EVALUATION OF PROGRAMS PROMOTING PARTICIPATORY DEVELOPMENT … · decentralization, (4) human rights, and (5) participation. The study synthesized in this report, Evaluation of Programs

EVALUATION OF PROGRAMS PROMOTINGPARTICIPATORY DEVELOPMENT AND GOOD

GOVERNANCE

SYNTHESIS REPORT

DAC EXPERT GROUP ON AID EVALUATION

1997

1

FOREWORD

The Development Assistance Committee of the Organisation for Economic Co-operation andDevelopment adopted orientations on participatory development and good governance at its High Level Meeting inDecember 1993. The preamble to those orientations described them as “a work in progress”. Such acharacterisation could also be applied to any of the other DAC principles and guidelines. However, it wasconsidered especially appropriate for guidelines concerning a field of development co-operation about which manydonors then had only limited experience.

In approving the orientations, the 1993 High Level Meeting of the DAC also approved the creation of anad hoc Working Party on Participatory Development and Good Governance. The mandate of the Working Partywas to help bring “PDGG” into the mainstream of development co-operation through a three-year program ofactivities.

At the same time, the DAC Expert Group on Aid Evaluation formed a PDGG Steering Committee. Oneof the most innovative and valuable activites of the subsequent work of the DAC in cementing the role of PDGGas an integral element in development co-operation was the collaboration between these two subsidiary bodies --the Working Party on PDGG and the Expert Group on Aid Evaluation.

The PDGG Working Party has now completed its work, leaving an impressive legacy of studies,seminars and pilot activities that will be of immeasurable value to the international community. The “Final Reportof the Ad Hoc Working Group on Participatory Development and Good Governance” is now available as aseparate report.

The Expert Group on Aid Evaluation undertook to develop a review in the following five main areas ofPDGG activitiy:

• legal systems;

• public sector management;

• decentralisation;

• human rights;

• participation.

Each of these themes corresponded to an important area of activity by the PDGG Working Party. Thus,the experience and analysis reflected in each theme paper produced by the evaluation contributed to the forward-looking work of the Working Party. For example, the evaluation paper on human rights provided a basis fordiscussion and review at a 1996 human rights seminar sponsored by the OECD Development Centre as acomplement to the PDGG Working Party’s program.

This productive collaboration between subsidiary bodies of the DAC has contributed to increasedeffectiveness by both groups. Learning from experience and planning future activites have been brought together,helping to assure that the key elements of good governance and participation will be more knowledgeably andsensibly addressed in future development co-operation efforts.

In 1989 the DAC first acknowledged the existence of “a vital connection between open, democratic andaccountable political systems, individual rights and the effective and equitable operation of economic systems”.1

Seven years later the 1996 DAC High Level Meeting reaffirmed that the “investment of development resources in

1 Policy Statement by DAC Aid Ministers and Heads of Aid Agencies on Development Co-operation in the 1990s, reprintedin the 1989 DAC Development Co-operation Report, OECD (1989).

2

democratic governance will contribute to more accountable, transparent and participatory societies conducive todevelopment progress”.2

The policy of DAC Members has remained constant over those seven years. What has changed is theircapacity to carry out effective programs and activities in the field of participatory development and goodgovernance, and the degree to which they have made such programs and activities an essential element of theirdevelopment co-operation efforts.

The present volume identifies many lessons learned which have contributed to enhanced capacities todesign and carry out sound programs. The publication of these lessons will provide a further tool for integratingparticipation and governance elements in coherent development partnerships. Like the orientations themselves, theevaluation of PDGG projects and programs remains a work in progress. But, as the following report demonstrates,much work has been done, many lessons have been learned, and a sound basis for further progress now exists.

This example of the importance of evaluation to the formulation and implementation of sound policiesand practices (which, as noted above, is also an example of the benefits of collaboration between subsidiary bodiesof the DAC) demonstrate the value that can be added by the DAC when its Members are willing to devoteadditional time and effort to joint activites that reflect their shared values and interests. All of us involved indevelopment co-operation owe a debt of gratitude to those who participated in the production of this evaluation --from the individuals and agencies who produced the theme papers and the synthesis, to the leadership of the ExpertGroup on Aid Evaluations and the leadership of the Working Party on PDGG, to the members of the OECDSecretariat who supported the entire process and helped assure the publication of this report.

James H. MichelChair, Development Assistance Committee

2 Shaping the 21st Century: The Contribution of Development Co-operation, reprinted in the 1996 DAC Development Co-operation Report, OECD (1997).

3

TABLE OF CONTENTS

SYNTHESIS REPORT OVERVIEW....................................................................................................5

STUDY PURPOSE, AUDIENCES, AND USES ...............................................................................5STUDY STRENGTHS AND LIMITATIONS....................................................................................5BACKGROUND: MILESTONES IN STUDY IMPLEMENTATION................................................6HIGHLIGHTS OF STUDY CONCLUSIONS AND LESSONS.........................................................7CROSS-CUTTING LESSONS AND ISSUES .................................................................................24

LEGAL SYSTEMS .............................................................................................................................28

INTRODUCTION............................................................................................................................28KEY CONCEPTS AND FRAMEWORKS.......................................................................................28SUMMARY OF EXPERIENCE ......................................................................................................29CONCLUSIONS AND LESSONS LEARNED................................................................................31

PRIVATIZATION ...............................................................................................................................39

INTRODUCTION............................................................................................................................39KEY CONCEPTS AND FRAMEWORKS.......................................................................................39SUMMARY OF EXPERIENCE, CONCLUSIONS AND LESSONS LEARNED............................39

INSTITUTIONAL STRENGTHENING..............................................................................................45

INTRODUCTION............................................................................................................................45KEY CONCEPTS AND FRAMEWORKS.......................................................................................45SUMMARY OF EXPERIENCE, CONCLUSIONS AND LESSONS LEARNED............................46

EVALUATION CAPACITY BUILDING ............................................................................................51

PURPOSE OF THE STUDY ...........................................................................................................51MAJOR FINDINGS OF THE STUDY ............................................................................................51CONCLUSIONS AND IMPLICATIONS........................................................................................52

DECENTRALIZATION......................................................................................................................55

INTRODUCTION............................................................................................................................55KEY CONCEPTS AND FRAMEWORKS.......................................................................................57SUMMARY OF EXPERIENCES....................................................................................................58CONCLUSIONS AND LESSONS LEARNED................................................................................66

HUMAN RIGHTS...............................................................................................................................73

INTRODUCTION............................................................................................................................73KEY CONCEPTS AND FRAMEWORKS.......................................................................................73

4

SUMMARY OF EXPERIENCE ......................................................................................................74CONCLUSIONS AND LESSONS LEARNED................................................................................82

PARTICIPATION...............................................................................................................................86

INTRODUCTION............................................................................................................................86KEY CONCEPTS AND FRAMEWORKS.......................................................................................87SUMMARY OF EXPERIENCE ......................................................................................................94CONCLUSIONS AND SUGGESTIONS FOR FUTURE WORK....................................................98

ANNEXES -REFERENCES/BIBLIOGRAPHY

5

SYNTHESIS REPORT OVERVIEW

This synthesis report represents a significant contribution toour understanding of what assistance programs andstrategies donor agencies can effectively use to promotegood governance and participatory development in theemerging democracies of the Third World. Each chapterfocuses on a specific theme area, summarizing donors'experiences and lessons learned in support of: (1) legalsystems, (2) public sector management, (3)decentralization, (4) human rights, and (5) participation.

The study synthesized in this report, Evaluation ofPrograms Promoting Participatory Development and GoodGovernance (PD/GG), was conducted by the OECD/DACExpert Group on Aid Evaluation. A PD/GG SteeringCommittee, chaired by the United States, was established in1993 to help plan and guide the work. Steering Committeemembers took the lead responsibility for preparing papersfor the different theme areas as follows:

__________________________________________THEME LEAD MEMBERS

1. Legal Systems United States

2. Public Sector United Kingdom Management

3. Decentralization Norway

4. Human Rights Netherlands

5. Participation Sweden(in collaboration with theOECD Development Center)

_______________________________________________

Final papers are now available on each of these themes.The executive summaries of these papers comprise theindividual chapters of this synthesis report. This overviewchapter outlines the study's intended purpose, audience anduses, provides some background on how the study wasimplemented, and highlights key findings and lessons.

STUDY PURPOSE, AUDIENCES, AND USES

The PD/GG study's purpose is to synthesize the experiencesof donor agencies in achieving participatory developmentand good governance objectives in the five theme areas,analyzing which intervention approaches and strategiesworked well and which did not in varying country contexts.Emphasis is on providing substantive lessons relevant forguiding donor agencies' policy and program strategydecisions and also for improving the design andimplementation of PD/GG activities. Some of the papers

also address monitoring and evaluation issuesin their theme areas.

Key audiences for the study findings andlessons are (1) the DAC and its Ad HocWorking Group on PD/GG, and (2) the seniorpolicymakers and operational managers in theDAC Member donor agencies. The study willalso be shared more widely with those in thedonor community working on PD/GG, such asother donors and developing countryorganizations, NGOs, foundations, universities,etc.

Some of the theme papers prepared as part ofthis study have already served useful purposesof stimulating discussion at several donorworkshops. For example, the legal systemspaper was discussed at a May 1994 meeting ofthe Ad Hoc Group on PD/GG on legal systemsdevelopment, and the human rights paper wasreviewed at a February 1996 seminar on humanrights sponsored by the DAC DevelopmentCenter. Moreover, the decentralization paperwas used in a seminar sponsored by the DACAd Hoc Working Group on PD/GG held in late1996. These forums provided excellentopportunities to share and disseminate thestudy results.

STUDY STRENGTHS ANDLIMITATIONS

A key strength of this study is that its findingsand lessons are based on the experiences ofmany donor agencies, and also on reviews ofthe general state-of-the-art literature. Thus, itrepresents a "collective wisdom" rather thanjust the experiences of a single donor agency.A limitation is that the theme papers are basedalmost exclusively on reviews of existingdocuments, and not on field-based evaluationsjointly sponsored by donors. Thus, some of themost recent experience of donors, if noevaluations already existed, could not becaptured. Also, the study is focused on the fivebroad theme areas that had the most existing,useful documentation; other subjects of interestto PD/GG, for example -- election support,legislative strengthening, civil/militaryrelations, etc., are not covered.

Nevertheless, the findings and lessonsemerging from this study are a majoraccomplishment in a program area of growingimportance and priority within many DACMember and observer donor organizations. For

6

the first time, lessons and best practices based on thecommon experiences of the donor agencies have beencollected, analyzed and synthesized. The emergingcollective wisdom and consensus about "what works"comprises a major contribution to the DAC's learningprocess, and may be helpful in guiding members' PD/GGprogram strategy decisions and intervention choices.

BACKGROUND: MILESTONES IN STUDYIMPLEMENTATION

This study of donor experience with participatorydevelopment and good governance had its origins at a May1992 DAC meeting on PD/GG in which it was decided torequest that the DAC Expert Group on Aid Evaluationshould take up subjects in the areas of PD/GG as part of itswork program. This idea was subsequently approved by theDAC Senior Level Meeting in June, and at the October1992 meeting of the Expert Group on Aid Evaluation(EGE), it was agreed to undertake a three-year evaluationprogram of PD/GG under the leadership of the UnitedStates. A Steering Committee, created to plan, guide, andconduct the PD/GG study, held its first meeting in March1993.

In September 1993, the United States undertook aninventory of existing and planned PD/GG-relatedevaluations, studies and research supported by the EGEmembers, in order to determine which themes or topics hadsufficient documentation on experiences to make a reviewworthwhile. Five general themes emerged from this surveyas most promising, including human rights, legal systems,decentralization, participation and public sectormanagement. (Other topics, such as elections, labor unions,civil/military relations, and media, had a smaller number ofstudies available).At the October 1993 EGE meetings, several membersagreed to take the lead responsibilities for conductingstudies on the five key themes identified in the inventory.It was also decided that approach papers would be preparedfor each of the five theme areas by the lead members. Toguide preparation of these papers, the United Statesprepared a Framework Paper (finalized in March 1994).Each approach paper was to (a) define the theme, (b)review available documents and assess the feasibility ofsynthesizing lessons learned from donor experience (To theextent possible, a substantive synthesis of experience andlessons learned was also to be attempted in this section),and (c) make recommendations/plans for future work in thetheme area. All EGE members had been requested at theOctober 1993 meetings to submit relevant documents toeach of the theme lead members as soon as possible.

At the March 1994 EGE meetings, lead members sharedtheir initial draft approach papers among themselves. Issuesarose concerning the boundaries of the five themes, withseveral areas of overlap becoming evident. A May 1994

progress report prepared by the United Statesidentified and proposed possible solutions forthese areas of overlap among the five themeareas.

Agreements clarifying the theme boundarieswere reached by the Steering Committeemembers and approved at the October 1994EGE meetings. For example, the public sectormanagement theme was subdivided into severalsub-topics. The United Kingdom agreed toprepare papers covering (1) public sectorinstitutional strengthening, and(2) privatization. Evaluation capacity building(ECB) was also identified as a special aspect ofthe public sector management theme ofimportance to the EGE. Denmark later agreedto conduct a study on this topic. Leadmembers were encouraged to co-ordinate andcollaborate as their work progressed to helpensure consistency in findings and to avoidduplication of effort. The October 1994 EGEmeetings also included discussions settingMarch 1996 as the target date for completingthe overall study effort.

Before the October 1995 EGE meetings, thelead agencies completed drafts of their themepapers and submitted them for review andcomments. Lead members were asked toconfer to ensure that between them, all relevantaspects of the themes were covered, that therespective reports were consistent in theirfindings, and no important gaps remained. EGEmembers were requested, once more, to sendany recent evaluations or other relevantdocumentation to lead members for them toincorporate into updated papers. Based onDAC Members' feedback and new materials,revisions were made during the months thatfollowed, with final papers completed inFebruary 1996.

Whereas the initial approach papers typicallyemphasized theme definitions, conceptualframeworks, and plans for further work, thefinal theme papers concentrate more onpresenting the actual substantive findings,conclusions and lessons based on analysis andsynthesis of the documents.

At the October 1995 meetings, it was decidedthat this synthesis paper be prepared by theUnited States to summarize the study findingsand lessons learned in a single volume. Itincorporates the executive summaries of thevarious theme papers and provides an overviewchapter.

7

The draft synthesis report was reviewed at the April 1996EGE meetings. A decision was made to hold a jointworkshop between the EGE and the Ad Hoc WorkingGroup on PD/GG on October 16, 1996 to enable membersto discuss the report's findings and lessons more fully. Inpreparation for the workshop, the U.S. finalized revisions tothe synthesis report and adding some cross-cutting lessonsand issues, during July 1996. Following the completion ofthe privatization study, some further final revisions weremade early in 1997.

HIGHLIGHTS OF STUDY CONCLUSIONS ANDLESSONS

This section introduces each of the theme papers,highlighting some of the key conclusions regarding theperformance (e.g. effectiveness, impact and sustainability)of commonly used intervention approaches, and drawingfrom this experience lessons and best practices for thefuture.

1. LEGAL SYSTEMS

The theme paper on the development of legal systems isbased on a United States Agency for InternationalDevelopment, Center for Development Information andEvaluation (USAID/CDIE) assessment, Weighing in on theScales of Justice: Strategic Approaches for Donor-Supported Rule of Law Programs (February 1994). Itspurposes are to assess recent donor experience in rule oflaw (ROL); to develop criteria for initiating ROL programs;and to propose a strategic framework for setting ROLpriorities and designing country programs. The assessmentapproach included fieldwork in six countries reviewingROL activities and strategies of USAID and the Asia andFord Foundations.

The paper suggests that donors consider a number ofcriteria for determining whether or not to initiate ROLprograms in a particular country context. Unless minimallyfavorable preconditions exist, investment in ROL may notbe warranted. Criteria include: sufficient support forreform among the political elite, existence of reformistconstituencies, level of judicial independence, level ofjudicial probity, freedom of speech, and donor leverage andinfluence.

An analytical framework is then presented for guidingdonors' ROL strategic choices, once the decision to initiateROL activities is made. Donor-supported ROL activitiesare divided into four main strategies: (1) constituency andcoalition building, (2) structural reform, (3) access creation,and (4) legal system strengthening. The frameworksuggests these strategies be implemented sequentially,although not necessarily rigidly or mechanically. Theframework is intended as a tool to help donors determine

which of the four strategies should dominateunder what country conditions. For example, itshould help avoid commonplace errors made inthe past, such as making large investments instrengthening formal legal systems when thepolitical will to reform is absent.

Conclusions and Lessons

Some of the key conclusions about theperformance of ROL strategies and lessonslearned include:

Preconditions for ROL Investment

In some countries, minimal preconditionsneeded for ROL programs to be successful (e.g.sufficient levels of political support, reformistconstituencies, etc.) do not exist. ROL effortsare not appropriate everywhere.

8

Constituency and Coalition Building

Donors need to devote more attention to constituency andcoalition building strategies that mobilize indigenoussupport for desired ROL reforms (e.g. increasingaccessibility, fairness, transparency). Until recently theyhave played only a minor role in donor ROL efforts,although returns on this investment can be large.

Where political leadership support for ROL reform is notsufficiently strong, an initial strategy of constituency andcoalition building may be needed, before other strategiesare introduced.

Constituencies vary in their potential for being sources ofsupport for ROL reform. For example, bar associations arerarely sources for reform initiatives, whereas thecommercial sector and NGO-based coalitions can beimportant forces for reform.

A free and effective media is needed to successfullyimplement this strategy. Only then can public debate beimproved and people be mobilized to hold the legal systemaccountable.

Reliable statistics on court management are needed toinform public debate on the justice system's inner workings.

Opinion surveys are invaluable for assessing public demandfor judicial reform.

Structural Reform

Structural reform is the boldest and most difficult ROLstrategy to undertake, because it seeks to alter infundamental ways the basic rules governing the judicialsystem.

The impact of structural reform is often diluted by theabsence of pressures for accountability and enforcement.Some kind of constituency needs to keep a persistent watchto hold the government and judiciary in compliance withpromised reforms.

Introducing new structures may provide more returns thanreform of older, entrenched institutions. Alternative DisputeResolution (ADR) mechanisms, for example, have provento be very effective new modalities to replace traditionalsystems that are unresponsive, expensive, or corrupt. WithADR mechanisms, disputes are removed from the regularcourt system and channelled into other structures.Examples include: mediation boards, neighbourhoodcounselling centers, binding arbitration schemes, andcommercial arbitration bodies.

Access Creation

Rural and low-income urban populations tendto be woefully undeserved by legal services.Donor approaches to help make legal servicesmore accessible and affordable to the poorinclude legal aid, alternative dispute resolution,legal literacy campaigns, and support for legal-advocacy NGOs.

ADR mechanisms, such as mediation councils,show significant promise as a low-cost way toprovide rapid, accessible services for settlinggrievances.

Legal advocacy NGOs represent perhaps themost promising variant of the access strategies.Not only do they aggressively use the law toassist disadvantaged groups, their advocacy andlobbying activities also make them animportant constituency for reform in general.

Conventional legal aid activities are frequentlylimited in their impact and scope, due to thehigh cost of skilled legal staff time. Suchprograms are often underfunded and reach onlya small portion of the population.

Legal literacy efforts can be very extensive,reaching large numbers of people, but theirpractical value is quite limited in terms of whatcan be imparted in a few hours to semi-literatepeople. Campaigns targeted at specificconstituencies and followed-up withprofessional legal aid may be more effectivethan generic, non-targeted campaigns.

Legal Systems Strengthening

Institutional strengthening of the judicialsystem is not necessarily the best place to begina ROL program. Such strengthening efforts willalmost certainly be unproductive unless there isa sufficiently committed political leadership, asound legal structure, and reasonablywidespread access. Once these logical priorsteps have been taken, legal systemstrengthening can be very productive.

In countries where legal system strengtheningis warranted, traditional institution buildingapproaches such as provision of commodities(e.g. computers), training, and improvedmanagement systems, are often less importantthan changing long-standing organizationalprocedures, structures and subcultures.

Introducing court statistics and databasesystems involves more than just counting cases.Quantitative data needs to be complemented by

9

better understanding of the "why's" of bottlenecks, delays,and backlogs, to make the information useful.

2.a PUBLIC SECTOR MANAGEMENT:PRIVATIZATION

The Synthesis Study of the Impact of Privatization andDivestiture in Developing and Post-Communist Countries(February 1997) was prepared by the United Kingdom'sOverseas Development Administration (ODA), EvaluationDepartment. The objective of the study is to review theimpact of privatization in developing and post-communistcountries by synthesizing existing literature. From thissynthesis, conclusions are drawn about the success ofprivatization overall, the success of various differentapproaches to privatization, the importance of the policyenvironment, and other matters.


The study assesses the impacts of privatization:(a) on theenterprises themselves; (b) on government finances; (c) onsociety, especially on consumers and employees; (d) oneconomic effects, particularly on strengthening of capitalmarkets, widening of ownership of capital, on competition,and on private investment. Also, the study analyzes thevarying effectiveness of different approaches toprivatization in achieving results, and draws lessons aboutbest practices.

Enterprises

The performance of companies improved after privatizationin almost all developing country cases. Specifically,profitability of firms increased after privatization in the vastmajority of cases whereas investment, productivity andefficiency increased in most cases. Where performance didnot improve, the reason was often poor handling of theprivatization process, such as buyers being chosen withoutregard for their ability to run enterprises or even to meet thepurchase price. Another reason for post-privatizationdifficulties was continued government interference in theenterprise.

Evidence on enterprise performance in the post-communistcountries is less voluminous and more clouded by otherfactors. The main external force affecting enterpriseperformance in these countries was the extreme recessionand severe short-term economic disruptions at theintroduction of market economic relations. Nevertheless,the evidence is that privatized companies weathered thisstorm much better than companies that remained state-owned.

Government Finances

Very substantial sums have been raised for governmentsfrom privatizations, representing up to 15 per cent of total

annual revenue in some Latin Americancountries, for example. Such large sums havehelped with macroeconomic stabilization andwith repayment of state debts. Privatizationhas also enabled governments to cease payinglarge subsidies to state enterprises, therebyimproving fiscal health. Profit sharingmechanisms attached to the terms of sale in afew cases enabled governments to share in thefinancial benefits of improved postprivatization performance. In only a few casesdid a government experience modest fiscallosses from privatization, due to discontinueddividend payments from profitable stateenterprises. In cases where state ownedenterprises were performing poorly,governments obviously did much better fiscallyby selling them and getting the loss-makers offtheir books. In some cases sales proceeds wereinsufficient to pay off liabilities that thegovernment had to assume at the time ofprivatization, but at least privatizing stoppedlosses from getting bigger.

Consumers

Consumers benefited from privatization in themajority of cases. Efficiency improvements inregulated industries were passed back to theconsumer in the form of lower prices.Consumers also benefited from efficiencyimprovement in competitive industries.Release of the investment constraint enabledmany privatized enterprises to greatly increasethe availability of services, which of coursebenefited consumers who were previouslydenied service.

The impact of privatization on consumers inpost-communist countries is more mixed.Prices have increased greatly sinceliberalization, but in the past products wereoften not available at all. And such priceincreases are more attributable to theintroduction of market economics as a wholethan to privatization in particular. There issome evidence that privatized enterprises areseeking to improve product quality andintroduce new products to meet consumerdemand. However, the continued monopolypowers of many post-communist enterprises,particularly in the former Soviet countries, haslimited pressures on enterprises to meetconsumer needs. It could have been greaterhad competition been taken into greateraccount when privatization was beingimplemented.

10

Employees

Surprisingly, evidence shows that employees benefitedfrom privatization, though not in every case. Theybenefited in three ways: (1) employment levels tended toincrease; (2) wages tended to improve after privatization;and (3) many employees were able to buy shares in theenterprise being privatized, and thus benefited fromincreases in the value of the shares.

In post-communist countries, employment levels havegenerally fallen, even though wage levels have tended toincrease. However, employment levels in state enterprisesthat were not privatized have tended to fall even faster,suggesting that privatization has, relatively speaking,helped preserve employment.

Capital Market Strengthening andWidening of Capital Ownership

Privatization has done much to strengthen capital marketsand widen capital ownership. But such effects dependlargely on the methods of privatization used. Countries thatconcentrated on sales to foreign investors were unable tocapture these benefits, whereas countries that sought to putshares of privatized companies in the hands of a largenumber of citizens strengthened their capital marketsconsiderably. However, where ownership has passed to avery large number of small shareholders, owner pressure onmanagement has been more difficult to exercise. Sale ofshares to employees has been another means of wideningthe ownership of capital.Competition

Privatization has had largely beneficial effects oncompetition, although many liberalization measuresintroduced at the time of privatization theoretically couldhave been introduced without it. In practice the two gotogether. Even where liberalization was delayed to give theprivatized company a period of protection, competition wasintroduced eventually. Without privatization, it probablywould not have been.

New Investment

Privatization has been an important means for countries toattract foreign investment. In post-communist countriesprivatization accounts for a large proportion of total foreigninvestment. Some countries have failed to attract muchforeign investment into privatization, often because ofrestrictions placed on such investment. Privatization hasalso encouraged greater foreign investment indirectly, by"signalling" the government's commitment to freer markets.

This overall success of privatization efforts in recent yearssuggests that donors should continue to give it emphasis inthe future.

Critical Success Factors

However, there were important differences inthe performance of the individual privatizationcases surveyed, and an analysis of thesevariations revealed critical success factors. Keylessons for improving performance ofprivatization efforts are as follows:

Critical to long-term success is establishing aproper balance between objectives. Theobjective of raising revenue often conflictswith objectives of increasing efficiency,competition and consumer choice: forexample, when governments allow some of anenterprise's monopoly power to continue whentransferring it to the private sector, in order toextract a higher sales price. Governments toooften ignore the importance of competition forlong-term success, in their short-term efforts ofgetting a loss-making company off their books.

Trade-offs exist between the quantity andquality of privatizations. While countries withlarge amounts of assets to privatize need tomove fast, they must allow time for properpreparation and planning or results will bepoor. It is very important to improve the designof mass privatization schemes to deliver moreeffective enterprise corporate governance.Preconditions for successful privatizationsinclude: clear political commitment; thoroughplanning and preparation of an institutionalframework for implementation; andeducational work to promote understanding ofthe need for privatization.

The policy environment is a critical factorinfluencing the success or failure ofprivatization. The worse the policyenvironment, the poorer the performance.More attention needs to be directed toimproving the policy frameworksimultaneously with privatization efforts. Thisincludes introducing regulatory frameworks(such as modern corporate law, shareholderrights, competition policy, utility regulations,capital market laws, and trade policyliberalization) and financial sector reform.

2.b PUBLIC SECTORMANAGEMENT: INSTITUTIONALSTRENGTHENING

The Synthesis Study on Public SectorInstitutional Strengthening (March 1996) wasprepared by the United Kingdom's OverseasDevelopment Administration (ODA),

11

Evaluation Department. Topics include donors' experienceswith institutional strengthening of central governmentagencies, civil service reform, state owned enterprises, andlocal government, as well as with the role of technicalassistance and training. It is based on review of donoragencies' studies and evaluations on the subject, and also ofthe general literature.


Institutional Strengthening of Central GovernmentAgencies

Until recently, institutional development has not been verypopular with donors for a number of reasons. It is oftenpolitically sensitive, lacks prestige, is difficult to defineand understand, is difficult to deal with in operational terms(there are no "blueprints" for success), and does not lenditself to quantitative measurement. Moreover, donors aregenerally better at "hardware" projects.

Recently, however, there has been a growing concern withinstitutional development because research indicates thatproject sustainability is directly related to the success oftheir institution-building components. Yet donor capacitiesin this field are often deficient.

There is an emerging consensus among donors, and in theliterature, regarding which approaches to institutionalstrengthening and development are more successful andwhich are not. Some of the key lessons follow:

• Avoid simplistic, supply-sided approaches that injecttechnical assistance, training, and equipment as the solesolution. Instead, apply a demand-oriented approachthat emphasizes locating and encouraging stakeholderswho have an interest in the organization, and who placeon it performance demands, pressures and discipline.

• Evidence suggests that organizational performance isalso improved by factors such as high-specificity of thetask (i.e. clear and focused objectives); built-inperformance incentives and sanctions against poorperformance; and increased competition between andwithin organizations.

• It is important to understand the country-context. Manyof the factors influencing organizational performance issituation-specific, and lacks "blueprints" and predictablepatterns. Thus, more emphasis is needed on country-oriented analysis.

• Support networks of organizations, promote pluralismand co-operation. Involve different types oforganizations which can contribute in a complementaryway to the overall goal of an intervention.

• Recipient commitment and "ownership" is akey factor influencing the success ofinstitutional strengthening efforts.Therefore it needs to be carefully assessed.In some cases, lack of commitment may bean immutable constraint, and institutionaldevelopment should not be promoted. Inother cases, designs may need to beadjusted to reflect existing levels ofcommitment, or strategies need to beadopted for building higher commitmentthrough education, facilitating stakeholderworkshops and participation, dialogue, etc.

• Use existing institutions wherever possible,and avoid the temptation to establishparallel "project management units" if theaim is institutional development.Experience indicates that while establishingPMUs may enable faster projectimplementation than if existing bureaucraticstructures are used, they usually do notprovide a sound basis for long-terminstitutional development or forsustainability of project activities afterdonor involvement ends.

Civil Service Reform

Most donor experiences with civil servicereform (CSR) began in the 1980s in Africa,where extreme economic difficulty led to thescrutiny of oversized, inefficient public sectors.Structural adjustment programs of the late1980s included CSR components. Thesereforms sought short-term cost-containment(e.g. numbers reductions, recruitment freezes,enforced retirements, etc.) and medium-termcapacity building for cost containment and forincreases in efficiency (e.g. integration ofpayroll and budget functions, provision ofincentive structures, personnel policy andtraining, etc.). These early CSR effortsfocused on the size of the civil service,promoted retrenchment, were seen as "donor-driven” and had disappointing results. Impactson numbers of civil servants and on total wagebills were "modest" at best.

The 1990s has seen growing recipientgovernment "ownership" of the CSR processesin their countries with better results. Newerdonor approaches involve active participationof recipient country officials and otherstakeholders. Moreover, donors' capacitybuilding activities are no longer targetedexclusively on management services divisionsof the civil service departments, but are being

12

broadened to include line ministries, departments and localgovernments.

Several factors appear to have positively influenced theresults of CSR efforts to date:

Political will and commitment to reform within the topranks of the civil service is critical to successfulimplementation. Donor conditionality in major structuraladjustment programs has been an important vehicle forencouraging this.

Older approaches emphasizing "numbers" have given wayin the 1990s to more comprehensive approaches that beginwith assessments of the appropriate role of the state in theeconomy. Also, a new "governance" approach to CSRseeks the views of clients as well as civil servants in orderto get a better picture of public service deficiencies.

Reform should be seen as a process; resistance to civilservice reform is normal and should be planned for andmade more politically feasible by accompanying thereforms with compensation packages. Costs and benefits ofalternatives need to be calculated early on.

Adequate time is required for reforms to take place,especially changes in bureaucratic culture and attitudes.Training components of CSR are crucial, but too oftenignored or not followed-up to ensure impact.

Experience indicates that donor advocacy and backing of"home-grown" CSR initiatives can keep them moving.

The ultimate aims of the CSR effort -- public serviceimprovement at an affordable cost -- need to be kept inview, and progress monitored. Then these specific gainscan be used to sell the program to the public and to defendit against critics.

Donors might improve communication and publicinformation about the aims, progress and achievements ofCSR, so that reforms become more propelled by publicpressure. Public attitudes about the civil service could beinvestigated and publicized in devising better governancestrategies in the future.

Finally, some lessons about "best practices" in designingCSR programs:

Ensure that reliable estimates of civil service numbers andrelated costs are known at an early stage.

Build capacity for reforms from the start, across the rangeof agencies involved in the CSR program.

Minimize the number and complexity of new administrativeprocedures or structures, recognizing that capacity torespond to reforms will be limited. Be clear about

priorities. Impose a strict hiring freeze;without it, much of the apparent benefits ofretrenchment on numbers and payroll willquickly evaporate.

Scrutinize proposals for measures to improvethe absorption of retrenches into the privatesector. Many of the first to go may already be"moonlighting" and thus already familiar withthe private sector. There may be scope foreasing access to credit, on commercial lines.Remember decentralization programs as ameans of improving public accountability andservice effectiveness. CSR withoutdecentralization will be of limited value.

Emphasize leadership training and supervisoryskills for staff who remain. Studies indicatemajor returns in service efficiency and moralefrom such training.

Multidisciplinary teams from diverse originsare best for conducting the "big picture"analyses of CSR programs.

Strengthening State Owned Enterprises

While State Owned Enterprises (SOEs) remainimportant in terms of GNP, gross fixed capitalformation, and employment in manydeveloping countries, they have oftenperformed poorly and have represented abudgetary burden on governments.

According to World Bank and ODA documentsreviewed, early donor efforts attempted tostrengthen individual SOEs, often as acomponent of capital/infrastructure projects inspecific sectors. The approach failed to takefull regard of the sectoral institutionaldevelopment context. Emphasis was onimproving the hardware and managerialsoftware of the organization. It did notacknowledge the many external"environmental" factors affecting performanceoutside of the control of SOE managers, suchas political interference in pricing policies,decision-making and staffing.

By the early 1990s, a new World Bankapproach to SOE support emphasizedsimulating private sector conditions, that is,maximization of profits, in a competitivemarket, under the control of managers withcapacity, authority, and motivation, faced withthe threat of bankruptcy if the concern cannotcompete. Specific aspects included:rationalizing the size and scope of the SOE

13

sector, including abolishing patently uncommercialoperations and encouraging privatization of those that arecommercially viable; influencing the market environmentby influencing prices, interest rates, input prices, and tradebarriers; and providing incentives to improve managers'calibre and skills.

Institutional reform efforts commonly introduced haveincluded: redefining clearer objectives for SOEs; holdingmanagers accountable for achieving results and providingthem with autonomy and implementation flexibility;improving selection procedures and compensation formanagers, with the key criterion competence to operate in acommercial environment; providing incentives to produceresults and sanctions for non-performance; developingperformance monitoring systems with quantitative targetsto track accountability for results; and limiting the role ofgovernment and its interface with SOEs.

The results of introducing these measures have beenbroadly positive, at least in terms of reduction in SOEnumbers. Impact on performance of SOEs is less certain,although there is some evidence of performanceimprovement, increased profits and productivity gains, andreduced budget transfers.

Review of a number of ODA case studies (of poorlyperforming SOE projects), suggest some further lessons fordesigning SOE strengthening activities:

Study the institutional context, including the overallsectoral context; the relationship between the SOE and thegovernment, the SOE and other sectoral institutions, andthe SOE and its customers; and the competence ofmanagers.

Discuss the capital and institutional aspects together.Capital funding should not be committed before aninstitutional strengthening strategy has been mutuallyagreed.

Baseline information for planning revenue collectionshould be gathered to help ensure financially viableinvestments.

Take account of and stimulate recipient initiativesregarding approaches to institutional strengthening.

Try for a mix of technical assistance (TA) inputs, includingexperienced operators as well as management consultants.

Use a process, not a blueprint approach to institutionalstrengthening; one which allows for an evolving agenda andwhich emphasizes flexibility and learning from feedback.

Training efforts need to take account of the SOE's culture,training policy, incentives and disincentives to training,personnel regulations, trainer capacities, and other factors.

Training objectives should be clear, simple andpractical. Trainees' supervisors should beinvolved in the training's design and in follow-up.

Assessments of corruption within the SOEsshould be conducted and measures for itscontrol discussed with SOE managers.

Strengthening Local Government

This section complements findings of theNorwegian theme paper on decentralizationdiscussed below.

Decentralization is popular; of 75 developingcountries, 63 claim to have started transferringpowers to local governments. A number ofdonors have supported decentralization effortsin order to make the public sector moreeffective, accessible and responsive to localneeds.

In theory, some of the possible advantages ofdecentralization include: encouraging localofficials to tailor projects and services to localconditions; motivating them to understand andreport on local conditions; better informinglocal people about plans; stimulating localparticipation and decision-making; relievingtop management of routine tasks; facilitatinglocal political education and training;encouraging local resource mobilization; andimproving accountability and legitimacy.

In reality, however, decentralization is not initself always a good thing. Some donors haveconcluded that the success of decentralizationdepends on the incentives it creates, the localcapabilities it mobilizes or stimulates, and thecosts it imposes. Thus, it is necessary toanalyze what must be administered locally, andto transfer only those powers and resourcesneeded to take over those functions. It is alsoimportant to examine the history andcharacteristics of the country situation, andassess if in fact government will become moreaccessible, responsive, and efficient ifdecentralization takes place.

The paper points to some fundamental barriersto decentralization. The "paradox of power"requires that for reforms to be successful, theremust be diffused political support andparticipation. Yet those in power, whosecommitment is needed, see this diffusion as athreat. This paradox may obstructdecentralization reforms and innovations.

14

Limitations on capacity building for decentralization arealso noted. To gain legitimacy and sustainability, localgovernments need to function effectively. Yet some recentreviews indicate that the path to effective decentralizationis fraught with difficulties. And the provision of TA toremove knowledge or skill deficiencies -- the usual supply-side approach of the past -- did not address the real natureof the problem.

The key to success is understanding the structure ofincentives facing politicians and officials at thelocal/municipal level, resulting from the relationship ofcentral and local government. Without such analysis, thereis danger that decentralization will not occur as an orderlysequence of reforms aimed at improved service delivery,but rather will be a disorderly series of concessions bycentral governments attempting to maintain politicalstability.

Some other lessons drawn from donor experiences include:functional responsibilities between the different levels ofgovernment require unambiguous clarification;corresponding revenue sources need to be authorizedaccording to these responsibilities; a system ofaccountability is needed that encompasses both regulationby central government and incentives to local constituents;service delivery requires as much attention as provision ofinfrastructure; local governments need freedom to adapt tolocal circumstances; and effort is best invested where thereis a real interest and commitment to reform.

Technical Assistance and Training

There is growing disillusionment among donors andrecipient countries with technical assistance (TA) as theprimary vehicle for capacity building, particularly in theAfrican context. There is consensus that TA is not veryeffective in achieving self-reliance in recipient countries bybuilding institutions and strengthening local capacities.Another conclusion is that TA is very expensive. Residentexpatriate TA is the most contentious form of TA, ongrounds of poor cost-effectiveness. TA is often viewed asdonor-driven. The practice of donors of hiring local staff orgiving salary supplements is destabilizing. Dependency ondonors is increased by these approaches.

Poor TA is also explained by unclear objectives and termsof references, too much emphasis on expatriate TA, poorselection of expatriate consultants, and inadequatesupervision.

Some suggestions made to improve TA include: changingthe mix from reliance on long-term expatriates to short-term advisers as coaches in intermittent visits; increasingthe number of local consultants used; pursuing long-terminstitutional "twinning" arrangements; improving localmanagement of TA; making costs of TA more transparent

and "untying" TA (so recipients do not have tohave TA in order to obtain equipment ortraining).

Training is too often invoked as the panacea tocomplement TA and develop skills.Inappropriate training, which is unrelated toeither needs or the operational context oftrainees, can have potentially wasteful,negative, or demotivating effects.

Some lessons of experience, based mainly onODA materials, include:

See training as a means to an end, and part of awider set of objectives; consider not only off-job courses but also distance-learning, coachingor action-learning techniques.

Conduct analyses of the trainees' organizationalenvironment, jobs and career prospects beforeembarking on training programs. Examinequestions of training policies and plans, plansfor re-integrating trainees, evidence thattrainees will return to relevant posts, will haveresources to implement training, etc.

Train in groups rather than individuals, as away to counteract the effects of high staffturnover and to provide trainees withopportunities to mutually reinforce each other'slearning. This implies more training needsanalysis and longer-term relationships betweendonors and organizations to implement stafftraining efforts.

More effort should go to third-country and in-country training.

2.c PUBLIC SECTORMANAGEMENT: EVALUATIONCAPACITY BUILDING

The report Issues in Aid Evaluation: EvaluationCapacity Building - Donor Support andExperiences (Copenhagen, 1996) was preparedby the Danish International DevelopmentAssistance (DANIDA), Ministry of ForeignAffairs, for the DAC Expert Group on AidEvaluation. It reviews donor experiences withevaluation capacity-building (ECB), a topic ofrenewed interest to the Expert Group emergingas part of the PD/GG evaluation. Specifically,improving capacity to monitor and evaluateprograms is viewed as an important aspect ofstrengthening public sector management andaccountability. The study is based onresponses to a survey of donor agencies' ECB

15

activities, supplementary documentation, and lessonsemerging from a full day of discussions between EGEmembers and developing country participants at theinformal seminar held in Canberra, in February 1996.


The World Bank and the regional development banks havebeen most actively supporting national evaluation systemsand stimulating demand for evaluation in the context ofpublic sector reforms and good governance initiatives. TheUnited Nations Development Program (UNDP) hascontributed to understanding of ECB by producing a seriesof country monographs on monitoring and evaluation.

ECB support provided by bilateral donors is concentrated ina few agencies, including USAID, CIDA, Sida, and theNetherlands DGIS. The DGIS is one of a very few bilateralagencies that has formulated a policy in support of ECB.Most bilateral ECB assistance is provided as a componentof a project or program containing a more comprehensivepackage of development interventions. Whereasmultilateral donors have tended to focus their ECB effortsat the national level supporting overall evaluation systems,bilateral donors have typically concentrated their ECBactivities at the department or project/program level,sometimes with the added motive of having their ownassistance activities monitored and evaluated. While somedonors limit their ECB support to training, others have alsoused joint evaluations as a means to support capacitybuilding.

Despite more than two decades of ECB support, there hasbeen limited systematic assessment of the effectiveness andsustainability of these activities. Drawing conclusions andlessons is thus difficult and preliminary.

It appears that establishing a useable evaluation function, atboth national and sub-sectoral levels, is a long-term processinvolving donor support activities including consultancies,staff training and equipment. However, host countrycommitment and a sense of ownership at the seniormanagement and policy level as well as the legalfoundation of the evaluation function is probably moreimportant for its success than the supply of donor inputs.

In summary, progress has been made in achieving ECB butmore work is needed in the areas of political advocacy,local level commitment, and commitment of donors tosupport a long-term strategy of ECB.

Key lessons learned are summarized below:

i) Approaches to strengthen evaluation functions

• sustainable evaluation institutions need politicalcommitment and support at the highest policy andmanagement levels, and should be able to

demonstrate their usefulness to theselevels. The design of evaluationsystems also needs to take intoaccount the specific government andadministrative culture in the hostcountry/organisation.

• Political commitment and seniormanagement demand should bepre-conditions for ECB supplyactivities, and have to be linked to thegovernance issue. A long-termstrategy is needed for effectiveinterventions.

• The scope of national levelperformance evaluation andperformance auditing systems aremoving closer to each other, althoughthe former is likely to be more closelyintegrated in the planning processwhile the latter system tends to focusmore on accountability at the policylevel. The choice of approach may,however, depend on other factors suchas political commitment, legalframework and institutionalcapabilities.

• Development policy and aid tend toshift from a project/program to asector/policy focus setting newdemands for host country evaluationinstitutions.

• Sustainable and effective evaluationsystems must have a legal foundationor a firm statutory organisationalregulation.

• An evaluation unit's independencefrom line management is important aswell as the security of careerpossibilities for evaluation staff andmanagers.

• Regional, sectoral and program/projectevaluations become more useful ifthey are based on a co-ordinatedapproach linked to a nationalevaluation system particularly withrespect to methodologies and dataneeds.

ii) Elements of donor support strategies

• Duration and scope of support shouldbe flexible and balanced between

16

needs for long-term relations and ownership byhost institutions.

• Consideration of support to either a national levelevaluation or a performance auditing systemshould include policy demand for its use andlegislative backing of the system.

• Efforts to institutionalise training in evaluation(including training of trainers) particularly onmethodological aspects of evaluation.

• Long-term twinning arrangements will supportprofessionalism. Increased use of the evaluationtool in developed country governments increasethe possibilities for making long-term twinningarrangements with specialised evaluationinstitutions in donor countries.

• Support to training institutions and curriculumdevelopment which on a broad base can strengthenevaluation capabilities in government and civilsociety.

iii) Role of donor evaluation units and the DACExpert Group

Areas where donor evaluation units may play an active roleinclude:

• Promote an agency ECB support policy or strategyparticularly in view of new aid forms beingintroduced including program support to institutionand capacity building as part of good governanceinitiatives at national and sectoral levels.

• Advocate and stimulate interest in evaluation incountry dialogues and sector program assistance.

• Provide technical advice to operational unitsresponsible for ECB support activities.

• Support the establishment of twinningarrangements between other domestic evaluationinstitutions and host country institutions.

• Arrange joint-evaluations with a true participatoryapproach where the needs of both parties areincorporated from the start, and where the capacitybuilding element is considered specifically.

• Co-ordinate its evaluation program with hostcountries and other donors in order to optimise useof resources and constrained capacity of recipientcountries' evaluation systems.

• Assist in securing consistentevaluation methodologies andterminologies in the ECB supportactivities of the agency.

• Advise on training facilities andmaterials on evaluation.

The Expert Group can:

• Continue to promote commonprinciples, methodology andterminology shared by donor and hostcountries.

• Encourage and facilitate members tohave sector and country evaluationsco-ordinated and shared with hostcountry evaluation institutions.

• Continue supporting regionalnetworking to promote exchange ofexpertise and share evaluationinformation among recipient anddonor countries.

• Support regional and south-southseminars and training.

• Facilitate exchange of information ondonor ECB support activities andpromote assessment of these activities.

DECENTRALIZATION

The Norwegian Ministry for Foreign Affairsprepared the Approach Paper onDecentralization (February 1996).Decentralization refers to attempts to changethe balance of power from the centralgovernment to local, regional or othersubnational levels. It can take the form oftransfers of power to govern, to tax, and to planand implement projects. Topics coveredinclude developing countries' experiences withdecentralization strategies aimed at improvingdemocracy and governance, managementefficiency in service delivery, and financialperformance. The study is based on review ofdocuments on decentralization in a selectednumber of countries, used as illustrations ofvarious models of political, administrative andfinancial decentralization.

The paper presents a framework thatdistinguishes between four major forms ofdecentralization: (1) deconcentration, (2)

17

delegation, (3) devolution, and (4) economic deregulation.

Deconcentration involves the transfer of functions withinthe central government hierarchy through shifting theworkload from central ministries to field offices.

Delegation involves transfer of responsibility forimplementing sector duties to regional or functionaldevelopment authorities, parastatals and other semi-autonomous agencies that operate independently of centralgovernment control. This usually occurs in sectors with arelatively sound income-generating basis.

Devolution involves transfer of authority to legallyconstituted local governments such as states, provinces,districts or municipalities. In devolved systems, the localgovernments have authority over a wide range of sectoraloperations, limited only by broadly defined national policyguidelines. Local level staff are responsible to locallyelected officials rather than to sector ministries.

In economic decentralization, efforts are to deregulate thecentral government's economic control and promotestrategies for private sector development, communityparticipation and private-public partnerships.


Key conclusions and lessons are summarized belowregarding the impacts of decentralization on governance, onthe delivery of services, and on public finances, and alsoabout donors' involvement in decentralization.

Governance

The study demonstrates that the contribution ofdecentralization to improving democracy and governance ispromising, especially in countries with a long history ofnation-building and a bureaucratic history. However, inmany of the least developed countries the impact ongovernance has been more limited. There has beenconsiderable ambiguity in the willingness to transferpolitical power from the central government in many ofthese countries. Even when legal powers, functions, andtasks have been allocated, adequate personnel and financialresources are often not provided. The failure to delegate ispartly rooted in the unwillingness to share power andauthority, but is also rooted in a number of weaknesses oflocal governments and problems at the sub-national level,such as low capacities, lack of administrative competence,weak planning and control systems, etc. There are alsodangers of ethnic or religious conflicts, especially if thedecentralization reforms involve control of resources,income or employment opportunities.

In many developing countries, local councils have notestablished themselves as credible institutions forarticulation of local interests. Their lack of legitimacy

results in low political activity andparticipation at the local level. A different,more promising pattern has emerged in LatinAmerica in the past decade, where pressure hasbeen generated at the local level fordecentralization and improved local politicalauthority after the fall of authoritarian regimes.

In some countries, national political leadershave used decentralization schemes to try toavoid the responsibility for the delivery ofservices by shifting blame for poorperformance to local authorities. Devolution offunctions and tasks to locally elected councilswithout resources has had a negative impact ongovernance. Again, Central and SouthAmerica show a different picture whenincreased financial resources have enabledlocal councils to deliver government servicesmore efficiently and governance has improved.

Delivery of Government Services

Two aspects of decentralization's impact onmanagement efficiency in the delivery ofgovernment services are: (1) the capacities oflocal governments, and (2) the division of rolesbetween central and local governments.

Low capacities of local governments frequentlyprevent them from implementing policies andusing available resources for their intendedpurposes. There is also often no accountabilityfor funds, due to lack of trained personnel,audit regulations and enforcement mechanisms.Local governments' budgeting and planningmodels are often inadequate.A current shift in decentralization strategiesargues that government should limit itsinvolvement to covering basic serviceprovision and establish an environment thatpromotes private economic activity andproduction. This could substantially reducemany of the problems related to inadequatecapacity at local government levels. However,there is little systematic experience withdevolving government services to the privatesector and to NGOs and communityorganizations. More systematic studies ofprivatization and public/private partnershipsare needed.

Most countries have dual systems of localgovernment; local authorities and adeconcentrated staff from the centralministries. There is often tension betweenthem; for example, lack of clarity about rolesand responsibilities; local representatives being

18

overruled by central government officials; and differencesin development priorities.

Lessons from recent analyses of decentralization stress theimportance of vertical linkages; that is, of reorganizing andreorienting ministries at the central level to better supportdecentralization. Decentralization policies and localgovernment reform should therefore always be given dueattention in civil service and public sector reform programs.

Also, when social services are devolved to local authorities,there are implications for lines of communication to thecentral ministry. A challenge for future devolution of theresponsibilities of social services to local authorities will beto work out proper arrangements of authority andcommunication between local councils and the respectiveministries, including division of responsibilities andcapacity building efforts.

Financial Performance

The importance of local revenues and incomes has beenpointed out in almost all reports on decentralization.Equally apparent is the scarcity of financial resources thatcharacterizes local government institutions in general. Thefinancial aspects of decentralization relate to both revenuesand expenditures. Many African countries lack systems orcapacity for financial planning and budgeting at the locallevel. In Central and South America, on the other hand,decentralization has improved the financial base and qualityof locally delivered services.

Effectiveness of local governments depend on their abilityto generate necessary financial and staff resources. Whereefforts to strengthen revenue collection at the local levelhave been successful, it results in significant redirection ofresources towards those areas.

Whereas some analysts reason that provision of servicesand revenue collection should be equally decentralized, thestudy argues that the decision to decentralize taxes shouldinvolve consideration of principles of fairness andefficiency. And if the decentralization of expenditures ismore desirable than decentralization of taxes, then it mustbe concluded that transfers and subsidies to localgovernments are necessary.

A key lesson is that the design of financial decentralizationshould be a major component of all decentralizationprograms. In the past, financing decentralization has oftenbeen treated rudimentarily, and not dealt with asmeticulously as the administrative aspects.

Donor Support and Involvement

The experience with donors' involvement indecentralization has been mixed. Donor support has to alarge extent been focused on administrative structures and

has not paid sufficient attention to politicalforces and processes, especially at the locallevel.

When analyzing donor assistance anddecentralization reforms, all aid, not just thesmall proportion of aid going to localgovernment strengthening, has to be taken intoaccount. Aid assistance and sectoral supporthas gone to strengthen central governmentinstitutions. Because of this, centralgovernment institutions have become lessdependent on local groups and structures, andtherefore less interested in establishing adialogue and mutual co-operation with them. Itis therefore important to relate decentralizationreforms to general public sector reforms,including ministerial reforms.

The paper also addresses donor support for(1) institution building versus sectoralassistance, (2) assisting the central governmentin the decentralization process, and (3) the roleof technical experts and training.

On the first point, the conclusion is that donorsshould support both institution building anddirect support for sectoral programs. On theone hand, sustainable programs presupposewell functioning institutions with sufficientcapacity to take over project activities atproject end. On the other hand, capacitybuilding works best if institutions havesubstantial tasks and responsibilities. Givenlocal governments' lack of funds anddependence on donors, a continued balancebetween assistance to institutionalstrengthening and to direct support of sectors iscalled for.

Secondly, donor assistance might be consideredin some cases to assist a central government inco-ordinating and implementing adecentralization program (for example bysupporting a central co-ordination unit).

Regarding the use of technical assistance, manydonors are shifting to "twinning arrangements"-- long-term co-operative arrangementsbetween institutions in donor and recipientcountries -- instead of the traditional model ofplacing foreign experts in positions indeveloping country institutions. Although"twinning", in principle, has many advantages,it also has potential and identified problems ofcapacity and relevance. A final lesson is thatto improve institutional performance, lessfunding should go to traditional interventions

19

such as vehicles, equipment, training and study tours.Instead, more efforts should go to demand-driven strategiesand to building political and organizational cultures whichare more conducive to accountability and transparency.

4. HUMAN RIGHTS

The paper, Donor Experience in Support of Human Rights:Some Lessons Learned (April 1996), was prepared by theOperations Review Unit of the Netherlands Ministry ofForeign Affairs. It synthesizes experiences and lessonslearned concerning human rights issues in three areas:(1) donor strategies supporting the implementation ofhuman rights, (2) the management of those activities, and(3) evaluation and monitoring of human rightsinterventions. Sources of information include donors'evaluations, research papers, and various human rightspublications.

In the past, theories related aid only indirectly todemocracy and human rights, through a chain linking aid toeconomic growth which led to prosperity which led todemocracy. During the 1970s this was modified, with the"basic human needs" strategy considered as identical toachieving economic and social rights. Civil and politicalrights were generally not pursued by donor agencies duringthe 1970s and 1980s. Since the end of the 1980s,development theories began to emphasize the need forsimultaneous support for human and democratic rightsalong with economic development.


Donor Strategies Supporting the Implementationof Human Rights

Three main ways are identified in which donors and NGOssupport recipient governments and indigenous NGOs: bytechnical and financial assistance, by persuasion, and bypressure.Technical and financial assistance can help promotehuman rights by (a) integrating human rights issues intomainstream development projects, (b) by projects andprograms which specifically support human rights, and (c)by orienting overall development programs towards reform.

Persuasion involves means such as internationalconferences, policy dialogue, informal meetings, jointresearch and public debates. Pressure can be brought tobear through public declarations to withdraw personnel,imposing sanctions, or conditionality .

Integrating Human Rights into MainstreamDevelopment Programs

The study concludes that the implementation of economic,social and cultural rights is not identical with either "basic

human needs" or with a "welfare state".Implementation of these rights involves notonly legislative measures but also institutionbuilding activities.

Donors are increasingly declaring protectionand promotion of human rights as a priority oftheir development programs. However, only afew donor agencies have incorporated humanrights into their operational guidelines. Donoragency staff generally do not knowinternational human rights law nor the work ofUN organizations responsible for human rightsprotection, standards and monitoring.

The principal types of interventions used insupport of the rights of vulnerable anddisadvantaged groups include: humanitarianaid, empowerment, legal enforcement, socialenforcement, and public and politicalenforcement. Experience has shown that theseapproaches work best when integrated into anoverall strategy, rather than when used inisolation. Integration may call for more inter-agency work and co-ordinated decision-making.

Human Rights Projects and Programs

The main goal of human rights projects andprograms is to strengthen respect for civil andpolitical rights. They often work incombination with efforts to support democracy,and include a broad variety of activities such asstrengthening civil society, assisting legislative,judicial and executive systems, and supportingtransition processes.

Some lessons relate to the different phases ofpolitical development. For example, humanrights support in authoritarian and semi-authoritarian systems is only possible if thereare channels or "niches" of entry. It is bestconducted either above or below the nationallevel; that is supporting global informationnetworks or local efforts, thus avoiding thegovernment-to-government route. Supportingcountries in transition to democracy mayconcentrate on supporting civil society andpolitical processes. Support for formaldemocracies generally provide aid for civilsociety and is structured to address politicalsystem reforms necessary to maintaincredibility and legitimacy.

Other lessons learned relate to support ofspecific institutions:

20

Legislative system support can help gain governmentallegitimacy by supporting elections, including informing theelectorate about their rights, election monitoring, analysisand reform of laws governing elections.

Executive branch support should help guarantee efficientdelivery of government services and promote transparency,sound management, eradication of corruption, and respectof human rights by the army and police. Support should befocused on institutions seriously committed to reform, andshould also support local NGOs in the same sector toengender constant pressure for reform of the governmentinstitutions concerned. Support for decentralization canfoster wide participation in local democratic institutions,but care must be taken to avoid control by self-servingelites.

Judicial system support has had limited impact according tothe study. Lessons for donor agencies include: undertakingjudicial system support only if the government is seriouslycommitted to reform; focusing on key problem areas in theperformance of the system; considering possibilities for"de- judicialization"; and allowing adequate time for newinstitutions (like an ombudsman for human rights) todevelop.

Depending on the country situation, support forindependent media may include: denunciating inadequatelaws, drafting new laws guaranteeing freedom of the press,or focusing on practical application of professional codes.

Support for human rights NGOs should help ensurecomplementarily and co-operation between governmentaland non-governmental organizations. When targeting anddesigning support programs, donors need to understand howNGOs function and the different circumstances in whichthey work. How they work depends in part on their ownstrategic choices; some use co-operative, juridical, and non-participatory strategies that foster reliable relationships withauthorities, while others use confrontational, non-juridicaland participatory (GRASSROOTS initiatives) strategiesthat tend to have unpredictable results.

Development Programs and Political Reform

To make political reform a constituent element in anagency's total program of aid to any given country calls foran analysis of how far that program is geared towards somecombination of human rights and democratisation. Thepurpose of such an analysis would be to re-orient thecomposition of the aid so as to support political reform.

Several strategies can be identified: (a) In countries inwhich the human-rights situation is deteriorating, budgetand commodity aid can be cancelled. 'Fragile democraticreformers' may be supported by programs in which aid isdisbursed quickly. (b) Aid can be concentrated on thosesectors of public administration which are less corrupt than

others and which show some commitment toreform. This concentration may be on a fewcentral ministries or may mean a shift fromcentral to regional or municipal government.In both cases, sectoral concentration may alsoresult in increased bargaining power and allowbetter co-ordination by the donor's program co-ordinators. (c) In countries with authoritarian,military, repressive or corrupt regimes, aid maybe channelled through NGOs, thusstrengthening civil society (so far as possible)while reducing or ending direct support for thegovernment. In countries in transition todemocracy, these strategies can be reversed.

Persuasion

Persuasion is an effort to convince agovernment that altering policies is in its ownlong-term interests. It is demanding in terms oftime, knowledge, experience and commitment.But compared to financial aid or punitivemeasures, persuasion can be very cost-effectiveand, if successful, can result in increasedcommitment and ownership of reformprograms. Surprisingly, the documentsreviewed do not provide any substantivelessons.

Conditionality

Conditionality during the 1980s related mainlyto economic structural adjustment programs.Only since the 1990s has politicalconditionality emerged, focused on promotingdemocracy, rule of law, human rights and goodgovernance. Emerging conclusions and lessonsabout political conditionality relate tonormative (legitimacy) and to instrumentalaspects.

Because of the growing importance accorded topopular decision-making, donors areincreasingly considering internal legitimacy asa factor when designing conditionalitymeasures against authoritarian, repressive, orself-serving regimes.

Experience shows a mixed record of successwhen conditionality is imposed. Perhaps themain effects of political conditionality willemerge slowly, as donors' rhetoric is followedby channelling aid allocations away fromauthoritarian and repressive regimes, and asthese governments begin to adapt to donordemands. One lesson is the need to payattention to possible linkages between politicaland economic conditionality, especially in

21

cases where both are applied by different donors ormultilateral agencies.

Donor Management of Human Rights Activities

Lessons relate to policy formulation, programming andprogram management, project design, and selectingorganizations through which aid is to be channeled.

Experiences with policy formulation vary considerably.One approach calls for clear country or regional humanrights policy statements, with clarity in goal-setting,consistency between broad goals, timeframes and capacity-building. The alternative approach argues for diffuse policystatements that allow for broad interpretation whendesigning projects in different country contexts, and thatenable field offices to maintain a low profile.

In the past, many donor agencies preferred "reactive"human rights programming approaches -- with low risk andmodest investments. Now there is a growing recognition ofthe need for long-term "pro-active" approaches to meet themajor challenges of democratization, although skepticismremains about their feasibility.

When choosing between reactive and pro-activeprogramming, donor agencies should consider their overallpolicy approach. If their policy is to support human rightsand democratic development in a given country and it isrecognized as a long-term commitment, then more pro-active programming approaches are appropriate. If instead,overall policy goals are less clear and fixed, and are basedon uncertain commitment, then reactive programming withloose funding frameworks allowing flexibility are the betterchoice.

When designing projects, a set of clear and realisticmedium-term objectives (and performance indicators)should be formulated, and grandiose goals that cannot beevaluated should be avoided. In repressive, authoritarianregimes, assistance should be channelled through localNGOs, not government institutions. Reportingrequirements need to be kept very flexible, recognizing theneed for NGOs to camouflage their actual strategies withvague objectives. Within democratic systems,combinations of support to governmental and non-governmental organizations are most effective. Standardsfor reporting can be more stringent.

Aid channelled through international NGOs benefits fromthe influence these organizations have in a large number ofcountries, their professionalism and neutrality. Localhuman rights groups may gain credence and protectionthrough their association with respected internationalorganizations.

The impact of human rights interventions can often beincreased by clustering. For example, electoral support

might include support for government'sorganization of elections and NGOs's ability tomobilize and train voters. In countries where adonor supports many interventions, clusteringmay call for, or lead to, the development of acountry strategy.

Several possible linkages exist between humanrights projects and policy dialogue. Policydialogue can be used to initiate new humanrights projects. It can also be used tolegitimize human rights activities. Problemsidentified in the projects can be followed up indialogue. The impacts of human rights projectsmay be reduced because of adverse policyconditions. Projects should be designed to takethis into account, and policy dialogue shouldinvolve other donors as well as the hostgovernment.

There is a need for co-ordination amongdifferent actors in conducting policy dialogue.This implies closer co-ordination and co-operation between the aid agency and otherdepartments of the donor government; betweendonor agencies and human rights NGOs; andamong donors.

Evaluation and Monitoring of HumanRights Interventions

The review of documents reveal that evaluatorsof human rights interventions use a broad rangeof disciplines, terminologies and techniques,and no generally accepted framework foranalysis exists. Nevertheless, three broadapproaches to evaluation are distinguished:assessing project effectiveness; understandinghuman rights issues by means of consultativeevaluations; and assessing program impactsagainst the background of different politicalsystems.

Assessing a human rights project'seffectiveness requires development of clearperformance indicators at the project level, andcollection of baseline data against which tomeasure progress. Focus should be onmanageable objectives such as impacts of theproject on direct beneficiaries and otherstakeholders, rather than trying to link it withchanges in the overall human rights situation.Donor agencies should co-ordinate with oneanother more in carrying out analyses ofhuman rights, baseline studies and parallel orjoint evaluations.

22

Another lesson is that consultative (i.e. participatory)evaluations have several advantages for evaluating humanrights activities, including paying attention to process,dynamics of change, learning and strategic orientation, andgiving voice to the claims and concerns of partners andstakeholders.

A new challenge is building a methodological frameworkfor assessing the impact of human rights programs againstthe background of different political systems. This wouldinvolve four tasks: identifying different phases of politicaldevelopment; identifying the relevant sectors in public lifeand civil society; categorizing projects supporting thesesectors; and the development of performance indicators forboth overall and sector goals. Experience with thisapproach to evaluation of human rights interventions is stillvery limited.

5. PARTICIPATION

Evaluation and Participation: An Approach Paper (March1996) was prepared by the Department of Evaluation andInternal Audit of the Swedish International DevelopmentAgency (SIDA), and benefited from collaboration with theOECD Development Center. It reviews and synthesizesdonor agencies' experience with participation, based onevaluation reports and other relevant policy, procedure, andproject documents. It focuses on two sets of issues: (1)donor agency experience with support to participation, and(2) methodologies for the evaluation of participation andfor participatory evaluation.

Participation or participatory development is fairly new asan explicit policy and program focus, although it hasorigins in small-scale "community development" projectsdating back to the 1960s and 1970s. Since the mid-1980s,participation received greater attention due to emphases onsustainability, institutional development, and policy reform.In the 1990s, issues such as poverty alleviation, genderequality, and enhancement of civil society stimulated evenmore support for advocacy groups with an empowerment orpolitical agenda.

While donors have supported projects and programsrelevant for participation for several decades, participationhas not been "operationalized" in development procedures.Thus, practice has lagged behind donor awareness, policydeclarations and rhetoric about participation. Therefore,evaluations explicitly addressing participation are only nowbeginning to emerge.

Definitions of Participation

The review demonstrates a wide variety of definitions andconceptual frameworks concerning participatorydevelopment. Some of its definitions, for example, are:

IBRD "A process by which people,especially disadvantaged people,influence decisions that affect them."

DAC "A process by which people take anactive and influential hand in shapingdecisions that affect their lives."

The various definitions of participation can beviewed along a continuum from the more far- reaching with respect to empowerment,influence and control over projects on the partof grassroots participants or other stakeholders,to the more conventional conceptions wheredonors essentially retain decision-makingpower and control with respect to key projectand planning functions.

Donors also differ in their views ofparticipation as a means or an end. Somepromote participation as a means to increaseproject effectiveness and sustainability,whereas others see it as both a means and anend in itself.

Dimensions of Participation

Different levels, dimensions, or kinds ofparticipation can be distinguished. They mayrefer: (a) to where in the project cycleparticipation occurs (planning, implementation,monitoring and evaluation, take-over); (b) tothe type of participation (i.e. participants aspassive beneficiaries, informants, cost-sharers,or stakeholders with a voice in managementdecision-making); or (c) to societal levels(local, regional, national).

An operational dimension of participationconcerns specific project functions. In manyprojects, participation takes place inconstruction, operation and maintenance tasks,but more seldom in project planning,management, control over resources anddistribution of benefits. For projects to achievelong-term sustainability, participants or theirorganizations must gradually take over all thesefunctions.

Participation and DevelopmentOrganizations

Another aspect of participation relates todifferent categories of developmentorganizations. Promoting participation is acomplex task that must be dealt with at local,intermediate and national levels. One commonform of donor assistance promoting

23

participation involves small-scale, community-specificprojects aiming at social objectives that is typicallyoperated by NGOs. NGOs are seen as working effectivelyat the grassroots level. However, the paper points out thatnot all NGOs work in a participatory manner, and that theyface constraints in terms of "scaling-up" successful projectsbeyond the community level or in influencing broaderpolicy formulation or civil institution-building. Donorsshould also work with governments, encouraging themthrough policy dialogue to become more open toparticipatory approaches. Governments are the most directactors for promoting a favorable environment forparticipation among public sector organizations thatoperate, in co-operation with NGOs and grassrootsorganizations, at regional and local levels.

Costs and Benefits of Participation

The paper reports evidence that the benefits of participationoutweigh the costs. For example, the World Bank notesthat participation can improve the quality, effectiveness andsustainability of projects, and strengthen ownership andcommitment of government and stakeholders. Systematicevaluations of the costs and benefits of participation, whilescarce, indicate that the costs, in terms of time and moneyspent, tend to be relatively higher for participatory projectsin their early phases, but that they pay-off in terms ofgreater effectiveness and sustainability in later phases.

Evaluation and Participation

A distinction is drawn between two issues: (1) evaluationof participation and (2) participation in evaluation.

The first involves evaluating development processes andoutcomes in terms of participation with respect to specificprojects. Data collection methods for evaluation ofparticipation include traditional quantitative methods(sample surveys) to "measure" quantifiable aspects ofparticipation, for example, indicators like the number ofproject beneficiaries, the frequency of project meetings,number of beneficiaries attending meetings, beneficiarycontributions in terms of labor, money or materials,distribution of benefits, etc. However, becauseparticipation is a process of complex social change,quantitative indicators give only a very incomplete pictureof participation. To understand the nature of participation,how or why it takes place, complementary qualitativeanalyses must be conducted, using methods such asparticipant observation, group discussions, key informants,and field workshops. It is proposed that four categories ofphenomena be routinely monitored using quantitative andqualitative techniques: (1) project or group activities; (2)changes in project group behavior; (3) group action andarticulation; and (4) project-group relationship.

The second issue relates to participatory techniques inevaluation. This refers to the degrees of involvement of

different categories of stakeholders (e.g.beneficiaries, project staff, grassrootsorganizations, etc.) in the evaluation process.

During the 1980s, rapid rural appraisalmethods evolved from an initial emphasis onobtaining data rapidly and cost-effectively. Itinvolved outside evaluators "listening" to avariety of beneficiaries and stakeholders. Thiswas followed by participatory rural appraisal,which reduced the role of the outside expert to"catalyst" or "facilitator" enabling involvedgroups of people to conduct their owninvestigations and analysis, as well as to planand take action.

In participatory evaluation, a variety ofbeneficiaries and other stakeholders activelytake part in determining the evaluationobjectives; in selecting procedures and datacollection methods; in analysis andinterpretation of data; and in makingrecommendations and taking action decisions.Participation by the beneficiaries and otherstakeholders in the evaluation processtransforms it into an opportunity to negotiate,to learn, and to be empowered to take action.

In practice, however, participation in donoragencies' evaluations has generally beenlimited to including a few rapid appraisaltechniques. Donor evaluations continue to bedesigned largely with donor agencymanagement and accountability considerationsin mind and conducted by outside evaluators.Despite participation rhetoric, there is verylittle evidence of genuine participation inevaluations.

The low incidence of true participatoryevaluation may reflect unease on the part ofdonor agencies with these innovativeapproaches, or the limitations imposed bydemands for accountability-for-results andobjectivity that imply use of external evaluatorsand traditional methods.Some positive trends are noted, however.Among these are increasing emphasis on theimportance of participation, especially inpolicy documents; an increase in the use ofsome participatory techniques in evaluations(e.g. focus groups, key informant interviews,workshops); and the current development bysome donors of participation guidelines,procedural notes, "tool boxes", and resourcematerials.

Recommendations

24

To strengthen these positive trends, future work is requiredat two levels: (1) participation must be mainstreamed bydonor agencies into their operations; and (2) evaluationtools need to be further developed. At a general level,mainstreaming participation into operations -- i.e. dialogue,planning, implementation, monitoring and evaluation ofdevelopment activities -- is still a major task ahead for mostdonors. At a technical and procedural level, furtherdevelopment of operational guidelines, evaluationframeworks, methods, procedures and indicators that betteraccommodate participatory processes are also needed.

CROSS-CUTTING LESSONS AND ISSUES

A number of lessons cut across the individual PD/GGtheme papers. In addition, several unresolved issues areraised repeatedly and deserve special mention.

Lessons Learned

Reform efforts require political commitment and localconstituencies supportive of change and able to holdofficials accountable. Political will is a requisite forsuccessful and sustainable PD/GG reforms. "Governmentinterference", "poor policy environment" and "lack ofpolitical commitment" are the most often cited reasons forfailure of PD/GG assistance efforts. Donors cannot forcereforms to take place, but assistance can help facilitatechanges already being sought by the host country. Wherepolitical will does not exist, donor assistance shouldprobably be limited to helping create that political will.

If these conditions do not exist, donor assistance effortsshould first emphasize constituency and coalition-building.This includes locating and encouraging stakeholders whowill exert performance demands and discipline upon publicsector institutions, and/or policy reform efforts.

Donor advocacy of "home grown" initiatives is likely to bemore successful than donor-driven reform efforts. PD/GGdevelopment needs to be an indigenous effort, not anexternally imposed condition. Host country commitmentand ownership is a key success factor. Especially in caseswhere there may be some negative impacts (e.g. civilservice reforms, privatization), participatory strategies areneeded that build indigenous support for reforms, such asstakeholder workshops, educational and informationcampaigns. Projects need proper balance among, andclear definition of their objectives. Effective PD/GGprojects and organizations have clear and attainableobjectives, performance monitoring systems to trackprogress towards those objectives, and incentive systemsthat hold managers accountable for results. Focus shouldbe on setting reasonable objectives that projectmanagement can attain, such as impacts of the project ondirect beneficiaries and other stakeholders, rather than

trying to link it with overly broad national-level changes such as the overall human rightssituation. If there is more than one objective,trade-offs may exist and should be properlybalanced (For example, in privatization efforts,the objective of raising revenues often conflictswith objectives of increasing efficiency,competition and consumer choice). Use aprocess, not a blueprint, approach that allowsfor an evolving agenda, emphasizes flexibility,and learning from experience. Most of thetheme papers discussed evidence of the meritsof flexible "process" -- in contrast to"blueprint" -- approaches in the planning,design and implementation of PD/GGassistance activities. While some lessonslearned about PD/GG efforts can begeneralized across countries, nonetheless thereis also a uniqueness for each country. PD/GGassistance should be planned and analyzed inthe specific context in which it is delivered.Moreover, progress in democratization istypically relatively unpredictable and non-linear, thus a flexible approach that allows forfrequent information feedback, learning andadjustment is desirable.

Participation and participatory approachesshould be important ingredients in all donorPD/GG assistance efforts. Changes in PD/GG,to be meaningful, require effective involvementof local citizens and groups at all stages ofdonor activities. Likewise, local ownership ofthese activities -- an outcome of suchparticipation -- facilitates success.

Most of the theme papers discuss participationand participatory approaches as factors, or evenpre-conditions, for successful and sustainabledonor efforts to build democracy and promotegood governance.

Adequate participation and consultation withclients and stakeholders in planning,implementation, monitoring and evaluation ofPD/GG assistance programs will provide aconstant feedback loop of information, enablelearning to take place, and facilitate programadjustments, innovations, and improvements.Moreover, it builds ownership and commitmentto the reform efforts that are essential forsustainability.

Traditional institution-building approaches(e.g. commodities, training, long-term residentexpatriate technical assistance) are lessimportant and effective than changingorganizational procedures, structures and

25

cultures. Transfers of commodities and building humanresource capacities do not -- by themselves -- enhanceparticipation, democratization or better governance.Instead, they need to be an integral part of a broaderassistance strategy.

In particular, more attention needs to go to demand-drivenstrategies and to building political and organizationalcultures conducive to accountability and transparency.This requires dealing not only with physical commodity andinformation/ knowledge needs, but more importantly withstructural and incentive barriers to change. PD/GGassistance is a long-term process in which a democraticculture is built, strengthened and reinforced; it is not aseries of training, technical assistance, and commoditytransfer activities.

Too much attention has gone to "administrativestructures" and not enough to "political processes" inPD/GG assistance. There is a real need to examine thepolitical nature of assistance as well as the technical nature.To stimulate evolution of a more democratic culture in agiven country context, donor analysis must look carefully atpolitical issues -- i.e. political shifts, negotiations, deals,opportunities -- and not just at technical issues andcommodity imports. Introducing new structures may, insome cases, be more successful than reforming old ones.For example, in legal systems strengthening, creatingalternative dispute resolution (ADR) mechanisms is moreeffective than trying to reform traditional systems that areunresponsive, expensive or corrupt. However, this does notmean donors should create organizations for their ownproject purposes (e.g. "project management units"), as thesetend to undermine public sector institutional strengtheningand are not sustainable.

PD/GG reform often requires appropriate assistance atboth the local and central level -- and paying attention tothe linkages. For example, there are numerous inter-relationships between civil service reform anddecentralization efforts. Decentralization programs are ameans of improving public accountability and serviceeffectiveness when supporting civil service reforms. Civilservice reform without decentralization will be of limitedvalue. Successful decentralization requires willingness totransfer political power from the central government, clarityabout what responsibilities are being delegated, andcorresponding personnel and financial resources needed tocarry out these responsibilities. Vertical linkages and openlines of communication are therefore important; centrallevel ministries need reorganizing and reorienting to bettersupport decentralization.

Donors should consider "clustering" PD/GG activities inways that are supportive of each other and of broaderstrategic objectives. In the past, many donors favored low-risk, short-term, and "reactive" approaches to PD/GG thatgave them maximum flexibility and a low profile.

Activities tended to be isolated targets ofopportunity. Now, more donor agencies areconsidering the advantages of developinglonger-term country strategies that clearly statepolicies and objectives and that pro-activelysupport integrated packages of activities. Theimpact of interventions can often be increasedby clustering mutually supportive activities intoan overall, integrated strategy. Effectiveness ofindividual activities can also be improved withbetter inter-agency/donor co-ordination in thePD/GG area.

In repressive, authoritarian regimes, however, acase can be made for maintaining flexibility,low profile, and vague or even camouflagedobjectives and strategies.

Donors should pay more attention to"mainstreaming" PD/GG approaches,especially in promoting human rights andparticipation. Donors have traditionallyemphasized separate or specific projectactivities to promote participation,democratization, or human rights. Less hasbeen done to operationalize or integrate theseconcerns more broadly into mainstreamdevelopment projects. For example, donorsshould consider preparing guidelines thatoperationalize human rights/participatoryapproaches into their overall project/programplanning, implementation, monitoring andevaluation procedures.

Donor experience with using policy dialogueto promote PD/GG is not well documented.Similarly, little information is available onexerting "conditionality". Politicalleadership's commitment and a favorablepolicy environment are often cited as keyconditions for successful PD/GG projects andreform efforts. One of the papers asserts thatpolicy dialogue or persuasion -- i.e. donorefforts to convince a government that attainingpolicies is in its long-term interest -- can be"very cost-effective and if successful can resultin increased commitment and ownership ofreform programs." Policy dialogue efforts tendto be more effective when other donors as wellas the host government are involved.

Donor conditionality in major structuraladjustment programs was found to be animportant vehicle for encouraging civil servicereforms. But donor experience with morerecent political conditionality (e.g. in humanrights) is more limited and mixed. Linkingpolitical and economic conditionality and

26

improving co-ordination among donors may increase itssuccess. Also, with the growing importance accordedpopular decision-making, donors are increasinglyconsidering internal legitimacy as a factor when designingconditionality measures.

Ultimately internal political pressures from constituenciesand local groups are probably more important forsustainability of reform efforts than external donorpressures. However, surprisingly little documentation isavailable on donor experiences with these modalities forpromoting PD/GG.A number of lessons have been learned regarding thespecial roles and advantages of involving NGOs inimplementing PD/GG activities. Effective democraciesand good governance require that citizens participate invarious associations and organizations ("civil society") thatwill advocate their interests vis-a-vis government. NGOscan be very effective in this role. For example, legaladvocacy NGOs were found to be especially successful inincreasing access of the poor to legal services while alsopromoting legal reform.

In authoritarian regimes where government-to-governmentassistance is not desirable, working directly with NGOs topromote democratization and reform is advised. Even insituations where donors work directly with governmentinstitutions to improve their effectiveness, transparency andaccountability, it still pays to support local NGOs in thesame sector to stimulate constant pressure on thegovernment for reform.

Local NGOs are typically seen as effective with small-scale, participatory project activities at a GRASSROOTSlevel, although their limited capacity often preventsreplication or "scaling-up" successful projects beyond thecommunity level.

While there may also be a role for NGOs in "devolving"formerly government responsibilities and services, littleinformation on this subject was found.

Donor monitoring and evaluation (M&E) of PD/GGactivities should consider both quantitative measures andcomplementary qualitative measures, and should takebetter advantage of participatory evaluation techniques.Effective M&E of the performance of PD/GG projectsrequires clearly stated objectives, development ofperformance measures/indicators, and regular collection ofquantitative data on progress. But there is also a need tocomplement the quantitative data with qualitative analysesand interpretations of the numbers. For example, the legalsystems paper points to the importance of gathering reliablestatistics on court case management, but cautions thatanalyses to better understand why there may be bottlenecksand delays is also essential to make the information useful.Qualitative evaluation techniques such as rapid appraisalmethods are especially good at answering "why" questions

and are thus complementary to morequantitative performance monitoring systems.One problem that is especially prevalent inPD/GG types of activities (e.g. process-oriented, institution-building) is the difficultyof finding appropriate quantitative indicatorsfor measuring performance. This iscompounded by lack of theory or conceptualframework for PD/GG assistance that makes itdifficult to clearly articulate programobjectives. Often the most easily quantifiedindicators are either too "low" (i.e. are simpleoutputs like numbers of people trained) or too"high" (i.e. overall measures of human rightsviolations, etc.) to appropriately representattainable project objectives.

Truly participatory evaluation techniques havemany advantages that donors should consider,such as promoting learning and encouragingclient/stakeholder ownership of and use of theevaluation results. While donors' rhetoricfavors using more participatory evaluation,actual practice is still limited if not rare.Donors often use rapid appraisal techniquesthat involve "listening" to customers andstakeholders, but they typically stop short ofbringing them fully into the evaluation process.Donor concerns for independence, objectivity,accountability, and "control" over evaluationsmay put some practical limits on use ofparticipatory evaluation techniques in somecases.

Some Remaining Issues

One issue is the positive and negative linkagesbetween the different types of PD/GGinterventions. This issue is important becausethe responsibility for dealing with unintendednegative impacts often does not lie with theagency directly responsible for theimplementation of that particular interventionbut become the responsibility of other actors to"repair the damage". For example

• Civil Service Reform andParticipation. How should donorsdeal with participation in developmentprocesses in which it is very clear thatthere is a negative impact on thoseinvolved? For example, realparticipation of those in the civilservice threatened by the effects ofreform processes is very difficult, ifnot unlikely.

27

• Human Rights and Privatization. While theprivatization study mentions mostly positiveeffects, individual employees of privatizedcompanies and their families may lose healthbenefits and other services that in effect cause adeterioration in their human(economic/social/cultural) rights.

• Human Rights and Democratization. Within thedomain of human rights, one can promote politicalrights through democratization programs.However, in countries with religious and ethnictensions or conflicts, the process ofdemocratization may lead to the disintegration ofthe state and violation of human rights includingeconomic, social and cultural rights.

Another issue is the interconnection of the various PD/GGprograms. Are some a necessary pre-condition for others?For example:

• Relation between Legal Systems,Human Rights and Privatization. Afunctional system of rule of law can beconsidered a pre-condition for theimplementation of human rights aswell as privatization programs. Thelegal systems study stipulates thatcreating a situation in which rule oflaw is adhered to may prove to be along-term process. The importantquestion is: How in the meantime(under conditions without rule of law)can donors implement interventionssupporting human rights orprivatization?

1

This overview chapter was prepared for the DAC ExpertGroup on Aid Evaluation by Annette Binnendijk, SeniorEvaluation Advisor to the Center for DevelopmentInformation and Evaluation (CDIE), USAID.

28

LEGAL SYSTEMS1

INTRODUCTION

Support for the Rule of Law (ROL) has emerged as a majorcomponent of an expanding portfolio of USAID democracyprograms. USAID investments in law programs date backto the 1960s, but the current resurgence of activities in thisarea began in the mid-1980s with USAID's initiation of theAdministration of Justice program in Latin America. Sincethe early 1990s, USAID ROL programs have spread to Asiaand are starting up in Africa and in Eastern Europe and theNew Independent States of the former Soviet Union.

In May 1992, the U.S. Agency for InternationalDevelopment (USAID), Center for DevelopmentInformation and Evaluation (CDIE) initiated an assessmentof donor-supported Rule of Law programs in Argentina,Colombia, Honduras, the Philippines, Sri Lanka, andUruguay. During the subsequent 18 months CDIE teams ofthree to five people spent one month in each country2

collecting data and interviewing observers of andparticipants in judicial reform efforts sponsored by USAIDand the Asia and Ford Foundations. While the assessmentis limited to the experiences of these three organizations,the general lessons should have broader applicability forother donor agencies supporting legal systems development.

The CDIE assessment was both prospective andretrospective, with its central purposes to:

• Assess recent donor experience in ROL.

• Develop criteria for initiating ROL programs.

• Propose a strategic framework for setting ROLpriorities and designing country programs.

Criteria for Country Investments

A range of generic criteria for determining whether acountry's environment might support ROL programsemerged from the six case studies. These criteria shouldform the basis for assessing whether donor investment inROL is worthwhile. The assessment should be done beforea firm decision to proceed is made. For example, knowingthe potential for support or opposition to legal reformamong political elites and organized constituencies (such asbar associations, commercial organizations, and non-governmental organizations [NGOs]) is especially crucialfor deciding whether investments in legal and judicialreform can yield significant positive results. Similarly, suchfactors as judicial autonomy, corruption, media freedom,

and donor leverage are critical in determiningthe prospects for successful donor supportedreform.

KEY CONCEPTS ANDFRAMEWORKS

The case studies also facilitate the developmentof an analytical framework for USAID plannersto use to identify the investment priorities andstrategies for effecting sustainable ROLprograms. The studies identified four essentialneeds and matching strategies for addressingthese needs. In sequential order these needs andstrategies are:

• Host country political leadership insupport of ROL programs. If politicalleadership is weak and fragmented,donors will need to supportconstituency and coalition buildingstrategies to strengthen political andpublic pressure for reform.

• Adequate legal system structures. Ifsufficient political support exists butthe legal system structures are weak,donors will need to emphasize astructural reform strategy.

• An accessible and equitable legalsystem. Where political will and legalstructures are relatively adequate butthe accessibility and equity of thelegal system are deficient, donors willneed to focus on access creationstrategies.

• Institutional capacity. Once the firstthree strategic conditions are judgedfavorably, emphasis should be placedon the institutional capacity of existinglegal structures to perform theirintended functions. Where capacity isinadequate, donors will need to engagein legal system strengtheningstrategies.

The analytical framework is intended as a toolto help donors set program priorities. Theresults of the analyses conducted using theframework will rarely dictate a simple positiveor negative recommendation to pursue one oranother of the strategies for ROL programs. Forthis reason, donors will likely pursue more thanone ROL strategy. The framework helps todetermine when each of the four strategies

29

should predominate. The experiences in the six countriesstudied suggest that a proper sequencing of the fourstrategies is important. For example, in many countriesbuilding constituencies and coalitions to create demand forstructural reform should take place before early and heavyinvestments are made in supply of legal systemstrengthening endeavors.

Other characteristics of the framework should behighlighted as well. First, the framework indicates that theformulation of ROL strategies should be problem driven;that is, program planners should identify the host countryweaknesses in ROL that seriously constrain democraticdevelopment. Second, the framework defines reform as apolitical process that cannot simply be reduced toconventional technical assistance or to institutionaldevelopment strategies. Finally, because ROL programs arepolitical, donors must often devote more attention todesigning strategies that facilitate host country demand forreform instead of the more traditional supply side assistancestrategies.

SUMMARY OF EXPERIENCE

Although constituency and coalition building is the first ofthe four ROL strategies discussed in the previous section -and was used eventually in five of the six countries studied- it was carried out as the first ROL strategy in only onecase, Colombia. Projects in Argentina and the Philippinesadopted constituency and coalition building as a fall backstrategy after other approaches proved unworkable. InHonduras and Sri Lanka some elements of the strategyappeared during the course of the ROL enterprise, but theywere not major parts of the effort to mold constituencies forreform. The argument made here - that the constituencyand coalition building strategy should be considered first -is one of the study's major conclusions.

The reasons why constituency and coalition building wasnot the initial ROL strategy in most of the sample countriesare complex. The late 1980s were a time of great optimismfor democracy as a way of political life. Argentina,Honduras, the Philippines, and Uruguay were just emergingfrom periods of sustained dictatorship or authoritariangovernment with legacies of serious human rights abusesand with judiciary systems whose independence had longbeen compromised. A newly elected political leadership ineach country showed a renewed interest in democratizationand in energizing weakened judicial systems. It was a timeof great hope for ROL development. Some countriesstudied pursued one strategy, whereas others undertookmultiple strategies simultaneously, but usually gave priorityto one. In several cases the initial strategy provedunproductive and was replaced by a different approach. Thefollowing analysis of each of the four strategies and its

impact on ROL development providesimportant insights into ROL program design.

Constituency and Coalition BuildingStrategies

The assessment's major finding concerns theneed for USAID and other donors to devotemore attention to constituency and coalitionbuilding. Constituency building refers to donorsupport for citizen, commercial, andprofessional groups engaged in mobilizingpublic pressure for legal reform and in helpingoversee government performance in executingreform measures. Similarly, coalition buildingrefers to donor efforts to help forge reformcoalitions and alliances among NGO leadersand senior government managers.

Until recently constituency and coalitionbuilding strategies played only a minor role indonor ROL reform efforts. In the late 1980s itwas assumed that the newly emergingdemocracies of Argentina, Honduras, thePhilippines, and Uruguay would demonstratethe political will necessary to move directly tostructural reform and legal systemstrengthening. It soon became clear, however,that except for Uruguay host governmentcommitments to legal reform were weak anduncertain. Fortunately, in two countries -Argentina and the Philippines - it was possibleto move away from these stalled efforts toconstituency building strategies aimed atincreasing public pressure and political supportfor legal reform.

The returns on investments in constituency andcoalition building and the overall importance ofthis strategy in ROL efforts are wellexemplified in Colombia. Colombia is the onecountry where USAID undertook a concertedand protracted effort to bring together reformelites who then became leaders in bringingmajor changes to the judiciary.

Similar results can be found in the Philippines,where in the face of weak governmentcommitment to legal change the Asia and FordFoundations focused on mobilizing newconstituencies to pressure for reform. The AsiaFoundation helped an NGO coalitionrepresenting urban poor communities to lobbyvigorously and win legislative support forurban housing rights for the poor.

Structural Reform Strategies

30

Structural reform strategies refer to the rules governing thelegal system that are reflected in constitutional provisionsand laws. Undertaking a donor supported structural reformstrategy can be rewarding, although it presents a formidablechallenge because it requires constitutional changes orlegislation. These initiatives are time-consuming endeavorsand will encounter opposition from entrenched politicalinterests. Furthermore, structural reform is only an earlystep in ROL development and requires follow up as part ofan access creation or legal system strengthening strategy toimplement the reform.

All of the countries studied have engaged in structuralreform of one kind or another, in some cases with USAIDassistance and in other cases independently. For example,four countries sought to change their judicial personnelsystems to encourage merit-based appointments andpromotions. The adoption of merit systems is particularlyimportant for raising professional standards and enhancingthe independence of the judicial branch.

Access Creation Strategies

In several of the counties studied, donor-supported accesscreation strategies have helped make legal services moreavailable and affordable to low income people who lack themeans and knowledge for seeking resolution of disputes orredress of grievances when their rights have been violated.These efforts have included legal aid, paralegal training,alternative dispute resolution, legal literacy campaigns, andlegal advocacy non-government organization (NGO)support.

Alternative dispute resolution (ADR) strategies are the mostwidespread. In five of the six countries USAID or the AsiaFoundation is supporting ADR mechanisms designed todivert cases away from the regular court system tomediation boards, neighborhood counseling centers, andbinding arbitration schemes. These ADR measures are newbut are showing promise as a low cost measure forproviding expeditious and accessible services for settlinggrievances, particularly for lower income people.

In three countries, Argentina, the Philippines, and SriLanka, support has been provided for legal aid programsand in the latter two for legal literacy campaigns andparalegal services as well. These activities are limited intheir reach and impact when they are pursued as discreteefforts. Legal aid programs tend to deal with individualcases, depend on scarce pro bono lawyer services, andfrequently cannot afford to pursue court litigation.Similarly, although disputants may be motivated by whatthey learn through legal literacy campaigns and theparalegal efforts carried out on their behalf, theiracquaintance with the law would still be too rudimentary to

empower them to act effectively or have accessto individuals who can act for them.

Legal aid and literacy efforts are more effectivewhen they are developed around specific needsand issues and are linked to organizations thathave the professional competence to engage inlitigation and to provide legal representation.The assessment revealed that legal advocacyNGOs are performing this role only in SriLanka and the Philippines. Such NGOs employlawyers who seek out and engage in classaction litigation, public interest suits, and testcases on behalf of disadvantaged groups whosuffer from a common infringement of theirrights. Legal advocacy NGOs can be highlyeffective because they target specific issues andgroups. They seek, through legal means, toreform structures perpetuating poverty andoppression, to empower communities to takeaction in defense of their rights, and to breakthe bonds of passivity and dependency. Forthese reasons investments in legal advocacyNGOs have the potential for yielding highreturns.

31

Legal Systems Strengthening Strategies

USAID and other donors have supported strategies forstrengthening legal systems in all six countries studied.Such strategies generally included traditional institutionbuilding activities and focused on enhancing the capacity ofhost government judicial institutions to render justice moreeffectively and efficiently. These strategies were directedtoward introducing new systems of court administrationsuch as improved record keeping and budget and personnelmanagement, the design and conduct of pre-entry and post-entry training programs for judges, court staff, and lawyers,and the acquisition of modern technology such ascomputers for case tracking.

The record of achievement with regard to legal systemstrengthening is mixed. In Uruguay and Colombia, forexample, these strategies have contributed to importantimprovements in judicial performance. In Uruguay theintroduction of oral procedures to supplement and replacemuch of the traditional written approach to civil caseprocessing has considerably reduced the time required tomove cases through the courts. In Colombia revampedPublic Order courts for handling terrorism cases haveincreased the conviction rate from 30 per cent to around 70per cent in the first year of operation. Still unknown is theimpact of court efficiency on human rights in a country thathas long had a poor record in this area.

Progress in the other four countries studied has been moresporadic. In Honduras USAID assisted in upgrading theskills of prosecutors and public defenders. These officialsare pursuing their roles more vigorously but are stillconstrained by inefficient judicial procedures. In Argentinaefforts to improve the legal system were unsuccessful at thenational level, but found a receptive audience in theprovincial courts. Court improvement efforts in thePhilippines have received Asia Foundation support but haveyielded few results.

CONCLUSIONS AND LESSONSLEARNED

Lessons in Constituency andCoalition Building

• A strong civil society is an effectivebase for launching efforts to mobilizeconstituencies to support ROLdevelopment.

• There are few examples of barassociations serving as major sourcesfor reform initiatives.

• The commercial sector can be animportant reform constituency.

• Although NGO-based coalitions mayprove difficult to build, they can forma strong force for legal reform.

• Free and effective media are neededto support constituency building.

• Reliable court statistics are needed toinform public debate on ROL.

• Opinion surveys are invaluable forassessing public demand for judicialreform.

• Donors have more to learn aboutcrafting coalition building strategies.

Insights in constituency building must betentative because the strategy is a relativelyrecent development. But it is clear that havinga strong and vigorous civil society helps as afoundation for mobilizing constituency support.

Because the Philippines is well known for itsrobust NGO environment, the ROL challengewas mostly to inspire active NGOs to take onnew work. In Argentina a vibrant civil societyhad become well established between therestoration of democracy in 1983 and thebeginning of USAID assistance for ROLdevelopment at the end of the decade. The taskwas to redirect energies already in use. In bothHonduras and Sri Lanka, on the other hand,civil society is weaker, particularly inHonduras; constituency building thereforebecomes more difficult.

Four sectors stand out immediately as mostreceptive to constituency building: bar

32

associations, the commercial sector, the NGO community,and the media. A review of the characteristics of thesesectors in the six countries reveals a mixed pictureregarding their potential contributions to judicial reform.

Bar Associations

To many observers, host country bar associations are animportant constituency to press for legal reform. Forexample, the Integrated Bar of the Philippines (IBP) hasproposed referring all commercial cases to arbitrationinstead of to the courts to avoid the delays and corruptionencumbering litigation in the courts. But IBP is the onlyexample of such efforts by bar associations in the sixcountries studied. Accordingly, other than the Philippines,there are no case study examples of bar associations servingas major sources of reform initiatives.

Commercial Sector

Another potential constituency for legal reform is thecommercial sector, where there is real incentive to press forproperty and contract-rights enforcement as the cornerstoneof an effective legal system. Here the country studiesprovide a mixed picture. In Honduras businesses have notactively promoted legal reform, largely because the moresizeable firms - with their political and economic clout -have not been inconvenienced by a weak legal system. InSri Lanka large firms avoid the courts at all costs andfrequently use political connections to evade litigation. Inboth countries the business community remains at themargin of the legal reform arena.

Non Governmental Organizations (NGOs) andLegal Aid

Another potential constituency for legal reform is thatportion of the NGO sector engaged in legal aid and legaladvocacy. Although NGOs represent an important resourcefor extending access to legal services, there are severalfactors inhibiting their emergence as advocates for legalreform.

• NGOs represent relatively small constituencies.• It is difficult for NGOs to form coalitions to

champion reform agendas.

• Leadership styles also impede the development ofNGO coalitions. Many NGOs are personalexpressions of dynamic leaders who, havingfounded an organization, are reluctant to sharepower with or subordinate their identity to acoalition involving other NGOs.

Although problems in building NGO constituencies andcoalitions can be discouraging, donors should not shy awayfrom supporting such efforts. The power of a coalition in

achieving basic reforms has been welldemonstrated in the Philippines.

The Media

Free and effective media are needed forconstituencies to build their base of support andto generate public pressure for legal reform. Inthe absence of effective media, coalitions andconstituencies advocating reform work inisolation and are deprived of the opportunity toinfluence and mobilize public opinion. Mediathat are free and have the professional capacityto investigate and report on deficiencies injudicial performance and the legal system aregenerally a critical ingredient of the reformprocess.

In the three LDC case-study countries - thePhilippines, Sri Lanka, and Honduras -government exercises direct or indirect controlsover television broadcasts and limits thebroadcasting of reports that might reflectpoorly on the legal system or highlight majorsocial issues that touch on legal matters. Theregulation of radio broadcasts is somewhatdifferent. Strict controls on the content ofradio broadcasting exist in Sri Lanka, but inHonduras and the Philippines radio stations,particularly talk shows, have assumed animportant role in allowing citizens to voicetheir opinions about government programs andservices. In the Philippines some of the talkshows have been hosted by legal resourceNGOs. Radio journalism is not without risks,however; more than 30 broadcast journalistshave been killed in the Philippines during thelast decade.

The print media are another matter. Newspaperreporting has been short on substantivereporting and long on sensational or superficialjournalism in the three LDC countries. InHonduras and Sri Lanka (as well as in thePhilippines during the Marcos era), newspapershave been careful not to report on items thatmight reflect negatively on the government andparticularly on more powerful public figuresbecause the government can ration or withholdnewsprint or withhold advertising. (The latteris particularly injurious where commercialadvertisers are few.) And if these measuresprove ineffective, intolerant governments canresort to intimidation or worse againstoffending journalists. All three countries havea history of suppressing the press.

33

Court Statistics

One crucial foundation for informed public debate on ajustice system is sound data and analyses on the system'sinner workings. The impact of investigative journalism andlegal analyses will necessarily be limited if no one has afirm idea of the actual dimensions of court congestion,average time to process a case through the legal system,and so on. This problem is not confined to developingcountries. Only in recent years has such applied researchbeen undertaken in the United States. In none of the threesample LDCs has applied research been initiated, while inthe other three it is of quite recent origin. Accordingly, aswas the case in the United States, there are misconceptionsof what is wrong with the judicial system. Prescriptions areoffered that may be irrelevant and wasteful of publicresources, such as hiring more judges to ease courtcongestion, when in fact the problem lies elsewhere.

Opinion Surveys

Public opinion surveys reveal little that is not alreadyknown or strongly suspected by leaders who make an effortto keep in touch with the public. But surveys do makeknown opinions about public issues in a way that is difficultfor leaders to deny or ignore. The open existence of theArgentine Gallup poll data and the Philippine businesscommunity opinion survey, for example, make it harder forthe Supreme Courts of these two countries to act as if theyenjoyed complete public support. Such data alone cannotforce reform, but they contribute to a climate in whichpolitical will for reform is easier to find.

Other Lessons

Coalition building is labor intensive and donors should beprepared to provide enough staff to support such anendeavor. This is demonstrated in the successful USAIDcoalition-building effort in Colombia. Conditions werefavorable in Colombia, but less so in the other countries.Where coalitions did not form, it is frequently unclearwhether the problem was a consequence of unfavorableconditions or a function of deficiencies in the donor'sapproach.

Lessons in Structural Reform

• Structural reform is the boldest and most difficultROL strategy to undertake.

• The impact of structural reform is frequentlydiluted by the absence of pressures foraccountability and enforcement.

• Introducing new structures may provide morereturns than reform of older entrenchedinstitutions.

Structural reform is perhaps the boldest andmost difficult strategy to undertake in an ROLprogram because it seeks to alter infundamental and profound ways the basic rulesgoverning the judicial system. This is mostconspicuous, for example, in cases where ahost government tries to initiate a transitionfrom a patronage to merit system forappointing judges and judicial staff, as is beingattempted in one way or another in four of thesix study countries.

A move toward depoliticization is bold becauseit calls for a major reconfiguration of power inboth the external and internal dynamics of thejudicial system. Externally, merit systemsprovide the judiciary with greater independencefrom the executive and legislative branches;internally, meritocracies diminish personalisticrule and favoritism, fostering more regularizedand rational procedures.

Honduras has carried the merit system idea thefarthest, since the system there comprises notonly recruitment but also promotion. TheHonduras experience illustrates how difficultreform can be, because it challenges thetraditional basis of political power. This is sofor three reasons. First, patronage in Honduras,as in other countries, is a critical resource and amedium of exchange in the political system.Political power does not derive from holdingformal positions of authority in fragilegovernment institutions or adherence toembryonic norms of democracy; power comesfrom the capacity of elites to compete withrivals in building alliances with patrons andclients. Patronage is the glue that leaders useto build and hold coalitions together. A merit-based judiciary means one less system tomobilize in such maneuvers.

Second, political control of the judiciaryimplies an ability to bend the rules of behaviorin one's favor. Compliant courts are not asconcerned with state corruption andsuppression of rights as are independent ones.Executive and legislative branches -accustomed to operating above and beyond thelaw - would prefer not to deal with theconstraining influence of an independentjudiciary. By controlling appointments andpromotions, by ensuring that judicial terms areof short tenure, by saddling the judiciary withinadequate budgets, and passing legislation

34

restricting court jurisdiction, the executive and legislaturekeep the judiciary in check.

A third (though comparatively minor) rationale formaintaining patronage in the judiciary is that HonduranGovernment employees are a primary source of financialsupport for the two major political parties. The incumbentpolitical party dispenses the rewards of governmentemployment to its followers and then "taxes" their salarieson a regular basis to build party coffers. Indeed, evenpersons hired to the judiciary under the USAID-supportedcareer merit system still pay levies to the incumbentpolitical party.

Launching a structural reform effort will encounter passiveor active resistance from vested interests and politicalfactions most likely to lose power and resources because ofthe reform agenda. From a donor's perspective, however,when an opportunity arises to initiate basic reforms, thepositive gains to be reaped justify the modest investment,even in the face of risk. Since four of the six case studygovernments have embarked on some sort of structuralreform, this indicates that there often is considerablereceptivity to such initiatives.

A legal reform process that does not include strengtheningmechanisms to ensure enforcement will likely proveineffective. Many things can go wrong in structural reform.Unremitting political opposition and general bureaucraticinertia can gradually deplete commitment to the continuousenforcement of newly adopted reform measures. Changesand rotations in government leadership mean that reformistcoalitions will lose members and eventually their elan andpolitical strength. Without an enduring internal coalition, aconstituency needs to maintain a persistent watch to holdthe government and judiciary in compliance with promisedreforms.

The assessment uncovered several cases in which importantreforms were introduced, with or without donor support, butthey either never got off the ground or faltered after short-lived compliance. Thus, in Sri Lanka, because so manyprison inmates were detained for long periods awaiting trial(because they were too poor to pay bail), the legislatureenacted a bill specifying a time limit after which the courtshad to release prisoners on bond. However, numerousgovernment officials have reported that the law is not beingenforced. In some cases the Supreme Court was issuingwrits for release of detainees, but the police and militaryauthorities were ignoring them.

Introducing a new structure may provide more immediatereturns than trying to reform older, entrenched institutions.The use of alternative dispute resolution (ADR) techniquesis an obvious choice. In Sri Lanka, there has been a highlevel of enthusiasm and commitment for the rapidintroduction of a nationwide mediation system to replace

the older structure. Abolishing, rather thanreforming a highly politicized mediationstructure in this case allowed the new structuresto begin afresh, unimpaired by pastcommitments and poor performance.

In Argentina and Colombia, ADR mechanismsrepresent new opportunities for litigants whosee the traditional court system asunresponsive, time consuming, and expensive.Following through in developing ADRenterprises in these two countries will likelybring greater success than trying to revamp theregular court system.

Finally, in the Philippines and Uruguay,commercial ADR promises to provide accessfor the many litigants who perceive the formalcourt system to be closed to them. In thePhilippines binding arbitration is an attractivealternative to delays, corruption, andunpredictability. In Uruguay the issue is moresimply that judges do not know commerciallaw.

35

Lessons in Access Creation

• Conventional legal aid, legal literacy, andparalegal activities are frequently quite limited intheir impact.

• ADR is a low-cost measure that can provideexpeditious and accessible services in settlinggrievances.

• Legal advocacy represents the most promisingaccess strategy.

The experience with the impact of access creation strategiesis mixed. Conventional legal aid activities have a limitedimpact. Whether funded from private voluntary or publicsources, programs are under funded and reach only a smallportion of the population. For example, in Sri Lanka thegovernment-operated Legal Aid Commission receives$10,000 annually and has only one office, located inColombo. There are a number of privately funded NGOsthat provide legal advice, but they do not have the money totake cases to court.

Legal literacy efforts can be extensive and reach largenumbers of people, but have limited value because so littlecan be imparted in two or three hours to scores ofsemiliterate people. To be sure, some country studiesindicate that, once informed of their rights and availablelegal services, people are motivated to seek assistance inaddressing their grievances. The experience of the SriLanka program suggests the need for backup professionallegal services to help counsel these individuals and, ifneeded, process their claims through the judicial system.Legal aid is labor intensive, and its reach is severelyrestricted because there are so few lawyers willing toprovide pro bono services.

The Sri Lanka experience further suggests that paralegalcampaigns targeted to specific constituencies and combinedwith follow-up professional legal aid are more appropriatethan investments in generic non-targeted campaigns. SomeNGOs in Sri Lanka plan to follow this approach inaddressing the lack of worker rights in tea plantations andexport processing zones. In another example, the TaskForce for Detainees of the Philippines targeted its legal aidefforts to counteract the human rights abuses of the Marcosregime in the early 1980s.ADR mechanisms, such as mediation councils, are a low-cost measure for providing more expeditious and accessibleservices in settling grievances. Experience in the fivecountries indicates that mediation can effectively settledisputes for many who cannot afford litigation. Indeed,under the new mediation law in Sri Lanka, disputantscannot go to court until they have first tried a mediationcouncil.

There are limitations to the use of ADR. Inmany instances, such as in Sri Lanka,mediation councils are not mandated to dealwith disputes that arise between governmentagencies and the citizens and communities towhom they are presumably accountable. Inaddition, both parties to a dispute must appearbefore the mediator. Frequently the defendantdoes not make an appearance, which happenedin approximately 50 per cent of the Sri Lankacases. To respond to this problem, grantingmediators summons power is being discussed,but opponents argue that this would violate thevoluntary character of mediation.

Legal advocacy NGOs represent perhaps themost promising variant of all the accessstrategies. They aggressively use the law toassist disadvantaged groups and their advocacyand lobbying activities make them an importantconstituency for reform. Legal advocacyNGOs frequently serve a dual purpose inbridging access and constituency-buildingstrategies.

Legal advocacy NGOs were not a factor inHonduras and in USAID's reform programs inArgentina, Colombia, or Uruguay. There are afew legal advocacy NGOs in Sri Lankareceiving Asia Foundation or USAID support,but they have yet to develop strongGRASSROOTS linkages to groups needingtheir assistance. Some of their leaders havepressed the courts for use of class action suitson the model promoted by the Indian SupremeCourt under the banner of "social actionlitigation" (SAL). SAL allows legal-resourceNGOs to file cases on behalf of groups, such asbonded laborers, whose rights have beenviolated. However, the Sri Lanka SupremeCourt has resisted its introduction to thatcountry.

A range of advantages and benefits sets legaladvocacy apart from other access strategies.Taken together these advantages and benefitsmake a very appealing investment. Legaladvocacy NGOs can be highly effective in:

• Extending benefits widely. Legaladvocacy strategies seek to maximizethe scarce supply of legal services forthe poor and dis-empowered byfocusing on issues involving groups ofpeople rather than individual clients.

36

• Achieving structural change. In many instanceslegal advocacy strategies address structuralconditions that perpetuate poverty and oppressionrather than simply litigate ameliorativesettlements.

• Effecting targeted outreach. Legal advocacyprograms are frequently targeted to specific groupsor issues, rather than to generic or diffuse needs,such as legal literacy, thereby funnelingorganizational energies toward well-defined needs.

• Pursuing integrated strategies. Legal advocacyfeatures integrated application of a range of accessstrategies (e.g., legal literacy, paralegals, legal aid,media) that can be synergistically combined andtargeted around achieving manageable and visibleresults.

• Empowering citizens. Legal advocacy seeks toempower communities and groups to take action indefense of their rights and to break bonds ofpassivity and dependency on outside resources.

• Building constituencies. As mentioned above,successful legal advocacy can produceconstituencies that pressure government agenciesand legislatures for legal reform.

• Enforcing accountability. Once groups andcommunities are mobilized as self-sustainingconstituencies, their continuing vigilance can serveto keep government agencies responsible forimplementing laws that would otherwise remainonly on the books.

Lessons in Legal SystemStrengthening

• Legal system strengthening may notbe the best place to begin for an ROLdevelopment program, but it can be ahighly effective strategy.

• Successful components of legalsystem strengthening strategies varywidely among countries.

• Understanding clearly thequantitative aspects of court delay isdifficult.

The most obvious lesson - alluded to severaltimes in this report - is that legal systemstrengthening is not necessarily the best placeto begin an ROL development program. If theprior steps are not in place, legal systemstrengthening will almost certainly beunproductive, as was the case in Argentina andthe Philippines. On the other hand, when thereis determined political leadership, the legalstructure is sound, and access is reasonablywide, legal system strengthening can yieldpositive results, as was observed in Colombiaand Uruguay.

This may not be a palatable lesson for USAIDor other international donors to digest, withtheir long experience in institution building.Given decades of development work inbuilding institutions in fields ranging fromagricultural credit to family planning to wastewater treatment, it is scarcely surprising that asimilar strategy was adopted for promotingROL development in Argentina, Honduras, andthe Philippines. Fortunately, it proved feasibleto change approaches midstream in Argentinaand the Philippines and to transfer theinstitutional venue in order to maintain theoriginal legal system strengthening approach inArgentina.

Another lesson is that the most successful legalsystem strengthening strategies in each countrywere peculiar to the particular legal systemenvironment found there - a pattern that can becontrasted to access creation strategies, whereit was observed that ADR approaches found awarm reception in five of the six countriesstudied. For legal system strengthening therewas a much greater difference between whatseemed to work in one place and whatappeared successful in another. In Argentina it

37

was a variety of small institution building activities at theprovincial level, whereas in Colombia it was the PublicOrder Courts. In Uruguay it was training in new oralprocedures, while in Sri Lanka it was helping to establishthe national mediation program.

Finally, it appears that introducing court statistical anddatabase systems involves more than just counting cases.Statistical exercises launched in Argentina produced largequantities of data in the first few years of work, but so farthis mass of information has been useless in creating anunderstanding of the "why" and "where" of bottlenecks,delays, and backlogs. Uruguayan statistical work hasproduced more coherent and accessible reports, butconsiderably more is needed to make the informationuseful.

Cross Cutting Lessons: Tentative Imperatives

Aside from the specific lessons identified for each of thefour strategies, several cross-cutting, suggestive insights arehighlighted here in concluding this report. The lessons aretermed "tentative", because USAID and other donoragencies are still low on the learning curve of whataccounts for success and failure in ROL projects. Theassessment highlighted several implications that crosscutthe four strategies:

• Preconditions for undertaking an effective ROLprogram may be marginally present at best inmany countries; thus ROL development efforts arenot appropriate everywhere. Of course in somecases, donors may be directed to invest in ROLprograms without such preconditions; where thisoccurs, the risk of failure is high.

• In countries with both favorable and unfavorableconditions for reform, an initial strategy ofconstituency and coalition building may be neededbefore other strategies are emphasized.

• Where ROL programs are constrained to engage inlegal system strengthening efforts, even thoughpolitical will appears weak or absent, such effortsmay be a transaction cost of initiating constituencyand coalition building activities.

• A hierarchy of institution building problemsexists, and difficulty increases with each ascendingstep. Commodity drops, human resource training,and improved management systems are leastdifficult; changes in organizational structures andsubcultures are most difficult.

• Donor ROL projects are often cost efficient butstaff intensive. Most often, although ROL projectsdo not require large outlays of financial assistance,

they are demanding of donor staff infacilitating processes of dialogue andchange within host countryinstitutions.

• Holding the justice systemaccountable for its actions or inactionis essential to democraticsustainability. The two most importantingredients in maintainingaccountability are activeconstituencies and coalitions thatdemand a high quality of justice and afree press that can point to lapses inthe system.

• The most popular of the strategiesemployed in the countries studied wasthe alternative dispute resolution(ADR) mechanisms, withrepresentation in five of the six cases.This pattern suggests that ADR shouldbe given increased attention in futuredonor planning in the ROL sector.

• Donors can serve effectively in apioneering or trailblazing capacity inthe ROL field, acting as anexperimental, risk-taking innovator todevelop approaches that can, whenproved, be taken over by other donorswilling to make substantialinvestments in this sector.

• ROL development programs receive aconsiderable boost when there is apolicy convergence between hostcountry government priorities andthose of the donor nation.

• Using intermediary organizations asROL program managers can be highlyeffective. The type of intermediaryused varied among the six countriesfrom U.S. or host country NGOs to theUnited Nations Development Program.

38

Suggested Further Research

Donor agencies could benefit from further analysis of therange of strategies for coalition building and the attendantlevel and kinds of support that might be needed. Analysis ofcoalition building strategies should be a significant elementin future USAID-assisted efforts in ROL development.

Another area that requires further investigation is thesequencing of activities in an overall ROL strategy. Eachstrategy is essentially field determined as has beenrepeatedly pointed out in this paper. Yet there are commonproperties to any political system that might argue for anemphasis in one strategy before undertaking another.

More research is needed on the impact of alternativedispute resolution. ADR appears to be the most promisingapproach to relieve the pressures on existing legal systems.It introduces the practice of third-party mediation withoutrecourse to courts and the formal judicial system. Becausewe have seen that many disputants are suspicious of theformal legal system, ADR may be a way of promoting ROLamong the population most needing redress of grievances.

One last lesson offers a cautious though positive note.Although USAID, along with the Asia and FordFoundations, has learned a good deal about buildingconstituencies and coalitions for judicial reform, there isstill much to learn, particularly in the coalition buildingsphere.

A Final Word on USAID and Other DonorSupport for Rule of Law

In the early 1990s, USAID broadened the geographic andanalytic perspectives of its law programs. Because supportfor democracy has emerged as a major objective, USAIDMissions worldwide are including law projects in the designand implementation of country democracy programs.Furthermore, the programmatic focus and content of theseefforts are encompassing a wider array of objectives,strategies, and activities. In some countries USAID'sapproach involves focusing on issues of access, legal aid,and mobilization of public demand for legal reform. Inother instances more emphasis is on institution buildingwithin the formal judicial system.

Because USAID is entering an era in which a wider array ofapproaches to law and democracy programs is being used, amore systemic perspective is needed than that implied bythe older term "Administration of Justice." This generalapproach is likely to have import for other foreignassistance donors that are working or are contemplatingworking in other social and economic domains in thedeveloping nations.

Building on the USAID experience but movingbeyond it as well, work in this new era isdistinguished by the application of a broaderrange of strategies to enhance the Rule of Lawand by corresponding refinements in ourunderstanding of what strategies areappropriate under variable sociopoliticalconditions. In that sense, this evaluationsynthesis is offered to advance the process ofstrategic thinking on how to design Rule ofLaw programs.

1

This chapter was prepared by Russell Stout of DevelopmentAssociates for the Center for Development Information andEvaluation, USAID. It is based upon the CDIEStudy,"Weighing in the Scales of Justice: StrategicApproaches for Donor Supported Rule of Law Programs",by Gary Hansen (February 1994)

2

Argentina and Uruguay were visited during a single onemonth period.

39

PRIVATIZATION1

INTRODUCTION

Study Objective

The objective of this study is to review the impact ofprivatization in developing and post-communist countriesby pulling together a synthesis of existing literature. Fromthis synthesis a number of conclusions have been drawnabout the success of privatization overall, the success ofvarious different approaches to privatization, theimportance of the policy environment, and other matters.

KEY CONCEPTS AND FRAMEWORKS

Main Issues in Assessing the Impact ofPrivatization

The main issues in assessing the impact of privatizationwhich this study seeks to address, are the following:

• What has been the effect on the privatizedenterprises themselves, particularly in terms ofprofitability, productivity/efficiency andinvestment?

• What has been the impact on government finances,primarily in terms of funds raised for state budgetfrom sale, subsidies no longer paid, tax revenueraised after, versus before privatization, and use offunds to repay domestic or external debt?

• What has been the social impact, measured largelyin terms of impact on consumers and onemployees?

• What have been the overall economic effects,particularly in terms of the impact onstrengthening of capital markets, widening ofownership of capital, on competition, and onprivate investment, domestic and foreign?

• What are the effects of differentapproaches to privatization on the endresults, and what lessons can belearned?

SUMMARY OF EXPERIENCE,CONCLUSIONS AND LESSONSLEARNED

Main Findings

Impact on Enterprises Privatized:(i) Developing Countries

The performance of companies improved afterprivatization in almost all cases. Galal et al.found that performance improved in 8 out of 9developing country cases studied. A largersample of companies examined by Megginsonet al found substantially improved performancein around 75 per cent of cases and to someextent in almost all cases. Other cases show asimilar picture.

Looking at individual performance measures,profitability of firms increased in the vastmajority of cases. Also, in most casesinvestment increased substantially afterprivatization. This release of the resourceconstraint that is common to many countries'state owned enterprises (SOEs) partly explainswhy even efficient state companies operating incompetitive or regulated markets under thesame conditions as private companies (forexample the Chilean state enterprises) wereable to turn in a better performance afterprivatization. It also explains why theMalaysian airline, MAS, was able to produce abetter performance after partial privatization,even though management did not change.Privatization removed constraints on newinvestment.

Productivity and efficiency improved in mostcases where it was specifically studied, oftenby output growing faster than labor and otherinputs. In the Chilean electricity generationcase cited, there was much more efficient useof the main input, coal. In the case of MAS,however, productivity did not increase, becauseprivatization did not result in the enterpriseundergoing any changes, other than havingimproved access to sources of capital forinvestment. The privatized Bangladesh textilemills were able to improve operating or staticefficiency, but not longer-term or dynamic

40

efficiency. This was because of poor handling of theoriginal privatization process, in that debt and other issueswere not resolved, and the poor business environment inBangladesh.

Other cases where efficiency and performance as a wholedid not improve, such as instances in Mali and Zaire, arealso characterized by poor handling of the privatizationprocess itself. Buyers were chosen without regard to theirability to run their enterprises, or even to meet the purchaseprice, which was payable in installments. Chile's first roundof privatizations failed for similar reasons, even thoughthere is no evidence that the performance of enterprises didnot improve after privatization. Buyers were sold the firmson credit and were unable to continue to meet paymentsafter Chile entered a recession.

There are a number of cases, largely from the poorerdeveloping countries, where performance can be consideredto have improved dramatically after privatization, in thatthe enterprises concerned were actually closed before theywere privatized. In the public sector they had been unableto continue operating. In many of these cases, profitabilityand the long-term future of these enterprises afterprivatization were not assured, simply because theenterprises should probably never have been created in thefirst place. Those enterprises whose production could moreeasily be adapted by the new private owners to meet a realdemand had a better future than those which could not. TheTogolese enterprises affected are an example of this. Mostcould adapt their production, with the sole exception of adetergent factory.

Another reason for post-privatization difficulties iscontinued government interference in the enterprise. Forexample, the Jamaican edible oils company Seprod initiallydid well after privatization until the government reimposedprice control on its products.

In one case, that of Mexicana de Aviacion, the companyfaced difficulties because its new owners misjudged themarket and overinvested, an example which reinforces theobvious point that private owners are not infallible.However, as stated initially, in the vast majority of casesenterprise performance improved greatly after privatization.

Impact on Enterprises Privatized:(ii) Post-Communist Countries

Evidence from the post-communist countrieson enterprise performance after privatization isless detailed, less voluminous and moreclouded by other factors. The main externalforce affecting company performance in post-communist countries was the extremerecession, which occurred as a result of theeconomic collapse of the Soviet bloc and thevery rapid introduction of market economicrelations which in many cases caused severeshort-term economic disruption.

Nevertheless, the evidence is that privatizedcompanies have been able to weather this stormmuch better than companies which remainedstate-owned, although there are significantdifferences between post-communist countriesdue to initial economic circumstances and theprivatization methods chosen.

Polish privatized enterprises have been able toincrease profitability in most cases, or at leastreduce losses, and have undertaken extensiveproduct improvement and marketing exercises.However, they have generally not increasedinvestment, for reasons of either unwillingnessor lack of capital. Enterprises privatized bymanagement employee buy-out tend to havethe most difficulties in this regard, followed bythose sold by public offer, whereas those soldby trade sale, particularly to western firms,have had the best access to investment.

A similar picture is presented in the Czech andSlovak Republics. Those enterprises with astrategic western investor have done best. Inthe Czech and Slovak Republics as in Poland,access to securities markets has helped someenterprises raise new capital. Those enterprisesin the Czech and Slovak Republics privatizedmainly through the voucher method do not yetseem to have improved enterprise efficiency,although the whole mass privatization scheme,applied fully in the Czech Republic, hasenabled that country to create a full privatemarket economy much more quickly than othercountries.

There is evidence in most post-communistcountries that privatized companies havestarted restructuring to some extent, althoughsome more so than others. With someexceptions, such as Uralmash, Russianenterprises have not undergone significant

41

restructuring, although they do seem to have improvedproduct quality and marketing in most cases. Pressure toimprove performance in Russian enterprises comes morefrom a desire to survive than from shareholder pressure.The majority of enterprises are more than 50 per centowned by management and employees, who seek to protecttheir own positions. This contrasts with East European post-communist countries.

Impact on Government Finances

Very substantial sums have been raised for Governmentsfrom privatizations, representing up to 15 per cent of totalannual revenue in some Latin American countries, forexample. Such large sums have helped considerably withmacroeconomic stabilization, as in Mexico, and haveenabled the repayment of large proportions of state debts,as in Mexico, Argentina and Honduras.

Privatization has also enabled governments to cease payinglarge subsides to state enterprises, and in several countriesthis has had a significant impact on fiscal health. In the caseof Jamaican hotels that were privatized, not only was theGovernment able to cease subsidies and receive large sumsfrom the sale of the formerly loss-making hotels, but it wasable to cancel income tax concessions to the hotel sectorand increase its tax revenues.

In two Malaysian cases the Government receivedparticularly large sums from the privatizations because itattached profit sharing mechanisms to the terms of thesales. It was able to share in the financial benefits of thegreatly improved post privatization performance.

In two of the three detailed Chilean cases cited, theGovernment was likely to experience a modest fiscal lossfrom privatization, in that if public ownership hadcontinued it would likely have received slightly more fromthe continued dividend payments from the two efficient andprofitable state enterprises. However, the modest fiscallosses in these two cases are outweighed by the overallpositive welfare effect of the privatizations.

In those cases where SOEs had ceased operating prior toprivatization, or were performing very poorly indeed, theGovernment obviously did much better fiscally by sellingthem and getting the loss-makers off their books. Howeverin some cases sales proceeds were insufficient to pay offliabilities that the Government had to assume at the time ofprivatization. Privatization was still of course wise from thefiscal point of view because it stopped the liabilities andlosses getting bigger.

Impact on Consumers and Employees

Consumers benefited from privatization in the majority ofcases. Efficiency improvements in regulated industries were

passed back to the consumer in the form oflower prices. For example, the Chileanelectricity distribution company ENERSIS wasable to reduce prices by sharply lowering theamount of electricity that was stolen orunbilled. Consumers also benefited fromefficiency improvement in competitiveindustries.

Release of the investment constraint enabledmany privatized enterprises to greatly increasethe availability of services, which of coursebenefited consumers who were previouslydenied service.

It is difficult to comment sensibly on theimpact of privatization on consumers in post-communist countries. Prices have increasedgreatly since liberalization of those economies,but before the products were often not availableat all. Such price increases are moreattributable to the introduction of marketeconomics as a whole than to privatization inparticular. There is evidence that privatizedenterprises are seeking to improve productquality and introduce new products to meetconsumer demand. In this respect post-communist consumers are benefiting directlyfrom privatization. However, given themonopolistic power that is in whole or in partretained by many post-communist enterprises,particularly in former Soviet countries, thepressure on enterprises to meet consumer needsis often not great. It could have been greater ifcompetition had been taken into greateraccount when privatization was beingimplemented.

Surprisingly, the evidence shows thatemployees benefited from privatization,although not of course in every case. Twodetailed studies were found to show thatemployees benefited from privatization in mostcases. Most other evidence points to a similarconclusion.

Employees tended to benefit in three differentways. Firstly, employment levels tended toincrease after privatization. Secondly,remuneration packages tended to improve afterprivatization, often with performance andprofit-related pay becoming available. Thirdly,many employees were able to buy shares in theenterprise that was being privatized, andbenefited from the increase in the value of theshares.

42

Obviously in some cases, such as that of Aeromexico,employment levels were reduced after privatization. Thesecases were not in the majority, however. In the cases ofenterprises that had ceased operating prior to privatization,privatization and the restarting of operations were clearlybeneficial to employees. In other cases, such as theprivatized hotels in Jamaica, the success of the hotels andconsequent boost to tourism in Jamaica had a majorbeneficial effect on indirect employment.

In post-communist countries, employment levels inprivatized enterprises have generally fallen, even thoughwage levels have tended to increase after privatization.However, employment levels in enterprises which have notbeen privatized have tended to fall even faster, suggestingthat privatization has helped preserve employment, at leastin relative terms.

Overall Economic Effects: (i) Capital MarketStrengthening and Widening of Capital Ownership

Privatization has done much to strengthen capital marketsand widen the ownership of capital, although such effectsare closely related to the methods of privatization pursuedby individual countries. Those countries which haveconcentrated on trade sales to foreign investors have beenunable to capture such benefits. However, those countrieswhich have sought to put shares of privatized companies inthe hands of large numbers of citizens, such as the citedexamples of Jamaica, Chile, Mexico, Nigeria, Poland andthe Czech and Slovak Republics, have been able tostrengthen their capital markets considerably and create orextend a large group of share-owners.

Sale of shares to employees has been another means ofdemocratizing the ownership of capital that has beensuccessfully pursued in both developing and post-communist countries. Of course, the mass privatizationschemes of the Czech and Slovak Republics, and of Russia,have created more share owners than any other approach.The ability of share-owners in Russia to exercise fully theirownership rights is in some doubt, however.

Overall Economic Effects: (ii) Impact onCompetition

Privatization has had a largely beneficial effecton competition, although many liberalizationmeasures were introduced at the time ofprivatization and theoretically could have beenintroduced without it. In practice the two tendto go hand in hand. Even where liberalizationhas been delayed to give the privatizedcompany a period of protection fromcompetition, competition has been introducedeventually. Without privatization it probablywould not have been. This is the case with thetwo Latin American telecommunicationscompanies, Telmex and CTC.

In post-communist countries, privatization hasresulted in greater competition as the previousmonopolistic structure of the economy isbroken down and smaller privatized unitsemerge from large agglomerations. However,in many cases there was no coherent strategyfor introducing greater competition, and littleco-operation between bodies responsible forcompetition and bodies responsible forprivatization. Gains from competition couldhave been much greater had coherent strategiesexisted.

Overall Economic Effects: (iii) Impact onNew Private Domestic and ForeignInvestment

Privatization has become an important meansfor countries to attract foreign investment. Inpost-communist countries privatizationaccounts for a large proportion of total foreigninvestment, for example 86 per cent inHungary and 64 per cent in Poland. Incountries such as Peru, Venezuela, Argentinaand Jamaica it has accounted for between 30and 40 per cent of total foreign investment.Some countries, such as Nigeria and Brazil,have attracted little foreign investment intoprivatization, often because of restrictionsplaced on such investment. Other countrieshave merely placed less emphasis on sales toforeigners.

Privatization has also had an important'signalling effect,' demonstrating governmentalcommitment to freer markets, and encouraginggreater greenfield investment and other formsof investment not directly related toprivatization. Such a signalling effect can alsohelp to reverse capital flight.

43

A World Bank study by Frank Sader states thatprivatizations have a particularly strong influence overdecisions to invest and calculates that each dollar ofprivatization revenue generates an extra 38¢ in newinvestment. The study also states that financial andinfrastructure privatizations have the most positive effect onother foreign direct investment.

Main Lessons and Critical Success Factors

A main lesson that can be drawn from this study is thatprivatization in the vast majority of cases is a verysuccessful and beneficial reform for developing and post-communist countries. The results of privatization in thesecountries have been in general very good in terms ofenterprise performance, fiscal impact, impact on consumersand employees, and wider economic impact on criticalfactors such as increased private investment.

These findings support the emphasis that donor agencieshave put on promoting privatization in recent years. Ifanything they suggest that an even greater emphasis shouldbe placed on privatization in the years ahead. However,important differences can be discerned in the quality of theresults of the individual privatizations surveyed. In ahandful of cases the privatizations did not bring beneficialresults. In many others, while the overall impact waspositive, and the impact on most individual indicators alsopositive, the quality of the privatization and the outcomescould have been improved, in many cases substantially.There is therefore much to be gained from an effort toencourage best practice in privatization and to improve theability of individual countries to learn from experience.

The primary lessons and critical success factors that canbe derived from this study are as follows:

• Establishing a proper balance between objectives iscritical to long-term success. The objective of raisingrevenue often conflicts with the efficiency, competitionand consumer choice increasing objectives. Pressurizedby fiscal concerns, too many governments succumb tothe temptation of preserving some of an enterprise'smonopoly power when transferring it to the privatesector in order to extract a higher sale price.

• Similarly, too many governments fail to distinguishproperly between short- and long-term objectives. Theshort-term objective may be to get a loss-makingcompany off the government's books and ensure that itstarts to function more effectively. But this is sometimesdone in such a way as to minimize long-term pressuresfor efficiency improvement. In general, the frameworkfor competition in an industry is often ignored byprivatizing governments. Proper attention tocompetition issues would bring better long-term results.

• There is an obvious trade-off betweenquantity and quality in privatizations.Countries with large amounts of assets toprivatize obviously need to move fast, but ifthey do so without proper preparation andplanning, the results will be poor. Acontrast can be made between the Czechand Russian mass privatization programs.The results of the former are clearlysuperior to the latter, where many issuesconcerning such matters as proper title toshare ownership, and corporate governance,were unresolved.

• Clear political will, accompanied bythorough planning to establish a clearinstitutional structure for privatization andproper procedures for its implementation,are important preconditions for success.Educational work designed to achieve abetter understanding of the need forprivatization, as well as time spentpreparing an effective institutionalframework for its implementation, areworthwhile investments of effort.

• The quality of the policy environment is avital factor in determining the extent of thebeneficial impact of privatization. Theworse the policy environment the lessbenefits will come from the privatization.This suggests that much more attentionshould be devoted to improving the policyenvironment at the same time asprivatization is being developed. Thatmeans the introduction of general andsector regulatory frameworks, includingmatters such as modern corporate law,shareholder rights, competition policy,utility regulatory frameworks, capitalmarkets laws, and liberalization of tradepolicy. Financial sector reform is animportant precondition for success inprivatization in many countries.

1

This synthesis study was commissioned by the EvaluationDepartment, Overseas Development Administration (ODA),London, on behalf of the OECD Development AssistanceCommittee's Steering Committee on Evaluation ofParticipatory Development and Good Governance, a sub-committee of the Development Assistance Committee'sExpert Group on Aid Evaluation. It is one of a group ofstudies commissioned by the Steering Committee.

The report was prepared by the International Division of theAdam Smith Institute, London. The views expressed in thereport do not necessarily represent those of the ODA.

44

45

INSTITUTIONAL STRENGTHENING1

INTRODUCTION

Approach

The Synthesis Study on Public Sector InstitutionalStrengthening is based on two documents: theApproach Paper from the Evaluation Department of theODA dated September 1994, and the Framework Paperpresented to the DAC Expert Group on Aid Evaluationin March 1995. The broad outline of topics to becovered were agreed at that meeting. The paper coversthe general theme of institutional strengthening indevelopment co-operation, civil service reform,strengthening state owned enterprises (SOEs), localgovernment strengthening, the role of technicalassistance and training, and some basic conclusions ofparticular relevance to aid donors.

The analysis is based upon available documentationfrom DAC Members, plus relevant publicationsobtained after perusal of several recent bibliographieson the subject, together with other pertinentpublications and case studies known to the author. Acomprehensive bibliography is presented in theAppendix of the full synthesis report.


Institutional Strengthening in DevelopmentCo-operation

The consensus in the literature on institutionalstrengthening in development co-operation points awayfrom an exclusive focus on a particular agency, andinstead points to the merits of country- and sector-wide purviews before deciding on counterpart agencies,and towards collaboration between all parties to theintervention.

Definitions have preoccupied some branches of theliterature. The World Bank's Institutional Development(ID) Handbook distinguishes concept from process:"The concept of an institution means the rules, rolesand structures organized by people to conduct theirjoint activities. The process of ID means to increasethe capacity of institutions to perform their functions;often this process means strengthening or clarifyinginter-institutional relationships". But pragmatistsacknowledge that "the practical core of institutionbuilding is organization-building". A recurring themein the literature is that institution building is a means to

wider ends: effective service delivery forexample, not an end in itself.

Tackling institutional issues indevelopment co-operation has not untilrecently been a popular activity because ofthe difficulties of handling institutionalfactors. The reasons for this include the"state of the art" being poorly developed.Furthermore, the conclusion is widely heldthat "there are no blueprints", or ID is, inIsrael's phraseology, "low-specificity".Hardware projects are easier to define anddirect. As one would expect, the literatureindicates that success rates are higher withtechnical components or projects than withinstitutional development components orprojects.

However, recent research by the WorldBank indicates that sustainability ofprojects is directly related to the success oftheir institution-building components andthe extent to which such factors wereconsidered in design. Yet donor capacitiesin this particular analytical field have beenshown to be deficient.

Approaches to institutional strengtheninghave gone through various phases, withdifferent results. Donor efforts havemoved from supply-side injections oftechnical assistance, equipment,information technology and trainingtowards injecting additional resources onlyinto an environment in which theorganization is subject to demands,pressures and discipline (from customers,stakeholders etc.) Thus the tendency is toconduct stakeholder analysis; to supportinterrelated networks of institutions(instead of just one); to assess critically thecommitment of the recipients; to makelong-term commitments of aid resources;to adopt a "process" rather than precise"blueprint" approach, and assumeresistance to change in at least somequarters.

Therefore donors have had to start:

• assessing the roles, relationshipsand capacities of the principalactors in a sector beforecommitting resources, andchoosing institutional partners;

46

• encouraging participation of the target groupand institution in identification and throughoutall stages of the cycle, to build (mutual)understanding and commitment;

• setting realistic time-frames and objectives;

• simplifying design to fit host-countrycapacities;

• building-in flexibility and close monitoring tofacilitate adjustments of projects as theyproceed;

• stimulating inter-organizational linkages;

• using existing institutions wherever possible,and avoiding the temptation to establishparallel "project management units" if the aimis ID;

• reviewing carefully selection criteria, choicemechanisms, and terms of references (ToRs)of technical assistance personnel.

SUMMARY OF EXPERIENCE,CONCLUSIONS AND LESSONS LEARNED

Civil Service Reform

The key points of donors' experience in civil servicereform are related to the above general analysis. Thus,interventions have moved away from ambitious wide-ranging attempts at wholesale reform, focusing onshort-term cost containment and medium-term capacitybuilding for better efficiency, in a much-reduced scale.The disappointing results of the approach have led to asearch for others: for example the more selective"governance approach" to civil service reform espousedby the World Bank.

The factors which have appeared to influence outcomespositively include: commitment; political will withinthe top ranks of the civil service; approaches which donot just focus on numbers, but instead review first therole of the state; seeing reform as a process; allowingtime for reforms, and concentrating on the considerabletraining implications; backing up "home-grown"initiatives; keeping the ultimate aims of the civilservice reform (CSR) exercise in view: focusing onpublic service improvement, at an affordable cost;including improvements in public information andimproved governance as a means to improvingaccountability of the civil service; and modifying thedesign of civil service programs, particularly theircomponents and sequencing.

The following approaches have workedbest:

• getting the numbers in the civilservice clear from the start;

• building capacity early, across therange of agencies involved in theCSR program;

• minimizing the number andcomplexity of new administrativeprocedures or structures, butimposing a hiring freeze;

• scrutinizing critically proposalsfor measures to improve theabsorption of retrenches into theprivate sector;

• using decentralization programsas a means of improving publicaccountability and serviceeffectiveness;

• emphasizing leadership andsupervisory skills for themanagers who remain in the civilservice;

• using multi-disciplinary technicalassistance, both expatriate andlocal in origin.

State Owned Enterprises

While State-Owned Enterprises (SOEs)have remained important in terms of GNP,gross fixed capital formation, andemployment in the economies of manydeveloping countries, they have oftenperformed badly and represented abudgetary burden on the state. Early donorforays into the SOE domain attempted tostrengthen individual enterprises, but thisapproach did not acknowledge that thereare many external environmental factorsoutside of the control of SOE managers:political interference in pricing policies,decision-taking and staffing was common.Credit and labor market policies andrigidities were often imposed on SOEs.Budget constraints would be evaded.

After reflection by donors in the late1980s, it was concluded that the

47

institutional and political environment of SOEs shouldbe taken as important datum in formulating theapproach towards not just the SOEs in question, buttowards the public sector as a whole (especially in Sub-Saharan Africa where SOEs are extremely important innational economic terms). Simulation of private sectorconditions is seen as the key to this approach. Thesummary of conditions to be simulated (as far aspossible: difficulties with state monopolies areacknowledged) are: "maximization of profits, in acompetitive market, under the control of managers withcapacity, authority and motivation, faced with thethreat of bankruptcy if the concern cannot compete."

This has involved:

• redefinition of clearer objectives for SOEs;

• provision of autonomy for managers, who areheld accountable for results;

• improving selection procedures andcompensation for managers, the key criterionbeing competence to operate in a commercialenvironment;

• the provision of incentives to produce results,and sanctions for non- achievement.

Unfortunately, the impact of the institutionalcomponents of the SOE case studies in the studysample (all from ODA) has been negligible. Thelessons of experience point to the importance of:

• studying the institutional context;

• discussing the capital and institutional aspectsof a prospective SOE project together;

• collecting baseline information, then planningfor revenue collection;

• stimulating and taking account of recipients'initiatives;

• mixing technical assistance inputs, byexperimenting with twinning relationships;

• giving training priority;

• assessing the seriousness of the problem ofcorruption, and including it in any feasibilityanalysis.

Local Government Strengthening

Decentralization is popular. Of 75 developing countrieswith more than 5 million inhabitants, 63 claim to have

started transferring powers to localgovernment. But much depends what ismeant by "decentralization". Under"deconcentration", decision-taking remainsat center, local branches execute centralpolicies, and are accountable to the center."Devolution" provides for local-levelauthority to decide use of resources, andaccountability to local constituents."Delegation" implies transfer of decisionmaking and management authority to anorganization not controlled by centralgovernment.

The possible advantages of decentralizedmanagement and administration include:

• encouragement of local officialsto tailor projects and services tolocal conditions;

• local officials may be moremotivated to find out about localconditions if they are expected toreport on them;

• local people may be betterinformed about what is beingplanned and the reasons for theseplans;

• it can therefore stimulate localparticipation;

• it may relieve top management ofdetailed tasks;

• it may facilitate local politicaleducation, and encourageconstructive attitudes to change,mobilization of more localresources, and betteraccountability.

The question remains as to whether it istherefore always a good thing. TheInternational Bank for Reconstruction andDevelopment (IBRD), the SwedishInternational Development Co-operationAgency (Sida), and Dutch aid areskeptical. They conclude that the successof decentralization depends on theincentives it creates, the local capabilitiesit mobilizes or stimulates, and the costs itimposes. Therefore it is necessary toanalyze what must be administered locally,and to transfer only those powers and

48

resources needed to take over those functions. It is alsoimportant to examine the history and characteristics ofthe situation, and assess if in fact government will bemore accessible and efficient if decentralization takesplace. Therefore it all depends on the political,historic, socio-economic, financial and legal context.

There are some fundamental barriers todecentralization. "The paradox of power", inRondinelli's words, requires that for reforms to besuccessful, there must be diffused political support andparticipation; yet those in power, whose commitment isneeded, see this diffusion as a threat. This paradox mayobstruct innovation in governments and within donoradministrations. Also, local government must functioneffectively if it is to gain sustainability and legitimacy.The key is the structure of incentives facing politiciansand officials at municipal level; this pattern depends oncentral/local government interaction. There is a realdanger that decentralization can become a series of"random concessions" by central governments trying tomaintain political stability.

Furthermore, as can be observed in particularly theLatin American context, if local government is tied tocentral government mandates too closely, localgovernment leaders will be judged on how successfullythey obtain funds from higher levels of government,not on how well they solve local service problems.

The lessons of (largely negative) experience of donor-sponsored attempts at local government developmentworldwide include:

• "decentralization" comes to have manymeanings in different contexts;

• responsibilities between the different layers ofgovernment require unambiguousclarification;

• revenue sources need to be authorizedaccording to these responsibilities;

• accountability needs to encompass regulationby central government, and incentives to localconstituents;

• effort is best invested where there is realinterest in reform;

• service delivery needs as much attention asprovision of infrastructure (keeping in mindthe ends of local government, not just themeans);

• local governments need thefreedom to adapt structures tolocal circumstances, without over-prescription from centralgovernment;

• poor local services often lead torefusal of consumers to pay forservices.

49

Technical Assistance and Training inInstitutional Strengthening

Growing disillusionment with Technical Assistance(TA) as the primary vehicle for capacity building isapparent on the part of donors and recipient countries.This is particularly acute in the African setting. In arecent review by the World Bank, only 25 per cent ofID components of technical assistance projects couldbe considered successful. Another conclusion is thatTA is very expensive. Notwithstanding someconstructive suggestions for enhancing theeffectiveness of TA (improving the definition ofobjectives, diminution of the selection and proportionof expatriate TA by pursuing intermittent staffingpatterns, increasing attention to supervision, and tolocal management of TA), in the words of one observer"there is no clearly-lit path to renewed public sectoreffectiveness". More recent suggestions includemaking the pricing of TA more transparent, therebyclarifying the opportunity cost of TA, and untying the"bundle" so that it is no longer tied to other projectinputs.

Training - often seen as a panacea - is justifiablycoming to be viewed more critically in the literature.Some of the lessons which have been drawn point tothe following approaches:

• seeing training as a means to an end, wherethere is a choice of training means;

• examining trainees' working environments toensure training is not likely to becounterproductive or demotivating;

• seeking out existing policies within targetorganizations regarding training, and assessingthe likelihood of the effective utilization ofnewly acquired skills;

• training in groups;

• better needs analysis;

• capitalizing on the advantages of a longrelationship between a donor and a targetinstitution;

• improving training evaluation practices;

• pursuing in-country or third-country training,to fill information gaps;

• considering "twinning" as a cost-effectiveapproach to organizational development,which is less likely in many cases to result in

the sometimes hostile reactionsevoked by managementconsultants.

Constraints Faced by Donors:and Recent Trends

The academic literature demonstrates alack of awareness of the constraints underwhich donors work. These relate to theirstaffing (and their experience in thecountries to which aid is given);geographical distribution of offices;pressures from domestic constituencies foraccountability in the use of aid funds,while maintaining spending andcommitment rates; the conceivablypolitically sensitive nature of ID work, anddomestic pressures to tie aid to donorcountry suppliers.

However, it is clear that, notwithstandingthese constraints on donor abilities andfreedom to learn, the experience of the lastdecades has been salutary. The languagenow heard is of the need for donors to mapout the institutional terrain at country andsector level. Standard "recipes" have beenshown to be inapplicable. Sharedcommitment - on both recipients' anddonors' sides - has been shown to be vital:conditionality can be a blunt tool inforcing institutional change. Participationfrom the earliest stages of a project'sidentification by those who will be partiesto it is one of the most effective ways ofgenerating commitment. Rapid donorresponse to well-considered initiatives onthe recipients' side is another approachlikely to engender commitment. Yet theliterature is thin on this very subject. Fromthe donors' side, time horizons for programcommitments to agencies or sectors mustlengthen. It is axiomatic that "ID takestime".

Learning is slowly becoming respectablein the aid field. It is explicitly a part of thelatest IBRD version of the project cycle.One point to be learned is thatsustainability of projects depends onpredominantly institutional factors and thestrengthening of institutions involved,which is best achieved not through theapplication of various "supply-side"means, but by understanding more clearlydemands for one of the ends of institution-

50

building: improved service delivery, and greaterawareness amongst the consumers or users of servicesor outputs as to their rights, and the standards they canexpect.

There is also a growing realization amongst the donorcommunity that their response to the imperatives of theID agenda is constrained by their own institutionalsituation. The process of remolding professionalattitudes amongst aid technical staff will take time.

It is to be hoped that a renewed willingnessof donors to learn from their own and eachothers' experiences, to pay more attentionto the impact in relation to the costs ofwhat they undertake, and to do more topublicize the positive results of what hasbeen achieved with the co-operation ofprofessionals and citizens in developingcountries, will result in a gradual sea-change of public opinion in favor ofdevotion of a pattern of resources to andthrough aid agencies towards bettercollaboration in the field of public sectorinstitutional strengthening.

1

This synthesis study was commissioned by theEvaluation Department, Overseas DevelopmentAdministration (ODA), London, on behalf of theOECD Development Assistance Committee's SteeringCommittee on Evaluation of ParticipatoryDevelopment and Good Governance, a sub-committeeof the Development Assistance Committee's ExpertGroup on Aid Evaluation. It is one of a group ofstudies commissioned by the Steering Committee.The report was prepared by David Watson, a privateconsultant. The views expressed in the report do notnecessarily represent those of the ODA.

51

EVALUATION CAPACITY BUILDING 1

PURPOSE OF THE STUDY

At the meeting of the DAC Expert Group on AidEvaluation (Expert Group) in October 1994, it wasdecided to make an analysis of donor experiences withEvaluation Capacity Building (ECB) activities. Therenewed interest for ECB emerged as part of theevaluation of programs supporting ParticipatoryDevelopment and Good Governance (PD/GG). TheExpert Group decided that the review of ECB activitiesshould be a study of its own since ECB is a taskspecified in the Group's mandate. Danida tookresponsibility to be the lead agency for this task.

The study was set up to address the following

questions:2

• What are the most successful donor ECBpolicies and strategies?

• Who are the most appropriate developingcountry partner organisations for ECB efforts?

• What are effective donor activities andorganisational approaches to implementing?

As part of the study, a survey of donor agencies’ ECBactivities was conducted during the first half of 1995 towhich 24 agencies responded. Supplementarydocumentation was received from several agencies,particularly multilateral donors which have been veryactively involved in dialogues with recipient countrieson ECB, including in the organisation of severalregional seminars and workshops. A first report wasdistributed to the Expert Group in September 1995.

At the informal seminar in Canberra, February 1996,Expert Group members together with developingcountry participants discussed ECB as a full day's topicwith presentation of papers including three countrycase studies on support to ECB in Indonesia,Zimbabwe and the Philippines. The major lessonsfrom the presentations and discussions at the seminarhave been included in the study.

MAJOR FINDINGS OF THE STUDY

The DAC report from 1988 already observed that therehad been a significant strengthening of the evaluationprocess in many developing countries.3 Since then, it

has, in particular, been the World Bankand the regional development UNDP bankswhich have been most actively supportingnational evaluation systems andstimulating the demand for evaluation inconnection with public sector reforms andgood governance initiatives. ECBactivities by the United NationsDevelopment Program (UNDP) haveincluded the production of a series ofmonitoring and evaluation countrymonographs.

Support provided by bilateral donors isconcentrated on a few agencies, includingUSAID, the Canadian InternationalDevelopment Agency (CIDA), theNetherlands Directorate-General forInternational Co-operation (DGIS) and theSwedish International DevelopmentCo-operation Agency (Sida). DGIS has, asone of the very few bilateral donors, aformulated policy for support to evaluationcapacity building. Sida's support toperformance audit in several Africancountries is the best example of long-terminstitutional twinning arrangements; in thiscase between the Swedish National AuditBureau and the respective host countrycounterpart institutions.

Most of the bilateral ECB assistance hasoften been part of a project or programcontaining a more comprehensive packageof development interventions. Multilateraldonors have tended to focus theirinterventions at the national levelsupporting overall evaluation systemsoften as part of broader public sectorinitiatives. Bilateral donors have, on theother hand, been more inclined to supportevaluation functions at the department orproject/program level sometimes with themotive to have their own assistancemonitored and evaluated. Many donorsalso see their efforts to make jointevaluation as a means to support capacitybuilding. Other donors limit their ECBsupport to training.

In spite of more than two decades supportthere has been limited systematicassessment of the effect and sustainabilityof the assistance. It is therefore difficult todraw very specific conclusions aboutlessons learned. It appears that acombination of support activities including

52

consultancy, staff training and equipment is needed inmost cases, and that the establishment of a usableevaluation function on national as well as sub-sectorallevel is a long- term process. The duration and input ofdonor resources for ECB support varies considerably.However, where the duration of the intervention hasbeen short there is often a need for follow-up measures.

There is broad agreement that the commitment andsense of ownership made at policy and seniormanagement level as well as the legal foundation of theevaluation function is more important than supply ofdonor resource inputs. However, from an assessmentpoint of view a number of issues remain regarding theexperiences of the ECB support. For example there isvery limited knowledge of the sustainability ofsupported evaluation institutions at project and sectorlevel which are not linked up to a national or regionalevaluation system.

Experiences with joint evaluations have been mixed,and it appears that careful design is needed if jointevaluations are to give a satisfactory outcome for boththe donor and the host country, and at the same timecontribute to capacity building with a lasting effect.

CONCLUSIONS AND IMPLICATIONS

In summary, most agree that some progress has beenmade in achieving ECB but more work is needed in theareas of political advocacy, local level commitmentand commitment of donors to support a long-termstrategy of ECB.

The experience of recent renewed interest forevaluations in developing countries provide lessonswhich can be summarised under:

i) approaches to strengthen evaluation functions;

ii) elements of donor support strategies, and;

iii) role of donor evaluation units and the ExpertGroup.

Although the available material is limited, thefollowing observations are broadly supported; however,opinions may differ on specific issues.

An important element of the recent concern about theevaluation capacity in developing countries is theacknowledgement that donors and host countries haveeach their legitimate but sometimes different interestsin evaluations. Secondly, a sincere long-termcommitment to strengthen evaluation capacity indeveloping countries also means leaving more of theinitiative and design of evaluations to the recipients.

While much of the assessment of effectsand impact should be left to the recipientcountry, donors will continue to haveneeds for evaluation both foraccountability purposes and to providelessons on the adequacy of their aiddelivery systems.

Approaches to strengthenevaluation functions

• sustainable evaluation institutionsneed political commitment andsupport at the highest policy andmanagement levels, and should beable to demonstrate theirusefulness at these levels. Thedesign of evaluation systems alsoneeds to take into account thespecific government andadministrative culture in the hostcountry/organisation.

• Political commitment and seniormanagement demand should bepre-conditions for ECB supplyactivities, and have to be linked tothe governance issue. A long-term strategy is needed foreffective interventions.

• The scope of national levelperformance evaluation andperformance auditing systems aremoving closer to each other,although the former is likely to becloser integrated in the planningprocess while the latter systemtend to focus more onaccountability at the policy level.The choice of approach may,however, depend on other factorslike political commitment, legalframework and institutionalcapabilities.

• Development policy and aid tendto shift from a project/program tosector/policy focus setting newdemands for host countryevaluation institutions.

• Sustainable and effectiveevaluation systems must have alegal foundation or a firm statuaryorganisational regulation.

53

• An evaluation unit's independency from linemanagement is important as well as thesecurity of career possibilities for evaluationstaff and managers.

• Regional, sectoral and program/projectevaluations become more useful if they arebased on a co-ordinated approach linked to anational evaluation system particularly withrespect to methodologies and data needs.

Elements of donor support strategies

• Duration and scope of support should beflexible and balanced between needs forlong-term relations and ownership by hostinstitutions.

• Consideration of support to either a nationallevel evaluation or a performance auditingsystem should include policy demand for itsuse and legislative backing of the system.

• Efforts to institutionalise trainingin evaluation (including trainingof trainers) particularly onmethodological aspects ofevaluation.

• Long-term twinning arrangementswill support professionalism.Increased use of the evaluationtool in developed countrygovernments increase thepossibilities for making long-termtwinning arrangements withspecialised evaluation institutionsin donor countries.

• Support to training institutionsand curriculum developmentwhich on a broad base canstrengthen evaluation capabilitiesin government and civil society.

Role of donor evaluation units andthe Expert Group

While the above list provides possibleways for donors to strengthen support toECB, the need for action by the ExpertGroup and donor evaluation units mustalso be considered. While multilateralagencies' evaluation units may, to a certainextent, be able to provide support toevaluation capacity building, it is veryunlikely that bilateral evaluation units cantake operational responsibilities for supportto ECB. Their independency and capacityoften set limitations for such aninvolvement. However, that does notprevent bilateral evaluation units fromusing their professional competence toadvise and support operational units'activities and promote ECB in guidelinesand policy formulation.

Areas where donor evaluation units mayplay an active role include:

• Promote an agency ECB supportpolicy or strategy particularly inview of new aid forms beingintroduced including programsupport to institution and capacitybuilding as part of goodgovernance initiatives at nationaland sectoral levels.

54

• Advocate and stimulate interest in evaluationin country dialogues and sector programassistance.

• Provide technical advice to operational unitsresponsible for ECB support activities.

• Support the establishment of twinningarrangements between other domesticevaluation institutions and host countryinstitutions.

• Arrange joint-evaluations with a trueparticipatory approach where the needs ofboth parties are incorporated from the start,and where the capacity building element isconsidered specifically.

• Co-ordinate its evaluation program with hostcountries and other donors in order to optimiseuse of resources and constrained capacity ofrecipient countries' evaluation systems.

• Assist in securing consistent evaluationmethodologies and terminologies in the ECBsupport activities of the agency.

• Advice on training facilities and materials onevaluation.

The Expert Group can:

• Continue to promote commonprinciples, methodology andterminology shared by donor andhost countries.

• Encourage and facilitate membersto have sector and countryevaluations co-ordinated andshared with host countryevaluation institutions.

• Continue supporting regionalnetworking to promote exchangeof expertise and share evaluationinformation among recipient anddonor countries.

• Support regional and south-southseminars and training.

• Facilitate exchange of informationon donor ECB support activitiesand promote assessment of theseactivities.

Most of the possible activities outlined forthe Expert Group can be combined withother initiatives taken by donor agencies.

1

The report on Evaluation Capacity Building wasprepared by Henrik Schaumburg-Muller, consultant toDANIDA (Danish International DevelopmentAssistance).

2

No distinction is made in this report betweenevaluation capacity building and evaluation capacitydevelopment. The latter terminology is preferred bythe World Bank and the United States Agency forInternational Development (USAID).

3

Evaluation in Developing Countries - A step in adialogue, Paris: DAC/OECD 1988.

55

DECENTRALIZATION1

INTRODUCTION

Purpose and Scope of the Paper

The DAC Expert Group on Aid Evaluation has defineddecentralization as one of five themes in its discussionof Programs Promoting Participatory Development andGood Governance (PD/GG). Decentralization, it isbelieved, might promote efficiency, equity andpolitical participation. It is therefore of relevance forthe development of PD/GG.

In many developing countries decentralization is animportant political issue, and most countries haveadopted strategies in this respect. It is usually not aquestion whether decentralization should beundertaken; rather, it is a question of how todecentralize and what to decentralize; what powers canbe allocated to local governments; which functions canbe delegated to local institutions; what expendituresand taxes can be decentralized; what subsidy or transferprograms can and should be developed; what kind ofadministration and co-ordinating mechanisms can beutilized; and how can decentralization be co-ordinatedwith other reform programs?

Topics Covered

A series of objectives and potential benefits can belinked to decentralization. It is useful to distinguishbetween four major objectives:

i) To improve democracy and political equity;

ii) To improve management efficiency;

iii) To improve financial performance throughincreased revenue generation and rationalexpenditure decisions; and

iv) To provide a better environment for privateenterprise and responsiveness to local needs.

Democracy and Political Equity

Decentralization of government is generally seen as anelement for strengthening democracy, a way ofbringing decisions closer to the people most affected bythe decisions, and a way of achieving the participationof ordinary people in decision-making at the locallevel. Decentralization may help reduce regional

inequalities in terms of development, andcontribute to an equitable distribution ofresources and opportunities.

Politically, decentralization meanstransferring power to people andinstitutions in the periphery who otherwisewould not have much influence ondecision-making at local or national levels.Decentralization is an importantinstrument for improving democracy andachieving better governance. The tendencyof central governments or elites to becomeall-powerful can also be controlled orcounteracted by stronger local or regionalgovernments.

It should also be observed, however, thatdecentralization sometimes may be a wayfor the state to penetrate and control the(rural) society (Mutizawa-Mangiza 1990).The intent of decentralization strategiesmay thus be to enhance the leading role ofthe dominant party or the government.

Competing political interests can motivatesupport or opposition to decentralizationprograms. It is therefore important toidentify and assess the motivation fordecentralization among different groups orpolitical actors. Economic andadministrative rationales for or againstdecentralization are often advanced toconceal the primacy of political interests orissues. These political considerations areparticularly intricate when decentralizationefforts are undertaken to build nationalunity. Decentralization often results in theallocation of resources and benefits to aparticular region, ethnic group, or othersubset of a country's population.

Decentralization also contributes tomaintaining economic and otherdifferences between regions by reducingthe central government's obligations orefforts to subsidize or stimulate lessdeveloped areas.Changes in Management Efficiency -Improved Service Delivery

One important argument fordecentralization is that it can enhance andstrengthen a country's management andadministrative efficiency. Administrativeeffects that one wants to achieve bydecentralization are to:

56

• extend public services to rural areas, andimprove administrative and managerialcapacity;

• increase efficiency and effectiveness ingovernment operations;

• enhance economic and social developmentprograms;

• reduce overload and congestion in channels ofadministration and communication;

• facilitate more effective integration ofgovernment programs, and improve thetechnical capacity to deliver public services atlocal level;

• improve opportunities for governmentaccountability.

It is widely agreed that decentralization is necessary toimprove public management, economic performance,and income-distribution. However, decentralization ofeconomic management functions can also result inmaintaining old-fashioned or historically-rooted publicsector inefficiencies. Local governments or authoritiescan monopolize productive sectors and distort theterms for private sector participation and involvement.Institutions established at the local level frequentlyhave a poor organizational structure and are often notable to deal appropriately with the duties andresponsibilities assigned to them (Silverman 1992).Thus, transferring authority and power to the publicsector at the local level does not necessarily result ininstitutional strengthening that fosters greaterefficiency and equality. A separate effort and fundingfor local administrative capacity building is oftennecessary.Financial Performance and EconomicEfficiency

One justification for decentralization is that it willimprove economic development, equity, and incomedistribution. In a number of developing countries therehas been government involvement and intervention ineconomic activities and production of services.Economic efficiency in these government enterpriseshas been very low. An important discussion in thedeveloping countries is what the proper size and scopeof the public sector should be.

Local councils often experience serious difficulties andare inefficient in their efforts to collect levy. Therevenue basis is generally low, both centrally andlocally, and resistance to taxes is widespread. The

central government frequently fails tocome up with the funding for activities thatthe local authorities depend upon it toprovide. Decentralization of functions andtasks to local authorities without funds iscommon. Financial decentralizationremains controversial and a highly politicalissue in developing countries.

Sources of Information andMethods Used

This paper is based on the documentationon decentralization in a selected number ofdeveloping countries. The documentscover country experiences with politicaland administrative decentralization inAfrica, South Asia, Central and SouthAmerica. The country cases are used asillustrations of various models of political,administrative and financialdecentralization and not a representation ofdecentralization in each continent.

Interviews were not carried out indeveloping countries nor with donororganizations. The members of the teamresponsible for this paper, however, dohave extensive experience withdecentralization in developing countriesthrough relevant research projects or aidprograms.

57


Definition

Decentralization refers to attempts to change thebalance of power from the central government to local,regional, or other subnational levels. Decentralizationthus relates to the role of, and the relationship between,central and local institutions, both public and private.Decentralization can take the form of transfers ofpower to govern, to tax, and to plan and implementprojects.

When a country adopts a decentralization policy it isconcerned with two broad strategies for action, that is,political and administrative strategies to decentralize.Recently, financial decentralization has also beenemphasized (Agarwala 1992, Chole 1994).

The term "decentralization" is used to describe avariety of institutional structures and arrangements. Itis, however, common to distinguish between four majorforms of decentralization (Blair 1994):

• Deconcentration;• Delegation;• Devolution; and• Economic deregulation

A. Deconcentration

Deconcentration involves the transfer of selectedfunctions "... within the central government hierarchythrough the shifting of workload from central ministriesto field officers, the creation of field agencies, or theshifting of responsibility to local administrative unitsthat are part of the central government structure."(Rondinelli 1983). At the local level, governmentoperates in separate ministry offices or line-agencies.In deconcentrated systems there is little or nohorizontal integration or co-ordination of work betweenthe different sector ministries and agencies at the locallevel.

B. Delegation

Delegation involves transfer ofresponsibility for maintaining orimplementing sector duties to regional orfunctional development authorities,parastatals and other semi-autonomousgovernment agencies, that operateindependently of central governmentcontrol. Delegation usually occurs insectors that have a relatively soundincome-generating basis, such as energyproduction and supply,telecommunications, and publictransportation, etc.

C. Devolution

Devolution involves the transfer ofdiscretionary authority to legallyconstituted local governments, such asstates, provinces, districts ormunicipalities. In devolved systems,responsibility for a wide range ofoperations, encompassing more than onesector, are assigned to local governments.An essential characteristic of discretionaryauthority is that the overseer role ofcentral government is limited to ensuringthat local governments operate withinbroadly defined national policy. To theextent that local governments havediscretionary authority, they can do whatthey want, bound only by the broadnational policy guidelines, their financial,human, and material capability andcapacity.

In devolved systems, local level staff isresponsible to local elected councils, ratherthan to sector ministries. The managementof projects is integrated into theestablished structure of local governmentinstitutions.

D. Economic Decentralization andSelf-Governance

An important aspect of decentralization inthe last 15-20 years is the drive to shiftresponsibilities for economic productionand activities from the public sector toprivate or quasi-public organizations.Central in this are efforts to deregulate thecentral government's economic control andpromote strategies for private sector

58

development or community participation and private-public partnership.

SUMMARY OF EXPERIENCES

In this section we examine decentralization reforms indeveloping countries in Africa, South Asia, and LatinAmerica. Our focus is on country experience ratherthan donor experiences. Decentralization is a politicalissue that is rooted in a country's history and politics,and donors are considered subsidiary to countryexperiences. Donors play a supporting role in nationaldecentralization. It is therefore important to record anddiscuss donor experiences to make this support morerelevant and efficient.

Sub-Saharan Africa

In African countries the decentralization of the 1970smeant a deconcentration of administration andstrengthening of regional administration. Thedominance of the one-party state producedcentralization of power and a weakening of the localpower base. Decentralization was again promoted bynational governments and external donors as a result ofthe economic stagnation and crisis of the 1970s and1980s. Locally elected councils were reintroduced inmany countries, but they remain underfunded andunable to assume the tasks that have been transferred tolocal government. The central government remainsoverburdened and the political influence andimplementation capacity at the local level remains low(Laleye and Olowu 1990). In some casesdecentralization was an integral part of public sectorreform programs (Zambia), while the Civil ServiceReform Programs in others were expanded to includedecentralization (Tanzania), or merged with adecentralization reform (Uganda).

Tanzania

We can distinguish three phases in the development oflocal government in independent Tanzania (Max,1991). In the first phase, until 1972, there were electedorgans at the local level (District and Town Councils).In the years after independence, however, local electedauthorities gradually lost influence and capacity, asmeasured by reductions in tax revenues and decliningmotivation at the local level. In the second phase(1972-1982), local elected bodies were abolished andall administration was taken over by the central state.In this period, which was labelled "decentralization",local taxation was abolished and the regionaladministrative level was strengthened.

The present system of government wasintroduced in 1984. In this system, councilelections and local taxation werereintroduced and the tasks previouslyundertaken by the developmentcommittees were transferred to the DistrictCouncils. The administrative apparatusthat had developed at the regional levelwas preserved. The autonomy of theDistricts has therefore been more formalthan real.

Tanzania's system of local government hasbeen in a state of deep crisis (UNDP, 1993,NORAD, 1995). In response to this crisis,the structural adjustment program wasexpanded to include an ambitious publicsector reform program consisting of threeareas: (i) civil service reform, (ii)parastatal sector reform, and (iii) financialand planning system reform. The CivilService Reform Program consists of sixmajor elements:

• ministerial organization andefficiency reviews;

• pay reform;

• personnel control andmanagement;

• administrative capacity building;

• retrenchment and redeployment ofstaff;

• local government reform.

When the program started there were onlyfive components. The sixth, localgovernment reform, was added in the1994 Action Plan and is by far thesmallest. Little has happened in localgovernment reform in Tanzania.Autonomy of local political authorities isseverely limited. The local governmentreform component has been underpreparation for some time and some reviewpapers have been produced, but as a reformprocess it is only expected to start up in1996.

It is too early to evaluate the effects of thereform process because it is still in apreparatory phase. We can, however, state

59

what a successful reform program will require:

i) economic improvements, so that state financesare improved;

ii) favorable political conditions, with firmsupport for the reform process; and,

iii) an expansion of program priorities, beyond thenarrow goal of cost cutting and retrenchment(although this is also necessary), to includefactors such as relations between differentgovernment levels and agencies, personneland recruitment policies, work conditions,staff qualifications and improvements inservice delivery.

Zambia

At independence in 1964, decision making was highlycentralized in Lusaka. The Local Administration Act of1980 provided for the merger of party and governmentadministrative organs at the district level into acommon structure called the district council. In 1991,local political institutions were reintroduced. The firstlocal elections under the new system were held in1992, giving the ruling party, MMD, a dominantposition.

Under the new system, local governments have fairlywide formal responsibilities for housing, water supply,sanitation, roads, fire services and town and countryplanning. Local authorities are also empowered toappoint local officers and other employees. In practice,councils differ greatly in the range of functions theydischarge (Tordoff and Young 1994). This reflects bothvariations in financial and manpower resources anddifferences in legal status.

All Zambian local authorities are currently in adesperate financial situation (Tordoff and Mukwena1995). This is related to the limited revenue actuallycollected at the local level and to blockages andirregularities in the flow of funds from the center(GoZ/LOGOSP 1994).

In addition to the limitations described above,decentralization is facing severe problems because of:

• failure of the central government, and theMinistry of Local Government and Housing inparticular, to provide guidance, support andinformation

• lack of an employment and training policy inlocal government

• delays in implementing the newpolicy

• persistence of legal ambiguities

• no system for local governmentfinancing.

(Tordoff and Mukwena 1995).

Zambia is implementing a Public SectorReform Program (PSRP) wheredecentralization is the third component(GoZ 1993). ODA/UK is supporting adecentralization secretariat (LOGOSP -Local Government Support Project) inZambia. ODA has put an emphasis onmonitoring and regular review missionsduring project implementation. Nodecision has been taken yet to have an ex-post evaluation of the project. LOGOSP isalso supporting the development of asystem for local government performanceappraisal. These appraisal systems will beutilized by the district staff in collaborationwith LOGOSP and will be closely relatedto the training to be done in the districts(LOGOSP, 1993).

Parallel to the district councils, a separatedecentralization exercise has beenconducted by the Ministry of Health.Under this program, Health ManagementBoards have been created at all majorhospitals and at the district level. Thedistrict health boards will be chaired by thedistrict medical officer and will have 5-10members, drawn from a variety ofprofessional and community backgrounds.Each district council is entitled to have 2-3representatives on the boards. Thefunctions of the boards include preparationof annual plans and budgets and themanagement and operation of the districthealth service. The center, however,retains control over general policy making.The implementation of the healthmanagement reform program has beendelayed, although the law governing thereform, the National Health Service Act,and the guidelines were issued last year. Itis therefore too early to assess the results.The World Bank, UNICEF and DANIDAare the main donors. DANIDA's supportincludes funds and technical assistance forthe Health Reform Implementation Team.There will be a major evaluation

60

conducted in the summer of 1996, co-sponsored by allthree donors.

Uganda

As a part of this reconstruction of governmentinstitutions after the civil war, the National ResistanceMovement (NRM) introduced an administrative reformprogram in 1986. This system was based on ResistanceCouncils, operating at five levels, from the village tothe District (DANIDA 1995b, Brett 1992). Thesecouncils did not control administration, but were seenas community development agencies that could serveas watchdogs of officials. They were weak, however,and had no control over budgets. They could complainabout abuses, but were unable to ensure that action wastaken in response to the complaints.

A Task Force was set up in 1990 to prepare a proposalfor decentralization reform. The main objectives of theTask Force's proposal, which was passed in 1992, wasto devolve democratic power to local authorities.Decentralization was considered a necessary element ina reform program to bring power closer to the people,to increase the range and authority of elected officials,and to improve the efficiency and accountability of theadministrative system. This agenda had widespreadsupport, since it appeared to be the only effectivemechanism for ensuring that the state would neveragain be able to abuse its powers as it had done in the1970s and early 1980s.

The new system has produced a fundamental change inthe institutional arrangements through which authorityhas been managed and services delivered (DANIDA,1995). The changes include:

• Transferring decision-makingauthority from administrators (theDistrict Administrator, centralministries) to elected DistrictCouncils;

• Granting autonomy to councilsbelow the district level;

• Reforming the planning anddecision-making process;

• Establishing a DecentralizationSecretariat, funded by DANIDA,with responsibility for organizingand co-ordinating theimplementation of the reformprogram;

• Financial decentralization in theform of block grants from centralgovernment to district councils.

The reform program is to be implementedin stages. A wide range of supportiveactivities are planned, such as training,production of information materials andfinancial management reform. Parallel tothe decentralization reform, a Civil ServiceReform Program was started up in 1990(Langseth 1995), and the two reforms havegradually merged.

The only component which has beensubject to a substantial review so far is theDecentralization Secretariat. A DANIDAmission states: "The decentralizationsecretariat has generally performed verywell since it started in 1992. It has mademajor contributions to decentralizationpolicies and played an important role intheir implementation." (DANIDA 1995b).Some results have also been achieved atthe ground level, most notably in the fieldof local level financial management. It isimportant to note, however, that thestarting level in this area was extremelylow and that enormous improvements willbe required. The report is relativelyoptimistic about the prospects of the otherelements of the reform program, althoughit states that a "successful" outcomedepends on a significant strengthening ofthe capacity of local governmentinstitutions in several areas.

61

Ghana

The system of local government in Ghana has gonethrough several phases since independence (Crook1994). Inherited at independence was a British systemof local government. This system was the subject ofintense conflict and substantial political manipulationduring the 1950s and 1960s. The present system ofgovernment in Ghana was created in 1989, with theapproval of the Local Government Act (Naustdalslid1992, World Bank 1993). The 1989 Act reintroduceddemocratic institutions at the local level, making theelected District Assemblies the highest body at thedistrict level. Formal political authority in the districtsis vested in the district assembly, which is responsiblefor development planning and a wide range of othergovernment services. Membership in the assembly isdetermined partly through local elections in singleparty constituencies and partly through appointment bythe central government.

The 1989 reform embodied an array of competingprinciples - representative democracy, GRASSROOTSpopulism, CDR (Committees for the Defence of theRevolution), managed democratic centralism, anddeconcentrated development planning. Thesecompeting principles resulted in a half-heartedimplementation of the decentralization system. Anassessment of the performance of the DistrictAssemblies (Crook 1994) points to the followingfactors:

i) Joint implementation of the structuraladjustment program and civil service reformshas led to a strong emphasis on cuttinggovernment expenditures.

ii) The continued political importance of thedeconcentrated state administration, togetherwith the CDRs, has meant that local politicalinstitutions have remained weak.

iii) The introduction of local elections in Ghanadid achieve some success in political terms.Popular enthusiasm for the assemblies whenthey were introduced was high, but has wanedas the expectations have not been met.

Ghana's experiment with decentralization has been amixed experience. It has not led to any significantimprovement in government capacity. But initially,political participation increased, and the legitimacy ofthe state (at least at the local level) appeared to beimproving. However, as a result of a lack of resources,insufficient local political autonomy and no supportfrom other important actors, these achievements havenot been sustained.

Ivory Coast

From the 1950s until 1980, localgovernment in Ivory Coast did not exist.Local elections were not held between1956 and 1980, and state institutions at thelocal level were under total central control.In the same period, the colonial prefecturalsystem of territorial administration wasexpanded and strengthened, making thecountry one of the most centralized inAfrica.

A reform process was started in 1980 andcompetitive elections were introducedwithin the one party system at central andlocal levels. A new policy of"communalization" was also announcedand new municipal authorities werecreated in the same year (Crook and Manor1992).

There are still major problems to beovercome in the Ivorian system of localgovernment. At the political level, threemain problems can be identified:

i) The role of the mayors has madethe communes an instrument ofdomination for local elites.Mayors have been accused ofignoring both the elected councilsand the local population.

ii) The level of political mobilizationand participation has been low.The lack of real decentralizationand the long traditions ofcentralized rule through theprefectural system, have beeninvoked to explain this apathy.

62

iii) The elected councils have remained in a weakposition vis-a-vis the administration. Both thedeconcentrated state administration and theline ministries view the local elected councilswith deep suspicion.

At the administrative level, there have been problemsof institutional conflict and bureaucratic co-ordination.The continued presence of the powerful and prestigiousprefectural system, exercising the authority of theMinistry of the Interior, remains an unresolved issue.

Central ministries, particularly the Ministry of Finance,have been strong proponents of decentralization, seeingit as a way of reducing the financial burden on thecentral government. By moving the financialresponsibility for rural development to the districts, it ishoped that substantial savings can be made. This"decentralization of costs" is not in the interest of localCouncils, since it means that they would lose the bulkof their income if central government grants areremoved.

South Asia

India - the State of Karnataka

The state of Karnataka, in the south-western part ofIndia, introduced a new system of local government in1985. The new system, which has been described as"one of the most radical in the entire third world",decentralized powers and resources from the state levelto the districts (Crook and Manor 1992). The newsystem also sought to increase political control overadministration by strengthening elected councillors vis-a-vis the administration.

The core institutions in the new systems, the DistrictCouncils, have no power to tax. They can raiseresources by making public investments and byborrowing, but they remain dependent on financialallocations from the state government.

The new system has had two main benefits (Crook andManor 1992):

i) Popular political participation has increased.This applies to a variety of forms ofparticipation, such as voting, activeparticipation in electoral campaigns,membership in organizations, contacting localpoliticians and taking part in meetings.

ii) The responsiveness of government institutionshas also been enhanced. This applies both tothe administration's responsiveness topoliticians and to politicians' responsiveness to

voters, where there has been themost marked improvement.

But problems remain:

i) Opposition among higher levelpoliticians and administrators.Civil servants and legislators onthe statelevel have been accused of takingdecisions on subjects that belongto District Councils and ofimposing their preferred policiesby financial and administrativecontrols.

ii) Insufficient local governmentautonomy. Although the formalresponsibilities of the DistrictCouncils were substantial, thestate government has beenreducing the Districts to mereimplementing agencies of policiesdetermined elsewhere.

iii) Distributional consequences of thereform. The District Councilshave been dominated byrepresentatives of local elitegroups, such as landowners. Thishas been the case despite theexistence of special quotas fordisadvantaged groups. InKarnataka, socially disadvantagedgroups have had considerablepolitical influence at the statelevel. Little knowledge exists onthe impact on the position ofweaker groups, when power hasbeen moved from a level wherethey have influence (the statelevel) to a level where they havenone.

63

Bangladesh

Bangladesh has implemented several decentralization"packages" during the last decades. One of the majordecentralization schemes was the upazila system,introduced in 1983. These reforms have not been verysuccessful. Ingham and Kalam (1992) write that thereis little ground for optimism about the outcome ofdecentralization policies in Bangladesh. Fieldworkcarried out in three different rural districts revealedwidespread dissatisfaction with decentralizationmeasures.

A survey revealed that government officials, electedrepresentatives, and the local elites respondedpositively to the idea of the upazila local governmentsystem. There was widespread frustration anddissatisfaction, however, about how the upazilaadministration was operating. The upazila institutionsand officers usually just respond to decisions andinitiatives taken at the central level. The centralgovernment rarely consulted - or invited suggestionsfrom - the upazila level, and the centrally madedecisions usually had to be accepted. Conflict, lack ofco-operation and corrupt practices have also beenmajor problems (Ingham and Kalam 1992).

The upazila chairmen are considered unaccountable tothe people. Use of their powers to gain materialbenefits and social prestige for themselves rather thanbenefiting the poor has been widespread (Khan 1987,Rahman 1986). The central government has beenaccused of "having shown only a marginal commitmentto power sharing and of having distorted andmanipulated decentralized institutions, in order to buildup a political power base in rural areas" (Ingham &Kalam 1992). This is reaffirmed in another study of theperformance of local government and NGOs in selectedregions, which concluded that the system did notfunction properly (Alam, Huque and Westergård 1994).The system had a high degree of central control, withlittle political will to implement reforms.

Decentralization is aimed at helping the poorestgroups, but a survey revealed that the majority ofrespondents had no detailed knowledge of the upazilaadministration. Few had attended meetings and therewas little access to information (Ingham and Kalam1992). The goal of increasing public participation andstrengthening democracy through decentralization hasnot been achieved through government reforms inBangladesh.

Sri Lanka

Significant changes have taken place in localgovernment and administration in Sri Lanka over the

last two decades. One of the mostfundamental and important changes is thedevolution of political and administrativepowers from the central to the provinciallevel. In 1988 Provincial Councils wereestablished under the authority of a ChiefMinister to carry out policy formulation,development planning, and financialmanagement at the Provincial level. Thecentral government has maintained controlover several key functions or areas ofresponsibility, such as highways and majorirrigation. Work linked to these areas isdone directly by the government's lineagencies (Departments, Boards andAuthorities), which have branch offices inthe provinces or the districts.

Divisional Secretariats were introduced in1992 as an additional administrative unit,with dual duty to co-ordinate tasksinitiated by the central government and theProvincial Councils.

The argument for these decentralizationreforms has been to provide local peoplewith a better opportunity to participate inlocal level planning and developmentactivities. Below the province and districtlevels, Local Councils (Pradeshiya Sabhas)have been set up in connection with theDivisional Secretariats.

The Sri Lanka decentralized politicalsystem is ambitious and complex, and itsimplementation has not been withoutproblems. First of all, there has beenconsiderable confusion regarding thedevolution and decentralization processamong ordinary citizens and publicofficers. This has been linked partly to thetransformation procedure and thedifficulties of getting people settled intonew offices and positions. Moreover, theinstructions and regulations about issuessuch as lines of command, authority,responsibilities, etc. have been unclear.The complex arrangements regardingownership, rights to use, and obligations tomaintain public assets are unclear. Thislack of clarity in responsibilities andauthority has reduced the efficiency of thelocal institutions (Dale 1992).

Latin America

64

During the last decade most of the countries in theregion have pursued a policy of decentralization tolower political levels (Winkler 1994, Murphy 1995,Wiesner 1994). There is still considerable reluctance,however, to surrender political and fiscal power tolocal political structures (Bidus 1995) and there aresignificant variations in national systems for financinglocal government in the midst of economic crises inmany of the countries (NACLA 1995). Most countrieshave undergone macroeconomic reform programs withrestructuring of the public sector as the most salientfeature. The movement towards decentralization seemsclear in most countries, but each country has shownunique experience both regarding the instruments usedand the pace of change.

The countries reviewed have developed their ownmodel of decentralization, varying from the centralizedstructure of Chile with central control by the state overthe municipalities, the federal governance structure incountries like Argentina, Brazil and Venezuela, to theincreased independence of municipalities in Honduras.In Central America the locally elected leaders have,through national and regional municipal associationslike the Federacion de Municipios del IstmoCentroamericano (FEMICA), efficiently brought theissue of decentralization and municipal autonomy tothe national political agenda (Bidus 1995).

Many of the reforms are linked to peace andreconciliation processes, and since they have only beenenacted two to five years ago, no conclusiveassessment of the benefits of these changes can bemade. There is, however, a clear vision among themunicipal organizations in Central America that theyhave an important role to play both in local andnational development. By 1996 all Central Americanmunicipalities will directly elect their mayors. Whilepolitical and fiscal autonomy is high on the CentralAmerican agenda, fiscal transfers play a critical role inthe decentralization drive in South America; in somecountries transfers amount to more than double thelocally generated tax revenue. Little is known,however, about the effects of transfers on thetransparency of the local political process, ongovernance in general, and whether the end result willbe a "simple fiscal decentralization model" or a"developmental decentralization strategy" (Wiesner1994).

Local revenue generation has historically been low inLatin America. The current financial decentralizationreforms include enhanced financial transfers fromcentral to local government, rather than an increase inlocal revenue generation (Winkler 1994). Taking themunicipalities’ share of total government spending asan indicator of the importance of the subnational

government as provider for public services,and the share and evolution of localgovernment's own source revenues as anindication of their fiscal autonomy, we canget an approximation to the degree offiscal decentralization in some LatinAmerican countries. The national taxrevenue in Chile declined in relative termsin the period from 1980-92 from 95.6 percent to 93.3 per cent while the revenue ofthe subnational level grew from 4.4 percent to 6.9 per cent during the same period.After the adoption of the new "PopularParticipation Law" in Bolivia, the ruralmunicipalities now control over 20 percent of the national budget. It is, however,necessary to look further into how theinter-governmental tax revenue andexpenditure is shared by the differentlevels of government, to be able to tellmore about the effects on resourceallocation efficiency, distributionaloutcomes and the degree of localautonomy over local expenditure comparedwith transfers.

In Latin American countries there is ageneral trend for the national share ofexpenditure to decline and subnationalshares of expenditure to increase. Transferstend to grow steadily, reducing thesubnational share of expenditure financedby their own tax resources. The politicalimplication seems to be "if localjurisdictions are able to export taxes andenjoy largely unconditional transfers, howcould local accountability, public sectormanagement and efficiency and equityobjectives be more attainable throughfiscal decentralization" (Wiesner 1994). In1991 municipalities in Brazil were onlyfinancing 29 per cent of their expenditurefrom locally generated taxes, while inChile there was a decline from 76 per centin 1970 to 60 per cent in 1992 (Murphy1995).

The fastest growth of expenditure is at themunicipal level and if the trend continues,municipal expenditure will exceed that ofthe middle levels (region, federal states).In most cases there has been a politicaldecision to enhance revenue transfers,before analysis and decisions are taken onwhich functions to decentralize and whatlocal institutional capacity should be built.This raises the classical problem with

65

transfers; they weaken the principle of correspondencebetween revenue generation and service delivery, withthe possible undermining of local tax efforts. Thetransfer of revenue from the state to the municipalitieshas increased considerably and the bulk of this increasehas been in the health and educational sector. Since1979 the Municipal Common Fund has redistributedresources between municipalities on the basis ofrelative poverty indicators.

There is a relatively weak correspondence betweenlocal revenue generation and local activities. Thismight reduce the total revenue of the state since it hasdiscouraged municipalities from raising their ownrevenues, as experienced in Guatemala (Bidus 1995).More than 41 per cent of all municipal revenues inCentral America are currently in the budgets of the fivecapital cities while the remaining 1,100 municipalitiescontrol the remaining 59 per cent (Wiesner 1994). Theredistribution process has significantly reduced povertyin the most deprived areas. In El Salvador the closecorrespondence between locally raised revenue andlocal project planning has shown that municipalprojects on average cost 45 per cent less than centrallyfunded and implemented projects.

In comparison, Honduras has been seen as leading theway in the region in terms of devolution of power,authority, and resources to the local level. The 1990Municipal Law and the electoral reforms have giventhe citizens greater local control and the possibility ofparticipating in local affairs. The Honduran MunicipalAssociation has been central in the reform process,leading to strengthening of local autonomy.

Studies of Central American countries show thatcitizens believe that the local government should havemore responsibility and resources, and that they aremore skilled at resolving community problems than thecentral government (Bidus 1995). Donor initiatedprograms devoted to supplying credit to infrastructureprojects promoted by municipalities on the basis ofmatching funds, such as the Municipal InfrastructureFinance Program designed by USAID, have been animportant confidence building supplement togovernment transfers. The Municipalities have takenon a more progressive political role in some of theCentral American states (NACLA 1995).

A study of decentralization in Colombia concluded thatthe traditional supply-driven donor support programsfor capacity building at the local government levelhave a poor track record (World Bank 1994). Theyargue that "sustainable development at the local level ispossible only when there is effective demand by localadministrations and communities." The report arguesfor a demand-driven approach, where technical

assistance follows local demand and istailored to local needs, and whereinformation exchange betweenmunicipalities on best practices can beencouraged to promote local leadershipthat will work for locally innovativesolutions, including improved communityparticipation such as user involvement inservice delivery boards.

Similar positive results have been observedfrom support provided by USAID over thepast two decades for the Peruviangovernment's decentralization effortsthrough IRDPs, disaster relief,rehabilitation and reconstruction (DRR)projects, and program development andsupport (PD&S) activities. The projectsfocused on the situation of the individualfarmer and used private contractors formuch of the work. This reinforcement ofthe private sector was an essential part ofthe projects. The projects helped the localinstitutions develop planning andimplementation capabilities that wereflexible, efficient, and responsive to localneeds (Schmidt 1988, 1989).

In Nicaragua a number of bilateral andmultilateral donors are supporting publicsector and decentralization reform.Decentralization is a key component of theNicaraguan Government's PolicyFramework Paper (1994-97) and of publicsector reforms. DANIDA is supportinglocal government capacity building inproject planning and implementation andgeneral administrative and planningcapacity. A recent review shows that theproject has achieved best results in theformer area, while it has been difficult toachieve general administrative andplanning capacity. (DANIDA 1995a)Decentralization in Nicaragua hasachieved considerable results, given theshort time-frame of the experience. Thereare, however, a number of problemsconfronting local governmentstrengthening: (i) there are no data andinformation on local governmentfinancing; (ii) a number of funds areavailable for projects at local level, buteach has its own rules and procedures,causing confusion at the local level; and(iii) there is no communication betweencentral agencies involved in

66

decentralization, which creates difficulties for locallevel decisions.

CONCLUSIONS AND LESSONS LEARNED

Transfer of Political Power - Impacts onGovernance

One of the general lessons that can be drawn from theassessment of decentralization reforms is that there isconsiderable ambiguity in the willingness to transferpolitical power and influence from the centralgovernment. Even when legal powers, functions andtasks have been allocated, adequate personnel andfinancial resources are not provided. An assessment ofthe current situation is as much about the impact offailed and muted decentralization as it is about theimpact of decentralization.

The failure to delegate or devolve powers is partlyrooted in a number of weaknesses of local governmentsand problems at the sub-national level. When designingand carrying out schemes for decentralization reforms,there is always a danger that conflicts along ethnic orreligious lines, or along other differentiationmechanisms, may emerge. In particular, conflicts mayarise when the reforms are also linked to control overscarce resources, employment opportunities, incomes,etc. Multi-party systems have increased thisvulnerability in many African countries, where thereoften are tensions between the party in centralgovernment and opposition parties that might bedominant in certain regions of the country (Zambia,Mozambique).

In many developing countries local councils have notestablished themselves as credible institutions forarticulation of local interests. People often tend toconsider them more as local agents for state power thanas institutions representing local interests. The lack oflegitimacy often expresses itself in low politicalactivity and low public participation at the local level.A different pattern seems to have emerged during thelast decade in Latin America, where pressure has beengenerated at the local level for decentralization andimproved local political authority after the fall ofauthoritarian regimes.

In some countries, national political leaders have useddecentralization schemes to try to avoid theresponsibility for the delivery of services by shiftingthe blame for poor performance to local authorities.(Silverman 1992)

• In practice, the transfer of real power to localauthorities is often more rhetorical than real;

• Control over funds and personnelat the local level is limited;

• The capacity of localadministrative institutions is low;

• The co-ordination of planning andimplementation of developmentprojects is inadequate.

Devolution of functions and tasks tolocally elected councils without theresources has had a negative impact ongovernance. Again, Central and SouthAmerica show a different picture whenincreased financial resources have enabledthe local council to deliver governmentservices more efficiently and governancehas improved.

Division of Roles BetweenCentral and Local Governments -Impact on Delivery ofGovernment Services

Most countries have dual systems of localgovernment, with a system of localauthorities, and a deconcentrated stafffrom the central ministries. There is oftentension between these levels ofdecentralization:

i) It is not clear which tasks andfunctions should be handledcentrally and what should be dealtwith at local level.

ii) Elected representatives at thelocal level are frequentlyoverruled by central governmentofficials, who in practice havemore power. Technical expertisemeans superior positions andprestige.

iii) Local horizontal co-ordinationamong the central government'sline agencies is difficult. Theofficers tend to fight for resourcesfor their department instead ofpromoting co-operation.

iv) The central government's officialshave a tendency to be moreconcerned with long-termeconomic projects, while local

67

representatives and the people are moreinterested in short-term social issues andprograms.

Recent analyses of decentralization in developingcountries stress the importance of vertical linkages.Agencies at the central level must be reorganized andreoriented to be in a better position to supportdecentralization. Proper decentralization also impliesreorganization of ministerial organization for servicedelivery. That is why decentralization reforms have tobe co-ordinated with ministerial reforms under thecurrent reform programs in developing countries.

When social services are devolved to local authorities,there are implications for lines of communication to thecentral ministries. In Botswana the responsibility forpublic health was decentralized to the local authoritiesand staff integrated under the authority of the localcouncils, i.e., under the authority of the Ministry forLocal Government and Communal Lands (Lauglo andMolutsi 1994, 1995). At the same time, the staffcontinued to report for professional matters to theMinistry of Health. The Ministry of Local Governmentwanted to build up the necessary competence in publichealth, but in small economies it is difficult to provideand sustain the necessary resources for capacitybuilding on health matters in two ministries. Achallenge for future devolution of the responsibilitiesfor social services to local authorities will be to workout proper arrangement of authority andcommunication between district councils and therespective ministries, including division ofresponsibilities and capacity building to take place inthe Ministry for Local Government.

Capabilities of Local Governments -Impact on Delivery of GovernmentServices

Inefficiency in local government often manifests itselfin an inability to implement policies and to use theresources available for their intended purposes. Thereis also no accountability; equipment and materials arefrequently diverted for private use, and authorities failto produce audited accounts. This condition may berelated to a lack of adequately trained personnel and toinadequate regulations and enforcement mechanisms. Itis also related to the structural relations between localand central authorities. There may well be alliancesbetween the central and local elites, and the localpopulation may not have the power or the resources tocontrol the elite.

Local governments' budgeting and planning models areoften inadequate. In several countries district planstend to be presented as aggregated 'shopping lists' made

up of suggestions submitted by villagersand district councils, as well as the centralgovernment's line agencies. (Sri Lanka,Tanzania, Bangladesh).The policy shift from attempting to controleconomic behavior through participation inthe production of goods and services toproviding an enabling environment forprivate sector production and investment isone of the most important elements of therecent decentralization strategies ofdeveloping countries.

It is important to distinguish betweenprovision and production. The currentideology is that government should limitits involvement to cover service provisionand establish an environment thatpromotes private economic activity andproduction. This may substantially reducemany of the problems related to inadequatecapacity at subnational government levels,but for local authorities to exercise theirresponsibilities well, they need to carry outessential management functions.

There is little systematic experience withdevolving government services to theprivate sector and to NGOs andcommunity organizations. The greatvariations in economic development andorganizational level of developingcountries make it impossible to providegeneral conclusions. The Africanexperience shows that there are fewpartners in the provision of social services.Nor is there a market for services in thetechnical infrastructure sector, whichmakes it difficult to set up private/publicinstitutions for road building, water supply,etc. This is an area were there is a need forsystematic studies of the currentexperience of privatization andpublic/private partnerships.

Revenues and Expenditures -Impact on the Economy andPublic Finances

The importance of local revenues andincomes has been pointed out in almost allreports on decentralization. Equallyapparent is the scarcity of financialresources that characterizes localgovernment institutions in general. Thefinancial aspects of decentralizationpolicies relate both to revenues and

68

expenditures. Many African countries do not havesystems or capacity for financial planning andbudgeting at the local level. On the other hand, thecurrent decentralization movement in Central andSouth America has improved the financial base andquality of the services delivered locally.

The exercise of effective discretionary authority bylocal governments depends on their ability to generatethe necessary financial and staff resources. If efforts tostrengthen revenue collection at the local level issuccessful, it results in significant redirection ofresources towards these areas.

Some analysts argue that provision of services andrevenue collection should be equally decentralized.There is a fallacy in this reasoning. For example, sometaxes are more suited for decentralization than others.When considering whether to decentralize taxes, policymaking should be guided by two fundamentalprinciples: efficiency, and fairness.

According to the principle of fairness, tax bases thatare unevenly distributed between local governments orregions are not suited for decentralization. Forexample, taxes based on natural resources shouldremain under the control of the central government.Import taxes or value added taxes, where the burden ofthe tax imposed in a given jurisdiction can be borne bytaxpayers established in another jurisdiction, are notsuited for decentralization. For efficiency reasons,taxes that can induce people or companies to moveaway from high rate areas to low rate areas are notapplicable for decentralization. It would lead tomisallocation of resources. The most typical examplehere is the personal income tax (WB & IIA 1990).

This logic partly ruins a much needed responsibilityand legitimacy mechanism for taxing. In a balancedsystem, the cost of taxing has to be compared with thesocial benefits of spending (the money collected). Theexpenditure level will be controlled and limited by thetaxes available. However, in an unbalanced system,where local governments will have to spend more thanthey collect in tax, how can they legitimate their needto do so, and what would be the mechanisms forcontrolling spending?

If it is accepted that the decentralization ofexpenditures is more desirable than decentralization oftaxes, then it must be concluded that transfers orsubsidies to local governments are necessary.Transfers or subsidies should therefore be considered asan integral part of decentralization policies andstrategies. This is also the reason why decentralizationis so important in most countries. A relevant questionis: what types of subsidies should be utilized, and

according to what criteria should they beallocated? The design of financialdecentralization should therefore be amajor component of all decentralizationprograms (WB & IIA 1990). Financingdecentralization is often treatedrudimentarily and is not dealt with asmeticulously as the political andadministrative aspects.

Donors' Support andInvolvement

Since the mid 1970s donors have beensupporting decentralization reforms andschemes in developing countries. Foreigndonors, however, cannot establish a well-performing local government indeveloping countries. Both strong politicalcommitment and existing capacity toimplement reforms are conditions forsuccessful reform programs. If theseconditions are fulfilled, donors could - iftheir programs are well designed andimplemented - act as catalysts for a processleading to improved local governmentperformance.

Donors should also be aware that bychannelling funds directly to thestrengthening of state institutions, theyinevitably take on a more political role.Projects aiming at institution building willby definition seek to improve the capacityof institutions, which in this case meansthe capacity of the state. When donorsprovide a substantial proportion ofgovernment funds, their support could bedecisive in determining the outcome ofinternal political struggles. Much of thecurrent aid programs actually have acentralizing effect on developing countries,especially in those countries where donorfunds make up large part of the investmentbudgets. Aid programs historically havestrengthened central governments in therecipient countries and orientedaccountability toward the external donorcommunity. This has implicitly weakenedaccountability to national and localpolitical constituencies. The supportivecapabilities of donors and the time-framefor donor involvement should be anexplicit concern in designing decentralizedprojects and programs (World Bank, 1993,UNDP 1993, Schmidt 1989).

69

Institution Building vs. SectoralAssistance

Foreign donors have historically been involved insupporting local government in three ways,

i) donors support sectoral projects to beimplemented directly by local authorities(health, agriculture, water supply, ruraldevelopment, etc., often through districtdevelopment programs),

ii) support is given to projects aimed specificallyat institution building. This includes supportfor the design of relations between local andcentral authorities, assistance in theimplementation of decentralization programsand support to reform programs and institutionbuilding at the local level.

iii) support is given to a combination of sectoralassistance and institution building; the donorsassist local authorities to implement sectoralprojects (whether donor funded or not).

In projects whose specific aim is to strengthen localgovernment institutions, the perspective becomessomewhat different. It makes no sense to bypassofficial channels in order to increase project efficiency,since a major purpose of the project is to makegovernment institutions more efficient.

It should be kept in mind, however, that thejustification for giving support to decentralization andinstitutional building is that improvement ofgovernment performance is seen as a necessarycondition for better service provision and efficient andlegitimate governance. One problem with such projectsis that it is very difficult to assess their impact with anycertainty. This makes it all the more important thatproject objectives are clearly stated, though notnecessarily in quantitative terms, and that projectactivities are designed in accordance with theseobjectives.

There has been a debate among donors about therelative importance that should be given tostrengthening government institutions as opposed tosectoral support. On the one hand, sustainable sectoralprograms presuppose fairly well functioninginstitutions, with sufficient capacity to take overproject activities after the end of the project period. Anargument could be made for concentrating on supportto improve government capacity.

On the other hand, given local governments' lack offunds and the donor dependence of many regions, a

reorientation of donor support fromsectoral programs to institution buildingcould leave local authorities with no othertasks than developing themselves throughdonor funded capacity-building projects.

Capacity building works best if institutionshave substantial tasks and responsibilities.The prospect of successful decentralizationand improvement of local governmentperformance will probably be enhanced ifit is combined and co-ordinated withsectoral support.

Assisting the CentralGovernment in theDecentralization Process

Firm support and commitment from thecentral government is a condition forsuccessful decentralization. The centralgovernment also has an important role inco-ordinating and implementing adecentralization program.

A central co-ordination unit may thereforebe required. This could be a division in theMinistry of Local Government (or itsequivalent) or an independent unit.DANIDA has funded such a unit inUganda and, although the program is stillunder way, the indications are that the co-ordinating unit has been a success. InTanzania the World Bank funded thenational secretariat for the Civil ServiceReform Program, including the secretariatfor the local government reformcomponent. In Zambia the ODA/UKfunded the decentralization secretariat.

The World Bank has raised the issue ofwhether the funding of reform secretariatsshould not be the responsibility of thedeveloping countries, with donorssupporting activities under the programs.In many of the least developed countriesthis is not an option. Whether there is aneed for a central co-ordinating unit mayvary between countries, but if there aredoubts about the capacity of the centralgovernment for funding such units, thenassistance should be considered.

The Role of Technical Expertsand Training

70

One instrument that donor organizations have used topromote institution building is the so-called expert-counterpart arrangement, in which foreign expertsoccupy positions in developing countries at aninternational salary level for a fixed period. Recentreports have concluded that such arrangements areexpensive and that they generate adverse effects as aresult of the enormous difference in salary levelsbetween the foreign experts and the local counterparts.Even more importantly, there is little evidence thatthey are effective as training arrangements.

As a consequence, many donors are now shifting to"twinning arrangements" or long-term arrangements forco-operation between institutions in donor countriesand recipient countries. This is an arrangement that, inprinciple, has several advantages, namely, greateracceptance of foreign "donor-side" personnel, whocome as fellow professionals with similar problems;and flexibility in the type and timing of assistance andthe possibility of long-term relationships. There are,however, a number of potential and identified problemsassociated with such arrangements: the number ofrelevant and committed donor country organizationsmay be limited; their knowledge may not be relevant ina developing country context; developing practicalarrangements specifying the role of the donor countryorganization could be difficult; and administrativecosts could be high.

Projects that aim to improve local governmentperformance must address the structural andinstitutional factors that influence performance. Lessfunding should go to traditional forms of aid, such asvehicles, equipment and study tours. There have beenseveral recent studies that have criticized traditionaltraining programs for capacity building (Moore 1992,Grindle and Hilderbrand 1995). These studiesrecommend efforts to build more political andorganizational cultures, which are more conducive toaccountability and transparency.

A demand-driven strategy where support is given as aresponse to local demands is recommended to promotelocal innovative and responsible leadership (WorldBank 1995). Such aid is much more difficult toprogram and implement than traditional interventions,and there are bound to be failures. But there are noreadily apparent alternatives if the intention is improveassistance to decentralization and governance.

Monitoring and Evaluation

Various donors are now active in supportingdecentralization, both through support to localgovernment reforms and capacity building. Donors alsosupport decentralization of services especially in

sectors like water and sanitation andprimary health care. Guidelines are beingdeveloped for (i) designingdecentralization programs and training(Smith 1993), and (ii) assessing localinstitutions (Therkildsen et al. 1992).Evaluations by OECD countries over thecoming years should therefore to a greaterextent reflect on, and include in theanalysis, institutional issues and thedecentralization which is taking place inthe policy environment of the aidinterventions.

More recent studies have shown thatevaluations have little impact on aid, andthat the learning effect in the aidadministrations has been small. In someprojects efforts have therefore beeninvested in making the evaluations morerelevant, and tailored to the needs to makeinformed choices of adjustments andprogram changes in the intervention.

Receiving relevant information duringproject implementation seems to be amajor concern for the donors. Earlierevaluations of decentralization andinstitutional development have often beenhampered by poor monitoring andreporting systems during projectimplementation, as well as lack of anybaseline studies.

If properly designed, evaluations ofinstitution building, capacityimprovements and decentralization shouldbe of relevance to improve our knowledgeof how outside aid interventions cansupport and strengthen national and localpublic administration. A number of suchreviews and evaluations have already beenperformed (World Bank 1990a, DANIDA1995a, 1995b, SIDA 1989, 1990, 1991a,1993, 1995, UD 1993, 1995), and more areexpected as this type of aid assistanceincreases.

Efforts have also been made in the aidorganisations to discuss issues and topicsthat should be included in designing orevaluating decentralization and institutionbuilding (Moore 1994, SIDA 1991b, WorldBank 1990b, 1992, 1993, 1994).

UK/ODA in their support fordecentralization in Zambia has put

71

emphasis on proper monitoring and regular reviewmissions during project implementation. The LOGOSP(Local Government Support Project) secretariatproduce a comprehensive six monthly report, andoutside experts (researchers and consultants) areregularly called in to assess specific issues. No decisionhas yet been taken to have an ex-post evaluation ofLOGOSP. LOGOSP is also supporting thedevelopment of a system for local governmentperformance appraisal. These appraisal systems will beutilized by the district staff in close collaboration withLOGOSP, and will be closely related to the trainingwhich will take place in the districts (LOGOSP, 1993).

DANIDA does not attempt to make impact assessmentsof their support to decentralization in Uganda. Insteadthey regularly send expert teams to do evaluations oforganization and implementation of the program atregular intervals, to check progress and identifyproblems to be addressed (DANIDA 1995b).

The World Bank in their support to Civil ServiceReform Program and Local Government Reform inUganda, is making an attempt to develop a monitoringand evaluation system to measure impact of the reformon service delivery (Langseth 1995).

Using the evaluation to increase the ownership of thereform program has also been tried by some projects.An example is the World Bank "Governance Approachto Civil Service Reform", and their efforts to developan "Institutional Environmental Assessment" method.Instead of having an outside expert team doing theassessment, the various sectors of the publicadministration were themselves taking part in theassessment, with the aid of a facilitator or projectleader (World Bank 1994).

This was done in order to increase their knowledge,reflection and ownership of the problems identified inthe institutional environmental assessment, and shouldmake for less hostility to new proposals fororganizational changes, and more rapid implementationof recommendations. Organizing assessments andevaluations with the participation of a great number ofpeople was, however, not without its problems, soparticipatory assessments were therefore notrecommended as a substitute to expert evaluations andassessments, but as a supplement.

These considerations show that there are a number oftypes of evaluations, and that the selection of type ofevaluation should reflect the types of information andknowledge needed in the specific situation:

i) Expert evaluation (traditional ex-postevaluation)

ii) Process evaluation (includingreviews and information feedbackduring project implementation -important for pilot projects)

iii) "Problem oriented" evaluations,with selected issues (for exampleinstitutional issues, gender,poverty, etc.)

iv) Participatory evaluations, usingfacilitators to run participatory(monitoring and) evaluation(some-times using project staff asfacilitators, when the monitoringand evaluation is part of theproject itself).

Conclusions andRecommendations

First, the study demonstrates that thecontribution of decentralization toimproving democracy and equity ispromising, especially in countries with along history of nation building and abureaucratic history. However,decentralization has limited impact ongovernance in many of the least developedcountries, where decentralization is carriedout in a period of economic crises andstagnation and under considerable externalpressure.

There is considerable ambiguity in thewillingness to decentralize real power andresources from the central government inmany of these countries. This is partlyrooted in differences in interest, but also inweak capabilities and a number of otherproblems at the local level such as lack ofadministrative competence, weak planningand control systems, and lack of coherentlocal mobilization. In most developingcountries decentralization reforms areambiguous and often create confusion anduncertainty at the local level about therules and policies governing thedecentralization. Local governments andlocal population are often ill-informedabout current decentralization policies.

Second, the study concludes thatdecentralization has improvedmanagement efficiency and financialperformance in Asia and Central and SouthAmerica. However, most local

72

government systems, particularly in Sub-SaharanAfrica have been hampered by unclear authority-relations and roles, detailed central intervention, weakaccountability, and lack of funds. The potential at thelocal level for promoting resource mobilization,planning and management has rarely been utilized.

Third, the experience with donors' involvement indecentralization has been mixed. Donor support has toa large extent been focused on administrative structuresand they have not paid sufficient attention to politicalforces and processes, especially at the local level.When analyzing aid assistance and decentralizationreforms, all aid, not just the small proportion of aidgoing to local government strengthening has to betaken into account. Aid assistance, sector support etc.,has gone to strengthen central government institutions.Because of this, central government institutions havebecome less dependent on local groups and structures,and therefore less interested in establishing a dialogueand mutual co-operation with them. It is thereforeimportant to relate decentralization reforms to generalpublic sector reforms, including ministerial reforms.

There is a mixed picture on the performance ofdecentralization reforms, varying between countriesand regions. The material consulted for this paperclearly indicate that

i) Generalizations about all developing countriesdo not provide much insight or knowledge.There is a need to continue to promote countryand region based studies.

ii) It is difficult to measure the impact ofdecentralization on governance. Central andSouth America score high on pressure frombelow for decentralization, enhanced fiscaldecentralization, and improved governance.This circumstance might have been set inmotion by improved central governance,reflecting economic and political development(including the fall of authoritarian regimes).

There is a need for more systematic, comparative,studies of:

• Relations between decentralization reformsand other public sector reforms in developingcountries.

• Economic deregulation and the use of privatesector and community based organizations forservice delivery, i.e. local innovative solutionsfor service provision, including running andmaintaining service infrastructure.

• Financial decentralization andsystems of promotingaccountability at the local level.

• Decentralization and its impact onpotential regional and ethnicconflict. Systems for powersharing in divided countries,including decentralization andlocal governments in post-conflictsocieties.

• Institution building at localgovernment level and systems forcapacity building.

The studies could be carried out as jointeffort between several donors that aresupporting similar reforms and programs.As a minimal form of donor co-operation,reviews and evaluations reports should beactively distributed and used by otherdonors.

1

This paper was prepared by a team at the NorwegianInstitute of Urban and Regional Research on behalf ofthe Norwegian Ministry for Foreign Affairs.

73

HUMAN RIGHTS1

INTRODUCTION

This study, undertaken within the framework of theDAC Expert Group on Aid Evaluation's researchprogram on Participatory Development and GoodGovernance (PD/GG), synthesizes experiences in issuesof human rights. Its purpose in doing so is to providesubstantive lessons with three central themes:

• donor strategies intended to support theimplementation of human rights;

• the management of donor's support for humanrights;

• the evaluation and monitoring of human rightsinterventions.

Desk research forms the basis of this study and thematerials that have been reviewed can be classified asfollows:

• documents from evaluations, by donors, ofhuman rights projects and programs;

• research papers on good governance,democratization and human rights indeveloping countries;

• scientific publications dealing with humanrights, institutional change and relationsbetween state and society in developingcountries (a limited number).

The most important sources were documents from thedonor's evaluations; unfortunately only a few wereavailable. Most donor agencies have started to evaluatehuman rights activities fairly recently and this, togetherwith the principle that the issues must be treated withgreat discretion, may have restricted some disclosures.

Issues surrounding the implementation of human rightsare complex and much debated; there is a wide varietyof opinion about the contribution to be made by donors'interventions. Historical context is important and thetime for "universal and timeless recipes" has not yetcome. These caveats should be kept in mind whenreading this summary.


Human Rights

Defining, conceptualizing andimplementing human rights activities havebeen central to the UN system for morethan forty years. This study concentrateson the International Bill of Human Rights,which comprises the UniversalDeclaration of Human Rights, theInternational Covenant on Civil andPolitical Rights, and the InternationalCovenant on Economic, Social andCultural Rights. Other instruments dealingwith human rights have been taken intoaccount as far as possible.

Human Rights, Democracy, andDevelopment

Current political thinking that links humanrights, democracy, and development co-operation should be understood in itshistorical and theoretical context.

In the past, theories that have related aid todemocracy and human rights did so onlyindirectly. A theoretical chain linked aidwith growth (growth theories), growth withgeneral prosperity (trickle-down theories),and socio-economic prosperity withdemocracy (theories about economicprerequisites for democracy). Theseconcepts were derived mainly from thehistory of democracy in the West, wherehuman rights emerged as an integral partof the democratization process. During theSecond UN Development Decade (1971-80) these concepts were slowly modified.First, the basic-needs, poverty-orientedstrategy combined growth and generalprosperity under the concept "growth withequity". It was generally understood thatdevelopment co-operation enableseconomic and social rights to be realized,thus suggesting that development work andthe realization of these rights are almostidentical. Second, issues of civil andpolitical rights came slowly to the surface.During the 1970s and 1980s, few donors oragencies pursued policies that activelyfostered civil and political rights, eventhough many NGOs, and so-called politicalfoundations, did.

Since the end of the 1980s, developmenttheories have emphasized the need forhuman and democratic rights to go intandem with economic growth. Some go so

74

far as to argue that democracy is the parent of growthrather than the other way round. This thinking has nowbecome current in the donor community and isreflected in a 1989 statement of the DAC aid ministers:"There is a vital connection, now more widelyappreciated, between open, democratic andaccountable political systems, individual rights and theeffective and equitable operation of economicsystems." (OECD, 1989). Acceptance of this principlewould mean that aid must be directed simultaneously ateconomic development, human rights and democraticdevelopment.

External Support for the Implementation ofHuman Rights

In using the term external support, this study, while notexcluding persuasion and pressure as forms ofintervention, is chiefly about technical and financialassistance. Peace-keeping and humanitarian aid areexcluded. Implementation is used becausestandard-setting, monitoring and the supervision ofhuman rights, by the United Nations and otherinternational or regional Human Rights fora, are alsoexcluded.

Donors, agencies and NGOs can support and influencerecipient governments and indigenous NGOs in threemain ways: by technical and financial assistance, bypersuasion and by pressure.

Technical and financial assistance, for governmentsand NGOs, can help to build the capacity to design andimplement reform programs and to defray their costs. Itcan also help to promote human rights, (a) byintegrating human rights issues into mainstreamdevelopment projects, (b) by projects and programswhich specifically support human rights and (c) byorienting overall development programs towardsreform.

Persuasion, which must include deepening mutualunderstanding between both partners, is intended toinfluence the development of the recipient's approach.Its means range from international conferences throughbilateral and multilateral policy dialogue to informalmeetings and private conversations. Persuasion canalso take less direct forms such as joint research orpublic debate in journals and other media.Pressure may take several forms and can be exercisedthrough various channels. A wide variety of measuresare available to donor governments, ranging fromdiscrete confidential demarches, through publicdeclarations to withdrawing personnel and imposingvarious sanctions. This study considers one particularform of pressure - conditionality.


Human Rights and MainstreamDevelopment Programs

Efforts to address the relationship betweenhuman rights and development co-operation have often led policy-makers,legal experts and development workersinto a labyrinth of definitions andconceptual knots. This labyrinth can beillustrated conceptually, institutionally andoperationally.

Concepts

Human rights and development assistanceare sometimes dealt with as two distinctspheres, one in which some actors lookafter human rights while others areresponsible for development. This divisionis reflected in discussions of "growth withequity" where it is generally understoodthat development co-operation enableseconomic and social rights to be realized.Those whose concepts arise from the UNDeclaration on the Right to Development(1986) claim that human rights anddevelopment are not divisible, but thatdevelopment should properly be seen as asubset of human rights. For them,development co-operation has nolegitimate task beyond the achievement ofhuman rights.

A related issue is often discussed under thecatch-phrase "human rights versus basichuman needs". Some agencies interprettheir poverty reduction programs(employment, public services, targetedtransfers, social safety nets and so on) astheir contribution to the realization ofeconomic and social rights. Othersmaintain that development programs thatrest primarily on a perception of basicneeds subtly reinforce the powerlessness ofrecipients. A human rights approach todevelopment, from this point of view,should enable even the most marginalizedand powerless of people and groups tomake a legal claim against the state. Aclaim would be considered legal insofar asit is sustainable under international, aswell as national, law and practice.

75

A few preliminary conclusions may be drawn from thedocuments reviewed for this study:

• the implementation of economic, social andcultural rights is not identical with therealization of basic human needs. It alsoencompasses respect for and the protection ofindividual freedoms and liberties;

• the implementation of economic, social andcultural rights is not identical with a "welfarestate" program. In many areas the state shouldconfine itself to creating general conditionsthat allow individuals and social forces thefreedom to develop initiatives to meet theirparticular needs. At the same time, states areobliged to alleviate the distress of those forwhom no minimum conditions can beguaranteed without state assistance;

• the implementation of economic, social andcultural rights implies not only legislativemeasures but also, to a large extent, institutionbuilding activities.

Institutions

Members of the DAC are increasingly putting theprotection and promotion of human rights among thepriorities in their development co-operation, at least interms of their declarations of intent. There aresubstantial differences between them, however, overwhich activities should be included under the heading.Should all development co-operation be described assupport for human rights? Should humanitarian aid andsocial development be included? Should all women'sprojects be so described?

Only a handful of agencies have tried to incorporatehuman rights into their operational guidelines andadministrative procedures. One reason for this is thatmany actors involved in this field (lawyers, diplomats,officials in international organizations) do not knowhow to tackle socio-economic and institution buildingproblems. Development workers and officials in aidagencies, for their part, do not know internationalhuman rights law. Observers agree that, in addition tobeing adopted into the programs of aid agencies,economic, social and cultural rights have to win a placein the national policies of developing countries.

Another factor affecting human rights institutions is theabsence of effective bridges between bilateral andNGO aid agencies on the one hand, and theinternational organizations responsible for settingstandards and monitoring economic, social and culturalrights on the other. To a large extent, the work, among

others, of the UN Committee onEconomic, Social and Cultural Rights, theUN Sub-Commission on Prevention ofDiscrimination and Protection ofMinorities, and the Special Reporters, isnot known, or is ignored by aid agencies.There have been apposite suggestions forintegrating the proposals of these UNbodies with the decision-making processesof the aid agencies (country strategypapers, policy dialogue meetings and soon). No feasibility studies of the practicalimplications of these suggestions werediscovered in the research for this report.

Operations

The principal interventions in support ofthe rights of vulnerable and disadvantagedgroups (women, children, indigenouspeople, landless peasants, migrant workersand others) are addressed both to thegroups and to the wider society where theylive. They can be summarized thus:

• humanitarian aid - material andfinancial (sometimes provided asemergency aid);

• empowerment - awareness raising,organization, resource building,income generation, human rightstraining;

• legal enforcement - the juridicalapproach based on professionallegal knowledge, access tomagistrates, courts etc;

• social enforcement - organizedcollective action by which publicagencies and private power-holders are made aware of, andpushed to concede, the legalclaims of deprived groups;

• public and political enforcement -advocacy and lobbying to raisethe awareness of the generalpublic and of official power-holders in order to change theirattitudes to vulnerable groups andto remedy deficiencies in theexisting political and legal system(law reform, economic policies).

76

Experience has shown that these approaches, whileuseful as isolated interventions, should be integratedinto an overall strategy to improve their effectivenessand to strengthen their sustainability. Theempowerment of vulnerable groups is often a necessaryprecondition for legal, social and political enforcement.Public and political enforcement are preconditions forthe successful application of legal and socialenforcement. Humanitarian, emergency andrehabilitative aid can be used to counteract the negativeeffects of human rights struggles, such as, for example,loss of seasonal work, dismissals, cancellations oftenancies or discriminatory practices by governmentofficers. All this may seem self-evident, but practicehas proved difficult. One of the main difficulties informulating and applying integrated strategies lies inthe compartmentalizing of the decision-makingprocess. In place of separate agencies that concentrateon achieving agency objectives (for example,humanitarian or legal assistance, or incomegeneration), all agencies need a common focus on thespecific situation of the group they are setting out tohelp. This calls for effective inter-agency attention andco-ordinated decision-making.

Specific Human Rights Projects andPrograms (Positive Measures)

The main goal of human rights projects and programsis to strengthen respect for civil and political rights.They often work in combination with efforts topromote and to support democracy in developingcountries and are usually referred to as "positivemeasures" to distinguish them from "negative"conditionality. They also cover a very broad area,from, for example, strengthening the organizations ofcivil society, through assisting the legal, judicial andexecutive systems, to supporting the processes oftransition.

There is evidence that the impact of human rights anddemocratization programs depends on three interrelatedand interlocking processes: phases in politicaldevelopment, phases in economic adjustment programsand phases in ethnic, nationalistic or religious conflicts.As yet, emerging lessons relate only to differingpolitical systems.

Positive Measures in Relation to DifferentPhases of Political Development

A major finding of several different studies is that, to aconsiderable degree, human rights projects andprograms are relevant to, and feasible in any country,depending on the phase of its political development.The lessons are summarized below.

Authoritarian and Semi-AuthoritarianSystems

In authoritarian and semi-authoritariansystems, support for human rights anddemocracy is only possible if there arechannels, or "niches" of entry. This meansthat the donor's field offices and embassiesshould actively identify them.

A regional or sub-regional approach mayhelp to create awareness, even inauthoritarian countries. Several programsof international non-governmentalorganizations (INGOs) are based on theconcept of contributing to the emergenceof a global civil society, a continuum ofdemocracy and civil society players frommunicipal to global levels. These INGOsare global information networks. It isargued that for human rights work to beeffective, especially in authoritariansystems of government, it is bestconducted above or below the nationallevel, that is, regionally, or at local andmunicipal levels. Networking, byelectronic and other means, is a method oflinking the two. By supporting programs ofthis kind, the bilateral, government-to-government route can be avoided.

Countries in Transition to Democracy

Building democratic institutions is a long-term process that must aim to transformcomplex patterns of institutionalizedbehavior. There is a distinct probabilitythat democratic changes will be effected,in the short run, more by political eventsthan by external support. This does notpreclude the possibility, however, that suchsupport will have an influence in the longrun.

During the transition phase, externallysupported projects usually concentrate oncivil society. This is justified since theorganizations of civil society have often, atbest, been neglected and, at worst,frequently controlled by the state, asituation which leaves vulnerable groups inurgent need of support. At the same time,the political process should be approachedmore actively. The most important aspectof institution building for democracy isthat the substance of democraticdevelopment, not just the forms or outward

77

appearances, should be transmitted to the hostcountries.

Formal Democracies

A problem in many countries that are formallydemocratic is that political actors, government partiesand parliament may lose credibility while,simultaneously, continuing to control the politicalsystem. Under these circumstances, aid for civil societywill have little impact unless it is structured to addressreform of the political system. Care must be taken,however, not to favor particular parties or politicaltrends.

Structural adjustment and economic modernizationmay impede efforts to realize constitutionally andlegally secured rights in newly established democraciesand may weaken state capacities to enforce laws and sodeadlock democratic institutions. For instance, theprivatization of public property, the deregulation ofindustrial and labor markets, free trade zones and thecommercialization of communal land, may violateeconomic, social and cultural rights. Thus, western-oriented political and economic policies may facilitatethe establishment of formal democratic institutions,while simultaneously weakening the political-economicfoundations for the enforcement of human rights.

Positive Measures Supporting SpecificInstitutions and Organizations

There are lessons to be learned about support fromdonors for the legislative and executive bodies, thejudicial system, and civil society institutions -independent media and human rights NGOs, forexample. Civic education programs in preparation forelections are crucial to success.

Support for the Legislative System

Governments in many countries have, for the first time,achieved legitimacy by gaining office as a result of thefirst or second national elections, in which theyinformed the electorate about their rights and theirability to participate.

Countries in transition to democracy need support, andnot only in election monitoring. This means that donorsmust ensure that preparatory missions have taken place,that the laws governing elections have been analyzedand, where necessary, adjusted.

Support for the Executive Branch

Support for the executive branch should help guaranteeefficient government services delivery for the people

and should, to this end, promotetransparency, sound management, theeradication of corruption and respect forhuman rights from the army and the police.

Areas for co-operation with state agenciesshould be carefully selected to preventsupport for those institutions that lack aserious commitment to reform. If stateagencies are supported, local NGOsworking in the same sector should, ifpossible, also be supported, so as toengender constant pressure for reform onthe state institutions concerned.

Decentralization is an important mattersince it often has a direct link todemocratization, that is, wide participationin local democratic institutions. There is apossibility, however, that decentralizationwill result in localized discrimination iflocal society is controlled by exclusive andself-serving elites. Similar problems canalso arise between different regions in acountry if decentralization is notaccompanied by appropriate financialsupport.

Support for the Judicial System

Evaluations offer little evidence that donorprograms have had a significant positiveimpact on judicial systems or on publicattitudes about their fairness. In general,ambitious goals for the reform, ortransformation, of judicial systems havesimply not been met. From this fourlessons can be identified.

First, support for a judicial system shouldonly be undertaken if the government, theleadership of the judiciary and importantelements within professional organizationsare all seriously committed to reform.Second, projects should not be directed atmarginal areas like the minor codes or atmerely providing administrative andtechnical aid; instead they should bedirected at key problem areas in theperformance of the system. Third, supportfor projects intended to reform the judicialsystem should always be accompanied byan analysis of the possibilities for "de-judicialization"; that is, a consideration ofthose areas that could be removed from thesystem and taken on, instead, by socialservices agencies and NGOs. Fourth, a new

78

institution, like that of an ombudsman for humanrights, needs time to develop a record of performance.

Support for Independent Media

In countries where the freedom of the press facessevere threats, the political denunciation of the lawsand regulations in question is appropriate. In countrieswhere democratization has begun, it is important tocontribute to the drafting of laws which guarantee andpromote freedom of the press. In countries with a moresatisfactory situation, the focus should be on theelaboration and practical application of professionalcodes.

Support for Human Rights NGOs

It is important to ensure complementarity and co-operation between governmental and non-governmental organizations. When donors target anddesign support programs and establish reportingsystems, it is essential that they understand how NGOsfunction and the different circumstances in which theywork.

The ways in which human rights NGOs work depend,in part, on their own strategic choices. Those that useco-operative, juridical and non-participatory strategieshave reliable relationships with the authorities, operatein bureaucratic institutions and evade unpredictablegrassroots initiatives. Their courses are predictable, andbureaucratic competence and specialized legalknowledge are important to them.

On the other hand, the most unpredictable results andsituations are generated by NGOs that useconfrontational, non-juridical and participatorystrategies. These produce dynamic power struggles,operate in informal settings and follow a process-oriented way of preparing, implementing andevaluating initiatives with the target group. NGOs ofthis type value field-oriented competence, motivationand imagination, courage, strategic sense andfamiliarity with a broad spectrum of everydayproblems within the target group.

Persuasion and Pressure

Persuasion

Persuasion is an effort to convince the targetgovernment (or the anti-reform elements within it) thataltered policies are in its own best long-term interests.It may include combined efforts to recognize anddefine new problem areas; to search for, analyze andevaluate the consequences of alternative solutions(joint problem-solving); to implement those that

balance the benefits of both parties and toclarify their respective roles in theframework of the partnership.

Persuasion is demanding in terms of time,knowledge, experience and commitment.If it is undertaken over too brief a periodand without commitment, it is likely to failor simply to result in compliance charades.In comparison with financial aid or withpunitive measures, however, it can be verycost-effective and, if successful, can resultin increased commitment to, andownership of, reform programs.

A synthesis of the lesson learned,therefore, could be extremely valuable inaddressing, for example, two questions:

• Under what external conditionsand in which issues is persuasionlikely to be feasible and effective?

• What are the essentialpreconditions for persuasion?Among them could be timing,staff experience, commitment andcredibility.

Surprisingly, apart from a few remarksabout the lack of agency staffcommitment, the documents reviewed forthis study do not provide substantivelessons.

Conditionality

Two generations of conditionality can bedistinguished. The first, propagated by theBretton Woods Institutions since the early1980s, is related to structural adjustmentprograms that have, as their primeobjectives, administrative reform, budgetbalance and market liberalization. In the1990s, aid donors have also increasinglymade official development aid conditionalon political reform within recipientcountries. The objectives of this secondgeneration of political conditionality are topromote democracy, the rule of law,human rights and good governance(accountability, transparency andpredictability).

Emerging lessons relate to the normativeaspects (legitimacy) as well as to the

79

instrumental aspects of political conditionality.

Normative Aspects (Legitimacy)

The principle of not intervening in the internal affairsof other sovereign states, although still broadlyaccepted, has been weakened. Recent developments,including the conditionality debate, have contributed tothis process. For example, the principle of non-intervention does not distinguish between a regimebased on control by means of coercion and one basedon popular consent. Because of the importance that isnow accorded to popular participation in governmentaldecision making, donors are increasingly consideringinternal legitimacy as a factor when designingconditionality measures against authoritarian,repressive, or self-serving regimes.

Instrumental Aspects

Practical experience shows a mixed record of successwhen conditionality is imposed. It is quite possible,however, that the main effects of politicalconditionality will emerge slowly. If donors follow uptheir rhetoric in future aid allocations, then aid flows toauthoritarian and repressive regimes will drain away.The governments of these regimes may then prefer toadapt to the demands made on them by donors andtheir own citizens, rather than to confront them.

Some general propositions have been indicated in theuse of power plays and brinkmanship. In order to assessthe risks and the probability of prevailing, donorsshould take into account:

• the internal power base of the recipientgovernment;

• the recipient government's ability to useexternal intervention to strengthen its popularsupport;

• the extent of the recipient country'sdependence on aid;

• the scope and relative importance of thebilateral relationship;

• the probability that a unilateral action mayhave a snowball effect;

• co-ordinated action by several donors;

• possible negative side-effects;

• multiple conditionality.

Careful attention should be given topossible links between political andeconomic conditionality, especially incases where both are applied by differentdonors (or groups of donors) or multilateralagencies. Political boundaries to economicconditionality become apparent at thepoint where the aid-receiving regime feelsthat the continued implementation ofeconomic reform entails more politicalcosts and risks than would non-implementation. Economic boundaries topolitical conditionality are reached whenthe donors (or groups of donors) feel thatcontinued insistence on politicalconditionality (for example, ondemocracy) jeopardizes their owneconomic interests. Boundaries may alsobe created where the economy of therecipient country is unable to sustain thepolitical reforms on which aid is madeconditional. For example, it is uncertainwhether democracy and human rights canbe sustained in conjunction withwidespread poverty. The issue, therefore,is not whether donors should encouragerecipients to be democratic or to observehuman rights, but that they should bear inmind economic imperatives and, ifnecessary, provide substantial assistance.

Managing Donor Support

Human Rights Project and ProgramManagement

The emerging lessons refer to thefollowing themes: policy formulation,programming and program management,project design and channel selection.

Policy Formulation

Experiences with the formulation ofhuman rights policies vary considerably. Itis argued that there is a strong case forachieving greater clarity in goal-settingand more consistency between broad goals,time-frames and human rights capacitybuilding. All this could be facilitated byshort and clear country or regional humanrights policy statements such as a "countryconcept" or "country paper".

It is alternatively argued that policystatements should be diffuse, especially for

80

purposes of approving projects. Then, during projectdesign, a broad interpretation, reflecting the conditionsof human rights in individual countries, could be made.This would also allow field offices or embassies tokeep a low profile in the human rights arena and woulddiminish the likelihood that the donor is seen to beinterfering in the affairs of the recipient country.

Programming and Program Management

In the past, because many agencies preferred low risksand modest investments, initiatives in human rightsprogramming have been reactive rather than pro-active.Now, there is a growing recognition of the need forlong-term approaches. Democratization, thestrengthening of civil society and of institutionalstructures to allow for human rights protection, arehuge and long-term challenges. Because of the rapidpace of political and other changes in developing aswell as in industrialized countries, and the difficulty ofpredicting such changes, there is skepticism about thefeasibility of long-term approaches. Few programs arebased on making long-term commitments to majorpartners.

When choosing between reactive and pro-activeprogramming, agencies should take into account thefollowing considerations. If their policy is to supporthuman rights and democratic development in a givencountry, and if it is recognized that democratizationand institutionalization of both the rule of law andregimes for the protection of human rights may takedecades to take full effect, then reactive programmingseems to be inappropriate. A pro-active approachwould be more instrumental. Alternatively, if overallpolicy goals are less clear, less fixed and based onuncertain commitment, then reactive programmingwith loose funding frameworks, allowing forflexibility, adaptability and opportunistic initiatives,will be the rational response. It will not, however,address central issues, but will simply indicatesolidarity with certain organizations and provide themwith limited assistance. Expectations of results shouldbe scaled down accordingly.

If an agency's programming is essentially reactive, thenthe need for information about human rights is limited.In considering applications for support, the mostimportant consideration will be the characteristics ofthe applicant organization: mission goals and program,governance and the composition of its board, thecommitment and competence of its staff, its degree ofautonomy from, and its degree of acceptability to, thehost government.

If, however, an agency has decided to move to morepro-active programming, then it is faced with a need to

invest in (a) a range of cost-effectivemethods for undertaking rapid analyses ofthe contexts of human rights (base-linestudies), and (b) simple dynamic modelsfor the analyses of interactive processes tobe used for monitoring purposes.

Project Design and Channel Selection

When designing projects and selectingchannels, a set of medium-term andmiddle-range objectives that will providerealistic points of reference, should beformulated. Grandiose goals will makeevaluations difficult because progresstowards them will be dependent on toomany external factors. In addition to clearand realistic goals, unambiguousperformance indicators are needed. Theearly identification of such indicatorspermits a systematic assessment ofeffectiveness and of impacts.

In repressive, authoritarian systems,assistance should, in principle, bechannelled through local NGOs as they area key fulcrum for change. Reportingrequirements should then be very flexible,partly for security reasons because localcounterparts are often hesitant to giveinformation, and partly because they mightnot be able to collect or collate theinformation required. Under theseconditions, donors should recognize thatlocal organizations often need tocamouflage their actual strategy withneutral and vague objectives. Withindemocratic systems, combinations ofsupport for governmental and non-governmental organizations are thought tobe much more effective. Standards forreporting in these systems should bedefined more stringently.

The impact of human rights interventionscan often be increased by clustering. Forexample, electoral support should not onlyconcentrate on the government'sorganization of elections, but also on theNGOs' ability to mobilize the populationand to train potential voters. Clusteringincreases the aid agency's commitment in agiven sector, but it may also increase therisks to the agency by raising its profile. Inthose countries where a donor supportsmany interventions, clustering may call

81

for, or lead to, the development of a country strategy.

Aid channelled through INGOs may benefit from therelations and influence these organizations have in alarge number of countries. In addition, INGOs thatsupport human rights often have the comparativeadvantage of being both professional and neutral. Localhuman rights groups may be given credence and,eventually, protection by virtue of their associationwith a recognized and respected internationalorganization. The views of local NGOs should be takeninto account when choosing an INGO, because it mustfirst be established that the INGO has sufficientcredibility in the country or region.

Donor Capacity Building for Human RightsSupport

Many agencies are still in the process of improving thecapacity of their staff to analyze and monitor humanrights, good governance and political participation.Special issues like the demobilization of troops, theredesign of civil-military relations or public security,are areas in which only a few aid agencies and foreignministries have extensive experience and expertise.

Because most agencies are now facing limited, or evenreduced, funding, it is necessary to identify existingcapacities and to specify objectives for strengtheningthem.

Key decisions about human rights programs will haveimplications for resources. Therefore, when assessingneeds for capacity building, particular attention shouldbe given to:• the anticipated mode of programming

(reactive or pro-active);

• the costs and benefits of different forms ofpresence in the field;

• the anticipated intensity of relations betweenthe donor and the implementing agency;

• the role and responsibilities of theimplementing agency.

Institutional Frameworks for Policy Dialogue

Although the documents reviewed for this study do notprovide a comprehensive analysis of institutionalframeworks for policy dialogue or of their comparativeeffectiveness, a few lessons about issues of co-ordination between different actors have beenidentified.

Co-ordination between Departments ofthe Donor Government

No human rights intervention is purelytechnical. The design of programs inhuman rights, and the timing of theirannouncements, carry a clear politicalmessage. Close co-ordination between aidagencies and the respective departments ofForeign Affairs is crucial for strategicprogramming in human rights anddemocratic development. Coherencebetween the aid agency and otherdepartments of the donor government isalso essential for an effective presence inthe host country.

Co-ordination Between Donor Agenciesand NGOs

Donor agencies must clearly articulatetheir human rights policy, analysis andstrategy to their multiple stakeholders;stakeholder's judgements should also betaken into consideration. When preparingdialogue meetings on bilateral policy, it isimportant to involve Northern (and, wherepossible, Southern) human rights NGOs.Donors, however, must ultimatelydetermine their own policy and not have itdetermined by the NGO agenda.

Bilateral Policy Dialogue

There are three possible links betweenhuman rights projects and policy dialogue:

• policy dialogue can be used toinitiate new human rights projectsand programs, but experienceseems to indicate that this methodis rarely used;

• policy dialogue can be used as ameans to legitimize human rightsactivities. Even small humanrights projects provide foreigndonors with concrete knowledgeof the local situation and,eventually, the opportunity toinclude human rights issues in thedialogue with the host country.For example, discussing therelationship between the donorand marginalized groups duringpolicy consultations may be

82

instrumental in protecting such groups fromadverse interference;

• problems identified in the project can befollowed up in dialogue. Aid agencies do notalways follow up on problems identifiedthrough their human rights programs withinterventions at policy level. Yet withoutchanges in policy, project results may be lostor undercut. The impact of human rightsprograms may be reduced because ofnationally and internationally adverse policyconditions. Nationally, for example, theimpact of a program supporting marginalizedindigenous groups or community-based co-operatives may be limited by land titleproblems. Internationally, an example is thenegative effects of Structural AdjustmentPrograms on some human rights programs.

The main conclusion is that a neutral, if not favorable,policy environment is important for consolidating theshort-term results of human rights projects. Projectsshould be designed to take this into account, and policydialogue should involve other donors as well as thehost government.

Co-ordination between Donors

Lack of co-ordination between donors causes confusionand wastes limited resources. It also sometimes permitsanti-reform groups in the recipient countries to exploitdiffering perspectives among donors in order to diluteor evade reforms.

As yet, there seems to be little agreement about whichinstitutional frameworks should be used for co-ordinating donors in human rights programs,democratization and good governance. This issuecannot be separated from the question of whichfunctions should be attributed to a multilateral ratherthan a bilateral framework of policy dialogue. Anotherunanswered question is: when should pressure(conditionality) be applied? Many aid administrationshave established bilateral relations with a number ofrecipient countries on the basis of country programsthat advance through an annual process of consultationand negotiation. Objectives are set in continuingbilateral dialogue. In this context issues known to becontroversial and that may have a negative influenceon development co-operation are likely to be avoided,or, at best, addressed in very general terms. This mayexplain why bilateral donors often use a dual strategy:supportive (positive) measures are brought to the forein the bilateral dialogue, while pressure and negativeconditionality are left to the multilateral fora. Thus,smaller donors as well as donors with vested strategic

or economic interests in given developingcountries can enjoy a free ride. If,however, this results in acompartmentalization in which multilateralfora are mainly used for pressure andconditionality, and bilateral frameworksfor the discussion of positive measures, thelinks between them may be lost.

Evaluation of Human RightsProjects and Programs

Because aid agencies have started toevaluate human rights activities fairlyrecently, the practice of drawing lessons isnew. An analysis of the availabledocuments shows that evaluators use awide range of disciplines, terminologiesand approaches. For example, the samehuman rights topic may be analyzed inlegal, sociological or political terms.Human rights institutions may also beexamined in a number of different ways:administrative capacity might be assessed,so might the policy environment,stakeholder or public choice analysesmight be carried out, and so on. Since nogenerally accepted framework for theanalysis of human rights projects andprograms is available, the comparison andintegration of conclusions is difficult, ifnot impossible.

Nevertheless, three broad approaches toevaluation can be distinguished:

• assessing project effectiveness;

• understanding human rights issuesby means of consultativeevaluations;

• assessing program impacts againstthe background of differentpolitical systems.

CONCLUSIONS AND LESSONSLEARNED

Assessing Project Effectiveness

Evaluators must assess the organization ofproject management, the effectiveness ofits institution building, improvements inthe human rights situation in the countryconcerned and the sustainability of the

83

project's results. A summary of the emerging lessonsfollows.

Difficulties in evaluating human rights projects springfrom unclear definitions of progress indicators and alack of base-line data. It would, therefore, be advisableto develop progress indicators at project level;however, the collection of base-line data often cannotbe justified for individual projects that are modestlyfunded, have few beneficiaries or little outreach.

Evaluators often try to analyze the links between asingle project and changes in the overall human rightssituation. This is an extremely difficult task because ofother variables. They should, instead, concentrate on amore manageable task and analyze and assess theconcrete links between the project and directbeneficiaries and other stakeholders. Depending on thescope of the project, such links could be studied at themicro- or the macro-level.

None of the evaluations reviewed seems to take intoaccount the programming of other agencies and theimpacts of their projects. It is necessary for donoragencies to co-ordinate with one another in carryingout analyses of human rights, base-line studies andparallel and combined evaluations.

Apart from these general issues, there are specificproblems in assessing the effects of interventions inhuman rights. First, considering the responses and thecounter-strategies of other actors would call for anevaluation of the process, not just of the outputs.Second, evaluators need to know the right places inwhich to look for the impacts of human rights anddemocratic initiatives. Identifying impacts is a difficulttask, because the unique shape of a society and of itspolitics usually affect the way in which it responds tosuch initiatives. This may entail the study of processesthat are not readily observable - for example, how anindividual feels and acts in terms of his or her ownrights and responsibilities vis-à-vis those of the state. Itis suggested that rapid appraisal methods, such as focusgroup or key informant interviews, could beinstrumental in this study.

Learning about Human Rights by Means ofConsultative Evaluations

The purpose of the consultative evaluations that areused by Northern NGOs is to improve partnershiprelations and, most importantly, to understand thenature and characteristics of human rights work indeveloping countries. Evaluators often use aninterpretative and critical approach, in which attentionis paid to process, dynamics of change, learning and

strategic orientation, and to the claims andconcerns brought forward by thestakeholders.

Several lessons are emerging from theexperience with this approach. In essence,human rights activities are intended tocorrect an imbalance of power. Thus,human rights evaluation involves politicalanalysis and is political. The results of theevaluation will, therefore, depend, at leastin part, on the ideological presuppositionsof the evaluators. Conducting evaluationsin a consultative, (that is, participatory)way has the advantage of including thevoice of human rights partners indeveloping countries and, hence, theirconcepts of human rights, partnership andevaluation methodology.

One question often arises: Who evaluateswhom and for what purpose? Northernagencies should ask themselves whethertheir institutional or bureaucratic needstake precedence over the needs of theirSouthern partners. Several Southern NGOsmention the desirability of reciprocalevaluations. According to them, evaluatingthe performance of Northern partners intheir work of fund-raising, administration,public information, policy dialogue andadvocacy in the North, "...would pushNorthern funding organizations from theirback stage position into the limelight."

Impact Assessment in PoliticalContext

A new challenge is presented in learninghow to assess the impact of human rightsprograms that support sectors of public lifeor the organizations of civil society,against the background of differentpolitical systems. Four tasks have to betackled in building the methodologicalframework for responding to the challenge:

• the identification of differentphases of political development;

• the identification of the relevantsectors in public life and theorganizations of civil society;

• the categorization of projectssupporting these sectors;

84

• the design, for both overall and sector goals,of concrete indicators for use in impactassessment.

A number of other lessons have emerged. Evaluatorsacknowledge the difficulties in measuring impactswhich can only indirectly be achieved. It is obviousthat a multitude of factors contributed to, or opposed,the overall objectives of the interventions examined inthis research, and a separate measurement provedimpossible. Evaluators, however, stress the point thatthere is an urgent need for impact studies that lead,relatively quickly and reliably, to more informed policydecisions. This would be preferable to the rigorousexplanatory studies of a small number of projects thatdo not deliver important general information.Evaluators should try to decide whether changesoccurred, and whether they can plausibly be attributedto donor-supported interventions. An appropriatemethodology needs to be developed.

The assessment of the impact of programs in humanrights and democracy against the background ofdifferent political systems is a major step forward. Thereal challenge still lies ahead, however. There isevidence that the impact of human rights programs,directed towards human rights, democratization andgood governance, not only depends on developments inthe political system, but on three inter-related, andoften interlocking, processes: stages in politicaldevelopment, stages in economic adjustment and stagesin ethnic, nationalist and religious conflicts. Furtherresearch is required in these preliminary findings.

Conclusions: Work to be Done

A major lesson to date is that the challenge ofsupporting the implementation of human rights is muchmore complex than once thought. Thus, this studypoints to the need to develop a number of practicalmethodological and organizational tools.

1) Develop an analytic framework for theassessment of human rights policies andprograms.

Aid agencies, consultants, and researchers areconfronted with the task of developing an analyticframework to be used as a practical instrument for:

• the analysis of short-and long-term links,positive and negative, between policy goals inhuman rights, democratization, rule of law,good governance, and sustainabledevelopment.

• the analysis of three interrelatedprocesses: stages in politicaldevelopment, stages in economicadjustment, and stages in ethnic,nationalist or religious conflict.

This framework should be instrumental infilling the gap between policy formulationand program implementation.

2) Create appropriatemethodological tools forprogramming, monitoring, andevaluating human rightsactivities.

There is an urgent need for:

• a range of cost-effective methodsfor making rapid baseline studiesof human rights;

• simple models of the dynamics ofhuman rights suitable foranalyzing the interactiveprocesses between projectbeneficiaries, state agencies, andother stakeholders;

• a system of concrete indicators forprogress or performance.

Progress indicators should encompass themonitoring of actual violations of humanrights, but should also cover the broaderfield of the capacity and the incentives ofall actors and agencies involved in theenactment, implementation and monitoringof human rights, the administration ofjustice and the enforcement of humanrights.

3) Strengthen co-ordinationbetween funding andimplementing agencies.

There are strong arguments for aidagencies:

• to co-operate in carrying outanalyses of the context of humanrights, making baseline studiesand parallel and combinedstudies. Depending on thesituation of a given developingcountry, concrete initiatives couldbe taken within the most

85

appropriate institutional framework(consultative group, round table, or NGOplatform);

• to create a network for learning about bestpractices in human rights and democratization.

Learning from evaluation documents aloneis time consuming and tends to bepiecemeal. A network for learning shouldestablish a system for the exchange ofinformation between aid agencies,practitioners in the field and researchinstitutes. Consideration could be given tobuilding a Development AssistanceCommittee (DAC) network.

1

Prepared by Walter Stolz, consultant, in co-operationwith Anjet Lanting, research assistant, on behalf of theOperations Review Unit, Netherlands Ministry ofForeign Affairs.

86

PARTICIPATION1

INTRODUCTION

Purpose and Scope of the Paper

This text is an Executive Summary of a more extensivepaper on Evaluation and Participation that reviews andsynthesizes the experiences of donor agencies withregard to support for participation or participatorydevelopment in policy work and through program andproject funding. It attempts, firstly, to provide anoverview of the present situation with regard toevaluation and participation and to obtain a clearerpicture than presently available of commonalities anddifferences in current evaluation thinking and practiceamong DAC agencies. A second objective is to drawout lessons learned on the basis of the review, and tomake suggestions regarding strategies andmethodological issues to be pursued in future work onevaluation and participation. The paper is an non-exhaustive state-of-the-art report and is fairly selectivein focus. It represents a contribution to the process ofrethinking and dialogue to which all DAC ExpertGroup on Aid Evaluation (EGE) members contribute.The main focus is on donor/funding agency experiencewith participation as revealed in evaluation reports andother relevant project documents as well as within thecontext of development policies. Donor/fundingagencies' use and experiences with assessmentmethodologies for the evaluation of participation arealso discussed.

Topics Covered

In section II (Key Concepts and Frameworks Used),terminology, concepts and definitions of participationare discussed. This section also incorporates insightsand clarifying conceptual frameworks from relevantresearch and recent analytical work on participationwith regard to development processes generally, and toinstitutional development in particular. A discussion ispresented of participation in evaluation, and ofmethods and approaches involved in participatoryevaluation, based on innovative and experimentalapproaches and experiences reflected in recentresearch, analytical and reference literature onparticipation and evaluation rather than on submittedagency documents. Subsequently, the evaluations,project documents, and procedural manuals providedby agencies are reviewed with respect to the use and/orapplication of the concepts, definitions and indicatorsof participation discussed in section II. A special focusis on NGOs as potentially important actors and

intermediaries for support to andpromotion of popular participation.

Importance and Rationale for theTopics

The point of departure for the selection oftopics was the Framework for Evaluationof Programs Promoting ParticipatoryDevelopment and Good Governance(November 1993) established by EGEmembers. This framework has alsodetermined the boundaries of participationwith respect to other thematic areas,particularly Decentralization and HumanRights, but also the Legal Systems andPublic Sector Management themes. Thediscussion on the role of NGOs was alsoincluded in accordance with memberagreements reported in the EGEFramework paper.

Participation is a relatively new concernfor evaluation. There is also a gapbetween participation as rhetoric andelement in current development policydiscourse, on the one hand, andparticipation as operational practice on theother.

There continue to be few evaluations ofparticipation owing to the relative noveltyof the issues involved. Evaluationsexplicitly addressing participation, orapplying participatory methods, are onlyjust now beginning to emerge. However, asthe review demonstrates, donors havesupported programs and projects clearlyrelevant for participation since the 1960s.These considerations all point to theusefulness of reviewing the presentsituation with respect to these topics, inorder to take the learning process a stepforward.

There are also some difficulties indefinitions of popular participation andparticipatory development2.. Theseconcepts imply various analyticaldimensions and aspects that tend to beformulated and interpreted in differentways, in different settings and by differentactors according to the administrative levelinvolved in a particular developmentintervention at a specific point in time.Definitional problems with regard toparticipation result from the fact that not

87

only conceptual/theoretical, policy and operationaldimensions are at issue, but also ideological and moralones. Definitions may also vary with respect to how'operational' they are in terms of applicability andrelevance in actual projects. The significance andoutcomes of participation in the practice ofdevelopment co-operation also vary in accordance withsuch interpretations and conceptualizations.

Sources of Information and Methods Used

This chapter is based on a qualitative review of 43agency evaluation reports, project documents andrelevant policy or procedural papers received fromeight bilateral donors. About the same number ofanalytical documents and publications on participation,some of them from The World Bank or commissionedby bilateral donors, were also perused.3 The paucity ofevaluations and other project documentation onparticipation is due partly to the fact that this topic, asan explicit focus, is relatively new. It may be due alsoto agencies’ distinction between assessments of therelatively few recent projects in which participation hasbeen an explicit objective or strategy and most earlierprojects in which participation was defined obtusely orpursued more implicitly.

The sample of documents represents the work of only asmall number of bilateral donor agencies and does notprovide conclusive evidence regarding such issues asconnections between participation and sustainability(or impact on governance, government services,economy and public finance). Despite suchreservations, however, we have been able to get a fairlygood picture of development agencies' collective stateof knowledge and intentions with regard toparticipation and evaluation at this time. Findingsinclude certain tendencies and trends that were noted,which we feel will serve well as a basis for future DACwork.


Although it has its origins in the concept of communitydevelopment in the 1960s and 1970s, participationreceived new attention as an explicit goal indevelopment assistance in the late 1980s due toincreased emphasis on project sustainability,institutional development and policy reform. It wasinitially promoted and applied mainly by NGOs and insmall-scale projects, but also by multilateralorganizations such as FAO, ILO and UNRISD, as wellas some bilateral agencies. Popular participation asoperational development practice at the project orprogram level has, however, lagged far behind generalawareness of its benefits and donor advocacy of its

principles. Its use as rhetoric in policydiscussions and public declarations ofnational development objectives andstrategies is widespread.

The end of the Cold War and thetransformation of authoritarian regimesand political systems in many of thecountries in Africa, Asia and LatinAmerica during the last few years focusedrenewed attention on participation in thecontext of political as well as economicdevelopment. Popular participation hasbeen viewed as one of the ways ofstrengthening democracy, civil society,decentralization, human rights and thedevelopment of forms of good governance.The challenge of escalating environmentaland social problems has meant thatparticipation is increasingly beingrecognized also as an essential componentof sustainable development strategies. It isnow also obvious that in order for'participation' to become more useful andapplicable as a management concept and inactual development practice, it is necessaryto define it in operational terms, withregard to primary beneficiary/stakeholderlevels in particular social and politicalcontexts.

88

Concepts and Definitions of Participation

The various definitions of participation applied byactors and agencies may be viewed along a continuumfrom more far-reaching or profound with respect toempowerment, influence and control on the part ofgrassroots participants, to more conventionalconceptions where agencies or project staff stillessentially retain decision-making power and controlwith respect to key project functions. Definitions mayalso vary with respect to how 'operational' they are interms of applicability and relevance in actual projectswithin specific social, cultural and politicaldevelopment contexts. UNRISD's Popular ParticipationProgram, 1985, defined participation as "the organizedeffort to increase control over resources and regulativeinstitutions in given social situations on the part ofgroups or movements hitherto excluded from suchcontrol".4 Although rather general, this definitioncaptures the wider meaning of the participation conceptand stresses its empowerment, control and decision-making aspects. The World Bank's Learning Group onPopular Participation defined popular participation as"a process by which people, especially disadvantagedpeople, influence decisions that affect them".5

The term "popular" refers not only to the absolute poorbut also to a broader range of people who aredisadvantaged in terms of wealth, education, ethnicgroup or gender structures, and "participation"connotes influence on development decisions andproject or program design, not simply involvement inthe implementation or benefits of a developmentactivity. In subsequent World Bank documents, afterinternal review and discussion, the term "popularparticipation" is replaced by the more abstract andgeneral "participatory development", while "popular","poor" or "disadvantaged" groups are subsumed underthe broader and more inclusive "stakeholder" concept.

According to this modified definition "participatorydevelopment is a process through which stakeholdersinfluence and share control over developmentinitiatives, decisions and resources which affect them.""Stakeholders" range from the ultimate beneficiary in agiven society or setting to individuals or institutionswith indirect interest. "Key stakeholders" are thoseintended to be directly affected by a proposedintervention, i.e., those who may be expected to benefitor lose from Bank-supported operations or who warrantredress from any negative effects of such operations,particularly among the poor or marginalized. It shouldbe added that the Bank recognizes itself as astakeholder with its own objectives, policies andinstitutional responsibilities.6

The World Bank Learning Group onPopular Participation also endeavored tospecify and operationalize participation. Itproposes a classification of instruments ofparticipation, referring to institutionaldevices which organize and promote thesustainability of popular participation, suchas local level development workers, NGOs(local, intermediary, apex organizations),local government units, centralgovernment agencies and private sectormechanisms.7 The World Bank maintainsthat its Articles of Agreement prohibit itsintervention in political affairs, and givenits focus on economic development, itsinterest in participation is primarily as ameans to improve the results of itsinvestments. Several bilateral agencies(e.g. CIDA, GTZ, Sida, USAID), however,refer to participation as both an end and ameans, and frequently view it as anexplicit aspect of objectives such asdemocratization, equity, human rights andsustainable development.

The OECD/DAC definition ofparticipation8 approximates the WorldBank definition cited above and in a recentOECD Development Center report"Participatory development stands for apartnership which is built upon the basisof a dialogue among the various actors(stakeholders), during which the 'agenda' isset jointly, and local views and indigenousknowledge are deliberately sought andrespected. This implies negotiation ratherthan the dominance of an externally setproject agenda. Thus people become actors

instead of being simply beneficiaries".9.

This definition also implies as a mainobjective the empowerment of the localactors (as individuals or groups andinstitutions) to make participationsustainable. Entry points for the processshould be sought both from below (localorganizations, NGOs) and from the top(policy dialogue).

Levels and Dimensions ofParticipation

Different dimensions and levels, degrees orkinds of participation can be analyticallydistinguished10. These terms are usedslightly differently by different authors andorganizations. They refer basically to

89

where in the project cycle participation occurs(planning, implementation, monitoring and evaluation,take-over), to the quality, intensity or extent ofparticipation (as passive beneficiaries, informants, cost-sharers, or as colleagues or counterparts with a voice inmanagement, decision-making and control), and tosocietal levels (local, regional, national).

An operational dimension of participation concernsspecific project functions or tasks such as construction,operation, maintenance, management and distributionof benefits.11 In many projects, participation takesplace with regard to tasks such as construction work,operation and maintenance (contributions in the formof money, labor or material), but more seldom inproject formulation, management, control overresources and distribution of benefits. If resources arecontributed by the "intended beneficiaries" or "primarystakeholders", they should obviously also have a say inthe management and distribution of project resourcesand benefits. Participation in these respects must beintroduced and prepared already from the beginning.As a project or development activity reaches a certainlevel of maturity, the beneficiaries or theirorganizations should gradually take over responsibilityfor management and other key project functionsbecause this is a fundamental prerequisite for finaltake-over and project sustainability in the long run.

Links between participation and gender have beennoted by some development administrators andresearchers. To a certain extent, attention to the oneencourages attention to the other, but these links haveyet seldom been systematically explored. Participationand gender do not necessarily exist in a dynamic,mutually reinforcing relationship. Much evidenceshows that a focus on participation does notautomatically result in attention to gender.12 TheWorld Bank, OECD Development Center and others,discuss as a particular focus, gender aspects and theactive participation of women at various phases ofprograms. It has also been suggested that representationand participation in terms of gender could be used inevaluations as an indicator of the nature and overalldegree of participation in projects13 because the samefactors that prevent projects from being or becomingparticipatory in general are those that also contribute tothe exclusion of women.14

Evaluations provide opportunities for testing thetranslation of development policy into implementation.Gender issues as well as participation requirerethinking of assessment methods particularly withrespect to equality and empowerment aspects,recognition and analysis of roles, responsibilities,resources and interests of female and malestakeholders. And since both monitoring and evaluation

tend more readily to assess inputs orphysical outputs, the need for innovativeapproaches and better indicators forassessing impact and effects is alsosomething that evaluation of gender andevaluation of participation share.

Participation and DifferentCategories of DevelopmentOrganizations

Small-scale, community-specific projectsaiming at social objectives, frequentlyoperated by NGOs, comprise a commonform of donor assistance to the promotionof participation. NGOs are seen by donors(and by themselves) as working primarilywith well-defined, grassroots-level,beneficiary groups. Since NGOs do notwork within governmental structures, theyare less restrained by bureaucraticobstacles and political resistance thanbilateral and multilateral agencies.Advocates of support to participationthrough NGOs often maintain that theycontribute to more effective achievementof project objectives because of NGOability to reach people at the local level.

The NGO label includes widely differingkinds of organizations, some of which areworking in a participatory manner, whileothers do not. Additionally, NGOs'representativeness and accountability oftenremain unclear.

Some problems related to induceddevelopment cannot be resolvedexclusively at the local level. Programsmust often make decisions and considertrade-offs regarding, for example,allocation of limited resources andsequencing of activities, that affect widerareas. While inputs into the planning andoperational processes of such undertakingscan be provided through consultations atthe local level, some strategic, overalldecisions must be taken at national,regional or program level.

Definitions and approaches to participationadopted by multilateral organizations,bilateral agencies and NGOs differ due totheir different politico-organizationalmandates and the contexts or settings inwhich they are embedded and operate. Akey role for multilateral and bilateral

90

agencies is to encourage governments, through policydialogue, to become more open to participatoryapproaches. Governments are the most direct andinfluential actors for promoting a favorableenvironment for participation among the intermediarypublic sector organizations which operate, frequently inco-operation with NGOs and grassroots' organizations,at regional and local levels.15

With regard to agency field presence, deemedimportant for support to participatory projects, bilateralagencies tend to have more extensive and continuousfield presence than most multilateral agencies. NGOsnormally have more presence at the local level thanbilateral agencies. The different preconditions andcharacteristics and the complementarity of theseorganizations mean that there is considerable space andneed for collaboration between them in order to fullyexploit their respective advantages.

Costs and Benefits of Participation

A main finding of the World Bank Learning Group onParticipatory Development is that "there is significantevidence that participation can, in many circumstances,improve the quality, effectiveness and sustainability ofprojects, and strengthen ownership and commitment ofgovernment and stakeholders. Community participationstrategies are found to be particularly important inreaching the poor".16 Systematic evaluations or"measurements" of the costs and benefits ofparticipation are scarce, but generally indicate that thecosts, in terms of time and money spent, tend to berelatively higher for participatory projects in the courseof their early phases. The initial investments inparticipation, however, tend to pay off in terms ofincreased efficiency and sustainability and in savingtime in subsequent phases.17 Delays in disbursementduring early phases may occur in projects wherecommunities are given the responsibility to select,design and implement project activities because of thetime required to build up sufficient communityawareness and capacity. Rapid increases in thesubsequent numbers of communities involved and intheir capacity to manage the activities, as well as betterprospects for long-term success, generally compensatefor the time spent on initial preparation. For poorpeople, the costs of participation are generallymeasured in terms of added time spent onorganizational matters as well as in terms of cost-sharing contributions they may make, and such costsmay be considered by them as high.

Statistical analysis of 121 rural water supply projects inAsia, Africa and Latin America found that "beneficiaryparticipation" was the single most important factor indetermining overall quality of implementation, a

significant contributing factor to projecteffectiveness, maintenance of watersystems, overall economic benefits,percentage of the target populationreached, and environmental benefits.Participation also resulted in communitymembers acquiring new water-related andorganizational skills, and strengthenedcommunity organizations that went on toundertake other development activities.18

Other benefits observed in case studies ofWorld Bank-financed projects include: anincreased uptake of project services,decreased operational costs, an increasedrate of return, and increased incomes ofprimary producers.19 It has also beenpointed out that in conventional(quantitative) evaluations of participationits costs are generally weighed only againstestimated benefits and not against the costsof not encouraging and assistingparticipation.20 Several of the benefits ofparticipation presented above would thusconstitute costs associated with non-participatory approaches (lack of use andmisuse of facilities, poor maintenance, lowrates of sustainability, etc).

Evaluating Participation RequiresMethodological Adjustments

In discussing evaluation and participation,it is essential to distinguish amongdifferent types of projects, as well asdifferent types of evaluation. An importantdistinction is to be made betweenevaluation of participation andparticipation in evaluation. Evaluation ofparticipation in development projectsrefers to the assessment of a specificobjective or outcome of an activity,whereas participation in evaluation refersto the degrees of involvement of differentcategories of social actors (e.g. agencies,project staff, grassroots groups) in theevaluation process. Evaluation ofparticipation could, of course, take place inany type of development project todetermine its extent and outcome, or itsimplications with respect to conventionalevaluation criteria such as relevance,efficiency, effectiveness, replicability andsustainability of project activities.

Evaluating development process andoutcome in terms of participation involvesa consideration of the concepts,

91

definitions, dimensions, levels and forms ofparticipation discussed above, with respect to specificprojects. As indicated above beneficiary participationin evaluation, or participatory evaluation, requireattention and preparation already at the very beginningof the project cycle. Data collection methods for theevaluation of participation include traditionalquantitative methods (questionnaires, sample surveys)to "measure" quantifiable aspects of participation.Conventional procedures involve the identification ofsome basic criteria of participation. On the basis ofthese, a set of appropriate indicators (and interviewquestions) may be selected that are assumed to reflectextent, intensity and changes in participation atdetermined stages of a program or an activity.

Such quantitative indicators are, for example, numberof project beneficiaries as a proportion of totalpopulation, frequency of project meetings, proportionsand total numbers of beneficiaries attending meetings(or taking part in different project components oractivities), recruitment of leadership as a function ofsocial position or stratification pattern, rotation ofleadership over time, distribution and circulation of keyfunctions or tasks within the project or beneficiaryorganization, beneficiary contributions in the form oflabor, money or material, distribution of benefitsresulting from the project.

Participation is a process of complex social change andquantitative indicators provide us with only a veryincomplete understanding or picture of participation.When the objective is to determine the character ofparticipation and how it takes place, identification ofqualitative indicators is necessary. Such indicators maybe used to describe or characterize the relationsbetween leaders and general members of anorganization, forms of organization, forms anddynamics of decision-making, group solidarity,community spirit, conflict and problem-solvingcapacity, etc. Such properties, qualities, attitudes,relationships or behavior, are almost impossible tomeasure in quantitative terms. They must be perceived,described and analyzed through qualitative interviewsand direct observation. Evaluation of participation willthus always require some combination of quantitativeand qualitative approaches to data collection, analysisand interpretation.

Quantitative aspects can be dealt with by measurementthat leads, through application of quantitative methodsof data collection and analysis, to judgement.Qualitative aspects require descriptions (of propertiesor processes) leading, through interpretation, tostatements about their nature or consequences.Evaluation of qualitative aspects therefore requiresdifferent indicators and methods of data collection and

analysis. Oakley suggests, inter alia, thefollowing key principles or characteristicsof qualitative evaluation: it is heuristic inthat the evaluation approach is subject tocontinuous redefinition as knowledge ofthe process and its outcome increases; it isholistic and sees the program as a wholethat needs to be understood and analyzedfrom many different perspectives; it isinductive in the sense that the evaluatorseeks to understand the outcome of adevelopment project without imposingpredetermined expectations; it is alsointerpretive, built up through description ofthe significant facts, figures, andcharacteristics of the project that are anaccurate reflection of its overallcomplexity; it implies a close contact withthe participants of a program in their ownenvironment to understand the realities anddetails of their everyday life.21

The selection and development ofqualitative indicators of participation is atan initial stage, and much work andexperimentation remain to be done. Oakleysuggests three broad areas of qualitativeindicators of participation particularlyrelated to the changes occurring in thenature, growth and behavior of the projectgroup as a result of the project activities:organizational growth, internal structuringof the project group, allocation of specificroles to group members, emergingleadership structure, formalization ofgroup structure; group behavior; changingnature of involvement of project groupmembers, emerging sense of collectivewill and solidarity, involvement in groupdiscussions and decisions, ability toanalyze and explain issues and problems;group self-reliance; increasing ability of aproject group to propose and to considercourses of action, group members'knowledge and understanding ofgovernment policies and programs,changing relationships of group withproject staff/group facilitator,formalization of independent identity ofthe group, independent action undertakenby the group.

These indicators must also be observed andrecorded. Furthermore, qualitativeindicators are mostly intangible,manifesting themselves over time, theymust be related to some observable

92

phenomena or activities and be part of a system ofcontinuous monitoring. "Monitoring, therefore, hasemerged as the key to the evaluation of participationand certainly as the only way to ensure a continualsupply of relevant data and information".22 Fourcategories of concrete phenomena are suggested which,if monitored continually, should provide relevantinformation and data for both quantitative andqualitative indicators: (1) project or group activities,economic or production activities,physical/construction work, collective project groupwork, project group internal structuring; (2) changes inproject group behavior; nature of project groupmeetings, levels of explanation and discussion, people'sinvolvement in group discussions, incidence ofconsensus and disagreement, emerging patterns ofleadership; (3) group action and articulation,independent action by project group, levels and natureof contacts with outside officials, levels and nature ofcontacts with other groups or organizations: (4)project-group relationship; nature of initialrelationship, building up of the project group, natureand changes in relationship between project and group,project withdrawal.

Methods that have been employed in judging certainqualitatively observed phenomena include: grouprecords, log-books, and diaries providing a continualaccount of events and the process of participation as itunfolds; group discussions, key informants, and fieldworkshops.23 Participant observer evaluation,conducted by a neutral qualified observer to assess orevaluate project processes and effects in localcommunities, employs various of the qualitative andquantitative information gathering techniquesmentioned above. Although the main responsibility for,and direction of, the evaluation process remains withthe participant observer, it allows for a high degree ofinfluence and participation from different categories ofstakeholders.24

Participatory Techniques in Evaluation

Parallel to the increasing attention given to popularparticipation in the 1980s, Rapid Rural Appraisal(RRA) methods evolved from an initial emphasis onrapid and cost-effective data "extraction" by "outsiders"towards Participatory Rural Appraisal (PRA). PRAreduced the role of the "outside" researcher or "expert"to that of "convener", "catalyst", or "facilitator"enabling people to undertake and share their owninvestigations and analysis, as well as to plan and takeaction.25 The basic epistemological or methodologicalpremises underlying PRA, such as value-pluralism andthe rejection of objective "reality" or "truth" (there aremultiple layers of "realities" depending on positions incultural and social settings) and thereby the importance

of local knowledge, are also shared by thedifferent versions of "participatorymonitoring" or "participatory evaluation".

In participatory evaluation a variety ofstakeholders should actively take part inthe determination of evaluation objectives,the selection of procedures and datacollection methods, the analysis andinterpretation of data, as well as decisionsregarding measures and action based onrecommendations produced as a result ofthe evaluation process.

In practice, however, participatoryevaluation is most often limited to theinclusion of a few techniques aimedexplicitly at project participants orbeneficiaries, such as group consultationsor key informant interviews. Theseconsultations usually take place within theconventional type of donor evaluationconceived of largely as an end-product orneatly-bound, one-off type of anessentially quantitative exercise.Beneficiaries or stakeholders have thus noparticipation in the determination ofevaluation objectives, modalities, analysisand interpretation of evaluation data. Norhas evaluation usually been an aspect ofplanning and monitoring throughout theproject or program process. The extremelylow incidence of true participatoryevaluation reflects partly a lack of ease onthe part of donors with the more innovativeapproaches that such evaluation requires.As Marsden, Oakley and Pratt26 point out,few projects adopt an authenticparticipatory approach to evaluation.

Participatory evaluation is applicable indifferent types of projects and settings. It isnot intended to completely replaceconventional evaluation methods. Rather,it should be used as a complement to makesuch methods more appropriate andeffective, and to ensure that the reality andclaims of the beneficiaries are really takeninto account. Our impression is thatinclusion of participatory techniques inmore or less standard donor evaluations isincreasing considerably.

Beneficiary participation in monitoringand evaluation should occur in projectsclaiming to use a participatory approach.But participatory monitoring and

93

evaluation could be employed also in other types ofprojects. The introduction of participatory techniquesin projects that have not followed a participatoryapproach in planning may have to find ways to dealwith certain inherent contradictions and practicaldifficulties. These would involve, for example,reorientation of conventional time-frames, re-scheduling resource allocation, and probably new waysof decision-making. Investments of time and resourcesfor participation, however, tend to pay off in the formof firmer beneficiary commitment, more efficientcapacity building, and ultimately enhancedsustainability.27 As we shall see below, the increasinginterest and work concerning participation inevaluation and evaluation of participation have notbeen limited to the field of development co-operation,but have been influenced and inspired by a broaderscientific debate on the theory and practice ofevaluation in general.

Participatory Evaluation: An Opportunityfor Negotiation?

In a recent comprehensive review of methods andapproaches to social evaluation, Marsden, Oakley andPratt28 describe the "traditional type of evaluation" or"first generation evaluations" as generated by thedemands of management from the donor perspective,and usually serving the purpose of justifying agencyspending. This type of evaluation was conducted by"external" or "independent" evaluation expertsapplying quantitative approaches to data collection("measurement-oriented") and relying on a positivistscientific paradigm where the belief in the existence ofuniversal and objective "truths" constituted acornerstone.

A modified version of "traditional evaluation" is the"second generation evaluation" characterized bydescription of patterns of strengths and weaknesseswith respect to stated objectives, whereas the "thirdgeneration" was characterized by efforts to reachjudgements. The evaluator assumed the role of a judge,while retaining the earlier technical and descriptivefunctions as well. Guba and Lincoln identify thepervasive weaknesses of these three generations ofevaluation as a tendency toward managerialism, afailure to accommodate value-pluralism, and an overcommitment to the scientific paradigm of inquiry. Asan alternative they suggest "fourth generationevaluation" which is based on two key elements:responsive focusing and constructivist methodology.

Responsive focusing sets the boundaries of theevaluation by interaction with its stakeholders and theconstructivist methodology provides the widerframework in which "truth" and "fact" are recognized

for their subjectivity. The positivismparadigm and its belief system arerejected.29 The outcome of this process isnot conclusions based on "objective" valuejudgements, but an agenda for negotiationbased on the claims, concerns and issuesthat were not resolved in the evaluationdialogue. Fourth-generation evaluationthus constitutes a forum for debate.Evaluation is a learning process involvingdifferent sets of understandings that needto be negotiated. It provides importantmoments in the lives of all developmentprojects, when opportunities for the"negotiation of values" might be centrallyaddressed.

Participation by the beneficiaries in suchnegotiation transforms the evaluation intoa process of empowerment that offersopportunities for furthering ourunderstanding of the operationalization ofa participatory agenda. Further, suchnegotiation comprises building blocks inthe development of effective partnershipsbetween what tend to be essentiallyunequal partners. As a complement totraditional approaches to evaluation aninterpretive approach, consisting of thefollowing stages, is suggested:

1. Identify the full range ofinterested parties.

2. Find out how the evaluation isperceived - stakeholder claimsand concerns.

3. Provide context and amethodology through which thesecan be understood, taken intoaccount and constructivelycriticized.

4. Generate as much consensusabout different interpretations aspossible.

5. Prepare an agenda for negotiation.

6. Collect and provide theinformation requested in theagenda.

7. Establish and mediate a forum ofstakeholders in which negotiationcan take place.

94

8. Develop a text available to all.

9. Recycle evaluation to take up unresolvedissues.


The documents received from eight bilateral agencies(DANIDA, DGIS, FINNIDA, GTZ, NORAD, ODA,Sida and USAID) were reviewed qualitatively, in termsof three main concerns: (i) participation orientation orsensitivity overall; (ii) whether and how participationwas reflected in the evaluation criteria selected torepresent output or impact; and (iii) participation asreflected in evaluation procedures and methodology.Where possible, the Terms of Reference were alsoexamined. The reviewed documents includeevaluations and project documents, as well as somepolicy papers, analytical reports and proceduralmanuals. In addition to traditional evaluation criteria(relevance, cost-effectiveness, impact, efficiency/cost-benefit relation, sustainability, replicability), thedocuments were examined against the background ofthe conceptual and analytical issues discussed above.

Forms of assistance to participation commonly are: (1)co-financing; (2) framework support or Block-Granttype of funding for NGO support, usually to donorcountry NGOs; (3) mainstream - usually bilateral coreprogram - project or program support e.g. in IRDPs orarea programs (4) direct support funds for small-scaleprojects, usually administered through donor embassiesor development co-operation offices; and (5) policy,research or analytical work.

The evaluations focused variously on activities,projects, sectors or national development efforts andare clearly influenced by the policy, planningprocedures and traditional evaluation practice of therespective agencies. Evaluations specifically focusingon participation in development activities are few andthere is little evidence of participation in evaluation,even in evaluations of NGO support. NGO evaluations,however, did employ participatory techniques to agreater extent than evaluations of core-support,bilateral programs.

Agency policies and experiences concerningparticipation exhibit some differences but many pointsof commonality can also be noted. A main one is thatin a general sense all of the documents clearly showawareness of participation issues. The nature anddegree of awareness, and its translation into explicit,systematic attention to participation vary considerably.Another general observation concerns the more

pronounced focus on participation in thedocuments representing interventions inthe agricultural, rural or villagedevelopment sectors, than in thoserepresenting other sectors. Although wesuspect that this relationship may begeneral, no firm conclusions can be drawnsince there is a clear predominance ofagricultural, rural and village developmentreports in the documentation submitted forreview, just as there is a bias towardsinterventions in Africa.

Another striking point of commonalityconcerns overt donor attempts in the 1990sto transcend the confines of top-downapproaches by adopting participatorystrategies to an increasing degree. There isa clear trend not only to plan better in ageneral sense through formulation ofplanning frameworks and conceptual tools,but also to promote participatory planning,as is evidenced in the documents of severalagencies. Current donor/funding agencydiscussions as revealed in the documentsdisplay a receptivity to participatorytechniques and notions of participation inevaluation. Although the reports describeevaluation procedures essentially in termsof conventional, donor-steered agendas andmechanisms, they also incorporateimpulses from analytical and researchwork on participation. This is important inbridging the persistent gap betweenparticipation as rhetoric, and asdevelopment practice and partnership.

The reports note and discuss a number ofperceptions, beliefs, and assertions aboutNGO initiatives ("articles of faith") held byfunding agencies, and by the NGOsthemselves, that support the positive lightin which they are increasingly viewed.These include the idea that NGOs are ableto reach the poorest, or at least grassrootslevels, with much less difficulty thanmainstream development assistance. TheNGO evaluations also employ somewhatmore innovative criteria and methods thanmainstream program evaluations. Theyraise such issues as how to constructivelyassess exclusion, the culture of theimplementor at operational levels versusthe cultures of the beneficiary groups, etc.

Institutional arrangements forimplementation include national

95

governments, line ministries, local government units,parastatals, various kinds of NGOs (or "borderline"associations such as trade unions), community or adhoc village or small local organizations. Thedocuments reviewed here do not represent more than asmall sample. We feel, however, that many of theirfindings and conclusions are largely concurrent, andthat they thus demonstrate central issues, problems aswell as indications of progress made in the area ofparticipation and the assessment of participation. Onequestion that cannot be answered definitively concernsdifferences in donor approaches, and differencesbetween the donor community and the approaches toparticipation and planning models adhered to bydifferent categories of NGOs ("learning process"models as opposed to "blueprint" or "top-down"models). Large NGOs exhibit both models in theirapproaches to participation.

The evaluations commonly referred to general policiesconcerning prioritized development principles or issues(poverty alleviation, sustainability, institutionalcapacity building, etc.) in assessing attainment orrelevance of support and activities with respect tostated objectives. Prescriptive donor manuals that detailplanning or procedural methods also containdiscussions of participation, although these tend toreflect exclusively the perspective of the donor.Analytical papers emphasize interpretive or other non-conventional evaluation principles and procedures withregard to assessing participation. According to theselatter, evaluations should not only focus onparticipation, but should themselves be participatory inthe sense discussed above.

Most evaluations followed a conventional constitutionas "a pre-structured exercise with prescribedprocedures ... by external evaluators, which iscommented upon by the project manager and providesthe basis for decisions by the commissioning - i.e.funding - party." The challenge of a participatoryevaluation procedure that "includes the same steps butfollows different rules, since it is considered a learningprocess for all involved, and in which criteria andindicators are not prescribed in advance by'outsiders'"30 is seldom taken up.

Lessons learned from the evaluations in terms ofpopular participation are often not formulated as suchin the reports. Despite common conceptions of NGOsas operating closer to poor people, the performance andimpact of NGO projects varies greatly. The FINNIDAand Sida NGO support evaluations indicate that evenwhere short- to medium-term performance is good,sustainability is weak as are participatory elements.NGO reporting is also normally weak, except forfinancial reports which are generally rigorous.

Substantive reporting and monitoring ofdevelopment progress and impact arefrequently lacking. Though formallyrequired by donors, in practice it is neitherstrictly followed by the NGOs evaluated,nor followed up by the funding agencies.

Donor Experiences with Respectto Specific Evaluation Issues

Definitions of participation and ofproject stakeholders

The difficulty of operationally definingparticipation emerges clearly in thedocuments. Definitions vary a great deal,both in projects supported and inevaluation studies intended to assessparticipation. Participation in projects isoften defined either very generally orinterpreted to mean a range of stakeholderroles, few of which actually involve "anactive and influential hand in shapingdevelopment decisions that affect theirlives" for primary stakeholders. Inevaluations, participation is defined or atleast practiced in terms of "consultation",the use of PRA methods or otherparticipatory techniques. Virtually none ofthe evaluations, even the mostcomprehensive ones, demonstrated muchevidence of participatory evaluationmethods.

In large scale, multi-sectoral ruraldevelopment programs, support toparticipation has involved local publicsector institutions rather than grassrootsstakeholder groups. There are, however,clear indications in the donor materials ofa movement towards more exactingplanning methods in which a basis for themonitoring and evaluation of participationat different levels can be established earlyon. Some project documents from veryrecent years refine "target group"designations by disaggregating stakeholdercategories into a number of specifiedgroups.

Evaluation methodology

Analytical and evaluative documentsdiscuss the difficulties of measuring slow,long-term, qualitative or structural/politicalchange, such as in attitude and behavior. Amajor issue raised in the methodological

96

discussions of the NGO evaluations was thedevelopment of parameters for judging performance ofNGO-funded programs. Criteria for evaluating theseand other social dimensions are underdeveloped.Impact and effectiveness are assessed and discussedprimarily in quantitative terms (e.g. numbers of NGOs,projects, funds allocated, people reached, etc).

A prominent feature of the NGO material is theemphasis on awareness, organization and inadequaciesof most evaluations due to the fact that they reflectalmost exclusively the perspective of the fundingagency and project management, and not that of thestakeholders. What happens to women and to genderrelations as an indication of participation is treated inmost of the documents. Gender information isgenerally requested in Terms of Reference forevaluations. The extent to which NGOs really are ableto secure the participation of the poorest women andmen is, however, difficult to determine, as is the issueof sustainability of beneficial impact and institutions.This may be due to the fact that NGO funds often go tosmall-scale projects adopting fairly short-termperspectives rather than directly addressing linkages tothe wider politico-economic context. In addition,designating stakeholders may be quite tenuous, interms of desegregation of target populations accordingto gender roles, status etc. Since the late 1980s, therehas been greater emphasis on careful planning inanticipation of impact assessments and evaluationthrough variations of logical framework analysis.Whether such models are or can be renderedparticipatory in orientation is an important question.Some agencies stress planning for (quantifiable) resultsin this context, and this may hamper attention to thelearning process aspects that enhance participation. Wenote also a trend toward an increasing incorporation ofparticipatory evaluation techniques. Some interventionswhich are exploring adoption of a learning processapproach see evaluation as part of an ongoing processof project reporting and monitoring for purposes ofself-management as well as for periodic assessment ina funding agenda.

Conventional Evaluation Criteria

Relevance

The evaluations tend to assess relevance in two ways,namely, against the objectives formulated with respectto a particular intervention and against the widerdevelopment objectives of the organization. Relevancewas also assessed against an analysis of the particulareconomic, political and organizational contexts locallyand nationally, within which an intervention wasframed. This is particularly true for the NGOevaluations. Analytical reports and some project

documents from the late 1980s and the1990s relate assessment of relevance tomeeting the prioritized needs of differentunderprivileged groups.

Effectiveness and Impact

The more comprehensive NGO evaluations(DGIS, FINNIDA, Sida) discuss theanalytical and operational aspects of thesecriteria in nuanced or multidimensionalterms. Effects are discussed in descriptiveterms intended to denote the nature andextent of change promoted by the projectsin a qualitative as well as in a quantitativesense. Concerning participation,distinctions are made in NGO evaluationsbetween direct and indirect impact(FINNIDA), and between immediateproject goals and wider developmentobjectives (Sida). The NGO evaluationsfocused also more clearly on actorperspectives, rather than on projectbeneficiaries as a target group.

Other qualitative, interpretive indicatorsapplied and discussed in the NGOevaluations are responsiveness of projectsand NGOs to pressures "from above" and"from below" and to unforeseen events orproblems as well as linkages to popular orlabor movements, professionalorganizations, etc. The development ofsuch linkages could be indicative ofinstitutional development in a broadersense and thus, of project or NGOeffectiveness. Likewise, the existence ofstrategies explicitly attentive to the activeparticipation of the populations concernedthroughout the project includingmonitoring and evaluation also constituteseffectiveness in that problems may bedetected before they develop into majorsources of conflict. Effectiveness is alsorelated in most of the NGO evaluations tosuch dimensions as internal project groupstructure and functioning and thus toorganizational development.

For several donors, evaluations of supportto the NGO sector are quite new. Thereports evince a conscious focus onpolitical as well as on social or economicgoals. The Terms of Reference for theNGO evaluations are actually moredetailed when it comes to participationthan those for the core-support, bilateral

97

programs. This probably reflects overt donor attentionin recent years to participatory and governance issues,but doubtless also, donor expectations of NGO projectsand programs in terms of participation. The Terms ofReference and other documents compiled in connectionwith the NGO evaluations seem also to allow for agreater degree of flexibility and innovation in terms ofmethods than those for the bilateral programs. Theseevaluations were allowed to take place over anextended period of time, involved a series of relativelylong field visits, and included some participatorytechniques, even if in no instance could they becategorized as fully participatory.

Particular problems concerning impact that wereidentified include the determination of good qualitativeassessment measures, evaluating the relation of studiesand research to promotion activities for those NGOsthat are involved with both kinds of activities, andthose involved in assessing the impact, in amount andkind, of operational advice disseminated from trial andagricultural extension activities. Other problems whichchallenge existing tools for assessing impact are:participation as indicator of a process of humanresource development, degree of partnership betweenprojects and stakeholders, as well as questions of howto meaningfully apply what the authors of the DGISNGO evaluation termed the actors' matrix, (i.e. state;grassroots organizations; local permeativeorganizations or institutions such as the church, unions,political parties; donor agencies; other big NGOs) in anevaluation methodology intended to determine, forexample, the impact of big NGOs at macro levels.

The reputed greater effectiveness of NGOs as analternative to mainstream bilateral programs was notborne out in any general way. Impact in a veryconstrained sense of attaining immediate objectiveswas considered to be achieved by the NGOs, althoughnot in a broader sense of more widespread or sustaineddevelopment. The track record of NGOs as regardseffectiveness is certainly not worse than bilateralprograms and may be better depending on the time-frame and extent of social, economic or politicalimpact concerned.

Efficiency

All the evaluations attempted some kind of adaptedcost-benefit analysis. The CPT Norte II NGOevaluation discussed the difficulty of using budget,number of staff, communities and people reached ascriteria for judging efficiency. Type of programimplemented was used as an assessment marker and theobservation is made that basic services delivery andmanagement seem to lend themselves more readily tolarge-scale support programs than production activities,

particularly in a gender perspective(DGIS/NOVIB). NGO production projects(intended to impact directly on livelihoodsand income-generating capacity) costmuch more than social service ones, but itis not possible to say whether the cost-benefit relation is more favorable for theone or the other kind. Also here, the use oflinkages as an evaluation criterion in termsof multiplier effects and making use of asmall number of staff to achieve widegeographical and structural coverage wasdiscussed in terms of efficiency.

Sustainability and Replicability

The importance of participation in the pre-project planning stage is stated in the NGOevaluations and in various analyticalpapers. Such participation, whichcommonly did not take place according tothe evaluations, was used as oneassessment indicator of potential forsustainability. Sustainability, particularlyof institutions, was found to be low exceptwhere these had some degree of pre-existence, were headed by professionals, orhave been linked into a collaborativenetwork, or even into government. This isdue partly to the contained nature ofproject support and, for NGO support, tothe short time-frame commonly appliedand to the harsh economic and politicalenvironments in which the projectsoperated. Intermediary NGOs may be saidto have a basic operational modality thatmay be replicable to a certain extent withinthe same country. However, the fact thatthese NGOs try to respond and adaptinterventions to the local context meansthat such "replicability" is limited andqualified.

Additional Evaluation Criteria andIssues Related to PopularParticipation

The analytical reports, researchdocuments, and case study materialsyielded criteria in addition to theconventional ones. Combinations of thesewere applied particularly in some of themore innovative evaluations, usuallyinvolving support to NGOs.

Features of a Learning Process

98

Nearly all the evaluations noted an absence of bothbaseline or other useful planning information, and ofactive participation in intervention design, whethermainstream programs or NGO projects. Even wherethere existed some form of appraisal, feasibility orsimilar study, there was little evidence that this wasused and built upon as a conscious monitoring effortwithin a programming system that could incorporateand generate useful new information. Weakest in thisrespect were most of the NGO forms of support. Here,we note the "articles of faith"-type of perceptions ofdonors and of NGOs themselves, as well as the lack ofdonor funding and rigor regarding self-evaluation.Considerably more effort is put into the financialreporting. A trend toward adoption of a more process-oriented approach as opposed to a top-down, blueprintor delivery approach in agency planning andimplementation is noticeable. Demands for "results-oriented" planning monitored in terms of quantifiable,immediately verifiable indicators may, however,weaken such innovations with respect to participation,unless partnership and negotiation in formulatingproject objectives and in the planning of projectcontent are achieved.

Donor Evaluation Culture

Donor agency evaluation culture as defined inprocedural and evaluation manuals can imposeconstraints on participatory evaluation. Renderingconventional project planning cycles more flexible toincorporate participation throughout could improve thissituation. Heretofore, evaluating participation andparticipatory evaluation remain an approach ratherseparate from the mainstream of agency support leftlargely to "the few staff members skilled in the socialsciences or possessing the kind of practical fieldexperience that would allow them to do the analysisrequired"31, to special assessments such as"beneficiary" or "participatory poverty assessments",and to local community and field project levels. In themainstream of evaluation, conventional quantitativemethods are thought to exhibit a high degree ofprecision and objectivity. Participatory evaluation isimplicitly discouraged through associating it with lessreliable evaluation instruments, rather than viewing itas an integral part of a process of partnership indecision-making and ownership.

Donor agencies are, however, attempting to loosen theconfines of conventional evaluation wisdom, as theincreasing incorporation of participatory techniquesdemonstrates. Consultations with "focus groups", "keyinformants", workshops and "sensitization" training areamong the most commonly used techniques. Someagencies are also exploring the possibilities of

participatory planning and projectformulation, as a foundation for sharedmonitoring and evaluation.

CONCLUSIONS ANDSUGGESTIONS FOR FUTUREWORK

A very mixed picture is given by thedocuments reviewed, largely due to aconsiderable diversity of concepts, theabsence of clear concepts of participationor the implicit nature of such concepts.This mixed picture hampers the evaluationof participation. But some positive trendscan also be noted: acknowledgement of theimportance of participation, especially inpolicy documents; and use of someparticipatory techniques in evaluationsreviewed (e.g. focus group discussions, keyinformant interviewing, workshops andsensitization training).

To strengthen these trends, participationneeds to be mainstreamed by developmentagencies (see DAC Orientations, OECDDevelopment Center and World Bankpublications on participatory development)and evaluation tools need to be furtherdeveloped. In other words, the empiricalbase, i.e. the practical experience withevaluation of participation is rather limitedbut there are strong elements (based onpolicy intentions and analytical material)to develop prescriptions for:

How to evaluate participation, namely:

• In a favorable context whichacknowledges the importance ofparticipation, and builds theconcept into planning andimplementation practice.

• As a process with earlystakeholder involvement,distinguishing projects accordingto their participatory or non-participatory origins, whereevaluation is part of continualmonitoring taking placethroughout the projects.Beneficiary participation shouldbe sought in all evaluations, but ismore feasible and can be moreprofound in projects designed for

99

and planned with participation from thebeginning. Here, evaluation is part of alearning and negotiation process involvingdonor agencies, governments and primarystakeholders.

• On the basis of clear concepts and analyticframeworks in which stakeholders areappropriately defined and possibly mistakenassumptions, e.g. about the participatory roleof NGOs, are challenged.

• With indicators reflecting these conceptsand the results of systematic, collaborativeanalysis. Evaluation of participation shouldinclude quantitative and qualitativeassessments in accordance with the levels,dimensions, instruments and indicators ofparticipation identified in the conceptualdiscussion above.

• Preferably in a participatory way, and bymaking participatory evaluation an integralpart of evaluation (as suggested in DACPrinciples for Evaluation pp.23-25).

Future work lies at two levels:

a) at a general level, mainstreamingparticipation in agencies and partnercountries is still a major task to whichevaluators are well suited to contribute (i.e.through intra-agency dialogue on participatorydevelopment), but which needs broadinstitutional support of both leadership andstaff;

b) at a technical and procedural level, throughfurther developing evaluation frameworks,procedures and indicators that betteraccommodate participatory processes.

Promoting donor and partnership learning through trialswith specially conceived evaluations embodying thefollowing elements could be a step in developing theevaluation of participation: policy and operationaldefinitions of participation; the actual use ofparticipatory evaluation; partnership and an activedialogue between the different involved parties, with adeparture point in the primary stakeholders.

Other elements include the identification ofcontextually relevant monitoring and evaluationcriteria and indicators of participation related to:gender, involvement of poor, marginalized categoriesof people, negotiation and the resolution of conflicts

and differences both internally and withrespect to other organizations andinstitutions.

Establish and systematically apply criteriaconnecting participation to the attainmentof autonomy, expansion and diversificationof project activities, such as:

1) the extent to which financing andmanagement are taken over locally, bypartner governments and/or primarystakeholders and their organizations;

2) the expansion or replication of a givenactivity without further project support;

3) diversification, i.e. the degree to whichcapacities created by a project for a givenpurpose are also used for other activities.___________________________

1

The approach paper has been prepared for theEvaluation Division of the Swedish InternationalDevelopment Co-operation Agency (Sida) by AndersRudqvist and Prudence Woodford-Berger at theDevelopment Studies Unit, Stockholm University. Ithas benefitted greatly from the ideas and comments ofHartmut Schneider, OECD Development Center.

2

The terms "popular participation" and "participatorydevelopment" are frequently used synonymously, andas interchangeable concepts. However, proponents ofthe former term argue that it focuses more directly andexplicitly on disadvantaged groups at the grassrootslevel, than the latter more general term. According tothis view, popular participation more clearly reflects apreferential option for the poor in developmentprograms.

3

These analytical documents appear in the Notes andReferences section.

100

4

Fortin, C. and M. Stiefel, 1985, Of People, Power and Participation:An Overview of the Popular Participation Project, Geneva: UNRISD,p.11; and Stiefel, M. and M. Wolfe, 1994, A Voice for the Excluded:Popular Participation in Development, Utopia or Necessity? Genevaand London, UNRISD and Zed Books.

5

Bhatnagar, B. and A. Williams (eds.), 1992, ParticipatoryDevelopment and the World Bank: Potential Directions for Change,World Bank Discussion Papers 183, p.177. Washington D.C.: TheWorld Bank.

6

World Bank Operations Policy Department, 1994, The World Bankand Participation, pp.1-2, Washington D.C.: The World Bank.

7

Bhatnagar and Williams, Op.cit., pp.179-180.

8

DAC Orientations on Participatory Development and GoodGovernance, OECD/GD, 1993.

9

Schneider, H. and M-H. Libercier, 1994, "Concepts, Issues andExperiences for Building Up Participation" (p.3), in ParticipatoryDevelopment From Advocacy to Action, Paris: OECD DevelopmentCenter.

10

Rudqvist, A., 1987, Popular Participation: Levels and Dimensions,Stockholm: Development Studies Unit, Popular Participation Program,Stockholm University; and 1991, Guidelines for Consultations andPopular Participation in Development Processes and Projects,Development Studies Unit, Popular Participation Program, StockholmUniversity; as well as Bhatnagar and Williams, Op.cit., p. 177ff.

11

Development Studies Unit, 1991, Op.cit., pp 10-11, as well asAppendix 2.

12

The issue of whether WID and gender aspects and perspectives areadequately addressed in discussions of participation has been treatedin a number of works. Some of these which argue most effectivelyagainst the assumption that 'participation' logically or automaticallyincorporates gender, or that gender equity is merely a component ofparticipation are: Bamberger, M., M. Blackden and A. Tadesse"Gender Issues in Participation", a paper prepared for the World BankWorkshop on Participatory Development, Washington D.C., May 17-20 1994; Gender Issues in Bank Lending: An Overview. WashingtonD.C.: The World Bank Operations Policy Department, 1994; and V.Siddharth, 1995, "Gendered Participation: NGOs and The WorldBank", in IDS Bulletin, Vol. 26, no. 3, pp 31-38.

13

Gezelius, H. and D. Millwood, 1988, NGOs in Development andParticipation in Practice: An Initial Inquiry, Stockholm: DevelopmentStudies Unit, Popular Participation Program, Stockholm University.Working Paper 3.

14

Weekes-Vagliani, W, 1994, "Participatory Development andGender", in Participatory Development From Advocacy to Action, H.Schneider, and M-H. Libercier (eds.), Paris: OECD DevelopmentCenter.

15

Both public sector organizations and NGOs exist in a variety ofmanifestations and are obviously significant social actors forparticipation. Space and the mandate for this approach paper do notallow for a review here of different categories of organization. Werefer the reader to the classifications and detailed discussions in thefollowing works: M.J. Esman and N. Uphoff Local Organizations -

Intermediaries in Rural Development, Ithaca N.Y.:Cornell University Press, 1984; N. Uphoff LocalInstitutional Development: An Analytical Sourcebookwith Cases, West Hartford, Connecticut: KumarianPress, 1986; and S. Burkey People First - A Guide toSelf-Reliant Participatory Rural Development,London: Zed Books, 1993.

16

World Bank Operations Policy Department, 1994,The World Bank and Participation, p.i.

17

Ibid., p. 21ff.

18

D. Narayan, 1994, The Contribution of People'sParticipation: 121 Rural Water Supply Projects,Washington D.C.: The World Bank.

19

The World Bank and Participation, pp. 23-24.

20

N. Uphoff, 1992, "Monitoring and EvaluatingPopular Participation in World Bank-AssistedProjects", in B. Bhatnagar and A. Williams (eds.)Participatory Development and the World Bank:Potential Directions for Change (World BankDiscussion Papers 183), p. 144. Washington D.C.: TheWorld Bank.

21

Oakley, P. et al, 1991, Projects With People, ThePractice of Participation in Rural Development,Geneva: ILO, pp.243-245.

22

Ibid., p.250

23

Huizer, G., 1983, Guiding Principles for People'sParticipation Projects, Rome: FAO, (pp.55-65), and

Oakley, Op.cit., p. 256-257.

24

L. Salmen, 1987, Listen to the People: ParticipantObserver Evaluation of Development Projects, TheWorld Bank. New York: Oxford University Press.

25

Chambers, R., 1992, Rural Appraisal: Rapid,Relaxed and Participatory, Sussex: IDS, DiscussionPaper no. 311, pp. 12-13.

26

Marsden, D., P. Oakley, and B. Pratt, 1994,Measuring the Process: Guidelines for EvaluatingSocial Development, Oxford: Intrac Publications, p.96.

27

Operations Policy Department, 1994, The WorldBank and Participation, pp. 21-24. Washington D.C.:The World Bank.

28

Marsden, D., P. Oakley, and B. Pratt, 1994,Measuring the Process: Guidelines for EvaluatingSocial Development, Oxford: Intrac Publications (p.16); and Guba, E.G., and Y.S. Lincoln, 1989, FourthGeneration Evaluation, London: Sage.

29

Guba and Lincoln, Op.cit., pp. 35-38; Marsden,Oakley and Pratt, Op.cit., p. 30.

101

30

Schneider, H. Participatory Development From Advocacy to Action,Paris: OECD Development Center OECD, pp.21-22.

31

Bhatnagar, B and A. Williams (eds.), 1992, ParticipatoryDevelopment and the World Bank: Potential Directions for Change,World Bank Discussion Papers 183, p.171. Washington D.C.: The

World Bank.

ANNEXES

ANNEX 1 - LEGAL SYSTEMSREFERENCES

BLAIR Harry, MILLSAP William, MUDGE Arthur, and SAID Mary, 1993, A Strategic Assessment of Legal SystemsDevelopment in Colombia, USAID, Washington D.C.

BLAIR Harry, SAID Mary and SILLIMAN Sidney, 1993, "A Strategic Assessment of Legal Systems Development in thePhilippines", Draft, USAID, Washington D.C.

DARNTON John, 1993, "England's Judges (and Hiring System) Under Fire", New York Times, 18 July.

DE SILVA Rangita, and SANJEEVA Jayawardena. 1993. "Women and Violence: A Socio Legal Study" Colombo, Sri Lanka:Legal Aid Centre, Faculty of Law, University of Colombo.

FREEDOM HOUSE, 1993, Freedom House Review: 24(1): 3-22

GEORGE Terrence, 1991,"Toward Justice: Foundation Programs in Human Rights and Social Justice and Governance andPublic Policy in the Philippines", Unpublished paper, Ford Foundation, 20 February.

GREGORIO Carlos G, 1993, "Programa de Mejoras en la Administracion del Justicia en la Republica Argentina: InvestigacionPiloto Sobre Duracion del Proceso Judicial", Final report, USAID and Fundacion la Ley, February, Buenos Aires, Argentina.

GUNAWARDHANA Anandi, NISHAN Muthukrishna, PARVEEN Mubarak, and SAVITHRI Walatara, editors, 1993. Lawyersas Social Engineers: A Report by South Asian Law Students, Colombo, Sri Lanka: Legal Aid Centre, Faculty of Law, Universityof Colombo.

HANSEN Gary, MILLSAP William, SMITH Ralph, and SAID Mary, 1993, A Strategic Assessment of Legal SystemsDevelopment in Honduras, Technical Report No. 10, USAID, Washington D.C.

HANSEN Gary, 1993, "Doing Democracy in the Third World: Developing an Applied Framework for Legal Development",Paper presented for the annual meeting of the American Political Science Association. Washington D.C. 2-5 September 1993.

HANSEN Gary, SAID Mary, OBERST Robert, and VAVRE Jacki, 1994, A Strategic Assessment of Legal Systems Developmentin Sri Lanka, USAID, Washington D.C.

HEIN Gordon R., 1993, "The Asia Foundation's Law Programs", Paper presented at the Association of American Law Schools1993 Annual Meeting, San Francisco, 8 January.

JENSEN Erik G, 1993, "Views on Strategic Law Programming: Challenges to Strengthening Democracy Through the LegalSystem", Draft, Manila, Philippines: Asia Foundation.

LECCI Gail, 1991, "El Salvador Judicial Reform Update: 1984 to the Present".

LEGADA Ric Tan, and DEMAREE JAB Ravel, 1988, Administration of Justice in the Philippines: Organizational Structureand Management of the Judiciary, Final Report on Module III, Queen City: Institute of Judicial Administration, University ofthe Philippines.

MACHETE BUSINESS CLUB, 1993, "MBE members bullish about 1993 but cite power, police and judiciary as problemareas", Press release, 4 March.

PODER JUDICIAL, COURT SUPREME DE JUSTICIA DE LA NATION, 1993, "Bolton Estadistico", Buenos Aires, Argentina:Secretaria letrada de Estadisticas.

ROMULO Ricardo J, 1985. "Results of the BBC-PBA Nationwide Opinion Survey on the Legal Profession and the Judiciary",PBA Newsletter, Manila, Philippines.

SOCIAL WEATHER STATIONS, 1992, "Social Weather Report Survey", Manila, Philippines, December.

STEINBERG David I., 1992, "The Administration of Justice in Asia: A Conceptual and Project Survey", Draft, Arlington,Virginia: Development Associates, Inc.

UNITED NATIONS DEVELOPMENT PROGRAMME, 1991, Human Development Report, New York: Oxford UniversityPress.

USAID/GUATEMALA, 1991, "Stocktaking of 1986-1991: Administration of Justice Program", USAID/Guatemala.

VELEZ B. Eduardo, GOMEZ DE LEON Patricia, and GERALDO A Jaime, 1987, Jueces y Justicia en Colombia,Bogota, Colombia: Instituo SER, October.

ANNEX 2 - PRIVATIZATIONBIBLIOGRAPHY

ADAM Christopher, CAVENDISH William and MISTRY Percy S., 1992, Adjusting Privatization, James Curry, London.

ASPE Pedro, 1991, "Thoughts on structural transformation in Mexico: The case of privatization of public sector enterprises”,Address given at the World Affairs Council in Los Angeles.

BARNES Guillermo, 1992, Lessons from Bank Privatization in Mexico, World Bank, Washington D.C., November.

BELKA Marek, ESTRIN Saul, SCHAFFER Mark E. and SINGH I.J, 1994, “Enterprise adjustment in Poland”, paper presented atthe Workshop on Enterprise Adjustment in Eastern Europe, World Bank, PRDTE, Washington D.C., September.

BOBEVA Daniela, 1995, "Post-privatization behaviour of enterprises in Bulgaria", paper presented to the OECD Advisory Groupon Privatization, Moscow 29-31, March.

BOUIN O. and MICHALET Ch-A., 1991, Rebalancing the public and private sectors: Developing country experience, OECDDevelopment Center.

BOYKO Maxim, 1995, "Performance of enterprises in the Russian Federation", paper presented to the OECD Advisory Group onPrivatization, Moscow 29-31, March.

BOYKO Maxim, SHLEIFER Andrei and VISHNY Robert, 1995, “Privatizing Russia”, Institute of Technology, Massachusetts.

BOWEN Sally, 1995, “Interbanc makes the most of new-found freedom", Financial Times, August.

BUKOWSKI Jacek, 1995, "Performance of Polish Small and Medium-Sized Enterprises Privatized through Leveraged Buyout", apaper presented to the OECD Advisory Group on Privatization, Moscow, 29-31, March.

CALLAGHY Thomas M. and WILSON III Ernest James, 1988, "Africa: Policy, reality or ritual", in The Promise ofPrivatization: A Challenge for US Policy, Raymond Vernon (ed.), Council on Foreign Relations, New York.

CAPEL James Inc., 1995, Telecoms in Latin America, May.

CARANA CORPORATION, 1991, “Evaluation of project 522-0289, Privatization of State-Owned Enterprises in Honduras”,USAID, Washington D.C.

CARLIN Wendy, VAN REENEN John and WOLFE Toby, 1995, “Enterprise Restructuring in early transition: the case studyevidence from Central and Eastern Europe”, Economics of Transition, Volume 3(4), 427-458.

CESKA Roman, 1995, "Performance of Enterprises in the Czech Republic", paper presented to the OECD Advisory Group onPrivatization, Moscow, 29-31, March.

CSEPI Lajos and ILLES Maria, 1995, "Performance of Privatized Enterprises in Hungary", Conference paper, OECD AdvisoryGroup on Privatization, Moscow, March.

CSEPI Lajos and ERZSEBET Lukacs, 1994, "Privatization in Hungary - 1994" in Privatization in Central and Eastern Europe,CEEPN.

LOPEZ-DE-SILANES F., 1993, "Determinants of privatization prices", Mimeo, National Bureau of Economic Research, HarvardUniversity.

DIEWERT W.E., 1992, Efficiency measures for diesel electric power plants, IENED, The World Bank.

EARLE John S., ESTRIN Saul and LESHENKO Larisa L., 1995, “Ownership structures, patterns of control and enterprisebehaviour in Russia”, Unpublished final draft, August.

EBRD, “Transition Report 1995”.

GONZALEZ FRAGA Javier A., 1991, "Argentine Privatization in Retrospect", in Privatization of Public Enterprises in LatinAmerica, International Center for Economic Growth, San Francisco.

FREELAND Chrystia, 1995, "Factory of factories offers hope for reforms", Financial Times, April 15/16.

FUKUI Ryu, 1992, “Study on the historical performance of the public enterprise sector in Guinea”, Japan Economic ResearchInstitute.

“Gesamtwirtschaftliche und unternehmerische Anpassungsprozess in Ostdeutschland. Sechster Brericht”. In: Wochenberichtdes DIW, Nr. 39/1992 and Nr. 15/1994

GALAL Ahmed, JONES Leroy, TANDON Pankau and VOGELSANG Ingo, 1994, Welfare Consequences of Selling PublicEnterprises: An Empirical Analysis, World Bank, Washington D.C.

GALAL Ahmed and SHIRLEY Mary et al, 1995, Bureaucrats in Business, World Bank, September.

HACHETTE D. and LUDERS R., 1993, Privatization in Chile, San Francisco, ISC Press.

HARWOOD John and RUFF Jacquelyn, 1995, "Launching modern telecommunications systems", in Global Telecoms Business,February/March.

JONES, L.P., TANDON Pankaj, and VOGELSANG Ingo, 1990, Selling Public Enterprise: A Cost Benefit Methodology, MITPress, Cambridge, Massachusetts.

KIKERI Sunita, NELLIS John and SHIRLEY Mary, 1992, Privatization: The Lessons of Experience, World Bank, WashingtonD.C.

Konjunkturbericht 3/94 des Institut fur Wirtscaftsforschung Halle.

KPMG Management Consulting, 1994,"Evaluation of assistance to the Czech and Slovak Voucher Mass Privatization Schemes",ODA Evaluation Department, November.

KPMG Management Consulting, 1994, "Polish pilot privatization project evaluation", ODA Evaluation Department, November.

LORCH Klaus, 1991, "Privatization through private sale: The Bangladeshi textile industry", in Privatization and Control ofState-Owned Enterprises, Ravi Ramamurti and Raymond Vernon, World Bank, Washington D.C.

LUDERS Rolf, 1990,"Privatization in Chile: Lessons from a massive divestiture programme in a developing country".

LUDERS Rolf and HACHETTE Dominique, 1987.

MAZUR Marek, 1995, "Assessment of the situation of privatized enterprises and banks in Poland", paper presented to the OECDAdvisory Group on Privatization, Moscow 29-31, March.

MEGGINSON William L.,.NASH Robert C. and VAN RANDENBORGH Matthias, 1994, "The Financial and OperatingPerformance of Newly Privatised Firms: An International Empirical Analysis", in The Journal of Finance, Vol. XLIX, No. 2,June.

NANKANI Helen, 1988, Techniques of Privatization of State-Owned Enterprises, Vol. 2, Selected Case Studies, World BankWashington D.C.

PIETRZAK Edmund, 1992, The Development of the Capital Market in the Process of Transformation of Poland's Economy intoa Market System, Gdansk Institute for Market Economics, Gdansk.

PINERA Jose and GLADE William, 1991, "Privatization in Chile", in "Privatization of Public Enterprises in Latin America",International Center for Economic Growth, San Francisco.

SADER Frank, 1993, "Privatizations and Foreign Investment in the Developing World": 1988-1992, World Bank InternationalEconomics Department, Washington D.C.

SISTEMA ECONOMICO LATINAMERICANO, 1995, The Latin American and Carribean region's bet on privatization,Caracas, July.

WEBSTER Leila M., FRANZ Juegen, ARTIMIEV Igor & WACKMAN Harold, 1994, Newly Privatized Russian Enterprises,World Bank, Washington D.C., October.

WORLD BANK, 1994, “Assistance to Privatization in Developing Countries”, Operations Evaluation Department, World Bank,Washington D.C., August, Volume II, Annexes.

WORLD BANK, 1995, Performance Audit Report: Mexico - Road Transport and Telecommunications Sector Adjustment Loan,Operations Evaluation Department , Report No. 14400, April.

YOUNG Peter and MOORE John, 1991, “The Great White Elephant Sale - A Review of Privatization in Togo”, (unpublishedmanuscript)

YOUNG Peter and REYNOLDS Paul, 1994, The Amnesia of Reform, Adam Smith Institute, London.

YOUNG Peter, 1995, "Dial C for Choice", Worldlink magazine, September/October.

ANNEX 3 - INSTITUTIONAL DEVELOPMENT (ID)REFERENCES

ACTION AID, 1990, The Effectiveness of British Aid for Training, Development Report, Hulme D. for Action Aid.

ADAMOLEKUN, L., 1993a, “Note on Civil Service Policy Reform in Sub-Saharan Africa”, International Journal of PublicSector Management, 1993. 6(3).

ADAMOLEKUN, L., 1993b, “Institutional Perspectives on Africa's Development Crisis”, International Journal of Public SectorManagement 1993 3(2).

ADB, 1994, Report of Task Force on Improving Project Quality, Manila, January.

BERG, 1993, Rethinking Technical Cooperation: Reforms for Capacity Building in Africa, Berg E. for UNDP/DAI New York.

BUYCK, B., 1991, The Bank's Use of Technical Assistance for ID, IBRD.

CHOPRA, 1990, “Aid in the Nineties: Institutional Development - Lesson of Experience”, R K Chopra (Director OED, IBRD)paper delivered to National Advisory Council for Development Cooperation, The Hague, August.

COOPER, L., 1984, The Twinning of Institutions: Its use as a Technical Assistance Delivery System, IBRD.

DANIDA, 1988, Institutional Aspects of Danish Project Assistance: Issues and recommendations on ID.

DANIDA, 1992, Institutional Issues in Danida projects: A Synthesis, Evaluation Report.

DANIDA, 1993, Institutional Development: Issues Paper, January.

DAVIDSON, B., 1992, The Black Man's Burden: Africa and the Curse of the Nation-State, New York Times Books.

DAVIDSON, F., 1993, “Gearing up for Effective Management of Urban Development”, Cities Vol. 1993.

DECKER and VAN LAAR, 1995, Policy administration and development: a select bibliography, Dekker A. and van Laar J.(compilers), ECDPM Maastricht and Rijks University, Leiden, March.

DGIS, 1988, Program Evaluation, Western Province, Zambia 1979-87 IOV, Dutch Ministry of Foreign Affairs, October.

DGIS, 1992, Final Evaluation Report on Human Resource Development Project, Dutch Ministry of Foreign Affairs, August.

DGIS, 1993a, Report of Evalution Mission on Strengthening of Planning and Development Department Project, Baluchistan,Pakistan, Dutch Ministry of Foreign Affairs, August.

DGIS, 1993b, Report of Evaluation Mission on Strengthening of Planning and Development Department project, NWFPPakistan, September.

DGIS, 1995, “Institutional Development in DGIS: Theory and Practice”, P. de Haan, et al. Paper prepared for anOECD/IBRD/UNDP Technical Cooperation Workshop on Institutional Development in Aid Cooperation, Eschborn, May.

ECDPM, 1993, Capacity Building for Governance and Public Sector Management: Programme Description, November.

ECDPM, 1994, A Governance Approach to Civil Service Reform: Operational Considerations, Report of a consultativemeeting, Corkery J. and A. Land ECDPM, June.

EC, 1995, Institutional Appraisal and Development Methodology Project: Phase One - Synthesis Report (Four Volumes), BMBManagement Consulting for Development for the Commission of the European Communities.

FRERKS G. and OTTO J.M., 1995, “Decentralization for development? Between Myth and Panacea”. Paper presented for theSeminar on Decentralization and Development, organized by WBIO, Netherlands, June.

GARY, 1990, Institutional Development Work in the Bank, A Review of 84 Bank Projects Gray C., et al. IBRD.

GTZ, 1992, Institutional Pluralism: Focus of German Technical Cooperation, Eylers H. and Sulzer R., Tokyo, December.

IBRD, 1989, The Reform of State Owned Enterprises: Lessons from Bank Lending, Shirley M., Policy and Research Series, No.4.

IBRD, 1991, The Reform of Public Sector Management: Lessons from Experience, Policy and Research Series No. 18.

IBRD, 1992a, “Indigenous Management Practices: Lessons for Africa's Management in the 90s”, Concept Paper for RegionalStudy, M. Dia, April.

IBRD, 1992b, Effective Implementation: Key to Development Impact, Report of the World Bank's Portfolio Management TaskForce (the Wapenhans Report).

IBRD, 1993a, Governance: The World Bank's Experience, November.

IBRD, 1993b, “A Governance Approach to Civil Service Reform in Sub-Saharan Africa”, Technical Paper 225 M. Dia.

IBRD, 1993c, “Indonesia - Civil Service Issues”, Informal Discussion Paper, October.

IBRD, 1994a, Projectizing the Governance Approach to Civil Service Reform: An Institutional Environment Assessment forSecal Preparation in the Gambia, Pinto R.F. and A.J. Mrope, May.

IBRD, 1994b, Evaluation Results 1992, Operations Evaluation Department.

IBRD, 1994c, Improving the Management of Technical Assistance for Institutional Development, Organization Design Staff,August.

IBRD, 1994d, World Development Report 1994, Infrastructure for Development.

IBRD, 1995, Aide Memoire - Zambia Public Sector Management Review Mission, July.

ISLAM, 1993, “Public Enterprise Reform: Managerial Autonomy, Accountability and Performance Contracts” , PublicAdministration and Development, Vol 13.

ISRAEL, A., 1987, Institutional Development, OUP/IBRD.

ISRAEL, A. and HEAVER, R., 1986, “Country Commitment to Development Projects”, IBRD Discussion Paper 4.

KAIJAGE, G.J., 1993, Management Consulting in Africa: Utilizing local expertise, Kumarian Press, West Hartford.

LETHEM, F. and COOPER, L., 1993, Management Project-Related TA: the Lessons of Success, IBRD.

MUTAHABA G. AND BALOGUN M. (eds), 1992, Enhancing Policy Management Capacity in Africa, Kumarian Press.

MUTAHABA G., BAGUMA R., HALFANI M., 1993, Vitalizing African Public Administration for Recovery and Development,Kumarian Press.

MCGILL, 1993, “Institution Building for a Third World City Council” by McGill R International Journal of Public SectorManagement 1993 6(5).

NARAYAN ASUWAMI, 1991, “Institutional Building for Development: Lessons Learned and Tasks Ahead”.Narayanasuwami, Asian Development Review No 2.

ODA, 1990, “Evaluation of the Training Project for the Indian Railways Modernization of Workshops” Report EV472 ODAEvaluation Department, March.

ODA, 1993a, Evaluation of ODA Project in Support of Ghana Civil Service Reform Programme, London, September.

ODA, 1993b, Evaluation of Uganda Public Administration Training and Institutional Development Project, September.

ODA, 1993c, Evaluation of ODA Assistance to Kenya Railways in the Second Railway Project, September.

ODA, 1994a, Synthesis Study on Institutional Strengthening Projects and Experience, May.

ODA, 1994b, “Civil Service Reform Experience”, Record of Discussions and Suggestions for Further Review, Government andInstitutions Department, May.

ODA, 1994c, Interim Evaluation: Cairo Wasterwater Project EV539, March.

ODA, 1994d, Greater Dhaka Power Transmission and Distribution Project Phases II and IIA EV550, January.

ODI, 1993, “Aid Effectiveness: Impact on Institutional Capacity (Section 4); Policy Conditions and Institutional Reform(Section 5)” of paper on Aid Effectiveness for IBRD (unpublished) for IBRD.

OECD/DAC, 1993, DAC Orientations: Participatory Development and Good Government .

PADCO, 1985, Final Report on the Local Government Training Project, Washington D.C.

PAUL, S., 1990, Institutional Development in World Bank Projects, IBRD

P/W, 1991, Recent Activities in Combatting Fraud and Corruption in Government, Conference on New Developments inGovernment Financial Management, Price Waterhouse, April

QUARLES VANUFFORD, 1988, The Hidden Crisis in Development: Development Bureaucracies, Quarles Vanufford, P., DKruijt, and T Downing Eds. Free University Press Amsterdam

RONDINELLI, D, 1992, “UNDP Assistance for Urban Development: An Assessment of Institution Building Activities inDeveloping Countries”, International Review of Administrative Sciences 58(4).

RONDINELLI, D, 1993, Development Project as Policy Experiments (Second Edition) Routledge, London.

SIDA, 1991, Making Government Work.

SIDA, 1994a, Annual Report.

SIDA, 1994b, Institution Building as a Development Assistance Method: A Review of the Literature and Ideas Moore, M.,Stewart, S. and Hudock, A., IDS, Sussex, for SIDA, September.

UNDP, 1990, Foreign Aid Accountability: Perspectives of Donors and Recipients UN, September.

UNDP, 1992, A Study of Government Monitoring and Evaluation Systems: The Case of Uganda, Central Evaluation Office,UNDP, December.

UNDP, 1994, Process Consultation: Systemic Improvement of Public Sector Management.

UNDP/HABITAT, 1993, “Elements of Urban Management” Davey K. J. for Urban Management Programme (No. 11).

UNDP/HABITAT, 1993b, “Decentralization and its Implications for Urban Service Delivery” W. Dillinger for UrbanManagement Programme (No 16).

UNDP/MDP, 1995a, A Review of the Civil Service Reform Programme in Ghana, I. Mackinson, M. A. Bentil and A. KordaUNDP/MDP, March.

UNDP/MDP, 1995b, Uganda Civil Service Reform Case Study, K. Brown, K. Kiragu, and S. Villadsden, June.

UNDP/MDP, 1995c, Civil Service Reform in Tanzania A Case Study prepared by NORAD, May (Draft).

USAID, 1975, Project Appraisal Report on the General Public Administration (Comptroller general) Project, Peru, July.

USAID, 1977, Public Sector Implementation: An Overview and a Bibliography, Syracuse University for USAID.

USAID, 1979, Improving Public Services Report of a Conference.

USAID, 1982a, “Effective Institution Building”: USAID Programme Evaluation Discussion Paper, March.

USAID, 1982b, Impact Study on Institute of Public Administration, Liberia, April.

USAID, 1982c, Impact Study on Civil Service Administration, Liberia, April.

USAID, 1983, Local Government Trends and Performance: Assessment of Aid's Involvement in Latin America.

USAID, 1985, Implementation Policy and Institutional Change via Performance Disbursement: USAID Occasional paper No 1,July.

USAID, 1986a, Experience with Public Management Training in LDCs: Paper submitted to DAC, February.

USAID, 1986b, Managing Development Programmes: Management Strategies and Project Interventions in Six AfricanAgricultural Projects: January.

USAID, 1986c, Development Management in Africa: Context and Strategy: A Synthesis of Six Agricultural Projects, January.

USAID, 1987, Measures to Control Bureaucratic Corruption in Asia: A Review of the Literature, M.H. Bidus for USAID.

USAID, 1988, Report on Civil Service Reform Efforts with particular reference to Sub-Saharan Africa: An Overivew withLessons for Liberia, July.

USAID, 1990, USAID and African Capacity Building: Report by HIID, September.

USAID, 1992a, The Internal Organization of Government: Institute of Policy Reform, Washington D.C., June.

USAID, 1992b, Credibility, Rent-Seeking and Political Instability: IRIS/USAID Working-Paper No 31, October.

USAID, 1992c, Respondacon II: Second Inter-American Conference on the Problems of Fraud and Corruption in Government.

USAID, 1992d, Concept of Governance and its Implications for AID's Development Assistance Programme in Africa, June.

USAID, 1994, Decentralized Public Service Provision in Sub-Saharan Africa: A False Start: Lessons from theDecentralization: Finance and Management Project.

ANNEX 4 - EVALUATION CAPACITY BUILDINGREFERENCES

ALBERT, Hugues and ABOUTABIT, El Mostapha (UPDATED BY ABOUTABIT E.M.); A Study of Government Monitoringand Evaluation Systems: The Case of Morocco, OESP/UNDP, New York, 1995.

ASIAN DEVELOPMENT BANK; Regional Seminar on Performance Evaluation in Asia and the Pacific - Summary ofProceedings, ADB, Manila, 1992.

ASIAN DEVELOPMENT BANK; Post-evaluation and Feedback: Realities and Challenges in the Asian and Pacific Region,Manila 1995.

AUSAID; Proceedings of the DAC Expert Group on Aid Evaluation Canberra Seminar, Canberra, 1996.

BINNENDIJK, Annette; Managing for Results: A Case Study of the "Ecuador Experiment", PPC/CDIE, Washington, 1993.

BOWDEN, Peter; National Monitoring and Evaluation: Development Programmes in the Third World, Avebury, 1988.

CARRIBBEAN DEVELOPMENT BANK & INTER-AMERICAN DEVELOPMENT BANK; Regional Seminar on Monitoringand Evaluation in the Carribbean: Strengthening Evaluation Capabilities for Sustainable Development, Barbados, 1994.

CENTRAL AMERICAN EVALUATION ASSOCIATION; INTERNATIONAL FUND FOR AGRICULTURALDEVELOPMENT & INTER-AMERICAN DEVELOPMENT BANK; First Evaluation Seminar for Central America, Panamaand the Dominican Republic, San Jose, 1994.

COMMANDER, Simon; Government Monitoring and Evaluation Systems in Africa - What are the Questions?, CEO/UNDP,New York, 1987.

CUMMINGHAM, I an; The Wisdom of Strategic Learning: The Self Managed Learning Solution, McGraw Hill, London, 1994.

INTER-AMERICAN DEVELOPMENT BANK & OECD; Regional Seminar on Monitoring and Evaluation in Latin Americaand the Carribbean: Strengthening Evaluation Capabilities for Sustainaable Development, OECD, Paris, 1994.

KHAN, M. Adil a.o.; A Study of Government Monitoring and Evaluation Systems: The Case of the Philippines, CEO/UNDP,New York, 1993.

KHAN, M. Adil; Generic Issues in Monitoring and Evaluation: What Works and What Does Not?; CEO/UNDP, New York1993.

KHAN, M. Adil; Initiatives and Efficiency: Experience of Malaysia in Monitoring and Evaluation, CEO/UNDP, New York1991.

KHAN, KATUNZE and TWODO; A Study of Government Monitoring and Evaluation Systems: The Case of Uganda,CEO/UNDP, New York, 1992.

KHAN, M. Adil and MAHLAHLA, S.; Search for Efficiency: Evaluation of Zimbabwe, CEO/UNDP, New York, 1991.

KHAN, M. Adil and PARTADIREDJA, Ace; A Study of Government Monitoring and Evaluation Systems: The Case ofIndonesia, CEO/UNDP, New York, 1993.

KHAN, M Adil and WANGCHUK, Namgyal; A Study of Government Monitoring and Evalution Systems: The Case of Bhutan,CEO/UNDP, New York, 1993.

KRUSE, Stein-Erik, Can Projects Learn? - Lessons from a Monitoring & Evaluation Project in Kenya, (mimeo) Niarobi, 1994.

LEEUW, RIST and SONNICHSEN; Can Government Learn? Transaction Publishers, New Brunswick and London, 1994.

MAYNE, J. et Al; Advancing Public Policy Evaluation: Learning from International Experiences, North-Holland, theNetherlands, 1992.

OECD; Evaluation in Developing Countries: A Step in a Dialogue, OECD, Paris, 1998.

RIST, Ray C.; Programme Evaluation and the Management of Government - Patterns and Prospects Across Eight Nations,Transaction Publishers, New Brunswick and London, 1990.

ROVANI, Yves; Presentation for the Brazilian Government's Seminar on a National System of Integrated Evaluation, WorldBank, 1989.

ROY, Roger F. and SAVANÉ, Djiguiba Sy; Revue des systémes de suivi et d'évaluation gouvernementaux: Le cas de la Guinée,CEO/UNDP, New York, 1993.

ROY, Roger F. and VELLOSO, Paul W.R.; A Study of Government Monitoring and Evaluation Systems: The Case of Brazil,CEO/UNDP, New York, 1992.

ROY, Roger F. and WIJAYASURIYA, P.M.W.; A Study of Government Monitoring and Evaluation Systems: The Case of SriLanka, CEO/UNDP, New York, 1992.

SIDA; Training for Public Service - An Evaluation of Sweden Co-operation with Zimbabwe in Public Service Training 1982-86.Dufborg, Jones and Lewin. Sida Evaluation Report 1987/5, Stockholm, 1987.

UNDP; National Monitoring and Evaluation in India, CEO/UNDP, New York, 1992.

USAID; “Managing for Results: Experience in Two USAID Missions”, USAID Managing for Results Working Paper No. 1,1994.

USAID Evaluation News, Various issues, Center for Development Information and Evaluation (CDIE).

WORLD BANK, Lessons & Practices, Various issues, Operations Evaluation Department (OED).

WORLD BANK; “A System for Evaluating the Performance of Government-Invested Enterprises in the Republic of Korea”,Park, Young C. World Bank Discussion Paper No. 3, Washington D.C., 1986.

WORLD BANK; “ From Macroeconomic Correction to Public Sector Reform - The Critical Role of Evaluation”, EduardoWiesner D. World Bank Discussion Paper No. 214, Washington D.C., 1993.

WORLD BANK; Monitoring and Evaluation Development Projects: The South Asian Experience. Ahmed, Viqar andBamberger, Michael, EDI Seminar Series, Washington D.C., 1989.

WORLD BANK; “ Performance Evaluation for Enterprises”, Jones, Leroy P. World Bank Discussion Papers No. 122,Washington D.C., 1992.

WORLD BANK; Evaluation Capacity Development; Report of the Task Force, Washington D.C., 1994.

ANNEX 5 - DECENTRALIZATIONREFERENCES

AGARWALA, R., 1992, “China: Reforming Intergovernmental Fiscal Relations”, World Bank. Discussion Papers No. 178.Washington D.C.

ALAM, HUQUE and WESTERGAARD, 1994, Development through Decentralization in Bangladesh, The University PressLimited, Bangladesh.

BIDUS, M.H., 1995, A review of Decentralization and Municipal Development Initiatives and their Effect on Democratization inCentral America, International City/County Management Association, Municipal Development and Management Program.

BLAIR, 1995, “Assessing Democratic Decentralization: A CDIE Concept Paper”, Paper prepared for USAID, Draft. WashingtonD.C., September.

BRETT, 1992, Providing for the Rural Poor: Institutional Decay and Transformation in Uganda. Institute of DevelopmentStudies, Brighton Sussex BN1 9RE, England.

CHOLE, 1994, Fiscal Decentralization in Ethiopia. Addis Ababa, 1994.

CROOK, R., 1994, “Four Years of Ghana's District Assemblies in Operation”, Public Administration and Development, vol 14,October 1994.

CROOK and MANOR, 1994, Enhancing Participation and Institutional Performance: DemocraticDecentralization in South Asia and West Africa, London, Report to ESCOR, the Overseas Development Administration, to bepublished by Cambridge University Press.

DALE, R., 1992, Organization of Regional Development, Sarvodaya, Ratmalana.

DANIDA, 1995a, Joint Review of Municipal Strengthening in Region I, Nicaragua, Copenhagen.

DANIDA, 1995b, Review of Danish Assistance to Decentralization of Government in Uganda, Copenhagen.

GOZ, 1993, The public sector reform program (PSRP), Government of Zambia. Lusaka.

GOZ/LOGOSP, 1993, Terms of Reference for a Consultancy to Implement the Proposed Local Government Support Project(LOGOSP) in Zambia. Lusaka.

GRINDLE & HILDERBRAND, 1995, "Building sustainable capacity in the public sector: what can be done?" In PublicAdministration and Development, Volume 15, Number 5, December 1995.

INGHAM, B. and KALAM, 1992, “Decentralization and development theory and evidence from Bangladesh.” In: PublicAdministration and Development. Vol. 12. No. 4. London.

LALEYE, O.M and OLOWU, D., 1990, “Decentralization in Africa”, in: WB & IIA 1990.

LANGSETH, 1995, "Civil service reform in Uganda: lessons learned", in Public Administration and Development. Volume 15,Number 4, October 1995.

LAUGLO and MOLUTSI, 1994, Decentralization and Health Systems Performance: The Botswana Case Study. DIS (Oslo)/TheUniversity of Botswana.

LAUGLO AND MOLUTSI, 1995, Transferring Regional Health Teams to Local Councils: Decentralization and Health SystemsPerformance. The Botswana Case Study. DIS (Oslo)/The University of Botswana.

LOGOSP, 1993, Terms of reference for a consultancy to implement the proposed Local Government Support Project (LOGOSP)in Zambia. Lusaka.

MANOR, J., 1995, ”Democratic Decentralization in Africa and Asia”, in IDS Bulletin, no 2.

MAX, J.A., 1991, The development of local government in Tanzania, Dar es Salaam

MOORE, M. et Al., 1994, Institution Building as a Development Assistance Method: A Review of the Literature and Ideas.Report for SIDA. September.

MURPHY, R.L. et Al, 1995, Fiscal Decentralization in Latin America. Johns Hopkins University Press, Washington D.C.

MUTIZWA-MANGIZA, N., 1990, “Decentralization and district development planning in Zimbabwe”, in: PublicAdministration and Development. John Wiley & Sons, Ltd. Sussex.

NACLA, 1995, “"Introduction to Hope" The Left in Local Politics”, in Report on Local Politics, Volume XXIX,Number 1, July/August 1995.

NAUSTDALSLID, J., 1992, “Decentralization Policies in Ghana”, NIBR-Working paper 1992:120. Oslo.

NORAD, 1995, Civil Service Reform in Tanzania: Consultancy to prepare and make recommendations on a Nordic Initiative forassistance to civil service reforms, restructuring and institutional development, Oslo.

OECD, 1993a, DAC Orientations on Participatory Development and Good Governance. OECD/GD (93)191. Paris.

OECD, 1993b, Good Governance and Development Assistance: A Background Paper. OECD. Paris.

RONDINELLI, D.A. et Al., 1983, Decentralization and development: Policy implementation in developing countries. BeverlyHills, Calif.: Sage.

SCHMIDT, G.D., 1988, Effective Donor Support for Decentralization: Some Insights from USAID Experience in Peru, USAID/CDIE Working Paper. No. 93.

SCHMIDT, G.D., 1989, Donors and decentralization in developing countries. Insights from aid experience in Peru. WestviewPress. Boulder, San Francisco & London.

SIDA, 1989, Swedish Public Administration Assistance in Tanzania, By Samhoff and Wuyts, Stockholm.

SIDA, 1990, Central Authority blocks Local Autonomy. An evaluation of SIDA Support to Local Government Development inZimbabwe 1983-1989, SIDA Evaluation Report 1990/2. Stockholm.

SIDA, 1991a, Keeping the Goals in Sight. An Evaluation of Swedish Support to Public Administration in Zimbabwe,.SIDAEvaluation report 1991/4. Stockholm.

SIDA, 1991b, Making Government work. Guidelines and Framework for SIDA Support to the Development of PublicAdministration, Stockholm, May.

SIDA, 1993, Shifting the Balance? Towards Sustainable Local Government, Decentralisation and District Development InBotswana. SIDA Evaluation Report 1993/4. Stockholm.

SIDA, 1995, Evaluation of the SIDA supported Beira-Gothenburg Twinning Programme, Draft Report, prepared by InterconsultSweden.

SILVERMAN, J.M., 1992, “Public Sector Decentralization - Economic Policy and Sector Investment Programs”. World BankTechnical Paper # 188, Africa Technical Department Series, The World Bank, Washington D.C.

SMITH, B.C., 1985, Decentralization. The Territorial Dimension of the State. London. George Allen & Urwin.

SMITH, B.C., 1993, Choices in the Design of Decentralisation. An overview and curriculum for central government officialsresponsible for reorganisation at the local level. Commonwealth Secretariat. London.

THERKILDSEN, O. et.Al., 1992, Assessing Local Insitutions and Local Government, Center for Development Research (CDR).Copenhagen.

TORDOFF and MUKWENA, 1995, Decentralization and local government reform: An assessment, Government of Zambia,Lusaka.

TORDOFF and YOUNG, 1994, “Decentralization and public sector reform in Zambia”, in Journal of Southern African Studies,1994.

UD, 1993, Capacity Building in Development Cooperation: Towards Integration and Recipient Responsibility, by AsplanAnalyse. Royal Norwegian Ministry of Foreign Affairs, Evaluation Report 4/1993. Oslo, 1993.

UD, 1995, Rural Development and Local Government in Tanzania. A Study of Experience with RUDEP and KIDEP, and aDiscussion of Alternative Strategies for District Development in Tanzania. By NIBR. Royal Norwegian Ministry of ForeignAffairs.

UNDP, 1993, Human Development Report 1993. UNDP. Oxford University Press. New York. Oxford.

VAN PELT, S., 1994, Decentralization in Central America: A Policy Overview, a publication of the Regional InformationClearinghouse.

WALLICH, C.I, 1992, Fiscal Decentralization. Intergovernmental Relations in Russia. The World Bank, Studies in Economiesin Transformation, No. 6, Washington D.C.

WIESNER, E.D., 1994, Fiscal Decentralization and Social Spending in Latin America: The Search for Efficiency and Equity.Working Paper Series 1994. Inter-American Development Bank.

WB and IIA , 1990, Strengthening Local Governments in Sub-Saharan Africa. Proceedings of Two Workshops Held in PorrettaTerme, Italy, March 1989, World Bank and Istituto Italo-Africano, An EDI Policy Seminar Report Number 21.

WORLD BANK, 1990a, Institutional Reforms in Sector Adjustment Operations. The World Bank's Experience. By Samuel Paul,World Bank Discussion Papers 92, Washington D.C.

WORLD BANK, 1990b, Public Sector Management Issues in Structural Adjustment Lending. By Barbara Nunberg, WorldBank Discussion Papers 99, Washington D.C.

WORLD BANK, 1992, Public Sector Decentralization. Economic Policy and Sector Investment Programs, by Jerry M.Silverman, World Bank Technical Paper Number 188, Washington D.C.

WORLD BANK, 1993, A Governance Approach to Civil Service Reforms in Sub-Saharan Africa, by Mamadou Dia, World BankTechnical Paper Number 225, Washington D.C.

WORLD BANK, 1993, Ghana Strengthening Local Initiative and Building of Local Capacity. WB. Western Africa Department.

WORLD BANK, 1994, Projectizing the Governance Approach to Civil Service Reform. An Institutional EnvironmentAssessment for Preparing a Sector Adjustment Loan in the Gambia. World Bank Discussion Papers 252. Washington D.C.

WORLD BANK , 1995, Colombia. Local Government Capacity: Beyond Technical Assistance. Country Department III. ReportNo. 14085-CO. Washington D.C., July.

YÁÑEZ H.J. and LETELIER S.L, 1995, "Chile." In: R.L. Murphy (Ed): Fiscal Decentralization in Latin America. JohnsHopkins University Press. Washington D.C.

Documents consulted, but not referred to in the text:

ALLEN, H.J.,1990, Cultivating the Grass Roots. Why Local Government Matters. Bombay: All India Institute of Local Self-Government.

ANGEL, H.S, 1990, Descentralizacion, o la Devolucion de Poderes del Gobierno Central al Gobierno Local. USAID/Honduras.

BENNETT, R., 1989, Territory and Administration in Europe, London, Printer Publishers.

CHAMBERS, R., 1983, Rural Development: Putting the Last First, Longmans, London.

JOHNSON, R.W., 1995, Decentralization Strategy Design: Complementary Perspectives on a Common Theme. ResearchTriangle Institute, paper prepared for the Implementing Policy Change Project, funded by USAID, August.

KOHLI, A., 1990, Democracy and Discontent, Cambridge University Press.

MARGLIN, F. AND MARGLIN, S., 1990, Dominating Knowledge, Development, Culture and Resistance, Clarendon Press,Oxford.

RONDINELLI, D.A., 1981, “Government Decentralization in Comparative Perspective: Theory and Practice in DevelopingCountries”, in: International Review of Administrative Sciences, Vol. XLVII.

RONDINELLI, D.A., 1986, Overview of Decentralization in Developing Countries. Academy for Educ. Dev. Research TriangleInstitute, North Carolina.

RONDINELLI and NELLIS, 1986, “Assessing Decentralization Policies in Developing Countries: The Case for CautiousOptimism”, in: Development Policy Review. Vol. 4, SAGE, London.

RONDINELLI Et al., 1987, Decentralizing Public Services in Developing Countries: A Framework for Policy Analysis andImplementation, Research Triangle Institute, North Carolina.

SHAH A., 1991, The New Fiscal Federalism in Brazil, World Bank Discussion Papers, No. 124, Washington D.C.

SHAH A., 1994, “The Reform of Intergovernmental Fiscal Relations in Developing and Emerging Market Economies”, TheWorld Bank Policy and Research Series, No. 23. Washington D.C.

SHAH A. & QURESHI, 1994, “Intergovernmental Fiscal Relations in Indonesia”, Issues and Reform Options, World BankDiscussion Papers. No. 239. Washington D.C.

SMITH, B.C., 1993, Choices in the Design of Decentralization. An overview and curriculum for central government officialsresponsible for reorganization at the local level, Commonwealth Secretariat. London.

STÖHR, W. and D. TAYLOR (Eds.), 1981, Development from Above or Below, John Wiley & Sons, Ltd., London.

USAID, 1985, Decentralization, local government institutions and resource mobilisation, Bangladesh Academy for RuralDevelopment. Dhaka.

USAID, 1991a, Decentralization: Improving Urban Management in Asia, USAID - Working Paper, Research Triangle Institute,North Carolina.

USAID, 1991b, Decentralization and Democratic Governance. A Review of Latin American Experience and Lessons for Sub-Saharan Africa. USAID -Working Paper, The Urban Institute, Washington D.C.

WORLD BANK, 1994, Improving the Management of Technical Assistance for Institutional Development, Mimeo. September.

ANNEX 6 - HUMAN RIGHTSREFERENCES

1. Evaluation Documents

AID, 1988, An Assessment of the Inter-American Institute of Human Rights.

AID, 1990, Experience with Democratic Initiatives.

AID, 1993, Near East Human Rights Evaluation.

AID/THE ASIA FOUNDATION, 1987, AID Human Rights Programs in Asian and Near East Countries, 1987.

AID/NAPA, 1989, Evaluation of the Civic Education and Human Rights Grant.

BMZ, 1995, Evaluierung der Aktivitäten der politischen Stiftungen in Chile.

BMZ, 1995, Evaluierung der Aktivitäten der politischen Stiftungen in der Republik Südafrika.

BMZ, 1995, Förderung der Menschen-und Bürgerrechte in Kolumbien.

CIDA, 1994, Lessons Learned in Human Rights and Democratic Development.

CIDA, 1994, Accompanying Democracy: Lessons Learned in El Salvador.

CIDA, 1994, Evaluation of CIDA's Human Rights Program Mechanisms in Sri Lanka.

CIDA, 1994, Disabled Peoples' International Institutional Evaluation.

DANIDA/DanChurchAid, 1992, Interim Evaluation of the Human Rights Programme of DanChurchAid.

DANIDA, 1994, Review of DANIDA Support to the Consolidation of Democracy in Nepal.

DGIS/Cebemo/Icco, 1992, The role of NGOs in the building of a just and democractic South Africa.

DGIS/Hivos, 1992, Hivos and Human Rights in Central America.

DGIS/Novib, 1992, International Human Rights Organization.

EUROPEAN COMMISSION, 1995, Evaluation of EC Positive Measures in Favour of Human Rights and Democracy.

2. Annuals

AMNESTY INTERNATIONAL, Annual Report, London: Amnesty International Publications.

BAEHR, Peter; HEY, Hilde; SMITH, Jacqueline; SWINEHART, Theresa; (eds.) Human Rights in Developing Countries.Yearbooks. Oslo: Nordic Human Rights Publications. The Hague, Boston, London: Kluwer Law International 1990-1995.

HUMANA, Charles. World Human Rights Guide, Oxford OUP, 1984-1992.

GASTIL, Raymond D. Freedom in the World. Political Rights and Civil Liberties, Westport, Conn.: Greenwood Press. (Annualsince 1978).

UNDP, Human Development Reports, New York and Oxford: OUP.

UNITED STATES DEPARTMENT OF STATE. Country Reports on Human Rights Practices.

3. Special Studies and Documents

ADVISORY COMMITTE ON HUMAN RIGHTS AND FOREIGN POLICY c/o Ministry of Foreign Affairs, The Hague:Economic, Social and Cultural Human Rights. (Advisory Report No. 18) 1995.

ASIAN DEVELOPMENT BANK, Guidelines for Social Analysis of Development Projects, 1991.

BAEHR, Peter; VAN HOFF, Fried; LIU NALAI; TAO ZHENGHUA (eds.) Human Rights: Chinese and Dutch Perspectives,The Hague, Boston, London: Kluwer Law International, 1995.

BALL, N. Pressing for Peace: Can Aid induce Reform? Washington D.C.: Overseas Development Council, 1992.

BEIGBEDER, Y. International Monitoring of Plebiscites, Referenda and National Elections. Dordrecht, Boston, London:Martinus Nijhoff Publishers, 1994.

BLAIR, H. and HANSEN, G. Weighing in on the Scale of Justice. Strategies Approaches for Donor Supported Rule of LawPrograms, Washington CDIE, 1995.

BREWSTER, G., The International Labor Organization and Human Rights, Indianopolis, U.F.S.I. Reports, 1986/87.

CENTER FOR DEVELOPMENT STUDIES, Swansea, Guidance on Ethnicity, Ethnic Minorities and Indigenous Peoples,London, ODA, 1995.

CODEHUCA, El Habeas Corpus en Centro America, Agosto, 1992.

COOMANS, A.P.M., The international protection of the right to education, Maastricht, 1992.

COOMANNS, F. Economics, Social and Cultural Rights, SIM Special No. 16, 1995.

DEUTSCHES INSTITUT FUR ENTWICKLUNGSPOLITIK, Positive MaBnahmen zur Förderung von Demokratie undMenschenrechten als Aufgabe der Entwicklungszusammenarbeit, Berlin, 1994.

DONNELLY, J., International Human Rights, Boulder, Westview Press, 1993.

EGGER, Ph. Freedom of Association, Rural Workers' Organizations and Participatory Development, OECD, 1994.

EIDE, A., OSHAUG, A. and EIDE, W.B. Report on the right to adequate food as a human right (E/CN.4/Sub, 2/1987/23).

EUROPEAN POLITICAL CO-OPERATION (EPC), Human Rights Reference Handbook. (Main editor: Th. R.G. van Banning),The Hague, 1993.

FORSYTH, D. (ed.), Human Rights and Development, London, 1989.

GALENKAMP, M., Collective Rights, SIM Special No. 16, 1995.

GILLIES, D., “Human Rights, Governance and Democracy: The World Bank's Problem Frontiers”, in: Netherlands Quarterly ofHuman Rights, 1993.

HEINZ, W.F. Ursachen und Folgen von Menschenrechtsverletzungen in der Dritten Welt, Saabrücken; Verlag Breitenbach,1986

HUMAN RIGHTS WATCH, The Human Rights Watch Global Report on Women's Human Rights, New York, London andBrussels, 1995.

ILO, Developing Countries and ILO Standards, Geneva: ILO, 1986.

INTERNATIONAL COMMISSION OF JURISTS, Development, Human Rights and the Rule of Law, Oxford, 1981.

INTERNATIONAL COMMISSION OF JURISTS, States of Emergency: Their Impact on Human Rights, Geneva ; 1983.

INTERNATIONAL COMMISSION OF JURISTS, The dependence of judges and lawyers, a compilation of internationalstandards, Geneva ; 1990.

MCGARY J. and O'LEARY Brendan (eds.), The Politics of Ethnic Conflict Regulation, London, New York: Routledge, 1993.

MADSEN, H.L., Towards Human Rights Assessments of Development Projects, Working Paper for the Chr. Michelsen Institute,Bergen, 1991.

NELSON, J.M. (Ed.), Fragile Coalitions: The Politics of Economic Adjustment, New Brunswick (USA) and Oxford (UK):Transaction Books, 1989.

NELSON, J.M. and EGLINTON, S.J., Encouraging Democracy: What Role for Conditioned Aid? Washington D.C.: OverseasDevelopment Council, 1992.

NELSON, J.M. and EGLINTON, S.J., Global Goals, Contentious Means: Isues of Multiple Aid Conditionality, WashingtonD.C., Overseas Development Council, 1993.

NETHERLANDS INSTITUTE OF HUMAN RIGHTS (SIM), Training Course on International Human Rights Law for Judgesand Lawyers of South America, Utrecht, 1991.

NETHERLANDS MINISTRY OF FOREIGN AFFAIRS, Human Rights Instruments, The Hague, 1993.

OECD, Development Co-operation, 1984 Review. Paris.

OECD, DAC Policy Statement on Development Co-operation in the 1990s, Paris, 1989.

OECD, Principles for evaluation of development assistance, Paris, 1991.

OECD, DAC Orientations on Participatory Development and Good Governance, Paris, 1993.

OECD. Update of the Survey of DAC Members' Policies and Programmes in Participatory Development and Good Governance,Paris, 1993.

OECD, Evaluation des Politiques et Programmes des Membres du CAD concernant la Participation des Femmes auDéveloppement, Paris, 1994.

OECD, Survey Report on DAC Members' Bilateral Development Co-operation Activites for the Promotion of Human Rights,Paris, 1996.

REHOF, L.A. and GULMAN, C. (eds.). Human Rights in Domestic Law and Development Assistance Policies of the NordicCountries, Dordrecht, 1989.

REINERMAN, D., The human rights criterion on German development co-operation, Bonn: BMZ, 1995.

SCHMID, A.P. Research on Gross Human Rights Violations, Leiden: P.I.O.O.M., 1989.

SCHMID, A.P. and JONGMAN, A.J., Monitoring Human Rights Violations, Leiden: P.I.O.O.M., 1992.

SCHMID, A.P. and JONGMAN, A.J., Monitoring Human Rights. Manual for Assessing Country Performance, Utrecht:Netherlands Institute of Human Rights (SIM), 1994.

SCOTT, G. “The Interdependence and Permeability of Human Rights Norms”, in: Osgoode Hall Law Journal, Vol. 27, 1989.

SMITH, A.G., Human Rights and Choice in Poverty, Bergen: CHr. Michelsen Institute, 1989.

STOKKE, O. (ed.), Aid and Political Conditionality, EADI Book Series 16, London: Frank Cass, 1995.

THE HUMAN RIGHTS COUNCIL OF AUSTRIALIA INC, The Right Way to Development: A Human Rights Approach toDevelopment Assistance, Marrickville, 1995.

THE WORLD BANK, Governance. The World Bank's Experiences, Washington D.C., 1994.

THE WORLD BANK, The Social Dimensions of Structural Adjustment in Africa: A Policy Agenda, Washington D.C. 1988.

THE WORLD BANK, Making Adjustments Work for the Poor - A Framework for Policy Reform in Africa, Washington D.C.,1990.

THOOLEN, H. and VERSTAPPEN, B., Human Rights Missions. A study of the fact-finding practices of non-governmentalorganizations, Dordrecht, Boston: Martinus Nijhoff Publishers, 1986.

THORNBERRY, P., Minorities and Human Rights Law, London: Minority Rights Group Ltd., 1987.

TOMASEVSKI, K., Development Aid and Human Rights Revisited, London, 1993.

TÜRK, D., The Realization of Economic, Social and Cultural Rights, Second Progress Report, UN Commission on HumanRights (E/CN.4/Sub.2/1991).

UNDP, Public Sector Management, Governance, and Sustainable Human Development, NewYork. 1994.

USAID, Strengthening Human Rights Monitoring Missions. An Option Paper, Washington, 1995.

VAN DE SAND, K., Menschenrechte in der praktischen Entwicklungszusammenarbeit, NORD-SUD aktuell. Bonn, 1995.

WALLER, P.P., “Aid and Conditionality: The case of Germany” in Stokke, O. (ed.) Aid and Political Conditionality, EADIBook Series 16, London: Frank Cass, 1995.

WHITE, B. Children, Work and Child Labour: Changing Responses to the Employment of Children, The Hague: Institute ofSocial Studies, 1994.

WISSENSCHAFTLICHER BEIRAT BEIM BMZ, Die Bedeutung sozialer Menschenrechte für dieEntwicklungszusammenarbeit, BMZ-aktuell Nr. 55. Bonn, 1995.

4. "Development Co-operation and the Promotion of Human Rights."(Informal Experts Workshop organised by the OECD Development Centre, February 1996.)

Room Documents3:

ACCOLLA, P. (USAID), Child Labour.

HÉLIE-LUCAS, M.A. (Women Living Under Muslim Laws). L’Internationalisme dans le Mouvement des Femmes.

MENDELSON FORMAN, J. (USAID), From Documentation to Deterrence: The Operation of Human Rights MonitoringMissions.

PHILIPS, A. (Minority Rights Group International), Minority Rights and Development: Critical Issues in Human Rights Today.

SGANGA, C. (Annesty International), Human Rights Education.

5. Selected General Bibliography

BATES, R.H., Towards a Political Economy of Development, Berkeley and Los Angeles: University of California Press, 1988.

BRINEKERHOFF, D.W. ‘Institutional Development in World Bank Projects: analytical approaches and intervention designs’,in: Public Administration and Development, Vol. 14, 1994.

BUSHNELL, P., SHLAPENTOKH, V., VANDERPOOL, C. and SUNDRAM J. State Organized Terror. The Case of ViolentInternal Repression, Boulder: Westview Press, 1991.

CALLAGHY, Th. M., The State-Society Struggle, New York: Columbia University Press, 1984.

DIA, M.A., Governance Approach to Civil Service Reform in Sub-Saharan Africa, World Bank Technical Paper No. 225.Washington D.C., 1993.

GOODMAN, L.W., MENDELSON, J.S.R. and RIAL, J., The Military and Democracy, Lexington (USA), Toronto (Canada):Lexington Books, 1990.

HUNTINGTON, S.P., The Third Wave. Democratization in the Late Twentieth Century, Norman: University of OklahomaPress, 1991.

KOHLI, A., MIGDAL J.S. and SHUE, V. State Power and Social Forces. Domination and Transformation in the Third World,Cambridge: Cambridge University Press, 1994.

MCGARRY, J. and O'LEARY, B. (eds.). The politics of ethnic conflict regulation. London, New York: Routledge Press, 1993.

MIGDAL, J.S., Strong Societies and Weak States. State-Society Relations and State Capabilities in the Third World, Princeton:Princeton University Press, 1988.

NELSON, J.M. (ed.), Fragile Coalitions: The Politics of Economic Adjustment, Washington: ODC, 1989.

NORTH, D.C., Institutions, Institutional Change and Economic Performance, Cambridge: Cambridge University Press, 1992.

STOHL, M. and LOPEZ, G.A. (eds), Government Violence and Repression, An Agenda for Research.

3'Room documents' is a term peculiar to the OECD and refers to unpublished documents from seminarsand workshops.

VANBEL, R. and WILLETT, Th. D., The Political Economy of International Organizations. A Public Choice Approach,Boulder. Westview Press, 1991.

ANNEX 7 - PARTICIPATIONDOCUMENTS REVIEWED

DANIDA, Forestation in Sudan, 1993.

DANIDA, Training/Community Planning, Bolivia, Sri Lanka, Zambia, 1993.

DANIDA, Institutional Development in Africa, Asia and Latin America, 1993.

DANIDA, Support to Agriculture in Asia and Africa, 1994.

DGIS, Dutch assistance, 1991.

ISS/CEBEMO(DGIS), NGO support CPT Norte II, Brazil, 1991.

DGIS/CEBEMO, Non-formal education in India, 1993.

DGIS, Dutch assistance, 1993.

DGIS/NOVIB, Big NGOs in Latin America, 1993.

FINNIDA, Integrated Community Development, Nicaragua, 1993.

FINNIDA, Forestry, Zanzibar, 1993.

FINNIDA, Forestry and Participation, Zanzibar, 1993.

FINNIDA, PRA in Rural Integrated Project Support, Tanzania, 1993.

FINNIDA, NGO Support Program, Cases Ethiopia, Nepal, Nicaragua, Uganda, 1994.

FINNIDA, Rural Integrated Project Suport, Tanzania, 1994.

FINNIDA , Microenterprises, 1994 Tanzania, 1994.

GTZ, Project planning manual, 1991.

NORAD, Socio-cultural issues in Aid, 1987.

NORAD, IRDP, Sri Lanka, 1992.

NORAD, IRDPs and District development, Tanzania, 1995.

ODA, Stakeholder participation, 1995.

ODA, Stakeholder participation, 1995.

ODA, Assessment of stakeholder, 1995.

Sida, Integrated rural development, Guinea-Bissau, 1993.

Sida, NGO sector support, case studies Bolivia, India, Kenya, Zimbabwe, 1994.

Sida, Evaluation Manual, 1994.

Sida, NGO support, Bangladesh, 1994.

Sida, LFA Guidelines, 1995.

USAID, Education for participation in Latin America, 1989.

USAID, Community participation in Water Supply and Sanitation, 1991.

USAID, Participation in Development Fund Programs, Africa, 1992.

USAID, Participatory Development, 1993.

USAID, “Best practices” for participation, Africa, 1994.

USAID, Participation in Economic Policy Reform, Africa, 1994.