Pushed too far, this type of selectivitycan undermine the balanced approach called forby the CDF, or force a tradeoff with capacitybuilding and ignore dynamic comparative ad-vantage. Supporting the creation of project man-agement units, for example, may undermineoverall capacity building.

• Country performance. Selectivity, in the sense of areward for work well done, may be an alternativeto conventional ex ante conditionality, which is anincentive to do a good job. In a donor’s globaloperations, selectivity implies a bounded applica-tion of the partnership principle: where countrycommitment is lacking, recipients must first qualifyfor partnership by meeting a minimum level ofperformance. Practical difficulties have arisen withselectivity based on performance. The chairman ofthe DAC warns against donors trying to invest in“winners” and withdraw from “losers.” Thiswould require, he warns, “better forecasting abilitythan anyone claims to possess” (DAC 1999: 4).

The Road to Partnership: Strategy for Government-LedAid CoordinationPartnership requires strong institutions. It also requireswell-developed skills in negotiation, communication,economic and social analysis, information technology,and diplomacy. Participants at a workshop for seniorgovernment officials engaged in aid coordination, spon-sored by OED and the Swiss Agency for Development(SDC) in February 1999, concluded that more investmentwas needed in the skills required for effective aidcoordination, particularly government’s ability to designand carry out economic and social studies and to analyzedonor studies.

The Bank provides considerable funding for capac-ity building through loans, grants, and trust funds, butmost of it is not woven into a national strategy in theway macroeconomic analysis is. Donor efforts tostrengthen country aid management and coordinationcapacity have tended to be expensive, supply-drivenand ineffective, and aid coordination activities havehad little positive effect on capacity or partnership(World Bank 1999m). There is a need for real change.Much can be learned from the corporate world aboutpartnerships and strategic alliances, particularly aboutdefining and exploiting comparative advantages (boxA7.1). The Bank, in consultation with other donors andaffected countries, needs to give force to the concept of

A n n e x 7 . P a r t n e r s h i p a n d S t r a t e g i c S e l e c t i v i t y


6 8

country responsibility, putting the country in thedriver’s seat. The Bank should work with the govern-ment and other development partners to formulate amultiyear strategy to support countries in assumingleadership of CG and consortium meetings and inbuilding aid coordination capacity at the local level.

The Bank and the UN should jointly make capacitybuilding and aid coordination part of the nationalstrategy dialogue with countries at apex aid coordina-tion meetings. Both the Bank and the UN call forownership, harmonization, cost-effectiveness, subsid-iarity (dispersal of authority as close to the grassrootsas government allows), and institutional and indi-vidual commitment and discipline. Partnership wouldbe stronger and the burdens on government lighter ifthe Bank and the UN could follow a single path, rather

Source: World Bank data; UNDP 1996, p 16.

IBOX A7.2 CAPACITY BUILDING ASSISTANCE TO KAZAKHSTAN AND VIETNAM FOR AID COORDINATION

n 1992 the Bankgave an Institu-tional Develop-

its numerous reorganiza-tions, there was consider-able delay in contractingthe long-term adviser.Once on board, however,the adviser established astrong working relation-ship with the staff andcreated a database ofexternally financed tech-nical assistance andinvestment projects. Theadviser worked withNAFI staff to preparedocuments for the firstand second CG meetings.Both the managementand staff of NAFI cred-ited the adviser withstrong knowledge of thesubject and success intransferring knowledgeand skills in coordinatingforeign aid and establish-ing a solid foundation forthe aid coordination

agency. The grant wasfollowed by a technicalassistance loan to supportaid coordination andmanagement.

The UNDP initiated athree-year technical assis-tance project in Vietnamin 1993, also providing along-term adviser, short-term consultants, train-ing, study tours, andsome equipment for theaid coordination unit, theForeign EconomicRelations Department(FERD). Although housedadministratively in theCentral Planning Com-mission, the projectestablished ambitiousgoals for improving theoverall aid managementsystem, from line minis-tries down to the provin-cial level, and for con-

tributing to a “pro-cess” for improved aidcoordination.

The project evalua-tion report noted thatthe project strength-ened the capacity ofFERD and helped thegovernment preparedocumentation for thefirst donor meetingand subsequent CGmeetings. It did not,however, achieve itsambitious goals foroverall improvementof the aid managementsystem. The WorldBank has taken overaid coordinationresponsibility from theUNDP.

The similarity ofBank and UNDP expe-riences in these twocases is striking.

ment Fund grant toKazakhstan’s aid coor-dination unit, theNational Agency forForeign Investment(NAFI), to improve itssystem of aid manage-ment and coordination.The grant wasdesigned to finance thecosts of a long-term aidcoordination adviser, anumber of short-termconsultants, trainingand study tours foragency staff, andprocurement of equip-ment and facilities forthe aid coordinationunit.

Because of NAFI’slack of familiarity withBank procedures and

than moving along parallel and duplicate trails. Effortsby the Bank and the UNDP to strengthen aid coordina-tion capacity in two distinctly different countriesunderscore the potential for greater success throughjoint efforts (box A7.2).

A related issue is the location of CG meetings andother apex-level aid coordination meetings. Most CGmeetings take place in donor capitals—especiallyParis—but a growing number are being held in recipi-ent countries, as recommended by the PartnershipPaper and by the 1998 evaluation of the SpecialProgram of Assistance for Sub-Saharan Africa (WorldBank 1998k: 17). Other recommendations to encouragebroader participation include having the host govern-ment chair aid coordination sessions and frame theagenda. Box A7.3 highlights the advantages and

6 9

Advantages• Participation by the highest levels of govern-

ment leadership• Increased sense of ownership by

government and subsequent commitment• Greater public scrutiny and collaboration with

civil society and private sector• Reduced perception of donor domination• Broader participation by government

agencies• First-hand view of the partner country for

donor headquarters representatives.

BOX A7.3 ADVANTAGES AND DISADVANTAGES OF IN-COUNTRY AID COORDINATION MEETINGS

Disadvantages• Possible downgrading of the level of donor

representation (but not necessarily across theboard)

• Government preference in some cases formeeting outside the country because of thesensitivity of CG issues

• Less free expression of opinions by local donorrepresentatives, who also may not have thepower to commit their governments

• Loss of ability to take advantage of Paris’scentral location and Bank facilities designedexpressly to meet the needs of CG meetings.

disadvantages of in-country aid coordination meetings.None of this is to suggest that the Bank disengage

from the aid coordination process or from efforts tostrengthen its country programs. On the contrary, byencouraging countries to exercise leadership in aidcoordination, the Bank will be better able to free

A n n e x 7 . P a r t n e r s h i p a n d S t r a t e g i c S e l e c t i v i t y

resources to assist its members in the building of long-term capacity. For the country, the challenge ofresponsibility for leadership in aid coordination shouldstrengthen commitment and ownership—and ulti-mately development outcomes.

7 1

As global integration deepens, the number of develop-ment problems that require supranational policy re-sponses grows. These cross-border challenges arisefrom combinations of market, government, and sys-temic failures. Thus, a new development frontier isemerging, and with it a new role and complementaryrationale for development assistance. A country focuswill continue to be important, but official developmentfinance will likely be needed to meet a large deficit ininternational public goods.

Global forces (including technological change) arecreating not only new, far-reaching opportunities, but alsoa host of potential problems—“international publicbads.” These problems include the possibility of financialcontagion, the spread of disease, the loss of biodiversity

ANNEX 8. INTERNATIONAL PUBLIC GOODS AND AID EFFECTIVENESS

and cultural heritage, cross-border (if not global) environ-mental problems, migrants and refugees, and many formsof illicit transboundary behavior. Another problem—therisk of protectionist backlash—could crop up in industrialcountries in response to rising domestic inequality, laborinsecurity, and disagreements over appropriate environ-mental standards.

Solutions to these problems are similar to interna-tional public goods (box A8.1). Their reach can beglobal (like global public goods), regional (affecting asubcontinent, continent, or hemisphere), or local (af-fecting a small number of neighboring countries).Solutions may affect industrial countries, sets of devel-oping countries (in the case of some tropical diseases),or, increasingly, both poor and rich countries.

ABOX A8.1 PUBLIC GOODS PRIMER

public goodis a commod-ity, service,

of externality. An exter-nality occurs when thewelfare of an agentdepends directly notonly on what that agentdoes, but also on whatothers do or fail to do.The motivation toinvest in public goodsarises from the desire tobring out positive exter-nalities or to correct orcompensate for negativeones. Collective (orgovernment) action isnecessary to producepublic goods becauseprivate solutions oftenfail. In particular,markets are unable tosupply nonexcludablegoods.

Public goods arecritical to development.Rules and standards,infrastructure, institu-tions in public service,

property rights, law andorder, and, more gener-ally, functional socialand political cohesionare developmentresources with a publicgood character. Societ-ies at different levels ofdevelopment distinguishthemselves by theiraccumulated wealth ofpublic goods (the non-physical componentsreferred to as socialcapital by someauthors).

Public goods differaccording to theirpublic benefit on acontinuum betweenpure public goods andpure private goods.Private commoditiesand services are rivaland excludable. In be-tween these polar oppo-sites are club goods

(toll roads), which areexcludable butnonrival, and com-mon pool goods orcommon property(groundwater ormineral deposits),which are nonexclud-able but rival. Com-mon property tends tobe overused in theabsence of rules. Aswith pure publicgoods, the prudent orsustainable use ofcommon property is amatter of collectivechoice. Governmentaction (for example,in the form of regula-tion) may also beneeded to ensure equi-table and competitiveaccess to club goods.

or resource whoseconsumption by oneuser does not reduceits availability toother users. Publicgoods are nonrival inconsumption andnonexcludable; thatis, the provider of thegood cannot preventsomeone from con-suming it, regardlessof whether the userpays for the good.Because of this char-acteristic, publicgoods—such as cleanair, national defense,and street lighting—tend to be under-supplied.

A close relative ofthe concept of a pub-lic good is the notion


7 2

Aid Effectiveness LinkagesSeveral strategic and process-related considerationsthat change the focus and resources of internationalpublic goods could change the effectiveness of aid.First, if a development problem such as a banking crisisarises from transnational spillover, it cannot be re-solved by the traditional approach of an uncoordinatedset of national overtures. “In contrast to conventionalforeign aid that focuses on individual countries,transnational problems demand a multicountry, prob-lem-oriented approach to development cooperation”(Gwin 1999). Although intervention at the nationallevel may also be necessary, sector-specific nationalaction is usually most effective when embodied in amulticountry framework.

Second, systemic crises—international conflict;cross-border environmental, financial, food, or naturalresource crises; and disruptions brought about byexcessive social inequality—can destroy foreign aid–financed development achievements. Crises canquickly spoil the fruits of past economic growth inemerging markets and commodity-exporting develop-ing countries. Some of this growth may have beenfinanced by foreign assistance. Such a setback wouldforce future official flows to affected countries into lessfavorable initial conditions.

Third, the aid process can undermine its effectiveness.

The CDF addresses several defects in the (nationallyfocused) mainstream aid delivery pattern, including:

• A tendency toward top-down and spending-oriented approaches at the expense of localcapacity building and ownership

• Fragmented aid delivery with large numbers ofinsufficiently coordinated sources of assistanceand projects relative to absorption capacity

• Questionable aid allocation patterns.

The principles by which the framework seeks toincrease development effectiveness—such as partner-ship; country ownership and involvement; a resultsfocus; and a long-term, holistic view—are preciselythose that make for successful, global, collaborativeprograms.

Two long-standing international public goods ven-tures, the Onchocerciasis Control Program in WestAfrica (box A8.2) and the Consultative Group onInternational Agricultural Research (CGIAR), confirmthe potential effectiveness of highly focusedmulticountry and multiactor partnerships. Both pro-grams have existed since the early 1970s and showexceptionally high rates of return to investment. Thesuccess of these programs demonstrates that, withproper leadership, issue-oriented international cam-

nchocersiasis,or riverblindness, is

highly successful, withan overall economic rateof return of 20 percent—600,000 cases have beenprevented, 34 millionpeople have beenprotected, 5 millionyears of productive laborhave been added, and 25million hectares of landhave been freed of thedisease. Partners in theprogram include govern-ments, local communi-ties, international organi-zations, bilateral donors,

corporations, founda-tions, and NGOs. Inter-vention focuses on vectorcontrol, drug distribu-tion, and capacity build-ing in national healthprograms.

The program demon-strates that partnershipsare complex and difficultto establish and main-tain, particularly giventhe mix of corporatecultures. To replicate theprogram’s success, it isimportant to identify the

comparative advan-tages that each partnerbrings and to allocateresponsibilities accord-ingly, with preciseobjectives. Leadership,personal relationships,and trust are crucial.All partners mustperceive a payoff toparticipation. Creditshould be shared liber-ally and frequently.Flexibility and compro-mise are fundamental.

OBOX A8.2 PROGRESS IN CONTROLLING RIVER BLINDNESS

a painful and debili-tating disease thatinfects 20 millionpeople in Sub-SaharanAfrica and places 120million others at riskof contracting the dis-ease. A multipartnerOnchocersiasisControl Program hasoperated in WestAfrica for 25 years.The program has been

7 3

paigns can produce strong results. For example, theprogram to eradicate river blindness appears to havefostered a disciplined process and motivated partici-pants to persevere.

An assessment of the CGIAR reaches similarconclusions (Anderson and Dalrymple 1999). (Thisassessment cautions, however, that “in reporting re-search accomplishments, the basic problems are aggre-gation and attribution.”) The CGIAR—a partnership ofgovernments, multilateral institutions, and founda-tions—has catalyzed international collective action inthe service of world food security. The program isdevoted to sustainable crop improvement, especiallyfor staple foods consumed by the poor. The CGIAR hasgenerated impressive global externalities and can beviewed as a model of transnational standard-settingand governance in its field. It has often been suggestedthat something similar be created to address tropicaldiseases or, more specifically, the underresearched andunderattended “orphan” diseases that account for mostof the disease burden in poor countries. A recent movein that direction is the Global Forum for HealthResearch, established in 1997 as an independent,multiactor foundation hosted by the World HealthOrganization to correct the “10/90 disequilibrium.”(Only 10 percent of annual global spending on healthresearch in the private and public sectors is devotedto the health needs of 90 percent of the world’spopulation.)

Based on examples from the Onchocerciasis Con-trol Program and the CGIAR, it appears that mission-oriented transnational networks that address highlyvisible and urgent human priorities can serve as arallying device for the coordination of contributingpartners. Selectivity is ensured up-front through thechoice of public goods to be created. Fragmentedbehavior and free-riding are hindered by the visibilityof the program and the public support associated withits objectives. Shared learning occurs (or should occur)as a matter of course. Motivation and coordinationamong donors and partners are easier to achieve, andconditionality and allocation of donor funds are lesscontentious than in multiobjective, multisector countryassistance programs. There is also less scope forpolitics to interfere with technical integrity. Developingcountries should be strengthened by the presence ofthese programs in the Networks—a presence that isindispensable to finding policy solutions to cross-border externalities within countries.

More thought needs to be given to the interactionamong international and national public goods.Transnational policy should address the synergiesamong investments in development resources at theinternational and national levels, which could be a newdimension of aid coordination. The implication is notthat investments in international public goods shouldwait until conditions are right for their application inmost countries, but that conditions on the ground mustbe nurtured so that international development goodscan be put to use. Current practices, even among suchsuccessful programs as the Onchocerciasis ControlProgram and the CGIAR, could do more to promotenational public goods. Aid coordination partnershipshold the promise that improvements to domestic pro-grams will reveal that norms and standards can besuperior alternatives to conditionality.

Institutional Implications and OutlookThe domestic and international dimensions of thedevelopment challenge are becoming increasingly in-terrelated. Issue-focused international partnerships fordevelopment are mushrooming and are likely to con-tinue to grow in number. Both domestic and interna-tional measures can help address undesirable spilloverand secure ownership for domestic reform and increasethe effectiveness of sectoral interventions at the na-tional level. Reform becomes less dependent on intru-sive conditionality as the policy gap between theregional and global levels closes. It is easier toencourage voluntary compliance with generally ac-cepted international standards than to introduce top-down conditionality. However, voluntary compliancerequires that developing countries be adequately repre-sented in the partnerships that develop the standardsand that coordinated national capacity and institutionbuilding strengthen the ability to implement sectoralreform. Unfortunately, there is a deficit in both areas.

The CDF provides a formula to address thesechallenges. The framework’s key tenets of inclusivenessand wholeness should be respected when partnershipsat the regional and global levels are being built. Withguidance from the framework, prioritization at thenational level can help identify areas where interna-tional programs are needed to supplement nationalefforts. Capacity building in national and local institu-tions (state and nonstate) is critical to the effectiveimplementation of coordinated efforts.

A n n e x 8 . I n t e r n a t i o n a l P u b l i c G o o d s a n d A i d E f f e c t i v e n e s s

7 5

ANNEX 9. RESULTS-BASED MANAGEMENT GLOSSARY

This Glossary defines key terms as they are used in thisreport, and builds on definitions in the report “Gover-nance in Transition” (OECD 1995).

ACCOUNTABILITY is the obligation to demonstrateand take responsibility for performance in light ofagreed expectations. It can take place in relationshipsother than the hierarchical, even when there is noactual “conferring” of responsibility. With a movetoward a partnership model of programming anddelivery, this new concept allows for mutual account-ability, and thus a more mature relationship betweenthe Bank and the borrower or other codeliverers. Inorder for such a definition to be effective, partners mustjointly clarify and set goals and responsibilities; perfor-mance expectations that are balanced by the commen-surate resources of each party; credible reportingmechanisms to demonstrate performance achieved andwhat has been learned; and reasonable review andadjustment systems to ensure that feedback on theperformance achieved and difficulties encountered canbe recognized and corrected as necessary.

EFFECTIVENESS refers to the extent to which objec-tives (of an organization, policy, or program) areachieved, or the relationship between the intended andactual effect of outputs in the achievement of objectives(for example, the extent to which the condition ofhospital patients improves as a result of treatment).

EFFICIENCY refers to the relationship between re-sources (inputs) used and outputs produced (for ex-ample, nurse hours per occupied hospital bed day). Anefficient activity maximizes output for a given input, orminimizes input for a given output. Efficiency mea-sures take the form of output–input ratios (productivity)and expenditure–output ratios (unit cost).

EVALUATION is the assessment, in as systematic andobjective a manner as possible, of an ongoing orcompleted project, program, or policy and its design,implementation, and results. An evaluation shouldprovide information that is credible and useful, en-abling the incorporation of lessons learned into thedecisionmaking process.

IMPACTS are the long-term effects and changes thatresult from the outcome of an activity. They are theultimate criterion for development effectiveness. Forexample, an impact of a microenterprise developmentprogram might be an “increase in the rate of employ-ment.”

INPUTS are the total resources available to carry outan organization’s activities, including the materialgoods, financial resources, and human time and effort.

OBJECTIVES usually serve three different functions: todescribe the future the organization is trying to achieveand give guidelines for the organization’s activities; tojustify the organization’s existence; and to provide thebasis for evaluation. Ideally, objectives and strategiesform a shared, consistent, and integrated hierarchicalsystem, moving from a general vision to a morespecific and concrete direction for organizationalactivities.

OUTCOMES are the immediate effects and changesachieved in relation to objectives (for example, usingfewer resources compared with plans, previous perfor-mance, or the performance of other organizations). Theoutcome of a health publicity campaign might be a 5percent increase in awareness among those targeted.

OUTPUTS are the direct products of an organization’sactivities in goods or services (for example, number oftraining persondays by type of training course). Thissays nothing about the actual outcome (such as skillsabsorbed, or whether the skills helped gain long-termemployment).

ORGANIZATIONAL CULTURE can be described as theshared understanding and interpretation of the world thatemerges in an organization when its members interactwith each other and their surroundings. The cultureprovides the basis for the informal aspects of organiza-tional life—values, attitudes, and behaviors. The bestway to energize an organization is to create andadminister culture. What makes a manager a goodmanager is her or his ability to foster culture.


7 6

PERFORMANCE INDICATORS are used to proxyquantitative measures when output or performance isnot directly measurable. They do not necessarily coverall aspects of performance, but they can providerelevant information toward the assessment of perfor-mance (for example, qualifications obtained through atraining scheme, or hospital admission rates for infec-tious diseases).

PERFORMANCE MEASUREMENT is the compara-tive assessment of policy outcomes, outputs, andinputs; performance measures are most useful whenused for comparisons over time or among units per-forming similar work.

REACH refers to the process of client orientation. Itmeans:

• Identifying the client of the program or service,and the target group within the client group (suchas older women or poor women)

• Identifying results desired in terms of client needsor problems to be solved

• Designing indicators that can be tested by clientsatisfaction or changes in client behavior.

RESULTS-BASED MANAGEMENT (RBM) is an ap-proach to management that has been adopted by manyleading private corporations and government adminis-trations worldwide, for the purpose of providing acoherent framework for strategic planning and man-agement based on learning and accountability in adecentralized framework. All RBM systems are charac-terized by the following features:

• Clear corporate goals and objectives• A performance measurement system focusing on

results• A learning culture grounded in evaluation• Stakeholder participation at all stages of pro-

gram design and implementation• Clear accountabilities in a decentralized frame-

work• Links among results, planning, and resource

allocation• Client orientation (reach).

THE RESULTS CHAIN reflects the RBM framework—from inputs to outputs, through reach (client/benefi-ciary orientation), to outcomes, and ultimate impacts.

7 7

ANNEX 10. MANAGING DEVELOPMENT EFFECTIVENESS: AN OVERVIEW FROM THE CODE CHAIRPERSON

Despite the aftershocks of the 1997 financial crisis, thequality of the active portfolio improved significantlyover the past year, due in part to an increased numberof poor-performing closures, a significant achievement.Currently, 20 percent of commitments and 19 percent ofactive projects are considered at risk, by the end ofFY99, compared with 21 percent and 25 percent,respectively, a year earlier. The large reduction inactive projects at risk reflects not only a broad-basedimprovement in the estimated prospects of ongoingprojects, but also the addition of new projects approvedafter the financial crisis and the portfolio cleanup.Significant differences remain though across sectorsand regions. Performance in the EAP and LCR regionshave been restored to pre-crisis levels, but ECA perfor-mance fell sharply due to the decline of the Russiaportfolio. The electric power and energy sector remainsthe riskiest in the Bank, while weaknesses in publicsector management performance are a matter forconcern given the importance of capacity building.

Portfolio cleanup in FY98 and FY99 has caused aplateauing in quality at exit trends below the StrategicCompact target of 75 percent satisfactory. The ARDEreports the proportion of exiting projects with satisfac-tory outcomes, as evaluated by OED, declined slightlyfrom a peak of 74 percent in FY97 to 72 percent inFY98 and 70 percent for a sample of FY99.Sustainability and institutional development impact forthese projects, while improving, remain far too low.Results for adjustment lending, led by sector-specificoperations, are a notable exception to this stagnantperformance—reaching 85 percent satisfactory forFY98–99 exits. There is a growing gap betweencomplexity and capacity, confirming the need forincreased emphasis on capacity building as proposedby the President in his Annual Meetings speech.

Individual CDF principles, solidly grounded in develop-ment experience, are valid on their own, but it is theirdemanding combination that promises to deliver highrewards. The ARDE presents ample evaluation evidencein support of the individual CDF elements. It alsohighlights the challenges inherent in implementing theCDF principles in a synergistic way. Promising ap-proaches to managing these tensions exist, but theirapplication will require considerable retooling, increased

flexibility, along with new skills and attitudes, a listeningand adaptive mode, and far greater strategic selectivity incooperation with partners.

Improving Performance Management TensionsUnder the Strategic Compact, considerable progresshas been made in clarifying the Bank’s values, reorient-ing its mission, and realigning its organization andskills. At the operational level, the existing set ofinstruments for portfolio monitoring is proving ad-equate. However, the easy gains in portfolio perfor-mance have been realized and further improvementswill become increasingly difficult. While it should befeasible for realism and proactivity to hit StrategicCompact targets, key questions remain on how toaccelerate the pace of improvements in quality at entry,supervision, and economic and sector work. Concernsalso exist with respect to the effective management ofsafeguard and fiduciary policies, particularly their lessstringent application in adjustment lending.

Managing Operational Risks and Country ManagementThe CDF is an apt response to the need for greaterdevelopment effectiveness in an environment of stag-nant aid flows and greater public pressure for account-ability and development impact. The changing envi-ronment has increased the demands on the Bank. In thiscontext, CDF implementation offers the promise of highrewards while creating new demands and pressurepoints in an already stretched organization. The ARDEhighlights the key challenges and tensions, as well aspromising approaches that can serve as a starting pointfor addressing them.

Evaluation AgendaSince the new evaluation framework was endorsed bysenior management and the Board in 1997, considerableprogress has been achieved toward its goals. Completionreporting has continued to improve, self-evaluation ofanalytical and advisory activities has continued to ex-pand, and evaluation is now better connected to theknowledge management system. In contrast, oversight ofsafeguard policy compliance requires continued strength-ening, monitoring and evaluation has made negligibleprogress, and program coordination among evaluationand control groups remains weak. Finally, evaluationcapacity development in borrowing countries is not being


7 8

Jan PiercyChairperson, CODE

given the priority it deserves. In sum, the 1997 evaluationstrategy remains relevant, but significant enhancementsin independent and self-evaluation practices are stillneeded to fulfill the organizational learning potential ofthe function.

In particular, four items require priority attentionin the months ahead: a) review and reform of thequality assurance system; b) improving monitoring andevaluation through capacity development inside andoutside the Bank; c) harmonization of performanceindicators at the project, country, and sector levels; andd) improved coordination and user friendliness ofevaluation and control programs.

For OED, the priorities are: a) a greater focus ongetting results from performance audits and evaluationstudies, to be achieved through increased participationof borrowers and the poor in evaluations; b) promotionof implementation of the ICR reform by the Opera-tional Services Board, OCS, and the Regions forincreased learning; c) more attention to financialaccountability, social development, institutional as-pects, safeguards, and private sector development; andd) more effective outreach with respect to evaluationcapacity development, harmonization of evaluationmethods, coordination of evaluation programs, andevaluation alliances.

7 9

ENDNOTES

Chapter 11. Table A3.2. India and China are excluded because they

represent exceptional cases; that is, the Bank’s performance islikely to have a modest influence on country results (World Bank1997g).

2. For example, Borrower Ownership of Adjustment Programsand the Political Economy of Reform (Johnson and Wasty 1993).

3. The background papers for the 1999 ARDE are identified bythe addition of “BP” following the year of publication. All are listedin the “Background Papers” section of the Bibliography.

Chapter 21. See Annex 1 for a graphical treatment of the trend

analysis, along with a discussion of outstanding projects.

2. Assessments of project characteristics come from OEDproject evaluations. Demanding refers to the extent to which theproject could be expected to strain the economic, institutional,and human resources of the government or implementing agency.Complexity refers to such factors as the range of policy andinstitutional improvements contemplated, the number of institu-tions involved, the number of project components and theirgeographic dispersion, and the number of cofinanciers. Riskinessrefers to the likelihood that the project, as designed, would beexpected to fail to meet relevant project objectives efficiently.

3. Consistent with the trends in the APPI, outcome perfor-mance over the decade is higher when weighted by disbursements,with more than 80 percent of projects exiting in fiscal 1998showing satisfactory outcomes.

4. This comparison is limited to sectors with at least 10projects exiting in FY98–99 and with changes in average outcomeof more than 3 percentage points.

5. For reference, investment performance levels are largelyconsistent with Bankwide aggregates.

6. The Quality Assistance Group’s recent quality-at-entryassessment suggests similar concerns, with institutional capacityaspects and risk assessment and sustainability receiving two ofthe three lowest project subratings (World Bank 1999k).

7. The sample of countries with more than 10 Bank-supported projects implemented in 1990–99 was isolated andbroken into groups by project performance in institutional devel-opment. For the countries in each group a time profile of anindependent measure of national bureaucratic quality was re-viewed. The analysis showed a positive time-trend in nationalbureaucratic quality in most countries in the group with highinstitutional development performance, but no clear pattern incountries in the other groups.

8. The sampling strategy provided 50 percent coverage ofadjustment operations and 33 percent coverage of the largergroup of investment operations. Coverage of both was stratifiedacross sectors through random selection.

Chapter 31. A 1998 OED review of World Bank assistance to financial

sector reform found that of 23 countries where the Bank had

provided support since 1985, only 12 had satisfactory perfor-mance. The East Asia crisis is expected to affect the performanceratings in 3 of the 23 countries found to have satisfactory overallratings (Indonesia, Korea, and the Philippines) (Mathieu 1998).

2. The donors urged that by June 30, 1991, or at the latestbefore the end of the IDA9 period, environmental action plans becompleted for all IDA recipients, with priority given to countrieswhere major problems have been identified, and that the results beincorporated into country lending strategies.

3. Other partners perceive a tendency for the Bank to sendlarge missions for too short a time, to produce bulky aidememoires and then disappear, leaving a shell-shocked localgovernment to make sense of the contents and action plans.Rather than facilitating partnership, this increases the transac-tion costs for clients and partners.

Chapter 41. In size and income level, the group is very close to the

middle-income countries. But low-income countries and thosewith large populations are overrepresented. The group is fairlyrepresentative of the Bank Regions, except that Latin Americaand the Caribbean is underrepresented and South Asia is overrep-resented.

2. Investment efficiency is defined as the GDP growth ratedivided by the investment-GDP ratio. This is the inverse of theconventional incremental capital-output ratio (ICOR), adjustedfor changes in terms of trade between investment goods and GDP.

3. The CAE for Côte d’Ivoire suggests that teachers’ wageswere abnormally high relative to per capita GDP. The mediansalary of high school teachers in 1979 was 300,000 CFA francs amonth, equivalent to $30,000 a year at 1996 prices. The reportobserved that internal efficiency suffered from high repetitionand dropout rates, and many students were poorly prepared forentry into the working world.

4. The review examined Bank experience with 83 restructur-ing operations in 46 countries between fiscal 1980 and 1994(Mathieu 1996).

5. The report concluded that “the twin emphases on policyreform and privatization, together with a major retreat fromfinancial intermediation operations (due to their poor results),also meant that the Bank was left with few instruments (andoften little interest) to deal with institutional issues affectingrestructuring. This was also the case with issues of technologydevelopment, labor market legislation, tax incentives, investmentregulations, and trade policy, particularly at the sector and sub-sector levels.”

6. The review covers 19 poverty assessments and updates (allthose done in fiscal 1997 and 1998, and a sample from fiscal 1996).The 1999 review is a follow-up to OED’s first review of povertyassessments, conducted in 1996, which examined poverty assess-ments completed through December 1994 (World Bank data).

Chapter 51. For example, “New Lessons from Old Projects: The Workings

of Rural Development in Northeast Brazil” (World Bank 1993b).


8 0

Within the framework of flexible program design, the better per-forming activities consistently departed from their original design,led by dynamic managers and local involvement (World Bank data).

2. That study also identified an inward-focused Bank cultureas a source of complexity, a culture that emphasized analyticalcomprehensiveness, rewarded larger and more visible interven-tions, discouraged risk-taking and making tradeoffs, diffusedaccountability, and downplayed the need for external feedback(World Bank data).

3. Some types of social fund projects, such as family plan-ning, are not demanded by the community, although theseactivities may have high social returns. The poor in a communitydo not come forward with proposals because they have limitedcapacity to propose projects.

Chapter 61. The Bank, IMF, UNDP, and others should exercise caution

in introducing too many planning and programming instruments.From a developing country viewpoint, these tools may be top-down and confusing foreign concepts and may divert localresources and managerial talents away from building and im-proving existing local strategy development and planning pro-cesses. Developing countries should have a voice in shaping andharmonizing the design and use of these tools.

2. For example, Seeing Like a State (Scott 1998). Suchhegemonic planning and social engineering approaches reflectlittle confidence in the skills, initiative, intelligence, and experi-ence of the beneficiaries. The precision and authority of suchapproaches depended not only on bracketing contingency, butalso on standardizing the subjects of development.

3. Some of the recent organizational changes in the Bank mayactually increase the tensions and challenges in implementing theCDF. For example, some budgetary and personnel policies and thecontracting out of technical services may reinforce rather thanalleviate short-term orientation and incentives, promote specializa-tion rather than integration skills, and link budgetary resources moreclosely to lending rather than recognizing the growing importance ofknowledge, facilitation, and advisory services.

Annex 11. OED has reviewed all implementation completion reports

delivered to date by Regional staff, covering 118 of the 277 projectsexiting the portfolio in fiscal 1999 (43 percent coverage). Comple-tion reports for the remaining projects exiting in fiscal 1999 areexpected to be delivered to OED and reviewed by spring 2000.

Annex 31. Investment efficiency is defined as the GDP growth rate

divided by the investment-GDP ratio. This is the inverse of theconventional incremental capital-output ratio (ICOR), but isadjusted for changes in terms of trade between investment goodsand overall GDP.

Annex 51. As proposed in PREM Note 27 (World Bank 1999j).

2. This is supported by PREM Note 25 (World Bank 1999a).

Annex 61. The Bangalore, India, case cited above is one such

example (Paul 1998).

Annex 71. This annex draws on OED’s forthcoming Aid Coordina-

tion Study.

2. The other guiding principles were selectivity, client orienta-tion, results-orientation, cost-effectiveness, and financial integrity.

3. The United Nations articulated a similar framework atabout the same time. Also see Robert Picciotto, The Logic ofPartnership. A Development Perspective. (OED, World Bank,September 29, 1998).

4. The 1997 ARDE measured policy performance as aproject-weighted index of three components: inflation, fiscalbalance, and openness. For institutional quality, it draws on ameasure employed by the World Development Report 1997(World Bank 1997g), based on a set of responses by foreigninvestors on red tape, the regulatory environment, and the degreeof freedom from political pressure (Evans and Battaile 1998: 15,20). OED evaluative research (Johnson and Wasty 1993) showsthat country commitment and policy performance are closelyrelated.

5. Other characteristics can also have a bearing on theeffectiveness of aid coordination, such as country size and itsstrategic or geopolitical significance. Other things being equal,the larger the country and the greater its strategic significance,the more likely it is that donors will take aid coordinationseriously. But these characteristics are fixed (country size) ornearly impervious to efforts at deliberate modification (strategicimportance). The desire of government officials to maintain thestatus quo has also been cited as an obstacle to country leadership(World Bank 1998i: 34). This is part of institutional capacity,broadly defined to include incentive structures.

6. Another donor characteristic is diversity of procedural andaccountability requirements, which can also impede greatercountry leadership. However, it is plausible that the greater thedevelopment orientation, the greater the willingness of donors toharmonize procedural requirements. The same comment appliesto the internal incentive structures of donor agencies that workagainst coordination (World Bank 1998i: 34). Several replies tothe joint questionnaire sent by UNDP and the Bank to donorsmentioned development orientation and procedural diversity aspotential barriers to effective aid coordination.

7. For a fuller discussion of the subject see ODI, Chapter 2:Partnership and Strategic Selectivity (Maxwell, Simon, Foster,Naschold, and Conway 1999 BP).

BIBLIOGRAPHY

8 1

This report has drawn on a wide range of World Bankdocuments and numerous outside sources. World Banksources include OED’s major studies, Country Assis-tance Evaluations, and project audits, as well asongoing research. These and other sources are listedalphabetically by author or organization in twogroups: Background Papers commissioned for thisreport and a Selected Bibliography.

Background PapersThe background papers are identified in the text by theaddition of “BP” following the year of publication.Selected background papers are forthcoming.

Blackburn, James, Robert Chambers and JohnGaventa. 1999. “Learning to Take Time and GoSlow: Mainstreaming Participation in Developmentand the Comprehensive Development Framework.”Institute of Development Studies, Sussex. Photocopy.

Branson, William, and Nagy Hanna. 1999. “Policy Re-form, Ownership, and Conditionality.” Photocopy.

Crook, Richard C., and James Manor. 1999. “The Po-tential Contribution of Democratic Decentralizationto the Comprehensive Development Framework.”Institute of Development Studies, Sussex. Photocopy.

Ellerman, David. 1999. “Helping People Help Them-selves: The Challenge of Autonomy-Compatible As-sistance.” Photocopy.

Ferroni, Marco. 1999. “Reforming Foreign Aid: TheRole of International Public Goods.” Photocopy.

Fox, James W. 1999. “Applying the CDF to USAIDExperiences.” Photocopy.

Jones, Stephen, and Andrew Lawson. 1999. “ScalingUp: Issues in Moving from Projects to ProgrammaticAid.” Oxford Policy Management. Photocopy.

Maxwell, Simon, Mick Foster, Felix Naschold, andTim Conway. 1999. “Operationalizing the Compre-hensive Development Framework: Evidence fromContemporary Research.” Overseas DevelopmentInstitute. Photocopy.

Selected BibliographyAlbouey, Y., and R. Bousba. 1998. The Impact of

IPPs in Developing Countries—Out of the Crisisand into the Future: Viewpoint. Washington, D.C.:World Bank.

Anderson, J., and D. Dalrymple. 1999. The WorldBank, the Grant Program, and the CGIAR: ARetrospective Review. OED Working Paper Series No.1. Washington, D.C.: World Bank.

Andrews, K. 1987. The Concept of Corporate StrategyHomewood, IL.: Irwin.

Arboleda, J. 1999. Participatory Country AssistanceStrategy in Colombia: A Case Study. Social Develop-ment Papers No. 33. Washington, D.C.: World Bank.

Argyris, C., and D.A. Schön 1978. OrganizationalLearning: A Theory of Action Perspective. Reading,MA: Addison-Wesley.

Aycrigg, M. 1998. “Participation and the WorldBank: Successes, Constraints and Responses: Draftfor Discussion.” World Bank Development PaperNo. 29, November. Washington, D.C., World Bank.

Blackburn, J., and Jeremy Holland. 1998. WhoChanges? Institutionalizing Partnership and Devel-opment. London: Intermediate Technologies.

Bower, J. L., and Yves Doz. 1979. “Strategy Formula-tion: A Social and Political Process.” In StrategicManagement. Boston, MA: Little-Brown.

Buckley, R. 1999. 1998 Annual Review of Develop-ment Effectiveness. Report No. 19062, OED. Wash-ington, D.C.: World Bank.

Buono, A. F. 1991. “Managing Strategic Alliances:Organizational and Human Resource Consider-ations.” Business and the Contemporary World 3(4):92–101.

———. 1997. “Enhancing Strategic Partnerships: Inter-vening in Network Organizations.” Journal of Orga-nizational Change Management 10(3): 251–66.

Chambers, R., and G. Conway. 1992. Sustainable Ru-ral Livelihoods: Practical Concepts for the 21stCentury. IDS Discussion Paper 296. Brighton, U.K.

Commission on International Development. 1969. Part-ners in Development: Report. New York: Praeger.

DAC (Development Assistance Committee, OECD).1999. “Overview of the DAC Chair: Staying theCourse toward Development Results through Partner-ship.” In James H. Michel, ed., Development Coop-eration: Efforts and Policies of the Members of the De-velopment Assistance Committee. Paris: OECD.

Dasgupta, P. 1997. “Social Capital and Economic Per-formance.” Paper presented at World Bank confer-ence: Social Capital: Integrating the Economist’sand the Sociologist’s Perspective. World Bank,Washington, D.C.


8 2

Datta-Mitra, J. 1997. Fiscal Management in Adjust-ment Lending. OED Report. Washington, D.C.:World Bank.

DFID. 1998. Sustainable Rural Livelihoods: What Con-tribution Can We Make? Paper presented at the DFIDNatural Resources Advisors’ Conference, July 1998.London.

———. 1999. Sustainable Livelihoods GuidanceSheets Issues. DFID Series. Available fromhttp:www.oneworld.org/odi/

Dollar, D. 1999. The Comprehensive DevelopmentFramework and Recent Development Research.Washington, D.C.: World Bank.

Dollar, D., and Jakob Svensson. 1998. What Explainsthe Success or Failure of Structural Adjustment Pro-grams. World Bank Policy Research Working Paper.Washington, D.C.

Doz, Y., and Gary Hamel. 1998. Alliance Advantage.Boston: Harvard Business School.

Ekbom, A., and J. Bojo. 1997. Mainstreaming Envi-ronment in Country Assistance Strategies. Discus-sion Paper 1, World Bank, Environment Group, Af-rica Region. Washington, D.C.

Elster, J., C. Offe, and others. 1998. Institutional Designin Post-Communist Societies: Rebuilding the Ship atSea. Cambridge, U.K.: Cambridge University Press.

Estrella, M., and John Gaventa. 1998. Who CountsReality? Participatory Monitoring and Evaluation:A Literature Review. IDS Working Paper No. 70.Brighton, U.K.: IDS.

Evans, A., and William Battaile. 1998. 1997 AnnualReview of Development Effectiveness. Report No.17705, OED. Washington, D.C.: World Bank.

Farrington, J., D. Carney, C. Ashley, and C. Turton.1999. “Sustainable Livelihoods in Practice: EarlyApplications of Concepts in Rural Areas.” NaturalResources Perspectives No. 42. London: ODI.

Fukuyama, F. 1995. Trust: The Social Virtues and theCreation of Prosperity. New York: Free Press.

Gibbs, Christopher, Claudia Fumo, and ThomasKuby. 1998. Nongovernmental Organizations inWorld Bank–Supported Projects. OED Report.Washington, D.C.: World Bank.

Gwin, C. 1999. “The New Development CooperationParadigm.” In ODC Viewpoint. Washington, D.C.:Overseas Development Council.

Hamel, G., and C. K. Prahalad. 1994. Competing forthe Future. Boston: Harvard Business School Press.

———. 1999. “Strategic Intent.” Harvard BusinessReview 89: 63–76.

Hanna, N. 1985. Strategic Planning and Manage-ment: A Review of Recent Experience. World BankStaff Working Paper No. 751. Washington, D.C.

———. 1991. “The Information Technology Revolu-tion and Economic Development.” World Bank Dis-cussion Paper No. 120. Washington, D.C.

———. 1999. “A Proposed Role for Aid Agencies inHelping Developing Countries to Design National In-formation Technology Strategies.” Technology Man-agement: Strategies and Applications 4: 299–302.

Hanna, N., and Sandor Boyson. 1993. InformationTechnology in World Bank Lending. World BankDiscussion Paper 206. Washington, D.C.

Hanna, N., Sandor Boyson, and ShakuntalaGunaratne. 1996. The East Asian Miracle and In-formation Technology: Strategic Management ofTechnology Learning. World Bank Discussion Paper326. Washington, D.C.

Hirschman, A. O. 1958. The Strategy of EconomicDevelopment. New Haven: Yale University Press.

———. 1970. Exit, Voice, and Loyalty: Resources toDecline in Firms, Organizations, and States. Cam-bridge, MA: Harvard University Press.

Jayarajah, C., W. Branson, and B. Sen. 1995. Struc-tural and Sectoral Adjustment Operations: WorldBank Experience, 1980–92. OED Report. Washing-ton, D.C.: World Bank.

———. 1996. Social Dimensions of Adjustment:World Bank Experience, 1980–93. OED Report.Washington, D.C.: World Bank.

Johnson, J., and Sulaiman Wasty. 1993. BorrowerOwnership of Adjustment Programs and the Politi-cal Economy of Reform. World Bank Discussion Pa-per No. 199. Washington, D.C.

Johnston, R., and P. R. Lawrence. 1988. “Beyond Ver-tical Integration—The Rise of the Value-addingPartnership.” Harvard Business Review 66 (4): 94–101.

Jones, S. 1997. Ethiopia Social Sector DevelopmentPrograms. Report to ODA. Oxford, U.K.: OxfordPolicy Management.

———. 1999. Increasing Aid Effectiveness in Africa?The World Bank and Sector Investment Programs.Oxford, U.K.: Oxford Policy Management.

Kanter, R. M. 1989. When Giants Learn to Dance:Mastering the Challenge of Strategy, Management andCareers in the 1990s. New York: Simon and Schuster.

8 3

B i b l i o g r a p h y

———. 1994. “Collaborative Advantage: The Art of Al-liances.” Harvard Business Review 72 (4): 96–108.

Korton, D. 1980. “Community Organization and Ru-ral Development: A Learning Process Approach.”Public Administration Review 40 (5): 480–511.

———. 1988. “From Bureaucratic to Strategic Orga-nization.” In F.F. Siy and S.Y. Siy, Jr., eds., Trans-forming a Bureaucracy: the Experience of the Phil-ippine National Irrigation Administration. WestHartford, CT: Kumarian.

Kreimer, A., and others. 1998. The World Bank’s Ex-perience with Post-Conflict Reconstruction. ReportNo. 17752, OED. Washington, D.C.: World Bank.

Kuhn, Thomas S. 1970. The Structure of ScientificRevolutions. Vol. II.

Lindblom, C. 1990. Inquiry and Change. New Haven:Yale University Press.

Mackay, K. (ed). 1998a. Evaluation Capacity Devel-opment in Zimbabwe. ECD Working Paper, OED.Washington, D.C.

———. 1998b. Public Sector Performance: the Criti-cal Role of Evaluation. ECD Working Paper, OED.Washington, D.C.

Marcel, M. 1999. “Effectiveness of the State and De-velopment Lessons from the Chilean Experience.” InG. Perry and D. Leipziger, eds., Chile: RecentPolicy Lessons and Emerging Challenges. Washing-ton, D.C.: World Bank.

March, J.G., and H.A. Simon. 1958. Organizations.New York: Wiley.

Margulis, S., and J. Bernstein. 1995. “National Envi-ronmental Strategies: Learning from Experience.”World Bank Environment Department Dissemina-tion Notes: 30. Washington, D.C.

Margulis, S., and T. Vetleseter. 1998. “EnvironmentalMatters at the World Bank: Annual Review — Fall1998.” Report No. 18540. Washington, D.C., WorldBank.

Mathieu, N. 1996. Industrial Restructuring: WorldBank Experience, Future Challenges. Report No.15707, OED. Washington, D.C.: World Bank.

———. 1998. Financial Sector Reform. Report No.17454, OED. Washington, D.C.: World Bank.

Maxwell, S. 1997. “Implementing the World Food Sum-mit Plan of Action: Organizational Issues in Multi-sectoral Planning.” Food Policy 22 (6): 515–31.

Murphy, J. 1997. Mainstreaming Gender in WorldBank Lending: an Update. Report No.17169, OED.Washington, D.C.: World Bank.

Najmabadi, F., and S. Lall. 1995. Developing Indus-trial Technology: Lessons for Policy and Practice.OED Report. Washington, D.C.: World Bank.

Narayan, D. 1999. “Voices of the Poor.” Washington,D.C., World Bank. Photocopy.

OECD. 1995. Governance in Transition: Public Man-agement Reforms in OECD Countries. Paris.

———. 1996. Shaping the 21st Century: The Contri-bution of Development Cooperation. Paris.

———. 1997a. Evaluation of Programs PromotingParticipatory Development and Good Governance:Synthesis Report. Paris.

———. 1997b. In Search of Results: PerformanceManagement Practices. Paris.

———. 1999. Developing Country Scoping Workshopon National Strategies for Sustainable Develop-ment, November 18–19, 1998. Sunningdale.

OECD and DAC. 1996. Shaping the 21st Century: TheContribution to Development Cooperation. Wash-ington, D.C.: OECD Washington Center.

OED. 1997. Lessons of Fiscal Adjustment: SelectedProceedings from a World Bank Seminar. Washing-ton, D.C.: World Bank.

Ohiorheunan, L. 1998. National Long-term Perspec-tive Studies—Lessons from Experience. Prepared forWorkshop in Entebbe, Uganda.

Paul, S. 1998. Making Voice Work: the Report Card onBangalore’s Public Service. World Bank Policy Re-search Working Paper No. 1921. Washington, D.C.

Picciotto, R. 1995. Putting Institutional Economics toWork: From Participation to Governance. WorldBank Discussion Paper 304. Washington, D.C.

———. 1998. Logic of Partnership. OED Report.Washington, D.C.: World Bank.

———. 1999. Poverty Reduction and InstitutionalChange. Inaugural Wolf Lecture. Washington, D.C.:World Bank.

Picciotto, R., and Eduardo Wiesner. 1998. Evaluationand Development: The Institutional Dimension. NewBrunswick, NJ, and London: Transaction for theWorld Bank.

Porter, M. 1980. Competitive Strategy: Techniques forAnalyzing Industries and Competitors. New York:Free Press.

———. 1985. Competitive Advantage: Creating andSustaining Superior Performance. New York: FreePress.

———. 1996. “What is Strategy?” Harvard BusinessReview, November-December: 61–78.


8 4

Purcell, D., and Jock R. Anderson. 1997. AgriculturalExtension and Research: Achievements and Prob-lems in National Systems. Report No. 16851, OED.Washington, D.C.: World Bank.

Rodrik, D. 1999. The New Global Economy and De-veloping Countries: Making Openness Work. Wash-ington, D.C.: Overseas Development Council.

Sabel, C. 1994. “Learning by Monitoring: The Institu-tions of Economic Development.” In Lloyd Rodwinand Donald Schon, eds., Rethinking the Develop-ment Experience: Essays Provoked by the Work ofAlbert O. Hirschman. Washington, D.C.: BrookingsInstitution.

Sachs, J. D., and A. M. Warner. 1995. “Economic Re-form and the Process of Global Integration.”Brookings Papers on Economic Activity 1: 1–118.

Schumpeter, J. A. 1983. The Theory of Economic De-velopment. New Brunswick, NJ: Transaction.

Scott, J. C. 1998. Seeing Like a State: How CertainSchemes to Improve the Human Condition HaveFailed. New Haven: Yale University Press.

Sen, A., and J. D. Wolfensohn. 1999. “Development:A Coin with Two Sides.” May 5. International Her-ald Tribune.

Senge, P. M. 1990. The Fifth Discipline: The Art andPractice of the Learning Organization. New York:Doubleday.

Shepherd, A. 1999. Evaluation of DFID Support toPoverty Reduction: Lessons from the Evaluation ofa Bilateral Program. Working Paper for a WorldBank Conference on Evaluation and Poverty Reduc-tion, June 14-15, 1999, University of Birmingham.

SIDA (Swedish International Development Agency).1999. Making Partnership Work on the Ground.Workshop Report, Ministry of Foreign Affairs.

Stiglitz, J. 1998. Towards a New Paradigm for Devel-opment: Strategies, Policies, and Processes.Prebisch Lecture, UNCTAD, Geneva, October 19.

———. 1999a. Participation and Development: Per-spectives from the Comprehensive DevelopmentParadigm. Speech given February 27, 1999. WorldBank, Washington, D.C.

———. 1999b. “Whither Reform? Ten Years of Tran-sition.” Paper prepared for the Annual Bank Confer-ence on Development Economics, April 29, 1999.World Bank, Washington, D.C.

Subbarao, K., and others. 1997. “Safety Net Pro-grams and Poverty Reduction: Lessons from Cross-country Experience.” Directions in Development.World Bank, Washington, D.C.

Thomas, V., and others. 1999. Quality of Growth.Background Paper, 1999/2000 World DevelopmentReport. World Bank, Washington, D.C. Photocopy.

UNDP. 1996. Developing Capacities in Aid Coordina-tion and Management. Vietnam.

———. 1997. Measuring and Managing Results: Les-sons for Development Cooperation. New York.

United Nations. 1998. World Investment Report 1998:Trends and Determinants. New York.

Wack, P. 1985. “Scenarios: Uncharted Waters Ahead.”Harvard Business Review, September-October 1985:73–89.

Williamson, J. 1997. “The Washington Consensus Re-visited.” In L. Emmerij, ed., Economic and SocialDevelopment into the XXI Century. Washington,D.C.: Inter-American Development Bank.

Wolf, C., Jr. 1997. Markets or Governments: Choos-ing Between Imperfect Alternatives, 2d ed. Cam-bridge, MA: MIT Press.

Wolfensohn, J. D. 1997. “Annual Meetings Address:The Challenge of Inclusion.” World Bank, Washing-ton, D.C.

———. 1998. “The Other Crisis: 1998 Annual Meet-ings Address.” World Bank/IMF Annual Meetings,October 6, 1998. World Bank, Washington, D.C.

———. 1999a. “A Proposal for a Comprehensive Devel-opment Framework.” World Bank, Washington, D.C.

———. 1999b. “Coalitions for Change.” World Bank/IMF Annual Meetings Address, September 28, 1999.World Bank, Washington, D.C.

World Bank. 1993a. Implementing the World Bank’sStrategies to Reduce Poverty: Progress and Chal-lenge. Report No. 11826. Washington, D.C.

———. 1993b. New Lessons from Old Projects: TheWorkings of Rural Development in Northeast Bra-zil. OED Report. Washington, D.C.

———. 1994. A Review of World Bank Experience inIrrigation. Report No. 14908, OED. Washington,D.C.

———. 1995a. Enhancing the Quality of Life in Ur-ban Indonesia: The Legacy of Kampung Improve-ment Program: An Impact Evaluation. Report No.14747, OED. Washington, D.C.

8 5

———.1995b. Lending for Electric Power in Sub-Sa-haran Africa. Report No. 14961, OED. Washing-ton, D.C.

———. 1995c. Poverty Reduction at the World Bank:Progress in Fiscal 1994. Report No. 14021. Wash-ington, D.C.

———. 1996a. Development of Housing, Water Sup-ply and Sanitation in Nairobi. Report No. 15586,OED. Washington, D.C.

———. 1996b. Pakistan: On-Farm Water ManagementProject and Irrigation Systems and RehabilitationProject. Report No. 15863, OED. Washington, D.C.

———. 1996c. Poverty Reduction at the World Bank:Progress and Challenges in the 1990s. Report No.15950. Washington, D.C.

———. 1996d. Reforming Agriculture: the WorldBank Goes to Market. Report No. 1670, OED.Washington, D.C.

———. 1997a. Brazil—Secondary and Feeder RoadsProject and Second Feeder Roads Project. ReportNo. 16738, OED. Washington, D.C.

———. 1997b. India—A Comparative Review ofHealth Sector Reform in Four States: An Opera-tional Perspective. Report No. 15753, OED. Wash-ington, D.C.

———. 1997c. Korea—Delayed Development of theCholla Region: An Institutional Study. Report No.16211, OED. Washington, D.C.

———. 1997d. Poverty Reduction at the World Bank:Progress in Fiscal 1996 and 1997. Report No.17380. Washington, D.C.

———. 1997e. Sri Lanka—Kurunegala and SecondRural Development Projects. Report No. 16418,OED. Washington, D.C.

———. 1997f. World Bank Support for Small-ScaleEnterprises in Ecuador: An Impact Evaluation. Re-port No. 16790, OED. Washington, D.C.

———. 1997g. World Development Report 1997: TheState in a Changing World. Report No. 17300.Washington, D.C.

———. 1998a. The Brazil Health System. Report18142, OED. Washington, D.C.

———. 1998b. East Asia: The Road to Recovery. Re-port No. 18457. Washington, D.C.

———. 1998c. The Impact of World Bank Support tothe HNP Sector in Zimbabwe. Report No. 18141,OED. Washington, D.C.

———. 1998d. Jamaica Country Assistance Note. Re-port No. 19356, OED. Washington, D.C.

B i b l i o g r a p h y

———. 1998e. Lessons from Experience in HNP. Re-port No. 18642, OED. Washington, D.C.

———. 1998f. Malawi—The World Bank and the Ag-ricultural Sector. Report No. 17898, OED. Wash-ington, D.C.

———. 1998g. Mauritania: Second Livestock Project.Report No. 18133, OED. Washington, D.C.

———. 1998h. Paraguay: Community-based RuralWater Systems & Development of Village Commit-tees. Report No. 17923, OED. Washington, D.C.

———. 1998i. Partnership and Development: Pro-posed Actions for the World Bank. World BankPartnership Group, Strategy, and Resources Man-agement Discussion Paper. Washington, D.C.

———. 1998j. Recent Experience with InvoluntaryResettlement: An Overview. Report No. 17538,OED. Washington, D.C.

———. 1998k. The Special Program of Assistance forAfrica (SPA): An Independent Evaluation. OED Re-port. Washington, D.C.

———. 1998l. Sri Lanka—Community Water Supplyand Sanitation Project. Report No. 18113, OED.Washington, D.C.

———. 1998m. The World Bank and the AgriculturalSector in Kenya: An OED Review. Report No.18088, OED. Washington, D.C.

———. 1998n. World Development Report 1998/99:Knowledge for Development. Washington, D.C.

———. 1998o. Assessing Aid: What Works, WhatDoesn’t, and Why. Washington, D.C.: World Bank.

———. 1999a. “Assessing Borrower Ownership UsingReform Readiness Analysis.” PREM Note 25, June.Washington, D.C.

———. 1999b. Case Study of World Bank Activitiesin the Health Sector in India. Report No. 19537,OED. Washington, D.C.

———. 1999c. Civil Service Reform: A Review ofWorld Bank Assistance. Report No. 19599, OED.Washington, D.C.

———. 1999d. Development Effectiveness in Health,Nutrition, and Population: Lessons from WorldBank Experience. Report No. 19266, OED. Wash-ington, D.C.

———. 1999e. Evaluating Health Projects: Lessonsfrom Literature. World Bank Discussion Paper No.356. Washington, D.C.

———. 1999f. “Evaluation of Social Fund Projects.”OED. Washington, D.C. Photocopy.


8 6

———. 1999g. Health Care in India: Learning FromExperience. OED Precis 187. Washington, D.C.

———. 1999h. Health Care in Mali: Building onCommunity Involvement. OED Precis 188. Wash-ington, D.C.

———. 1999i. Higher-Impact Adjustment Lending(HIAL): Initial Evaluation. Report No. 19797,OED. Washington, D.C.

———. 1999j. Lessons from Large AdjustmentLoans. PREM Note 27, August. Washington, D.C.

———. 1999k. Poverty Assessments: A Follow-up Re-view. Report No. 19630, OED. Washington, D.C.

———. 1999l. “Poverty Study.” OED. Washington,D.C. Photocopy.

———. 1999m. Review of Aid Coordination and theRole of the World Bank. Report No. 19840, OED.Washington, D.C.

———. 1999n. Tanzania, Agriculture, and the WorldBank: An OED Review. Report No. 18111, OED.Washington, D.C.

———. 1999o. World Development Report 1999:Poverty and Development 1999/2000. Washington,D.C.

8 7

OPERATIONS EVALUATION DEPARTMENT PUBLICATIONS

Ordering World Bank PublicationsCustomers in the United States and in territories notserved by any of the Bank’s publication distributorsmay send publication orders to:

The World BankP.O. Box 960Herndon, VA 20172-0960Fax: (703) 661-1501Telephone: (703) 661-1580The address for the World Bank publication databaseon the Internet is: http://www.worldbank.org (selectpublications/project info).E-mail: [email protected] number: (202) 522-1500Telephone number: (202) 458-5454

The World Bank InfoShop serves walk-in customersonly. The InfoShop is located at:

701 18th Street, NWWashington, DC 20433, USA

All other customers must place their orders throughtheir local distributors.

Ordering by e-mailIf you have an established account with the WorldBank, you may transmit your order by electronic mailon the Internet to: [email protected]. Please in-clude your account number, billing and shippingaddresses, the title and order number, quantity, andunit price for each item.

The Operations Evaluation Department (OED), anindependent evaluation unit reporting to the WorldBank’s Executive Directors, rates the developmentimpact and performance of all the Bank’s completedlending operations. Results and recommendations arereported to the Executive Directors and fed back intothe design and implementation of new policies andprojects. In addition to the individual operations andcountry assistance programs, OED evaluates theBank’s policies and processes.

Summaries of studies and the full text of the Précisand Lessons & Practices can be read on the Internet athttp://www.worldbank.org/html/oed

How To Order OED PublicationsOperations evaluation studies, World Bank discussionpapers, and all other documents are available from theWorld Bank InfoShop.

Documents listed with a stock number and pricecode may be obtained through the World Bank’s mailorder service or from its InfoShop in downtownWashington, D.C. For information on all other docu-ments, contact the World Bank InfoShop.

For more information about this study or OED’sother evaluation work, please contact ElizabethCampbell-Pagé or the OED Help Desk.

Operations Evaluation DepartmentPartnerships & Knowledge Programs (OEDPK)E-mail: [email protected]: [email protected]: (202) 473-4497Facsimile: (202) 522-3200

1999 ANNUAL REVIEW OF DEVELOPMENT EFFECTIVENESS · William Battaile, Ruchira Corcoran, Deepa...

Documents

Transcript of 1999 ANNUAL REVIEW OF DEVELOPMENT EFFECTIVENESS · William Battaile, Ruchira Corcoran, Deepa...