waiting for democracy— the politics of choice in

W R I R E P O R T

Worl d

Resources

I nsti tute

WAITING FOR DEMOCRACYThe Politics of Choice in Natural Resource Decentralization

JESSE C. RIBOT

World

Resources

Institute

10 G Street, NESuite 800Washington, DC 20002www.wri.org

WA

ITIN

G F

OR

DE

MO

CR

AC

Y

Th

e P

olit

ics

of

Ch

oic

e in

Na

tu

ra

l Re

so

ur

ce

De

ce

nt

ra

liza

tio

nW

RI R

EP

OR

T

WAITING FOR DEMOCRACYTHE POLITICS OF CHOICE IN

NATURAL RESOURCE DECENTRALIZATION

JESSE C. RIBOT

World

Resources

Institute

W A S H I N G T O N , D C

ii

WRI: WAITING FOR DEMOCRACY

MARTHA SCHULTZ AND EMILY YAGHMOUREDITORS

HYACINTH BILLINGSPUBLICATIONS DIRECTOR

MAGGIE POWELLLAYOUT

MOR GUEYECOVER ART

Each World Resources Institute report represents a timely, scholarly treatmentof a subject of public concern. WRI takes responsibility for choosing the studytopics and guaranteeing its authors and researchers freedom of inquiry. It also

solicits and responds to the guidance of advisory panels and expert reviewers.Unless otherwise stated, however, all the interpretation and findings set forthin WRI publications are those of the authors.

Copyright © 2004 World Resources Institute. All rights reserved.

ISBN 1-56973-564-6

Printed in the United States of America. Text stock has a recycled content of30%; cover stock has a recycled content of 10%.

The cover image was designed by the author and painted by the Senegalesereverse glass-painting artist Mor Gueye. As a government officer organizes a“participatory” meeting, a child is conjugating on the wall “I participate, youparticipate, he/she participates, we participate, you (plural) participate, theyprofit.” This graffiti was first brought to my attention by Lisa Peattie (personalcommunication April 2004), who saw this writing on a wall in Paris in 1968.The graffiti was written following President Charles De Gaulle’s May 1968speech in which he spoke of the need for more “participation” and proposed areferendum on the topic (Rohan 1988; Atelier Populaire 1969; Gunn and Iain1998 at www.marxist.com/1968/may68.html).

Figure 3 © The New Yorker Collection 2001 Charles Barsotti fromcartoonbank.com. All Rights Reserved.

iii


CONTENTS

Section 2: Local Accountability—Institutional Choices in Practice ................. 25

Elected Local Government ................................... 26Local Administrative Bodies: Deconcentration .... 30Customary Authority ............................................. 31NGOs and Community Groups ........................... 34Crosscutting Institutional Issues .......................... 36

Single-purpose vs. multi-purpose bodies ................. 36Privileging marginal and poor groups .................... 38Elite capture and patronage ................................... 41Role of a strong central state ................................... 43Accountability of the central state and

donors to local government .................................. 43Institutional Choice—Asking the

Right Questions ................................................. 45

Section 3. Responsiveness of LocalInstitutions—Power and Capacity ............. 47

Power Matters ....................................................... 47Power transfers in practice ..................................... 47Privatization is not decentralization ...................... 51Geographic scale ..................................................... 52Interjurisdictional equity ........................................ 52Security: The “means of transfer” problem .............. 53Setting the boundaries on local power:

Planning versus standards .................................. 54Local control over outside industries ....................... 59

The Capacity Conundrum .................................... 59Power Choices—Asking the Right Questions ...... 65

Acronyms ...................................................... vii

Acknowledgments ....................................... viii

Executive Summary ........................................ 1Main Findings ........................................................ 2Recommendations .................................................. 4Implementation ..................................................... 6

Introduction—Democracy andNatural Resources ......................................... 7

What Is Decentralization and Why Decentralize? .... 8Democracy and Natural Resources: Mutual

Reinforcement, Sustainability, and Scaling Up . 12Fettered Reforms .................................................. 14Measuring Decentralization: Cases,

Methods, Limitations .......................................... 15Organization of the Report ................................... 16

Section 1: Theoretical Foundationof Decentralization’s PotentialEnvironmental Benefits ............................... 17

Democratic Decentralization andRepresentation: A Policy Process ...................... 17

Accountability and Principles of Institutional Choice .. 18Powers: Discretion and Subsidiarity ..................... 21Mediating Factors: Decentralization

is No Panacea ..................................................... 23

iv


Section 4: Lessons Learned ........................ 67Accountability and Institutional Choice ...............67

Interinstitutional accountability, plurality,and coexistence—Choose democracy .................. 68

Institutional choice for scaling upand sustaining representation ............................ 70

Institutional choice and citizenship ...................... 70Subsidiarity: Choosing Powers,

Sequencing Transfers ........................................ 70Subsidiarity principles ........................................... 70Sequencing points ................................................... 71

Beyond Actors, Powers, and Accountability:Other Factors Affecting Outcomes .................... 72

(En)countering Resistance to DemocraticDecentralization: Seizing Opportunities ........... 74Resistance to the transfer of powers ........................ 74Institutional choices as a form of resistance ............ 75Resistance by local actors ....................................... 76Countering resistance ............................................ 76

Section 5: Recommendations ..................... 79Establish the Institutional Infrastructure

of Democratic Decentralization ......................... 79Decentralize the Management of Natural

Resources ........................................................... 82Counter Resistance and Seize Opportunity .......... 82Further Research .................................................. 83

Annex A: WRI Research onDecentralization and the Environment ...... 85

Annex B: Case Study Descriptions ............. 87Country Case Studies ............................................ 87Crosscutting Cases .............................................. 107

Annex C: Accountability Mechanisms ......109

Works Cited ................................................. 115

Endnotes ...................................................... 133

Boxes1. Defining Decentralization and Related Terms ............ 92. The Logic behind Decentralization’s Promises ............ 123. Measuring Decentralization’s Outcomes ............ 164. Sustained Forest Management in India through

the Transfer of “Means of Regulation” ....................... 225. Trees Saved by Local Elections in Uganda .................. 266. Elected Councillors in Senegal Give Away

Forests against the Will of Local People ...................... 287. Elected Authorities Lose Legitimacy

When They Can’t Deliver .......................................... 308. King vs. Council in Uganda—

Who Better Manages the Forests? ............................. 329. Manufacturing Success ............................................. 37

10. Elite Participation with Committee Proliferation ......... 3811. Timber Concessions vs. Local Management

and Matters of Scale .................................................. 4012. The Loss of Local Jurisdiction

in Uganda’s Masindi District ..................................... 5013. Overly Complex Management Planning ..................... 5514. Greater Local Equity and Revenues,

Fewer Outside Loggers in Bolivia .............................. 5815. Capacity as an Excuse for Retaining

Central Power in Nicaragua ...................................... 6016. Double Standards or Minimum

Standards in Senegal ................................................ 6317. Perverse Incentives in Uganda .................................. 6418. Decentralizing Over-exploitation

in Indonesia—Get It While It Lasts ............................ 7319. Proposed Environmental Subsidiarity Principles ........ 81

Figures1. Formal Definitions: Decentralization,

Not Decentralization ................................................. 102. Accountability Defines Effective Decentralization ....... 193. The Silent Majority ................................................... 394. Accountability Relations Mediated

by Local Government ............................................... 69

v


This report is dedicated to the memory of

Dr. Nyangabyaki Bazaara, who directed the

Centre for Basic Research in Kampala, Uganda.

Dr. Bazaara was an intellectual pillar of WRI’s

decentralization research program. Reading this

report you will see reflections of his insights and spirit

throughout. He died in Kampala in August 2003

at the age of 42.

vi


vii


ACRONYMS

CBNRM Community-based natural resource

management

CBO Community-based organization

CBT Community-based targeting

CIFOR Center for International Forestry

Research

GGAOF Gouvernement Général de l’Afrique

Occidentale Française

GOC Government of Cameroon

NGO Nongovernmental organization

NRM Natural resource management

NWP Nature, Wealth, and Power

PVO Private voluntary organization

RDM République du Mali

RDS République du Sénégal

UNCDF United Nations Capital Development

Fund

UNCED United Nations 1992 Conference on

Environment and Development

WRI World Resources Institute

viii


ACKNOWLEDGMENTS

Naughton, Linda Norgrove, Doug Porter, OumarSissokho, Marja Spierenburg, Boubicar Thiam, andMatt Turner.

I also am grateful to the participants in the BellagioConference on Decentralization and the Environ-ment, all of whose research and observations contrib-uted to this document. In addition to Bazaara,Kassibo, Ntsebeza, and Oyono, mentioned above, theBellagio participants are Arun Agrawal, AmitaBaviskar, Christian Brannstrom, Ben Cousins, MairiDupar, Fernando Melo Farrera, David Kaimowitz,Anne Larson, James Manor, Robin Mearns, JamesMurombedzi, Pablo Pacheco, Nancy Peluso, Ida AjuPradnja Resosudarmo, Roberto Sanchez, UraivanTan-Kim-Yong, and Xu Jianchu. Special thanks toMehr Latif for being an excellent rapporteur at theBellagio conference; many of the insights shecaptured are used in this report.

I deeply appreciate Jon Anderson, Diana Conyers,David Jhirad, Anne Larson, Crescencia Maurer,Carmen Revegna, Frances Seymour, and Peter Veitfor taking the time to provide careful and invaluablecomments on this document. Many thanks toCatherine Benson for coordination and assistanceduring writing, editing, rewriting, reediting, andproduction. Thanks to Emily Yaghmour and MarthaSchultz for their careful editing and to Maggie Powellfor the cover design and layout.

I want to thank CIFOR in Bogor, Indonesia, forcontributing its experience and in-depth research tothe Bellagio meeting by generously financing the

The research and reflection of colleagues fromaround the world provided case material and insightsfor this document. First, I owe great thanks to themembers of WRI’s Accountability, Decentralization,and the Environment research initiative. They areNyangabyaki Bazaara, Diana Conyers, Thierno Diallo,Patrice Etoungou, Juliet Kanyesigye, BréhimaKassibo, Naffet Keita, Patrice Bigombe Logo, AloisMandondo, Everisto Mapedza, Frank EmmanuelMuhereza, Eugene Muramira, Agrippinah Namara,Xavier Nsabagasani, Lungisile Ntsebeza, Phil RenéOyono, and Alda Salomao. I would also like to thankthe Center for International Forestry Research(CIFOR) and the Centre for Basic Research—thehomes of our Cameroon and Uganda teams—fortheir collaboration in this effort.

Many other people who also contributed to WRI’sdecentralization research project deserve sincerethanks. In particular, I want to acknowledge themembers of our international advisory group,Cheibane Coulibaly, Olivier Dubois, Cyprian Fisiy,Gerti Hesseling, Mahmood Mamdani, James Manor,and Pauline Peters, for their constructive feedback onthe research design and their comments on theresearchers’ working paper drafts. I want also tothank the following peer reviewers for the detailedcomments they gave to each researcher: TorBenjaminsen, Peter Bolch, David Brown, Jean-PierreChauveau, Diana Conyers, Chimère Diaw, VupenyuDzingirai, Susanne Freidberg, Mara Goldman, JaneGuyer, Rebecca Hardin, David Hughes, VeroniqueJoiris, Randi Kaarhus, Alain Karsenty, Jeremy Lind,Christian Lund, Paul Mathieu, Rebecca McLain, Lisa

ix


participation of Larson, Pacheco, and Resosudarmo. Ialso want to thank CIFOR-Cameroon in Yaoundé, andthe Center for Basic Research in Kampala, Uganda,for contributing their findings from highly productivelong-term institutional research collaborations withWRI. I am grateful to the Economic Growth, Agricul-ture, and Trade Bureau of the U.S. Agency forInternational Development (USAID), to USAID’sAfrica Bureau, and to USAID’s Mali and Ugandamissions, and to the Dutch government for gener-ously funding WRI and our African colleagues toconduct the research that informs this document. Iam also indebted to the Rockefeller Foundation forproviding the funds and an inspiring venue for the

conference at Bellagio, which generated many of theideas in this publication. Thanks also to JamesMurombedzi and the South Africa office of the FordFoundation for generously funding the writing andpublication of this document.

I want to express special thanks to my colleaguesPeter Veit and Jon Anderson for their steadfastencouragement and support throughout this project.

My deepest thanks are to Allyson Purpura for beinga supportive source of critical and constructivefeedback, and for putting up with my many absenceswhile doing research abroad.

JESSE C. RIBOT

x


1


EXECUTIVE SUMMARY

The majority of developing nations claim to beundertaking some form of decentralizationinvolving natural resource management. This

report examines cases from a subset of the countrieswhose decentralizations are considered the mostadvanced: Bolivia, Brazil, Cameroon, China, India,Indonesia, Mali, Mexico, Mongolia, Nicaragua,Senegal, South Africa, Thailand, Uganda, andZimbabwe. This report

● assesses the degree to which proclaimed decen-tralizations involving natural resources are beingestablished in law and in practice,

● evaluates how these decentralization reforms affectsocial and environmental outcomes, and

● recommends ways that these reforms could bedesigned and implemented to improve their socialand environmental outcomes.

Theorists promote decentralization reforms basedon the following proposition: IF institutional arrange-ments include local authorities who represent and areaccountable to the local population and who holddiscretionary powers over public resources, THENthe decisions they make will lead to more efficientand equitable outcomes than if central authoritiesmade those decisions.

Theory indicates that downwardly accountable orrepresentative local actors with significant discretion-ary power constitute the necessary infrastructure foreffective decentralization. Decentralization advocates,such as governments, donors, and NGOs, aim to “getthe institutions right” so as to improve development

and environmental outcomes. By analyzing actors,powers, and their accountability, and by identifyingother factors that shape outcomes, this reportevaluates the above proposition.

The research in this report addresses publicmanagement of local public resources. It does notevaluate private-property regimes or private-sectorreforms—except where privatization is taking place inthe name of decentralization. The report focusesmainly on decentralizations related to forests locatedoutside of parks or conservation areas. Such reservesusually are centrally managed, and the vast majorityof forests and forest-based communities are locatedoutside of these areas. Nevertheless, the findings arerelevant to other environmental arenas and to broaderdecentralization processes.

The audience for the report includes

● international and national policy makers andinternational donors and development agencies;

● activists, including local-government associations,federations, and other groups representing ruralinterests, and environment- and justice-oriented NGOs;

● practitioners involved in civic education of localpeople and public authorities working with naturalresources, as well as practitioners involved in theimplementation of decentralized natural resourcemanagement policies and programs; and

● researchers wishing to deepen our knowledge ofthe effects of decentralizations and the best waysto leverage, structure, and sequence decentraliza-tion reforms.

2


Based on worldwide research and experience, andon a sample of the vast and growing literature, thisreport summarizes findings and recommendationson decentralized management of natural resources.A policy brief from WRI’s decentralization researchprogram was published under the title DemocraticDecentralization of Natural Resources: InstitutionalizingPopular Participation (see Ribot 2002a). This reportdetails, nuances, and extends the findings of thatbrief.

MAIN FINDINGS

Case studies from around the world indicate that theinstitutional arrangements necessary to bring aboutdecentralization are rarely established in so-calleddecentralization reforms. Instead, many reforms resultin privatization and/or deconcentration—the transferof powers to central government agents in the localarena. In other reforms, governments simply proclaimthat they are decentralizing and enact a theatricalimage of reform for their donor audiences.

Most decentralizations are not being established inlaw or they are not being implemented in practice.The failure to create decentralization can be ex-plained, at least in part, by the entrenched resistancethat decentralization reforms encounter at every step.While “getting the institutions right” would probablylead to better outcomes, groups that fear losing powerin decentralization reforms pose staunch resistance toadequate policy making and implementation. Decen-tralization is not even getting to IF in the IF-THENproposition.

Despite the incompleteness of most decentraliza-tion reforms, the many case studies of reformsunderway provide evidence of their potential out-comes and provide insights into how they might bebetter structured and sequenced. Positive outcomesattributed to decentralizations involving naturalresources include the following:

● local governments have been able to demonstratecapacity and initiative in natural resource manage-ment;

● local people have been empowered to protect theirforests from outside commercial interests;

● local councils and local people have increased theirrevenues from resource use;

● marginal and disadvantaged groups have played agreater role in natural resource management andhave benefited more from local resources; and

● some cases of sustained forest management havebeen observed.

Negative outcomes associated with decentralizationinclude elite capture and conflict. Many more pooroutcomes are associated with incomplete decentraliza-tions. They include

● increased vulnerability of local people whenmanagement burdens are transferred withoutresources;

● the tendency of “decentralized” projects to mobi-lize local people as mere labor rather than empow-ering them to make decisions for themselves;

● the creation of new forms of exclusion throughdouble standards that require complex manage-ment plans from local communities while allowinglarge-scale commercial interests to enter and usethe resource with little planning and even lessmonitoring;

● increased public exclusion through privatization ofpublic resources such as forests to individuals,corporations, NGOs, and customary authorities;and

● the delegitimizing of fledgling local democraciesby failing to give them discretionary powers and bycreating competition when public resources areprivatized or managed through donor-led partici-patory processes.

Case studies and literature on decentralization reformsprovide valuable lessons on how to improve institutionalarrangements and outcomes. Effective decentralizationrequires creating a realm of local discretionary powersunder representative authorities. To establish, sustain,and scale up these arrangements requires keen attentionto the institutions and powers being chosen for localcommunities by governments, donors, and develop-ment practitioners in decentralization reforms. Mostcurrent choices are not enfranchising communities to

3


manage their own resources and development. They arenot providing the basic institutional infrastructure—the politicial-administrative environment—that wouldenable demand-driven democratic natural resourcemanagement.

Most institutions currently being chosen to receivenatural resource management powers are notaccountable to the local population as a whole. Fordecentralization to be equitable and efficient, theremust be a clear line of accountability from decisionmakers to the local population. When powers overpublic natural resources or any other public goods areexercised by private, nongovernmental, or traditionalorganizations, there must be a line of accountabilityfrom these organizations to local representativeauthorities. Many reforms are now encouraging aplurality of local institutions—such as committees,associations, NGOs, and customary authorities. Butsuch proliferation of institutions that are unaccount-able to representative authorities may be a formulafor elite capture.1 For more integrated and representa-tive public decision making, governments, donors,and NGOs should insist that all organizationsexercising public powers be accountable to represen-tative authorities. Presently, publicly accountableintegrated decision making is rarely being established.

Residency-based forms of belonging and citizenshipprovide a strong basis for democratic participation innatural resource management. Customary authoritiesand NGOs are often chosen by decentralizationdesigners and practitioners as if they represent or canspeak for local people. Customary authorities,however, often base inclusion on identity and interest.They represent only people of certain origins,ethnicities, lineages, and religious identities, whileNGOs represent interested parties. Identity- andinterest-based inclusion in public decision makingmay reinforce differences, fragment community, andproduce conflict. When governments, donors, andNGOs arrange public inclusion through customaryauthorities, they subjugate local people to theseauthorities and their forms of belonging—which areoften highly unequal and usually severely genderbiased. In the process they legitimize these authori-ties, delegitimizing democratic authorities that mightotherwise have exercised these public powers.

Presently, inclusive and integrative forms of citizenshipare often being undermined.

Donors, NGOs, and governments also often chooselocal institutions that are not easily scaled up orsustained. They base inclusion and representation onparticipatory processes, or membership in NGOs orlocal committees. These interventions are usuallyshort lived. They are also expensive and labor-intensive. When projects end, efforts at inclusioncollapse. They are not sustainable and cannot bereplicated or institutionalized across large territories.Basing inclusion on democratic local governmentwould institutionalize participation in empoweredrepresentative authorities. These authorities operateacross national territories and are sustainable overtime and through changes in leadership and regimes.Presently, scaling up and sustaining broad-basedparticipation remain significant limiting factors forcommunity-oriented natural resource management.

Local representative decision makers rarely receivemeaningful discretionary powers. Instead of transfer-ring lucrative opportunities or powers that wouldenable local authorities to make decisions over thedisposition and use of forests, central authoritiesoften transfer burdens of management with little orno funding. When decisions and responsibilities aretransferred, they are usually accompanied by exces-sive oversight. Most powers are also transferredthrough insecure mechanisms and can thereforeeasily be taken back by central agents. Such insecuretransfers are a formula for manipulation from above.In addition, powers transferred to local authorities areoften limited to small cordoned-off areas while themajority of forests remain under the control of thecentral government or are privatized. Presently,insufficient and inappropriate powers turn mostdecentralization reforms into charades.

Providing representative public decision making isan essential role of good government. In the past twodecades, development agents have favoredprivatization and civil-society-based interventions. Inthe process, they have avoided, weakened, anddelegitimized representative local-governmentinstitutions. Empowering local governments withnatural resource management authority could reverse

4


this trend. To test the proposition that decentraliza-tion leads to positive outcomes, governments, donors,and NGOs should choose to institutionalize naturalresource management in elected local government orin institutions that account to local democraticauthority.

Democratic decentralization is progressing slowly.Not only is it challenged by privatization and civil-society development approaches, it also faces resis-tance because it threatens entrenched patronagerelations enjoyed by the highest-level central officials,as well as by local merchants, local bureaucrats, andfront-line forestry agents. It threatens powerful, oftenurban-based commercial interests who fear losingaccess to productive resources. It threatens the rolesof local elite, traditional authorities, and governmentadministrators already present in the local arena.

Decentralization also evokes fears among environ-mentalists and environmental agencies that localpeople will not adequately manage important naturalresources. The most commonly expressed fears arethat decentralization will lead to destruction of theresource, or that it will lead to conflict among partiesinterested in and using the resources. Environmental-ists also argue that local officials are corrupt and willuse their control over natural resources for personalprofit. They often feel that local authorities lack thecapacity to adequately manage resources and con-flicts. At times these fears are founded, but moreoften they are not.

To mitigate the feared effects of decentralizationreform, these actors choose inappropriate institutionsand powers, often justifying their choices withspecious capacity and environmental-menace argu-ments. Choosing inappropriate institutions andpowers is in itself an act of resistance. Their choicescompromise legislation and block implementation.These tactics lead to the partially implementedreforms that rarely merit the label of decentralization.Under these circumstances, it is not surprising thatthe decentralization experiment is in fetters.

Partial implementation has led some policy makersto believe that decentralization of natural resourcemanagement is not working. A backlash is forming.

Those who are threatened by decentralization arequick to declare its failure—even where it has neverbeen implemented. It would be a shame if slow anddisingenuous implementation were to consolidate abacklash before these reforms have had a chance tobe tried.

RECOMMENDATIONS

This report provides two main recommendations: (1)work with democratic local government to institution-alize downwardly accountable representation and (2)transfer environmental powers to create local discre-tion. In short, downwardly accountable local authori-ties with discretionary powers must be established sothat the decentralization experiment can proceed.Democratic local government appears to be the most-promising local institution for establishing durableand widespread decentralization. During the reformprocess, resistance must be identified and countered.Measures to help achieve these recommendations aredescribed below and are presented in more detail atthe end of this report.

(1) Work with democratic local government tostrengthen democracy and citizenship, and toestablish the basic institutional infrastructure ofinclusion that can be sustained and scaled up. Centralgovernment agents, donors, and NGOs engaged innatural resource management and use can do thefollowing:

● Develop guiding principles for choosing andstrengthening representative local institutions.These principles might include– Work with and strengthen local democratic

government wherever it exists.– Insist that governments establish democratic

local government where it does not exist.– Insist that governments admit independent

candidates in local elections.– Work to establish multiple mechanisms—in

addition to or in the absence of elections—toachieve downward accountability for localgovernment and for other institutions.

5


– Ensure public participation by making localinstitutions that hold public powers—includingNGOs, traditional authorities, committees, andprivate bodies—accountable to democraticauthorities.

– Avoid creating a proliferation of new single-purpose local institutions—NGOs, committees,etc.

– Help develop a universal form of inclusion bypromoting residency-based citizenship ratherthan identity-based forms of belonging.

(2) Develop environmental power transfer guidelines toencourage the transfer of sufficient, equitable, andsecure discretionary powers to democratic localauthorities. To accomplish this, governments, donors,and NGOs should take the following steps.

● Establish environmental subsidiarity principles toguide the transfer of sufficient discretionarypowers.– Subsidiarity principles are needed to help policy

makers and practitioners identify the powersthat are most appropriate for transfer to specificlevels of political-administrative authorities.There are many kinds of powers that can betransferred. Draft subsidiarity principles areproposed in the conclusion of this report.Governments, donors, and NGOs should usesuch principles to leverage decentralizationefforts. Advocacy groups and popular move-ments could use such principles to demand thattheir representatives be appropriately empow-ered.

● Establish uniform minimum environmentalstandards to help protect the environment, delimitthe domain of local freedom, and increase equity.– Minimum standards can be a transparent way to

codify the domain of local discretionary actionwithin a set of rules designed to protect andsustain resources. Minimum environmentalstandards specify the minimum requirementsfor the use and management of the resourcebase. These standards also provide a domain offreedom or discretion in which people canmanage and use surrounding resources. Under

current regulatory systems in the developingworld, almost all resource use requires approvalof the central authorities, giving ultimatedecision-making power to these central actors.The shift to minimum standards complementsdecentralization by creating space for localdiscretion, which is essential to decentralizedmanagement systems. Governments seriousabout decentralization should consider aminimum-standards approach.

● Remove double standards in forest management toincrease equity and justice.– Double standards that favor commercial indus-

try are common. In many countries, forestrylaws require more arduous management bylocal communities than by large-scale commer-cial industries, creating unjust barriers forcommunity entry into forestry. These barriersprevent transfer of environmental managementrights and powers to local authorities. Makingminimum standards uniform—treatingcommunities and industries equally—willreduce unjust double standards now applied byforest services around the world.

In addition to these measures, sequence decentraliza-tion of natural resources strategically to minimizeconflict and resistance. Policy makers and practitio-ners can use sequencing guidelines to push centralenvironmental agency authorities to devolve powersprogressively rather than holding everything backbased on specious excuses about local capacity orrisks to the environment. Some sequencing prin-ciples might include the following:

● Establish representative and/or downwardlyaccountable authorities first.

● Transfer political decisions related to allocation ofaccess and benefit distribution before transferringhighly technical resource management decisions.

● Transfer decisions and powers that are not likely toincite violence or lead to environmental damage.

● Transfer finance and revenue-raising powersbefore transferring management obligations.

● Transfer additional funding with all centrallymandated obligations.

6


● Transfer powers before requiring demonstration ofcapacity.

● Establish minimum environmental standardsbefore requiring management planning.

IMPLEMENTATION

Decentralization reforms involve the relocation ofpower, producing winners, losers, and resistance. Toimplement any of these recommendations, promot-ers of decentralized natural resource management,including many environmentalists, local and interna-tional NGOs, and both bilateral and multilateraldonor agencies, will have to confront and counter thisresistance within their own ranks and throughout thelarger set of institutions affected by decentralizationreforms. Decentralization is about asking the rich andpowerful to give up some of their wealth and powerto poor and marginalized people. Even if there arelong-term benefits for the wealthy, resistance, out offear and self-interest, is inevitable. Few powerfulbodies are willing to give up their power for the sakeof equity and efficiency, which is probably why mostdecentralization takes place during times of economicand political crisis, not through well-planned reforms.Nevertheless, two strategies may move the decentrali-zation experiment forward: countering resistance andopportunism.

Resistance to decentralization in the naturalresources sectors needs to be challenged by everyoneinterested in decentralization. Supporters of decen-tralization must be informed by grounded researchabout the advantages and shortcomings of decentrali-zation reforms and the requirements and opportuni-ties for change. Local populations and federations ofinterested parties, such as peasant organizations and

local government associations, will need to knowwhat their governments are saying and what they aredoing. Advocacy NGOs and other activists, as well asdonors, must point out the contradictions in currentreforms and counter the interests that are sabotagingchange.

Opportunities for change arise continually. Some-times opportunities arise when politicians declaretheir support for local democracy. Whether theirsupport is genuine or a mere ploy to please constitu-ents and donors, it can become a catalyst for positivechange. People, hearing the rhetoric, come to expectlocal democracy and begin to demand it. Opportuni-ties often arise during economic and political crisesthat force governments to make reforms. Sometimesprogressive politicians and bureaucrats become activewithin government, creating channels of influence onthe inside. These are opportunities that can be takenadvantage of if the constituencies interested indemocratic decentralization know what they want andunderstand how to create it. Advocates need to beprepared for opportunity.

“Getting the institutions right” or “getting to IF” isan ongoing challenge. It is a process where theobjective is never attained or is attained and lost.Institutional change cannot be removed from itspolitical-economic context. The institutional arrange-ments of decentralization are embedded in a largerpolitical-economic set of struggles and relations thatmust be understood and engaged. The promoters ofdecentralization can only move toward getting thelocal institutions right by confronting resistance andby identifying and seizing opportunities—again andagain.

7


INTRODUCTION

DEMOCRACY AND NATURAL RESOURCES

determination (Mair 1936, 13). The language of earlierdecentralizations included laudable goals that wewould be happy to see in today’s policy objectives. Forexample, in 1936, Mair (p. 12) wrote, “The basic aimof Indirect Rule is the development of an Africansociety able to participate in the life of the modernworld as a community in its own right.”4 Thesepolicies were lauded as liberal and liberating. History,however, has judged them as cruel and coercive(Mamdani 1996a).

The most recent wave of decentralizations, which isthe focus of this report, is introducing a newemancipatory language of democracy, pluralism, andrights—implying that local people will be enfran-chised and therefore transformed from subjects intocitizens. Indeed, most decentralizations are beinglabeled democratic. Natural resource managementalso is moving toward premises that are moredemocratic and rights-based.5 The new and hopefullanguage is based on ideals of efficiency, equity, andjustice that are believed to stem from local democraticdecision making.6 These enfranchising words seemeven more progressive than their colonial precursors:similarly idealistic, but ultimately cruel and oppres-sive. If put into action, this discourse could create theconditions that decentralization proponents believewill lead to more effective and just outcomes inservice delivery, development, and natural resourcemanagement.7

The colonial experience should be taken as acautionary tale, however; emancipatory language isnot enough. Is the new language of decentralization

Most developing and transitional countrieshave launched decentralization reformsover the past decade (United Nations Capital

Development Fund 2000, 5–11).2 Many claim toinclude decentralization of natural resource manage-ment in these reforms (Agrawal 2001). Whether or notdecentralization reforms directly involve naturalresource transfers, these reforms are important toenvironmental management because—complete orincomplete, well or poorly designed and executed—they are transforming the local institutional infrastruc-ture on which local natural resource managementdepends. Decentralization is changing the kinds ofauthorities that make decisions over natural resources;the kinds of decisions that these authorities areempowered to make; and the relations of accountabil-ity between the central state, local government, otherlocal institutions, and the local population. Thesereforms have implications for the way in which localpeople can derive benefits from natural resources, andhow they value, manage, and use them. In short,decentralizations are changing the local institutionallandscape for better or for worse.

The current decentralization trend follows severalwaves of decentralization reforms during the pastcentury.3 The earlier reforms emphasized nationalcohesion, effective rule, and the efficient manage-ment of rural subjects—particularly under Britishpolicies of Indirect Rule and the French policy calledAssociation (Buell 1928; Mair 1936; Mamdani 1996a).These earlier reforms, despite being instruments of“European penetration,” were described by theirarchitects in terms of local development and self-

8


being codified in law and translated into practice? Arecurrent reforms integrating the basic elements thatwould justify calling them democratic decentraliza-tion? What are the effects of these new reforms onsociety and on the environment? Plausible theoriestell us that decentralization of authority over naturalresources should produce more efficient and equi-table outcomes. While theorists and practitionershave high expectations for the current global wave ofdecentralization reforms, the theory must be ap-proached with caution since surprisingly littleresearch has been done to assess whether the neces-sary conditions exist or whether they lead to thedesired outcomes.8 This report (1) assesses whetherproclaimed decentralizations involving naturalresources are being established in law and in practice;(2) evaluates how the institutional arrangements ofthese decentralization reforms affect social andenvironmental outcomes; and (3) recommends waysthat these reforms and their social and environmentaloutcomes might be improved.

The research conducted for this report set out to testthe hypothesis that decentralization (defined in Box 1)leads to improved efficiency, equity, service delivery,development, and natural resource management. Theresearch found that few cases exist where the institu-tional arrangements necessary for real decentraliza-tion reform have been established. The centralhypothesis, therefore, can hardly be tested. For thisreason, the report is organized around the analysis ofinstitutional arrangements being established inpractice; that is, the accountability of the local actorsto whom power is being transferred and the kinds ofpower they are receiving. The outcomes that havebeen associated with various elements of decentrali-zation are also presented and discussed. But due tothe partial implementation of most reforms, thisreport focuses more on why decentralization reformsare being blocked and how best to overcome theobstacles than on the outcomes of implementedefforts.

This introduction briefly defines the basic termsand elements of decentralization (addressed in depthin Section I); discusses some mutually reinforcingrelations between local democracy and naturalresource management; sketches the state of current

fettered decentralization reforms; describes the datasources, research methods and their limitations; andprovides an outline for the remainder of the report.

WHAT IS DECENTRALIZATIONAND WHY DECENTRALIZE?

Decentralization is usually defined as any act bywhich central government formally cedes powers toactors and institutions at lower levels in a political-administrative and territorial hierarchy (Mawhood1983; Smith 1985) (see Box 1).9 Decentralizationreforms typically are promoted to achieve six broadobjectives:

1. dismantling or downsizing central government,2. consolidating national unity,3. consolidating central power (by shedding risks and

burdens),4. increasing local participation and local democ-

racy,10

5. improving the efficiency and equity of local servicedelivery, and

6. strengthening local government (Ribot 2002a).

Environmentalists, natural resource managers, anddevelopment agents are also promoting decentraliza-tion as a way of increasing both efficiency and equityin natural resource management (Ribot 2002b).More recently, natural resource decentralization isbeing promoted as a means for giving substance topolitical rights (Anderson 2002; Kaimowitz and Ribot2002).

Decentralization comes in two primary forms:democratic decentralization (also called politicaldecentralization or devolution11), which involvestransfers of power to elected local authorities, anddeconcentration (also known as administrative decen-tralization), which involves transfers of power to localoffices of central government agencies, such as toappointed district officers or local offices of lineministries. Democratic decentralization enables localpeople to make decisions for themselves throughtheir representative local authorities. Whiledeconcentration does move government decisionmakers closer to the people they serve, allowing them

9


BOX 1 DEFINING DECENTRALIZATION AND RELATED TERMS

Decentralization is any act by which a centralgovernment formally cedes powers to actors andinstitutions at lower levels in a political-administrative and territorial hierarchy.

Democratic Decentralization (often also calledPolitical Decentralization or Devolution) occurswhen powers and resources are transferred toauthorities representative of and accountable to localpopulations. These are typically elected localgovernments. Democratic decentralization aims toincrease public participation in local decisionmaking. Democratic decentralization is aninstitutionalized form of the participatory approach.Of the two primary forms of decentralization,democratic decentralization is considered thestronger and the one from which theory indicatesthe greatest benefits can be derived.

Deconcentration or Administrative Decentralizationconcerns transfers of power to local branches of thecentral state, such as prefects, administrators, orlocal technical line ministry agents. These upwardlyaccountable bodies are appointed localadministrative extensions of the central state. Theymay have some downward accountability built intotheir functions, but their primary responsibility is tocentral government. Deconcentration is consideredthe weaker form of decentralization becausedownward accountability is not as well established asin the democratic or political form ofdecentralization.

Fiscal Decentralization is a term often used todescribe the transfer of funds (block grants orportions of national tax revenue) or fund-raisingpowers (ability to tax, charge fees, receive grants, orimpose fines) to local authorities. But in ourframework, this is not a form of decentralization.Although funding is essential in decentralizations, it

is merely a kind of power that is decentralized.Fiscal decentralization is therefore not an analyticcategory parallel to democratic and administrativedecentralization as it is often used. Fiscal transfersto democratic local government would simply bedemocratic decentralization, as would be the transferof regulatory responsibilities to the same body.

Comanagement is when government or donors formpartnerships with local bodies for the purpose ofsoliciting participation in decision making orimplementing management programs. Becausepowers are not transferred with comanagement, it isnot a form of decentralization. It is more akin to acontracting arrangement for implementation ofprograms or the provision of services. (Similar tocoadministration, described below.)

Coadministration is the transfer of central mandatesto elected local government or other local bodies insuch a way that the central state uses theseinstitutions as administrative bodies for centralgovernment objectives. This kind of transfertransforms democratic decentralization into a formof deconcentration.

Privatization is the transfer of powers to any non-state entity, including individuals, corporations, andNGOs. Privatization, although often carried out inthe name of decentralization, is not a form ofdecentralization. It operates on an exclusive logicrather than on the inclusive public logic ofdecentralization.

Subsidiarity refers to the principle that the relevantlevel for decisions is the most-local-possible level atwhich decision making will not result in negativeeffects at higher social or political-administrative levels.

Sources: Resosudarmo 2002; Ribot 2002b.

10


to better understand and serve local needs, it does notempower local people as democratic decentralizationdoes, but rather serves as an extension of centralgovernment into the local arena.

Figure 1 illustrates the various institutions beingempowered in the name of decentralization. As thediagram demonstrates, not all transfers of powerfrom central government to local institutions consti-tute decentralization. Local institutions function on acontinuum from public to private. All types ofinstitutions can serve both public and private ends.Local government bodies can be captured by elite to

become their private domains of operation, andprivate industries can be highly dedicated to publicservice. For the most part, however, local governmentbodies are intended to serve public interests whileprivate industries serve private interests.

The institutions in the middle of the spectrum, suchas customary authorities and NGOs, are public-private hybrids. When state powers are transferred tothese institutions, there may be some mechanismsfor public accountability, but that public may be asubset of the general population, such as a profes-sional or ethnic group. Also, these groups are

Central GovernmentDonors

Big NGOs

Democratic Local Government

Administrative Local Authority

Customary Authority

NGOPVOCBO

Individual or Corporation

DEMOC

RATIC

DECE

NTRA

LIZAT

ION

Government

Decentralization Not Decentralization

Public Public-Private Spectrum Private

Ministries:HealthEnvironmentEducation

3rd Sector

NON-MARKET PRIVATIZATION

HYBRIDS?

DECO

NCEN

TRAT

ION

PRIVATIZATIONPower Transfers

Contracts and Grants

In current reforms, powers are being transferred to many institutions in the name of decentralization. Formally defined, however, decentralization only refers to transfers of power within the political-administrative hierarchy of the state.

FIGURE 1 FORMAL DEFINITIONS: DECENTRALIZATION, NOT DECENTRALIZATION

11


dedicated largely to the interests of their membersonly. Transferring public resources to these bodiesconstitutes a form of non-market privatization sincethese groups may not be commercial but they areusually oriented toward their own private interests(even if those interests include public service).

A decentralization is effective to the degree

that it meaningfully represents the local

public—that is, the degree to which local

authorities and institutions are empowered

and downwardly accountable to the local

population.

Figure 1 is also organized with stronger forms ofdownward accountability on the left and weaker formson the right. The likelihood that a decentralizationreform will achieve the benefits that theory predictsdepends on the inclusiveness and public accountabilityof the institutions being established. A decentralizationis effective to the degree that it meaningfully repre-sents the local public—that is, the degree to whichlocal authorities and institutions are empowered anddownwardly accountable to the local population. Themore accountable local authorities are to the popula-tion, the more inclusive their decisions are likely to be.In theory, democratic decentralization is considered astronger form of decentralization than deconcentrationis; it is expected to result in greater equity and effi-ciency since representation is through elected authori-ties who are (at least through some procedure) system-atically accountable to the population. Deconcentrationis the weaker form since inclusion is through ap-pointed institutions that, while mandated to serve localpopulations, are less systematically accountable tothem.

Why should decentralization be expected to lead togreater efficiency in achieving certain environmentaland social outcomes? According to the logic ofcurrent decentralization, efficiency is linked to localenfranchisement or the broad-based inclusion of localpeople in public decision making. Recent theorists ofdecentralization assert that “underlying most of the

purported benefits of decentralization is the existenceof democratic mechanisms that allow local govern-ments to discern the needs and preferences of theirconstituents, as well as provide a way for theseconstituents to hold local governments accountable tothem” (Smoke 1999, 10).12 The underlying efficiencyand equity argument for decentralization is that localinstitutions can better discern and are more likely torespond to local needs and aspirations.13 This abilityis believed to stem from local authorities havingbetter access to information and being more easilyheld accountable to local populations due to theirclose proximity.14 Representative, or downwardlyaccountable and responsive, local authorities arecentral to this formula.15 When such local authoritieshave powers to make meaningful decisions over localmatters, people are enfranchised and there is reasonto believe that greater equity and efficiency will follow(see Box 2). Efficiency is also linked to inter-sectoralcoordination by the multi-sectoral decision making oflocal authorities and to increased flexibility that localintegrated decision making affords (Diana Conyers,personal communication, January 2004).

It is important to distinguish between formal andeffective democratic decentralization (see Agrawaland Ribot 1999). Formal democratic decentralizationrefers to power transfers within the state—that is, tolocal elected or appointed government authorities andline ministry personnel. Effective democratic decen-tralization refers to the transfer of power to anyauthority who is accountable to the population.Formal decentralization may be more likely to resultin effective decentralization than other forms sincesome accountability mechanisms are likely to be builtinto local government systems. Power transfers toelected local authorities who are upwardly account-able to the central state constitute effectivedeconcentration, whereas power transfers to custom-ary authorities who are downwardly accountable tothe population constitute effective democraticdecentralization. Examples of a spectrum of suchcases are presented in this report. The accountabilityof elected authorities, NGOs, and customary authori-ties varies widely—from effective democratic decen-tralization to effective privatization. The accountabil-ity—and nature—of local authorities must be deter-mined empirically.

12


DEMOCRACY AND NATURALRESOURCES: MUTUALREINFORCEMENT, SUSTAINABILITY,AND SCALING UP

What are the links between democracy and naturalresource management? How might they havecharacteristics that are mutually reinforcing? Ander-son (2002) and Kaimowitz and Ribot (2002) haveargued that local democratic decision making canimprove natural resource management while naturalresource management can enable local democracy.Democratic decision making has positive effects onnatural resource management for several reasons. For

instance, democratic forms of decision making are agood fit for the requirements of natural resourcemanagement. Environmental characteristics (e.g., thespecies of trees or the probability of erosion) varygreatly from area to area and change over time, as dothe multiple and overlapping claims and socialmeanings attached to natural resources. As a result,natural resource management often requires morelocal knowledge than other sectors, such as health,education, and infrastructure.16 Democratic represen-tation can help incorporate local knowledge andmultiple local voices, and hence the implications ofactions, into decision making about resource use.Inclusive local decision makers may, for example,

BOX 2 THE LOGIC BEHIND DECENTRALIZATION’S PROMISES

Democracy: Greater participation in public decisionmaking is often argued to be a good in and of itself.Democratic decentralization is believed to promoteincreased participation.

Equity: Decentralization is believed to help improveequity within local jurisdictions through greaterretention and more democratic distribution ofbenefits from local activities.

Efficiency: Economic and managerial efficiency isbelieved to increase by

Accounting for costs in decision making: Whencommunities and their representatives makeresource-use decisions, they are believed to bemore likely to take into account (or internalize)the entire array of costs to local people. Resourcewaste may result when outsiders orunaccountable individuals make decisions basedon their own benefits without considering coststo others.

Increasing accountability: By bringing publicdecision making closer to the citizenry,decentralization is believed to increase public-sector accountability and therefore effectiveness.

Reducing transaction costs: Administrative andmanagement transaction costs may be reduced

due to the proximity of local participants andaccess to local skills, labor, and information.

Matching services to needs: Bringing localknowledge and aspirations into project design,implementation, management, and evaluationmay better match actions to local needs. Bringinggovernment closer to people increases efficiencyby helping to tap into the knowledge, creativity,and resources of local communities.

Improving coordination: Decentralization is alsobelieved to increase effectiveness of coordinationand flexibility among administrative agencies andin development/conservation planning andimplementation.

Providing resources: Sharing the benefits of localresources also can contribute both todevelopment and to environmental managementagendas by providing local communities withmaterial and revenues.

Producing regulatory agents: Participation in theexercise of environmental regulatory powers cantransform local people into “environmentalsubjects.” It is an important part of thedevelopment of environmental consciousness.Local people and regulators are also more likelyto respect rules they have had a role in creating.

Sources: Agrawal 2001; Ribot 2002b; Larson 2000a.

13


better weigh the value of harvesting and selling aresource with the value of conserving it.

Just as democratic decentralization may improvenatural resource management, equitable naturalresource management and public control of naturalresources can be a lever to promote local democracy.For two reasons, natural resource management isparticularly well suited for reinforcing representativelocal authorities. First, the great variability of naturalresources privileges certain kinds of local knowledgeover outside “expert” knowledge. Harnessing this localknowledge requires greater local discretion in decisionmaking over natural resources so that local people canbring this knowledge into decisions—such discretionis an essential element of democracy (see Section I).Second, natural resources can support local authoritiesby providing them with substantive powers. Naturalresources can provide local authorities with financesthrough taxation and fees and with powers overlucrative activities and subsistence goods that aremeaningful and necessary to local people and theirlivelihoods. By providing a domain of discretion andsubstantive powers, natural resources have thepotential to strengthen, legitimize, and sustain localauthorities. They enable local authorities to act in waysthat are useful to local people, giving local peoplereason to engage local government. When governmenthas meaningful powers and is open to influence by thepeople they represent, local people are transformedfrom managed subjects into engaged citizens(Mamdani 1996a). (Kaimowitz and Ribot 2002.)

In the long run, the linkage between local democ-racy and natural resources may also help us to sustainand scale up efforts at greater local participation. Todate, most attempts to promote more inclusive andjust natural resource management have depended onthe following participatory methods facilitated byoutside agencies:17 participatory rural appraisal;participatory mapping; palavers; town meetings;stakeholder jamborees; terroir, commons manage-ment, adaptive comanagement, and network ap-proaches; and pluralist18 decision-making processes.These approaches may involve institutions such aseconomic interest groups, single-purpose and multi-purpose appointed or elected committees, joint-management and comanagement arrangements,

private voluntary organizations (PVOs), community-based organizations (CBOs), nongovernmentalorganizations (NGOs), government-organized NGOs(GONGOs), and World Bank-organized NGOs(known as BONGOs).19

Just as democratic decentralization is believed

to improve natural resource management,

equitable natural resource management and

public control of natural resources may be a

lever to promote local democracy.

Participatory approaches and the institutions theycreate have been used primarily to solicit input oflocal populations and groups into a decision-makingprocess that they may or may not control. Often theyare used to mobilize people to perform specific tasks.These approaches often impose outside agendas andmanipulate local populations rather than empowerthem (Cooke and Kothari 2001). But even when theyare used as a means of empowerment or of bringinglocal voice into decisions, they are difficult to sustainor scale up. These methods require labor-intensiveoutside facilitation,20 usually limiting them to thelifetime and location of the outside intervention.

Democratic decentralization, when well structured,provides institutional infrastructure—representationand meaningful powers—for sustained inclusion. Itis an institutionalized form of popular participation.Because local government, however imperfect it maybe, serves multiple purposes and is a permanentpublic institution, it can be sustained beyond the endof focused projects and outside interventions. Localgovernment can also be established across wholenational territories; therefore, it could be the basis forscaling up popular participation in natural resourcemanagement and use. Like direct democracy, thesenon-institutionalized participatory approaches areimpractical, as John Stuart Mills has argued, in all butthe smallest of communities (in Green 1993, 3).

Unfortunately, natural resource management canalso be set up in ways that undermine democratiza-

14


tion processes. Natural resource managementprograms often empower NGOs, other local organiza-tions and associations, and customary authorities inthe name of participation and decentralization. Butthere is little reason to believe that these local institu-tions systematically represent or are accountable tothe local population.21 Some are and some are not,depending on the particularities of the local historyand social relations (Ribot 1999a). Further, stake-holder approaches to participation, in which inter-ested parties are included in decision making, are nota form of democratic popular participation. Democ-racy is about accountability to the population as awhole, not just the inclusion of “stakeholders” withinterests. Empowering these alternative processes,institutions, and authorities with public powersundermines fledgling local democracies by creatingcompetition and denying them essential resources.

Incomplete or misdirected decentralization reforms

may ultimately undermine efforts to create sustain-

able and inclusive rural institutions.

In short, natural resources can be used to leveragethe development of local democracy. They can also beused in ways that undermine local democraticinstitutions. Local democracy, in turn, can supportthe sustained and scaled up forms of local participa-tion in natural resource management and use.Advocates of both local democracy and of just andsustainable natural resource management can usenatural resources for these joint long-term objectivesby establishing institutions for democratic manage-ment of all local resources and sectors.

FETTERED REFORMS

With all of its theoretical accolades, decentralizationhas progressed slowly. In practice, there is considerableconfusion and obfuscation about what constitutesdecentralization. In the name of decentralization, thepowers over natural and other resources that are beingallocated to local authorities are extremely limited andhighly circumscribed by oversight and unnecessary

restrictions (Ribot 1999a; Ndegwa 2002, 15). Further,they are often transferred to a variety of local institu-tions and authorities that may not be downwardlyaccountable. Most reforms in the name of decentraliza-tion are not structured in ways likely to deliver thepresumed benefits of decentralization and participa-tion. In a World Bank survey of thirty African coun-tries, decentralization was seen to have achievedmoderate levels of implementation in only half of thecases; nevertheless, all claimed to be decentralizing(Ndegwa 2002). Such incomplete or misdirectedreforms may ultimately undermine efforts to createsustainable and inclusive rural institutions.22

Non-implementation of decentralization is largelydue to government resistance (see Mansuri and Rao2003; Ribot and Oyono forthcoming).23 The leantoward democratic forms of decentralization isconfronting governments with the contradictions of acentury of extractive production-oriented policiestoward the rural world. As democratic decentraliza-tion is legislated and applied, the procedural objec-tives of inclusion and enfranchisement conflict withinstrumental objectives such as production targets ofcentral ministries or environmental objectives ofdonors and international NGOs (Shivaramakrishnan2000; Ribot 2002a; Baviskar 2004).24 Because localchoices are often inconsistent with central objectives,many ministries are designing their sectoral “decen-tralization” reforms to override local choice by usingnew elected local institutions to pursue their centrallydefined objectives. In doing so, they are devolvingobligations and creating requirements that diminishrather than enhance local discretion. Elected localauthorities often become administrators of centralmandates rather than executors of local interest. Suchapproaches undermine the core tenets of decentrali-zation.

Decentralization also threatens entrenched patron-age patterns from the highest level of central govern-ment to the front-line park rangers and forestryagents that are part of its implementation. It threat-ens powerful, often urban-based commercial interestswho fear losing access to productive resources. Itthreatens the roles of local elite, traditional authori-ties, and government administrators already presentin the local arena. In addition, decentralization evokes

15


fears among environmentalists and environmentalagencies that local people will not adequately manageimportant public natural resources. The mostcommonly expressed fears are that decentralizationwill lead to destruction of the resource, or that it willlead to conflict among parties interested in and usingthe resources. Environmentalists also fear that localofficials are corrupt and will use their control overnatural resources for personal profit. They alsofrequently feel that local authorities lack the capacityto adequately manage resources and conflicts. Manytimes these fears are founded. Often they are not.

It is not surprising that decentralizations areprogressing slowly. Their partial implementation hasoften led to a perception that decentralization ofnatural resource management is not working. Thosewho are threatened by decentralizations are quick todeclare their failure—even where they have neverbeen implemented. It would be a shame if slow anddisingenuous implementation were to consolidate abacklash before these reforms have had a chance tobe implemented and tested. Effective decentralizationis about creating a realm of local discretion defined byinclusive local processes and local authorities empow-ered with decisions and resources that are meaning-ful to local people. These conditions are not beingestablished.

MEASURING DECENTRALIZATION:CASES, METHODS, LIMITATIONS

The findings in this report derive from case-basedcomparative research conducted under WRI’sDecentralization and the Environment in Africaproject in Cameroon, Mali, Senegal, South Africa,Uganda, and Zimbabwe. The report also draws oncase studies from Bolivia, Brazil, Cameroon, China,India, Indonesia, Mali, Mexico, Mongolia, Nicaragua,South Africa, Thailand, and Uganda, which weredeveloped for the WRI-organized Conference onDecentralization and the Environment, held inBellagio, Italy, in February 2002. WRI’s Asia-basedResources Policy Support Initiative also provided casematerial from countries in the Mekong region. Theseresearch programs and the Bellagio meeting aredescribed in Annex A. Information on the locations

and themes of the case studies are provided in AnnexB. The report also relies on the available literature onthe topic. Most of the cases focus on forest re-sources.25 Examples are also borrowed from othersectors wherever they shed light on the dynamics inwhich we are interested.

It would be a shame if slow and disingenuous

implementation were to consolidate a backlash

before these reforms have had a chance to be

implemented and tested.

Had the research for this report focused the empiri-cal analysis solely on decentralization reforms thatproduce institutional arrangements fitting thedefinition of democratic decentralization, there wouldbe few cases to study. In decentralization reforms, thediscourse is often very progressive, with nationalpresidents and ministers proclaiming their intentionto create and empower new local democratic institu-tions. The laws, however, are often insufficient tocreate the institutional arrangements that the dis-course calls for. Even when the laws are sufficient tocreate some degree of decentralization they are oftenonly partly implemented or they are transformed inpractice, resulting in very little real decentralizationon the ground. This disjuncture between discourse,law, and practice makes decentralization and itsoutcomes extremely difficult to study. Rather thanbelieving the proclamations of governments, thelabels given to reforms, or the letter of the laws madein the name of decentralization, it is critical tocharacterize decentralization reforms empirically.

Most of the case studies presented in this reportfollowed Agrawal and Ribot’s (1999) “actors, powers,and accountability” approach to empirically character-ize decentralizations (see more detailed discussion inSection I). In this approach, researchers first focus onwhich actors are receiving new powers, what thosepowers are, and the kinds of accountability relationsin which those actors are located. The researchersthen identify outcomes that may be associated withthe institutional arrangements they observed. Box 3

16


BOX 3 MEASURING DECENTRALIZATION’S OUTCOMES

Has decentralization really occurred? Can social andenvironmental change be associated withdecentralized institutional arrangements?Characterizing decentralization involves evaluatingchanges in laws and their implementation, as well aschanges in local institutions and their powers,abilities, and accountability. Measuring outcomesrequires adequate data before and afterdecentralization or direct observation of processesaffected by new institutional arrangements. Theseinstitutional, social, and environmental changes aredifficult to identify and measure.

Connecting outcomes to decentralization andseparating these outcomes from other ongoingchanges such as natural variability of biophysicalprocesses is difficult. In Mongolia, for example, it ishard to attribute livelihoods and changes in naturalresource management to decentralization because ofthe many simultaneous overlapping sets of policyreforms. In China, simultaneous liberalization,logging bans, restrictions on steep slopewoodcutting, and changes in ownership all affectforest use, management, and local livelihoods,making it impossible to identify the unique effectsof decentralization. How does one know whetherdecentralization is responsible for these outcomes?What are the effects of other phenomena? Inaddition, measuring outcomes requires past baseline

data with which to compare them. If decentralizationhad not occurred, would the outcomes have beenless equitable, more equitable, or about the same?Some local violence in Mali may be caused bydecentralization, but has local violence increasedoverall since decentralization began? Finally,aggregating outcomes is another problematic aspectof measuring decentralizations. Some changes mayact in countervailing ways, such as whendemocratization may lead to reduced forest cover iflocal people value income over conservation. How dowe assess overall outcomes when some are positiveand others are negative?

Further, most decentralizations, even the mostlauded ones, are recent or have only been partlyimplemented. The Indonesian decentralizationreform laws were enacted only two-and-a-half yearsbefore they were studied, and implementation onlyofficially began a year before. The Maliandecentralization began with elections for localauthorities only three years before the study, but theenvironmental department has not yet officiallydecentralized any significant powers. This too makesthe measurement of outcomes tenuous. For thesereasons, the results presented in this report arepreliminary. More in-depth research is needed.

Sources: Bazaara 2002, 2003; Kassibo 2002a; Latif 2002;Mearns 2002; Resosudarmo 2002; Xu 2002.

describes some of the difficulties in measuringdecentralizations and their outcomes. The Africaresearch teams based their analysis on nine monthsof fieldwork during 2001 and 2002 using thisapproach. The participants in the Bellagio Conferenceon Decentralization and the Environment alsodeveloped their case studies using the actors, powers,and accountability framework.

ORGANIZATION OF THE REPORT

This report is organized into four main parts. Section Isketches the theoretical arguments for why decentrali-zation is believed to generate positive outcomes andlays out the framework that guided the research andanalysis in this report. The next two sections examinethe empirical data on the critical elements of decen-tralization: Section II considers institutional choice andaccountability, and Section III examines powertransfers and capacity. Section IV discusses lessonslearned, and Section V presents recommendations.

17


1

alleviation (Crook and Sverrisson 2001, iii);26 relief offiscal crisis (Meinzen-Dick and Knox 1999, 5; Olowu2001); political and macro-economic stability (WorldBank 2000, 107; Prud’homme 2001, 14); andnational unity and state building (Mamdani 1996a;Conyers 2000a, 7; Bazaara 2001, 7–13).27 Most of thelocal benefits from decentralization are believed tocome from increased popular participation, which, inturn, leads to increases in democracy, efficiency, andequity (see Agrawal and Ribot 1999; Ribot 2002b).28

Like decentralization advocates, natural resourcemanagement theorists and practitioners also empha-size the need for local participation as a means forincreasing management effectiveness and equity.29

Representative local authorities are a means forestablishing permanent (institutionalized andtherefore sustainable) and large-scale (coveringwhole national territories) popular participation.30

Representative local authorities can be a mechanismby which the knowledge, needs, and aspirations oflocal people are brought into public decision-makingprocesses and translated into policy. Pitkin (1967)and Manin, Przeworski, and Stokes (1999, 2) definerepresentation to be when authorities act in the bestinterest of the public.31 Representation is themechanism that makes decentralized institutionseffective (Agrawal and Ribot 1999; Smoke 1999).Representation is part and parcel of democracy,which can be substantively defined as leadershipthat is accountable to the people (Moore 1997). Butrepresentation also includes the means—powersand abilities—to transform popular needs andaspirations into policy (see Manin, Przeworski, and

THEORETICAL FOUNDATION OF

DECENTRALIZATION’S POTENTIAL

ENVIRONMENTAL BENEFITS

Is decentralization good for natural resourcemanagement? Theories tell us it may be. Thissection discusses why theory indicates that

certain kinds of decentralized institutional arrange-ments lead to positive outcomes. Theory helps us toidentify the main elements of decentralizations andthe causal relations between these elements andpredicted outcomes. With a basic understanding ofdecentralization theory, the empirical findings of thisstudy can be used to explain why the necessaryinstitutional reforms do or do not materialize, toevaluate the effectiveness of some of the institutionalarrangements of decentralization, and to identifyadditional factors that must be integrated into ourunderstanding and theories of decentralization. Fromthese findings we can develop new theory and we canalso develop recommendations so that policy makers,practitioners, and activists can seize opportunitiesand avoid pitfalls in their current efforts.

DEMOCRATIC DECENTRALIZATIONAND REPRESENTATION: A POLICYPROCESS

Theorists, practitioners, and advocates believe thatdecentralization can lead to a number of positiveoutcomes. These include: democratization andparticipation (Rothchild 1994, 1; Mbassi 1995, 23;Ribot 1996; Crook and Manor 1998); rural develop-ment (Uphoff and Esman 1974, xx; de Valk 1991; Roe1995a, 833; United Nations Development Programme1999; Helmsing 2001; Ribot 2002b); public serviceperformance (World Bank 2000, 107); poverty

18


Stokes 1999, 2). Representation is an institutional-ized democratic form of participation.

Manin, Przeworski, and Stokes (1999, 2) breakrepresentation into responsiveness and accountability.32

They model the policy process as a chain in whichpreferences expressed through various signalsbecome mandates and are translated into policies andthen outcomes. Responsiveness is the relationbetween signals and policies. Accountability is therelation between outcomes and sanctions. “A govern-ment is ‘responsive’ if it adopts policies that aresignaled as preferred by citizens,” they write. “Gov-ernments are ‘accountable’ if citizens can sanctionthem appropriately. . . .” A government is representa-tive because it is responsive and/or accountable.Accountability and responsiveness are the basicelements of representative local institutions andauthorities in strong or democratic decentralization.

An institution is democratic to the degree it is

accountable to society.

Accountability is the exercise of counterpower tobalance arbitrary action (Agrawal and Ribot 1999,478). It is manifested in the ability to sanction(Manin, Przeworski, and Stokes 1999, 2). Theaccountability relation is established through anensemble of sanctions. Accountability is constitutedby the set of mechanisms that, in theory, ensure thatpolicy outcomes are as consistent with local needs,aspirations, and the best public interest as policymakers can make them.

Responsiveness is a function of the multiplefactors that enable authorities to translate localneeds and aspirations into policy. These include thepowers to act on behalf of the people and the abilitiesto analyze and to translate signals into policy.Responsiveness is also a matter of will. It may bemotivated by the fear of sanctions or it may be afunction of ideology—that is, public spiritedness ora belief in public service. What makes democraticsystems unique is that they do not rely on ideologyalone to ensure that the full policy cycle is represen-

tative. Democratic systems rely on a mix of ideologyand sanction, with the sanction as the guarantee.(Manin, Przeworski, and Stokes 1999, 2.)

Accountability provides signals. Powers and abilitiesenable local authorities to translate signals intopolicies. Hence, representation, and thereforeeffective decentralization, requires local authoritieswho are accountable to the local populations and whoare empowered and able to conduct public affairs.This framework tells us that empowering localinstitutions that are not accountable to local popula-tions may not produce outcomes that decentralizationpromises, and creating accountable local authoritieswithout appropriate powers and abilities will certainlynot deliver the goods.

In the analysis that follows two central aspects ofdecentralization will be analyzed: (1) accountability,and (2) powers of local actors. To simplify the analy-sis, this report does not treat abilities separately frompowers.33 Issues of “capacity”, which is a kind ofinternal aspect of ability, will be treated along withmatters of power in Section III. Accountability andpower are each discussed in more detail below,followed by a discussion of other mediating factorsthat shape the effects of decentralization. Thesemediating factors ultimately need to be brought intoany model of decentralization.

ACCOUNTABILITY AND PRINCIPLESOF INSTITUTIONAL CHOICE

Accountability, as defined in the preceding section, isa defining characteristic of institutions. Institutionsare categorized by whom they are accountable to—andoften by how they are held accountable. We often callan institution democratic if it is accountable throughelections. Moore, however, defines democracy “as asub-species of a broader concept: the accountability ofstate to society” (1997, 3). With this substantivedefinition, he extends democracy far beyond elec-tions, to include any system based on a large array ofpossible accountability mechanisms (see Annex C fora discussion of accountability mechanisms). Hence,an institution is democratic to the degree it is ac-countable to society.

19


By the formal definition, decentralization consists oftransfers within the political-administrative structureof the state, which is a body that ostensibly has someaccountability to the people whether directly (as indemocratic or political decentralization) or indirectly(via the central government, as in administrativedecentralization or deconcentration). Democraticdecentralization and deconcentration are the twokinds of institutional arrangements that fit thisdefinition. They are based on elected and appointed

administrative authorities and have varying degreesof downward accountability.

Many other kinds of transfers are made in the nameof decentralization. Figure 2 provides a schematicdesigned to classify reforms where transfers takeplace to elected local government, local administrativebodies, NGOs, community groups, customaryauthorities, or private corporations and individuals.Transfers to such bodies as NGOs, user groups, and

Accountability

Power Transfers

Contracts and Grants

Central GovernmentDonors

Big NGOs



Customary Authority

NGOPVOCBO


DEMOC

RATIC

DECE

NTRA

LIZAT

ION


DECO

NCEN

TRAT

ION

PRIVATIZATION

Local Populations

STRONG WEAK? ?

NON-MARKET PRIVATIZATION

HYBRIDS?

Effective decentralization is when powers are transferred to downwardly accountable local institutions. In this diagram, the accountability arrows reflect a mix of what is expected and observed about different institutions. Decentralization is more effective when downward accountability is stronger (Agrawal and Ribot 1999). Democratic decentralization is the strongest form—when it has strong electoral mechanisms of downward accountability. Deconcentration is a weaker form because downward accountability is less systematic. The degree to which customary authorities and other local institutions are downwardly accountability is uncertain and varies greatly.

FIGURE 2 ACCOUNTABILITY DEFINES EFFECTIVE DECENTRALIZATION

20


customary authorities create public-private hybridsand represent something between decentralizationand privatization. The accountability of these hybridsis unsystematic, indeterminate, or ambiguous vis-à-vis the powers transferred to them. Transferringpowers to these institutions is a kind of non-marketprivatization because, although they are not marketoriented, they are only accountable to their members;such transfers, therefore, do not represent effectivedecentralization.

All institutions and individuals, public and private,are accountable to regulatory agencies and areobligated to operate within the bounds of existinglaws. Regulatory bodies circumscribe the domain offreedoms for each type of institution. Institutions aredefined by the direction of their accountabilityconcerning discretionary actions within their domainof freedom. With respect to their legal discretionarypowers, private bodies are not accountable to thepublic at large. They may use a private forest orprivate land as they wish without consulting any-one—provided they are not breaking any laws orregulations. This differs from a democratic authority,which owns or holds in trust a forest or plot of land.Public authorities must, like private operators, respectexisting regulations, but democratic authorities areaccountable to the local population with respect tohow that property is used. Since private bodies arenot accountable to the population as a whole, trans-fers to private bodies constitute privatization, notdemocratic decentralization (see discussion under“Privatization is not decentralization” in Section III).

In practice, all institutions operate within a mix ofaccountability relations. Elected officials are oftenmore accountable to their party than to the peoplewho elected them. Administrators may be profoundlydedicated to serving the needs and aspirations of thepeople (see Tendler 1997). So-called customaryauthorities in some places and times may be perfectlydemocratic (for example see Spierenburg 1995).These categories are not fixed. Rather, they provide aheuristic device for analyzing whether institutionscan be considered to be substantively democratic,which is the foundation of effective decentralizationreforms. This framework allows us to focus on theelements of decentralization that theory tells us will

lead to improved outcomes. It allows us to evaluateempirically whether the institutional conditions are inplace—the mix of downward accountabilities anddiscretionary powers—that are expected to generatepositive outcomes.

Principles of institutional choice can be developed toguide the selection of local institutions withindecentralizations. Downward accountability is thefirst criterion. The measure of accountability iswhether institutions are subject to sufficient account-ability mechanisms, such as elections, public meet-ings, public reporting requirements, recall, account-ability-inducing magic, protest, etc. (see Annex C).Other important institutional characteristics includeamenability to influence via legislation, which makesa given choice a policy option; replicability andgeneralizability for scaling up; and sustainability overtime. The advantages of elected local government areamenability to legislative manipulation, national-scalereplicability, systematic accountability to the popula-tion (depending on the structure of its electoralsystem), and sustainability as institutionalizedrepresentation. While some forms of customaryauthority may be representative and sustainable, theycannot be replicated and spread across the nationalterritory. While some NGOs may serve the bestinterests of local people, their accountability is notsystematic and they may not be sustainable. Whilethese principles point toward elected local govern-ment, in many places democratic local governmentdoes not exist—or it is captured by central authoritiesmaking it upwardly accountable. In these instancesMandondo (2000a) recommends working withwhichever institutions are making public decisionswhile applying multiple accountability mechanismsto these institutions.

Two other issues emerge in local institutionalchoice when there are multiple institutions involved.Too many local institutions or too many levels of localgovernment can diffuse authority among too manyactors, making them all ineffective and underminingtheir legitimacy to act in the public domain. Competi-tion among local institutions is also believed by someto create greater efficiency in serving public interests,but the evidence does not support this (Tendler2000). When there are multiple institutions, who

21


depends on whom is important when public re-sources are involved. NGOs, traditional authorities,and other private bodies holding public powersshould be accountable to representative institutionsfor decentralization to be effective (Blair 2000).

POWERS: DISCRETIONAND SUBSIDIARITY

For any democratic process, discretionary powers arenecessary so that elected authorities can be respon-sive to their constituents. Unlike mandates, whichrequire local authorities to act on behalf of themandating agency, discretion enables representativesto act on behalf of their constituents. The domain oflocal discretion is the basis for local democraticdecision making (Ribot 1999a, 2001). Withoutdiscretionary powers, local government is merely anadministrative extension of central government.Having discretion is what defines meaningfulauthority in local democracy. Discretion defineseffective decentralization.34

Effective decentralization of natural resourcemanagement requires that local authorities have adomain of discretion related to the management oftheir natural resources. First, discretion is needed sodecision makers can adapt to the great spatial,temporal, and social variations of use and manage-ment. Second, discretion is often needed in makingallocative decisions over use and management due tothe lucrative nature of the many opportunities thatnatural resources present. Discretion in allocation canhelp optimize use by allowing local authorities tosanction abusive users and reward good managers.Discretion in allocating lucrative opportunities alsogives local authorities the power to bargain withoutside users. Third, discretion is necessary forflexible and locally informed negotiation among themultiple claims by pre-existing authorities, citizens,outside commercial interests, and local government.

The domain of local discretion is important since itis the domain of freedom around which democraticlocal government, citizenship, and civil society canform and develop. Even if elected, local authoritieswho do not hold discretionary powers are not demo-

cratic since they cannot be responsive. Nor do theyhave a meaningful domain of action within which thelocal population can judge and respond to theirperformance. It is the ability of local people to selecttheir leaders based on performance that makeselections meaningful. It is also the exercise of powersby local leaders that allows them to develop locallegitimacy and to cultivate a positive relation withsociety.

Without discretionary powers, local govern-

ment is an administrative extension of central

government. Having discretion is what defines

meaningful authority in local democracy.

Citizenship crystallizes around empowered localauthorities. It is with respect to the powers of localauthorities that citizens are motivated to becomeengaged in self-government and civil society ismotivated to organize and apply influence.35 Peopledo not organize to influence authorities that cannotrespond. In addition, Agrawal (2001) argues that localauthorities themselves engage positively in manage-ment when they control the means to regulate—inother words, it is by wielding power that local authori-ties become identified with and take up resourcemanagement and regulation (see Box 4).

The kinds of powers that can be devolved toconstitute a domain of discretion are the classicpowers of government: decision making, implemen-tation and enforcement powers (executive), power tomake rules or bylaws (legislative), and dispute-resolution powers (judicial) (see Agrawal and Ribot1999). But devolution of these powers does notnecessarily create discretion. Powers can be devolvedto give local authorities independent choices, or asmandates to implement policies handed down byhigher levels of government. These powers may comefunded or unfunded by central government, or theymay include powers to acquire funds—such asthrough loans, taxation, fees, fines, and sale ofconfiscated goods.

22


Participation or representation always involvespowers—there must be something in which toparticipate. Which powers local people should holdremains a difficult question in most participatory anddecentralized natural resource management. Theview that local participation is beneficial for naturalresource management was strongly promoted at theUnited Nations Conference on Environment andDevelopment in Rio de Janeiro (Meinzen-Dick andKnox 1999; Twyman 1998). The conference’s Agenda21 states that “one of the fundamental prerequisites

for the achievement of sustainable development isbroad public participation in decision making”(1992a, sect. 23.1). The Rio Declaration also empha-sizes that “environmental issues are best handledwith the participation of all concerned citizens, at therelevant level” (United Nations Conference onEnvironment and Development 1992b, Principle 10).

The words “at the relevant level” are critical. Howcan the relevant level for decision making be deter-mined? The principle of “subsidiarity” offers guid-

BOX 4 SUSTAINED FOREST MANAGEMENT IN INDIA THROUGH THE TRANSFER OF“MEANS OF REGULATION”

In Kumaon, India, local people became engaged inenvironmental management when the tools andpowers of regulation, the means of regulation, wereplaced in their hands. Kumaon presents one of thelongest-standing cases of decentralizedenvironmental management. After a series of localrebellions against British timber extraction in theforests of Kumaon in the 1930s, the British bannedcommercial timber extraction and devolved therights to manage forest use to van panchayats(elected local forest councils). Through theexperiment in Kumaon the forests have providedlivelihoods for more than seventy years.

Agrawal observed that in Kumaon “once the centralgovernment created local centers of decision makingand granted them the authority to regulate forests, itbecame possible to use available informationeffectively in the service of environmentalconservation.” The transfer to communities of whatAgrawal calls “means of regulation,” such as rulemaking and enforcement, leads to the engagement oflocal people, along with their knowledge and skills, inthe management and use of resources. In addition,Agrawal argues, “communities accomplish localregulation at significantly lower costs than anycentral government can.”

It is important to note that in the Kumaon case,local autonomy was limited to management of non-

commercial uses. Much of the environmentalstability could be attributed to the timber ban. Inthis case the forests were protected from outsidecommercial loggers as well as from local interests.But would there have been such a success hadthere been more local autonomy? Would theforests have been in good condition if localactivities had not been limited to raising revenuefrom fodder and dead trees, but had beenextended to deriving revenue from timber sales?This is unclear.

Not all of the communities in the KumaonHimalya were able to enforce forest managementregulations. Agrawal observed that the failure ofsome of the elected forest councils “to enforceforest-related rules is most evident when villagersdo not have sufficient forests under their control,when villages are very small or very large, whenlevels of migration from specific village to theplains are very high, and when governmentofficials provide little or no support to councilmembers trying to protect the forest.”Nevertheless, “the seventy-year history ofdecentralization in Kumaon demonstrates thepossibility that local regulatory institutions canmanage resources successfully.”

Source: Agrawal 2001, 209–10.

23


ance. According to this principle, the relevant level forany given decision is the most-local-possible level,provided that making the decision at this level doesnot cause negative effects at higher social or political-administrative scales (Føllesdal 1998; Rocher andRouillard 1998). In other words, according tosubsidiarity notions, decisions that can be madelocally without jeopardizing social and ecological wellbeing should be located at the local level—notprecluding some framework of standards andregulations.36

Other important questions must be asked indeveloping criteria to determine which powers belongat which level and under what kinds of authorities.Which powers should be public, private, or collectivein nature (Ostrom 1990, 30)? Which should belocated at national, regional, district, or local levels?The answers to these questions are technical andpolitical—two dimensions that are often conflated(Bazaara 2003). Some issues have to do with thenature and scale of a natural resource—protecting awatershed or a wildlife species may require somemanagement functions to be decided and imple-mented at a large political-administrative scale. Somedecisions may have to do with economies of scale intechnical knowledge or equipment. Some have to dowith the effects of local decisions on the broaderpopulation—such as when an upstream forest usercauses erosion, flooding, and siltation for districtsdownstream. Others may relate to the need forleveling mechanisms to effect interjurisdictionalequity at a national level. Ultimately, all powerallocations are political in so far as they createwinners, losers, and conflicting interests.

Developing subsidiarity guidelines should take all ofthese considerations into account. Most fundamen-tally, these principles must emphasize that discretionis essential to local democracy—which is the key tothe ostensible benefits of decentralized management.That discretion, however, must be well chosen andbounded so that its effects do not harm society or thenation as a whole (see “Setting the boundaries onlocal power” in Section III).

MEDIATING FACTORS:DECENTRALIZATION ISNO PANACEA

Even perfectly representative and downwardlyaccountable local authorities may over-exploitresources and ignore minority interests if given theunbridled power to do so. Decentralization is not astand-alone panacea for natural resource manage-ment—or for the management of any public re-source. When it is profitable, collective decisionmakers are likely to exploit natural resources ratherthan conserve them, especially if they do not bear theindirect costs, such as when deforestation by up-stream users leads to downstream flooding and damsiltation, or when the present needs are urgent andlocal costs of exploitation accrue later. Ensuringpositive environmental and social outcomes requiressome regulations and standards. It requires meansfor ensuring that nationally defined environmentaland social concerns are taken into account. It is naïvepopulism to assume otherwise.

Even perfectly representative and downwardly

accountable local authorities may over-exploit

resources and ignore minority interests if given

the unbridled power to do so.

The ways in which devolved powers are useddepends on the incentives, including, but not limitedto, sanctions, that decision makers face. If it isprofitable to cut down the forests, then there is littlereason to believe that a local collective will or shouldhave different values than an individual. They may,however, weigh those values differently than dooutside interests or wealthy individuals. The advan-tage of collective decisions is that the broader spreadof interests may bring certain marginal values intodecision making that may be forgotten in a privatescheme—helping to “internalize externalities.”Collective local decision making may also bring more-detailed local knowledge to decisions and may alsoweigh the costs of labor differently. Further, thepublic, as opposed to wealthy corporations, especiallyif the public consists of poor primary producers, does

24


not have an exit option that would allow it to exploitthe resource and then leave (Bardhan, Ghatak, andKaraivanov 2002). If given the power, local decisionmakers are likely to block extractive activities thatprofit non-local actors at the expense of the localpopulation, as they have done in Nicaragua andSenegal (Ribot 2000; Larson 2002). There is alsogood reason to believe that local people are morelikely to respect rules that they have had a role increating (Agrawal 2001; Jon Anderson, personalcommunication, January 2004; Larson 2004a).

Nobody reasonably advocates devolution of all

powers.

However, if alternative development opportunitiesare not created or if the benefits to the community asa whole are greater than without over-exploitation,local people are likely to over-exploit the resource(Baviskar in Latif 2002, 38). Decentralization isexpected to increase inclusion, equity, and effi-ciency—even if that leads to democratic, equitable,and efficient transformation of natural resources intoother, more-locally-preferable investments. Hence, ifthere are values that government or society at higherlevels wants protected, rules, regulations, andstandards are needed—hopefully elaborated incooperation with local populations. Nobody reason-ably advocates devolution of all powers. Local discre-tion with a minimum regulatory infrastructure mustbe established to bound and protect local freedom.

Larson (2003) points out that “the perfect combina-tion of formal powers and downward accountability[i.e., some ideal decentralization] may never actuallyexist in practice or, at best, will only exist in a state ofdynamic tension. At the same time, it is unlikely thatthese two factors alone would necessarily lead to goodmanagement decisions.”37 She argues that additionalfactors must be brought in to explain the relationbetween decentralization and outcomes (also seeOyono 2003).38 These include “power relationsamong local actors [i.e., among various elite andmarginal groups], the overall structure of incentivesregarding the resource in question [prices, values,regulatory arrangements], environmental and socialideology [beliefs that shape how resources are valuedand used], and local government capacity [its ability toassess and respond to local needs]” (Larson 2003).39

These factors mediate between decentralization andoutcomes (also see discussion in Latif 2002).

The mediating factors between decentralization andoutcomes still need further theorizing. The wholerange of factors that shape outcomes need to be takeninto account in preserving greater public good—butwithout allowing the greater public good to becomean excuse to avoid decentralization. This is one axis ofLarson’s (2003) “dynamic tension.” From thesefactors, measures for protecting threatened values canbe developed to accompany decentralization reforms.Some of these, including minimum environmentalstandards, are discussed in the analysis section of thisreport.

25


2

approaches, though widespread, do not constitutedemocratic decentralization since they do not transferpublic powers nor do they create or empower repre-sentative local authorities. They may at times beinclusive and may even integrate elected authoritiesto represent local populations, but participatoryapproaches and comanagement arrangements are notdecentralization because, although they solicit adviceon the exercise of powers and may even share powers,they do not clearly transfer powers.40

The choice of institutions and the ways in whichthose institutions are accountable have profoundimplications for the effectiveness of decentralization.Central actors generally choose to devolve powers tolocal actors with whom they already have relation-ships. In addition, they choose policy instrumentsthat allow them to allocate state-controlled resources(such as subsidies) rather than those that operatewithout central redistribution (such as taxation)(Bates 1981). Having the power to allocate stateresources allows central state actors to create ormaintain patronage networks. Indeed, it is for thisreason that the “means of transfer” is so important(see Section III). Means of transfer that permitcentral agents to allocate resources permit centralgovernments to cultivate patronage networks and tomaintain control over politically valuable resources. Itis therefore important to note the choices centralgovernments and projects are making and how theyconstruct the authority and the accountability of thelocal actors to whom they transfer powers (Ribot1996, 1999).

LOCAL ACCOUNTABILITY—

INSTITUTIONAL CHOICES IN PRACTICE

Accountability of local governance institutionsconstitutes a major problem of democraticprogrammes.

Olowu 2001, 51

The case that decentralised political bodies can deliverservices more efficiently and more responsibly dependson adequate mechanisms for political and financialaccountability. Wealthy countries with a long demo-cratic tradition possess elements of accountability(educated population, media, administrative andjudicial capacity, regularity of elections, etc.) that donot necessarily operate well elsewhere. In the absenceof effective accountability mechanisms, there aredangers that decentralised resources will be improperlydiverted and/or that the benefits of decentralizationwill be captured by insiders. Lack of administrativecapacity and of complementary resources may alsoprevent political efficiency benefits of decentralisationfrom being realised in practice. Assuring accountabil-ity is critical. . . .

PMA Grant Study 2001, 29–30(Plan for the Modernisation of Agriculture, Uganda)

Natural resources have been under administrativemanagement for a long time. This report is con-cerned with the ways that current decentralizationsare moving away from this top-down form of man-agement. In many instances, popular or stakeholderparticipatory processes and comanagement ap-proaches are used to mobilize local populations toparticipate in the management and use of localresources. As discussed in the previous section, these

26


Truly democratic local institutions, which areaccountable to local populations, may be less likely tosupport central objectives—either for the patronagepurposes of central state actors or for the purposes ofthe maintenance and reproduction of local elite. It isfor this reason that central states consistently chooseor construct upwardly accountable local institutions;they do this by using party-list electoral systems, bytransferring the most significant powers to theappointed local administrators rather than electedbodies, by creating ad hoc committees or NGOs, andby resuscitating customary authorities. In thismanner, elite capture and patronage are linked.Central authorities gravitate toward upwardly ac-countable institutions and local elite to forge mutuallysupportive arrangements.

This section discusses the accountability of actorstypically chosen as the local recipients of decentral-ized powers in current reforms. These include electedlocal government, customary authorities, NGOs andcommittees, and private bodies. The section also willexamine crosscutting institutional-arrangementproblems that affect the inclusion of marginalized orpoor groups in decision making, issues related to therepresentation of local populations in the nationalarena, and the role of the central state.

ELECTED LOCAL GOVERNMENT

Elections are one of the most common means forestablishing downward accountability of localauthorities.41 Hence, theorists view elected localauthorities as the most appropriate recipients ofdecentralized powers (Mawhood 1983; Crook andManor 1998; Agrawal and Ribot 1999; Smoke 1999and 2001). There is some evidence that elected localauthorities can improve natural resource manage-ment. Bazaara (2002a, 29) reported that in Ugandaelections led to the conservation of some protectedtrees (see Box 5). 42 Mehta (1996) reported a case inIndia in which villagers replaced an old unresponsiveelite with new representatives willing to set up a jointforest management scheme.

Evidence from other sectors of the importance ofelections is encouraging. Crook and Sverrisson (2001,

50) observed in their study of decentralization’seffects on poverty reduction, “Fair and competitiveelections were a key factor in developing publicaccountability in the most successful cases.”Echeverri-Gent (1992) observed that competitive localelections in West Bengal, India, helped make policymore responsive to the poor. Similarly, in ColombiaFiszbein (1997a) found that competitive electionswere important in making decentralization effective.In Bangladesh, local council leaders dominated localdecision making and used their position for self-promotion, but in elections ninety percent of themwere not reelected (Parker 1995, 26). Parker arguesthat with open elections, locally controlled resourceswere redirected towards more micro-level publicworks—precisely because open elections make localpoliticians more likely to meet felt local needs.

Electoral systems do not always create downwardaccountability. While openness and competition are

BOX 5 TREES SAVED BY LOCALELECTIONS IN UGANDA

Local democratic checks and balances in theMbale District of Uganda appear to haveprevented corruption by elected officials, savingprotected mahogany trees. A Chinese forestrycompany asked the Mbale District council toallow it to set up a sawmill in the district. Thecouncilors happily rented the company idlebuildings to house the mill.

When the mill was set up, the companyapproached the forestry officer who theninformed them that they had not followed properprocedure and that commercial cutting was notallowed in the district. The councilors triedpressuring the forestry officer, but the officerinstructed them to ask the people. Sinceelections were looming, the councilors did notdare pursue the matter. As Bazaara points out,“the trees were saved because of elections.”

Source: Bazaara 2003, 29.

27


important, another key aspect of electoral systems isthe ability of the electorate to choose leaders and toreject incumbents (Manin, Przeworski, and Stokes1999, 2). But not all electoral systems allow such achoice. As a key accountability mechanism, electionsneed to be carefully scrutinized. While they may beimportant (where they exist), they are not always wellstructured or sufficient in and of themselves toguarantee participation. Electoral systems must beanalyzed for how candidates are chosen, suffrage,term lengths, and means of recall (Ribot 1999a;Prud’homme 2001). Just because local officials are“elected” does not make them downwardly account-able. Their accountability is at least partly a matter ofhow they are elected.43 For example, systems orga-nized around party slates that do not admit indepen-dent candidates appear less downwardly accountablethan systems with independent candidates.

In countries with independent candidates in localelections, authorities are more likely to be account-able to the people. Unfortunately, independentcandidates are relatively uncommon. From amongthe countries represented at Bellagio, only India,Mali, Mexico, Uganda, and Zimbabwe admit indepen-dent candidates in local elections (Agrawal 2002;Baviskar 2002; Bazaara 2002b; Kassibo 2002; MeloFarrera 2002). Nevertheless, the degree to which localauthorities in these countries are downwardlyaccountable in practice varies greatly (Alavi 1973;Parker 1995; Mehta 1996; Ribot 1996, 1999; Crookand Manor 1998). Admissions of independentcandidates should, however, help to produce down-wardly accountable authorities.

Most African governments have created localgovernments that are upwardly accountable to thecentral state (Mawhood 1983; Wunsch and Oluwu1995; Ribot 1999a; Oyugi 2000; Crook andSverrisson 2001). Around the world, upward account-ability of local elected governments is producedthrough party-list elections, as in Bolivia, Brazil,Burkina Faso, Cameroon, Indonesia, Ivory Coast,Mongolia,44 Nicaragua, Niger, Senegal, South Africa,and Zimbabwe. Candidates are effectively appointedby central political parties—usually the party inpower. The candidates are often viewed by localpeople as accountable to the party, rather than to the

local population (Ribot 1999a; Resosudarmo 2002).45

In places where only the party in power can organizelists across the country and there is no real competi-tion among parties, these systems leave little chancefor local populations to choose their own representa-tives (Ribot 1999a).

There are many examples where representation isdysfunctional due to electoral systems based on partylists or slates (see Ribot 2002a). Under pressure fromthe political party that put them on electoral lists,local councilors in Senegal allow forest exploitation tocontinue against the expressed wishes of the peoplewho voted for them (see Box 6). In Zimbabwe, ruralcouncilors overseeing environmental projects havenot reinvested in the community; local peopleattribute this to their accountability to their politicalparty (Mapedza and Mandondo 2002; Mandondo andMapedza 2003). A similar electoral system in Indone-sia also may help to explain forest over-exploitation(Resosudarmo 2002).46

Under some conditions, party-list systems can servetheir constituents relatively well. Dupar andBadenoch (2002, 11) observed that in Thailand, whichhas embarked on some of the most ambitiousdecentralization reforms in mainland Southeast Asia,political parties have sought to strengthen theirsupport in rural areas by increasing the voice of ruralconstituencies. Democracy theorists believe that thesekinds of results are likely when there is significantcompetition among political parties (Schumpeter1976[1943]). The problem appears to be that in mostdeveloping countries, opposition parties have insuffi-cient resources to organize at the level of localelections; therefore, competition remains low.

Even when the electoral system is well crafted,many elected officials are not accountable to theirconstituencies. Elite capture and other problemsconstantly emerge (Alavi 1973; Mehta 1996; Smoke1999, 17–9; World Bank 2000, 109, 121; Crook andSverrisson 2001; Ribot 2001a). There are many waysthat local elite or political parties are able to capturethe electoral process, bringing the local accountabilityof leaders into question (Crook and Sverrisson 2001).Members of the rural elite manipulate candidacies,electoral processes, and persons in elected or other

28


leadership positions (Mehta 1996). Local people oftenvote for elite candidates due to their economic andsocial dependence on members of the elite forsurvival (Alavi 1973).

The effectiveness and legitimacy of local authoritiescan be undermined in numerous ways. Elite captureand upward accountability can compromise theeffectiveness of representation and service delivery. Thefailure of central government to empower elected localauthorities also undermines their effectiveness andlegitimacy. Local authorities need discretion overmeaningful matters in order to serve populations. Whenthey do not or cannot deliver, people lose confidence andinterest in elected authorities (see Box 7).

Despite their limitations, elected representatives areessential building blocks for accountable local govern-ment. But even where there are well-structured electedlocal authorities, governments and donors often avoidthem in favor of other kinds of local organizations(Romeo 1996). Often, they are avoided for being too“political” (Baviskar 2002; cf. Ferguson 1996) or forbeing inefficient or lacking in capacity. While thesearguments are often wrong (see capacity discussions inSections III and V), powers are instead transferred tolocal administrators, customary authorities, commit-tees, NGOs, or other private actors.

While local governments may be more likely to bedownwardly accountable, depending on the kind of

BOX 6 ELECTED COUNCILLORS IN SENEGAL GIVE AWAY FORESTS AGAINST THEWILL OF LOCAL PEOPLE

In the rural community of Makacoulibantang, inthe Tambacounda region of Senegal, villagers havebeen fighting against commercial charcoalproduction since the early 1990s. For years,production continued because the forest serviceinsisted that the rural population had no rights overthe forests. Foresters told local people that theforests belonged to the nation as a whole andneeded to be exploited to supply the capital citywith fuel.

In 1998, a new decentralized forestry law gave ruralcommunities and their rural councils (the most-local level of government) three key new rights: (1)the right to manage their forests—albeit followingoverly elaborate management plans outlined by theforest service, (2) the right to determine who couldproduce within their forests, and (3) the right toapprove, through the president of the rural council,any production within their community before ittakes place. The forest service also was supposed tochange from a system of nationally fixedproduction quotas to a system where productionquotas would be allocated based on the ecologicalpotential of each forest. A three-year transitionperiod ending in February 2001 was allowed for thefull implementation of the new laws.

But in 2003 the national forest service was stillmaking all key decisions. For the 2003 productionseason, the forest service set a national charcoalproduction quota and allocated quota shares amongprofessionally licensed forestry firms. The quotawas based on the previous year’s production ofeach forestry firm, not on the ecological potentialof each rural community’s forests. The annualcharcoal production quota in Senegal has longbeen set far below the level of national demand—causing a supply crisis every year that must beresolved by allocating supplementary quotas late inthe season. The minister announced at a nationalmeeting that he was lowering the 2003 quota inorder to protect the nation’s forests.

The head of the national union of commercialforesters called the low quota a “political quota”intended to make donors happy. But it was evenmore political than that. Setting the quota far belowdemand creates a patronage resource that the forestservice and ministry gets to allocate as extra quotasto make up the difference between the initialnational quota and consumption. Contrary to thenew laws, after setting and allocating the quota atthe national level, the quota is passed to the regionalcouncils to distribute to the merchants in each

29


electoral system and on the particular local culturaland historical context, there are also other advantagesto working through local elected authorities. Com-pared to other local institutions, local governmentsare formally institutionalized and therefore can besustainable over time and scaleable over nationalterritories. Sustainability and the ability to scale upneed to be explored further, but they appear to be twoof the comparative advantages of working with localgovernment.

In short, powers transferred to authorities chosenthrough competitive local elections that admitindependent candidates appear most likely to be

exercised in a locally accountable manner. In addi-tion, when decentralizations are based on democraticlocal government, they may be more sustainable andscaleable across national territories. Local govern-ment is still, nevertheless, subject to elite capture andcorruption; also, it often lacks some of the skillsnecessary to respond effectively to local needs.

Authorities chosen through competitive localelections that admit independent candidates appearmost likely to be locally accountable; in addition, theymay be more sustainable and scaleable acrossnational territories.

BOX 6 CONTINUED

region. The merchants then launch production byinstalling their woodcutters in forest villages. Therural councils are not even involved.

Despite widespread popular resistance to charcoalproduction in Makacoulibantang, the president ofthe rural council gave his authorization forproduction when the forestry officer asked for it.Indeed, all of the rural council presidents in theregion did the same. When asked why they areallowing charcoal production in their communitieswhen almost everyone is against it, the threecouncilors interviewed said they had no choice.When pressed further, it became evident that theywere under pressure from the ruling party to allowcharcoal production. When asked why he did notexercise the right given in the new forestry laws,one council president said, “The law is the state, theparty is the state. What can we do?”

This outcome can be at least partly explained by theorganization of the electoral system. The localelectoral laws in Senegal do not admit independentcandidates. Further, they ensure that councils arecontrolled by only one party by giving half of thecouncil seats to the party with the most votes anddividing the other half proportionally. Hence, if the

leading party gains twenty percent of the vote, theyget sixty percent of the seats on the rural council—that is half of the seats plus twenty percent of theother half. The ruling party usually leads. Thepresident of the rural council is caught between thelaw, which is the state, and the minister, who is thestate and the party. In short, the president actedagainst the will of the population that elected himbecause he is beholden to the party, which selectsthe candidates on the party slate, and to the state,which allocates the resources his rural communityneeds. The result is that the forests of the ruralcommunities continue to be cut by migrantwoodcutters installed in their forests by urbanmerchants who have great influence over powerfulpolitical actors.

This situation was further exacerbated by a failure ofcivic education. Based on interviews, rural residentsare unaware that the presidents of the rural councilshave the power to stop charcoal production in theircommunities. Armed with this knowledge, theymight try to hold the rural councils accountable.

Sources: République du Sénégal 1998; Ribot 2000; Thiawand Ribot 2003; Interviews by author of rural councilors,village chiefs, and villagers, April 2003.

30


LOCAL ADMINISTRATIVE BODIES—DECONCENTRATION

Deconcentration accompanies all decentralizationsand often substitutes for them. Deconcentration is apoor second choice to democratic decentralization. It

does not establish democratic forms of decisionmaking. In addition, it relies on the perceptions andmotives of centrally appointed agents to discern theneeds of local peoples. The deconcentration optiondoes not provide representative authorities withdiscretionary power. Without empowered representa-tives, the system must rely on the good will ofcentrally appointed agents and their superiors tomake decisions on behalf of local populations, andtheorists do not expect to get as effective internaliza-tion of externalities. Deconcentration is not verydifferent, however, from decentralization to upwardlyaccountable party-chosen local representatives. Thereis, nevertheless, evidence that deconcentration canserve local interests well. In Brazil, for example,Tendler (1997) observed that a system of performanceawards led local civil servants to better serve localneeds. In India, Joshi (2000) found that front-lineforesters and the foresters union helped establish andspread joint forest management arrangements.

It is important to note that empowering localadministrative branches of the central state—administrative units and line ministry local officesand agents—is not the only way to create effectivedeconcentration. As noted in the above section, theway the electoral system is designed matters; someelectoral systems can establish downwardly account-able local authorities, while others force them to beaccountable to the central state or to a politicalparty—what Resosudarmo (2002) calls“coadministration.” Second, the accountabilty of localauthorities often follows the kind and source of theirpowers. If elected authorities are given only the powerto implement central mandates, then they areeffectively turned into elected central administra-tors—not local representatives. In colonial times,customary authorities—such as chiefs and head-men—were officially administrators for the colonialpowers and they were backed by colonial militaryforce. Coadministration in the name ofcomanagement can also be a form of central adminis-tration of local affairs. So while administrativeauthorities can be accountable to local populations,many other kinds of authorities can be made up-wardly accountable to the central state and, therefore,can be used as administrative units rather than asindependent local decentralized political authorities.

BOX 7 ELECTED AUTHORITIES LOSELEGITIMACY WHEN THEYCAN’T DELIVER

Ghana decentralized powers to elected districtassemblies with the expectation that doing sowould liberate people. Assemblies were createdbut given very little power. The money that theyhad was spent to transport politicians to andfrom meetings. People were excited about thepromise of democratic opportunities, but theoutcomes did not meet their expectations andthey became cynical and angry.

In one case in Uganda, decentralization led tobroad participation in the beginning. However,members of the old village elite learned to assertthemselves in local council affairs and startedappropriating resources. Local people feltexcluded and became angry.

In Senegal, elected rural councilors in the 1980sfelt that nobody respected them because they hadno significant powers. They were embarrassed tobe elected officials. They could do nothing forthe people. Since the mid-1990s their powersunder the law have increased—but they are notgetting to exercise them in practice.

When leadership has nothing to offer, peoplelose interest. When the leadership is nottransparent, people lose trust. The resultingcynicism can poison relations between localgovernment and the local population, but it alsocan prompt change.

Sources: Hesseling and Smit n.d.; Bazaara and Manor inLatif 2002, 32–3, 45.

31


Authorities chosen through competitive local

elections that admit independent candidates

appear most likely to be locally accountable; in

addition, they may be more sustainable and

scaleable across national territories.

In sum, deconcentration is a weak form of decen-tralization. Local administrative authorities can beaccountable to local people when faced with properincentives or when ideologically dedicated to theirrole as public servants. Their downward accountabil-ity, however, is usually less-systematically structuredthan is that of elected authorities.

CUSTOMARY AUTHORITY

While chiefs clearly must not be ignored in environ-mental management, their representation asuncontentious “custodians” of natural resources seemsunwise.

Porter and Young 1998, 523

Across Africa, so-called traditional or customaryauthorities, such as chiefs, big men, kings,marabouts, and other religious leaders, have beenreasserting their authority in recent years (vanRouveroy van Nieuwaal 1987 in Therkildsen 1993,84; Brock and Coulibaly 1999, 152; Ntsebeza 1999,2002; Manor 2000).47 Traditional authorities mayalso be gaining strength elsewhere—as in Guatemala,Indonesia or in China’s ethnic regions. They play animportant role in many decentralization programs. Attimes, chiefs are evoked as a reason not to decentral-ize. Sometimes they are identified as the actors towhom decentralized powers should be given. Chiefsfind support in international donor agencies as wellas in national governments. But are they representa-tive, accountable, and legitimate? Are they appropri-ate institutions to conduct local public affairs?

In the past several years, customary authorities,with help of their allies in government, and oftenwith donor support, are reemerging as a political

force against local democratization in Mozambique,South Africa, Uganda, and Zimbabwe (Ntsebeza1999; Jeter 2000, A1; Manor 2000; Bazaara 2003).Muhereza (2001a) has pointed out that “decentraliz-ing” control over forests to customary authorities inUganda contributes to the strengthening of custom-ary kingdoms at the expense of the democraticallyelected local council system (see Box 8). In Mali,donors are reconstituting customary authorities asforest managers at the expense of democraticallyelected local councils (Kassibo 2002a). Kassibo(2001) has argued that traditional authorities are alsoreemerging in reaction to local democracy and to thewoman’s movement in Mali. Chiefs in Burkina Faso,Mali, and Niger are also often evoked by members ofcentral government as a threat to democratic localinstitutions, or as a reason not to decentralize orestablish democratic local institutions. Chiefs arethreatened by transfer of control over land allocationto democratic institutions and the more generalundermining of their authority by alternative repre-sentative local institutions. Because of this, they andtheir allies in central government—who may also losetheir rural power base through local democratiza-tion—pose a serious threat to decentralizations.

Van Rouveroy van Nieuwaal and van Dijk (1999, 6)have argued that land and natural resource manage-ment are being renewed as arenas for chiefly power.NGOs “appear to have turned chiefly office into anarena of brokerage, thus opening new perspectivesand avenues for entrepreneurial activity.” Naturalresources and land allocation are described asdomains in which chiefs’ “nostalgic claims to authen-tic ritual power are effectuated in terms of realpolitical power. . . . In most cases chiefs succeed ininvoking ritual rights from the ‘past’, which they thentranslate into instruments for ‘hard’ political broker-age. Chiefs negotiate their positions in the context ofglobal discourse on sustainability, environmentalawareness and national and international interest inecological preservation.” In essence, chiefs use thisdiscourse of authenticity to their advantage—particu-larly with respect to gaining leverage within naturalresource management and land allocation.

There are plenty of examples of chiefly influence innatural resource management. Oyono (2002, 12)

32


BOX 8 KING VS. COUNCIL IN UGANDA—WHO BETTER MANAGES THE FORESTS?

Uganda’s Forest Service transferred several foreststo the Bunyoro-Kitara Kingdom in the district ofMasindi. Following the transfer, these forests werepoorly managed. The management trust set up bythe king was accused of mismanagement forallowing increased pit-sawing. The district forestofficer, a centrally appointed official, observed:

The kingdom was selling trees like cows. They soldstanding trees without undertaking an inventory toestablish the volume of wood. This had partlycontributed to the current over-exploitation of trees inkingdom forests. The kingdom officials refused toallow field extension staff to access their forests andeven issued their own licenses for harvested timber,which created a lot of confusion in the department.

In addition, forest villages expressed resentment fornot being consulted on the trust’s establishment oron the formulation of policies—in which they wouldhave been involved under district jurisdiction.Villagers showed their opposition and disapproval byignoring the trust’s regulations and by setting forestfires to burn trees. In an attempt to redress thesituation, the king cancelled all licenses andconcessions and fined license and concessionholders who did not stop production. The situationremains unresolved. It is interesting to note,nevertheless, that villagers were able to use sabotageto bring the king to be more accountable to them.

In stark contrast, the neighbouring forests are bettermanaged under local-council management. Underthe pre-decentralization forestry regime in MasindiDistrict, the licensing fee for charcoal producers wasvery high (over sixty percent of the producer price ofcharcoal). Charcoal makers cut more trees tocompensate for the high tax. Truckers, however,paid a very low licensing fee (only eleven percent ofthe sale price for charcoal). Because of this inequity

between producers and truckers, “many charcoalproducers grew hostile toward the local and centralgovernment authorities [who set the fees], in somecases refusing to pay licensing fees and failing tocooperate with government in other areas”(Muhereza 2003, 8). Because of these problems, thedistrict received less than thirty percent of therevenues expected from charcoal fees. More treeswere being cut, and the district was losing almostU.S.$1.4 million per year.

To resolve these problems and increase the revenues,the district executive committee resolved to drasticallyreduce the fee for producers to about one percent ofits original level, and the fee for truckers wasincreased by forty percent. To further raise revenues,the committee formulated and facilitated the passingof the Masindi District Production and EnvironmentOrdinance of 2002. The ordinance requireslandowners to set aside at least ten percent of theirland for tree planting and enables the council topenalize delinquent landowners. The ordinance alsoempowers the district council to make bylaws toensure sustainable use and revenue collection fromthe charcoal trade. Together, these measures increaseddistrict revenues while improving forest management.Income from charcoal production revenue in MasindiDistrict is used by subcounty councils to subsidizetheir struggle to deliver services.

Unfortunately, although elected councils appear tomanage the resources better and to invest the incomein the community, councils control a small portion ofthe forests in Masindi District. Most of the forestshave been privatized and are now controlled bycustomary authorities, or they have been cordonedoff as central reserves. The domain over whichcouncils have jurisdiction has been severely limited,undermining the discretion and ultimately thelegitimacy of local councils.

33


observed that forestry decentralization in Cameroonis allowing village chiefs and management committeemembers to collude in establishing themselves as anew “forestry elite.” In Zambia and Lesoto,Shackelton and Campbell (2001, 8–9) noted thatchiefs exerted disproportional influence as wildlifemanagement unit chairs, which allowed them to“divert some benefits from CBNRM [community-based natural resource management] to build theirown power base.” In Namibia, however, their exclu-sion resulted in conflict and delays. The chiefs werelater coopted into the committees. Shackelton andCampbell (2001, 9) feel that the key to success is toachieve a balance with respect to the power of chiefsin CBNRM by providing traditional leaders with ex-officio or non-executive roles. Oyono (2003) believesthat customary authorities need to be included innatural resource management committees in order tohelp articulate the actions of these new managementinstitutions with pre-existing institutional relations,but he cautions that their inclusion must be subordi-nated to decision making by elected and representa-tive local authorities.

Customary authorities can at times be a positiveforce. Manor (in Latif 2002, 41) points out that inMozambique some chiefs are relatively progressive,and since civil society is weak, chiefs are the onlylocal people with whom government and outsiderscan work (also see Alexandre 1970). Under theseconditions, even the most progressive organizationswork with chiefs if they want to get anything done.Spierenburg (1995) describes highly democratic,locally accountable traditional decision-makingsystems in Zimbabwe. As mentioned earlier, there aremany possible kinds of accountability systems builtinto chieftaincies or councils of elders. As Manin,Przeworski, and Stokes (1999) point out, evenwithout any accountability mechanisms in place,some authorities will be responsive, and thereforerepresentative, due to their ideology. They maybelieve their role is one of public service and loyalrepresentation. But, without a systematic form ofaccountability, the responsiveness of customaryauthorities is more likely to be based on the personal-ity and beliefs of individual leaders.

Under some conditions, customary authorities canpose a threat to the development of local democraticgovernment and to effective decentralization. InSenegal, village chiefs are being used as intermediar-ies between elected rural councilors and villagers as ifthey represent the village48 (Thiaw and Ribot 2003).In this manner, villagers are being disenfranchisedand separated from the democratic process—espe-cially because many village chiefs are making deci-sions that are against the interests of the villagers(Ribot 2000). In the name of legitimacy and tradi-tion, they take powers that would otherwise go todemocratic authorities. Ironically, they are notnecessarily popular public figures, as is often as-sumed. Traditional political authorities have oftenbeen viewed as the extended arm of the state in thelocality, and usually are regarded as inefficient,corrupt, undemocratic, and excluding of women(Therkildsen 1993, 87). Perhaps customary authori-ties or customary councils are better viewed as privateinterests, just another elite group, or as members ofcivil society—a kind of NGO or interest group. Likeany other actors, some chiefs are dedicated to publicservice and some are not.

Whether a chieftaincy accurately represents peopleconcerning public natural resources will depend onthe history and personality of the particular personand place. Of course the same is true of even well-crafted electoral systems; however, in electoralsystems there are always sanctions that can beapplied. The lack of systematic accountability acrosschieftaincy systems makes it difficult to rely on chiefsand other customary authorities as a generalizedform of representation. While some are definitelydemocratic and some are highly accountable, othersare not. Further, customary rulers are often chosenbecause someone believes them to be legitimate. Thistoo may not be the case, or their legitimacy may comefrom fear rather than from respect as in one case inSouth Africa (Ntsebeza 2004).49 Legitimacy may bean instrumental reason for choosing a local authority,but it is not necessarily a democratic one. In addition,chiefs are receiving public powers that could go tolocal elected authorities, thereby undermining theestablishment of local democracy.

34


In short, traditional authorities are not a primordialcategory of ideal legitimate leaders. Closer scrutiny oftraditional authorities shows that: they are notnecessarily representative, legitimate, or even liked bylocal populations;50 they are often constructions ofthe central state and in many places are still officialadministrative auxiliaries of central authorities (asunder Indirect Rule); they are not necessarily account-able to the local population; and empowering orworking with them may not serve the efficiency,equity, or development aims of many decentralizationadvocates. So-called customary or traditional authori-ties are sometimes representative and locally account-able, sometimes administrative, and sometimesprivate actors. Perhaps, as in the 1993 Local Govern-ment Transitional Act in South Africa, local chiefsshould be defined and treated as just another “inter-est group” (Ntsebeza 2004). The key question, whoseanswer will differ from place to place, is how todecentralize so as to draw on the strengths andwisdom of traditional authorities while reinforcingand legitimizing local democracy.

NGOS AND COMMUNITY GROUPS

The danger is that a small group, probably urbanelites of local origin, will use their high-level contactsto have a community forest approved in the name ofthe whole community but through an institutionwhich actually represents only a small part of it.

Brown 1999, 47

According to a definition provided by the World Bank,nongovernmental organizations (NGOs) are “privateorganizations that pursue activities to relieve suffer-ing, promote the interests of the poor, protect theenvironment, provide basic social services, or under-take community development.” The defining charac-teristics of an NGO are “altruism and voluntarism”(World Bank 2003).51 NGOs are valued by differentpeople for different reasons. Conyers (1990, 25)points out that some NGO supporters “are motivatednot by ‘developmental’ concerns but by the desire toreduce public spending and support capitalistenterprise.”52 Others believe “that NGOs, andespecially local people’s organizations, can oftenprovide a more participatory and responsive basis for

development than any bureaucracy can ever hope todo. . . . Many supporters of the latter view envisagesuch organizations working in partnership withdecentralized government organizations, not replac-ing them” (Conyers 1990, 25).

The key question, whose answer will differ

from place to place, is how to decentralize so

as to draw on the strengths and wisdom of

traditional authorities while reinforcing and

legitimating local democracy.

NGOs can play a significant positive role in decentrali-zations by making people aware of their rights as votersand by monitoring elections (Smoke 1999, 11). Theyalso can provide services to local populations and canorganize to lobby government to provide greater servicesto local populations. In Cameroon, for example, NGOshave helped rural communities develop complex forestmanagement plans required by their government(Etoungou 2003; Oyono 2003). NGOs can act to achievecollective ends; but, as Kassimir (2001) cautions, theyare themselves “fields of power and authority” and notmerely collective actors supporting democracy (see alsoK. von Benda-Beckmann 1981; Fox 1990; Ferguson1994; Lund 1998).

NGOs can also have negative effects on local democ-racy and therefore on the foundations for effectivedemocratic decentralization. Across Asia, NGOs havebeen observed trying “to undermine the developmentof local governments, which are seen as a threat totheir powerful position in the community” (Smoke1999, 11). International NGOs, donors, and other suchorganizations often choose to work with local NGOsand community groups in order to avoid local govern-ments. This takes from local representative govern-ments valuable opportunities that could support theirdevelopment and legitimacy. In addition, NGOs canundermine grassroots organizations by dominatinglocal agendas. In a case in eastern Cameroon, the vicepresident of a village development committee ex-plained to Etoungou (2001) that an NGO had told hisvillage that it had come to follow the village’s case. The

35


vice president had replied that while “we must saythanks to this NGO, we hope that it neverthelessunderstand that we can fly with our own wings.”

Many projects also approach sub-village interest-defined groups—such as fishers, woodcutters,pastoralists, farmers, or women’s or youth groupsorganized into associations, NGOs, or cooperatives.These groups do not necessarily reflect the concernsof a village as a whole—particularly in mattersconcerning public resources such as forests, streams,pastures, or public works. While they are oftentreated as if they were representative, they are not.They represent their particular interests, and theirrepresentatives or leaders are accountable to theirparticular constituencies—and often only to them-selves. There is no systematic basis for them to speakon behalf of the community as a whole. Spokesper-sons for various local movements or organizationsare often self-appointed or sponsored by outside aidagencies or NGOs and are, consequently, nonrepre-sentative (National Research Council 1992, 35; Guyer1994, 223; Mazonde 1996, 56). As Utting (1999, 170)points out, we need to ask to whom NGOs areaccountable, if indeed they are accountable to anyone.

In Senegal, for example, forestry cooperativepresidents—usually powerful notables—treat theircooperatives as private property, often filling themwith family or dummy members to obtain stateservices to which cooperatives are entitled (Cruise-O’Brien 1975, 128; Ribot 1993). In Cameroon, thecommunity forestry law devolves powers to localforest management committees (Government ofCameroon 1998; Brown 1999, 47). To be recognizedofficially, these committees must be registered by theadministration as “common initiative groups” or“associations” (Oyono 2003, 17). They do not neces-sarily represent the community, however. While thelaw requires that the committees consult representa-tives of all segments of the community, it is not clearby whom these “representatives” are chosen, and theresults of the consultation are not binding. Theconsultation is just another process—a kind of ritualto appease the powers that be—that must take placein setting up a community forest (Government ofCameroon 1998, 12). In Mongolia, in reaction to poorpastureland-management laws and circumstances,

civil-society groups are mobilizing to experiment withgroup pasture management. There remain manyquestions, however, concerning these groups’representativeness and their legitimacy among non-group members53 (Mearns 2002, 10–1; 2004).

Governments too are giving NGOs official roles in“representing” local populations. In Uganda’sdecentralization laws, NGOs have been given anofficial role on the newly established district and localenvironmental committees (Cappon and Lind 2000).By over-representing NGOs in public decisions, thiskind of legislation conflates the importance of havinga plurality of voices within a society with the idea ofrepresentative democracy. There are reasons toreserve seats on councils for women, as is done inUganda and in India, who are roughly half of anygiven population and who are systematically excludedfrom elected positions in many places (Vijayalakshmiand Chandrashekar 2002; Bazaara 2003).54 Thesemechanisms help ensure more even gender represen-tation. There are also good arguments for systematicmechanisms to include poor and marginalizedpopulations (see, for example, Crook and Sverrison2001). While some NGOs and other interest groupsmay be composed of marginal actors, many areconstituted of segments of the population that areneither marginalized nor under-represented, andwhether or not marginal, they are unlikely to berepresentative.

International or national NGOs not only createproblems for local governments when they bypassthem in favor of working with local NGOs andcommunity groups, but they sometimes also createproblems when they choose to work with localgovernments. One problem is that NGOs sometimestry to push representative authorities to pursue theNGOs’ agendas. In Nicaragua, NGOs not only askedlocal governments to make decisions on issues thatlocal governments were not interested in, but theyeven asked local governments to take actions thatwere not legal (Larson 2002, 15). NGOs also causedproblems for local governments in Nicaragua whenthey did work in the name of local government butlocal government could not follow through becausethe NGOs did not provide for a way for the work tocontinue after their project ended. However, despite

36


the difficulties that NGOs created for local govern-ments in Nicaragua, Larson (in Latif 2002, 40)observed that environmental outcomes were betterwhere there were more NGOs in place.

In short, NGOs and community groups certainlycan have productive roles in development. It isimportant, however, to ensure that these roles do notcompete with or undermine those of democratic localauthorities.

CROSSCUTTING INSTITUTIONALISSUES

There are several crosscutting issues that are impor-tant to consider when determining which institutionsshould receive decentralized powers. This subsectionexamines

• the problems associated with working with single-purpose as opposed to multi-purpose local organi-zations,

• the factors that shape equity and inclusion ofmarginal and poor groups,

• the problems of elite capture and patronage,

• the role of a strong central state, and

• the accountability of the central state and donors tolocal government.

Single-purpose vs. multi-purpose bodies“Participatory” or “decentralized” natural resourcemanagement activities are often organized throughmanagement committees (Conyers 1990, 23–6;Therkildsen 1993, 75; Engberg-Pedersen 1995, 4;Ribot 1995a; Alden Wily 2000a; Etoungou 2001;Muhereza 2001a; Baviskar 2002; Conyers 2002;Cousins and Kepe 2002; Namara and Nsabagasani2003; Oyono 2003). These committees are usuallyoriented toward a single objective, such as naturalresource management (Manor 2004). They may beelected, self-selected, or appointed. They may betemporary or registered as an NGO or cooperative.They are typically under the close supervision ortutelle of a technical agency (such as the forestry

ministry), local administrators, donor programs, orNGOs.

Manor (2004) suggests that earlier decentraliza-tions in the mid-1980s and 1990s were more devel-oping-country driven and were characterized by aleaning toward multi-sector local institutions. Themore recent decentralizations, he suggests, are moretechnocratically driven, and lean toward single-purpose bodies. This appears to be a global change.Manor (in Latif 2002, 74; also see Manor 2004)observes that local bureaucrats like multi-purposecouncils because they encourage cross-sectoralcoordination. Manor (in Latif 2002, 74) gives theexample of a minister in Bangladesh who insistedthat fisheries officials and other relevant ministriesget involved in a fisheries project to help explore themulti-dimensional problems and achieve bettercoordination. Technocrats—including environmental-ists—however, seem to gravitate toward single-purpose committees in order to achieve their specificaims (Manor and Brannstrom in Latif 2002, 74; alsosee Manor 2004 and Brannstrom 2004) (see Box 9).

According to Manor (2004), donors, too, favorsingle-purpose committees because they give theappearance of local consultation while liberating thedonor’s project from micro-management by govern-ment agencies. Such committees are often excessivelyfunded (with mandates that are too ambitious forthese committees to carry out), while general-purposelocal councils are starved of funds. Dupar (in Latif2002) also observed this phenomenon in AsianDevelopment Bank-funded river-basin committees inareas where subdistrict governments have fewresources. This over-funding undermines the legiti-macy of other local authorities. The same phenom-enon of over-funding is seen with NGOs.

Government and NGO implementing agencieschoose to work with management committees inMadhya Pradesh, India, even though panchayats(elected local governments) already exist. Baviskar(2002, 4) points out that this choice is ostensibly madeby agencies because officials “perceive tensionsbetween the idea of democratic decentralization as aprocess that could become ‘too political’ and thewatershed mission as a project that needed to be kept

37


away from politics. . . . Anxiety about keeping themission ‘free from politics’ is used by state officials tojustify the political move of separating the watershedcommittee from the panchayat.” But free from politics(which is impossible) would also mean free from thepolitical debates that are the essence of democracy(also see Ferguson 1994). In a forest managementproject in Guinea, the U.S. Agency for InternationalDevelopment also chose to avoid local democraticinstitutions in favor of project-created committees. Thereason the agency gave was that the local councilorswere inefficient and slow (Ribot 1998a).

Manor (in Latif 2002, 19; Manor 2004) points outthat single-purpose committees usually are notelected but rather are appointed by governmentofficials. To appear representative, a single-purposecommittee often includes a village chief or a memberof an elected local council.55 Even assuming that thesemembers do represent their communities, however,these members are usually only one person amongmany members on the committees and do not have adeciding vote (Ribot 1996, 1999). For example, Mali’slocal government representative sits on a committeeto allocate quotas for access to commercial forestryresources. The representative, however, is only oneamong four members, and he is not guaranteed acontrolling role. The real decisions about quotaallocation are made by the regional governor, who isan appointee of central government (République duMali 1994, art. 6, 11). The system is very similar inBurkina Faso (Delnooz 1999; Ribot 1999a). Inclusionof a single local representative does not make acommittee representative or accountable to thepopulation. In Uganda, community-level committeesare occupied by a small elite, also bringing intoquestion who they represent (see Box 10).56 Further,many of these committees are presided over by thetechnical service agent who mobilized them; there-fore, even if they are representative, they are usuallyonly advisory in nature (Ribot 1999a).

Single-purpose committees appear to be less ac-countable to local populations, more instrumental, andmore prone to elite capture than are multi-purposeinstitutions (Working Group 3 of the Bellagio Confer-ence, Latif 2002, 47). Even when elected, however,single-purpose user committees are not necessarily

democratic or equitable. In a case from Uganda, abeekeepers committee was elected, but its mandatewas so narrow that only interested parties participatedand these beekeepers excluded others from forest use.This divided the community. There was no recoursesince the committee was established by a one-timeelection. Further, the committee’s powers and obliga-tions were spelled out by the wildlife authority thatcreated the committee. Therefore, the committeemembers were upwardly accountable to the wildlifeauthority, not to the population in the village. (SeeNamara and Nsabagasani 2003.) Bazaara (in Latif2002, 75) adds that one reason that single-purposecommittees are less accountable to the community andmore likely to produce conflict is that their membersdo not allow others outside of the committee to have

BOX 9 MANUFACTURING SUCCESS

In Madhya Pradesh, India, a “decentralized”watershed management project avoided existingdemocratic local institutions, set up electedcommittees of its own, and devolved powers tothese committees. The project then circumscribedthe committee’s powers with guidelines, norms,and overbearing supervision. The project,structured in this way, channeled resources to thebetter-off farmers in a number of ways. Becausethe project needed to show success, it focused onthe farms that were most likely to succeed. Also,because those with the skills to participate indecision making and project activities wereeconomically best placed, they were more able totake advantage of the project. “Left out from thisprocess are other, less favorably situated villagers,who cannot create networks of mutually beneficialtransactions with state officials,” observesBaviskar. “These include those adivasis [scheduledcastes], whose membership in politicalorganizations demanding control over naturalresources such as forests and land threatens toundermine state control.”

Source: Baviskar 2002, 8–9.

38


access to the information they control. This partlydefeats the objective of democratic decentralization.

Single-purpose committees can, however, bedownwardly accountable (Murombedzi in Latif 2002;Conyers 2002). They can be fairly elected or ap-pointed by a locally elected body. Manor (in Latif2002; also see Larson 2002) argues that the munici-pal forestry units in Bolivia have advantages over usercommittees that are accountable to ministries andforest services because the municipal forestry unitsare constituted by and report to local elected govern-ments. Manor’s point is similar to Blair’s (2000)

argument that committees and administrative bodiesshould be accountable to elected bodies.

Single-purpose committees can also serve theinterests of marginal populations. In the village ofMoangué-Le Bosquet in Cameroon, the democraticprocess behind the establishment of single-purposeforest-management committees has allowed themarginalized Baka to create a niche for themselves(Oyono 2002). Nevertheless, multi-purpose councilsmay have some advantages over single-purposecommittees. Manor (in Latif 2002, 34) asserts thatmulti-purpose councils (which he equates withdemocratic decentralization) usually increase popularparticipation, legitimacy, accountability, and transpar-ency. They are often more equitable and tend to havespecial provisions for embracing marginalized groups.

The wisdom on this matter indicates that the bestarrangement for single-purpose committees is to beaccountable to a multi-purpose democraticallyestablished authority—whether they are popularlyelected or appointed by that body.

Privileging marginal and poor groupsHow can development programs reach poor andmarginal groups to include them in decision making?Poverty mapping can be used to identify poor com-munities (Henninger and Snel 2002), but identifyingpoor within communities is a more difficult task.Central targeting systems are better at the prior thanthe latter (Mansuri and Rao 2003, 12). In Peru, forexample, Paxson and Schady (2002 in Mansuri andRao 2003, 14) found that funds that were distributedusing geographic targeting methods typically reachedthe poorest communities but had trouble reachingthe poorest households within those communities. Amethod called community-based targeting (CBT),which relies on information provided by communi-ties themselves, can be used to identify the poorwithin communities. Studies have shown that CBTproduces better information than surveys andquestionnaires administered by outside organizations(Mansuri and Rao 2003, 12).

BOX 10 ELITE PARTICIPATION WITHCOMMITTEE PROLIFERATION

In recent years, local committees have become verypopular in development circles. Although they maybe useful for capacity building and for achievingspecific outcomes, they can also have the negativeeffect of competing with democratic institutionswithout promoting democratic or inclusiveprocesses. In Uganda, a class of local elitedeveloped whose role was to participate in thevarious committees being set up by outside donors,projects, and by local government. These“professional” committee members emerged fromthe proliferation of local committees and thedemand for “participation” by the many single-purpose programs being established in the areaaround Mabira forest. The situation was convenientfor projects that wanted to impress donors bytrotting out their highly skilled committeemembers. This new cadre of committee membersalso gained significant management andnegotiation skills. However, it also created anappearance of participation while involving only asmall number of people. In addition, theproliferation of local institutions diffused authorityamong too many local institutions, taking powersaway from democratic bodies.

Sources: Namara and Nsabagasani 2002; Manor in Latif2002; Manor 2004.

39


CBT was found to be less effective in more stratifiedcommunities, in communities that lacked schools,and in communities with low aggregate funding(Galasso and Ravallion 2001 in Mansuri and Rao2003, 12). Other factors that influence the effective-ness of community-based targeting include thefollowing: (1) How effectively a given community canmobilize information and monitor disbursement.These tasks can be costly and can furnish opportuni-ties for elite capture or corruption. (2) How interestedthe community is in equity. Do they want to help thepoor? (3) How heterogeneous the community is. CBTwill be less effective in communities where multipleand conflicting identities present competing incen-tives. (Conning and Kevane 2002 in Mansuri and Rao2003, 13.)

Many factors conspire against including the poorand their concerns in local decision making.Brannstrom (in Latif 2002, 32) points out thatincreasing the participation of one group may crowdout other groups (also see Cooke and Kothari 2001,8). For example, the participants who join a project inits early stages may have more voice and opportuni-ties than those joining in later stages, and the pooroften arrive late if at all. Bazaara (in Latif 2002, 33)presents an example of a decentralization reformaround parks in Uganda in which some local groupshave been helped but others have not. In the decen-tralization reform in Madhya Pradesh, India, the needto produce successful outcomes led the project toinvest in better-off members of the community eventhough the project was intended for the poor (see Box9, above) (Baviskar 2002; also see Mansuri and Rao2003, 27–8).

Another dynamic marginalizing poor groups is thatthey have little ability to express themselves or tosanction decision makers. Larson (2002, 16) arguesthat stakeholder approaches that bring everyone tothe same table are not sufficient because some actorsare not willing or able to engage in the process (seeFigure 3). Marginal groups must hold real power ifnegotiations and their outcomes are to be moreequitable. Therefore, this form of participation is notenough. Some measures to correct power imbalancesare also needed. Society is highly differentiated andaccess to accountability measures is also unequally

distributed. Marginal groups are less able to takeadvantage of accountability mechanisms than moreprivileged groups. Because of this, mechanisms areneeded that privilege marginalized or disadvantagedgroups if they are ever to be part of the politicalprocess (Young 2000, 8).

Decentralizations in Bolivia, Cameroon, Nicaragua,and Zimbabwe have led to greater inclusion of somemarginal populations in forestry decisions (Conyers2002; Larson 2002; Mapedza and Mandondo 2002;Oyono 2002; Pacheco 2002). The outcomes, how-ever, are mixed. In Bolivia, simultaneously strength-ening municipalities while decentralizing forestsector powers “has created new opportunities forindigenous people, small farmers, and small-scaletimber producers by increasing their bargainingpower with other social actors” (Pacheco 2002, 10).The reform also has weakened outside concessionholders and large ranchers. However, in addition toassisting marginal groups, the reform has reinforcedthe power of local elite merchants, professionals,ranchers, and sawmill operators, who continue todominate municipalities (Pacheco 2002, 10). Indig-enous peoples have consolidated their land claims inmunicipalities where their candidates have wonelections, but they have been marginalized wheretheir candidates have not won. Further, the financial,administrative, and technical obstacles to entry into

FIGURE 3 THE SILENT MAJORITY

Source: © The New Yorker Collection 2001 Charles Barsottifrom cartoonbank.com. All Rights Reserved.

Any other objections?

40


timber exploitation are leading indigenous groups topartner with outsiders (see Box 11). This may not beso unlike the experience in Nicaragua and Cameroonwhere insiders become intermediaries or stand-insfor outside operators (see Bigombe Logo 2001;

Larson 2002; Oyono 2002). Pacheco concludes that“in most cases, previously marginalized groups canbenefit from forest resources to improve theirlivelihoods, but they are constrained by limited accessto productive assets and markets” (2002, 14).

BOX 11 TIMBER CONCESSIONS VS. LOCAL MANAGEMENT AND MATTERS OF SCALE

Timber concessions are a long-standing means forenclosing forested areas for particular uses andusers—usually outsiders. Colonial administrators inthe tropical hardwood forests of the southern coastof West Africa cornered the lucrative timber marketsfor their European clients.

The colonial historian R. L. Buell reported that“before 1924, natives held [forest] concessions andsold wood upon the same basis as Europeans. Butthe competition became so keen and native cuttingso difficult to control that in a 1924 administrativeorder, the government declared that a native couldnot cut and sell wood except for his own use withoutmaking a deposit with the government of twenty-fivehundred francs—a prohibitive sum.” He goes on tosay that the same system of concessions is used inthe French Congo, resulting in “the locking up ofthe resources of the territory in the hands of a dozenlarge companies” (Buell 1928, vol. II: 256). In 1930one such company, the Compagnie ForestièreÉquatoriale, held 100,000 hectares of leaseconcessions (Hailey 1938, 1005).

In addition to the unfair allocation of concessions,large fees, which large companies could afford andrural dwellers could not, also were used to preventrural dwellers from competing with largecommercial users. Within the colonies, nocommercial exploitation permits were free (GGAOF1935, art. 79). Concession licenses called temporarypermits—lasting twenty-five years, but in specialcases up to fifty years—could be allocated after thesale of the plots by public auction. A minimum saleprice would be fixed by the lieutenant-governor andthe sale would go to the highest bidder. Lieutenant-governors and ministers of the colonies held the

power to allocate permits and to set the fees forexploitation rights. Given the scale and duration ofthe concessions, these prices were beyond the meansof local populations and most African merchants.

Large-scale timber concessions today are justified onthe grounds that they allow for economies of scale inplanning, management, monitoring, and control.But does this justification hold? First, there is noreason that small-scale operations cannot abide bythe same kind of guidelines as large-scale operationsif the planning is done by local districts (the localgovernment) rather than by the woodcuttersthemselves. In this manner, areas where extractionis permitted can be designated, and permits for aslittle as one tree stem can be issued. Remote sensingtechnologies can also help to monitor cutting outsideof designated areas. Second, there may be alternativemeans of control that are not territorial in nature; forexample, monitoring to check the origin of harvestedtimber and to ensure that woodcutters have theappropriate papers can occur on transport roads andat sawmills and ports. Third, the rules of extractionset out in national laws rarely are followed by large-scale operators; therefore, it cannot be argued that itis more efficient or effective to control at that level.

In Bolivia, eighty percent of the forest is allocated bycentral government rather than by municipalgovernment. Does allocation—which is primarily apolitical activity, not a technical one—really need tooccur at the central level? Can allocation ofconcessions be devolved with positive social andenvironmental consequences?

Sources: Buell 1928; GGAOF 1935; Hailey 1938; Larson inLatif 2002, 28. Also see Pacheco 2002.

41


In some cases, marginal groups have acquired powerthrough local electoral processes. In Cameroon, whenlocal communities were allowed to create communityforests, the Baka acquired the right to manage a forestecosystem in the village of Moangué-Le Bosquet(Oyono 2004a). Larson (2002, 13) describes twoinstances in which marginal populations in Nicaraguahave been able to make local government work forthem. In one area local indigenous groups managed toput their concerns on the municipal agenda throughparticipation in the traditional political system. Inanother, the mestizo (mixed descent) populationmanaged to elect a mayor who took the side of peasantcolonists who were invading indigenous territories.The latter case is another example of a mixed out-come––while the reform benefited one marginalgroup, it allowed that group to further marginalizeanother (also see Olowu 2001, 54).

In any decentralization process, there is always therisk that some groups and individuals will be excluded.Baviskar (in Latif 2002) points out that decentralizedmanagement carries the risk of creating “new enclo-sures” by excluding non-local actors such as migrantsand pastoralists from the power transfer. The exampleabove of the beekeepers committee in Uganda is a casein point (Bazaara in Latif 2002, 15; also see Namaraand Nsabagasani 2003). People excluded from theparticipatory comanagement scheme later came tosabotage the resources. In Bolivia, Pacheco (in Latif2002, 18) reports that enforcement against illegalpractices has been biased against non-resident smallfarmers. These examples suggest that additionalmeasures and institutions are needed to ensureinclusion of the most vulnerable groups.

Manor (in Latif 2002, 54) points out that centralgovernments tend to be more generous toward thepoor than are local governments. Local elite are moreprejudiced against the poor than actors at higherlevels. Crook and Sverrisson (2001) point out that forserving the poor, “the most successful cases were theones where central government not only had anideological commitment to pro-poor policies, but wasprepared to engage actively with local politics. . . , tochallenge local elite resistance if necessary and toensure implementation of policies” (2001, 48). Manyof the other accountability mechanisms mentioned

above can also help to include marginal groups inpublic decisions.

In sum, if decentralization is to include marginalpopulations in decision making, additional effortsbeyond standard electoral representation will berequired. Poverty mapping is one important tool(Henninger and Snel 2002) and CBT is another(Mansuri and Rao 2003). For inclusion of women,measures might include reserving seats on electedcouncils, as in India and Uganda (see Vijayalakshmiand Chandrashekar 2002; Bazaara 2003). For othermarginalized groups, these might include participatoryprocesses that solicit the needs of marginal popula-tions, and earmarked funds and projects targeted atserving such needs (Galasso and Ravallion 2001;Alderman 2002; and Conning and Kevane 2002 inMansuri and Rao 2003). It is important to note,however, that such participatory processes should beused as a complement to and not a substitute forrepresentation—they are a tool that representativeauthorities can use to engage and become informedabout marginal groups. Processes that privilege theinclusion of these populations must be developed andimplemented (see Crook and Sverrisson 2001, 4). Inaddition, central governments can promote local equityby issuing mandates that require local communities toinclude marginal groups in decision making.

Elite capture and patronageIn many instances, it is local elite rather than themost vulnerable that capture decentralised power—which is then utilised to repress local minorities—including women and other marginal groups.

Dele Olowu 2001, 54

The objective of decentralizations is to enable allmembers of society to participate in decision making,but all too often members of elite groups manage toderive most of the benefits of decentralizationreforms (Smith 1985; Crook and Manor 1998; Utting1999, 174). According to the World Bank (2000,109), “conceding power to local governments is noguarantee that all local interest groups will berepresented in local politics. It may simply mean thatpower is transferred from national to local elites.”

42


Elite groups were observed to have captured thebenefits of decentralization efforts in Bolivia, Brazil,Cameroon, India, Indonesia, Mali, Mexico, Nicaragua,Senegal, Uganda, and Zimbabwe.57 In Uganda it isexpensive to run for election under the current system;therefore, leaders are members of elite groups (Saito2000, 11). In Cameroon, members of the elite havebeen able to capture forest resources under newcommunity-based forestry laws by fronting fictivecommunity forests for their own benefit or preemptingthe establishment of community forests by acquiringmore-lucrative commercial licenses (vents de coupe) intheir place (Brown 1999, 46).58 Oyono (2002) reportsthat in Cameroon chiefs and management committeemembers are colluding to establish themselves as anew “forestry elite.” In both Indonesia and Cameroon,local people are being used as proxies for outsidecommercial interests to gain access to timber (Oyono2002; Resosudarmo 2002). Similarly, members of thelocal elite in Senegal captured control of forestrycooperatives in the 1970s and ’80s by registering themin their names and filling them with fictive members(Ribot 1993). It is no surprise that Conyers (2000a, 8)concludes that decentralized decision making mayonly benefit a minority of the population.

But why does elite capture occur? Why do centralgovernments and outside organizations alike so oftenchoose to transfer decentralized power to members ofthe local elite? Crook and Sverrisson (2001) describeelite capture as the product of both local agrariansocial and political-economic structures and of local-central relations. To what degree can elite capture beattributed to the status of the elite within the localarena? To what degree can it be attributed to thedesire of central actors to use their resources tomaintain local patronage networks—in other words,to create local elite to serve as their agents in the localarena? To what degree are other dynamics at play?

Elite members of society may be favored becausethey are more able to take advantage of opportunities.Rao and Ibanez (2002 in Mansuri and Rao 2003) andGugerty and Kramer (2000 in Mansuri and Rao 2003)found that the better networked and educated mem-bers of communities derived the benefits of WorldBank social funds. They found that “the formation andtraining of village groups increased entry of wealthier

and more educated men and women into leadershippositions within the group because of the attractive-ness of outside funding” (Mansuri and Rao 2003, 19).In addition, projects (like the one in Madhya Pradesh,India, mentioned above) are likely to favor the mostable members of society so that the projects are morelikely to succeed (Baviskar 2004).

Which forms of decentralization are more or lessprone to elite capture in Africa remains an importantquestion. According to Bardhan and Mookherjee(1999, 2), political scientists commonly presume that“the lower the level of government, the greater is theextent of capture by vested interests, and the lessprotected minorities and the poor tend to be. . . . Ifthe conventional presumption is correct, the advan-tage of decentralizing delivery mechanisms to localgovernments with access to superior local informa-tion would be compromised by greater capture ofthese programs by local elite. The case for suchreforms of decentralization would then depend on theresulting tradeoff between these two effects.”Bardhan and Mookherjee point out, however, thatthere has been little empirical study of the relativevulnerability of local and national governments toelite capture. They cite a large survey in India thatdoes not support the presumption of greater captureof local governments.

Domination by elite may not be all bad. After all, the“local elite are required for the success of local govern-ment systems—they bring resources, knowledge [and]networks that make these systems become fullyoperational and effective,” writes Olowu (2001, 57). Andelite capture does not necessarily mean that all of thebenfits of a reform go to the elite. Evidence suggests thatthe poor are also likely to benefit (Mansuri and Rao2003, 33). However, the more unequal the society,particularly when members of the elite concentrate theirpowers, the less effective are efforts to target the poor.Olowu proposes public participation, accountability, andtransparency as safeguards to prevent the elite fromexcluding the masses (2001, 57). Crook and Sverrisson(2001, 4) propose central intervention “to ensureprogressive or pro-poor outcomes.” Although centralintervention may be necessary to protect the poor, it isimportant to remember that central actors are oftenresponsible for local elite capture.

43


In sum, we need to better understand the relativerole of local and central forces in creating elitecapture. We need to understand the magnitude ofelite capture, the effects of it, and ways to avert it orharness it for public well being.

Role of a strong central stateStudies of improved local government indicate thatdecentralization benefits from a strong central state.59

Ironically, structural adjustment programs, which areimposed upon developing countries by the WorldBank Group to streamline governments, promote thedownsizing of central governments. Downsizingappears to undermine the establishment of soundlocal government by depriving central governmentsof the funds and staff that are needed to supportsuccessful local reforms (Crook and Sverrisson 2001).

Crook and Sverrison (2001) argue that marginalpopulations are best served by decentralized authori-ties when the central state supports poverty allevia-tion and other programs to serve the interests of thepoor. Peluso (2002, 7) points to arguments madeafter World War II and in the U.S. intervention inAfghanistan that a strong state is needed “to defusethe power of local elite and to neutralize their vestedinterests.” Otherwise, one set of central elite may justbe replaced by another more-local elite. Peluso (2002)also points out that violence sometimes accompaniesboth decentralization and democratization.60

In decentralization efforts, violence sometimeserupts over control of natural resources; in democrati-zation efforts, violence can erupt between customaryand democratic authorities. A strong central govern-ment can help mitigate such violence.

In addition, the external support that a strongcentral government can provide local communitiescontributes to the sustainability of decentralizationreforms. While community involvement can makeprojects more sustainable, after a project is com-pleted, the failure of line ministries to providesupport can be the Achilles’ heel of sustainability(Mansuri and Rao 2003, 19). Studies in Malawi(Kleemeier 2000 in Mansuri and Rao 2003, 19) and

in countries around the world (Sara and Katz 2000and Newman et al. 2003 in Mansuri and Rao 2003,19) also found that poor sustainability is largely dueto lack of institutional support from external agen-cies. In addition, Mosse (Mansuri and Rao 2003, 19)finds that the maintenance of community infrastruc-ture depends on external agents, “hence the need fora well functioning state apparatus does not disappearwith CDD [community-driven development, a form ofdecentralized/participatory development interventionof the World Bank].”

A strong central government is also important inensuring equity among local districts. According tothe World Bank (2000, 110), interjurisdictional equitycomes from the willingness of the central state toengage in redistribution among regions. Redistribu-tion can only be accomplished with central govern-ment assistance (Smoke 1999 and 2001, 16). Withoutit, decentralization can result in a situation whereregions or localities endowed with good natural,financial, or technical resources prosper at theexpense of those without (Conyers 2000a, 8).

In short, a strong central state—with an interest inpromoting democratic decentralization and servingmarginal groups—can greatly contribute to the successof a decentralization reform. It can protect the reformfrom vested interests and members of the elite in andout of government, give administrative and technicalsupport to local authorities, provide finances to localauthorities, ensure interjurisdictional equity, ensurethat marginal populations have a voice and are served,mitigate local conflict, and ensure the sustainability ofinterventions. Central states may also need to be strongin order to transfer powers to lower tiers of govern-ment—central governments need strength andsecurity to delegate and give up certain functions.

Accountability of the central state anddonors to local governmentFor decentralization to be successful, not only mustlocal institutions be downwardly accountable, butother levels of government must also be accountableto local government. Crook and Sverrisson (2001, 32)report that poor performance by local government is

44


partly the result of the failure of higher officials andappointees to be accountable to local government.Onyach-Olaa and Porter (2000, 1, 9) argue that localgovernment can only be downwardly accountable ifcentral government agencies are accountable to localgovernment by delivering timely and accurate policyguidance, monitoring, mentoring, complianceverification, and so forth. They make the sameargument for the accountability of donors to localgovernment. When local government needs theassistance of donors or of district, regional, or centralgovernment, either for coordination of larger-scaleactions or for expertise and equipment, mechanismsmust be in place to help ensure that these serviceswill be delivered in a timely manner.

Locally elected representatives—under electoral rulesthat encourage participation and representation—canmake central government more responsive to localneeds (Parker 1995, 27; Crook and Manor 1998; WorldBank 2000, 121). For example, citing Crook andManor, Parker (1995, 27) reports: “In Karnataka,decentralization brought more elected representativesinto the process of RD [rural development]. Thequantity and quality of work undertaken by lineagencies was closely monitored and problems reportedearly, and it became harder for bureaucrats to get awaywith corrupt acts. The result was enhanced institu-tional effectiveness and improved coordination amongcivil servants working for different line agencies.”When local authorities make demands on centralgovernment to secure or improve services, centralgovernment can become more responsive.

The actions that central governments take (or fail totake) also can affect the legitimacy of local govern-ments. Local government needs the support of centralgovernment in order to carry out its functions and togain the confidence of local people. In Transkei,South Africa, central government administratorsfailed to even acknowledge local government in theirconstitutionally sanctioned role in land allocation andin integrated rural planning. The failure of electedlocal government to perform this role undermined

the authority of local representatives and forced localresidents to turn back to tribal authorities for accessto basic resources—although many residents wouldhave preferred to work with their elected authorities.(Ntsebeza 2004.) In China, democratic local govern-ments are the only level of government that isdemocratic. Because of this, they have difficulties inmaking higher levels of government accountable tothem for financing and services (Xu and Ribot 2004).

Relations of accountability between administrativeand representative branches of local government arealso important in holding local authorities accountableto local populations (Doug Porter, personal communi-cation, 1999; cf. Onyach-Olaa and Porter 2000).Central government is made more accountable to localauthorities when deconcentrated branches of centralgovernment are required to account to the elected localauthorities. These vertical and horizontal ties withingovernment can shape the relation of accountabilitybetween local government actors and their constituen-cies (Blair 2000). Similarly, the relations betweencustomary authorities and their administrative superi-ors in government can positively or negatively shapetheir relations of accountability (in many countrieschiefs or headmen hold the official status of adminis-trative assistants to the sous-préfet, préfet, or districtofficer and should be viewed as government adminis-trators) (see Ribot 1999a).

Making central government and its agenciesaccountable to local elected authorities is an impor-tant aspect of decentralization (Blair 2000). Manor(2000, 2) argues that there must be mechanisms forlocal representatives to hold bureaucrats accountableto them and for people to hold elected representativesaccountable to the population (also see Ntsebeza2004). In practice, however, the central governmentusually exercises excessive oversight and control overlocal governments—making local government,elected or appointed, upwardly accountable. This is amajor problem in natural resource managementwhere almost all decision-making powers are attenu-ated through required approval from above.

45


INSTITUTIONAL CHOICE—ASKING THE RIGHT QUESTIONS61

c. Is inclusion interest-based (such as by profession orcause)?d. Is inclusion narrow and interest-based like stakeholderapproaches or broad-based, as in representative systems?

9. Are the institutions favorable toward marginal and poorpopulations?a. Do marginal and poor populations have influence overand voice in the institutions?b. Do the institutions provide mechanisms to ensure theinclusion of marginal and poor populations in decisionmaking and benefits?

10.Whose interests are ultimately served by the institu-tions?a. Are the institutions serving the patronage interests ofcentral actors?b. Are they serving only the interests of their members?c. Are the institutions serving only a subsector or fraction ofthe population?d. Are they serving the population as a whole?

11. Can multiple institutions freely function in the localarena?a. Do citizens have rights to organize?b. Do citizens and local organizations have rights to lobbygovernment?c. Can groups and organizations easily attain legalrecognition and status?

12. Are lines of accountability over public decisionsmediated through representative authorities?a. When non-representative institutions are given publicdecision-making powers, are they accountable to representa-tive authorities concerning the exercise of these powers?b. Do these institutions compete with and underminerepresentative authorities, or do they strengthen representa-tive authorities?

13. What are the long-term implications of the choice ofthese institutions for justice, sustainability, scaleability,and the formation of citizenship?a. Do these institutions encourage broad-based involvementof local people?b. Do they enfranchise people as citizens?c. Do they give local people voice and agency?d. Do they enable long-term stability?e. Are they replicable across territories?

1. What kinds of institutions are receiving powers in thename of decentralization? Are they elected localgovernment bodies, local administrative bodies, localbranches of line ministries, traditional authorities,NGOs, private voluntary organizations, community-based organizations, associations, appointed commit-tees, elected committees, etc.?

2. If the local institutions are elected, do the electoral ruleshelp make them representative?a. How long in advance are elections announced?b. Is suffrage universal? Is it based on residency?c. How are candidates are chosen? Do electoral laws admitindependent candidates?d. What are term lengths?e. Are there means for recalling elected officials?

3. To whom are the local institutions accountable withrespect to the exercise of the transferred powers?a. Through what mechanisms are the local institutionsaccountable?b. Are there multiple mechanisms of accountability?

4. How do the sources of the institutions’ funding orpowers affect their accountability?a. Can the institutions raise revenues locally?b. Are the institutions dependent on grants and fundingfrom outside agents?

5. How do the mechanisms through which resources aretransferred affect the institutions’ accountability?a. Are the powers the institutions receive transferred assecure rights?b. Are they transferred as privileges that can be takenaway?

6. Are the institutions integrative across sectors?a. Are the institutions single-sector or multi-sector oriented?Are they single-purpose or multi-purpose?b. Do the institutions play a mediating role among sectors?c. Are the institutions involved in allocation of resourcesamong sectors?

7. Do the institutions favor procedural matters of democ-racy or a specific set of instrumental objectives?

8. What basis of inclusion do the institutions use?a. Is inclusion residency-based?b. Is belonging identity-based (as in ethnicity, race, religion,lineage)?

Practitioners, donors, policy makers, and activistsneed to ask some straightforward questions in orderto evaluate whether the institutions being chosen toreceive decentralized powers are likely to provide the

equity, efficiency, development, and environmentbenefits that decentralization promises. Thesequestions include the following:

46


47


3

determining how and to whom to transfer power. Byexamining these decentralizations, we can deduce theimplicit principles that have guided the decisions thatleaders have made about power transfers. It isnoteworthy that the principles of power distributionbeing followed in practice are nothing like what mostpromoters of decentralization would advocate.

It is possible to observe how powers are allocated ina given reform by studying the decentralization andsectoral laws, but one must also study the allocationof powers in practice. As one study in Uganda notes,“the statutory description of powers and responsibili-ties may be a poor guide to how things actually workon the ground. Moreover, practice itself evolves overtime” (Plan for the Modernisation of AgricultureGrant Study 2001, 28). In some places, laws areambiguous, leaving interpretation to the more-powerful actors. In Indonesia, the laws that specifycontrol over natural resources are a case in point.Resosudarmo writes, “The explanation of section 7 ofthe law states that utilization of natural resourcesremains with the central government, while section10 states that regions are authorized to managenatural resources within their territories.” Laws mayindicate the intentions of policy makers—withambiguity often being among those intentions. In thefinal analysis, however, practice is the best indicatorof the principles in play.

The most commonly transferred positive powers arethe powers to extract tax and fee revenues fromnatural resources (Agrawal 2001; Conyers 2002; Latif2002; Mapedza and Mandondo 2002; Pacheco 2002;

RESPONSIVENESS OF LOCAL INSTITUTIONS—

POWER AND CAPACITY

Having local leaders who are willing (whethermotivated by sanctions, inspired by personal commitment, or both) to serve the

people well is not enough to ensure that local peopleare well served. Local leaders also must have both thepower and the ability to serve the people. Withoutpower and the ability to use it, they cannot be respon-sive. This section considers the powers that centralauthorities can and should transfer to local govern-ment. It also discusses the issue of local capacity andshows how a discourse on the lack of capacity is oftenused to justify retaining powers at the central level.

POWER MATTERS

Power transfers in practiceNo decentralization or community-based naturalresource management advocates should be calling forthe transfer of all decisions over natural resources tolocal populations. As discussed in a previous section,advocates of subsidiarity maintain that the most-appropriate level for any given decision is the most-local-possible one, provided that making the decisionat this level does not negatively impact the greatercommon good. They have developed subsidiarityprinciples to help determine the levels at whichcertain decisions should be made.

While any well-conceived decentralization reformshould involve careful consideration of subsidiarityguidelines, few, if any, have done so. That doesn’tmean, however, that the makers of current decentrali-zation reforms have not followed certain principles in

48


Resosudarmo 2002; Bigombe Logo 2003; Muhereza2003).62 Income from natural resources has beeninvested in the development of other sectors, as cross-sectoral subsidies to schools, grain mills, and otherpublic projects (Conyers 2002; Bigombe Logo 2003;Muhereza 2003, 8–11). Sanchez and Mearns (in Latif2002, 43) insist that fiscal resources are needed forlocal authorities to fulfill their new roles. Whilenatural resources should not be the sole source ofrevenue for a district, they can be important. Unfortu-nately, the fiscal powers transferred to local authori-ties often do not cover the cost of their mandates,leaving little for investment elsewhere. Withoutadequate fiscal resources, local authorities havedifficulties being effective.63 It is important to keep inmind that although fiscal powers are important, theymay not be the most important powers. Otherpowers, including allocative powers, decisions overthe disposition of significant public resources, rule-making powers, and adjudication powers can alsocreate significant discretion.

The principles of powers distribution being

followed in practice are nothing like what most

promoters of decentralization would advocate.

While the most commonly transferred positivepowers are fiscal, the power to allocate resources hasalso been transferred in many cases. For example, inBolivia, Cameroon, Indonesia, Uganda, and Zimba-bwe, the power to allocate small concessions has beentransferred to local governments (Mapedza andMandondo 2002; Pacheco 2002; Resosudarmo 2002;Bazaara 2003; Muhereza 2003). But, as illustrated bycases in Bolivia, Mali, Nicaragua, and Uganda, whilethese transfers represent a great advance in decentrali-zations, the powers are often attenuated through limitson domain and by excessive oversight (Ribot 1999a;Kassibo 2002a; Pacheco 2002; Muhereza 2003).

In Nicaragua, local governments (municipalities)have been given considerable responsibilities. Theyare required “to develop, conserve, and control therational use of the environment and natural resourcesas the basis for the sustainable development of the

municipality and the country, promoting localinitiatives in these areas and contributing to theirmonitoring, vigilance, and control in coordinationwith the corresponding national entities” (Municipali-ties law, quoted by Larson 2002, 4). Municipalgovernments have also been granted a twenty-fivepercent share of tax receipts on exploitation contracts.Further, central government is required to considerlocal government opinion on exploitation requests. Inpractice, however, municipal governments do notreceive the revenues due, and their opinion is oftenignored or not requested. Most notable, however, isthat “the central government has retained controlover those aspects of natural resource managementthat generate income” (Larson 2002, 5).

In Senegal and Bolivia, similar patterns can be seen.Senegal gave rural council members the right to pre-approve any exploitation, but this right has not beenrespected. Members of rural councils have beenpressured into allowing production in their areaswhile the councils and local populations have ben-efited little from exploitation.64 In Bolivia, Pacheco(2002, 7, 10) explains, the central government hasdevolved use rights to indigenous peoples but hasreserved for itself the right to allocate at least eightypercent of all timber concessions. In addition, whilelocal municipal governments must approve conces-sions in their twenty percent of the domain, it is stillthe central superintendent of forests who allocatespermits. Indeed, while some exploitation rights havebeen devolved, indigenous groups and peasants areconstrained by a prohibition for small producers onthe use of chain saws. This prohibition limits theability of poor forest users to benefit from theresource—it constitutes an entry barrier. In addition,financial and technical-skill limitations make itdifficult for them to engage in harvesting.

In practice, we see that many powers remaincentralized that could be devolved to local authoritieswithout threat to the environment (Fairhead andLeach 1996; Ribot 1999b; Goldman 2001; Conyers2002, 29). Forestry and wildlife agencies transferusufruct rights while retaining central control overthe lucrative aspects of the sector (Ribot 2001b,2001c, 2002a; Bazaara 2002b). Management require-ments are set by central governments that far exceed

49


necessary measures (Fairhead and Leach 1996; Leachand Mearns 1996; Ribot 1999c). Forest agenciescommonly establish complex prescriptive systems offorest management planning, requiring “expert”forestry agent approval before local governments canmake any decisions as to how, when, or where forestsare used and commercialized; this creates more entrybarriers for local people. In most decentralizations,few discretionary powers over natural resources aretransferred to local authorities.

Frequently, mandates and burdens, rather thandiscretionary powers, are transferred to local authori-ties. These are often activities that central governmentwishes to rid itself of, such as powers to implementmanagement programs defined by the environmentalministries, as in Cameroon, Senegal, and Bolivia(Ribot 1996; Oyono 2002; Pacheco 2002), andpowers to enforce, as in India and Mongolia (Agrawal2001; Mearns 2004). Sometimes those activitiescome as funded mandates and sometimes as un-funded mandates, but in either case they provide verylittle discretionary power for local leaders. Undersuch administratively driven decentralizations, thefocus is implementation and enforcement (seeMamdani 1996a). The discretionary decision-makingaspect of the executive is diminished. When suchburdens are transferred to local democratic authori-ties, local authorities are effectively turned into anadministrative branch of the central governmentagency that gave them the mandate. This arrange-ment is closer to deconcentration than to democraticdecentralization. Resosudarmo (2002, 2) has usedthe term “coadministration” to describe this verycommon arrangement. “The lack of discretionarypowers at the local government level de-legitimizesauthorities in local government who are rightlyviewed with skepticism as impotent or as purveyorsof state interests” (Muhereza 2003, 3).

In most decentralizations highly lucrative opportu-nities are retained for central government authorities,while degraded, non-commercial, low-value lands andresources used primarily for subsistence tend to betransferred to local authorities (Ribot 1999b, 2000;Bazaara 2002b, 2003; Resosudarmo 2002; Xu 2002;Muhereza 2003). One way that central governmentsretain control over lucrative resources is to create

large concessions (see Box 11, above). In Indonesiaand Uganda the power to allocate small timber orwoodfuel licenses has been devolved, but largelicenses and concessions have been retained forcentral government allocation (Resosudarmo 2002;Muhereza 2003). In Cameroon, local communitieshave no access to vast concessions that the centralgovernment continues to allocate to foreign anddomestic commercial forestry corporations (BigombeLogo 2003; Etoungou 2003; Oyono 2003). InUganda, the central government is involved inallocating mining and timber concessions, issuinghunting licenses, and managing tourist parks, buthas little interest in regulating non-timber forestproducts for household consumption, firewood, andlocal fishing resources (Bazaara 2003). As theparticipants in the Bellagio workshop observed, “usevalues rather than exchange values are being decen-tralized” (Bellagio Working Group 1, in Latif 2002,53).

Sometimes central governments refuse to decentral-ize natural resource management powers becausethey are concerned (or claim to be concerned) aboutthe resource. They rightly point out that devolvingcertain technical decisions can be problematic, butthey use this as justification not to devolve naturalresource management decisions at all. But allocativedecisions, which relate only to who can cut trees, arenot technical decisions. They can be made at the locallevel without jeopardizing environmental resources.Principles of environmental subsidiarity are neededso that debates over technical necessity and theinstrumental—environmental or political—aims ofinterest groups do not dominate the setting of laws.For example, Bazaara found that in deciding whichresources to devolve authority over, environmentalministries often conflated technical decisions thatshould be made by the ministry with political deci-sions of allocation. Technical decisions concerninghow much to cut, which species to cut, where to cut,techniques for cutting, and when to cut, may be subjectto technical guidelines. Allocative decisions concern-ing who can cut trees can be devolved withoutconcern for the resource if the technical decisionshave already been made. Either over-specifying thetechnical or retaining the political allocative decisionsblocks transfer of appropriate powers.

50


Which technical decisions can be devolved andwhich cannot depends on many factors. Khwaja(2001 in Mansuri and Rao 2003, 16) found thatcommunity participation in technical projects innorthern Pakistan adversely affected outcomes whilein non-technical projects community participationimproved outcomes significantly. He also found thatcommunity participation led to better maintenancethan projects built by local government—not that thetwo are mutually exclusive. Finsterbusch and vanWincklin (1989 in Mansuri and Rao 2003, 16) foundthat in fifty-two U.S. Agency for InternationalDevelopment projects with participatory elements,the smaller and less-technically complex projectswere more effective.65 But is this difference due to thenature of the participation? What were the factorsthat made large, technical projects difficult: hightransaction costs, inequality in open social fora, thehigher stakes of the more-complex technical projects,or the use of technical complexity by outside agentsand members of the elite to retain control?

Limiting the territory over which local authoritieshave jurisdiction is often used as a strategy to limitlocal powers. In Uganda, the right to revenues andthe ability to allocate concessions and productionpermits were restricted by establishing centralreserves and privatizing forests to customary users(see Box 12). In Cameroon, there is a progressivecommunity forestry law, but community forestswhere communities are permitted to engage incommercial activities can only be established in a verylimited set of areas determined by the forest service(Ekoko 2000; Oyono 2002). In Mali, the very strongdecentralization laws that give local councils consid-erable powers over local forests are underminedbecause the central state has not established a localdomain in which to exercise them (Ribot 1995a;Kassibo 2002a).

In other instances, powers devolved are hemmed inby overly restrictive management requirements oroversight. Often, this right is also shrouded inextensive planning requirements and tight oversight,restricting local discretion—these restrictions arefurther discussed in the subsection titled “TheCapacity Conundrum.”

BOX 12 THE LOSS OF LOCALJURISDICTION IN UGANDA’SMASINDI DISTRICT

In Masindi District of Uganda, the jurisdiction oflocal governments was severely restricted by thereclassification of forests under the 1998 ForestReserves Order. The order placed centralreserves under direct central government controland limited district management to the localforest reserves. It reclassified seventeen forestsin Masindi as central forest reserves anddesignated eight forests as local forest reserves.

Only two of these, Kirebe (49 hectares) andMasindi Port (18 hectares), remained underdistrict council jurisdiction after May 2000,when the others were returned to the Kingdomof Bunyoro-Kitara. Then, in 2001, the NationalForest Authority gave the kingdom the MasindiPort eucalyptus plantation, leaving only theKirebe Forest to the Masindi district council.One village forest was established in 1999 in avillage called Alimugonza with the help of aconservation and development project funded bythe U.S. Agency for International Development.This village forest was declared by resolution ofthe local government, and was at the time of thisstudy still being managed in an ad hoc manner.

These changes left local governments with verylimited areas and almost no forests over which toexercise the powers transferred to them by thedecentralization laws. Most of the forests overwhich they had had jurisdiction were centralizedor turned over to commercial interests andcustomary authorities.

A similar phenomenon is seen in Mali, where thenew forestry laws have given new powers to localauthorities, but the government has yet to give thema domain over which to exercise these powers.

Sources: Ribot 1999a; Muhereza 2001a, 17–9; Muhereza2003, 22–3.

51


A careful examination of current decentralizationsin law and practice reveal the following:

● Few discretionary powers are being devolved, andoverbearing oversight is reducing that discretionfurther.

● Mandates to execute centrally defined programsare being devolved.

● The odium of management is being transferred.● The most-lucrative opportunities are retained for

central control while the least-lucrative activitiesare being devolved to local authorities.

● Allocation decisions concerning who shouldreceive powers are being retained by centralgovernment—even though these decisions couldbe devolved with little consequence for naturalresources.

● Control over subsistence resources (use values)and over resources with little commercial value isdevolved.

● Income-generating resources such as forests andagricultural land are being privatized.

● The extent of the territorial jurisdiction over whichdecentralized bodies have authority is beinglimited;

● The largest and most lucrative areas (farmland,pasture, and forests) or resources (minerals andoil) are being reserved for private interests orcentral management.

● Fiscal resources are often insufficient to cover thecost of mandated service provision and manage-ment burdens.

The overriding principle that most governmentsseem to be following is: keep everything of valuecentralized and transfer centrally defined obligationsto lower-level authorities. This is a clear formula forcentral consolidation. It does not produce democraticdecentralization.

To determine whether central authorities aretransferring the powers that are necessary for demo-cratic decentralization, analysts should answer severalkey questions. A list of suggested power choicequestions is provided at the end of this section.

The overriding principle that most governments

seem to be following is: keep everything of value

centralized and transfer centrally defined

obligations to lower-level authorities. This does

not produce democratic decentralization.

Privatization is not decentralizationWhile private sector participation in development isnow accepted as a reality and a desirable option,privatization raises questions that are radicallydifferent from those of decentralization. For one thing,privatization targets allocation efficiency and does notshare decentralization’s concern for equity through theredistribution of power. For another, the former focuseson the contribution of a narrow circle of economicactors (investors, entrepreneurs, and individuals withaccess to capital or information), whereas the primaryobjective of decentralization is the dispersal of eco-nomic and political power in such a way that themass of the people have a sense of participating indecisions affecting their life and well being. Thirdly,once the state surrenders the ownership and manage-ment of an enterprise, it (the enterprise) ceases to bepart of the “public business” for which governmentministers are individually or collectively responsible.In any case, by no stretch of the imagination canprivatization be equated with mass participation indevelopment, or regarded as a major landmark inpolitical decentralization.

Balogun 2000, 155

There is considerable confusion in decentralizationsconcerning which powers belong in the public sphereand which should be private. For example, in Bolivia,Cameroon, Indonesia, Nicaragua, and Senegal, lawspecifies that forests are under central governmentmanagement (Ribot 1996; Larson 2002; Pacheco2002; Resosudarmo 2002; Oyono 2003). This placesthem in the public domain. But in many places,natural resources are being privatized in the name ofdecentralization (Ribot 1999a; Alden Wily 2000a, 7–8; Muhereza 2003). In Uganda, for example, forests

52


are being privatized to customary authorities andcustomary users (Muhereza 2003) (see Box 8, above).Some resources are appropriately private, but manyshould remain in the public domain. Better guide-lines are needed to determine which resources shouldremain public and which should be privatized. This isa critical topic for public debate.

As discussed earlier in the report, privatization isnot a form of decentralization. Decentralization ispredicated on an inclusive logic, while privatizationoperates on a very different, exclusive logic (Agrawaland Ribot 1999; Balogun 2000, 155; Oyugi 2000, 6).Privatization involves transferring resources in thepublic domain to private bodies. While it is usuallythought of as a transfer to individuals or privatecommercial interests, this report considers transfersto non-commercial private entities, such as NGOsand management committees organized by donororganizations and forest services, also to be a form ofprivatization—”nonmarket privatization.” This formof privatization is effectively the same in that itremoves decisions from the public domain. Bothtypes are oriented toward individual or memberobjectives—which are not necessarily those of thepublic.

Transfers to private bodies, whether commercial ornoncommercial, do not constitute democraticdecentralization. Meinzen-Dick and Knox (1999, 4)point out that the private sector can include usergroups, but, in their definition, they limit thesegroups to “individuals or firms, who are accountableto their shareholders, and NGOs, who are account-able to their donors.” Neither user groups who areaccountable to donors nor user groups who areaccountable to shareholders are likely to be represen-tative of, and accountable to, the local community asa whole. This alone means that transfers to them donot constitute democratic decentralization. Account-ability to donors or the state may make them intoeffective administrative bodies, but transfers toadministrative bodies certainly do not constitutedemocratic decentralization. Transfer of powers toother authorities, such as local chiefs, may alsoconstitute privatization, depending on whether thechief is, as is often the case, an administrativeauxiliary of the state (hence upwardly accountable),

an elected official or a person dedicated to the localpopulation (downwardly accountable), or just anautocratic local actor (effectively a private body).

NGOs, autocratic chiefs, and other private bodiescan serve the public interest or only the interest oftheir members (Fox 1990; Guyer 1994; Fox andBrown 1999; Conyers 2000b, 21). They may beupwardly, downwardly, or horizontally (to their peers)accountable within certain regulatory and moralbounds, but their objectives are determined by theirmembers—not the public as a whole. Privatization tothese bodies can lead to more exclusion than partici-pation and to less public accountability (cf. Balandand Platteau 1996; Woodhouse 1997; Hilhorst andAarnink 1999). For example, in the name of decen-tralization, Namibia has created the CommunalLands Reform bill. According to Alden Wily (2000a,7), the intention of the bill is “to open local commonsto individual entitlement,” but doing so “also threat-ens local forest tenure security.” In other places,privatization of public lands takes place throughforestry concessions to corporations, as in Cameroon(Oyono 2003), Mozambique (Alden Wily 2000a, 8),Senegal (Ribot 1996; République du Sénégal 1998),and Uganda (Muhereza 2003).

In addition to reducing the public domain, transfers toprivate bodies take resources that could otherwise go tostrengthening and legitimizing democratic authorities.

Geographic scaleGeographic scale also matters for the distribution ofpowers (Dupar and Badenoch 2002). For purposesof accountability, representation and participation inforestry, some political or administrative jurisdic-tions may be too large to really be considered local—such as the most-local level of local government inBurkina Faso.66 Which scale is most appropriate forwhich kinds of decisions? Cousins (in Latif 2002,23) points out a disjunction between types of naturalresources concerning the scales of management.Different resources have different scales andboundaries. There is not one appropriate level.Watersheds and forests may not match local govern-ment political or administrative jurisdictions.

53


Matching jurisdictions with ecological formationscannot always be accomplished.

This research program focused on the most-locallevel of decentralized local government and thereforedid not explore higher-scale jurisdictions. Neverthe-less, some approaches have been suggested. Oneapproach to multiple geographic scales is to encour-age the formation of local government federationsand networks so that upstream and downstreamconstituents can work together for mutual benefits.This may be better than establishing new special-purpose districts (Ostrom, Schroeder, and Wynne1993). The impulse to recentralize into higher-scaledistricts makes sense for some resources, but it mustbe cautiously approached. Adding more scales ofgovernance can take powers away from the localarena, concentrating them at higher levels (seeMandondo and Mapedza 2003). Proliferation ofscales, being advocated by polycentric governanceproponents, may diffuse powers among too manyactors, rather than creating a cogent managementsystem.67 The formation of federations and networksmay keep governance local while attending to multi-scale problems (see Section V).

The “landscape” approach to environmentalmanagement, now becoming popular among envi-ronmentalists and donors, should be approached withcaution. It should not be used as one more excuse tomaintain or re-centralize control over natural re-sources. Perhaps landscapes can be viewed as zonesin which different minimum standards apply (seediscussion of minimum standards in Section III). Inthis manner, it could be made to be consistent withprinciples of decentralization.

Interjurisdictional equityBoth the decentralization of natural resources and thatof social services and infrastructure pose issues ofinterjurisdictional equity. In the cases of social servicesand infrastructure, a strong argument can be madethat financing these activities from local revenues leadsto a highly unequal—and hence unjust—distributionof benefits. This implies that rich areas should bemade to transfer some of their resources to the poorer

areas. This argument also arises in the context ofnatural resources with high commercial value, such aspetroleum, minerals, and timber. Ironically, in manyinstances the people living in regions that are rich innatural resources are quite poor precisely becausecentral governments and large companies havetraditionally appropriated the great majority of benefitswithout reinvesting in the local area. In addition,resources are often put into equalization funds orcentral government agencies for redistribution, butnever return to the local arena (see Xu 2002).

In some cases, environmental conservation hasbeen threatened by interjurisdictional inequity. Whenthe Indonesian government transferred to localgovernments the power to allocate exploitationlicenses and permits and to collect taxes on forestproducts, concerns about interjurisdictional inequityarose because the forests in some jurisdictions areprotected while the forests in others are not. Thosedistricts whose forests are protected for conservationfeel cheated of income opportunities and tax revenueswhen they observe neighboring districts profitingfrom the decentralization reforms. As a result, localdistricts with conservation areas are seeking toconvert them to production areas, threatening theirprotected status. (Resosudarmo 2002, 18–9.) Inshort, interjurisdictional equity needs to be taken intoaccount when designing subsidiarity principles.

Security: The “means of transfer”problemThe “means of transfer” problem is another criticaldimension of power in decentralization (Conyers1990, 20; Ahwoi 2000). The security andsustainability of decentralization reforms dependlargely on the means used to transfer powers fromcentral government to other entities. Means oftransfer can be constitutional or legislative, or theymay take place through ministerial decrees oradministrative orders.

Constitutional transfers are the most secure orsustainable (see Conyers 2000a). Ethiopia, Ghana, Mali,Senegal, South Africa, and Uganda all have constitu-tional clauses that ensure some degree of government

54


decentralization (United Nations Capital DevelopmentFund 2000, 6). While these clauses do not specifywhich powers are decentralized, they provide leveragefor lawmakers to establish and maintain decentralizedgovernance arrangements. The specification of thepowers to be decentralized—whether or not there isconstitutional support for decentralization—usuallytakes place through decrees and orders, which are less-stable forms of transfer. Powers transferred by ordersand decrees can change with the balance of powersamong parties or with the whims of the party oradministrators in power (Ahwoi 2000).

In Burkina Faso, Cameroon, Guinea, Mali, Senegal,South Africa, Zimbabwe, and certainly elsewhere,decisions concerning the allocation of importantpowers are made by ministerial or administrativedecree. In Mali, for example, decentralization is writteninto the constitution, while decentralization of powersover natural resources are specified in environmentallegislation, such as the 1996 forestry code. However,within the forestry code, the powers to be devolved areleft to be specified by decree of the minister respon-sible for forests. The procedures to resolve disputesover forestry matters are to be specified by order of thestate-appointed governor of each region. Like manyenvironmental agencies around the world, Mali’senvironmental service has yet to officially transfer anypowers to local authorities. The decentralization inMali’s environmental sector is thus a discretionarymatter for the ministry responsible for forests and itsadministrative staff. In this manner, what appears tobe a constitutional guarantee is transformed intoexecutive-branch discretion (Ribot 2002b). Retaininginordinate discretion in the executive branch alsocreates opportunities for allocation along political andsocial lines to serve the interests of central agents.Secure means of transfer may help to reduce suchopportunities for corruption.

Like land-tenure security, the security of the transferof decision-making powers from central governmentto local institutions affects the sustainability of thereforms and the willingness of local people to believeand invest in the reforms. Until people believe thatthe rights they have gained are secure, decentraliza-tion may cause over-exploitation of resources. Whenlaws change frequently, people do not believe or

invest in them. New rights to exploit forests may beexercised with urgency by people who believe that thegovernment will take these rights away in the nearfuture. Such over-exploitation following decentraliza-tion reforms was observed in Kumaon, India, in the1930s before the situation stabilized, and exploitationhas been intense under Indonesia’s recent decentrali-zation (Resosudarmo 2002; Resosudarmo andAgrawal in Latif 2002).

Insecurity also discourages the formation of citizen-ship. Local people may not engage or organize ascitizens to influence local government when localgovernment has little or only temporary power. Also,central authorities can subject local people to their willby threatening to withdraw powers—they can trans-form even elected local authorities from instruments ofenfranchisement into central administrative agents.

Setting the boundaries on local power:Planning versus standardsLocal people do not necessarily manage local resources“better” than others. Just as non-locals, “local people”may be driven by the profit motive or the desire forpersonal enrichment; they may have economic needsthat promote over-exploitation of resources. It is mostlikely that there will be different groups of people withdifferent resource needs and interests. To name a few,these may include indigenous populations; NGOs,which may be oriented more toward preservation ormore toward sustainable-use conceptions of “conserva-tion,” and which may represent broad sectors of localpeople or only a small élite; groups of peasants orotherwise poor resource users; and capitalist enterprises,which may be run by local people or represented bylocal intermediaries. Under these circumstances whatdoes it mean to promote local participation?

Anne Larson 2002, 15–6

How can decentralization be structured so that thecomplex mix of local objectives are achieved withoutcompromising national well being or causing nega-tive outcomes for segments of society upstream anddownstream or at larger socio-political and ecologicalscales? Conservation can contribute to, or be atradeoff with, subsistence use, poverty alleviation,

55


vulnerability reduction, or wealth generation. Decen-tralization should enable local people to exploitresources—for subsistence or cash. The question is,within which management parameters and whodecides what those parameters should be? Powers notonly need to be transferred, they must also bebounded. But what is the best way to set boundarieson power with the least compromise to local au-tonomy and to well being at all levels? This is apolitical question that needs to be technically in-formed and publicly debated.

Centrally determined laws, standards, and guidelines,while clearly necessary to protect some public values,put bounds on local discretion. If they are highlyspecified and restrictive, then the degree of discretionis limited. The setting of regulations and standards is alocus of tension between diverse interests at thenational level where they are made, as well as betweennational and local interests. National regulations andstandards circumscribe the domain of local autonomy.Decentralization becomes effective when the domainof local autonomy is high. This is when local authori-ties have significant discretionary powers.

Since certain forms of regulation may favor localdiscretion over others, they may better fit the require-ments of decentralization. In current decentraliza-tions, central agencies have set highly specifiedrequirements for local authorities to follow inmanaging their natural resources. Resource usetypically is circumscribed by highly elaborate rulesand tight central-agency oversight. In current decen-tralizations, almost all new rights are accompanied byrequirements for elaborate management-planningexercises prior to any exploitation of the resource. Analternative to the management planning approach iscalled minimum environmental standards. Planningand standards are discussed here in turn.

Management planningDecentralizations often go wrong when they devolverisks and costs that are too great for local groups. Thedifficulty of complying with complex managementplans and bureaucratic procedures required by thegovernment and the cost of the staffing and equipmentneeded to carry out management and monitoring

requirements may be too great. In Bolivia, the cost ofcreating land-use plans was beyond the means of smallfarmers, making it impossible for them to comply withthe laws (Latif 2002, 67). In Cameroon, the process forcommunity engagement in forest management wastoo complex for villagers to navigate (see Box 13).

In many instances, management plans are notneeded or could be simplified. They are often hurdles

BOX 13 OVERLY COMPLEXMANAGEMENT PLANNING

In Cameroon, communities wishing to set up acommunity forest under the 1994 forestry lawmust navigate a complex process. They arerequired to

● create and register a community forestrymanagement committee with a writtenconstitution,

● make maps of and demarcate their traditionalterritories,

● compare those boundaries with allowable zonesin a government-approved forest-use plan,

● determine the extent of forest to be set asideas a community forest,

● establish a simplified forest managementplan, and

● seek approval for the management plan by thelocal prefecture administrative authorities andthe central government.

In 2001, however, under pressure from donorsconcerning the difficulties these unnecessaryrequirements pose, the Ministry of Environmentand Forests set up a community forestrydevelopment unit to provide implementationassistance. Nonetheless, to date, no communityhas been able to establish a community forestwithout extensive external assistance from NGOsand donors.

Sources: Ribot 1995a; Etoungou 2001; Graziani andBurnham 2002; Oyono 2003, 2004a, 2004b, 2004c.

56


to prevent local people from exploiting the resource(see Box 13). They also make local communitiesdependent on central government or an externalNGO for assistance in elaborating plans before theycan access surrounding resources. In Mali andSenegal, complex plans have been required for localauthorities to use or exploit woodfuels (Ribot 1995b).While many governments, such as Senegal, are nowintroducing simplified management plans, theseplans remain a barrier for local decision making andmaintain dependence through approval processes(field interviews in Senegal by author, July 2004). InSouth Africa’s land reform, business plans arerequired that are not necessary (Cousins in Latif2002, 61). In all of these cases, management plansbecome a sticking point in the transfer of powers. Ineach case, plans are a means for unnecessary central-ization of decisions and approval processes.

Management planning can be useful. In forestry, forexample, planning can be a tool for minimizing theeffects of certain kinds of extraction or for increasingthe natural regeneration of trees. Managementplanning can also serve as a protection for localpeople since it can codify their forest use rights. Anelaborate management plan can also be a negotiatingtool for local populations over the use of surroundingresources. At present, however, rather than being atool to enhance people’s security and their ability tobenefit from the resource, management plans are ahurdle because they are required for all exploitationactivities. While there are many activities that can beconducted without management plans, forest servicescontinue to refuse to acknowledge this.

Minimum standardsMany activists and practitioners believe that localcommunities are more conservative than are outsid-ers toward natural resources. While this may be truein some cases, there is still good reason to expect thatlocal communities will convert natural wealth intofinancial wealth, especially where cash is in shortsupply, and where cash is viewed as more valuablethan, for example, standing forests. Communitiesmay act like private corporations when benefits ofresource exploitation outweigh immediate costs—asthey well should.

In an example of community choosing exploitation,Larson (2002, 16) describes some cases whereincreasing incomes of the poor increases environ-mental degradation—particularly deforestation wherepeasants consider ranching to be a better means thanforestry for escaping poverty. Poverty alleviation andconservation are often trade-offs (see Kaimowitzforthcoming, cited in Larson 2002, 16). Of course,this is just fine where local, national, and societalgood is not threatened by the choices being made.Local forest use, or even conversion to agriculture,can be positive: having local people evaluate thenecessary tradeoffs is the procedural objective ofdemocratic decentralization.

Minimum environmental standards are more

consistent with local autonomy than are

complex management plans. They are a

logical complement to decentralization.

In an example from Cameroon, Oyono (2002, 12)explains that younger members of communities feltthat they had the right to consume forests as ameasure of intergenerational justice. The creation ofcommunity (village-level) and communal (localgovernment) forestry has caused rifts between youngand old over forest exploitation. Young people want tobenefit from the forests. As one youth said, “the state,those who are dead, the old persons, and our fathershave eaten too much from the forests; now it is ourturn” (Oyono 2002, 12).68

Part of the incentive structure that communitiesshould face in making these consumption and forest-conversion decisions should include a minimum set ofrules designed to protect the resource base. Environ-mental standards are typically a set of rules that specifywhat activities are not permitted and the kinds ofmeasures that must be taken when engaging in others.For example, in forestry, a set of minimum standardsmight specify which species are not to be cut, theminimum or maximum diameter allowed for thecutting of other species, the height at which a treemust be cut to encourage coppicing, the percentage of

57


the forest that may be cut each year, the percentage of agiven species that may be taken in a given year, etc.

Under a standards approach, anything that is notspecified in the law is permitted. This contrasts with amanagement planning approach, where only activi-ties that are specified are allowed (it is also the polaropposite of Napoleonic code law found in mostformer colonies of France) . A minimum environ-mental standards approach would replace the micro-management approach currently exercised throughelaborate plans and planning processes. A set ofminimum standards specifies a set of restrictions andguidelines for environmental use and management.Any local government or individual operating withinthose restrictions needs no management plan to useor manage resources.

Minimum environmental standards are moreconsistent with local autonomy than are complexmanagement plans. They are a logical complement todecentralization. The minimum-standards approachcomplements decentralization by specifying theboundaries to the domain of local autonomy withoutrestricting discretion within those boundaries.Standards can allow more room for local flexibilityand adaptation, and they eliminate layers of requiredapproval that can block local initiative. Standardscould enable local communities to exercise the rightsto exploit and profit from forests, or to convert foreststo other uses, without causing the kinds of social andenvironmental damages commonly associated withcentrally allocated commercial concessions.

The use of minimum standards does not implyblind trust. It requires central authorities to deter-mine and set the limits of local action in the firstplace and to monitor and sanction. Local representa-tive authorities must be given incentives to enforcethese standards, make public management and usedecisions, and mediate disputes among users. Thismight mean that if standards are not respected, thecommunity will lose its right to exploit the resourcefor a specified period or until they produce a plan forrestoring the damages done by breaching existingstandards. It may mean that the community orindividual who breaches the standards is fined.

However, great care must be taken in creatingpunishments so that they do not become a means toexploit local communities. Otherwise, fines might beimposed and restoration of resources might berequired even when the actions do not merit a fineand the damage is not significant. To protect againstabuses, the standards must be clear and as unrestric-tive as possible. If they restrict activities that localshave practiced in the past and know do not damagethe resource base, local people will not respect thestandards and will not comply with them. It isimportant to ensure that the setting of minimumstandards does not become an opportunity for a newform of coercion and extraction.

The ability of the minimum-standards approach toempower local populations is undermined whentechnical standards are set too high, as they often arein many sectors (Tendler 1980; Ribot 1995a; Oyono2003). Excessively high standards can prevent localpopulations from entering into activities and canimpede service delivery by raising costs beyond localmeans. In some cases, high standards are a meansfor a technical ministry to impose its control overresources (usually lucrative ones); in some cases, highstandards are an artifact of unadapted regulationshanded down by colonial governments; and in somecases, high standards are the result of misperceptionsabout the resource itself.

In natural resources the potential domain of localdiscretion, in which national good is not threatenedand local knowledge is sufficient for management, isvery large. It is much larger than environmentalministries are typically willing to admit (Fairhead andLeach 1996; Leach and Mearns 1996; Ribot 1999c).More research and public debate must go intoidentifying the boundaries between what can andcannot be done without the direct intervention ofenvironmental agencies.

The minimum-standards approach is a goodcomplement to democratic decentralization because itcreates the space for a high degree of local autonomywhile ensuring that local actions do not have negativeconsequences for others. A shift from a management-planning to a minimum-environmental-standardsapproach is in order. This does not mean that man-

58


agement planning should be eliminated. Some kindsof actions may require plans. But unless the activitiesviolate or require modification of the minimum

standards, central environmental ministries shouldnot require permission, direct approval, or plans.

The minimum-standards approach is a goodcomplement to democratic decentralization because it

BOX 14 GREATER LOCAL EQUITY AND REVENUES, FEWER OUTSIDE LOGGERS INBOLIVIA

Bolivia has enacted one of the most laudeddecentralizations in Latin America. The 1994popular participation law and 1995 administrativereform initiated the transfer of responsibilities fromcentral to departmental and municipalgovernments, expanded local governmentjurisdiction beyond urban areas, and establishedtax-revenue sources for local governments. (InBolivia, the municipality is the most-local level oflocal government.) The forestry reforms of 1996transferred research, extension, conservation, andmunicipal-level capacity-building responsibilities todepartment-level governments and gavemunicipalities the right to monitor forestry activitiesand to allocate up to twenty percent of the forests intheir jurisdictions as community reserves andconcessions.

Each municipal government has also beenempowered to appoint a municipal forestry unit(MFU) to carry out its new functions. The MFUdevelops management and use plans that must beapproved by the national forest superintendent.Local forest user groups also have been given theright to up to twenty percent of the national lands ineach municipal district. (The other eighty percent ofthe forest remains under national governmentcontrol.) The local governments are elected byuniversal suffrage from candidates presented byparty lists. Local officials serve five-year terms.Councils choose, and can recall by vote, a mayorfrom among themselves.

Bolivia’s changing institutional arrangements haveprovided several social benefits. Foremost are thecreation of opportunities for a more equitableallocation of resources, greater control by marginal

groups over local resources, and increasedparticipation in local government and themanagement of natural resources. Members of theelite have been forced to recognize and negotiatewith marginal groups. Also, indigenous groups andsmall farmers have been able to gain authorizationsfor small-scale logging operations and alternativeland use due to the election of indigenous people tomunicipal councils through alternative parties. Inreaction to deteriorating livelihoods in pastoralzones, civil society groups have mobilized toincrease downward accountability of more receptivelocal governments. There have also been somepositive environmental outcomes. Localgovernments are helping control non-resident andlarge-scale timber exploitation, and the chain sawrestrictions enacted may be good for the forests.Expenditures on forestry are sharply up due to theportion of forest taxes that municipalities nowreceive.

The experience has also had its downside. Land-claim processes are too bureaucratic and slow, land-use plans are not very participatory, vigilancecommittees are not working, public investment isstill too urban, and decentralization has opened newopportunities for elite capture. Despite greaterrevenues, local governments are investing very littlein natural-resource projects, timber royalties areinsufficient to support the new municipal forestryunits, there is little control over illegal activities bylocals, chain saw restrictions limit access to forestsby poor users, and land-use plans that justifyclearing are less expensive than forest managementplans for poor farmers.

Sources: Larson in Latif 2002, 27–8; Pacheco 2002, 9.

59


creates the space for a high degree of local autonomywhile ensuring that local actions do not have negativeconsequences for others.

Local control over outside industriesMany of the problems faced by local authorities andlocal people are caused when outside commercialinterests enter the area, cut their forests, and leave.As Bardhan, Ghatak, and Karaivanov (2002) pointout, large, wealthy operators have an exit option. Theycan leave when the resource is exhausted, causingover-exploitation, and inefficiency and inequity inresource use. Extra measures are needed to disciplinelarge and wealthy operators.

In some instances local authorities have beenempowered to confront and stop this kind of exploita-tion (see Box 14). But in most cases local authoritieshave little power over extractive activities. Thesepowers are further diminished by the small domainover which they are given authority; most forests aremanaged by central government, and local authoritieshave no power over these forests. Local authoritiesshould have more control over the forests within theirterritorial jurisdictions and more control over whocan enter and use the resource.

The right of local authorities to say NO to exploita-tion in their domain would be one way of addressingthis issue (see Ribot 1995a). This does not mean thatlocal authorities automatically should have the rightto exploit the resource themselves. But this is onemeasure of local control that has been established inlaw in Mali and Senegal—although in Senegal, ruralcouncil presidents have all been pressured into saying“yes” to production, despite widespread disapprovalby local populations (Thiaw and Ribot 2003 andfieldwork in Senegal by author, April 2003 and July2004). It empowers local populations to at leastconserve the resources around them (Ribot 1995b;République du Sénégal 1998).69

Unfortunately, central governments are oftenunwilling to empower local authorities to denyoutside industries access to local forests. Donors and

international and local activists can help by pressur-ing central governments to transfer this power.

THE CAPACITY CONUNDRUM

Capacity is an illusive quality—linked to skills,knowledge, and legitimacy. Local institutions receiv-ing powers they have never held may not have theskills to exercise those powers effectively. Capacity isessential to the exercise of power and to responsive-ness. To be responsive to the needs of its constitu-ents, local government may require the following:technical knowledge and skills, financial manage-ment skills, ability to call for technical expertise,ability to negotiate, and ability to use courts (Baviskarin Latif 2002, 69).

In decentralization reforms, governments oftenargue that the choice of local institutions and powersmust be a function of the capacity of local institu-tions. Typically, local authorities are cited as lackingfinancial management skills and technical knowledgeabout resource management and use. The capacitiesneeded cannot be separated from the kinds of powersbeing transferred and the kinds of obligations andmandates accompanying those power transfers.Capacity is necessary for the powers that requireskilled action to exercise. Deciding how to conserve orexploit a forest may require some skills or at leastsome guidelines. However, many powers, such as thepower to decide who should exploit a forest that hasbeen earmarked for exploitation by a forest service, donot require special skills. Deciding whether to exploitor conserve surrounding forests also may requirelittle technical knowledge. This subsection exploresrelations between powers and capacities as well ashow capacity arguments are used to disable, ratherthan to enable, the process of decentralization.

Capacity is a means for achieving a combination ofobjectives. It can be viewed as the tools that enablethe pursuit of a goal. “In that sense, one knows thatthere is a capacity deficiency when particular organi-zations or institutional arrangements are under-performing relative to what could be achieved giventhe available resources” (Fiszbein 1997a). Capacity,therefore, should be evaluated with respect to specific

60


goals. “When faced with the question of whether localcapacity constitutes a bottleneck to successfuldecentralization reforms, before attempting to offeran answer, one should ask ‘capacity to do what?’”(Blair 1997).

Two common fears are expressed by central govern-ment agencies involved in decentralization of naturalresources. First is that the transfer will lead to thedestruction of the resource. Second is that thetransfers will lead to local conflict. These problemsare often characterized as “risks to manage” bytechnical agencies. Sometimes these fears arefounded. Usually they are not, or they are greatlyexaggerated (Fairhead and Leach 1996; Ribot 1999b;Swift 1999; Baviskar 2002). Nevertheless, these fearsare used to argue against the transfer of powers tolocal authorities. Arguments that too much decen-tralization has caused damage or over-exploitationhave led to backlashes against decentralizations inUganda in 1993 and more recently in Ghana and

Indonesia (Bazaara 2002b; Aaron de Grassi, personalcommunication, 2002; Resosudarmo 2002). Argu-ments that there will be conflict with customaryauthorities have been used in Burkina Faso, Mali,Mozambique, Niger, Senegal, South Africa, andZimbabwe (Ribot 2001b).

Based on these expressed fears, arguments thatlocal populations lack the capacity to manage re-sources and conflict are common affronts againsttransferring resources. These capacity arguments arebacked by technical arguments concerning manage-ment requirements and by stories of conflicts anddeaths from resource disputes since decentralizationbegan. While these fears may not be groundless,there are certainly management and use rights andpowers that can be transferred without threateningthe resource base or causing new conflicts. Further,according to researchers at the Bellagio conference(Latif 2002), there is insufficient evidence thatconflicts increased after any of the decentralization

BOX 15 CAPACITY AS AN EXCUSE FOR RETAINING CENTRAL POWER IN NICARAGUA

The lack of capacity to make technical decisions wasthe excuse long given by the Nicaraguan ForestryInstitute (INAFOR) for not transferring powers tolocal governments (municipalities). INAFOR officialsstated that municipal governments’ opinions onmilling contracts would not be taken into accountunless backed by “rigorous technical argument.” Fewmunicipalities, however, had technical personnel.

Then, in 2000, transfers of decision-making powersbegan to take place when a new director was hiredwho was committed to deconcentration anddecentralization and to eradicating corruption withinINAFOR. The new director fought for the transfer oftwenty-five percent of the budget to municipalities,arguing that this would get municipalities to take thenew legal framework seriously. She met regularlywith municipal leaders, providing them withinformation on logging and the forestry industry. Shegained the support of the agroforesters’ organizationby simplifying the permitting process and bystrengthening INAFOR municipal delegations.

Then, in 2001, after making these positive changesand winning the acknowledgement andappreciation of local people, the new director wasfired. According to the minister, she was firedbecause she allowed local governments to over-exploit the forests without controls. According tothe former director herself, she was fired becauseshe refused the minister’s request to administerillegal logging permits. According to others, shewas fired because she married a logger. In anyevent, INAFOR officials soon after reversed herreforms.

This story indicates two important points. First isthat the notion of capacity is used as an excuse toretain central control over lucrative aspects ofnatural resources. Second, capacity arguments canbe a cover for maintaining the corrupt rent-seekingopportunities for central agency actors. These kindsof excuses are heard in almost every country.

Source: Larson 2002, 6.

61


reforms that have been studied. This issue requiresfurther research.

Capacity arguments often are used as a tool tojustify retaining central control and excluding localpeople from decision making and benefits. Theargument that powers cannot be devolved without thecapacity to receive them is common and oftenspecious (Conyers 1990, 30; Clauzel 1995, 49; Ribot1996; Oyugi 2000, 10). While it is clear that capaci-ties matter, the argument by technocrats that localpeople lack capacity to manage forests is a keyobstacle to decentralization. The forest services ofmany countries—Burkina Faso, Indonesia, Mali,Nicaragua, Niger, Senegal, and Uganda—have usedthis argument merely to maintain central control(Ribot 1995a, 1995b, 1995c, 1999a; Bazaara 2002b,4; Larson 2002, 6; Resosudarmo 2002, 6–7; Bazaara2003, 57; Larson 2003, 3) (see Box 15).70

In the vast majority of cases, a lack of capacity is notthe real reason for the failings of local governments. Inmost of the cases in which local people fail to meetcertain objectives, they fail for one or more of thefollowing reasons. First, the objectives may not beparticularly important to local people. They may be theobjectives of central government or other outsidegroups. In this case, capacity is not the issue—theissue is that central authorities have not respected localpriorities. A second reason that local people may fail tomeet certain objectives is that local authorities andpeople lack funds and materials to achieve the objec-tives. Again, capacity is not the issue—the issue is thatcomplementary resources need to be supplied. And athird reason that local people may fail to meet certainobjectives is that the standards or tasks required bygovernment to execute the objectives are unnecessarilytoo complex for the skills of local people. In this case,capacity may not be the issue either—the technicalhurdles need to be rethought: are they justifiably highor are they set high simply to provide an excuse not totransfer powers to the local level?

Many hurdles are erected that prevent local peoplefrom managing and using resources. In Bolivia,municipalities have the right to exploit twenty percentof the forested domain. However, because municipali-ties do not have the means to develop management

plans, these areas are allocated to others by theenvironmental service. The fact that local authoritieshave not been able to fulfill the conditions stipulatedin the regulations justifies allowing outside producersto clear-cut local forests. (Pacheco 2002, 11.) Thissame strategy for overriding local initiative is alsoused by Senegal’s and Cameroon’s forest services (seeBox 13, above). Because these planning approachesare more complex than necessary and much simplermethods could be used, it is the complexity thatneeds to be reduced, and perhaps not the capacitythat needs increasing.

Uganda provides other examples in which centralgovernment has erected large hurdles for localgovernments to overcome. The 1961–62 annualreport of Uganda’s forestry department states, “Theadministration of crown forests (i.e., central forestreserves) shall be transferred to local governmentswhen the central government is satisfied that the localgovernments have the resources and staff to adminis-ter them properly. Meanwhile, they will continue tobe administered by the central government.” Bazaarapoints out that administration was never transferredto local government. The 1963–64 report states that“no action was taken to transfer administration ofcentral forest reserves in Buganda or elsewhere to theappropriate local authorities because it is still beingconsidered that they do not have the money and staffwith which to manage these important resources”(Bazaara 2001, 33).

In natural resource management in Africa, it hasbeen a long-standing practice of governments todescribe local populations as ignorant and unable tomanage their resources. The colonial foresters werequite clear about this. French forester André Bertin(1919, 125–7) compared Africans to “children ten totwelve years old” and advised that “the employermust not be a distant boss but must behave with hisworkers like the father of the family.”71 Unfortunately,these attitudes persist. In 1994 Mali’s central plan-ning unit responsible for coordination of environ-mental activities reported that “the state shouldbehave as a good father who assists, advises, andcontrols” (Degnobl 1995, 10).

62


A related ideological parallel is technocratic pater-nalism, which is rooted in these same infantilizingattitudes. In 2002, a Malian forest service technicaladvisor said, “We cannot transfer responsibilities tothe blind”; when asked who is blind, he responded,“The illiterate elected councilors and the villagers whoelected them” (author was present at this meeting inBamako in 2002). In participatory forestry, forestvillagers are all too often seen as ignorant, land-hungry peasants, lacking the “capacity” to make but afew highly monitored decisions over the dispositionof forests, which they would destroy if not con-strained (see Blaikie 1985; Guha 1990; Peluso 1992;Thomson 1995, 3; Fairhead and Leach 1996). As inthe colonial period, forests are seen as being threat-ened by the actions of unregulated local populations(Hubert 1920, 421–2, 462–3; Delevoy 1923, 471;Aubréville 1939, 486–7). Foresters have consistentlyargued that their “expertise” is needed to control thisdangerous situation.

The idea that local people are ignorant and lackcapacity, however, has been well challenged over thepast two decades (see Chambers 1983; Moorhead1989; Able and Blaikie 1990; Ostrom 1990; Peluso1992; Fairhead and Leach 1996, among many others).Farvar (1997, 4) argues that “local communities, ifassisted in suitable ways and trusted, can ensure thesustainable use and development of local resources.”They are capable of “identifying their own problemsand needs, analyzing and categorizing them, andidentifying priorities. They are also often capable ofprogramming and designing projects.” Farvar providesexamples from around the world where local commu-nities demonstrate natural resource managementcapacities. Indeed, there are many cases that demon-strate that local communities have more managementcapacities than are attributed to them.

In Mali, for example, Kassibo (in Latif 2002, 62)argues that the state gives out permits to harvestforests but no decision-making powers because theybelieve that the local people will put the forests atrisk. State actors (officials and agents) are reluctant tobuild the capacity of local elected bodies because theyare afraid of the enormity of the value they may lose iflocal bodies become empowered. However, theelected rural council in Mali’s Baye Forest began

protecting surrounding forests when it heard that thenew decentralization laws would eventually allow thecouncil to manage the forest (Kassibo 2002a). In thiscase a local community took the initiative and wasable to protect a forest without outside assistance.Promises of decentralization created empoweringlocal expectations.

In an example from the early 1990s, the ruralcommunity of Makacoulibantang in the Tambacoundaregion of Senegal proposed in a meeting with theforest service a set of measures to protect forests thatthe forest service was allocating to outside merchantsfor exploitation. The rural council and some chiefs hadheard that the coming decentralization reforms were toallow them to manage their forests. The rural popula-tion requested the following: protection of the Forest ofBokko for village usufructuary and pastoral uses;protection of a band along the Gambian frontier toprevent border conflicts; channeling of woodcuttersinto specified areas; more reforestation; and benefitsfor the rural population. Mixed in with these manage-ment proposals, the local people expressed strongdisapproval of commercial production in their region.The people also expressed disapproval that the authori-ties were not taking their wishes into account (Ribot2000). The people of Makacoulibantang clearlydemonstrated their capacity to plan, but to date theforest service has not permitted them to take over theseaspects of management.

Management plans, controlled by forest servicesbased on arguments that local capacity is lacking,often are used more to deprive local users of theresource than they are to conserve the resource. Forexample, in the Tambacounda region of Senegal, theforest service continues to allocate charcoal produc-tion permits to merchants without requiring them touse management plans. Meanwhile, locals, who feelthat the charcoal industry is damaging and would liketo see production stopped, are not permitted toengage in production on the grounds that manage-ment plans are necessary (see Box 16) (Ribot 1995a,1998). This kind of double talk and double standardillustrates how management plans are used to allowsome actors to benefit from a commercially lucrativeresource while depriving others of the opportunity.These double standards give forest services patronage

63


resources that they can allocate selectively to theirwealthy urban merchant clients.

Ironically, the capacity of central governments tomanage and protect conservation areas is ofteninadequate. The numbers of personnel are insuffi-cient to monitor and enforce policies over vast areaswith difficult terrain; in addition, central governmentemployees often lack basic equipment such asvehicles (Resosudarmo 2002, 15). According toOnyach-Olaa and Porter (2000, 3), “most often it isargued that until there has been a marked improve-ment in local government capacities and institutionalpractices, they will fail to respond to needs expressedby their constituencies, will tend to neglect nationalpolicy priorities, and, in the main, behave in unac-countable and inefficient ways.” But, they continue,important research in Uganda has made it “increas-ingly evident that local government performance is

greatly dependent on and is actually being con-strained by inability of central government agenciesand their donor partners to deliver on their mandatedresponsibilities.” In short, local government perfor-mance may be more a function of central governmentand donor accountability than local capacity per se.

Conyers (1990, 30) also points out that “it is fre-quently argued that decentralization should not takeplace until the necessary capacity exists; but this tendsto be a ‘chicken and egg’ type of argument, since moreoften than not it is only the pressure of decentraliza-tion which motivates the action necessary to improvecapacity—and motivates the existing staff at the locallevel to recognize their own potential and demonstratetheir real abilities.” Conyers (1990, 30) and Fiszbein(1997a) suggest that the relation between the ability toreceive power and local capacity is not unilinear. Theyalso argue that the “capacity” argument is often evoked

BOX 16 DOUBLE STANDARDS OR MINIMUM STANDARDS IN SENEGAL

Across Africa rural communities are held to adifferent standard of forest management than areurban-based commercial forest users. Ruralcommunities are required to create complexmanagement plans before they can use and benefitfrom forests. Commercial users are given easyaccess. This double standard creates an unevenplaying field on which local villagers aredisadvantaged and commercial merchants arefavored.

In Senegal rural communities have the right tomanage their own forest resources. But to do so,they are required to develop complex forestmanagement plans. Most rural communities inSenegal lack the skills to develop such plans.Meanwhile, large-scale commercial charcoalproducers are given licenses, quotas, and permits toconduct exploitation with absolutely nomanagement plans required. Agents in Senegal’snational forest service argue that management plansare needed to protect the forests from the localpeople; yet local people believe the real threat to the

forests are the large commercial producers. Indeed,the local opposition to commercial production isintense. It has led to violence in several areas wherethe forest service is giving exploitation licenses tocommercial charcoal makers. In a recent incident inthe rural community of Misira, several local villagerswho wanted to stop the charcoal production werethrown in jail for attacking the charcoal makers.

The high management planning hurdles for villagersand easy access for merchants is a double standardapplied by the forest service in Senegal. Ruralpopulations are not being allowed to benefit from theirforests, while urban-based forestry merchants engagein lucrative production. Unfortunately, this is what“decentralization” amounts to in Senegal’s forestrysector. Uniform minimum environmental standardswould remove the double standard, leveling theplaying field and decentralizing meaningfulopportunities to rural communities.

Source: Field interviews and analysis of forestry code byauthor, Senegal, April 2003.

64


to avoid transferring powers or to avoid reducing thedegree of tutelle or oversight.

Using his research in Colombia, Fiszbein (1997a, 1)shows that “what appears to some analysts and policymakers as lack of capacity, might in fact be thereflection of a conflict in the objective function used,on the one hand, by those analysts/policy makersand, on the other hand, by the local people.” He

attributes what is often perceived as “lack of capacity”to this conflict between national and local prefer-ences. He reports, “Many of those local governmentsmight have unusual or perverse preferences—at leastfrom the national perspective—but they sure had nolack of capacity to achieve their objectives” (1997, 3).

A lack of capacity should not be conflated with a lackof motivation and a lack of interest in central priorities.

BOX 17 PERVERSE INCENTIVES IN UGANDA

While 1998 legislation transferred the managementof reserves under 100 hectares to local councils, thisresponsibility came with insufficient funding. Localcouncils receive only forty percent of the feescollected from those reserves. Sixty percent goes tothe central forestry department. Because the fortypercent is not sufficient for them to train and hiretheir own officers, the local governments inMukono, Mbale, and Masindi are compelled todepend on the “forestry officials employed by andaccountable to the central government.”

But the forestry department expected the localcouncils to pay for their services from these samelimited revenues. Because they felt that the divisionof revenues was skewed toward central government,the local council in Mbale refused to pay the forestrydepartment for assistance. In the end, themanagement of the reserves suffered.

In addition, local governments are expected totransport confiscated illegally harvested timber toKampala for auction by the forestry department. Theforestry service neither pays the cost of transport(vehicles, fuel, or personnel), nor do the local councilsreceive any of the sale revenues. According toBazaara, the local councils “regard this policy asanother reason they could not carry outresponsibilities such as environmental improvement.”

Of the three districts studied in Uganda, only twomanaged to set up environmental institutions, suchas district or subcounty environmental committees,

required by the decentralization laws. Mukono,which was one of the earlier districts to bedecentralized, has not established environmentalinstitutions. Mukono does not even have sufficientfunds to pay allowances for committee members.Masindi and Mbale were able to establish thesecommittees because they received funding from theNational Environmental Management Authority.Bazaara concludes that “where there has been noexternal financial support, districts have notestablished the institutions stipulated in the law,”putting “into question the long-term sustainabilityof the various environmental sensitization exercisesand programs for improving the environmentcurrently being implemented using donor funds.”

Council members in Uganda are not investing innatural resource management for a couple ofreasons. First, while council members are mandatedto maintain and improve the environment, otherpriorities, including their own allowances anddevelopment projects, use up available funds.Secondly and most importantly, elected officials facea conflict of interest between the needs of localpeople and the mandates handed down by theforestry department. “Councilors want to remain inthe good books of their electorate,” writes Bazaara.“Therefore, their main interest is in ‘development’or ‘poverty-alleviation’ oriented projects thatenhance their possibility of being re-elected comethe next round of elections.”

Source: Bazaara 2003, 25–8.

65


In some cases, local populations do not carry outenvironmental management and other functionsbecause these are not high priorities for them(Engberg-Pedersen 1995, 2; Bazaara 2002b). This doesnot mean they cannot perform these functions or lackthe capacity. It simply means they have other objectivesto which to apply their time and skills. Decentraliza-tion should be structured so that the instrumentalobjectives of the center are not being transferred justso that local authorities can implement them. Thiskind of responsibility transfer amounts to delegationand at times corvée (see Ribot 1995a). The language inMali recently changed from transfer of “competence”(which means “jurisdiction” in French) to transfer of“missions” (Bamako 2002). The question is “Whosemission?”72 By transferring central missions orresponsibilities to local authorities, central govern-ments are using elected local bodies to implementcentral agendas. This tension between central and localagendas is persistent. In any event, lack of implemen-tation of external agendas should not be taken toindicate the level of local capacity.

Fiszbein further points out that the perception ofpoor capacity is fostered by poorly designed incen-tives for local government or “an inadequate designof incentives in a principle-agent relationship”(1997a, 3). For example: “When fully accounted, thecombination of earmarking and unfunded mandatesrepresented for many municipalities [in Colombia]more than 100 percent of the untied portion of theautomatic inter-governmental grant they werereceiving. . . . Thus, the observation that few munici-palities were complying with those mandates . . . wasmore a reflection on the absurdity of the policy thanon the local capacity” (1997, 3). (Also see Box 17.)

Members of the Colombian congress have arguedthat “no real benefit would be derived from transfer-

ring funds and responsibilities to local governmentsif their lack of capacity would not allow them tomanage them effectively” (Fiszbein 1997a, 1). Asopponents, however, have argued, these capacitieswill only develop “if fiscal resources and responsibili-ties for service delivery [are] transferred to localgovernments” (1997, 2). The lessons learned inColombia are relevant and help us pose importantresearch questions. To what degree is lack of capacityactually a binding constraint on decentralization? Towhat degree is the capacity question really aboutpolitical struggle over resources?

Complex mandates for planning and managementthat require higher capacity than is available ornecessary, lack of funds to match mandates, and lackof motivation because natural resources are not apriority investment area should not be conflated withlack of capacity. All of these problems—separately orin combination—block or are used to fetter naturalresource management decentralizations. Lack-of-capacity arguments should be viewed with suspicion.Much capacity is needed, but its absence should notbe an excuse for inaction. As the literature suggests,even when local governments genuinely lack capaci-ties that are necessary for natural resource manage-ment, capacity sometimes will follow power—that is,people learn from exercising power (Fiszbein 1997a;Agrawal 2001).

The minimum environmental standards approachdiscussed in the previous section is one way toaddress the technical requirements problem. It helpsidentify the minimum necessary management criteriafor a given set of objectives. It does not, however,resolve the political fight over whose objectivesshould be served, nor does it supply the material orfinancial means for implementation.

66


POWER CHOICES—ASKING THE RIGHT QUESTIONS

When confronted with the choice of powers to decentralize, policy makers and analysts should ask several keyquestions. The following list of questions is based on the observations in this section.

1. Does the transfer of power result in discretion for localdecision makers?

2. Are the powers being transferred meaningful to localpeople?

3. Are the powers being transferred significant enough toengage local people with local decision makers?

4. Are mandates (obligations) being transferred?a. Are those mandates sufficiently funded?b. Are they within the capacity of local authorities toimplement?

5. Are political choices—such as who can use a resourceand who can benefit—being retained at the center ortransferred to local decision makers?

6. Are resources (such as lands, pastures, forests, fisher-ies, etc.) that have been accessible to the public andserve public interest being privatized?

7. Are transfers of power made in a secure manner, or canthey be taken away at the whim of central authorities?

8. Is there appropriate separation of executive andlegislative powers in the local arena and within agenciesof central government?

9. Is there sufficient power—executive, legislative,judicial—in the local arena to balance and fight centralinterests?

10.If exploitation is allowed, are local authorities receivingthe right to determine who exploits the resource?a. Can they decide who has subsistence access to theresource?b. Can they decide who can exploit the resource commer-cially—i.e., can they allocate exploitation rights?

11. Do local authorities have the right to say NO to com-mercial exploitation of local resources—that is, do theyhave the right to conserve the resource?

12. Are the powers transferred well matched with thepolitical-administrative scale and the ecological scale ofthe resource?

13. Is the environmental service using a system thatrequires approval for every decision, or are theredecisions that can be made locally under an environ-mental standards framework?

14. Do environmental laws exclude local communities fromdecision making and benefits by treating communitiesand commercial interests differently?

15. Is it really necessary from an ecological perspective forlocal governments to demonstrate capacity before thepowers are transferred?

67


4

ACCOUNTABILITY ANDINSTITUTIONAL CHOICE

Accountability is an essential element of decentraliza-tion. It is partly a matter of the kinds of institutionsbeing chosen and partly a matter of the kinds ofaccountability mechanisms applied to these institu-tions—although the two cannot be entirely separatedsince institutions differ largely by how and to whomthey are accountable. Accountability mechanisms arenot discussed here—they have been addressed in themain text and are outlined further in Annex C. Thissubsection focuses on institutional choice.

Implementing decentralization means ensuringthat the appropriate local authorities or institutionsare chosen and that they are given sufficient andmeaningful discretionary powers. In some placespowers have been transferred to local institutions. Inothers, democratic institutions have been crafted. Butrarely do governments establish downwardly account-able authorities with powers. Power without account-ability is not representation, and neither is account-ability without powers. Partial arrangements do notestablish democratic decentralization, and they willnot deliver the benefits decentralization promises.Implementation requires powers and accountability.

Despite the importance of giving local authoritiesdiscretionary powers, in so-called decentralizationreforms around the world, public powers are beingtransferred not to local democratic institutions, but tonon-state institutions. This effective privatization ofpublic powers is depriving new democratic institutionsof the powers that—according to the logic of demo-

LESSONS LEARNED

There are many lessons to be drawn fromcurrent research. Only a few of the mostsalient will be discussed here. Others are

highlighted throughout the text and are indicated asrecommendations in Section V. This section focuseson insights related to

● accountability and institutional choice,● subsidiarity principles and the choice and sequenc-

ing of powers,● other factors that affect outcomes, and● resistance.

How to ensure the accountability of institutionsreceiving power transfers, how to select the institu-tions to receive power transfers, how to select thepowers for transfer, and how to sequence powertransfers are design problems associated with thecore elements of decentralization—actors, powers,and accountability. Other factors, such as the legalenvironment in which decentralized powers are to beexercised, the civic education local people receive, thetechnical assistance that local people receive, and theincentives that motivate local people, also are dis-cussed. Since decentralization reform is oftenthwarted by the fierce resistance it encounters, thefinal portion of this section focuses on resistance. Itsummarizes the kinds of resistance decentralizationfaces and suggests ways to overcome this resistance.

Power without accountability is not representation,

and neither is accountability without powers.

68


cratic decentralization—they should have received.Instead of going to and strengthening local democraticauthorities, powers are being exercised through donor-led participatory processes and are being given tocustomary authorities, NGOs, and other quasi-privateor private institutions and individuals.

Powers are being siphoned away from the

representative bodies that could otherwise

become the institutionalized form of popular

participation that the participatory develop-

ment movement was striving for.

The institutional “pluralism” and institutionalproliferation that has grown out of populist andantigovernment civil-society movements is, ironically,creating competition with and undermining thedevelopment of local democratic or elected authori-ties. Local democracy is in crisis. Powers are beingsiphoned away from the representative bodies thatcould otherwise become the institutionalized form ofpopular participation that the participatory develop-ment movement was striving for. The institutionalarrangements that are creating and could help headoff this crisis are discussed here.

Interinstitutional accountability,plurality, and coexistence—Choose democracyIn the local arena there is typically a mix of institu-tions. These institutions can compete or cooperate.They can complement or displace each other. Theexistence of multiple local authorities is not aninherent threat to local representative government,provided these authorities do not undermine localgovernment power and legitimacy. For local institu-tional pluralism to reinforce democratic decentraliza-tion, public resources should be under representativeauthority (see Figure 4). Blair (2000) argues thatother institutions with a role in using or managingpublic resources should be accountable to the publicthrough the local representative authorities. If they

are not, there is likely to be a weakening anddelegitimation of the representative authority by thediffusing of powers over public resources amongmultiple institutions. Meaningful representativeauthority requires powers over public resources. Ifthese powers are diffused across many institutions,representative authority can be undermined. Ifinstitutions with control over public resources are notaccountable to some form of representative authority,then the best-placed and most-powerful authoritiesand institutions are likely to capture power overresources—whether they are chiefs, NGOs, busi-nesses, or user groups. Without the mediation ofrepresentative authority, a plurality of institutionscould be a formula for elite capture.

Without the mediation of representative

authority, a plurality of institutions could be a

formula for elite capture.

Diana Conyers (personal communication, January2004) argues that a plurality of independent institu-tions can be positive because it diffuses the effects ofelite capture. Since institutions are likely to becontrolled by different members of the elite, the moreinstitutions with control over public resources, themore people (patrons and their clients) will havesome access to resources. If all institutions have to beresponsible to a representative authority, she argues,there is a risk that they will all become dominated bythe elite members who control that authority. Shepoints out that the outcome may depend on thenumber and relative influence of the various elitegroups within a particular society. However, protect-ing against elite capture does not necessarilystrengthen and legitimize local elected authorities.Rather than simply asking how elite capture can beavoided, it is of use to ask how local institutions cansupport the development of local representativegovernment so that elite members have moredifficulty dominating public affairs.

Creating or supporting democratic institutions doesnot reduce the plurality of institutions or require theelimination of other parallel or subordinate institu-

69


tions. These institutions coexist with and playimportant roles in helping local government do itsjob. For example, if powers over forest access aretransferred to democratic local institutions, electedbodies can depend on customary authorities toallocate access, provided the customary authorities donot take actions that conflict with the interests ofother segments of local society. There is also noreason that people cannot make decisions or resolveconflicts over forest or land use through the head of ahousehold, a customary authority, a religious leader,or an NGO before or instead of bringing them to theelected local government.73

In addition, the existence of elected authorities orcourts to which people can turn will influence theways in which these other institutions behave. Themere option of appeal to elected authorities or courtsmay increase the likelihood that decisions will bemade fairly by non-elected authorities and mayreduce the likelihood of conflict. It may discipline thehousehold head, customary authority, or imam tomake more equitable decisions and to be more even-handed in arbitrating. The creation of democraticlocal institutions is not meant to eliminate other localauthorities. It is meant to create an alternative thatlocal people can choose.

FIGURE 4 ACCOUNTABILITY RELATIONS MEDIATED BY LOCAL GOVERNMENT

Accountability

Power Transfers

Central Government





Local Populations

Customary Authority

NGO/PVOCBO

In an ideal democratic decentralization, representative local governments have public powers, and their primary accountability is downward to the local population. Other institutions holding public powers are accountable to the representative local government. In this manner, accountability to the people is channeled through representative local authorities. The central government also must be accountable to local government so that local authorities can demand support services such as equipment, information, and technical expertise.

70


However, when these other local institutionsreceive power transfers that should have been givento local representative governments, problems canarise. Transferring power to these other groupsreduces the domain of authority of local democraticinstitutions, forcing them to compete with otherinstitutions for legitimacy. It may make localdemocracy less likely to succeed (Summers 2001 inMansuri and Rao 2003, 2). Local accountability ismediated through representation. Therefore,representative authorities should be the hub of theinstitutional landscape that governs the use ofpublic resources. The choice of local institutions andhow they depend on each other is critical for theestablishment of effective decentralization.

Representative authorities should be the hub of

the institutional landscape that governs the

use of public resources.

Institutional choice for scaling up andsustaining representationBefore power is transferred to an institution, it iscritical to ensure that the institution is capable ofscaling up representation or replicating it on a largerscale; it is also critical to ensure that it can sustainrepresentation over time. Participatory approaches,which have been and continue to be used to includelocal populations in rural development and naturalresource management activities, have severe limita-tions in both regards. Participatory approaches do notwork well at larger scales (Green 1993; Clark 1998),and they often do not survive once the projects thatmobilized them are gone. In addition, they are labor-intensive for both those facilitating them and thoseparticipating in them, and they are often neitherdemocratic nor just (see Cooke and Kothari 2001). Asargued in the introduction of this report, local govern-ment appears more likely than any other institution tohave the characteristics of generalizability and institu-tional stability, which are necessary for scaling up andsustaining representation.

Institutional choice and citizenshipDecentralization is believed to be most efficient andequitable when inclusion in decision making is basedon the whole population. Democratic systems operateon residency-based citizenship. Customary authori-ties, however, base belonging on origin, lineage,ethnicity, or religious identity, and NGOs basebelonging on interest. Identity- and interest-basedinclusion in public decision making can reinforcedifferences and fragment community. These forms ofinclusion may lead to conflict, and they are moreprone to political manipulation. Current conflictsover land in the Ivory Coast have been exacerbated bythe government’s appeal for support to certain groupsbased on their identities of ethnicity or origin, ratherthan residency (Chauveau 1994). Examples acrossAfrica of ethnic conflict are common (Geschiere2003). In Uganda, local professional-based commit-tees caused divisions within communities (Namaraand Nsabagasani 2003). Residency-based citizenshipappears to be the most inclusive of the differentforms of inclusion (Mamdani 1996a).74 For thisreason, democratic authorities whose elections arepredicated on residency-based universal suffragecould be a better choice for representation than aremany other kinds of local institutions. Mechanismsstill must be crafted, however, to ensure representa-tion of the most-marginal and disadvantaged groups.

SUBSIDIARITY: CHOOSING POWERS,SEQUENCING TRANSFERS

Powers must be chosen and transferred. This subsec-tion outlines a few guiding principles that haveemerged from this study for identifying powers totransfer and for sequencing their transfer to localauthorities.

Subsidiarity principlesClear environmental subsidiarity principles will helpguide the choice of powers to allocate to differentlevels of authority in and out of government. Prin-ciples that might be helpful for activists, practitioners,and policy makers are listed and explained in Box 19,located in the recommendations section.

71


Decisions as to which powers can be transferredcannot be made solely by interested parties withinenvironmental ministries. Procedures for determin-ing the optimal division of powers must involvemultiple parties. Capacity arguments and technicalarguments for retaining powers at the center shouldbe weighed with great care. At the center there ismuch fear of losing power. There are also many falsescientific-sounding arguments that can misguidethese decisions.

Sequencing pointsThere is no overall sequence in which a decentraliza-tion should take place. Clearly, it helps to have lawsthat create truly democratic local authorities and thatenable the transfer of significant powers to them. Thefirst step is to analyze laws to see if they are adequate.Where adequate laws do not exist, the first step is towork toward establishing them. Beyond this, thereare several sequencing guidelines that can be derivedfrom the observations and analysis in this report.Most of these concern the transfer of powers. Theyare described below.

Which powers first?The order in which powers are devolved can matter.Powers that are less conflictual should be devolvedbefore those that are more conflictual. Powers thatare less likely to threaten long-term sustainabilityshould be devolved before those that are certain tocause damages if misused. Powers that do not requireany funding should be transferred before powersrequiring funds—if funds are not available. Powersrequiring little local skill should be transferred beforethose requiring high technical or financial manage-ment capacity. Powers that can be transferred withoutrequiring planning should be transferred withoutplanning conditionalities. The two big issues insequencing appear to be capacity and managementplanning requirements.

Power before capacityGovernments are waiting to transfer powers untillocal people demonstrate that they have capacity. This

may sound reasonable, but it is absurd to think thatpeople can develop skills without having the powersto exercise and practice them. Many powers can betransferred before capacity is demonstrated becausetransfer of these powers pose little risk to the re-source (Ribot 1999b). Some powers can be trans-ferred because the risks are worthwhile. Once thepowers are transferred, local institutions may acquirethe capacity (Fiszbein 1997a; Agrawal 2001). The riskthat mistakes will be made must be weighed into thecost of decentralization. Usually those mistakes arenot so grave. Many forests do regenerate with time.Which powers can be transferred will depend onwhether the potential damages are reversible and onthe benefits versus the risks of trying.

A power that local people can exercise without

having any special capacities is the right to

say no to extractive industries.

Standards before planningThe practice of requiring local people to create complexmanagement plans before they are allowed to manageand use local resources is blocking environmentaldecentralizations. Requiring local users to adhere tominimum standards may be a better option thanrequiring them to develop complex managementplans. The minimum standards approach would allowthem to conduct benign activities of extraction and usewithout having to develop plans. For those who desirehigher productivity than natural regeneration affords,guidelines could be available. But requiring multi-stepplans that no local community has the technical orfinancial means to implement fetters local control. Oneexample of a power that local people can exercisewithout having any special capacities is the right to sayNO to extractive industries; in fact, two Africancountries, Senegal and Mali, have already createdlegislation giving local people this power. This does notnecessarily give local authorities any rights to exploitthe resource, but, if enforced, it gives them plenty ofnew power to negotiate and to conserve the resource.This power cannot be argued to threaten the resourceor to require an elaborate plan.

72


Give decentralization timeBefore we judge the impacts of decentralization, weneed to be sure that a real decentralization has beenestablished and that the outcomes we observe can beattributed to it. There is a backlash brewing againstthe decentralization of natural resources, but it is outof sequence. First we need decentralization beforeany reaction—good or bad—is justified. Even if thedecentralization is in place, it needs time to stabilizeand produce outcomes. Decentralizations should notbe judged in their first years.

Most cases studied involve relatively recent reforms.Indonesia’s reform is less than three years old. SouthAfrica’s reforms are five to seven years old. It is tooearly to see the results in either of these cases. Yetbacklashes against environmental decentralizationsare already forming in Uganda and Ghana (Ribot2002b). In both countries, the government claimedthat people were over-exploiting and attacking theforests. They may have been. Over-consumptionimmediately following reform has been observed inIndonesia; Kumaon, India; and several other places.Agrawal argues that this instability is likely untilpeople gain confidence that the decentralized controlover and access to forests is not just temporary. Theyneed to have a sense of security in the laws. Thetendency is to consume while the opportunity is theresince it is not likely to last.

The Asia Group at the Bellagio conference de-scribed a “crisis of expectations” phenomenon inIndonesia, in which people over-exploited the forestresource because they did not have confidence thatthe reforms would last (see Box 18). They thereforetried to take advantage of the exploitation opportuni-ties while they lasted (Latif 2002, 67). Short timehorizons and uncertainty as to whether reforms willlast created opportunism and over-extraction. AtBellagio, Baviskar argued that the Kumaon caseillustrates that “we expect too much too soon fromdecentralization” (Latif 2002, 11). She suggested thatconflict and frustration are part of the nature oftransition. Agrawal then pointed out that Kumaon’sarchives show that in the first years of the reform, theforesters talked of the reforms as if they were tempo-rary and unlikely to succeed. They did not believe the

councils could manage the forests. But a more stableregime emerged gradually and the forest councils arestill active seventy years later. If decentralizations areto unfold, they will need time to stabilize and forpeople to have confidence that the powers theyreceived will stay.

BEYOND ACTORS, POWERS, ANDACCOUNTABILITY: OTHER FACTORSAFFECTING OUTCOMES

There are many factors, aside from those associatedwith the actors, powers, and accountability model,that affect the social and environmental outcomes ofdecentralization reforms. Some of these factors canbe legislated or provided by central government.These include the following: attention to inter- andintra-jurisdictional equity; legal recognition for localorganizations; legal protections for organizing,lobbying, and filing suit; availability of technicalsupport from central government; ability of localgovernments to mobilize technical assistance andequipment that central government has to offer; civiceducation; government education; the right andability to federate local authorities and organizations;and the right of local authorities to control outsideextractive industries. The legislation associated withthese factors is in addition to the legislation that isbasic to creating decentralized authorities, such asappropriate electoral codes and laws that transferexecutive, legislative, and judicial powers.

Some other factors that directly shape environmen-tal outcomes in decentralized arrangements include:the kinds of environmental standards in place, thematch between geographic scale of natural resourcesand the scale of political/managerial jurisdiction,quality of technical assistance from outside NGOs,and non-environmental priorities of local authoritiesand communities. Environmental standards and thescale issue are discussed with powers problems inSection III and in the subsection on resistance, below.Technical assistance from NGOs can clearly be apositive force in improving environmental outcomesand should be encouraged.

73


One issue of particular concern to environmental-ists is the engagement of local authorities in environ-mental issues. To ensure good environmentalmanagement through local authorities, those authori-ties need incentives to become engaged. According toLarson, local authorities can be motivated by thedesire to generate municipal revenue, by a specificconflict or crisis, by pressure from projects or anNGO, or by recognizing a particular problem(Agrawal 2001; Larson 2002, 13–4; Bazaara 2003, 24;Namara and Nsabagasani 2003, 47). Larson foundthat a mix of the first three incentives was common.She claims that in some cases these led to the fourth.This last factor appears to be part of a learningprocess resulting from the other three factors.Agrawal (2001) makes a similar point when heobserves that better management in Indian vanpanchayats came with the transfer of regulatorypowers. These powers induced the development ofenvironmental consciousness and concern. Along thesame lines as Larson, Bazaara (2002a, 24) writes thatthere are four factors that explain why local govern-ments do not exercise the environmental powers theyhold: lack of concern about environmental issues,financial constraints, lack of clear institutionalrelationships, and conflicts of interest.

Local authorities need specific incentives to getinvolved in environmental issues. They have compet-ing problems to attend to. In interviews conducted bythe author in the winter of 2003–04, World Bankproject team leaders reported that local communitiesrarely choose natural resource management projectsif they have a choice of alternative investments. Theirfirst priority is to invest in productive activities thatmeet immediate income needs. For local communi-ties to invest in natural resource management, localleaders must see the interest in it, or they must havesufficient funds to meet their first priorities, as wellas natural resource management needs. Centralgovernments, donors, and NGOs can all play a role increating incentives for local authorities to invest innatural resource management.75 Reasonable environ-mental standards also create an incentive (fear ofsanction) for local leaders and people to bettermanage the environment.

BOX 18 DECENTRALIZING OVER-EXPLOITATION IN INDONESIA—GET IT WHILE IT LASTS

As part of decentralization reforms, Indonesia’sforest service made it possible for localauthorities to allocate small-scale timber licensesto create an incentive to curb illegal logging.Because these local permit holders bear the lossfrom fraud, they were expected to be more likelyto control and report illegal logging. Thisoutcome, however, has not been demonstrated.

In fact, these changes have led to the effectivelegalization of illegal timber harvesting. Ratherthan confiscating and auctioning illegal timber,the district officials are now assessing a fee onthe timber and issuing documentation allowingresale and transport. These practices could havenegative effects on the controllability of theindustry. (Resosudarmo 2002, 13–4.) Peluso(2002, 8) also observes a greatly increased levelof potentially damaging illegal logging due to theemergence of new mechanisms of illegalextraction and the participation of military forcesin such exploitation.

These negative consequences may be associatedwith two shortcomings of Indonesia’sdecentralization reform. First, the localauthorities in Indonesia are elected from partylists and may not be downwardly accountable tothe local population. Peluso points out that italso may just be that some leaders in Indonesiaare using decentralization as an excuse todisobey national law and to make their ownpolicies. In either event, the leaders do notappear accountable to their constituents. Second,the national ministry responsible for forests isalready threatening to re-centralize permitallocation powers. It may be that because localofficials do not believe they will be able to keepthese powers, they are attempting to capitalize onthem while they last.

Sources: Peluso 2002, 7–8; Resosudarmo 2002, 13–4.

74


If democratic decentralization takes place, accompa-nying measures will help ensure that democraticdecentralization will serve not only the interests oflocal communities, but also the long-term and larger-scale concerns of society.

(EN)COUNTERING RESISTANCE TODEMOCRATIC DECENTRALIZATION:SEIZING OPPORTUNITIES

Historically, decentralization initiatives have notenjoyed great success, largely for two reasons: despitetheir rhetoric central governments have often notreally wanted to devolve any real power to the locallevel; and, when significant authority has beendevolved, a disproportionate share of the benefits hasbeen captured by local elites. The new democraticvariant of decentralization, however, promises toovercome these problems by introducing increasedparticipation, accountability, and transparency inlocal governance, along with empowerment forpreviously marginal groups.

Harry Blair 1997

Decentralization advocates were hopeful thatdemocratic decentralization would produce moreequitable and just outcomes than earlier decentraliza-tion reforms had done, but, unfortunately, many of theproblems of earlier reforms are occurring in the newreforms. States are still using decentralization toconcentrate power rather than as a means of redistrib-uting power to a greater portion of society. This sectionsummarizes some of the key fetters on decentraliza-tions (also see Ribot and Oyono forthcoming). As lawsare written and implemented around the world, manyactors who are threatened by decentralization areholding it back. Central actors do so by hanging ontopowers and by choosing local institutions they canmost easily control. Members of the local elite do so bycapturing the powers that are being transferred.

Resistance to the transfer of powersResistance to power transfers is justified through anumber of often specious arguments and practiced

through multiple means. Some of the most commonarguments are summarized here.

Capacity arguments are commonly used as an excusenot to devolve powers to local authorities. By arguingthat local people lack capacity, central governmentshave managed to retain control over many decisionsthat could be transferred to local people.

Means of transfer are commonly used that enablegovernments to retain powers in the executive branch(see Conyers 2002). Rather than transferring powerthrough legislation, legislators are making laws thatinclude “delays,” leaving decisions to ministerial oradministrative decrees and orders. In this way, thetransfer of powers is retained as a discretionarydecision of executive-branch ministers and adminis-trators. When powers are transferred, they areallocated as privileges that can be withdrawn atexecutive will rather than as secure rights for localpeople to invest in.

Failure to create a local domain, or the reduction oflocal jurisdiction, hems in local powers. InCameroon, there is a progressive community forestrylaw, but community forests can only be established ina very limited set of areas determined by the forestservice (Ekoko 2000; Oyono 2002). In Mali, the verystrong decentralization laws that give local councilsconsiderable powers over local forests are under-mined because the central state has refused toestablish a local domain in which to exercise them(Ribot 1995a; Kassibo 2002a). In Uganda,privatization and reserves leave few forests underlocal government control (Muhereza 2003, 8-11).

“Scientistical” arguments—specious, technical-sounding arguments—are another common meansof retaining central control over natural resources.Even though we now know that there are manyenvironmental use powers that can be transferredwithout threat to the resource base—and evenwithout any need for enhanced capacities (see Ribot1999b; Fairhead and Leach 1996), forest servicessystematically refuse to acknowledge this. Further,these kinds of arguments have led to the establish-ment of overly complex management-planning

75


requirements that have often made it impossible forlocal communities to use or manage surroundingresources with any degree of independence.

Choice of powers being transferred is another way toresist. Burdens and mandates are being transferredwithout or in place of positive powers, and non-commercially valuable uses are transferred in lieu oflucrative opportunities. In Uganda, Bazaara (2002a)observed that central authorities were treating techni-cal and political decisions alike, and since they couldargue that some technical decisions are more appropri-ately made at the central level, they were retainingpower over political decisions, as well as technicaldecisions. Bazaara argues that acting as if commercialallocation decisions are technical decisions for theforest service to make hides the fact that the centralagencies are reserving for themselves what are ulti-mately political choices. In addition, central authoritiesoften conflate transferring the instrumental objectivesof the state with transferring discretionary powers.There is a big difference between the two: discretion-ary powers give local authorities real power, while theinstrumental objectives of the state saddle local govern-ment with central mandates, turning them into effectiveadministrative branches of central government.

Local institutions differ in the degree to which

they represent and are accountable to local

populations; therefore, the degree to which

transfers to them can be considered decentrali-

zation differs.

Excessive oversight and overbearing approval processesare used almost universally to reduce local discretion.Examples include management committees presidedover by foresters, management planning that requiresforest service approval before any exploitation canoccur, and the withholding of transfers until certaincapacity criteria are met as judged by central authori-ties. Communities are told that they will get powerswhen their plans are approved or when they demon-strate capacities defined by the technical services.

These are the most common ways in which powertransfers are restricted, stopped, or reversed. Theresult is that decentralizations are rarely creating thespace for discretion, which local democracies musthave to be meaningful.

Institutional choices as a form ofresistanceChoosing the appropriate local institutions is criticalfor establishing decentralization. Local institutionsdiffer in the degree to which they represent and areaccountable to local populations; therefore, the degreeto which transfers to them can be considered decen-tralization differs. Even when elected local institutionsexist, less-democratic institutions are being chosen.This undermines decentralization. Governments anddonors are often choosing deconcentrated bodies,customary authorities, NGOs, elected and appointedlocal committees, or private corporations and individu-als in lieu of elected local bodies. These choicesattenuate the effects of decentralization by takingpowers that would otherwise have been given toelected local institutions. These choices create competi-tion for local democracy—competition that localdemocracy may not be capable of standing up to.

Institutional choice is also important because powerimparts legitimacy to those who acquire it, and powercreates capacity in those who acquire it. Electedbodies without powers will never be legitimate.Governments often use the argument that they aregiving powers to customary authorities becausecustomary leaders are legitimate and the electedbodies are not. In addition, they often allocate powersto administrative bodies or to committees of theirown creation because they argue that the electedbodies lack capacity to manage resources and conflict.These choices defeat the democratic aspect ofdecentralization and prevent democratic authoritiesfrom functioning. Central authorities often refuse toacknowledge that local representative authoritiescannot acquire legitimacy and capacity if they are notgiven any powers. Instead, they use a lack of legiti-macy or capacity as justification for avoiding placingsecure discretionary powers in the hands of indepen-

76


dent local authorities. These tactics are used to fetterthe transfer of powers to democratic local institutionsthat may not be easily controlled by central minis-tries.

Resistance by local actorsElite capture, which has been common in decentrali-zations, may attenuate the expected positive out-comes of decentralization. This phenomenon, inwhich members of the local elite capture the benefitsof decentralization reform, cannot be separated fromthe institutional choices and patronage relationsbeing cultivated by central actors. In other words,members of the local elite are not alone responsiblefor elite capture. Central actors often encourage it fortheir own purposes. The prevalence of elite capturereinforces the need for multiple accountabilitymechanisms and well-structured accountabilityrelations (see Annex C for a list of such mechanisms).

Countering resistanceResistance to decentralization can easily be seen inthe arguments that are used to justify not transferringpowers and not choosing or creating representativelocal institutions. To counter this resistance, thesearguments must be met with counterarguments.These counterarguments can become tools forconstituencies to understand and then challenge thediscourses and practices of obfuscation and obstruc-tion. The dominant arguments used against decen-tralization and some counterarguments that canchallenge them are outlined here. These are followedby brief discussions of education and voice asadditional means to counter resistance.

Countering arguments1. Argument: Transferring natural resource manage-

ment powers to local people leads to environmen-tal destruction.Counterargument: Many fears of environmentaldestruction are unfounded. Environmentalsubsidiarity principles based on grounded research

can help identify the powers that do not threatenlong-term sustainable use of natural resources.Further, local people should have the right to useand change the resources around them. Destruc-tion to one observer may be a positive change toanother. What is and is not environmental destruc-tion should be a point of interpretation andpolitical negotiation.

2. Argument: Transferring natural resource manage-ment powers to local people will incite conflict andviolence among competing interests, and localauthorities will be unable to control the violence.Counterargument: Some power transfers carry littlerisk of sparking conflict. It is imperative todetermine when conflict arguments are real andwhen they are excuses not to transfer powers. Thepowers that can be transferred without creatingconflict should be identified and transferred. Asfor power transfers that might result in conflict,central authorities might accompany thesetransfers with mediation devices to prevent conflictfrom becoming violent.

3. Argument: Local authorities lack the capacity tomanage natural resources.Counterargument: This argument is often specious.There are many examples in the environmentalsector, as well as other sectors, that reveal how thisargument has been used as an excuse not totransfer power over lucrative resources. Manypowers can be transferred that do not require anyspecial local skills. Many skills that are ostensiblyneeded can be hired—not everyone need be aforester to cut down trees or to decide who shouldbe involved in forest management (no more thanone needs to be a mechanic to drive a car).

4. Argument: Local authorities have insufficient fiscalresources to manage their natural resources.Counterargument: Many powers over naturalresource management and use do not requirefiscal resources. These powers could be identifiedand transferred first. If central governmentauthorities are really as concerned about manage-ment activities as they claim to be, they cantransfer the fiscal resources to local governmentsto carry them out. If central government does nothave these resources, then this argument is anabsurd basis for recentralizing.

77


5. Argument: Customary or traditional authorities aremore legitimate than local government, and custom-ary authorities, NGOs, and other local institutions aremore efficient than local government.Counterargument: Legitimacy follows power.Customary authorities may be legitimate if theyare given powers to yield. However, their legiti-macy does not necessarily mean that people likethem. It may derive from fear or just from habit.In addition, central authorities often are lessinterested in transferring powers to local institu-tions that are legitimate and efficient than they arein transferring powers to institutions that are likelyto do the bidding of central agents. In short, thisargument is often used as an excuse to make aninstitutional choice that would facilitate elitecapture or external manipulation.

Armed with well-grounded arguments, constituen-cies can organize to hold government accountable forthe pronouncements they make in their decentraliza-tion discourses. These local, national, and interna-tional constituencies can organize to help ensure thatdemocratic decentralization will be codified in lawand established in practice.

Civic and government educationTo counter local resistance—that is, the resistancethat comes in the local arena after the laws are passedand implementation has begun—people need toknow their rights (if and when they have any), andthey need to know the obligations that governmenthas toward them (Shrestha 1999, 34). Civic educationcan help. In addition to educating local constituents,it is also necessary to educate local governmentrepresentatives about their powers and their obliga-tions to local people. Providing civic and governmenteducation may involve publishing manuals written inlocal languages that explain the laws.

Awareness of rights can create popular demand formore-responsive government or for nationallyrecognized local rights. In Mali’s decentralization, thelaws require the government forestry agency totransfer powers to newly elected rural councils. Theagency has refused to do so. However, local councilsin the Forest of Baye, having heard that they have

rights over the forest, its management, and use,began to organize forest protection in anticipation ofthe formal transfer of powers. Dupar (in Latif 2002,78) points out that in Cambodia, donor agenciessupported a legal literacy campaign that emboldenedlocal people, especially women. When rights exist,civic education can encourage people to exercisethose rights as full citizens.

Informing people of their rights, explaining newlaws in clear, accessible language, and translatingcivic educational materials into local languages canencourage citizen engagement and local governmentresponsibility. It can encourage citizens to demandthat government serve them. Central government,donors, and NGOs can support these local civiceducational efforts.

Multiple channels for local voice: Federationsand legislative representationThe weakness of decentralization in the rural areas—which is where natural resources are located—reflectsthe weaknesses and marginality of rural people in thedeveloping world. Rural people have rarely had astrong voice in the making and implementation ofnational policy. They have rarely had control overtheir own local affairs. Democratic decentralizationrepresents a radical change. If it is to succeed, it willneed a broad constituency with leadership that isresponsive to the needs of rural people. Creatingdemocratic local institutions is not enough. Multiplechannels of voice and recourse are needed so thatnational policies favor rural needs and so that there ispressure to implement these policies. In a democraticsystem where influence on government is possibleand representatives are accountable and responsive,multiple voices and channels of influence can haveleverage.

Associations of local representatives and federationsof peasant organizations and farmers unions may beproductive channels. Federations of NGOs and oflocal governments already are emerging in the studycountries. In Mali and Senegal, there are nationalassociations of local elected representatives. InBotswana, the Botswana Community-BasedOrganisation Network (BOCOBONET) is a federation

78


of representatives of community-based organizations(CBOs). BOCOBONET facilitates informationexchange, advocates to government for CBOs,coordinates capacity building and training of mem-bers, and ensures that technical services are deliveredto the organizations of its members (BotswanaCommunity-Based Organisation Network 2003).

Deputies and members of parliament are alsopotential channels for local voice in national policymaking. Veit et al. (forthcoming) argue that fourfactors determine whether deputies and members ofparliament will represent the interests of local popula-tions. These factors are: downward accountability,

autonomy from the executive to perform their environ-mental representation responsibilities, authority orpower to fulfill their representation roles, and ambi-tion. While most parliamentary representatives do notattend to the environmental concerns of their constitu-encies, there may be measures that can be taken toencourage representatives to be more responsive tothese concerns. Courts may be another channel of localinfluence up the political-administrative hierarchy.

In this set of studies, federations, legislativerepresentation, and courts were not explored. How-ever, they are worth exploring as a way of enhancingthe voice of local people in the national arena.

79


5

1. Work through elected local government to scale up andinstitutionalize popular participation. Governments,donors, and NGOs who wish to create sustainableparticipatory local institutions that operate locallyacross whole national territories should workthrough elected local government.

2. Foster downward accountability of all local authori-ties. Governments, donors, and NGOs can fosterlocal accountability by● choosing to work with and build on elected local

governments where they exist,● insisting on and encouraging their creation

elsewhere,● encouraging electoral processes that admit

independent candidates since most do not,● insisting that all administrative and private

institutions and authorities dealing with publicresources in the local arena are accountable toelected local government, and

● applying multiple accountability measures to allinstitutions making public decisions.

3. Ensure that NGOs and traditional authorities areaccountable through democratic authorities. Everyoneinvolved in the design and implementation ofdecentralizations should insist on democratic linesof accountability. For decentralization to beeffective, there must be a clear line of accountabil-ity from those who make decisions over publicresources to local people. On public decision-making matters, nongovernment and traditionalinstitutions must be accountable to representative

RECOMMENDATIONS

The most general recommendations of thisreport can be summed up as follows: 1) Designand implement reforms that merit the name of

democratic decentralization; 2) If these reforms areestablished, study the positive outcomes to identifythe constellation of measures and circumstances thatmake those outcomes possible.

Two sets of recommendations are presented below.The first recommendations concern the establish-ment of democratic decentralizations writ large.Democratic decentralization establishes the necessaryinfrastructure for decentralizing natural resourcepowers. The second set of recommendations is aimedat improving natural resource decentralizations.These lists, however, are by no means exhaustive.Many more recommendations are implied by the casematerial discussion in this report.

ESTABLISH THE INSTITUTIONALINFRASTRUCTURE OF DEMOCRATICDECENTRALIZATION

IF the institutional arrangements of democraticdecentralization are established, THEN we can testwhether decentralization will lead to improvedoutcomes and under what conditions. Establishingdecentralization is key. The institutional infrastruc-ture of democratic decentralization is also thenecessary infrastructure for the decentralization ofnatural resource management. To establish thisinfrastructure, the following recommendations maybe helpful.

80


authorities who are accountable to the people(Blair 2000).76

4. Base inclusion on residency rather than identity.Those involved in decentralization should workwith elected local authorities because theseauthorities generally represent people through asystem of residency-based citizenship, rather thanother more-exclusive forms of belonging. Basingrepresentation and inclusion on whole popula-tions, as residency does, is believed to producemore efficient and equitable decentralizations.Customary authorities base belonging on origin,lineage, ethnicity, or religious identity, and NGOsbase belonging on interest. Identity- and interest-based forms of inclusion in public decisionmaking can reinforce differences and fragmentcommunity, sometimes leading to conflict and agreater susceptibility to political manipulation.Representing citizens based on residency—provided the special needs of mobile populationsare taken into account—may be more inclusive.

5. Privilege procedural over instrumental objectives. Toestablish democratic and inclusive processes in thelocal arena, some of the instrumental objectives ofenvironmental management may have to besubordinated to the procedural objectives ofinclusion. Favoring democratic process in theshort run will help build the institutions that in thelong run could serve to make and implement justand sustainable policies. Environmentalists shouldsupport the procedural objectives of institutionaliz-ing democratic and representative authorities.

6. Transfer sufficient and appropriate discretionarypowers. Governments, donors, and NGOs musttransfer significant and meaningful discretion tolocal authorities. Discretionary powers must betransferred if decentralization is to interest localpeople, emancipate them, and empower localauthorities.

7. Transfer powers as secure rights. So that localinstitutions and people will invest in new arrange-ments and so that local people can be enfranchisedas citizens rather than managed as subjects,

governments should use secure means to transferpowers to local authorities. Secure transfers cancreate the space for local people to engage theirrepresentatives as citizens. Transfers made asprivileges subject local people to the whims of theallocating agencies and authorities.

8. Support equity and justice. Central governmentintervention may be needed for redressing inequi-ties and preventing elite capture of public decision-making processes. Central government also mustestablish a legal environment that enables people toorganize, appeal to their representatives, demandtheir rights, and take recourse when needed.

9. Establish fair and accessible adjudication. Govern-ments should establish accessible independentcourts, channels of appeal outside of the govern-ment agencies involved in natural resourcemanagement, and local dispute resolution mecha-nisms. Donors and NGOs can also supportalternative adjudication mechanisms to supple-ment official channels—not to replace them.

10.Support local civic education. Governments, donors,and NGOs can inform people of their rights, writelaws in clear and accessible language, and translatelegal texts into local languages to encouragepopular engagement and local governmentresponsibility. When there are meaningful rights,it is critical for people to know them. Educatinglocal authorities of their rights and responsibilitiesalso can foster responsible local governance.

11. Support multiple forms of local representation in thenational arena. Governments, donors, and largerNGOs can support organizations such as● national associations of local elected authorities;● federations of farmers, fishers, forest dwellers,

and pastoralists, as well as federations of localNGOs;

● farmer and peasant unions; and● local representatives in national assemblies and

parliaments.These groups and individuals can help to representthe needs and protect the rights of local popula-tions in the formulation of national policies—

81


BOX 19 PROPOSED ENVIRONMENTAL SUBSIDIARITY PRINCIPLES

The research suggests some principles ofsubsidiarity to guide the transfer of power. Manyothers might be appropriate. Subsidiarity principlescan be helpful, but ultimately where powers arelocated within society is a social and politicaldecision. The following are some subsidiarityprinciples designed to enhance local enfranchisementin control over natural resources:

● Create discretion: Transfer discretionary powers tolocal authorities to give them someindependence. Without discretion, localauthorities cannot be democratic.

● Provide significant powers: Devolve powers such ascontrol over land and forest access. These powersare meaningful to local people and, as a result,they reinforce the authority of those who holdthem.

● Fund mandates: Provide sufficient fiscal resourcesand technical support to fulfill mandates. Inaddition, mandates should never be the onlypowers transferred to local authorities.

● Transfer funds and fund-raising powers: Earmarkfunds for local authorities and transfer revenue-raising rights to local authorities. Transferpowers to borrow, to tax, to charge fees, etc.

● Transfer lucrative opportunities: Transfercommercially valuable resource-use opportunitiesto local authorities, as well as subsistence-oriented usufruct rights.

● Do not conflate technical with political decisions: Donot conflate technical decisions, such as whichspecies to protect or cut, with political decisions,such as who should have access to resources.

Some technical decisions need to be made at thecentral level, whereas political decisionsconcerning use of resources may be made at thelocal level.

● Maintain the public domain: Keep publicresources—including most forests, fisheries, andpastures—within the public sector. They shouldnot be privatized.

● Ensure security: Transfer powers to representativelocal authorities as secure rights and not asretractable privileges.

● Separate powers: Separate and balance executive,legislative, and judicial powers at each level ofgovernment.

● Balance powers in government: Balance powersgiven to each level of government with those ofother levels so that each level of authority has theability to use its powers and to negotiate withother levels of authority.

● Balance powers over commercial resources: Givelocal authorities the power to restrict the accessof outside industries to local resources. Outsidershave an exit option and therefore do not treat theresource with respect (Bardhan et al.).

● Match powers to scales: Take into account the scaleof a resource with respect to the financial andtechnical resources available to manage thatresource. Transfer powers to the most-local leveland encourage those local jurisdictions to formfederations to manage resources they share.Creating federations to manage resources thatspan multiple local districts may be preferable tocreating special districts or giving powers tohigher levels of authority.

82


concerning such critical matters as the setting ofminimum standards and the selection of powers totransfer to local governments. These coordinatedvoices help to redress the gross power imbalancebetween national and local actors.

12.Give decentralization time. Judge decentralizationonly after it has been tried. Give it sufficient timeto stabilize and bear fruit.

DECENTRALIZE THE MANAGEMENTOF NATURAL RESOURCES

The following recommendations are more specific tothe decentralization of natural resource managementand use.

1. Develop environmental subsidiarity principles.Governments, donors, NGOs, and the researchcommunity should work to develop environmentalsubsidiarity principles (see Box 19) to guide thetransfer of appropriate and sufficient powers tolocal authorities.

2. Establish minimum environmental standards.Governments should shift from the management-planning approach to the minimum-environmen-tal-standards approach. Broad minimum standardscan facilitate ecologically sound independent localdecision making without saddling local peoplewith the burden of creating and implementingelaborate management plans. Establishing just andsound standards will help balance efforts to protectthe environment with the needs and aspirations oflocal people. To establish truly minimum stan-dards, however, will require an analytic review ofthe poorly evidenced environmental orthodoxy thatpromotes over-regulation and over-protection ofthe environment and of natural resources.

3. Remove double standards in forest management.Foresters and environmentalists must apply manage-ment standards and requirements uniformly. Localcommunities typically are required to produce more-detailed management plans and to conduct reforesta-tion that is more rigorous and labor-intensive thancommercial industry is required to do. They are alsomore closely monitored.

4. Transfer to local authorities the right to say NO toextractive activities. New environmental legislationmust give local communities veto power overresource extraction. This right does not threatenthe resource. It is a simple way of giving localpopulations access control. It also gives themleverage to negotiate with outsiders.

5. Develop indicators for monitoring and evaluatingdecentralization and its outcomes. Donors and NGOsshould develop and monitor indicators of progressfor decentralization legislation, implementation,and outcomes to provide needed feedback to keepdecentralization experiments on track.

6. Sequence decentralization of natural resourcesstrategically. Policy makers and practitioners cansequence decentralization to minimize conflictand to avoid excuses to block the transfer ofpowers. Consider the following proposed sequenc-ing principles:● Establish representative and/or downwardly

accountable authorities first.● Transfer political decisions concerning the

allocation of access and benefits before transfer-ring highly technical decisions concerning howto manage resources.

● Transfer decisions and powers that do not threatento cause violence or environmental damage beforetransferring more problematic powers.

● Transfer decisions that do not require financebefore transferring those that do.

● Transfer finance and revenue-raising powersbefore transferring management obligations.

● Transfer sufficient finance with any centrallymandated obligations.

● Transfer powers before requiring capacity.● Establish minimum environmental standards

before requiring management planning.

COUNTER RESISTANCE ANDSEIZE OPPORTUNITY

Two strategies may help to implement the aboverecommendations and to move the decentralizationexperiment forward: countering resistance andseizing opportunity.

83


1. Resistance to decentralization in the naturalresources sectors needs to be challenged byeveryone with an interest in decentralization.Grounded research must be available that identi-fies the successes and shortcomings of decentrali-zation reforms, as well as the requirements andopportunities for change. Communities, peasantorganizations, federations of interested parties,and local government representatives will need toknow what their governments are saying and whatthey are doing. Decentralization advocates,including NGOs and donors, will need to point outthe contradictions in current reforms and tocounter the interests sabotaging them.

2. Opportunities for change arise continually.Sometimes opportunities arise when politicianstalk of local democracy—whether or not theyintend it to take root—leading people to expect anddemand it. Sometimes opportunities arise whengovernments have economic and political crisesthat force them to make reforms, opening newspaces for policy influence. Sometimes progressivepoliticians and bureaucrats become active withingovernment, creating channels of influence on theinside. These are all opportunities that can betaken advantage of if the constituencies interestedin democratic decentralization know what theywant and understand how to achieve it. Advocatesneed to be prepared for opportunity. “Getting thelocal institutions right” is still the hardest chal-lenge of all; it requires confronting resistance andidentifying and seizing opportunities.

FURTHER RESEARCH

Research is essential for policy making, monitoring,and evaluation. Policy research is a critical tool forlearning and providing constructive feedback aboutthe implementation and effects of policies. Else-where, the author (Ribot 2002a) has developed adetailed decentralization research agenda. This reportbrings into focus several arenas that require furtherresearch: Institutional arrangements, incentives, localaccess to lucrative markets (“economic decentraliza-tion”), accountability, and state building. These arediscussed briefly below.

Institutional arrangements: The effects of institutionalchoices being made by policy makers and practitio-ners in decentralization reforms are still poorlyunderstood. What are the effects of empoweringdifferent kinds and different mixes of institutions inthe local arena on such values as democracy, equity,efficiency, conflict, citizenship, and sustainability?What are the effects of competition among multiplelocal institutions? How does consolidation of servicefunctions in one institution serve local populations?Are single-purpose or multi-purpose institutionsbetter placed to manage natural resources? How dosingle-purpose institutions work most productivelywith multi-purpose democratic institutions? What arethe implications of different institutional arrange-ments for the long-term prospects of democracy andsocial well being? What are the effects of interestgroups and traditional authorities in the local democ-ratization process? How does empowerment ofcustomary chiefs shape citizenship and belonging?These issues must be researched, debated, and betterunderstood.

Incentives: Local authorities generally do not invest innatural resource maintenance or protection. Manyproject, NGO, and government personnel attributethis lack of interest to: local people have other moreimmediate needs; natural resources being publicgoods, hence people do not want to invest in some-thing from which others also will reap benefits; thewait for economic and subsistence returns being toolong; and the scale of environmental managementbeing perceived to be too large for meager localmeans. These assumptions need to be explored.Alternative hypotheses also must be tested. Perhapslocal people do not engage in natural resourcemanagement because they are not compensated.Unlike in health, education, or infrastructure, localpeople are often expected to invest their labor innatural resource management with no financialcompensation—as if environmental work is not labor.In some places, the reason for the lack of interest inenvironmental protection may be cultural, theologi-cal, or cosmological.77 In other places people may notinvest in natural resources because they know thatmost of the benefits will be appropriated by govern-ment and commercial interests—as has been theirwidespread experience in the past. Different people

84


hold different ideas about the bounty of nature thatmay make them more or less likely to invest. Manyhypotheses can be spun out. Grounded field researchis needed to understand why local authorities andlocal people generally do not invest in this arena.Research is also needed to understand why and underwhat circumstances local people are motivated tomanage surrounding resources.

Access to lucrative markets: As decentralizations takeplace, local communities are ostensibly gaining morepowers to manage the resources around them. Butare they gaining access to lucrative markets for theseresources? Many activists assume that decentraliza-tion will increase local benefits and profits. Althoughsome public powers are being devolved, profits fromlocal resources are still out of reach for most ruralcommunities and individuals. This is because accessto lucrative opportunities and to rural-urban andinternational resource markets are still controlled bygovernment and private monopolies, monopsonies,oligopolies, and oligopsonies—usually secured byregulatory systems that require management plans,licensing, permitting, and production quotas thatenable extractive industries to capture resourcemarkets (Ribot 1998b; Peluso and Ribot 2003). Ruralpeople may control land and forests in some places,but they still do not reap the benefits because they aredependent on outside merchants who controlmarketing channels. Local people need to gain accessto the great wealth that continues to be extracted fromtheir territories so that they can invest in their ownpriorities and development. If local people are toparticipate in profit, “economic decentralization”—the opening of access to lucrative opportunities—willbe essential. Research is needed to better understandhow political and administrative decentralizationshapes economic decentralization and on how thedistribution of access to profits is changing underdecentralization reforms.

Accountability mechanisms: Downward accountabilityis an important ingredient in representation, localdemocracy, and decentralization. Work is needed toidentify and better understand accountability mecha-

nisms that can be applied to local institutions holdingpublic powers (some of which are identified in AnnexC). It is clear that elections are insufficient to makelocal governments accountable. Other mechanismsthat can be used systematically must be identified,and it is necessary to identify the institutions to whichthese mechanisms can be applied. Which mecha-nisms are appropriate for holding elected andappointed local government authorities accountable?Which mechanisms should be generalized to allactors and institutions whose actions affect publicresources or affect public well being? More researchis also needed on the effects of centralized financingsystems, unitary states, federalism, and politicalparties on the accountability of local elected authori-ties. In particular, local elected authorities are usuallymore accountable to the party that backed them thanto the people who elected them. This phenomenonneeds to be explored. Is it due to the lack of inter-party competition in poor countries? Is it a functionof the nature of presidential regimes where power isconcentrated in the executive branch and resourcesare allocated strictly along party lines? Do indepen-dent candidates resolve these problems?

State building: For decentralized democracy to emerge,the state will need to be a central part of the project.Boone (2003, 9–13) poses several important questions:What conditions in the local arena make local democ-racy possible? How do patron-client relations, historiesof leadership forms, the wealth and strength of localelite, property relations, and difference in the interestsand bargaining powers of rural leaders shape theestablishment of local democracy? Most importantly,what are the conditions that make local democracy partof the state-building ambitions of central authorities?Local democracy could be a means to legitimize theState—starting locally and building toward the center.Responsive government may draw in constituents andhelp to strengthen and legitimize government. Ifresearch can show that local democracy has a positiveeffect on the way local people perceive the central state,it might help motivate central authorities to promotelocal democracy.

85


Mozambique: Alda SalomaoSouth Africa: Lungisile NtsebezaTanzania: Mara GoldmanUganda: Nyangabyaki Bazaara, Juliet Kanyesigye,Frank Emmanuel Muhereza, Eugene Muramira,Agrippinah Namara, and Xavier NsabagasaniZimbabwe: Diana Conyers, Alois Mandondo, andEveristo Mapedza

The Bellagio Conference on Decentralization and theEnvironment examined research findings related todecentralization and the environment around theworld. The conference considered cases from Africa:Cameroon, Mali, South Africa, Uganda, and Zimba-bwe; from Asia: China, India, Indonesia, Mongolia,and Thailand; and from Latin America: Bolivia,Brazil, Mexico, and Nicaragua.

The conference provided a forum for inter-regionaland interdisciplinary dialogue among researchersworking on the decentralization-environmentrelationship. The objective of the conference was toexamine how decentralizations are unfolding in orderto improve, mainstream, and sustain the positiveaspects of decentralizations both for the environmentand for local livelihoods.

Papers were contributed by Arun Agrawal, AmitaBaviskar, Nyangabyaki Bazaara, ChristianBrannstrom, Ben Cousins, David Kaimowitz,Bréhima Kassibo, Thembela Kepe, Anne Larson,James Manor, Robin Mearns, Fernando Melo Farrera,Lungisile Ntsebeza, Phil René Oyono, Pablo Pacheco,

ANNEX A: WRI RESEARCH ON

DECENTRALIZATION

AND THE ENVIRONMENT

Many of the findings in this publicationderive from WRI’s Decentralization and theEnvironment in Africa comparative re-

search project and cases presented at the Conferenceon Decentralization and the Environment, which wasorganized by WRI and held in Bellagio, Italy, inFebruary 2002. WRI’s Asia-based Resources PolicySupport Initiative also provided case material.

The Accountability, Decentralization, and the Envi-ronment Initiative is part of the Institutions andGovernance Program’s Environmental Accountabilityin Africa project at WRI. As part of this initiative,policy research and analysis have been conductedover the past three years to better understand theeffects of decentralization on ecological and socialchanges in Cameroon, Mali, Senegal, South Africa,Uganda, and Zimbabwe. Working papers on the topicwere also contributed by researchers working inMozambique, Tanzania, and the Central AfricanRepublic. The goal of the project is to improve thedesign and implementation of decentralizationpolicies currently sweeping through sub-SaharanAfrica so that they promote social equity and environ-mental sustainability. Following is a list of theresearchers who have contributed to this project,according to the country for which they conductedresearch. Their papers are listed in the Works Cited.

Cameroon: Patrice Bigombe Logo, Patrice Etoungou,and Phil René OyonoCentral African Republic: Rebecca HardinMali: Thierno Diallo, Bréhima Kassibo, andNaffet Keita

86


Nancy Peluso, Ida Aju Pradnja Resosudarmo, Jesse C.Ribot, Uraivan Tan-Kim-Yong, and Xu Jianchu. Theircontributions are listed in Works Cited (most arepublished in Ribot and Larson 2004).

The Resources Policy Support Initiative at WRIprovided research exchange forums and conductsresearch on decentralized and transboundary natural

resource management in the Mekong River basin.Research was conducted in Cambodia, China, Laos,Thailand, and Vietnam by Dang Thanh Ha, HoangHuu Cai, Le Van An, Nguyen Quang Dung, PhamThi Huong, Sith Sam Ath, Tran Duc Vien, and ZuoTing. Their research has been summarized by MairiDupar and Nathan Badenoch, referenced in WorksCited.

87


ANNEX B:

CASE STUDY DESCRIPTIONS78

This annex presents descriptions and abstractsof the case studies produced under WRI’sAccountability, Decentralization, and the

Environment research initiative. These case studiesserve as the base documents for this research report.

The twenty-five cases in fifteen countries are listed inalphabetical order by country. They are followed bytwo crosscutting studies. Many other case studieswere drawn on and are cited in the text.

COUNTRY CASE STUDIES

COUNTRY AND SITE BOLIVIA, Bolivian lowlands

FOCUS OF CASE STUDY Forestry

CASE DESCRIPTION This article discusses the implications of decentralization on forest resources in theBolivian lowlands. Bolivia’s forest management systems are the most decentralizedof any Latin American country. Deforestation is also lower in Bolivia than othercountries, and revenue from the forest sector is less than two percent of thecountry’s GDP. Bolivia covers 1,098,581 square kilometers and has approximatelytwo million inhabitants in the lowland areas. Eighty percent of the country’s fifty-three million hectares of forest is located in these lowlands. The Bolivian lowlandsregion has a heterogeneous set of actors and land-tenure systems. The populationincludes cattle ranchers, collectors of non-timber products, forest concessionaires,indigenous farmers, larger landholders, small-farm colonists, and small-scale timberoperators. The land-tenure systems include private property regimes and indigenouscommunal systems. The lowland economy consists of commercial farming andranching, small-scale food production, cocoa cultivation, mining, and petroleum andnatural gas production. Although Bolivia has made progress, true democraticdecentralization is still incomplete.

ABSTRACT Bolivia has undertaken important policy reforms since the mid-1990s aimed atinstitutionalizing popular participation and promoting democratic decentralization.In the forestry sector, municipal governments have received responsibilities, andvarious mechanisms have been established to hold these governments accountable tolocal populations. Although Bolivia’s reforms are among the most advanced forestrysector decentralizations, the democratization of decision making is limited, with

88


local governments being primarily responsible for monitoring forest managementand illegal activities, and for promoting forest management by local users. Thenational government has retained the right to define standards and allocate forestresources. Still, the decentralization has created conditions for local forest users andmunicipal governments to become stronger players in natural resources governance.Outcomes of decentralization are mixed, mainly due to municipal resources andcapacity, local power relationships, and the degree to which local economies dependon forest resources.

SOURCE Pacheco 2004

COUNTRY AND SITES BRAZIL, Bahia, Paraná, and São Paulo

FOCUS OF CASE STUDY Water resources

CASE DESCRIPTION This study examines Brazil’s decentralization of water resources to determine ifparticipation of municipalities was the best indicator of downward accountability.This theory was tested using three case-study areas, each of which defineddecentralization differently. In the state of Bahia, a deconcentration policy is in place.Regional district municipalities in the Grande River valley (32,200 sq. km.), locatedin the far western portion of Bahia, remain upwardly accountable to the state. InParaná’s Tibagi River valley (24,712 sq. km.), the state has empowered firms andmunicipalities as water users to the exclusion of civil society. In the Sorocaba-TietêRiver valley (12,099 sq. km.), located near the city of São Paulo, the state mandatesequal participation of state officials, municipalities, and civil society representatives.

ABSTRACT Recent studies of decentralization of natural resource management have advancedtwo related claims: participation of municipalities encourages downwardaccountability to local populations and single-issue decentralization encouragesupward accountability to higher levels of government. This article examines a case atthe intersection of these arguments: participation of municipalities in single-issuedecentralization. Evidence from Brazil’s water-resources decentralization indicatesthat participation of municipalities is not the best indicator of downwardaccountability; rather, the nature of social and business groups and the mechanismsthat encourage their participation better predict downward accountability. Single-issue decentralization may have unintended positive effects on downwardaccountability.

SOURCE Brannstrom 2004

89


COUNTRY AND SITES CAMEROON, Twenty forest villages in South Province and East Province

FOCUS OF CASE STUDY Forestry tax system

CASE DESCRIPTION This study was conducted in five districts in the southern forest zone of Cameroon,specifically in East and South Provinces. In East Province, areas included Dimako,Lomié, and Yokadouma Districts. In Lomié District, the villages of Eschiembor,Kong, and Ngola were covered, together with the Lomié rural council. Two areaswere studied in South Province: Ebolowa and Kribi. In Ebolowa, the study wasconducted in the villages of Afanenegong, Feeyop, and Ma’amezam in the Ebomamsector, and in the villages of Mvoula and Nselang in the Mbilbekon sector. In Kribi,the study was conducted in the villages of Ebondja I and II, Nlende-dibe, and Lolabe,together with the Kribi rural council. The region is part of the Congo Basin, thesecond largest forested area in the world. The East Province of Cameroon alonegenerates seventy percent of southern Cameroon’s total forest revenue. This regionwas chosen as a research area for its diverse ethnic mix, including the Badjoué, Baka,Bakola-Bagyéli, Batanga, Bulu, Konabembe Kwassio, Maka, and Nzimé. These sitesare the largest forestry tax-generating areas in Cameroon and will soon receiverevenue from the Chad-Cameroon pipeline.

ABSTRACT The study examines the changes in the local management of forestry revenue inSouth Province and East Province resulting from the decentralized forestry taxationsystem introduced in 1994; it considers the political, socio-economic, and ecologicalimpact at the local level. The 1994 forestry reforms introduced new procedures foraccess to, and local management of, forestry revenue. Unfortunately, thedecentralization process implemented in this context is authoritariandecentralization. Imposed from above and ignoring the real needs and expectationsof the local communities, it retains many of the powers of the central state. Statecontrol is both direct and indirect, imposed through the rural councils and forestryfees management committees. It represents predatory and neo-patrimonial alliancesbetween the central state, the decentralized bodies, and the forestry feesmanagement committees, and between the authorities of the central state, the localadministration, and local political figures. Efficient and transparent management offorestry revenue only can be guaranteed in a dynamic of democratic decentralization,in which real powers over the local management of forestry revenue are devolved tolocal institutions and actors who are accountable to the local populations for theexercise of those powers. The study demonstrates the need to see thedecentralization of forest management in general, and the local management offorestry revenue in particular, as part of the overall framework of political andadministrative decentralization in Cameroon.

SOURCE Bigombe Logo 2003

90


COUNTRY AND SITES CAMEROON, Messamena and Lomié Counties in eastern Cameroon

FOCUS OF CASE STUDY Community forests

CASE DESCRIPTION Community forests in Messamena and Lomié Counties, located in easternCameroon’s Upper Nyong region. According to the Koozimé people, who inhabitthese counties, tension exists between the local populations and the NGOcommunities.

ABSTRACT Cameroon’s 1994 forestry law introduced a new approach to natural resourcemanagement. The 1996 constitution introduced decentralized authorities, whoserole is to enable the economic, social, and cultural development of its peoples.Although the regulatory laws for this decentralization reform are still forthcoming,the new legal framework for environmental policy and the overhaul of theconstitution demonstrate the government’s willingness to decentralize and toimprove forest resource management. At the same time, though decentralizedmanagement might not seem to be inappropriate, it may be a foreign notion to forestdwellers. This study examines the community forests in Messamena and LomiéCounties to analyze factors that have not been fully taken into account in the currentprocess of establishing community forests. These include exorbitant technical,financial, and human costs; the ambiguous role of some international and localNGOs; virtual, rather than substantial, common initiative groups and associations;and the disregard of traditional law, despite its legal primacy in African states.Because decentralization and forest management have been tied to the“development” of forest populations, many misunderstandings have resulted. Thesefindings suggest the question: are community forests the best ways to achievedecentralized management in eastern Cameroon?

SOURCE Etoungou 2003

COUNTRY AND SITES CAMEROON, East, South, Northwest, and Southwest Provinces

FOCUS OF CASE STUDY Local forest management

CASE DESCRIPTION This study tests the hypothesis that improvements in standard of living, equity, andecological sustainability are both compromised and delayed when local managementsystems are implemented without the proper institutional arrangements already inplace. The research included central, regional or mid-level, and local actors in thirtyvillages, located in five different sites: Dimako/Mbang and Lomié (both in EastProvince); Ebolowa (South Province); Mount Cameroon (Southwest Province); andOku (Northwest Province). The Dimako site contains a council forest, while Lomié isan experimental site for the establishment of community forests. Work on thedecentralized management of forestry fees was carried out at Mbang, Dimako, andEbolowa. Oku, in Northwest Province, was selected for comparative purposes and isoutside the forest region where the study was focused. It was chosen because of itsdistinctly different political culture and its traditional social systems.

91


ABSTRACT Decentralization in Cameroon’s forestry sector has opened opportunities forperipheral actors to participate in forest management. But local governancestructures created to represent local communities are not grounded in local realities.Instead, these committees obey instrumental and administrative priorities, resultingin weak structures that are not supported by sound institutional arrangements orlocal collective action. This weak organizational infrastructure has generatedinefficiency in local management of forests and related benefits, a new socialstratification, and the hijacking of committees by a self-interested elite connected tostate authorities. The process also has led to ecological uncertainty, due to localopportunistic behavior.

SOURCE Oyono 2004a, 2004b, and 2004c

COUNTRY AND SITE CHINA, Yunnan Province

FOCUS OF CASE STUDY Forest Management

CASE DESCRIPTION This article focuses on the effects of decentralization policies on forest managementin Yunnan Province, located in southwestern China. Yunnan, or “South of theClouds,” borders Laos, Myanmar, and Vietnam. The mountainous region is relativelypoor and ethnically diverse. Yunnan has a population of 42 million, which includestwenty-five distinct ethnic groups, such as the Bai, Dai, Hani, Jingpo, Lisu, Miao, andYi. The majority of the population is Han Chinese. China’s transition to a market-based economy has affected all of these groups. Forest livelihoods play a complexand central role in Yunnan. China’s new economic system allows the populationgreater market access, but at a price to China’s forests. The forests are critical forconservation, including erosion and climate control. Construction in this area is alsothreatening the ecological health of Yunnan.

ABSTRACT Over the last two decades, China has introduced forest resource managementreforms aimed at protecting forests and enhancing rural livelihoods. As part of thesereforms, some powers over forestry management have been decentralized to village-level institutions. However, this decentralization has so far failed to give localcommunities adequate control over forest resources, especially in areas like Yunnan,where the ethnic minority population is heavily dependent on these resources for itslivelihood. Insufficient powers have been decentralized, the village-level institutionsare not sufficiently accountable to the public, and some of the reforms designed toprotect forest resources have had a negative impact on rural livelihoods.

SOURCE Xu 2002; Xu and Ribot 2004

92


COUNTRY AND SITES GHANA, Accra, the Eastern Region, and Kumasi

FOCUS OF CASE STUDY Subsidiarity principles

CASE DESCRIPTION Ghana is touted as the model of democracy in Africa. Successful power devolutions,some of which are more than twenty years old, have made Ghana an ideal location toevaluate decentralization. Interviews for this study were conducted with academics,government officials, international donors, and local and international NGO staffmembers in Accra, the Eastern Region, and Kumasi. While the fieldwork andinterviews were carried out in these places, the arguments in the paper cover theentire southern forest zone and the transition zone. Cocoa, oil palm, cassava, andmaize are the main crops in the study areas. Land quality is good. Absolutepopulation densities are relatively high, and many communities (dominated by Akanpolitico-ethnic structures) include people from local areas as well as people from thenorthern regions who migrated to the area at different times throughout thetwentieth century.

ABSTRACT Despite Ghana’s reputation for having established collaborative and community-based management of its forests, most forest management power in Ghana remainsin the hands of a centralized forestry administration. Throughout the twentiethcentury, Ghanaian regimes have allowed forest resources to be used by state agents,traditional authorities, and domestic and international firms. When challenged todevolve forest management powers to local people, these regimes have misconstruedagro-ecological processes to justify retaining control. They essentially have usedsubsidiarity principles to argue that forest management powers must remaincentralized for the good of the public. This paper provides recommendations forstopping this misuse of subsidiarity principles. It recommends conductingparticipatory empirical studies of forest agro-ecological management and creatinginclusive processes for formulating and interpreting policies and laws.

SOURCE de Grassi 2003

COUNTRY AND SITE INDIA, Kumaon, in the Indian Himalayas

FOCUS OF CASE STUDY Forest communities

CASE DESCRIPTION This article examines forest councils in Kumaon, in the Indian Himalayas.Kumaon’s forest councils were formed in 1931, making them one of the oldestexamples of decentralization of forests.

ABSTRACT As a new strategy to conserve resources, decentralization of political authority hasdisplaced earlier coercive conservation policies in many countries. More than sixtycountries claim to have decentralized forest control. In these countries, communitiesare supposed to be involved in joint strategies to conserve forests. Decentralizationaims to achieve one of the central aspirations of equitable political governanceHumans should have a say in their own affairs. Given the ubiquity ofdecentralization initiatives, two questions are given critical attention in this study

93


(1) What accounts for decentralization of political authority toward local decisionmakers? Voluntary relinquishing of powers seems to fly in the face of expected statebehavior. (2) Do the actual effects of decentralization policies match claims thatdecentralization is more efficient and equitable and provides more politicalempowerment for local people than centralized policies?

SOURCE Agrawal 2001

COUNTRY AND SITE INDIA, Madhya Pradesh

FOCUS OF CASE STUDY Watershed development

CASE DESCRIPTION The article focuses on Jhabua District in Madhya Pradesh, India. Caste-Hindus and atrading class of caste-Hindus and Muslims dominate the majority Scheduled Tribesof the Bhil and Bhilala populations. This region is extremely poor, with seventy-fivepercent of the population living below the poverty line. Many of Jhabua’s populationuse this land in addition to their own small plots, resulting in tension between thepopulation and forestry officials.

ABSTRACT This article examines the watershed development mission in Madhya Pradesh, India,where the pressure to show tangible “success,” combined with the bureaucraticimperative to retain control, subverts the decentralized structure of participatoryresource management. Project and funding imperatives have helped to underminethe very processes that they purport to support. Yet decentralized management is notsimply empty rhetoric; its success or subversion depends on the active collaborationof the state government and villagers. Villagers and lower-level bureaucrats bringdiverse agendas and perspectives to development projects, co-opting new institutionsand assimilating them into ongoing individual and collective projects of socialsurvival and gain. In conclusion, the article suggests ways to improve accountabilityin resource management.

SOURCE Baviskar 2004

94


COUNTRY AND SITES INDONESIA, East Kalimantan, West Kalimantan

FOCUS OF CASE STUDY Forestry

CASE DESCRIPTION Examples for this case are drawn from the district level, including Kapuas District,Kutai Barat District located in East Kalimantan, Sintang District located in WestKalimantan, and Malinau District. The Indonesian archipelago consists of 1.9million square kilometers and contains 17,000 islands. Indonesia is characterized bydiverse economic, geographic, and social disparities, particularly between theresidents of the highly populated Java island and the larger outer islands ofKalimantan, Papua, Sulawesi, and Sumatera. The Ministry of Forestry controlsapproximately 120 of the 190 million hectares of Indonesia’s land area. Forestscontribute substantially to both local and national economies. The forest estate isclassified into areas for agricultural production, for conservation and protection, andfor timber production.

ABSTRACT For more than thirty years, Indonesia’s central government controlled its forests, thethird largest area of tropical forests in the world. Driven by serious political,administrative, and economic demands for reforms, the central government hasbegun to decentralize, transferring new powers to the district and municipal levels.Decentralization in the forestry sector has included transferring income frompermits and logging and reforestation fees, as well as transferring the right for theselower levels of government to issue logging permits. This sudden new access toIndonesia’s lucrative timber market has led local peoples and governments to rush totake advantage of a resource to which they previously had little right. The result hasincluded the proliferation of permits with little regard for the effect on forestresources. Large areas, including some protected areas, are being destroyed andthreatened with conversion to other uses. Local peoples, however, appear not to havebeen the ones receiving the primary benefits; they have been taken instead by thosewho have the required capital for permits and logging.

SOURCE Resosudarmo 2004

COUNTRY INDONESIA

FOCUS OF CASE STUDY Violence

CASE DESCRIPTION The discussion in this paper is based on observations made by journalists, otherresearchers, NGOs, public officials, and human rights observers in newspapers,reports, papers, Internet resources, and other primary and secondary documents.The paper is exploratory in the sense that it is not based on the author’s own,systematically planned field research intended to compare violent and non-violentcases of the transition to decentralized resource management. Nevertheless, theauthor draws on her knowledge of and extensive field experience in Indonesia—spanning the past twenty-five years—and her recent professional work on thepolitical ecology of violence and its relationship to environmental management.

95


ABSTRACT Indonesia, like many other countries, has experienced violence either preceding theimplementation of a decentralization reform, following the implementation, orduring the implementation process. This is an exploratory paper about therelationships between decentralization of resource management and violence, withparticular reference to Indonesia. The author considers a few of the many instancesin Indonesia where decentralization has been preceded, encouraged, or accompaniedby violence. In other words, violence has either been a cause, an effect, or acharacteristic of the decentralization process. The author hopes to raise somequestions about the kinds of circumstances in which we might expect violence tooccur and at the very least to be provocative about the pros and cons ofdecentralization in contemporary Indonesia.

SOURCE Peluso 2002

COUNTRY AND SITES MALI, Mopti region, including the Baye and the Youwarou communes

FOCUS OF CASE STUDY State of decentralization

CASE DESCRIPTION This study examines approaches to environmental management through Mali’s first,second, and third republics. The concept of decentralization was introduced in thethird republic, and this report tests its effectiveness in environmental managementwith two case studies. The two areas analyzed are the forestry sector in Bayecommune and pastureland in the Youwarou commune.

ABSTRACT This study outlines the legal, political, and economic context in which majorinstitutional changes in post-colonial sub-Saharan African states were carried out.Further, it shows the relation between political changes and changes in thephilosophies of those in control of the state at various epochs. This work establishesdirect links between the concepts of liberal democracy, the decentralization ofenvironmental management, and democratic participation. The report definesdemocratic decentralization as the institutional expression of the participatoryapproach; it measures the powers transferred by the central state to local institutionsand the accountability conferred upon them in the arena of natural resourcemanagement. It analyzes the functioning of the participatory approach as the basisof institutional arrangements. Specifically, it examines arrangements at two sites inthe Mopti region, one focusing on forestry and the other on herding. The reportdemonstrates how power is allocated to the key players in Mali in the naturalresources domain and identifies the institutional arrangements that determine howthese powers are utilized in natural resource management. It is clear from the studythat the state retains the dominant role in management of the environment in spiteof the legislative innovations of Mali’s third republic, which are designed toencourage broad popular participation but remain inoperative due to a lack ofconcrete application.

SOURCE Kassibo 2002a; 2004

96


COUNTRY MONGOLIA

FOCUS OF CASE STUDY Pastureland management

CASE DESCRIPTION This case study examines pastoral livelihoods in rural Mongolia throughout the1990s. The author’s ongoing management of the World Bank’s support for “pro-poor” rural development in Mongolia, as well as published secondary sources,provide additional insights. Mongolia is one of the most sparsely populated countriesin the world. Although its land area is approximately the size of western Europe, thetotal population is only 2.4 million. There is approximately one person to everysquare kilometer in Mongolia. Consequently, high quality service delivery andgovernance is difficult to achieve with such a dispersed population. The study areasrepresent diverse ecological and market-access conditions.

ABSTRACT Mongolia’s post-socialist transition since 1990 has included, among other changes,reforms toward democratic decentralization. For natural resource governance, andpastureland management in particular, decentralization has been at best incompleteand at worst “empty.” It has created an institutional vacuum that herders and othershave sought to fill with recourse to formal and informal, new and old arrangements.Herding households are also rapidly increasing against a background of economichardship and vulnerability. The effects include an altered distribution of grazingpressure, with discernibly adverse impacts on the pastoral environment, andacceleration of already rising inequality. Democratic decentralization could help torestore environmental and social justice.

SOURCE Mearns 2004

COUNTRIES MOZAMBIQUE, with comparison to Namibia, and Zambia

FOCUS OF CASE STUDY Wildlife and forestry legislation

CASE DESCRIPTION This study examines the role played by multiple actors, the accountability ofrepresentatives, and the delegation of power in community wildlife management andforest user rights in Mozambique at the national level.

ABSTRACT The objective of this working paper is to assess whether the legal provisions forcommunity-based natural resource management (CBNRM) contained in theMozambican environmental laws take into account the various factors that shape theeffectiveness of participation and decentralization. It also assesses whether the legalprovisions take into account the elements that promote the achievement of the dualobjective defined for CBNRM. This paper focuses on the provisions of the Republicof Mozambique’s Wildlife and Forestry Law of 1999 and draws comparisons withprovisions of the wildlife management acts of Namibia and Zambia. It analyzesdecentralization studies that propose an assessment of actors, powers, andaccountability mechanisms in the context of decentralization and participatorynatural resource management. Granting natural resource management powers andthe right to accrue benefits from such management to local communities is a

97


complex task; the political, legal, and institutional issues must be addressed in orderto enable the effective exercise of management powers by local communities. Powersdevolved to local communities not only help communities to meet their economicand social needs but also to achieve the desired conservation results. This papershows that what are commonly called “rights” are nothing more than simpleprivileges given and taken at the discretion of state authorities without real transferof decision-making powers to local communities. Representation mechanismsproposed in laws, as well as procedures established for CBNRM processes, still needto be better clarified for these initiatives to achieve the declared objectives. Whileupward accountability mechanisms for local actors can be found in most laws,downward accountability mechanisms are lacking.

SOURCE Salomao forthcoming

COUNTRY AND SITES NICARAGUA, north Atlantic, southern Atlantic, and north Pacific

FOCUS OF CASE STUDY Forest-management authority

CASE DESCRIPTION This article draws from research on forest management authorities located in 21 ofNicaragua’s 151 municipalities. These municipalities represent all of the main agro-ecological zones of Nicaragua, but particularly the forested areas. Six of themunicipalities included in the study are located in the Bosawas rainforest reserve inthe north Atlantic, two are found in the Indio-Maiz reserve in the southernmostAtlantic regions, and four are in the Las Segovias pine forests. The remaining ninemunicipalities are located in the more deforested north Pacific regions, four in theplains and five in the drier inland areas.

ABSTRACT This article argues that decentralization of natural resource management is apolitical process resisted by the central government due to the fear of losing powerand/or economic resources to local governments. In Nicaragua, although the formalprocess of power transfers largely stagnated from 1997 to 2003, decentralization“from below” continued to advance thanks to political pressure from civil society andmunicipal governments and the increasing legitimacy of local authority. At the sametime, many municipal governments have little interest in resource managementwhere there are few apparent economic benefits. However, these local governmentscan be motivated to take on natural resource management responsibilities; pressurefrom constituents and NGOs, for example, can motivate them to do so. Local andgrassroots processes are necessary conditions to make formal decentralizationdemocratic and responsible.

SOURCE Larson 2004b

98


COUNTRY AND SITES SOUTH AFRICA, Transkei region, Eastern Cape Province

FOCUS OF CASE STUDY Land and local government reform

CASE DESCRIPTION Two districts were examined in the case studies the Tshezi area, located in the Transkeiregion’s Wild Coast, and the Xhalanga area. The areas have different backgrounds, buteach demonstrates the unresolved function, issues, and roles of traditional tribalauthorities, particularly related to land issues. Tshezi has a stronger tradition ofchieftainship and traditional power structures, while the Xhalanga communityhistorically has shown more resistance to state-controlled local governments.

ABSTRACT Post-apartheid South Africa embarked on an important democratization process thatincluded reforms in local government and land administration in the formerBantustans. A new conception of “developmental local government” introduced thenotion of elected local leadership and an emphasis on improving the quality of life ofpreviously disadvantaged sectors. Nevertheless, this democratization process risksserious compromise due to the concessions being made to traditional ruralauthorities—the same autocratic local authorities who enjoyed significant powersunder apartheid. In particular, the legal transfer of land administration to electedofficials has not occurred, leaving this important power under the authority ofunaccountable local chiefs and headmen. Is democratic decentralization compatiblewith traditional authorities (chiefs of various ranks) who inherit their posts? Whilesome chiefs and headmen may choose to promote local participation, the freedom tochoose ones’ leaders is automatically excluded, seriously limiting downwardaccountability. The article argues that the central government must recognize thisincompatibility and support the rights of rural residents to full citizenship.

SOURCE Ntsebeza 2003, 2004

COUNTRY AND SITES SOUTH AFRICA, Transkei region, Eastern Cape Province

FOCUS OF CASE STUDY Ecotourism

CASE DESCRIPTION This article examines a failed community-based ecotourism development project inthe Mkambati area, one of the poorest rural regions of South Africa, located in theWild Coast of Eastern Cape Province. In the Mkambati project, both local and non-local actors and agencies have been involved in the complex politics of land andnatural resource rights, and the question of who decides (and therefore who benefits)has been central to the many-sided conflicts that have erupted. The project is locatedwithin the former Transkei “homeland” of South Africa, in northeastern Pondoland;it is one of several projects targeted for high profile spatial development initiatives.The area is inhabited by the mPondo people, who speak a Xhosa dialect (IsiMpondo).The case study focuses on three sub-areas within the wider Mkambati area that areunder different tenure regimes communal tenure settlements to the west; 11,000hectares of state land in the center, formerly used by a now defunct parastatalagricultural project; and the 7,000-hectare state-owned Mkambati Nature Reserve to

99


the east. The communal tenure area comprises seven villages that make up theThaweni Tribal Authority.

ABSTRACT A failed ecotourism project in South Africa illustrates the weakness ofdecentralizations that place insufficient emphasis on democracy. The projectpromoted “empowerment” of local communities (and entrepreneurs) andgovernment facilitation of community “partnerships” with private-sector investors.Neglected were the devolution of effective powers and the accountability of localbodies to community. Project planning failed to promote local participation, whichwould have allowed a clear understanding of local livelihoods and of deep-seated landand resource tensions. Entrepreneurial elite attempted to capture project benefits,and ordinary community members resisted. Severe conflicts erupted, stalling theproject. To overcome these challenges, democratic decentralization requires a strongcentral state to provide resources for land-tenure reform and to promote devolutionof powers and the accountability of local bodies.

SOURCE Cousins and Kepe 2004

COUNTRY AND SITES TANZANIA, Maasai ecosystem in northern Tanzania (multiple sites)

FOCUS OF CASE STUDY Community-based conservation

CASE DESCRIPTION The boundaries of the Northern Maasai Ecosystem were designed to enclose themovements of migratory wildlife species, which are found outside of the nationalparks for six months every year. This area highlights the ecological compatibility ofMaasai pastoralism with wildlife conservation. It also demonstrates the importanceof wildlife conservation outside of park boundaries. Land management plans thatcan be changed easily are critical; however, zoning and land use planning andportioning are still employed. The area covers 370,000 hectares and includesTarangire National Park, Lake Manyara National Park, the Simanjiro Plains, and theKwakuchinja corridor. There are 263 villages across five districts inhabited mainly bythe Maasai people and Dorobo hunter-gatherers.

ABSTRACT This working paper analyzes the evidence surrounding government policy supportingthe building of a new community-based conservation (CBC) program in Tanzania. CBCshifts the focus of conservation from nature as protected exclusively through state controlto nature as managed through inclusive, participatory, community-based endeavors. Toeffectively make this shift, CBC devolves natural resource management to localcommunities and hence is often referred to as community-based natural resourcemanagement. This paper outlines how, despite the rhetoric of devolution andcommunity participation, conservation planning in Tanzania remains a top-downendeavor in which the knowledge-claims of local people remain relegated to the margins.This paper also addresses how the challenges posed by an inclusive and participatoryCBC are particularly salient in the Maasai ecosystem in northern Tanzania where manyof the consequences or “constellations” of unsuccessful conservation projects furtherchallenge the implementation of CBC initiatives. The paper discusses the need to

100


address these challenges and engage the Maasai as active knowing agents in theconservation process; doing so will produce a better match between the political,rhetorical, social, and ecological goals of CBC. It also suggests that in addition to thedifficulties associated with the transfer of power from state to community hands, thereare also complex challenges that CBC poses to the culture or institution of conservation.The intended (and at times unintended) landscapes of conservation are crafted forlegibility, manageability, and foreign scientific expertise, leaving little room for theinclusion of indigenous or local knowledge claims, even though “the community” hasbecome the catchall solution for effective conservation and development. CBC initiativesthat favor the inclusion of complex local knowledge are difficult to administer and do notfit into the managerial categories of conservation and, therefore, challenge the currentlandscape of conservation. Local communities, therefore, have become viewed as toolsfor, or “commodities” of, conservation rather than as active knowing agents who must begranted substantive decision-making power.

SOURCE Goldman 2001

COUNTRY AND SITES THAILAND, Chom Thong, Sam Muen, and Mae Tho-Omlong in Chiang MaiProvince

FOCUS OF CASE STUDY Watersheds

CASE DESCRIPTION This paper examines three distinct arrangements of actors and accountability in thewatersheds of Chiang Mai Province. Chom Thong, located in the southern district, ishome to Doi Inthanon National Park. The area contains both mountains and valleysand more than thirty villages in the upland and highland areas. A growing populationin the area’s thirty villages and increasing agriculture use have led to competingnatural resource agendas. Karen farmers who have been in the area for more than onehundred years formed a coalition to protest Hmong immigrants to the area. (TheKaren are an ethnic group originally from Burma.) The Mae Klang watershed supplieswater for vegetables cultivated for city markets in northern Thailand and Bangkok.Both the lowland and the upland Karen farmers enjoy good relationships. Sam Muenalso contains multiethnic communities who comanage forests. Omlong, located on thewestern range of the Thai-Burma border to the south of Chiang Mai City, is the centralcultural area for the Karen. The Karen’s sustainable agriculture and forest practiceshave been used as successful models of development. Now, with the addition of theHmong in the Me Tho area, the dense green forests have been deforested and land usechanges have resulted in conflict over overlapping uses.

ABSTRACT In Thailand decentralization reforms have been underway for almost two decades,starting from the early steps of public debate and implemented in several pilot schemes.In 1997, the enactment of the “New Constitution” has marked a radical change inpolitical reform. Decentralization has been a “politics-led process.” Through three cases,this paper discusses (1) whether the recent political process is creating conditions fordecentralization, (2) how decentralization is practiced and constrained at local levels, and(3) the preliminary impacts of decentralization at the local level.

SOURCE Tan-Kim-Yong 2002

101


COUNTRY AND SITES UGANDA, Masindi, Mbale, and Mukono Districts

FOCUS OF CASE STUDY Decentralization

CASE DESCRIPTION This study used a case-study approach in three districts in Uganda Masindi, Mbale,and Mukono. Masindi’s population is 479,000 and the district is forty-nine percentprotected. The district has three kingdom reserves, one local forest reserve, andfifteen central forest reserves. It is surrounded by large lakes. One of Uganda’s mostpopulated districts is Mbale, home to 826,000 people. Within these districts, land isin great demand and often dominates local politics. Mbale’s few forest reserves aredivided among central forest reserves, local forest reserves, and national parks.Mukono District was one of the first thirteen districts in Uganda to experiencedecentralization reforms. Swamps, wetlands, and open water make up sixty-eightpercent of the district. Fifty-eight central forest reserves and ten local forest reservesare located in Mukono. Human encroachment and agriculture have caused extensiveland degradation and soil erosion. Mbale District has only 13 square kilometers ofofficial forest reserves, compared to 570 square kilometers in Mukono and 1,030square kilometers in Masindi.

ABSTRACT The devolution of decision-making powers over natural resources to publiclyaccountable local authorities is frequently advocated as a means of achieving socialdevelopment and enhancing environmental management. The experience ofUganda’s current decentralization reforms, however, suggests that the extent towhich such benefits occur depends on the character of the decentralization.Uganda’s decentralization is renowned internationally for its local origins,participatory character, and political commitment. However, an analysis of thereforms from an environmental perspective indicates three major problem areas.First, there has not been an effective or consistent devolution of powers over naturalresource management. In the case of forestry, for example, powers have beendecentralized and re-centralized several times since the reforms began in 1993. Thereason for this is that decentralization has been used primarily to resolve thegovernment’s financial and legitimacy problems, rather than as a means of achievingeither public participation per se or improved environmental management. Second,local governments have failed to exercise the limited powers they do have sincecontrol over the necessary financial and human resources has remained centralized.Case studies of three districts—Mukono, Mbale, and Masindi—reveal that they haveonly been able to exercise such powers when donor assistance has been available.Third, when local governments have attempted to influence environmental matters,the social and environmental outcomes have not always been positive due to conflictsof interest among the actors involved in natural resource utilization at the local level.The study suggests that there is a need for the central government to devolveeffective environmental powers to local governments, for local governments toincrease their revenue-raising capacity in order to achieve greater financialautonomy, and for the introduction of checks and balances to prevent the misuse ofpowers to achieve personal gain.

SOURCE Bazaara 2002a, 2003, 2004a

102


COUNTRY AND SITE UGANDA, Masindi District

FOCUS OF CASE STUDY Forest management powers

CASE DESCRIPTION This study focuses on the transfer of forest management powers to local jurisdictions inMasindi District, an area rich in natural wealth located in western Uganda. Research wasconducted at study sites in the district, including the Budongo, Karujubu, and Pakanyiareas; in addition, interviews were conducted with members of the central government.The population of Masindi District has an annual growth rate of five percent; 466,204people live in the district, a figure nearly double that of ten years ago. The forest cover ispatchy as a result of the burgeoning population and the diversification of land use.Today, seven of Masindi’s eight reserves have been designated as local reserves.

ABSTRACT Uganda’s decentralization of natural resource management has become a model forenvironmental decentralizations elsewhere in Africa. This study set out to determine theextent to which significant discretionary powers have shifted to popularly elected anddownwardly accountable local government actors. Effective political or democraticdecentralization depends on the transfer of discretionary powers to local governments.The research showed that very little forest management power was transferred todemocratically elected and downwardly accountable local governments. Instead, thedecentralization reform returned forests to unelected traditional authorities; it alsoresulted in the privatization of limited forest management powers to licensed usergroups. The forestry department was interested in transferring only those powers thatserved to increase its revenues and reduce its expenditures. The limited transfer of forestmanagement powers through a combination of devolution and privatization satisfied thestate’s desire to increase its legitimacy and support among pivotal rural constituencies; toa large extent, it accomplished this by transferring powers as a form of patrimony. Theshift of powers through decentralization, therefore, was carefully measured toconsolidate the base of support upon which the state drew its legitimacy and power. Itdid little to increase the power and legitimacy of local governments, which were left in anuncertain and weak bargaining position in relation to the central state. Rather thanincreasing local discretion over the management of natural resources, Uganda’sdecentralization reforms actually prolonged state control and power over natural wealth.However, the reforms did not wholly produce the outcomes that central authorities hadhoped for. While privatization did result in higher forestry department revenues, thetradeoff was greater involvement of private sector actors in the department’s decisionmaking. In fact, to the dismay of the forestry department, private sector user groups wereable to influence decision making up to the highest levels of the forestry sector. Therecipients of the new forest management powers began to challenge state decisionsregarding the management of forest reserves and to lobby for the transfer of furtherpowers. Ultimately, however, the state retained significant powers to manage forests,while selectively “decentralizing” limited powers to district and sub-county councils. Overtime, powers shifted upwards and downwards. The forestry department regained controlof the coveted larger central forest reserves established in 1998. The unsteadyprogression of decentralization reforms in Uganda points to an unwillingness to transferreal discretionary powers over the management and use of forest reserves to popularlyelected and representative district and sub-county councils.

SOURCE Muhereza 2003

103


COUNTRY AND SITE UGANDA, Bwindi Impenetrable National Park

FOCUS OF CASE STUDY Protected area management

CASE DESCRIPTION Bwindi Impenetrable National Park is located in southwestern Uganda, near theDemocratic Republic of the Congo. It was the first national park in Uganda to test thecollaborative management approach. In addition, a defined institutional frameworkpilot-tested the principles of participation and local government decision making inBwindi and its sister park, Mgahinga National Park. The park is located on the edge ofthe Western Rift Valley, on the Kigezi Highlands; it comprises the districts of Kabale,Kanungu, and Kisoro. The diverse forest area is home to rare and endemic flora andfauna. The surrounding areas compose some of Uganda’s most densely populatedareas, with 102 to 320 persons per square kilometer. The majority of the populationoriginally was attracted to the area because of the enormous opportunity provided bythe forest, much of which has now been cleared, partly due to population pressure.Agriculture is the main livelihood in the area although animal husbandry andbeekeeping are also popular. The park has had a varied history as a forest reserve, ananimal and game sanctuary, and, most recently, as a World Heritage site.

ABSTRACT This working paper focuses on the roles of local institutions involved in naturalresource management under the community conservation approach to protected areamanagement in Bwindi Impenetrable National Park. It deals specifically withquestions of competing interests and distribution of decision-making powers amongthe actors. It also describes and analyzes the current institutions involved in protectedarea management and investigates the extent to which decentralized institutionalarrangements guarantee effective local participation in decision making. Findingsshow that community conservation and collaborative management as practiced inprotected areas in Uganda today do not offer sufficient opportunity to producedemocratic governance of natural resources. Community participation undercommunity conservation and collaborative management does not adequately andeffectively translate into community empowerment and control over resources. Thereare several contributing factors. Central authorities and supporting agencies remain incontrol both of resources and of the policy framework, which was formulated withinsufficient community and local government input. Also, the principles upon whichcollaborative management is based were not developed out of mutual agreementbetween the communities and the other partners. The conservation and tourisminterests of the Uganda Wildlife Authority and the donor-funded agencies oftenoverride the interests of the local communities who have no mandate to deal with themost critical problems affecting them. The authors conclude that local governmentinterests lie mainly in activities that generate revenue and enhance human rights andbenefits, not in environmental conservation. The national government overestimatesthe readiness, willingness, and capacity of local governments to assume responsibilityfor conservation of the environment; they fail to recognize that local populationsshould not take on responsibilities that surpass their decision-making powers. Theabove weaknesses in collaborative management have grossly undermined downwardaccountability of local institutions to the communities they serve; these weaknesseshave limited the democratic governance of natural resources.

SOURCE Namara and Nsabagasani 2003

104


COUNTRY AND SITES ZIMBABWE, Binga District

FOCUS OF CASE STUDY CAMPFIRE Wildlife program

CASE DESCRIPTION Binga District, one of fifty-seven administrative districts in Zimbabwe, is located innorthwestern Zimbabwe, near the Zambian border. The estimated population is100,000, consisting mostly of the Tonga ethnic group. Wildlife is abundant in BingaDistrict, an area in which the CAMPFIRE program has played a substantial role. TheBinga people have had an unusual history in the area. Originally, the Binga werecharacterized as living in “subsistence affluence” by the river and enjoying a richcultural and economic life. However, in the 1950s, the Kariba Dam was built and thearea where they were living was flooded, forcing the Binga to relocate. Consequently,they lost many of their social and cultural ties; a national survey rated the district assuffering from the most severe poverty in all of Zimbabwe. More recently, the Bingahave become actively involved in trying to ameliorate their fate; they votedoverwhelmingly for the opposition party in the last election and are increasingly seenas an important force in national politics.

ABSTRACT This working paper details the efforts to decentralize wildlife management inZimbabwe since 1980 and uses the well-known CAMPFIRE program to illustratesome of the major challenges in implementing such programs effectively. Whileearly efforts created an interest in decentralization within local communities andwith ministry officials, very little real decentralization took place. This was primarilydue to a lack of financial support and to political resistance to changing powerstructures. These past decentralization efforts set the stage for the creation of theCAMPFIRE program, the primary goals of which are to foster sustainablecommunity management of wildlife and to reduce human/wildlife conflicts. Toaccomplish these objectives, CAMPFIRE delegates responsibilities to regionalcouncils who are required to follow strict managerial guidelines. AlthoughCAMPFIRE has had a number of positive effects, the program has failed to achieveits goals and to effectively decentralize wildlife management. The CAMPFIREprogram effectively restricts councils and local communities to managing wildlifepopulations within the guidelines set by, and therefore under the control of, nationalauthorities. These guidelines have been ignored largely because they were imposedon communities in a top-down manner. One of the reasons that guidelines wereimposed at the central level is that national authorities distrust the ability of localcommunities to manage wildlife effectively; in addition, local councils failed to exertpressure on national authorities to grant them the power necessary to makedecisions and set guidelines. While these are important factors, the most significantobstacle to decentralizing wildlife management in Binga is that leaders in theDepartment of National Parks and Wildlife Management have vested financial andpolitical interests in the management of wildlife, especially in the awarding oflucrative safari contracts. As a result of this resistance, there have been, and will be,no net positive effects from CAMPFIRE’s efforts, and the program continues to be atrisk for hijacking by local elite who often use program funds and dividends forpersonal gain.

SOURCE Conyers 2002

105


COUNTRY AND SITES ZIMBABWE

FOCUS OF CASE STUDY Forests

CASE DESCRIPTION Zimbabwe consists of 390,759 square kilometers. Its population is divided into twomain ethnic and linguistic groups, the Ndebele, who mostly inhabit the southwest,and the Shona. This study explores the political economy of the allocation of naturalresource governance powers among actors at a variety of scales of organization inZimbabwe. It examines natural resource legislation along with a broad set of otherrelevant laws since control over natural resources is intricately linked with othercomponents of local governance. Much of the study is a review of the 1988 RuralDistrict Councils Act, which is based on the ideology of decentralization andprovides structural and operational arrangements for rural local governance. The actprovides the legal basis for, among other things, the demarcation and establishmentof distinct spatial units for rural administrative purposes, designated rural districts,and the establishment of rural district councils to preside over such districts. It alsoprescribes the structure of these councils and the process for constituting theirmembership; it specifies the roles and responsibilities of the councils, which includethe planning and implementing of development programs and related taxing andfiscal authority; and it bestows on them minor legislative powers that enable them toenact bylaws.

ABSTRACT Based on an analysis of relevant legislation, this study reviews the political economyof the allocation of powers over natural resource management between the centralstate, the rural district councils, and the local communities. The allocations arecritically examined in relation to citizen empowerment. An analytical toolkit, derivedfrom a review of relevant literature, is provided for holistically scrutinizing the extentand operation of accountability between and within decentralized governance unitsand other units above and below them. These theoretical tools are then used to arguethat, contrary to establishing the institutional infrastructure for decentralized naturalresource management in Zimbabwe, most legislation re-centralizes power at thedistrict level, doing so at the expense of meaningful citizen participation in naturalresource governance. Only the “privilege” of initiating development plans appears tohave trickled down to the grassroots level. The state and its closest actors still retainthe complementary roles of approval, implementation, and fiscal control. The studyconcludes by considering ways for addressing the contradiction between thedecentralization that the state claims to have achieved and the re-concentration ofpower at the district level that is actually taking place, assessing the scope for bothincremental and radical approaches.

SOURCE Mandondo and Mapedza 2003

106


COUNTRY AND SITES ZIMBABWE, Mafungautsi State Forest in Chemwiro-Masawi and Batanai

FOCUS OF CASE STUDY Resource Management Committees

CASE DESCRIPTION The Mafungautsi State Forest, located in west central Zimbabwe, is in Gokwe SouthDistrict. The forest makes up nearly ten percent of Zimbabwe’s forest reserves andcovers 82,100 hectares. The two study sites, Chemwiro-Masawi and Batanai, borderthe Mafungautsi State Forest. Each site is represented by a resource managementcommittee formed by Zimbabwe’s forestry commission. Timber is commerciallyextracted from Chemwiro-Mawai, and few people live in the forest. In contrast, theBatanai area was formerly forest land from which many people were evicted.Batanai’s resource management committee is considered successful whileChemwiro-Mwai’s committee is considered a failure.

ABSTRACT Although governance innovations that involve moving powers closer to the citizensare receiving increasing policy support, their implementation is not withoutproblems. This study uses a review and case-study approach to critically examine thecontradictions and ambiguities of “peasant empowerment” in a comanagementventure between Zimbabwean foresters and peasant communities. The institutionalinfrastructure for comanagement was derived from and superimposed upon acomplex web of local power bases, further fragmenting existing networks of interest,affection, and association, and thus limiting the scope for comanagement. Thelegislative environment, at least during the pre-2000 period, supported theexpropriation and control of the land and resources of peasant communities, thuscontradicting the underlying principle of comanagement, which is that of equalpartnership. Powers over natural resources have remained centralized in the nationalstate; the little power that has been decentralized has been transferred to levels thatare not close enough to the citizens. Furthermore, there is no legislation that gives alegal mandate and fiscal autonomy to units closer to the citizens than the districtlevel. The comanagement venture is supply led rather than demand driven,originating in international development assistance circles and implemented on theterms and conditions of their allies in the state bureaucracies responsible for naturalresource management. However, in spite of their marginalization, peasantcommunities have a wide repertoire of tools that enable them to significantlypenetrate local and national political processes. The study identifies the need forfundamental changes in the comanagement system, including the creation ofdownwardly accountable institutions and experimentation with new comanagementrelations. It argues that such changes require related reversals in the ways thatresearchers, policy makers, civil society organizations, and other facilitators havetraditionally conducted their business. The central thesis is that the state and otherexternal actors have sought to mould and discipline local institutions in order toachieve top-down conservation objectives.

SOURCE Mapedza and Mandondo 2002

107


CROSSCUTTING CASES

The following cases do not focus on any one particular region or country; instead, each draws lessons ondecentralization from multiple countries or regions.

FOCUS OF CASE STUDY Natural Resources as a Lever for Local Democracy

CASE DESCRIPTION Multiple cases

ABSTRACT This article compares service and infrastructure with natural resource managementin order to bring attention to unique aspects of natural resource management thatmake it well suited to support more democratic approaches to decentralization. Theobjective of this article is to show that natural resources can foster local democracy—as an end in and of itself and in order to improve the effectiveness ofdecentralizations in all sectors. The management of local natural resources hasvarious characteristics that make it difficult to approach as a centrally mandated andfunded sector and that make it difficult to ignore legislative and judiciary functions.The differences among sectors are all matters of degree and not kind. The authors’aim is to use natural resource management’s particular mix of characteristics tothrow aspects of all decentralization reforms into relief. Those are the aspects thatneed attention in order to balance the immediate technical or instrumental objectiveof meeting national social and economic ends with the longer-term participatory,inclusive, and empowering procedural objectives of local democracy.

SOURCE Kaimowitz and Ribot 2002

FOCUS OF CASE STUDY Committees

CASE DESCRIPTION Multiple cases

ABSTRACT Decentralizations in the 1980s transferred powers to multi-purpose local governments.In recent years international donors and central governments are increasingly turningtoward single-purpose user committees. Although these committees appear to be lessdemocratically accountable and less representative than local government, donors viewuser committees as a mechanism to give local peoples greater say over thedevelopment decisions that affect them. Central government officials establish usercommittees at the insistence of donors but then manipulate them by selectingcommittee members and by reining in their powers. The article explores how theseproliferating single-purpose committees are undermining the democratic processesthat were presumably institutionalized with the creation and strengthening of electedlocal governments in developing countries. This new approach fragments localparticipation, reducing its coherence and effectiveness and leaving the poor quitepossibly worse off than before. These committees appear to usurp local governmentfunctions and deprive local governments of revenues. These myriad problems result indestructive conflicts and the undermining of local government authority.

SOURCE Manor 2004

108


109


ANNEX C:

ACCOUNTABILITY MECHANISMS

Africa, Tanzania, and Zimbabwe (Therkildsen 2001,27, fn. 26, 29–30).79

There is often a failure to separate powers of thejudiciary, legislative, and executive, particularly at thelocal level. Without separation of powers, there is nobalance of powers within government and no alterna-tive routes for people to challenge representatives andadministrators or to change or even enforce the rulesby which all branches of government operate. This iswhat O’Donnell (1999, 38, in Brinkerhoff 2001, 3)calls “horizontal accountability.” In the colonialperiod, the judiciary and executive branches werefused in the local arena both through the indigenatcourts of the French system and in the British systemof Indirect Rule. This failure to separate powersdenies recourse, as it does in Uganda’s local govern-ments. It is a problem in many other countries aswell. The failure to separate powers often deniesrecourse to individuals who have disputes withtechnical services. (Mamdani 1996a; Oloka-Onyango1994, 463.)

Ostrom (1999) argues that polycentricity of govern-ment and the balance of powers are importantstructural aspects of accountability. A balance ofpowers in which there are counterpowers to thecentral government can increase accountability byincreasing the number of actors with a voice inpolitics and the ability of non-central actors toscrutinize central institutions. The World Bank(2000, 112) suggests that there is a need to institu-tionalize the balance of powers between national andlocal governments through rules that protect and

Below is a list of accountability mechanismsthat emerge in the literature. Brinkerhoff(2001) provides an excellent analysis, making

accountability issues much more legible.Hirschman’s (1970) analysis of exit, voice, and loyaltymay be another frame for organizing an analysis ofaccountability. This list, however, is not systematizedaround either framework. Nevertheless, it illustratesthat there are multiple mechanisms that policymakers can use to establish greater public account-ability of government and of other local authoritiesthat hold public powers.

Legal recourse through courts is an importantmeans of accountability. Independent judiciaries arecritical for holding public figures accountable. Oftenthe judiciary is not independent for numerousstructural reasons. For example, local authorities whomay very appropriately have the power to adjudicateamong local citizens should not adjudicate inter-jurisdictional cases or cases between themselves andothers (Oloka-Onyango 1994, 463; Mamdani 1996a;Agrawal and Ribot 1999). Regardless of whethercourts are well structured, they are inaccessible tomany citizens in Africa. They are often too far awayor too expensive or complex for average citizens touse. Public interest law associations can help toensure that poor and marginal populations havegreater access to legal recourse (Veit and Faradayforthcoming). Rothchild cites the office of thetribunal administratif, which was introduced by theFrench, and the office of the ombudsman as twouseful recourse structures (Rothchild 1994, 6). Suchcomplaint officers have been established in South

110


limit the rights of sub-national governments. Ostrom(1999) supports the notion of polycentricity ofgovernment, suggesting it introduces such a balance.She argues that multiple loci of power, combinedwith higher levels of government whose role is toprotect people from the excesses of power of localelite actors, are necessary for balancing power.

Independent or third-party monitoring by electedcomptrollers, NGOs, or the associative movements canhelp construct downward accountability. In somecountries, such as the United States, there are electedtown comptrollers. These officials monitor the affairsof local government for the local community. NGOsand other associations can also play a monitoring role.While they should have no powers over communityresources or decisions (since NGOs are private bodiesand are not necessarily accountable or representative),they can monitor local and national government toensure they are meeting their legal obligations. Theycan also inform the local population and/or file suit ifthe government is not living up to its requirements.NGOs and associations can also, of course, lobby onbehalf of the portion of civil society that they represent.

The literature contains good examples of indepen-dent monitoring. Blair (2000, 24) describes Bolivia’svigilance committees, which were created to monitorelected bodies; its members are “selected” fromtraditional local governance systems, includingpeasant unions and neighborhood councils. InCanada, Native American women recently blamedtribal chiefs of “rigging elections, stealing govern-ment money, and going on fancy gambling vacationsin the States, while their people live in third worldpoverty” (Brooke 2001, A3). These women haveformed the First Nations Accountability Coalition.One member has used hunger strikes to demandaccountability of the Native American chiefs and ofthe members of Canada’s government who supportthem. The group also has put together a documentrecounting numerous instances of corruption andhave held “hearings” on the matter around thecountry. They also have delivered their findings toCanada’s Senate.

Confederations, federations, and unions of associa-tions and other groups within society can constitute an

associative movement. One example is Senegal’sFONG (the Federation of NGOs), which is a nationallyconstituted lobbying group representing a variety ofrural associations around Senegal. Such associativemovements can be supported by enacting legislationpermitting associations, federations, and confedera-tions to form, and by organizing assistance. They canfoster accountability by monitoring, informing, andlobbying. While lobbying has been supported as anactivity to hold governments to account, it can be verydifficult and risky in the absence of recourse and otherlegislation that allows people to organize and pressuretheir governments. Of course, lobbying also can reduceaccountability toward the less powerful and be highlyskewed toward more powerful interest groups.

Transparency—that is, openness to public scru-tiny—is an accountability mechanism frequentlycalled for by international organizations.80

The members of working groups at the 1993international conference titled “Local Self-governance,People’s Participation, and Development” held inKampala “stressed again and again the necessity forgreater transparency and accountability at all levels ofgovernment to ensure efficiency and honesty….Democratic reforms and checks and balances at alllevels of government are needed to ensure that goodgovernance can be sustained over time” (Rothchild1994, 6). Another means to increase transparency is toprovide information on the roles and obligations ofgovernment. This information can be provided by themedia, by NGOs, and even by government as indicatedin public reporting requirements.

It is important, however, to remember that transpar-ency is not enough. In many countries corruption istransparent. It is often plainly visible that ministersand other government officials drive cars and live inhouses that they could never afford on their salaries.But nothing is done. Transparency does not work byitself. Other accountability mechanisms enablingpeople to sanction government in reaction to newinformation must also be available.

A free media can also play a monitoring role. A freemedia can shape public action. Sen (1981) argues thatfree media can avert famine. The media also can

111


disseminate important information to local popula-tions on what their rights are and what they canexpect from their local government leaders (cf.Tendler 1997, 15). The participants in a 1994 confer-ence on decentralization in Ouagadougou agreed thatwhile representation is necessary, “citizen participa-tion in local affairs is necessarily limited when thereare constraints on the freedom of association and theliberty of expression” (Mbassi 1995, 28). Freedom ofspeech more generally, then, is an important issue. Itis needed at the national scale in terms of free mediaand legal protections of freedom of expression. It isdirely needed in the most local arenas, where womenand minorities are often unable to express themselvesfreely—particularly in public meetings and throughrepresentatives.

The practice of public discussion also can increasetransparency. This practice is often used in Africawhere elders gather to discuss community matters inpublic. Also, in the Mhondoro cults of Africa, spiritmediums speak and are interpreted in public toexpress the will of ancestors (Spierenburg 1995). Thepractice of holding public meetings with representa-tives to discuss budget and policy decisions also canincrease transparency. In Uganda, many local peopledid not feel listened to by their representatives.Nonetheless, people felt it was very important to gatherand to exchange views with each other at local councilmeetings. One survey reported that sixty-three percentof Ugandans claim to participate in the local councildecision-making process (Saito 2000, 8). Public forahave serious limits since marginalized groups, such aswomen, religious minorities, or immigrants, may notfeel comfortable voicing their opinions. As a result,public fora should not provide the only opportunitiesfor people to express themselves. A requirement ofpublic reporting, such as the weekly posting of budgetsin local papers or just on the wall of the local govern-ment building, is another mechanism to rendergovernment accountable. If budgets, decisions, andplanned programs and spending are publicly posted,people will have an easier time discerning whetherlocal government is serving their interests. This is avery easy mechanism to legislate.

Participatory processes also can be employed toimprove dialogue between government and people.

Through facilitated processes of participation inwhich information is exchanged, people can learnwhich services government can provide, and expecta-tions can be built. In this manner, people can learn tomake more demands on their representatives.Orchestrated participation also can increase publicparticipation in decision making in a way thatcomplements or strengthens other representativeorgans and adds to the public’s ability to makedemands on local authorities. The danger is when“participatory” methods are used as an alternative torepresentative and accountable government—indeed,as a way of avoiding government. This use of partici-pation can undermine the legitimacy and accountabil-ity of local government.

Civic education is important for building account-ability of government. It empowers people to knowtheir rights and to know the powers and obligations oftheir representatives. In Uganda, many local people donot know how the elected local councils work. Manyare suspicious that “the local leaders are eating [their]money.” Explaining democratic local government topeople can engage them in governing processes. OneUgandan woman to whom decentralization wasexplained for the first time said, “In the past I wasunwilling to pay my tax because I was not clear howthe money was used. Now I am willing to pay it sinceI now know how it is used” (Saito 2000, 10). Educa-tion writ large is also important for forming a criticalcitizenry. General education, which includes literacy,numeracy, analytical skills, history, and other infor-mation dissemination, is another way to empowerpeople to make demands on their representatives.

Placing discretionary powers in the hands of localleaders can increase the accountability of governmentand increase the engagement of civic organizations inpublic affairs. Cofinancing, in which communities arerequired to provide a portion of the funds for a givenprogram, has been proposed as a way to give commu-nities “ownership” of local programs. While “experi-ence so far is that this is both patronizing and empiri-cally difficult to substantiate,” cofinancing mayproduce some positive outcomes, according to Onyach-Olaa and Porter (2000, 25). They observed that wherethe funds transferred were truly discretionary and seenas a supplement to local funds, local councils “have no

112


difficulty in principle contributing a major share oftheir revenue,” and where there is no required contri-bution, local councils “have begun to utilise the fundsin very inventive ways” with higher efficiency and“multiplier” effects.81 “If truly discretionary resourcesare made available, experience so far suggests thatroom begins to be created for better ‘downwardaccountability,’ to local constituents who expect anddemand that councilors will stretch the resources as faras possible” (Onyach-Olaa 2000, 25). Joshi (personalcommunication, Institute of Development Studies,Sussex 1999) observed that in India civil societyorganizations begin to form around and lobby stronglocal governments; this supports the argument thathaving discretionary powers within the local arena isone way to make local government accountable.Without powers, there is actually no reason for anyoneto even try to hold local government to account.

Proximity of leaders to their community and theirembeddedness in local social relations can also makesome difference. Community-based natural resourcemanagement is believed to increase accountabilitybecause the community is nearby and is closelyinvolved in the formulation and implementation ofnatural resource management (Hue 2000, 4). Theway authorities are embedded in social relationswithin their communities may help to make themmore accountable to the local population. Authoritieswithin the communities they govern must live withthe consequences of their decisions on people theyknow and who know them. This fact may influencetheir decision making. Clearly, different ideologies orbelief systems of leaders and their communities canalso have an influence on accountability relationsalthough these may be less amenable to policyintervention. For example, the Fon (Ghanaian chief)feels that the community is in his hands and that thisis a responsibility he has inherited.

Civic dedication can play an important role too.Tendler (1997, 15) shows that public sector workerscan be highly dedicated to their jobs. This was in thecontext where civil servants were given greaterautonomy than usual and performed quite well attheir jobs. “On the one hand, workers wanted toperform better in order to live up to the new trustplaced in them by their clients and citizens in

general. The trust was a result of the more custom-ized arrangements of their work and the publicmessages of respect from the state. On the otherhand,” Tendler continues, “the communities wherethese public servants worked watched over themmore closely. The state’s publicity campaigns andsimilar messages had armed citizens with newinformation about their rights to better governmentand about how public services were supposed towork.” Along these lines, awards for communityservice can be another accountability mechanism.Codes of conduct for politicians and civil servants canbe another means of moral leverage; these have beenapplied in South Africa, Tanzania, and Uganda(Therkildsen 2001, 27, fu. 26).

Reputations that societies hold people to and peoplewant to maintain can also shape their public andprivate behavior (Bordieu 1977). Trust is anotherelement that, if developed, is believed to improve theaccountability of local governments. Putnam (1993)argues that environments with numerous civil societyorganizations; social networks that link people togovernment, society, and business; and relationsbased on shared values and trust enjoy greater levelsof mutual accountability within state, market, andcivil society. This in turn is argued to lead to moreefficient government and a more synergetic relationbetween state and civil society. (Evans 1997;Bebbington and Kopp 1998, 13)

Administrative dependence on local elected authori-ties can increase administrative accountability to localpopulations. Blair (2000) points out that administra-tive bodies should be accountable to elected authori-ties and elected authorities should be accountable tothe people. Entrusting local government to manageservice and development activities within the publicdomain involves making public-resource users andpublic-service providers accountable to local govern-ment. This can be accomplished by enabling localgovernment to contract out service provision, toprovide it in-house when appropriate, and to havecontrol over the advisors and experts they hire or whooffer them services from NGOs or the central state.Creating mechanisms so that local government cancontract competitively with line ministries or privateservice providers and experts may effect the account-

113


with no graduation of taxation there is no officialtheory of inequality and no way for the poorer majorityto demand higher contributions from their wealthybrethren; with no property tax there is no basis for …growing outside business in the area to support itsdevelopment. . . . The material basis for a [Westernstyle] form of democratic struggle for accountabilityand control is more or less defunct” (Guyer 1992, 57,in Therkildsen 1993, 85).

Central government can play the role of ensuringaccountability of local government; it also must beaccountable for some services to local government.Central government’s oversight of local government canhelp ensure that local government is accountable to localpopulations (cf. Tendler 1997, 15). Uphoff and Esman(1974, xx) state, “Sanctions to control the acts of leadersof local organizations should be balanced both fromabove and from below to get the best performance”(emphasis in original). Parker (1995, 35) also argues forcentral monitoring and sanctions to “penalize institu-tions that do not carry out their functions appropriately.”Tendler (1997), however, cautions against this sanction-based approach, pointing out that greater degrees oflocal autonomy can improve government performanceof community services. (Cf. Evans 1997.) A system ofinternal performance audits can also ensure account-ability (Rothchild 1994, 6).

Social movements are another effective tool forholding governments accountable to their people(Mamdani and Wamba-dia-Wamba 1995). Resistanceand threats of resistance can motivate governments toact on behalf of concerned populations when otheraccountability mechanisms fail. Social movements,resistance, sabotage, and other forms of rebellion canbe effective ways for local populations to create adomain of local autonomy or to make governmentresponsive (Ribot 2000; cf. Scott 1976). However, asthe participants in a 1994 conference on decentraliza-tion in Ouagadougou agreed, while representation isnecessary, “citizen participation in local affairs isnecessarily limited when there are constraints on thefreedom of association and the liberty of expression”(Mbassi 1995, 28). Such freedoms are critical forenabling the formation of groups ranging from NGOs,peasant organizations, and vigilance committees, towhole social movements.

ability of these service providers (cf. Therkildsen2001, 27, fn. 26). This strategy attempts to createcompetition among service providers and incentivesfor providing better services. For this purpose it mayrequire central government to commit to providinglocal government with a budget—a budget drawnfrom the resources central government wouldordinarily have spent directly supporting line minis-tries—or it may require taxation powers.

Administrative dependence may also involve makingprivate organizations and NGOs accountable to localgovernment by requiring local government approvalfor the use and management of commonly held publicresources. For outside projects involving ad hoc orpermanent planning and coordination committees,local government authorities would have the finaldecision-making power over their activities to ensurethat these activities were under representative commu-nity control. The law can make local administration(even deconcentrated services) subordinate to electedlocal government. In this manner, local electedgovernment can have decision-making powers, vetopowers, or other forms of control over the decisions ofcentral government ministries in matters determinedto be within local jurisdiction.

According to some analysts, taxation arrangementscan seriously affect the relations of accountabilitybetween people and their governments. Moore (1997)has forcefully argued that governments that depend ontaxes derived from the earned income of their popula-tions are more likely to have populations that makedemands on government and hold their governmentsaccountable. (Therkildsen (2001, 30) has found thatuser charges serve a similar function.) Moore also hasargued the converse, that governments dependent onoutside assistance are not likely to be accountable totheir populations. (Also see Guyer 1992; Yates 1996.)In much of Africa, local governments have haddifficult times collecting taxes. Local tax rates andcollection have declined in countries such as Kenya,Nigeria, Tanzania, and Zimbabwe (Therkildsen 1993,84–5). In southern Nigeria, rates fell from eightpercent of income in 1968 to less than two percent forpoorer farmers by 1988. “With such low contributions… ,” writes Guyer, “financial management becomes apoor basis for people’s demands for accountability;

114


115


WORKS CITED

Alden Wily, Liz. 2000a. “Democratic woodland manage-ment in eastern and southern Africa: What’s happen-ing?” Mimeo. (Later published as issue paper no. 99,Drylands Programme, International Institute forEnvironment and Development, London.)

———. 2000b. “Sharing products or power? Intentions,meanings, and approaches to community involvementin forest management in east and southern Africa.”Draft paper, 7 Nov. Mimeo.

Alex, Gary, Chris Gerrard, Derek Byerlee, C. Annor-Frempong, Daniel Moreau, David Neilson, DeyGasasin, Eliseo Ponce, Graham Kerr, JaakkoKangasniemi, Joko Budianto, Mahmood Kamara,Marie-Hélène Collion, Matthew McMahon, MelissaWilliams, Solomon Bekure, and Willem Zjip. 2000.“Decentralizing agricultural extension: Lessons andgood practice.” Rural Development Family, Agricul-tural Knowledge and Information Systems, GoodPractice Note, August, World Bank.

Alexandre, Pierre. 1970. “The problems of chieftaincies inFrench-speaking Africa.” First published as “Leproblème des chefferies en Afrique noire française.”Notes et Etudes Documentaires, no. 2508 (10 Feb. 1959).Reprint in West African chiefs: Their changing statusunder colonial rule and independence, edited by MichaelCrowder and Obaro Ikime, 2–13. New York: AfricanaPublishing Company.

Anderson, John, Jean Clément, and Loy Van Crowder.1998. “Accommodating conflicting interests inforestry: Concepts emerging from pluralism.”Unasylva 49, no. 194: 3–10.

Anderson, Jon. 2002. Nature, wealth, and power: Emergingbest practice for revitalizing rural Africa. Washington,D.C.: United States Agency for International Develop-ment.

Able, N.O.J, and Piers Blaikie. 1990. “Land degradation,stocking rates and conservation policies in thecommunal rangelands of Botswana and Zimbabwe.”Pastoral network paper no. 29a, Overseas Develop-ment Institute, London.

Adams, William M., Dan Brockington, Jane Dyson, andBashkar Vira. 2003. “Managing tragedies: Under-standing conflict over common pool resources.”Science 302, 1915–6.

Agrawal, Arun. 2001. “The regulatory community:Decentralization and the environment in the vanpanchayats (forest councils) of Kumaon.” MountainResearch and Development 21, no. 3 (August): 208–11.

———. 1997. “Community in conservation: Beyondenchantment and disenchantment.” Draft of conserva-tion and development forum working paper no. 1, 18Mar., University of Florida, Gainesville.

Agrawal, Arun, and Clark C. Gibson, eds. 2001. Communi-ties and nature: Ethnicity, gender, and the state incommunity-based conservation. Piscataway, N.J.: RutgersUniversity Press.

Agrawal, Arun, and Jesse C. Ribot. 1999. “Accountabilityin decentralization: A framework with South Asianand African cases.” Journal of Developing Areas 33,summer: 473–502.

Ahwoi, Kwamena. 2000. “Challenges facing local govern-ments in Africa in the 21st Century: An executivesummary.” Local Government Perspectives 7, no. 3:insert, 1–7.

Alavi, Hamza. 1973. “Peasant classes and primordialloyalties.” Journal of Peasant Studies 1, no. 1 (October):23–98.

Alcorn, Janice B. 1999. “Community governance ofbiodiversity resources: Policy reforms that can make adifference.” The Good Society 9, no. 2: 42–6.

116


Andersson, Krister P. 2002. “Explaining the mixed successof municipal governance of forest resources in Bolivia:Overcoming local information barriers.” Center for theStudy of Institutions, Population, and EnvironmentalChange, Indiana University, Bloomington. Draft: 28Jan. Mimeo.

Atelier Populaire. 1969. Posters from the revolution: Paris,May 1968. London: Dobson Books, Ltd.

Aubréville, August. 1939. “L’utilsation totale des boixtropicaux de l’ouest Africain.” Revus des Eaux et Forets,June: 485–95 and 599–608.

Baland, Jean-Marie, and Jean-Philippe Platteau. 1996.Halting degradation of natural resources: Is there a role forrural communities? Oxford: Clarendon Press.

Balogun, M. J. 2000. “The scope for popular participationin decentralization, community governance, anddevelopment: Towards a new paradigm of centre-periphery relations.” Regional Development Dialogue 21,no. 1 (spring): 153–73.

Banerjee, Ajit, Gabriel Campbell, Chona Cruz, SheltonDavis, and Augusta Molnar. 1994. Participatoryforestry paper presented at a workshop on participa-tory development, The World Bank, Washington, D.C.,17–20 May.

Bardhan, Pranab. 1997. “Corruption and development: Areview of issues.” Journal of Economic Literature 35,September: 1320–46.

Bardhan, Pranab, and Dilip Mookherjee. 1999. “Relativecapture of local and central governments: An essay onthe political economy of decentralization.” First draftof paper presented at the American EconomicAssociation’s 2000 session on political economy,governance, and development, 30 Nov. Mimeo.

Bardhan, Pranab, Maitreesh Ghatak, and AlexanderKaraivanov. 2002. “Inequality, market imperfections,and voluntary provision of collective goods.” Depart-ment of Economics, University of California atBerkeley. 20 Apr. Mimeo.

Bassett, Thomas J., and Donald E. Crummey, eds. 1993.Land in African agrarian systems. Madison: Universityof Wisconsin Press.

Bates, Robert. 1981. Markets and states in tropical Africa.Berkeley: University of California Press.

Baviskar, Amita. 2004. “Between micro-politics andadministrative imperatives: Decentralization and thewatershed mission in Madhya Pradesh, India.” InDecentralization of natural resources: Experiences inAfrica, Asia and Latin America, edited by Jesse C. Ribotand Anne M. Larson. London: Frank Cass. Also

published in the European Journal of DevelopmentResearch 16, no. 1.

———. 2002. “Between micro-politics and administrativeimperatives: Decentralization and the watershedmission in Madhya Pradesh, India.” Paper presentedat WRI’s Conference on Decentralization and theEnvironment, Bellagio, Italy, 18–22 Feb. [Laterpublished as Baviskar 2004.]

———. Baviskar, Amita. Forthcoming. “The dreammachine: The model development project and theremaking of the state.” In Waterscapes: The culturalpolitics of a natural resource, edited by Amita Baviskar.Delhi: Permanent Black.

Bayart, Jean-François. 1993. The state in Africa: The politicsof the belly. London: Longman.

Bazaara, Nyangabyaki. Forthcoming. “Subjecting nature tocentral authority: The struggle over public goods in theformation of citizenship.” Africa Development.

———. 2003. “Decentralization, politics and environmentin Uganda.” Environmental Governance in Africaworking paper no. 7, WRI, Washington, D.C.

———. 2002a. “Ugandan decentralisation and thepromise of popular participation in environmentalmanagement: Byoya bya nswa (form without sub-stance)?” Working paper for Decentralization,Accountability, and the Environment ResearchProgram, Institutions and Governance Program, WRI.Jan. draft.

———. 2002b. “Actors, powers and environmentalaccountability in Uganda’s decentralization.” Paperpresented at WRI’s Conference on Decentralizationand the Environment, Bellagio, Italy, 18–22 Feb.

———. 2001. “From despotic to democratic decentraliza-tion in Uganda: History of control over nature.” Draftreport for the Programme on Environmental Account-ability in Decentralized Contexts, WRI and Centre forBasic Research, Kampala. Mimeo.

Bebbington, Anthony, and Adalberto Kopp. 1998. “Net-working and rural development through sustainableforest management: Frameworks for pluralisticapproaches.” Unasylva 49, no. 194: 11–18.

Beck, Linda J. 2001. “Reining in the marabouts? Democra-tization and local governance in Senegal.” AfricanAffairs, no. 401: 601.

Beinart, William, and Peter A. Coates. 1995. Environmentand history: The taming of nature in the USA and SouthAfrica. London: Routledge.

117


Benjaminsen, Tor A. 2000. “Conservation in the Sahel:Policies and people in Mali, 1900–1998.” In Producingnature and poverty in Africa, edited by Vigdis Broch-Due and Richard A. Schroeder. Stockholm: ElandersGotab.

———. 1997. “Natural resource management, paradigmshifts, and the decentralization reform in Mali.”Human Ecology 25, no. 2: 121–43.

Berry, Sara. 1993. No condition is permanent: The socialdynamics of agrarian change in sub-Saharan Africa.Madison: University of Wisconsin Press.

———. 1989. “Social institutions and access to resources.”Africa 59, no. 1.

Bertin, André. 1919. La Question Forestière Coloniale:Mission d’Études Forestières Envoyée dans les ColoniesFrançaise par les Ministères de la Guerre, de l’Armementet des Colonies. Paris: Émile Larose, Libraire-Éditeur.

Bhagwati, Jagdish N. 1982. “Directly unproductive, profit-seeking (DUP) activities.” Journal of Political Economy90, no. 5: 988–1002.

Bigombe Logo, Patrice. 2003. “The decentralized forestrytaxation system in Cameroon: Local management andstate logic.” Environmental Governance in Africaworking paper no. 10, WRI, Washington, D.C.

———. 2001. “Decentralisation de la gestion forestière etdéveloppement local au Cameroun: Economiepolitique de l’accountabilite et de la performance dansla gestion locale des revenus forestiers au sud-estCameroun.” Draft report to the WRI/Centre forInternational Forestry Research program on decen-tralization and the environment, Yaounde, February.Mimeo.

Bish, Robert L., and Vincent Ostrom. 1973. Understandingurban government. Washington, D.C.: AmericanEnterprise Institute for Public Policy Research.

Biya, Paul. 1986. Communal liberalism. London: MacMillan.

Blaikie, Piers. 1985. The political economy of soil erosion.London: Longman.

Blair, Harry. 2000. “Participation and accountability at theperiphery: Democratic local governance in sixcountries.” World Development 28, no. 1: 21–39.

———. 1997. “Spreading power to the periphery: Anassessment of democratic local governance—Executivesummary.” Food and Agriculture Organization of theUnited Nations, Rome.

Blundo, Giorgio, and Ronald Jaubert. 1998. “La gestionlocale des ressources—décentralisation et participa-tion: Rupture ou continuité des modes de gestion?” InDiscours et réalités des politiques participatives de gestion

de l’environnement: Le cas du Sénégal, edited by PeterUtting and Ronald Jaubert. Geneva: United NationsResearch Institute for Social Development.

Blundo, Giorgio, and Roch Mongbo, eds. 1999.“Décentralisation, pouvoirs sociaux et réseauxsociaux.” APAD Bulletin, no. 16.

Boone, Catherine. 2003. Political topographies of the Africanstate: Territorial authority and institutional choice.Cambridge: Cambridge University Press.

Botswana Community-Based Organisation Network(BOCOBONET). 2003. File:///Documents andSetting/chrisc/Local Settings/Temp/BOCOBONET.html.

Bourdieu, Pierre. 1977. Outline of a theory of practice.Cambridge: Cambridge University Press.

Bowles, Samuel, and Herbert Gintis. 1999. “Communityand reciprocity.” The Good Society 9, no. 2: 47–51.

Brannstrom, Christian. 2004. “Decentralising waterresource management in Brazil.” In Decentralization ofnatural resources: Experiences in Africa, Asia and LatinAmerica, edited by Jesse C. Ribot and Anne M. Larson.London: Frank Cass. Also published in the EuropeanJournal of Development Research 16, no. 1. [Laterpublished as Brannstrom 2004.]

———. 2002. “Decentralizing water-resources manage-ment in Brazil.” Paper presented at WRI’s Conferenceon Decentralization and the Environment, Bellagio,Italy, 18–22 Feb.

Brinkerhoff, Derick W. 2001. “Taking account of account-ability: A conceptual overview and strategic options.”Draft report for the Implementing Policy ChangeProject, Phase 2, Center for Democracy and Gover-nance, U.S. Agency for International Development.Abt. Associates, Inc., Washington, D.C. Mimeo.

Brock, Karen, and N’golo Coulibaly. 1999. “Sustainablerural livelihoods in Mali.” Research report no. 35,Sustainable Livelihoods Programme, Institute ofDevelopment Studies, Brighton.

Brooke, James. 2001. “Canada’s tribal women fight (mostlymale) graft.” New York Times, 1 Jan., A3.

Brown, David. 1999. “Principles and practice of forest co-management: Evidence from west-central Africa.”European Union tropical forestry paper no. 2,Overseas Development Institute, London.

Bruch, Carl, Wole Coker, and Chris VanArsdale. 2001.“Breathing life into fundamental principles: Imple-menting constitutional environmental protections inAfrica.” Environmental Governance in Africa workingpaper no. 2, WRI, Washington, D.C.

118


Brunner, Jake, and François Ekoko. 2000. “Cameroon.” InThe right conditions: The World Bank, structuraladjustment, and forest policy reform, edited by Frances J.Seymour and Navroz K. Dubash. Washington, D.C.:WRI.

Buell, Raymond Leslie. 1928. The native problem in Africa.Vols. 1 and 2. New York: MacMillan.

Cappon, Jan, and Jeremy Lind. 2000. Realities or rhetoric?Revisiting the decentralization of natural resourcesmanagement in Uganda and Zambia. Nairobi: ACTSPress.

CARE-Mali. 1993. “Ogokana.” Agriculture and NaturalResources technical report series, Projet d’Agro-Sylviculture Villageoise de Koro, December.

Cernea, Michael M. 1989. “Nongovernmental organiza-tions and local development.” Regional DevelopmentDialogue 10, no. 2: 117–42.

Cernea, Michael, ed. 1985. Putting people first: Sociologicalvariables in rural development. New York: OxfordUniversity Press.

Chambers, Robert. 1983. Rural development: Putting the lastfirst. Essex: Longman Scientific and Technical.

Chanock, Martin. 1991. “Paradigms, policies, and property:A review of the customary law of land tenure.” In Lawin colonial Africa, edited by Kristin Mann and RichardRoberts. Portsmouth, N.H.: Heinemann.

Chauveau, Jean-Pierre. 1994. “Participation paysanne etpopulisme bureaucratique. Essai d’histoire et desociologie de la culture de développement.” In Lesassociations paysannes en Afrique: Organisation etdynamiques, edited by Jean-Pierre Jacob and PhilippeLavigne Delville. Marseille: APAD, and Paris: Karthala.

Cheema, G. Sabbir, and Dennis A. Rondinelli, eds. 1983.Decentralization and development: Policy implementationin developing countries. Beverly Hills: Sage Publications.

Clark, Sherman J. 1998. “A populist critique of directdemocracy.” Harvard Law Review 112, no. 434.

Clauzel, J. 1995. “Données, enjeux et techniques de ladécentralisation en Afrique.” In La décentralisation enAfrique de l’Ouest, edited by J. P. Elong Mbassi. Reportof the Conference on Decentralization in West Africa,Ouagadougou, 5–8 April. Cotonou, Benin: PDM andCEDA.

Cohen, John M., and Norman Uphoff. 1977. “Ruraldevelopment participation: Concepts and measures forproject design, implementation, and evaluation.” RuralDevelopment Committee, Cornell University, Ithaca.

Cohen, John M., and Stephen B. Peterson. 1997. “Adminis-trative decentralization: A new framework for im-proved governance, accountability, and performance.”Development discussion paper no. 582, HarvardInstitute for International Development, HarvardUniversity.

Coleman, James S. 1988. “Social capital in the creation ofhuman capital.” American Journal of Sociology 94:supplement, 95–121.

Conyers, Diana. 2002. “Whose elephants are they?Decentralization of control over wildlife managementthrough the CAMPFIRE programme in Binga District,Zimbabwe.” WRI working paper draft, May. Mimeo.Later published as Environmental Governance inAfrica working paper no. 4, WRI, Washington, D.C.

———. 2000a. “Decentralisation: A conceptual analysis.Pt. 1.” Local Government Perspectives: News and Views onLocal Government in Sub-Saharan Africa 7, no. 3: 7–9,13.

———. 2000b. “Decentralisation: A conceptual analysis.Pt. 2.” Local Government Perspectives: News and Views onLocal Government in Sub-Saharan Africa 7, no. 4: 18–24.

———. 1991. “Decentralization and development plan-ning: A comparative perspective.” In Decentralizing forparticipatory planning: Comparing the experiences ofZimbabwe and other anglophone countries in eastern andsouthern Africa, edited by Peter de Valk and Kadmiel H.Wekwete. Aldershot: Avebury Press.

———. 1983. “Decentralization: The latest fashion indevelopment administration?” Public Administrationand Development 3, no. 1: 97–109.

Cooke, Bill. 2001. “The social psychological limits ofparticipation?” Chap. 7 in Participation: The newtyranny? edited by Bill Cooke and Uma Kothari, 102–21. London: Zed Books.

Cooke, Bill, and Uma Kothari, eds. 2001. Participation: Thenew tyranny? London: Zed Books.

Cooper, Frederick. 1996. Decolonization and African society:The labor question in French and British Africa. Cam-bridge: Cambridge University Press.

Cousins, Ben. 2002. “Decentralization when land andresource rights are deeply contested: A case study ofthe Mkambati eco-tourism project on the Wild Coast ofSouth Africa.” Paper presented at WRI’s Conferenceon Decentralization and the Environment, 18–22 Feb.,Bellagio, Italy. [Later published as Cousins and Kepe2004]

119


Cousins, Ben, and Tembele Kepe. 2004. “Decentralisationwhen land and resource rights are deeply contested: Acase study of the Mkambati eco-tourism project on theWild Coast of South Africa.” In Decentralization ofnatural resources: Experiences in Africa, Asia and LatinAmerica, edited by Jesse C. Ribot and Anne M. Larson.London: Frank Cass. Also published in the EuropeanJournal of Development Research 16, no. 1.

Cowan, Richard L. Gray. 1958. Local government in WestAfrica. New York: Columbia University Press.

Crook, Richard C., and James Manor. 1998. Democracy anddecentralization in Southeast Asia and West Africa:Participation, accountability, and performance. Cam-bridge: Cambridge University Press.

———. 1994. “Enhancing participation and institutionalperformance: Democratic decentralization in SouthAsia and West Africa.” Report on research project,Overseas Development Administration. January.

Crook, Richard C., and Alan Sturia Sverrisson. 2001.“Decentralization and poverty-alleviation in developingcountries: A comparative analysis, or is West Bengalunique?” Working paper no. 130, Institute of Develop-ment Studies, Brighton.

Crowder, Michael, and Obaro Ikime, eds. 1970. WestAfrican chiefs: Their changing status under colonial ruleand independence. New York: Africana PublishingCompany.

Cruise-O’Brien, Donald. 1975. Saints and politicians: Essaysin the organization of a Senegalese peasant society.London: Cambridge University Press.

Dahl, Robert. 1981. “The city in the future of democracy.”In Politics and government of urban Canada, edited by L.D. Feldman. London: Methuen.

de Grassi, Aaron. 2003. “Constructing subsidiarity,consolidating hegemony: Political economy and agro-ecological processes in Ghanaian forestry.” Environ-mental Governance in Africa working paper no. 13,WRI, Washington, D.C.

de Tocqueville, Alexis. 1835. Democracy in America. Reprint,New York: Vintage, 1945.

de Valk, Peter. 1991. “State, decentralization, and participa-tion.” In Decentralizing for participatory planning:Comparing the experiences of Zimbabwe and otheranglophone countries in eastern and southern Africa,edited by Peter de Valk and Kadmiel H. Wekwete.Aldershot: Avebury Press.

de Valk, Peter, and Kadmiel H. Wekwete, eds. 1991.Decentralizing for participatory planning: Comparing theexperiences of Zimbabwe and other anglophone countries ineastern and southern Africa. Aldershot: Avebury Press.

Degnobl, Tove. 1995. “The terroir approach to naturalresource management––Panacea or phantom?—TheMalian experience.” Papers of the fifth annualconference of the International Association ofCommon Property.

Delevoy, G. 1923. “La Conservation des Forêts Coloniales.”Bulletin de la société Centrale Forestiere de Belqiue,September: 465–87, and October: 542–78.

Delnooz, Pascale. 1999. “Gestion des ressourcesforestières: La communauté, l’etat et le marché: Étudede projets d’aménagament au Burkina Faso.” Ph.D.dissertation, Fondation UniversitaireLuxembourgeoise, Louvain-la-Neuve, Belgium.December.

Delville, Philippe Lavigne. 1999. “Harmonising formal lawand customary land rights in French-speaking WestAfrica.” Issue paper no. 86, Drylands Programme,International Institute for Environment and Develop-ment, London.

Dembélé, Eli, and Mamadou Augustin Dembélé. 2001.“Communication de la DDRS/CMDT–MALI :Experience de gestion de terroir villageois ‘SIWAA,Brousse Seche.’” WRI Workshop on Policy Research,Analysis and Findings of: The Program on Account-ability, Decentralization and the Environment: LocalDemocracy and Natural Resources in Sub-SaharanAfrica, 15–19 Oct. 2001, Cape Town, South Africa.

Deschamps, Hubert. 1963. “Et maintenant, Lord Lugard?”Africa: Journal of the International African Institute 22,no. 4: 293–306.

Diallo, Mahamadou A. 1994. “Problématique de ladécentralisation au Niger.” Preparatory document forthe Praia Conference, CILSS, Republic of Niger.Financed by the Netherlands and the U.S. Agency forInternational Development–Niger. May.

Diaw, Mariteuw Chimère, and Phil René Oyono. 1998.“Instrumentalité et déficit des itinéraires dedécentralisation de la gestion des ressources naturellesau Cameroun.” Bulletin Arbres, Forêts et CommunautésRurales, no. 15/16 (December): 20–25.

Dillinger, W. 1994. “Decentralization and its implicationsfor urban service delivery.” Urban ManagementProgram discussion paper no. 16, World Bank,Washington, D.C.

Dove, Michael R., and Daniel M. Kammen. 2001. “Ver-nacular models of development: An analysis ofIndonesia under the ‘new order.’” World Development29, no. 4: 619–39.

120


Downs, R. E., and S. P. Reyna, eds. 1988. Land and societyin contemporary Africa. Hanover, N.H.: UniversityPress of New England.

Dupar, Mairi, and Nathan Badenoch. 2002. “Environment,livelihoods, and local institutions: Decentralization inmainland Southeast Asia.” Washington, D.C.: WRI.

Echeverri-Gent, John. 1992. “Public participation andpoverty alleviation: The experience of reform commu-nists in India’s West Bengal.” World Development 20,no. 10: 1401–22.

Ekoko, François. 2000. “Environmental adjustment inCameroon: Challenges and opportunities for policyreform in the forestry sector.” Draft paper for WRI.

Elbow, Kent M. 1996. “Legislative reform, tenure, andnatural resource management in Niger: The new ruralcode.” The Land Tenure Center, Madison, Wis., forCILSS and the U.S. Agency for International Develop-ment. May.

Engberg-Pedersen, Lars. 1995. “Creating local democraticpolitics from above: The ‘gestion des terroirs’ approachin Burkina Faso.” Issue paper no. 54, DrylandsProgramme, International Institute for Environmentand Development, London.

Ergas, Zaki, ed. 1987. The African state in transition. NewYork: St. Martin’s Press.

Escobar, Arturo. 1995. Encountering development: Themaking and unmaking of the third world. Princeton:Princeton University Press.

Etoungou, Patrice. 2003. “Decentralization viewed frominside: The implementation of community forests inEast Cameroon.” Environmental Governance in Africaworking paper no. 12, WRI, Washington, D.C.

———. 2001. “L’impense des forêts communautaires:Décentralisation à l’est du Cameroun.” Draft paper forthe Programme on Decentralization and the Environ-ment, WRI and Center for International ForestryResearch–Cameroon, June. Yaoundé. Mimeo.

Evans, Peter B. 1997. “The eclipse of the state? Reflectionson stateness in an era of globalization.” World Politics50, October: 62–87.

Fairhead, James, and Melissa Leach. 1996. Misreading theAfrican landscape: Society and ecology in a forest-savannamosaic. Cambridge: Cambridge University Press.

Farvar, M. Taghi. 1997. “Decentralization and naturalresource management—issues in rural development,”Paper presented at the Technical Consultation onDecentralization, Rome 16-18 December 1997. WorldBank/FAO.

Ferguson, James. 1996. “Transnational topographies ofpower: Beyond ‘the state’ and ‘civil society’ in the studyof African politics.” Article manuscript, April 1996.Mimeo.

———. 1994. The anti-politics machine: “Development,”depoliticization, and bureaucratic power in Lesotho.Minneapolis: University of Minnesota Press.

Fisher, Michael H. 1991. Indirect rule in India: Residents andthe residency system, 1764–1858. New Delhi: OxfordUniversity Press.

Fisher, William F. 1997. “Doing good? The politics andantipolitics of NGO practices.” Annual Review ofAnthropology 26, 439–64.

Fisiy, Cyprian F. 1995. “Chieftaincy in the modern state:An institution at the crossroads of democraticchange.” Paideuma 41: 49–62.

———. 1992. “Power and privilege in the administrationof law: Land law reforms and social differentiation inCameroon.” Research report no. 48, African StudiesCentre, Leiden, Netherlands.

Fiszbein, Ariel. 1997a. “Decentralization and local capacity:Some thoughts on a controversial relationship.” Paperpresented at the Food and Agriculture Organization/United Nations Capital Development Fund/WorldBank technical consultation on decentralization,Rome, 15–18 Dec.

———. 1997b. “The emergence of local capacity: Lessonsfrom Colombia.” World Development 25, no. 7: 1029–43.

Føllesdal, Andreas. 1998. “Survey article: Subsidiarity.” TheJournal of Political Philosophy 6, no. 2: 190–218.

Foltz, William J. 1965. From French West Africa to the Malifederation. New Haven: Yale University Press.

Fomété, T. 2002. “La fiscalité forestière et l’implication desmommunautés locales à la gestion forestière auCameroun.” Rural Development Network paper no.25b, Overseas Development Institute, London.

Fortes, M., and E. E. Evans-Pritchard. 1940. African politicalsystems. Reprint, London: KPI Press, 1987.

Fox, Jonathan. 1990. Editor’s introduction. The challenge ofrural democratization: Perspectives from Latin Americaand the Philippines. Special issue of The Journal ofDevelopment Studies 26, no. 4 (July): 8–18.

Fox, Jonathan, and L. David Brown, eds. 1998. The strugglefor accountability: The World Bank, NGOs, and grassrootsmovements. Cambridge, Mass.: Massachusetts Instituteof Technology Press.

Gellar, Sheldon. 1995. Senegal: An African nation betweenIslam and the West. Boulder, Colo.: Westview Press.

121


Geschiere, Peter. 2003. “Autochthony and the struggle overcitizenship: New modes in the politics of belonging inAfrica and elsewhere.” First draft of discussion paperfor new Social Science Research Council programme.Mimeo.

———. 1993. “Chiefs and colonial rule in Cameroon:Inventing chieftaincy, French and British style.” Africa63, no. 2: 151–75.

GGAOF (Gouvernement Général de l’Afrique OccidentaleFrançaise). 1941. No. 2092 FOR.-Arrêté foxant lesmodalites d’application dans la colonie du Sénégal del’arrêté general ni. 3782 du 15 Novembre 1938,reglementant les exploitations industrielles de bois defeu et de charbon, Saint-Louis le 28 Juillet 1941.Journal Officiel du Sénégal, 443–5.

———. 1935. No. 1704 A.P.—Arrêté promulgant enAfrique occidentale française le décret du 4 juillet1935, sur le régime forestier de l’Afrique occidentalefrançaise. Journal Officiel du Sénégal, 24 July 1935, pp.599–606.

Goldman, Mara. 2001. “Partitioned nature, privilegedknowledge: Community-based conservation in theMaasai ecosystem, Tanzania.” Draft paper for theProgramme on Decentralization and the Environment,WRI. Madison, June. Mimeo. Later published asEnvironmental Governance in Africa working paperno. 3, WRI, Washington, D.C.

Gore, C. 1984. Relations in question. London: Methuen.

Government of Burkina Faso. 1993. Loi no. 007/93/ADP,Portant sur le régime électoral des conseillers devillage, de secteur communal, de départment et deprovince. Government of Burkina Faso.

Government of Cameroon. 1998. “Manual of the proce-dures for the attribution, and norms of the manage-ment, of community forests.” Ministry of Environ-ment and Forests. Editions CLE, Yaoundé.

Granovetter, Mark. 1985. “Economic action and socialstructure: The problem of embeddedness.” AmericanJournal of Sociology 91, no. 3: 481–510.

Graziani, Monica, and Philip Burnham. 2002. “Legalpluralism in the rain forests of southeasternCameroon.” In Rural resources and local livelihoods inAfrica, edited by Katherine Homewood. Oxford: JamesCurry. (Based on 2001 draft, mimeo).

Green, Philip, ed. 1993. Democracy: Key concepts in criticaltheory. New Jersey: Humanities Press.

Guha, Ramchandra. 1990. “An early environmental debate:The making of the 1878 Forest Act.” The IndianEconomic and Social History Review 27, no. 1.

Guyer, Jane. 1994. “The spatial dimensions of civil societyin Africa: An anthropologist looks at Nigeria.” In Civilsociety and the state in Africa, edited by John W.Harbeson, Donald Rothchild, and Naomi Chazan.Boulder, Colo.: Lynne Rienner.

———. 1992. “Representation without taxation: An essayon democracy in rural Nigeria, 1952–1990.” AfricanStudies Review 35, no. 1: 41–79.

Hailey, Lord. 1938. An African survey: A study of problemsarising in Africa south of the Sahara. London: OxfordUniversity Press.

Hardin, Rebecca. 2002. “Concessionary politics in thewestern Congo basin: History and culture in forestuse.” Environmental Governance in Africa workingpaper no. 6, WRI, Washington, D.C.

Heller, Patrick. 1996. “Social capital as a product of classmobilization and state intervention: Industrial workersin Kerala, India.” World Development 24, no. 6: 1055–71.

Helmsing, A. H. J. (Bert). 2001. “Local economic develop-ment: New generations of actors, policies, andinstruments.” Summary report for the United NationsCapital Development Fund Symposium on Decentrali-zation and Local Governance in Africa, Cape Town,South Africa, 26–30 Mar.

Henninger, Norbert, and Mathilde Snel. 2002. Where arethe poor? Experiences with the development and use ofpoverty maps. Washington, D.C.: WRI.

Hesseling, Gerti. 1996. “Legal and institutional incentivesfor local environmental management.” In Improvednatural resource management: The role of formalorganisations and informal networks and institutions,edited by Henrik Secher Marcussen. Occasional paperno. 17, International Development Studies, RoskildeUniversity, Denmark.

———. 1994. “Legal and institutional conditions for localmanagement of natural resources: Mali.” In Landtenure and sustainable land use, edited by R. J. Bakema,31–46. Royal Tropical Institute, Amsterdam, Bulletin332.

Hesseling, Gerti, in collaboration with M. Sypkens Smit.N.d. (c. 1984). “Le droit foncier au Sénégal: L’impactde la réforme foncière en Basse Casamance.” Mimeo.

Hildyard, Nicholas, Pandurang Hegde, Palu Wolvekamp,and Somasekhare Reddy. 2001. “Pluralism, participa-tion, and power: Joint forest management in India.”Chap. 4 in Participation: The new tyranny? edited byBill Cooke and Uma Kothari, 56–71. London: ZedBooks.

122


Hilhorst, Theta, and Nettie Aarnink. 1999. “Co-managingthe commons: Setting the stage in Mali and Zambia.”Royal Tropical Institute, Amsterdam, Bulletin 346.

Hirschman, Albert O. 1970. Exit, voice, and loyalty:Responses to decline in firms, organizations, and states.Cambridge: Harvard University Press.

Hoben, Allan, Pauline Peters, and Dianne Rocheleau.1998. “Participation, civil society, and foreign assis-tance to Africa.” In Africa’s valuable assets: A reader innatural resource management, edited by Peter G. Veit.Washington, D.C.: WRI.

Hubert, Henry. 1920. “Le desséchement progressif enAfrique Occidentale.” Comité d’Études Historiques etScientifiques de l’Afrique Occidentale Française.

Hudock, Ann C. 1999. NGOs and civil society: Democracy byproxy? Cambridge: Policy Press.

Hue, Le Thi Van. 2000. “Institutional arrangements forcommunity-based mangrove forest management inGiao Lac village, Giao Thuy District, Nam DinhProvince, Vietnam.” Paper presented at the Workshopon Institutions and Uncertainty, Institute of Develop-ment Studies, Brighton, 6–8 Nov.

Huther, Jeff, and Anwar Shah. 1998. “Applying a simplemeasure of good governance to the debate on fiscaldecentralization.” Policy research working paper no.1894, World Bank, Washington, D.C.

Hyden, Goran. 1980. Beyond Ujamaa in Tanzania:Underdevelopment and an uncaptured peasantry.Berkeley: University of California Press.

Jeter, Jon. 2000. “In South Africa, tribal ways vs. moderngovernment.” Washington Post, 18 Dec., A1.

Joshi, Anuradha. 2000. Roots of change: front line workersand forest policy reform in West

Bengal. Doctoral Dissertation, Cambridge, MA: Depart-ment of Urban Studies and Planning, MassachusettsInstitute of Technology.

Kaimowitz, David, Gonzalo Flores, James Johnson, PabloPacheco, Iciar Pavéz, J. Montgomery Roper, CristianVallejos, and Roger Vélez. 2000. “Local governmentand biodiversity conservation: A case from theBolivian lowlands.” Draft report for the BiodiversitySupport Program, World Wildlife Fund, Washington,D.C.

Kaimowitz, David and Jesse C. Ribot. 2002. “Services andinfrastructure versus natural resource management:Building a base for democratic decentralization.”Paper presented at WRI’s Conference on Decentraliza-tion and the Environment, Bellagio, Italy, 18–22 Feb.

Karlström, Mikael. 1996. “Imagining democracy: Politicalculture and democratization in Buganda.” Africa 66,no. 4.

Kassibo, Bréhima. 2004. “Historical and political founda-tions for participatory management and democraticdecentralization in Mali: A synthesis of two casestudies.” Draft Environmental Governance in Africaworking paper, WRI, Washington, D.C.

———. 2002a. “Participatory management and democraticdecentralization management of the Samori Forest inBabye Commune, Mopti Region, Mali.” Paperpresented at WRI’s Conference on Decentralizationand the Environment, Bellagio, Italy, 18–22 Feb.

———. 2002b. “Décentralisation et responsabilité dans lesmodes de gestion des ressources naturelles et desredevances y afférentes. Les Bourgoutières Yallarbe deYouwarou.” Draft Environmental Governance in Africaworking paper, WRI, Washington, D.C. Mimeo.

———. 2001. “Résurgence des autorités coutumières ausein de l’etat moderne africain:

Le cas des chefferies traditionnelles au Mali.” Researchproposal for WRI, May. Mimeo.

______, ed. 1998. “La décentralisation au Mali: Etat deslieux.” APAD Bulletin, no. 14.

Kassimir, Ronald. 2001. “Producing local politics: Gover-nance, representation, and non-state organizations inAfrica.” In Intervention and transnationalism in Africa:Global-local networks of power, edited by ThomasCallaghy, Ronald Kassimir, and Robert Lutham.Cambridge: Cambridge University Press.

Keeley, James, and Ian Scoones. 1999. “Understandingenvironmental policy processes: A review.” Workingpaper no. 89, Institute of Development Studies,Brighton.

Kini, Rémi. 1995. “An evaluation study of natural forestmanagement for wood fuel production in the Sahel: ABurkina Faso case.” Document submitted toward M.S.degree, Department of Resource Development,Michigan State University.

Krueger, Anne O. 1974. “The political economy of rent-seeking society.” American Economic Review 64, no. 3(June).

Larson, Anne M. 2004a. “Democratic decentralization inthe forestry sector: Lessons learned from Africa, Asia,and Latin America.” Paper prepared for the UnitedNations Forestry Forum, Interlaken, April 2004.

———. 2004b. “Formal decentralization and the impera-tive of decentralization ‘from below’: A case study ofnatural resource management in Nicaragua.” InDecentralization of natural resources: Experiences in

123


Africa, Asia and Latin America, edited by Jesse C. Ribotand Anne M. Larson. London: Frank Cass. Alsopublished in the European Journal of DevelopmentResearch 16, no. 1.

———. 2003. “Decision making under decentralization:Local governments in natural resource management.”December 2003 draft. Mimeo.

———. 2002. “Decentralization and natural resourcemanagement: A Nicaraguan case study.” Paperpresented at WRI’s Conference on Decentralizationand the Environment, Bellagio, Italy, 18–22 Feb. [Laterpublished as Larson 2004b]

Larson, Anne M., and Jesse C. Ribot. 2004. “Decentraliza-tion through a natural resource lens: An introduction.”In Decentralization of natural resources: Experiences inAfrica, Asia and Latin America, edited by Jesse C. Ribotand Anne M. Larson. London: Frank Cass. Alsopublished as the introduction to a special issue of theEuropean Journal of Development Research 16, no. 1.

Latif, Mehr. 2002. “Minutes of the World ResourcesInstitute Conference on Decentralization and theEnvironment,” Bellagio, Italy, 18–22 Feb. Environmen-tal Governance in Africa working paper, WRI,Washington, D.C.

Leach, Melissa, and Robin Mearns, eds. 1996. The lie of theland: Challenging received wisdom on the Africanenvironment. Oxford: James Currey Publishers Ltd. andHeinemann.

Lewis, B. D. 1998. “The impact of public infrastructure onmunicipal economic development: Empirical resultsfrom Kenya.” Review of Urban and Regional Develop-ment Studies 10, no. 2.

Lissu, Tundu A. 2000. “Repackaging authoritarianism:Freedom of association and expression and the right toorganize under the proposed NGO policy for Tanza-nia.” Lawyers’ Environmental Action Team researchreports, August. http://www.leat.or.tz/publications/authoritarianism/.

Little, Peter D. 1994. “The link between local participationand improved conservation: A review of issues andexperiences.” In Natural connections: Perspectives incommunity-based conservation, edited by DavidWestern, R. Michael Wright, and Shirley C. Strum.Washington, D.C.: Island Press.

Lonsdale, John. 1986. “Political accountability in Africanhistory.” In Political domination in Africa, edited byPatrick Chabal. Cambridge: Cambridge UniversityPress.

Lund, Christian. 1998. Law, power, and politics in Niger:Land struggles and the rural code. Hamburg: Lit Verlag.

Mair, Lucy P. 1936. Native policies in Africa. New York:Negro University Press.

Makara, Sabiti. 2000. “Decentralization for good gover-nance and development: Uganda’s experience.”Regional Development Dialogue 21, no. 1 (spring): 73–92.

Mamdani, Mahmood. 1996a. Citizen and subject: Contem-porary Africa and the legacy of late colonialism.Princeton: Princeton University Press.

———. 1996b. “Indirect rule, civil society, and ethnicity:The African dilemma.” Social Justice 23, no. 1/2(spring/summer): 145–6.

Mamdani, Mahmood, and Ernest Wamba-dia-Wamba.1995. African studies in social movements and democracy.Dakar, Senegal: Council for the Development of SocialScience Research in Africa.

Mandondo, Alois. 2000a. “Forging (un)democraticresource governance systems from the relic ofZimbabwe’s colonial past.” Institute of EnvironmentalStudies, University of Zimbabwe, and Center forInternational Forestry Research. Mimeo.

———. 2000b. “Situating Zimbabwe’s natural resourcegovernance systems in history.” Occasional paper no.32, Center for International Forestry Research, Bogor,Indonesia.

Mandondo, Alois, and Everisto Mapedza. 2003. “Allocationof government authority and responsibility in tieredgovernance systems: The case of environment-relatedlaws in Zimbabwe.” Environmental Governance inAfrica working paper no. 11, WRI, Washington, D.C.

Manin, Bernard, Adam Przeworski, and Susan Stokes.1999. “Elections and representation.” In Democracy,accountability, and representation, edited by Manin,Przeworski, and Stokes. Cambridge: CambridgeUniversity Press.

Mann, Kristin, and Richard Roberts, eds. 1991. Law incolonial Africa. Portsmouth, N.H.: Heinemann.

Manor, James. 2004. “User committees: A potentiallydamaging second wave of decentralization?” InDecentralization of natural resources: Experiences inAfrica, Asia and Latin America, edited by Jesse C. Ribotand Anne M. Larson. London: Frank Cass. Alsopublished in the European Journal of DevelopmentResearch 16, no. 1.

———. 2000. “Local government in South Africa:Potential disaster despite genuine promise.” Paperprepared for the United Kingdom’s Department forInternational Development. Institute of DevelopmentStudies, Brighton. Mimeo.

124


———. 1999. The political economy of democratic decentrali-zation. Washington, D.C.: World Bank.

Mansuri, Ghazala, and Vijayendra Rao. 2003. “Evaluatingcommunity-driven development: A review of theevidence.” First draft report, Development ResearchGroup, World Bank, February.

Mapedza, Everisto, and Alois Mandondo. 2002. “Co-management in the Mafungautsi state forest area ofZimbabwe—What stake for local communities?”Environmental Governance in Africa working paperno. 5, WRI, Washington, D.C.

Mawhood, Philip. 1983. Local government in the third world.Chichester: John Wiley.

Mazonde, Isaac Ncube. 1996. “The Basarwa of Botswana:Leadership, legitimacy, and participation in develop-ment sites.” Cultural Survival Quarterly, fall. Specialissue on participation, edited by Pauline Peters.

Mbassi, J. P. Elong, ed. 1995. La décentralisation en Afriquede l’Ouest. Report of the Conference on Decentraliza-tion in West Africa, Ouagadougou, 5–8 April. Cotonou,Benin: PDM and CEDA.

McCay, Bonnie. 1998. Oyster wars and the public trust:Property, law, and ecology in New Jersey history. Tucson:University of Arizona Press.

McIntosh, Alastair. 1990. “Rethinking chieftaincy and thefuture of rural local government: A preliminaryinvestigation.” Transformation (Durban, South Africa),no. 13: 27–45.

Mearns, Robin. 2004. “Decentralization, rural livelihoods,and pasture-land management in post-socialistMongolia.” In Decentralization of natural resources:Experiences in Africa, Asia and Latin America, edited byJesse C. Ribot and Anne M. Larson. London: FrankCass. Also published in the European Journal ofDevelopment Research 16, no. 1.

———. 2002. “The consequence of post-socialist transi-tion and decentralization for rural livelihoods andpastureland management in Mongolia.” Paperpresented at WRI’s Conference on Decentralizationand the Environment, Bellagio, Italy, 18–22 Feb. [Laterpublished as Mearns 2004.]

Mehta, Ajay S. 1996. “Micro politics of voluntary action:An anatomy of change in two villages.” CulturalSurvival Quarterly, fall. Special issue on participation,edited by Pauline Peters.

Meinzen-Dick, Ruth, and Anna Knox. 1999. “Collectiveaction, property rights, and devolution of naturalresource management: A conceptual framework.”Draft paper for workshop, 15 July. Mimeo.

Meinzen-Dick, Ruth, and Rajendra Pradhan. 2000.“Implications of legal pluralism for natural resourcemanagement.” Paper presented at the Workshop onInstitutions and Uncertainty, Institute of DevelopmentStudies, Brighton, 6–8 Nov. Mimeo.

Melo Farrera, Fernando. 2002. “Local diversity to contendwith globalization in Oaxaca, Mexico.” Paper presentedat WRI’s Conference on Decentralization and theEnvironment, Bellagio, Italy, 18–22 Feb.

Meynen, Wicky, and Martin Doornbos. 2004. “Decentraliz-ing natural resource management: A recipe forsustainability and equity?” In Decentralization ofnatural resources: Experiences in Africa, Asia and LatinAmerica, edited by Jesse C. Ribot and Anne M. Larson.London: Frank Cass. Also published in the EuropeanJournal of Development Research 16, no. 1.

Mish, Frederick C., ed. 1991. Webster’s ninth new collegiatedictionary. Springfield, Mass.: Merriam-Webster.

Mkandawire, Thandika. 1995. “Beyond crisis: Towarddemocratic developmental states in Africa.” Paperpresented at Crises, Conflicts, and Transformations:Responses and Perspectives, the eighth generalassembly of the Council for the Development of SocialScience Research in Africa, 26 June–2 July.

Moore, Mick. 1997. “Death without taxes: Democracy, statecapacity, and aid dependence in the fourth world.”Draft. Later published in Towards a democraticdevelopmental state, edited by G. White and M.Robinson. Oxford: Oxford University Press, 1998.

Moorhead, R. 1989. “Changes taking place in commonproperty resource management in the inland delta ofMali.” In Common property resources: Ecology andcommunity-based sustainable development, edited by F.Berkes. London: IUCN/Belhaven Press.

Mosse, David. 2001. “‘People’s knowledge,’ participation,and patronage: Operations and representations inrural development.” Chap. 2 in Participation: The newtyranny? edited by Bill Cooke and Uma Kothari, 16–36.London: Zed Books.

Muhereza, Frank Emmanuel. 2003. “Environmentaldecentralization and the management of forestresources in Masindi District, Uganda.” Environmen-tal Governance in Africa working paper no. 8, WRI,Washington, D.C. Mimeo.

———. 2001a. “Concept note on the need to rethink theissue of private forests.” Kampala, Uganda. February.Mimeo.

———. 2001b. “Environmental decentralization and themanagement of community forests in Pakanyisubcounty, Masindi District.” Draft report for the

125


Programme on Decentralization and the Environment,WRI and Centre for Basic Research, Kampala.February. Mimeo.

Musgrave, Richard A., ed. 1965. Essays in fiscal federalism.Washington, D.C.: The Brookings Institution.

Mutizwa-Mangiza, Naison D. 2000. “Comment.” RegionalDevelopment Dialogue 21, no. 1 (spring): 23–5. (Thiscommentary is on “Decentralization for good gover-nance and development: The unending debate,”written by Walter O. Oyugi. Regional DevelopmentDialogue 21, no. 1 (spring): 3–22.)

Mwebaza, Rose. 1999. “How to integrate statutory andcustomary tenure: The Uganda case.” Issue paper no.83, Drylands Programme, International Institute forEnvironment and Development, London.

Namara, Agrippinah, and Xavier Nsabagasani. 2003.“Decentralization and wildlife management: Devolv-ing rights or shedding responsibility? Bwindi Impen-etrable National Park, Uganda.” EnvironmentalGovernance in Africa working paper no. 9, WRI,Washington, D.C.

———. 2001. “Decentralised governance and the wildlifemanagement sector: Whose interests matter? The caseof Bwindi Impenetrable National Park, Uganda.” Draftworking paper for the Programme on Decentralizationand the Environment, WRI and Centre for BasicResearch, Kampala. June. Mimeo.

National Research Council. 1992. Democratization inAfrica: African views, African voices, edited by Sahr JohnKpundeh. Washington, D.C.: National Academy Press.

Ndegwa, Stephen N. 2002. “Decentralization in Africa: Astocktaking survey.” Africa Region working paper no.40, World Bank, Washington, D.C.

Ngaido, Tidiane. 1996. “Redefining the boundaries ofcontrol: Post-colonial tenure policies and dynamics ofsocial and tenure change in western Niger.” Ph.D.dissertation, University of Wisconsin, Madison.

Nkrumah, Stephen A. 2000. “Decentralization for goodgovernance and development: The Ghanaian experi-ence.” Regional Development Dialogue 21, no. 1 (spring):53–67.

Nsibambi, Apolo, ed. 1998. Decentralisation and civil societyin Uganda: The quest for good governance. Kampala:Fountain Publishers.

Ntsebeza, Lungisile. 2004. “Democratic decentralizationand traditional authority: Dilemmas of land adminis-tration in rural South Africa.” In Decentralization ofnatural resources: Experiences in Africa, Asia and LatinAmerica, edited by Jesse C. Ribot and Anne M. Larson.

London: Frank Cass. Also published in the EuropeanJournal of Development Research 16, no. 1.

———. 2003. “Local governance, power and naturalresources: A perspective from the rural areas of SouthAfrica’s former Bantustans.” Environmental Gover-nance in Africa working paper no. 14, WRI, Washing-ton, D.C. Mimeo.

———. 2002. “Decentralization and natural resourcemanagement in rural South Africa: Problems andprospects.” Paper submitted to the Conference onDecentralization and the Environment, Bellagio, Italy,18–22 Feb. [Later published as Ntsebeza 2004.]

———. 1999. “Land tenure reform in South Africa: Anexample from the Eastern Cape Province.” Issue paperno. 82, Drylands Programme, International Institutefor Environment and Development, London.

Oates, William E. 1972. The political economy of fiscalfederalism. Lexington, Mass.: Lexington Books.

Oloka-Onyango, Joe. 1994. “Judicial power and constitu-tionalism in Uganda: A historical perspective.” InUganda: Studies in living condi-tions, popular move-ments, and constitutionalism, edited by MahmoodMamdani and Joe Oloka-Onyango. Journal fürEntwicklungspolitik and Centre for Basic Research,Kampala.

Olowu, Dele. 2001. “Local political and institutionalstructures and processes.” Summary report preparedfor the United Nations Capital Development FundSymposium on Decentralization and Local Gover-nance in Africa, Cape Town, South Africa, 26–30 Mar.

Onibon, A., B. Dabiré, and L. Ferroukhi. 1999. “Localpractices and the decentralization and development ofnatural resource management in French-speakingWest Africa.” Unasylva 199, no. 50.

Onyach-Olaa, Martin, and Doug Porter. 2000. “Localgovernment performance and decentral-isation inUganda: Implications for central governments anddonors.” Draft paper, mimeo.

Ostrom, Elinor. 1999. “Polycentricity, complexity, and thecommons.” The Good Society 9, no. 2: 37–41.

———. 1990. Governing the commons: The evolution ofinstitutions for collective action. Cambridge: CambridgeUniversity Press.

Ostrom, Elinor, Larry Schroeder, and Susan Wynne. 1993.Institutional incentives and sustainable development:Infrastructure policies in perspective. Boulder, Colo.:Westview.

126


Ouali, Firmin, Paul Kiemdé, and Dénise Yaméogo. 1994.“Etude de base sur l’état de la décentralisation auBurkina Faso.” Summary report for the CommissionNationale de la Décentralisation, Premier Ministère,Burkina Faso, June.

Ouédraogo, Hubert M. G. 1994. “Les coutumes relatives ala gestion des ressources naturelles au Burkina Faso.”Technical Cooperation Programme, Food and Agricul-ture Organization (FAO), FAO-LEG: TCP/BKF/2352,Rome, March.

Overdevest, Christine. 2000. “Participatory democracy,representative democracy, and the nature of concen-trated interests: A case study of public involvement ona national forest district.” Society and Natural Resources13.

Oyono, Phil René. 2004a. “Institutional deficit, representa-tion, and decentralized forest management inCameroon.” Environmental Governance in Africaworking paper, WRI, Washington, D.C. Forthcoming.

———. 2004b. “One step forward, two steps back?Paradoxes of natural resources management decen-tralization in Cameroon.” Journal of Modern AfricanStudies 42, no 1. Forthcoming.

———. 2004c. “Social and organizational roots ofecological uncertainties in Cameroon’s forest manage-ment decentralization model.” In Decentralization ofnatural resources: Experiences in Africa, Asia and LatinAmerica, edited by Jesse C. Ribot and Anne M. Larson.London: Frank Cass. Also published in the EuropeanJournal of Development Research 16, no. 1.

———. 2003. “Institutional infrastructure of local forestmanagement in Cameroon.” Draft EnvironmentalGovernance in Africa working paper, WRI, Washing-ton, D.C. Mimeo.

———. 2002. “Forest management, systemic crisis andpolicy change: Socio-organizational roots of ecologicaluncertainties in Cameroon’s decentralization model.”Paper presented at WRI’s Conference on Decentraliza-tion and the Environment, Bellagio, Italy, 18–22 Feb.[Later published as Oyono 2004c.]

Oyugi, Walter O. 2000. “Decentralization for goodgovernance and development: The unending debate.”Regional Development Dialogue 21, no. 1 (spring): 3–22.

———. 1990. “Civil bureaucracy in East Africa: A criticalanalysis of role performance since independence.” InPublic administration in world perspective, edited by O.P. Dwivedi and Keith M. Henderson. Ames, IO:Blackwell Publishing.

Pacheco, Pablo. 2004. “What lies behind decentralization?Forest, powers, and actors in lowland Bolivia.” InDecentralization of natural resources: Experiences inAfrica, Asia and Latin America, edited by Jesse C. Ribotand Anne M. Larson. London: Frank Cass. Alsopublished in the European Journal of DevelopmentResearch 16, no. 1.

———. 2002. “The implications of decentralization inforest management: Municipalities and local forestusers in lowland Bolivia.” Paper presented at WRI’sConference on Decentralization and the Environment,Bellagio, Italy, 18–22 Feb. [Later published as Pacheco2004.]

Painter, Thomas, James Sumberg, and Thomas Price.1994. “Your terroir and my ‘action space’: Implicationsof differentiation, mobility, and diversification for theapproche terroir in Sahelian West Africa.” Africa 64, no.4.

Parker, Andrew N. 1995. “Decentralization: The wayforward for rural development?” Policy researchworking paper no. 1475, Agriculture and NaturalResources Department, World Bank, Washington,D.C.

Peluso, Nancy. 2002. “Some questions about violence anddecentralization: A preliminary exploration of the caseof Indonesia.” Paper presented at WRI’s Conferenceon Decentralization and the Environment, Bellagio,Italy, 18–22 Feb.

———. 1993. “Coercing conservation: The politics of stateresource control.” Global Environmental Change 3, no.2 (June): 199–218.

———. 1992. Rich forests, poor people: Resource control andresistance in Java. Berkeley: University of CaliforniaPress.

Perham, Margery. 1960. Lugard: The years of authority,1898–1945. London: Collins Press.

Pitkin, Hannah Fenichel. 1967. The concept of representa-tion. Berkeley: University of California Press.

Plan for the Modernisation of Agriculture (PMA) GrantStudy. 2001. “Study to define modalities for theproposed PMA non-sectoral conditional grant.” Draftfinal report submitted to the PMA steering committee,February. This study was financed by the U.S. Agencyfor International Development, the Food and Agricul-ture Organization of the United Nations, Ireland, UK,Netherlands, Norway, and Denmark. The authors andconsultants include Brian van Arkadie, FredBitanihirwe, Joseph Robert Bugembe, Robert Clausen,Ssewankambo Emmanuel, Gerhard van’t Land,

127


Stephen Lister, Milton Chwa Ogeda, Ray Purcell, andNite Baza Tanzarn.

Poffenberger, Mark. 1994. “Resurgence of communityforest management in Eastern India.” In Naturalconnections: Perspectives in community-based conserva-tion, edited by David Western, R. Michael Wright, andShirley C. Strum. Washington, D.C.: Island Press.

Porter, Doug, and David Craig. 2000. “Devolving thehydra, respecting local domains: Decentralisedapproaches to indigenous governance.” Draft paper 30Apr. Not to be cited without permission.

Porter, Doug, and Martin Onyach-Olaa. 1999. “Inclusiveplanning and allocation for rural services.” Develop-ment in Practice 9, no. 1/2 (February).

Porter, Gina, and Einir Young. 1998. “Decentralizedenvironmental management and popular participationin coastal Ghana.” Journal of International Development10: 515–26.

Prud’homme, Remy. 2001. “Fiscal decentralization andintergovernmental fiscal relations.” Summary reportfor the United Nations Capital Development FundSymposium on Decentralization and Local Gover-nance in Africa, Cape Town, South Africa, 26–30 Mar.

Putnam, Robert D. 1993. Making democracy work: Civictraditions in modern Italy. Princeton: PrincetonUniversity Press.

Ranger, Terrance. 1993. “The invention of traditionrevisited: The case of colonial Africa.” In Legitimacyand the state in twentieth-century Africa: Essays inhonour of A. H. Kirk-Greene, edited by Terrance Rangerand Olufemi Vaughan. Oxford: MacMillan.

Republic of Niger. 1992. Ordonnance no. 92-037 du 21Août 1992, Projet d’ordonnance portant surl’organisation de la commercialisation et du transportde bois dans les grandes agglomérations, et la fiscalitéqui lui est applicable.

République du Mali. 1996. Loi no. 96-050, Portant sur lesprincipes de constitution et de gestion du domaine descollectivités territoriales. Adopted 16 Oct., Bamako.

———. 1995. Loi no. 95-034, Portant sur le code descollectivités territoriales en République du Mali.Adopted 27 Jan.

———. 1994. Loi no. 94-__/AN-RM, Portant surl’organisation de l’exploitation, du transport et ducommerce du bois. National Assembly, Republic ofMali.

———. 1991. Ordonnance no. 91-074/P-CTSP du 10Octobre 1991, Code électoral.

———. 1977. Ordonnance no. 77-44/CMLN du 12 Juillet1977, Portant sur le réorganisation territoriale etadministrative de la République du Mali.

République du Sénégal. 1998. “Code Forestier,” Loi no.98/03 du 8 Janvier 1998, Décret no. 98/164 du 20Février 1998. Ministere de l’Environnement et de laProtection de la Nature, Direction des Eaux, Forêts,Chasse et de la Conservation des Sols.

———. 1993. Loi no. 93-06 du 4 Février 1993, Codeforestier. Ministry of the Environment and theProtection of Nature.

———. 1972. “Decret no. 72-636 du 29 Mai 1972 relatifaux attributions des chefs de circonscriptionsadministratives et chefs de village.” Journal Officiel dela République du Sénégal, 17 June.

———. 1964. Loi no. 64-46 du 17 Juin 1964, Relative audomaine national. Journal Officiel de la République duSénégal.

Resosudarmo, Ida Aju Pradnja. 2004. “Closer to peopleand trees: Will decentralization work for the peopleand the forests of Indonesia?” In Decentralization ofnatural resources: Experiences in Africa, Asia and LatinAmerica, edited by Jesse C. Ribot and Anne M. Larson.London: Frank Cass. Also published in the EuropeanJournal of Development Research 16, no. 1.

———. 2002. “Closer to people and trees: Will decentrali-zation work for the people and the forests of Indone-sia?” Paper presented at WRI’s Conference onDecentralization and the Environment, Bellagio, Italy,18–22 Feb. [Later published as Resosudarmo 2004]

Ribot, Jesse C. 2003. “Democratic decentralization ofnatural resources: Institutional choice and discretion-ary powers in sub-Saharan Africa.” Public Administra-tion and Development 23: 53–65.

———. 2002a. “African decentralization: Local actors,powers, and accountability.” Democracy, Governance,and Human Rights working paper no. 8, UnitedNations Research Institute for Social Development andthe International Development Research Centre,Geneva.

———. 2002b. “Democratic decentralization of naturalresources: Institutionalizing popular participation.”Washington, D.C.: WRI.

———. 2001a. “Integral local development: ‘Accommodat-ing multiple interests’ through entrustment andaccountable representation.” International Journal ofAgricultural Resources, Governance, and Ecology 1, no. 3/4: 327–50.

128


———. 2001b. “Notes from the Meeting on the Re-emergence of Chiefs and Custom,” Program on Landand Agrarian Studies, University of Western Cape, May.

———. 2001c. “Science, use rights and exclusion: Ahistory of forestry in francophone West Africa.” Issuepaper no. 104, Drylands Programme, InternationalInstitute for Environment and Development. Alsopublished in French translation.

———. 2000. “Rebellion, representation, and enfranchise-ment in the forest villages of Makacoulibantang,eastern Senegal.” In People, Plants, and Justice: ThePolitics of Nature Conservation, edited by CharlesZerner. New York: Columbia University Press.

———. 1999a. “Decentralization, participation, andaccountability in Sahelian forestry: Legal instrumentsof political-administrative control.” Africa 69, no. 1.

———. 1999b. “A history of fear: Imagining deforestationin the West African dryland forests.” Global Ecologyand Biogeography, no. 8: 291–300.

———. 1999c. “Integral local development: Authority,accountability, and entrustment in natural resourcemanagement.” Working paper for the RegionalProgram for the Traditional Energy Sector in the Africatechnical group (AFTG1-Energy), World Bank, April.

———. 1998a. “A political economy of cooperative lawreform in Mali: The state is the best hen.” Report forManagement Systems International and U.S. Agencyfor International Development–Mali.

———. 1998b. “Theorizing access: Forest profits alongSenegal’s charcoal commodity chain.” Development andChange 29: 307–41.

———. 1996. “Participation without representation:Chiefs, councils, and rural representation.” CulturalSurvival Quarterly, fall. (Special issue on participation,edited by Pauline Peters.)

———. 1995a. “From exclusion to participation: TurningSenegal’s forestry policy around.” World Development23, no. 9: 1587–99.

———. 1995b. “Local forest control in Mali: An institu-tional analysis of participatory policies.” Forestry sectorpolicy report/working paper, Africa Technical Divisionof the World Bank, January.

———. 1995c. “Local forest control in Burkina Faso, Mali,Niger, Senegal, and Gambia: A review and critique ofnew participatory policies.” Africa Region, WorldBank.

———. 1995d. “Local forest access control in Senegal:Toward participatory forestry policies.” Forestry sectorpolicy report/working paper, Technical Division of theWorld Bank, January.

———. 1993. “Market-state relations and environmentalpolicy: Limits of state capacity in Senegal.” In The stateand social power in global environmental politics, editedby Ronnie D. Lipschutz and Ken Conca. New York:Columbia University Press.

Ribot, Jesse C., and Anne M. Larson, eds. 2004. Decentrali-zation of natural resources: Experiences in Africa, Asiaand Latin America. London: Frank Cass. Also pub-lished under the title Democratic decentralizationthrough a natural resource lens: Cases from Africa, Asiaand Latin America as a special issue of the EuropeanJournal of Development Research 16, no. 1.

Ribot, Jesse C., and Phil René Oyono. Forthcoming.“Resisting democratic decentralization in Africa: Somestate and elite strategies for holding onto power.”Chapter in a book on the state of affairs in Africa,edited by Ben Wisner and Camilla Toulmine. London:EARTHSCAN.

Riker, William H. 1964. Federalism: Origin, operation,significance. Boston: Little Brown.

Roberts, Richard. 1997. “Conflicts over property in theMiddle Niger Valley at the beginning of the twentiethcentury.” African Economic History 25: 79–96.

———. 1991. “The ambiguities of legal jurisdiction in earlycolonial French Sudan.” In Law in colonial Africa,edited by Kristin Mann and Richard Roberts. Ports-mouth, N.H.: Heinemann.

Rocher, F., and C. Rouillard. 1998. “Decentralization,subsidiarity, and neoliberalism in Canada: When thetree hides the forest.” Canadian Public Policy-Analysede Politiques 24, no. 2.

Roe, Emery M. 1996. Except-Africa: Remaking development,rethinking power. New Brunswick: Transaction Publish-ers.

———. 1995a. “Except-Africa: Postscript to a specialsection on development narratives.” World Develop-ment 23: 1064–9.

———. 1995b. “More than the politics of decentralization:Local government reform, district development, andpublic administration in Zimbabwe.” World Develop-ment 23, no. 5: 833–43.

Rohan, Marc. 1988. Paris ‘68: graffiti, posters, newspapers,and poems of the events of May 1968. London: ImpactBooks.

Romeo, Leonardo. 1996. “Local development funds:Promoting decentralized planning and financing ofrural development.” Policy series, United NationsCapital Development Fund.

129


Rondinelli, Dennis A., James S. McCullough, and RonaldW. Johnson. 1989. “Analysing decentralization politicsin developing countries: A political economy frame-work.” Development and Change 20: 57–87.

Rosenau, James N. 1993. “Environmental challenges in aturbulent world.” In The state and social power in globalenvironmental politics, edited by Ronnie D. Lipschutzand Ken Conca. New York: Columbia University Press.

Rothchild, Donald. 1994. “Strengthening African localinitiative: Local self-governance, decentralisation, andaccountability.” Institut für Afrika-Kunde, Hamburg.

Saito, Fumuhiko. 2000. “Decentralisation in Uganda:Challenges for the 21st century.” Paper presented atthe workshop on Uganda, Institute of CommonwealthStudies, 7 April. Paper delivered at the seminar series,Centre for Basic Research, 25 May. University ofLondon. Mimeo.

Salomao, Alda. forthcoming. “Participatory naturalresource management in Mozambique: Legal andinstitutional issues in community-based naturalresources management.” Environmental Governancein Africa working paper, WRI, Washington, D.C..Forthcoming.

Satish, Chandra, and Mark Poffenberger. 1989. “Commu-nity forestry management in West Bengal: FPC casestudies.” In Forest regeneration through communityprotection committees, edited by K. C. Malhotra andMark Poffenberger, 22–47; Appendix 1, 1–6. Calcutta,21–22 June.

Schilder, Ard, and Bram Boeve. 1996. “An institutionalapproach to local economic development: The case ofsublocational initiatives in three districts in Kenya.”Regional Development Dialogue 17, no. 2: 94–117.

Schroeder, Larry, and Matt Andrews. 2001. “Key issues insectoral decentralization.” Summary report for theUnited Nations Capital Development Fund Sympo-sium on Decentralization and Local Governance inAfrica, Cape Town, South Africa, 26–30 Mar.

Schroeder, Richard, A. 1999. “Community, forestry, andconditionality in the Gambia.” Africa 69, no. 1,January.

Schumacher, Edward J. 1975. Politics, bureaucracy, and ruraldevelopment in Senegal. Berkeley: University ofCalifornia Press.

Schumpeter, Joesph A. 1943. Capitalism, socialism, anddemocracy. Reprint, London: George Allen and UnwinLtd., 1976.

Scott, James C. 1998. Seeing like a state: How certainschemes to improve the human condition have failed. NewHaven: Yale University Press.

———. 1976. The moral economy of the peasant: Rebellionand subsistence in Southeast Asia. New Haven: YaleUniversity Press.

Selznick, Philip. 1949. TVA and the grass roots: A study ofpolitics and organization. Reprint, Berkeley: Universityof California Press, 1984.

Sen, Amartya. 1981. Poverty and famines: An essay onentitlement and deprivation. Oxford: Clarendon.

Sewell, David O. 1996. “‘The dangers of decentralization’according to Prud’homme: Some further aspects.” TheWorld Bank Research Observer 11, no. 1 (February): 143–50.

Shackelton, Sheona, and Bruce Campbell. 2001. “Devolu-tion in natural resource management: Institutionalarrangements and power shifts—A synthesis of casestudies from southern Africa.” U.S. Agency forInternational Development, World Wildlife Fund,European Union, Center for International ForestryResearch, and the Council of Scientific and IndustrialResearch. June.

Shah, Anwar. 1998. “Balance, accountability, and respon-siveness: Lessons about decentralization.” Revision ofpaper presented at the World Bank Conference onEvaluation and Development, April 1997. (Laterpublished as policy research working paper no. 2021,World Bank, Washington, D.C.)

Sharma, Keshav C. 2000. “Popular participation for goodgovernance and development at the local level: Thecase of Botswana.” Regional Development Dialogue 21,no. 1 (spring): 177–91.

Sharpe, Barrie. 1998. “‘First the forest’ conservation,‘community’ and ‘participation’ in South-westCameroon.” Africa 68, no. 1: 25–45.

Shipton, Parker. 1995. “Luo entrustment: Foreign financeand the soil of the spirits in Kenya.” Africa 65, no. 2:165–96.

Shiva, Vandana. 1989. Staying alive. Atlantic Highlands,N.J.: Zed Books.

Shivaramakrishnan, K. 2000. “Crafting the public spherein the forests of West Bengal: Democracy, develop-ment, and political action.” American Ethnologist 27,no. 2: 431–61.

Shrestha, N. Kaji. 1999. “Community forestry in Nepal indanger.” Forest, Trees, and People Newsletter, no. 38(March).

Singh, Naresh, and Vangile Titi. 1995. Empowerment:Towards sustainable development. Atlantic Highlands,N.J.: Zed Books.

130


Smith, Brian C. 1985. Decentralization: The territorialdimension of the state. London: George Allen.

Smoke, Paul. 2001. “Fiscal decentralization in developingcountries: A review of current concepts and practice.”Paper no. 2, Programme on Democracy, Governance,and Human Rights, United Nations Research Institutefor Social Development (UNRISD), Geneva. (Forth-coming in Public sector reform, governance, andinstitutional change, edited by Yusuf Bangura. Geneva:UNRISD.)

———. 1999. “Understanding decentralization in Asia:An overview of key issues and challenges.” RegionalDevelopment Dialogue 20, no. 2 (autumn): 1–17.

Smoke, Paul, and Blane Lewis. 1996. “Fiscal decentraliza-tion in Indonesia: A new approach to an old idea.”World Development 24, no. 8 (August): 1281–99.

Social Science Research Council (SSRC). 1999. Investing inreturn: Rates of return of African Ph.D.’s trained in NorthAmerica. New York: SSRC.

Spierenburg, Marja. 1995. “The role of the Mhondoro cultin the struggle for control over land in Dande (north-ern Zimbabwe): Social commentaries and theinfluence of adherents.” Center for Applied SocialSciences, University of Zimbabwe, Harare, andAmsterdam School for Social Science Research,Amsterdam. October.

Steffensen, Jesper, and Svend Trollegaard. 2000. “Fiscaldecentralisation and sub-national government financein relation to infrastructure and service provision:Synthesis report on six sub-Saharan African countrystudies.” The National Association of Local Authoritiesin Denmark, directed by the World Bank, financed bythe Danish International Development Agency andthe U.S. Agency for International Development. May.

Steins, N. A. 2001. “New directions in natural resourcemanagement: The offer of actor-network theory.”Institute of Development Studies Bulletin 32, no. 4:18–25.

Suret-Canale, Jean. 1970. “The Fouta-Djalon chieftaincy.”In West African chiefs: Their changing status undercolonial rule and independence, edited by MichaelCrowder and Obaro Ikime, 2–13. New York: AfricanaPublishing Company.

———. 1966. “La fin de la chefferie en Guinée.” Journal ofAfrican History 7, no. 3: 459–93.

Swift, Jeremy 1999. “Wildlife, pastoralists, and science inTanzania.” In The lie of the land: Challenging receivedwisdom on the African environment, edited by Leach andMearns. Oxford: James Currey Publishers Ltd. andHeinemann.

Tan-Kim-Yong, Uraivan. 2002. “Constructing politicalprocess and reform for decentralization in Thailand:Three case studies in decentralized natural resourcemanagement.” Paper presented at WRI’s Conferenceon Decentralization and the Environment, Bellagio,Italy, 18–22 Feb.

Tendler, Judith. 2000. “Why are social funds so popular?”In Local dynamics in the era of globalization, edited byShahid Yusuf, Weiping Wu, and Simon Evenett.Companion volume to the World Development Report1999/2000. Oxford: Oxford University Press.

———. 1997. Good government in the tropics. Baltimore,Md.: Johns Hopkins University Press.

———. 1980. “New light on rural electrification: Theevidence from Bolivia.” Project on Managing Decen-tralization, Institute of International Studies, Univer-sity of California at Berkeley, publication no. R80.5,September.

Therkildsen, Ole. 2001. “Efficiency, accountability, andimplementation: Public sector reform in east andsouthern Africa.” Draft of paper no. 3, Programme onDemocracy, Governance, and Human Rights, UnitedNations Research Institute for Social Development,Geneva. February. Mimeo.

———. 1993. “Legitimacy, local governments, and naturalresource management in sub-Saharan Africa.” InInstitutional issues in natural resources management:With special reference to arid and semi-arid areas inAfrica, edited by Henrik Secher Marcussen. Occasionalpaper no. 9, International Development Studies,Roskilde University, Denmark.

Thiaw, Sagane, and Jesse C. Ribot. 2003. “Insiders out:Forest access through village chiefs in Senegal.” Paperpresented at the International Conferenceon Competing Jurisdictions: Settling Land Claims inAfrica, held at the Vrije Universiteit, Amsterdam, 24–27 Sept. 2003.

Thomson, Jamie. 1995. “Community institutions and thegovernance of local woodstocks in the context of Mali’sdemocratic transition.” Paper presented at the 38thannual meeting of the African Studies Association, 3–6 Nov., Orlando, Fla.

Tiebout, Charles M. 1972. “A pure theory of local expendi-tures.” In Readings in urban economics, edited by M.Edel and J. Rothenberg. New York: MacMillan.

Tötemeyer, Gerhard K. H. 2000. “Decentralization forempowerment of local units of governance andsociety: A critical analysis of the Namibian case.”Regional Development Dialogue 21, no. 1 (spring): 95–118.

131


Tukahebwa, Geoffrey B. 1998. “The role of district councilsin decentralisation.” In Decentralisation and civil societyin Uganda: The quest for good governance, edited byApolo Nsibambi. Kampala: Fountain Publishers.

Twyman, Chasca. 1998. “Rethinking community resourcemanagement: Managing resources or managingpeople in western Botswana?” Third World Quarterly19, no. 4.

United Nations Capital Development Fund (UNCDF).2000. “Africa: Decentralisation and local governanceconference concept paper.” Draft paper for theUNCDF Symposium on Decentralization and LocalGovernance in Africa, Cape Town, South Africa, 26–30 Mar.

United Nations Conference on Environment and Develop-ment. 1992a. “Agenda 21.” Rio de Janeiro, 3–14 June.

———. 1992b. “The Rio Declaration: Report of the UnitedNations Conference on Environment and Develop-ment.” Rio de Janeiro, 3–14 June.

United Nations Development Programme (UNDP). 1999.“Taking risks: United Nations Capital DevelopmentFund.” UNDP, New York.

United Nations Environment Programme. 1992. “RioDeclaration on Environment and Development.”www.unep.org/unep/rio.htm.

Uphoff, Norman T., and Milton J. Esman. 1974. “Localorganization for rural development: Analysis of Asianexperience.” RLG no. 19, special series on rural localgovernment, Rural Development Committee, CornellUniversity, Ithaca.

Utting, Peter, ed. 1999. “Forest policy and politics in thePhilippines: The dynamics of participatory conserva-tion.” Manuscript. Mimeo. (Later published by theUnited Nations Research Institute for Social Develop-ment and Ateneo de Manila University Press. QuezonCity, Philippines. 2000.)

Utting, Peter, and Ronald Jaubert, eds. 1998. Discours etréalités des politiques participatives de gestion del’environnement: Le cas du Sénégal. Geneva: UnitedNations Research Institute for Social Development.

van Rouveroy van Nieuwaal, Emile. 1987. “Chiefs andAfrican states: Some introductory notes and anextensive bibliography on African chieftaincy.” Journalof Legal Pluralism, no. 25/26.

van Rouveroy van Nieuwaal, E. Adriaan B., and Rijk vanDijk, eds. 1999. African chieftaincy in a new socio-political landscape. Hamburg: Lit Verlag.

Veit, Peter, G. Banda, A. Brownell, M. Dhliwayo, P. Galega,G. Kanja, R. Nshala, B. Ochieng, A. Salomao, and G.Tumushabe. 2004. “Serving constituents, servicingnature: Legislative representation and natural resourcepolitics in Africa.” Unpublished manuscript, 7 Jul.Mimeo. WRI, Washington, D.C. Forthcoming.

Veit, Peter G., and George Faraday. 2001. “Beyond thepresidents’ men.” Draft paper for WRI. Mimeo.

Vijayalakshmi, V., and B. K. Chandrashekar. 2002.“Authority, powerlessness, and dependence: Womenand political participation.” Working paper no. 106,Institute for Social and Economic Change, Bangalore.

von Benda-Beckmann, Keebet. 1981. “Forum shopping andshopping forums: Dispute processing in aMinangkabau village in West Sumatra.” Journal ofLegal Pluralism 19: 117–59.

Walker, Peter A. 2000. “Democracy and environment:Congruence or contradiction?” Commons SouthernAfrica 2, no. 1: 4–6.

Wassenich, Paul and Katherine Whiteside. 2003. “CDDImpact Assessment Study: Optimizing EvaluationDesign Under Constraints,” Final draft report, 19December 2003, Community Driven Developmentand Social Capital Anchor and the Social Funds Teamin the Social Protection Anchor of the HumanDevelopment Network of the World Bank. See http:/lnweb18.worldbank.org/ESSD/sdvext.nsf/09ByDocName/Evaluations.

Watts, Michael J. 1993. “Idioms of land and labor: Produc-ing politics and rice in Senegambia.” In Land inAfrican agrarian systems, edited by Thomas J. Bassettand Donald E. Crummey. Madison: University ofWisconsin Press.

———. 1987. “State, oil, and agriculture in Nigeria.”Institute of International Studies, Berkeley.

Webster, Neil. 1992. “Panchayati Raj in West Bengal:Popular participation for the people or the party?”Development and Change 23, no. 4 (October).

Weimer, Bernhard. 1996. “Challenges for democratizationand regional development in southern Africa: Focuson Mozambique.” Regional Development Dialogue 17,no. 2: 49–50.

Weinstein, Brian. 1972. Éboué. New York: Oxford Univer-sity Press.

Western, David, and R. Michael Wright. 1994. “Back-ground to community-based conservation.” In Naturalconnections: Perspectives in community-based conserva-tion, edited by David Western, R. Michael Wright, andShirley C. Strum. Washington, D.C.: Island Press.

132


Western, David, R. Michael Wright, and Shirley C. Strum,eds. 1994. Natural connections: Perspectives in commu-nity-based conservation. Washington, D.C.: IslandPress.

Wolfire, Deanna Madvin, Jake Brunner, and Nigel Sizer.1998. “Forests and the Democratic Republic of Congo:Opportunity in a time of crisis.” Forest FrontiersInitiative, WRI.

Wollenberg, Eva, David Edmunds, and Jon Anderson.2001. “Pluralism and the less powerful: Accommodat-ing multiple interests in local forest management.”International Journal of Agricultural Resources, Gover-nance, and Ecology 1, no. 3/4.

Woodhouse, Philip. 1997. “Governance and local environ-mental management in Africa.” Review of AfricanPolitical Economy, no. 74: 537–47.

World Bank. 2003. “Nongovernmental Organizations andCivil Society/Overview,” located on World Bankwebsite (http://wbln0018.worldbank.org/essd/essd.nsf/NGOs/home). Accessed 8 June 2001.

World Bank. 2000. World development report 1999/2000:Entering the 21st century: The changing developmentlandscape. Oxford: Oxford University Press.

———. 1996. The World Bank participation sourcebook.Washington, D.C.: World Bank.

Wunsch, James S., and Dele Olowu. 1995. The failure of thecentralized state: Institutions and self-governance inAfrica. San Francisco: Institute for ContemporaryStudies Press.

Xu Jianchu. 2002. “Decentralization and accountability inforest management: A trend and its policies on‘natural forest protection’ and ‘upland conversion.’”Paper presented at WRI’s Conference on Decentraliza-tion and the Environment, Bellagio, Italy, 18–22 Feb.[Later published as Xu and Ribot 2004.]

Xu Jianchu, and Jesse C. Ribot. 2004. “Decentralizationand accountability in forest management: Case fromYunnan, southwest China.” In Decentralization ofnatural resources: Experiences in Africa, Asia and LatinAmerica, edited by Jesse C. Ribot and Anne M. Larson.London: Frank Cass. Also published in the EuropeanJournal of Development Research 16, no. 1.

Yates, Douglas A. 1996. The rentier state in Africa: Oil rentdependency and neocolonialism in the Republic of Gabon.Trenton: Africa World Press.

Young, Iris Marion. 2000. Inclusion and democracy. Oxford:Oxford University Press.

133


1. It is important to note that a plurality of institutionsand views is a fact and can be a very positive thing.However, this plurality needs mechanisms of media-tion if it is not to become a “marketplace” of institu-tions and ideas in which the strongest actors domi-nate. Adams et al. (2003, 1916) and others have calledfor mediation among this plurality of “stakeholders,”institutions, and views as a way to achieve better policyoutcomes. Representation can be such a mediatingmechanism because it privileges the broadest cross-section of society by recognizing citizens as theprimary stakeholders.

2. Globally, all but twelve of the seventy-five developingand transitional countries with populations over fivemillion claim to be transferring political powers tolocal units of government (Dillinger 1994, 8, in Crookand Manor 1998; World Bank 2000). Today, Ethiopia,Ghana, Mali, Namibia, Nigeria, Senegal, South Africa,and Uganda have constitutions that are explicitly pro-decentralization and formally recognize the existenceof local government (Therkildsen 1993, 83; Tötemeyer2000, 95; United Nations Capital Development Fund2000; Bazaara in Latif 2002, 14). “There is not asingle country in Africa in which some form of localgovernment is not in operation,” and the statedobjective of virtually all of these reforms is tostrengthen democratic governance and serviceprovision (Oyugi 2000, 16). In a World Bank survey ofthirty African countries, all claimed to be decentraliz-ing (Ndegwa 2002). Also see Shackelton andCampbell (2001) for cases of decentralizations insouthern Africa.

3. The current global pursuit of decentralization is notthe first time governments have attempted to transferfiscal responsibilities or institutional powers to lower-level institutions. In francophone West Africa, forexample, beginning in 1917 at least four distinct waves

can be discerned. In each case, decentralizationaccompanied and perhaps was sparked by fiscal crisesof the state (Ribot 1999a). (For some history ofdecentralization in Africa, see Buell 1928; Cowan1958, 60; République du Sénégal 1972; Weinstein1972, 263–6; Schumacher 1975, 89–90; Républiquedu Mali 1977; Mawhood 1983; Conyers 1983;Rondinelli et al. 1989; de Valk and Wekwete 1991;Therkildsen 1993; Crook and Manor 1994; Diallo1994; Ouali et al. 1994, 7; Rothchild 1994; Gellar1995, 48; Mbassi 1995, 23; Mamdani 1996a; Ahwoi2000, 2; United Nations Capital Development Fund2000; Hesseling n.d., 15.) In southern Asia, at leastthree waves of decentralization are visible since themid-1800s. The first of these occurred as early as the1860s according to Fisher (1991). The second wave ofdecentralization took place after independencethrough efforts at community development (see Ribot1999a; Shackelton and Campbell 2001). A third waveis currently underway. In Indonesia, decentralizationswere taking place in the 1950s and ’60s (Resosudarmo2002, 2).

4. This description is suspiciously similar to descriptionsof modern-day participatory approaches: “The broadaim of participatory development is to increase theinvolvement of socially and economically marginalizedpeoples in decision-making over their own lives”(Guijt 1998, 1, in Cooke and Kothari 2001, 5).

5. Anderson, Clément, and Crowder 1998; Utting 1999;Alden Wily 2000a; Meinzen-Dick and Pradhan 2000;Wollenberg, Edmunds, and Anderson 2001.

6. Uphoff and Esman 1974, xx; Agrawal and Ribot 1999;Smoke 1999, 10; Crook and Sverrisson 2001, 5.

7. There is plenty of evidence that local people canmanage their own resources sustainably (see Onibon,Dabié, and Ferroukhi 1999; Adams, Brockington,

ENDNOTES

134


Dyson, and Vira 2003, 1915). Community-basednatural resource management experiments fromaround the world also provide indications that localpopulations can successfully and equitably managenatural resources. First, they demonstrate that localcommunities have the capacity to protect forests andfoster regeneration (Dembélé and Dembélé 2001).Second, they demonstrate that local democraticprocesses can lead to the protection and managementof resources for local community use (Conyers 2002).Lastly, they demonstrate that local management cangenerate revenues for local public works (BigombeLogo 2001 and 2003; Larson 2002; Pacheco 2002).But it is important to keep in mind that these project-based approaches occur under close outside supervi-sion and with intensive assistance and, therefore, donot reflect what would occur under a more generalizeddecentralization scenario. It is difficult to determinefrom project-based experiences the degree to whichsuccess is engineered from above or derives from localconditions and initiative (see Schroeder 1999;Baviskar 2002).

8. For notable exceptions, see Ribot 1995a, 1995c, 1999a,1999c; Tendler 1997; Crook and Manor 1998; Saito2000; Crook and Sverrisson 2001; Therkildsen 2001;Anderson 2002. The Plan for the Modernisation ofAgriculture Grant Study (2001, 29) points out thatwhile some analysts argue that decentralization is agood in itself because it fosters participation, thesearguments are still subject to verification: “in practicelittle research has been done.” On the lack of substan-tive research on the effects of The World Bank’scommunity-driven development programs, also seeMansuri and Rao (2003) and Wassenich andWhiteside (2003).

9. Note that the formal transfer of power to lower levelsof government may sometimes be a centralizing act ifthe powers being devolved were earlier exercisedinformally by non-state actors.

10. On participation and democracy as motives, see Bishand Ostrom 1973; Conyers 1990, 16; de Valk 1991, 11;Engberg-Pedersen 1995, 1; Weimer 1996, 49–50;Meinzen-Dick and Knox 1999, 5; Balogun 2000;Oyugi 2000, 4; Sharma 2000; Cooke and Kothari2001.

11. I prefer to avoid the term “devolution” because it isused in so many different ways.

12. Uphoff and Esman (1974, xx) imply that accountablerepresentation enhances development performance.Crook and Sverrisson (2001, 5) point out that there aretwo steps in linking local participation or democracy to

outcomes. The first step is to measure the “responsive-ness” of local decision-making processes or authoritiesagainst local needs; the second step is to measure thedegree of responsiveness against the effectiveness ofoutcomes. Crook and Sverrisson define institutionalresponsiveness as “the achievement of ‘congruencebetween community preferences and public policies’such that the activities of the institution are valued bythe public. . . . Responsiveness of policies is a matterof the process through which they are defined, thedegree of empowerment and ‘ownership’ which is feltby those affected by them. . . and, therefore, thegeneral legitimacy of the institution and the proce-dures by which it allocates resources” (2001, 5). This isanalogous to what I have called accountability of thelocal authorities.

13. The term “developmentalist” refers to decentraliza-tions that take place for the purposes of local develop-ment. This includes decentralizations whose objec-tives are better service provision, better local manage-ment, and more-democratic local processes.“Developmentalist” does not include those decentrali-zations that are a byproduct of downsizing centralgovernment, reducing central fiscal burdens, ornational unification through the integration of splintergroups or break-away regions.

14. They have argued that greater local knowledge isneeded in natural resource management to under-stand both the complex physical characteristics ofresources and the culturally specific patterns ofresource management and use. Because naturalresources sustain local livelihoods and are also animportant source of national wealth, they are the locusof potential conflict and tension. Extra sensitivity tolocal livelihood strategies and political dynamics isrequired in the making and implementation of anyenvironmental interventions. Further, given the needfor local labor and knowledge in implementation,many authors have argued that local people must beincluded at every stage in order to build their sense ofownership of reforms. This will help garner localsupport and prevent sabotage and disaffection. Also,because natural resources are a source of wealth,many authors have argued for participation as a way ofincreasing local incomes (See Ostrom 1990;Poffenberger 1994; Western, Wright, and Strum 1994;Baland and Platteau 1996; Agrawal and Ribot 1999;Agrawal and Gibson 2001; Plan for the Modernisationof Agriculture Grant Study 2001, 28).

135


15. Ribot 1996; Crook and Manor 1998; Shah 1998;Agrawal and Ribot 1999; Manin, Przeworski, andStokes 1999, 2; Blair 2000; Brinkerhoff 2001; Olowu2001; Prud’homme 2001; Therkildsen 2001.

16. The fine adaptation of service provision to localconditions and in response to local needs andpriorities is one of the reasons behind decentraliza-tion. Nevertheless, to a certain extent, a school, aclinic, or a bridge in one location is not very differentfrom a similar item in another location. Thus, thesocial service ministries and social investment fundsachieve substantial economies of scale by designingnational curricula, health programs, and constructionprototypes for the entire country. In contrast, naturalresources are inherently much more location specific.

17. The democratic impulse behind the current globalwave of decentralizations was already present in theparticipatory development movement of the 1970s,’80s, and ’90s, when development theorists andpractitioners used a variety of participatory methods toincrease the equity and the effectiveness of develop-ment. By 1978, the International Development and FoodAssistance Act required all U.S. agencies to conductdevelopment for the poor “through institutions at thelocal level, increasing [poor people’s] participation inthe making of decisions that affect their lives.” (Theact contains definite Schumaker-type language onfavoring labor-intensive investments. I believe that theimpulse in the U.S. toward “participatory” approachesgrew out of the post-inner-city riots of the 1960s’“advocacy planning” movement.) Popular participationwas also promoted by the United Nations Conferenceon Environment and Development in its Agenda 21and the Rio Declaration (1992a and 1992b). (Agenda21 (chapters 23.1–3, and 27.1–4) calls for “broad publicparticipation in decision-making” but, in contradictionto this statement, goes on to speak almost exclusivelyof NGO participation. The Rio Declaration calls for the“participation of all concerned citizens.”) Over the pastthree decades, the developing world has been inun-dated with participatory projects, programs, and laws(Cohen and Uphoff 1977; Cernea 1985; Satish andPoffenberger 1989; Peluso 1992; Banerjee et al. 1994).Such participation is argued to result in increasedequity and efficiency (Cohen and Uphoff 1977; Cernea1985; National Research Council 1992, 35; Baland andPlatteau 1996; World Bank 1996). This is a question-able assumption. Participatory approaches are oftenunjust (Cooke and Kothari 2001).

18. An important distinction between pluralist and mostparticipatory approaches is that pluralist approaches donot seek consensus (Jon Anderson, personal commu-nication, January 2004).

19. Chambers 1983; Fisher 1997; Steins 2001; Cooke andKothari 2001. These approaches can be divided intotwo major camps: popular and stakeholder ap-proaches. The participatory movement originated withthe desire to achieve popular inclusion—along thelines of popular democracy. The movement laterspawned “stakeholder” and “pluralist” approaches,which created fora for interested (and sufficientlyorganized and funded) individuals and civil societygroups to negotiate and inform decisions. This reportis concerned with popular forms of inclusion sincethese are the forms that decentralization theoryexpects to represent the broadest cross-section ofsociety. Stakeholder approaches, which favor thosewho are organized or interested, rather than all thecitizens of a given jurisdiction, are not consideredrepresentative or popular forms of inclusion.

Even the popular approaches spawned by theparticipatory movement have drawbacks. In the publicfora of direct democracy and participatory approaches,many of the most marginalized people are notcomfortable or safe expressing their opinions (Cooke2001). Villages and settlements and other smallcommunities are usually highly stratified along age,gender, class, caste, ethnic, livelihood, race, religious,and political lines, as well as along lines of indigenousversus migrant or foreign origins (Ribot 1995a; Cookeand Kothari 2001, 5). Local elite often dominate socialand political discourse, making it difficult for marginalpopulations to participate. Representative forms maybe better than direct forms of democracy, but they toounder-serve marginal groups (see Crook andSverrisson 2001).

20. Cohen and Uphoff 1977; Chambers 1983; Cernea1985; Satish and Poffenberger 1989; World Bank1996.

21. Van Rouveroy van Nieuwaal and van Dijk 1999;Kassibo 2002; Ntsebeza 2003.

22. Ribot 1995a, 1999; Benjaminsen 1997, 2000; Agrawaland Ribot 1999; Oyugi 2000, 10; Crook andSverrisson 2001. The World Bank (2000, 107)understates the problem, observing that “decentraliza-tion is often implemented haphazardly.” Equallydamaging, the term “decentralization” is also appliedto programs and reforms that ultimately are designedto retain central control (Mawhood 1983; Alcorn 1999,44; Ribot 1999a; Conyers 2000a). Oyugi (2000, 10)

136


goes so far as to say that “the legal-political design oflocal government in Africa tends to weaken thecultivation of a democratic culture at the local level aswell as weaken the ability of local authorities to takeinitiative in the field of service provision.”

23. See Mansuri and Rao 2003; Ribot and Oyono forth-coming.

24. For example, decentralization has left Senegal’s forestservice face to face with local people who want to usetheir new forest-management rights to stop commer-cial cutting. The forest service, however, wants them tomanage the forests to provide the capital withwoodfuel.

25. Forestry was chosen as a focus because it is a classiccollective resource for which monitoring and exclusionare difficult, and on which rural livelihoods around theworld depend. As a diffuse resource, forests are a goodcandidate for the kind of diffuse control and manage-ment that decentralization offers.

26. Crook and Sverrisson (2001, iii), however, “concludedthat responsiveness to the poor is quite a rare out-come.”

27. Design and implementation of decentralizationsfollows from the motives for decentralizing. Why theytake place helps explain why they take the particularforms they do and the kinds of outcomes theyproduce. The theoretical efficiency or equity benefitsare rarely the only reasons for a central government togive up central powers. Decentralizations haveresulted from

● pressure from economic crises (Olowu 2001, 53;Therkildsen 2001, 1; Muhereza 2003; Bazaaraforthcoming);

● the need for central governments to shed fiscaland administrative burdens (Nsibambi 1998, 2;Bazaara forthcoming);

● management failures of central governments(Wunsch and Olowu 1995);

● emulation of reforms in other developingcountries (Therkildsen 2001, 1);

● populist political success (Heller 1996; Olowu2001, 53);

● donor pressures and conditions as part ofstructural adjustment and other programs imposedfrom the outside (Mutizwa-Mangiza 2000, 24;World Bank 2000; Therkildsen 2001, 1);

● pressure from subnational splinter groups andlocal elite (Brock and Coulibaly 1999, 30; UnitedNations Capital Development Fund 2000, 3;World Bank 2000, 108–9; Olowu 2001, 53); and

● particular configurations of relations betweencentral and local authorities (Crook andSverrisson 2001, 2).

Kaimowitz et al. (2000, 8) show that newdecentralization laws in Bolivia were inspired by (1)growing pressure from regional groups, (2) decentrali-zations occurring in neighboring countries, and (3)support from international development agencies. Ofcourse, they are most likely a conjunctural result ofthese and other global, national, and local forces.....

As Conyers (2000a, 9) has so eloquently pointedout: “Most decentralisation efforts have both explicitand implicit objectives. Those objectives likely toappeal to the general public, such as local empower-ment and administrative efficiency, are generallyexplicitly stated, while less popular ones, such asincreasing central control and ‘passing the buck’, areunlikely to be voiced.”

28. I use the term “popular” to modify participation toindicate that I am not talking about stakeholdermethods and to ensure that I am not misunderstoodas advocating participatory techniques, which havebeen duly critiqued (e.g., Cooke and Kothari 2001).

29. Although this is often not the case, participatory jointforest management has been observed to creategreater marginalization and loss of livelihoods fordisadvantaged groups (Hildyard, Hegde, Wolvekamp,and Reddy 2001). The basic assumption is one ofequality in the process. If this does not hold, thetheories do not stand up.

30. Arguments defending decentralization on the basisthat it encourages greater participation of citizens indemocratic governance are given by Alexis deTocqueville (1945), John Stuart Mill (in Green 1993),and, more recently, by Robert Dahl (1981, 47–9). NeilWebster (1992, 129) is only one of the later figures toargue that decentralization is “seen as a means bywhich the state can be more responsive, moreadaptable, to regional and local needs than is the casewith a concentration of administrative powers” (seealso Bish and Ostrom 1973 and Weimer 1996, 49–50).For a counterview that does not see in decentralizationany necessary benefits of responsiveness or freedom,see Riker 1964, 142; he writes, “To one who believes inthe majoritarian notion of freedom, it is impossible tointerpret federalism as other than a device of minoritytyranny.” Schilder and Boeve (1996, 94–117) arguethat decentralization deserves support for the effi-ciency that it provides. For efficiency arguments in thecontext of public choice, see Tiebout 1972 and Oates1972, 11–2.

137


31. Also see discussion of public environmental hearingsin Crook and Sverrisson 2001 and Overdevest 2000.

32. Accountability needs to be separated from responsive-ness. Manor (in Latif 2002, 52) makes the observationthat responsiveness can be measured by surveys.Responsiveness reflects the degree to which govern-ment actions conform to popular preferences. (Alsosee discussion of public environmental hearings inCrook and Sverrisson 2001 and Overdevest 2000.)

33. Powers in this framework are the classic executive,legislative, and judicial arenas of action—and theresources needed to execute actions. Ability is anaspect of power composed of internal and externalfactors. Internal factors include the skills of the peopleand institutions that exercise powers. External factorsinclude the physical, legal, and social environment inwhich the powers are to be exercised, the accountabil-ity of the central state to local government, and theresponsiveness of both central and local governmentin providing assistance to local people. In this report,the internal and external aspects of ability are foldedinto discussions of power and are discussed wherethey emerge throughout the report. For simplicity, theyare not treated as a separate category. The internalaspect is dealt with, at least in part, in discussions of“capacity.” Capacity merits additional discussionbecause it is at once a necessary aspect of responsive-ness and it is used by policy makers and practitionersas an excuse not to empower.

34. Mansuri and Rao (2003, 16–7) present evidence thateducation projects where participants are given greaterdiscretion also produce more effective outcomes. Forexample, school management in El Salvador, Nicara-gua, and the Philippines was found to be moreeffective when greater autonomy was given to schoolsand to parent teacher associations.

35. Anu Joshi (personal communication, 1999) hasargued that civil society organizes and crystallizesaround empowered representative authorities. InMongolia, Mearns (2002, 4) reports a pattern in whichthere is substantial political decentralization with littleadministrative and fiscal transfer. Under thesecircumstances, he observes, “there is little opportunityfor elected assemblies to exert any influence overtechnical department staff and quality of servicedelivery at their respective level of government, andtherefore little incentive for local populations toexpress their views to elected representatives.”

36. Of course, what can be decided locally without socialor ecological harm depends also upon the constraintsof nationally set social and environmental standards. Ifstandards are set for the type, extent, and techniquesof timber harvest, then harvesting decisions can bedevolved to local authorities contingent on respect forthose standards.

37. Italics in original.

38. Oyono (2003) also argues that in Cameroon, conserva-tion is not ensured without accompanying measures.In many of Cameroon’s forests, local communitiesprefer to exploit surrounding forests as commercialconcessions. One villager explained that people want“to eat and drink now with the money from forests.”Based on statements of this kind, Cameroonianresearchers feel that decentralization without protec-tive measures may well be “ecologically counter-productive.”

39. Power shapes how non-government actors makedemands and act as counterpowers to local govern-ment—the strongest and best organized are likely tohave more influence. Incentive structures (theeconomic needs of local government, the value of theresource economically and for local uses, and environ-mental and social ideologies, such as environmental orfrontier ethics) can drive use patterns. Local govern-ment capacity to effect change also matters. (Larson2003, 20–3)

40. Indeed, “participatory approaches have provedcompatible with top-down planning systems, and havenot necessarily heralded changes in prevailinginstitutional practices of development” (Mosse 2001,17). Some authors have also argued that participatoryapproaches and the rhetoric of participation are usedto mask centralization in the name of decentralization(Cooke and Kothari 2001, 7).

41. Despite the importance of local elections, “comparedto national elections, local elections have not beenseriously studied” (Olowu 2001, 57).

42. In Uganda’s local council system, “universal adultsuffrage has increased the people’s sense of involve-ment in policy-making process” (Saito 2000, 4).

43. In Kenya, people line up behind candidates’ pictures;however, this does not constitute true voting (Manor inLatif 2002, 44). In Nicaragua, where elections are alsoby party slate, some mayors are responsive to constitu-ents, others just to the nation’s president (Latif 2002,56).

138


44. Upward accountability can be crafted into representa-tive systems in different ways. In Mongolia, thepowerful governors, appointed by the higher-levelassemblies, are primarily upwardly accountable andare the key decision makers in the management oflocal space. The elected assemblies at each level, whoare more accountable to their constituencies, hold verylimited powers of oversight and play a primarilyadvisory, rather than decision-making, role. (Mearns2002, 3–4)

45. In Indonesia, powers have been devolved, but theaccountability of local authorities is questionable(Resosudarmo 2002).

46. In Indonesia, elections are not effective accountabilitymechanisms because of the party-list system. Peopledo not know who the candidates are for a given party,and the candidates are inaccessible. These local electedauthorities in Indonesia are reportedly over-exploitingthe forests (Resosudarmo 2002; Latif 2002, 47).

47. Examples can be found in Burkina Faso, Chad, Côted’Ivoire, Ghana, Mozambique, Niger, Nigeria, Senegal,Sierra Leone, South Africa, Tanzania, Zambia, andZimbabwe (Ribot 2002a, 2002b). In South Africa,traditional chiefs are fighting for powers over landtenure arrangements. These individuals are notrepresentative, and their empowerment is inconsistentwith the democratic principles of elected representa-tion enshrined in the constitution. (See Ntsebeza1999, 1004.)

Land is being transferred from the state toindividuals and others. The implementation has beenweak and incomplete. There are many ambiguitiesaround the appropriate unit for land ownership—private individuals, customary authorities, etc. It is notin the interests of the state to resolve these ambigu-ities because people within the central governmentwant the votes that chiefs can garner. (Latif 2002, 55)

It would be worth exploring whether suchtraditional authorities have been reasserting theirauthority in Asia and Latin America.

48. This observation is based on fieldwork that the authorconducted in Senegal in 2003 and 2004.

49. Ntsebeza (2004) points out that many people considerchiefs to be “legitimate” because people still go tothem for various kinds of decisions—such as landallocation. But he also points out that this logic isflawed since these people have no alternative. They arenot choosing to go to the chiefs, they are forced to.Hence, their actions cannot be construed as indicatingthe legitimacy of chiefs.

50. Engberg-Pedersen (1995) points out that in BurkinaFaso “the two most important institutions for decision-making and implementation at the village level—thechieftainship and the working groups [village commit-tees registered as ‘groupement villageois’]—are notdemocratic. There are no traditions for open discus-sion of village affairs, for criticism of decisions madeby leaders, for public elections of leaders, etc. Not onlydo leaders try to monopolize the right to makedecisions, but others do not consider themselves in aposition to make proposals regarding village affairs.They do not even seem to have opinions on thesematters, because as some argued in [the village of ]Nahirindon, it would be offensive towards villageleaders if they had.”

Peluso (2002, 9) reports that in Indonesia villageleaders are bribed by illegal timber extractors, causingsplits within villages.

51. Lissu (2000) also defines them as groups acting in the“public interest or interest of a group of individuals ororganizations.” Hence, they are private in nature butcan act on behalf of public or private collective interest.

52. These motives, Conyers (1990, 25) points out, are“closely associated with the ‘Reaganite’ and‘Thatcherite’ policies which have permeated bothdeveloped and less developed countries in the last fewyears. . . .”

53. Mearns (2002, 11–2) reports that in Mongolia’spastoral areas, “a ‘tragedy of open access’ has emergedin the institutional vacuum created by incompletedecentralization. . . .” According to Mearns, socialvulnerability in the face of natural hazards hasincreased at least partly due to the devolution ofmanagement responsibilities without financialresources to carry them out and without accountabilityreforms that would subject local government to publicscrutiny.

54. But even this form of inclusion, while it may increasewomen’s presence, does not necessarily translate intopower due to broader issues that hinder women’sparticipation (Vijayalakshmi and Chandrashekar 2002;Bazaara 2003).

55. Baviskar (Forthcoming, 21) describes a case in India inwhich the committee’s procedures were dictated by theproject, and the committee and other participatoryprocesses were created as a charade to create a “recordof ‘participation,’” rather than to induce greaterinclusion or representation in decisions.

139


56. Hildyard, Hedge, Wolvekamp, and Reddy (2001) haveargued that “unless participatory processes take intoaccount the relative bargaining power of so-calledstakeholders, they are in danger of merely providingopportunities for the more powerful.”

57. See Ribot 1999a; Baviskar 2002; Bazaara 2002;Brannstrom 2002; Kassibo 2002a; Larson 2002;Mapedza and Mandondo 2002; Melo Farrera 2002;Oyono 2002; Pacheco 2002; Resosudarmo 2002;Muhereza 2003.

58. “An additional danger [in Cameroon] is that once acommunity has invested in the inventory, and timberresources of high value [have been] confirmed,unscrupulous government officials might be temptedto find evidence of an infraction (breach of the rules)on the part of the community in question, allowing alogging company to step in and exploit the area as aconventional vente de coup [small-scale commerciallicense], benefiting from the inventory work alreadydone at no extra cost to itself” (Brown 1999, 47).

59. See Mbassi 1995, 24; Evans 1997; Tendler 1997; Crookand Manor 1998; Conyers 2000b, 22; Mutizwa-Mangiza 2000, 23; Peluso 2002, 7; Xu 2002, 14.

As Conyers (2000b, 22) points out, “Ironically,decentralisation policies are most likely to be imple-mented effectively in situations where the governmentis politically secure and power is concentrated in thehands of a relatively small group of people. A securegovernment can afford to decentralise a substantialamount of power without threatening its ownexistence, while the centralisation of power enablesthe key leaders to make and implement policydecisions without undue interference from otherinterest groups within government.”

60. Decentralizations change the distribution of powers incomplex ways, creating winners and losers. Violenceamong resource users has been reported during theimplementation of decentralization reforms inCameroon, Indonesia, and Mali (Kassibo 2002a; Latif2002; Peluso 2002; Resosudarmo 2002). In Mali,conflict among farmers and pastoralists has reportedlyincreased, but it is not clear whether this increase isdue to decentralization or to other changes in landoccupation and herd management (CheibaneCoulibaly, personal communication, Bamako, Mali,2001; Bréhima Kassibo, personal communication,Bamako, Mali, 2002).

61. I owe great thanks to Crescencia Maurer of WRI forsuggesting the development of these questions.

62. Increased local revenues have also been generatedthrough environmental decentralizations in Bolivia,Cameroon, India, Indonesia, Nicaragua, and Zimba-bwe (Agrawal 2001; Conyers 2002; Larson 2002;Mapedza and Mandondo 2002; Resosudarmo 2002;Bigombe Logo 2003). In Cameroon, for example, newforestry laws earmark timber stumpage fees for electedlocal councils.

63. For more in-depth discussion on this topic, seeMusgrave 1965; Huther and Shah 1998; Shah 1998;Smoke 1999, 2000; Conyers 2000a; Oyugi 2000;Steffensen and Trollegaard 2000; and Prud’homme2001. Also see Ribot 2002a for a summary discussion.

64. This observation is based on fieldwork that the authorconducted in Senegal in 2003 and 2004.

65. Finsterbusch and van Wincklin (1989 in Mansuri andRao 2003, 16) also found that participatory projects inmore-developed countries were more effective than inless-developed areas.

66. Based on work in India, Agrawal (2001) points outthat decentralization success can be fettered by thesize of jurisdictions. Jurisdictions that are too large oreven too small can hamper decentralization success.

67. This observation was inspired by Joel Barkin’scomments on a decentralization panel at the Novem-ber 1996 annual meeting of the African StudiesAssociation, which was held in San Francisco, Calif.

68. Oyono (2002, 14) points out that in many places thereis a local majority that want to consume forests. Hecites Fomété (2001) who argues that “local communi-ties largely express the wish to see forests surroundingthem transformed into concessions and exploited atonce.” Based on this observation, Oyono argues thatdecentralization may well be “for the momentecologically counterproductive.”

69. Some municipalities in Nicaragua have negotiated theprivilege to reject commercial interests; however, thisprivilege is not protected by law (Anne Larson,personal communication, January 2004).

70. In Uganda, the central government decentralizedpowers over forests in 1993, but recentralized them afew years later because they claimed that the localpeople were not “psychologically, technically andfinancially prepared to manage them on a sustainablebasis” (Bazaara 2002, 4; 2002a, 57).

140


71. Bertin, concerned about disciplining laborers, goes onto ask, “How, in effect, can one pursue a native who,most of the time has nothing to confiscate, andescapes in this manner all manner of civil constraint?”The answer Bertin provides is that “one must treat theun-educated natives like children, which they are inreality.” It was the very lack of markets in land—created by the systems of Association and Indirect Ruleand their need to collectivize land to give their localcadres a basis for power—that also drove the coloniz-ers to use coercive means to obtain and disciplinelaborers.

72. The question of “whose mission?” is very similar tothe question of “whose resource?” Both questionssuggest a more basic question: who should makedecisions on behalf of local people with respect toresources and other projects that affect them?

73. In Europe and the U.S., there are many “customary”institutions and authorities that have not beenreplaced by democratic institutions. When a crime iscommitted among siblings, the siblings may go first totheir parents for help. If this fails, they may go to theirsynagogue or mosque or even to binding arbitrationwell before they even consider going to the courts.

74. Of course, provisions also have to be made for migrantand nomadic communities (Ngaido 1996; Bazaara inLatif 2002, 15).

75. World Bank Community Driven Development teams,for example, have lowered copayments on naturalresource management investments and requirednatural resource management measures as part ofother investments. These incentives, however, reflectoutside priorities that may best be inscribed in publiclydebated standards and regulations, rather thanimposed in an ad hoc manner.

76. The independent exercise of public powers by aplurality of local institutions does not guaranteepopular inclusion. While pluralism is a good thing,pluralism unmediated by accountable representationallows the strongest institutions, not necessarily themost efficient or just, to rule. The articulation betweenstate and non-state, formal and informal, institutionsmust be carefully worked out. Democracy is notintended to replace tradition. It is meant to representall people and their traditions.

77. See, for example, Adams et al. 2003, 1915.

78. This annex was composed by Catherine Benson.

79. “The Zimbabwe ombudsman handles on average 100complaints a month, while the Tanzanian handled200 cases per month during its first 20 years ofexistence” (Therkildsen 2001, 29).

80. According to Therkildsen (2001, 27), “transparency,translated into clear and explicit managerial targetscombined with increased managerial autonomy andincentives to perform, makes it easier to establish thebasis for managerial accountability and to achieveoutputs (without which the notion of accountabilitybecomes irrelevant . . . ). This, in turn, increasespolitical accountability in two ways. By making targetsexplicit, it is easier for managers—in dialogue withpoliticians—to match them with political priorities.And by monitoring the extent to which targets are met,politicians can, in turn, hold managers accountable fortheir performance or lack of it.”

81. This observation may challenge Moore’s (1997)claims, discussed below.

141


ABOUT THE AUTHOR

Jesse C. Ribot is a Senior Associate at the World ResourcesInstitute where he currently directs a policy research programon decentralization and the environment. Before coming toWRI, Dr. Ribot was a research associate at the Harvard Centerfor Population and Development Studies, a fellow at the YaleProgram in Agrarian Studies, and a lecturer in Urban Studiesand Planning at Massachusetts Institute of Technology. He hasconducted research and published widely on local democracy,decentralization, environmental justice, social vulnerability,access to natural resources, and the effects of resource marketson local livelihoods. You may write to the author [email protected].

142


ABOUT WRI

World Resources Institute is an environmental researchand policy organization that creates solutions to protectthe Earth and improve people’s lives.

Our work is concentrated on achieving progress towardfour key goals:

● protect Earth’s living systems● increase access to information● create sustainable enterprise and opportunity● reverse global warming.

Our strength is our ability to catalyze permanent changethrough partnerships that implement innovative, incen-tive-based solutions that are founded upon hard, objec-tive data. We know that harnessing the power of marketswill ensure real, not cosmetic, change.

We are an independent and non-partisan organization.Yet, we work closely with governments, the private sector,and civil society groups around the world, because thatguarantees ownership of solutions and yields far greaterimpact that any other way of operating.

W R I R E P O R T

Worl d

Resources

I nsti tute

WAITING FOR DEMOCRACYThe Politics of Choice in Natural Resource Decentralization

JESSE C. RIBOT

World

Resources

Institute

10 G Street, NESuite 800Washington, DC 20002www.wri.org

WA

ITIN

G F

OR

DE

MO

CR

AC

Y

Th

e P

olit

ics

of

Ch

oic

e in

Na

tu

ra

l Re

so

ur

ce

De

ce

nt

ra

liza

tio

nW

RI R

EP

OR

T