Aid Coordination and Aid Effectiveness - OECD · number of such coordination activities has grown...

Aid Coordination and Aid Effectiveness

Arne Disch

A Report submitted to the Norwegian Ministry of Foreign Affairsby ECON Centre for Economic Analysis

The Ministry does not accept any responsibility for the information in thisreport nor the views expressed, which are solely those of ECON Centre for Economic Analysis

TABLE OF CONTENTS 3

Table of Contents

ACRONYMS AND ABBREVATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

1 EXECUTIVE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

2.1 MANDATE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102.2 METHODOLOGY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102.3 THE LARGER PICTURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

3 PARTNERSHIP AND AID COORDINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

3.1 AID COORDINATION AND PARTNERSHIP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113.1.1 OECD/DAC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113.1.2 UNDP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123.1.3 Sweden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123.1.4 United Kingdom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133.1.5 The World Bank . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

3.2 NORWAY AND RECIPIENT RESPONSIBILITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133.2.1 Recipient Responsibility vs. Partnership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

3.3 SUMMING UP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

4 AID COORDINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

4.1 INTERNATIONAL COORDINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174.2 COUNTRY-LEVEL COORDINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174.3 CONTENT OF COORDINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184.4 DEGREES OF COORDINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184.5 OTHER COORDINATION DIMENSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184.6 SUMMING UP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

5 COORDINATION EXPERIENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

5.1 GENERAL COORDINATION MECHANISMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205.2 COUNTRY LEVEL COORDINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

5.2.1 Overall Coordination Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215.2.2 Donor Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215.2.3 Best Recipient Practice: Botswana . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

5.3 RECIPIENT PERSPECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235.4 NORDIC COORDINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

5.4.1 Policy Coordination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245.4.2 Operational Coordination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

5.5 AID COORDINATION IN MOZAMBIQUE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265.5.1 Achievements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265.5.2 Conflict and Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

5.6 SUMMING UP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

6 LESSONS FROM COORDINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

6.1 CONTENT OF COORDINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296.2 DEGREES OF COORDINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306.3 COORDINATION AT NATIONAL LEVEL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326.4 SECTOR COORDINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336.5 GEOGRAPHIC COORDINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356.6 SUMMING UP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

4 TABLE OF CONTENTS

7 THE UNFINISHED AGENDA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

7.1 MEASURING AID COORDINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377.2 DESIGNING LEAST COST AID COORDINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377.3 THE CDF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

ANNEX I: MOZAMBIQUE EXPERIENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

1.4 DONOR COORDINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381.1.1 Lessons from Macro Coordination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381.1.2 Lessons from Sector Coordination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401.1.3 Trade-offs: Scope versus Focus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421.1.4 Lessons from Implementation/Procedural Coordination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

1.2 DONOR STRATEGIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461.2.1 Perceptions of Development Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461.2.2 Donor Development Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471.2.3 Donor Strategies and Sector Priorities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471.2.4 Donor Strategy Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

1.3 CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

ACRONYMS AND ABBREVIATIONS 5

Acronyms and Abbreviations

BoP Balance of PaymentCDF Comprehensive Developement FrameworkCG Consultative GroupDAC Development Assistance CommitteeDAF Debt Alleviation FundDFID Department for International DevelopmentECA Economic Commission for AfricaEU European UnionGCA Global Coalition for AfricaGCPI Gabinete de Coordenacão dos Programas de ImportacãoIMF International Monetary FundJEM Joint Evaluation MissionLDC Least Developed CountriesLMG Like Minded GroupMAE Ministry for Public Administration (Mozambique)MinAg Ministry of Agriculture (Mozambique)MinEd Ministry of Education (Mozambique)MoH Ministry of Health (Mozambique)MONAP Mozambique - Nordic Agricultural ProgrammeMPF Ministry of Pfarming and Finance (Mozambique)NMFA Norwegian Ministry of Foreign AffairsNORAD Norwegian Agency for Development CooperationNORSAD Nordic SADCC FundOAU Organization for African UnityOECD Organization for Economic Cooperation & DevelopmentPROAGRI Mozambique’s Agricultural Sector Development ProgrammeROCS Road and Coastal ShippingRT Round Table meetingsSADC Southern African Development CommunitySADDC Southern African Development Coordination CommitteeSAP Structural Adjustment ProgramSIP Sector Investment ProgramSPA Special Program for AfricaSWAP Sector Wide Approach to ProgrammingSWAPO South West Africa People’s OrganizationTRA Tanzania Revenue AuthorityUN United NationsUNDAF United Nations Development Assistance FrameworkUNDP United Nations Development Programme

6 ACRONYMS AND ABBREVIATIONS

1 EXECUTIVE SUMMARY 7

1 Executive Summary

The so-called “Pearson Commission” presented itsreport entitled Partners in Development. Report of theCommission on International Development 30 yearsago. Key themes in that report remain valid today: therecipient must be responsible for its own developmentpolicies but donors have a right to be consulted; theinteraction between the parties should be based on apartnership where predictability and performance-based resource flows are integral parts.

Aid coordination has subsequently become a keydimension in the operationalization of the partnershipconcept. Donors and recipients alike are allocatingincreasing resources for aid coordination activities,with the understanding that efficient aid coordinationgenerates efficiency gains. It is not clear, however,how to measure these efficiency gains, nor is there agood understanding of what a “least cost” strategy foraid coordination improvements should look like.

This study was commissioned by the Planning andEvaluation Unit of the Ministry of Foreign Affairs toprovide an update on the aid coordination debate. Itwas primarily to be a desk study, but supplemented byinterviews of resource persons and the use of relevantcase studies. Particular attention was to be paid torecipient experiences and viewpoints where thesecould be identified.

Partnership, Recipient Responsibility and AidCoordinationThe partnership concept is given different contents bydifferent actors, but overall it is an important step inthe direction of a more equitable relationship betweendonors and recipients. In particular, it recognizes thecentrality of the recipient in the relationship, and thedesirability of the recipient to manage it. But the real-ity is that fundamental asymmetries remain betweenthe two sets of actors, not only in terms of economicresources, but also political. A key aspect here is thelack of credibility and legitimacy on the side of manyrecipient governments that undermines their politicalbargaining power with respect to donors. There is thusa serious gap between rhetoric and reality regardingthe content of partnership concept where the resource-rich partners have a particular responsibility for notdominating and taking over the relationship.

Instead of partnership, Norway has used a recipientresponsibility concept, to underline the need fordonors to step back and have donor coordination sub-ordinated to recipient priorities. Norway risks becom-ing marginalized in the development of the partnership

concept if it insists on maintaining a separate concept,since both partnership and recipient responsibility areto address systemic dimensions of the donor-recipientrelationship. Norway may make a greater contributionby bringing the experiences gained from the recipientresponsibility approach into the partnership discus-sion.

The Structure of Aid CoordinationTwo kinds of aid coordination take place at the interna-tional level. Focused events such as the internationalconferences help generate consensus on overarchingpolicy issues, largely on specific issues like the envi-ronment (Rio), education (Jomtien) etc. On the otherhand, more permanent institutions like OECD/DAC orprocess-oriented fora like the SPA work with systemicand implementation problems. Most of the latter insti-tutions are donor-driven, however, with recipientcountries having limited influence.

Coordination at the recipient country level usuallyincludes overarching development issues and financ-ing of national priorities through Roundtable or Con-sultative Group meetings, held abroad or in-country.There are often a variety of inter-donor coordinationmechanisms in place as well as government-donorfora, both at macroeconomic and sector levels. Thenumber of such coordination activities has grown rap-idly over the last years without clarity as to what theoutcomes of them have been. There thus seems to be alack of focus and results-orientation regarding the aidcoordination activities themselves, which represents amajor challenge both to donors and recipients.

Dimensions of Aid CoordinationAid coordination can be de-composed along severaldimensions: geographic (international coordination ongeneral issues versus recipient country-specific); con-tent (policies, principles and priorities versus proce-dures versus practices); degrees of intensity/commitment (general consultation versus cooperationat the strategic level versus collaboration at implemen-tation level); national issues versus sector and sub-sec-tor issues; at geographic/regional level within a givencountry; and along functional lines (technical assist-ance versus general balance of payments support, forexample).

By de-composing aid coordination efforts along thesevarious dimensions it may be possible to identifywhich cells in the multi-dimensional coordinationmatrix yield the higher pay-off to effort.

8 1 EXECUTIVE SUMMARY

Lessons from CoordinationWhile all actors in principle agree that it is importantthat recipient governments manage external assistanceand thus should be responsible for aid coordination,most of the poorer countries lack the technical andadministrative capacity and skills to handle this effi-ciently. A more serious shortcoming may be their lackof political credibility and legitimacy, however, whichis a dimension that requires more attention. But donorpractices fragment and thus weaken aid coordinationpossibilities and outcomes as well, and this is at leastas much to blame as lack of local implementationcapacity and will.

Nordic ExperiencesThe Nordic countries have a high degree of common-ality with respect to overarching policies and a strongcommitment to flexible and coordinated developmentefforts. The experience has been, however, that effec-tive coordination is difficult to achieve. Strong contin-uous political commitment is required, since thebureaucratic interests of the various aid administra-tions will otherwise tend to fragment efforts alongmore particularistic agendas. Since donor country pol-icy agendas change over time, the outcomes of inter-Nordic coordination efforts have therefore varied con-siderably.

Recipient Politics and Aid Coordination Successful aid coordination implies structural changesto resource flows – in particular that central agencieslike ministries of finance take on a coordinating andthus a stronger role. This may often generate resist-ance from those within the local administration thatfeel they will lose out, such as aid-favored units in lineministries – a problem that should not be underesti-mated. Understanding the politics of aid coordinationis therefore as important as understanding the econom-ics of it.

Donor Strategy Coordination and National Develop-ment ProgramsIf there is no coherent government development pro-gram and strategy in place, donors cannot substitutefor this by coordinating their own strategy processes,but should rather support government’s efforts atdeveloping a credible program. This position arrived atfor example in the context of a Mozambique survey, isnot accepted by all donors, who believe that manyrecipient governments are not serious about addressingkey structural issues and that donor conditionality isstill required. Recent studies note the failure of condi-tionality as a means for attaining policy changes, how-ever, and this points to the need for more imaginativeand constructive support for recipient-managed strate-gies.

Coordination of Policies versus Procedures andPracticesCoordination of policies may not be so important,however, since it is relatively easy to reach agreementon policies and priorities. The real disagreements tendto appear with regards to procedures and practices. Itis at this level, therefore, that coordination gains arepotentially the greatest, since differences in how activ-ities are to be implemented generate enormous admin-istrative and transaction costs to both donors andrecipients. Establishing a program for identifying“best practices” at this level would seem to be impor-tant.

Aid Coordination MechanismsRegarding mechanisms for aid coordination, the mostcommon one - informational meetings – may also bethe least efficient and could be streamlined/downsized.At the other extreme, formal contracts often representcost-efficient coordination because while they may becostly to establish – investment costs are high – theygenerate lasting and potentially substantial reductionsin recurrent transaction cost, and these are the realproblems for local administration.

Most Interesting Current EffortsCoordination at national level is required to achievesystemic changes, which in turn generate major bene-fits to the host country. The most interesting coordina-tion experiences right now seem to be taking place atthe sector levels, however, where innovativeapproaches such as SWAPs are being tried out in anumber of countries. At both levels, however, patienceand realistic capacity building are required. In thisconnection it is important to note that institutionaldevelopment is key to successful aid coordination,since this establishes national “rules of the game” thatare managed by national authorities and which allactors must follow. This is the only means throughwhich sustainable reductions in transaction costs canbe attained and maintained.

The Unfinished AgendaThis assessment was primarily to be a desk-study anddone within strict resource ceilings. It therefore repre-sents a first limited effort at analyzing the aid coordi-nation field, and thus suffers from a number of short-comings:

• The Nordic experiences were not an explicit partof this study, and thus only a few examples havebeen looked into. There are, however, a number ofcases that ought to be reviewed for more carefullessons: the Nordic UN initiatives, several of thelarger SADDC programs including the NORSADfund, as well as more careful reviews of theNamibia and Tanzania experiences.

1 EXECUTIVE SUMMARY 9

• Over the last year, considerable work has beencarried out regarding the experiences of sectorcoordination/sector budget support. This areashould be paid more attention in the future.

• Another area that merits more work is donor prac-tices. Norway’s fragmentation of its cooperationfinancing according to own priorities is one exam-ple, and the degree to which this is common oughtto be looked into. Sources could be national bud-get documents but also the DAC donor reviews.

Conceptual and empirical issues need to be developed.At the conceptual level, it is important that someagreement can be reached on how to measure costsand benefits of aid coordination. Right now, vast andincreasing resources are going into this area withoutthere being a clear idea about whether the coordination

efforts are being done in a rational way or not. A trans-action costs approach to this problem appears to beone of the more promising avenues.

In the empirical field, a program of case studies shouldbe developed that would track some interesting casesand analyze what characterizes what are considered tobe “best practice” examples. Without some solidempirical foundations, it is difficult to see how aidcoordination efforts can be steered in better directions.

Finally, the launching of the Comprehensive Develop-ment Framework, CDF, provides a unique opportunityfor donors and recipients alike to test out and apply ina systematic way the lessons from many years of aidcoordination. The question is if enough is currentlyknown to specify what are the most cost-effectiveways of coordinating aid resources.

10 2 INTRODUCTION

2 Introduction

Both donors and aid recipients are spending considera-ble resources on aid coordination activities. The trendseems to be towards an increase in these levels, yet rel-atively little is known about the outcomes and impactof these efforts. In particular, there does not seem to bemuch of a strategy in place for how to improve theeffectiveness of the aid coordination resources them-selves.

Norway’s Minister for Development and HumanRights has made it clear that she sees this as an areafor improvement. As a step towards understandingwhere the debate on aid coordination is today, thePlanning and Evaluation Unit in the Ministry of For-eign Affairs commissioned the current study. It was tobe primarily a desk study, but was at the same timerather ambitious in terms of identifying lessons thatmight be relevant for the future. Given the limited timeand resources available, on the one hand, and theagenda to be covered, on the other, the strategy chosenhas been to prepare a report that is as much a discus-sion paper as just a review. The reason is that there aresome issues that seem worthwhile to present fordebate, but the material underlying them may be lim-ited and uncertain, so it is not clear how valid or gen-eral the conclusions really are. As a way of promotingthe debate, however, it was found that presenting thetentative conclusions might be the most helpfulapproach.

2.1 MandateThe mandate for the study divided the task in twoparts. The first was to be a review of the studies availa-ble on aid coordination, including the experiences withConsultative Group meetings, of sector coordination,etc. It should in particular look at material that con-tains information on recipients’ experiences. Thereview was to cover both Norwegian and internationalstudies, and could be supplemented with interviews.The other part of the study was to look at the develop-ment of Norwegian policies and practices with respectto main channels of Norwegian assistance.

Based on the above work, the study is to summarizethe experiences with aid coordination, and identify and

assess the factors that have contributed to or weakenedaid coordination. Attention was to be paid to donor-recipient coordination, and particularly which factorshave held back recipient capacity for managing coor-dination. A comparison should be made between Nor-wegian and international experiences in these areas,and an analysis of what may have caused the diver-gences should be presented. Finally, if at all possible,the study should say something about costs and bene-fits from aid coordination.

2.2 MethodologyThe study has been based on four sets of inputs. Thefirst one was the desk study, where the attached bibli-ography contains the material reviewed. The secondone was a limited series of interviews with aid officialsin Oslo and Stockholm. The third component con-sisted of case material from Mozambique and Tanza-nia, taking advantage of two major reviews that ECONcarried out in these two countries over the last twelvemonths, and that generated material relevant to thisstudy. Finally, the Ministry financed the participationof ECON as an observer at a workshop organized bythe World Bank in Washington in February 1999. Thisworkshop was for senior officials in developing coun-tries who have aid coordination responsibilities in theirhome countries, where the purpose was to criticallyreview their experience in this area, and look at waysof making aid coordination more efficient.

2.3 The Larger PictureThe World Bank, the UNDP, and the DevelopmentAssistance Committee, DAC, are also currentlyengaged in reviews of the issues surrounding aid coor-dination. These other studies are considerable largerefforts and will thus cover overarching issues more in-depth. Taking this into account, this review thuspresents some more case material that may not be usedin the other reports. Some of the experiences with aidcoordination among the Nordic countries are pre-sented, as is material from Tanzania and Mozambique– two aid dependent countries with somewhat differentexperiences regarding aid coordination.

3 PARTNERSHIP AND AID COORDINATION 11

3 Partnership and Aid Coordination

The mandate asks that particular attention be paid tothe interaction between donors and recipients withregards to aid coordination – that is, review the struc-tural dimension of this interaction. This is done alongtwo dimensions. Chapter 3 gives an overview of actorsand roles within aid coordination, whereas chapter 4looks at dimensions of aid coordination.

3.1 Aid Coordination and PartnershipCentral to donor-recipient coordination is the conceptof partnership – which is hardly a novel concern. Thefirst major commission established by the internationalcommunity dealt with the problems of developmentand aid. Headed by Lester Pearson, the commission’sreport Partners in Development. Report of the Com-mission on International Development, was presentedin 1969. Much of what it said then is still relevanttoday:

“The formation and execution of development pol-icies must ultimately be the responsibility of therecipient alone, but the donors have a right to beheard and to be informed of major events anddecisions. This calls for a new partnership basedon an informal understanding expressing thereciprocal rights and obligations of donors andrecipient. The precise arrangements will andshould differ from country to country, but, broadly,aid-providers, including international organiza-tions, should be able to expect periodic consulta-tions on matters of economic policy … Recipients,on the other hand, should be entitled to a promptand reasonably steady aid flow at the level agreedand allocation of additional aid according toexplicit criteria emphasizing economic perfor-mance” (Pearson et. al. 1969, pp. 127-128).

The title of the Pearson Commission report was care-fully chosen. And the title and arguments for partner-ing are clearly as topical today as when the report wasproduced. The World Bank presented a discussionpaper in 1998 entitled “Partnership for Development:Proposed Actions for the World Bank”. The focus ofthe paper is “on the single question of how partner-ships can provide a tool to deliver more effectivedevelopment assistance to developing countries”(World Bank 1998d, p. 2).

The Pearson Commission’s call for a partnershipbased on reciprocal rights and obligations betweendonors and recipients has thus been a constant themeover the last 30 years, and aid coordination has been akey dimension in its operationalization. Over the last

several years, there has been greater consensus con-cerning the need for and content of the partnership. Atthe same time, work has intensified regarding aid coor-dination, with both the donor community and develop-ing countries intensifying efforts to improveperformance. A wide range of agencies are involved,and there are more efforts than ever at identifying“best practices” and understanding what makes themsuccessful.

3.1.1 OECD/DAC The Development Assistance Committee, DAC, of theOECD carried out a series of assessments in the late1980s and early 1990s, leading to the DAC Principlesfor Effective Aid – Development Assistance Manual(OECD 1992). The 1992 document states as the first“compelling lesson” from its review of 25 years ofdevelopment cooperation that “aid can only be aseffective as the policy, economic and administrativeenvironment in which it works…The fundamental les-son is that aid has to be more concerned with creatingthe fundamental conditions for its effectiveness”. Fur-thermore, “Greater emphasis should be given to ensur-ing the commitment of recipients’ executing agenciesand the motivation of local target groups through theactive involvement in selection, design and implemen-tation”. (OECD 1992, pp. 5-6). Regarding the specificissue of aid coordination, the document has as the firstoverriding principle that “donor coordination [is] tohelp developing countries establish and implementimproved policies and carefully appraised investmentand expenditure programs” (OECD 1992, p. 25). It haslittle to say in terms of how this is to be achieved, how-ever.

Subsequently, the DAC has produced a policy paper,Shaping the 21st Century: The Contribution of Devel-opment Co-operation (OECD 1996). It establishes aseries of specific objectives for the development coop-eration that are related to poverty reduction. It goes onto say “Acceptance of the partnership model, withgreater clarity in the roles of the partners, is one of themost positive changes we are proposing in the frame-work for development co-operation”. It then notes that“Paternalistic approaches have no place in this frame-work” (OECD 1996, p. 13). The document defineswhat is termed “a compact for effective partnership” inthree parts: joint responsibilities, developing countryresponsibilities, and external partner responsibilities.On the developing country side, responsibilitiesinclude appropriate macroeconomic policies; socialdevelopment with increased participation and genderequality, accountability and the rule of law, strengthen-ing of human and institutional capacity. External part-

12 3 PARTNERSHIP AND AID COORDINATION

ners should provide reliable and appropriateassistance, adhere to guidelines for effective aid, sup-port capacity building to avoid aid dependency, andwork for better coordination among the external part-ners (OECD 1996, pp. 14-15).

3.1.2 UNDP Within the UN system, UNDP has a mandate for coor-dinating programming and to some extent even fund-ing. In addition, UNDP co-chairs the Round Table, RT,meetings with recipient governments. UNDP has car-ried out several studies to review its overall aid coordi-nation role and experiences. The last study, done fiveyears ago (COWIconsult 1994), made a numberpoints:

• UNDP’s mandate, presence and tools can bothservice the donor community and ensure transferof ownership of aid coordination to the nationalinstitutions.

• UNDP can perform three key roles: (i) be a neu-tral provider of aid coordination services, (ii) pro-vide technical services for aid coordination andmanagement; and (iii) provide substantive inputsin key areas, such as sustainable human develop-ment and good governance, that would promotelocal capacity for aid coordination.

• UNDP should focus its efforts in the LDCs, whereit has its main presence, has the largest coopera-tion program, and already plays an importantcoordinating role within the UN system.

• In middle-income countries, UNDP should pro-vide technical assistance in select core and lineministries, avoiding duplication or conflicts withother large bilateral or multilateral programs andagencies that might already be providing similarservices to other national institutions.

UNDP is currently carrying out a review of the RoundTable mechanism, where the full cycle of preparation,conference, sector meetings, monitoring and follow upwill be looked into. Several case studies will be doneand subsequently discussed with development partnersin workshops both at the country and sub-regional lev-els. Other UN mechanisms like the UN DevelopmentAssistance Framework, UNDAF, and Human Devel-opment Reports will be looked at in terms of theirpotential usefulness and relevance for aid coordina-tion1.

3.1.3 SwedenAmong the bilaterals, Sweden seems to have carriedout the most thorough review of its cooperation princi-ples. Focus was on its relations with Africa, since mostSwedish financial resources are channeled there, andthis is where Sweden – and other donors – have expe-rienced the greatest difficulties in attaining statedobjectives. Sida produced a series of studies under the“Project 2015” umbrella, including a major report onaid dependency (Sida 1996). This report contained twostudies – one on the problem of aid dependency ingeneral (Riddell 1996), while the other one was a casestudy of aid dependency in Tanzania, but seen from arecipient-country perspective (Sobhan 1996). Thereport presented an action program for combating aiddependency in a number of points, the most importantof which were: (i) reward the quality/impact of aidrather than levels of aid, using systematic monitoring;(ii) induce a genuine program approach with recipientmanagement and pooled resources; (iii) build nationalrather than project capacity, including policy environ-ment, civil society; (iv) make policy dialogue broadand transparent, including academia, NGOs, repre-sentative entities; (v) support domestic resource mobi-lization.

This was followed by a major report to the Ministry ofForeign Affairs by a panel of external resource personsentitled Partnership with Africa. Proposals for a NewSwedish Policy towards Sub-Saharan Africa (SwedishMinistry of Foreign Affairs 1997). The partnershipconcept is given considerable space: “Here, ‘partner-ship’ denotes the aspiration to establish, jointly withAfrican partners, a more equal and respectful relation-ship” (ibid, page 22). The need for monitoring imple-mentation of the principles is proposed, a code ofconduct for donor countries is presented, (pre-) condi-tions for Swedish support and what are assumed tohave to be shared values for a partnership to be mean-ingful, are laid out.

Based on this report, the Ministry itself presented aformal policy paper, Africa on the Move. RevitalizingSwedish Policy towards Africa for the 21st Century(Swedish Ministry of Foreign Affairs 1998). Partner-ship is again discussed, where it is explained that part-nership “primarily involves not donors and recipients,but two parties collaborating to attain jointly formu-lated aims”. It goes on to state that “Obviously, theinequality that is inherent in the aid relationship can-not be eliminated all at once. But it can be tackled invarious ways while, at the same time, contacts areextended into areas where mutual interest results inself-funding exchange” (ibid, page 9). Genuine part-nership requires a basis of common values and mutualtrust; clearly formulated objectives, conditions, obliga-tions, roles and responsibilities among the parties; a

1. Reviews of the UNDAF have already taken place leading to anumber of revisions to both the structure and organization of theUNDAF compared with the original concept. The UNDPreviwe will presumably bring the various issues surrounding theUNDAF to the table for further discussion.


shared aspiration to bring about increased equalitywhere the weaker party is to be strengthened throughthe partnership. The operationalization of partnershipis seen to consist of African leadership and ownershipof the dialogue; improved donor coordination prefera-bly under recipient government management; well-developed sectoral and budget support; simplified pro-cedures; contractual clarity and transparency;increased coherence between different areas of policy;rewards for progress; and extraordinary debt relief foreligible countries.

Finally, in its budget for 1999 presented to Parliamentin October 1998, the Swedish Government states thataid coordination is of increasing importance, and thatone consequence is that Sweden will perhaps notdevelop any further country strategies. The reason isthat the proliferation of donor strategies is creatingsubstantial problems for recipient countries. Swedenwill therefore support attempts to work out commongovernment-donor country strategies.

3.1.4 United KingdomThe new Labor Government presented its developmentassistance vision in a White Paper entitled EliminatingWorld Poverty. A Challenge for the 21st Century(United Kingdom 1997). The partnership concept isonce again central to the policy presented, where it isstated that “(We shall) work closely with other donorsand development agencies to build partnerships withdeveloping countries to strengthen the commitment tothe elimination of poverty” (ibid, page 6). The partnerconcept developed seems more traditional than Swe-den’s, however, as the White Paper first lays out theobjectives for poverty reduction in various fields, andthen goes on to state that “Countries with which weare prepared in principle to embark on a deeper, long-term partnership, involving all forms of assistance …will be countries where the UK is wanted as a partner,has the influence to play a positive role, and a compar-ative advantage in being able to make a strategic con-tribution to poverty reduction” (ibid., panel 14 page39). A series of commitments expected by the recipi-ent government are then given, focusing on agreedinternational development targets, pro-poor policies,and good governance.

3.1.5 The World Bank The World Bank has recently launched its partnershipinitiative, referred to above. In a subsequent review ofaid coordination experiences it notes that “If the Bankis to make headway with its partnership initiative, aidcoordination must come first and foremost. The poten-tial contribution of aid coordination to developmenteffectiveness has long been recognized within thedevelopment community.” The purpose of the reviewis to “enhance the development effectiveness of the

Bank’s participation in aid coordination processes,support for building country capacity, and interactionswith other donors” (World Bank 1999b, p. i). Aidcoordination efforts are therefore analyzed within aresults-based framework, using a fairly classic set ofevaluative criteria for development effectiveness: rele-vance, efficacy, efficiency, and sustainability, but add-ing the dimension of institutional development.

The on-going World Bank review consists of twophases, where the workshop with officials from devel-oping countries in February 1999 (see section 5.4) sig-naled the end of the first phase. Like the UNDP study,the World Bank now intends to go to the field and dis-cuss with partners and verify through more carefulanalyses what kinds of empirical basis there might befor improving aid coordination mechanisms, and whatthe role of the World Bank should be.

3.2 Norway and Recipient Responsibility Norway has always been a strong supporter of themultilateral system – 50% of Norwegian aid should inprinciple be channeled through multilateral channels –and also of the principle of aid coordination. In aWhite Paper from the Ministry of Foreign Affairs,MFA, in 1971, the goal of allocating one percent ofNorway’s GDP for development assistance was sub-mitted for the first time. It also refers to the work inDAC to reduce tied aid. Norway states that as a matterof policy it will support this with respect to its own aidfinancing, and will also support the coordinationefforts to promote the proposal as such (NorwegianMinistry of Foreign Affairs, NMFA, 1971).

A White Paper from 1986 makes reference to discus-sions within DAC on the role regarding conditionali-ties in connection with Structural AdjustmentPrograms, SAPs. On the one hand, there weredemands for donor adherence to the SAP frameworks,which had been controversial, as the Nordic countrieswere not convinced that all the conditionalities of theSAPs were appropriate. On the other hand, the WhitePaper notes that there was a growing acceptance of theposition that one cannot force a reluctant recipient toaccept policies they do not agree with – a point thatwas important to Norway’s approach to developmentcooperation.

The document then states that Norway must under-stand the macroeconomics of development assistancebetter and develop a clearer and more consistent policyat this level. It should bring these into the debates inthe IMF, World Bank and the Paris Club. The Ministrythus underlines the point that if Norway is to partici-pate in the development dialogue, this must be basedon in-depth knowledge not only of the particular issues


at hand, but also the overarching framework questions.Furthermore, Norway needs to pursue this dialoguesystematically, and the question of coordinating thevarious dimensions of Norwegian development coop-eration and work closer with the other partners in theseareas is brought to the fore (NMFA 1986).

In the 1991 White Paper, the principle of recipientresponsibility was introduced. This will from now onunderlie all Norwegian development cooperation. Lit-tle is explicitly said about aid coordination. Insteadthere is considerable attention paid to Norway workingthrough the multilateral system, and in particular theUN (NMFA 1991).

In 1995, the Ministry of Foreign Affairs commissioneda major study on Norwegian development cooperation.The issue of aid dependency was reviewed where adistinction was drawn between the economic, adminis-trative and psychological dimensions. The latter wasseen as perhaps the most serious aspect as recipients atall levels developed a mentality of letting donors takethe real decisions when it came to the utilization ofexternal financing. This problem was driven in largepart by donors wishing to direct the use of their aidaccording to own priorities, often by entering into dia-logue with implementing agencies like line ministriesand thus circumventing national systems and institu-tions. The latter tended to be weak to begin with, butthe problem was compounded by donor behavior. Thetypical outcome of this was that as aid continued to befragmented, it was rational for local actors such as lineministers to invest time and resources mobilizing aidthrough bilateral discussions rather than support acoordinating exercise run by finance or planning min-istries. The donors would in fact be structuring thefield such that few stakeholders on the recipient sidewould have an interest in overarching national coordi-nation (NMFA 1995a).

In the subsequent White Paper, the Ministry notes thatthere are five main conditions for successful aid, out ofwhich three are linked to issues of recipient responsi-bility: (i) aid must be structured so that it acts as a sup-plement to own efforts. In countries where thepolitical, economic and other framework conditionsare not conducive to such development, aid will havelittle or no effect; (ii) the division of labor of roles andresponsibilities must be clear. The sustainability of agiven activity is dependent on the recipient’s ability toplan and implement it. If this capacity is lacking, theactivity has little chance of surviving the aid phase. Ifthe recipient is not made responsible at the outset, adependency-relationship will easily develop; (iii)skills and institutional capacity are fundamental pre-requisites for developing countries taking responsibil-ity of own development (NMFA 1995b).

Finally, in the budget proposal presented to Parliamenton 5 October 1998, the Ministry notes that the work infavor of social justice and sustainable development is ajoint responsibility of all countries. The Governmenttherefore will intensify its work to improve aid coordi-nation among the donors, ensuring that developmentcooperation is based on recipients’ own premises.Norway wants to contribute to this by supportingefforts at country level for recipients to take a strongerlead in aid coordination locally. It will also work withDAC, the Nordic countries and other like-mindeddonors, and with the multilateral agencies to promotethese views (NMFA 1998b).

3.2.1 Recipient Responsibility vs. PartnershipNorway has consciously chosen to stick with the con-cept of recipient responsibility rather than partnership.When the Ministry and NORAD reviewed perform-ance and options for achieving improved impact in theearly 1990s, the aid dependency issue, and in particu-lar the lack of recipient ownership as a stumblingblock to sustainability, required a major and system-atic change to the donor-recipient relationship.

At the same time, there was considerable concernabout new relationships between donors and recipientsremaining donor driven. While aid dependency ischaracterized by an absence of recipient governmentpro-active leadership, this had not come about simplybecause recipients did not wish to assume this task(though this clearly also played a role), but alsobecause many donors aggressively were pursuing theirown agendas. The space left for recipient authorities totake decisions was often tightly circumscribed. Repro-ducing a new form of patron-client relationship in theguise of partnership therefore had to be avoided. Sonot only had recipients to step forward – donors had tostep back.

The way of achieving this was seen to be to emphasizethe recipient role and maintain the focus on this.Whether recipients were able to successfully assumethis role or not was seen to be a function of several fac-tors: political will on the side of the recipient to takeon this role; the willingness of donors to step back;and the organizational capacity in the public sector toactually implement. The latter problem was seen as ajoint responsibility, where Norway was committed toassist with necessary resources when asked.

In a discussion regarding what the content of the recip-ient responsibility concept should be, an internalNORAD document from August 1998 spells out somekey issues (NORAD 1998). It notes that the partner-ship concept is not concretized in terms of the relation-ship. In particular a number of donors use it withoutdiscussing the problem of power and responsibilities


between the parties – there is a strong assumption ofharmony of interests behind the usage (for examplethat both parties understand and wish to fight povertyalong similar lines). The document spells out howNorway’s aid administration, particularly in the field,is to act at the macroeconomic, sector and activity lev-els so that it would be in line with recipient responsi-bility. And it underlines the need for Norway toactively engage in dialogue with recipient govern-ments and the donor community to ensure that theseprinciples are applied more broadly in every country.

Some IssuesHow much difference is there actually between thepartnership and the recipient responsibility concepts?Some of the practical distinctions may at first seemtrivial. There is, however, a more explicit – somewould claim realistic – analysis of role conflicts andpower relations behind the recipient responsibilityconcept that leads to a better strategy for actuallyattaining genuine partnership. First it notes the poten-tial conflict of interests between the parties that entailsthat the donor must genuinely step back and listencarefully to what recipient interests are. This generatesa strong argument for the recipient retaining primaryresponsibility for the cooperation, while partnershipmay legitimize too much donor interference. Secondly,it puts the recipient-donor relation at the center – theinternal donor coordination must be subordinated tothis.

There is a danger that recipient responsibility entailstoo passive a role on the side of the donor, however –that when the donor steps back, it vacates the scene.The recent evaluation of Norwegian developmentcooperation with Tanzania claims that Norway has notbeen active enough as a dialogue partner in key areassuch as gender, and that this has hurt impact (ECON1999). At the same time, Norway has always beenactive in aid coordination fora and strongly supportedcoordination initiatives, including the close donorcoordination that takes place many places through theLike-minded Group, LMG. The difference betweenSweden’s partnership concept and Norway’s recipientresponsibility is in practice often negligible.

The Norwegian wariness of partnerships being “oldwine in new bottles”, despite the DAC’s strong state-ment against paternalism in this area, is not withoutreason, however. Maxwell and Riddell, in an issue ofthe Journal of International Development entirelydevoted to analysis and comments on the UK Govern-ment’s White Paper, warn that the change in rhetoricdoes not necessarily herald in a different era in donor-recipient relations (Maxwell and Riddell 1998).Despite the DAC principles being embraced by allback in 1996, donor behavior is still largely defined by

own agendas and policies. The list of challenges tomore effective aid coordination that senior staff fromnearly 20 developing countries presented at a work-shop in Washington in February 1999, were largely thesame that a similar group could have been expected totable ten years earlier (see 5.4).

3.3 Summing UpThere are several asymmetries in the donor-recipientrelationship that make the partnership concept compli-cated. The first one is obvious: one party is providingresources, the other is receiving – and this is the funda-mental reason for the partnership in the first place, asthe concept is used by a number of actors.

The Swedish notion of partnership goes beyond devel-opment cooperation, however, as this is only part ofthe total interaction. It emphasizes the two-way flowof benefits form broader interaction patterns, such astrade, financial flows, and cultural cooperation, notingthat there is no true partnership till a situation of moreequity in perceived benefits has been established.Focusing on aid structurally forces the debate toremain skewed in favor of donor perceptions, and thisin itself puts recipient countries at a disadvantage.

Another asymmetry exists in the field of politics. Theissue is not so much the skewed political power –though this is important – but the fact that the donorpolitical system generally has a lot more credibilityand legitimacy. Both their policies but also their sys-tems have clear political foundations due to represent-ative parliamentary debates and decisions. Donorshence have a totally different basis from which todemand negotiation outcomes that are acceptable tothemselves: “our parliament cannot accept this”. Lackof “good governance” thus undermines bargainingpower.

Strengthening the political system is thus importantfor a number of reasons. But this is an area wheredonors both have problems finding their proper roleand implementing well. Norway’s development coop-eration with Tanzania is a case in point where the strat-egy developed was quite innovative and relevant,emphasizing the need for support to political transfor-mations. Implementation, however, focused on formalmulti-party politics, whereas the real forces foraccountability and representative democracy are com-ing from the free media, representative organizationslike labor unions, employers’ associations and profes-sional organizations, and advocacy groups like NGOs(ECON 1999).


Conclusions• The partnership concept is given different con-

tents, but overall is an important step in the direc-tion of a more equitable relationship betweendonors and recipients. In particular, it recognizesthe centrality of the recipient in the relationship,and the desirability of the recipient to manage it.

• The reality of partnership today is poor, and fun-damental asymmetries between the two sets ofparties maintain the gap. This reinforces remain-ing vestiges of paternalism on the side of thedonors, including in attitudes. A more embracingconcept in line with Sweden’s usage would helpbridge the gap and in particular is more supportiveof providing the recipient voice.

• The gap between the two parties is not only interms of economic resources, but also political.Donor assistance in the area of political develop-ment has been poor, with an over-emphasis onformal multi-party politics. Less support has beenprovided for the development of enabling frame-works for competitive politics and to representa-tive and advocacy groups that are struggling for

increased accountability and representativedemocracy.

• While Norway’s “recipient responsibility” con-tains a more realistic political analysis and a moreexplicit set of guidelines for implementing it thanmost partnership concepts, Norway may still con-sider whether it is worth maintaining it as a sepa-rate concept when dealing with its developmentpartners. Partnership and recipient responsibilityboth address systemic dimensions of donor-recipi-ent relationship. Having one donor maintain itsown concept means necessarily fragmentation:there can be no genuine “recipient responsibility”if it only defines the bilateral relationship betweenNorway and recipient when all other actors areadhering to a slightly different framework. Nor-way may instead focus on highlighting both thepolitical foundations and the practical experiencesfrom the recipient responsibility approach whenparticipating in the partnership discussions. Thiswould be a way of enriching it and perhaps a moreconstructive contribution to the development ofgenuine partnership in the context of recent initia-tives like the Comprehensive DevelopmentFramework.

4 AID COORDINATION 17

4 Aid Coordination

The partnership concept provides the frameworkwithin which aid coordination is to take place. Butwhile the structure and principles of partnership areimportant, it is largely the practical experiences of aidcoordination that are giving content to the partnershipconcept. In order to understand the lessons of aid coor-dination better, it is helpful to de-compose these activ-ities along some key dimensions.

4.1 International CoordinationTwo different mechanisms are used to coordinate aidat the global level. The first is the large internationalmeetings addressing particular issues, and where theobjective is to establish general consensus on keyquestions. The first half of the 1990s saw a largenumber of such gatherings, such as for education(Jomtien 1990), the environment (Rio 1992), popula-tion (Cairo 1994), gender (Beijing 1995), and socialdevelopment (Copenhagen 1995). These conferenceswere important for reaching agreements on certainprinciples and standards against which all govern-ments in principle can be measured and held accounta-ble for. These conferences are organized by UNagencies and provide recipient countries considerablevoice, particularly at the formal representational/vot-ing level.

The second are the more permanent institutionalarrangements that work on systemic and implemen-tation issues. The OECD/DAC, the Special Programfor Africa (SPA), the Paris Club, and the coordinat-ing multilateral agencies UNDP and World Bank,all are engaged in improving aid coordination at theglobal level. Regional bodies like the Global Coali-tion for Africa (GCA), the Southern African Devel-opment Community (SADC), the EconomicCommission for Africa (ECA) and the Organizationfor African Unity (OAU) are used to harmonize aidcoordination approaches – some of them in donor-recipient fora (GCA, SADC), others as more inter-nal recipient deliberative bodies (ECA, OAU). Thedivision of labor between the various internationalcoordination mechanisms is largely clear, thoughthere has been some overlap between SPA and DACdiscussions.

The most important fora like the DAC and SPA areclearly donor dominated. The ones that the developingcountries themselves manage – institutions like ECA,OAU, fora like GCA and SADC – have little influence.This is in part due to recipient countries themselvesnot using them actively and empowering them to actmore aggressively – which in turn is largely a function

of the realistic assessment that these agencies haveneither the resources nor the credibility to make themgenuine playmakers. The conclusion is that at the levelof international coordination, there is a long way togenuine partnership.

4.2 Country-Level CoordinationWhile international fora are key to improving over-arching principles for aid coordination, the country-level mechanisms are the ones that determine towhat extent impact is actually achieved. In manycountries there is a bewildering array of activitiesthat in itself may constitute a problem for nationalmanagement:

• Consultative Group, CG, and Round Table, RT,meetings are the “apex” fora for aid coordinationwith regards to a particular recipient country.These meetings generally cover both policy issuesand resources mobilization.

• Many governments have established local aidcoordination fora, usually chaired by ministries offinance or planning, or the central bank. The con-tent and structure of such meetings vary, fromquarterly meetings with background documenta-tion distributed beforehand, to ad hoc sessionscalled to address a particular topic, such as pre-paring for a CG meeting or reporting back after aCG meeting has taken place.

• The most intensive coordination work often takesplace at the sector level, where donors and lineministries or government working groups workjointly on issues like sector budget support (basketfunding and other mechanisms), sector policies,implementation problems, etc.

• Internal donor meetings are of a wide variety. In anumber of countries, the UNDP and/or WorldBank chair meetings with the donor community toshare information, raise policy questions, and har-monize positions or procedures in particularfields. The UN system has established its owninternal coordinating mechanism, where a formalUN Development Assistance Framework,UNDAF, is used to streamline and harmonize UNagency activities and priorities. The EU has a for-mal mandate to coordinate the activities of itsmember-states. The so-called “Like-mindedGroup” (LMG) is an informal grouping of donorsthat in some countries comes together to harmo-

18 4 AID COORDINATION

nize positions with a focus on improving on-the-ground coordination2.

Total time spent preparing for and participating inthese various activities takes a considerable share ofavailable time both for local policy makers and donoroffices and embassies. Yet in no country this study isaware of is there a good overview of all of these coor-dination activities in terms of participation, frequency,agendas, much less regarding outcomes/results andcosts.

4.3 Content of CoordinationWhile most aid coordination takes place at the nationallevel, the content of the coordination efforts variesconsiderably:

• Policies, principles and priorities: At this level,the challenges usually consist of establishing con-sistency between recipient and donor overarchingpolicies and priorities. The international confer-ences referred to earlier (Jomtien on educationetc.) deal almost exclusively with this set ofissues, while different in-country fora are used toclarify what the priority concerns of the differentactors are, and how they can be harmonized tomake up a consistent set of activities. The discus-sions thus focus on “doing the right things” ” andgetting all the actors to agree on these.

• Procedures: All actors have established a set ofprocedures that must be followed, particularly inthe field of financial disbursement, accountingand auditing, but also covering areas like procure-ment, personnel policies, tax and customs treat-ment, reporting, and policy dialogue. These aredefined by each actor’s own deliberative body athome, which means that once in the field, the con-fusion of non-harmonized procedures createsenormous transaction costs of various kinds, mostof which fall on the recipient authorities. Sectorworking groups often spend a lot of their time onreaching agreements on these “doing the thingsright”-issues.

• Practices: While procedures are formal “rules ofthe game”, practices are the more informallybased ways of doing business, very often part ofthe “corporate culture”. While formally not so dif-ficult to address as procedural differences, theyoften have practical implications for how tasks are

carried out. Typical areas concern informationsharing, the practical relationship between fieldstaff of externally funded agencies and localauthorities, how external agencies interact withlocal communities, what activities an agency willand will not address, etc.

4.4 Degrees of CoordinationCoordination can also be of different degrees of inten-sity/commitment:

• Consultation: This generally focuses on informa-tion sharing, either between host governmentsand donors, or within the donor community. Usu-ally there are no formal commitments or decisionstaken at this level, though there is often the inten-tion or desire that the consultation will lead atleast to informal understandings of improvedpractices along some defined dimension.

• Cooperation: This is of a more strategic naturewhere policies, priorities, principles are discussedwith the intention of arriving at some form of har-monization. This requires a degree of consensusas well as trust that may not cover all donors,though coordination across the entire donor com-munity is being seen more and more, particularlyin emergency situations.

• Collaboration: This addresses issues of proce-dures and practices, where there is a consciouseffort to ensure that implementation of activitiesruns as smoothly and seamlessly as possible,independent of funding source. The typical casesare the early joint import support programs andnow sector budget support in the form of basketfunding where donors accept one set of disburse-ment, reporting, accounting and auditing proce-dures and where there is complete fungibilitybetween each individual donor’s contribution aswell as public funds in that sector.

4.5 Other Coordination Dimensions

Geographic/Regional CoordinationIn a number of countries, donors are either being askedby local governments or on their own focusing theiractivities in particular geographic areas, but based on adivision of labor amongst themselves. As will be notedbelow, geographic or country area coordinationpresents some serious questions regarding what aregood ways of donors coordinating their resourceswithin a recipient country.

2. The LMG composition varies somewhat from one developingcountry to another, but usually consists of the Nordic countries,the Netherlands, Canada and sometimes Switzerland, the UKand Ireland.

4 AID COORDINATION 19

Functional Coordination There is some functional coordination taking place,where different actors focus on activities based ontheir formal mandates or comparative advantage. Typi-cal cases are the UN agencies, which have a particularmandate for technical cooperation in certain fields, orthe multilateral lending institutions that are particu-larly concerned with macro-economic or financial/debt issues. Among bilaterals, there are also cases ofdivision of labor, for example among the Nordic coun-tries where at times there have been agreements thatonly one or two of them would be active in a sector ina given country (see 5.1.2).

4.6 Summing Up• At the international level the UN-sponsored glo-

bal conferences have been important for arrivingat agreements on international standards in a num-

ber of sectors. Recipient country participation andinfluence is substantial. More permanent institu-tions that deal with systemic and implementationissues – SPA, DAC etc. – tend to be much moredonor dominated, but are opening up to broaderparticipation and some more partnership – thoughthere is a considerable way to go in this area.

• In-country coordinating mechanisms are multi-dimensional, covering policies, procedures andpractices based on consultations, cooperation andcollaboration along national, sectoral, geographicand functional dimensions. The complexity andsheer number of such coordinating activitiesposes a serious strain on both donor and recipientresources on the ground, making it difficult toselect and make rational choices among alterna-tive activities to participate in and focus ownefforts on.

20 5 COORDINATION EXPERIENCES

5 Coordination Experiences

While the structure of aid coordination has becomequite complex, the lessons from participating in vari-ous kinds of activities are even more fragmented. Notonly is the range of activities expanding, but the wayin which interaction takes place is changing andbecoming more diverse as the actors gain experienceand try to adjust the specific aid arrangements to theparticular circumstances in a given host country.

Most experiences from aid coordination are not docu-mented. Of those that are, many are today less relevantbecause the experiences described have been super-seded by new approaches. In order to capture the fla-vor of recent practices, this study has therefore had torely on perhaps less well documented but more rele-vant verbal information and less formal and moredescriptive material such as working papers and con-sultancy studies

The first part of this chapter looks at the reviews ofsome of the mechanisms that have been put in place bythe international community for overall aid coordina-tion. The second part looks at country-level coordina-tion, including both donor behavior and what isconsidered “best practice” by a recipient country,while the third section discusses the limited informa-tion regarding recipient views on aid coordination.

Section four reviews some of the attempts by the Nor-dic countries to coordinate their aid over the last 20years, for some lessons on multi-donor coordination.Mozambique, one of the world’s most aid dependentcountries, is trying to improve the efficiency of exter-nal resources through better coordination, and this isdiscussed in the last section.

5.1 General Coordination MechanismsThe entire multilateral system – the UN agencies, theBretton Woods institutions and regional developmentbanks – contain important aid coordination dimen-sions. All the major organizations have been subject toa number of reviews and evaluations, where the NordicUN projects have in fact analyzed the total field activi-ties of the UN system with a view to reforming andimproving what is seen as inefficient management ofinternational assistance due to duplication, excessivebureaucracy, unclear divisions of labor between someof them, etc. The relationship between the World Bankand regional development banks has also been the sub-ject of a recent discussion by the Nordic countries,again with a view to identifying the various institu-tions’ comparative advantage within the larger exter-nal resources mobilization framework.

One noticeable and somewhat contradictory feature ofthese various agencies is that while they often are seento behave differently, their membership/ownership islargely the same: the member states of the interna-tional community. A major reason for this is that theinfluence of donors versus recipients varies across thedifferent agencies since the voting structure is differ-ent. But another important issue is the lack of coordi-nation among different agencies within a donorcountry. In a given organization a donor can be repre-sented by the ministry of foreign affairs, or ministry offinance, the aid agency, the central bank, or a line min-istry (particularly an issue with the various UN agen-cies). Foreign debt and trade-offs between macro-economic stabilization and poverty reduction havebeen classic areas of disagreement within donor coun-tries and led to individual donors providing inconsis-tent signals on similar issues in different fora. Whilethis remains a problem, it seems less of one towardsthe end of the 1990s than earlier, in part because thereis greater international consensus on many key issues,both among the donors but more importantly also withrecipient governments. At the same time, donors seemto have addressed the internal coordination issue bet-ter. One of changes over the last several years in anumber of countries is to integrate aid and trade poli-cies more tightly into ministries of foreign affairs. Thishas usually also led to closer organizational andadministrative integration of aid agencies into the for-eign policy administration.

DAC The OECD/DAC is first and foremost a secretariat tothe donor community, and as such has no basis fromwhich to impose or enforce changes to donor prac-tices. Instead, it relies on persuasion and negotiation toencourage donors to modify attitudes and actions.Considerable consensus-building work goes on in theDAC High-Level Meetings and the technical commit-tees that prepare these meetings, where decision-mak-ers from the donor countries meet to review analysesof current policies and practices and decide on possi-ble changes. The interest in identifying and adopting“best practices” in a number of fields as general guide-lines for the donor community has been helpful. Cou-pled with this are the annual peer reviews of donorpractices, where representatives from other membersof the donor community assess a given donor’s pro-grams. In addition to the specific findings that are pro-duced, the process itself generates a lot of lessons onboth sides of the table, providing additional instru-ments and incentives for harmonization of donor prac-tices in the direction of the “best practices” identified.

5 COORDINATION EXPERIENCES 21

Despite being an agency that can only arrive at conclu-sions through consensus, the DAC has proposed aseries of standards that are being accepted by thedonor community, such as in the Shaping the 21st Cen-tury document.

SPAThe Special Program of Assistance, SPA, developed inresponse to particular problems being faced by debt-distressed African countries in the early 1980s. SPAwas to assist in mobilizing quick-disbursing funds,usually to support Bretton Woods-led SAPs. Overtime, the need for harmonization of policies and proce-dures in areas related to these quick-disbursing fundsled the SPA to develop working groups to addressissues like standardization of procurement policies,untying of aid, etc. A recent evaluation of the SPA(World Bank 1998c) found that the SPA had been suc-cessful in mobilizing and disbursing the additionalresources required, it helped channel these resourcesto the better-performing (in adjustment terms) coun-tries, and managed to untie an ever larger share of suchbalance of payments support. SPA was successful instreamlining procurement practices and ending ear-marking of counterpart funds, but did not achievemuch in the area of local salary supplements – a prac-tice that is extremely detrimental to the developmentof local markets for skilled labor. The policy debatesthat developed in the context of the SPA were helpfulin improving mutual understandings of a series ofissues related to the adjustment programs. One surpris-ing finding was that much of the work being done inthe context of the SPA was not widely disseminated –the impact of the learning was much less than couldhave been expected. Another area of concern has beenthat the SPA has very much been a “donors’ club” andwhere the agenda therefore has been totally donordriven. There has so far been little partnership withrecipient governments.

5.2 Country Level Coordination

5.2.1 Overall Coordination Mechanisms

CG Meetings The World Bank has reviewed the Consultative GroupMeetings as an instrument on several occasions. Themain findings are summarized in the recent OED doc-ument on aid coordination. One comment is that “theBank’s CG role is widely appreciated by other aidcoordination partners, but some fault the Bank’s lackof transparency and timeliness in sharing relevantinformation and lack of openness to their participationin processes relevant to more effective coordination”(World Bank 1999b, page 21). Another typical obser-vation from bilaterals is that the CG meetings are suc-

cessful in terms of resource mobilization, but are tooformal fora for genuine policy exchanges.

More and more CG meetings are now being held inhost country capitals rather than in Paris, which is seenas positive. Among other things this permits a muchwider local representation at the meetings, both mak-ing the CG process itself more transparent to localsociety, but also enabling resource poor but importantcivil society actors to participate (NGOs, labor unions,religious organizations, etc.). The more substantiveinvolvement of host country governments in the prepa-ration and hosting of the meetings is also an importantstep in the right direction. The highly visible WorldBank hand in preparation of key policy documents hasraised questions as to the local ownership of contentsand interest in implementing the policy changes. Hav-ing local finance ministries manage key policy proc-esses like PFP preparations for the CG meetingstherefore is reducing this concern considerably.

RT and Paris ClubRT Meetings and Paris Club have not been looked at tothe same degree as the CG meetings. The RT meetingsare generally considered more participatory in theirmanagement than the CGs, due to UNDP’s quite dif-ferent style in its relations to recipient governments.The UNDP is currently engaged in a review of theirexperiences with the RTs. The Paris Club, on the otherhand, is considered extremely closed, and totallydonor controlled.

5.2.2 Donor PracticesEarly reviews of the experiences with aid coordinationoften pointed to the donors as constituting a problem.Several recipient governments had tried to establishprocedures and structures that were in line with thedemands from donors, yet when donors were asked touse the new approaches, they generally did not do so(Calhoun and Whittington 1988; Clift 1988; Adams1989).

The problem of donor-driven processes and donor-determined outcomes is still a major issue. In a reviewof government-donor relations in Tanzania, a group ofindependent advisers established by the Governmentof Tanzania and the donor community jointly, pointedto donor attitudes and actions that clearly underminedTanzania’s possibilities for taking ownership. Thegroup had as their fourth recommendation that “sub-stantial changes are needed in the operational cultureof bilateral donors …” A number of other points in thelist of 19 recommendations point to the need forchanges in donor practices (Helleiner et. al. 1995). Thegroup noted that technical assistance is an area wheredonors consistently “over-supply” compared withwhat recipient governments want. This is in line with


the conclusions in UNDP’s review of the experienceswith technical cooperation (Berg 1993). A series ofcase studies on aid effectiveness, organized by theOverseas Development Center, brought out similarstories as two of the four main factors undermining aideffectiveness were seen to be (i) poor coordination ofaid, and (ii) the proliferation of stand-alone projects –i.e., donors funding activities outside the governmentprograms (van de Walle and Johnston 1996). Therelated case study of Danish aid in Tanzania gives adetailed description of how poor donor coordinationfragments and thus undermines Tanzania's possibilitiesfor managing aid effectively (Bagachwa et. al. 1998).The evaluation of Norway’s assistance to Tanzaniafound that the major problem was the extreme frag-mentation of the program caused by Norwegian deci-sions. This undermined Tanzania’s possibilities formanaging it (ECON 1999).

It is not only in relations to host governments thatdonor practices fall well short of ideals. As was notedmany years ago, “donor agencies are not all that goodat avoiding repetition of their own mistakes; and theyare worse at avoiding each other’s, since there is verylittle communication about failure from one agency toanother… There is a strong case for major efforts toimprove the learning process, and for strengtheningwithin agencies the incentives to improve the qualityof lending, not just its quantity” (Cassen 1986, repro-duced in Cassen 1994, p. 12). A recent survey of howdonors go about developing their country strategies inMozambique found that these shortcomings still verymuch exist (ECON 1998).

One controversial aspect of donor practices is the useof ex ante requirements for support – the much-debated “conditionalities”. Recent studies have con-cluded that such pre-conditions as a way of achievingsustainable policy changes simply do not work (Kil-lick 1998 for general analysis, Selbervik 1999 for acase study on Tanzanian-Norwegian relations). Themain reason given is that donors are not consistent interms of applying sanctions when cases of non-com-pliance of conditionalities occur. The expected costs torecipients of not implementing certain policies they donot like, is thus seen to be low. In some sense this canbe seen as a positive aspect for recipient countries –they are in effect able to exert a “negative veto” onnon-preferred policy prescriptions through non-imple-mentation without undue costs. On the other hand, onecan turn this argument around and note that with veryfew exceptions, recipients do not have the political(and economic!) strength to systematically reject aid-funded activities that are not seen as important –donors are almost always able to “buy” an outcomethat is agreeable to their own objectives. The fullimplications of this do not seem to be appreciated by

either donors or recipients in terms of the changes thatthe integrated sector work (like in sector SWAPs) andthe Comprehensive Development Framework maylead to. This will be discussed more extensively in thenext chapter.

5.2.3 Best Recipient Practice: BotswanaGeneral principles for efficient aid administration havebeen well known for a long time, and Botswana is usu-ally referred to as a country that has consistentlyapplied them for over 20 years. It has a political sys-tem that has both legitimacy and credibility and hasbeen able to set national priorities and stick with them.It has established a budget and planning system thatoperationalizes the priorities in transparent fashionwithin realistic resource ceilings3. It has created astrong Ministry of Finance and Development Planningthat has the technical skills to develop the budgets andmonitor implementation, and has the political backingto take action if line ministries overstep boundaries.There is an accounting and auditing system that gener-ates credible reports in a format that allows donors totrack resource usage (see Nordås et. al. 1998 for agood description of Botswana’s situation and prac-tices). Most important of all, Botswana has had thepolitical will to say “No” to donor proposals that werenot part of its own national plans.

Why Others Have Not Been Able to Emulate Bot-swanaOther countries have attempted to establish similarsystems, but as noted above donors did not respect andfollow them. Tanzania, for example, had such an expe-rience in the early 1990s. Donors insisted that theGovernment develop and present an aid coordinationstrategy, since a number of evaluations and WorldBank-led Public Expenditure Reviews had revealedmajor problems in the utilization of external financing.The Ministry of Finance presented a plan in 1992 thatindicated which donors it wanted to have participate inwhat sectors. This plan was largely ignored and neverbecame operational.

The reasons for this negative outcome were several.The document itself was poor in a number of fields. Itdid not have a realistic macro frame that set overallresource limits. There was no strategy behind thechoice and levels of assistance requested, but simplyan extrapolation of existing activities that was based

3. Norway in fact provided assistance as of the mid-1970s for theestablishment of a macro-economic model that was used to esti-mate medium-term growth, and from this derive projections forpublic expenditure ceilings, investment versus recurrent costbudgets, etc. This work was picked up 15 years later whenNamibia became independent and sought out “best practice”examples in the region for its own planning and budgetingstructures. The MACMOD model being developed in Tanzaniatoday also is being developed with support from the samemilieu at the Christian Michelsen Institute in Bergen.


both on poor data but also unrealistic with respect toaid availability and with an implied prioritization thatdonors had difficulties accepting. There were no realpolicy changes being proposed that would have sug-gested more realistic framework conditions for theactivities, and thus no reason to believe that the pro-ductivity and sustainability of investments wouldimprove. The suggested sector concentrations for thevarious donors had not been discussed with the donorcommunity and often did not correspond to where theyreally wanted to focus their resources. More impor-tantly, however, there had not been a good politicalprocess inside the government, so neither line minis-tries nor central offices like the Planning Commissionwere on board. The relationship to the overall politicalsystem was also weak, meaning that the legitimacyand credibility – in addition to the realism – of thewhole exercise was questioned.

It was also obvious, however, that a number of donorssimply were not interested in being coordinated in thisway. As is clear from for example the Mozambiquesurvey (ECON 1998), some donors have guidelinesfrom their capitals that make it difficult for them toparticipate fully in such coordination processes.Unless there is an unusually strong and clear leader-ship on the recipient side that can legitimately call thedonors to order – as in Botswana – the political strug-gle over resource allocations will be determined by theowners of the resources. Those donors who wish topursue their own objectives may then often simplyelect not to be present – at least as donors – in coun-tries where such strong government leadership is inplace.

5.3 Recipient PerspectivesThere is relatively little written material on aid coordi-nation produced by recipient authorities themselves.Most of it is presumably to be found in formal state-ments presented at CG meetings by government offi-cials, where resource mobilization considerationsskews the message towards the positive, and wherefocus is on good intentions as an incentive to mobilizeadditional funds, rather than on areas of disagreementand conflict. The most important source of informa-tion is therefore probably more informal statements bythe politicians and the viewpoints of senior officialsinvolved in aid coordination.

The World Bank organized a workshop in Washingtonin February 1999 where senior officials from about 20developing countries were invited to discuss aid coor-dination. When reviewing their experiences, nineissues were seen as constituting the main obstacles tomore successful coordination: (i) there were conflict-ing goals and stakes between the donors, with compe-

tition between them at several levels of thedevelopment hierarchy; (ii) donors have different fun-damental paradigms regarding development whichcreates conflicts with respect to where and howresources should be used; (iii) donors have differentprocedures, generating procedural overload for thelocal administrations; (iv) in addition to being differ-ent, many of the procedures are also seen as unneces-sarily complicated; (v) recipient governments are oftendependent on donors for the basic data needed forgood aid management such as debt levels, aid flows,etc. In addition, donors often do not provide full infor-mation on their assistance or in categories that are notcompatible with local usage, or not in synchronizationwith local programming cycles; (vi) many donors haveconditionalities with the assistance – but these areoften not harmonized, and may in fact conflict witheach other; (vii) many of the activities that the donorswish to support are not really sustainable at the levelsand with the technical solutions that donors introduce;(viii) too little attention is paid to overall nationalcapacity and too much to having particular capacity inplace that will support donor-specific activities; (ix)many of the donor staff negotiating with the localadministration are not at the technical forefront of theissues that are being discussed, yet recipients must lis-ten to them since they make decisions regardingresource allocations. The rapid turnover of donor staffcompounds this problem, since it takes time to under-stand the specific constraints of the local environmentswhere activities are to be implemented (World Bank1999c).

These viewpoints are in line with the studies previ-ously referred to which either were country specificexperiences (Adams 1989; Calhoun and Whittington1988; Clift 1988), case studies on aid effectiveness(van de Walle and Johnston 1996; Carlsson, Somole-kae and van de Walle 1997; Bagachwa et. al. 1998), ormore general studies on donor-recipient relations(Panday and Williams 1990; Helleiner et. al. 1995;Baehr 1997).

The Tanzania study done for Sida (Sobhan 1996) is aninteresting case, as a leading social scientist from onecountry where aid relations have generated considera-ble controversy (Bangladesh) uses those insights tostudy a case of considerably more serious aid depend-ency. One point Sobhan raises is the lack of localcapacity to counter the donor onslaught in terms ofstudies, data and viewpoints. He notes the efforts inBangladesh to produce their own annual developmentreview that among other things can challenge the Bret-ton Woods-orthodoxy in terms of analyticalapproaches and assumptions. This is seen as a criticalstep for a country to be able to develop a level of


knowledge and self-assertion required for meaningfuldialogue with development partners.

A common impression one gets from recipient offi-cials about aid coordination, however, is a mixture ofresignation, cynicism and some hope. The resignationis due to the overwhelming demands that donors putforward regarding how the aid coordination processshould be structured, how fast it should move, etc.This is coupled with the experience of having to dealwith donors one-on-one in bilateral annual meetings,the countless donor identification, monitoring, andevaluation mission, and then afterwards meeting thedonors as a group where the messages tend to besomewhat different. The cynicism comes from hearingso many messages regarding aid coordination, andseeing that when choices are made, donors tend to gofor their own solutions rather than joint efforts. Butthere is also some hope, as the trend is seen as being inthe right direction. There are a lot more genuine coor-dination efforts taking place, more donors are willingto join – but are also expecting better performancefrom host governments.

5.4 Nordic CoordinationThe Nordic countries have tried to improve the effec-tiveness of their development assistance through coor-dinating efforts closely with each other. Despite quitesimilar attitudes and policies, not all the attempts havebeen successful.

5.4.1 Policy CoordinationThere is a very close dialogue between the Nordic cap-itals on a wide range of development assistance issues.Under the aegis of the Nordic Council, the ministersresponsible for development cooperation meet at regu-lar intervals to discuss common positions on issuesthat have been identified as of major importance.Below the level of ministers, the political secretaries,senior officials and the senior economists in both theministries and the aid agencies meet on a fairly regularbasis. Ahead of major international conferences, bothtechnical and political management meets to discusscommon positions on key issues. In general, there is agood understanding of each other’s positions and a lotof trust. There are, however, also differences, which attimes make genuine collaboration difficult or evenacrimonious. Denmark has a more commercial bent toits development cooperation with a parliamentary rulestating that 50% of its bilateral aid should be used toprocure Danish goods and services. Finland alsoemphasizes areas where it has a comparative advan-tage. An overview for the DAC of where the differentdonors focus their resources shows that Finland is

engaged in the forestry sector in 24 countries, in 19 ofwhich it is the only donor (Danida 1999).

Coordinating with the Bretton Woods InstitutionsThe Nordic countries, along with the Baltic states,share an Executive Director position in the WorldBank, and a similar arrangement is in place for theIMF. The country that provides the Director acts asfocal point back in the Nordic countries, and organizesa small secretariat to coordinate the communicationwith the Executive Director’s office. The officials ineach capital that are responsible for relations with theWorld Bank have a weekly telephone conference thatis to provide consistency and full information amongthe Nordics on their positions on the various issuesthat come up. On operational matters, each country isresponsible for liaising with their respective fieldoffices where this is seen as appropriate. The localfield offices in turn often communicate amongst them-selves on matters that are coming up dealing with thatparticular country, such as major issues at CG meet-ings. The principle and structure for this cooperationis thus quite clear, and this collaboration is seen aslargely successful and well functioning.

There are strong external and internal incentives thatmake this system work. The overall structure is exter-nally imposed since the Bank sets the organization ofthe Executive Directors’ offices. If one particulardonor does not like it, there are no alternatives availa-ble. The position of Executive Director gives the Nor-dics the possibility of having an important voice insidethe Bank, and if this voice represents a joint mandate,the voice is heard much clearer. Since the World Bankis so important, there is a strong interest in participat-ing and mobilizing support for own positions. Therotation among the Nordics for providing the Directoracts as an incentive as well. Finally, there is a lot ofinterest from minister and political parties with regardsto Norwegian positions in the Bretton Woods institu-tions. There is therefore considerable pressure forresults – accountability – from the politicians on thepublic officials involved in this area.

General Policy CoordinationThere is considerable collaboration among technicalstaff of the Nordic aid agencies and ministries, both athead office and in the field. Over the last couple ofyears, more systematic programs of discussion havebeen put in place, to strengthen the Nordics’ role andcontributions to key debates. There are often prepara-tory sessions before SPA and DAC meetings, and top-ics like sector budget support have been addressed attwo larger seminars where external resource personshave been commissioned to prepare background stud-ies (Boesen et. al. 1999; Sida 1998).


5.4.2 Operational Coordination

Nordic Research InstitutesThe Nordic countries jointly set up the Nordic AfricaInstitute in Uppsala in Sweden and the Nordic AsiaInstitute in Copenhagen in Denmark (there was sup-posed to be a Latin America institute in Norway, butthis was never established). An evaluation of the Nor-dic Africa Institute, NAI, is largely positive in the rolethe institution has played for Nordic research andunderstanding of Africa and African issues. It acts as afocal point for a large network of institutions in Africaand the Nordic countries; as a center for “knowledgemanagement” on Africa where its library is the pre-mier documentation center on Africa in the Nordiccountries; and as a center for quality research onAfrica and with African academics (McNeill et. al.1997). The report raises some critical questions aboutwhat the primary mission of the NAI should be, but asa collaborative effort, it is seen as positive. In the caseof both institutions, however, the host country hasassumed a larger share of the budget costs, whichundermines the “Nordic” nature of them.

Coordination in the Field: MONAPIn the 1970s and 1980s, the Nordic countries tried tofully integrate their efforts with respect to some largedevelopment efforts in Eastern and Southern Africa. In1967, they started up a joint program for the develop-ment of the cooperative movement in Kenya, and afew years later a similar program in Tanzania as wellas a large-scale agricultural program in Mbeya. In1978, they established the most ambitious effort, theMozambique-Nordic Agricultural Program, MONAP.

A formal decision on Nordic collaboration was pre-sented in a Nordic Council White Paper in 1979, andthis was further elaborated in an Activity Program(“handlingsplan”) adopted in October 1986. There wastherefore considerable political support for theseefforts.

MONAP was an attempt not only at coordination, butalso at broad-based long-term development for anentire sector. It covered institutional and organiza-tional development as well as research, extension andtraining. It established a network of field stationsacross the country to cover the various agro-climaticzones and different socio-economic producer groups(small-scale peasantry as well as more modern familyfarming). It covered agriculture, livestock, forestry andfisheries. In short, a comprehensive program with con-siderable financial resources, flexibility in terms ofwhat to do, and a long-term perspective to allow insti-tutions and processes develop at their own pace.

Yet twelve years and USD 200 million later, MONAPwas disbanded, and many of the activities that hadbeen initiated were discontinued. What went wrong?

There were a number of factors that contributed to thisoutcome. The most important was the rapidly deterio-rating policy and macro-economic environment. Asthe internal strife in Mozambique increased in inten-sity during the 1980s, the field-based activities had tobe abandoned. As the fiscal crisis of the state devel-oped, the resources in the Ministry of Agriculture,MinAg, dwindled, forcing the Nordic countries to takeon an ever increasing share of recurrent costs, contraryto the original concept. MinAg itself was a weak insti-tution, and there quickly developed a rent-seeking atti-tude where MONAP was used to finance activities ofinterest to those who had access to the MONAP funds.The Nordic countries, in response to increasing prob-lems of mismanagement and simple lack of MinAgcapacity, built up an implementation unit that at onepoint had more high-level technical staff on its payrollthan MinAg itself, while it simultaneously built up ahuge administrative infrastructure to micro-managevehicles, housing, field allowances, etc. As with anumber of Nordic projects elsewhere, the technicalsolutions chosen often were qualitatively good butextremely costly both on the investment and opera-tions side, undermining the cost-effectiveness and sus-tainability. Ownership of many of the activities waslow – MONAP projects were clearly seen as Nordiccreatures.

On top of the deteriorating situation on the groundcame the problems caused by a top-heavy and exceed-ingly bureaucratic decision-making structure in theNordic countries. At the top was the Nordic Council,which defined the priorities for the joint Nordic devel-opment efforts. Then came an Advisory Committeethat was to support the Nordic Council in its policydecisions, and finally a committee of senior officials(“Embetsmannskomiteen”) that was key regardingmajor operational decisions. Below this three-tierstructure, one of the Nordic aid agencies was given theoperational secretariat function both in the Nordiccountries and the field, while in the field a consultativegroup consisting of all the Nordic aid agencies met towork with the focal agency. The different Nordiccountries had given different degrees of freedom totheir local representatives to take decisions, so in theend a lot of decisions had to be taken back in the Nor-dic countries. This created an administrative quagmire,with lots of time used to prepare documentation, argueboth in towards the coordinating groups as well as theindividual country representatives in these coordinat-ing bodies. The flexibility and cost-effectiveness thathad been intended was quickly lost. The problemswere compounded when different interests came to the


fore, for example when countries became concernedwith what share of procurement flowed by to theircountry (Lindstad 1987, various MONAP annualreports).

In the end, the combination of worsening frameworkconditions, lack of local resources, commitment andownership; inappropriate technology choices and thuslow sustainability; top-heavy and inefficient decision-making and administrative structures; and increasingconcerns by some of the partners to manage own aidresources according to own criteria led to all the jointNordic programs collapsing. In the 1990s, the inte-grated implementation model has not been pursued.

Coordination in the Field: Namibia and TanzaniaAs the independence of Namibia approached, inten-sive discussions took place between the Nordic coun-tries and SWAPO to ensure that the good politicalcooperation and economic assistance that had devel-oped could be translated into efficient support forNamibia’s development efforts.

The basic idea was to ensure a well-coordinated Nor-dic program, but avoiding the costs and problems ofearlier joint efforts. The solution arrived at was for theparties to agree to a division of labor in terms of secto-ral engagement. An analysis was carried out looking at(a) Namibia’s development strategy and resultantdevelopment cooperation needs, (b) what the non-Nor-dic donors were likely to provide, and (c) what areasthe Nordics were good at. Based on this, Norway wasasked to assist in the fisheries and energy sectors - afocus that has largely been maintained. This is a pro-gram that Norway would normally have great difficul-ties accepting, since both sectors are capital-intensivemodern-sector growth oriented and not in line withNorway’s “softer” direct poverty-focus in many othercountries. To the extent that Norway was to providesupport to other sectors, it would be by co-financingother donors’ activities.

At one point there was a suggestion that there shouldbe a joint embassy, to ensure close operational coordi-nation. This was abandoned, for a number of reasons,but one of them was that there really was no need forthis kind of coordination because of the clear up-frontdivision of labor.

The Nordics also began coordinating their sector workmore globally. The idea was that one of the donorswould concentrate on health while another might doeducation or agriculture, and that while all donorscould participate financially in a sector in a givencountry, it would be through the focal donor. Theintention was to build up a critical mass of sectoraltechnical and policy skills so that the Nordics could

participate more actively in framework and sector dis-cussions. While this agreement worked to some extentinitially, it never became fully operational, in part dueto resistance by affected technical staff in the variousaid agencies. Over time, this collaboration has fal-tered.

In Tanzania, as a follow-up to the report by the groupof independent advisers on Government-donor rela-tions (Helleiner et. al. 1995), the so-called “Nordic Ini-tiative” was developed. It partly developed someprinciples for government-donor relations that empha-sized both Tanzania’s rights and obligations. Theseprinciples were later adopted by the entire donor com-munity at a government-donor workshop in January1997. But there was a commitment among the Nordiccountries to carry the collaboration further. Agreementhas in principle been reached that the Nordic countrieswill fund parts of the local government reform througha pooling of funds (“common basket”). One of theNordics will be the lead donor and handle the adminis-trative tasks. Several studies and joint missions haveclearly laid out the principles for the joint operations(Lund et. al. 1995; Lund and Rimberg 1997), and theembassies have been given considerable leeway bytheir head offices to take operational decisions. Thereis also the idea that the Nordics should have jointannual meetings and joint programming exercises,though so far this has not happened due to the reluc-tance of one of the four Nordic countries to actuallyforego its own programming.

5.5 Aid Coordination in MozambiqueAccording to a number of criteria, Mozambique is themost aid-dependent country in the world. More thanothers, it would therefore benefit from an efficient aidcoordination system. While it has not been able toestablish Botswana’s political and administrative man-agement of external financing, it has taken a series ofsteps over the last years that are clearly strengtheningits systems and improving its credibility vis-a-vis thedonor community. Annex 1 gives a more detaileddescription of the developments.

5.5.1 Achievements Mozambique has established a rich web of coordina-tion mechanisms. CG meetings are generally held withannual intervals, where the most recent one was organ-ized in Maputo. The Central Bank organizes quarterlymeetings on balance of payments support and generalmacro policies where documentation is circulated todonors beforehand. There are joint Government-donorsector working groups in five key sectors, pursuing anumber of different models, which thus constitute arich source of learning. Finally, there is a committedand accessible political leadership that is open to dis-


cussing modifications and improvements to existingarrangements.

On the donor side, there is a (bewildering?) array ofcoordinating activities that supplement the govern-ment-donor dialogue. The World Bank and UNDPchair regular meetings with the donors; the EU andUN agencies have their internal coordination meet-ings; the Like-minded group is active; the UN systemis sponsoring around 18 small sector working groups;there is an increasing number of joint donor missionsin a number of fields. The costs of aid coordination arethus escalating as activities expand “horizontally” tocover more areas and issues, and also “vertically” asdonors delve more deeply into difficult frameworkissues.

Most progress has been achieved in the field of bal-ance of payments support. While about a dozen un-coordinated import support programs some twelveyears ago created enormous transaction costs toimporters, donors are today providing budget supportthrough market-driven mechanisms. The major gainhas, however, probably been on the institutional devel-opment side, where national systems have been devel-oped to respond to market as well as donor pressuresfor efficiency and accountability, and where the credi-bility of the national instruments is now established,though the quality still has some ways to go.

The political developments since the elections in 1994have improved the legitimacy of government positions.Parliament in particular has strengthened its role insetting national priorities, establishing greateraccountability within what was formerly a one-partycentrally managed state. The policy instruments, how-ever, are still poorly integrated, so the ability to steerexternal financing towards priority areas remainsweak.

The sector working groups in the roads, health, agri-cultural and education sectors have quite different his-tories. The roads program was originally prepared aspart of a World Bank-led sector investment program.Due to its history and thus the investment profile thathas been established, it is still seen as largely a WorldBank-driven program. In agriculture, a very lengthydevelopment and discussion process lay behind theidentification of the PROAGRI program, where finally14 key donors agreed to a 48-point program as thebasis for the collaboration. While the negotiationshave at times been heated, agreement has formallybeen reached on the key issues to be addressed –though it is unclear as yet if the capacity and commit-ment to implement is in place. In the health sector, aslower process of identifying “champions of change”

within the health system has been pursued, whereoperational improvements in a number of areas havebeen attained. There are, however, problems getting inplace the overarching strategic vision and action planthat will improve the overall efficiency of the healthcare delivery system. In education, a low-key andrather slow process has ensured local ownership of thetotal process, but at the cost of agonizingly slowprogress.

5.5.2 Conflict and ChangePartly by default, Mozambique is putting together asystem that builds on a number of building blocks. Insome processes, donor involvement has been over-whelming and has led to conflicts during the process(both PROAGRI and ROCS have had some bumpystretches); on the other hand, in others progress hasbeen extremely slow when donors have not pushed(education) or uneven despite donors pushing (health).Strategic comprehensive sector agreements haveyielded impressive action plans with focus on keyissues – but uncertain ownership and capacity toimplement (PROAGRI). Learning-intensive participa-tory processes build coalitions and insights as integralparts of the approach – but are often not able to tacklesystemic issues, at least in the short run (healthSWAP). Which approach is most appropriate is a func-tion of what is feasible, what the existing situation inthe sector is and thus at what level problems can beattacked.

One of the lessons from the Mozambique experience isthat donors underestimate the resistance to the changesaid coordination lead to. Successful aid coordinationmeans that access to external financing goes through acentral government mechanism, and that public offi-cials have to compete for these resources through amore transparent political process. Many officials havebuilt a strong position within the public sector due totheir direct access to donor financing that bypassessuch central mechanisms. Their ability to resist changecan be substantial (often thanks to donor-fundedprojects), underlining the importance of identifying“champions of change” that can successfully carrythrough these transformations from within the system.Understanding the politics of aid coordination is there-fore often more important than understanding the eco-nomics of it, yet donors tend not to be very good at it.This leads to exaggerated expectations and unrealisticbenchmarks for changes, which means that the processbecomes donor-driven (see Dixit 1996 for a good anal-ysis of economists’ frustrations in understanding poli-ticians’ criteria for decisions).


5.6 Summing Up• There exists a fairly dense structure of aid coordi-

nation mechanisms and institutions at the interna-tional level, but where overlaps and inefficienciesexist. Many of the more influential actors are stillheavily donor-driven, though the tendency istowards more open and participatory processes.

• The national apex coordination mechanisms,Roundtable and Consultative Group meetings, areunder review. The UNDP is looking into the RTmeetings and the UN Development AssistanceFramework, while the CG meetings are becomingmore recipient oriented, both in terms of prepara-tion, hosting and siting of the events. There is stillsome ways to go in terms of genuine participatoryownership and management – in part a function ofcapacity and political will on the side of recipientgovernments.

• Donor practices have historically been a majorfactor in blocking more efficient aid coordinationon the ground. Though improvements are takingplace, particularistic donor practices still createconsiderable obstacles and costs to host govern-ments.

• Recipient ability to manage aid coordination in amore efficient way is as much a function of politi-cal legitimacy and credibility as it is of technicaland administrative skills and capacity.

• Recipient perspectives on aid coordination are ingeneral somewhat less euphoric than donor views,though there is a sense that more genuine partner-ships and active listening is developing, thus alsogenerating more optimistic views on the side ofrecipient officials.

Nordic Experiences• There is considerable agreement among the Nor-

dic countries with regards to basic principles fordevelopment cooperation: putting the recipient inthe driver’s seat, poverty reduction, etc. Theagreement on principles has led to interest in har-monizing aid activities – yet experiences so farfall well short of intentions. When there is astrong, direct and continued political interest inhaving coordination work, it does (coordinationwith World Bank).

• Simple, robust systems work best (up-front divi-sion of labor – no complicated administrative andcontinuous political management required –Namibia). If this is coupled with interests ofimporting stakeholder groups in the donor coun-try, the structure may continue (Namibia). If it isin contradiction with interests of stakeholdergroups in the donor country, it is more likely tofail (sector division of labor between Nordics).

• The degree of coordination is more determined bydonor than recipient wishes, though when therecipient is strong and unambiguous, this providesimportant backing to those donor staff who sup-port coordination (Namibia).

• The Nordic experiences indicate that even whenthere is a high degree of commonality amongstthe donors, efficient aid coordination that is afunction of donor consensus will only be sustain-able under certain circumstances. This arguesstrongly in favor of recipients being the determi-nant for continued and successful coordination.

6 LESSONS FROM COORDINATION 29

6 Lessons from Coordination

To summarize the lessons from the various coordina-tion experiences, it may be helpful to look at thesefrom the point of view of the different dimensions thathave been used in chapter 4.

6.1 Content of Coordination

Policies and PrinciplesThe point of departure for the Mozambique survey(ECON 1998) was the wish on the side of the donorcommunity to identify ways to improve aid coordina-tion. One hypothesis was that if donors could cooper-ate on the preparations of their assistance strategies,then a lot of other coordination issues would be takencare of as well: a joint strategizing process would gen-erate consensus on what the major objectives for exter-nal assistance should be, and once this was in place, itwould be easier for each donor to see how their partic-ular assistance could fit within the larger picture. Thestudy’s basic conclusion on this point was rather nega-tive, for two sets of reasons:

• As a matter of principle, the donors cannot ontheir own define overarching development objec-tives for a country. Either the recipient govern-ment has an acceptable development plan orstrategy in place – or it should receive assistancedeveloping these, if it so wishes, but clearly underits own management and in a transparent fashion.Too much donor involvement or interference cancause serious problems for the national politicaldialogue if the government is seen as listeningmore to the donor community rather than tonational stakeholders. The risk of damage tonational political accountability is seen as over-whelming.

• On the practical side, the guidelines the variousdonors have for elaborating strategies are so dif-ferent that the costs of actually harmonizing thework would be extremely high. But it was also notclear what the product of such an exercise wouldbe. The one example of joint programming inMozambique was the first UN DevelopmentAssistance Framework. While UNDP was satis-fied with the outcome, several UN agencies felt itwas a time-consuming and formalistic processthat produced a political compromise rather than afocused and operational program. This problemwould be compounded if the exercise were toinvolve a number of bilateral donors, as thiswould require political clearance from a series of

headquarters – a process that would be time-con-suming and prone to break-down.

The main conclusion of the Mozambique study wastherefore that the most cost-efficient and politicallyviable approach would be for donors to assist the gov-ernment develop its vision and the necessary tools toimplement it: budgets and plans at the national andsector levels. What the donors could bring to the tablein terms of principles and priorities would be the DACShaping the 21st Century-document.

This was a controversial conclusion that was notshared by all. There were those who felt a moredirected dialogue on policies and priorities not onlyis feasible but required in order for the efficacy ofexternal financing to reach acceptable levels. If thisdid not happen, it would be difficult to justify contin-ued aid.

Part of the disagreement could be traced to differentunderstandings of the partnership concept, wherejoint responsibility for policy choices clearly gavedonors a role and legitimacy in discussions, if not indecisions.

One issue was the problem of “ganging up”: at whatpoint does donor coordination go from being a con-structive attempt at sorting out internal differences sothat interaction with government is simplified, toestablishing donor concurrence on controversial issuesand thus “ganging up”? For some donors this repre-sents a real dilemma: the pressures towards internaldonor coordination and consensus may lead to a hostgovernment facing a united but possibly hostile donorbloc.

A key question is how far aid coordination should go,and who determines the limits. The conclusion fromKillick’s study – that conditionalities do not workbecause donors in fact do not impose the sanctions thathave been threatened – can be used to argue for closercoordination on the donor side to ensure that condi-tionalities are in fact respected. In 1986, then-VicePresident of the World Bank, Ernest Stern, severelycriticized bilateral donors for not in fact respecting theSAP frameworks since they provided funding undercircumstances where the SAP criteria for doing sowere not fulfilled.

But that leaves the problem of what to do in a situationwhere the government in principle is pro-developmentbut has severe capacity problems, or where line minis-tries are not interested in pushing the development

30 6 LESSONS FROM COORDINATION

agenda4. If donors do not push for change, it will notcome about very fast due to the forces of inertia andresistance to removal of privileges. Without a goodgovernment policy in place, donors have to compen-sate for this by spending relatively more resources onidentifying where their own financing should be used.Doing this in tandem with other donors is clearly morerational than going it alone, but then comes up againstthe problem that at the end of the day this wouldimpose solutions on the recipient. But with no pro-active programming in place, issues like povertyreduction, gender and environmental degradation willnot be addressed, so the costs of non-action are notborne by the government but by priority beneficiarygroups.

There is no easy solution to the various contradictionsthis problem poses. The answer provided in theMozambique case was based on a mixture of principleand pragmatic notions that donor-imposed solutions,no matter how well designed and well intentioned, willin the end be less sustainable and cost effective thanthe ones generated locally, and that donors will have tolive with that.

Procedures and PracticesHaving a common set of policies and priorities withthe government is key to improving the relevance andefficacy of external financing. It does not necessarilyimprove the cost efficiency of using aid resources,however, because the highly different procedures andpractices within the donor community put an enor-mous burden on the recipient administration. The listof donor sins is long and well known (see section 5.3).And it is at the level of procedures and practices thatthe real costs to the recipient appear. Disagreementsover policies and priorities lead to negotiations overwhat is to be supported – the procedures and practicesdimension deals with the nuts and bolts of the how.And the devil is in the details: 20 donors with supportto an agricultural program may disagree on what ismost important to do, but the government can usuallyfind an overarching program where most interests canbe accommodated. There is no basis upon which tofind a least common denominator on procedures, how-ever, since each donor is accountable back to her ownhead office for following administrative requirementfor the use of public funds.

In the case of Mozambique, pooling of funds wasmade possible in a couple of areas when a small group

of donors was willing to take some risks and accept theMozambican procedures. In return, more intensiveconsultations and joint monitoring was agreed to sothat problems could be quickly identified andaddressed. While this principle was fine, in practiceproblems arose. This was partly because Mozambicancounterparts did not have the additional capacity toimplement the more demanding coordination and bet-ter quality tasks that were now being asked of them.But also some of the local partners in the system werenot interested in the changes. While the hoped-forimprovements did not materialize as quickly asexpected, it was felt that these new ways of workingtogether still represented progress compared with themore fragmented ways of doing business. With theprinciples for the new procedures in place, the imple-mentation of them could be addressed over time as ajoint responsibility.

Two lessons were drawn from this experience. Thefirst was that not all donors are able to able to adhereto a recipient standard for legal reasons: home countrylaws demanded that own procedures were adhered to.But if the government along with a core group ofdonors could establish a track record of acceptableresults in terms of reporting, accounting etc., thenother donors are often able to modify their own proce-dures so that they can take advantage of and becomemore or less integrated into the national practice with-out formally signing on.

But the second lesson was that if there were agreementon policies but problems with procedures, the formal-istic regulations would tend to win over principles andpolicies. It was extremely frustrating for the Mozam-bicans to go into negotiations on policy matters andbelieve the two parties had arrived at a consensus forexample on pooling of funds, only to discover thatimplementation became next to impossible due to pro-cedural differences on the donor side that individualdonors were not able to overcome because of guide-lines or instructions from their home offices. Muchmore attention thus needs to be paid to this set ofissues.

6.2 Degrees of Coordination

Consultations/Informational Meetings Meetings where information is shared is the mostcommon form of aid coordination. Such meetings takeplace both in government-donor and inter-donor set-tings, across a wide range of issues and sectors. Due tothe large number of such meetings, this form of coor-dination is the single largest user of time and yet theone where donors seem least satisfied. In the Mozam-bique survey, the typical comments were that the

4. It should be noted that the above conclusion was reached in thepositive context of a country where the political commitment todevelopment and change is quite strong, though capacity toimplement is limited. In cases where such commitment is lack-ing, the conclusions from recent studies (Killick 1998, WorldBank 1998a) are quite clear: neither the carrot nor the stick -funding and conditionality - will ensure genuine policyimprovements.


meetings are poorly prepared with no operationalagenda. There is an unstructured and unfocused pres-entation of large amounts of disconnected information.Many donors are not open and honest about what theyintend to do but rather provide information about deci-sions already taken, so there is little information giventhat can help planning and genuine ex ante coordina-tion. In general there are too many such meetings withlittle value-added produced.

Despite the general grousing, most donors attend whatthey consider to be an excessive number of such meet-ings. When challenged as to why they continue mis-allocating their own time in this way, two reasons areusually given: (i) there is a fear that at this meeting,something important may come up that they mightotherwise miss, (ii) political pressures to be good cor-porate citizens by being present and thus show supportfor the general objective of coordination.

The main conclusion was that this form of aid coordi-nation is rather ineffective. It yields little in terms ofimproved performance or focus of resources, there islittle registration of any kinds of commitments or deci-sions so nobody knows what the actual outcomes ofthe meetings are, and this lack of structure leaves mostactors quite frustrated. Such informational meetingsdo play an important role in terms of involving all par-ties, however, because a number of donors have diffi-culties joining more operational or strategiccooperation exercises.

The main recommendations that were made in the caseof Mozambique was to (a) reduce the number ofgroups, (b) define some objectives or targets for thegroups to achieve over the next six to 24 months, sothat some performance monitoring was put in place,and (c) put in place a budget that would reveal whatthe cooperation exercises really involved in terms ofresources. The most important cost is undoubtedlystaff time, both among donors and recipient, so onechallenge would be to come up with a good pricingsystem for the time of various parties. An alternativewas to provide a hierarchy of coordination mecha-nisms, where ceilings in terms of number or length ofmeetings for the various areas was established, so as tokeep time costs down and force more content into thefewer meetings that would be held.

Cooperation and CollaborationMore strategic forms of cooperation, either to harmo-nize policies, priorities and principles or implementa-tion procedures and practices, take place in a numberof different fora, from CG meetings, local aid coordi-nation sessions chaired by the government or by focaldonors, sector working groups, down to smaller group-

ings such as the internal EU and UN meetings or theLike-minded Group.

During the interviews for the Mozambique study, itbecame clear that different groups of donors had quitedifferent views of how closely donors ought to collab-orate. The Like-minded Group is quite positive aboutcoordination, as long as the recipient is clearly manag-ing the process. What could be termed the Anglo-Saxon bloc – the Bretton Woods institutions, USAIDand to some extent DFID – were more aggressive interms of “getting basics right”: there was a feeling thatmany recipient governments were not sincere in pursu-ing pro-poor policies and that donor conditionalitywas required to improve aid efficiency. Finally, the“Latin bloc” of southern European donors was moreconcerned with ensuring that the formal donor-recipi-ent relations were in place. To them, donor coordina-tion smacked of paternalism and interference in asovereign nation’s right to define the relationship itwanted with any particular country. At the same time,the southern Europeans were quite open about the factthat their aid programs were more tightly integratedwith strategic, political and commercial interests.

These divergences result from fundamental differencesin worldviews in the different societies. As such theyare not amenable to negotiations, but must be taken asfundamental “facts of life”. There has been conver-gence through the 1990s, however, so the gap inapproaches and priorities among the donors is nar-rower than it was ten years ago. But there are stillsome differences that need to be recognized and thatset limits to how far many actors are able to go interms of formal commitments.

As noted above, it is usually easier to reach agree-ments on overarching concerns like pollicies and pri-orities, than on how to implement them. The problemsdo not arise simply because some donors are particu-larly difficult, but often because well-intentioned aidagencies pursue a variety of development objectivesthrough managing the structure and size of the pro-gram and how it is to be implemented. The evaluationof Norway’s development cooperation with Tanzaniafor the 1994-1997 period noted that if all donorsbehaved like Norway, there would have been 500 bigand 2,000 small projects during any one year. This wasat a time when Tanzania was being pushed very hardby the donor community to reduce its investmentbudget to a “super core” of only 125 projects.

The administrative costs to Tanzania of such a pro-gram would have been enormous, forcing the countryto allocate scarce management resources to addressthe planning, monitoring, reporting, accounting andauditing needs of such a highly fragmented program. It


points to what may be the single largest source of effi-ciency gains from aid coordination: reducing adminis-trative and transaction costs to both donors andrecipient. In the Mozambique study, the claim wasmade that the most efficient mechanism for aid coordi-nation was formal contracts. This would reduce oreven eliminate the more pernicious transaction costs,namely uncertainty and information costs by makingobligations and rights for both parties explicit. Thisgets rid of the ad hoc adjustments that tend to ariseduring the lifetime of an activity, and in particular dis-ciplines donors who have a tendency to change theirways of doing business when it suits them.

Establishing such contracts can be extremely costly.The Mozambique study noted that it took one year ofquite tough negotiations before the pooling agreementon technical assistance to the health sector was inplace. While the investment costs of such formal con-tracts therefore may be quite high, they reduce therecurrent transaction costs, and these in the long runare much more important. The health sector case notedthat not only did efficiency of external resourcesincrease dramatically, but the institutional develop-ment impact was also positive. The parallel systemsfor recruiting and paying expatriate health personnelwere done away with, external interference in review-ing medical qualifications no longer took place, etc.(see Annex 1 section 1.1.4). These, however, are quiteimpressionistic conclusions, but worth following upwith more careful studies, since a transaction costsapproach to understanding coordination costs and ben-efits may prove to be a key step forward.

6.3 Coordination at National LevelThe major gains to coordination come at the nationallevel when systemic changes are arrived at. The historyof balance of payments, BoP, support in Mozambiqueis instructive. With the introduction of the country’sfirst structural adjustment program in the mid-1980s,donors were encouraged to provide BoP support in theform of import support programs. The incredible inef-ficiencies introduced by a dozen different donorsregarding procurement rules, documentation, report-ing, accounting and auditing led to the establishmentof a government coordination office that received tech-nical assistance from a number of donors, but in awell-structured fashion. Joint Evaluation Missions,JEMs, were undertaken to sort out problems together.This turned out to be extremely constructive, as donorswere confronted with the operational problems the dif-ferent import regimes led to. Agreements werereached to standardize and harmonize a series of mat-ters, reducing the costs considerably and strengtheninglocal systems and processes. As confidence in localmanagement and reporting grew, more donors were

willing to provide un-tied BoP support as “basketfunding” under the common set of rules.

The financial sector was restructured. The commercialbanking parts of the central bank were hived off andthe state banks privatized while banking supervisionand other policy and regulatory functions werestrengthened. The sector was opened up to foreignbanks, and the financial sector deepened by introduc-ing new savings and credit instruments. This provideda different environment for BoP support, opening uppossibilities for more market-determined resourcetransfers. One donor thus begun providing free foreignexchange to the central bank on the proviso that theforeign exchange was sold under free and competitiveconditions, that the countervalue funds thus generatedwould be transferred to the Treasury for budget sup-port, and that monitoring and reporting systems wouldbe improved to allow for better tracking of the flow offunds. The development of the new system proved sosuccessful that other donors followed suit, with theresult that there have now been several joint BoP mis-sions working with the national authorities partly toaudit performance but even more to work with thelocal institutions to find ways of further improving thesystem. As a result of this, there was a very rapid shiftfrom various forms of tied BoP financing – commod-ity aid, import support, restricted BoP financing – toun-tied budget support. While the ratio of the blackmarket to official market exchange rate was 40:1 in themid-1980s, there is now a unified market-determinedrate. The fall in administrative costs to donors, recipi-ent and not least the purchasers of foreign exchange isimpressive.

The key gains were in the field of institutional devel-opment. If one makes a distinction between institu-tions as representing “the rules of the game” whileorganizations are “the actors who play the game”,there has been major capacity building or organiza-tional development within the various actors: the cen-tral bank, the commercial banks, the coordinatingoffice for import support, etc. But the institutionaldevelopment has been more important. Due to restruc-turing of the financial sector, market mechanismsdetermine more efficient allocations of foreignexchange. The roles of the different actors are moreclearly defined providing transparency and thus betterpossibilities for accountability in the transactions.Removing the distortions in the market has reducedthe incentives for rent-seeking and corruption. Thereporting that is being produced is primarily for thenational authorities to manage the macro-economybetter, but is the basis for reporting back to donors aswell. While this has reduced the accounting, reportingand auditing costs enormously, the major benefit isthat it has shifted government attention away from


reporting to the donors and instead on developing ownreporting systems based on national managementneeds.

The point is that it is not simply individual organiza-tions that have improved – the major gains have comefrom the fact that there are now improved systems andprocesses in place. It has taken considerable time toachieve this because it was not easy to convince actorsto change behavior and rules. At times it was the gov-ernment that resisted, at other times donors, but asmutual trust was built based on actual performance, afew donors were willing to take some risks and try outnew procedures, which would then push the experi-ence frontier forward. This step-by-step approach alsoallowed the government to build capacity at a morerealistic pace. The systems that were developed neces-sarily were Mozambican as donors’ own proceduresand approaches had to give way to the consensus solu-tion. While the new systems may have contained con-siderable inputs from external sources, they werecodified and passed in their final form by Mozambicanauthorities, which gave them an improved basis fromwhich to demand compliance, by the donor commu-nity. Any parallel management and reporting systemssuch as by donor-funded project implementation unitswould structurally be undermining this institutionaldevelopment.

In parallel to the reforms in the financial sector,improved budgeting and planning has improved over-all transparency and accountability regarding resourceallocations, budget control and auditing. ThoughMozambique still has quite a way to go before publicfinance management has reached acceptable stand-ards, the trend is very positive. Donors have thereforecontinued to provide large-scale assistance, much of itin the form of un-tied or weakly tied program aid.While some of this is in recognition of fungibility andthus the futility of earmarking resources, the increas-ing reliance on national mechanisms to determineresource use has reduced administrative costs to bothparties considerably.

As a response to these structural changes, donorbehavior has changed. There is an increasing share ofdonor funding in the form of program aid rather thanproject funding. In the case of Mozambique, there ismore predictability in the resource flows as donorschannel more through the negotiated programs ratherthan ad hoc short-term allocations. An increasingshare of funds is going to finance recurrent costs ratherthan investments, getting a better balance in the publicbudget. A greater share of donor resources is reflectedin government budgets, so there is more transparencyas well as predictability with respect to externalfinancing.

Emergency programs tend to provide the most com-prehensive and consistent coordination programs. InMozambique, the UN system took a lead first in theemergency relief program, but during the transitionalprogram from when the peace agreement between thebelligerents had been signed in 1992 and till thenational elections were held in 1994, it nearly func-tioned as a shadow government. The key to this rolewas the legitimacy it had as a neutral and non-inter-ested party plus the political legitimacy conferred bythe UN Secretary General. Once the extraordinary sit-uation was over and an elected government was inplace, the elaborate coordination mechanisms disap-peared quickly. While some regretted losing some ofthe technical skills and oversight capacities that theUN system had built up, it was also clear that it hadbecome too intrusive for a peacetime situation. Theelected government needed to establish itself and putin place a dialogue with national stakeholder groupswithout the interference of external actors. There arethus limits to how far donor coordination should go.

6.4 Sector CoordinationMany of the more interesting coordination experiencesare taking place at the sector level. Sector InvestmentPrograms, SIPs, and Sector Wide Approaches to Pro-gramming, SWAPs, have generated considerable inter-est as vehicles for improved coordination of local andexternal resources. One definition of what defines aSIP or SWAP is that it must (i) be sector-wide inscope; (ii) based on a clear sector strategy, (iii) localstakeholders must be fully in charge; (iv) all the maindonors must participate; (v) there are common imple-mentation arrangements; and (vi) local capacity ratherthan external technical assistance should be used(Jones 1997, p. 1). Other variations on these criteriaexist (i.e., Cassell 1997, which is more detailed), butthe fundamentals are the same: the approach is com-prehensive in scope, time, number of actors, and link-ages.

The above criteria are quite stringent, and a central les-son is that moving to a full-fledged SWAP is a time-consuming process that should not be rushed. The keyissues are – as usual – to ensure that local politicalcommitment and implementation capacity is in place;to get donors to adhere to common and transparentprocedures; and to find mechanisms that are mutuallyacceptable for addressing problems and conflicts.

One important challenge is the integration of sectorwork within macro frameworks. Ensuring that sectorplans are sustainable given the overall resource availa-bility and priorities determined by national authoritiesis often ignored. Sector working groups have turnedout to be good fora for raising these kinds of issues,


given that it is often more meaningful to discuss pro-ductivity and sustainability of donor-funded activitiesat the level of the sector rather than at general macrolevels. The goal is often to develop activity-basedbudgeting with clear outcome criteria that will permitintegrated financial and performance audits.

One key lesson from sector work is that central bodies– ministries of finance or planning, the Prime Min-ster’s office, or the central bank – should be involved.These are central government institutions that have anobvious interest in improving public finance manage-ment, and are often key allies or “champions ofchange” for sector development.

Another finding is that it is often more important tofocus on implementation procedures rather than theformal sector policies. Policies are more prone tochanges whereas implementing modalities tend to berelatively stable (Boesen et. al. 1999). Furthermore, itis often how policies are implemented that will deter-mine actual outcomes. This is another reason why har-monization of donor actions is important.

Questions have been raised why, in an era of fungibil-ity, one should bother with SWAPs and sector budgetsupport instead of just general BoP support? - Theanswer is in part that each sector faces a particular setof institutional issues – the structure of the transactioncosts to be addressed varies. Sector budget support isthe appropriate vehicle for addressing systems issuesat ministry level and ensuring that for example minis-terial budgeting and accounting is linked to nationalsystems (surprisingly enough not always the case!).Macro programs sometimes contain objectives relatedto health and education sector performance, which isnot appropriate: if the agreement is with the Ministryof Finance, it should only contain issues for which theministry itself can be held accountable. Likewise, sec-tor budget support should probably not containdemands regarding national budget priorities (increasein the share going to social sectors, for example), sincethis is dealt with at a level above the individual lineministries.

There is a lively debate going on regarding what thepre-conditions for sector support should be. Manydonors feel they are facing difficult choices regardingmoving from project to program aid, but without clearguidelines as to when this is appropriate and even lesswhat the monitorable variables are that should be fol-lowed to see if the program is on track. While manyagree that in principle program aid is better, thisshould somehow be revealed in practice, so it shouldbe possible to identify some verifiable dimensions tosee if it is correct or not.

One response to this has been that the sector approachis very much a process- and learning-based one, forboth parties. It simply is not possible for an aiddependent country to come up with satisfactory per-formance along the key criteria defined for a SWAP asa pre-condition for SWAP support. Instead, these char-acteristics will be outcomes of the successful SWAPprocess itself. Donors should therefore (once again) bewilling to take more risks than they currently do – butnot blindly so. Realistic improvements along a limitedset of dimensions should be established, with flexiblenegative and positive sanctions established as incen-tives for performance improvements. Work is thusrequired to understand the existing situation (bench-mark study) so that performance improvements can beidentified and subsequently monitored.

A recent study done for Sida on the practical issuesregarding sector budget support points out the difficul-ties and costs of such work actually being successful(Sjölander and Brobäck 1999). The study notes thework that the World Bank and IMF had to put intomany of the African countries in order to come upwith an acceptable understanding of the macroecon-omy and public finances, and be able to track this overtime. Even after many years of work, it is recognizedthat the factual basis for policy decision-making maybe fragile. And the annual consultations and monitor-ing visits are still required and remain costly. In veryfew instances have donors been willing to invest simi-lar resources to address sector issues. Without this,however, the consultants believe that the sector workwill not generate the kinds of outcomes that areexpected.

The real challenge, however, is that this level ofinvolvement leads to commitments on the side of bothparties that in particular the donors do not seem tohave fully considered. Sector programs entail long-term economic and political engagements – genuinepartnership in fact – that mean that donors must getmore heavily involved in national sector policies thanmany of them feel comfortable with. The level ofintrusion becomes much higher than when individualprojects are being discussed. At the same time thecoordination that joint sector imposes also reducesflexibility on both sides of the table: individual donorscan no longer avoid the responsibility for joint deci-sions. When mechanisms like basket funding are inplace, various forms of conditionality will in fact alsobecome effective since other financing mechanismswill have been reduced, removed or at least stronglyfrowned upon by the donors involved in that sector.


6.5 Geographic CoordinationCoordination along geographic lines is also beginningto take place. In Mozambique, the EU member-stateshave agreed amongst themselves that in each province,one EU country will have a particular responsibilityfor coordinating all EU-funded activities in that prov-ince. There is even a suggestion that the embassiesestablish a provincial office in “their” province, toensure operational coordination on the ground. InUganda, various aid agencies also have been asked toprovide assistance to particular districts.

What is of considerable concern is that some of theseinitiatives are taking place at the same time that vari-ous forms of decentralization of the public sector arebeing promoted. Creating a geographic “division oflabor” between donors during a process of decentrali-zation may create problems. The reason is that there ismuch less accountability and countervailing forces atlevels below the national. While donor-driven proc-esses at the national level are problematic, at regionalor local levels they can become very distortive. Duringthe war in Mozambique, there were NGOs that hadlarger operating budgets in a province than the provin-cial administration. This of course created major prob-lems, not least of all in terms of the dependence of thelocal administration on NGO resources. At districtlevel it is much worse: locally elected officials canoften only stand by and watch as NGOs and otherdonor-funded activities move ahead according to theirown agendas. While there are many cases of goodcooperation and coordination, this is much less astructural phenomenon than a case-by-case outcomelargely dependent on the individuals involved.

When this is combined with decentralization, prob-lems increase. In the Mozambique case, there are pro-posals that the EU country coordinating activities in aprovince be given a particular responsibility for sup-porting the local administration. Given the very differ-ent administrative procedures and traditions in the EUcountries, one can imagine the impact on the consist-ency of provincial administrations if one provincereceives help from Sweden and another one fromFrance, for example.

This problem has already come up in the health sector,where different donors provide assistance to differentprovinces. One donor has provided quite massive capi-tal assistance to the Manica province for rehabilitationof infrastructure, but has managed most of the pro-gram through own consultants. Another has tried tobuild up large parts of the health system in Tete, but aspart of a larger assistance in the province, creating anew set of dependency distortions. A third donor hastried to strengthen the overall health finance adminis-

tration, but comes up against problems of identifyingwhat the real resource availability is due other donorsnot providing full information, so that its assistancemay in fact further distort resource allocations.

The question is if donors may be re-producing some ofthe mistakes of the first structural adjustment pro-grams. The challenge at that time was first seen to beto remove distortions to economic markets, so that the“level playing field” could be established. It turned outthat getting to the situation where the principle of “onedollar – one vote” in fact worked required not onlyremoving the structural barriers to more competitivemarkets, but also to put in place implementable pro-competitive regulatory frameworks. If not, the highlyskewed income and wealth distributions coupled withhighly concentrated political power tended to createnew forms of market distortions that worked againstthe poor.

Similar problems exist in the political markets. Gettingto a situation where the principle of “one person – onevote” becomes reality requires probably even moretransformation and pro-competitive regulatory andoversight capacities than was the case with the eco-nomic system. Yet donors’ knowledge of and insightsinto local political processes and power constellationsis considerable poorer than was the case when eco-nomic markets were tackled.

Because the pressures for getting results are considera-ble, the easy and most obvious things are done first,with the danger that the process is done somewhatbackwards. The typical first step is to identify what isnot working well, and then assume that a restructuringwill address the issue. Since central government oftenis not good at delivering services, decentralizationseems like a good move, based on the notion that if“government comes to the people”, local politicalprocesses will be more participatory and democratic.Once it becomes clear that local politics may in fact bemore manipulative and corrupt, it can become quiteexpensive and difficult to sort things out, as strength-ened local political stakeholders can mount a strongdefense of their privileges.

In Tanzania, the Government established the TanzaniaRevenue Authority, TRA, to improve tax collections atthe center, as this was a major problem in the area ofpublic finances. The TRA is now seen as doing a cred-ible job. At the same time, a research team has foundthat as much as 90% of locally collected revenue maynever find its way into the public coffers (Fjeldstadand Semboja 1999). At the local level there are usuallyno bodies that have oversight functions and the politi-cal clout to address such issues.


Aid coordination along geographic dimensions pointsto the downside risk of coordination. If the principlesupon which the coordination is done are not good, thecosts to the recipient can become extremely high.Donor coordination based on a division of labor byarea coupled with decentralization may be a recipe fordisaster. One possible outcome is a fragmentation oflocal public administration without systems ofaccountability and transparency in place. Nationalinstitutional capacity is thus weakened or even under-mined while no complementary or supporting systemsare put in place locally – the worst of both worlds.

6.6 Summing Up • If the host government does not have a develop-

ment program and strategy in place, donors can-not substitute for this by coordinating their ownstrategy processes, but should rather support gov-ernment’s efforts at developing a credible program

• It is easier to establish agreement on policies andpriorities than on procedures and practices. It isthe latter, which address implementation differ-ences, where coordination gains are potentiallythe greatest, however, since these differences gen-erate enormous administrative and transactioncosts to both donors and recipients. Policies fur-thermore are more subject to change than proce-dures, which tend to remain amazingly stable, yetit is often procedures that determine impact on theground.

• It is not necessary to have all donors in an areaagree to new procedures and practices. It is usu-ally enough that a core group of donors take thelead, assume some of the risks/experimentationcosts to work with government to develop the newapproaches. Once a track record has been estab-lished, other donors are often willing to modifytheir own procedures to come as close as possibleto the new government/donor standard.

• The most common mechanism for aid coordina-tion - informational meetings – may also be themost ineffective. Streamlining and downsizingsuch meetings can free up scarce time for moreproductive aid coordination activities.

• Formal contracts may represent one of the mostcost-efficient mechanisms of coordination. Theyare costly to establish – investment costs are high– but they generate lasting and potentially sub-stantial reductions in recurrent transaction cost,and these are the real problems for a local admin-istration.

• A wide range of different approaches to aid coor-dination can be helpful in a first phase of develop-ing overall aid coordination instruments andstructures, as experience on the ground is requiredto identify least-cost solutions. Differences inapproaches may also reflect different political andeconomic realities in various sectors. But overtime national authorities need to impose coher-ence and discipline on external resource flows andmake aid coordination more transparent through aconsistent set of rules and mechanisms.

• Successful aid coordination implies structuralchanges to resource flows. This will generateresistance from those within the host country thatfeel they will lose out. Since these actors usuallyoccupy central positions, they can mount effectiveresistance, a problem that should not be underesti-mated. Understanding the politics of aid coordina-tion is therefore as important as understanding theeconomics of it.

• Institutional development is key to aid efficiencyimprovements as this is the means through whichsustainable reductions in transaction costs areattained. Organizational development is helpfulbut in itself of less importance (technical-adminis-trative capacity and skills within a given organiza-tion).

• Coordination at national level is required toachieve systemic changes, which in turn generatemajor benefits to the host country. The most inter-esting coordination experiences right now seem tobe taking place at the sector levels, however,where innovative approaches such as SWAPs arebeing tried out in a number of countries.

• Recent reviews of what it takes to implementcomprehensive SWAPs highlight potential down-sides since a genuine SWAP may imply a level ofjoint interaction at the policy and financing levelsthat some parties may not be comfortable with.On the one hand, a SWAP demands considerableown administrative and policy resources that fewdonors have indicated so far that they are willingto commit on a long-term basis. But it may alsorequire tying own financial resources to largerjoint frameworks in a way that some donors mayfind troubling. These issues merit very carefulconsideration.

• Geographic coordination in connection with pub-lic sector reforms/decentralization is also seen ascontaining serious downside risks, which mayamong other things undermine institutional devel-opment at the national level.

7 THE UNFINISHED AGENDA 37

7 The Unfinished Agenda

While aid coordination is considered only a dimensionwithin the partnership concept, the purpose of partner-ship and thus also of aid coordination is to improvedevelopment effectiveness – that is, the extent towhich intended purposes behind the provision ofexternal financing have been attained.

7.1 Measuring Aid CoordinationThe World Bank study on aid coordination provides ananalytical frame for studying these questions consist-ing of five dimensions: Relevance, Efficacy, Effi-ciency, Sustainability and Institutional Development(World Bank 1999b). The preliminary conclusions arethat aid coordination is highly relevant and improvesefficacy regarding development impact. The resultswith respect to efficiency, sustainability and institu-tional development are considered more uncertain.

One reason for this dichotomy in the findings may bethat it is easier to agree that some of the principles foraid coordination are being attained while it is difficultto measure the more operational dimensions of it.

One way of addressing this problem may be to look atsome of the links that can be established between thethree dimensions of efficiency, sustainability and insti-tutional development. One assumption behind some ofthe analysis in this report is that institutional develop-ment is key to sustainability. Only when improvednational systems, processes and capacities – institu-tional development - are in place will local activitiesbecome more sustainable. Institutional improvementscan often be measured as reductions in transactioncosts of various kinds. The challenge is to operational-ize transaction costs so that they can be monitored (notnecessarily measured at the cardinal level, but at leastmonitored for change over time). These transactioncosts are of various kinds – money, time, information,uncertainty – but as soon as they are identifiable, thenone also has a measure of efficiency gains.

7.2 Designing Least Cost Aid CoordinationSubordinated to this general problem of aid efficacy ishow to make aid coordination itself more cost effi-cient. This is a more manageable problem, as aid coor-dination activities are more easily identifiable andmeasurable (here again a transaction cost approachmay prove to be helpful).

More work should go into looking at some case stud-ies. In the case of the Nordic experiences, relevantcases would be the larger Nordic programs in theSADDC region during the last 20 years or so. Thiscould include the NORSAD fund, support to some ofthe regional secretariats like the Transport Commis-sion in Maputo, as well as the country experiences ofNamibia and Tanzania.

A number of bilateral donors have looked at the expe-riences with sector support in various forms. A criticalreview of the lessons so far could also generate moreinsights into the pros and cons of various forms of aidcoordination. The recent Sida study that looked on theone hand at sector program aid and macroeconomicdevelopment and on the other at the need for a func-tioning administration and financial management sys-tems (Sjölander and Brobäck 1999) poses a series ofcritical questions that may have been overlooked sofar.

Donor practices deserve considerable more analysis inthis context. The DAC country reviews provide awealth of material that can be used, but nationalbudget documents will also reveal a lot about how dif-ferent donor governments think about their aidresources.

As of now there is not a strategy in place to develop a“least cost” program for aid coordination improve-ments. Given all the material that is potentially availa-ble, some more work in this area may yield highreturns in terms of more impact from the very consid-erable resources that today are being invested in aidcoordination both by donors and recipients.

7.3 The CDF Many of the issues and questions raised above willundoubtedly become even more urgent as the discus-sions on the Comprehensive Development Framework,CDF, proceed. The CDF may in particular clarify theextent to which more intensive aid coordination atnational and sectoral levels poses downside risks andcosts as well as benefits. This may set the stage formore careful empirical work regarding aid coordina-tion, since one of the conclusions from this review isthat we perhaps know a lot but in fact understand verylittle about aid coordination: much has been said aboutcoordination activities but little seems to be knownabout systemic outcomes, and how better to achievethem.

38 ANNEX I: MOZAMBIQUE EXPERIENCES

Annex I: Mozambique Experiences

(Modified excerpts from ECON [1998]: “Donor Coor-dination in Mozambique: Background Report”. ECONMemorandum 45/98, August. Oslo)

1.1 Donor CoordinationThere was clear consensus among the donors thateffective coordination among all actors would ensurebetter outcomes. This is being pursued in a number ofways, from the import/budget support quarterly meet-ings chaired by the Banco de Mocambique, throughsector working groups, to internal donor coordinationmeetings. Within the donor community, the Develop-ment Partners Group covers external financing agen-cies - bilateral and multilateral - and major UNexecuting agencies. The UN and EU systems haveinternal coordination mechanisms, while the so-calledLike-minded Group also convenes coordination meet-ings. A large number of working groups have been setup, which in turn generate countless meetings. Coordi-nation is particularly demanding on senior decision-makers’ time, which is the scarcest resource both forthe national authorities and the donors. Since the timeavailable is limited, and as all forms of coordinationare not equally important, it is of interest to review theexperiences with coordination in terms of reducingcosts.

MeetingsThe most common coordination mechanism is meet-ings. Most are for information sharing, but the data areusually not processed and hence do not provide muchoperational assistance. The result is that much time isspent debating general matters but without clearexpectations in terms of what is to come out of theinformation shared. Most donors complained about thelarge number of meetings and limited useful out-comes. Still, most seemed not to have clear criteria orprioritization regarding participation in meetings, butattempted to attend more than they actually felt wasnecessary.

Appropriateness of CoordinationSome donors do not focus on coordination with otherdonors since their major concern is that activitiesshould be discussed with Government. Their time isthus spent ensuring that the collaboration with thenational authorities is in place. There is also the feel-ing that too much internal coordination leads to “gang-ing up”, since so much time is spent on agreeingamongst the foreign actors that the Government comesup against a donor bloc, rather than different partnerswho can bring different but complementary contribu-tions to the solution of a given problem.

This divergence in attitudes towards coordination isreflected in other ways as well. Some donors engage indialogue with local stakeholders, inviting NGOs, keyinformants in civil society and leaders of civic organi-zations to discussions about their strategy. Others feelthat this is an inappropriate intrusion into the dialoguebetween a government and its civil society, and thatdonors are trespassing onto grounds that really is nottheirs to enter. These differences are not linked to theindividuals in the various donor offices - they resultfrom quite different traditions and fundamental world-views in the home countries.

SovereigntyEffective coordination has costs in terms of loss ofsovereignty as well. It reduces the ability of differentactors - both on the donor side and Government side -to move ahead with own projects and priorities. Coor-dination requires taking into consideration legitimateconcerns of other actors, thus accommodating intereststhat may often conflict. The reluctance of some actorsto coordinate seems in part to derive from this issue aswell, in some cases due to commercial interests.

Whether the Government or the rest of the donor com-munity should accept this reason for retreating fromcoordination is a tricky issue. Pursuing narrow com-mercial concerns comes at the expense of somebodyelse. Forcing more transparency and accountability onthe side of the donors is thus potentially an importantpart of the unfinished coordination agenda.

Levels of Coordination Given such basic divisions amongst the donors, twodifferent approaches can be considered. One is to finda program of coordination that all actors can agree to.This would be a “least common denominator” model,where to some extent the agenda is constrained bythose who wish the lowest level of coordination. Theother is that groups that want more coordination andfeel hamstrung by this approach can have a moreintensive dialogue amongst themselves.

Both models are in place in Mozambique today. Thelarge information sharing meetings with perhaps fewoperational conclusions and focus are acceptable forafor all participants. The Like-minded Group is estab-lished to develop closer operational collaboration.

1.1.1 Lessons from Macro Coordination

Import Support ProgramsAs Mozambique began restructuring its economy inthe second half of the 1980s, liberalization of the trade

ANNEX I: MOZAMBIQUE EXPERIENCES 39

regime was a key feature. Many donors funded importsupport programs, all of which had slightly differentprocedures. For importers, accessing these funds wasoften a nightmare. The time delays that the varioussteps led to meant that the typical importation proce-dure would take six to nine months. Oftentimes, animporter would have to make renewed requests for theforeign exchange, sometimes from a different donorwith a slightly different set of regulations, because thevalidity of the supplier’s offer had expired, and theimporter then had to start the process all over again.The transaction costs in terms of waiting, uncertainty,fees and complying with different donors’ reportingrequirements made real costs of importing skyrocket.In the early days of the import support programs, thesecosts were still worth accepting since the local cur-rency was so over-valued that any foreign exchangerepresented a windfall gain. As the exchange ratebecame more realistic, however, these access costsbecame a real impediment to the growth of a moreefficient private sector.

To harmonize the import support programs, the Gabi-nete de Coordenacão dos Programas de Importacão,GCPI, was established in the then-Ministry of Trade in1987. The GCPI received considerable technicalassistance from a range of donors: Sweden assisted onthe importation side, the UK with setting up internalmanagement systems, etc. In 1991, the World Bank,with the participation of four bilateral donors organ-ized a first Joint Evaluation Mission, JEM. Two yearslater the next JEM took place, this time with almosttwice the number of donors on-board.

The JEMs were costly to organize, but in the endturned out to be well worth the effort. Donors whowere not fully aware of the issues that GCPI andMozambique were facing, got a much better under-standing. General policy issues could be raised, and agenuine discussion regarding the various options wereput on the table and acceptable solutions negotiated.The outcome was a streamlining of procedures aroundstandards that had been used in other countries. Issueslike lower boundaries for when International Competi-tive Bidding and Limited Competitive Bidding had tobe used were clarified, and as practices and reportingimproved, these floor levels were increased. Five yearslater, these standards are still largely in place, andthere are few discussions regarding the content ofthese procedures.

The coordination in this area thus covered policies,institutional development (GCPI), technical assist-ance, and joint monitoring - quite an impressive pack-age.

One lesson is that as the Mozambican economy devel-ops, simpler and more direct ways of doing businessare becoming possible. The transaction costs to privatesector importers of getting their goods into the countryare much lower today than they were ten years ago,because so many of the procedures have been simpli-fied and standardized. The administrative costs to boththe Mozambican authorities and donors have alsofallen to a fraction of what they used to be.

Another lesson is that the JEMs were allowed to die assoon as it was felt that the benefits did not justify theconsiderable expenses. A key cost was the prepara-tions and coordination efforts, where the World Bankas lead donor spent considerable time putting togetherthe team, ensuring that background information wasavailable to all, etc. As with all such exercises, someaspects were not seen as satisfactory by all the partici-pants, so for some the up-front planning should per-haps have been better. But this points to the need forhaving a budget that allows one to have some sense ofhow much a task costs. Without this it is difficult toknow if an intervention makes sense or not.

Another reason the JEMs were discontinued was thatthe Government took a more aggressive role in manag-ing these funds. It established an Import SupportWorking Group, which is still functioning, so the largeexternal missions were substituted by more continuouson-site monitoring by the local donor offices.

Finally, the share of BoP resources that go to importsupport programs is falling in favor of more un-tiedforms. The need for policy assessment and monitoringhas shifted to the “new frontier” of budget support andthe Debt Alleviation Fund.

Balance of Payments SupportThe GCPI experience lies behind the more structureddialogue that is taking place regarding various formsof balance of payments support at the macro level. TheBanco de Mocambique chairs quarterly meetingswhere written reports on the opening and closing bal-ances plus the use of funds during the period in ques-tion of the different donor contributions are discussed.The MPF provides reporting on contributions to theDebt Alleviation Fund, DAF.

The structure of these two sets of reports, the periodic-ity, level of detail, etc., are different. This points to theneed for some internal coordination on the governmentside, since for the donors and from a macro-economicperspective contributions to the DAF and budget sup-port are nearly perfect substitutes. The importantpoint, however, is that there is an established forumwhich convenes at predictable intervals where bothcontent and process issues can be discussed. Financial


documentation is prepared and distributed beforehand,so the possibilities for a genuine dialogue have beenput in place by the Government.

One important lesson is how some risk-taking by theSwiss helped develop the budget support mechanismthat is largely in place today. The Swiss provided freeforeign exchange to the Central Bank, which then soldit to the commercial banks at the prevailing exchangerate. The countervalue funds were allocated to the gen-eral budget with no earmarking, but with general con-ditionalities regarding how the budget structure was todevelop. This mechanism has been refined by the Cen-tral Bank, and is the one in use today.

By taking a calculated risk - essentially trusting theGovernment to deliver on some up-front promiseswithout having any firm guarantees that things wouldwork out the way both parties hoped - a new and moreefficient form of donor support was established. Thiswas achieved not through a long drawn-out negotiationprocess where all the pieces had to be put in place first,but by experimenting and allowing capacity andanswers to develop as the challenges appeared. It alsorevealed a lot of trust in Government. The authoritiesresponded to this at the political level by payingenough attention to ensure that it would be a success.Those who were involved at the technical/operationallevels felt a responsibility for the new mechanism thatensured that actual implementation also went well.

The institutional development/capacity building effectof this is critical. It was left to the national authoritiesto work out how exactly the various mechanisms wereto function, so responsibility was placed squarelywhere it belonged. No outside technical experts werecalled in to do this - this was worked out among thevarious entities that had to be involved. The learningwas thus system-wide and system developing, as therewas no formal teaching to some heads of departments,but rather involved all those who had to take a part inthe successful completion of the operations. This gen-erated a lot of confidence among those involved, andhas ensured that the system in place is sustainable.

All reporting on the use of these funds is based on thedata and systems generated internally by the Mozam-bican institutions. These reports are part of the man-agement information system used to keep an eye onMozambican public finances. They have thus on theone hand strengthened Mozambican systems by beingused and hence respected by the donors (a very impor-tant signal from the donor community!), and at thesame time the additional demands that the donors puton the reports has led to quality improvements of pub-lic finance reporting. These become constructive insti-tutional development impulses that are built on using

existing procedures, and then taking them seriously byputting quality demands on them - at a pace and levelthat local capacity can handle.

Another lesson comes from a mission in May 1998 byfour bilateral donors that provide budget support. Themission noted that there were considerable efficiencygains to be had from looking at the various forms ofBoP - budget support, debt alleviation contributions,sector budget support, and import support - and estab-lishing consistent criteria for what are largely substi-tute forms of BoP support. Most of these lessons arefor the donors, because the different support programshad differing conditionalities, not because of funda-mental disagreements with goals, but for simple lackof inter-donor coordination and lack of consultingeach other’s documentation.

1.1.2 Lessons from Sector CoordinationThe sector coordination efforts exhibit quite differentcharacteristics as far as donor and government rolesare concerned, and also to what extent they are “bot-tom up” or “top down” exercises. There are well-func-tioning working groups in a number of sectors: roads,water, agriculture, health, education, and statistics.Only some of these will be referred to below.

ROCSIn the roads sector, the ROCS program, ROads andCoastal Shipping, was the first Sector Investment Pro-gram, SIP, in Mozambique. ROCS is central for the re-establishment of the national transport network, whichis needed for the re-integration of local markets into afunctioning national market. This in turn is a prerequi-site for any broad-based economic growth, whetherunderstood geographically or across sectors and socialgroups.

The elaboration of the program was donor-led, thoughthere were intensive consultations with the relevantnational and to some extent local authorities. Thisstructure was in part due to the need for putting inplace a national program that could allocate resourcesrationally across a vast territory and where the needswere virtually limitless, and where it was felt that tech-nical expertise based on international experiencewould be of major assistance. While the national trunkroad program thus was prepared at head office level,local authorities were requested to provide inputs asfar as the secondary and tertiary road networks wereconcerned. To that extent there were locally-basedinputs to the overall program.

The scaling of the program, the technology choicesmade, the way contracts were structured and awarded,and other aspects of the implementation of the pro-gram have generated debate. Over time, as capacity


has been built on the Mozambican side and as lessonshave been learned, adjustments are being made and theMozambican voice given more prominence. ButROCS is still considered by many to be donor-driven,which is problematic given its size, strategic impor-tance for Mozambique’s development, and the sustain-ability challenges it faces.

PROAGRIPROAGRI can trace its “roots” back to 1992, whenUNDP began funding a “pre-program”. This was toreview the agricultural sector activities and build upinstitutional capacity that could develop sector plansand policies. Progress was, for a number of reasons,fairly slow. At the same time, resolving the problemsthe sector is facing is critical to the possibilities ofaddressing poverty. The process was therefore acceler-ated by the donors taking a more active - to someextent perhaps aggressive - role in its development. Apre-appraisal was carried out early 1997, where anumber of weak points were identified and where theMinistry of Agriculture and Fisheries, MinAg, wasasked to address those before the full appraisal couldtake place.

The donors resolved to tackle some of the stumblingblocks arising from different donor procedures andpolicies. A workshop involving 14 key foreign part-ners - bilateral donors, multilateral credit institutionsand UN agencies - was held in Copenhagen in October1997. All 14 present agreed to work together to solve alist of 48 topics that were seen as key to the successfulimplementation of PROAGRI. This donor coordina-tion workshop was an important event, because manypractical problems in sector program implementationare the result of different agencies not being willing tofollow standardized procedures but rather continuingto use their own internal rules and regulations.

On the other hand, such events point to the problemsof “ganging up”, whereby donors agree to coordinatetheir approach and in particular their conditionalitiesregarding the discussions with the Government. In thecase of PROAGRI, there have been important differ-ences between the two parties with respect to the roleof the public sector in agriculture, the restructuring ofthe Ministry, how to address the land issue, etc. Donorrepresentatives must assure their own taxpayers, how-ever, that public funds are being put to best use. Identi-fying where “dialogue” ends and “ganging up” beginsis clearly difficult and highly subjective, but is an issuethat should be explicitly on the agenda regarding les-sons learned from sector working groups.

PROAGRI has been a broad strategic planning exer-cise where overriding problems have been raised in thecontext of setting overall priorities for the ministry.

The dialogue has thus largely been at the level of theheads of the various parties involved. Only towards theend of the process was there some involvement of civilsociety.

It has been a drawn-out process, which in somerespects has been costly. But it seems also at the end ofthe day to have delivered on an impressive range ofissues in a sector that clearly contains a lot of powerfuland conflicting interests.

Ministry of EducationIn the education sector, the Ministry of Education,MinEd, has taken a much stronger lead in defining theprogram and managing the dialogue. This has some-thing to do with the role the focal donor in the sectorhas taken on, where it has tried to keep a low profileand respond to MinEd requests and needs rather thanaccelerate the process. This has at time caused delaysand frustrations, and the technical work may not havebeen as good as if more outside technical expertise hadbeen used. On the other hand, there has been no doubtregarding MinEd’s ownership of the process and thecontents of it.

The problems faced by MinEd are in many ways sim-pler than in the agricultural sector, so it has been easierfor the ministry to take a strong position. There arefew changes foreseen to the role of the public sector,so the structural changes to the ministry are drivenmore by a need for improved efficiency than for hand-ing over important functions to the private sector orother actors. The agenda is reasonably clear withstrong support across the spectrum of actors, so thereis not the same problem of generating consensus.There is considerable continuity in the ministry’s workprogram, so resource re-allocations are less dramatic.And finally the ministry covers essentially an entireand easily defined sector, so the problems of relation-ships to other sectors and programs is less trouble-some, allowing MinEd to focus more exclusively onits own responsibilities.

Ministry of HealthThe Ministry of Health in 1992 had developed anacclaimed sector strategy for the post-war period5.Based on an analysis of the main medical challengesand probable resource availability, it mapped out aprogram for expanding its human and institutionalcapacity. With this as the point of departure, a HealthSector Recovery Program was prepared by Ministrystaff with the help of the World Bank. This formed the

5. The strategy was considered a “best practice” example by theWHO, which published it in its Macroeconomics, Health andDevelopment Series in 1994. See Abdul Razak Noormahomedand Malcolm Segall, The Public Health Sector in Mozambique:A Post-War Strategy for Rehabilitation and Sustained Develop-ment. WHO, Geneva, May 1994. 94 pp.


basis for a World Bank-funded Health Sector Invest-ment Program, SIP, where a number of bilateraldonors are providing co-financing. Parts of the strat-egy are now out of date, and the ministry has to comeup with a more relevant and cost-efficient response tothe ever-increasing demands on the sector.

In 1997, the public budget for health was around USD25 million, while donors provided an additional USD90-110 million, of which USD 27 million alone wasfor central drug imports. The need for coordination haslong been evident, and the MoH has been an early sup-porter of sector working groups. An intensive dialoguewith the donors has led to the establishment of severalworking groups addressing particular problems: drugsimports, technical assistance, etc. More importantly,the focal donor has been willing to invest resources intesting out new and innovative ways of makingresources available to the sector, assuming the risksinvolved.

This has led to the establishment of sector budget sup-port programmed at the national level but being madeavailable to provincial directorates of health for recur-rent costs that the government budget could not cover,to development of pooling agreements for technicalassistance and drug imports, and joint audits of donorresources. These activities are now organized in amore structured and focused dialogue with MoH in theform of a Sector-Wide Approach to Programming,SWAP.

A Technical Advisory Group visited Mozambique inFebruary 1998 to review the progress on developing ahealth SWAP. The group noted that the Ministry ofHealth, MoH, had put in place a number of key build-ing blocks necessary for a successful SWAP. Thedevelopment of a full-scale SWAP was seen as amedium term proposition. For the time being, the rec-ommended strategy was to follow an incrementalprocess, where new instruments are introduced and theexisting ones are consolidated. In this way, institu-tions, communication, accountability, transparencyand thus trust and credible reporting are developedalong the way. The report noted a series of lessons thathad been learned from health coordination efforts sofar:

• Coordination requires strong political commit-ment and considerable senior management timeby both donors and Government to achieve con-crete results. Since many actors are involved withmany head offices that must be consulted, the pro-cess will require patience and understanding onall sides.

• A lot of detailed work is required to understandthe procedures of each actor, identify where thedifferences lie, and find possible options that canbe acceptable to all parties.

• Any change generates resistance. Where resis-tance is encountered, it is important that manage-ment steps in, and that agreements in principle arefollowed up by solidarity in practice.

• Formal agreements are preferable because withwritten contracts the various parties cannot resortto ad hoc solutions every time an issue arises: theparties have to work things out in terms of theframework already agreed to.

• The trend should be to move from administrativecontrols and permissions prior to implementation(“red tape”), to verification and audits afterwards,with resource incentives and penalty clausesspelled out. If reporting improves and losses arereduced, aid resource levels can be increased anddonor conditions relaxed. If targets are missed,more audits and controls may be put in place andresources may be reduced or control transferredback to donors.

The concern for institutional development as perhapsthe most important spin-off of the Government-donorcollaboration is a hallmark of the health sector work,and one that should be studied more carefully. The rea-son why this has been so important, is that the donorsinvolved felt that implementation is a greater problemthan formal policy consensus. If this is a correct con-cern, one implication may be that while PROAGRI hasgenerated agreement on an impressive range of keyissues, the ability to actually carry out all the activitiesthat these agreements imply may not have beenstrengthened much by the process, and that the follow-up to the agreements may thus be lagging.

1.1.3 Trade-offs: Scope versus FocusIn-depth sector work is time intensive. This forcesdonors to choose between putting a relatively highshare of their own time into a few limited areas, or par-ticipate in a wider range of activities to get the broadpicture. There clearly is a trade-off here, and the ten-dency till recently by most donors was to participate ina wide range of sectors rather than focus on a fewmore in-depth.

The role of focal donor is particularly time demanding.Most donors are hesitant about taking on such a com-mitment because the task seems not only very time-intensive at the moment, but it appears as almost anendless responsibility. One reason is that there is oftennot a clear agenda on the table, so even the in-depth


tasks widen out and take on a whole range of dimen-sions that had not been envisaged initially. Many timesthis is unavoidable: if one wants to look into how toestablish donor pools of resources for the recurrentbudgets within a sector, and particularly ensure thatthe funds are made available at provincial level, anumber of collateral issues need to be addressed: howdoes the internal management of Provincial Depart-ments of Planning and Finance function; how canaccounting at district level be improved so that thereimbursement procedures for renewed recurrentbudget funds can be released in a timely manner;where are the key constraints on the audit side so thatobvious cases of fraud can be prosecuted and malprac-tice reduced; etc.?

Another reason donors hesitate, is that it requires con-siderable technical insights. It is not enough to have ageneral feel for the issues - one must contribute qualityand value-added to the debate. Particularly for smallerdonor offices in Maputo, this is a problem becausethey often do not have the staff to both cover the over-all program and at the same time have specific sectorskills.

The trend now seems to be, however, that more donorsare focusing their financial and therefore also theiradministrative resources on fewer sectors. The reviewof donor strategies and their financial assistance showsthat three or four sectors often take 80-90 percent of adonor’s aid. Much of this is a pragmatic administrativedecision: projects are being merged into fewer butlarger programs, since this makes both strategic sensebut also reduces organizational overhead. But it alsoreleases time to focus on policy framework issuesrather than project implementation detail, so that boththe ability and interest in participating in policy dis-cussions seems to be increasing.

1.1.4 Lessons from Implementation/Procedural Coordination

Whether donor coordination is successful or not isrevealed when words are to be turned into actions.Experience shows that if there is a conflict between aprinciple or a policy that a donor has agreed to, and aprocedure or regulation that somehow makes the fol-low-through of the policy difficult, procedure willwin! Donors are aware of this dilemma, and they aremore flexible and willing to adjust to local rules andprocedures if these are seen to provide a similar guar-antee of control and insight as their own would. Inorder for them to be able to judge this, detailed under-standing of local procedures is required, since “thedevil is in the details”. While moving support awayfrom individual projects towards sector budget supportis a major step forward that opens up for local donorcollaboration to a much greater extent than project

coordination can, the attention is moved from projectimplementation towards public financial managementproblems. In the short run, this does not make donors’life easier!

Local ResistanceThe real coordination costs, to some extent, rear theirhead when the nuts and bolts are to be put in place.The negotiations over the pooling of resources fortechnical assistance in the health field took about ayear, due to problems on the Government side as wellas time required to find solutions acceptable to thedonors involved. The streamlining of finances andintegrating donor resources into regular Governmentprocedures removed the access some managers had tothe donor resources. Now they have to go to the centerand fight for resources in competition with other man-agers within their own political system that they couldavoid previously. There were personal benefits fromthese relations as well: study tours, computers andvehicles, possibilities for consultancies, and so forth.The donor community has created a public service thatin some areas has a dual allegiance - to the manage-ment of the ministry, and to the donor community.And when the donors try to change what have so farbeen the rules of the game, there is at times strongresistance to this, which donors should not underesti-mate.

Local versus National Coordination: What is Cost-effective?Achieving coordination at the district level is often notso difficult - but the benefits in terms of savings to thesystem are also slight. The big systemic gains are gen-erated when most of the donor resources in an area arepooled and integrated into standard procedures atnational level.

Efficient Coordination: Resource PoolingResource pooling is taking place with respect to medi-cal assistance. Previously, a number of donors pro-vided various categories of medical personnel basedon requests coming both from the Ministry but oftenalso from provinces - particularly provinces where thatparticular donor might be funding other health activi-ties. The pay levels and other conditions of work werelargely determined by donor policies, the personnelsometimes were from the donor country, the donoroften did the recruiting and thus had the real say interms of the professional profile of the candidates, thedonors decided whether they wanted to continue thissupport in that post or not, etc.

The pooling means that the Government now identifieswhich positions are priority, what the qualificationsare that they are looking for, offer remuneration pack-ages that are standard and transparent to all, and have a


reasonably predictable flow of funds that permit themto scale back over time if this becomes necessary.

The efficiency gains are substantial. The average costof medical staff funded by donors has fallen. In addi-tion the total share of each donor dollar available tohire staff has gone up because the real costs of recruit-ing, monitoring etc. have fallen. Donors did a lot ofthis previously using high-cost donor administrativestaff, both in the home country and in the resident mis-sions. Now the steps required to recruit, monitor andevaluate external staff is part of the regular routines ofthe ministry, so the marginal cost to the system is low.The medical staff is hired by the Government, whichthus can assess the medical qualifications and - if needbe - get rid of staff that does not answer to expecta-tions without this creating a diplomatic problem vis-à-vis a given donor.

There have been a number of practical problems, sothe transition to a purely Mozambican administrationof the funds has not been all smooth. But the process isunderway, and it is now possible to identify where bot-tlenecks and inefficiencies remain, so that progress canbe monitored and improvements demanded. But thisalso points to the need for being willing to stick to aprogram and “walk the extra mile” in order to get theresults.

Changes: Creating Winners and LosersThe higher up in a system the changes are being intro-duced, the greater the resistance will be by those whosee that they will lose power and direct benefits. Thosewho benefit from the changes are often not in a posi-tion to fight effectively for them: they are often eitherlower down or further out in the system, so their politi-cal power is less. This is particularly true of changesthat are linked with the decentralization, and wheretherefore the major “winners” inside the system willbe provincial and district level managers while the los-ers are at ministry level. The gains will be generatedover time so it is difficult to see or predict them today,hence difficult to fight for, while the losers see theirlosses immediately. Those who lose from the changesare usually small, highly articulate, and - due to thedonor resources - often powerful groups within thesystem, and can therefore create serious impedimentsto change.

In the case of technical assistance, the donors stuck totheir guns and refused to accept “business as usual”. Inthe end, the pooling agreement was signed. In the areaof drugs imports, resistance to changes in how drugsand vaccines are to be imported - streamlining pro-curement with a central pool of donor funds - has notcome as far. This has given rise to some managers try-ing to circumvent the new policies by relying on the

good and direct contacts established with individualdonors over the years. This has led to incidents whereliterally hundreds of thousands of dollars have beenwasted in drugs procurement, because what wasrational and made sense from an individual programpoint of view was irrational and duplication for theoverall system. The reason this could happen is thatdonors were not coordinated and did not respect Gov-ernment policies. This is due to lack of knowledge ofwhat is happening overall in the sector.

Coordination versus ContractsWhile donors are in principle willing to join forces,there are often problems if a formal agreement has tobe signed. This is because if a contract is to be signed,it has to be cleared by head office legal departments.Trying to find wording that is acceptable to a numberof different donor countries is time-consuming, evenwhen everybody involved is doing their best to be pos-itive and constructive.

Avoiding these headaches by setting up informal orparallel systems is often seen as a lesser evil than try-ing to get through the home country bureaucracy, andin many cases this undoubtedly is the only solution. Itleaves a messy situation on the ground, however,where Mozambique is stuck with the costs of trying tomake many different systems compatible.

When a formal agreement can be signed, it is thereforea major step forward, as it signals among other thingsthat local procedures and regulations are taken seri-ously also by the donors. In principle it allows donorsto “sit back” and take advantage of the Mozambicanreporting system, whether it is on the financial or per-formance sides. In practice, all the weaknesses of theMozambican systems then become in part a donorproblem as well.

This, however, is probably how it should be - thatdonors and Government jointly struggle to get theexisting institutional arrangements to function better,to satisfy the quality demands that a public administra-tion has to face today. But it requires a level of under-standing of how local institutions are supposed tofunction that many donors do not currently have.

Focal DonorThis is where focal donors are of great help. It may notbe necessary for all donors in a sector to know thatlevel of detail if one or a group of donors on behalf ofthe community can delve more deeply into the myster-ies of local administration.

This requires considerable coordination built on trust,and this coordination should be spelled out more care-fully and explicitly than is the case today. Some


donors feel frustrated on the one hand by some playinga more central role in the dialogue with the authorities,and on the other realizing that they themselves do nothave the capacity to take on such a time-intensive dia-logue.

There may be a need to separate the policy and techni-cal dialogues more clearly. What is causing some fric-tion is that focal donors are perceived to “block theentrance” at the policy level, which is where a focaldonor should open doors but not monopolize the entry.At the technical level, however, focal donors can beexpected to do a lot of the detailed work required tocome up with specific proposals. These kinds ofarrangements need some thinking through, and areview of experiences so far might help to better definesuch a division of labor.

Public Financial Management: The Weak LinkThere are clearly some systemic dimensions of publicfinance management that require a qualitativeimprovement before donors are going to have greatertrust in the financial reporting that is being produced,and the public sector audit system should perhapsbecome an area of priority concern over the next years.In principle, the system has three levels, starting witheach ministry’s own internal inspectorate. The Inspec-tor-General’s office of the Ministry of Planning andFinance has a mandate to audit all public funds,whether at ministerial, provincial or district levels.Finally, the Administrative Court has the responsibilityto close the public accounts and also to prosecute sus-pected cases of fraud. All three levels are weak today,which means that the ability of the system to monitoruse of public funds, detect problems, correct them andeffectively pursue cases of mismanagement, is not sat-isfactory.

The Objective: Un-tying AidStrengthening these functions will have an importantimpact on the ability of the Government to convincedonors to un-tie more of their aid. And un-tying aid isstill an important point on the agenda. As Mozam-bique moves towards a more market-based economy,inefficiencies that exist with respect to accessing fundsbecome impediments to the structural transformationof the economy.

Summing Up:

Cost Effectiveness• There has, till recently, been limited concern

regarding what kinds of coordination is most effi-cient to the various parties. Coordination, how-ever, is expensive! There should be some costingof resources being used, to get a picture of whatthe total costs of various key activities in fact is.

• A costing exercise can assist in re-directing atten-tion from high-cost, low-impact activities to morehigh-impact areas. In particular, mechanisms thathave out-lived their cost-effective usefulnessshould be abandoned. Inefficient coordinationvehicles are being maintained for reasons of iner-tia or prestige. Perhaps a maximum of six or tenworking groups should be active at any one time,with clear agendas defined for the following 6 to24 months, with some monitoring of performanceand costs.

• There is considerable potential for administrativecost reductions once new ways of doing businessare agreed to. These are linked to capacity build-ing and institutional development on the Mozam-bican side. One criterion for successfulcoordination may be “did it lead to sustainablequality improvements within Mozambican institu-tions?”. If one cannot answer an unqualified ‘Yes’to this question, one should perhaps ask what hasin reality been attained.

• Coordination at the programming/policy settingstage is usually more cost-effective than trying tomobilize around already defined programs,because there is little ownership and understand-ing of the total program by those who did not par-ticipate in its formulation. On the other hand, jointprogramming can become costly if not carefullyplanned and managed. Good understanding of thetasks and divisions of labor between actors maylessen the coordination costs considerably.

• Donors should be more willing to take risk whennew or difficult issues come up. Many donors arerisk averse, pushing the costs of this onto Mozam-bique by holding back on potentially more effi-cient implementation modalities.

Policy versus Technical Coordination• There should be a clear distinction between coor-

dination at the policy setting and technical imple-mentation levels. Some of the friction betweendonors may stem from this. The mechanismsrequired for coordination to function at these twolevels is different. At the first level, all actors areessentially equal and have as much legitimacywith respect to their viewpoints, whereas techni-cal coordination can be more entrusted to focaldonors with particular skills or capacity in thatarea.

Resistance to Change• There is often resistance to change from the

Mozambican public service. This is in part due tofear of loss of resource control and power by indi-


vidual managers. It may be important to identify“champions of change” within the individualinstitutions, and show the entire agency that per-formance is rewarded. This requires that donorsare able to respond flexibly to superior perfor-mance in sectors, and may be one argument forstep-by-step implementation rather than tryingwholesale change. If the latter approach succeeds,gains are spectacular - if it fails, it may be costlyto get the change process moving again.

Role of Government• There is a need to hear from the Mozambican

authorities how they view the different experi-ences, and what they see as preferable modes ofcollaboration in the future.

• Sector working groups are producing integratedactivity plans and budgets, where the MPF isplaying a central and constructive role in assess-ing ceilings on financing levels.

• Sector working groups also represent dangers interms of “ganging up” by donors in areas wherethere are policy disagreements with Mozambicanauthorities. This issue requires some explicit anal-ysis, so as to agree on “rules of the game” thatallow Mozambique to have a clear voice in suchdebates.

Differences among Donors• There are important differences regarding how

some donors view coordination. These are rootedin factors that are not subject to negotiation atfield level, and thus are important to note in termsof how far coordination in fact can go.

Differences between Sector Working Groups• Different working groups have followed very dif-

ferent processes. There should be a more carefulreview of what the lessons - positive and negative- have been in each case, to ensure that there issystematic learning. One interesting comparisonis between PROAGRI and the Health SWAP,including the different points of departure: PROA-GRI being a comprehensive strategic exercise,while the Health SWAP has been based on incre-mental learning.

Coordination of Implementation Procedures• The major gains are had when real institutional

development takes place, which usually requiresformal contracts or agreements between the par-ties. This is the only way to ensure that ad hocarrangements do not appear, as these underminethe gains otherwise being realized. Such formalagreements are time-consuming and often costly

to attain. The more actors involved, the more diffi-cult, because each donor must return to its headoffice for legal clearance. One strategy is todevelop a contract between Government and a fewdonors who are willing or able to engage in rea-sonably quick dialogue and reach closure. After aperiod of implementation, other donor offices canthen point to actual practice as an argument vis-à-vis their own head offices as an argument for join-ing existing arrangements. There is then no longeran issue of negotiating everything from scratch,but of getting dispensations from the head officeto join what in practice has proven to be anacceptable alternative to home-country proce-dures.

• Public accounting and audit functions mustreceive more assistance in the period to come, tostrengthen credible public finance managementthat extends to effective pursuit of identified casesof corruption and mismanagement.

1.2 Donor StrategiesDonor assistance strategy documents vary considera-bly in size, content and form. Some donors do not havea formal strategy document, others have a one pagesummary of what the donor is doing, prepared by theembassy in Maputo. Others have documents of severalhundred pages, following detailed instructions laid outin manuals and guidelines prepared by their headoffices. Some documents contain careful reviews ofsectors or issues that are of particular relevance to theircountry’s aid policies, while others emphasize localparticipation and ownership of the stated objectivesand strategies.

1.2.1 Perceptions of Development ChallengesMost donors take as their point of departure somedescription or reference to the situation in Mozam-bique, the major development challenges the countryis facing, as well as government policies and priorities.The thoroughness of these analyses varies, with only afew donors referring to the Five-year Plan or theannual Economic and Social Plan. More use the Pov-erty Reduction Strategy and the Policy FrameworkPaper. Donors’ familiarity with these documents alsovaries considerably. Some donors were not aware ofthe existence of the Five-year Plan, and gave differentanswers as to which Mozambican planning documentshould be considered the most important.

To the extent that strategy documents include adescription or analysis of the development challenges,there is a lot of common ground, with particular focuson:


• The transition from war and emergency to peace,and the resulting implications for changes in theneeds for, and role of, development assistance inMozambique.

• The encouraging political developments, but a sit-uation which is considered fragile and therefore aneed to further build and strengthen democraticinstitutions.

• The need to improve living conditions, in particu-lar for the poorest in the population.

• The need to create a basis for economic growthand private sector development.

• The need to gradually reduce aid dependency.

Even if there is much agreement, it tends to be at ageneral level, and is not often used for the develop-ment of the specific strategy. Continued engagement ina sector seems to be more a function of having estab-lished a presence and thus have knowledge of the sec-tor rather than flowing from the strategic analysis.Some donors, however, have carried out a review,looked at the assistance program in place at the time,and have decided to aggressively change their aid pro-file to make it fall into line with their situation analysisand new strategy.

1.2.2 Donor Development ObjectivesIndependent of the written analyses or references toMozambican Government documents, most donorshave formulated objectives for their developmentcooperation program in Mozambique. The most com-mon overall objectives are

• poverty reduction;

• sustainable human development;

• peace and stability;

• democratic development/good governance;,

• improved living conditions;

• economic growth;

• creation of a market economy;

• reduced aid dependency.

This list is quite general and uncontroversial, butdonors weight the individual components differently.Very few donors have a conscious strategy withrespect to peace and stability, for example, except sup-

port to formal multi-party activities and elections.Only one donor seems to have developed an explicitprogram for working with and strengthening civil soci-ety.

Many focus on economic growth through private sec-tor development, but where the term “private sector”covers quite different concepts. In some cases itfocuses on the firm sector, while in others it clearlyincludes small-holder farmers and petty traders, so thatit becomes virtually everything that is “non public sec-tor”. In some cases, it seems like the agency in factwanted to talk about “civil society” rather than “pri-vate sector”.

Investments in human resources have different ration-ales. Some speak of “human capital formation” in animplicit capital accumulation model. Others arguemore from the point of view of the realization of indi-vidual potential, as a pre-condition for the develop-ment of democracy and participation, and that humanresources development has value in itself since educa-tion and health have important direct welfare/“con-sumption” dimensions in addition to economicinvestment ones.

Donor objectives are generally consistent with Gov-ernment ones. This is largely because the objectivespresented in Government documents do not constitutea very focused development agenda that forces donorsto avoid certain activities. It is thus not difficult to pickand choose a range of objectives that suit each donor.

This is not necessarily a bad thing, since many donorsare constrained by their head offices in terms of wherethey are to provide support. The Nordic countries mustpay attention to poverty reduction, gender and envi-ronment, the US has provided a lot of support to devel-opment of a market economy, etc. But there has been atrend towards convergence of views within the donorcommunity, to some extent sparked by the workinggroups established by the SPA program. Here donorshave over the years worked on a range of key issuesthat the African economies are facing, and have comeup with a series of pragmatic answers to these thathave been discussed with the partner countries inAfrica. Mozambique has for a number of years beenan active participant in this dialogue, and it shouldtherefore come as no surprise that there has been aconvergence also between donors and host govern-ment during the 1990s in Mozambique.

1.2.3 Donor Strategies and Sector PrioritiesAll donors have defined and to some extent explicitlyformulated sector priorities in their assistance strate-gies. These are often linked with the country analysis,but other factors also come into play. These include


historical involvement in certain areas; commercialinterests in the donor country; comparative advantagein terms of the skills and other factors they can offerMozambique as part of the total aid “package”; thedesire to focus on areas that are relatively easy to man-age in terms of following impact and how resourcesare being used. In some cases, priorities are extendedto cover geographic areas. Within the EU system, thishas gone so far that there is a “focal EU donor” foreach province that is given a particular responsibilityfor ensuring that all EU-country funded activities in agiven province are, to a greater or lesser extent, coordi-nated, with the expectation that the donor in questionwill also concentrate a significant share of ownresources in that province.

An assessment of how to achieve the greatest impacthas often led to a combination of sector and geo-graphic focus. Finland and Denmark both support thehealth sector, and chose to work in a particular prov-ince since their experience till then had been that weakcapacity at the provincial level had held back success-ful implementation of critical health programs. Swit-zerland saw the same problem but decided somewhatlater to attack it differently, by making resources avail-able at the national level for programming to provincesthat needed additional resources the most. Switzerlandwould then in addition provide technical assistance forkey activities that would strengthen management andimplementation capacities.

Different strategies have been developed at differentpoints in time, so that what was seen as useful threeyears ago is no longer seen as helpful. This reflects inpart the quickly changing situation in Mozambique,where certain systems and institutions are developing,permitting less intrusive and direct management by thedonors. Donors are therefore also willing to provide agreater share of their aid financing in more un-tiedforms, either as direct budget support at the macrolevel, or shifting funding for certain sector programsfrom project aid to sector budget support. The intensedialogue taking place in some of the sectors is contrib-uting to this shift in how donors think of their sectorsupport, as confidence is being built that certain tasksthat previously were not carried out in a satisfactorymanner can now be entrusted to national entities. Thisalso implies that there is less need to earmark such agreat share of funds to particular sectors or programs,but instead let the overall budget and programmingprocess on the Mozambican side determine more ofwhere resources ought to go.

Many donors mention health, education, rural devel-opment and agriculture. Some of these strategiestherefore are “politically correct”, but may not be fol-lowed up on the funding side. Other areas are hardly

mentioned, including key issues such as financial sec-tor reform, employment creation, public sector restruc-turing and modernization, and strengthening andreform of the legal system.

Some strategy documents are only available in thedonor country’s language, one donor stating that it wastheir understanding that they were not at liberty toshow the strategy to Mozambican authorities. Thisunderlines the fact that a key audience is home countrydecision makers - Parliamentarians and ministers. Thestrategy must show that current activities and fundingare in line with political decisions back home, toensure continued home country support.

This political task is clearly important, since continuedhigh support for Mozambique by the donor commu-nity will only be possible as long as donor head officesare satisfied that funds are going to the areas they havebeen involved in deciding upon. Changing aid profilesis thus a delicate political process where the local aidrepresentatives have the best understanding of whatwill and what will not sell “back home”.

Given that this is an important task of the strategy doc-ument, it is clear that trying to coordinate the contentof such documents is of little avail. While one canalways find some innocuous general objectives - as isoften the case in strategy documents in Mozambique -the wording is actually important in terms of the “sell-ing” job in the donor country. This is situation andcountry specific, and is not something that can benegotiated or for which there are “objective” criteriathat can decide what is an appropriate phrasing or pri-ority.

1.2.4 Donor Strategy ProcessesIn the same way as strategy objectives and contentvary considerably among the donors, so do the proc-esses leading to the formulation of country assistancestrategies. Some donors have detailed guidelines forthis purpose, and the process can take a couple ofyears. Some donors base their work on general over-views, while others carry out detailed studies of sec-tors or issues, using embassy staff, local orinternational consultants, involving headquarters, theembassy and to a varying degree the recipient govern-ment. In some cases, dialogue with the host country isimportant, where dialogue partners can be governmentofficials, or representatives from beneficiary groups,or important spokespersons for civil society. In a cou-ple of cases, the country strategy goes to the donorcountry Cabinet and is thus official policy, while inother cases it is simply cleared by the donor countryministry responsible for aid funding. This clearancecan be at the political or even higher technical level.


The cycle for strategy production differs, with somedonors following fixed cycles established by the headoffice. Others produce the documents on an “asneeded” basis: when the old strategy is seen as notbeing an adequate basis for current aid policies, theprocess of putting a new one in place is initiated.Instructions regarding what the strategy needs to con-tain, and thus the level of detail and with whom onemust work, is highly donor-specific. In only somecases do strategy documents contain resource levelswith indicative allocations across key areas.

In most cases, donor offices see the strategy documentas a useful reference tool which allows them to priori-tize resource use. For many, an important aspect of thestrategy is that it allows them to say “No” to certainproposals, because there is a formal document clearedby their head office that they can refer to.

Some donors discuss the strategy in draft form withthe Mozambican authorities. This may be simply toinform the Mozambicans of any changes, and receivetheir comments and reactions before final decisionshave been taken. As the strategy is first and foremostthe donor’s own document, there is often not a needfor any formal clearance. Some donors, however,make a point of having the strategy formally acknowl-edged by the Mozambican authorities, for example byhaving the strategy presented and discussed during anannual consultation. This is to ensure that the newstrategy is known and will be respected by the localauthorities when it comes to future project requests.

Summing Up:• The use of Mozambican Government documents

as a point of departure for donor strategies is oftenpoor, in part because some of them are not wellknown or seen to be lacking in operationality andfocus.

• Despite this, there is a high degree of convergencebetween donors, and between donors and theGovernment, on general objectives and develop-ment challenges.

• The relative consistency in terms of wordingwhen it comes to choice of priority sectors may infact hide important differences in terms of priori-ties and implicit models of how a society (oughtto) develop. These differences can only be sortedout at the level of sector policies and programs,however, when concrete choices of implementa-tion are made.

• Geographic focus and choice of sectors is theresult of a complex set of factors, including whatimplementation modalities donors believe will

yield best results. Few of these choices resultdirectly from the rigorous country situation analy-sis, but are more pragmatic, and thus in principleopen to negotiation and change.

• Many of the strategy documents have a donor-country audience as their primary target groups.In these cases, there are limited possibilities forstrategy development coordination, as improvedefficiency of commonly programmed resources isnot a primary concern. In other cases, however,there is clearly an interest in improving relevanceand focus of the strategy document, opening upfor perhaps group-wise coordination.

• Donor strategy processes vary enormously interms of structure, guidelines, breadth of partici-pation, time, cost, involvement of local authoritiesand representatives of civil society, and use ofbackground studies and analyses.

• Donors seldom use each other’s materials whenpreparing own strategies, and there is thus a lackof knowledge of what other donors are doing andhow they might be addressing similar problems.

• The learning process is thus highly internal toeach donor: they accumulate experiences fromtheir own strategy work - whether earlier ones inMozambique, or similar ones in other countries.But they do not learn much from each other,which is another area for potential coordinationand savings.

1.3 ConclusionsAs long as the Government documentation is consid-ered unsatisfactory, the donors will compensate forthis by putting what at the aggregate level are consid-erable sums into producing their own documents.While donors will always have a need for an assistancestrategy that provides the explicit links between homecountry priorities and local objectives, the costs ofproducing these will be much lower the better therecipient government documents are. Good local strat-egy documents can thus generate considerable savingson donor aid administrative budgets, which in theaggregate is also to Mozambique’s advantage.

Strengthening Government Planning InstrumentsDonors can make at least two important contributionsto improve the current situation. In the first place, thefiscal management reforms require better informationin a timely manner regarding donor resource availabil-ity and expenditures. This would make a big differenceto the quality of the resource documentation that theGovernment can produce.


The second is to support the processes required forimproving the strategic documents - both those that areproviding overarching guidelines (like “Mozambique2020”) and the ones that will create the buildingblocks (the various budget instruments). This can beboth in the form of financing, and technical inputs inthose areas where external expertise can providevalue-added.

Donor Coordination: Division of Labor?Donors are discussing possible divisions of laboramongst them as a means of achieving improved effi-ciency. Several dimensions are considered:

• Individual donors could focus on a few sectorswhere they have particular skills or insights tooffer (“comparative advantage”, “supply capac-ity”). This would be in line with the concept offocal donor at the technical level;

• Individual donors could focus on a few geo-graphic areas, to ensure symbiosis between differ-ent activities they support in a given zone(“synergies”). This is the EU concept of aregional focal donor;

• Individual donors should focus on particularforms of assistance where they have a comparativeadvantage (i.e., UN system in technical assis-tance, lending institutions in assessing risk ininfrastructure investments, etc.) (specialization).This in fact flows from existing mandates forsome of the more specialized agencies.

To some extent, these changes are necessary. If donorsare going to engage in meaningful policy dialogue,they will have to focus on a few areas and specialize,perhaps by taking on some in-depth technical or policyfocal role in the bargain.

On the other hand, some donors clearly are better incertain areas than others. Reaching agreements withthe Government and partners in the donor communityon such area concentrations can obviously generateefficiency gains both to the individual agency but alsoto the system as a whole. With fewer but betterinformed and perhaps more committed donors in asector, the coordination costs drop - though the riskexposure for Government may have increased. If a keydonor suddenly has to pull back from a commitment,for example because of aid funding cut-backs, it maytake considerable time for the Government to findalternative sources. It is therefore important that donorstrategies contain serious medium-term financingcommitments that give the Government the certaintyrequired to carry out long-range planning.

But one should distinguish between coordination atthe operational level, and at the planning and resourcesallocation level. For allocating resources, donors cannever have the same amount of information and thesame accountability to the public regarding prioritiesas the host government does. The factual basis uponwhich donors can try to plan resource allocations isthus faulty, and will always be a second-best option.

One form for coordination that raises serious concern,is the geographic. In a situation where Mozambique isdecentralizing its public administration, any move tohave different donors provide assistance to individualprovinces can have serious distortionary impact onhow the overall public administration develops. Itshould be an absolute pre-condition for this kind ofdonor coordination that the ministry responsible forpublic administration, MAE, has a strong voice andcoordination and supervision capacity in place - pre-requisites that do not seem to be in place today.

Summing Up• There is no sense in trying to coordinate donors’

strategy formulation. The strategies are often forhome audiences, and thus serve different func-tions from one country to another. It may makesense to harmonize certain components, however,such as background information, share sectorstudies, etc. Sector working groups should pro-vide information on what is available.

• Some donors may not be in a position to coordi-nate much with others, partly because they arepursuing an agenda largely defined by head office(commercial interests is a typical case), partlybecause they politically do not feel it is correct tocoordinate with other donors rather than withGovernment. This mixture of principle and oppor-tunism would make an embracing exercise poten-tially frustrating and costly, with presumably fewtangible results to show for it.

• Identifying possibilities of division of laboramongst donors may generate efficiency gains,but should be done with caution: what would bethe criteria for “assigning” certain sectors to par-ticular donors (or vice-versa)? What would be therole of Government in such discussions? SinceGovernment clearly should have an importantvoice in this, what is the purpose of organizingthis in terms of a donor grouping rather than hav-ing a more structured Government-donor dia-logue?

• The critical fact is that funds are fungible.Whether donor A finances health or roads, andwhether the money is for drugs or a cancer spe-


cialist, the end result is that it impacts on the totalresource envelope available to the public sector.What should be of concern to the donors is not ifown funds are being spent in a good way or not,but if the overall policy framework guarantees thatthe public sector as such is providing value formoney, and what the distributional outcomes ofthese expenditures are.

• The most cost-efficient coordination mechanismavailable to both donors and Mozambican authori-

ties as far as resource allocations is concerned, isone integrated Government strategy and budgetdocument. The efficiency gains would be great,not just in terms of where and why externalfinancing should be allocated, but even more as asignaling instrument to local and foreign inves-tors. As private sector development accelerates,private investments, both local and foreign, willbe much more instrumental in pushing the overalldevelopment efforts forward.

52 BIBLIOGRAPHY

BIBLIOGRAPHY

Adams, Martin E. (1989): “Aid Co-ordination inAfrica: A Review”. Development Policy Review,Vol. 7 No. 2, June, pp. 185-192.

Baehr, Peter (1997): “Problems of aid conditionality.The Netherlands and Indonesia”. Third WorldQuarterly, Vol. 18 no 2, pp. 363-376.

Bagachwa, Mboya; Godwin Mjema, Joseph Shitunduand Samuel Wangwe (1998): A Study of AidEffectiveness. The Case of Danish Aid to Tanza-nia. Economic Research Bureau, University ofDar es Salaam; Economic and Social ResearchFoundation, Dar es Salaam; Centre for Develop-ment Research, Copenhagen, Denmark.

Berg, Elliot J. (coord) (1993): Rethinking TechnicalCooperation: Reforms for Capacity Building inAfrica. UNDP and Development Alternatives,Inc, New York.

Boesen, Jannik; Poul Engberg-Pedersen and OleTherkildsen (1999): “Sector Assessment for Sec-tor Support – An Issues Paper”. Centre for Devel-opment Research, draft mimeo, January.Copenhagen.

Calhoun, Craig and Dale Whittington (1988): “WhoReally Wants Donor Co-ordination?”. Develop-ment Policy Review, Vol. 6 No. 3, September, pp295-309.

Carlsson, Jerker; Gloria Somolekae and Nicolas vande Walle (1997): Foreign Aid in Africa. Learningfrom Country Experiences. Nordic Institute forAfrica Studies, Uppsala.

Cassell, Andrew (1997): “A guide to sector-wideapproaches for health development. Concepts,issues and working arrangements”. Working doc-ument prepared with funding from WHO,Danida, DFID and EC. Mimeo, London.

Cassen, Robert and associates (1994): Does Aid Work?Second revised edition. First edition originallypublished 1986. Oxford University Press.

Clift, Charles (1988): “Aid Co-ordination: Are ThereAny Lessons to be Learnt from Kenya?”. Devel-opment Policy Review, Vol. 6 No. 2, June, pp115-137.

Danida (1999): “Sector Matrix”. DAC Working Partyon Aid Evaluation, Room Document No. 7,Mimeo, January. Copenenhagen/Paris.

Department for International Development/DFID(1998): “Emerging Lessons from Sector

Approaches in Africa”. Report of a meeting ofAfrica Directorate economists, 30 September.London.

Dixit, Avinash K. (1996): The Making of EconomicPolicy: A Transaction-Cost Politics Perspective.The MIT Press, Cambridge/Massachusetts andLondon.

ECON Centre for Economic Research (1997): “Expe-rience with Norwegian Aid Evaluations”. Report6/97. Oslo, Norway

ECON Centre for Economic Research (1998): “DonorCoordination in Mozambique”. Memorandum no.45/98. Oslo, Norway.

ECON Centre for Economic Research (1999): “Evalu-ation of the Tanzania-Norway DevelopmentCooperation 1994-1997”. Report 10/99. Oslo,Norway.

European Commission (1996): “Various Concepts andApproaches to Sector Support”. Directorate-Gen-eral VIII, Brussels.

Fjeldstad, Odd-Helge and Joseph Semboja (1999):“Local Government Taxation and Tax Adminis-tration in Tanzania”. CMI Working Papers, Chris-tian Michelsen Institute, Bergen.

Forrester, Paul and Bernard McLoughlin (1998): “Sec-tor Wide Approaches and Financial Accountabil-ity”. Report for Irish Aid, Department of ForeignAffairs. Mimeo, December. Dublin.

Forss, Kim; John Carlsen, Egil Frøyland, Taimi Sitariand Knud Vilby (1988): “Evaluation of the Effec-tiveness of Technical Assistance PersonnelFinanced by the Nordic Countries”. A study com-missioned by DANIDA, FINNIDA, NORAD andSida. March.

Helleiner, Gerald K.; Tony Killick, Nguyuru Lipumba,Benno J. Ndulu and Knud Erik Svendsen (1995):Report of The Group of Independent Advisors onDevelopment Cooperation Issues between Tanza-nia and its Aid Donors. Ministry of ForeignAffairs-Danida, June. Copenhagen, Denmark.

Jones, Stephen P. (1997): Sector Investment Programsin Africa. Issues and Experience. World BankTechnical Paper No. 374, Africa Region Series.Washington D.C., September.

Killick, Tony (1998): Aid and the Political Economy ofPolicy Change. Routledge, with Overseas Devel-opment Institute, London and New York.

BIBLIOGRAPHY 53

Krueger, Ann O. (1981): “Loans to Assist the Transi-tion to Outward-Looking Policies”, World Econ-omy, Vol 4 no. 3, September, pp.

Lindstad, Olav (1987): “Betenkning om administras-jonen av fellesnordiske prosjekter” [Somethoughts on the administration of joint Nordicprojects]. NORAGRIC, Oslo, February.

Lund, Bernt H., Göran Andersson, Lene M. P. Hansenand Jon Naustalslid (1995): “Civil ServiceReform in Tanzania. Consultancy to Prepare andmake Recommendation on a Nordic Initiative forAssistance to Civil Service Reforms, Restructur-ing and Institutional Development”. March.NORAD, Oslo, Norway.

Lund, Bernt H. and Kjeld Rimberg (1997): “Civil Ser-vice Reform in Tanzania. Consultancy to assistthe Civil Service Reform Secretariat in develop-ing a Governance, Management Developmentand Leadership Programme”. August, Kjeld Rim-berg & Co, Oslo, Norway.

Maxwell, Simon and Roger Riddell (1998): “The Con-cept of Partnership in the White Paper”, Journalof International Development, Vol. 10 no. 2(March-April).

McNeill, Desmond; Eduardo Archetti, Rie Odgaardand Nancy Frank (1997): Evaluation of the Nor-dic Africa Institute, Uppsala. Evaluation Report8.97, Norwegian Ministry of Foreign Affairs,Oslo.

Netherlands Ministry of Foreign Affairs (1996): Sec-toral Budget Support. Economic Structure andEmployment Department, Macroeconomic Anal-ysis Division. The Hague, November.

NORAD (1998): Mottakeransvar og praktisering avgiverrollen [“Recipient responsibility and theapplication of the role of the donor”]. Mimeo,Oslo, August.

Nordås, Hildegunn Kyvik; Gilbert Sekgoma and Glo-ria Somolekae (1998): Managing Good Fortune:Macroecnomic management and the role of aid inBotswana. Evaluation Report 6.98, NorwegianMinistry of Foreign Affairs, Oslo.

Norwegian Ministry of Foreign Affairs (1971): WhitePaper no. 29 (1971-72), Om enkelte hoved-spørsmål vedrørende Norges samarbeid medutviklingslandene [Some key issues regardingNorway’s collaboration with the developingcountries]. Oslo.

Norwegian Ministry of Foreign Affairs (1974): WhitePaper no. 94 (1974-75), Norges økonomiskesamkvem med utviklingslandene [Norway’s eco-

nomic relations with the developing countries].Oslo.

Norwegian Ministry of Foreign Affairs (1986): WhitePaper no. 34 (1986-87), Om hovedspørsmål inorsk utviklingshjelp. Tilleggsmelding til St. meld.nr. 36 (1984-85). [Key issues regarding Norwe-gian development assistance. Supplementaryreport to White Paper no. 36 (1984-85)]. Oslo.

Norwegian Ministry of Foreign Affairs (1991): WhitePaper no. 51 (1991-92), Om utviklingstrekk iNord-Sør forholdet og Norges samarbeid medutviklingslandene [Development trends in North-South relations and Norway’s cooperation withdeveloping countries]. Oslo.

Norwegian Ministry of Foreign Affairs (1995a): Norsksør-politikk for en verden i endring. Rapport fraNord-Sør/Bistandskommisjonen. [NorwegianSouth policies in a changing world. Report of theNorth-South/Cooperation commission]. NorgesOffentlige Utredninger, NOU 1995:5 [OfficialPublic Studies, no. 5, 1995], Oslo.

Norwegian Ministry of Foreign Affairs (1995b): WhitePaper no. 19 (1995-96), En verden i endring.Hovedtrekk i norsk politikk overfor utviklingslan-dene [A changing world. Main aspects of Norwe-gian policies with respect to the developingcountries]. Oslo.

Norwegian Ministry of Foreign Affairs (1998a):“Focus on Norwegian development cooperation.Statement to the Storting on development cooper-ation policy. Report on Norwegian developmentcooperation in 1997”. Royal Norwegian Ministryof Foreign Affairs, Oslo.

Norwegian Ministry of Foreign Affairs (1998b): “Stprp nr 1 (1998-99), for budsjetterminen 1999”[Parliamentary Proposition number 1 (1998-99),for the fiscal year 1999]. Royal Norwegian Minis-try of Foreign Affairs, Oslo.

OECD (1992): DAC Principles for Effective Aid.Development Assistance Manual. Paris.

OECD (1996): Shaping the 21st Century: The Contri-bution of Development Co-operation. Paris.

Panday, Devendra Raj and Maurice Williams (1990):“Aid Coordination and Effectiveness: LeastDeveloped Countries 1981-1989”. DocumentUNCLD.C. II/4, preparatory process for SecondUnited Nations Conference on the Least Devel-oped Countries, 8 March 1990. Geneva.

Pearson, Lester B. et. al. (1969): Partners in Develop-ment. Report of the Commission on InternationalDevelopment. Praeger, New York.

54 BIBLIOGRAPHY

Riddell, Roger (1996): Aid Dependency. In Sida(1996). Stockholm, Sweden.

Ross, David J. (1990): “Aid Co-ordination”. PublicAdministration and Development Vol. 10 No. 3July-Sep 1990, pp. 331-342.

Selbervik, Hilde (1997): Aid as a tool for promotion ofhuman rights and democracy: what can Norwaydo? Evaluation Report 7.97, Norwegian Ministryof Foreign Affairs, Oslo.

Selbervik, Hilde (1999): “Aid and Conditionality. Therole of the bilateral donor: A case study of theNorwegian-Tanzanian aid relationship”. Draftfinal report, April. Christian Michelsens Institute,Bergen.

Serageldin, Ismail (1995): Nurturing Development.Aid and Cooperation in Today’s Changing World.Direction in Development-series, World Bank,Washington D.C.

Sida (1995): Sector Programme Support. BackgroundDocument to Sida Policy. Department for Policyand Legal Issues. Stockholm, July.

Sida (1996): Project 2015. Aid Dependency – Causes,Symptoms and Remedies. Stockholm, Sweden.

Sida (1998), “Report from the seminar on practicalplanning and implications of sector-wideapproaches for education and health develop-ment”. Meeting held at Lejondal, March. MariaStridsman, DESO. Mimeo. Stockholm.

Sjölander, Stefan and Ulrika Brobäck (1999): “Sektor-programstöd och makroekonomisk utveckling.Behovet att en fungerande förvaltning ochfungerande finansiella styringssystem – ett under-lag för riktlinjer” [“Sector program aid and mac-roeconomic development. The need for afunctioning administration and functioning finan-cial management systems – background forguidelines”]. Swedish Aid and Agencies Consult-ants SAAC. Stockholm, February.

Sobhan, R. (1996): Aid Dependence and Donor Policy.In Sida (1996). Stockholm, Sweden.

Swedish Ministry of Foreign Affairs (1997): Partner-ship with Africa. Proposals for a New SwedishPolicy towards Sub-Saharan Africa. Stockholm

Swedish Ministry of Foreign Affairs (1998): Africa onthe Move. Revitalizing Swedish Policy towardsAfrica for the 21st Century. Stockholm.

United Kingdom, Secretary of State for InternationalDevelopment (1997): Eliminating World Pov-erty: A Challenge for the 21st Century. WhitePaper on International Development presented toParliament, November.

van de Walle, Nicolas, and Timothy A. Johnston(1996): Improving Aid to Africa. Policy Essay no.21, Overseas Development Council, WashingtonD.C.

World Bank (1998a): Assessing Aid. What works, whatdoesn’t, and why. Policy Research Unit, Septem-ber. Washington D.C.

World Bank (1998b): 1997 Annual Review of Develop-ment Effectiveness. Operations EvaluationDepartment, April. Washington, D.C.

World Bank (1998c): The Special Program of Assis-tance for Africa (SPA): An Independent Evalua-tion. Operations Evaluation Department, March.Washington, D.C.

World Bank (1998d): “Partnership for Development:Proposed Actions for the World Bank”. Discus-sion Paper, May. Washington, D.C.

World Bank (1999a): 1998 Annual Review of Develop-ment Effectiveness. Operations EvaluationDepartment, February. Washington, D.C.

World Bank (1999b): “Aid Coordination and the Roleof the World Bank. An OED Review (Phase I)”.Operations Evaluation Department, January,mimeo. Washington, D.C.

World Bank (1999c): “Aid Coordination in an Era ofPartnership”. Proceedings from a Workshop forGovernment Aid Coordination Officials February10-11 1999. Operations Evaluation Department,February, mimeo. Washington, D.C.

Aid Coordination and Aid Effectiveness - OECD · number of such coordination activities has grown...

Documents

Transcript of Aid Coordination and Aid Effectiveness - OECD · number of such coordination activities has grown...