From War to Democracy

Attempts to introduce democracy in the wake of civil war face a critical

problem: how can war-torn societies move toward peace and democracy

when competitive politics and hard-fought elections exacerbate social and

political conflict? Through a study of six themes (peacekeeping, manage-

ment of violence, power sharing, political party transformation, elections,

civil society, and international reactions to democratization crises) this

volume considers the dilemmas that arise in pursuing peace after civil

war through processes of democratization. The contributors’ research high-

lights the complex relationship between democratization, which is compe-

titive, and peacebuilding or efforts to achieve reconciliation. The book

offers insights into more effective action in peacebuilding in light of the

short-term negative effects that democratization can introduce. It is a

thought-provoking work that seeks both to advance theory and to provide

policy-relevant findings to facilitate more effective and durable transitions

from war to democracy.

A N N A K. J A R S T A D is Assistant Professor and Coordinator for the Conflict

and Democracy Program in the Department of Peace and Conflict

Research at Uppsala University, Sweden.

T I M O T H Y D. S I S K is Associate Professor in the Graduate School of

International Studies at the University of Denver.

From War to Democracy

Dilemmas of Peacebuilding

Edited by

A N N A K . J A R S T A D

a n d

T I M O T H Y D . S I S K

CAMBRIDGE UNIVERSITY PRESS

Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo

Cambridge University PressThe Edinburgh Building, Cambridge CB2 8RU, UK

First published in print format

ISBN-13 978-0-521-88566-9

ISBN-13 978-0-521-71327-6

ISBN-13 978-0-511-39492-8

© Cambridge University Press 2008

2008

Information on this title: www.cambridge.org/9780521885669

This publication is in copyright. Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press.

Cambridge University Press has no responsibility for the persistence or accuracy of urls for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

Published in the United States of America by Cambridge University Press, New York

www.cambridge.org

paperback

eBook (NetLibrary)

hardback

Contents

List of tables page vii

About the authors viii

List of acronyms xiii

Acknowledgments xvi

Introduction

ANNA K. JARSTAD AND TIMOTHY D. SISK 1

Part I The perils of war-to-democracy transitions 15

1 Dilemmas of war-to-democracy transitions: theories

and concepts

ANNA K. JARSTAD 17

Part II The security context 37

2 Peacekeeping and democratization

VIRGINIA PAGE FORTNA 39

3 Violence in war-to-democracy transitions

KRISTINE HOGLUND 80

Part III The political process 103

4 Power sharing: former enemies in joint government

ANNA K. JARSTAD 105

5 When rebels change their stripes: armed insurgents

in post-war politics

MIMMI SODERBERG KOVACS 134

6 Post-war elections: uncertain turning points of transition

BENJAMIN REILLY 157

v

7 Civil society in war-to-democracy transitions

ROBERTO BELLONI 182

Part IV International engagement 211

8 International responses to crises of democratization

in war-torn societies

PETER WALLENSTEEN 213

9 Peacebuilding as democratization: findings and

recommendations

TIMOTHY D. SISK 239

References 260

Index 284

vi Contents

Tables

1.1 Recent cases of peacebuilding after civil war page 33

2.1 Effects of peacekeeping on the duration of peace 53

2.2 Mean democratization scores (Polity), by peacekeeping 55

2.3 Democratization one year out 58

2.4 Democratization two years out 59

2.5 Democratization five years out 60

2.6 Democratization one year out, by peacekeeping

mission type 66

2.7 Democratization two years out, by peacekeeping

mission type 67

2.8 Democratization five years out, by peacekeeping

mission type 68

2.9 UN vs. non-UN peacekeeping 72

Appendix: List of cases 77

4.1 Examples of recent power-sharing accords 112

vii

About the authors

ROBERTO BELLONI is Assistant Professor at the School of Politics,

International Studies and Philosophy at Queen’s University, Belfast.

He was most recently a Fellow with the International Security Program

and the WPF Program on Intrastate Conflict at Harvard University. He

received his Ph.D. from the Graduate School of International Studies at

the University of Denver (2003). An Italian national, Belloni has exten-

sive work experience in Southeast Europe, in particular Bosnia-

Herzegovina, and also Kosovo and Croatia. His research focuses on

peacebuilding and post-war democratization, with particular emphasis

on the Balkans. He has published in the Review of International

Studies, International Peacekeeping, the Journal of Peace Research,

Ethnopolitics, Civil Wars, the Whitehead Journal of Diplomacy and

International Relations, International Studies Perspectives and in edi-

ted volumes. His recent work includes the volume State Building and

International Intervention in Bosnia (2007).

VIRGINIA PAGE FORTNA is Associate Professor at the Political Science

Department and the Saltzman Institute of War and Peace Studies at

Columbia University. Her research focuses on the durability of peace

in the aftermath of both civil and interstate wars. She is the author of

Does Peacekeeping Work? Shaping Belligerents’ Choices after Civil War

(2008) and Peace Time: Cease-Fire Agreements and the Durability of

Peace (2004) and has published articles in International Organization,

International Studies Quarterly, International Studies Review, and

World Politics. She is currently working on a research project on long-

term historical trends in war termination.

Fortna has been a Fellow at the Hoover Institution at Stanford

University (2004–2005) and a Visiting Fellow at the American

Academy of Arts and Sciences in Cambridge, MA (2002–2003). She

viii

was a pre-doctoral and then a postdoctoral fellow at the Center

for International Security and Cooperation at Stanford University

(1997–1999). Her graduate work was done at the Government

Department at Harvard University (Ph.D. 1998). Before graduate

school, she worked at the Henry L. Stimson Center, a think tank in

Washington, DC. She is a graduate of Wesleyan University.

KRISTINE HOGLUND is Assistant Professor at the Department of Peace

and Conflict Research, Uppsala University, Sweden. She has specialized

in the inter-linkages between conflict resolution and violence. Her

research has covered issues such as the dilemmas of democratization

in countries emerging from violent conflict, the importance of trust in

peace negotiation processes, and the role of international actors in

dealing with crises in war-torn societies. Cases analyzed include

Guatemala, Northern Ireland, South Africa, Kosovo, and Sri Lanka.

Hoglund has been a pre-doctoral visiting fellow at the Conflict

Management Program, ‘‘Nitze School of Advanced International

Studies’’ (SAIS), Johns Hopkins University, Washington, DC, and at

the International Institute for Applied Systems Analysis (IIASA),

Laxemburg, Austria. She holds a Ph.D. in peace and conflict research

from Uppsala University. Recent publications are ‘‘‘Sticking One’s

Neck Out’: Reducing Mistrust in Sri Lanka’s Peace Negotiations’’

(2006), with Isak Svensson, in the Negotiation Journal; ‘‘Violence by

the State: Official Spoilers and Their Allies’’ (2006), with I. William

Zartman, in John Darby (ed.), Violence and Reconstruction; and

‘‘Violence and the Peace Process in Sri Lanka’’ (2005), in the journal

Civil Wars.

ANNA K. JARSTAD is Assistant Professor and Coordinator for the Conflict

and Democracy Program at the Department of Peace and Conflict

Research at Uppsala University, Sweden. She specializes in power

sharing as a form of conflict management, in large-N studies, and in

in-depth studies, for instance in Cyprus, New Zealand, Kosovo, Bosnia

and Herzegovina, and Macedonia. Her work also includes research on

peacebuilding, democratization, and conflict management in ethnically

divided societies. A prior research project focused on dilemmas in

nation building, political integration, and ethnic identities in Cyprus.

That project was supported by the Swedish Institute of International

Affairs and focused on the consequences for conflict resolution given

About the authors ix

the EU accession of the Republic of Cyprus. She holds a Ph.D. in peace

and conflict research from Uppsala University. Her doctoral disserta-

tion, Changing the Game: Consociational Theory and Ethnic Quotas

in Cyprus and New Zealand (published 2001) investigated how ethnic

quotas can contribute to changes in the actor’s ranking order of pre-

ferences by upgrading the value of cooperation, thereby contributing to

viable peace. Recent publications are ‘‘To Share or to Divide?

Negotiating the Future of Kosovo,’’ Civil Wars 9 (3), 2007, and

‘‘Peace by Pact: Data On the Implementation of Peace Agreements,’’

with Ralph Sundberg, in Ashok Swain, Ramses Amer, and Joakim

Ojendal (eds.), Globalization and Challenges to Building Peace (2007).

BENJAMIN REILLY is Director of the Center for Democratic Institutions

at the Crawford School of Economics and Government, Australian

National University. His work focuses on democratization, political

institutions, and conflict management, and he has advised numerous

governments and international organizations on these issues. He has

held visiting appointments at Oxford, Canterbury, and Harvard uni-

versities, and his work has received financial support from the Carnegie

Corporation of New York, the United States Institute of Peace, the

East–West Centre, and the Australian Research Council.

His publications include Democracy and Diversity: Political

Engineering in the Asia-Pacific (2006), Democracy in Divided

Societies: Electoral Engineering for Conflict Management (2001),

Electoral Systems and Conflict in Divided Societies (1999), Democracy

and Deep Rooted Conflict: Options for Negotiators (1998), and The

International IDEA Handbook of Electoral System Design (1997). He

has also written for academic journals such as Comparative Political

Studies, the Journal of Democracy, International Security, The National

Interest, Party Politics, Electoral Studies, the International Political

Science Review, the Australian Journal of International Affairs,

International Peacekeeping and the Asian Survey. He holds a Ph.D. in

political science from the Australian National University.

TIMOTHY D. SISK is Associate Dean and Associate Professor at the

Graduate School of International Studies (GSIS), University of Denver,

and Director of the Center for 21st Century Global Governance, a

research and policy development institute at GSIS. He also serves as an

Associate Fellow of the Geneva Centre for Security Policy in Geneva,

x About the authors

Switzerland. Sisk specializes in peace processes and international con-

flict resolution, especially negotiation, mediation, and international

intervention in contemporary wars. His recent research has focused

on systematic approaches to conflict assessment and institutions and

processes for conflict management. He is currently finishing a book

titled Bargaining with Bullets. Violence, Negotiation and International

Mediation in Civil Wars.

A former Program Officer and Research Scholar at the federally

chartered United States Institute of Peace in Washington, DC, Sisk was

a Washington-based scholar and analyst of international relations and

US foreign policy for fifteen years. He is the author of five books

and many articles, including Democratization in South Africa (1995)

and Power Sharing and International Mediation in Ethnic Conflicts

(1995). Sisk earned a Ph.D. ‘‘with distinction’’ in political science (com-

parative politics, research methods) from the George Washington

University, in 1992, and an MA in International Journalism (1984)

and a BA in Foreign Service and German (1982) from Baylor University.

MIMMI SO DERBERG KOVACS is Assistant Professor at the Department of

Peace and Conflict Research at Uppsala University in Sweden. She holds

a M.Sc. in Peace and Conflict Studies from Uppsala University (1999)

and a M.Sc. in International Relations from the London School of

Economics and Political Science (2000). She has been a pre-doctoral

fellow with the Belfer Center for Science and International Affairs at

Harvard University. Her research interests include non-state actors in

civil wars, conflict resolution processes, and post-war democratization.

In her doctoral thesis she examined why some rebel groups but not

others successfully transform to viable political parties following peace

agreements in intrastate armed conflicts, with a focus on cases from

Cambodia, El Salvador, Mozambique, and Sierra Leone. She has written

the research report Democratization and Armed Conflicts in Weak

States for the Swedish International Development Cooperation Agency

(Sida) and has published in the journal Civil Wars.

PETER WALLENSTEEN holds the Dag Hammarskjold Chair in Peace and

Conflict Research at Uppsala University, Sweden (since 1985) and is

the Richard G. Starmann Sr. Research Professor of Peace Studies at the

University of Notre Dame, Indiana, USA (since 2006). The second,

updated edition of his book Understanding Conflict Resolution: War,

About the authors xi

Peace and the Global System was published in 2007. Wallensteen

directs the Uppsala Conflict Data Program (www.ucdp.uu.se) which

publishes annual updates on political conflict in the SIPRI Yearbooks,

the Journal of Peace Research, and the Human Security Reports.

Recently data on conflict prevention in low-intensity conflicts have

been added. Wallensteen also follows UN affairs on issues relating to

prevention, negotiation, and resolution of conflict. An assignment is to

lead the Special Program on the Implementation of Targeted Sanctions

which analyzes concrete experiences of UN and EU sanctions

(www.smartsanctions.se). A report was presented to the UN Security

Council in 2003 (Making Targeted Sanctions Effective: Guidelines for

the Implementation of UN Policy Options, with Carina Staibano and

Mikael Eriksson). He is co-editor of International Sanctions: Between

Words and Wars in the Global System (2005).

xii About the authors

Acronyms

ACS Assembly for Civil Society (Guatemala)

ANC African National Congress (South Africa)

ASEAN Association of Southeast Asian Nations

AU African Union

CNDD-FDD National Council for the Defense of Democracy-

Forces for the Defense of Democracy

(Conseil National pour la Defense de la Democratie-

Forces pour la Defense de la Democratie, Burundi)

CODES Convention for a Democratic South Africa

CPP Cambodian People’s Party

DDR disarmament, demobilization, and reintegration

DPA Dayton Peace Agreement

DPI Department of Public Information (United Nations)

DRC Democratic Republic of the Congo

ECOMOG Economic Community of West African States

Monitoring Group

ECOWAS Economic Community of West African States

EPL Popular Liberation Army (Ejercito Popular de

Liberacion, Colombia)

EPLF Eritrean People’s Liberation Front

EPRDF Ethiopian People’s Revolutionary Democratic Front

EU European Union

FARC Revolutionary Armed Forces of Colombia (Fuerzas

Armadas Revolucionarias de Colombia)

FMLN Farabundo Martı National Liberation Front (Frente

Farabundo Martı para la Liberacion Nacional, El

Salvador)

FRELIMO Liberation Front of Mozambique (Frente de

Libertacao de Mocambique)

FRETILIN Revolutionary Front of Independent East Timor

(Frente Revolucionaria de Timor-Leste Independente)

xiii

FUNCINPEC National United Front for an Independent, Neutral,

Peaceful, and Cooperative Cambodia (Front Uni

National pour un Cambodge Independant, Neutre,

Pacifique, et Cooperatif)

GAM Free Aceh Movement (Gerakan Aceh Merdeka,

Indonesia)

GDP gross domestic product

HDZ Croatian Democratic Union (Hrvatska demokratska

zajednica)

ICG International Crisis Group

ICTY International Criminal Tribunal for the Former

Yugoslavia

IDEA International Institute for Democracy and Electoral

Assistance

IFOR Implementation Force (Bosnia and Herzegovina)

IRA Irish Republican Army (Northern Ireland)

KFOR Kosovo Force

KPU National Elections Commission (Indonesia)

LRA Lord’s Resistance Army (Uganda)

LTTE Liberation Tigers of Tamil Eelam (Sri Lanka)

LURD Liberians United for Reconciliation and Democracy

MINURSO United Nations Mission for the Referendum in

Western Sahara

MINUGUA United Nations Verification Mission in Guatemala

MINUSTAH United Nations Stabilization in Haiti

MODEL Movement for Democracy in Liberia

MPLA Popular Movement for the Liberation of Angola

(Movimento Popular de Libertacao de Angola)

NATO North Atlantic Treaty Organization

NGOs non-governmental organizations

NPFL National Patriotic Front of Liberia

NPP National Patriotic Party (Liberia)

OAS Organization of American States

OECD/DAC Organization for Economic Cooperation and

Development / Development Assistance Committee

OHR Office of the High Representative

OLF Oromo Liberation Front (Ethiopia)

ONUMOZ United Nations Operations in Mozambique

ONUSAL United Nations Observer Mission in El Salvador

xiv List of acronyms

OSCE Organization for Security and Cooperation in Europe

Palipehutu–

FNL

Party for the Liberation of the Hutu People–Forces for

the National Liberation (Parti pour la liberation du

peuple Hutu – Forces nationales de liberation)

PDK Democratic Party of Kosovo (Partia Demokratike e

Kosoves)

PR proportional representation

PSA power-sharing accord

RENAMO Mozambic National Resistance (Resistencia Nacional

Mocambicana)

RUC Royal Ulster Constabulary (Northern Ireland)

RUF Revolutionary United Front (Sierra Leone)

SDA Party of Democratic Action (Stranka Demokratske

Akcije, Bosnia and Herzegovina)

SDS Serb Democratic Party (Srpska Demokratska Stranka,

Bosnia and Herzegovina)

Sida Swedish International Development Cooperation Agency

SFOR Stabilization Force (Bosnia and Herzogovina)

SLMM Sri Lanka Monitoring Mission

SNTV single non-transferable vote

SPLM Sudan People’s Liberation Movement

SRSG Special Representative of the Secretary-General

(United Nations)

SSR security sector reform

SWAPO South West Africa People’s Organization (Namibia)

UCK Kosovo Liberation Army (Ushtria Clivimtarc e Kosoves)

UN United Nations

UNAMSIL United Nations Mission in Sierra Leone

UNAVEM United Nations Angola Verification Mission (I, II, III)

UNDP United Nations Development Program

UNITA National Union for the Total Independence of Angola

(Uniao Nacional para a Independencia Total de Angola)

UNMIK United Nations Mission in Kosovo

UNTAES United Nations Transitional Administration for Eastern

Slavonia

UNTAET United Nations Transitional Administration in East

Timor

UNTAG United Nations Transition Assistance Group (Namibia)

URNG Guatemalan National Revolutionary Unity (Unidad

Revolucionaria Nacional Guatemalteca)

List of acronyms xv

Acknowledgments

This project has been a truly collective endeavor where all the authors

have commented and contributed to each other’s work. The book’s

editors, Anna Jarstad and Timothy D. Sisk, are deeply grateful to all

authors in the project for making it such a rewarding task. Our first

meeting was held in April 2005 in Sandhamn, Sweden, where we

discussed the theoretical framework and each contribution to the

project.

The project was generously supported by research awards from the

Swedish Vetenskapsradet (Research Council) and the Grant Program

of the United States Institute of Peace, and the editors thank these

organizations for their kind contributions.

The project participants would like to acknowledge a number of

scholars who have contributed comments and insights on the project

papers, including panel chairs and discussants at the 2006 Annual

Meeting of the International Studies Association, namely Larry

Diamond, Roland Paris, and Donald Rothchild. Thania Paffenholz

also deserves a special thanks for her comments on the chapter on civil

society. We are grateful for the recommendations of the two anonymous

reviewers who read the draft manuscript for Cambridge University

Press.

Additionally, International IDEA (the Institute for Democracy and

Electoral Assistance) and the University of Denver’s Graduate School

of International Studies also graciously supported the March 26–27,

2006, Vail Symposium ‘‘Dilemmas of Democratization in War-Torn

Societies’’ at which the papers in this project were presented. Judith

Large at IDEA is thanked especially for her ongoing support of the

project and for comments and insights at an early design phase of the

research.

The authors are grateful to the participants at Vail for their especially

valuable contributions to the project’s designs and especially its fram-

ing and conclusions. They are: Pauline Baker President, the Fund for

xvi

Peace; Elizabeth Cousens, Vice President International Peace Academy;

Tom Farer, Dean, Graduate School of International Studies, University

of Denver; Karen Feste, Professor, Graduate School of International

Studies and Director, Master of Arts Program in Conflict Resolution,

University of Denver; Matthew Hodes, Director, Conflict Resolution

Program, the Carter Center; Carlos Juarez Dean, College of Interna-

tional Studies, Hawai’i Pacific University; Sakuntala Kardirgamar-

Rajasingham, Head, South Asia Program, International IDEA; Brett

Lacy, Former Officer in Charge, Liberia, the Carter Center; Johanna

Mendelson Forman, Senior Program Officer, Peace, Security and Human

Rights, United Nations Foundation; Hiroko Miyamura, Senior Poli-

tical Officer, Electoral Assistance Division, United Nations; Lawrence

Robertson, Conflict Specialist, Office of Conflict Management and

Mitigation, US Agency for International Development; Ambassador

Gordon Smith, Director, Centre for Global Studies, University of

Victoria; Massimo Tommasoli, Director of Operations, International

IDEA. Participants of the research seminar at the Department of Peace

and Conflict Research, Uppsala University, especially Cecilia Albin,

Thomas Ohlson and Mats Hammarstrom, have contributed with valu-

able comments on the theory chapter, and Ralph Sundberg has com-

piled data on peace agreements. Additionally, several graduate

students at GSIS have made especially helpful comments on the book:

they are Kris Bauman, Andrew Barwig, Sumani Dash, and Arturo

Lopez-Levy.

The editors and authors would also like to pay tribute to the late

Donald Rothchild, who passed away in February 2007 following a

brief but aggressive illness. Don read and commented on several of the

chapters in this book, in his usual collegial style, and his prolific work

on peacebuilding in divided societies was formative in our collective

learning. His life as a scholar will continue to be one that we aspire to

emulate.

Acknowledgments xvii

Introduction

A N N A K . J A R S T A D A N D T I M O T H Y D . S I S K

Introducing democracy in the wake of civil war raises a stark question:

How can societies shattered by war, with all the deep social enmity,

personal suffering, and economic devastation that war brings, simulta-

neously move toward peace and democracy when competitive politics

and hard-fought elections exacerbate social and political conflict? This

book explores this question from two somewhat disparate strands of

scholarly research: democratic transition theory and practice, which

emphasizes, in the move from authoritarian rule to more democratic

politics, elite-negotiated democratization pacts, popular mobilization,

political party transformation, constitution making, electoral design,

and resurrection of civil society; and theory and practice of post-war

peacebuilding, with its emphasis on the elite and public negotiation of

comprehensive peace agreements, the search for security through cease-

fires, demobilization of armed forces, inclusion and reconciliation, exter-

nal security guarantees, and long-term conflict transformation.

Introducing democracy in the wake of war has become a standard

practice: since the 1990s, democratization is an integral part of inter-

national peacebuilding missions in the wake of civil war. Democracy

and peace – two often-desired goals – are promoted in war-torn socie-

ties shattered by war. However, today’s headlines – from Afghanistan,

Kosovo, Nepal, and the Democratic Republic of Congo (DRC) to name

but a few – reflect the evident dilemmas of war-to-democracy transi-

tions. In these and other cases, democracy and peace do not always

move forward hand in hand: sometimes, advances in democratization

threaten peace, and the compromises necessary for peace restrict or

defer democratization.

This book is about the dilemmas that arise in pursuing peace after

civil war through processes of democratization. We seek to identify

and evaluate the core dilemma of peacebuilding versus democr-

atization, and several manifestations of this dilemma, in six issue

areas: peacekeeping, management of violence, power sharing, political

1

party transformation, elections, civil society, and international reac-

tions to democratization crises. This research effort enhances the

understanding of the complex relationship between the two processes,

with an eye toward more effective action in peacebuilding.

The book also outlines ways to accommodate the negative effects

that occur when the processes of democratization and peacebuilding

clash. It proposes designs of peace missions that avoid creating dilem-

mas, but also identifies situations where dilemmas cannot be overcome

and where a choice has to be made between efforts to promote peace or

democracy. In this way, the research presented in this volume seeks

both to advance theory and to provide policy-relevant findings to

facilitate more effective and durable war-to-democracy transitions.

War-to-democracy transitions: patterns and rationales

In bringing armed conflicts in the 1990s and early 2000s to an end, a

critical question emerged for policymakers seeking to secure sustain-

able peace: How can the international community assist societies

wracked by internal war to transform in a way that deep-rooted social

conflicts can be ameliorated through non-violent means? Two con-

tending answers to this question have been put forward: one is to

separate warring parties by creating new sovereign states, especially if

the war has been fought among territorially distinct ethnic, linguistic,

or national groups (Kaufmann 1997). The other is to end the war by

encouraging the parties to negotiate a settlement and to undergo a war-

to-democracy transition within an existing state, in which conflicts on

the battlefield or the street are ended through the sequenced introduc-

tion of democracy: elections, parliamentary politics by political parties,

independent judicial institutions, and resuscitation of civil society all

underpinned by a basic floor of human rights usually enshrined in

newly negotiated constitutions.1

No post-Cold War civil war has been terminated by a peace agree-

ment stipulating partition, in line with the first plausible outcome.

Indeed, of the instances of partition of existing states since the end of

the Cold War, none were the outcome of a negotiated peace agreement

between the government and the armed opposition group (although

1 Recent evidence suggests that today’s wars are much more likely to end at thepeace table than on the battlefield; see Eriksson and Wallensteen (2004).

2 Anna K. Jarstad and Timothy D. Sisk

some settlements involve extensive decentralization that approaches

internal partition). The breakup of the former Soviet Union was a

disintegrating empire, and it occurred not as a result of a civil war.

The breakup of Yugoslavia was a result neither of a peace deal, nor of a

civil war. On the contrary, the wars in former Yugoslavia occurred

after the declaration of independence by the respective former repub-

lics. The independence of Eritrea (de facto in 1991 and de jure in 1993)

was never agreed in a peace agreement, but proclaimed after the military

victory of the opposition, and a subsequent referendum. Timor Leste

(formerly East Timor) achieved its independence after an agreement

between Portugal and Indonesia; no negotiations took place between

the Revolutionary Front of Independent East Timor (FRETILIN) and

Indonesia. Also in this case, independence was approved in a referendum

(UCDP 2007). Indeed, many of the internal armed conflicts involving

territorial claims have been ‘‘terminated’’ by cease-fire agreements but

have not advanced to comprehensive peace settlements in part because

the international bias against the creation of new states as an outcome of

civil war usually takes partition off the table: Cyprus, Azerbaijan

(Karabagh), and Georgia are all examples.

The second alternative, to encourage the warring parties to reach a

comprehensive, negotiated settlement featuring a transition to democ-

racy, is essentially the default approach of the international community

in its response to end contemporary wars. The bias against partition of

existing states in the international system is one reason, but it is not the

only one. The other is that partition in itself does not solve the problem

of contending social groups living together; it only rearranges territor-

ial borders but does not solve the problem of managing social conflict

(Chesterman, Farer, and Sisk 2003). Thus, the move toward democ-

racy after war is the imperative even in those instances where prior

historical legacies have led to newly independent states such as Bosnia

and Herzegovina or Timor Leste. War termination today is principally

about building anew or to rebuilding functioning, secure, stable, and

democratic (or ‘‘republican’’) states (Barnett 2006).

Data generated from the Uppsala Conflict Data Program demon-

strate that the most common provision in peace accords for resolving

conflicts over government is the holding of elections, while peace

accords after conflicts over territory often establish local governance

over the disputed territory. After a peak during 1991–1992 when fifty-

one armed conflicts were active, the number of conflicts has for the last

Introduction 3

couple of years decreased to a level equivalent to the levels of the 1970s

(around thirty armed conflicts). Since 2004, all conflicts have been

recorded as some type of intrastate conflict.2 Peace agreements were

concluded in one-third of armed conflicts that have been active since

the end of World War II, and more conflicts are being settled than new

ones are emerging (Harbom and Wallensteen 2005; Harbom,

Hogbladh, and Wallensteen 2006).

It is not surprising, then, that comprehensive peace agreements in

civil wars today – from Namibia in 1990 to more recent cases of

Afghanistan, Liberia, or the Ivory Coast – generally envisage democ-

racy as the end-state of a peacebuilding transition process, replete with

promises for the full protection of human rights, for electoral processes

in lieu of battlefield encounters, for transitional justice mechanisms

often lenient to those who have waged violence, and for the promised

arbitration of disputes through law instead of the rule of the gun.3

Dilemmas: international community perspectives

The actors involved in war-to-democracy transitions view the dilemma

of peacebuilding versus democratization from different perspectives.

For whom is this dilemma experienced, and how?

The external motives for post-war democratization are compelling. In

the cold reality of negotiated peace agreements following civil war today,

where the international community’s normative and material levers of

inducement are ubiquitously brought to bear, the war-termination

choice for a process of democratization is today a preferred choice.

For the international community democratization is a process by which

the root causes and articulated grievances of the parties can be

2 According to Uppsala Conflict Data Program (UCDP), the conflict between Iraqand the coalition of the United States, the United Kingdom, and Australia wascoded as interstate in 2003 and as internationalized internal armed conflict for theyears 2004, 2005, and 2006. An armed conflict is defined by UCDP as a contestedincompatibility that concerns government or territory or both, where the use ofarmed force between two parties results in at least twenty-five battle-relateddeaths. Of these two parties, at least one has to be the government of a state. Fordetailed definitions of the different categories of armed conflicts, seewww.ucdp.uu.se. In this book, the term civil war is used more broadly than theUCDP definition of intrastate war. Here it also refers to minor internal armedconflicts that do not meet the UCDP criteria.

3 For an evaluation of post-war peacebuilding, see Paris 2004.

4 Anna K. Jarstad and Timothy D. Sisk

negotiated without recourse to bloodshed and, ideally, consistent with

the norms and principles outlined in international law (Franck 1992).

Likewise, democratization is increasingly linked to state building, for

without an electoral process there is no mechanism for generating

internal legitimacy for peace agreements. As Benjamin Reilly appro-

priately observes, ‘‘In any transition from conflict to peace, the creation

or restoration of some form of legitimate authority is paramount . . . the

support of the citizenry must be tested and obtained’’ (Reilly 2003a:

174). The faith-like belief in an ‘‘internal’’ democratic peace in the post-

Cold War era is as strong as international liberalism’s devotion to an

international democratic peace. Kofi Annan, the seventh Secretary-

General of the United Nations (UN), succinctly described the connec-

tion between democracy and peace:

At the center of virtually every civil conflict is the issue of the State and its

power – who controls it, and how it is used. No conflict can be resolved

without answering those questions, and nowadays the answers almost always

have to be democratic ones, at least in form . . . Democracy is practised in many

ways, and none of them is perfect. But at its best it provides a method for

managing and resolving disputes peacefully, in an atmosphere of mutual trust.4

For the international community, a war-to-democracy transition has a

certain undeniable appeal: the alternatives of authoritarian control or

partition are most often shelved as untenable outcomes for the interna-

tional community. But at the same time, democratization and peace-

building introduce acute dilemmas for external actors. Pauline Baker

insightfully summarizes the inherent tensions in international action in

war termination, stemming from the countervailing pressures within the

international community (and, conceivably, within individuals such as

policymakers who are internally weighing alternative approaches to war

termination):

[c]onflict managers tend to concentrate on short-term solutions that address

the precipitous events that sparked the conflict; above all, they seek a swift

and expedient end to the violence. Democratizers tend to concentrate on

longer-term solutions that address the root causes of the conflict; they search

for enduring democratic stability. The former see peace as a precondition for

4 UN Secretary-General Kofi Annan, ‘‘Why Democracy is an International Issue,’’Cyril Foster Lecture, Oxford University (UK), June 19, 2001 (available atwww.un.org/News/ossg/sg/stories/statements_search_full.asp?statID=11).

Introduction 5

democracy; the latter see democracy as a precondition for peace. (Baker

2001: 760)

External actors face perplexing problems, for example whether to

include perpetrators of violence in power sharing, to hold elections

despite insufficient security (with the hope that violence will ebb), to

bargain mostly with elites or to try to engage a wider group of parties

(such as political parties or civil societies), or whether to engage rebel

forces with a view toward their transformation into political parties.

These issues also arise when international actors are considering the

extent of their involvement in civil war termination efforts, in how to

engage (such as helping parties design the course of events in a war-to-

democracy transition, or ‘‘sequencing,’’ and when to leave).

Peacebuilders in war-torn societies face the difficult challenges of

providing security, fostering resuscitation of civil society, transforming

armed actors into human-rights-abiding democrats, providing basic

humanitarian relief and ‘‘peace-divided’’ development, and breaking

the rent-seeking ties of political economy that fueled the war for states

and rebel forces alike (Collier et al. 2003). Perhaps the most difficult

dilemma faced by international actors, particularly in UN peace opera-

tions, are challenging questions over the use of coercive measures such

as force. Use of force by peace operations to buttress a negotiated

settlement, especially when the legitimacy of action by the international

community is disputed or resisted (see the respective chapters by Virginia

Page Fortna, Kristine Hoglund, Roberto Belloni, and Peter Wallensteen

in this volume), is risky and prone to backfire. Should the UN be in the

business of ensuring democracy at the barrel of a gun?5

When international actors engage, their interests may be insuffi-

ciently aligned causing a disconnection among the various types of

international actors who – generally with good intent – engage in the

efforts to bring peace to war-shattered states. Roberto Belloni shows

that coordination problems among international actors have been

central barriers to the deepening of peace in efforts to engage civil

5 Joanne Mariner of Human Rights Watch argues that in the case of Haiti, forinstance, it was important for the UN to use its military clout to prevent violentdisruption of the country’s elections. She writes that ‘‘It is crucial for the electionsto be credible in the eyes of the Haitian people. Otherwise, instead of advancingmuch-needed stability they could trigger yet another crisis.’’ See Haiti: Secure andCredible Elections Crucial for Stability, available at http://hrw.org/english/docs/2006/02/06/haiti12611.htm.

6 Anna K. Jarstad and Timothy D. Sisk

society in the war-to-democracy transition process in Bosnia and

Herzegovina. Peter Wallensteen shows that international responses

are episodic in the face of crises in war-to-democracy processes, and

that problems of coordination and will inherently limit the ability of

the international community to help parties negotiate successfully to

overcome the turbulence of such transitions. Moreover, they face

difficult choices over the instruments of support and coercion that

could potentially be brought to bear.

Finally, there is increasingly concern about the issue of authority in

war-to-democracy transitions. The international community has

assumed a more authoritative role through extensive international over-

sight or even transitional administration in cases either where the local

authorities fail to prevent crimes against humanity and mass violence

(e.g., Sierra Leone, Timor Leste), or when the state itself has been culp-

able in committing atrocities (e.g., in Cambodia and Kosovo, where

government authorities were accused of genocidal crimes). A difficult

challenge arises as a result of the need for firm international control of

the situation to manage problems of spoiler violence, or to organize

elections, while engaging in such a way that empowers local actors

(such as electoral management bodies), affirms state sovereignty, and

leads to a capable, functioning state when the international community’s

oversight ends. Simon Chesterman has shown that this challenge is

inherent in transitional administrations and that the United Nations,

especially, faces the problem of building democracy in war-torn societies

in ways that are fundamentally undemocratic (Chesterman 2004).

Dilemmas: protagonist perspectives

The endogenous motives for civil war protagonists is simple: democra-

tization provides a set of rules under which conflict can continue to be

waged through formal, rule-oriented institutions such as electoral and

parliamentary processes that offer a fundamental floor of human rights

in the event one party or another finds itself on the losing side of

collective decision-making processes. In John Rawls’ classic book

A Theory of Justice (1971), he postulates hypothetical negotiations

among individuals seeking to establish anew a political community

while ignorant about their future positions and status. In peace talks,

the protagonists negotiate the future through what is essentially a new

Rawlsian social contract, albeit without a fully obscured ‘‘veil of

Introduction 7

ignorance.’’6 As Viktor Vanberg and James Buchanan (1989: 61) have

argued: ‘‘Cooperation can replace conflict only if the differing interests,

held with different intensities by different persons, can be traded-off or

compromised, actually or symbolically, in a social contract.’’

Civil wars by definition feature factions that have some degree of

coherence; indeed, the coherence of a protagonist group, such as a rebel

force that seeks to represent an ethnic interest, is a key variable in

explaining the likelihood of negotiated settlements to civil wars in the

first place. Civil war protagonists view the conclusion of peace agree-

ments through democratization as attractive, but risky; likewise, they

view peacebuilding approaches such as power-sharing pacts as less

risky, but unattractive. As Mimmi Soderberg Kovacs and Anna K.

Jarstad claim in their respective chapters of this volume, protagonists’

aversion to democratization and peacebuilding poses severe obstacles

for a war-to-democracy transition. Rebels are not always interested

and able to transform into political parties, because if they emerge as a

political party, they risk losing the election. Power sharing can provide

guaranteed inclusion – and thereby an incentive for such transforma-

tion – but how does inclusion of former warring parties affect the

quality of democracy that emerges?

In turning to the war-to-democracy formula for war termination,

protagonists in civil wars face difficult challenges: because the interna-

tional system fails to adequately and consistently provide for external

security guarantees, protagonists face difficult dilemmas of uncer-

tainty. Comprehensive peace agreements do not end conflict (or even

violence, as Kristine Hoglund demonstrates); they simply set up pro-

cesses that give peace a chance to unfold over time.

As civil war negotiations ensue, state incumbents or rebel forces find

obstacles of democratization due to grave uncertainty for the future:

they have an insufficient capacity to determine whether through demo-

cratic processes – notably but not only elections – they will be enabled

to protect their vital interests into the future. The turbulence of war

does not offer a safe place from which to make judgments about

whether peace agreement guarantees, constitutional guarantees, laws

and institutions will be sufficient protection over time. While there may

6 In this sense, comprehensive peace agreements can be considered incipient socialcontracts, based on the principal of reciprocity that links the pursuit of justice andfairness to the establishment of political institutions; see Rawls (1971: 99).

8 Anna K. Jarstad and Timothy D. Sisk

be interest in escaping a conflict trap through a democratization for-

mula,7 it is a shaky accord upon which to base future prospects. In sum,

as several scholars have artfully shown, the long shadow of future

competition in elections creates a classic security dilemma for civil

war protagonists over time, one that grips them in a thick pall of

uncertainty.

On the other hand, peace agreements that limit uncertainty in demo-

cratization processes – such as power-sharing pacts – also contain

challenges for protagonists in terms of their strategies. Just as during

the war parties faced choices over whether to talk or fight, a negotiated

peace pact does not alleviate trade-offs related to strategy, it only

changes them. That is, in post-war transitions, especially as elections

loom and mobilization of constituencies heightens, protagonists must

choose strategies that simultaneously can maximize their vote share –

often, by emphasizing lines of conflict and difference – while needing to

conciliate with opponents in implementation of a peace pact (such as

disarmament). Protagonists in war-to-democracy dilemmas face these

challenges on a daily basis: cooperation and conflict go side by side as

bargainers in implementing peace agreements pursue countervailing

pressures of mobilization and conciliation.8

The issues of timing and sequencing are both sources of the dilemma

between efforts to promote democracy and peace – and key to a possible

way to a synchronized war-to-democracy transition. As the chapter

by Virginia Page Fortna in this volume indicates, here protagonists

respond to the putative assurances of external parties that – through

7 Formulas are broad principles framed to narrow the parameters of a conflict’soutcome; a formula defines an overarching concept that frames the parameters ofthe solution and defines the terms of trade or establishes a principle under whichthe conflict can be cooperatively managed. In economist’s terms, it defines thecontract zone. In order for parties to accept a formula, often but not alwaysproposed by a mediator, it must be seen as just and satisfactory; cover all majorissues; incorporate all sides’ demands; and contain a basic vision of post-wararrangements. A formula is not a settlement, but rather a statement of the scope ofthe conflict’s outcomes and the general procedures to get the parties to settlement:a formula is an agreement on certain basic conceptual issues needed to be resolvedbefore the bargaining on details can begin in earnest, for example a generaldeclaration of principles or framework agreement. See Druckman (1986).

8 For a review of the international community’s experience in post-war governance,see ‘‘Governance in Post-Conflict Situations: Lessons Learned,’’ United NationsDevelopment Program and the Christian Michelsen Institute, May 2004, availableat www.cmi.no/events/?undp-2004-governance-in-post-conflict-situations.

Introduction 9

the deployment of a peace operation – some of their problems of

uncertainty and strategy can be mitigated by the security presence

and skillful mediation diplomacy of outside actors, augmented as it

normally is through humanitarian relief and pledges of long-term

development aid. While engagement with external, international actors

by civil war protagonists raises a number of problems and obstacles –

the primary one is certainly security. Are the external guarantees for

ensuring compliance sufficient to allay protagonists’ fears while choos-

ing a peace-through-democratization formula? The commitment pro-

blem is especially acute in the long term: today, protagonists well know

that peacekeeping missions do not last forever, and indeed there are

pressures on the United Nations, for example, to manage a transition

quickly so as to move resources on to the next crisis (today, shifting

from Liberia to Sudan). Thus, external capacities to resolve protago-

nists’ commitment problems are temporary: over time, this issue, too,

cannot be avoided.

Settlements in civil wars reflect the convergence point of the parties’

preferences over new rules structures, or institutions, for the state once

arms have been laid down. Waterman (1993: 292) argues that ‘‘civil

wars are conflicts over political order,’’ and settlements in them entail

the ‘‘re-creation of the conditions for a viable, common political

order.’’ Importantly, settlements do not end conflicts: they are simply

agreements to continue bargaining under consensually defined rules of

interaction. Not surprisingly, settlements in internal conflicts often

take the form of new constitutions or significant packages of amend-

ments to existing constitutions. In the course of formal substantive

negotiations, parties formulate their positions based on their expecta-

tions of how the structure of the new institutions will serve their

interests; they exercise ‘‘analytical imagination’’ about the costs and

benefits of alternative institutions, such as the electoral system (Sisk

1995). Therefore, settlements do not definitively end civil wars, but

instead they are promises to end conflicts by creating new democratic

rules of the game to which all parties at the table can agree.

Exploring the dilemmas

This book investigates the dilemmas of democratization in war-torn

societies. In the first chapter (Part I), Jarstad investigates the tensions

between peacebuilding imperatives and democratization more fully;

10 Anna K. Jarstad and Timothy D. Sisk

presenting in depth and in reference to the existing literature a typology

of dilemmas that captures their multidimensional nature. The four

types of dilemmas that may arise when the processes of democratiza-

tion and peacebuilding have adverse effects on each other are subcate-

gories of the overarching dilemma and pertain to trade-offs between

efforts to promote peace or democracy. The first type of dilemma, the

horizontal dilemma, regards the relation between the elites of warring

parties and democratic political parties that experience a trade-off

between inclusion and exclusion.

The vertical dilemma entails the difficult choice between efficacy and

legitimacy. It regards the relation between the elite and mass politics.

For the sake of legitimacy, involvement of the people in all phases of the

peace process is desirable. But for the sake of efficacy, certain negotia-

tions need to be held in secret. The systemic dilemma refers to the issue

of ownership, of local versus international control of the processes of

democratization and peacebuilding. International involvement may be

necessary to end violence and to facilitate negotiations. But, democra-

tization as well as peacebuilding needs to be driven by local motives

and actions. The citizens of the countries that receive support to demo-

cratization and peacebuilding must feel that they own the processes,

and that democracy is not imposed from the outside.

The temporal dilemma occurs when there are trade-offs between

short-term and long-term effects on democratization and peacebuild-

ing. Efforts to support democratization may in the short run increase

the risk of violence, and thereby in the long run undermine the chances

for democracy to take root. Peacebuilding may involve restrictions on

democratic freedoms such as freedom of the press and mass demonstra-

tions. In the long run, such constraints may cause unrest and not only

have a negative effect on democratization, but also turn into an obsta-

cle for implementation of the peace agreement.

Part II of the volume addresses the key security issues that arise in a

war-to-democracy transition, how they are experienced, and how they

are addressed. The management of security raises a whole host of

potential problems, challenges, and obstacles to democracy: the demo-

bilization, return, and reintegration of former combatants, lingering

threats of political violence, the need for election-related dispute reso-

lution, and solving the ‘‘credible commitment’’ problem through the

deployment of peace operations. In Chapter 2, Fortna explores the

dilemma that arises when peacekeepers help provide the security and

Introduction 11

stability necessary for initial steps toward democracy, while the pre-

sence of such foreign actors can undermine the long-term development.

She also investigates the possibility that precisely by providing peace

and security, peacekeepers may undermine some of the incentives that

the pressure of war-making provides to processes of state building and

democratization. The third chapter, by Hoglund, analyses the complex

relationship between democratization and violence. Democratization

creates new opportunities and motives for violence. And measures to

combat violence – for instance through the use of coercive measures

and the inclusion of spoilers in politics – can undermine the democra-

tization process. She argues that new forms of violence emerge as the

organized political conflict subsides, rendering old ways of dealing

with violence ineffective.

Part III is about political transitions. Because protagonists in

today’s war rarely lose everything on the battlefield – instead, the

war-to-democracy transition is negotiated – they usually end up

sharing power either in temporary, transitional arrangements or in

permanent, power-sharing constitutions. As Jarstad writes in

Chapter 4, power sharing is essentially a coalescence of elites at the

top, and one of the enduring concerns about power-sharing solutions

are their top-down orientation, their perceived inflexibility, and the

allegations that in many instances they reinforce ethnic, nationalist, or

extremist tendencies.

When rebels lay down their arms, they often do so in the expectation

that they might win at the ballot box what they fought for on the

battlefield: they seek to transform into political parties. Soderberg

Kovacs investigates this issue in Chapter 5 on rebel-to-political party

transformations, illustrating the conditions under which some forces

can adapt and evolve, whereas others simply cannot.

Likewise, during the war-to-democracy transition critical choices are

made over electoral processes: the sequencing of elections, the electoral

system formula, the nature of elections (e.g., to a legislature, constitu-

ent assembly, or both), and other critical election-related issues such as

application of citizenship laws. What are the principal findings from

research and practical lessons learned on sequencing, shaping, and

managing electoral processes in war-torn societies? Reilly’s contribu-

tion in Chapter 6 evaluates the track record on the pivotal issue of

elections, and in particular the conditions under which electoral

processes may promote democratization but undermine peace, when

12 Anna K. Jarstad and Timothy D. Sisk

peacebuilding undermines democratization, and the conditions under

which electoral processes may contribute to both goals. In Chapter 7,

Roberto Belloni focuses on civil society; he argues that there is an

urgent need to move beyond the platitudinous endorsement of ‘‘civil

society’’ as essential element of peacebuilding by asking the tough

questions about which elements of society may be civil, and worthy

of inclusion, and those that are not.

Part IV evaluates the role of the international community and, in

conclusion, provides the policy-relevant findings that emerge from this

research. In Chapter 8, Wallensteen explains the lateness of interna-

tional responses to democratization crises in war-torn societies,

arguing that much earlier action is required if such crises are to be

avoided and the dual objectives of peacebuilding and democratization

are to simultaneously advance.

In the book’s concluding chapter, Timothy D. Sisk returns to the

question of ‘‘dilemmas for whom’’ and pulls together the key findings

of the foregoing contributions on how international actors and prota-

gonists must face up to the reality that there are inherent trade-offs and

difficulties in war-to-democracy transitions. The clear message of this

research is that in some instances, the dilemmas of democratization

simply cannot be avoided: either the imperatives of peacebuilding and

conflict management, or the imperatives of democratization, must be

singularly chosen in a particular situation as the right path to peace.

Generally, the choice is made for peace over democratization (by

delaying or restricting elections, for example); this choice is under-

standable, but it is a myopic one. Rather than foregoing democratiza-

tion, this research reveals that it must be done more effectively, and

cunningly. In some instances, the dilemmas can be avoided: through

ingenuity, policymakers and protagonists in conflict may be able to

reconcile the divergent choices they face and design sequences of

change that can simultaneously advance war-torn societies on the

path to peace and toward the best hope for peacebuilding in the long

run: democracy.

Introduction 13

PART I

The perils of war-to-democracytransitions

1 Dilemmas of war-to-democracytransitions: theories and concepts

A N N A K . J A R S T A D

War-torn societies entering the path toward democracy and peace face

the challenge that efforts to achieve one of these desirable goals can

have negative effects on the other. This chapter puts forward the

concept of war-to-democracy transitions to capture the dynamic

between the two interacting processes of democratization and peace-

building. Democratization refers to the process of opening up political

space, including improvements regarding contestation, participation,

and human rights. Peacebuilding regards the post-settlement period

and includes the implementation of the peace agreement.1 The focus is

on the dilemmas that arise when the two processes have adverse effects

on each other. A dilemma is a trade-off situation, where usually the

choice is between two bad things. However, here the dilemmas regard

two goals widely held to be mutually reinforcing, namely democracy

and peace. Such dilemmas for post-war transitions occur when actions

taken in the name of democratization have negative effects on the peace

process. Dilemmas are also activated when actions taken in the name of

1 A broad understanding is that peacebuilding denotes the various efforts insupport of political, institutional, and social transformation necessary to bringabout lasting peace (Bertram 1995). Besides the commonly used definition ofpeacebuilding as employed in the Agenda for Peace, there are at least two otherinterpretations of peacebuilding. Drawing on Johan Galtung, peacebuilding hasbeen used to refer to non-elite processes, beyond and below the state. Otherscholars use the term peacebuilding broadly to refer to peacemaking,peacekeeping, and conflict prevention (Call and Cook 2003). Such efforts mayinvolve the local population and local elites as both initiators and recipients ofassistance, as well as intergovernmental organizations and non-governmentalorganizations (NGOs). For our analysis of war-to-democracy transitions,peacebuilding begins with a peace accord that settles at least one of theincompatibilities at stake in the conflict, such as control over territory orgovernment. Peace agreements do not always put an end to violence. In othercases, the fighting ends but an agreement regulating the incompatibilities is notreached until years later. The focus here is on post-war peacebuilding, i.e., theprevention of a relapse into conflict. Peacebuilding is thus defined as efforts toimplement and consolidate violent peace agreements.

17

peace have negative effects on democratization. In addition, the dilem-

mas are often interacting: there is seldom one dilemma at a time, and

one dilemma may make another dilemma even more critical. Such

dilemmas pose severe challenges for both local and international actors

engaged in peacebuilding and democratization.

The chapter discusses four types of trade-off situations where the

choice is between reforms to promote democracy versus efforts to

secure peace. These dilemmas are here referred to as the horizontal

dilemma (i.e., inclusion versus exclusion), the vertical dilemma (i.e.,

legitimacy versus efficacy), the systemic dilemma (i.e., local versus

international ownership of the processes), and the temporal dilemma

(i.e., long-term versus short-term efforts). Failure to deal with such

dilemmas can result in a return to war (e.g., Angola 1992 and Liberia

2000). Alternatively it can result in backsliding to authoritarianism as

in, for example, Haiti 1994–2005 and Ethiopia after the elections in

2005. It is suggested that a theoretical explanation of why democrati-

zation in war-torn societies succeeds or fails needs to include a simul-

taneous analysis of these four dilemmas.

The overarching purpose of this chapter is to enhance our under-

standing of why efforts to promote democracy and peace do not always

go together. This book builds on research stemming from previously

separate discourses on democratization, peacebuilding, and conflict

theory to construct a framework for the analysis of simultaneous

democratization and peacebuilding. The concept of war-to-democracy

transitions is developed and explored. This chapter discusses why the

combination of a legacy of war, reforms to democratize, and efforts to

build peace often result in dilemmas where peacebuilding and demo-

cratization have adverse effects on each other. The failure to deal with

such dilemmas can have devastating effects, thus undermining both

long-term democratization and peace. A broad conclusion is that when

the choice is between securing the peace and promoting democracy,

peace should be given priority.

To date, discourses on democratization and armed conflict have not

been integrated, leaving a gap in our understanding of potential trade-

offs between peace versus long-term democratization for societies

shattered by conflict. Democracy is commonly understood as a system

where diverse interests are managed through ongoing negotiations and

accommodated by accountable and legitimate institutions. Although

conflicts are seldom fully resolved, democracy supposedly manages

18 Anna K. Jarstad

them by peaceful means (Commission on Global Governance 1995;

Przeworski 1991). In this way democracy and peace reinforce each

other. Democracy is consolidated when peaceful means of conflict

management are accepted as ‘‘the only game in town’’ (Linz and

Stepan 1996: 5). Sustainable peace presupposes a system of governance

where diverse interests and grievances are accommodated by negotia-

tions and compromises (Licklider 2005: 35; Wallensteen 2002:

139–144). Democratization and peacebuilding have also been wishfully

thought of as parallel and mutually beneficial processes. However,

researchers and policymakers have identified an apparent paradox:

while democracy as a political system is associated with peaceful conflict

management both within and between states, the road to democracy is

often conflict-ridden.2

The conditions typical for war-torn societies, as well as the dynamics

of and interplay between the two processes of peacebuilding and

democratization, contribute to this contradiction. While non-violent

conflicts are healthy features of any democracy, violent conflicts under-

mine the foundations for a functioning democracy. Electoral violence,

political assassinations, violent riots, and extreme levels of crime are

threats to the new political order and to basic civilian security. Failure

to deal with violence can lead to escalation of violence and a vicious

circle of retribution and violations of human rights.

After a peace deal is reached, the legacies of war tend to linger.

Insecurity and unsolved grievances mean that political elites, as well

as civil society, remain polarized and that the basis for inclusive ideol-

ogies is weak. In combination with a shattered infrastructure, and an

economy structured on the spoils of war, this polarization implies that

democratization faces particular challenges in post-war societies. This

is why the core elements of democracy, such as popular participation,

mobilization of interest groups, and open competition between politi-

cal parties, increase the risk of violent conflict in societies entering a

democratization process. In addition, efforts in support of peace deals

constrain the process of democratization. For example, the inclusion of

former rebels in government for the sake of peace may undermine

democratic legitimacy and long-term stability. When this is the case,

2 See, e.g., Brass 1991; Brown, Lynn-Jones, and Miller 1996; Gleditsch and Hegre1997; Mann 2005; Mansfield and Snyder 1995; Maoz and Russett 1993;Snyder 2000.

Dilemmas of war-to-democracy transitions 19

the simultaneous processes of peacebuilding and democratization have

adverse effects on each other.

War-to-democracy transition

Many contemporary democratization processes take place in societies

shattered by war. Previous research has often focused either on demo-

cratization (often analyzing transitions taking place in societies that

have not experienced armed conflict) or peacebuilding after intrastate

armed conflicts without any analysis of democratization. A common

assumption is that democracy implies peace and, vice versa, that peace

implies democracy. Multidimensional peace operations set out to

achieve both peace and democracy. The expectation is that post-war3

transitions result in both peace and democracy. However, the obstacles

facing many societies undergoing such transitions suggest a need to

combine the experiences of war-shattered societies with an integrated

theoretical framework on the processes of democratization and

peacebuilding.

The transition from relatively stable authoritarianism in Yugoslavia to

conflict-ridden democratization in Kosovo and the resumption of war

after elections in Angola 1992 give an indication of the broad scope of

cases undergoing simultaneous peacebuilding and democratization.

These transitions vary a great deal, for example in terms of starting

point (e.g., previous history of governance and type of warfare), con-

duct of international engagement, and progress toward democracy

and peace. The plentitude of cases and efforts intended to promote

both democratization and peacebuilding gives rise to the need for a

3 The phenomenon of war-to-democracy transition takes place in societiesshattered by violent conflict. Such societies are often referred to as ‘‘post-conflict’’cases. However, the term ‘‘post-conflict’’ invites the interpretation that (1) therehas been a violent conflict and (2) the conflict is now solved and violence hasceased. This term is actually a misnomer. By convention, ‘‘post-conflict’’ usuallydenotes societies affected by armed conflicts, where only parts of the conflicts aresolved and where some organized violent behavior still lingers on. Moreover, it isproblematic to use the term ‘‘post-conflict’’ when, in fact, non-violent conflicts arepart of all societies – also so-called post-conflict societies. Thus, the term‘‘conflict’’ fails to distinguish those societies where conflicts are settled by peacefulmeans from societies where violent relations prevail. The term ‘‘post-war’’ is usedhere, for want of a better one, to refer to situations where the major warfare hasceased, but where some incompatible issues may remain unsolved.

20 Anna K. Jarstad

comprehensive understanding of post-war transitions where democra-

tization and peacebuilding are treated as separate phenomena, which

might or might not reinforce each other. The core question is how a

war-to-democracy transition can be achieved peacefully. The chal-

lenges of democratization in post-war societies can be studied by

simultaneously looking at the peace process and the democratization

process: how do these processes develop? when do they reinforce each

other? and when and why do they clash?

The two intertwined processes, from violent conflict to peace on the

one hand, and from authoritarian rule to democracy on the other hand,

are here labeled a war-to-democracy transition. The two processes are

treated as separate, but related, and the dynamic within and between

these processes is in focus. This means that while in practice democra-

tization and peacebuilding often overlap, they are two analytically

different processes. The perception of democratization and peacebuild-

ing as two logically separate processes facilitates our understanding of

the conditions under which efforts to promote democracy and peace

clash. This conceptual framework can be used to analyze the effects of

such dilemmas for war-to-democracy transitions and enhance our

understanding of how such transformations can be facilitated.

Dilemmas: horizontal, vertical, systemic, and temporal

Four types of dilemmas may arise when the processes of democratiza-

tion and peacebuilding have adverse effects on each other: the horizon-

tal, the vertical, the systemic and the temporal.4

Firstly, the horizontal dilemma concerns the issue of which groups

should be represented in the processes of peace and democratization.

This decision regards the horizontal relation between the elites of war-

ring parties and of democratic political parties. A selected group of elites

may more easily commit to difficult compromises, while comprehensive

peace negotiations may result in more lasting agreements by involving

4 The labeling of these four types of dilemmas is my own but builds on previousresearch. Such research identifies obstacles and dilemmas related todemocratization in war-torn societies, specifically to what is here labeled thetemporal dilemma (see, e.g., Cousens 2001a; de Zeeuw 2005); the systemicdilemma (see, e.g., Burnell 2005; Chandler 1999; Chandler 2004; Knaus andMartin 2003); the horizontal dilemma (see, e.g., Stedman 1997), and what is herelabeled the vertical dilemma (see, e.g., Cousens 2001a; Holsti 1996; Paris 2004).

Dilemmas of war-to-democracy transitions 21

all parties with a stake in post-war developments. Broad inclusion is

also in line with democratic theory on power sharing, which suggests

that the more groups represented in the process, the more democratic it

is. Some groups also have legitimate reasons to demand political power

after years of oppression and discrimination. Research furthermore

suggests that warring parties are more likely to sign a peace deal if

they are guaranteed a share in the future government (Lijphart 1977;

Walter 2002). However, when broad inclusion is extended to violent

parties, it may have negative effects on democratization. Such inclusion

can be seen as a reward for violence and thereby contradict the demo-

cratic principle of non-violence. This is particularly true when inclu-

sion implies amnesty for persons who have committed human rights

violations during the war. Thus, the horizontal dilemma involves a

trade-off between inclusion (e.g., for the sake of reaching a peace deal

or broad representation) and exclusion (e.g., for the sake of reaching a

compromise solution and perhaps also for long-term democratization).

This dilemma also affects the prospects for peace. A rebel group that

expects to be excluded from future governments and control over part of

the territory may find peace too costly. For this reason, a peace deal often

stipulates inclusion of the main warring parties in the political process.

Peace agreements providing for guaranteed positions in government

have been reached in cases such as Burundi 2000, Cambodia 1991,

and the Democratic Republic of the Congo 2002. However, it may be

difficult to identify which groups need to be included for the sake of

peace. If the peace agreement grants some warring parties seats in the

government, this may provide an incentive for other groups to use

violence to gain the same political status. For example, the power-sharing

agreement for Sudan 2004 did not include the rebels in Darfur and

fighting continues. Not even the May 2006 agreement included all rebels

and these groups demand concessions before laying down their arms.

Moreover, inclusion may not end violence: some groups pursue a

dual strategy of violence and politics. For democracy to take root,

actors mobilized for war have to abandon military methods for nego-

tiations and compromises. These leaders also have to convince their

followers that they should demobilize and be prepared for concessions.

Ideally, they should also abandon excluding ideologies, such as ethno-

nationalism, and strive for broad-based democratic support. But also

parties commonly labeled as terrorist organizations, such as Hamas,

can gain democratic legitimacy via parliamentary elections. Although

22 Anna K. Jarstad

the hope is that this will lead to a change in Hamas’ politics, democratic

institutions do not always produce peaceful democrats.

Exclusion of potential spoilers is an alternative strategy for promoting

peaceful democratization. It rests on the notion of excluding nationalists

and authoritarian actors for the sake of only allowing democratic move-

ments to develop into political parties and compete for power. This

strategy more clearly opens up for new actors. However, research has

demonstrated that excluded groups to a greater extent return to violent

tactics (Gurr 2000a; Stedman 1997).

Civil society is often excluded from power-sharing deals. The exclu-

sion of such segments of society leads to an uneven start for parties in a

democratization process. One possibility to overcome this negative

effect of power sharing is to include a broad range of actors in the

peace negotiations and also in the future government. This was done,

for example, in the Democratic Republic of the Congo, where the 2002

peace accord included a wide spectrum of society. Other cases include

the Ivory Coast 2003 peace deal, which in addition to the warring

parties included the main political parties, and the Liberia 2003 agree-

ment, which included not only all warring parties to the conflict, but

also representatives from the civil society (Nilsson 2006: 110; Nilsson

and Soderberg Kovacs 2005).

Secondly, the vertical dilemma entails the difficult choice between

efficacy and legitimacy. It pertains to the relation between elite and

mass politics. On the one hand, legitimacy is expected to increase when

the people are involved in all phases of the peace process, and also have a

chance of influencing the crafting of a new constitution. On the other

hand, the elites often have an interest in a non-public process. They want

to signal resolve – that they are not prepared to make concessions – in

order to get the best deal they can at the negotiation table. After a peace

agreement, however, elites are expected to be conciliatory toward

former foes and also urge their followers to demobilize and accept

concessions. At the same time, some elites use the demands of extremist

groups to push for additional concessions from the other parties to the

conflict also after a peace deal.

An alternative approach, or addition to broad inclusion during peace

negotiations, is to strive for public support after a peace deal or demo-

cratic constitution is drafted. One way is to hold a referendum to try to

ensure such legitimacy. The constitution of Iraq was approved despite

the vast Sunni boycott of the October 2005 referendum. However, if

Dilemmas of war-to-democracy transitions 23

the new constitution or peace deal does not receive sufficient popular

support, the whole process is delayed. This was the case, when the

Greek Cypriot majority voted against United Nations Secretary-

General Kofi Annan’s plan to unite Cyprus in April 2004. The majority

rule typically used in referenda may in divided societies exacerbate

polarization. Under such conditions, it might be necessary for new

leaders to emerge for negotiations to restart.

Thirdly, the systemic dilemma refers to the issue of ownership, that is

of international versus local control of the processes of democratiza-

tion and peacebuilding. Third-party engagement in peacebuilding

might generate a dilemma of peace versus democracy. On the one

hand, international involvement may be necessary to end violence

and to facilitate negotiations. Support for democratic developments

aims to promote stability and institutions for conflict management. But

on the other hand, both sustainable peacebuilding and democratization

depend on the commitment of local people and elites. Contemporary

peacebuilding sometimes includes temporary external control over

political processes. These structures are not formally accountable to

the citizens in these states. In such cases local ownership is weak, thus

risking to halt or reverse the process and even to alienate people from

democracy as an ideal. For example, when the Office of the High

Representative (OHR) stepped in to dismiss elected officials in Bosnia

and Herzegovina, some people lost trust in the legitimacy of the system.

Such international intervention risks increasing support for extremism

and ethnochauvinism. At the same time, implementation of peace

agreements includes politically risky steps that local politicians cannot

always take, for fear of alienating their own constituencies.

But in a post-war context, moderate political parties tend to

be lacking or marginalized. For this reason, international support can

be pivotal to strengthen the capacity of local moderate groups and

thereby facilitate democratization. Ideally, democracy promotion

entails support to a locally driven democratization process. It is often

the case that at least some local actors demand democratization. Today

it would be difficult to withhold from people the right to choose a

government – even in cases with unclear status of the state, such as

Kosovo and Palestine. At the same time, international actors are reluc-

tant to provide funds without conditions. Assistance is often ear-

marked for specific purposes, which do not always correspond to the

most pressing local needs. Also, international engagements are often

24 Anna K. Jarstad

short-term, thereby prompting donors to try to speed up the process of

democratization.

Fourthly, the temporal dilemma regards trade-offs concerning short-

term versus long-term effects on democratization and peacebuilding.

Efforts to support democratization may in the short run increase

the risk of violence, and thereby in the long run undermine the chances

for democracy to take root. Likewise, peacebuilding may involve

restrictions on democratic freedoms such as freedom of press and

mass demonstrations. In the long run, such constraints may cause

unrest and turn into an obstacle for the implementation of the peace

agreement.

The timing of elections also activates the temporal dilemma.

Democracy means rule by the people, and it is difficult to think of

another way to ensure democratic legitimacy than through elections.

After a war, a democratic election also serves the purpose of bringing a

decisive end to the war and of sealing the peace deal. Consequently,

elections have come to be seen as the crowning event of the peace-

building phase and an ‘exit strategy’ for organizations engaged in

international peace missions.

However, the first post-war election is often riddled with violence

and flawed election outcomes. One reason is that some actors expect to

lose political power or control over valuable resources as a conse-

quence of the peace and democratization processes. By threats and

intimidation, these actors may seek to disrupt the transition, overthrow

the election results, or prevent election campaigns or voters from going

to the polls. In the worst cases, elections trigger violent conflict and the

process of democratization is halted or reversed. For this reason, it has

been suggested that elections should be postponed until the conditions

are stable and democratic institutions are in place (Cousens 2001b;

Mansfield and Snyder 2002a; Paris 2004). The sequencing of peace-

building has caused dilemmas for Kosovo, for example, where demo-

cratic institutions were introduced before its international status was

settled, and before reconciliation between the warring groups and

democratic norms have taken root. It remains to be seen whether this

method to promote democratization will contribute to a successful

war-to-democracy transition.

Not dealing with these four types of dilemmas can have devastating

effects. The Freedom House ranking is a crude, and some would even

say deceptive, indicator of democracy, but nevertheless gives us an idea

Dilemmas of war-to-democracy transitions 25

about which countries are far from democratic, or ‘‘not free’’ as it is

termed by this organization. A quick look at the Freedom House

ranking 2005 for the fourteen conflict locations where major peace-

building missions were launched between 1989 and 1999 (Paris 2004)

shows that very few have achieved what Freedom House calls ‘‘free-

dom.’’ Only Namibia, El Salvador, and Croatia are ranked ‘‘Free.’’

Angola, Cambodia, Kosovo, and Rwanda are ranked ‘‘Not Free,’’

and the rest ‘‘Partly Free.’’5

The lessons learned from missions aiming for democratization and

peacebuilding in, for example, Liberia and Haiti demonstrate that

when there is a choice between promoting democracy and peace,

securing the peace is pivotal. To understand why dilemmas of war-to-

democracy transitions occur, and why they can have such devastating

effects, it is suggested that theoretical explanations can be found in

three areas of research: the efforts to promote peacebuilding, the con-

flictual character of democratization processes, and the legacy of war.

Potentially negative effects of peacebuilding

The ending of war does not always mean the end of violence. In fact,

peacebuilding can in extreme cases make things worse. Recently, the

notion that negotiated settlements are the best way to end civil wars

have been challenged. According to Monica Toft, civil wars that end in

a military victory, in particular those that end in a rebel victory, are

associated with higher levels of democracy in the longer perspective

than are wars that end in a cease-fire or peace agreement (Toft 2003).6

However, contemporary conflicts are increasingly ended by negotiated

deals – rather than on the battlefield – and it is therefore important to

analyze the obstacles involved in the implementation of peace settle-

ments. Several aspects of peacebuilding may give rise to dilemmas

between efforts to promote peace and support to democratization.

This is the case when peacebuilding includes multiple tasks with

5 The ‘Partly Free’ 2005 cases that are included also in Roland Paris’ analysis ofmajor peace missions 1989–1999 are Nicaragua, Mozambique, Liberia, Bosniaand Herzegovina, Guatemala, East Timor, Sierra Leone (Freedom House 2006;Paris 2004). For definitions and coding according to Freedom House, refer towww.freedomhouse.org/template.cfm?page=15year=2005.

6 See more on this issue in Fortna’s chapter in this volume.

26 Anna K. Jarstad

competing objectives, when the coordination between a multitude of

actors falter, when project-oriented and short-termed missions fail to

mitigate negative long-term effects, and when international engage-

ment makes the host society dependent on external support. For peace

to become viable, it is not only the conditions that generated the

conflict that need to be addressed. In addition, peacebuilding should

be designed in such a way that the above-mentioned dysfunctions can

be avoided. Furthermore, it is suggested that peacebuilding needs to

focus on security and the designing of self-sustaining institutions.

Otherwise there is a risk that peacebuilding gives rise to new conflicts

or that it undermines democratization.

Since the 1990s, democratization has become an integral part of the

conflict-prevention agenda. Multidimensional peace operations have

become the model for contemporary peace promotion.7 Such missions

seek not only to prevent violence, but also to address the root causes of

conflict. In An Agenda for Peace (1992), former UN Secretary General

Boutros Boutros-Ghali suggested that post-war peacebuilding was a

form of conflict prevention via social and political reconstruction. In

1996, the Agenda for Peace was supplemented by the Agenda for Demo-

cratization, and in the Framework for Cooperation in Peacebuilding

(2001) UN Secretary-General Kofi Annan further developed the notion

that democratization is part of peacebuilding (United Nations 2001:

Annex 1:1). In December 2005, the UN General Assembly and the

Security Council jointly agreed to set up a Peacebuilding Commission

to help countries emerging from conflict manage the transition to stabi-

lity and development.

War is costly, but much of the costs of war occur after it is over.

Economic reconstruction is often vital for sustainable peace, but peace-

building missions are often insufficiently funded (Collier et al. 2003;

Paris 2004; Woodward 2002). Efforts to ‘‘demilitarize politics’’ are

also expensive parts of peacebuilding. This includes disarmament

and demobilization of previously warring parties, destruction of

weapons, reformation of the security sector, and issues related to

7 The first generation peacekeeping was mainly a military exercise aimed atupholding cease-fires through the method of separating warring parties andthereby providing opportunities for negotiations. The second generationpeacekeeping, also labeled multidimensional peace operations, includes bothmilitary and civilian components.

Dilemmas of war-to-democracy transitions 27

democratization, such as the transformation of rebel groups into poli-

tical parties (Lyons 2005; Spear 2002).

Democratization in war-torn societies also includes support for con-

stitutional and legal reforms, the establishment of election administra-

tion, training of election staff and media professionals, political party

assistance, international election and human rights monitoring and

civil society aid. Experiences from countries such as Cambodia,

Ethiopia, Uganda, and Guatemala demonstrate that peacebuilding

does not always move societies toward democracy. In these cases,

malfunction of institutions is often seen as a key problem: these

institutions are largely non-transparent and unaccountable to members

or society in general, politically biased, and financially unsustainable.

The impact of international engagements may be obstructed by, for

instance, interagency rivalries or donors’ ambition to demonstrate

short-term results. International support can also foster a ‘‘culture of

dependence.’’ This is especially problematic for support for peacebuild-

ing and democratization, as these processes need to be based on local

needs and driven by the people in the recipient society (Chandler 2004;

de Zeeuw 2005; Diamond 1999b; Paris 2004).

The conflictual elements of democratization

Successful democratization requires a minimum level of security and

consensus on which territory and people constitute the state: without a

defined demos, how can you have democracy? While this stance was

advocated already thirty years ago (Rustow 1970), contemporary

democratization is nevertheless promoted where these conditions are

lacking.8 It is suggested that this old recommendation needs to be taken

seriously. However, the necessity of ensuring that these preconditions

are in place before embarking on democratization seemed to be for-

gotten as more and more countries began to hold regular elections.

Research commonly assumed that democratic transitions developed

gradually, from political liberalization toward a consolidated democ-

racy. Although some violence occurred in earlier democratization

waves, it was not seen as a serious threat to peace. On the contrary,

8 This has been noted also by, for example, Juan Linz and Alfred Stepan (1992) andMansfield and Snyder (2005).

28 Anna K. Jarstad

the notion of ‘‘democratic peace’’ – that democracies virtually never go

to war with each other – prompted both researchers and policymakers

to conclude that the expansion of the democratic zone would reduce

the risk of armed conflicts (e.g., Dahl 1971; Diamond 1997a; Diamond

et al. 1997; Huntington 1991; Linz and Stepan 1996; O’Donnell,

Schmitter, and Whitehead 1986).

At the same time, research has shown that democracy as well as

democratization both contain conflictual elements (e.g., Hegre et al.

2001; MacMillan 2003; Mann 2005; Mansfield and Snyder 1995a

and b; Mansfield and Snyder 2002b; Mansfield and Snyder 2005;

Maoz and Russett 1993; Oneal, Russett, and Berbaum 2003; Rosato

2003; Thompson and Richard 1997). Not only may the move toward

democracy fail; in addition, democratization can exacerbate violent

conflict. It is proposed that two aspects of democratization can activate

dilemmas in war-torn societies: the essence of the stipulated goal of

democratization, namely democracy itself; and the shifts involved in

the democratic transitions. Firstly, democracy by definition includes

conflictual elements that can have severe effects in societies polarized

by violence. In line with Robert Dahl’s conception of polyarchy, key

elements of democracy include contestation (including elections), par-

ticipation, and basic human rights (Dahl 1971). Although democracy

ideally stipulates conflict regulation through norms and institutions, it

also induces conflict via increased contestation and polarization. Public

contestation provides opportunities for replacement of elected offi-

cials. As the political candidates compete for votes, they emphasize

their differences rather than their common ground. Likewise, during an

election campaign the constituencies are mobilized, thus oftentimes

enhancing already high levels of polarization. Hence, democracy by

definition implies opposition and mobilization along distinctive lines

and a certain degree of polarization. This means that democracy pro-

vides both opportunities and incentives for conflict. Whereas conflicts

are most often managed peacefully in consolidated democracies, this

feature can have severe effects in a typical post-war society where

weapons abound and people remain polarized.

Secondly, the process of democratization entails particular features

which increase the risk of violence (Gleditsch and Hegre 1997;

Mansfield and Snyder 1995a and b). The movement toward democracy

entails changes and shifts of power. This increases the risk of use of

violence by those who lose or fear losing power and by those that feel

Dilemmas of war-to-democracy transitions 29

that they should gain more. In accordance with the definition of democ-

racy above, democratization refers here to improvements in contestation

(such as more political parties and candidates), participation (broad-

ening of the electorate) and human rights (for example, freedom of

speech and freedom to organize demonstrations).

An emerging field of research specifies the conditions and types of

political transformations highly correlated with violent conflict.9 One

factor that contributes to the high levels of violence during the begin-

ning of a democratization process is that public expectations tend to be

high. People often demand rapid and real improvement in the quality of

life. But democracy does not automatically result in other desirable

goals such as economic development and equality. Many of the obsta-

cles embedded in the democratization process also relate to the fact that

the different democratization components cannot be achieved all at

once. Initially there are typically great discrepancies between public

demands and the institutional capacity to deliver, as well as between

public loyalty to the state and the state’s capacity to control undemo-

cratic elements and make legitimate political decisions. When reforms

do not work in concert, risk of violence increases.

Given the conflictual elements of democratization, basic consensus –

concerning the legitimacy of the state, its territory and its citizens – is

necessary to prevent democratization from turning violent. However, it

is often disagreement on these specific issues which caused violent

conflict to erupt in the first place. Unfortunately, democracy does not

solve the issue of what constitutes the state; which territory should

belong to it and which people should be citizens of the state. Even after

a peace deal many actors continue to challenge the state. Also when

former warring parties become part of the government, some continue

to use violent tactics with the ambition to change the character of the

9 Some findings suggest that states that are becoming more democratic reduce therisk of interstate war by half. However, especially rocky and rapid transitions orreversals are associated with an increased risk of war (Ward and Gleditsch 1998).Other scholars suggest that because each move toward democracy orauthoritarianism entails a risk of violent conflict, and because democracy is themost stable regime type, rapid democratization is less risky than a gradualprocess. Furthermore, not only transitions, but also certain phases are associatedwith conflict. The initial phase of democratization as well as consolidated semi-democracies – cases where no significant political change has occurred for sometime – are more war-prone than consolidated democracies and autocracies (Hegreet al. 2001).

30 Anna K. Jarstad

state. Even in instances where the elites have decided to disarm, people

may remain polarized. This means that war-shattered societies are

particularly vulnerable to the risks of democratization.

The legacy of violent conflict

After a violent intrastate conflict, conditions conducive to democrati-

zation are typically absent and the legacies of conflict tend to linger.10

Arms are widely available and often used, even after a peace deal has

been signed. Owing to psychological trauma of violence and fear of

renewed violence, mass mobilization along extremist lines remains.

Political ideologies are based on exclusive group-based interest rather

than on universal, society-wide interests. Threats and violence prevent

political candidates from running for office and hinder voters from

going to the polling stations. Political trust is low, which hinders

cooperation across subcultures. Whereas many civil society organiza-

tions play a pivotal role in humanitarian assistance and reconstruction

after war, there are examples where voluntary organizations foment

intergroup violence. This was the case during the civil war in Lebanon

1975–1989. In the worst case, such activities can even contribute to

genocidal violence. In Rwanda, radio broadcasting was used to insti-

gate the genocide of Tutsis in 1994.

During and after ethnic wars, people tend to seek protection in areas

where the majority population belong to their own ethnic group and

thus become displaced within their own country (Posen 1993). In such

contexts, proponents of a moderate ideology face a high risk of becom-

ing targets both of extremist violence by people belonging to the same

ethnic kin, and members of other ethnic groups who portray all non-

members as enemies. To escape violence, moderates often have to leave

for other states and become refugees.

Competition for votes based on increasingly extremist rhetoric,

so-called politics of outbidding, can also enhance polarization and foster

ethnic tension (Rabushka and Shepsle 1972: 187). Such outbidding has

taken place in, for example, post-war Bosnia and Herzegovina, where

nationalist parties have cemented their early grip of power in successive

10 For a discussion on conditions conducive to democracy and democratization,see, for example, Robert Dahl (1971) and Samuel Huntington (1991: 37–38).

Dilemmas of war-to-democracy transitions 31

elections. Additionally, in many post-war societies the political party

organization is absent, weak, or fragmented. The 2005 Afghan parlia-

mentary elections demonstrate this point. The over 5,800 candidates

that ran for office were formally part of different political parties, but

personalistic attributes and clan politics largely substituted the role of

ideologies.11

In addition to the initially contested issues, such as control over

government or control of territory, new issues emerge during the con-

flict. At the same time, there are often fewer resources to share or divide

after a conflict. Typically, the economy is weak, the level of unemploy-

ment high, the infrastructure shattered, and natural resources

destroyed or inaccessible, for example due to land mines. Refugees

and internally displaced persons are often prevented from returning

to their pre-war homes because of new occupants or destroyed houses,

insecurity, and lack of economic resources. Thus, an agreement seldom

means that a conflict is resolved. Conflicting attitudes, behavior,

and issues remain to be transformed after the fighting has stopped

(Lederach 1997).

Simultaneous peacebuilding and democratization

The cases in focus are war-torn societies that undergo simultaneous

peacebuilding and democratization. For our purposes, it is important

to include both cases where a war-to-democracy transition has taken

place and cases where such transition is impeded by difficult dilemmas.

To recall, peacebuilding is here seen as the implementation of a peace

agreement. This means that the pertinent cases for our analyses are

post-settlement cases.12 Democratization refers here to improvements

in contestation, participation, and human rights. This minimal defini-

tion allows us to analyze also cases where only minor moves toward

democracy have occurred.

To give a snapshot picture of pertinent cases, the list below (see

Table 1.1) indicates forty-two conflict locations where peace

11 See Reilly’s chapter in this volume.12 In addition, some of the chapters make reference also to cases where no peace

agreement has been reached, in order to illustrate and compare processes thatincrease our understanding of democratization and peacebuilding in war-tornsocieties.

32 Anna K. Jarstad

Table 1.1 Recent cases of peacebuilding after civil war

Conflict location Name and year of latest peace agreement or process

Afghanistan Post-Bonn process (2001)

Angola Lusaka process (1994–2002)

Bangladesh (Chittagong

Hill)

Chittagong Hill Tracts Peace Accord (1997)

Bosnia and Herzegovina Dayton Peace Agreement (1995)

Burundi Arusha process (2000–2003)

Cambodia Paris Agreement (1991)

Chad Various peace and reconciliation processes

(1989–2006)

Colombia EPL-government peace process (1991) and

Common Agenda process (1999–2002)

Comoros (Anjouan) Federalization process (2000–2003)

Congo National Dialogue process (1999–2001)

Croatia Erdut Agreement (1995)

DRC Inter-Congolese Dialogue process (1999–2003)

Djibouti Accord de reforme et concorde civile 2001

El Salvador Geneva process (1990–1992)

Georgia (Abkhazia) Declaration on measures for a political

settlement (1994)

Guatemala Esquipulas/Oslo/UN processes (1987–1996)

Guinea-Bissau Abuja Peace Agreement (1998)

Haiti MINUSTAH peacebuilding mission (from 2004)

India (Bodoland) Bodoland Autonomous Council Act (1993)

India (Tripura) Memorandum of Settlement (1993)

Indonesia (Aceh) Memorandum of understanding, Indonesia and

GAM (2005)

Israel (Palestine) Oslo process (1993–2000)

Ivory Coast Accra process (2003–2006)

Liberia Accra Comprehensive Peace Agreement (2003)

Macedonia Ohrid Agreement (2001)

Mali (Azawad) Pacte National (1992)

Mexico San Andres Accords (1996)

Moldova (Dniestr) Memorandum on the Basis for Normalization of

Relations (1997)

Mozambique Rome process (1990–1992)

Nepal Comprehensive Peace Agreement (2006)

Niger (Air and

Azawad)

Agreement on a Definite Peace (1995)

Dilemmas of war-to-democracy transitions 33

agreements have been signed during the post-Cold War era.13 The

conflict location for civil armed conflicts, and the latest peace accord

for each conflict, is listed above. In most of these cases, there have also

been improvements in contestation and human rights after the conflict

Table 1.1 (cont.)

Conflict location Name and year of latest peace agreement or process

Papua New Guinea

(Bougainville)

Bougainville Peace Agreement (2001)

Philippines (Mindanao) Mindanao Final Agreement (1996)

Philippines General Agreement for Peace (1995)

Rwanda Arusha process (1991–1993)

Senegal (Casamance) General Accord between Senegal and MFDC (2004)

Sierra Leone Abuja Ceasefire Agreement (2000)

Somalia Cairo Declaration on Somalia (1997)

Sudan (Southern Sudan) Machakos process (2002–2005)

Sudan (Darfur) Darfur Peace Agreement (2006)

Tajikistan Moscow Declaration (1997)

Uganda Yumbe Peace Agreement (2002)

UK (Northern Ireland) Good Friday Agreement (1998)

Yugoslavia (Kosovo) Rambouillet Agreement (1999)

Yugoslavia (Slovenia) Brioni Agreement (1991)

Source: Uppsala Conflict Data Program, March 2007. Selection made with

assistance of Ralph Sundberg. The name of the contested territory is indicated in

parentheses after each conflict location. Conflicts without parentheses concern

contest over government.

13 The processes of democratization and peacebuilding are ‘‘moving targets’’ andthere is no existing database that captures these phenomena. The list includedhere is based on the Uppsala Conflict Data Program (UCDP). This database hasthe benefits of including all armed conflicts with at least twenty-five battle-related deaths and is a systematic and continuous collection endeavor. However,it does not capture all peace processes. The cases in the list are coded by UCDP asconflict locations with peace agreements. In addition, they fulfill the followingcriteria: they are intrastate peace agreements (interstate peace agreements areexcluded) and the peace agreement was signed during the period 1989–2006.Cases not included in the list, according to the criteria above, include, forexample, Iraq 2003 (not coded by UCDP as a peace agreement) and South Africa1994 (not coded as an armed conflict after 1989), East Timor/Indonesia,Ethiopia, Lebanon, Nicaragua, Namibia, and Cyprus. Although these cases do

34 Anna K. Jarstad

was formally ended. Where this is the case, a war-to-democracy transi-

tion can be considered under way.

A brief look at the list gives us an idea of the many war-torn societies

where challenges face the people that live in these locations, and the

international community organizations which aim to support the pro-

cesses of democratization and peacebuilding. There is no easy way to

achieve peace and democracy simultaneously after civil war. Not only

are the conditions for the initiation of a democratization process unfa-

vorable after intrastate conflict, in addition, the opening up of political

space aggravates these conditions. The two processes of democratiza-

tion and peacebuilding may clash, thus leading to negative effects on

each other. Such potential quandaries reflect the inherent conflictual

nature of democracy and democratization as well as the difficult pro-

cess toward peace.

Appreciation of the particular dilemmas that arise in each post-war

situation is necessary for the design of proper means to advance

synchronized democratization and peacebuilding. Ideally, such analy-

sis makes it possible to avoid several of the dilemmas and to properly

design means to support peace and democracy simultaneously.

However, at particular points in time the inevitable choice arises

between promoting efforts to democracy or peace. Without a minimum

level of peace it is impossible to achieve free political contestation,

popular participation, and human rights. This is why this book sug-

gests that the sequencing, timing, and design of peace missions are vital

for international support for war-to-democracy transitions.

In the long run, the central issue is not choosing between peace or

democracy, but rather what steps toward peace and democracy should

be taken when, and how are they best timed, sequenced, and combined?

This book suggests that a minimal level of security is important before

elections take place. Violence needs to be reduced to permit elections, if

legitimate government is to result. In this way, securing a minimal level

of peace is a first necessary step for successful war-to-democracy transi-

tions. At the same time it is important not to postpone elections for too

long, as elections performs various functions after war. The book also

not fulfill the criteria above, some of the chapters in this book make reference tothese and other cases, as they prove valuable for understanding the broader issueof war-to-democracy transitions. For UCDP data and definitions, refer towww.pcr.uu.se/research/UCDP/index.htm.

Dilemmas of war-to-democracy transitions 35

outlines ways to promote democratization short of elections, by opening

up space for actors outside the formal peace process.

Whereas the process of democratization is very vulnerable to a

breach of peace, peacebuilding often progresses despite setbacks in

democratization. To recall, efforts to promote security is only one

element of peacebuilding. Successful war-to-democracy transitions

involve political, institutional, and social transformation to bring con-

ditions that enable lasting peace and democracy.

36 Anna K. Jarstad

PART II

The security context

2 Peacekeeping and democratization

V I R G I N I A P A G E F O R T N A

The international community has endeavored to maintain peace and

foster stable, democratic polities in a host of war-torn countries since

the end of the Cold War. A primary tool in this effort has been the

practice of international peacekeeping – the deployment of interna-

tional personnel, generally under the auspices of the United Nations or

regional organizations, to oversee the transition from war to peace.

Maintaining peace in the aftermath of civil war is difficult enough, but

the added task of shepherding a country toward democracy is fraught

with challenges and dilemmas. States recently torn asunder by warfare

do not provide particularly fertile ground for the growth of democracy.

Nor is peacekeeping, on the face of it, an especially powerful tool.

Peacekeeping missions are habitually under-funded, under-equipped,

and understaffed. They are cobbled together, often at the last minute,

with begged-for personnel (among whom there may be no common

language). They are often given unrealistically short mandates and

overly ambitious time-lines and benchmarks. And they are asked to

perform miracles, turning countries in which political institutions,

economic infrastructure, and the very fabric of society have all been

devastated by civil war into stable, functioning, democratic states.

This chapter assesses how well they do at accomplishing this vir-

tually impossible task. More specifically, what role does international

peacekeeping play in the democratization of post-war societies? It

shows that peacekeeping has neither a clear positive nor a clear nega-

tive effect on democratization. Rather, positive and negative effects

appear to cancel each other out, reflecting inherent dilemmas in the

attempt to foster both stable peace and democracy in the aftermath of

civil war.

Peacekeeping enhances stability, that is, it helps to prevent the recur-

rence of war. Peacekeepers might usefully be thought of as a modern-

day equivalent of the podesta, hired by warring clans in medieval

Genoa to administer the city and keep the peace, too weak to take

39

power themselves, but strong enough to tip the balance against either

side should it attack the other (Greif 1998). But as this collaborative

project on the ‘‘dilemmas of democratization’’ emphasizes, not all good

things necessarily go together. That which promotes stability may not

promote, and may in fact undermine, democracy.

Second only to stable peace, democratization is a core goal of the

international community when it undertakes peacekeeping missions.1

Since the mission in 1989 in Namibia, which marked a turning point in

peacekeeping with the end of the Cold War, most peacekeeping mis-

sions involve significant democratization components, including mon-

itoring or even running elections, civil rights monitoring and training,

fostering civil society and democratic political institutions. But the two

main goals of peacekeeping, maintaining peace and stability on the one

hand, and fostering democracy on the other, may often be in direct

conflict.

The effectiveness of peacekeeping on democratization is open to

debate. Doyle and Sambanis (2006), Heldt (2007), and Pickering

and Peceny (2006) find that intervention fosters democratization.

Similarly, Wantchekon (2004) argues that impartial peacekeepers pro-

vide one of the conditions for democracy to emerge from civil war.

However, Bueno de Mesquita and Downs (2006) find no positive effect

of intervention. Weinstein (2005) argues that outsiders’ attempts at

state- and democracy building can impede the development of strong

and democratic political and economic institutions, that in some cases

at least, post-war societies would be better off left to their own devices

in a process of ‘‘autonomous recovery.’’2 Similarly, Marten (2004: 155)

proposes that peacekeepers should limit their goals to providing stabi-

lity, and not try to transform societies. ‘‘The notion of imposing liberal

democracy abroad is a pipedream,’’ she argues.

The widespread use of peacekeeping in civil wars since the end of the

Cold War has made it possible to end the fighting with a truce or

1 Paris (2004) argues that all of the fourteen major peacebuilding missions between1989 and 1999 have included democratization, along with marketization, as astrategy for consolidating peace. See also Andersson (2000).

2 Wantchekon and Neeman (2002) propose a model by which democracy canemerge in war-torn states without the intervention of outsiders, though theyacknowledge that some peacekeeping functions, such as monitoring and trustbuilding, might encourage democratization – a notion that Wantchekon (2004)develops further.

40 Virginia Page Fortna

settlement in a number of cases that, were peacekeepers not available,

would likely have dragged on for some time until a clear victor even-

tually emerged. Peacekeeping has thus curtailed the violence but has

also led to less decisive outcomes (Fortna 2005). These less decisive

outcomes may also make democracy less likely to emerge.3 The ques-

tion of whether peacekeeping furthers democratization in war-torn

states or hinders it is thus an open one. This chapter aims to explore,

both theoretically and empirically, the relationship between interna-

tional peacekeeping and domestic processes of democratization in

post-war societies.

I define peacekeeping as the deployment of international personnel

to help maintain peace and security in the aftermath of war.4 All

peacekeeping missions involve military personnel, though they may

or may not be armed, and many missions include substantial civilian

components as well. This definition includes both operations based on

the traditional principles of peacekeeping, specifically the consent of

the belligerents themselves and the passive use of force, as well as peace

enforcement missions that relax these conditions considerably.

However, I also distinguish among types of missions, for the effects

of enforcement missions may differ from those of consent-based peace-

keeping. And perhaps most important for a study of peacekeeping and

democratization, the effects of multidimensional missions that include

large civilian components engaged in election monitoring, human

rights training and monitoring, police reform, institution building,

economic development, and so on may be quite different from the

effects of more traditional types of observer or interpositional peace-

keeping missions. Peacekeeping is thus broken into the following four

categories:

3 Toft (2003) argues that civil wars that end in a victory for one side, especiallythose that end in rebel victory, are followed by higher levels of democracy over thelong term than are wars that end in a truce or peace settlement. She suggests thatthis is because clear military winners are better able to consolidate politicalinstitutions. Her empirical analysis on this includes only a bivariate relationship,and may be picking up regression toward the mean in democracy scores after acivil war. (See below for further empirical investigation of the relationshipbetween war outcome and democratization.)

4 For a fuller discussion of the definition and types of peacekeeping, see Fortna(2008).

Peacekeeping and democratization 41

� Observation missions are small unarmed deployments of military

and sometimes civilian observers to monitor a cease-fire, the with-

drawal or cantonment of troops, or other terms of an agreement,

such as elections. Their main tasks are simply to watch and report on

what they see. Examples include the peacekeeping missions in

Angola in 1991 (UNAVEM II) and in the Western Sahara

(MINURSO), and the New Zealand and then Australian-led mis-

sions in Papua New Guinea in 1997–1998 (the Truce Monitoring

Group and Peace Monitoring Group, respectively).

� Interpositional missions (also known as traditional peacekeeping mis-

sions) are deployments of lightly armed troops. Like observer missions,

they monitor and report on compliance with an agreement, but they

also often serve to separate forces or to help demobilize and disarm

military factions. The UN missions in Angola in 1994 (UNAVEM III)

and in Guatemala in 1996 (MINUGUA) are examples.

� Multidimensional missions include both military and civilian compo-

nents helping to implement a comprehensive peace settlement. In

addition to the roles played by observer or interpositional missions,

they perform tasks such as organizing elections,5 human rights training

and monitoring, police reform, institution building, economic develop-

ment, and so on. Examples include the missions in El Salvador

(ONUSAL), Mozambique (ONUMOZ), and Namibia (UNTAG).

� Peace enforcement missions are mandated to use force for purposes

other than self-defense and involve substantial military contingents

to provide security and ensure compliance with a cease-fire. Some

enforcement missions are also multidimensional, consisting of large

civilian components as well as relatively robust military forces. The

West African and United Nations missions in Sierra Leone in 1999

(ECOMOG and UNAMSIL) and NATO missions in Bosnia (IFOR

and SFOR) fall in this category.

The first three categories together encompass what are often referred

to as Chapter VI peacekeeping, or consent-based missions, while the

fourth category is sometimes referred to as Chapter VII peacekeeping.

Chapter VI missions depend on the consent of the belligerents them-

selves and are mandated to use force only in self-defense, while

5 Note that I distinguish between organizing or running elections and electionobservation, which is a task often mandated to observational or interpositionalmissions.

42 Virginia Page Fortna

Chapter VII missions do not necessarily require consent and, as noted

above, rely more centrally on the use (or potential to use) force.

This breakdown of mission types does not distinguish missions

that temporarily take over the administration of the country, as in

Cambodia, Kosovo, or East Timor, from other types of peacekeeping.

These transitional administration missions may have rather different

effects than missions that oversee national administration of the state

only during the transition to peace. There are too few such missions

(only these three) in the data examined here to evaluate their separate

effects (they are included in the quantitative analysis as multidimen-

sional enforcement missions, as appropriate).6 I do, however, include a

brief discussion of democratization in these three cases below.

This study includes both peacekeeping undertaken by the United

Nations (the majority of peacekeeping missions) and by regional orga-

nizations or ad hoc groups of states, although I draw distinctions

between UN and non-UN missions in some of the analyses below.7

Of course, to evaluate the effects of peacekeeping on democratiza-

tion, we must compare post-war societies that received peacekeeping

with those that did not, where belligerents were left to their own

devices after the war. And because peacekeepers are not sent to

6 Another case of transitional administration, though not of an entire country,is the UN’s mission in Eastern Slavonia (UNTAES), a region of Croatia. Thepeacekeeping missions in Bosnia, by various parties including the UN, NATO, theEU, and the OSCE, had some aspects of a transitional administration but are notgenerally included in that category. For a qualitative assessment of the three casesmentioned above, as well as others that took over state administration to a lesserdegree, see Chesterman (2004). Transitional administrations may be becomingmore common over time, such that a quantitative assessment of their effects willbecome possible.

7 In some cases, UN and non-UN missions work simultaneously or sequentially.For example, there have been both NATO and UN peacekeepers in Bosnia, US-ledand UN peacekeepers in Haiti, and Economic Community of West African States(ECOWAS) and UN peacekeepers in Sierra Leone and Liberia. There are notenough of these ‘‘hybrid’’ cases to evaluate their effects separately, though theymay be becoming more frequent. I also do not distinguish ‘‘Integrated Missions,’’a term that has recently been used to refer to the need for integration amongvarious bodies within the UN that provide peacekeeping, peacebuilding,humanitarian assistance, and development aid, as well as integration between theUN and other peacekeeping organizations. As a recent report on the topic puts it,‘‘There is no adequate definition of an integrated mission. Nor is there an exampleof an integrated mission that serves as a model for what an integrated missionshould be’’ (Eide et al. 2005, 9). I therefore code missions according to the moreaccepted typology described here.

Peacekeeping and democratization 43

conflicts at random, we also need to take into account other factors

that affect both where peacekeepers go and democratization, including

a country’s previous experience (if any) of democracy, economic con-

ditions, and so on.8

Two related dilemmas

Two fundamental dilemmas of security and democracy face would-be

peacekeepers and the societies they aim to help. The first is that outside

assistance may be needed to achieve stable peace, but this external

assistance may thwart domestic processes of democratization. The

second emerges from the paradoxical relationship between war and

democracy – democracy requires peace and stability, but war and the

pressures of war-making on the state can provide incentives for

democratization.

Internal vs. external control and legitimacy

As Jarstad notes in her introduction to this book, the emergence of

democracy requires low levels of violence, and low levels of mistrust,

conditions that are unlikely in the immediate aftermath of civil war.

International peacekeepers can help reduce violence – both the threat

of a full-scale return to war, and lower levels of day-to-day violence.

International monitors can also increase levels of trust between former

belligerents and competing political forces. In these ways, peacekeepers

can help foster the conditions for successful democratization.

Peacekeepers can also foster the initial establishment of democratic

institutions by helping to set up and monitor elections. On the other

hand, by taking a large measure of control over political processes in

the transition from war to peace, outside peacekeepers can undermine

the principles of accountability and domestic legitimacy that are bed-

rocks of democratization.

In the short term, democracy has no chance of taking root if a society

is plagued by high levels of political violence, or worse yet, if full-

scale war resumes. Democracy requires peace and stability, and war-

torn societies often require outside help in maintaining peace. War is

8 For systematic studies of where peacekeepers tend to deploy, see Gilligan andStedman (2003); and Fortna (2008, Chapters 2 and 3).

44 Virginia Page Fortna

much more likely to resume, all else equal, if belligerents are left to

their own devices (Doyle and Sambanis 2000; Fortna 2004, 2008).

However, the very thing that can help ensure lasting peace, outside

intervention, often reduces the political space available for the emer-

gence of home-grown, domestically legitimate and accountable politi-

cal institutions. As Chesterman (2004: 1) argues, there are fundamental

tensions and contradictions involved in attempting ‘‘to establish the

conditions for legitimate and sustainable national governance through

a period of benevolent foreign autocracy.’’ Peacekeeping may thus

help foster conditions for the initial emergence of democracy, but at

the same time undermine conditions for it to thrive over the longer

term.

The arrival of foreign troops as peacekeepers by definition under-

mines state sovereignty. It also introduces a temporary and potentially

powerful actor into the domestic political scene, one whose presence

affects the incentives and behavior of indigenous political, economic,

and military actors. Particularly with the advent of large multidimen-

sional peacekeeping missions, these outsiders play a large role in the

transitional administration of the state (at the extreme, taking over

state administration entirely, as in Kosovo or East Timor). Ideally,

when the peacekeepers depart, they leave behind them at least nascent

democratic institutions. Peacekeeping missions can help hold the gov-

ernment to promises of political reform and can assist rebels in the

transition from military organizations to political ones. They can pro-

vide security during the tense periods of election campaigning and

balloting. And they can help reform armies and police forces and

provide human rights training, protecting political rights that are fun-

damental in a functioning democracy. In other words, they can help

start a war-torn society on the road to democracy.

But the large footprint of peacekeeping missions may trample and

crowd out local democratization efforts as much as empower them.

Peacekeeping creates a precedent of control by actors who are not

domestically accountable, or perhaps accountable at all.9 They can

create an artificial political structure that collapses when they leave,

and they can create a situation where the state’s reliance on

9 Problems of peacekeepers’ accountability (or lack thereof) have been mostnoticeable in the area of prostitution and sexual abuse (Mendelson 2005).

Peacekeeping and democratization 45

international assistance interferes with its accountability to local popu-

lations for support and funds (Weinstein 2005: 27).

In short, while peacekeepers can help provide security and incentives

for initial steps toward democracy, the presence of foreign actors can

undermine the long-run development of domestically legitimate demo-

cratic institutions. In the terminology used elsewhere in this volume,

this is a systemic dilemma which can cause a temporal dilemma.

Democracy through war or peace?

A more fundamental dilemma concerns the relationship between

democracy on the one hand, and war or peace, on the other. For

democracy to emerge there must be some semblance of peace and

security. Political violence and the atrocities of civil war are fundamen-

tally antithetical to the norms of cooperation, non-violent resolution of

political conflict, and basic trust inherent in a functioning democratic

system. There is good reason for the conventional wisdom that democ-

racy cannot be established without peace.

On the other hand, there is reason to suspect that war itself can create

the conditions for democratization. It was the need to raise funds and

manpower for war making that led to political reform, and therefore

democratization, in the process of European state making (Tilly 1985,

1990). States and subjects traded political rights for the taxes and

soldiers required for war. Weinstein argues that this same logic applies

to contemporary civil wars – that ‘‘war itself generates strong incentives

for rulers to secure the consent of governed and build representative

institutions.’’ This creates a policy dilemma between stopping mass

killing in the short term and allowing ‘‘the processes of internal, insti-

tutional change that warfare reflects’’ (Weinstein 2005: 4, 12). While

Weinstein focuses more on effective state institutions than on demo-

cratic ones per se, his argument nonetheless suggests that ‘‘autonomous

recovery’’ may be better for democratization than ‘‘aided recovery’’

through peacekeeping and external financial aid.10

10 Weinstein explores this argument in three cases: (1) Uganda, which consolidatedstate institutions and developed a ‘‘no-party democracy’’ after Museveni’s rebelvictory in 1986; (2) Eritrea, where ‘‘a democratic decentralization of politicalpower, a participatory constitution-making exercise, and the building ofnational political institutions reflective of a new national identity were all

46 Virginia Page Fortna

Wantchekon (Wantchekon and Neeman 2002; Wantchekon 2004)

argues that democracy can emerge directly from civil war when warlords

have economic incentives to create peace (so as to extract resources more

efficiently from society).11 They will prefer empowering the citizenry

(i.e., democratizing) to inviting in an external enforcer because the latter

cannot credibly commit to neutrality. Wantchekon draws a distinction

between a neutral (and relatively weak) arbitrator such as a UN peace-

keeping mission, and a powerful external enforcer. The former, he

argues, can assist in the transition to democracy, while the latter is a

substitute for democracy. But this suggests that the moves away from

strict neutrality and toward more militarily robust peacekeeping, under-

taken in the late 1990s, may be detrimental to democratization. If

Wantchekon’s reasoning is correct, there may thus be something of a

dilemma between militarily effective peacekeepers, better able to main-

tain stability, and those of the relatively weak and neutral variety that

can assist democracy rather than impede it in his model.

While Wantchekon contends that limited peacekeeping is good for

democracy, the implication of his argument points to a similar dilemma

as that suggested by Weinstein. It is the cost of war (in this case the

economic inefficiencies of war) that makes militarily powerful actors (war-

lords) democratize. If peacekeeping is too successful in its primary task of

maintaining peace, these actors may not need to democratize to reap the

benefits of peace. Peacekeeping and the stable peace it helps produce can

thus become a substitute for democracy rather than an aid to it.

Empirical implications

These dilemmas suggest that the empirical relationship between peace-

keeping and democratization will not be a straightforward positive

remnants (and products) of the EPLF’s [Eritrean People’s Liberation Front] war-making effort’’ (p. 18); and (3) Somalia, where Weinstein argues effective sub-state political structures were rebuilt ‘‘after the UN pulled out’’ (p. 14). Whilenone of these countries is a paragon of successful democratization, Weinsteinwould argue that they provide examples of the development of effective andrelatively representative state institutions. He does not, however, compare thesecases to democratization processes in cases where the international communityhas intervened.

11 This will not be the case when warlords can finance their operations and linetheir pockets with proceeds from drugs or easily extractable mineral wealth suchas diamonds.

Peacekeeping and democratization 47

one. On the one hand, if stable peace is necessary for democracy, then

since peacekeepers improve the chances for lasting peace, the presence

of peacekeepers should lead to more democratic outcomes, all else

being equal. Furthermore, peacekeeping can minimize day-to-day vio-

lence, build trust between former enemies, oversee initial elections,

monitor human rights, and generally help put war-torn societies on

the road toward democracy. On the other hand, the second dilemma

explored here implies that we might see mixed or even negative effects

of peacekeeping. Meanwhile, the first dilemma suggests that we need to

distinguish between the short-term and long-term effects of peacekeep-

ing on processes of democratization. Specifically, it suggests that the

presence of peacekeepers should enhance democratization in the short

term, but undermine it in the long term.

The remainder of this chapter consists of a quantitative empirical

exploration of these hypotheses. Of the civil wars ending since the end

of the Cold War, how have the democratization trajectories of those

with peacekeepers compared to those without? Has peacekeeping led

to higher or lower levels of democracy in the short term and in the long

term? Have different types of peacekeeping had different effects on

democracy?

The data

In order to examine the effects of peacekeeping on democratization, we

need to examine both cases in which peacekeepers were deployed and

those where belligerents were left to their own devices in the aftermath

of fighting. I use a data set created for a project evaluating the effects of

peacekeeping on the stability of peace (Fortna 2008). The data consist

of ninety-five cease-fires or breaks in the fighting from 1989 through

1999 in almost sixty civil wars. The data build on those compiled

by Doyle and Sambanis (2000; 2006), but I have added a number of

short-lived cease-fires not included in their data or in other data on civil

wars. Inclusion of these ultimately unsuccessful attempts to maintain

peace is particularly important for a study of post-war stability and

democratization as their omission would truncate variation in the

dependent variables and introduce selection bias. The cases are listed

in the appendix to this chapter.

I restrict the analysis to wars ending between 1989 and 1999

for several reasons. I examine only the post-Cold War era because

48 Virginia Page Fortna

peacekeeping was only very rarely used in civil wars during the Cold

War. Moreover, before 1989, its purpose was to contain conflicts from

drawing in the superpowers more than to maintain peace within the

conflict itself. More critical for this study, it was rarely used in an

attempt to bring democracy to war-torn societies until after 1989.

I include cases only through the end of 1999 for more practical reasons.

Data for some important control variables are unavailable after that

time. More important, this allows me to observe both whether peace

lasts and trajectories of democratization for at least five years after the

point of a cease-fire for all of my cases. So while to be included in this

study, a break in the fighting must occur before the end of 1999,

observation of the main dependent variable – post-war democratiza-

tion – continues through 2004. Ideally, to study long-term effects on

democratization we would have more than five years of information

since the most recent cases. But without a crystal ball, that is not

possible.12

Data on democratization are taken from two sources: the Polity IV

data set (Marshall and Jaggers 2002) and Freedom House (2005). The

Polity data rate countries on an annual basis on a 20-point scale from

full autocracy (�10) at one end to full democracy (þ10) at the other,

while Freedom House ranks countries, on two 7-point scales, in terms

of their civil liberties and political rights.13 Both of these measures

allow us to discern fairly fine-grain moves toward (or away from)

democracy and ‘‘freedom’’ in post-war societies. While there is signifi-

cant overlap in these two measures (as one would hope), they empha-

size different features of democracy. The Polity data focus relatively

more on the institutions and procedures of democracy, including elec-

tions, institutional constraints on the executive, and forms of competi-

tiveness in political participation. The Freedom House data rate

countries based on a checklist of political rights (including those con-

cerning the electoral process, political pluralism and participation,

government functioning and corruption), and civil liberties (including

12 Similarly, while it will be important to evaluate the effects of currentpeacekeeping cases, such as those in the Ivory Coast, the Sudan, and Liberia,it is too early to judge their effects on democratization.

13 Here I use an aggregate of these two Freedom House scores, rather thanthat organization’s designations of ‘‘free,’’ ‘‘partly free,’’ and ‘‘not free’’since collapsing aggregate scores into only three categories entails a loss ofinformation.

Peacekeeping and democratization 49

freedom of expression and belief, freedom of association, rule of law,

and personal autonomy and individual rights).14 Using both the Polity

and Freedom House data allows me to check the robustness of findings

about peacekeeping and democratization across these two measures.

For each data source, I measure democratization one, two, and five

years after the fighting stops. Because I am interested in movement

along the continuum from autocracy to democracy, I focus on change

in democracy scores rather than their absolute level. Democratization

one, two, and five years out is therefore measured as the difference

between a country’s Polity or Freedom House score in the year before

the fighting stops and its score one, two, and five years after the war

ends.15 I invert the Freedom House scores so that higher numbers

represent greater moves toward ‘‘freedom’’ so as to match the direction

of the Polity measure of democratization.

Studying democratization in the immediate aftermath of war pre-

sents some complications because war often resumes. Because I include

even very short-lived cease-fires in these data, there are cases in which

war breaks out anew before our measures of democratization are

taken. To be more specific, war resumes in less than one year in

thirty-one cases (about a third of the total), within two years in another

eight, and within five years in a further thirteen (that is, peace lasts

fewer than five years in over half of the cases). It is not obvious how

these cases should be treated. On the one hand, we might drop these

cases from the analysis, focusing rather on democratization in those

14 The Freedom House data have come under more criticism than the Politydata for being somewhat arbitrary, and for entailing a political bias(for example, in favor of US allies and of free market economic policies).To the extent that the international community has shared this bias, as Paris(2004) argues it has, this may provide a closer measure of what it is trying toachieve in post-war transitions.

15 I use the year before the fighting stops as the baseline, rather than the year thefighting stops because the annual Polity and Freedom House scores mayreflect moves along the democracy continuum that occur between the date of acease-fire and the end of the calendar year. Lagging the baseline by one yearensures that I am not capturing post-war levels in this baseline. I could alsouse democracy scores from before the war as a baseline, but since in some caseswar has been raging for many years and democracy levels often fluctuate overthese spans, I would be capturing changes in democracy that occur during thewar as part of my measure of post-war democratization. Furthermore, Iinclude democracy at the start of the war as a control variable to captureeffects of prior history with democracy.

50 Virginia Page Fortna

cases where peace is holding.16 On the other hand, these cases are

substantively interesting for a project on the relationship between

post-war democratization and post-war stability. In the empirical ana-

lyses that follow I take both tacks, presenting findings both for the full

set of cases, and for the set restricted to cases in which peace is holding

one, two, and five years out. This allows me, to an extent, to distinguish

the direct effects that peacekeeping has on democratization from the

indirect effects that it has because it makes peace more stable.

Many post-war societies do experience democratization. In the full set

of cases examined here (that is, including those in which war has since

resumed), the average democratization score one year after the fighting

stops is a hair over one point along the twenty-point Polity scale. While

most cases experience no change in their level of democracy this soon

after the war, and a few experience moves toward autocracy, about a

third see positive developments in terms of democratization. Many of

these are modest, one or two point increases in Polity scores (e.g.,

Namibia and Djibouti), but others were more dramatic, including

Mali’s fourteen-point jump, despite another round of war. Not surpris-

ingly, the democratization story is rosier if we restrict the analysis to only

those sixty-four cases in which peace held for the first year. The one year

democratization average for these cases is 1.35 using the Polity measure.

The trends are similar in the Freedom House data.

Two years out, we see even greater strides toward democracy. This is

not surprising as many democratic institutions need time to begin to

take root.17 Examining all cases, including those where peace falters

before two years are up, the average level of democratization is 1.42

points on the Polity scale with close to 40 percent of post-war countries

making at least some moves toward greater democracy. As before,

those cases where peace has held fare even better. Their average demo-

cratization after two years is 1.67 Polity points, with fewer backsliders

16 In some cases, war resumes but ends relatively quickly so that our measure ofdemocratization is taken when the country is back at peace. Because this secondcease-fire is also included in the data, these cases will enter the analyses again,with democratization scores taken one, two, and five years after the secondbreak in the fighting. In a small number of cases (3), the second cease-fire occursin the same year as the first so that democratization scores are duplicated. Thesecases are Guinea–Bissau 1998, India–Assam 1991, and Russia–Chechnya 1996.

17 Also, because states at war tend to have low levels of democracy, a statisticalregression toward the mean would suggest some positive moves towarddemocracy as we get farther in time from the fighting.

Peacekeeping and democratization 51

and larger gains in democracy. This improvement over time is as visible

in the Freedom House scores as it is in the Polity scores.

Five years out, the picture is remarkably close to that only two years

out – if we look at all cases. The average democratization over five

years is 1.46 points in Polity, only 0.04 higher than after two years. But

the discrepancy between cases in which war resumes and those where it

lasts is much larger five years out than in shorter-run examinations. Of

cases where peace has held, the five year democratization average is

2.78, dramatically higher than after just one or two years. In only two

cases in which peace has lasted at least five years are democracy scores

lower than they were during the war: Azerbaijan and Mali, which had

managed to democratize significantly in the early 1990s despite

ongoing fighting (see above), but lost ground, only slightly, afterward.

Meanwhile almost half of the cases moved away from autocracy and

toward democracy, including quite dramatic gains in Ethiopia,

Croatia, and Mozambique. The Freedom House scores show even

more improvement over five years. By this time, almost half the cases

had moved toward ‘‘freedom’’ relative to the year before the war ended

(including very large gains in places like Mali and South Africa), while

the rest are about evenly divided between no change and some move-

ment away from ‘‘freedom.’’ Of the restricted set of cases that saw no

new war, 65 percent made gains in political rights and civil liberties,

with only a handful moving away from ‘‘freedom.’’

Clearly then, democratization often happens in the aftermath of civil

war. But does peacekeeping by the international community foster this

democratization, or is it just as likely to hinder it?

Stable peace is better for the growth of democracy than renewed

warfare; that is apparent in the comparison of the full set of cases

with those in which peace lasts. And we know from previous

research that peacekeeping helps sustain stable peace. The risk of

war recurring is much lower when peacekeepers are present than

when belligerents are left to their own devices (Fortna 2004, 2008).

Table 2.1 shows the results of hazard analysis of the duration of

peace after civil war. The hazard ratios in this table indicate the

effect of variables, including peacekeeping and a series of control

variables, on the relative ‘‘hazard’’ of war resumption.18 Hazard

18 Results are from Weibull regressions. Results are even stronger if a Coxproportional hazards model is used. Robust standard errors are calculated

52 Virginia Page Fortna

ratios are interpreted relative to 1.0, such that a hazard ratio of

2.0 would indicate a doubling of the risk of war, all else equal; a

hazard ratio of 0.6 would indicate a 40 percent reduction in the risk

of war. Effects are shown for two versions of the peacekeeping

variable. Column 1 shows the effect of whether peacekeepers are

currently present, while column 2 shows the effects of having had a

Table 2.1 Effects of peacekeeping on the duration of peace

Time-varying measure Time-constant measure

Hazard

ratio (RSE) P> |Z|

Hazard

ratio (RSE) P> |Z|

All peacekeeping 0.43 (0.15) 0.02 0.19 (0.09) 0.00

Victory 0.12 (0.07) 0.00 0.08 (0.05) 0.00

Treaty 0.25 (0.11) 0.00 0.30 (0.13) 0.01

Identity war 1.34 (0.51) 0.45 1.05 (0.35) 0.89

Cost of war 1.19 (0.10) 0.04 1.20 (0.09) 0.02

Factions 0.56 (0.17) 0.05 0.70 (0.21) 0.24

Democracy 0.97 (0.03) 0.35 0.99 (0.03) 0.61

Infant mortality 1.00 (0.00) 0.52 1.00 (0.00) 0.45

Failed past agreement 1.19 (0.49) 0.68 1.21 (0.51) 0.65

Government army size 1.00 (0.00) 0.53 1.00 (0.00) 0.63

Contraband

financing

2.42 (0.83) 0.01 2.83 (1.04) 0.01

Mountainous terrain 1.05 (0.17) 0.76 1.21 (0.20) 0.26

Perm-5 contiguous 0.58 (0.33) 0.34 0.35 (0.22) 0.09

Perm-5 former colony 1.27 (0.51) 0.60 0.94 (0.37) 0.88

Duration of war 0.92 (0.03) 0.01 0.92 (0.03) 0.01

Shape parameter P 0.61 (0.05) 0.00 0.66 (0.06) 0.00

No. of subjects 93 93

No. of observations 129 129

Log pseudo-

likelihood

�142.77 �137.83

Weibull Regressions. Cases clustered by country for calculation of robust standard

errors.

with cases clustered so that the model assumes independence between casesin different countries but not necessarily between cases in the same country.

Peacekeeping and democratization 53

peacekeeping mission, even if it has since departed.19 The former is a

much more conservative estimate of peacekeeping effects (because

peacekeepers are not given credit for peace that lasts after they

depart), while the latter represents a more realistic assessment of the

‘‘success’’ of peacekeeping. But in either case, it is clear that peace-

keeping has a large and significant effect on stability. The hazard

ratio of 0.41 for the measure in column 1 indicates that peacekeeping

reduces the risk of war by 59 percent, all else equal. The measure of

peacekeeping used in column 2 indicates an 82 percent decline in the

risk of another war when peacekeepers have deployed, relative to

when belligerents are left to keep peace on their own.20

But the key question for this project is whether peacekeeping pro-

motes democracy in post-war societies, or whether democracy is better

served in those cases where belligerents manage without outside

interference.

Peacekeeping and democratization: findings

Table 2.2 shows the mean democratization levels for cases with no

peacekeepers, cases where Chapter VI consent-based peacekeeping

missions were deployed, and cases where more robust Chapter VII

enforcement missions were deployed. The table also describes the

results of tests of the statistical significance of these differences. On

the face of it, democratization appears to be aided by the presence of

peacekeepers. Cases with consent-based peacekeeping experience, on

average, more post-war democratization than cases with no peace-

keeping, and cases with enforcement peacekeeping missions experience

higher democratization still.21 These differences are generally not

19 In the duration analysis jargon, these correspond to ‘‘time-varying covariates’’and ‘‘time-constant covariates’’ measures, respectively.

20 For explanation of the statistical method, and much more detailed analysis of theeffects of peacekeeping on the duration of peace, see Fortna (2008, especiallyChapter 5).

21 Results shown here use the Polity measure of democratization. The pattern ismuch the same when the Freedom House measure is used, with two exceptions:both one and five years out, in the restricted set of cases where peace has lasted todate, democratization is slightly lower on average for Chapter VI missions thanwhen no peacekeepers deploy. For this measure at least, the effect of consent-based peacekeeping on democracy is largely about maintaining stability ratherthan a direct effect on civil liberties and political rights.

54 Virginia Page Fortna

statistically significant, however, so while the preponderance of evi-

dence suggests that peacekeeping fosters democratization, we cannot

reject the null hypothesis of no effect with confidence.22

The bivariate relationship between peacekeeping and democratiza-

tion tells us only so much, however. Because peacekeeping is not

deployed to civil wars at random, and because there may be other

factors that affect both whether peacekeepers are present and the

prospects for democratization, multivariate analysis is essential for eval-

uating this empirical relationship. In the analyses that follow I control

for a number of variables identified as important in the general literature

on democratization (e.g., Barro 1999; Burkhart and Lewis-Beck 1994;

Table 2.2 Mean democratization scores (Polity), by peacekeeping

One year

out

Two years

out

Five years

out

All

cases

If peace

lasts

All

cases

If peace

lasts

All

cases

If peace

lasts

No peacekeeping 0.36 0.49 0.46 0.68 0.65 2.25

Consent-based PK 1.14 1.00 2.36 1.79 2.62 2.67

Enforcement PK 4.55 5.00 4.73 4.89 4.00 5.20

N 92 62 92 54 84 37

Statistical significance of difference in means (results were consistent whether

Bonferroni, Scheffe, or Sidak methods were used):

(1) In no case is the difference in means between no peacekeeping and consent-based

peacekeeping statistically significant.

(2) The difference between consent-based peacekeeping and enforcement is

statistically significant only for democratization one year out if peace lasts (i.e.,

column 2).

(3) The difference between no peacekeeping and enforcement is statistically

significant one and two years out, among all cases or only if peace lasts (columns 1–4)

but not five years out (columns 5–6).

(4) The difference in means between no peacekeeping and any peacekeeping

(combined means not shown) is significant one, two, and five years out for all cases,

but is only marginally significant (0.05< p< 0.10) for cases in which peace lasts one

and two years out, and is not significant for cases in which peace lasts five years out.

22 Note that the lower levels of statistical significance for differences when werestrict analysis to those cases where peace has lasted, and when we look atlonger-term effects, may be the result of smaller sample sets.

Peacekeeping and democratization 55

Diamond, Linz, and Lipset 1988; Inglehart 1997; Lipset 1959;

Londregan and Poole 1996; O’Donnell, Schmitter, and Whitehead

1986; Przeworski et al. 1996; Przeworski and Limongi 1997; Ross

2001). These include GDP per capita, illiteracy rates, oil exports, and

the level of democracy at the beginning of the war. I also include

characteristics of the country or the conflict that have been found to

affect either the likelihood that peacekeepers will deploy, the prospects

for peace, or both. These include the war’s outcome: whether the

fighting ended in a clear victory for one side, whether it ended in a

peace treaty (with the omitted comparison category here being wars

that end with only a truce or a cease-fire); the cost of the war in terms of

lives lost or people displaced from their home; and a number of dummy

variables marking in turn: wars in which rebel forces funded their fight

through contraband (drugs, diamonds, etc.); conflicts in or next to the

territory of one of the five permanent members of the UN Security

Council (Perm-5); former colonies of the Perm-5; and as a proxy for

levels of mistrust, cease-fires that take place after a previous, failed,

cease-fire agreement.23

I also checked the effects of several other variables, including

whether the conflict was identity-based, whether it was a secessionist

conflict, whether neighboring countries provided significant aid to the

rebels, whether there were only two sides involved or many factions,

the size of the government’s army, and the mountainousness of the

terrain in the country.24 None of these affected levels of democratiza-

tion, nor affected other results, so I do not include them in the results

reported here. Nonetheless, some of these negative findings are sub-

stantively important and interesting. They suggest, for example, that

contrary to what conventional wisdom might lead us to expect, coun-

tries rent by ethnic wars or other identity-based divisions fare no worse

in terms of post-war democracy than do other war-torn countries.

Similarly, countries afflicted by complex wars that pit many factions

against each other have no worse democratization records than those

emerging from simpler wars that involved only two sides.

23 For more information on these variables, see Fortna (2008). The data areavailable from the author (see www.columbia.edu/~vpf4/research.htm).

24 Mountainous terrain has been shown to be a good predictor of civil war,presumably because it eases rebellion and hinders government efforts to fightinsurgency (Fearon and Laitin 2003).

56 Virginia Page Fortna

Because the measures of democratization are continuous, I use ordin-

ary least squares (OLS) regression. However, because not all of the

cases in the data are independent of one another, I cluster cases by

conflict and calculate robust standard errors.25

Table 2.3 shows the effects of peacekeeping, as well as other poten-

tially important variables on post-war democratization one year after

the conflict. Democratization is measured using both the Polity (col-

umns 1 and 2) and Freedom House (columns 3 and 4) scores. The table

shows the effects on all cases (columns 1 and 3) and on the subset of

cases in which peace lasts at least one year (columns 2 and 4). Table 2.4

shows the same thing for democratization two years out; Table 2.5 for

five years out (with the ‘‘if peace lasts’’ columns referring to cases in

which peace lasts two and five years, respectively).

It is worth discussing some of the control variables first since many of

these results are quite surprising. There is a relationship between the

level of democracy at the start of the war and democratization after-

wards, but this relationship is negative. Antebellum experience with

democracy actually makes countries more likely to move toward auto-

cracy after the fighting stops. This does not appear to be the result of

the fact that more democratic countries simply have less room for

improvement after the war – the result holds even if we drop countries

with high pre-war democracy scores from the analysis.26 This surpris-

ing result is perhaps less so when we consider that civil wars are

watershed events in the politics of a country. They mark a clear break

from the politics of the past. Many civil wars turn autocratic states into

democratic ones, while those enjoying democracy before the war have

experienced a cataclysmic shock to that system.

The results for GDP per capita and illiteracy are similarly unex-

pected. One of the strongest findings in the democratization literature

is that a relatively high level of wealth is a precondition for democracy.

In the subset of cases that have experienced civil war, however, the

opposite seems to be the case; richer countries enjoy, if anything, lower

25 Thus, I assume that cases are independent across conflicts, but notnecessarily within them. That is, democratization in Angola is independentof democratization in Cambodia, but democratization in Angola after1991 and after 1994 are related processes.

26 This negative relationship is significant for the Polity measure, butnot for the Freedom House measure. Note, however, that the measure ofpre-war democracy is taken from Polity data in both cases.

Peacekeeping and democratization 57

Table

2.3

Dem

ocr

ati

zati

on

on

eyea

rou

t

Poli

tym

easu

re

Fre

edom

House

mea

sure

All

case

s

Ifpea

ce

last

s

All

case

s

Ifpea

ce

last

s

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

All

pea

cekee

pin

g0.6

43

(0.9

93)

0.5

20.8

02

(1.2

04)

0.5

10.0

58

(0.6

79)

0.9

3�

0.7

00

(0.8

71)

0.4

3

Dem

ocr

acy

atw

arst

art�

0.2

66

(0.0

98)

0.0

1�

0.2

30

(0.1

02)

0.0

3�

0.0

41

(0.0

57)

0.4

8�

0.0

29

(0.0

53)

0.5

9

GD

P/c

apit

a�

0.0

003

(0.0

002)

0.2

4�

0.0

005

(0.0

003)

0.1

1�

0.0

00

(0.0

02)

0.4

9�

0.0

00

(0.0

00)

0.3

7

Illi

tera

cy0.0

17

(0.0

25)

0.5

00.0

02

(0.0

29)

0.9

3�

0.0

07

(0.0

19)

0.7

0�

0.0

01

(0.0

22)

0.9

6

Vic

tory

1.8

47

(1.2

27)

0.1

40.9

55

(1.6

50)

0.5

7�

0.0

05

(0.6

97)

1.0

0�

0.9

48

(0.7

80)

0.2

3

Tre

aty

1.0

27

(0.9

47)

0.2

90.1

14

(1.3

01)

0.9

31.3

62

(0.7

32)

0.0

70.8

15

(0.7

96)

0.3

1

Cost

of

war

�0.6

35

(0.2

18)

0.0

1�

0.6

53

(0.2

85)

0.0

3�

0.1

82

(0.1

28)

0.1

6�

0.1

23

(0.1

70)

0.4

7

Contr

aband

financi

ng

1.4

35

(1.0

90)

0.2

01.3

91

(1.3

27)

0.3

0�

0.0

63

(0.8

83)

0.9

40.1

57

(0.8

94)

0.8

6

Oil

�2.5

47

(0.8

71)

0.0

1�

1.6

92

(0.9

75)

0.0

9�

0.2

52

(0.7

11)

0.7

3�

0.2

47

(0.8

00)

0.7

6

Fail

edpast

agre

emen

t�

1.9

57

(0.6

61)

0.0

1�

1.5

67

(0.7

17)

0.0

40.1

72

(0.5

22)

0.7

40.4

53

(0.5

66)

0.4

3

Per

m-5

conti

guous

1.6

61

(0.8

14)

0.0

50.9

41

(1.2

40)

0.4

50.1

91

(0.4

68)

0.6

9�

0.0

41

(0.7

33)

0.9

6

Per

m-5

form

erco

lony�

1.5

55

(1.0

54)

0.1

5�

2.7

52

(1.2

82)

0.0

4�

0.1

20

(0.9

27)

0.9

0�

0.6

95

(1.0

13)

0.5

0

Const

ant

8.9

29

(3.0

52)

0.0

110.7

56

(4.0

47)

0.0

12.4

15

(1.7

82)

0.1

82.8

39

(2.1

24)

0.1

9

N81

56

84

58

R2

0.3

80.4

30.1

10.1

2

Table

2.4

Dem

ocr

ati

zati

on

two

yea

rsou

t

Poli

tym

easu

reFre

edom

House

mea

sure

All

case

sIf

pea

cela

sts

All

case

sIf

pea

cela

sts

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

All

pea

cekee

pin

g0.7

69

(0.9

80)

0.4

40.2

38

(1.2

62)

0.8

50.5

89

(0.6

73)

0.3

90.3

42

(0.8

18)

0.6

8

Dem

ocr

acy

atw

arst

art�

0.2

44

(0.0

97)

0.0

2�

0.2

23

(0.1

05)

0.0

4�

0.0

69

(0.0

66)

0.3

0�

0.0

77

(0.0

56)

0.1

8

GD

P/c

apit

a�

0.0

003

(0.0

003)

0.2

4�

0.0

005

(0.0

003)

0.1

5�

0.0

002

(0.0

002)

0.4

8�

0.0

000

(0.0

003)

0.9

6

Illi

tera

cy0.0

35

(0.0

25)

0.1

70.0

20

(0.0

30)

0.5

2�

0.0

12

(0.0

16)

0.4

6�

0.0

02

(0.0

20)

0.9

1

Vic

tory

1.7

66

(1.2

57)

0.1

70.5

23

(1.7

16)

0.7

60.6

92

(0.8

16)

0.4

0�

0.6

84

(0.9

41)

0.4

7

Tre

aty

1.4

24

(0.8

90)

0.1

2�

0.0

12

(1.3

57)

0.9

91.9

32

(0.6

61)

0.0

10.6

34

(0.8

55)

0.4

6

Cost

of

war

�0.5

74

(0.2

25)

0.0

2�

0.5

18

(0.2

98)

0.0

9�

0.1

52

(0.1

33)

0.2

6�

0.0

41

(0.1

67)

0.8

1

Contr

aband

financi

ng

0.6

56

(1.1

60)

0.5

80.2

37

(1.4

86)

0.8

7�

0.3

22

(0.6

89)

0.6

4�

0.0

02

(0.8

02)

1.0

0

Oil

�2.9

87

(0.8

07)

0.0

0�

2.7

52

(1.1

19)

0.0

2�

1.0

78

(0.4

96)

0.0

4�

1.5

40

(0.7

54)

0.0

5

Fail

edpast

agre

emen

t�

1.9

26

(0.7

21)

0.0

1�

1.8

63

(0.8

25)

0.0

30.1

33

(0.4

62)

0.7

80.1

59

(0.5

72)

0.7

8

Per

m-5

conti

guous

1.8

67

(0.8

19)

0.0

30.8

83

(1.3

51)

0.5

20.4

07

(0.5

08)

0.4

30.0

31

(0.9

38)

0.9

7

Per

m-5

form

erco

lony�

2.0

52

(1.0

49)

0.0

6�

3.7

86

(1.2

57)

0.0

1�

0.0

33

(0.8

17)

0.9

7�

0.9

77

(0.9

08)

0.2

9

Const

ant

8.0

66

(3.0

43)

0.0

110.3

01

(4.3

45)

0.0

21.9

27

(2.0

60)

0.3

62.0

23

(2.5

90)

0.4

4

N81

49

84

51

R2

0.4

10.4

80.2

10.2

5

Table

2.5

Dem

ocr

ati

zati

on

five

yea

rsou

t

Poli

tym

easu

reFre

edom

House

mea

sure

All

case

sIf

pea

cela

sts

All

case

sIf

pea

cela

sts

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

All

pea

cekee

pin

g0.2

26

(1.1

56)

0.8

5�

0.4

59

(1.2

11)

0.7

10.0

14

(0.6

70)

0.9

8�

0.9

74

(1.0

26)

0.3

5

Dem

ocr

acy

atw

arst

art�

0.2

85

(0.1

04)

0.0

1�

0.4

71

(0.1

24)

0.0

0�

0.1

10

(0.0

66)

0.1

0�

0.0

73

(0.0

92)

0.4

4

GD

P/c

apit

a0.0

001

(0.0

004)

0.7

40.0

007

(0.0

004)

0.1

2�

0.0

003

(0.0

002)

0.2

10.0

001

(0.0

003)

0.8

9

Illi

tera

cy0.0

22

(0.0

36)

0.5

4�

0.0

13

(0.0

35)

0.7

1�

0.0

10

(0.0

15)

0.5

1�

0.0

10

(0.0

25)

0.7

1

Vic

tory

1.3

94

(1.1

42)

0.2

31.9

91

(1.3

28)

0.1

51.1

39

(0.7

88)

0.1

6�

0.1

92

(1.3

64)

0.8

9

Tre

aty

0.6

29

(1.0

50)

0.5

5�

1.1

57

(1.3

41)

0.4

01.3

62

(0.7

14)

0.0

6�

0.2

37

(1.0

18)

0.8

2

Cost

of

war

�0.6

69

(0.2

47)

0.0

1�

0.4

46

(0.2

93)

0.1

4�

0.1

38

(0.1

47)

0.3

6�

0.0

47

(0.2

04)

0.8

2

Contr

aband

financi

ng

1.0

73

(1.5

29)

0.4

92.6

68

(1.4

29)

0.0

7�

0.5

35

(0.7

93)

0.5

0�

0.9

26

(1.2

64)

0.4

7

Oil

�5.3

48

(1.6

28)

0.0

0�

9.1

90

(1.6

41)

0.0

0�

2.4

73

(0.7

32)

0.0

0�

3.3

35

(1.3

35)

0.0

2

Fail

edpast

agre

emen

t�

1.2

61

(0.8

40)

0.1

4�

0.9

98

(1.1

89)

0.4

10.2

68

(0.4

86)

0.5

8�

0.2

58

(0.8

72)

0.7

7

Per

m-5

conti

guous

0.6

75

(1.6

55)

0.6

9�

0.0

71

(1.4

27)

0.9

60.6

01

(0.5

19)

0.2

50.7

61

(1.1

37)

0.5

1

Per

m-5

form

erco

lony�

2.5

37

(1.0

06)

0.0

2�

2.9

34

(1.2

20)

0.0

2�

1.3

67

(0.7

86)

0.0

9�

2.0

88

(1.0

83)

0.0

7

Const

ant

10.4

58

(3.6

67)

0.0

110.2

82

(4.2

91)

0.0

23.3

50

(2.1

03)

0.1

24.2

56

(3.2

78)

0.2

1

N76

35

84

39

R2

0.3

70.6

70.2

60.3

9

levels of democratization (that is, smaller moves toward democracy or

greater moves toward autocracy) than poorer ones in the first few

years. These effects are not statistically significant (though in some

models they are close) and should thus be interpreted with caution,

but we certainly do not see the expected positive effect of GDP in the

immediate aftermath of war. This negative effect falls away by the fifth

year, however, suggesting that the short-term and long-term effects of

wealth are quite different.

The short-term effect of literacy is the opposite of what we would

expect when the Polity measure of democratization is used. Again these

coefficients are not significant, but countries with higher proportions

of people who cannot read appear to undergo, if anything, more

democratization than those with more literate populations. By the

fifth year out, this effect holds when we examine all cases but flips to

the expected direction in the subset of cases in which peace has lasted to

date. The relationship between literacy and the Freedom House mea-

sure is more in line with expectations, higher levels of illiteracy depress

moves toward ‘‘freedom,’’ but this relationship is never even remotely

significant. In short, literacy appears to have no systematic benefits for

post-war democratization.

The effects of cost of war are more in line with what we would

intuitively expect. The more people killed and displaced by the war,

the harder it is to democratize. This effect is statistically significant and

quite robust for the Polity measure but is not significant for the

Freedom House measure. For the Polity measure, the size of this effect

is fairly consistent whether we are examining all cases or only those in

which peace has lasted.27 This suggests that the negative effect of the

cost of war is not just the result of the fact that high-cost wars are more

likely to resume, which in turn depresses democratization. Rather the

cost of war appears to have an independent effect on efforts to move

toward democracy. However, for the Freedom House measure, the

effect of the cost of war is generally larger for the full set of cases

than for the restricted set. This difference between the two measures

is telling. It indicates that the direct effect of costly wars operates more

strongly on the institutional forms of democracy rather than on poli-

tical rights and civil liberties. For the latter, it is the detrimental effect

27 The measure used here is the natural log of the number killed or displaced.

Peacekeeping and democratization 61

on stability that undermines democracy. This may be because very

costly wars undermine a state’s political and physical infrastructure

in ways that make building institutions more difficult. Alternatively, it

may reflect the countervailing effects suggested by the second dilemma

outlined above; that more costly wars lead citizens to demand more

political rights in ways that offset otherwise negative effects on

democracy.

Our proxy measure for levels of mistrust between the recent belli-

gerents (whether previous agreements between them have failed) has a

similar negative effect on the establishment of the institutions and

procedures of democracy (as measured by Polity). Here, however, the

results appear to wear off over time. One and two years out, a failed

agreement in the past depresses democratization by about two points

(on the twenty-point scale), all else equal, and this effect is statistically

significant. By five years out, however, the effect is about half as big,

and no longer significant. This effect does not hold for the expansion of

civil liberties and political rights (as measured by Freedom House),

however.

Wars that end with a decisive victory for one side see, if anything,

larger moves toward the institutions and procedures of democracy

(Polity) than those that do not, consistent with earlier research (Toft

2003). However, this effect is not significant, though it is often close, so

should be treated with caution. Moreover, it does not hold for the

Freedom House measure of democratization. The effect of victory on

civil liberties and political rights is generally negative, though not

consistently so.

Remember that the comparison category here is fighting that ends

with only a cease-fire or truce in place – an outcome that is neither good

for stable peace, nor, clearly, for democracy. When we examine all

cases, we see that wars that end with a peace treaty also experience

positive levels of democratization (though again, this effect is not

always significant). This is true whichever measure of democratization

is used. This effect is generally smaller than that for decisive victory,

however, and it does not hold among only those cases for which peace

lasts. This suggests that while peace treaties help make peace more

stable, which in turn fosters democratization, they do not, in and of

themselves, foster democratization. This result is quite surprising, as

many peace treaties call for elections in an attempt to move conflict

from the battlefield to the ballot box.

62 Virginia Page Fortna

Natural resources also have interesting effects on attempts at post-

war democratization. Oil-rich countries are cursed by their valuable

assets. The negative effect of oil on democratization is large, statisti-

cally significant, and grows over time. On the Polity scale, oil reduces

democratization by about 2 points one year out, 3 points two years out,

and a whopping 5 to 9 points by the fifth year after the war ends. The

pattern is much the same in the Freedom House data.28 This result

confirms studies that find ‘‘oil and democracy don’t mix.’’ As Ross

(2001) argues, oil wealth (a) allows governments to use low taxes and

patronage to avoid pressures for reform, (b) allows them to fund efforts

to repress reformist movements, and/or (c) produces economic growth

that, unlike other forms of economic development, does not lead to

social and cultural changes, such as education and economic speciali-

zation, that tend to foster democracy.

On the other hand, and quite surprisingly, contraband financing for

rebels does not appear to hurt the prospects for democracy. In fact the

positive coefficient for this variable in the Polity data suggests just the

opposite. This effect is not significant (though it comes close in some

models), nor does it hold in the Freedom House data, so we should not

conclude that conflict diamonds and drugs are good for democracy.

But the fact that we see no detrimental effect and the possibility of a

positive effect is quite unexpected.

The relationship between a country and the great powers – the United

States, Great Britain, France, China, or Russia (also known as the

permanent members of the Security Council, or Perm-5) – also has an

effect on post-war democratization. Conflicts that are in (e.g.,

Northern Ireland, Chechnya) or next door to the Perm-5 see higher

rates of democratization one and two years out. But this effect is much

larger and only statistically significant when we examine all cases

rather than the subset for which peace has held. And in the Polity

data at least, the effect is largely gone five years down the road.

Interestingly, former colonies of these countries democratize several

points fewer on the Polity scale than other countries, an effect that is

consistent across models and generally statistically significant. While

this effect is much less pronounced, it is generally still negative in the

Freedom House data. This finding is particularly surprising as all of

28 The scale of the two measures is different, but the size of the effect of oil ondemocratization is somewhat smaller in the Freedom House data.

Peacekeeping and democratization 63

these countries are former colonies of three leading democracies (the

United States, Britain, and France) that we might expect to be particu-

larly interested in fostering democracy in their former wards.29 Why

former colonies of these three countries might fare so poorly in post-

war democratization is unclear – it cannot be because these cases start

at higher rates of democracy and so have less room for improvement,

since the analysis controls for democracy levels at the start of the war.

This is an interesting topic for further investigation.

Turn now to our main variable of interest, peacekeeping. Overall,

the resul ts in Tables 2.3–2. 5 are rathe r wea k: peaceke eping appears to

have little effect on democratization. There is no statistically significant

effect in either direction.30 In the short term, one and two years after

the fighting stops, there is more evidence for a positive effect than a

negative one, at least when the Polity measure of the institutions and

procedures of democracy is employed (columns 1 and 2). Peacekeeping

does not appear to thwart democratization efforts and may have a

modest positive effect. But this effect is not statistically significant –

we cannot place much confidence in this positive result.31 We can,

however, be fairly confident that peacekeeping does not undermine

democratization, at least in the short term. In the longer term, five years

out, the positive effect becomes even smaller, and when only cases in

which peace lasts are considered, we see a possibly negative effect. This

suggests at least weak support for the temporal dilemma described

above. Although, again, lack of statistical significance makes definitive

statements impossible to make.

The effects are even worse for peacekeeping when the Freedom

House measure of moves toward or away from ‘‘freedom’’ is used

(columns 3 and 4). Here there is no evidence of a positive effect one

year out and if anything peacekeeping depresses freedom among the

subset of cases in which peace lasts. Two years out the coefficients are

29 This surprising negative effect is stronger in former French and US colonies (theseveral cases in the Philippines are the only instances of the latter) than in formerBritish colonies (results not shown).

30 This means that the positive and negative coefficients could be the result ofchance alone. Note, however, that because the data used here represent all cease-fires in civil wars during this period, not a smaller sample, issues of inferencefrom the data to the larger population are much smaller than in many statisticalanalyses.

31 In models in which the control for former colonies of the Perm-5 is omitted, thispositive effect is often significant or close to it.

64 Virginia Page Fortna

positive for both sets of cases, but still not significantly so. Five years

out, we again see no effect (the coefficient is essentially zero) in the full

set of cases, and if anything a negative effect in the restricted set. This

suggests that the direct effect of peacekeeping is, if anything, to thwart

political rights and civil liberties, but that this is offset by the positive

effects of peacekeeping on peace and stability. In short, there is evi-

dence for the dilemmas proposed above – peacekeeping can help estab-

lish peace, which is good for democracy, but it also undermines the

establishment of that democracy.

Ta bl es 2.6–2.8 distinguish between the four types of peacekeeping

mission outlined above: observer missions, interpositional or traditional

peacekeeping missions, multidimensional missions, and enforcement

missions. As noted above, the first three categories are consent-based

missions with Chapter VI mandates, while the fourth does not rely on

consent and is authorized under Chapter VII of the UN Charter. As

before, results are shown for both Polity and Freedom House measures,

and for both the full set of cases and the restricted set of cases in which

peace holds one, two, and five years out, respectively. Examining sys-

tematic differences in the effects of different types of peacekeeping is

difficult given the small numbers of cases in each category, but this

breakdown yields some potentially interesting results.

Let us look first at the effect of mission type on the institutions and

procedures of democracy, as measured by Polity (columns 1 and 2 of

the tables). One year after the fighting stops, all types of peacekeeping

have a positive (if insignificant) effect on democratization except multi-

dimensional peacekeeping. This is particularly striking because it is

exactly these missions, with large civilian components for election

monitoring, human rights monitoring, and other attempts at institution

building that we would expect to have the largest positive impact on

democracy building. In the very short term, it appears that maintaining

security is more important for the establishment of democracy than

explicit attempts by peacekeepers to build its institutions.

By two years out the picture has changed, however. Here, both

multidimensional peacekeeping and more robust peace enforcement

missions have, if anything, a positive effect on democracy building,

while monitoring and traditional peacekeeping missions have negative

coefficients. Five years out, monitoring and traditional peacekeeping

continue to have a negative effect, and multidimensional peacekeeping

maintains a positive coefficient. The effect of enforcement missions is

Peacekeeping and democratization 65

Table

2.6

Dem

ocr

ati

zati

on

on

eyea

rou

t,by

pea

cekee

pin

gm

issi

on

type

Poli

tym

easu

reFre

edom

House

mea

sure

All

case

sIf

pea

cela

sts

All

case

sIf

pea

cela

sts

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

Monit

ori

ng

PK

0.7

17

(1.6

53)

0.6

71.2

21

(2.0

64)

0.5

60.6

02

(1.4

13)

0.6

7�

0.6

67

(1.8

87)

0.7

3

Tra

dit

ionalPK

0.3

22

(1.0

90)

0.7

70.4

34

(1.2

98)

0.7

4�

0.5

36

(0.9

37)

0.5

7�

1.1

62

(0.9

30)

0.2

2

Mult

idim

ensi

onalPK�

1.3

96

(1.0

48)

0.1

9�

1.4

48

(1.3

24)

0.2

8�

0.1

97

(0.9

66)

0.8

4�

1.2

11

(1.1

76)

0.3

1

Enfo

rcem

ent

PK

1.9

95

(1.7

66)

0.2

72.0

19

(2.0

03)

0.3

20.4

94

(0.8

66)

0.5

7�

0.1

15

(1.0

41)

0.9

1

Dem

ocr

acy

atw

arst

art�

0.2

40

(0.1

04)

0.0

3�

0.1

99

(0.1

15)

0.0

9�

0.0

40

(0.0

58)

0.5

0�

0.0

21

(0.0

56)

0.7

2

GD

P/c

apit

a�

0.0

003

(0.0

003)

0.3

2�

0.0

005

(0.0

003)

0.1

7�

0.0

00

(0.0

00)

0.5

3�

0.0

00

(0.0

00)

0.4

7

Illi

tera

cy0.0

20

(0.0

27)

0.4

70.0

09

(0.0

32)

0.7

9�

0.0

08

(0.0

19)

0.6

70.0

01

(0.0

23)

0.9

7

Vic

tory

1.6

33

(1.3

30)

0.2

30.6

66

(1.8

56)

0.7

2�

0.0

11

(0.7

09)

0.9

9�

1.0

71

(0.7

67)

0.1

7

Tre

aty

1.1

64

(1.0

13)

0.2

60.2

72

(1.3

75)

0.8

41.3

72

(0.8

21)

0.1

00.7

63

(0.8

52)

0.3

8

Cost

of

war

�0.5

71

(0.2

22)

0.0

1�

0.5

50

(0.3

09)

0.0

8�

0.1

77

(0.1

37)

0.2

0�

0.1

14

(0.1

94)

0.5

6

Contr

aband

financi

ng

1.2

11

(1.1

43)

0.3

01.1

82

(1.3

76)

0.4

0�

0.1

35

(0.8

75)

0.8

80.1

38

(0.9

08)

0.8

8

Oil

�2.5

96

(0.9

95)

0.0

1�

1.8

02

(1.1

02)

0.1

1�

0.3

91

(0.6

96)

0.5

8�

0.3

14

(0.8

25)

0.7

1

Fail

edpast

agre

emen

t�

2.1

01

(0.7

15)

0.0

1�

1.6

91

(0.6

87)

0.0

20.1

60

(0.5

19)

0.7

60.3

92

(0.5

68)

0.4

9

Per

m-5

conti

guous

1.5

61

(0.8

50)

0.0

70.8

69

(1.3

73)

0.5

30.2

30

(0.4

89)

0.6

4�

0.0

18

(0.7

84)

0.9

8

Per

m-5

form

erco

lony�

1.5

51

(1.0

29)

0.1

4�

2.7

43

(1.2

63)

0.0

4�

0.1

37

(0.9

30)

0.8

8�

0.6

75

(1.0

31)

0.5

2

Const

ant

8.2

17

(3.1

94)

0.0

19.5

14

(4.4

86)

0.0

42.4

25

(1.9

90)

0.2

32.7

35

(2.5

46)

0.2

9

N81

56

84

58

R2

0.4

10.4

80.1

20.1

4

Table

2.7

Dem

ocr

ati

zati

on

two

yea

rsou

t,by

pea

cekee

pin

gm

issi

on

type

Poli

tym

easu

reFre

edom

House

mea

sure

All

case

sIf

pea

cela

sts

All

case

sIf

pea

cela

sts

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

Monit

ori

ng

PK

0.2

60

(1.6

48)

0.8

8�

1.5

59

(1.9

05)

0.4

2�

0.6

98

(0.9

70)

0.4

8�

1.7

35

(1.8

89)

0.3

6

Tra

dit

ionalPK

�0.0

37

(1.1

16)

0.9

7�

0.8

13

(1.3

41)

0.5

50.6

17

(0.7

38)

0.4

10.8

14

(0.8

86)

0.3

6

Mult

idim

ensi

onalPK

0.3

48

(1.6

97)

0.8

40.3

97

(2.4

05)

0.8

71.2

62

(0.9

60)

0.2

00.7

40

(1.3

43)

0.5

9

Enfo

rcem

ent

PK

2.0

25

(1.6

76)

0.2

31.5

85

(2.1

88)

0.4

70.8

63

(0.9

57)

0.3

70.5

84

(1.1

50)

0.6

2

Dem

ocr

acy

atw

arst

art�

0.2

35

(0.1

03)

0.0

3�

0.1

94

(0.1

33)

0.1

5�

0.0

70

(0.0

71)

0.3

3�

0.0

64

(0.0

69)

0.3

6

GD

P/c

apit

a�

0.0

004

(0.0

003)

0.2

3�

0.0

007

(0.0

004)

0.1

3�

0.0

002

(0.0

003)

0.4

2�

0.0

000

(0.0

004)0.9

1

Illi

tera

cy0.0

34

(0.0

27)

0.2

00.0

19

(0.0

35)

0.5

9�

0.0

12

(0.0

17)

0.5

0�

0.0

02

(0.0

23)

0.9

1

Vic

tory

1.5

75

(1.3

33)

0.2

40.2

16

(1.7

76)

0.9

00.5

34

(0.8

54)

0.5

4�

.921

(1.0

14)

0.3

7

Tre

aty

1.3

55

(0.9

46)

0.1

6�

0.2

95

(1.4

65)

0.8

41.6

91

(0.7

67)

0.0

30.1

93

(0.9

25)

0.8

4

Cost

of

war

�0.5

69

(0.2

26)

0.0

2�

0.5

74

(0.3

42)

0.1

0�

0.1

78

(0.1

49)

0.2

4�

0.0

93

(0.2

26)

0.6

8

Contr

aband

financi

ng

0.4

97

(1.2

17)

0.6

90.0

12

(1.6

82)

0.9

9�

0.2

81

(0.7

18)

0.7

0�

0.1

14

(0.8

21)

0.8

9

Oil

�3.1

25

(0.8

64)

0.0

0�

2.9

23

(1.2

92)

0.0

3�

1.0

09

(0.5

57)

0.0

8�

1.5

54

(0.8

33)

0.0

7

Fail

edpast

agre

emen

t�

1.9

76

(0.7

77)

0.0

2�

1.7

77

(0.8

60)

0.0

50.0

85

(0.4

93)

0.8

60.2

38

(0.6

79)

0.7

3

Per

m-5

conti

guous

1.8

24

(0.8

68)

0.0

40.6

82

(1.5

15)

0.6

60.3

58

(0.5

15)

0.4

9�

0.2

45

(1.0

12)

0.8

1

Per

m-5

form

erco

lony�

2.0

43

(1.0

13)

0.0

5�

3.6

24

(1.2

05)

0.0

10.0

11

(0.7

98)

0.9

9�

0.7

31

(0.9

36)

0.4

4

Const

ant

8.2

42

(3.1

85)

0.0

111.4

61

(5.0

87)

0.0

32.3

86

(2.2

88)

0.3

02.8

91

(3.3

45)

0.3

9

N81

49

84

51

R2

0.4

20.5

00.2

30.2

9

Table

2.8

Dem

ocr

ati

zati

on

five

yea

rsou

t,by

pea

cekee

pin

gm

issi

on

type

Poli

tym

easu

reFre

edom

House

mea

sure

All

case

sIf

pea

cela

sts

All

case

sIf

pea

cela

sts

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

Monit

ori

ng

PK

�0.1

20

(2.0

61)

0.9

5�

2.5

02

(3.2

17)

0.4

41.1

84

(1.3

78)

0.4

0�

2.3

06

(2.1

69)

0.3

0

Tra

dit

ionalPK

�1.2

38

(1.0

40

0.2

4�

0.9

95

(1.4

59)

0.5

0�

0.5

87

(0.8

70)

0.5

0�

1.8

27

(1.0

16)

0.0

8

Mult

idim

ensi

onalPK

1.6

00

(2.0

43)

0.4

41.0

44

(3.0

28)

0.7

30.3

11

(0.9

07)

0.7

3�

0.3

48

(1.4

41)

0.8

1

Enfo

rcem

ent

PK

1.3

94

(1.7

46)

0.4

3�

0.0

56

(1.2

59)

0.9

7�

0.1

10

(0.8

97)

0.9

0�

0.2

57

(1.6

07)

0.8

7

Dem

ocr

acy

atw

arst

art�

0.3

00

(0.1

19)

0.0

2�

0.4

74

(0.1

39)

0.0

0�

0.1

19

(0.0

70)

0.0

9�

0.0

61

(0.0

86)

0.4

8

GD

P/c

apit

a�

0.0

000

(0.0

004)

0.9

90.0

006

(0.0

005)

0.2

1�

0.0

003

(0.0

002)

0.2

6�

0.0

000

(0.0

003)

0.9

2

Illi

tera

cy0.0

17

(0.0

38)

0.6

6�

0.0

21

(0.0

41)

0.6

2�

0.0

13

(0.0

15)

0.4

0�

0.0

09

(0.0

30)

0.7

6

Vic

tory

1.3

83

(1.2

16)

0.2

61.7

27

(1.5

11)

0.2

61.3

07

(0.7

79)

0.1

0�

0.5

06

(1.3

66)

0.7

1

Tre

aty

0.4

49

(1.1

51)

0.7

0�

1.8

53

(1.1

85)

0.1

31.4

73

(0.7

65)

0.0

6�

0.6

09

(1.0

63)

0.5

7

Cost

of

war

�0.7

33

(0.2

56)

0.0

1�

0.5

67

(0.3

18)

0.0

9�

0.1

44

(0.1

59)

0.3

7�

0.1

32

(0.2

28)

0.5

7

Contr

aband

financi

ng

0.9

89

(1.5

50)

0.5

32.5

34

(1.5

34)

0.1

1�

0.5

80

(0.7

80)

0.4

6�

0.8

17

(0.9

04)

0.3

7

Oil

�5.5

86

(1.6

83)

0.0

0�

9.4

27

(1.5

88)

0.0

0�

2.6

71

(0.6

89)

0.0

0�

3.4

76

(1.3

31)

0.0

1

Fail

edpast

agre

emen

t�

1.1

57

(0.8

54)

0.1

8�

0.6

85

(1.2

49)

0.5

90.3

67

(0.4

98)

0.4

7�

0.1

87

(0.8

52)

0.8

3

Per

m-5

conti

guous

0.7

34

(1.6

65)

0.6

6�

0.3

76

(1.6

22)

0.8

20.6

84

(0.5

41)

0.2

10.7

64

(1.0

74)

0.4

8

Per

m-5

form

erco

lony�

2.4

92

(1.0

39)

0.0

2�

2.6

27

(1.1

78)

0.0

3�

1.4

18

(0.7

95)

0.0

8�

1.9

36

(1.3

73)

0.1

7

Const

ant

11.5

70

(3.8

60)

0.0

112.3

53

(5.3

05)

0.0

33.4

68

(2.2

71)

0.1

45.5

49

(2.7

45)

0.0

5

N76

35

84

39

R2

0.3

90.6

90.2

80.4

3

positive for all cases, but not for the subset in which peace has lasted,

suggesting that this positive effect is all about helping to maintain

peace, not in a direct affect on democratization. This suggests at least

some evidence for the dilemma noted above for more robust military

peacekeeping. It may be better able to keep the peace effectively, but

may also impinge on democracy to a larger extent. Again, none of these

effects is significant, so any conclusions must be extremely tentative.

The story is somewhat different, and murkier, when we examine the

Freedom House measures of change in political rights and civil liberties

(columns 3 and 4 in Tables 2.6–2. 8). One year after the war ends, the

effect of all types of peacekeeping is negative among the restricted set of

cases for which peace is holding. Among the full set of cases, monitor-

ing and enforcement missions indicate a positive relationship with

democratization, while traditional and multidimensional missions

indicate a negative relationship. However, any difference among mis-

sion types is most likely simply noise in the data, as none of the

relationships is statistically significant. In other words, as above, peace-

keeping has, if anything a negative direct effect on ‘‘freedom,’’ but this

is offset by a positive effect on stability. Two years out, the coefficients

for all types of mission except monitoring missions are positive in both

sets of cases, but again, we see no significant effects. Five years out, the

indication of any direct effect of peacekeeping on democratization is

negative (as indicated in the restricted set of cases), while the effects in

the full set of cases are mixed. Overall, there is little we can conclude

about differences in the various mission types in terms of effects on civil

liberties and political rights. The results generally echo those for peace-

keeping in general, as discussed above. That is, the direct effects are

generally negative, as indicated in the restricted set of cases, while

differences among mission types in the full set of cases are inconsistent

across time.

As noted earlier, there are too few cases of transitional administra-

tion missions to examine their effects on democracy in the quantitative

analysis. But a quick look at the democratization records of three cases

in this category is revealing. Because the war in Cambodia ended so

soon after the Cold War wound down, it gives us a chance to examine

political changes over a relatively long time period (some fourteen

years). The UN mission in Cambodia ostensibly put that country on

the road to democracy after decades of brutality and war. In the few

years after the war ended in 1991, both Cambodia’s Polity scores and

Peacekeeping and democratization 69

its Freedom House scores show a slight improvement over its pre-war

record (though there is little change from its Polity scores in the year

before the war ends). But Cambodia’s democratization project soon

stalled. Ongoing political violence and intimidation, and then a bloody

coup in 1997 reversed any gains in democracy or freedom. Since then,

Cambodia has democratized only slightly.32

In neither East Timor nor Kosovo do we have as long a time span to

study post-war democratization. The early democratization news in

East Timor seems quite good as the UN transitional authority helped

East Timorese leaders set up fledgling democratic institutions.

Elections for a constituent assembly in August 2001 set the stage for

independence in 2002. In terms of both political rights and civil liber-

ties, there has been enormous improvement, especially relative to the

abuses surrounding the independence vote in 1999 that precipitated the

crisis. While there are no Polity scores or Freedom House rankings for

East Timor, as opposed to Indonesia as a whole, before the war, both

now code the country as relatively democratic. Note that the rest of

Indonesia has also democratized significantly in the last several years,

making it somewhat harder to claim that it has been the UN mission in

East Timor that accounts for democratization.33 However, the

Freedom House country reports give credit to the United Nations

Transitional Administration in East Timor (UNTAET) for helping to

32 Cambodia’s Polity score was as low as �7 and �5 before the war (in the 1970s)but was 1 by 1990. In the early 1990s it remained at 1, dipped sharply in 1997 to�7, then rose to 2 in 1998, where it has remained since. Cambodia’s FreedomHouse ranking for political rights/civil liberties, respectively, was 7/7 (as far ‘‘notfree’’ as the scale goes) during the war, improved slightly to 6/6 in 1991 and1992, moving to the ‘‘partly free’’ ranking of 4/5 for 1993 and 1994, beforesliding over 1995 and 1996 to 7/6 in 1997, and then rebounded only slightly, to6/6 from 1998 to 2000, and 6/5 since then. (Remember that in its originalcodings, higher numbers correspond to less freedom rather than more.) Thecountry reports that Freedom House has issued for Cambodia since 2002describe elections marred by intimidation and violence, a judiciary that is notindependent, and despite some improvements in human rights in recent years, anongoing culture of impunity. Only in 2005 has the formation of a coalitiongovernment put a slight dent in Hun Sen’s autocratic rule (Freedom House2002–2005a).

33 Polity gives East Timor a relatively high democracy score of 6 in 2002 and 2003.Indonesia’s score is 7 for the same years, a vast difference from the�5 in the yearbefore the war ended, 1998. The Freedom House rankings for East Timor beginin 2002 with political rights/civil liberties at 5/3. The situation improves to 3/3 by2003 because of the presidential elections and stays there through 2005.

70 Virginia Page Fortna

‘‘rebuild roads and buildings and set up legislative and other basic

democratic institutions’’ (Freedom House 2002–2005b).

Kosovo, which continues to be administered by the UN, has demo-

cratized much less than East Timor. Because Kosovo is not indepen-

dent, there is no Polity data available, but Freedom House has issued

reports on the territory since 2003. These reports note some improve-

ment in both civil liberties and political rights in 2002 ‘‘because post-

war ethnic discrimination and terror has largely subsided.’’ They note

‘‘generally free and fair [municipal] elections,’’ but limited freedom of

expression because of security concerns and problems with freedom of

movement for ethnic minorities. The ratings stay the same through the

2004 report, but the 2005 report demotes Kosovo slightly (with an

overall status of ‘‘not free’’ rather than ‘‘partly free’’ as before) because

of the Serbian boycott of parliamentary elections following ethnic

violence in 2004.

While it is far too early to tell what the democracy trajectories of East

Timor and Kosovo will be over the long haul, together these three cases

of transitional administration suggest much the same pattern as other

forms of peacekeeping. There is some evidence of moves toward

democracy in the short term, but this is often followed by backsliding

toward autocracy several years down the road.

The effects of peacekeeping on democratization discussed so far

make no distinction between UN missions and peacekeeping under-

taken by other regional organizations or ad hoc groups of states. It is

reasonable to ask whether there is a difference. UN peacekeeping might

be considered more legitimate, both by the belligerents themselves and

by the international community, than peacekeeping conducted by

others. There might also be differences in the efficacy of UN vs. non-

UN peacekeeping. Table 2.9 shows the effects of UN and non-UN

peacekeeping on democratization (in both cases, relative to no peace-

keeping mission). Because the effects of the control variables are sub-

stantially the same as in earlier tables, I omit them here in the interest of

space. However, the results that are shown control for the same set of

variables as in Tables 2.3–2.8. As before, results are shown for both the

Polity and the Freedom House measures, for both the full set of cases

and those in which peace has lasted one, two, and five years out.

In the short term, one and two years after the war ends, it appears

that non-UN peacekeeping is better (or in some cases less bad) for

democratization than UN peacekeeping. Peacekeeping performed by

Peacekeeping and democratization 71

Table

2.9

UN

vs.

non

-UN

pea

cekee

pin

g

Poli

tym

easu

reFre

edom

House

mea

sure

All

case

sIf

pea

cela

sts

All

case

sIf

pea

cela

sts

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

Coef

.(R

SE

)P>

|t|

A.D

emocr

atiz

atio

none

year

out

UN

pea

cekee

pin

g�

0.0

53

(1.1

04)

0.9

6�

0.2

91

(1.3

84)

0.8

4�

0.4

94

(1.0

16)

0.6

3�

1.6

25

(1.1

16)

0.1

5

Non-U

N

pea

cekee

pin

g

1.0

35

(1.1

49)

0.3

71.3

78

(1.4

62)

0.3

50.3

77

(0.6

52)

0.5

7�

0.2

24

(0.7

68)

0.7

7

[Contr

olre

sult

s

om

itte

d]*

——

——

Const

ant

8.6

60

(3.0

22)

0.0

110.0

4(4

.124)

0.0

22.2

00

(1.7

07)

0.2

12.2

15

(2.0

58)

0.2

9

N81

56

84

58

R2

0.3

80.4

50.1

20.1

6

B.D

emocr

atiz

atio

ntw

oye

ars

out

UN

pea

cekee

pin

g0.4

54

(1.1

94)

0.7

1�

0.7

20

(1.5

64)

0.6

50.6

00

(0.8

24)

0.4

7�

0.2

47

(1.1

93)

0.8

4

Non-U

N

pea

cekee

pin

g

0.9

46

(1.1

23)

0.4

00.6

77

(1.5

93)

0.6

70.5

83

(0.6

83)

0.4

00.6

03

(0.7

94)

0.4

5

[Contr

olre

sult

s

om

itte

d]*

——

——

Const

ant

7.9

44

(2.9

98)

0.0

19.6

86

(4.4

52)

0.0

41.9

31

(2.0

69)

0.3

61.6

44

(2.7

55)

0.5

5

N81

49

84

51

R2

0.4

10.4

90.2

10.2

7

C.D

emocr

atiz

atio

nfi

veye

ars

out

UN

pea

cekee

pin

g1.1

35

(1.6

47)

0.5

0�

0.2

93

(2.2

29)

0.9

0�

0.0

23

(0.9

57)

0.9

8�

1.6

89

(1.2

66)

0.1

9

Non-U

N

pea

cekee

pin

g

�0.3

55

(0.9

76)

0.7

2�

0.5

51

(1.1

13)

0.6

30.0

35

(0.6

88)

0.9

6�

0.6

26

(1.2

00)

0.6

1

[Contr

olre

sult

s

om

itte

d]*

——

——

Const

ant

10.8

1(3

.703)

0.0

110.3

9(4

.291)

0.0

23.3

36

(2.1

34)

0.1

33.8

51

(2.4

23)

0.1

2

N76

35

84

39

R2

0.3

80.6

70.2

60.4

1

*T

hes

ere

gre

ssio

ns

contr

olfo

r:D

emocr

acy

at

war

start

,G

DP/c

apit

a,Il

lite

racy

,V

icto

ry,T

reaty

,C

ost

of

war,

Contr

aband

financi

ng,O

il,

Fail

edpast

agre

emen

t,Per

m-5

conti

guit

yand

Per

m-5

colo

ny.Fin

din

gs

for

thes

evari

able

sare

subst

anti

vel

yth

esa

me

as

inpre

vio

us

table

s.

Res

ult

sare

om

itte

dfo

rth

esa

ke

of

space

.

organizations or groups of states other than the UN has either a positive

effect on peacekeeping while UN missions have a negative effect, or

non-UN peacekeeping has less of a negative effect or a greater positive

effect than does UN peacekeeping. In the medium term, five years on,

the same can be said for the Freedom House measure of democratiza-

tion, but the reverse is true for the Polity measure. For the latter, UN

peacekeeping has a positive (or in the restricted set of cases, a less

negative) effect on democratization, relative to non-UN missions. In

all of these regressions, neither the UN nor the non-UN peacekeeping

coefficient is significant, so we should be very careful about drawing

strong conclusions from these comparisons, but they provide at least

weak evidence that UN peacekeeping is no more effective, and may

actually be less effective, than peacekeeping by other groups at institut-

ing democracy in post-war societies.

Conclusion

Many countries do make substantial strides toward democracy in the

aftermath of civil war. But despite the fact that building democracy has

been a major goal of the international community when it deploys

peacekeepers to war-torn societies, peacekeeping has no clear or strong

positive effect on democratization, relative to cases where belligerents

are left to their own devices. This stands in stark contrast to its effects

on the stability of peace. Peacekeepers help keep the peace very effec-

tively, but they do not necessarily foster democratization. The weak

effects of peacekeeping in general on democratization provide evidence

for the dilemmas outlined earlier in this chapter. The negative and

positive effects of peacekeeping on democracy appear to cancel each

other out, as the discussion above of the democratizing effects of both

war and peace would suggest. There is also at least some evidence for

the systemic and temporal dilemmas discussed above. The presence of

outside peacekeepers can help foster democracy, for example, by pro-

viding stability, decreasing levels of mistrust and fostering democratic

institutions, but this external presence is unaccountable politically and

can crowd out the indigenous growth of democracy. Thus we see that

peacekeeping may help foster democratization in the short term but

tends to undermine it over the longer term.

The small number of cases makes it hard to draw distinctions among

the effects of different types of peacekeeping missions – almost none of

74 Virginia Page Fortna

the pe acekeeping effect s in Tables 2.6–2 .8 are statistical ly signifi cant.

But it is striking that we do not see positive effects on democracy for

multidimensional missions, or in the examination of the few cases in

this category, of transitional administrations, which represent the most

focused attempts by peacekeepers to democratize post-war states.

Overall, we see again that the direct effects of peacekeeping on democ-

racy may undermine the positive indirect effects of increased stability,

while short-term gains are often undone in the longer term. It is also

notable that UN peacekeeping efforts, which might be thought more

legitimate than non-UN missions, are no better and may in fact be

worse for democracy than peacekeeping by other organizations or

groups of states.

Examination of the effects of peacekeeping on the establishment of

civil liberties and political rights (the key components of the Freedom

House measure of democracy) suggests that peacekeeping can foster

democracy by fostering stable peace, but that its direct effects on

political rights tend to be detrimental.

It should be stressed that the lack of statistical significance for almost

all of these results means we should be very cautious about interpreting

these findings – the differences among types of missions or between UN

and non-UN peacekeeping could easily be the result of statistical

chance alone.34 Nonetheless, these findings are consistent with the

complex relationship between peacekeeping and democratization

hypothesized above.

It is, of course, difficult to reach strong conclusions about the long-

term effects of peacekeeping when in some cases we have data available

to examine only five years of post-war attempts at democratization –

only time will tell what happens over the next decades. There are also

limits to what can be gleaned from large-N statistical surveys. This

analysis has two key benefits over other types of research methods. It

allows us to cover all cases in a systematic way, and it allows for multi-

variate analysis, which is crucial for evaluating a treatment, such as

peacekeeping, that is not applied at random. But careful analysis of

individual cases is required to flesh out the complex relationships and

34 As noted above, however, because these analyses represent the full universe ofcases we can have more confidence in the results than if we were attempting toinfer patterns from a smaller sample of cases.

Peacekeeping and democratization 75

the dilemmas inherent in post-war democratization. The other chapters

in this volume take on that important task.

The empirical results presented here, while somewhat tentative,

suggest that there are indeed trade-offs and dilemmas inherent in trying

both to maintain peace and to build democracy in post-war societies.

There are trade-offs between the short term and the long term, and

between the need for outsiders to help put a country on the road to

stable peace and democracy on the one hand, and the need on the other

for outsiders to leave political space for indigenous actors to build

homegrown institutions.

The international community has professed two central goals when

it intervenes in war-torn states – to help maintain peace and to foster

the growth of democracy. These have been presented as complemen-

tary tasks: stability is meant to encourage democracy, and democracy is

thought to help maintain stability. The institution of peacekeeping has

been designed to help with both endeavors. But there are contradictions

between these strategies, and peacekeeping is not a cure-all.

The evidence is very strong that peacekeeping is effective at achieving

the first goal. Peacekeeping makes peace much more likely to last. But

peacekeeping has not contributed strongly to the second goal. To the

contrary, the direct effects of peacekeeping (that is, separate from its

effect on stability) have, if anything been negative. This is especially

true for the less institutional or procedural aspects of democracy such

as the expansion of political rights and civil liberties. This creates

fundamental dilemmas for the international community. Because

there is some truth to the conventional wisdom that stable peace fosters

democracy, especially in the short term, the contribution of peacekeep-

ing to stability indirectly fosters democratization. But by intruding in

the domestic politics of countries as they try to recover from civil war,

peacekeepers also crowd out local efforts to build democratic tradi-

tions over the long term. That is, the direct and indirect effects of

peacekeeping on the establishment of democracy are fundamentally

contradictory. The weak and often contradictory statistical results

presented above indicate these competing processes at work.

Peacekeeping is good for democracy, and bad for it at the same time.

Because peacekeeping is clearly and unambiguously good for peace,

countries emerging from civil war are better off when peacekeepers

deploy than when they are left to their own devices. But the interna-

tional community should be under no illusions that peacekeeping will

76 Virginia Page Fortna

be able to transform war-torn states into ideal democratic societies.

There are, rather, fundamental dilemmas and trade-offs between the

goals of peace and democracy, and between the short- and long-term

effects of peacekeeping on democracy.

Appendix: List of cases

Mission type key

0¼ none

2¼ observation

3¼ interpositional

4¼multidimensional

5¼ enforcement

Note some cases experienced more than one mission type. The highest

mission code is reported here.

Name Cease-fire date Peacekeeping mission type

Afghanistan–Mujahideen 25 Apr 1992 0

Afghanistan–Taliban 07 Mar 1993 0

Algeria–FIS/AIS 15 Oct 1997 0

Angola 31 May 1991 2

Angola 20 Nov 1994 4

Azerbaijan–Nagorno K. 31 Aug 1993 0

Azerbaijan–Nagorno K. 16 May 1994 0

Bangladesh–CHT 01 Aug 1992 0

Bosnia 01 Jan 1995 3

Bosnia 14 Dec 1995 5

Cambodia 23 Oct 1991 4

Central African Rep. 25 Jan 1997 3

Chad 11 Aug 1994 0

Congo–Brazzaville 30 Jan 1994 0

Congo–Brazzaville 15 Oct 1997 0

Congo–Brazzaville 29 Dec 1999 0

Congo Dem. Rep./Zaire 17 May 1997 0

Croatia 30 Mar 1994 4

Croatia 12 Nov 1995 5

Djibouti 28 Feb 1992 3

Djibouti 26 Dec 1994 0

Peacekeeping and democratization 77

Name Cease-fire date Peacekeeping mission type

Egypt 15 Sep 1997 0

El Salvador 16 Dec 1992 4

Ethiopia–Eritrea 21 May 1991 0

Ethiopia–ideology 21 May 1991 0

Georgia–Abkhazia 27 Jul 1993 0

Georgia–Abkhazia 14 May 1994 3

Georgia–Ossetia 14 Jul 1994 3

Guatemala 26 Apr 1991 0

Guatemala 20 Mar 1996 3

Guinea–Bissau 26 Aug 1998 0

Guinea–Bissau 02 Nov 1998 0

Guinea–Bissau 03 Feb 1999 2

Guinea–Bissau 07 May 1999 2

Haiti 18 Sep 1994 5

India–Assam 20 Apr 1991 0

India–Assam 17 Dec 1991 0

India–Sikh 31 Dec 1993 0

Indonesia–Aceh 31 Dec 1991 0

Indonesia–E. Timor 25 Oct 1999 5

Iraq–Kurds 01 Mar 1993 5

Iraq–Kurds 15 Oct 1996 5

Iraq–Shia 15 Dec 1993 0

Israel–Palestinians 13 Sep 1993 0

Lebanon 13 Oct 1990 5

Liberia 28 Nov 1990 5

Liberia 17 Aug 1996 5

Mali 06 Jan 1991 0

Mali 31 Mar 1995 0

Moldova 21 Jul 1992 5

Morocco/W. Sahara 06 Sep 1991 2

Mozambique 04 Oct 1992 4

Myanmar–Kachin 15 Oct 1993 0

Myanmar–Karen 28 Apr 1992 0

Myanmar–Karen 15 Jun 1995 0

Namibia 01 Apr 1989 4

Nicaragua 19 Apr 1989 3

Pakistan–Mohajirs 15 Oct 1999 0

Papua New Guinea 15 Mar 1990 0

Papua New Guinea 21 Jan 1991 0

Papua New Guinea 10 Oct 1997 2

78 Virginia Page Fortna

Name Cease-fire date Peacekeeping mission type

Peru 31 Dec 1996 0

Philippines–Communists 31 Dec 1995 0

Philippines–Mindanao 07 Nov 1993 0

Philippines–Mindanao 31 Dec 1990 0

Philippines–Mindanao 02 Sept 1996 0

Romania 23 Dec 1989 0

Russia–Chechnya 01 Jun 1996 0

Russia–Chechnya 23 Aug 1996 0

Rwanda 31 Jul 1992 2

Rwanda 04 Aug 1993 3

Rwanda 18 Jul 1994 3

Senegal 08 Jul 1993 0

Sierra Leone 30 Nov 1996 0

Sierra Leone 07 Jul 1999 5

Somalia 27 Jan 1991 0

South Africa 26 Apr 1994 0

Sri Lanka (Tamil) 07 Jan 1995 0

Sri Lanka (JVP II) 29 Dec 1989 0

Sudan 01 May 1989 0

Sudan 28 Mar 1995 0

Sudan 15 Jul 1998 0

Sudan 15 Apr 1999 0

Tajikistan 27 Jun 1997 5

Turkey–Kurds 20 Mar 1993 0

Turkey–Kurds 01 Sep 1999 0

Uganda–Kony 15 Feb 1989 0

Uganda–LRA 15 Jul 1992 0

UK–N. Ireland 31 Aug 1994 0

UK–N. Ireland 10 Apr 1998 0

Yemen 10 Jul 1994 0

Yugoslavia–Croatia 23 Nov 1991 0

Yugoslavia–Croatia 03 Jan 1992 3

Yugoslavia–Kosovo 12 Oct 1998 2

Yugoslavia–Kosovo 09 Jun 1999 5

Peacekeeping and democratization 79

3 Violence in war-to-democracytransitions

K R I S T I N E H OG L U N D

In the Democratic Republic of Congo, violent clashes in the eastern

Kivu region have continued unabated in spite of the transition process

that began in 2003. In Afghanistan, insecurity has been a prominent

feature of the post-Bonn period, and escalations in violence clustered

around the presidential and parliamentary polls in 2004 and 2005. In

Liberia, violent riots spread through the capital Monrovia in October

2004, in spite of the heavy United Nations peacekeeping presence on

the ground. These examples illustrate that the signing of a peace deal

and the introduction of democracy to a war-torn society do not prevent

killings and atrocities from taking place – at least not in the short term.

This chapter asks two basic questions: Does democratization unin-

tentionally create conditions conducive to violence? And, does the

management of violence sometimes undermine the process of demo-

cratization? By combining insights from the literature on peacebuilding

and democratization, the chapter brings to light central dilemmas

related to violence and security in countries emerging from war that

recently have entered a path toward democracy.

Two arguments about the problematic relationship between vio-

lence, democratization, and peace are put forward and discussed.

Firstly, I argue that democratization creates new opportunities and

motives for violence. The opening up of political space in the area of

political competition (elections), and freedom of expression (media), as

well as reforms of the security sector are crucial to democratization.

However, in the transition period, these institutions are flawed or

ineffective, providing opportunities for substantial maneuvering for

elites who are willing to take advantage of the situation to fan violence.

Secondly, I argue that measures to combat violence – for instance

through the use of coercive measures and the inclusion of spoilers in

politics – can undermine the democratization process. While it is

necessary for domestic and international actors to deal with violence,

the manner in which violence prevention and control is carried out can

80

vary substantially, ranging from the use of force to negotiated solu-

tions. A particular challenge stems from the fact that new forms of

violence emerge as the organized political conflict subsides, rendering

old ways of dealing with violence ineffective.

This chapter analyzes violence and security in relation to five issue

areas – elections, media reform, security sector reform, the use of force

to manage violence, and inclusion as a means to regulate violence. The

purpose is to identify the four types of dilemmas outlined in Chapter 2

of this book as they pertain to these issues. Horizontal dilemmas are

activated when considering whether to include or exclude violence

makers and spoilers in the democratic process as a way of dealing

with violence. By including armed actors in, for instance, transitional

governments, legitimacy is granted to violence makers, with the risk of

continued reliance on violence as a political tool. Vertical dilemmas

relate to the most effective ways of dealing with violence. Coercive

measures may be effective in the sense that they stop violence through

threat and deterrence. On the other hand, combating violence with

violence might undermine the legitimacy of state institutions or peace-

keepers. Systemic dilemmas concern the role of international actors.

A dilemma for state institutions might arise from the need for interna-

tional assistance both in terms of verifying human rights and rebuilding

the state institutions for the protection of human rights. While inter-

national actors are needed to oversee peace agreements and to create a

secure political environment and basic civilian security, the dependence

on international assistance might undermine trust in local institutions.

Finally, temporal dilemmas are particularly crucial when considering

democratization efforts. For instance, reform of the police and military

is needed to democratize the security institutions. However, security

sector reform takes time and might in the meantime leave society

vulnerable to violence. In terms of establishing a secure environment,

lack of action to deal with violence might aggravate instability with the

effect of creating new grievances, and thereby increase the risk of

escalation of violence.

Violence, democratization and democracy

Since the end of the Cold War, democracy promotion has been part of

the peacebuilding strategy of the international community. Also in

cases with comparatively little international involvement – as in

Violence in war-to-democracy transitions 81

South Africa’s transition – the process of democratization has often

overlapped with a transition from war to peace. The rationale for this

approach – to build peace through democracy – is straightforward: a

prevailing assumption both within theory and practice has been that

peace and democracy are mutually reinforcing. While this is true in the

long run an expanding body of scholarly work has contended that

democratization processes – in particular in post-war societies – are

highly conflictual. Indeed, under certain circumstances, democratiza-

tion will make a return to war more likely (Mansfield and Snyder 2001;

2005; Paris 2004; Snyder 2000).

From this perspective, a more reasonable starting point is to assume

that democratization processes are prone to recurring outbursts of vio-

lence constituting severe challenges to war-shattered societies. This pre-

mise is also in line with evidence from peace processes worldwide

indicating that transitions from war to peace are frequently accompanied

by violence – both in new forms and along the old conflict lines. Political

assassinations, violent riots, and extreme levels of crime are threats to the

new political order and to basic civilian security. In extreme cases, failure

to deal with violence can lead to a resumption of war – as was the case of

Angola in 1992 – or to genocide – as in Rwanda in 1994.

A growing literature on conflict resolution and war termination

addresses the problem of violence and security during peace negotiations

and in the post-settlement phase (e.g., Darby 2001; 2006; Hoglund

2005; 2008; Kydd and Walter 2002; Stedman 1997). In particular, the

issues of security sector reform (SSR) (e.g., Call and Stanley 2003;

Cawthra and Luckham 2003), and disarmament, demobilization, and

reintegration (DDR) (e.g., Gamba 2003; Salomons 2005; Spear 2002)

have been accorded a great deal of attention.

There is also an awareness that violence is commonly part of demo-

cratic transitions (Dahl 1971; Huntington 1991). However, while

definitions of democracy typically include some notion of non-violence,

the literature on democracy oftentimes fails to include an analysis of

the relationship between violence and democratization.1 As expressed

by Laurence Whitehead, ‘‘[m]uch theorizing – not only on demo-

cracy but also on democratization – takes the basic security of the

median citizen as a datum, a presupposition upon which liberal and

1 Important exceptions include the edited volumes by Rapoport and Weinberg(2001a), and Cawthra and Luckham (2003).

82 Kristine Hoglund

constitutional systems can be founded, rather than as problematic

social constructs’’ (2002: 163).

The following section brings together insights from these two strands

of research to pinpoint two key problems related to violence and

security in war-to-democracy transitions. Violence commonly refers

to the intentional killing, injuring, or coercion of people, or the destruc-

tion of property, and for analytical purposes a distinction is made

between political and criminal violence. However, in the context of a

war-affected society the boundaries of political and criminal activity

become obscured, making it difficult to differentiate the two forms of

violence clearly or easily. For instance, armed groups may use criminal

violence, such as kidnappings or extortion, to finance the armed strug-

gle. In such cases, criminal violence clearly has political consequences,

although these might not have been intended. It also means that vio-

lence which in non-warring societies would be considered a non-political

criminal act sometimes becomes politicized during the armed conflict.

Bearing these caveats in mind, I will seek to distinguish between crim-

inal and political violence in this chapter.

Democratization: political space for violence

When belligerents close peace deals to end an armed conflict, violence

rarely ends. There are several reasons for the continued existence of

violence in post-war societies, such as the enduring legacies of violence

and the uncertainties inherent in transition processes. In what ways

does the democratization effort itself increase the inclination of actors

to utilize violence?

Scholars have found both new opportunities and new motives for

violence makers in democratization processes (Cawthra and Luckham

2003; Gurr 2000a: 85; Snyder 2000).2 Ideally, increases in democracy in

2 The argument has been thoroughly developed by Jack Snyder (2000) in relationto the perils of democratization and nationalism. For instance, Snyder hasemphasized the tendency of elites to mobilize electorates along ethnic divisionsin order to gain support. The argument is also in line with some findings withinthe research on terrorism, indicating that democracies are more likely thannon-democracies to experience terrorism. William Lee Eubank and LeonardWeinberg, for instance, find that ‘‘the likelihood of terrorist groups occurring indemocracies is three and one half times greater than their occurrence in non-democracies’’ (1994: 426).

Violence in war-to-democracy transitions 83

terms of contestation and participation provide new channels for groups

within society to express discontent. However, under certain circum-

stances, the opening up of political space may result in additional chan-

nels through which violence can be instigated and organized.

The war-to-democracy transition may also create new motives for

violence. Ideally, the peace accord and the democratization process

address some of the grievances and conflict issues that the war was

fought over. However, democratization processes may threaten the

interests of elites: ‘‘democratisation raises political expectations, while

at the same time tends to be actively resisted, often forcibly, by those

whose power and position it threatens’’ (Cawthra and Luckham 2003: 6).

The literature on conflict resolution makes similar claims. The elite may

feel threatened by the peace accord and the peace itself. So-called ‘‘spoi-

lers’’ pose a particular problem: those actors who are not satisfied with

the emerging peace and therefore are intent on disrupting the implemen-

tation of a peace agreement (Stedman 1997). Dissatisfaction with a peace

agreement may stem both from fear about what the peace will bring and

from the extreme positions held by some sectors within society.

Opportunities and motives may also interact, increasing the propen-

sity to use violence among actors in the war-to-democracy transition.

As Ted Robert Gurr (2000a: 86–87) writes: ‘‘In democratizing auto-

cracies . . . national and minority peoples ordinarily feel a loss of

security simultaneously with a transient increase in opportunities for

mobilization and action. New democratic regimes usually lack the

resources or institutional means to make and guarantee the kind of

accommodation that typify the established democracies.’’

This section discusses measures activated as part of a democratization

process that inadvertently may increase opportunities or motives for

violence. It highlights that in a situation where people are used to dealing

with conflict violently and where there are many grievances left unsolved –

the new-won freedoms can be exploited for violent purposes. Violence in

relation to elections, media reform, and security sector reform – three

key features of a democratization process – will be discussed in turn.

Elections

Competitive elections are integral features of democracy.3 Most inter-

nationally sponsored peace agreements stipulate the holding of free and

3 See more on elections in Reilly’s chapter in this volume.

84 Kristine Hoglund

fair elections.4 These elections serve a twofold purpose: (1) to establish

a legitimate and representative democratic government, and (2) to seal

the agreement and to put a decisive end to the war (Kumar 1998c: 5;

Reilly 2003b). Furthermore, to the international community, elections

for new political institutions provide a suitable point in time for redu-

cing its presence in the country (Lyons 2002; 2004; Reilly 2003b).5

Many scholars would argue that elections in post-war societies are

‘‘fundamentally different from those organized under normal circum-

stances’’ (Kumar 1998c: 1). For instance, local conflicts may surface as

refugees return to their communities to vote.

The opening up of political competition, which is a prerequisite for

democratic elections, creates new opportunities to organize violence.

Competitive elections – to be deemed free and fair6 – require funda-

mental political rights such as freedom of movement and freedom of

speech. These rights can be misused as a vehicle for militant political

organization. In particular, the competitive nature of elections may

aggravate existing conflicts and societal cleavages. In countries which

have experienced protracted conflict, political mobilization is likely to

be along the conflict lines (Paris 2004). Differences rather than similar-

ities are brought to the surface to win votes. Moreover, in electoral

processes the use of military metaphors is common: ‘‘the parties wage

‘campaigns’, employing ‘strategies and tactics’. Party faithful are called

‘cadre’, and areas with many supporters are known as ‘strongholds’ or

‘citadels’’’ (Rapoport and Weinberg 2001b: 31). For this reason, the

rhetoric used during the war is easily available to be made use of in an

election campaign to mobilize voters for support.

In competitive elections, groups that have had a dominant position in

society run the risk of losing some of their power if the voters do not

4 Rafael Lopez-Pintor (1997) uses the term ‘‘reconciliation elections’’ to denotesuch elections. There is a large literature on elections and election monitoring ingeneral (e.g., Abbink and Hesseling 1999; Cowen and Laakso 1997; Elklit 1999;McCoy, Gerber, and Pastor 1991), and some studies – mainly case specific – onelectoral violence (e.g., Damm 2003; Fischer 2002; Pausewang, Trondvall, andAalen 2002; Rapoport and Weinberg 2001a). However, there are few studies thatspecifically apply the concepts to the context of war-shattered societies. Forimportant contributions on post-war elections more generally, see in particularthe edited volume by Krishna Kumar (1998c) and Terrence Lyons (2005).

5 This constitutes the so-called ‘‘exit strategy.’’6 For a discussion on the concept of ‘‘free and fair,’’ see Jørgen Elklit and Palle

Svensson (1997).

Violence in war-to-democracy transitions 85

lend them support. Thus, elections may create incentives for political

actors to influence the electoral process through threat, intimidation,

and violence. The situation immediately after the elections is particu-

larly sensitive: will the contenders accept the outcome, or will they

resort to violence unless they emerge as winners? Thus, elections them-

selves can produce outbreaks of violence (Rapoport and Weinberg

2001b: 20).7

Electoral violence has been understood as ‘‘any random or organized

acts that seek to determine, delay, or otherwise influence an electoral

process through threat, verbal intimidation, hate speech, disinforma-

tion, physical assault, forced ‘protection,’ blackmail, destruction of

property, or assassination’’ (Fischer 2002: 8). There are surprisingly

few studies that have focused on the issue of electoral violence in post-

war societies. However, elections in Afghanistan and Iraq in recent

years have brought attention to the difficult issue of establishing a

secure environment that can facilitate free and fair democratic elections

in countries emerging from violent conflict. In many cases the first

elections after a civil war take place in a situation where the parties

are not fully disarmed and demobilized.

There is an abundance of cases where elections have prompted

violence. For instance, in August 1999 the people in East Timor went

to the polls to have their say regarding independence in a referendum

organized by the UN. An overwhelming majority voted in favor of

independence from Indonesia. However, the vote unleashed a brutal

campaign of violence and looting carried out by militia groups

with the support of the government of Indonesia. Prompted by the

disorder, the UN Security Council authorized the deployment of a

multinational force under Australian command in September 1999,

to restore peace and security. In contrast, the elections to the constitu-

ent assembly in 2001 were conducted in a secure environment,

although there had been fears that the elections would cause a new

escalation in violence.

In a democratic process, it is important that elections take place

under conditions that are free from fear and intimidation. Violence

and insecurity may affect the election result, if large sections of the

7 There are different suggestions about what these conditions may be. According toJeong (2005: 105), ‘‘the military balance on the ground can determine whetherelections without demobilization can pose a serious threat to stability.’’

86 Kristine Hoglund

population refrain from voting. Threat and intimidation during the

election campaign may force political contenders to leave the electoral

process. For instance, in Ethiopia, the Oromo Liberation Front (OLF)

withdrew its participation in the elections in 1992 due to threats,

intimidation, and attacks by the Ethiopian People’s Revolutionary

Democratic Front (EPRDF), which subsequently took power.

Elections in 2000 and 2001 were also accompanied by harassment

and violence, causing the opposition to leave the campaign.

To summarize: elections in a post-war society may inadvertently give

rise to new opportunities and new incentives for violence makers to

take up arms or to use threat and intimidation to influence the political

process. For instance, those whose power is threatened by democratic

elections – often termed spoilers – may use violence to disrupt the

transition process or to overthrow the election result. Other actors

may use violence as a means to influence the polls and the election

outcome, by preventing some actors from participating in the election

campaign or by intimidating people from going to the polls.

Media reform

Freedom of expression and press freedom are generally considered

cornerstones of a democratic society. Functioning and diverse media

provide information to the citizens and critical scrutiny of political

issues, crucial to democratic accountability.8 Freedom of expression

is also formulated as a fundamental human right in Article 19 of the

Universal Declaration of Human Rights and in other international law

documents.

In the course of armed conflict, media commonly become polarized

and serve as a tool for propaganda to the conflict parties. Thus, in

countries emerging from war, the media scene is in need of reform. As

part of the transition, efforts are often made to create new laws guar-

anteeing freedom of expression. Reform of the media infrastructure

includes the formation of independent media and widening the range of

media outlets and ownership. These reforms are needed to promote a

diversity of perspectives and to make these accessible to the public.

However, in the words of Ross Howard (2005: 118): ‘‘[i]t is a contra-

diction but true that in the singular most accessible democracies, where

8 On media and economic development, see World Bank (2002a).

Violence in war-to-democracy transitions 87

the media has complete freedom, the strongest impulses of the news

media may still lead it to play a destructive role.’’9

Moreover, in a democratizing context, there is a risk that the reform

of media – in particular their liberalization – can lead to the creation of

irresponsible media outlets of negative and nationalistic propaganda

(Paris 2004; Snyder 2000; Snyder and Ballentine 1996). In the most

extreme case – Rwanda – sustained propaganda campaigns instigated

genocidal violence. In particular, the hate propaganda by the Hutu-

dominated radio station Radio des Milles Collines has been deemed

particularly important in the genocide. Moreover, the newly liberalized

press was extremely polarized, with Hutu-extremists fanning anti-

Tutsi violence and anti-regime press further adding to the fears of the

Hutu of being excluded from power (Snyder and Ballentine 1996).

In most cases the link between media and violence is indirect in the

sense that media create a political environment conducive to polariza-

tion and violence.10 However, there are examples of media in war-

shattered societies spurring violence more directly through biased

reporting. For instance, in 2004 some twenty people died in Kosovo

as violent anti-Serb riots spread across the province. These riots had

been stimulated by the inflammatory media reports about the pre-

sumed drowning of three boys chased down into a river by Serbs near

the ethnic flashpoint Mitrovica. These reports were filled with flaws

and the claim that Serbs had been involved in the drowning proved

inaccurate (Haraszti 2004; Temporary Media Commissioner 2004).

The international community has instructed many initiatives to sup-

port media reform in democratization processes. Training in what has

been termed ‘‘peace media’’ has been gaining ground in peacebuilding

efforts. However, the international community has been criticized for

doing too little to suppress hate media. One way to promote effective

media is to endorse ‘‘codes of conduct’’ and licensing systems (Paris

2004: 198), which are used to regulate media in established democra-

cies. Another tool for the international community is to interrupt

the transmission or dissemination of hate media. For instance, in the

summer of 1997, Serb Radio Television in Bosnia launched an

9 Gadi Wolfsfeld (2001; 2003; 2004) has examined the positive versus the negativeeffect that media can have on peace processes and has specifically analyzed theconflicts in the Middle East and Northern Ireland.

10 For a useful overview of the incitements of ethnic violence, see Donald L.Horowitz (2001).

88 Kristine Hoglund

anti-Nato campaign, which was interrupted through international

intervention (Metzl 1997).

To recapitulate, in the hands of some people, media can be a power-

ful tool to foster hatred and to mobilize people quickly for violence. In

particular, media can be influential in the development of violent riots

and communal violence – oftentimes with ethnic overtures. In the most

extreme cases media have played a crucial part in genocidal violence.

Security sector reform

In the course of an armed conflict, the country’s security forces tend to

gain extreme powers. It is often difficult to maintain the distinction

between upholding law and order within the country, and providing

protection against external threats. The security forces – including the

police, paramilitary units, and other special units – are frequently part of

the counter-insurgency and repression of the opposition. Establishing

civilian control over the military thus becomes important for moving

toward democratic governance.11 The reform of the police – in order

for it to be the sole provider of internal security – is crucial in this

process (Holm and Eide 2000). However, a dilemma arises when the

size of the security sector is reduced at the same time as new threats

surface.

A substantial threat in countries emerging from war is the often-seen

increase in violent crime, including armed robbery and looting, car

hijackings, and domestic violence.12 Several factors are important in

explaining the rise in violent crime in countries such as South Africa,

Guatemala, El Salvador and Iraq. Among those is the lack of effective

institutions to deal with crime, widespread poverty, and easy access to

arms. Profitable economic activities that funded the violent campaigns

11 A dilemma related to security sector reform is that the military might feelthreatened by the civilian initiative for reform and may take action against it(Kohn 1997). There are three areas in which civilian control over the militarymust prevail in order to be democratic: (1) ‘‘clear chain of command under allcircumstances, with the head of the government atop of that chain,’’ (2) ‘‘decisionto begin or end warfare lies in civilian hands,’’ (3) military policy more generally(Kohn 1997: 6–7). See also Huntington (1957) on ‘‘objective civilian control.’’

12 Few studies, have examined criminal violence in connection to war andtransition periods. However, more scholars are paying attention to the topic; see,for instance, Roger Mac Ginty (2006). There are also several studies on crime inspecific post-war societies, such as South Africa. See for instance, AntonietteLouw (1997) and Mark Shaw (2002).

Violence in war-to-democracy transitions 89

of the guerillas or rebels during the conflict are commonly turned into

networks of organized crime. There is also the hypothesis suggesting

that as a result of the war experience, there is a higher tolerance for the

use of violence more generally.13

Reform of the police and the military, in particular in the short term

and in connection to rising levels of crime, makes it difficult for war-

shattered societies to deal effectively with violent crime (Cawthra and

Luckham 2003: 314–315). In El Salvador, the ‘‘[d]isruption of the

internal security system took its toll. The eventual turnover of almost

the entire investigative units meant that networks of informants had to

be reconstructed’’ (Call 2003: 843). There was a similar lack of resources

in, for instance, the attorney’s office due to reform of the judicial system.

The problem was further aggravated by the fact the demobilization

of the armed forces left ‘‘a public security gap, the number of people

circulating under arms decreased from 60,000 (including combatants on

both sides) to only about 6,000 National Police officers’’ (Call 2003:

843). In other cases, as in Guatemala, the absence of an effective police

force has led to situations where the military has been called in to

cooperate with the police to deal with crime. This meant that instead

of much needed resources to reform the police, funds were channeled to

boost up the military budget (challenging the limits to military spend-

ing stipulated in the peace agreement).

Dilemmas related to democratization efforts

Democratization can provide new opportunities for those disgruntled

with the emerging post-war political order. Moreover, the intrinsic

uncertainties of a war-to-democracy process can create new motives

for violence. In fact, in the post-war period, new forms of violence may

emerge, such as electoral violence. The security vacuum resulting from

security sector reform may lead to dramatic increases in criminal

violence.

There are several dilemmas pertaining to reforms for democratiza-

tion, which may have the perverse effect of opening political space for

violence. Media reforms, the holding of democratic and competitive

13 Dane Archer and Rosemary Gartner (1976) made a study of homicide rates incountries that had taken part in combat during international conflict and foundthat violent killings increased substantially in such countries. An explanation forthis result is the so-called legitimization of violence hypothesis, which suggeststhat war legitimizes the use of violence in society beyond the realm of combat.

90 Kristine Hoglund

elections, and security sector reform, all give rise to temporal dilem-

mas. Media reform instigated to support democratic development in a

war-torn society – if changes are introduced too fast or without ade-

quate controls – may in the short term indirectly be used to instigate

violence. The vulnerability of war-torn societies going through changes

in the security sector which are required for long-term democratization

may undermine stability in the short term. The introduction of elec-

toral processes in a war-to-democracy transition also encompasses

issues of timing: should elections be held early with the risk that

extremists come to power, or should they be postponed until the society

is less polarized?

Postponements of reforms or the holding of elections until security is

established can activate the vertical dilemma. How can a legitimate

post-war government be created without elections? Different argu-

ments about the timing of post-war elections have been put forward.

Roland Paris, for instance, contends that elections should not be held

until the conditions for holding free and fair elections are institutiona-

lized. This means the development of moderate political parties and of

a ‘‘judicial mechanism to rule on election-related disputes’’ (Paris 2004:

190). Terrence Lyons, on the other hand, argues that if elections are

postponed, ‘‘opportunities to assist war-termination’’ may be lost

(2005: 96). Thus, from a peacebuilding perspective it may be counter-

productive to postpone elections. For instance, the first post-war elec-

tions in Bosnia – carried out in September 1996 – were not only a

matter of electing government officials but also served to legitimate the

Dayton Peace Agreement.

The vertical dilemma is a potential quandary also in relation to

security sector reform. The change of uniforms – from a repressive

regime or from a rebel group – does not automatically infer an increase

in legitimacy of the police and military. If the same people that have

committed atrocities become part of the new police force, legitimacy of

the new institutions may be undermined. However, at the same time,

the recruitment of former rebels and regime is needed to fill a security

vacuum which might otherwise emerge. Recruitment may also be

important to occupy former soldiers who are potential threats to

peace if left unemployed. The question of eligibility for recruitment

to the police and armed forces becomes crucial. Liberia after the peace

agreement in 2003 is a case in point. The new army does not include

any elements from the old military apparatus, and it is hoped that it

Violence in war-to-democracy transitions 91

therefore will be granted a high degree of legitimacy. However, the

establishment of a new force has been slow and it is therefore believed

that the international community will need to maintain its support to

Liberia’s security sector for a long period. On the other hand, the

recruitment to the police has been considered successful, but there are

fears that the newly trained policemen will immediately fall into the

patterns of the old police.14 The vertical and temporal dilemmas are

closely related in the sense that security sector reform means more than

downsizing, a new name, and a change of uniform. For security sector

reform to be successful, a change in norms and a development toward

more respect for human rights are needed. This takes time.

Moreover, underlying the temporal dilemma is a more fundamental

dilemma related to the legitimacy of the new post-war order. If public

insecurity remains one of the key problems after a peace agreement has

been reached, the ability of local actors to address the issue of violence

properly may be questioned. This quandary will be discussed next.

Countering violence but undermining democracy?

Democracy is built on principles of non-violence. However, violence is

never entirely eradicated from society, prompting reflection on ways in

which violence can be prevented and brought under control. In estab-

lished democracies, violence is controlled and managed by elected

governments, backed by the police and military as the executive insti-

tutions. The legitimate use of force is restricted and granted to specific

agencies. Those responsible for violence are held accountable through

the judicial system based on the rule of law. This section highlights the

difficult task of dealing with violence during a war-to-democracy

transition: a situation that is often characterized by a need for funda-

mental transformation of the institutions responsible for security, as

well as a weak adherence to the rule of law. It points specifically to the

problem that efforts to control and manage violence – both by domestic

and by international actors – can severely undermine the democratiza-

tion process.

The protection of human rights and the provision of basic civilian

security is a centrepiece of liberal democracy. During an armed con-

flict, human rights are abused as part of the violence campaign both by

14 See more on the issue in the ICG report (2005d: 17).

92 Kristine Hoglund

the government and rebels. The human rights abuses can take many

forms, for instance forced recruitment to rebel groups, or torture and

‘‘disappearance’’ of political opponents by the state. Human rights

violations often transcend the purposes of the political conflict and

may, for instance, be used more generally to suppress opposition and

regime critics. Thus, peace deals are struck between parties that lack an

encouraging record in respect for human rights.

In order to uphold basic human rights and security, the belligerents

(the government side and the opposition alike) have to undergo a

transformation both in their purpose and in their means to deal with

conflict without violating basic human rights. Such transformation

does not happen overnight. For this reason, security is a key issue

during both the implementation of peace agreements and the democra-

tization processes. International actors are often brought in to oversee

the peace implementation and to provide security guarantees, because

trust between the belligerents is low. Moreover, local actors frequently

lack the necessary resources to sustain a secure environment.

If violence and insecurity gain the upper hand in a war-to-democracy

transition, a return to armed conflict is one possible scenario. Excessive

levels of political and criminal violence will severely challenge the

legitimacy of the governing institutions. Surveys indicate, for instance,

that disproportionate levels of crime undermine people’s trust in the

state institutions and democracy. Moreover, situations of extreme crime

and delinquency ‘‘can stimulate authoritarian and anti-democratic

reflexes among significant sectors of the electorate’’ (Whitehead

2002: 181). Promises about a restoration of law and order thus appeal

to large sections of a society fraught with violent crime. In El Salvador,

polls indicated increased political support for the hardliners as crime

and violence escalated (Call 2003: 859). In Israel, Benjamin Netanyahu

came to power in 1996 on a political platform calling for toughness

against terrorists.

A consequence of the failure of public institutions to provide basic

civilian security is that in response to criminal violence, alternative

providers of security emerge.15 For instance, vigilante groups have

15 An important issue to consider is who it is that provides security in a society.According to John Darby, the state and its agencies have the primaryresponsibility for maintaining law and order even in cases where they arefraudulent (2001: 44). However, during protracted conflict, the armedopposition can take control over large sections of the territory. For instance, in

Violence in war-to-democracy transitions 93

surfaced in South Africa as a result of the extreme levels of crime.

Moreover, in societies with high crime rates, private security guards

have found a profitable market. Thus, they have taken over security

functions, which in a working democracy should be provided by state

institutions. This privatization of security is problematic for democra-

tization since it erodes the legitimacy of the state institutions.

In the following section I will discuss the precariousness of using

force to control and combat violence in a democratizing context, and

attempts to deal with violence makers through marginalization or

inclusion in the democratic process.

The use of force

In established democracies, the state has the legitimate right to use

coercive measures to enforce the rule of law and to protect the country

from external threats. However, in countries emerging from war, the

state institutions frequently lack legitimacy and the necessary resources

to deal effectively with violence. Moreover, in war-shattered societies,

where violence is often at high – although subsiding – levels, it is

important to find means to control violence, that in themselves do

not undermine the democratization process. It underlines the fact that

‘‘[g]etting violent with violence is . . . risky’’ (Keane 2004: 174).

In several cases, coercive measures to deal with violence have

severely called into question the government’s willingness to uphold

basic democratic principles. In Northern Ireland, a set of anti-terror

legislation was adopted after a bomb in Omagh in 1998 – planted by a

republican splinter group – killed twenty-eight people. The legislation

gave broad powers to the security forces in the pursuit of suspected

terrorists and evoked memories of past abuses committed during the

course of the conflict.

In war-to-democracy transitions, the local institutions to deal with

violence are often weak and in many cases local officials might be

perpetrators of violence and human rights abuses themselves

(Plunkett 2005). For this reason, abuses commonly persist after the

armed conflict has ended. Human rights violations have serious con-

sequences for democratization. For instance, if the violations are car-

ried out by the government agencies, confidence in the post-war order

the mid-1990s the rebel group LTTE in Sri Lanka had de facto control over thenorthern part of the country.

94 Kristine Hoglund

is likely to erode (Kumar 1997: 9). Persistent human rights abuses –

whether by the government or by paramilitaries – will run counter to

attempts to reconcile social cleavages in society. For instance, human

rights abuses in post-war Guatemala have clearly inhibited meaningful

dialogue between the human rights community and the military. The

result is a deeply polarized society.

In many cases, the international community is called in or has

intervened to restore and maintain security in the post-agreement

phase.16 In some cases, the peacekeepers stand mainly as a buffer

between the belligerents. However, if state institutions are weak,

peacekeeping operations take on policing activities to uphold basic

civilian security and to create conditions for peace implementation

and democratization (Viggo Jakobsen 2003). ‘‘Where there is no effec-

tive local independent judiciary or police – the peacekeepers will be

required to take on the task of arrest, prosecution, and trial of serious

offenders’’ (Plunkett 2005: 85). In such instances, the international

forces have to be strong enough to withstand challenges from spoilers

and to deal with violence.

James Fearon and David Laitin cite two examples of violence man-

agement by the international community in post-war societies. In

Tajikistan, Russian peacekeepers were successful in dealing with vio-

lence makers because they were willing and able to utilize violence and

coercive measures themselves. Police units with members recruited

from the Royal Ulster Constabulary (RUC) (with experience from

Northern Ireland) ‘‘were probably the most effective police unit in

Kosovo, although their methods may not always have been fully in

accord with human rights conventions’’ (Fearon and Laitin 2004: 23).

Kosovo also provides plenty of examples of where the international

actors have held individuals viewed as security risks in custody on very

loose grounds and without being allowed a trial. The Kosovo case also

underscores the point that international actors are oftentimes inade-

quately equipped to deal with the new forms of violence that emerge in

post-war societies. Analyses of the riots in Kosovo during the spring of

2004 (in which nineteen people lost their lives) report that the interna-

tional forces under the NATO-led Kosovo Force (KFOR) and the UN

Mission in Kosovo (UNMIK) had neither sufficient capacity in terms of

16 See more on the role of international actors in the chapter on peacekeeping byFortna.

Violence in war-to-democracy transitions 95

number of troops deployed, nor had adequate training and equipment

to respond effectively to the ensuing riots.17

Thus, the use of force is highly problematic in countries emerging

from war. There is a danger that the excessive use of force – by local

security institutions or by international police or peacekeeping mis-

sions – in particular if these are not considered legitimate, may under-

mine the democratization process.

Inclusion as a means to regulate violence

Key players have carried out political assassinations, bomb explosions,

and other deeds to challenge peace accords and a nascent democratiza-

tion process. Armed conflict commonly experiences the rise of orga-

nized groups, warlords, and politicians that gain a position of power

and profit from war-related activities. These actors will only accept

peace and democratization on terms under which their investments and

privileges are not severely threatened. They have often been referred to

as ‘‘spoilers,’’ defined by Stephen Stedman as ‘‘leaders and parties who

believe that peace emerging from negotiations threaten their power,

worldview, and interests, and use violence to undermine attempts to

achieve it’’ (Stedman 1997: 5). While the literature on spoilers generally

has focused on rebels and guerillas, actors connected to the state (the

military, militias and paramilitaries, etc.) can also use violence to

disrupt the new order (Hoglund and Zartman 2006). Spoiler violence

can be directed toward the opponent, in order to instill fear and under-

mine political support for the peace agreement. However, violence to

destabilize a peace agreement may also target moderates within one’s

own group, also referred to as ‘internal feuding’ (Darby 2001).

In many peace processes, deliberate efforts have been made to regulate

violence through the inclusion of violence makers in the political process,

in exchange for promises about the abandonment of violence.18 In

numerous cases, the calling of a cease-fire has been a prerequisite for

the inception of peace negotiations. In Northern Ireland, the British

Government demanded a cease-fire from the Irish Republican Army to

allow their political branch, Sinn Fein, to participate in the peace talks.

17 See, e.g., the reports by the Human Rights Watch (2004) and the InternationalCrisis Group (ICG 2004a).

18 A related issue is whether actors who have committed war crimes should beallowed to participate in the political process. See more in Soderberg Kovacs’chapter on this issue.

96 Kristine Hoglund

In addition, the Mitchell principles were established to regulate the

inclusion of parties in the negotiation forum. These constituted a set of

principles of democracy and non-violence that the parties had to com-

mit to, to be allowed to participate in the peace negotiation process.

Sinn Fein and parties linked to loyalist paramilitary groups were on a

few occasions expelled from the peace talks due to cease-fire violations.

After the peace agreement in 1998, which set up the self-governing

institutions in Northern Ireland, pledges to principles of non-violence

have continued to be a means to regulate violence. Ministers to the

Executive had to reaffirm their commitment to non-violence and a set

of sanctions – ranging from the exclusion of a minister from his post, to

non-payment of fiscal support to the party – were available if it was

judged that a political party was implicated in violence. For instance,

after it was established that the IRA was involved in the serious armed

robbery of a bank in Belfast in December 2004, the Northern Ireland

Secretary of State acted to remove financial assistance to Sinn Fein.

Formal criteria for inclusion in peace processes – often through

cease-fires – may result in new forms of violence, since it is in the

interest of the parties not to be associated with violations of the

principles agreed to. In Northern Ireland, for instance, concerns

about adherence to the cease-fire and the Mitchell principles made

the paramilitaries change violence tactics from the use of punishment

shootings to punishment beatings (Mac Ginty and Darby 2002: 89). In

the wake of the 2002 cease-fire in Sri Lanka, threats, intimidation, and

killings between Tamil paramilitaries and within the LTTE outnum-

bered the violence between the LTTE and the government forces

(Hoglund 2005). This has called into question the value of the cease-fire.

Thus, inclusion in the political process based on commitments to

peaceful means can be an important tool to prevent and manage

violence. However, trust in the inclusion mechanism can be under-

mined if new forms of violence arise which are not as clearly linked

to the parties.

Dilemmas emerging from violence management

The prevention, control, and management of violence in war-to-

democracy transitions through the use of force or through inclusions

give rise to several serious dilemmas for democratization.

The question of the utility and consequences of the use of force

relates both to the vertical dilemma and to the systemic dilemma. The

Violence in war-to-democracy transitions 97

vertical dilemma is highlighted by persistent human rights abuses and

the excessive use of force, which might undermine the legitimacy of

those in power and the state institutions in the post-war order.

The role of international actors in the control of violence may result

in a systemic dilemma, by questioning the ability of local actors to

perform security functions – such as upholding order and the rule of

law – functions they would carry out under normal circumstances.

Moreover, international peacekeepers have to be strong enough and

willing to deal effectively with violence. However, in their pursuit of

violence makers, measures are sometimes taken that undermine the

democratization effort. Moreover, ‘‘[o]nce PKOs leave the world of

neutrality . . . the risk of crime and human rights abuses committed by

internationally sanctioned agents increases’’ (Fearon and Laitin 2004:

34). These activities not only undermine the local actors’ belief in the

democratization process but may also undercut faith in the interna-

tional community.

A question of whether spoilers using violence should be included in,

or excluded from, the democratic process pertains in essence to the

horizontal dilemma. There is a risk involved in incorporating actors

that do not adhere to democratic principles of non-violence and respect

for human rights, because their inclusion might undermine the legiti-

macy of the democratic process. For instance, in the Democratic

Republic of Congo, the transitional government put in place in 2003

is made up of ‘‘former’’ armed groups. These groupings demonstrated a

continued reliance on violence as a means to pursue their economic and

political interests, with a dire security situation as a consequence, in

particularly in the Kivu region.

However, there is also the argument suggesting that the more parties

that are excluded from a peace agreement, the more likely a return to

war (Darby and Mac Ginty 2000; Hampson 1996; Zahar 2003).19

Moreover, violence makers might have been guaranteed a place in

power through the peace agreement. Excluding them from that posi-

tion might ensure a return to war.

19 Recent quantitative research supports the notion that if rebel groups areexcluded from a peace settlement, a return to armed conflict is more likely.Moreover, research also suggests that it is generally the excluded groups thatengage in post-agreement fighting (Nilsson 2006).

98 Kristine Hoglund

Conclusions

This chapter points to two central problems in countries emerging from

war that are simultaneously undergoing democratization and peace-

building. On the one hand, measures taken and reforms initiated to

promote democracy – such as the holding of elections, reform of media

and the security sector – might have the unintended effect of producing

conditions that facilitate or encourage polarization and violence. On

the other hand, efforts to deal with violence – in particular by using

coercive measures and including spoilers in the democratic process –

may undermine the transition to democracy.

The relationship between violence, democratization, and peace-

building covers the four dilemmas highlighted in Chapter 2 of this

book. While more systematic research is needed, this chapter reflects

a first attempt to reach tentative conclusions about the importance of

these dilemmas. In particular, the temporal dilemma – a trade-off

between short-term and long-term achievements – is strongly related

to issues of when reforms for democratization should be instigated. It

was argued that elections, media reform, and security sector reform – if

carried through prematurely – might inadvertently create conditions

conducive to violence. The vertical dilemma – entailing issues over

legitimacy versus efficacy – relates to the fundamental necessity to

manage violence. Unless violence is controlled, the state institutions

responsible for domestic security will lose credibility. If human rights

are abused in the pursuit of violence makers, the democratization

process may be undermined. The horizontal dilemma comes strongly

out in relation to the inclusion versus exclusion of violence makers in

the political process. The systemic dilemma concerns the functions

performed by international actors, which may have the unintended

effect of undercutting the ability of, and trust in, domestic actors and

institutions.

By merging the literature on peacebuilding and on democratization,

this chapter prompts reflection on ways in which democratization

processes in war-ravaged states can be designed in order to alleviate

or prevent the potentially damaging effects of democratization efforts.

Democratization processes have to be accompanied by ambitious efforts

to disarm and demobilize armed actors, to prevent opportunists from

using violence for economic or political purposes. Convincing argu-

ments have been put forward by scholars for the case that democratic

Violence in war-to-democracy transitions 99

reform has to be delayed until strong institutions have been established

that can address and control the hazards of democratization in war-

shattered societies (Paris 2004). At the same time, a war ending opens

up opportunities for democratic reform which may also help legitimize

the post-war political order. Ideally, the peacebuilding process and

democratization can be carried out in tandem and the question is rather

a matter of design than about timing. For instance, peaceful demonstra-

tions are a democratic outlet for frustration and demands for political

change. However, to prevent peaceful demonstrations from turning

violent, they have to be backed up by adequate security arrangements.

In addition, this chapter calls for fresh thinking with regard to ways

of dealing with violence. In particular, successful violence management

requires strategies that differ from those that are used in stable, demo-

cratic societies. The reason is twofold. Firstly, in post-war societies,

uncertainty is exacerbated by the fact that the rules of the game in the

political sphere are not set but are still under development. Secondly,

violence-induced crises become focal points where the spirit and sub-

stance of peace and democracy are put to test. Thus, strategies adopted

by the international community and the local actors to prevent and

manage violence will have to take into account the specific context in

which it emerges. Or in Jeong’s words: ‘‘[a]n assessment of the extent

and nature of local violence is necessary when intervention strategies

are being designed’’ (Jeong 2005: 58).

When the armed struggle comes to an end, new forms of violence

emerge or are reintroduced, as was seen in Kosovo, Sri Lanka, and

Northern Ireland. As the regular fighting between rebels and govern-

ment forces subsided, violence on the streets and communal violence

has remained a feature of these countries in the wake of cease-fires.

Often a shift from military thinking to policing is needed to tackle

violence in post-war societies. This means that local and international

actors involved in violence management require training and equip-

ment to use anti-riot and policing techniques. The deployment of

international police missions is one response to this need. However,

their usefulness shows mixed results and more research is needed on the

design and operation of international police missions.20

20 Important exceptions include the edited volumes by Oakley, Dziedzic, andGoldberg (2002) and Holm and Eide (2000). On related issues from a US policyperspective, see Robert M. Perito (2004).

100 Kristine Hoglund

One of the most important distinctions to consider is whether the

violence is committed for political or criminal purposes. Criminal

violence has to be countered with traditional police work firmly

anchored in the rule of law. Thus, if the police are to be successful in

their effort to combat crime, this has to be accompanied by a function-

ing judicial system. With regard to political violence, inclusion is one

way of dealing with it. Through participation in the political arena,

even the more extremist parties have a vested interest in upholding the

political system. However, as discussed in the chapter, inclusion may

also undermine faith in the new political system.

Negotiations at the local level can be important to deal with violence

efficiently and can be used as an alternative or complement to the use of

force. Negotiations can be institutionalized through the formation of

peace committees or monitoring missions as in the cases of South

Africa and Sri Lanka. The Sri Lanka Monitoring Mission (SLMM) –

a civil mission established to oversee the 2002 cease-fire between the

Liberation Tigers of Tamil Eelam (LTTE) and the Sri Lankan

Government – was initially instrumental in solving disputes regarding

violence at the local level. However, in 2006 Sri Lanka de facto

returned to a state of war.

A final comment relates to a more fundamental dilemma pertaining

to war-to-democracy transitions. Political violence is frequently a

response to too little democracy. A transition period to establish a

new political order is therefore needed to overcome violence. While

violence during the transition process in some cases may threaten to

derail the transition, in other cases the transition survives all attempts

aimed at preventing it from proceeding. Thus, sometimes the two

processes seem to be closely linked and can instigate degenerative

spirals, with more violence sending countries down a trajectory of

instability and crumbling democracy. For this reason, a key question

for future research is to understand when a transition can cope with

violence, and when violence will continue well beyond the transition

and threaten long-term stability.

Violence in war-to-democracy transitions 101

PART II I

The political process

4 Power sharing: former enemiesin joint government

A N N A K . J A R S T A D

More than a decade after the peace agreement for Bosnia and

Herzegovina was signed in Dayton, Ohio, the country remains divided.

The peace agreement served its main aim – to end the 1992–1995

conflict – but its power-sharing provisions are now seen as an obstacle

to peacebuilding and democratization. The political representatives of

the three main ethnic groups have not managed to govern jointly.

Instead, the international community’s High Representative has

imposed laws and removed several politicians from office. For the

sake of efficacy, these measures have substituted normal procedures

for accountability and democratic legitimacy. The extensive external

control has undermined local ownership. Furthermore, the political

system provides no incentives to form cross-ethnic and moderate

political parties. The international community has therefore suggested

a need for constitutional changes. The pertinent issues include the

tripartite presidency and decentralization to the two entities, the

Bosniak-Croat Federation and Republika Srpska. Revisions toward

an integrated, centralized non-ethnic parliamentary democracy with a

single president have been proposed. However, these changes upset the

power balance that has regulated the conflict. Removing guaranteed

governmental position threatens the power of the present elite. Hence,

Serbian leaders want to maintain Republika Srpska, and many Croats

believe that they should also get their own entity. In April 2006 the

parliament voted against constitutional changes. The case raises perti-

nent questions about governance in post-war societies. In what ways

does power sharing facilitate, or obstruct, the transition of divided

societies toward a secure democracy?

In recent years peace agreements have stipulated the inclusion of war-

ring parties in government in countries such as Burundi, the Democratic

Republic of Congo, Liberia, Sudan, and Nepal. Previous research pro-

vides explanations for why parties find power sharing acceptable.

Within the field of conflict management, it has been demonstrated

105

that combatants are much more likely to sign an agreement if it reduces

uncertainty by the inclusion of guaranteed positions in the future

government. Scholars on democratic theory have advocated power

sharing as a school in democracy, socializing opponents into compro-

mises and moderation, and as the only option for democratic govern-

ance for many divided societies. This leads us to expect that power

sharing contributes to peace as well as democracy.

At the same time, case studies illustrate that power sharing may be a

source of instability, ineffective governance, and violent conflict (e.g.,

Rothchild 2005b). How can this puzzle be understood? My interpreta-

tion is that previous research on power sharing has underestimated the

long-term negative consequences of power sharing on both democracy

and peace. To date, scholars of conflict management have perceived of

power sharing as a tool for short-term peace. It has been suggested that

because contending parties cannot trust that the other side will uphold

an agreement on democratic governance after a winner-take-all elec-

tion, parties to a peace deal are likely to demand some form of power

sharing (Walter 1999). However, after a peace deal is signed, power

sharing may have negative effects on democratization as well as

peacebuilding.

Scholars of democratic theory, on the other hand, suggest that power

sharing promotes democracy, while assuming that it also produces

peace. Advocates of power sharing as a means for improving democ-

racy suggest that government by many is more legitimate than mere

majority rule (Lijphart 1999). Iterative cooperation in a grand coalition

is expected to promote tolerance, increased trust, the development of a

democratic culture and societal peace (Lijphart 1994). However, this

field of research has failed to take into account how the special condi-

tions of war-torn societies affect the functioning of power sharing.

The lack of integration of the two discourses on power sharing has

led to a neglect of a deeper understanding of potential dilemmas for

simultaneous and long-term democratization and peacebuilding.1

1 One of the few studies that indeed discusses the dilemmas of power sharing inrelation to both democratization and peace is Roeder and Rothchild 2005.However, in that volume ‘‘power sharing’’ refers to a variety of arrangements forsharing and dividing power between groups, including both formal provisions forpolitical, military, economic, and territorial arrangements as well as informalpractices of cross-communal cooperation. In the present chapter, I focus on thedilemmas related to peace accords stipulating guaranteed positions in the future

106 Anna K. Jarstad

I argue that power sharing in war-torn societies entails choices between

promotions of peace or democracy. Four mechanisms of power sharing

condition the prospects for peace: inclusion of warring parties, intra-

group contestation, international dependence, and the leveling of

power relations. At the same time, power sharing can affect democra-

tization negatively in at least four ways: by exclusion of moderate

elites, by lack of popular support, by external intervention preventing

local ownership of the political process, and by freezing ethnic division

by group representation.

Each of these eight attributes of power-sharing accords corresponds

to one of the four dilemmas, outlined in Chapter 1 of the book. Firstly,

expanding the political center by ensuring inclusion in parliament for

key representatives of society may facilitate both peacebuilding and

long-term democratization. However, many power-sharing arrange-

ments after civil war are based primarily on inclusion of representatives

of the former warring parties, which do not always give up their violent

tactics despite a peace agreement. While key figures in the conflict are

often included in such power-sharing arrangements, moderate actors

are often excluded from a share of power. These aspects speak to the

horizontal dilemma of which parties to include – and exclude – to make

power sharing conducive to both peace and democracy.

Secondly, the focus on a small group of elites activates the vertical

dilemma of efficacy at the cost of legitimacy. This may give rise to

violent splinter groups, as well as weak support for the power-sharing

system among the civilian population, which may affect the prospects

for peace and democracy negatively. Thirdly, power-sharing accords

may entail a systemic dilemma of local versus international ownership

and responsibility to uphold the peace agreement. In some cases, a

third-party guarantor is deemed necessary for the functioning of a

power-sharing government and to prevent a return to war. At the

same time, such intervention may undermine local ownership of the

political institutions.

Fourthly, power sharing often implies a leveling of power between

contending groups. Over time, there is a risk that the agreement does

not reflect the perceived power relations and some actors may even

challenge it by force. Furthermore, group representation can be an

legislative and executive branch of government, i.e., a type of formal politicalpower sharing.

Power sharing 107

obstacle for individual choice by restricting the possibility to change

political orientation over time, or voting for candidates of different

ethnic affiliations. These aspects hamper the processes of democratiza-

tion and peacebuilding and give rise to the temporal dilemma of power

sharing. Below, I begin by discussing the concept of power sharing.

Then, the effects of power sharing on the conflict dynamic as well as on

democratization are discussed separately. The chapter concludes by

discussing the dilemmas of power sharing and trade-offs between

democratization and peacebuilding.

The meaning of power sharing

Conceptual confusion has hampered research on power sharing. Two,

actually separate, strands of research use the term ‘‘power sharing,’’

often without recognizing the differences in terms of democracy and

conflict management. However, power sharing stipulated in part of the

conflict-management literature differs from power sharing in accor-

dance with democratic theory. Because of different definitions, there is

little overlap between the characteristics, the cases and the mechanisms

of these two concepts of power sharing.

One strand of research stems from the field of conflict management.

In this discourse the main function of power sharing is to end violence.

To lay down arms, parties often demand guaranteed positions in the

future government.2 Barbara Walter explains the function of power

sharing as a mechanism for solving the commitment problem in a con-

text of severe distrust and vulnerability. She outlines several reasons why

combatants have difficulty in credibly committing to a peace agreement.

The concessions involved in a peace deal increase the parties’ vulner-

ability and limit their ability to enforce the treaty’s other terms. Once the

parties have laid down their arms and surrendered territory, they

become vulnerable to surprise attacks. Also, the parties cannot trust

each other to implement the agreement. Therefore, provisions for

democratic governance do not by themselves settle the issue of uncer-

tainty. With less control over resources, parties may fear that the other

2 Alternative demands often concern some form of transfer of power to smallerunits, such as decentralization, federalism, autonomy, separate legal systems, etc.Division of power based on territorial units is sometimes called ‘‘territorial powersharing’’ (see, e.g., Hoddie and Hartzell 2005). In this volume, however, powersharing refers to a specific type of political pact, namely power-sharing accords.

108 Anna K. Jarstad

party will form a one-party state if it wins the election, and that those in

power would exploit minority and opposition groups. In the absence of

a strong civic culture, the population may not reject parties advocating

authoritarian control to foster order and economic advancement. For

this reason, the losers of the first post-war elections cannot count on

another opportunity to gain power. Under such circumstances, Walter

(1999) argues, each side is likely to demand some form of power

sharing as the price for peace. She has demonstrated that warring

parties are 38 percent more likely to sign an agreement if it includes

guaranteed positions in the future government (Walter 2002: 80).

In this discourse, the main function of power sharing is to end vio-

lence, not to build democracy. Here, power sharing denotes all types of

sharing and dividing power between former foes, with less emphasis on

democratic representation and elections (Hartzell and Hoddie 2003;

Rothchild 2005a; Spears 2000; Walter 1999). Ian S. Spears writes that

power sharing ‘‘can be compatible with democracy’’ (2000: 105, my

italics). However, power sharing is sometimes constructed as an alter-

native to competitive elections (Spears 2000: 108).

Among quantitative research on war endings, Matthew Hoddie and

Caroline A. Hartzell’s works are well cited. They suggest that of the

total of thirty-eight civil wars ended by negotiated settlement between

1945 and 1998, only one did not include provisions for power sharing.

Furthermore, they conclude that the more power sharing, the higher

the likelihood that peace will endure (Hartzell and Hoddie 2003: 319).

However, in contrast to the most common notion of power sharing,

guaranteed positions in the government are not a necessary component

of this definition. Rather, power sharing denotes any type of institution

dividing or sharing political, economic, territorial, and military power.

Military power sharing refers, for example, to provisions allowing

antagonists to remain armed or retain their own armed forces. In

order to qualify as political power sharing, it is sufficient that the

peace deal provides for electoral proportional representation or admin-

istrative proportional representation (e.g., appointment of representa-

tives of warring groups to courts).3 This means that cases without

political pacts are included as cases of power sharing (Hartzell and

Hoddie 2003; Hoddie and Hartzell 2005). Consequently, many pivotal

3 Bumba Mukherjee (2006) also regards proportional representation as a sufficientcriterion for power sharing.

Power sharing 109

issues, such as the impact of guaranteed positions in government on

peace and democracy, or factors associated with the creation of poli-

tical pacts after civil war, cannot be assessed based on their findings.4

To the other strand of research, which takes its starting point in Arend

Lijphart’s theory on consociationalism, political pacts are central. Here

the main issue is how to make democracy function in societies divided

along ethnic lines. This means that the theory of consociationalism

primarily addresses conflicts that are clearly ethnic. Lijphart (1968)

coined the concept of consociational democracy in 1968 to denote an

institutionalized form of democratic conflict management for divided

societies. Majoritarian electoral systems are inapt, he writes, since they

presuppose shifting majorities in parliament and fairly similar policies

of major parties in order not to exclude the other parties’ interests.

However, in divided societies, the political parties diverge to a great

extent. Where people vote along ethnic lines, political parties repre-

senting ethnic minorities have no chance of ever forming a majority,

and shifting majorities in parliament are therefore unlikely. Under

such conditions, Lijphart holds that majoritarian rule is not only

undemocratic, but also dangerous and risks resulting in civil strife

(1999: 31–33). For countries such as Lebanon in 1985, Lijphart writes,

‘‘the choice is not between consociational and majoritarian democracy,

but between consociational democracy and no democracy at all’’

(1985: 13).

To avoid partition or majoritarian dictatorship, Lijphart advocates a

group-based form of democracy, which he calls consociational democ-

racy. The two main components are grand coalition, implying that all

rival groups should be included in government and autonomy for each

ethnic segment in all matters not of common concern. Two additional

features are mutual veto rights and proportionality in political repre-

sentation, civil service appointments, and the allocation of public funds

(Lijphart 1993: 188–189).

Donald L. Horowitz, Timothy D. Sisk, Benjamin Reilly, Andrew

Reynolds, and others have recognized that there are also other forms

of democratic power sharing (e.g., Horowitz 1985; Reilly 2001;

4 Power-sharing arrangements defined as political pacts remain largely unexploredin quantitative research. Walter’s (2002) study is an exception. However, thisstudy is limited in many regards as it does not cover the entire post-Cold Warperiod, excludes minor conflicts, and only includes those peace deals where all orthe main warring parties are signatories.

110 Anna K. Jarstad

Reynolds 2002; Sisk 1996). The integrative approach to power sharing

eschews ethnic groups as the organizing principle for democracy.5

Instead, incentives for moderation and cooperation across ethnic

divides are engineered. These constitutional designs can include major-

itarian as well as non-majoritarian forms of electoral design. However,

proponents of the consociational as well as the integrative approach

agree on the baseline that some form of joint rule is the only option for

democratic governance in divided societies.

Thus, in the conflict-management discourse, power sharing is seen as

a mechanism to manage the uncertainty involved in a peace process – if

need be, as a substitute for elections – while research based on demo-

cratic theory treats power sharing as a mechanism to foster moderation

and to improve the quality of democracy. This means that researchers

of both schools advocate power sharing for war-shattered societies,

albeit for different reasons. However, the lack of integration between

the two discourses means that there is limited knowledge of the long-

term consequences of power sharing in societies emerging from war.

Of particular importance for the issues of democratization and peace-

building in war-torn societies are provisions for guaranteed positions in

government, stipulated in a peace accord. Table 4.1 provides a list of

examples of such power-sharing accords concluded in the post-Cold

War era.6 A power-sharing accord (PSA) is a political pact between

contending parties which formally outlines how power is to be shared

in the legislative and/or the executive branch of a future government.

It can be noted that power-sharing provisions have been stipulated to

manage several types of conflicts. Some power-sharing accords reflect

ethnic divisions. For example, the Dayton agreement stipulated power

sharing among Serbs, Croats, and Bosniaks in Bosnia and Herzegovina.

The Burundi agreement provided for two types of ethnic power shar-

ing: during a transitional period ethnopolitical parties shared power,

whereas the permanent constitution fixed the ratio between Hutu and

Tutsi representatives. By contrast, other peace accords do not make any

5 ‘‘Centripetalism’’ denotes a similar approach. Sisk uses this term to refer toelectoral engineering (in line with Horowitz) aiming for moderation andcompromises by reinforcing ‘‘the centre of a deeply divided political spectrum’’(Sisk 1995: 19). Reilly uses a wider definition including electoral incentives,arenas for bargaining and centrist, aggregative political parties (2001: 11).

6 Less recent examples of power-sharing accords include the London Accords 1959(Cyprus) and the Lancaster House Settlement 1979 (Zimbabwe).

Power sharing 111

reference to ethnic groups, for example the peace accords in Angola,

Cambodia, the Ivory Coast, Liberia, and Sierra Leone. By integrating

the two discourses on power sharing, ethnic as well as non-ethnic PSAs

can be taken into account. The long-term consequences of many of

these PSAs are yet to be seen. However, the theoretical arguments

outlined below suggest that the main functions of power-sharing

Table 4.1 Examples of recent power-sharing accords

Conflict location Name and year of power-sharing accord

Afghanistan Mahipar Agreement 1996

Angola Lusaka Protocol 1994

Bangladesh Chittagong Hill Tracts Peace Accord 1997

Bosnia and Herzegovina Dayton Peace Agreement 1995

Burundi Arusha Agreement 2000

Cambodia Paris Agreement 1991

Chad Tripoli Accord 2006

Colombia Acuerdo Final Gobierno Nacional 1991

Comoros Agreement on the Transitional Arrangements 2003

DRC Pretoria Agreement 20027

Guinea-Bissau Abuja Peace Agreement 1998

Ivory Coast Linas-Marcoussis Agreement 2003

Kosovo Rambouillet Agreement 1999

Lebanon Ta’if Charter 1989

Liberia Accra Comprehensive Peace Agreement 2003

Mali The Pacte National 1992

Mexico San Andres Accords 1996

Nepal Comprehensive Peace Agreement 2006

Rwanda Arusha Accords 1993

Sierra Leone Lome Accord 1999

South Africa Interim Constitution Agreement 1993

Sudan Naivasha Comprehensive Peace Agreement 2005

Darfur Peace Agreement 2006

Tajikistan Moscow Declaration 1997

Uganda Yumbe Peace Agreement 2002

Source: Uppsala Conflict Data Program, March 2007. Selection made with

assistance of Ralph Sundberg.

7 The formal title is the Global and Inclusive Agreement on the Transition in theDemocratic Republic of the Congo.

112 Anna K. Jarstad

agreements – inclusion, elite-oriented regulations, international guar-

antee, and the leveling of power between groups – can have negative

consequences for peacebuilding as well as on democratization.

Power sharing: effects on conflict dynamics

The logic of power sharing as a way to manage violent conflicts can be

understood in terms of a basic two-player game. In brief, conflict is

then seen as a result of a situation where one party strives for total

political control whereas the other demands partition. If one of the

parties has its way, it will gain a hegemonic position. If the other party

has its way, the game will result in partition. Under such circumstances,

or when both parties strive for total political control, the only solution

to manage conflict is joint rule (Jarstad 2001).

For this reason, power sharing is attractive to peace negotiators. At

the same time, there are several reasons why the warring parties are

reluctant to end conflicts by power sharing and why PSAs fail.8 A first

reason concerns the issues of insecurity and trust. One interpretation of

Lijphart’s theory is that it assumes that the provisions of consociation-

alism change the value of cooperation for the parties involved. In a

similar vein, Walter suggests that power sharing solves the commit-

ment problem. However, in many war-shattered societies, power shar-

ing might be inadequate to mitigate the parties’ vulnerability during the

peace process.9 Under such circumstances, the parties cannot trust each

other to uphold an agreement on power sharing.

Secondly, power sharing entails great concessions for many parties.

Per definition, power sharing levels the power relations between con-

tending groups. A strong party is likely to be reluctant to accept a peace

deal which renders all parties equal powers. In the choice between

8 A recent data collection, the IMPACT (Implementation of Pacts) data set,demonstrates that in the conflicts where an agreement of political power sharingwas reached and implemented within five years, nearly half of all cases saw therecurrence of warfare within this time period (Jarstad and Sundberg 2007).

9 In game-theoretical terms, this is the case when the rank orders of the parties’preferences result in a Prisoner’s Dilemma. Consociationalism provides nomechanism to solve the Prisoner’s Dilemma. Therefore, as Walter demonstrates, acombination of third-party security guarantees and power-sharing pacts are criticalto the peaceful resolution of civil wars (Walter 2002). For a review of the role ofthird parties in relation to power sharing, see the introduction by John MacGarryand Brendan O’Leary 2004).

Power sharing 113

democratic options, the larger party is likely to consider a majoritarian

electoral system more democratic and legitimate than a power-sharing

system, as the latter gives small groups too a share of power, and in

several cases even veto rights.

But a minority group may also be reluctant to sign a power-sharing

deal. Territorial control of part of a state may be preferable to the

inclusion in a power-sharing government. This is particularly true for

regions with valuable resources such as oil or diamonds.10 Furthermore,

inclusion in central government implies accountability also for decisions

that go against your own group. Thus, leaders may fear that if they accept

a power-sharing deal, their followers would see them as traitors.11

Thirdly, even in cases where war ends in a power-sharing accord, the

risk of a resumption of violence remains. This is not always a conse-

quence of power-sharing provisions, but of conditions which obstruct

any form of peace deal. For example, lingering violence may prevent

power sharing from taking root. This was the case in Angola 1994,

where violence continued even after the Lusaka agreement was signed.

However, it would be incorrect to argue that power sharing led to the

resumption of war in 1998, as the agreement was not fully implemen-

ted until many years later. In Sierra Leone in 1999, on the other hand, it

was violence that caused power sharing to break down. In this case,

power sharing was a condition for the peace accord and the subsequent

international peace operation, which eventually contributed to the

marginalization of the Revolutionary United Front (RUF).12

10 In recent years there has been an upsurge of research on what has become knownas the greed and grievance motives for conflict (see, e.g., Ballentine and Sherman2003; Berdal and Malone 2000; Collier et al. 2003). Based on several case studies,Karen Ballentine and Jake Sherman demonstrate that the combatants’opportunities for mobilization and/or incentives for self-enrichment were not theprimary causes of these conflicts. Nevertheless, access to natural and financialresources does in some cases create serious impediments to conflict resolution.‘‘Lootable’’ resources such as diamonds and illegal narcotics have been found toprolong non-separatist insurgencies, while ‘‘unlootable’’ resources, such as oil, gas,and deep-shaft mineral deposits are associated with separatist conflicts (Ballentineand Sherman 2003). Hence, there is reason to believe that natural resources impacton the parties’ preferences, the type of power sharing agreed on, and the success ofpower sharing in terms of conflict management.

11 This argument is, for example, illustrated by the case of South Africa, where theNational Party withdrew from the power-sharing agreement in 1996 (Sisk andStefes 2005).

12 See the chapter by Soderberg Kovacs on the failed transformation of RUF fromwarring group to political party in this volume.

114 Anna K. Jarstad

Fourthly, power sharing affects the intergroup relations in war-torn

societies. Below, I nuance the analysis of how power sharing affects the

conflict dynamics. I discuss the impact of four features of power-sharing

accords: inclusion of warring parties, intra-group contestation, inter-

national dependence, and the leveling of power relations between

groups. Although these four mechanisms can have positive effects on

conflict dynamics, I propose that under certain conditions they can also

affect intergroup relations negatively.

Inclusion of warring parties

The horizontal dilemma of power sharing is triggered by the choice of

whom to include in a power-sharing arrangement. In many conflicts,

power sharing is the key for reaching a peace deal at all. A peace deal is

per definition an agreement between contending actors. Negotiated

settlements result from compromises between warring parties, who

often maintain the option of resumption to war should the costs of

peace outweigh the prospective gains from future outcomes. For war-

ring groups who see little chance of winning the war on the battlefield,

a stake in the government, stipulated by a PSA, may work as an

incentive to lay down arms.

The African National Congress (ANC) in South Africa is only one

example of a former rebel group that has agreed to give up violent

tactics in exchange for a transition to democracy that featured power

sharing, at least for a transitional period. The armed struggle against

the apartheid regime in South Africa was ended by a series of pacts that

included ‘‘sunset clauses’’ by which power sharing would eventually

give way to majoritarian democracy. However the transition period

was far from peaceful. In the transitional period from 1990–1994,

large-scale violence raged throughout the country. Although political

violence ended in most parts of the country when the ANC-dominated

Government of National Unity came to power in 1994, the KwaZulu-

Natal province continued to be marred by political violence as local-level

power struggles prevailed. The interim constitution stipulated transi-

tional power sharing among all political parties that won substantial

votes in elections, but the erstwhile governing party, the National Party,

voluntarily withdrew from the Government of National Unity in May

1996. Nevertheless, the reliance on sunset clauses provided for a

smooth end to formal power sharing and the permanent constitution

Power sharing 115

was adopted in 1996 that does not feature formal guarantees of power

sharing. Informal power sharing continues to be vital for moderation

and compromise. Thus, power sharing is seen as the key to the com-

paratively successful transition from violent political conflict to peace

in South Africa (Sisk and Stefes 2005).

In Northern Ireland lingering violence and paramilitary organizations

have been the main obstacles for power sharing stipulated by the 1998

Good Friday agreement. The political party Sinn Fein has never officially

recognized its military wing, the Irish Republican Army (IRA), but has

widely been held accountable for its violent tactics. After the agreement

was signed, and Sinn Fein became part of the provincial government, a

splinter group called Real IRA has been responsible for a number of

violent incidents, most notably a bomb in Omagh, which killed twenty-

eight people (Belloni and Deane 2005). The IRA has now officially

declared an end to its armed campaign and Sinn Fein, led by Gerry

Adams, and the Democratic Unionist Party, headed by Ian Paisley,

entered a power-sharing government in May 2007.

However, exclusion of violence makers can also trigger violence. An

agreement which stipulates guaranteed positions in government only

for some of the warring groups provides a strong incentive for excluded

groups to resort to violence, with the purpose of gaining leverage for

future negotiations or winning militarily.13 The number of sufficient

signatories is a particular problem in conflicts with many warring

groups where there are reasons to fear that groups excluded from

power sharing will use violence to fight their way into the power-

sharing deal.

The 2000 Burundi PSA is a case of partial inclusion. After the ‘‘Arusha

Agreement for Peace and Reconciliation in Burundi’’ was signed on

August 28, 2000, four non-signatories continued to fight. For several

years the rebel movement the National Council for the Defence of

Democracy-Forces for the Defence of Democracy (CNDD-FDD) refused

to participate in the peace talks. Under strong pressure from the media-

tion team, led by Nelson Mandela, CNDD-FDD eventually laid down

their arms. The Pretoria Protocol of October 8, 2003 allotted them,

among other provisions for power sharing, four seats in the transitional

cabinet and fifteen seats in the National Assembly. The Global Peace

13 On the risk that exclusion leads to violence, see Darby and Mac Ginty 2000;Hampson 1996; Nilsson 2006; Zahar 2003.

116 Anna K. Jarstad

Accords of November 16, 2003 stipulated CNDD-FDD should get three

ministry positions and the Minister of State (Bentley and Southall 2005;

Rothchild 2005b).14 The last remaining rebel group, Palipehutu–FNL,

was included in peace negotiations in 2006.

Another case of partial inclusion is the Sudan 2005 accord which was

designed to end the conflict in the southern part of Sudan. The accord

did not aim at regulating the conflicts in other parts of Sudan.

However, it was incorrectly assumed that other groups in the country

were represented either by the government or the Sudanese People’s

Liberation Movement (SPLM). The 2005 accord unintentionally exa-

cerbated the conflict in the Darfur region and increased the risk for

conflict also in the eastern part of the country by excluding these

regions from the political and economic power sharing. The Darfur

crisis erupted when the peace negotiations concerning southern Sudan

progressed. The peace talks – and the subsequent accord – included

regulations concerning allocation of valuable resources. In Darfur, as

well as in other parts of Sudan, there were fears of exclusion from the

agreement on wealth sharing, which was an important aspect of the

power-sharing deal. Also the 2006 Darfur agreement failed to include

all warring parties: Only one rebel group faction signed the accord with

the government, and violence in that region did not end. In the eastern

part of Sudan, the Beja Congress demands a fair share of power and

national wealth, similar to the agreement made for the southern part of

Sudan. If the issue is not settled, there is a risk of violence also in this

region (ICG 2006a; UCDP 2006).15

As these cases demonstrate, it is often difficult to settle all interlinked

conflicts at once. Inclusion of warring parties in a power-sharing

arrangement does not always end violence. At the same time, conces-

sions to some warring parties can provide incentives for other parties

to use violence in the pursuit of a share of power. When this is the

case, inclusion of warring parties in a power-sharing arrangement

may at best be a short-term solution to violence. In the worst case,

inclusion of warring parties can escalate violence and give rise to new

conflicts.

14 The election on August 19, 2005 rendered CNDD-FDD a massive victory for theformer rebel movement (ICG 2005a).

15 My thanks to Ken Menkhaus for comments on my interpretations of theseevents.

Power sharing 117

Intra-group contestation

Power sharing changes the dynamic of political contestation by what I

call the ‘‘eclipse’’ mechanism. By fixing the ratio of government posi-

tions for each contending group, the primary conflict is regulated – or

at least concealed. When the limelight is removed from the main

incompatibilities, there is a possibility to focus on other tangible poli-

tical issues that cut across the former conflictual lines. Ideally, this

provides space for moderate factions in support of the peace process

and increases vertical legitimacy.

In some cases, it takes time for power sharing to have this effect. In

Angola, the power-sharing provisions of the 1994 Lusaka Protocol

contributed to an intraparty split of the National Union for the Total

Independence of Angola (UNITA).16 When Jonas Savimbi, the leader

of UNITA returned to war, many other UNITA members saw power

sharing as more attractive and refrained from using violence.17 Partial

implementation of the power-sharing provisions began by the forma-

tion of the Government of Unity and National Reconciliation in

April 1997, when some UNITA deputies took up their seats in parlia-

ment (Hodges 2001: 58). However, Savimbi never assumed the posi-

tion of minister, nor did he take up one of the two positions as

vice-president as stipulated in a constitutional amendment (Ohlson

1998: 79, footnote 76).

The power-sharing provisions of the 1994 Lusaka Protocol were not

fully implemented until large sections of UNITA had been militarily

defeated and after Savimbi was killed in February 2002. A subsequent

cease-fire agreement reinforced the provisions of the Lusaka accord,

16 Another case where power sharing was used to pacify warring groups is NewZealand. The process of substituting warfare with the incorporation of Maori inthe political process took some thirty years. In 1867 the British Crown offeredMaori tribes reserved seats in the New Zealand parliament on condition thatthey agreed to lay down their weapons. Although several tribes continued theirwarfare, eventually all Maori were pacified. Thus, also in this case power sharingdid eventually work as an incentive for moderation (Jarstad 2001).

17 It was not power sharing per se that caused the peace agreement in Angola tobreak down, but rather the failure to implement the agreement. A moreconvincing explanation of why the conflict lingered on despite the peace accord,suggested by several authors, is UNITA’s control of the lucrative diamond mines(Hodges 2001; Zahar 2005a). Thus, for Savimbi the spoils of war outweighedany potential gains of political influence via power sharing.

118 Anna K. Jarstad

and UNITA assumed four ministerial posts and began the transforma-

tion into a political party. It could therefore be suggested that the

power-sharing provisions of the 1994 Lusaka Protocol have contrib-

uted to peace in Angola. However, vertical legitimacy is still weak as

the present power-sharing arrangement between UNITA and the

Popular Movement for the Liberation of Angola (MPLA) is not based

on open contestation.

In the case of Angola, the moderate faction within UNITA eventually

gained ground. However, intra-group contestation does not always

lead to moderation. Pierre M. Atlas and Roy Licklider have suggested

that societies in civil war are often driven by a whole variety of con-

flicts. When the primary conflict is handled, other conflicts may come

to the foreground (Atlas and Licklider 1999). For this reason, intra-

group competition can also have the negative effect of resulting in

extremist splinter groups. In severe cases, this can lead to a resumption

of violence.18 This was the case in Rwanda 1993, where the inclusive

Arusha Accords could not prevent the genocide. An extremist faction

of the Hutu government feared exclusion from power sharing stipu-

lated in the agreement and formed a splinter group, which aimed to

eliminate all Hutu moderates and Tutsi (Rothchild 2005b).

International dependence

Several PSAs are mediated and guaranteed by international actors. Such

third-party dependence activates the systemic dilemma. In Cyprus, the

constitution of 1960 was a result of the negotiations between

Great Britain, Turkey, Greece, a Greek Cypriot delegation headed by

Makarios, and a Turkish Cypriot delegation headed by Kutchuk. It

allotted the Turkish Cypriots more than their proportional share of

seats in parliament to be elected separately by the Turkish Cypriot

electorate, veto rights in vital matters, and a Turkish vice-president.

The consociational arrangement was a compromise between the

Turkish Cypriot demand for partition (taksim) and the Greek Cypriot

desire for unification of Cyprus with Greece (enosis), or at least Greek

hegemonic control of political power.

18 One example of this took place in Northern Ireland, where a splinter group fromthe IRA planted a bomb in Omagh in August 1998, after the peace deal wassigned.

Power sharing 119

However, shortly after the formation of joint government, Makarios

felt confident of international support for removing provisions enhan-

cing Turkish Cypriot political power. In December 1963, President

Makarios unilaterally announced revision of the constitution, which

in fact eroded its consociational character. The reason for Makarios’

defection from the agreement was that he believed that his bargaining

position had improved. As a consequence, ethnic conflict erupted all

over the island and since then Turkish Cypriots have no representation

in the government. Nevertheless, the United Nations decided to recog-

nize the now all Greek Cypriot parliament as the only legitimate

government of Cyprus. Still today, when a parallel Turkish Cypriot

administration in the north has assumed the features of statehood, the

government controlling the southern part of the island is the only

internationally recognized Cypriot government (Jarstad 2001).

In this case, the promise of international guarantors to intervene to

restore constitutional order was not followed through and power

sharing ended. In cases such as Bosnia and Herzegovina and Kosovo,

the international community has decided to stay on to ensure multi-

ethnic power-sharing governance. It is yet to be seen whether these

amendments will provide for sustainable peace also without interna-

tional security guarantees.

Leveling power relations

Power sharing entails the leveling of power relations between contend-

ing groups. Such ensured inclusion in government contributes to con-

flict management by reducing the uncertainty of elections. However,

changing power relations between the target groups in the power-

sharing arrangement activate the temporal dilemma. A PSA tends to

reflect the power relations at the time of negotiations. If power rela-

tions shift, there is a risk of defection from the agreement. Power

relations may shift as a consequence of changing military capabilities

or shifting demographic proportions.

Lebanon is an illustrative case in this regard. Lebanon has experi-

enced four power-sharing systems since 1861. As the power balance

has changed, each system has created different winners and losers

amongst the seventeen religious groups in the country. The power-

sharing system introduced in 1943 lasted until 1975. As the Muslim

share of the population in Lebanon increased, many Muslims saw the

120 Anna K. Jarstad

fixed Muslim–Christian ratio in parliament as unjust. Dissatisfaction

with the power-sharing balance was a major cause of the fifteen-year-

long civil war that broke out in 1975. The 1989 Ta’if Charter of

National Reconciliation contributed to ending the conflict. By this

agreement the constitution was amended to provide for an equal number

of Christian and Muslim members instead of the previous six to five ratio

in the National Assembly (Zahar 2005b).19 However, the distribution of

seats and the relative balance between the various groups is again an

issue of discussion. The Sunni–Shiite polarization constitutes the most

serious threat to the old equilibrium. In May 2000, Israel ended the

occupation of southern Lebanon after a decade of Hizbollah attacks.

Hizbollah’s military achievements were rewarded during the democratic

elections in 2005, when Hizbollah won 14 out of 128 seats in parlia-

ment. Hizbollah have demanded a larger share of influence in both the

legislative and the executive branch of government to adjust the power-

sharing balance to reflect the increased Shiite population (ICG 2006b).

To avoid a situation where power sharing over time gives rise to

recurrence of conflict, several peace accords include provisions for abolish-

ment of power sharing after a transitional period. The so-called sunset

clause of the transitional power-sharing arrangement in South Africa is

often cited as a model for designing power sharing after war (see, e.g.,

Sisk and Stefes 2005).20 A less known case is the transitional power

sharing in Sierra Leone. The Lome Accord of July 7, 1999 provided for

guaranteed inclusion of the rebel movement Revolutionary United Front

(RUF). The post of vice-president was allotted to RUF’s leader Foday

Sankoh, along with four cabinet positions and four deputy ministerial

positions (Francis 2000: 363). UN troops were deployed to the country

and in spite of repeated outbreaks of violence by the RUF during the

implementation period, by January 2002 the disarmament and

19 The prior consociational system held for thirty years and during this period itmanaged to prevent full-scale civil war and, according to Arend Lijphart (1977:150), also established a remarkable record of democratic stability.

20 Colombia 1958–1974 is another case where power sharing was agreed to last fora specified period of time. However, during the time of power sharing there wereno means for other actors to gain government positions. During this period,guerillas that still operate, such as the Revolutionary Armed Forces of Columbia(FARC), were formed. It has been suggested that the political blockage ofpolitical movements outside of the National Front regime was one of the majorexplanations for why such non-parliamentary groups were formed (see, e.g.,Hartlyn 1998: 191 ff.).

Power sharing 121

demobilization was completed and the war declared over. The transi-

tional period ended with the holding of elections in May 2002, in which

the political arm of RUF received only 1.7 percent of the votes. Without

any guaranteed seats in government, no domestic popular support, and

with a significantly diminished battlefield capacity, the RUF soon

thereafter disappeared as a significant military or political force in

Sierra Leone (ICG 2002; Soderberg Kovacs 2007). In this case, transi-

tional power-sharing provisions proved to be sufficient for the rebels to

sign the accord and thereby provided for peace.

However, in other cases, the abolishment of power sharing is a

source of conflict. In some cases the actors demand permanent regula-

tions regarding government positions, and it may therefore not be

possible to convince the warring parties to agree to temporary power

sharing. Also when the parties have agreed to end power sharing after a

transitional period, it may be difficult to abolish it. After a period of

transitional power sharing in Cambodia, stipulated in the Paris

Agreement in October 1991, elections were held in May 1993. The

Cambodian People’s Party (CPP) under Hun Sen refused to accept that

they lost the elections. A coalition government was formed between

CPP and the winning party FUNCINPEC (National United Front for

an Independent, Neutral, Peaceful and Cooperative Cambodia) with a

dual prime minister arrangement. This power-sharing arrangement

had no basis in either the 1991 peace agreement or the constitution of

1993 (Brown 1998: 90–100). In addition, the ad hoc nature of the

power-sharing arrangements probably contributed to the renewed out-

break of violence between the parties in 1997 and the attempt by the

CPP to oust FUNCINPEC from power in a coup in the run-up to the

next elections. Nevertheless, the CPP–FUNCINPEC coalition was

revived after the 1998 elections, due to the two-thirds constitutional

requirement for government formation. The 2003 elections again led to

a stand-off between the two major parties (Reilly 2007). As these cases

indicate, power sharing has effects on the prospects of peace as well as

democracy. The consequences of power sharing regarding issues

related to democratization are discussed more thoroughly below.

Power sharing and democratization in war-torn societies

In terms of the quality of democracy, Lijphart has demonstrated that

non-majoritarian democracy outperforms majoritarian democracies

122 Anna K. Jarstad

on indicators of macroeconomic variables, gender issues, voter parti-

cipation, welfare policies, environmental protection, criminal justice,

and economic aid to developing countries (1999).21 Lijphart holds that

the majority is dependent on the minority’s allegiance to the regime. He

suggests that broad participation, bargaining, and compromise are

essential for legitimate decision making. Therefore, according to

Lijphart, the principle of maximum inclusion is more democratic

than majority rule (1999: 2, 32–33).

In addition, Lijphart expects consociationalism to promote compro-

mise and conciliation (1994: 12). Lijphart is not clear on exactly how

this transformation comes about, but interdependence and recursive

negotiations seem to be key components in this process. One interpre-

tation of the logic of consociationalism suggests that recognition of

ethnic demands is expected to increase the perception of security,

which in turn promotes mutual trust and reduces grievances. Iterative

cooperation in a grand coalition is expected to promote an under-

standing of the other party.22 In the long run, the policies of the parties

are expected to conform, to avoid exclusion or discrimination of the

other group. In this way consociationalism is expected to depoliticize

ethnicity and allow development of a common national identity. In

other words, consociationalism is expected to produce moderation

(Jarstad 2001: 46–48). For this reason, power sharing could ideally

work as a catalyst for peaceful cooperation among contending parties

after a peace deal is signed.

However, Lijphart’s theory on consociational democracy is mainly

based on the analysis of reforms in already democratic societies, such as

the Netherlands. The theory fails to analyze how the particular condi-

tions of a war-shattered society affect the democracy-promoting aspects

of power sharing. Below, I discuss four types of potential demo-

cratic shortcomings that designers of a PSA confront: exclusion of

moderates, weak popular support, lack of local ownership, and group

representation.

21 In this work Lijphart analyzes consensus democracies. However, most of theconsociational democracies are included in this category. For the differencebetween consensus and consociational democracy, see Jarstad (2001).

22 Such repeated rounds of cooperation can be understood by the reciprocalstrategy of tit for tat by matching the other player’s move in iterative games(Morrow 1994: 264).

Power sharing 123

Exclusion of moderates

When consociational democracy develops as a result of reform in

countries with prior experience of democracy, it indeed broadens

participation in decision making. According to the consociational

logic, recurring rounds of bargaining and compromises foster modera-

tion. However, in many war-torn societies, power sharing may have

the adverse effect of preventing the development of moderate political

parties. This issue of inclusion versus exclusion activates the horizontal

dilemma of power sharing.

In some peace accords, warring parties decide to share power, and at

the same time effectively block other political movements from power.

When this is the case, the former warring parties control economic

resources as well as media and other channels to mobilize political

support. Even if the government eventually opens up the political

space, any opposition groups lag behind, due to the lack of resources.

This uneven start for parties in a democratization process can have

long-term effects on the prospects of multi-party democracy.

One possibility to overcome this negative effect of power sharing is

to include other actors in the peace negotiations and also in the future

government. This was done, for example, in the Ivory Coast, where the

2003 PSA also included the main political parties, in addition to the

warring parties. The Liberia 2003 PSA included not only all warring

parties to the conflict – the government and the two rebel groups

Liberians United for Reconciliation and Democracy (LURD) and the

Movement for Democracy in Liberia (MODEL) – and the main

political parties, but also representatives from the civil society.

Likewise, the 2002 PSA for the Democratic Republic of Congo included

government representatives, rebel groups, militias, opposition parties,

and different civil society organizations (Nilsson 2006; Nilsson and

Soderberg Kovacs 2005).

Although inclusion of a broad spectrum of society in the peace

agreement is likely to improve the legitimacy of the peace process,

inclusion of warring groups remains an obstacle to democratization.

A typical peace deal stipulating power sharing tends to reflect power

relations during negotiations, rather than the size of potential electo-

rates. By granting warring parties a stake in government, violence is

rewarded. This in itself violates a fundamental principle of democracy.

The Sudan 2005 power-sharing accord only includes warring parties in

124 Anna K. Jarstad

the executive branch of government. Under these conditions, the only

access to political power is via the battlefield. Such arrangements lack

any form of democratic legitimacy.

Also when power-sharing governments are elected, reserved seats for

warring parties may undermine democratic legitimacy. There is a risk

that the actors continue to use violent tactics to affect the outcome of

future elections. Thus, the inclusion of warring parties in government

gives rise to several questions regarding legitimacy: Will they conform

to democratic norms of non-violence, tolerance and compromise? How

representative are these actors?

Popular support

Increasing the number of political parties in decision making beyond a

mere majority (or even plurality, as is the case in some systems) may

enhance vertical legitimacy of the political system. However, this

depends to a great extent on the level of support for the included elites

among the population. When elites are appointed to positions in a

power-sharing government, it is difficult to know how representative

they are. But also elective power sharing runs into problems of vertical

legitimacy. In divided societies power sharing freezes the conflict lines,

and the parties do not need to compete for votes among their former

foes. This means that most parties have a narrow support base. In

Bosnia and Herzegovina, moderate parties receive little electoral sup-

port. Instead, nationalist parties prevail in elections. At the same time,

the voter turnout is low – less than 55 percent (OSCE 2006) – and the

trend demonstrates a decline during the last years. This reflects weak

legitimacy of the political system as a whole.

In November 2006, Nepal’s Maoist rebels reached a transitional

power-sharing agreement with the civilian government to end the

country’s ten-year-long civil war. The constitutional process has thus

far concentrated on building elite consensus in closed-door negotia-

tions at the expense of public debate. The challenge to ensure vertical

legitimacy remains (ICG 2007a).

Even when power sharing and peace negotiations are exclusively

elite-driven, there are ways to gain public acceptance of the agreement.

One way to ensure that the system as a whole is seen as legitimate, is to

make power sharing the object of a referendum. In Sudan the 2005

peace agreement stipulates a referendum on self-determination for the

Power sharing 125

southern part of Sudan to be held in early 2011 (ICG 2006a). Thus,

during the interim period the North has a chance of making unity

attractive to the South. For such a referendum to result in support for

the peace agreement, the political leaders have to ‘‘sell’’ the agreement

to the public, and explain why concessions are necessary and desirable

for sustainable peace. If not, the public is likely to remain polarized.

This was the reason why the Greek Cypriot majority voted against UN

Secretary-General Kofi Annan’s plan to unite Cyprus under a power-

sharing accord in April 2004.

An example of a referendum that rendered the peace deal legitimate,

is the one concerning power sharing in Northern Ireland, where the

Belfast Agreement won the support of both communities in 1998.

Belloni and Deane suggest that the approval in the referendum is an

important factor explaining why the peace deal has not collapsed,

despite the many obstacles (2005).

Local ownership

Internationally mediated power-sharing agreements activate the sys-

temic dilemma of local versus external ownership of the democratiza-

tion process. In extreme cases, where power sharing is imposed by third

parties, local ownership is weak. In Kosovo power sharing was

imposed by UN provisions, without a consensus among the formerly

warring parties. In June 1999, UN Security Council Resolution 1244

turned Kosovo into a UN protectorate. It is headed by the Special

Representative of the Secretary-General (SRSG), who has the ultimate

political authority in Kosovo, including the right to issue legislative

acts, and appoint and remove any person in the civil administration.

Despite provisions for inclusion in the political bodies of representa-

tives of the Kosovo Albanians, Kosovo Serbs, and the other commu-

nities in Kosovo, power sharing has to a very limited extent led to

political cooperation across the former conflict lines. This is a result of

the lack of ownership of the power-sharing arrangement. The provi-

sions for power sharing were not issued by the central government, but

by United Nations Mission in Kosovo (UNMIK) regulation 2001/9,

May 15, 2001. These provisions were not put up for general consulta-

tion among the wider public. Not even the political parties involved in

the UNMIK structures signed any decisions for the creation of the new

political institutions. Instead, a small closed group designed the

126 Anna K. Jarstad

institutions. The only input from minority groups included consulta-

tions with a few Serbs. Mass demonstrations suggest that also the plan

presented by United Nations Special Envoy Martti Ahtisaari in the

beginning of 2007 lacks local ownership (Baldwin 2006; Jarstad 2007).

Power sharing in Bosnia and Herzegovina suffers from a similar lack

of local ownership. During the negotiations in Dayton 1995 concern-

ing Bosnia and Herzegovina, no representative from the civil society

was included.23 Furthermore, the High Representative has used his

mandate to remove elected officials from politics on several occasions,

thus undermining local ownership as the democratic foundation for the

power-sharing government. Both in Kosovo and in Bosnia and

Herzegovina, there is a problem of democratic accountability. Real

power is not vested in the democratically elected governments, but in

the international administration. These structures are not formally

accountable to the citizens in these states. Such international interven-

tion risks increasing support for extremism and ethnochauvinism, thus

undermining democratization.

It has been suggested that power sharing is more likely to last when it

is arrived at indigenously (Sisk 1996: 118). Even when a PSA is

mediated and guaranteed by international actors, there are ways to

enhance the local ownership of the agreement. An example of such an

effort is the Pretoria Agreement 2002 (DRC),24 which was mediated by

a third party (South Africa) and involved a broad spectrum of society.

The agreement was reached after discussion under the so-called Inter-

Congolese dialogue, which brought together the government, civil

society groups, political opposition groups, and rebel movements

(UCDP 2005).

Group representation

It has been questioned whether power sharing at all can be democratic.

One reason for this is that power sharing emphasizes representation of

certain target groups, rather than representation of ideas and ideolo-

gies.25 All power-sharing systems have to settle the difficult issues of

23 See the chapter by Belloni on civil society in war-torn countries in this volume.24 The formal title of the agreement is the ‘‘Global and Inclusive Agreement on the

Transition in the Democratic Republic of the Congo.’’25 See, e.g., Phillips 1998 on the representation of ideologies versus groups.

Power sharing 127

defining which groups should be represented and the share of seats for

each group. Over time, both which groups to include and the relative

share of power need to be reconsidered to address the temporal

dilemma of power sharing. Democratic forms of power sharing have

to regulate who is eligible to vote for the group representatives (sepa-

rate or common electoral rolls), who can stand for elections, and the

design of the electoral system.

In addition, the issue of accountability is problematic for all types of

broad-coalition governments. When all groups are included in govern-

ment, it is difficult to hold politicians accountable and elections may

even seem meaningless. However, elections can make a difference also in

power-sharing systems, by allowing voters to choose between different

candidates. Cyprus 1960 and Burundi 2000 are examples of different

types of power-sharing constitutions. In Cyprus, the constitution stipu-

lated that voters could choose between candidates on separate ethnic

rolls. This meant that Greek Cypriots could only vote for Greek Cypriot

candidates and Turkish Cypriots could only vote for Turkish Cypriot

candidates. Political parties with candidates of different ethnic affiliation

could not stand for elections. In practice, there was a lack of political

alternatives, especially among the Turkish Cypriots.

However, there are ways to make power sharing more democratic.

In New Zealand as well as in post-transitional Burundi, the power-

sharing arrangements allow the voters to vote for candidates of any

ethnic origin. The 2000 Burundi PSA stipulates two types of power

sharing; during the transitional period, power was shared among eth-

nopolitical parties, while the permanent constitution stipulated quotas

for ethnic representation in the democratically elected National

Assembly. Both types are based on a 40/60 formula, to over-represent

the Tutsi minority. In the Transitional Government, 40 percent of the

seats were allocated to political parties then largely representing Tutsi,

and 60 percent were reserved for the Hutu parties (ICG 2004c: 3).

The permanent constitution stipulates a quota for Tutsi candidates,

not Tutsi parties. This means that people are allowed to vote for any

party and that the Hutu parties have an incentive to include Tutsi in

order to gain more seats. In the government elected on August 19, 2005

40 percent are Tutsi, but most of them belong to traditionally Hutu

parties and only two Tutsi parties remain in the official institutions.

This system has thus had tangible effects on ethnic relations. The

traditional Hutu–Tutsi interethnic conflict has been substituted by

128 Anna K. Jarstad

violent dispute over power among Hutu parties. Furthermore, fewer

exclusive Tutsi parties remain in power (ICG 2005a). Under the per-

manent constitution, no group has ensured ethnoexclusive representa-

tion. Typically, the minority has more to fear from such an

arrangement as it risks absorption by the majority. But it also means

that the extremists have no ensured inclusion in government. It is yet to

be seen whether Tutsi interests are sufficiently taken care of by this

government. This power-sharing arrangement results in the dilemma of

providing incentives for moderate parties to attract voters from both

ethnic groups, while increasing the risk that extremist Tutsi parties

challenge the power-sharing setup by military means.

It is often suggested that power-sharing arrangements are inefficient

and that the decision-making process is slowed down when all groups

have a say in the process. That is true: power sharing does entail the

risk of stalemates and problems of accountability. However, this is a

consequence of any coalition government and does not imply that such

a system is not democratic. On the contrary, after a civil war, it is of

particular importance that all interests are represented, in order to

render decisions a high degree of legitimacy and to facilitate

implementation.

It is no easy task to define the groups and who should represent them.

Even in societies where power sharing develops in the course of peaceful

and democratic procedures, there is no obvious principle for defining

target groups for reserved seats in parliament. In New Zealand, each

individual, no matter their ethnic affiliation, can now enlist either on the

Maori roll or the General roll prior to each election. This allows for some

flexibility regarding definition of politicized ethnicity. In the case of New

Zealand, it is clear that Maori are the indigenous people, and no other

ethnic group, such as the large Chinese community, is allotted a quota.

However, it is contested whether or not aboriginal status is a just basis

for special representation. The number of seats depends on the share of

registered voters on the Maori roll. In addition to quotas for Maori, all

main political parties include Maori among their top candidates. As a

result, in the last elections the number of Maori in parliament roughly

corresponds to their demographic share. There is no mechanism ensur-

ing proportional ethnic representation. At present, however, it is in the

interest of all political parties to appeal also to Maori voters and it could

therefore be argued that the quotas have become a self-enforcing institu-

tion of conflict management (Jarstad 2001).

Power sharing 129

Conclusions

Power sharing has become a tool for ending conflicts by the inclusion of

warring parties in government. In this chapter, I have outlined how

previous research expects power sharing to contribute to peace and

democratization. Although it is yet too early to evaluate the long-term

effects of many of the power-sharing accords discussed here, I have

argued that power sharing in war-torn societies can have negative

effects on war-to-democracy transitions. Power sharing may not only

fail to end the violent conflict. In addition, it may enhance the risk of

violence and undermine democratization. Possible negative conse-

quences for sustainable peace were discussed in relation to four

mechanisms of power sharing: inclusion of warring parties, intra-

group contestation, international dependence, and the leveling of

power relations. Firstly, while power sharing works as an incentive to

lay down arms for included groups, it entails at the same time an

incentive for excluded groups to continue fighting or to take up arms.

Secondly, the ‘‘eclipse’’ mechanism may give rise to intra-group con-

flict. In the short run, power sharing may take the heat off the major

conflict and allow other political issues to emerge. But simultaneously,

there is a risk that extremist splinter groups are formed. In war-torn

societies this risk is particularly acute, and its consequences may be

devastating. Thirdly, international guarantees to uphold the power-

sharing arrangement may turn into dependence. Fourthly, in the long

run, shifting power relations mean that it may be difficult to avoid a

breakdown of the agreed power-sharing balance.

Furthermore, four problems pertaining to democratization via

power sharing were discussed: exclusion of moderates, lack of popular

support, lack of local ownership, and group representation. Firstly,

power sharing only among warring parties risks undermining the

development of moderate political parties. PSA allotting government

positions to warring parties sharply contradicts the principle of demo-

cratic legitimacy and undermines long-term democratization by limit-

ing the space for moderate parties. Secondly, the elites appointed to

share power may lack popular support. This is an impediment to

vertical legitimacy. Thirdly, the lack of local ownership may hinder

the development toward a democracy. Fourthly, power sharing often

presupposes the definition of groups and may prove inapt to accom-

modate new groups that emerge.

130 Anna K. Jarstad

These eight negative consequences of power sharing cannot all be

addressed at the same time. Rather, the design of a PSA involves trade-

offs between the desirable goals of promoting conditions conducive to

peacebuilding, on the one hand, and democratization, on the other

hand. Below, these trade-offs are discussed in relation to the four

dilemmas outlined in the conceptual chapter of this book.

The horizontal dilemma of power sharing speaks to the issue of

inclusion versus exclusion. The issue of which parties are included –

and which actors are excluded – in a power-sharing arrangement has

pivotal implications for long-term peace as well as democratization.

A viable peace deal cannot be reached without the leaders of the warring

parties. However, for the sake of democracy, exclusion of violence

makers is warranted. While this is a genuine dilemma which cannot

be solved, the negative consequences can be mitigated by inclusion of a

broad range of sections of society in the peace process as well as in the

power-sharing government. Furthermore, deliberate measures need to

be taken to ensure that all parties abstain from violence, for example by

disarmament and reforms pertaining to the security sector.

The vertical dilemma entails the issue of legitimacy versus efficacy. In

the case of power sharing, the consequences of intra-group contesta-

tion and popular support interact. For power sharing to contribute to

both peacebuilding and democratization, the intra-group contestation

should result in elected moderate politicians. This is highly unlikely,

but not impossible, in societies shattered by war. The ways to increase

the democratic legitimacy of a power-sharing system include election

of the members of the power-sharing government, mechanisms allow-

ing new political parties to enter, and making the power-sharing system

the object of a referendum. Thus, under certain conditions the vertical

dilemma can be avoided.

The systemic dilemma entails the issue of internal versus external

ownership of the power-sharing agreement.26 Ideally, the local actors

arrive at a well-balanced power-sharing arrangement. However, in the

midst of a civil war, third-party mediation is often needed to reach a

negotiated settlement including a power-sharing deal. To ensure local

ownership of the peace process and of the power-sharing agreement

26 For an analysis of dilemmas related to third-party guarantee, see the chapterby Fortna in this volume (see also Walter 2002 on this dilemma in relation topower sharing).

Power sharing 131

itself, broad inclusion is desirable already during the negotiation phase.

This can be combined with third-party guarantees to ensure a peaceful

transition toward democracy.

The temporal dilemma of power sharing speaks to the issue that the

power balance agreed on in a peace deal may be challenged over time.

A PSA tends to reflect power relations at the time of the negotiations,

rather than the size of potential electorates. To end a war, it may be

necessary to include warring groups in government, despite the risk of

undermining the prospects of democratic legitimacy and restricting the

possibility for moderate parties to develop. In the long run it is likely

that new groups emerge, excluded groups demand a share of govern-

ment posts, or that the demographic character of the state changes. If

there are no mechanisms for adjustments to such changes, the system

may be challenged. In such cases, transitional power sharing is to be

preferred. However, to end violence, it might be necessary to guarantee

the parties involved a permanent share of power. Warring parties may

not be prepared to lay down their arms and sign an agreement if they

know that transitional power sharing will end, and that they risk losing

all political and military power once the transitional period is over. For

this reason, there is no simple solution to the temporal dilemma of

power sharing.

By bringing together research on power sharing from two previously

disparate strands of research – conflict management and democratization –

the potential negative consequences of power sharing become evident.

Under power sharing there is no easy way to avoid dilemmas: the

management of one dilemma often results in new trade-offs between

peacebuilding and democratization. At the same time, in many cases

the alternatives to power sharing are worse.

Although each power-sharing deal needs to be designed based on an

analysis of the specific context of the war-torn society in focus, a few

guidelines could be suggested. Firstly, strive for inclusion rather than

exclusion. Design a system that allows for existing moderate parties

and for new groups to join. One possibility would be to stipulate a

minimum quota, not a fixed maximum quota, and to leave some

positions in parliament open for election outside of the quotas.

Secondly, immediately after a war legitimacy is often considered

more important than efficacy. Ensure that voters have a choice between

several candidates in each election. Strive to reserve seats for political

parties, not for specific individuals. Thirdly, strive for long-term rather

132 Anna K. Jarstad

than short-term assistance. To last, some power-sharing deals need

credible international security guarantees. In other cases, difficult

negotiations continue and need international facilitation and media-

tion, also after a power-sharing deal is reached. Fourthly, the balance

between the need for a flexible system and provisions to guarantee

some actors’ influence over a long time period depend on the overall

peace process. External security guarantees and international monitor-

ing are often pivotal for a smooth transition from power sharing to

more open forms of elected governments. To conclude, this chapter

suggests that for power sharing to promote both peace and democracy,

it needs a carefully crafted design, as well as continuous monitoring. By

taking into account the adverse effects of power sharing on peace-

building and democratization, some of the worst pitfalls can thus be

avoided.

Power sharing 133

5 When rebels change their stripes:armed insurgents in post-war politics

M I M M I S OD E R B E R G K O V A C S

Following the announcement of the final results of the legislative elec-

tions in Afghanistan in 2005, there was widespread concern that

powerful warlords, former Mujahideen commanders, and Taliban

strongmen had been elected to power in this war-ravaged country.

The same year a peace agreement was signed between the Indonesian

government and the guerillas in the province of Aceh, aimed at ending

the prolonged civil war. The provisions of the agreement laid out

the political and legal conditions for the establishment of local poli-

tical parties and thus gave the amnestied rebel movement an oppor-

tunity, for the first time, to pursue its aims through the ballot box. In

Nepal, the Maoist rebels agreed to join the interim government fol-

lowing the peace agreement in 2006. But what are the prospects for

democratization and sustainable peace in war-scattered societies in

which formerly armed insurgents emerge as politicians? The purpose

of this chapter is to address this pivotal issue that is likely to be of

great relevance to researchers and policymakers alike concerned with

better understanding the conditions that facilitate and obstruct a

transition to both peaceful and democratic politics in intrastate

armed conflicts.

Because issues related to failed governance and the unequal distribu-

tion of political power and public goods often are at the core of the

causes of civil wars, conflict resolution in these contexts frequently

includes efforts to introduce, reintroduce, or reinforce political reforms

aiming at a transition to democratic politics. One such political reform

is the effort to transform warring armies to political parties, who are to

pursue their political goals peacefully and openly within the political

system. This trend ought to be encouraging, as the process of rebel

conversion has been identified as one of the key factors for the success-

ful implementation of peace agreements. In war-to-democracy transi-

tions, the inclusion of formerly armed groups in politics also fulfills an

additional purpose, namely to provide the agents for the emergence of a

134

multi-party democracy in previously autocratic or semi-democratic

states. Thus, in recent decades we have witnessed numerous former

rebels emerge as political parties following peace agreements in intra-

state armed conflicts.

However, this study argues that although the inclusion of formerly

armed and militarized groups in post-war politics has proven benefi-

cial, or even necessary, for the purpose of ending the war through a

negotiated settlement, it sometimes has negative consequences for the

emerging democratization process. The militant, hierarchical, sectar-

ian, and internally undemocratic nature of many of these groups work

counter to the development of peaceful, democratic, transparent, and

inclusive policies. In addition, the process of rebel inclusion itself may

hamstring the democratization process due to the use of amnesty

provisions in the peace process or by limiting the possibility for other

actors to emerge and influence the post-war political agenda. In the

long run, the lack of democratic progress might also impair prospects

for sustainable peace. Paradoxically, therefore, efforts to promote both

peace and democratization in war-shattered societies through the

transformation of armed groups to peaceful parties might undermine

precisely those values that it sought to encourage. In spite of its rele-

vance, this question remains largely unexplored in the scholarly litera-

ture, as it tends to fall between two different strands of research that are

rarely integrated, namely that on war termination and conflict resolu-

tion on the one hand, and that on democracy and democratization on

the other.

This chapter will set out by discussing the argument in favor of

transforming rebel groups to political parties in civil war peace pro-

cesses for the sake of peace. Thereafter, the critical role played by

political parties in democracies is briefly addressed. This is followed

by a closer look at some contradictory issues that may arise when

the two processes of democratization and peacebuilding meet, and

armed groups emerge as political parties in the aftermath of a nego-

tiated settlement. These issues will be analyzed in the light of the

four overarching dilemmas addressed in the conceptual chapter

of this book: the vertical, the horizontal, the systemic, and the tem-

poral dilemma. Empirical examples will be given from a range of

different cases to illustrate the issues under discussion. Finally, in the

concluding remarks, a few implications for policy on this area will be

discussed.

When rebels change their stripes 135

Rebel-to-party transformations: a remedy for ending the war

Since the end of the Cold War, peace agreements between the warring

parties have become a more common option for ending civil wars

(Wallensteen 2002: 3). Following in the footsteps of this trend,

researchers have attempted to identify the conditions under which

parties enter into negotiations, when negotiations are likely to lead to

settlements, and when peace agreements between former belligerents

are likely to hold and prove durable. Within this strand of literature,

two different, though sometimes interrelated, arguments have been put

forward in favor of efforts to transform former rebels to political

parties in civil war peace processes. The first argument is based on

the notion that if the previous war was the result of legitimate grie-

vances on the part of the armed opposition, the conflict resolution

processes should strive to address these injustices, in particular the

systematic exclusion of certain groups, and as far as possible try to

amend them. The second argument builds on the view that in order to

ensure the commitment of the rebels to the peace process, one should

strive to increase their incentives for peace through granting them a

legitimate political role in the new post-war order.

In line with the first argument, it has been suggested that because

internal conflicts begin with the breakdown of normal politics, the

conflict resolution process should entail the establishment of mechan-

isms that allow the conflict to shift from violence back to politics. The

return to normal politics should be on the basis of a new inclusive polity

that brings together those who felt deprived and discriminated against,

and those who were part of the old political system, to share power and

benefits in a new political system. Specifically, third-party actors

should strive to put to work settlements that attempt to engage the

rebel group in a new role involving participation, legitimization, and

allocation through, for example, a conscious effort to encourage the

former rebels to pursue their goals openly in the political system

(Zartman 1995: 337–338). Because civil wars ultimately concern the

distribution of power in society, one of the key problems in civil war

conflict resolution concerns precisely the construction of a social and

political system, which ‘‘gives reasonable social and political space to

all groups in a society’’ (Wallensteen 2002: 133). Democratization, in

this view, becomes a way to handle the participation of parties in a

post-war society and to give space to actors who have previously been

136 Mimmi Soderberg Kovacs

suppressed or excluded from influence. A conflict resolution process

based on the establishment of a democratic political system thus

encompasses the necessary procedures for transferring the struggle

into constitutional and non-violent forms (Wallensteen 2002: 139).

The inclusion and active participation of the former rebels may there-

fore symbolize the end to one-party politics and exclusive political

systems, and becomes the ultimate sign of democratic politics. Cerdas

Cruz argues in reference to the Central American peace processes that

‘‘the incorporation of insurrectional groups into the system, trans-

formed into duly legalized political parties, is critical for the future of

the transition and the formation of a truly representative party system’’

(1998: 15).

Hence, peacebuilding in post-war societies that merely attempts to

reestablish the pre-war political order has serious limitations as it

ignores the imbalances between groups in existing political and eco-

nomic structures. Peacebuilding strategies therefore, ‘‘must be geared

towards modifying social structures and processes associated with such

power imbalances’’ (Jeong 2005: 3). In addition, the formation of a

legitimate government is a necessary condition for order and stability

in the aftermath of war, and trust and confidence in the new govern-

ment can best be gained through the establishment of democratic

principles reflecting an inclusive representation that may overcome

the divisions created by wartime alignments (Jeong 2005: 83–84).

Thus, in accordance with this line of reasoning, peace processes should

aim to address and modify the underlying grievances, in particular the

politics of social, economic, and political exclusion of certain groups or

regions that produced the war in the first place, in order to create stable

and peaceful solutions to intrastate armed conflicts.

In accordance with the second line of argument, there is another

more immediate and pragmatic reason for creating and supporting

solutions that attempt to engage the former antagonists in the post-

war political system. It has been suggested that third-party actors

concerned with the successful implementation of peace agreements in

civil wars should prioritize the demobilization of soldiers and the

demilitarization of politics, through, for example, the transformation

of former warring armies to peaceful political parties. In fact, without

doing so, ‘‘civil wars cannot be brought to an end, and important

normative goals such as the creation and consolidation of democracy

and the protection of human rights have little chance of success’’

When rebels change their stripes 137

(Stedman, Rothchild, and Cousens 2002: 668). Lyons argues that the

risk of renewed conflict increases significantly if warring parties have

the option to return to armed struggle. The disbanding and disarma-

ment of former combatants thus leave few choices for the parties but to

accept the peace process and the outcome of the post-war elections.

Any failure of demobilization, on the other hand, leaves the possibility

of resuming fighting an option. In addition, if former warring parties

have a stake in the post-war political order and believe they have a

credible chance of organizational survival and political influence

through peaceful means, they are more likely to remain committed to

implementing the peace deal and have fewer incentives to return to

arms. The successful conversion of former insurgents to political par-

ties thus affects both the capacity and the willingness of these groups to

remain committed to the peace process (Lyons 2002; 2004).

Whether the rationale has been stated in light of the first or the

second argument, or perhaps based on a combination of the two,

international donors and concerned third party actors as well as local

peace custodians have actively supported the transformation of former

warring groups to political parties in several peace processes since the

end of the Cold War (Kumar 1998c).1 In some cases it has even been

explicitly stated in the peace agreement that the former rebels should be

allowed to form a political party and participate in post-war elections

as a legitimate political alternative.

The outcome of these processes has, however, varied considerably.

In some cases the former rebels emerged as the new government party

after decades of armed struggle, as happened in South Africa, in East

Timor, and in the Palestinian self-governing territories. In other cases,

for example in El Salvador, Guatemala, Cambodia, and Mozambique,

former rebel groups emerged as opposition parties in the post-settlement

period. In some instances, these attempts to transform rebels to statesmen

failed when the rebels returned to arms during the implementation of the

1 The phenomenon itself is not new. Prior to the end of the Cold War, there aremany examples of armed groups that later assumed governmental power andbecame political parties. Most of these, however, did so through the use of force,either through victory in wars against other groups in the country or followingprolonged armed conflicts with their former colonial powers. In the post-ColdWar period, many former rebels have instead converted to political parties as partof a negotiated settlement to end the war and a decision to introduce democraticpolitics.

138 Mimmi Soderberg Kovacs

peace accord, as was the case with the Khmer Rouge in Cambodia.2

Finally, some rebels emerge as political parties through the establish-

ment of power-sharing mechanisms. In Angola, following the death of

their long-time rebel leader, Jonas Savimbi, and the subsequent 2002

cease-fire agreement with the government, the insurgency group

National Union for the Total Independence of Angola (UNITA) com-

pleted its transformation into a political party and decided to take up

its seats in both the parliament and in the cabinet that it had been

granted in the 1994 power-sharing agreement.

Although the specific problems and prospects these groups face

when entering legal and democratic politics are quite diverse depending

on the particular circumstances of their new role and position, they also

share many similar challenges in terms of the expectations on them as

the central agents of democratization.

Political parties as agents of democratization

Because most contemporary peace processes include a simultaneous

process toward democratization as part of the conflict resolution

efforts, the transformation of former rebels to peaceful political actors

fulfills an additional purpose. They provide the post-war political order

with the political parties deemed necessary for the emergence of a

multi-party democracy in previously autocratic or semi-democratic

states.

In the democracy literature, political parties have traditionally been

regarded as the main intermediary organization of liberal democracy,

linking citizens with the state. Building on this notion of political

parties as the intermediary links between the people and the state, the

various functions that political parties are expected to perform in

democracies can be grouped into two main categories. First, political

parties are expected to perform a variety of representative functions

aiming at channeling the interests and preferences of the people into

policy proposals. Second, they also perform a variety of procedural or

institutional functions that serve to organize the political system of the

2 For an in-depth comparative study on why some rebel groups and not otherssuccessfully transform from violent protagonists to viable political partiesfollowing peace agreements in intrastate armed conflicts, see Soderberg Kovacs2007.

When rebels change their stripes 139

state, for example in parliament or in government, or when it comes to

the recruitment of leaders (Bartolini and Mair 2001: 332; Gunther and

Diamond 2001: 7–8).

Political parties are thus important for a number of tasks deemed

critical for the democratic political system. Hence, in most contempor-

ary democracies, political parties are considered ‘‘a sine qua non for the

organization of the modern democratic polity as well as for the expres-

sion and manifestation of political pluralism’’ (van Biezen 2003: 1). In

new democracies emerging from authoritarian regimes in particular,

the very establishment of democratic procedures has been identified

with the establishment of free competition between parties, and parties

have hence been attributed a pivotal role for the expression of plural-

ism and political participation (van Biezen 2003).

However, it is commonly recognized that the empirical reality of

political parties found in many weak and developing states sometimes

makes it difficult for these to function as the expected agents of demo-

cratization. According to Grugel one of the greatest problems facing

party systems in new democracies is their lack of institutionalization;

they enjoy low levels of legitimacy, have weak roots in society, are

poorly organized, and there are few opportunities for structured inter-

action between parties. Thus, ‘‘they operate quite differently than the

academic literature on democracy and democratization would like to

suggest’’ (2002: 74–75). Randall and Svasand (2002), in their discus-

sion of the performance of political parties in Africa’s new democra-

cies, point to these parties’ limited geographical spread, their lack

of both financial and human resources, and their poor organizational

underpinnings. Owing to these shortcomings, the representative role of

these parties is generally limited, with small bases of support and weak

links to the population. Moreover, they rarely have an aggregative

function or present clearly distinguishable policy platforms.

However, while most observers acknowledge the limitations of poli-

tical parties in new democracies, political parties ‘‘remain an indispen-

sable institutional framework for representation and governance in a

democracy,’’ and ‘‘only political parties can fashion diverse identities,

interests, preferences, and passions into laws, appropriations, polices,

and coalitions’’ (Diamond 1999b: 96–98). Hence, Diamond argues

that given the critical role played by parties in a democracy, ‘‘some

degree of party system institutionalization – of parties with effective,

autonomous organizations, and developed, relatively stable linkages to

140 Mimmi Soderberg Kovacs

voting blocs and social organizations – seems an important condition

for democratic consolidation’’ (1999: 96–98). Thus, although the lim-

itations on political parties in new democracies are generally acknowl-

edged both in theory and in practice, political parties are still

considered vital to the process of democratization. This is equally

true for democratic transitions that emerge from authoritarian rule

and those that take place in the wake of the ending of a civil war.

Dilemmas of armed insurgents in post-war politics

From the perspective of both war termination and democratization, the

transformation of formerly armed groups to political parties in war-to-

democracy transitions is thus generally considered a positive develop-

ment that ought to be encouraged and supported by local as well as

international peace custodians. The parallel processes of peace and

democracy are expected to reinforce each other in cases where former

warring armies become political parties. However, the emergence of

rebels as newborn democrats in post-war politics gives rise to a number

of challenges for both democratic progress and sustainable peace. This

chapter will explore a few issues that are illustrative of this pheno-

menon. These issues are analyzed in light of the dilemmas outlined in

the conceptual framework of this book: the horizontal, the vertical, the

systemic, and the temporal dilemma.

The horizontal dilemma: inclusion versus exclusionof rebel groups

In terms of the horizontal dilemma, this chapter will discuss two issues

that illustrate the tensions that may arise between the values of peace

and democracy when former rebels are included into the post-war

political system. First, the risk of a sedimentation of parties and issues

based on the structures of war, and second, the risk of undermining the

democratic principle of rule of law when the peace agreement includes

provisions of amnesty for crimes committed during the armed struggle.

First, the inclusion of warring parties into post-war politics often

takes place at the expense of the exclusion of other political parties and

new political issues emerging from the post-war democratic context.

This is problematic from a democratic perspective because democratic

consolidation in post-war societies is most likely to take place when

When rebels change their stripes 141

‘‘a new leadership emerges, seeking to organize politics in a different

way from those adopted by discredited parties and leaders in the past’’

(Clapham and Wiseman 1995: 226). As noted by Ottaway, democra-

tization requires a radical shift in the nature of political power.

Elections in weak and previously non-democratic states are therefore

unlikely to function as a vehicle for genuine political change. In fact

they may even contribute to the sedimentation of existing power struc-

tures through a ‘‘premature closure’’ of the transition process due to the

establishment of formal procedures and institutions before a real

change in the nature of power has taken place (Ottaway 1997: 3).

In post-war societies, this means sedimentation of the structures of

war. Through the inclusion of formerly warring groups, these parties

come to define which questions will dictate the post-war political

agenda. In many cases, these are the same, or similar, issues over

which the war was fought. Hence, there is a risk that post-war politics

will serve to reinforce those incompatibilities that the preceding peace

process sought to reduce or eliminate, and consolidate wartime alle-

giances and societal divisions. This is, for example, illustrative for the

situation in Bosnia, where nationalist parties like the Croatian

Democratic Union (HDZ), the Serb Democratic Party (SDS), and the

Bosniac Party of Democratic Action (SDA) play an important role in

hampering democratic development (Reilly 2002).

Second, the inclusion of formerly warring parties in post-war politics

often takes place at the price of granting these individuals and groups

amnesty for war crimes and human rights abuses committed during the

war. The rebels often declare that their participation in the negotiations

and their commitment to a peaceful solution to the armed struggle is

contingent on assurances of amnesty. In a large number of peace

processes, from Angola in 1991 to Aceh in 2005, armed groups have

thus been granted amnesty for war crimes in the peace accords along-

side recognition and legitimacy as political actors.3 As noted by Baker,

3 The general amnesty provisions are sometimes limited by certain exceptions.For example, in the 1997 peace accords in Tajikistan, the amnesty law calls forcriminal charges against former civil war combatants to be dropped withthe exception of certain crimes, such as the killing of non-combatants, rape,terrorism, and drug smuggling (ICG 2004b). Likewise, in the Aceh peace accordsof 2005, only those charges with offences relating to the political conflict, that is,for rebellion, will be eligible for amnesty, and not those charged with criminalcharges, such as rape, murder, or arson (ICG 2005c).

142 Mimmi Soderberg Kovacs

this practice places peace and conflict resolution before justice and

human rights and may be considered a trade-off with the principle of

rule of law that symbolizes a democratic political system (Baker 2001).

From a democratic point of view, this practice may therefore pose an

obstacle to the prospects of democratization in post-war societies. An

approach that instead places justice before peace and emphasizes the

need for prosecution for abuses committed during the conflict can serve

several functions that further a democratic culture (Kritz 2001:

808–809). It can provide victims with a sense of justice, put an end to

a culture of impunity, and provide an important focus for rebuilding

the judiciary and the criminal justice system in accordance with rule-of-

law principles. In addition, settlements that build on democratic foun-

dations ‘‘have a far better chance of achieving sustainable security,’’

while settlements that put conflict management before democracy and

human rights may have ended brutal civil wars but remain inherently

fragile (Baker 2001: 760–761).

However, it should be noted that although peace and justice may be

seen as theoretically separate goals, they are not necessarily mutually

exclusive in practice. There have been many attempts to reconcile

them, for example through the establishment of truth and reconcilia-

tion commissions. The most appropriate approach for the purpose of

furthering both peace and democracy will differ substantially from case

to case and is dependent on answers to questions such as who should be

held responsible? for what? and by whom? It is not necessarily the case

that bringing the offenders to justice is the strategy that is the most

conducive for either peace or democracy. The outcome hinges to a large

extent on the perceived legitimacy of the selected approach in the eyes

of the domestic population. As noted by Arzt, ‘‘if international criminal

courts are to achieve their aims – one of which is to contribute to the

consolidation of democracy and the triumph of the rule of law [. . .] –

perception of their legitimacy by the local population is a crucial

factor’’ (2006: 227).

For example, the decision of the International Criminal Tribunal for

the former Yugoslavia (ICTY) to issue an indictment for war crimes

against Kosovo’s democratically elected Prime Minister Ramush

Haradinaj in March 2005 was met with shock and anger among most

of Kosovo’s majority Albanians. The decision resulted in a number of

protests and campaigns that, according to observers, could have esca-

lated to violence had it not been for Haradinaj’s own public plea for

When rebels change their stripes 143

calm (ICG 2005b). Hence, the trade-off between amnesty and justice in

war-to-democracy transitions brings up issues that also strongly relate

to the other dilemmas discussed in this book, most notably the vertical

and the systemic dilemma.

In some cases, however, the question of amnesty for the sake of peace

versus justice seems to be tied directly to the issue of inclusion versus

exclusion. More precisely, its seems like participation equals protection

and non-participation equals prosecution. As long as the rebels comply

with the terms of the agreement that provide for their inclusion into

peaceful and democratic politics and succeed in their transformation to

political parties, they are unlikely to face trial no matter the extent or

severity of prior human rights abuses. When, however, the rebels fail to

live up to their prior commitment to peace, they are more likely to be

held responsible for crimes committed during the armed conflict in

spite of previous guarantees of amnesty.

The case of the Revolutionary United Front (RUF) in Sierra Leone is

illustrative. The 1999 Lome peace agreement granted the rebels ‘‘abso-

lute and free pardon and reprieve’’ for crimes and human rights abuses

conducted during the armed struggle. The agreement further stated that

the Government of Sierra Leone would ensure that no official or judicial

action would be taken against any member of the RUF in respect to their

actions during the war.4 However, the United Nations, which played a

vital role as mediator during the negotiation process, issued a disclaimer

at the time of the signing of the accord, which stated that the amnesty

provisions were not applicable to ‘‘international crimes of genocide,

crimes against humanity, war crimes and other serious violations of

international humanitarian law’’ (Francis 2000: 366). Subsequently, in

May 2000 the RUF ambushed and abducted hundreds of the newly

arrived UN peacekeepers and the peace process broke down. In a direct

response to this event, the Government of Sierra Leone turned to the

UN to ask for assistance in setting up a court ‘‘to try and bring to

credible justice’’ members of the RUF for crimes committed during the

armed conflict (ICG 2003: 2). The Special Court became operational in

2002 and the first indictments were issued in March 2003 against the

4 Article IX, Peace Agreement between the Government of Sierra Leone and theRevolutionary United Front of Sierra Leone, available at www.sierra-leone.org/lomeaccord.html. Last visited February 26, 2007.

144 Mimmi Soderberg Kovacs

leadership of several of the parties that had participated in the civil war,

including most of the top RUF leaders and commanders.5

A similar development took place in Cambodia following the Khmer

Rouge’s defection from the peace process and its failure to comply with

the terms of the Paris Agreement of 1991. Following the dictates of

pragmatism and realpolitik, the accord provided for the participation

of the former rebels as a legitimate political party in a post-war demo-

cratic system built on power sharing. However, in July 1994, after

Khmer Rouge’s decision to withdraw from the peace process, the

Cambodian National Assembly responded by outlawing the group.

Shortly thereafter, the United States Congress passed the Cambodian

Genocide Justice Act. These two events signaled the beginning of a

changed national and international approach toward the rebels. In

1997, at a time when the movement was at the verge of its final military

and political collapse, a process toward international justice was set in

motion by the Cambodian Government and the United Nations. In

2003, a tribunal was established to try the surviving leadership of the

Khmer Rouge for genocide and crimes against humanity during the Pol

Pot regime from 1975 to 1979 (Hammer and Urs 2005).

The vertical dilemma: popular legitimacy versus efficacyof governance

In cases where former rebels have entered post-war politics, a number of

issues can be identified that relate to the second overarching dilemma

that speaks to the trade-off between legitimacy and efficacy. These issues

center on the democratic deficits that often continue to characterize the

practices of these parties after entering democratic politics due to the

legacy of the armed struggle and their background as political-military

organizations. In addition, they relate to the changing relationship

between the former rebels and the population at large in war-to-democracy

transitions. The particular issues raised in each case depend, however, to

a large extent on whether the insurgent group capture government

power in the post-war elections or assume the role of opposition party.

These two different scenarios will be discussed in turn.

The problem of transforming armed movements into democratic

governments is not unique to the post-Cold War era. A large number

5 See www.sierra-leone.org/specialcourt/html. Last visited February 26, 2007.

When rebels change their stripes 145

of African liberation movements were declared political parties on the

eve of independence from colonial rule. As noted by Salih, many of

these parties and their leaders found it difficult to adjust to the account-

ability and transparency democratic politics is supposed to entail. They

hardly ever transformed on the basis of politics and remained essen-

tially reliant on personalistic and clientelistic mechanisms of internal

control within their parties and in their relationship with their electo-

rate. Interestingly, the experiences of the early post-colonial liberation

movements are in many ways echoed in more contemporary examples

of armed movements in Africa that have come to power through

negotiated settlements. These parties often blur the distinction between

the party and the state and they continue to be an embodiment of

nationalist/populist politics of the liberation struggle (Salih 2003:

12–14). What characterized these movements during the armed strug-

gle was the stress on unity and the need to speak with one voice for a

single purpose. Paradoxically, therefore, the more successful these

groups were in terms of the liberation struggle, the more difficult they

experienced the transition to become one political competitor among

many in a multi-party democracy (Ottaway 1991).

For example, while it has been acknowledged that the African

National Congress (ANC) has come a long way in its internal transfor-

mation since the first democratic elections in South Africa in 1994,

critical voices have been raised over some internal practices of the

party, and warnings have been issued concerning its electoral domi-

nance, which might indicate that South Africa is becoming a de facto

one-party state (Lanegran 2001: 99). Although the ANC cannot be

blamed for its electoral success, the combination of the party’s domi-

nance and its lack of internal party democracy is a cause for concern, as

this means that the fate of South Africa’s new democracy to a consider-

able extent is dependent on the party’s own democratic behavior

(Randall and Svasand 2002: 46). Likewise, it has been argued that

the South West Africa People’s Organization (SWAPO) in Namibia

still plays the tunes of the nationalist politics that brought it to power,

which is proving a serious obstacle for further democratic develop-

ments (Melber 2002).

In East Timor, the former resistance party the Revolutionary Front

of Independent East Timor (FRETILIN) won the Constituent Assembly

elections of 2001, following the vote for independence in 1999.

According to Hohe, the party drew heavily on indigenous values,

146 Mimmi Soderberg Kovacs

traditions, and symbolism in its electoral campaign. In doing so, the

party was able to appeal to traditional values of unity and hierarchical

political authority, and the election campaign served to rationalize the

emergence of a one-party state, which after the elections has been

rapidly consolidated by the party. ‘‘Fretilin’s linkage of these concepts

to the international multiparty exercise means, logically, that political

opposition can only be expressed in a hostile manner, and there is no

space for alternative and peaceful competition’’ (Hohe 2002: 85).6

After its entry into politics, FRETILIN has attempted to use the party’s

own symbols as symbols of the state, which may be interpreted as ‘‘a

dangerous precedent bespeaking an authoritarian tendency to conflate

the ruling party with the state itself’’ (Smith 2004: 153).

Former rebels and guerilla movements that have emerged as opposi-

tion parties following the outcome of the electoral process have also

displayed difficulties in adjusting to democratic politics. They need to

take into consideration the opinions and wishes of their former mem-

bers and loyal supporters and simultaneously attempt to attract votes

among a broader section of the population in order to survive in

politics. In these situations, many parties are more likely to remain

true to their wartime ideology and political program than to adjust

their political message to the new political and economic realities. In

many instances, their parliamentary behavior as opposition parties

often shows tendencies of following the principles of a ‘‘reversed

Clausewitz,’’ where politics becomes the continuation of war by other

means.

Manning has argued that the character and degree of commitment

that former armed groups make to the post-war political system

depends on the kinds of challenges that adaptation to the new environ-

ment presents for them as organizations. Specifically, she argues that

parties that are forced to make significant adjustments to their

6 According to Smith, the only real opposition exists in the form of PresidentXanana Gusmao, the former leader of the FRETILIN’s armed wing and later theleader of the united resistance front formed in 1998, and the loyalty he commandsfrom the armed forces. Since the landslide victory of Gusmao in the presidentialelections of 2002, the conflict between FRETILIN and Gusmao has seen noimprovement and open disagreement has surfaced on a number of political issues(2004: 154–155). In addition, unofficial security elements, possibly linked toFRETILIN, have emerged. Thus, ‘‘the conditions have been laid, therefore, for thestark alternatives of either one-party rule or violent political competition’’ (Hohe2002: 85).

When rebels change their stripes 147

collective incentive strategies in order to compete politically and sur-

vive in post-war politics also invest in necessary changes in their inter-

nal authority structures. In addition, they also make more investments

in the procedural rules of the democratic game and thus make durable

democracy more likely. In contrast, in situations characterized by

the lack of vigorous political competition, these changes rarely take

place and the prospect of democratic progress is much less likely

(Manning 2004).

For example, Manning argues, Mozambiquan National Resistance

(RENAMO) in Mozambique has been subject to very little pressure to

change its collective incentives strategies after its entry into parliamen-

tary politics over a decade ago. The social and political cleavages that

characterized the county during the war carried over into the post-war

period, and the apparent failure of the government to improve living

conditions for the large part of the population that lives in the areas

from which RENAMO draws its support reinforces this picture. In

addition, Mozambique’s essentially two-party system poses an obsta-

cle to the emergence of viable political competitors. Hence, RENAMO

continues to fight the armed struggle from within parliamentary politics

and remains more anti-government than in favor of any well-defined

and clearly spelled-out political or economic agenda. In addition, it

attracts votes mostly among its former wartime constituencies in the

central and northern parts of the country. This lack of incentives to

change its appeal to the population has led to little pressure to address

questions of intraparty nature, and the party has eluded any significant

changes concerning its own internal organization. The old leadership

continues to control the party in a hierarchical fashion that has seri-

ously hampered the party’s democratic performance (Manning 2004).

In contrast, the Kosovo Liberation Army (UCK), which became the

Democratic Party of Kosovo (PDK), was obliged to make significant

adjustments to its collective incentives strategies and broaden its mes-

sage and identity in order to compete in post-war politics, as the

political cleavages that defined the armed conflict no longer defined

the political arena. This also allowed new party competitors to emerge

and the PDK was forced to invest in changes to its internal authority

structures, for the purpose of strengthening its position toward other

parties in the political system. Consequently, the party has adapted

more quickly to electoral politics and adjusted itself to the rules of the

game (Manning 2004).

148 Mimmi Soderberg Kovacs

In the cases of the former communist guerillas in Central America,

one of their central ideological characteristics during the armed conflict

was precisely their anti-systemic nature. For this reason the process

of integration into the political system has often been both slow

and complex (Cerdas Cruz 1998: 46). For example, although the

former communist guerilla in El Salvador, the Farabundo Martı

National Liberation Front (FMLN), has been comparatively successful

in its electoral performances following its entry into parliamentary

politics, it has struggled with internal disagreements over issues

of ideology and party governance. The internal divisions of the

FMLN emerged shortly after the first post-war elections in 1994,

when two of the five organizations that made up the FMLN left the

party over such a dispute. According to Wade (2007), this split signaled

the beginnings of the party’s post-war difficulties in keeping a united

front. Since then, she argues, the FMLN has been characterized by

factionalism between those who want to remain true to the character

of the revolutionary movement and favor a vertical decision-making

structure demanding strict adherence to the party line, and those press-

ing for more pragmatic and pluralistic policies and increased intraparty

democracy.

In Guatemala, the Guatemalan National Revolutionary Unity

(URNG) has suffered from the lack of an ideological and programmatic

identity following its transition to a legal political actor. The party has

remained cautious on ideological issues and unsure of how to identify

itself in the political system in general and to the Guatemalan left in

particular. Vinegard suggests that this absence of ideological and poli-

tical clarity has contributed to the URNG’s difficulties in establishing

itself as an effective voice of opposition to the government. In addition,

the hold of URNG’s traditional leaders on the party structure, and its

seeming lack of tolerance for internal dissent, has given reason to

question the party’s capacity for internal democratization (Vinegard

1998: 223–225).

The systemic dilemma: domestic versus internationallegitimization of parties

The transformation of warring actors to legal political parties in war-

to-democracy transitions also raises the question of who has the right

to determine which formerly warring groups are to be considered

When rebels change their stripes 149

legitimate political actors and therefore should be granted a role in

post-war politics. This question relates to the systemic dilemma, which

concerns the trade-off between international and domestic ownership

over processes of peacebuilding and democratization.

Because the transformation of former rebels to political parties has

emerged as a commonly recognized tool for conflict resolution and

democratization in many internationally supervised peace processes,

international actors have spent a considerable amount of resources on

these processes of rebel conversion in cases where this has been deemed

necessary. For example, in the Mozambique peace process, a UN trust

fund was established and eventually US$17 million was raised for it to

help RENAMO to transform itself into a political party, as the inter-

national community realized that the rebels’ participation in post-war

politics and their continued commitment to the peace process depended

on the provision of external resources (Vines 1996). In addition

to providing financial assistance, donors have also frequently con-

tributed with technical assistance and training to armed groups to

build their capacity to select candidates, organize election campaigns,

and monitor election outcomes. Assistance to these new parties has in

some instances also continued in the post-election period (Kumar

1998c: 218).

In some cases, international processes of legitimization of armed

groups have been aborted or reversed during the implementation pro-

cess due to the rebels’ inability or unwillingness to comply with the

terms of the agreement. This was, for example, the case with the RUF in

Sierra Leone and the Khmer Rouge in Cambodia. International actors

thus play a critical role in determining the fate of these groups. The

international community’s considerable influence over the dynamics

and outcomes of these processes of rebel conversion raises issues of

concern over ownership. Does the process of international legitimiza-

tion (or marginalization) of certain warring groups always comply with

the perceptions and opinions of the domestic audience whose political

future is at stake?

The case of the Lome Accord in 1999 between the Government of

Sierra Leone and the RUF is illustrative of a case where the strategy of

the international community stood in stark contrast to the wishes of the

majority of the local population. International mediators and donors

put strong pressure on the Government of Sierra Leone to extend

political recognition and legitimacy to the RUF in order to facilitate a

150 Mimmi Soderberg Kovacs

negotiated settlement to end the armed conflict. This was in spite of the

fact that most government representatives, including the president,

initially were very reluctant to do so, and regardless of the fact that

most evidence from the prolonged civil war seemed to suggest that the

rebels had very little support among the population for their cause.

According to the terms of the agreement, the rebels were granted a

considerable amount of political influence over state affairs and a

general amnesty for war crimes (Bangura 2000: 564–565). It was not

until after the UN hostage-taking event in May 2000 that international

opinion shifted toward a strategy of military and political marginaliza-

tion of the rebels, a strategy that better reflected the preferences of the

Sierra Leonean population. This conclusion may be drawn on the basis

of the outcome of the post-war elections in 2002, where the fraction of

the former rebel group that had turned into a political party only

received 1.7 percent of the popular vote (Kandeh 2003).

The situation might also be the reverse. When the Office of the High

Representative (OHR) in Bosnia Herzegovina has stepped in and dis-

missed democratically elected political actors and parties with a

nationalist agenda because they have been considered to be an obstacle

or a threat to peace and democracy in the country, it clearly violates the

democratically expressed opinions of the domestic population. When

Hamas, who publicly opposes a peace deal with Israel, was asked to

form a new government in the Palestinian territories following its

electoral success in the parliamentary elections in early 2006, the

international community faced a similar problem. In spite of having

come to power through democratic elections, both Israel and the

United States, who has branded the group as a terrorist organization,

decided to withhold funds previously channeled to the Palestinian

administration. Some critics argue that such a strategy might be coun-

terproductive, as it risks alienating the organization and may push

it even further in a non-democratic direction (ICG 2007b). Such a

development would seriously hamper any future prospects for peace

in the region.

The dilemma concerning the trade-off between external versus inter-

nal control over critical choices in war-to-democracy transition gives

no easy clues as to which approach truly is the most beneficial for either

peace or democracy. However, without considering the legitimacy of

the choices made, the prospects for durable peace and self-sustaining

democratic progress are likely to be in jeopardy.

When rebels change their stripes 151

The temporal dilemma: short-term versus long-term effectsof rebels in politics

Beyond posing an obstacle to democratic development, a continued

lack of good governance and democratic progress in new democracies

poses a long-term risk for sustainable peace and may contribute to

renewed armed conflict. This problem speaks to the temporal dilemma

between efforts to advance both peace and democracy following an

armed intrastate conflict.

In post-war societies, the lack of democratic progress is especially

problematic, as the legitimacy of the peace process is contingent on the

display of significant improvement in people’s daily lives. Unless peace

brings significant peace dividends, both the people and the warring

parties will start to question the value of the current peace accord and

the political order that emerged from it. The shortcomings of the

democratic system can thus provide fuel for new grievances. Besides,

in addition to the often conflict-ridden transitional phase, semi-

democracies, or states that fail to move beyond the initial buildup of

democratic institutions and where the underlying structures of politics

remain autocratic or neo-patrimonial in nature, are more prone to

armed conflict than are both democratic and autocratic states (Hegre

et al. 2001). This further underscores the gravity of the lack of demo-

cratic progress in post-war societies, and its implications for war

recurrence.

In Liberia, the post-war elections of 1997 and the developments that

followed in the wake of former warlord Charles Taylor’s electoral road

to the presidency are ample evidence of the possibly of such a scenario.

In the elections, Taylor and his former army-turned-party, the National

Patriotic Front of Liberia (NPFL), converted into the National Patriotic

Party (NPP), won the national vote with a landslide. Although the

elections were supposed to mark the end of seven years of brutal civil

war in Liberia, they also marked the beginning of a deteriorating

political and security situation that would ultimately culminate in the

outbreak of a new civil war in early 2000. After assuming power,

Taylor openly declared that he no longer considered himself committed

to the conditions of the Abuja peace accords. He refused to carry out

the reconstruction of the national army, and he took advantage of

his new power platform to oppress the political opposition in the

country. Soon thereafter, the new regime began to show the same

152 Mimmi Soderberg Kovacs

authoritarian attributes and the same social and political demise that

had characterized the pre-war authoritarian regime.

Subsequently, in the early months of 2000, reports of the founding of

a new rebel movement began to circulate in Liberia. The group

announced itself as Liberians United for Reconciliation and

Democracy (LURD) and consisted mostly of political opponents to

Taylor; militia groups, former politicians, and Liberian refugees that

Taylor’s politics had forced out of the country. This marked the begin-

ning of a new civil war that did not come to a conclusion until August

2003, when Taylor was forced to leave the country and step down from

power. There is little doubt that the outbreak of the civil war in early

2000 was directly linked to Taylor’s misuse of his role in government

power and the lack of democratic reform that had characterized the

post-Abuja political order (Nilsson and Soderberg Kovacs 2005).

The example of Liberia under Taylor strengthens the notion that

efforts to advance sustainable peace must be accompanied by efforts to

advance democratic values in war-shattered societies. However, it also

points to the difficulties that may arise when the two processes of

conflict resolution and democratization are introduced simultaneously

in a war-scattered society. Under such circumstances, the value of

peace (in terms of absence of war) may clash with that of democracy

and sustainable peace in a longer time perspective.

Conclusions

The purpose of this chapter has been to address and discuss the pro-

spects for democratization and sustainable peace in war-shattered

societies in which armed insurgents emerge as political parties in the

post-war political order. In the last decades, the transformation of

armed groups into political parties has become an integral part of

peace efforts aiming at ending civil wars. This development has

reflected findings emanating from the war termination and conflict

resolution literature, in which the transformation of warring groups

to political parties has been identified as one of the contributing factors

to the successful implementation of peace agreements in civil wars. The

theoretical rationale for such an approach has been based on the notion

that if civil wars are seen as generally emerging from the mobilization

of legitimate grievances on part of the population, the peace processes

and the post-war system should aim to address and mitigate these

When rebels change their stripes 153

injustices in order to create a just, and therefore stable, peace. In

addition, the successful implementation of civil war peace agreements

is contingent on the continued commitment of the warring parties to

the peace process and their incentive for doing so increases consider-

ably when granted a legitimate political role. Because contemporary

peace processes in civil wars frequently include a parallel transition to

democratic politics as part of the conflict resolution efforts, this trend

has also fulfilled an additional purpose, namely to provide the post-war

state with the political parties necessary to introduce a political system

based on the principles of multi-party democracy. Although the limita-

tions of political parties in new democracies have been widely acknowl-

edged, political parties are generally still seen as the key agents of

democratization and are expected to perform a wide variety of func-

tions deemed essential to the democratic political system.

This chapter has pointed to some tentative yet intriguing findings in

regard to this trend of rebel reappearance in post-war politics.

Although the transformation of formerly armed groups to political

parties has proven critical to the success of many peace processes in

the last decades, the inclusion and participation of these groups in post-

war politics has in several instances proven problematic for the demo-

cratization process. In terms of the horizontal dilemma, two issues in

particular were identified. The inclusion of former warring parties in

post-war politics often takes place at the expense of including new

political parties into the political process. Immediately following the

end of a civil war, this might be inevitable for the purpose of ending the

armed conflict and due to the lack of viable alternatives. However,

there is a great risk that this leads to sedimentation of the political

cleavages that defined the society during the armed conflict and the

issues that created the wartime alignments. A possible strategy for

overcoming this problem could be found in the use of transitional

arrangements, for example in the establishment of a transitional gov-

ernment that includes the warring parties for a predetermined and

limited time period. Only toward the end of the interim period, post-

war elections are held to determine the future distribution of power in

the country. This was, for example, the approach chosen in Liberia

following the 2003 peace agreement.

Another issue of concern is the use of general amnesty provisions for

former combatants in exchange for laying down their guns and com-

mitting to the peace process. In many cases this has been considered a

154 Mimmi Soderberg Kovacs

necessary condition for ending the war. From a democratic point of

view, however, such exemptions from the rule of law and the principles

of human rights risk encouraging a culture of impunity. However, it is

equally true that an emphasis on justice may impinge on the emerging

democratization process if the procedure is not considered legitimate

by the population at large, which was illustrated by the case of the

ICTY indictment of Prime Minister Haradinaj in Kosovo in 2005. In

some cases, amnesty has in reality been contingent on the parties’ post-

settlement behavior and their commitment to a peaceful process. In the

cases of both Sierra Leone and Cambodia, the peace agreements pro-

vided the RUF and the Khmer Rouge respectively with controversial

amnesty provisions and a legal status as political parties. These provi-

sions were later revoked following these groups’ repeated failures to

live up to the terms of the agreements. What may be considered the

most appropriate approach will thus most likely have to be settled on a

case-to-case basis. However, a general strategy that equals participa-

tion with protection and non-participation with prosecution risks

undermining the very values that these processes of international crimi-

nal law are intended to uphold.

In terms of the vertical dilemma, the inclusion of formerly armed

groups into positions of political power has in some cases resulted in

democratic deficits due to these groups’ organizational legacies as

political-military organizations originally created for a different pur-

pose. Many groups with their origins in the armed struggle have been

forced to make critical trade-offs between keeping their legitimacy in

the eyes of their wartime constituencies and the need for making

broader appeals to a larger part of the population, and between keeping

the structures of the original group intact and adjusting to the demo-

cratic demands of accountability, transparency, and efficacy. These

issues may be difficult to overcome. However, an increasing awareness

of this dilemma and its implications for the post-war order might help

to adjust the democratic expectation on these groups as they enter

peaceful politics.

The transformation of warring parties also raises the critical ques-

tion of who has the right to determine which groups are to be consid-

ered legitimate political actors. This issue relates to the systemic

dilemma and the trade-off between international and domestic owner-

ship of the war-to-democracy transition. This chapter suggests that the

process of international legitimization or marginalization of certain

When rebels change their stripes 155

warring groups needs to reflect the perceptions and opinions of the

concerned domestic population in order to work to the benefit of both

peace and democracy in that country.

Finally, in regards to the temporal dilemma, the lack of democratic

progress in a post-war society is problematic not only for the sake of

democracy, but also for peace, as failed governance also serves to

undermine the long-term prospects for durable peace. The case of

Liberia following the elections in 1997 is illustrative. Shortcomings in

the democratic process and lack of visible improvements following the

transition to democracy are especially problematic in post-war socie-

ties, where there already is a high risk of war recurrence due to the

extreme polarization of society and the heightened sense of insecurity.

At the same time, the conditions for a successful democratization

process and democratic consolidation are less favorable in post-war

societies than elsewhere, due to the legacies of the war.

This dilemma speaks to the critical need for a long-term policy of

engagement from the side of concerned local and international peace

custodians in war-shattered societies. The construction of comprehen-

sive peace agreements which attempt to reconcile a multitude of differ-

ent and sometimes competing goals, including that of ending violence

and initiating a process toward greater political liberalization and

democratization, might not, however, be the most conducive strategy.

The construction of such all-encompassing peace and democracy

packages does not ensure the viability of these processes, nor their

outcomes. Perhaps a better strategy would be to introduce these differ-

ent goals (and the appropriate methods and tools for reaching them) in

a step-by-step manner in a phased process. There is no readily available

formula for such a process, but the purpose would be to avoid or

mitigate some of the dilemmas that may arise when the values of

peace and democracy are aimed at simultaneously.

156 Mimmi Soderberg Kovacs

6 Post-war elections: uncertain turningpoints of transition

B E N J A M I N R E I L L Y

Elections held as part of a peace deal following a violent conflict high-

light several crucial dilemmas of democratization in post-war societies.

Such ‘‘post-war elections’’ are now a feature of almost all efforts to

democratize war-torn regions, with peace agreements routinely includ-

ing provisions for elections to be held as part of the process of conflict

termination, often with the assistance, supervision, or sometimes direct

control of the international community. But while post-war elections

have become an integral element of contemporary peace agreements,

they can also themselves become the focus of increasing tension and

renewed violence. Taking a comparative perspective, this chapter

focuses on several inherent dilemmas of post-war elections, including

issues of timing, sequencing, mechanics, political parties, and the role

of the international community. In each of these areas, post-war

elections force difficult choices to be made between short-term versus

long-term priorities, representation versus stability, domestic versus

international legitimacy, and a range of other sometimes incompatible

objectives.

These dilemmas are reinforced by the competing discourses that

dominate both academic and policy discussions of post-war elections.

On the one hand, elections and democracy are often seen as a primary

means of conflict management, with theorists arguing for the benefits

of democratic competition in managing the tensions inherent in all

societies, including war-torn ones (see Przeworski 1991). On the

other hand, an increasing body of work points to the dangers of holding

elections in conflict-prone societies, and the empirical reality that

societies in the early stages of democratization are often more, not

less, conflict prone (Snyder 2000). The role of the international com-

munity is another important factor. Competitive elections in post-war

societies are often promoted by international actors for a range of

varying and even mutually contradictory reasons. Typically, such elec-

tions are expected to play a role in simultaneously terminating civil

157

wars; encouraging the transformation of warring armies into peaceful

political parties; stimulating the development of ‘‘normal’’ politics;

choosing members of a legislature or other kind of representative

assembly; forming a government; and conferring legitimacy upon

the new political order. For all of these reasons, post-war elections

are today widely seen as an integral part of the process of war termina-

tion, international disengagement, and nation building.

However, the success of post-war elections in achieving these goals

has varied considerably. In some cases, such as Namibia in 1989, El

Salvador in 1994, or Mozambique in 1994, elections clearly played a

vital role in making a decisive break with the past. In others, such as

Angola’s abortive 1992 elections held under the Bicesse peace accord,

or Liberia’s 1997 elections, flawed elections created more problems

than they solved. Haiti’s parliamentary and presidential elections in

1995 led to a first-ever transition of power but also highlighted admin-

istrative deficiencies which undermined the credibility of the broader

electoral process. By contrast, in Cambodia’s United Nations adminis-

tered polls of 1993, the technically successful elections were soon

overwhelmed by the realities of power politics as the ‘‘losing’’ party at

the elections returned to power through hard-line tactics. In post-war

Bosnia, successive elections held under the Dayton Peace Accords

helped nationalist parties cement an early grip on political power,

while in Kosovo and East Timor a more measured electoral timetable

played a constructive role in terms of political development.

More recently, in Papua New Guinea’s rebellious island province of

Bougainville, the combination of an extended electoral timetable,

international observation, and systemic innovations have helped secure

one of the world’s most successful, if little-known, cases of post-war

peacebuilding. Similarly, Liberia’s 2005 elections marked the end of

the transition following the country’s second civil war and resulted in

Africa’s first democratically elected female head of state, former World

Bank employee Ellen Johnson-Sirleaf. Finally, while it is still too early

to evaluate recent high-profile elections in Afghanistan and Iraq, in

both cases it is clear that elections themselves have not led to an end to

hostilities, and in Iraq may have contributed to the ongoing sectarian

conflict.

As this brief survey indicates, there has been a considerable variation

in the relative success of elections in meeting the twin goals of war

termination and consolidation of democracy from country to country

158 Benjamin Reilly

and from case to case. The tension between these two overarching goals

lies at the heart of this particular dilemma of democratization. As

Terrence Lyons (2004: 272) notes, because post-war elections typically

carry such tremendous burdens of expectation, they are often saddled

with multiple and often mutually contradictory objectives:

They are designated in peace agreements as a primary instrument of imple-

mentation and hence play critical goals with regards to war termination. At

the same time, they are designed to promote a process of democratization and

to serve as ‘‘breakthrough’’ elections that initiate a new set of rules and

institutions for competitive, multiparty politics . . . Success with relation to

one goal, say war termination, does not necessarily mark ‘‘success’’ relative to

another, such as democratization.

In this chapter, I address some of the core dilemmas confronting

post-war elections that flow from these multiple and sometimes contra-

dictory goals.1 I focus in particular on those specific dilemmas which

tend to recur across both space and time: the inherent tension between

competitive elections and conflict management; the incompatibility of

short- versus long-term electoral objectives; the trade-off between effi-

ciency and inclusion in terms of government structure; the merits of

sequenced versus simultaneous local, regional, and national-level elec-

tions; the choice between party-based and independent forms of elec-

toral administration; and the need to build local accountability while

encouraging the development of national party politics.

Democratic dilemmas and post-war elections

The overarching question facing all post-war elections is under what

circumstances they help in building a new, peaceful, democratic order,

and under what circumstances they undermine prospects for stable

democracy and pave the way for a return to conflict. As one survey

notes, the high expectations often placed upon post-war elections tend

to be accompanied by a weakness in the preconditions for their success:

‘‘most war-torn societies lack the political climate, social and economic

stability, institutional infrastructure, and even political will to mount

successful elections’’ (Kumar 1998a: 7). The international community

1 This chapter also builds upon some of my earlier work focusing on other aspectsof post-war elections: see, in particular, Reilly 2002 and 2004.

Post-war elections 159

has often not been sufficiently cognizant of the dangers in pushing for

early post-war elections, particularly in countries which have recently

emerged from civil war, or given sufficient attention to the capacity

of the host country to carry them out. On the other hand, the promise

of early elections is often essential in getting commitments from major

powers to deploy peacekeepers and fund post-war reconstruction.

There are fundamental difficulties with holding competitive elec-

tions following a period of violent conflict. In such situations, a com-

bination of acute coordination problems, information asymmetries,

hardening of societal divisions, and fears for the future typically con-

front voters. One consequence of this can be a profound ‘‘security

dilemma’’ which afflicts both voters and candidates, whereby compet-

ing ethnic, religious, and political actors will often mobilize against the

possibility of future threats, triggering a cascading tit-for-tat escalation

and polarization from other segments of society. In many cases, rising

levels of internal conflict have accompanied or been precipitated by

transitions from authoritarian rule toward democracy. Despite their

essential role, post-war elections have often fomented these tensions,

becoming a lightning-rod for popular discontent and extremist senti-

ments (Dahl 1971; Horowitz 1985; Rabushka and Shepsle 1972).

Bosnia’s repeated post-Dayton elections held in 1996, 1998, 2000,

and 2002 were an illustration of this process in action, as voters from

different ethnic communities persistently re-elected hard-line national-

ist leaders despite overt attempts by the international community to

encourage moderate, pro-Western victors instead. The victory of

Hamas, which the United States considers a terrorist organization, at

the January 2006 elections to the Palestinian Authority is another case

in point.

Electoral competition in such circumstances often turns on the poli-

tics of ‘‘outbidding’’ – that is, competition for votes on the basis of

reciprocally heightened extremist rhetoric – increasing ethnic tensions,

and the polarization of the political spectrum (Rabushka and Shepsle

1972: 187). Taken to an extreme, outbidding cycles can lead to the

outright failure of democracy, as minorities are excluded from power,

restrictions are placed on opposition movements, and the ‘‘rules of the

game’’ manipulated to benefit incumbents. For instance, political out-

bidding over access to higher education was an early motivator for the

Sinhalese–Tamil conflict in Sri Lanka, fomenting a cycle of escalating

ethnic hostilities which led directly to the tragic civil war and erosion of

160 Benjamin Reilly

democracy there (DeVotta 2004). In other cases, by electing hard-line

leaders committed to exclusionary visions of the country, post-war

elections can become little more than ethnic censuses which increase

the risk of conflict, threatening, ultimately, the failure of democracy

itself – a recurrent problem that has been well documented in ethnically

divided societies.2

In this view, the processes of electoral democracy in post-war socie-

ties can represent a danger to peaceful state building. Increasing recog-

nition of these dangers has led some to contend that democracy itself is

part of the problem in such highly fraught situations, and that post-war

societies are too fragile to be exposed to the competitive pressures of

the electoral process.3 But this oft-heard critique ignores several fac-

tors. First, elections can be purposively designed to encourage not zero-

sum, winner-take-all outcomes, but rather the sharing of power

between groups. Indeed, many would argue that some form of power

sharing is a primary requirement for successful democratization in

post-war situations.4 Second, post-war polities face a real need to

construct a legitimate governing authority. Not least because so many

of today’s conflicts take place within states, the overarching challenge

is thus to build (or rebuild) a state that can function without direct

international involvement. Elections can be a crucial element in achiev-

ing this, providing that sufficient forethought is given to their purpose,

timing, and likely effects.

Post-war elections and international policy

Post-war democracy building is a difficult and disruptive process.

Democratization by its very nature undermines established political

orders, provides a pathway for new entrants to access the political

system, highlights social cleavages, subverts existing power relations,

and threatens incumbent authority. For all of these reasons, transitions

to democracy in general and competitive elections in particular have the

potential to be deeply destabilizing events. As Roland Paris (2004: 1)

observes, ‘‘the process of political and economic liberalization is inher-

ently tumultuous: It can exacerbate social tensions and undermine the

2 Horowitz 1985. Recent research on India has improved our understanding of thisprocess: see Wilkinson 2004 and Chandra 2004.

3 For an example of this sentiment, see Chesterman 2004. 4 See Sisk 1996.

Post-war elections 161

prospects for stable peace in the fragile conditions that typically exist in

countries just emerging from civil war.’’ Put simply, post-war demo-

cratization is a difficult, uncertain, and often dangerous business.

Despite this, over the past decade many Western policymakers

adopted a facile and naıve interpretation of democratic elections as

being a natural and unproblematic form of conflict resolution, a pro-

cess which began in the 1990s and may have reached its nadir with the

truly extravagant claims regarding the beneficent impact of democra-

tization and free elections made by the Bush administration in the post-

September 11 era. Post-war elections in Iraq, for instance, were not just

a means of choosing representatives but also highly symbolic events

signaling the establishment of a new political order. In other cases such

as East Timor, post-war elections have even marked the assumption of

a new nation into the family of international statehood. However,

post-war elections can also be highly fraught exercises which highlight

many of the dilemmas of democratization examined in this volume.

While elections certainly may play a role in settling violent conflicts

and creating new political orders, they will not inevitably do so, and the

multiple goals that post-war elections are supposed to achieve means

that they are often overloaded with inconsistent and sometimes

mutually conflicting objectives, such as ending armed conflict and

simultaneously promoting vigorous political competition.

The uncritical alignment between democratic elections and national

peace in the minds of many policymakers has its recent origin in the

so-called ‘‘democratic peace’’ thesis much cited by former US President

Bill Clinton during his term in office, which maintains that consoli-

dated democracies both have not historically and will not in the future

go to war with each other (Russett 1993). This thesis, which has strong

empirical support but shaky theoretical foundations, was conjoined

with the separate but related argument that democracies were more

peaceful internally as well, and that intrastate conflicts were thus less

prevalent and severe in democratic than autocratic environments (see,

for example, Gurr 2000b: 52–64). Both arguments have since been

overtaken in US policy by a broader conflation between democracy,

liberalism, and security, which was elevated into an article of faith in

the Bush administration – a conflation of hopes and beliefs which rests

on many unspoken assumptions. One is the expectation that a move

from authoritarian to democratic governance will inevitably lead

to more peaceful inter-communal relations and lower levels of conflict

162 Benjamin Reilly

in societies divided along ethnic, religious, or other lines. Another is

that democratic elections are the most reliable means of generating

moderate governments which do not pose a threat to others, or at least

to the West.

The evidence for both claims is (unsurprisingly) more complex.

While there is little question that consolidated democracies are, on

average, both less prone to large-scale internal conflict and much less

likely to go to war with each other than their authoritarian counter-

parts, historical and comparative research has repeatedly found that

countries undergoing the wrenching process of democratization are

neither. As one such study concluded, ‘‘while mature, stable democra-

cies are safer, states usually go through a dangerous transition to

democracy. Historical evidence from the last 200 years shows that in

this phase, countries become more war-prone, not less, and they do

fight wars with democratic states’’ (Mansfield and Snyder 1995b: 79).

Many indicators of intrastate conflict also tend to rise in the initial

period of democratization (de Nevers 1993).

A major insight of political science scholarship in recent years thus

concerns the dangers of early democratization in fragile states.5 In such

circumstances, electoral competition can quickly come to be character-

ized by centrifugal pressures, in which the moderate political center is

overwhelmed by extremist forces, leading to zero-sum politics in which

some groups are permanently included and some permanently

excluded. One consequence can be the failure of democracy itself, as

majority group hegemonic ‘‘control,’’ often aided by the assumption of

martial law or outright military rule, is justified in part by the need to

restore order and stability. This is often accompanied by the exclusion

of minorities, changes to the rules of the game to benefit incumbents,

restrictions on opposition movements, and the undermining of demo-

cratic institutions. The failure of Indonesia’s first, abortive experience

of democracy in the 1950s period is a good example of this dilemma of

early democratization before civic institutions developed. Indonesia’s

fall into authoritarian rule was largely a response to the chaotic experi-

ence of democracy between 1950 and 1957, when a combination of

5 In a book-length analysis of this phenomenon, Snyder (2000) argues thatdemocratization is most likely to stimulate internal conflict when elites arethreatened by rapid political change and when the expansion of popularparticipation precedes the formation of strong political institutions.

Post-war elections 163

religious, cultural, and regional conflicts, combined with a fragmented

party system and weak institutions, led to the declaration of martial

law by President Sukarno. Shifting coalitions of secular, Islamic,

nationalist, communal, and regional parties had precipitated six

changes of government in seven years, providing a ready pretext for

the overthrow of democracy and forty years of authoritarian rule under

the Sukarno and then Suharto regimes (Liddle 1997: 311).

Rapid political change and insecurity are almost endemic to the

circumstances in which post-war elections are held. Domestic political

institutions are weak or non-existent, voters are suspicious, and elites’

hold on power is tenuous. This is a dangerous combination of factors.

At different times during the 1990s, for instance, post-war elections in

war-torn African states such as Angola and Sierra Leone led to a

resumption of warfare as a result, in part, of the threats these elections

represented to incumbent elites. Likewise, the prospect of forthcoming

elections and ethnic power sharing in Rwanda has been identified as a

factor in the 1993 genocide there.6 Elsewhere, in Cambodia, Bosnia,

and Liberia, post-war elections cemented in power essentially non-

democratic elites – often the very same individuals who had been

instrumental in the prior conflict. Finally, in post-war Rwanda,

Ethiopia and Uganda, the winning parties to the conflict held elections

to legitimize their victory – thus subverting the conflict-mediating

functions which such elections were supposed to perform. Cases such

as these lend some weight to the World Bank’s suggestion that elections

in war-torn societies should be deferred by up to a decade to allow state

building to occur (World Bank 2003a).

None of this should be taken to mean that democracy is a negative

factor for the management of internal conflicts. By providing an insti-

tutional framework for diverse social groups to gain access to govern-

ment, participate in decision making, and influence policy outcomes,

democracies are capable of responding to societal conflicts by accom-

modation rather than repression, in sharp contrast to authoritarian

6 Paris (2004: 75–76), for example, writes that ‘‘the effort to move Rwanda in thedirection of democracy did not have the pacifying influence that internationalpeacebuilders had hoped for and apparently expected. The internationalcommunity had presented the plan for power sharing followed by democraticelections as a means of resolving Rwanda’s civil war, but attempts to foster peaceand stability in Rwanda by promoting political liberalization ultimatelybackfired, in the worst possible way.’’

164 Benjamin Reilly

regimes. This is one reason that theorists like Adam Przeworski char-

acterize democracy as a political arrangement which processes, but

never definitely resolves, social conflicts (Przeworski 1991: 10–14).

Under this interpretation, a functioning democracy serves as a system

of conflict management, with potential conflicts channeled into con-

stitutional arenas, such as non-violent competition between political

parties, rather than armed conflict on the streets. These arguments have

been buttressed by empirical studies which emphasize the success of

consolidated democracies in accommodating social cleavages and ten-

sions through peaceful means (Hegre et al. 2001: 33–48).

However, Przeworski’s arguments also highlight one – indeed, per-

haps the most fundamental – of the dilemmas facing post-war elec-

tions: the problem of uncertainty. Consolidated democracy works to

manage conflicts precisely because electoral outcomes are, by defini-

tion, uncertain: the players of the game do not know the outcome,

but commit to the game with the understanding that they may lose in

the short term but still be winners later on. Democracy requires this

certainty of uncertainty to engender loyalty from all players and thus to

survive over the long term. But in post-war societies, the uncertainty of

election outcomes is itself a source of tension, and a major threat to

incumbent elites which can make them wary of committing to the game

at all. For this reason, some kind of pre-election bargain on post-

election outcomes is often required, as was the case in South Africa’s

transition from apartheid, when strong constitutional guarantees

helped to reassure the key players (Sisk 1995).

The aftermath of the 1993 elections in Cambodia is a good example

of this dilemma in action. The culmination of the largest and most

expensive UN peacekeeping mission to date, the massive international

reconstruction of Cambodia reached its apex in 1993 when the United

Nations both ran and oversaw an electoral process which was techni-

cally almost faultless, but which resulted in the election of two main

parties, the Cambodian People’s Party (CPP) and the National United

Front for an Independent, Neutral, Peaceful, and Cooperative

Cambodia (FUNCINPEC), each of whom had expected to control

power alone. Amid threats of renewed civil war from the incumbent

CPP if it was excluded from government, a clumsy post-election power-

sharing deal brokered by the United Nations saw a coalition govern-

ment featuring ‘‘co-prime ministers’’ from the two parties installed. As

it reflected neither the election results nor common policy ground

Post-war elections 165

between the two parties, this arrangement proved highly unstable in

practice: the CPP remained in effective control of most of the armed

forces, the bureaucracy, and the judiciary, while FUNCINPEC’s

attempt to gain a greater share of real power paralyzed the executive

branch and the National Assembly. After a prolonged succession of

political crises, the coalition fell apart completely in 1997 when the

CPP forces of the ‘‘second Prime Minister,’’ Hun Sen, attacked those of

FUNCINPEC and the ‘‘first Prime Minister,’’ Prince Ranariddh, and

claimed power alone.

As Cambodia returned to its familiar politics of intimidation and

authoritarian rule, Hun Sen proceeded to change the electoral system

to benefit his government and restrict opposition movements. The

electoral formula was changed so that seats were allocated according

to the ‘‘highest average’’ method at the provincial level, rather than the

nationwide ‘‘largest remainder’’ system introduced by the United

Nations in 1993 – a change which wiped out smaller parties. In

response to calls for greater local accountability, district boundaries

were adjusted and a number of new districts created with the result that

over one-third of all seats were chosen from single-member districts.

These technical changes were accompanied by widespread intimida-

tion of opposition politicians and their supporters. The net effect was

the elimination of most opposition parties, to the advantage of the

larger incumbents. At the 1998 elections, which the CCP won outright,

calculations suggest that up to ten additional parties would have gained

representation had the election been held under the 1993 electoral

laws.7 Despite this, a range of international observer groups gave

their stamp of approval to the 1998 elections, with the European

Union particularly keen to see Cambodia’s re-elected government

receive international blessing, regardless of the sharp decline in the

quality of its democracy.

Short-term versus long-term objectives

Despite such setbacks, there is also evidence of learning by the inter-

national community regarding the dilemma posed by the disjuncture

between the short-term and longer-term objectives of post-war demo-

cratization. We now have a much deeper understanding of the

7 My thanks to Michael Maley for the data on this point.

166 Benjamin Reilly

relationship between civil society, political institutions, and interna-

tional actors as agents of post-war peacebuilding than before. Scholars

and policymakers alike increasingly appreciate the importance of select

incentives in moving societies from war to peace, and the self-reinforcing

role of democratic procedures in helping them stay there. Yet one

crucial aspect of post-war democracy building continues to be over-

looked: the temporal dimension of democracy building. Given that

democracy is a long-term process of political development, a key

dilemma common to nearly all post-war elections is the trade-off

between the short-term goals of war termination and the longer-term

goals of democracy building.

Over the course of the 1990s, the United Nations and other interna-

tional bodies developed a kind of standard operating procedure for

post-war peacebuilding as part of the new global consensus on the

virtues of democracy (Newman and Rich 2004). Once a minimum

level of peace had been obtained (which did not necessarily mean a

full cease-fire agreement), and a basic level of infrastructure was in

place, the next step was usually to hold some kind of elections – often

within a year or two of the start of the mission – followed by a rapid

hand-over to the newly elected local authorities, and an even more

rapid departure of international troops and personnel. Under this

approach, elections came to be seen as the crowning event of the

post-war peacebuilding phase, enabling the reestablishment of legiti-

mate domestic authority, and allowing international forces to disen-

gage and, in most cases, depart.

Several dilemmas are inherent in this new model of state reconstruc-

tion. One is the question of election timing. How early a fragile state

should hold elections or referenda represents a fundamental choice

facing almost all post-war peacebuilding and democratization efforts.

Hasty or rushed ‘‘instant elections’’ have become common for several

reasons: the need to ‘‘do something’’ quickly, to start the process of

political development and, of course, to have an identifiable ‘‘exit

strategy’’ for international involvement. A classic case is the

November 1996 elections in Bosnia, a date dictated in large part not

by events on the ground but rather the desire of the Clinton adminis-

tration to show progress in the Balkans in time for mid-term elections

in the United States.

More diffuse temporal concerns also encourage premature elections.

Democracy, as the mantra goes, is a long-term process, but the domestic

Post-war elections 167

political pressures that weigh on the Western states that usually fund

and implement peacekeeping missions are almost all short term. Quick

results are required. An early establishment of home-grown institu-

tions, such as representative legislatures and multiethnic peace and

security forces, is needed to create both the shell of a state and to create

a legitimate body politic for the international community to deal with.

Financial pressures mean that many missions have an incentive to scale

back their immediate presence and reduce the levels of their assistance

as soon as it is minimally feasible to do so.

All of this places considerable pressures on the post-war election

timetable and frequently leads to elections being held as early as

possible in the life of a peacekeeping mission in order to create some

kind of legitimate government – a pressure which in Iraq led to elec-

tions being held in 2005 in the absence of popular security and in the

face of a boycott from one of the country’s main ethnic groups, the

Sunni. Such ‘‘premature elections’’ can also create multiple, ongoing

problems for the development of peacetime politics in deeply divided

societies even years after the war has ended – as demonstrated by the

regular re-election of ethnic hard-liners in post-war Bosnia, where

nationalist parties and elites not only have continued to be elected by

the voters, but have attempted to use the democratic political process to

press their sectarian aims. In general, the early application of elections

immediately following a conflict increases the likelihood that the con-

test will become a de facto contest between the former warring armies

masquerading as political parties. By contrast, an extended process of

consultations and local-level peacebuilding, in which some of the real

interests and concerns that provoked the conflict are addressed in a

step-by-step fashion before national elections are held, may offer better

prospects for a peaceful transition in post-war societies.8

A related dilemma is the virtue of referenda on deeply divisive issues,

such as independence or self-determination, compared to the virtues of

constructing a phased series of consultations rather than one take-it-or-

leave-it choice – the latter exemplified by the rolling series of indepen-

dence referendums in the early 1990s which precipitated the breakup of

Yugoslavia, or the 1999 autonomy plebiscite in East Timor. Because

they channel complex questions of political identity into an all-or-

nothing, zero-sum choice, the comparative evidence suggests that

8 For a survey of these, see Harris and Reilly 1998.

168 Benjamin Reilly

referenda are unsuited to solving deeply divisive issues and are parti-

cularly inappropriate for the combustible conditions which apply in

post-war societies (Reilly 2003a). These failures suggest the need to

give more attention to alternative models of post-war transition and

methods of self-determination, such as those employed in the power-

sharing and autonomy arrangements in the Bougainville conflict, in

which time periods were deliberately lengthened or left unspecified.

The success of the peacemaking process in Bougainville, an island at

the eastern extremity of Papua New Guinea that was the site of the

largest and most violent conflict in the South Pacific since World War II,

deserves more international attention than it has received to date. The

conflict centered around demands for Bougainville’s independence

made by rebel groups, a demand opposed not only by Papua New

Guinea but also many Bougainvilleans themselves. Before the war

began, Bougainville’s substantial contributions to the national eco-

nomy was disproportionate to its small size and population, mainly

due to an enormous open-cut copper, gold, and silver mine that operated

on the island from 1972 until the conflict caused its closure in 1989.

The cascading violence came to a head in 1997, when the Papua New

Guinea Government commissioned an international mercenary service,

Executive Outcomes, to attack the rebels. But in a surprise move, the

Papua New Guinean army’s chief commander announced the refusal of

his forces to work with the mercenaries, who were ejected from the

country, and the incumbent Prime Minister and several key ministers

involved in engaging the mercenaries stood down. Taking advantage of

these changes, rebel forces began to make direct contact with the central

government. Further developments resulted in the New Zealand

Government facilitating talks between the Bougainvillean leaders.

These talks resulted in a cease-fire agreement, followed by the deploy-

ment of an unarmed ‘‘Peace Monitoring Group,’’ led by Australia, on the

island, accompanied by a UN observer team. Successive agreements –

notably the so-called ‘‘Loloata Understanding’’ of March 2000 and the

Bougainville Peace Agreement signed at Arawa in August 2001 – paved

the way for the election in 2005 of an autonomous government and

agreement on a future referendum on Bougainville’s political status to be

held after an extended period of autonomy.

Importantly, both the election of the autonomous government and

the longer-term referendum plans involve an extended time period, in

sharp contrast to the rush to elections that has taken place in Angola

Post-war elections 169

and the Balkans. Bougainville’s first autonomous government elections

were not held until June 2005, four years after the signing of the

Bougainville Peace Agreement, while the timing of the independence

referendum is set at some unspecified time in the future, after some ten

to fifteen years of autonomous government. This drawn-out timetable

appears to have played a role in allowing many local-level peace

initiatives, which would otherwise have been subsumed to electoral

considerations, to flourish. It also allowed a series of innovative

reforms to the electoral process to be introduced: Bougainville’s new

autonomous parliament reserves seats for specific regions, former

combatants, and women (the last of whom have played a particularly

important peacemaking role). These various innovations have helped

to deliver one of the more successful post-war electoral processes of

recent years. In a further institutional innovation, future Bougainville

elections will be held under the alternative vote, a ‘‘vote-pooling’’

electoral system which some scholars advocate as a means of promot-

ing moderation and accommodation in divided societies by coercing

cooperation across ethnic lines (Horowitz 1985; Reilly 2001).

A final key temporal issue, which confronts many post-war electoral

processes, is the sequencing of local, regional, and national elections.

The coordination of election timing at the national and sub-national

level directly affects the development of local and national-level parties

and the extent to which national or regional consciousness takes hold

politically. Some scholars such as Juan Linz and Alfred Stepan argue

that new democracies should hold national elections first, before regio-

nal or local ones, in order to generate incentives for the formation of

national, rather than regional, political parties (Linz and Stepan 1996:

98–107). Others such as Larry Diamond believe that simultaneous

national and local elections ‘‘can facilitate the mutual dependence of

regional and national leaders. The more posts that are filled at the

regional and local level . . . the greater the incentive for regional politi-

cians to coordinate their election activities by developing an integrated

party system’’ (Diamond 1999b: 158). This was the approach taken in

post-Suharto Indonesia, with identical party-based ballots being pre-

sented to voters at simultaneous elections for national, provincial, and

local assemblies – a strategy which strengthened the position of nation-

ally focused parties.

In recent transitional elections in East Timor, as well as other post-

war cases such as Kosovo and Afghanistan, the decision was made to

170 Benjamin Reilly

start at the local level first: rather than leading with national elections,

the preferred sequence was to hold local or municipal elections as a

precursor to national ones, allowing steps toward democratization to

be taken gradually. The relative success of these cases suggests that

scholars such as Linz and Stepan are likely mistaken in advising that

national elections should be held before local ones, at least if the aim is

to inculcate voters in the routines of electoral politics and party politics

in new democracies.9 For transitional elections, a first-run test of

municipal or local elections before national ones appears to have

much to recommend it. However, if party building is the aim, then

simultaneous national, regional, and local elections are likely to be

more favorable to the development of strong nationwide parties.

There is evidence of genuine learning over time by the UN and other

international actors on most of these issues, with more recognition of

the need for sustained international involvement for several years after

a conflict rather than the rushed ‘‘in-and-out’’ approach of former

years. In major international assistance operations such as Kosovo,

East Timor, and Afghanistan, pressure to hold ‘‘instant’’ national elec-

tions has been resisted in favor of a two-year period of political develop-

ment as part of a much longer process of democratization. In both

Kosovo and East Timor, relatively peaceful national elections, which

had been preceded by successful municipal polls, were held in 2001. In

Afghanistan, presidential elections were held in 2004, almost three

years after the fall of the Taliban, while the parliamentary poll was

further postponed until September 2005. Even the 2002 Loya Jirga

process in Afghanistan – which brought Afghan tribal representatives

and elected delegates together to choose an interim government in a

process that was only partially democratic – can be seen as a kind of

local election. By contrast in Iraq, national elections to a constituent

assembly were pushed through within a year of the conflict being

declared over by the US president. In general, the comparative evidence

suggests that a bottom-up approach to electoral timing is probably the

best way to encourage the development of party politics and to incul-

cate voters in the routines of electoral politics.

9 Indeed, of the three cases cited by Linz and Stepan in support of their argument –Spain, Yugoslavia, and the USSR – only Spain held its first truly competitiveelections at the national rather than the regional level. My thanks to BethanyLacina for bringing this to my attention.

Post-war elections 171

Electoral mechanics: efficiency versus inclusion

A recurring dilemma animating the choice of political institutions in all

democracies, including post-war states, is the trade-off between stabi-

lity and governability versus inclusion and representation. Classically,

‘‘representational’’ institutions are thought to best ensure the direct

translation of popular preferences and cleavages into the political

sphere via political parties representing distinct social groups, propor-

tional elections to promote the representation of minorities, and low

thresholds or other barriers on the formation of new parties. Together,

these institutions should ideally lead to the development of a diverse

multi-party system in which all significant social groups and interests

are separately represented. By contrast, ‘‘efficient’’ institutions that can

deliver clear parliamentary majorities to disciplined political parties

offering distinct policy alternatives are more likely to be associated

with majoritarian electoral laws and the presence of ‘‘catch-all’’ parties,

which can command electoral support across social cleavages.

These scholarly debates have direct implications for institutional

choices – particularly the choice of electoral system, which has long

been recognized as one of the most important institutional choices

affecting the nature of democracy. They can have profound implica-

tions for the extent to which the voices of the poor and other marginal

groups can be heard and their power enhanced. For example, systems

in which the parliament is elected from many small geographically

defined electoral districts tend not to be as good at representing mino-

rity opinion than proportional ones, but may be better at building

links of local accountability. These choices can also influence other

aspects of the political system, such as the development of the party

system, linkages between citizens and their leaders, political account-

ability, representation, and responsiveness. Because of such impacts,

constitutional and electoral system choices have many long-term con-

sequences for the process of democratic governance, and the choice of

electoral system is one of the most important political decisions for any

country.

Electoral systems are often categorized according to how proportio-

nately they operate in terms of translating votes cast by electors into

seats won by parties. A typical three-way structure divides such systems

into plurality-majority, semi-proportional, and proportional represen-

tation (PR) systems. Plurality-majority systems typically give more

172 Benjamin Reilly

emphasis to local representation via the use of small, single-member

electoral districts than to proportionality. Amongst such systems are

plurality (first-past-the-post), runoff, block, and alternative vote sys-

tems. By contrast, proportional representation systems – which typi-

cally use larger, multi-member districts and deliver more proportional

outcomes – include ‘‘open’’ and ‘‘closed’’ versions of party list PR, as

well as ‘‘mixed-member’’ and ‘‘single transferable vote’’ systems. Semi-

proportional systems such as the single non-transferable vote offer yet

other approaches, as do various mixtures of plurality and proportional

models – such as the ‘‘mixed’’ models by which part of the parliament is

elected via PR and part from local districts, a common choice in many

new democracies over the past decade.10

Most of the major transitional elections conducted in recent years,

including almost all of those held under UN auspices, have utilized

some form of proportional representation. Prominent transitional elec-

toral operations in Namibia (1989), Nicaragua (1990), Cambodia

(1993), Mozambique (1994), Liberia (1997), Bosnia (1996, 1998,

2000, 2002), Kosovo (2001), Sierra Leone (2002), Rwanda (2003),

and Iraq (2005) were all conducted under PR rules. In particular, the

simplest form of proportional representation – party-list PR – appears

to have become the de facto norm for UN-administered elections. But

the adoption of such systems for post-war elections has usually been

dictated more by administrative concerns, such as the need to avoid

demarcating individual electoral districts and to produce separate bal-

lot papers for each district, than these wider political issues. Indeed, in

many post-war elections, national PR systems are the only feasible way

to hold an election, as a uniform national ballot can be used, no

electoral districts need be demarcated, and the process of voter regis-

tration, vote counting, and the calculation of results is consequently

simplified. In Liberia in 1997, for example, population displacement

and the lack of accurate census data led to the abandonment of the old

system of single-member majoritarian constituencies in favor of a

proportional system with a single national constituency. Indeed, the

inclusion of internally displaced peoples, refugees, and ex-combatants

in the electoral process is a recurring dilemma of post-war elections

generally (see Lacy 2004).

10 For a survey of these options, see Reynolds, Reilly and Ellis 2005.

Post-war elections 173

For Iraq’s crucial 2005 elections, experts favored a system based

around provincial boundaries, to ensure greater accountability and

representation of local constituencies. However, this would have

entailed a lengthy national census. In the interests of time, it was

therefore decided to fall back on a single, nationwide district elected

by proportional representation in which 1/275th of the vote was suffi-

cient to gain a seat in the constituent assembly. While this doubtlessly

facilitated the administration of the election itself, it also had the effect

of fragmenting the legislature and marginalizing numerically smaller

groups like the Sunni, while encouraging ethnic polarization amongst

the electorate. Many new democracies have therefore preferred

‘‘mixed’’ electoral systems, in which part of the legislature is elected

on a national level by proportional representation, and part at the local

level from single-member districts, so that both proportionality and

accountability are maximized. For example, at the August 2001 elec-

tions for East Timor’s 88-member constituent assembly, most seats

were elected on a nationwide basis by list PR, but there were also

separate single-member electorates corresponding to each of the coun-

try’s thirteen districts. A similar system in Iraq may have guaranteed

the Sunni minority a baseline of political representation at the provin-

cial level, thus helping to assuage the political alienation which is at the

root of Iraq’s insurgency (Diamond 2005: 269).

Perhaps the most unusual electoral system choice for a post-war

election in recent years has been the decision to use the single non-

transferable vote (SNTV) for the September 2005 parliamentary elec-

tions in Afghanistan. Under SNTV each elector has one vote, there are

several seats to be elected in the district, and the candidates with the

highest number of votes fill these positions. As a result, the number of

candidates a party nominates in each district becomes a critical choice:

too few, and parties miss out on valuable chances to win additional

seats; too many, and they risk splitting their vote too thinly and losing

winnable seats. Despite being structurally majoritarian, SNTV can thus

advantage smaller parties and deliver relatively proportional election

outcomes. However, by forcing candidates from the same party to

compete against each other for the same pool of voters, personalistic

attributes are emphasized over and above those of the party. The

resulting candidate-centered, intraparty competition has been widely

identified as a cause of factionalism, corruption, and clientelistic poli-

tics in states like Japan, where the abandonment of SNTV in 1994 was

174 Benjamin Reilly

fuelled by a series of corruption scandals linked to factional competi-

tion which damaged confidence in the political system. While having

the virtue of simplicity, these pathologies mean that it is also an

extremely risky choice for a new democracy, particularly if encoura-

ging cohesive national political parties is an objective. In an ethnically

complex, clan-based society such as Afghanistan’s, SNTV makes it

unlikely that a consolidated party system can develop in the short

term. Illustrating this, the 2005 Afghan parliamentary elections fea-

tured over 5,800 candidates – in Kabul alone the ballot paper displayed

over 400 names – resulting in a fractionalized and incoherent parlia-

ment which is likely to remain highly divided and unable to coordinate

around pressing policy challenges.11

Electoral administration: independent or party-based?

Another dilemma of post-war elections is the role and responsibilities

afforded to the bodies charged with running the elections themselves.

While constitutional and electoral reforms have attracted a voluminous

academic literature, issues of electoral administration remain under-

studied by scholars and under-rated in general in terms of their effect

on post-war polities. There are several models of election administra-

tion used around the world. Some countries locate responsibility for the

administration of elections within a government portfolio like the

interior or home affairs ministry. Others situate the responsibility for

administration of elections within government agencies such as the

public records office, the tax department, or even the postal service.

In some countries, the body responsible for running elections is created

anew before each electoral event. And in some cases, as in Cambodia in

1993 or East Timor in 2001, the United Nations itself takes responsi-

bility for running the elections.

Probably the most important administrative decision concerns the

composition of the body managing the elections, and specifically

whether the elections are run by the government of the day or by

some form of independent electoral commission. Their perceived neu-

trality and independence from political interference lends credibility to

the electoral process, which is a crucial determinant of the success of

any election. A truly independent commission is one that is able to

11 See ‘‘Democracy, sort of,’’ The Economist, September 24, 2005, 34.

Post-war elections 175

operate effectively without direct ministerial control, including in

terms of its financial and administrative functions, and is (ideally)

composed of non-partisan appointees. In practice, many independent

commissions around the world do not have complete financial inde-

pendence and may comprise party representatives rather than non-

partisan appointments. However, they are still able to operate free

from government interference or control.

By contrast, in some countries electoral management bodies are

composed not of independent civil servants, judges, or other officials,

but rather of the political parties contesting the elections themselves.

This practice can provide a form of non-partisan independence if the

composition of party representation is balanced in such a way as to

ensure genuinely neutral functioning. The influence of the United States

is particularly important here, as the American form of electoral

administration is based on political appointees and party representa-

tives, and many post-war democracies, particularly in Latin America,

have followed this model for their own elections. Some authorities

argue that, when there is no better tradition or an existing body of

widely respected independent civil servants, a party-based electoral

authority may be the only realistic choice (Lopez-Pintor 1998: 53).

But recent problems with this model in important transitional elections

such as Indonesia and Haiti, as well as in established democracies (most

notably, the 2000 presidential poll in the United States), emphasize its

propensity for politicization and deadlock.

Most established and emerging democracies have chosen non-parti-

san models of electoral administration. Indeed, since the world’s largest

democracy, India, adopted this model at independence there is a clear

trend toward the adoption of independent electoral commissions

staffed by non-partisan civil servants. The comparative evidence

strongly favors independent commissions run by apolitical civil ser-

vants, and the United Nations now explicitly advocates this model

wherever possible. Party-based commissions have an almost inevitable

tendency to split along party lines. In Haiti, for example, the

Provisional Electoral Council was made up of representatives of the

political parties but was also deeply divided along party lines, and

internal mistrust and divisions prevented it from working efficiently

(Nelson 1998: 76). In Cambodia, by contrast, a non-partisan electoral

commission was widely seen as one of the outstanding elements of the

entire United Nations transitional administration and elections of

176 Benjamin Reilly

1993. Non-partisan commissions were also a prominent and successful

part of United Nations missions in Namibia and in East Timor.

The dangers of using party-based electoral administrations in transi-

tional situations were graphically demonstrated by Indonesia’s transi-

tional elections in 1999. Amid the flowering of new political

movements that accompanied the democratic opening, a requirement

that all political parties be represented on the National Elections

Commission (KPU) resulted in a deadlocked and unwieldy body of no

fewer than fifty-three officials, most of them party representatives

(including some individuals who were also candidates for the election).

As a result, during the preparation for one of the most important

transitional elections of the 1990s, the body charged with running the

elections was almost completely dysfunctional, being deeply divided

along party lines and unable to take even basic decisions (at one stage,

fist-fights broke out between different members of the commission).

Following the elections, the Indonesians moved quickly to discard the

party-based KPU and replace it with a much smaller, non-partisan

body of eleven non-party and non-government representatives. Such

cases underline that independent and permanent electoral management

bodies are a clear best practice of electoral administration, a conclusion

which has also been reinforced by a global study of electoral manage-

ment bodies (Lopez-Pintor 2000).

Encouraging national party politics

A final dilemma confronting post-war elections is the nature of the

emerging party system and the extent to which party politics becomes

institutionalized. Scholars of democracy have long considered political

parties to play a crucial role not just in representing interests, aggregat-

ing preferences, and forming governments, but also in managing con-

flict and promoting stable politics. As the key agents of political

articulation, aggregation, and representation, political parties are the

institution which impact most directly on the extent to which social

cleavages are translated into national politics. Parties perform a num-

ber of essential functions in a democracy: ideally, they represent poli-

tical constituencies and interests, recruit and socialize new candidates

for office, craft policy alternatives, set policymaking agendas, form

governments, and integrate disparate groups and individuals into the

democratic process (Diamond 1997b: xxiii). These linking, mediating,

Post-war elections 177

and representational functions mean that political parties are one of the

primary channels for building accountable and responsive government

in new democracies.

This stylized depiction of the roles parties play in terms of demo-

cratic consolidation, however, can be undermined by the reality of

communalism, clientelism, and other forms of particularistic politics.

In post-war situations, party politics tends to reflect the social clea-

vages which created the conflict in the first place. If the conflict had a

strong ethnic dimension, for instance, then these differences will tend

to be reflected in the new democratic system, particularly if ‘‘ethnic

parties’’ are allowed to form freely. The presence of such parties, in

turn, can quickly incite intergroup competition while inhibiting coop-

eration. As Gunther and Diamond write, ‘‘The electoral logic of the

ethnic party is to harden and mobilize its ethnic base with exclusive,

often polarizing appeals to ethnic group opportunity and threat . . . the

ethnic party’s particularistic, exclusivist, and often polarizing political

appeals make its overall contribution to society divisive and even

disintegrative’’ (2001: 23–24). At the margins, the presence of such

parties can lead to what Sartori dubbed ‘‘polarized pluralism,’’ where

the ideological distance between the parties expands, to the detriment

of the political center. Indeed, in Western democracies, the presence of

parties with extremely divergent policies and preferences has histori-

cally been an important predictor of political instability.

For this reason, many scholars of ethnic conflict advocate the need

for broad multiethnic parties or coalitions of parties as a key mechan-

ism for ameliorating conflict in ethnically divided societies. In such

party systems, elections tend to be fought out between a small number

of relatively large and cohesive parties, and politicians ‘‘crowd the

center’’ in their quest for the median voter, avoiding sharp differentia-

tion with their competitors. As a result, ‘‘they tend to have a moderat-

ing influence on the way interests are aggregated’’ (Haggard 1997:

140). But forging centrist, programmatic, political parties in a post-

war society is easier said than done, as parties often spring from the

same cleavages and tensions which spurred the original fighting, lead-

ing to a continuation of the former conflict through the new democratic

process.12 Increasing awareness of the problems caused by such polar-

ized or otherwise dysfunctional party systems has lately spurred

12 See Soderberg Kovacs’ chapter in this volume.

178 Benjamin Reilly

multilateral bodies such as the United Nations – which have tradition-

ally been wary of direct involvement in politics, preferring more tradi-

tional kinds of development assistance – to take a more active role in

assisting political party development in some countries.13

The most ambitious actors in this field have been the international

democracy promotion organizations which have proliferated over the

past decade (Carothers 1999). Because they are not bound by the same

strictures as multilateral agencies, some of these have attempted to

intervene directly in the development of the party system in recipient

countries. In Bosnia, for example, the US National Democratic

Institute openly and actively promoted putatively multiethnic parties

such as the Unified List coalition in preference to nationalist parties

such as the Serb Democratic Party (SDS) or the Croatian Democratic

Union (HDZ) at the 1996 elections (see National Democratic Institute

1996). Also in Bosnia, a range of related reforms to the electoral system

and other areas introduced in recent years by the OSCE have attempted

to undercut nationalist parties by changing voting procedures and, in

some cases, barring candidates from election (Belloni 2004). Kosovo

also saw overt attempts by the international community to mandate

multiethnicity in the political system (Simonsen 2004). However,

despite some inflated claims to the contrary, the success of such inter-

ventions has been modest, and ethnic parties continue to dominate the

political landscape.

The vexed problem of transforming armies into parties after a pro-

tracted period of conflict continues to trouble international interven-

tions in this field. As one survey of post-war elections concluded,

‘‘Democratic party building is proving to be a slow process. In all the

[post-war] countries, political parties are organized around personal-

ities, narrow political interests, and tribal and ethnic loyalties’’ (Kumar

1998b: 218). Historically, the most successful example of such a tran-

sition is probably the armies-to-parties transformation wrought by the

United Nations in Mozambique, where a special-purpose trust fund

and some creative international leadership succeeded in bringing the

previous fighting forces of the Liberation Front of Mozambique

(FRELIMO) and particularly RENAMO into the political fold.

Financing political party development has been an important element

of a number of other post-war elections. One approach involves

13 For a survey of these approaches, see Reilly 2006.

Post-war elections 179

channeling technical or financial assistance from international donor

agencies, non-governmental organizations, or multilateral agencies to

party organizations in those states in which the international commu-

nity has taken a prominent role, such as countries emerging from a

period of violent conflict. Recent proposals for political party assis-

tance in Afghanistan have also focused on this kind of approach.

Despite the widespread agreement on the importance of strong parties

for a functioning democracy, there are dilemmas inherent in the encour-

agement of broad-based political parties too. Post-Suharto Indonesia,

for example, has seen an ambitious exercise in ‘‘political engineering’’ to

promote broad-based parties with a national focus and hamper separa-

tist groups through a complex collection of incentives and restraints on

party system development, which requires parties to establish local

branches in half of all provinces and municipalities. The bias in favor

of national parties was so strong that regional parties were even banned

from competing in elections to the regional assemblies, where again only

national-level parties were permitted. While encouraging broad-based

parties, these provisions also represent a restriction upon new entrants

into the political system and may end up encouraging extra-constitutional

action by aggrieved minorities who are unable to contest elections.

Exceptions have already been made. As part of the 2005 peace deal,

which appears to have ended the long-running civil war in Aceh, the

rebel Free Aceh Movement (GAM) was explicitly granted the right to

compete in elections. As a result, there is direct – and, at the time of

writing, unresolved – contradiction between the provisions of the peace

deal and Indonesia’s national party law (Reilly 2006).

Conclusion

The core problem facing post-war elections in the contemporary era is

the ideologically driven belief that all good things go together – a belief

which ignores a great deal of contrary evidence regarding the interac-

tion between post-war politics and democracy. While well-crafted

elections may indeed be important instruments of peacebuilding,

polls held in highly conflictual environments often have pernicious

consequences. They can act as a catalyst for the development of parties

based around cultural, linguistic, religious, or other kinds of ‘‘ethnic’’

cleavage. They can promote a focus on regional, rather than national,

issues. They are inevitably an unattractive option for those groups who

180 Benjamin Reilly

see themselves being consigned to a permanent minority status, and are

therefore likely to be violently opposed by the potential losers. They

can serve to place in positions of elected authority leaders committed to

exclusionary visions of the country – in many cases, the same charac-

ters who started or fought the conflict in the first place.

Despite these well-known dangers, the outcomes that post-war elec-

tions are expected to foster have become increasingly overloaded by

policymakers and politicians. Following the fall of the Berlin Wall and

the collapse of legitimate alternatives, democracy came to be seen by

the United Nations and the international community more generally as

an essential element of post-war reconstruction, with erstwhile con-

flicts to be transformed into peaceful electoral competition via ballots

rather than bullets. As a result, elections have become a standard part

of the prescription of contemporary peacebuilding. In addition, the

post-September 11 era saw the emergence in American foreign policy

of a grander rhetoric of democracy (and, by extension, elections) as

essential elements in building peaceful states and combating religious

fanaticism in the Middle East. Therefore, many transitional elections

are now saddled with unrealistic expectations to achieve goals that are

inconsistent and sometimes incompatible.

A more realistic and less ideological appraisal of elections is required –

one which recognizes that elections can be potentially advantageous or

injurious to post-war democratization – and that success is dependent on

a careful consideration of timing, sequencing, mechanics, and adminis-

tration issues. On the basis of experience to date, such an appraisal

would likely include a recognition that while elections cannot be post-

poned for more than a few years as part of a post-war peace deal, rushed

elections held in situations of insecurity will almost inevitably aid extre-

mist parties and candidates; that highly proportional PR systems may be

administratively convenient but have hidden and sometimes debilitating

political costs; that independent electoral commissions are demonstrably

preferable to party-based models for established and emerging democ-

racies alike; that a sequenced step from local to national elections is

optimal for most post-war societies, particularly those with little prior

experience of democracy; and that building coherent political parties

focused on the delivery of public goods is a pre-eminent challenge.

Policymakers need to pay more attention to these issues, and to the

link between institutional choices and the broader goals of building

stable and democratic post-war polities.

Post-war elections 181

7 Civil society in war-to-democracytransitions

R O B E R T O B E L L O N I

In the 1980s Serbia, like all other Eastern European countries, wit-

nessed a strong revival of civil society. Several non-governmental orga-

nizations (NGOs) and committees for the defense of different rights

from freedom of expression to ecological protection were created –

primarily among intellectual circles in Belgrade. In the second half of

the decade many members of these organizations and groups began to

move from demands for human rights and democratization to extreme

nationalism which denied basic rights to members of other non-Serb

national groups. Virtually all intellectuals and grass-root organizations

supported Slobodan Milosevic’s rise to power – identifying him as the

champion of the Serb national cause. The outbreak of war in the 1990s,

combined with Milosevic’s controversial role in the protection of Serbs

throughout the territory of the former Yugoslavia split the emerging

civil society. Non-nationalist opposition to the regime’s wars in

Croatia, Bosnia, and Kosovo was dubbed the ‘‘other Serbia.’’ At the

same time, however, extreme nationalist groups also criticized

Milosevic, blaming him for having abandoned Serbs in Croatia and

for having failed to support Serbs adequately in Kosovo – an area

inhabited by an overwhelming majority of Albanians. The peaceful

revolution of October 2000, ending thirteen years of Milosevic’s rule,

occurred because of a broad alliance among actors within civil society,

including both nationalist and non-nationalist groups. The ouster of

Milosevic, however, did not terminate the influence of uncivil, violent,

and illegal groups – one of the legacies of the Milosevic era (Bieber

2003). In March 2003 a member of the criminal Zemun Clan assassi-

nated pro-Western reformist Prime Minister Zoran Djindjic, compli-

cating the reform process which had begun in October 2000.

This chapter examines the role of domestic civil society, that is, the

set of voluntary organizations and groups not created by the state, in

the transition from conflict to peace and democratic consolidation. It is

structured in two main sections. First, it provides a brief analysis of the

182

state of civil society in war-torn countries. As in Serbia, different

groups and organizations coexist within the civil society realm. First,

even the most deeply divided societies maintain multiethnic and civic

constituencies. There exist associations and groups which recognize the

importance of respecting human rights and promoting compromise,

dialogue, and economic and social integration. These groups strive to

promote civic politics instead of ethnic politics, and social and political

spaces of dialogue instead of ethnic or national segregation. Not only

do they provide a space to articulate citizens’ participation in public

life, but also they can constrain the arbitrary exercise of state power.

Supporters of the importance of civil society in war-to-democracy

transitions implicitly or explicitly endorse these kinds of associations

and groups. Second, people organize not only around democratic and

liberal values, but also around values that can be seen as ‘‘uncivil.’’

Many civil society groups and organizations justify and engage in

violent and/or illegal actions. Mafia-like groups and paramilitaries

often thrive in the context of a national, ethnic, or religious divide

and in the absence of a functioning state guaranteeing the framework

for peaceful, non-violent coexistence. Third, between civil and uncivil

groups there exists a wide spectrum of organizations divided along

ethnic, religious, and national lines. Civil society in conflict areas is as

polarized as political society. Many groups and associations organize

around sectarian and particularistic identities. Although most of these

groups and associations may openly reject violence, they often endorse

and promote a worldview that considers groups’ relations in zero-sum

terms and resist compromise and cooperation. Their very existence

perpetuates the divisions within society and can contribute to political

polarization and continuing confrontation between groups.

The second part of this chapter explains how the confidence and

expectations placed upon civic, cross-cutting, and pro-democracy civil

society are often frustrated in practice. Of the four dilemmas outlined in

the introduction to this volume, three in particular affect civil society –

complicating its contribution to both democratization and peacebuild-

ing. A vertical dilemma between legitimacy and efficacy arises in the

process of the negotiation and implementation of a peace agreement.

The inclusion of civil society groups in this process improves the

legitimacy of an agreement but may come at the cost of efficiency

by increasing the number of actors involved and thus making the

decision-making process more burdensome. While this dilemma is

Civil society in war-to-democracy transitions 183

particularly acute in the peacemaking phase, the next two dilemmas

affect the post-settlement peacebuilding/democratization transition.

A systemic dilemma may take place when international interven-

tion attempts to support local, bottom-up peace constituencies.

International assistance may be indispensable to guarantee the survival

and further development of local civil society groups; but such assis-

tance can result in skewed local priorities by inducing local groups to

prioritize those activities likely to attract international funding. Finally,

a temporal dilemma appears when short-term and long-term interven-

tion strategies conflict with each other. International organizations

have a tremendous institutional pressure to demonstrate short-term

results. But ‘‘short-termism’’ limits the effectiveness of international

intervention, in particular by preventing the adoption of long-term

structural projects.

In conclusion, this chapter asks under what conditions civil society

can have a positive influence on both democratization and peacebuild-

ing, and how to reframe and restructure international engagement to

better sustain post-settlement transitions. The three dilemmas identi-

fied are genuine and difficult to resolve. They require a far-reaching

adjustment in bilateral and multilateral donors’ practices. Central to

this adjustment is the need to mobilize effectively existing domestic

resources as equal partners to international engagement, instead of

continuously identifying domestic gaps to be filled with the proper

dose of international assistance. So long as international engagement

is framed around the notion that individuals, groups, and local associa-

tions are objects of international engagement, rather than being active

agents with resources and assets, civil society’s contribution to both

democratization and peace will be limited.

Civil society between ideals and reality

The current popularity of civil society in democratization and peace-

building projects is due in large measure to the legacy of the democratic

struggle in Eastern Europe and Latin America in the 1970s and 1980s.

Individuals and groups fighting dictatorships in these regions viewed

the struggle as one between civil society against the state, with the two

confronting each other in a zero-sum relationship (see, for example,

Arato 1981). When the Berlin Wall fell, civil society became endowed

almost with a heroic quality. Since then, much of the literature on the

184 Roberto Belloni

topic either has stressed the qualities of the ‘‘good’’ society to juxtapose

against the ‘‘bad’’ state or the need of a strong civil society to make

democratic institutions work more efficiently.

The civil society ideal and its limits

Perhaps more authoritatively than any other scholar, Robert Putnam

(1993; 2000) has argued that a healthy civil society is key to making

democracy work. In his view, civil society is the microcosm for the

development of democratic norms and practices. Social interaction in

voluntary NGOs is a school of democracy. When members agree to

play by the rules and commit to the goals of a civil society organization

or community group, protracted, face-to-face interaction will develop

‘‘bonding social capital,’’ reinforcing the members’ mutual trust and

cooperative behavior. Members are socialized into democratic norms

through a process of learning by doing.

Not only can civil society nurture trust and reciprocity, but also it

fosters tolerance for diversity – a crucial aspect for societies torn by civil

conflict. According to Putnam, the interaction with people from differ-

ent ethnic, racial, religious, and political affiliations and social status

promotes ‘‘bridging social capital.’’ By participating in civil society orga-

nizations, individuals learn how to confront divergent opinions and tend

to develop greater tolerance for different interests and views. Crucially, it

does not matter what type of group or organization individuals belong

to. Social capital and toleration can develop regardless of the goals and

aims of these groups and organizations. Accordingly, Putnam argues

that choirs, hunting organizations, bowling leagues, and more broadly

sport clubs are all useful tools for developing trust among its members

and tolerance for diversity.

Others have confirmed the existence of a strong positive relationship

between social capital and democratic performance – expanding the

analysis beyond civil society as a ‘‘school of democracy.’’ Larry

Diamond (1999a: 230–234), for example, argues that civil society

performs many important functions, including monitoring and

restraining the exercise of power by the state, stimulating political

participation, developing a democratic culture of tolerance and com-

promise, creating additional channels for the articulation and repre-

sentation of interests, generating cross-cutting cleavages, recruiting

and training new political leaders, improving the functioning of

Civil society in war-to-democracy transitions 185

democratic institutions, disseminating information to citizens, and

producing favorable conditions for economic reforms. Francis

Fukuyama (1995) has taken Diamond’s last point further, pointing

out how the existence of social capital (‘‘trust,’’ in Fukuyama’s jargon)

underpins and facilitates economic activities in the market, making

trustful societies not only more democratic but also more prosperous.

In sum, civil society participation has a beneficial influence on indivi-

duals, who become better citizens of the democratic polity even when

they participate in recreational, non-political groups. As a whole, a

vibrant civil society is instrumental for a more democratic and prosper-

ous society.

Building on research conducted primarily in Western consolidated

democracies, practitioners and students of democratization and peace-

building have argued in favor of the positive role civil society can play

in conflict areas (Fitzduff 2004; Gidron, Katz, and Hasenfel 2002;

Paffenholz and Spurk 2006; van Tongeren, Brenk, Hellema, and

Verhoeven 2005; World Bank 2003b). According to its proponents,

civil society organizations are the answer to the problems confronting

societies transitioning from war to peace and democracy. Not only can

civic organizations increase trust within and between different com-

munities, but also they are key to providing public goods when the state

is too weak, divided, or indifferent to do so – which is often the defining

condition of regions torn by civil strife. According to Daniel Posner

(2004), the collapse of the state can be ‘‘liberating’’ for civil society – a

view squarely within the ‘‘good’’ society / ‘‘bad’’ state tradition. Groups

and organizations can patrol neighborhoods in the absence of a func-

tioning police; they can organize a rudimentary judicial system when

state courts are unable to administer justice; and they provide educa-

tion to young people when schools are not working. Moreover, even

where the state still maintains some capacity to deliver services to its

citizens, civil society can still complement the work of domestic institu-

tions by helping to improve economic and political performance, con-

trol crime and corruption, provide opportunities to former combatants

to demobilize after war, and support the process of post-war return

home of refugees and displaced persons.

This ideal vision must be weighed against the reality of a fragmented,

factionalized, and occasionally xenophobic version of civil society.

Even in consolidated democracies, civil society is a vague and general

concept, which can be filled with different contents – ranging from

186 Roberto Belloni

democratic to undemocratic actors and from peaceful to violent agents.

Civic action is often burdened by incivility and violence – leading at

least some observers to argue that civil and uncivil aspects coexist

within the civil society realm. John Keane, for example, has underlined

how the persistence or the possibility of violence within and between

countries survives alongside the spreading of civilizing politics aimed at

reducing the incidence of genocide, ethnic cleansing, murder, and rape.

As Keane (1998: 135) argues, ‘‘all known forms of civil society are

plagued by endogenous sources of incivility, so much so that one can

propose the empirical-analytic thesis that incivility is a chronic feature

of civil societies . . . a perennial barrier to the actualization of a fully

‘civilized’ civil society’’ (emphasis in the original). Not only can domes-

tic organizations promote cooperation and trust among its members

and society at large, but also they can foment discord and violence. To

refer to Putnam’s work, it does matter a great deal what kind of tunes

choirs sing.1

Deeply divided societies: community polarizationand state weakness

The ease with which civilized coexistence breaks down in the escalation

to war corroborates Keane’s thesis about the Janus-like nature of civil

society. Anti-civic and violent forms of associationism prevail over

those pro-democracy organizations celebrated by advocates of civil

society. In Lebanon during the 1975–1989 civil war and in Rwanda

prior to the 1994 genocide, local voluntary organizations fomented

intergroup violence. In these and similar cases of breakdown of civil

coexistence, while political authorities, intellectuals, and religious

leaders provided the rationale for violence, civil society fragmented

into opposite camps, while the media quickly turned into an instrument

for nationalist propaganda. In Croatia, Serbia, and Bosnia-

Herzegovina local media actively endorsed the war agendas of their

main nationalist parties (Thompson 1994). In Sri Lanka mainstream

media frequently promote Sinhalese nationalist perspectives while

1 Putnam has been much criticized for not considering enough the ‘‘dark side’’ ofcivil society in his 1993 book Making Democracy Work. Margaret Levi (1996:52), for example, points out that the Oklahoma City bombers were members of abowling league. In his later book, Putnam (2000: 350–363) addresses the issuedirectly.

Civil society in war-to-democracy transitions 187

ignoring Tamil views – thus hardly contributing to a dialogue between

the warring parties (Orjuela 2005).

The war itself further polarizes civil society and more broadly the

views of all those involved, leaving a bitter legacy of resentment and

mistrust that complicates the process of post-war democratization and

peacebuilding. The presence of divided communities suspicious of each

other’s intentions hinders human rights protection and more broadly

the development of the rule of law. For example, in the case of inter-

ethnic crimes, a member of one group is unlikely to report another

member to the police. Not only does this attitude hamper police work,

which relies heavily on the community, but also it complicates cross-

ethnic human rights advocacy. The predominance of national and

group identities can lead some to condone human rights violations,

and even to accept corrupt and patrimonial relationships as the alter-

native to citizenship rights.

Moreover, even after the signing of a peace settlement the warring

parties typically continue to contest the state, thus weakening its legiti-

macy and capacity to provide the framework within which meaningful

dialogue among human rights and democracy groups can be carried

out. Weak political institutions without the support of their citizens

lack the ability to make and enforce binding decisions. When the state

is weak and/or contested, the influence of uncivil and even xenophobic

forms of civil society becomes particularly strong. In a paradoxical

reversal of the ‘‘anti-political’’ mood popular in the 1970s and 1980s in

Eastern Europe and Latin America, and celebrating civil society against

the state, the lack of functioning and effective state institutions in

conflict areas provides an opportunity for uncivil groups to thrive.

A public sphere with weak or no protection opens the way for the rule

of the stronger to replace the rule of law and for patrimonial and

patriarchical relations to replace the benefits of citizenship. In a few,

limited cases, civil society building can be possible even in the absence of

a state (see Paffenholz 2001: 3, on Somalia). In general, however, under

such conditions civil society is more likely to succumb to sectarian and

often violent and illegal interests.

In a post-war context, at least three types of civil society groups

coexist in the public sphere. While the impact of the first and second

types on democratization and peacebuilding is quite uncontroversial,

the role played by the third one is more complex and will be the focus of

this section.

188 Roberto Belloni

First, there are groups defending and promoting a politics of inclu-

sion and civic principles, and stressing the importance of developing

multiple civic identities, pluralism, equal opportunity, tolerance, and

the government’s accountability. They attempt to defend and enlarge

spaces for negotiation and compromise – islands of civility and dialo-

gue in a context polarized by mutual fear and mistrust. They strive for

inclusion, participation and equal access and place emphasis on the

need to negotiate and cooperate in view of building consensus rather

than repeating adversarial, zero-sum positions. It is not difficult to

recognize in this version the positive kind of civil society commended

by many scholars and practitioners and authoritatively put forward by

Putnam and others in their study of the relationship between civil

society, democracy, and economic prosperity.

There is increasing evidence that multiethnic and diverse civil society

organizations bridging national, ethnic, and social divisions have a key

role in preserving and consolidating peace. In his research on

Hindu–Muslim relations in India, Ashutosh Varshney (2002) found

that ethnically integrated organizations, including business organiza-

tions, trade unions, professional groups, political parties, and sport

clubs, stand out as the most effective way of controlling violence

between ethnic groups. Intercommunal engagement in formal organi-

zations leads, in Varshney’s words (2002: 46), to an ‘‘institutionalized

peace system’’ moderating tensions and pre-empting violence. Earlier

research in Northern Ireland by John Darby (1986) anticipated

Varshney’s observation. In the mid-1980s Darby examined working-

class areas of Belfast finding that of four comparable neighborhoods,

the one with the most developed network of community associations

had the lowest level of violence. Similarly, Timothy Sisk and Christoph

Stefes (2005) found that cross-cutting civil society groups were essen-

tial to sustain cooperation and moderation among members of differ-

ent ethnic groups in the transition from apartheid to majoritarian

democracy in South Africa. Everywhere in deeply divided societies,

groups struggling to promote pluralism, tolerance, and inclusion are

widely recognized as instrumental to further democratization and

peace. Women groups in particular are extremely engaged in building

bridges across communities and demanding increased participation

and government’s accountability (Shoemaker 2005).

However, the legal and political vacuum left by a weak or failing

state often makes society at large subject to the second type of civil

Civil society in war-to-democracy transitions 189

society – uncivil organizations such as mafia-like groups and parami-

litary gangs engaged in illegal and violent activities. In most conflict

regions the main threat to stability and obstacle to peacebuilding and

democratization is not the conflict per se, but criminality. Mafia-like

groups and paramilitaries represent instances of uncivil society sustain-

ing inter-ethnic divisions or enforcing intra-ethnic private and illegal

rule. Mafia-like organizations can be very successful in establishing

networks of economic and political control. They often maintain a

close relationship with the political establishment, which consolidated

itself out of the disintegration of the previous political order and the

outbreak of war. These organizations’ continuing existence and suc-

cessful economic performance is predicated on the presence of ethnic

separation and division (Jung 2003). Similarly, paramilitaries have a

vested interest in fomenting and perpetuating ethnic and national

segregation and exclusion. In Northern Ireland, for example, parami-

litary groups enforce a tight control over their respective communities,

control illegal trade, and resist the establishment of any effective state

institution that would curtail their activities, particularly a multiethnic

police force. Groups such as the Ulster Defence Association and the

Provisional Irish Republican Army, to name just two, are uncivil orga-

nizations engaged in widespread illegal activities, such as the control of

arms and drugs trafficking, whose negative consequences are borne by

the members of the respective communities they claim to represent. The

presence of the paramilitaries is a continuing source of discord, inse-

curity, and fear, and an obstacle to the consolidation of democratic

politics.

More complicated is the case with the third type of civil society,

those organizations operating within the boundaries of the law, usually

not engaged in violent or illegal behavior, but divisive for society as a

whole. Civil society institutions in conflict areas, including religious

and community groups, veteran associations and NGOs are often as

divided as the society in which they are embedded. Membership in

these organizations is usually determined by ascriptive criteria, such as

race or religion, or by having fulfilled patriotic duties, such as military

service. Similarity and cultural homogeneity provide an initial basis for

trust, which in turn is the foundation of successful cooperation. Ethnic

and religious groups can perform a positive task in furthering democ-

racy and democratization. Putnam (2000: 65–79) considers these

groups as a major source of social capital in the United States, and an

190 Roberto Belloni

important reason explaining this country’s successful democratic tra-

dition. In Eastern Europe and Latin America, where the civil society

revival has begun, many churches and religious organizations fought

for democracy and freedom. Even in divided societies and war-torn

regions, ethnic, national, and community groups fulfill important

tasks. In Central Asia, traditional networks comprising of village elders

and councils and organized around clan membership have been indis-

pensable in preventing or mitigating conflict – with considerable stabi-

lizing effects in a potentially volatile region (Collins 2006). In Tajikistan,

for example, local councils were effectively mobilized by international

agencies to smooth the relationship between refugees returning to their

villages and other local groups (Mullojanov 2001).

More examples can be cited from other war-torn areas. In

Cambodia, Buddhist village organizations helped reintegrate demobi-

lized fighters and build social capital. In Guatemala, Mozambique, and

Uganda, veterans’ associations played a similar function (Colletta,

Kostner, and Wiederhofer 2004). Likewise, in Rwanda the gacaca, an

adapted citizen tribunal system at the community level, has contributed

to post-genocide reconciliation by relieving the local justice system of

the impossible task of trying tens of thousands of perpetrators (Uvin

and Mironko 2003). Furthermore, not only do community organiza-

tions provide an avenue for social reintegration after the war, but also

they may provide their members with a sense of ontological security,

thus creating the conditions for increasing bridging social capital across

the ethnic/national/religious divide. In some cases, civil society organi-

zations can even evolve and move from uncivil to civil politics. In

Serbia, the Serbian Resistance Movement started out as an ultra-

nationalist group pressuring the government into more anti-Albanian

policies in Kosovo, but it slowly evolved into an organization accepting

key values of civil society such as the formal acceptance of ethnic

diversity (Bieber 2003).

At the same time, however, ethnic, national, and religious organiza-

tions and groups sometimes uphold exclusivist values and norms that

are incompatible with democratic practices. Civil society built on

ascriptive criteria is frequently based on exclusion of and conflict

with others, particularly in regions with a history of war. A worldview

that regards identity as closed, fixed, and inherently conflictual under-

pins a politics of belonging. Separation and exclusion are put forward

in the name of national purity and as a precondition for neighborly

Civil society in war-to-democracy transitions 191

coexistence. When civil society organizations are not civic, multiethnic

and multi-religious, their contribution to democracy and peace might

be spurious. Intra-group cooperation based on a sense of belonging and

kinship may be directed toward anti-social ends. Bonding social capital

can become the vehicle of ethnic and parochial interests undermining

social cohesion, fragmenting society, and pitting one group against

another (Colletta and Cullen 2000). Many countries in Africa, for

example, have long been successful in bonding associational life struc-

tured along ethnic, national, religious, and clan lines, but have often

been poor in both democracy and peace (Kasfir 2004). This is true in

other contexts as well. For example, war-torn Sri Lanka exhibits a high

level of popular mobilization developed along racial and sectarian lines –

but many civil society groups are not particularly supportive of peace-

making. Indeed, Sinhalese groups linked to the Buddhist clergy have

coalesced into the most vociferous opposition to the peace process

(Orjuela 2005).

The potentially negative impact of ascriptive organizations is visible

in other regions. The Orange Order in Northern Ireland, to mention the

most gripping example, is an entirely legal association which proclaims

itself ‘‘primarily a religious organization,’’ which is ‘‘Christ centered,

Bible based, and Church grounded.’’ It stands officially for civil and

religious liberty but remains firmly sectarian. Its members must sign a

declaration before joining that both their parents are Protestants and

leave if they marry a Catholic. The Order’s very existence and activities

are a continuing source of tension and occasionally of violent conflict.

Among these activities, the Order each year celebrates the military

triumph of Protestant King William III of Orange over the Catholics.

Since the signing of the Good Friday Agreement in 1998, the organiza-

tion’s parades through Catholic neighborhoods have contributed to

preserving the tension between parts of the Protestant and Catholic

community. On the Catholic side it is difficult to find a similarly

contentious organization. But the broader reality of a society divided

along exclusivist and sectarian lines remains. Indeed, Catholic civil

society often reflects the same political preferences of political society.

The most popular Catholic organization – the Gaelic Athletic

Association – is firmly nationalist (Belloni 2007).

Thus, although ascriptive groups can contribute positively to peace-

making and peacebuilding, such groups may also reinforce national

and social differences and perpetuate and even strengthen the divisions

192 Roberto Belloni

and cleavages existing in society – instead of fostering and sustaining a

public sphere hospitable to democratic and civic life. When recruitment

is exclusive and based on ascriptive criteria, civil organizations tend to

exacerbate existing divisions in political life and perpetuate the exis-

tence of potentially hostile communities (Hadenius and Uggla 1996:

1623). Strong in-group loyalty, while strengthening bonding social

capital, may also undermine cross-cutting social capital and keep

alive out-group antagonism. Organizations with little or no national,

religious, or even class diversity tend to reinforce the views of its

members and strengthen people’s sense of difference from other com-

munities, dangerously perpetuating stereotypes about outsiders.

This brief overview reveals how the considerable variety of ascriptive

organizations, and the different contexts where they operate, compli-

cates generalizations about their role in democratization and peace-

building. Generally speaking, ascriptive organizations are more likely

to be mobilized for democracy and peace when they arise and develop

to address local issues directly pertaining to the community they are

attempting to serve. Often these organizations and groups are small

and locally rooted. They pragmatically attempt to mitigate the con-

sequences of conflict and mediate between opposite groups. By con-

trast, bigger, nationwide groups can provide an important electoral

base for sectarian political parties. For these reasons, organizations of

demobilized soldiers, churches, and even sport leagues can and often

are mobilized and instrumentalized on the basis of national, ethnic, or

religious criteria – reinforcing societal cleavages, hindering the demo-

cratization process, and sometimes even engaging in undermining the

peace process (see, for example, Bojicic-Dzelilovic 2006).2

Overall, civil society reflects and incorporates the divisions within

society. Because of the legacy of the conflict, post-war societies are

particularly likely to be dominated by sectarian groups. Moreover, the

lack of an effective and legitimate state often leaves civil society in

conflict areas prey to violent and illegal interests. In this context,

associations striving to further civic politics and political, economic,

and social integration may be either a minority or less influential than

uncivil groups – or both. Their positive influence on war-to-democracy

transition may be further hindered by the choices of international and

2 Many thanks to Roland Paris who pushed me to think harder about this issue.

Civil society in war-to-democracy transitions 193

domestic actors. Civil society groups present peacemakers with diffi-

cult dilemmas and trade-offs, to which I now turn.

The vertical dilemma: inclusion or exclusion?

The vertical dilemma entails a choice between legitimacy and efficacy.

The involvement of civil society in the pre-negotiation and negotiation

of a peace agreement is one of the first choices faced by negotiators.

Should civil society groups participate in the peacemaking process?

Which segments and interests of society should be represented? More

specifically, which individuals and groups should participate and why?

What criteria should be applied in the selection of civil society repre-

sentatives? Should sectarian civil society be kept out? If so, on what

grounds? In general, what advantages would representatives of civil

society bring to the table and which problems could it create? The

involvement of civil society’s representatives has both positive and

negative consequences, making the decision of whether such represen-

tatives should be included a difficult one.

There is no unique approach to the issue of whether and to what

extent civil society should be involved in peacemaking. Rarely do civil

society members directly participate in the process with a seat at the

negotiating table. Often international mediators exclude non-combatants

from peace negotiations. The reasons for this exclusion are multiple,

including the need to focus primarily on reaching an agreement among

those actors who control the means of violence (and thus can stop the

war) and to conduct negotiations efficiently and successfully. The

successful conclusion of peace negotiations may require a high degree

of confidentiality, precluding or complicating the opening of the pro-

cess to actors other than the main warring parties. This need for

confidentiality can lead to the negotiation of a peace agreement behind

closed doors, frequently in a foreign country, increasing the possibility

that its terms will respond to the demands of the participants and their

immediate constituencies, including the military and the political and

economic elites. Furthermore, international mediators might fear that

the involvement of a greater number of individuals could delay the

process of the negotiations or even irritate domestic political elites –

who often dismiss local civic groups as either instruments of improper

international interference, or political dissidents on the payroll of the

opposition. Finally, inclusion may also require a thorough assessment

194 Roberto Belloni

of local reality and actors – a difficult task in the context of an ongoing

war. Thus, civic groups are often excluded from negotiations.

International mediators prefer to pay lip service to civic groups, instead

of meaningfully involving them in the process.

The process leading to the negotiation of the Dayton Peace

Agreement (DPA) ending the war in Bosnia in late 1995 exemplifies

this approach perhaps better than any other peacemaking process.

Chief American peace negotiator Richard Holbrooke focused his shut-

tle diplomacy on reaching a compromise among those political leaders

in Bosnia, Serbia, and Croatia most responsible for the war – not ever

meeting with any actor who did not directly control the means of

violence. During the talks that preceded the peace conference in

Dayton, Holbrooke (1998: 135) endorsed the principle of secrecy to

avoid ‘‘public pressure’’ from Bosnian citizens and groups. He later

adopted the same approach at Dayton, where lack of communication

with the press was the ‘‘most important rule’’ of the peace conference

(Holbrooke 1998: 236). This ground rule was expressly aimed at

insulating the main political leaders from the demands of civil society

constituencies back home. The fact that no civil society member was

invited to Dayton to participate in the negotiation ensured the success

of the strategy.3

Although effective in reaching an agreement ending a bloodshed,

which cost more than 100,000 lives, the DPA also entrenched the

power of the same ethnonational elites responsible for the war, and

made it very difficult for Bosnian civil society to play a positive role in

the post-settlement transition (Belloni 2001). Furthermore, only one

woman participated in the conference – a representative of the British

delegation. Unsurprisingly, the Dayton Peace Agreement were gender-

blind – creating serious obstacles for the success of the post-settlement

democratization and peacebuilding process. In particular, because the

vast majority of those individuals displaced by the war were women,

often widows who might have been physically and mentally abused

during the war, failing to take into account explicitly and directly their

3 In his detailed account of the peacemaking process in southeastern Europe,Holbrooke cites domestic civil society only once. In December 1996, hundreds ofthousands of Serb citizens crowded the streets of Belgrade demanding greaterdemocracy in Serbia. As Holbrooke (1998: 345) points out, on that occasionAmerican diplomacy ignored the demonstrators and their leaders, missing animportant ‘‘chance to affect events.’’

Civil society in war-to-democracy transitions 195

specific needs has posed a serious hurdle to the implementation of the

human rights aspects of the peace agreement (Lithander 2000).

The absence of civil society in the negotiation of the peace settlement

resulted in a self-fulfilling prophecy, creating the conditions for future

constraints in the peace implementation period. Belief in local elites’

cooperation in the implementation of the DPA proved illusory. In

particular, Milosevic’s role quickly evolved from that of peacemaker

and guarantor of the peace settlement to the main obstacle to its

implementation.4 His political trajectory highlights the difficult

choices involved in the vertical dilemma. The reliance on pacts between

elites and the marginalization of civic groups might be indispensable to

reach a peace settlement but at the same time, by preventing the active

involvement of domestic constituencies, it can limit the legitimacy of

the agreement, and complicate its later implementation. As we shall see

with reference to the Democratic Republic of the Congo and the Ivory

Coast, civil society’s inclusion in peacemaking might not prevent diffi-

culties during the peace implementation, but it ensures broader domes-

tic support.

The Bosnian peace negotiations, conducted under the pressure of an

ongoing bloody war, excluded civil society actors in the name of

prioritizing the signing of an agreement among the main war makers.

But the exclusion of civil society actors from the negotiating table also

occurs in less dramatic situations. In South Africa, where the transition

from apartheid occurred without the much-feared bloodshed, the prin-

cipal forum for the negotiation of the ‘‘historical compromise’’ – the

Convention for a Democratic South Africa (CODES) – excluded actors

from civil society. Similarly, in Northern Ireland no civil society group

as such was part of the peace negotiation process, which was initially

held in secret and eventually led to the signing of the 1998 Belfast

Agreement. In arguing against the inclusion of civil society, the

British government raised the question of how representative and

accountable civil society groups could be considered. As a result, in

order to obtain a seat at the negotiating table, civil society members

were asked to achieve a clearer legitimacy by participating in the

4 Under this respect, the evolution of Yasser Arafat’s role in the Middle Eastpeace process resembles that of Milosevic, from guarantor of the 1993 OsloAgreement, negotiated in secret in Norway, to the real or perceived mainobstacle to its later implementation.

196 Roberto Belloni

formal political process. The Women’s Coalition took up the challenge,

transformed itself into a political party, contested the elections, and

won one seat. One of its proposals put forward during the negotiation

process led to the establishment of the Civic Forum comprising repre-

sentatives of business, trade unions, and the voluntary sector (Guelke

2003: 68).5

Of course, even when civil society groups are not formally allowed to

join the negotiating process, as in South Africa and Northern Ireland,

they can still provide both inputs through informal channels and legiti-

macy and support to the peacemaking effort. In both instances, activ-

ities of civic groups have contributed to increasing the number of those

in favor of the peace process and thus they have been instrumental in

creating the background conditions necessary to reach a peace agree-

ment. The proximity between the general population and civil society

actors made civil society a key factor in the effort to build a peace

constituency. In South Africa, community groups and progressive anti-

apartheid organizations were instrumental in diffusing the tension in

the difficult and violent period of transition from apartheid to power

sharing and, later, to majoritarian democracy (Collin Marks 2000; Sisk

and Stefes 2005). In Northern Ireland prior to the Belfast Agreement

two campaigns played an important role at key times. In 1992–1993

the Opsahl Commission involved the grass roots and changed the terms

of the political discourse by coming forward with one of the key

principles of later peace negotiations: ‘‘parity of esteem.’’ In the spring

of 1998 the nonpartisan ‘‘Yes’’ campaign helped achieve a victory in the

referendum that sealed the deal between Protestants and Catholics

(Guelke 2003). As these examples suggest, local involvement of civic

and pro-democracy groups can provide greater popular backing for a

peace agreement, and facilitate its later implementation.

At the same time, however, civil society’s support for peacemaking

cannot be taken for granted. Both in South Africa and Northern Ireland

intergroup antagonism during the negotiation process heightened ten-

sions and divisions within society and risked derailing the negotiations.

A similar pattern is currently being repeated in other conflict areas. In

Sri Lanka, a cease-fire agreement between the majority Sinhalese com-

munity and the Tamils concentrated in the north of the country was

5 Tellingly, the largest organization in civil society in Northern Ireland – theOrange Order – was denied representation.

Civil society in war-to-democracy transitions 197

reached in February 2002 – opening the way for several rounds of peace

talks. Several pro-peace civil society groups mobilized support for this

process and raised awareness among the masses for the need for it. At

the same time, Sinhalese nationalist groups joined forces with Buddhist

monks to condemn international involvement in Sri Lanka’s internal

affairs and the holding of peace negotiations. They staged vociferous

protests – often even more visible than demonstrations in support of

peace (Orjuela 2005: 133).

The formal exclusion of civil society from official ‘‘settlement nego-

tiations,’’ which took place in Bosnia, South Africa, and Northern

Ireland, is not true of all peacemaking processes. Without civic

ownership, peace is unlikely to be self-reinforcing and sustainable –

suggesting the need for broadening the negotiations. Because elites-based

pacts can be unstable, the involvement of civil society can establish the

process on more solid foundations (Wanis-St. John and Kew 2006).

Accordingly, political elites can attempt to achieve broader legitimacy

for peace deals by accepting and even encouraging civil society’s direct

involvement. In the Democratic Republic of Congo (DRC), the peace

settlement signed in December 2002 included not only government

representatives and rebel groups, but also opposition parties and dif-

ferent civil society organizations. Similarly, in the Ivory Coast the peace

deals reached in early 2003 incorporated a wide spectra of society

(Nilsson 2005) As the cases of the DRC and the Ivory Coast both

suggest, the inclusion of civil society is not a panacea, since there

remain serious constraints on the implementation of the agreements

reached.

Sometimes the government encourages civil society’s consultative

role, not to meaningfully involve broad societal constituencies but as a

form of ‘‘window dressing’’ (Barnes 2002). Guatemala is an example of a

case where civil society did not sign the peace agreement but provided

non-binding inputs to it. In the early 1990s, the government solicited

civil society’s contribution in an attempt to provide legitimacy to the

peace deal – even though it expected that its diverse composition would

prevent it from reaching a common negotiating position, and thus led

to only little or no substantive input (Krznaric 1999). The Assembly for

Civil Society (ACS) representing the views of the civic sector was

invited to contribute to the peace accord. Although the government

accepted the ACS’s participation because it assumed this newly created

body would be ineffective, its presence represented a breakthrough,

198 Roberto Belloni

and its influence on the outcome of the peace negotiation was signifi-

cant, as many of its proposals were included in the peace accords. At

the same time, however, the ACS’s proposals on socioeconomic and

agrarian reform – particularly land redistribution – were ignored.

Guatemala’s most important business association (the Coordinating

Committee of Agricultural, Commercial, Industrial, and Financial

Associations) successfully lobbied against it – leading one com-

mentator to describe this association as the key ‘‘uncivil’’ actor bent

on limiting citizenship rights in order to preserve economic privilege

(Krznaric 1999).

In sum, the vertical dilemma involves a choice between efficiency and

legitimacy. Effective negotiations might require minimizing the num-

ber of individuals involved, and preserving a certain degree of secrecy.

At the same time, however, the inclusion of civil society representatives

in peacemaking negotiations can increase the legitimacy of a peace

agreement and the prospects for its implementation – even when

domestic elites expect civil society involvement will have only a modest

impact on the drafting of the peace settlement. Whether or not civil

society participated in peacemaking, and in what form, in the post-

agreement phase promoting local civic participation and involvement

is a central strategy of the international support for post-settlement

peace processes. The dilemmas involved in this strategy are the focus of

the next section.

The dilemmas of peacebuilding and democratization

International organizations, multilateral and bilateral donors, and

international NGOs have invested considerable resources in building

and strengthening domestic civil society in conflict areas. One reason is

pragmatic. Years of conflict at best weaken state institutions and at

worst decimate their capacity to deliver services to the citizens.

Sometimes foreign occupation and rule contribute to the inexperience

of domestic elites in addressing effectively societal concerns. In

Northern Ireland for almost three decades direct rule from London

created a dearth of debate about social policy and prevented political

parties from developing public policy experience. In this vacuum, civil

society stepped up its work and often replaced local institutions in

service delivery. When state institutions do not function properly,

and the domestic political elite does not have the ability or willingness

Civil society in war-to-democracy transitions 199

to perform its duties, civil society presents international donors with a

valuable, cheap, and overall reliable alternative. The second reason for

supporting and developing domestic civil society reflects a hope, or an

aspiration. International actors believe that the stronger the civil

society, the easier the fulfillment of the post-settlement peacebuilding

and democratization tasks, and the easier the process of international

disengagement.

Because civil society in war-torn countries is typically extremely

weak and divided, the crucial question is whether international funding

can build pro-democracy civic-networks over the short-run, or whether

the existence of a thriving civil society depends on long-term historical

legacies hard to change by short-term human action – particularly in

war-torn regions.6 The war’s damage to social capital complicates the

capacity of communal groups and/or the state to recover after the end

of the hostilities. Conflict weakens the fabric of society, divides its

population, and undermines the interpersonal trust necessary for effec-

tive cooperation and collective action (Colletta and Cullen 2000). In

this context, while international funding may be indispensable to make

it worthwhile for individuals to cooperate in local groups, it can also

give rise to the systemic and the temporal dilemma. First, the absence or

low levels of international aid might prevent the emergence and/or

consolidation of bridging, pro-democracy civil society. At the same

time, however, international support comes with a cost – in particular

when local organizations become more accountable to international

donors than to the communities they are supposed to serve. Second,

while international donors recognize the need to support long-term

structural projects, institutional constraints push them toward a short-

term and top-down approach. The issues raised by the systemic and the

temporal dilemmas, respectively, are the focus of the next two sections.

6 In his research on civic tradition in modern Italy, Putnam endorses the long-term historical perspective. He argues that those Italian regions exhibitingcivic strength owe their success to the legacy of republican traditions duringthe Renaissance. Although a civic tradition can be beneficial to democracy,its contribution can also be overstated. In Italy, civil society has strong rootsin the north, but this is precisely one of the regions that supported Mussolini’sFascist Party in the 1920s and 1930s. Similarly, civic density did not preventdemocratic collapse in Weimar Germany, where pre-existing and well-organizedsocial networks were captured and mobilized by anti-democratic movements(Berman 1997).

200 Roberto Belloni

The systemic dilemma: too little or too much internationalsupport?

The potential drawbacks raised by heavy international involvement for

the sake of strengthening local civil society lie at the heart of the systemic

dilemma. To begin with, the exact boundaries between, on the one hand,

civic, inclusive, and democratic groups and, on the other hand, uncivil,

exclusivist, and xenophobic organizations can be difficult to determine

in theory, and even more so in practice. Intransigent nationalists may be

as likely to create local organizations to take advantage of foreign funds

as pro-democracy civil society activists. In the Serb Republic of Bosnia,

for example, there is evidence that some local NGOs have learned how

to talk the language of multi-ethnicity, dialogue, and compromise neces-

sary to obtain foreign funding, but they remain unabashedly nationalist

(Katana 1999). In some cases, opportunities in the civil society realm

may even present local politicians with a profitable alternative in times

of political decline. Prominent politicians can create their own NGO to

benefit from political connections in obtaining funding and influence.

The blurring between government and civil society is further confirmed

by the influence of family ties across the two realms. The wife of indicted

war criminal and former President of the Serb Republic of Bosnia

Radovan Karadzic served as the Director of the International

Committee of the Red Cross in the Serb Republic both during and

after the 1992–1995 Bosnian war. Similarly, the wife of Croatian

President Franco Tudjman was head of the Humanitarian Foundation

for the Children of Croatia during the 1990s. In both cases, these civil

society/humanitarian organizations enjoyed special privileges from poli-

tical authorities. Although formally independent from the government,

their political affiliation was never called into question. Occasionally,

similar government-sponsored organizations can even be created to

undermine the work of pro-democracy and advocacy groups. In

Rwanda the government created false NGOs for the purpose of attribut-

ing misleading propaganda to them. For example, the organization

Licrodhor spread outrageous anti-Tutsi propaganda – arguably to dis-

credit a respected human rights organization called Liprodhor.7

7 Personal communication with Benjamin Siddle, Peacebuilding Program Managerfor Trocaire in Rwanda (1999–2002), April 2005. The ruling Rwandan PatrioticFront regularly intimidates local organizations, most notably by accusing any

Civil society in war-to-democracy transitions 201

But even when inclusive and pro-democracy organizations can be

identified and supported, the financial relationship between the local

non-governmental sector and international donors may set into opera-

tion the systemic dilemma. International support is often indispensable

for domestic civil society groups, who face an enormous challenge in

raising funds in the context of post-war societies. Private resources are

rarely available, while the government is more concerned about

attracting international aid for its own activities than directing such

aid to civil society. However, while international support may be

indispensable, such support may come at a cost. Local participation

and ownership encouraged, financed, and sponsored by outside donors

is limited in its capacity to create domestic social capital and ownership

of the peace process. Civil society building strategies based on the

attempt to support the creation and development of civil society groups

from the outside face at least three major limitations.

First, international engagement in divided societies reinforces the

view that local communities have needs which can be met through

international funding channeled via domestic NGOs. The prevailing

approach to civil society development downplays the existence of local

resources and knowledge. Domestic community leaders learn that the

best way to attract international funding is to underline the problems

and disparage community members and their individual and collective

resources. Residents of conflict areas are frequently described as

severely traumatized individuals in need of outside experts and fund-

ing. Furthermore, when successful in attracting international funding,

local civil society groups develop an interest in perpetuating the system,

while presenting themselves as necessary transmission belts between

local communities, national governments, and international donors.

The negative consequences of this strategy are clear. Instead of devel-

oping existing human and social resources, increasing existing social

trust, and providing stronger foundations for democratization and

peacebuilding, civil society groups are given incentives to downplay

the existence of these resources and rely on external support (Mathie

and Cunningham 2002). Moreover, support from the outside may

decrease the need of leaders of an organization to rely on members’

contributions, with negative consequences on members’ voluntarism,

potential critics of ‘‘ethnic divisionism.’’ On these grounds, independent humanrights organizations may also be banned (Hampson 2004).

202 Roberto Belloni

internal accountability, and in general on the organization’s ability to

advance a democratic political culture.

Second, the international donor–local group relationship has shown

the dangers inherent in the financial dependency of the latter on the

former. Building and strengthening civil society is often conceived as an

externally driven process that is often entirely dependent upon inter-

national resources (Hadenius and Uggla 1996; Hann and Dunn 1996;

Pugh 2000). Despite much talk about partnership and consultation

between international agencies and domestic civil society actors and

the need to develop local ownership of peace processes, donors reg-

ularly ignore local organizations’ priorities and agendas. In order to be

funded, local organizations have to operate according to the frequently

changing preferences of international actors, instead of the needs of the

communities these organizations are supposed to serve. Rather than

creating local ownership, this practice may lead to domestic depen-

dency. As an old African proverb wisely states, ‘‘the man with his hand

in someone else’s pocket will move when he moves.’’ Accountability is

redirected toward the donor and away from the organization’s social

base, and the idea of political and social participation and empower-

ment is undermined by the reality of an externally driven process,

leaving domestic groups in a weak and subordinated position.

Third, the technical mould of international support strategies, cen-

tered on training local trainers and creating local capacities, limits their

effectiveness. Many citizens in post-war societies show antipathy

toward an approach to democratization and peacebuilding centered

on international experts and trainers who often do not speak the local

language, and have limited direct knowledge of local customs and

culture. Often the training itself is not so much in developing skills to

increase the impact of local activism but more pragmatically on how to

successfully write a project proposal and be funded. After domestic

organizations are socialized to the language and expectations of inter-

national donors, aid delivery creates a hierarchy. There is a big dis-

crepancy between the most developed organizations that work in the

bigger cities and take advantage of their contacts and exchanges with

international agencies and the larger number of small civic groups that

strive for visibility and funding. Furthermore, evidence across a range

of different cases, ranging from Bosnia (Belloni 2001) to El Salvador

(Foley 1996: 84–86), Timor Leste (Patrick 2001), and Sri Lanka (Orjuela

2005) shows how donors tend to fund moderate, middle-class groups,

Civil society in war-to-democracy transitions 203

whose interests might match their own, but that might not be too

representative of society at large. Ordinary people often avoid partici-

pating in NGOs because they do not feel these organizations reflect

their needs and priorities. As a result, as Beatrice Pouligny (2005: 501)

perceptively points out, international intervention pretends to ‘‘build

‘new’ societies while excluding the large majority of their members.’’

Not only does this externally driven strengthening of civil society

exclude and disempower, but it also can be interpreted by domestic

groups and citizens as a continuation of a post-colonial Western ‘‘civi-

lizing mission’’ (Kaldor 2003: 9), and because of this resented and

resisted.

But the main dilemma of international engagement in civil society

building may come from its very success. International resources and

opportunities made available for the third sector risk deterring talented

and motivated citizens from joining political parties and contributing to

the political process through institutional channels, driving them instead

to find a role in the voluntary sector. The decision to privilege NGOs as

intermediaries between citizens and the state – rather than, for example,

political parties or trade unions – ends up de-politicizing society. As Paul

Stubbs (1999: 31) has argued in the context of international assistance to

southeastern European countries, ‘‘donor support has created a kind of

‘anti-political political opposition,’ marginalizing formal opposition

parties, turning social movements into bureaucracies, and ultimately

buttressing the status quo while formally seeming to support ‘democracy

from below.’’’8 This suggests that, paradoxically, instead of getting

around the dominance of political elites to find new opportunities for

social and political change, international support may result in reinfor-

cing the existing ethnic, national, and sectarian divisions.

The promotion of ‘‘transnational democracy/human rights networks’’

is sometimes touted as a way to overcome these problems triggered by

international support of domestic civil society. These networks describe

horizontal, flexible, and fluid links allowing international organizations,

international and domestic civil society, and bilateral donors and foun-

dations (not all necessarily present in the network at all times) to

8 According to Julie Hearn (2000), this is not an unintended consequence ofinternational assistance. Rather, civil society support to transitional anddemocratizing states seeks to neutralize potential opposition within domesticeducated elites.

204 Roberto Belloni

cooperate almost as if they were part of one organization to pursue

common goals. Margaret Keck and Kathryn Sikkink coined the expres-

sion ‘‘boomerang effect’’ to describe how local groups, when blocked

politically at the national level, can find the support of network members

to clear the national situation: ‘‘Voices that are suppressed in their own

societies may find that networks can project and amplify their concern

into an international arena, which in turn can echo back into their own

countries’’ (1998: x). The key insight relevant also for post-war societies

is that when government is divided, ineffective, and not responsive to

society’s needs and unable or unwilling to deliver services to their

citizens, transnational networks allow for the meaningful participation

of local actors in decision making. In theory, local organizations and

populations can be involved in the process of institution building

and reconstruction, avoiding the pitfall of the traditional hierarchical

and top-down international support of domestic civil society (Holohan,

2005; Kaldor, 2003; Reinicke and Deng 2000).

While transnational networks provide an interesting and potentially

far-reaching way to sidestep the systemic dilemma, they are ‘‘not neces-

sarily harmonious, democratic or effective’’ (Kaldor 2003: 96).

Moreover, there is no reason to assume that the less powerful members

of the network (above all local populations and civic organizations) can

set the agenda against the wishes of the most powerful members – thus

undermining the network participatory ethos. In Kosovo, where Anne

Holohan (2005) found evidence of the efficiency of networks of democ-

racy, international actors took the lead. When they established the frame-

work for international intervention following the 1999 war, they ignored

the ‘‘parallel society’’ that dominated Kosovo’s social life throughout the

1990s, failing to capitalize on a wealth of human resources that would

have improved the legitimacy and efficiency of intervention. In sum, the

unequal relationship between powerful international actors and rela-

tively powerless domestic civil society groups is a structural condition

complicating civil society’s contribution to democratization and peace-

building. The problems raised by this unequal relationship also lie at the

heart of the temporal dilemma – to which I now turn.

The temporal dilemma: short-term or long-term intervention?

Civil society building strategies are also affected by a temporal dilemma.

Short-and long-term approaches clash in the attempt to support

Civil society in war-to-democracy transitions 205

exogenously the development of local civic organizations. Societies

coming out of conflict require immediate support to consolidate

peace, and prevent any non-state actor from challenging the post-war

transition. At the same time, conventional wisdom also advises in favor

of the adoption of long-term, structural, and ultimately more effective

projects. International donors face a dilemma between long-term

democratization and peacebuilding intervention and short-term initia-

tives with a strong crisis prevention focus.

More often than not, international democratization and peace-

building agencies apply a short-term time frame to their programs

and projects. All international agencies, including leading ones such

as the United Nations and the Organization for Security and

Cooperation in Europe (OSCE), have a tremendous institutional pres-

sure to minimize short-term risks of failure and maximize visible gains.

These agencies must be able to demonstrate immediate progress. UN

field operations are reviewed regularly. The Heads of Mission report to

the UN Security Council every six months. Similar reporting require-

ments are in place for other regional organizations, such as the

OSCE. Moreover, media and think tanks frequently challenge the

efficacy of multilateral international organizations such as the UN

and the OSCE, intensifying the pressure on these bodies to show that

peacebuilding and democratization ‘‘are working.’’ This pressure is

even more severe for bilateral donors whose resources are dependent

on a budget allocated annually by the respective national legislatures

and/or governments.

The need for immediate outcomes often prevails, pushing toward a

short implementation time frame and explaining three common pro-

blems of internationally driven civil society building programs. First,

international involvement focuses on the creation of domestic NGOs –

the easiest, short-term success to achieve for foreign donors, but also

the most superficial one. As Michael Edwards (2004: 95) points out, the

number of NGOs in society is the easiest thing to influence, but also the

least important. A study on civil society building in Bosnia confirms

that there is no direct correlation between affiliation with voluntary

associations and the amount of social capital (World Bank 2002b: viii).

Similarly, the presence in Rwanda of numerous local NGOs supported

by external aid did not prevent the 1994 genocide. For the most part,

these NGOs were apolitical, service-oriented, and did not bridge group

divides (Uvin 1998). Civic values and the commitment to a common

206 Roberto Belloni

life are notoriously difficult to change and require a long-term strategy

that extends considerably beyond the creation of local NGOs.

Second, the greater the international role and need for short-term

successes, the more international agencies fear that recognizing pro-

blems may be perceived as an admission of failure. Therefore, they

devote considerable attention to selling their achievements, rather than

addressing deep-rooted problems and undertaking those long-term

structural projects that might have a more profound impact. The

more subtle side of this tendency to ‘‘show results’’ in the short term

is the problematic inclination to blame local actors for delays, obsta-

cles, and drawbacks in the process. In southeastern Europe slow pro-

gress is frequently blamed on the ‘‘Balkan mentality,’’ the supposed

combined effects of socialism and war. Similarly, in the Middle East

international actors frequently point to the lack of established demo-

cratic traditions to explain continuing turmoil.

Third, the pressure to achieve short-term concrete results goes a long

way to explain ‘‘projectism,’’ the tendency to treat democratization and

peacebuilding as a set of discrete interventions incorporated into a

project with a relatively clear beginning, implementation, and evalua-

tion, usually with a six-month timespan, or at best one year. ‘‘Projectism’’

makes peacebuilding and democratization a top-down enterprise.

Because of the brevity of internationally funded projects, international

agencies have little scope to develop significant local partnerships and

include local actors in decision making. Instead, they make important

decisions about the priorities and allocation of international assistance in

the initial phases of a project, when international understanding of local

conditions is limited (Carothers 1999: 264). Seen in this light, the idea of

developing ‘‘local ownership’’ of the peace process sounds at best naıve.

The pressure to show immediate and concrete results advises against the

implementation of long-term structural projects. At the same time, short-

term and discrete projects militate against capacity building and domes-

tic ownership. Thus, the temporal dilemma perversely reinforces the

systemic one, making civil society building a weak element in the peace-

building and democratization strategy.

Conclusion: beyond civil society building

Civil society in deeply divided regions reflects the larger society in

which it is embedded. Inclusive, civic, bridging, and pro-democracy

Civil society in war-to-democracy transitions 207

organizations work alongside exclusivist, sectarian, and occasionally

even xenophobic groups. The sectarian divide, which is often institu-

tionalized in the peace agreement ending the war, may also be instru-

mental in preserving and entrenching the power of illegal and uncivil

groups, such as paramilitaries and mafia syndicates with extensive

connections at the political level. In this context, the attempt of

international organizations and donors to promote and support civil

society building is often disappointing. The hope that civil society

groups may play a useful role in sustaining democratization and

peacebuilding with a view to rebuilding failed states may prove mis-

placed. A healthy civil society is the sign of a well-functioning state,

not its cause. Instead of providing the foundations for a more stable

democracy and improving the functioning of state institutions, as

argued by civil society theorists such as Putnam, Diamond, and

others, a healthy civil society needs a functioning state, including

effective policing, an impartial judiciary, the rule of law, and the

capacity to guarantee personal and societal security. Only institutions

and social norms making possible debate, dialogue, and the non-

violent reconciliation of differences can sustain peacebuilding and

democratization. In other words, the source of social capital is not

simply attributable to civil society but can also be produced by the

state.

This point leads to a first policy prescription: the best avenue to favor

the emergence and development of domestic civil society in war-to-

democracy transitions is to strengthen the state by establishing stable

and efficient social and political institutions. Despite persisting clien-

telism and corruption in both established democracies and transitional

states, public institutions still maintain some degree of transparency

and political accountability. Efficient institutions have an indispensa-

ble role in securing the conditions for civic participation by keeping in

check illegal, uncivil, and xenophobic groups while ensuring legal

protection for domestic, inclusive, and pro-democracy organizations.

At the same time, however, state building is bound to be resisted.

Entrenched interests may work to prevent or slow down the process.

Often domestic elites benefit from a weak state. Paramilitaries and

mafia-like groups may see state building as a threat to their interests

and actively oppose it. In the anti-statist mood prevailing since the

1980s, even civil society advocates may oppose this strategy. As

Keane (1998) and Pouligny (2005) remind us, scholars of civil society

208 Roberto Belloni

and its activists tend to be uncritical about the potentiality of civil

society and privilege this realm over the idea of developing the state.

The establishment of a functioning, democratic, and responsive

state is only one part of the strategy to address constructively the

dilemmas of democratization and peacebuilding in conflict areas.

Civil society organizations need also to become progressively less

dependent on foreign resources for their funding. Civil society orga-

nizations are unlikely to be an expression of grass-roots democracy

and inter-ethnic reconciliation when they are more accountable to

their international donors than their members and local constituen-

cies. By providing local civil society groups with the money and

resources to address needs in their own communities, international

agencies might end up, albeit unwittingly, controlling local groups’

agendas. An alternative approach should seek to stimulate the

agenda, not to create or control it.

While there is no clear-cut solution to organizations’ dependency on

their donors, it is at least possible to imagine alternatives. One simple

but far-reaching initiative is to make funding available for projects with

a time frame longer than the current six months to one-year period.

Funding cycles should be at least three years, with periodic progress

assessments. Multilateral donors like the European Union are better

placed than bilateral donors to provide long-term, structural support.

From 1995–2000 the EU package known as the ‘‘Peace Programme’’

for Northern Ireland was so decisive in facilitating the rapid growth

and influence of the local civil society sector that it was expanded for

the period 2000–2006 (Williamson, Scott, and Halfpenny 2000). Over

time, however, the solution to the problem of civil society organiza-

tions’ dependency on foreign resources is to reform the very approach

international agencies adopt in their civil society support programs.

An effective civil society development strategy capable of overcom-

ing the dilemmas of democratization and peacebuilding requires both

the ability to identify and support inclusive, bridging, and pro-democracy

organizations and the capacity to acknowledge and mobilize local

knowledge and experience. The sound analysis and careful assessment

of the context, taking into account the conflict history, actors, issues,

and dynamics should become routine operational work for interna-

tional agencies and lead to framing options for international involve-

ment (Paffenholz 2001). Regional experts and anthropologists need

to be more systematically involved in the analysis of domestic

Civil society in war-to-democracy transitions 209

structures and in the identification of opportunities for engagement.

International support must be compatible with existing local realities

and careful not to legitimize uncivil and inequitable social structures

and organizations.

Moreover, building on reliable mapping of the local reality, interna-

tional policy must shift from a needs-based strategy to an asset- or

strength-based approach – in much the same way development agen-

cies have begun conceiving community development in Western coun-

tries (Mathie and Cunningham 2002). Even in the unfavorable

conditions of division and mistrust deepened by the experience of

war, divided societies do possess a network of community groups and

individuals committed to inclusive politics, dialogue, compromise, and

peaceful change (Colletta and Cullen 2000). As Varshney (2002:

289–297) demonstrated in the case of India, small-scale, localized

intervention based on local knowledge and resources can have a con-

siderable impact. Similar examples from Mozambique, Sierra Leone,

Cambodia, and Guatemala confirm that ordinary people can and often

do mobilize to make sense of and overcome the challenges posed by a

profoundly disrupted environment (Pouligny 2005: 502–503). Instead

of continuously identifying real or perceived gaps and needs, usually to

be met with the proper dose of international money and expertise,

international support should be based on local strengths and resources.

In sum, the role for international actors is not that of filling needs but

rather that of helping local civil society to mobilize resources that work

toward a self-defined community vision. International actors must

recognize domestic partners as agents with their own experiences and

knowledge, and work to improve the quality of domestic contribution

to democratization and peacebuilding. Domestic ownership is critical

to (re)building bridging social capital. Only when active citizenship in

civil society is placed at the center of international engagement, rather

than at its receiving end, will democratization and peacebuilding be

grounded on more solid foundations.

210 Roberto Belloni

PART IV

International engagement

8 International responses to crisesof democratization in war-tornsocieties

P E T E R W A L L E N S T E E N

Post-war situations are difficult periods of transition, involving many

simultaneous dilemmas for new and inexperienced governments.

Historically, the Weimar Republic in defeated Germany is a classical

case, where the first democratic government led by Social Democrats

had to administer the disaster created by the previous regime and also

do this while abiding by the punitive stipulations of the Versailles Peace

Treaty. Germany struggled through a turbulent period of the 1920s,

only to be severely hit by the Great Depression with mass unemploy-

ment and the exploitation of these conditions by a ruthless new chal-

lenge, the Nazis. Central issues in Hitler’s agitation were the Versailles

Treaty and the weakness of democratic governance. The international

community at the time chose to take a hands-off attitude, and allowed

the Treaty to be undermined parallel to the elimination of democracy.

A new, even more devastating, war began, twenty years after the end-

ing of the previous one. This example highlights the significance of

peace conditions, as well as the importance for international organiza-

tions to be concerned with their broader implications. The fate of the

Weimar Republic and the rise of Nazi Germany created a lasting guilt

feeling among democratic states.

Peacemaking today differs in many respects, but the historical les-

sons should always be kept in mind. The transition from war to peace is

cumbersome and the hope to build a functioning democracy should be

cautioned with a need for long-term thinking and an understanding of

the dangers of failure. The post-World War I experience also suggests

that it is necessary to survey one or two decades before trends become

clearer, whether negative, as in this case, or positive, which is the

experience of the Federal Republic of Germany after World War II.

This also illustrates that the dilemmas for the international community

are whether, when, for how long, and in what ways to engage in the

face of a crisis in the democracy building one is favoring.

213

Since the late 1990s, international donor agencies have taken up

democracy as a major theme, and there is an emerging literature on

development assistance and promotion of democracy and good gov-

ernance as part of conflict prevention (Collier et al. 2003; Sida 2005). It

has been demonstrated that democratic states rarely fight wars with

one another and that democracies have fewer internal wars than other

forms of governance (Russett 1993; Russett and Oneal 2001). Since the

Cold War, it is clear that peace agreements in internal conflicts over

government often resort to democratic constitutions and elections

as ways of ending conflict (Harbom et al. 2006). Supporting the devel-

opment of a sustained democracy appears a reasonable approach in

post-war societies. It might help to bring about stability in difficult

conditions in order to ensure that a conflict does not return. But the

possibilities of supporting peace through democracy building is an

avenue which has become questioned (see, e.g., Paris 2004).

This approach to peacebuilding may result in difficult dilemmas at

critical junctures in a democratization process. The stipulations of a

democracy-oriented peace agreement may be difficult to implement.

Election outcomes may favor groups challenging the peace. The pace

of democratization may be different in different parts of a society

(socially, regionally). Parliamentary stalemates may prevent reforms of

society. Some studies demonstrate that the transition from autocratic

conditions to democratic ones may be the most conflict- and war-prone

phase of all (Hegre et al. 2001; Human Security Report 2005). That

observation might be particularly applicable to post-war conditions.

Taking this together, the way in which democracy is promoted is one

of the most critical issues of post-war crises of implementation.

However, the literature on how international organizations (whether

governmental or non-governmental) or media react to particular crises

in such a transition is surprisingly limited. There is some insight suggest-

ing that such issues rarely enter the international agenda. For instance,

the armed conflicts that reach the UN Security Council are regularly less

than half of all ongoing armed conflicts (Malone 2004; Wallensteen

2007). They have to pass through the screening of governments (often

unwilling to allow internal conflicts in their own countries to become

internationalized), regional powers (preferring to deal with matters

bilaterally or in regional bodies under their control), and the veto powers

of the UN Security Council Permanent Members. If we add that post-

war democratization crises constitute a subset of such conflicts, it can be

214 Peter Wallensteen

asked why they should be of concern to, for instance, the United

Nations. They may not easily be understood as threats to international

peace and security, which is stipulated for Security Council action. Still,

there is the observation that the UN is involved in democracy promotion,

and particularly in smaller countries rather than larger ones (Newman

and Rich 2004). On the whole, the UN is probably more engaged in

weak states than in major states, and democracy promotion is no excep-

tion to this pattern. However, in the case of UN involvement there is an

interest in the long-term construction of democratic governance, for

instance as part of a peacekeeping mission (Ludwig 2004; Newman

2004). The focus in this chapter is on the crisis or short-term situations:

when there is an observed threat to democratization, how is that

handled? If such crises are not dealt with in a constructive way, it has

implications for the subsequent efforts of democratization.

Clearly, there is a doctrinal development in the international under-

standing of when international institutions should react: democracy

issues are seen to be part of the international agenda in principle. It

needs to be seen in practice as well. There is a record of the UN reacting

to military coups overthrowing elected governments (Haiti 1993,

Sierra Leone 1997; see Farer 2004). This is not consistent, however.

The military takeover in Thailand in September 2006 did not result

in any official Security Council action, whereas a similar event in Fiji in

November generated two press statements. This refers to a situation

where the democratic system is overturned by unconstitutional means

from ‘‘inside’’ the country. There is also the possibility of outside force

being used in overthrowing a regime, often without prior UN approval

(Panama 1989, Iraq 2003, interestingly neither case directly linked

to post-war conditions), sometimes with a global understanding

(Afghanistan 2001). However, the number of military coups is low,

as are the military interventions. There are few instances of this kind to

react to. In more common post-war situations the dilemmas may look

different. Continued development toward a sustained democratic

society may be disrupted, not by constitutionally external forces (mili-

tary factions, outside powers) but by the participants themselves, nota-

bly a government clinging to power or a well-organized opposition

using means beyond the democratic framework. These are the more

limited democratization crises to which this chapter is addressed.

The international community is likely to face difficult problems when

confronting such a democratization crisis. The amorphous international

International responses to crises of democratization 215

community consists of organizations and actors supporting the values of

democracy, human rights, welfare, and peace. Its shape will vary but

may include UN organs, the European Union (EU), donor states, and

international civil society organizations (Wallensteen 2007: 251–258). It

will, almost by definition, exhibit differences of opinion and thus be hard

to move in a uniform direction. The focus in this chapter is on the actual

behavior of this community and how it tends to deal with dilemmas it

faces when confronting a democratization crisis. The chapter does not

address what would be the ‘‘best’’ choice, but rather tries to understand

which courses of action are probable under what conditions.

More specifically this chapter deals with the four dilemmas of this

book in the following ways: First, there is the issue whether to react, the

systemic dilemma: does the international community react at all or

does it prefer to let the local actors ‘‘play it out’’ among themselves?

Second, it discusses when the international community is likely to react

and when to stop reacting – the temporal dilemma: for how long is the

international community sustaining the reactions that it chooses?

Third, it raises the choice of what means: which actions are legitimate

and are they also effective – the vertical dilemma – and how do the

targets of action react? Fourth, by analyzing the three first dilemmas

also the horizontal dilemma is brought out: which segments of the

international system are included in a particular pattern of reactions

and which ones are not, whether by deliberate exclusion or by their

own choice? The strength of international consensus is possibly a

significant determining factor for action and impact.

There is no claim that this chapter will provide definitive answers to

these questions. The framework of understanding post-war develop-

ments as dilemmas, however, is a fruitful way of approaching reality,

and some novel issues can, hopefully, be raised. Mostly this will be

done by way of illustrative examples.

A democratization crisis: Uganda 2005

Developments in Uganda constitute a plausible starting point. The

victor of the Ugandan civil war, Yoweri Museveni, received solid

international support for pursuing a fairly generous and inclusive

policy, aimed at economic reform and social reconstruction. In the

Constitution of 1995 it was stipulated that the president could stay in

office for only two terms. This was applauded and donor support

216 Peter Wallensteen

continued, in spite of signs of a less than democratic implementation of

policies in the country. By late 2004, however, Museveni faced the

possibility of having to step down, after being in power since his

military victory in 1986. He chose instead to amend the Constitution,

so as to allow himself to run for a third term. His main opponent,

Dr. Kizza Besigye, previously exiled, returned in November 2005 to

contest the elections, only to be arrested on charges of treason and

rape (Tangri 2005). The international donor community reacted in

December, when, for instance, Sweden and the United Kingdom decided

to freeze further disbursement of direct budgetary support to Uganda.

The respect for human rights and democracy was stipulated in the

bilateral agreements, and thus Uganda was seen to have breached that

stipulation. There was even a fear of a military takeover.

The international community began to respond, in what it saw as

significant ways to uphold the principles of democracy, and prevent a

new polarization in the country, which already was involved in armed

conflicts both internally and externally. These acts by donors could be

seen as promoting democracy and preventing escalation of conflict at the

same time. The strains on Uganda’s path to sustained democracy had by

then been visible for at least a year, but also the election campaign in

2001 could have been questioned. Clearly, donors preferred a quieter

dialogue with the regime, rather than making public pronouncements.

As the international community had been involved in promoting demo-

cratization, it would be difficult not to react at all. Its first choice was for

a less confrontational approach. When this clearly did not work, firmer

reactions were slow in coming. The constitutional change as such could

not be opposed, as it was regarded as a matter of Uganda’s internal

affairs. However, the arrest of a leading opponent (and accompanying

measures against his supporters) threatened the fairness of the electoral

campaign. These seem to be the events that triggered international

reaction. A free and fair electoral process is central for democratic

development. Without this, the elections themselves and their outcome

can be disputed. Not to react to the arrest might have reduced credibility

in the international stand on democratization in Uganda. This example

demonstrates how a series of events can trigger the international com-

munity’s engagement. Still, the reactions were far from concerted (indi-

vidual donors taking different steps at different times) and restricted to

some forms of development assistance, not all. Was this likely to be

effective in stemming the erosion of democracy in Uganda?

International responses to crises of democratization 217

With this example in mind, we may ask which are the issues that will

be seen as a democratization crisis? The discussion on good governance

suggests a number of such issues: the rule of law, free and fair elections,

free media, a vibrant civil society, independence of the judiciary, inde-

pendence of the representative body (parliament, national assembly),

secure human rights, widespread participation. With this wide under-

standing of good governance and democratization, there are a number

of events that could be seen as ‘‘crisis’’ or ‘‘warning signs.’’ Clearly, only

some of these will trigger international reactions, although there might

a logical sequence to many of the events. The removal of judges as a

way of controlling the judiciary is unlikely to be seen as such a crisis but

may be a first step for a government to maintain itself in power. The

recruitment of staff into the central administration, and even biased or

corrupt operations in the state machinery will seldom unleash con-

certed reactions. In the reporting – by diplomats and media – this will

no doubt be observed, but not seen to be enough to single out a

particular government or country in an environment where similar

events have taken place before or are common in the region. For

instance, the closing down of universities when students or teachers

have voiced criticism seems very seldom to engage international gov-

ernmental organizations. Even increased regime control over media or

over civil society tends not to stimulate reactions. However, the donor

community is more likely to react to this, for instance as part of the

negotiations on renewal of development assistance. Here, there is

another dynamics at work: civil society and media will have colleagues

in neighboring or donor countries that react, thus stimulating political

consideration. If there were more solidarity in the university commu-

nity, actions curtailing academic freedom might also enter the arena of

political reaction.

A hypothesis is that the type of democratization crisis that is more

likely to generate international action is the one that directly concerns

the distribution of power in a country, notably local (governmental or

non-governmental) actions against the parliament and against the elec-

toral process. Thus, it is here suggested that the electoral process is the

single most typical event that is seen as a carrier of democratization and a

significant sign of progression toward ‘‘good governance’’ and democ-

racy. Threats to this process may trigger international action.

There are several explanations for this. In a democracy, elections

have a particular standing as a way of determining the political strength

218 Peter Wallensteen

of actors. This is so, although most analysts would agree that election

outcomes may, in fact, be determined by a number of earlier actions.

Furthermore, threats and/or manipulation of elections during the

course of a public campaign are likely to draw considerable media

attention and thus result in demands for reactions. In addition, as the

international community we discuss here largely is constituted by

democratic states this community is likely to understand the threats

to elections and, thus, be more willing to react. These possible explana-

tions are illustrated by the events in Uganda. This case, however, also

suggests that responses to events that threaten elections might be too

limited a perspective and even too late a moment to react firmly.

Let us then proceed to discuss whether and when international

reactions are likely in democratizing processes – by what means, with

what degree of consensus, and for how long. In doing this, the different

dilemmas will be illustrated, and, as will be made clear, several dilem-

mas may emerge at the same time. This helps to show the complexities

facing political decision-makers when various triggering events make

decisions necessary.

International response: react or not?

The systematic dilemma refers to the issue of ownership and posits the

question of whether there is (or should be) international or local con-

trol over democracy building. Ideologically, the international commu-

nity would normally adhere to the position of giving primacy to local

ownership. However, many democratization processes are nurtured by

international actors, through, for instance, development assistance.

Thus, there is international involvement, making triggering events

not only a matter of local politics but also an international concern.

Furthermore, from the point of view of the international system, a post-

war crisis in implementation of a peace agreement or of an agreed

democratizing constitution can be seen as an early warning sign of a

possible return to armed conflict. To react also becomes a matter of

conflict prevention. A common lesson from the 1990s is the need to

react early to have a maximum impact. The primary responsibility for

the UN is to prevent war. Security Council actions under Chapter VII

can only be taken to deal with a threat to international peace and

security. The idea of conflict prevention, however, highlights the pos-

sibility of reacting under the less demanding Chapter VI which asks for

International responses to crises of democratization 219

the pacific settlement of disputes. Democracy promotion per se is not in

the UN Charter, and action in this field will require broad agreement to

be pursued. Since the end of the Cold War, the UN Security Council has

regularly concerned itself with internal affairs and also with demo-

cratic conditions. The report by the UN Secretary-General called

‘‘An Agenda for Democratization’’ in 1996 is credited with giving the

UN a firmer role in this. The UN practice since then has made democ-

racy a significant concern (Newman 2004: 194). These are factors that

would generate a willingness to respond in the face of a crisis in

democratization.

At the same time, democratization is a highly sensitive process.

International reactions are likely to be interpreted in terms of both

who pursues them (a matter of the West versus others, i.e., a horizontal

dilemma) and their impact on the local power struggles in the con-

cerned society (who gains, who loses, i.e., affecting the vertical

dilemma). The international bodies can easily be seen as partial – by

intention or in effect – and supporting certain social forces. This con-

tradicts a basic tenet in international governmental donor policy to

remain ‘‘above,’’ ‘‘beyond,’’ or ‘‘outside’’ such domestic power strug-

gles. It is an issue that has faced all peacekeeping operations in internal

conflicts since the first one in the Congo in 1960. Today, the inhibitions

against acting and even being partial may be lower, as long as it is

justified in terms of promoting democracy.

The case of Uganda demonstrates this. The threat of deteriorating

conditions in Uganda’s democracy has not been a matter for the UN

Security Council. It has not even passed a resolution on the ongoing

armed conflict in Northern Uganda, although the issue has been

addressed in other ways. The UN Security Council has taken up

Uganda’s relations with the neighbors to the east and south: The

Democratic Republic of Congo and Rwanda. It is not likely that the

electoral question in Uganda will come to the Council’s agenda unless it

can be seen to lead to a serious international situation. Thus, the UN is

constrained and the principle of non-involvement in internal affairs is

likely to prevail. A broad international coalition behind actions is less

probable. The horizontal dilemma of who should intervene was actu-

ally highlighted and the outcome in this case has been that the UN did

not act at all. Consequently, the actors dealing with Uganda’s demo-

cratic crisis were instead major donors, who have their own reasons for

this (not the least that it was their taxpayers’ money that was supplied

220 Peter Wallensteen

based on agreements that were contradicted). In other words, the

question of whether and when to react, for an international body,

will be limited by its mandate, which is an expression of the underlying

understanding of why the organization exists in the first place. In

bilateral relations other paths can be followed.

A hypothesis is, however, that matters will turn out differently if the

international body is already involved in the situation. As noted, the

civil war in Uganda had not robustly entered the UN Security Council

agenda, and as a consequence, there was no record of a UN Security

Council commitment to internal democratic development. This con-

trasts other cases. Rich (2004) provides a list of twenty such situations,

all relating to wars that have had considerable and long-term UN

commitment, going from Namibia (first resolution in 1978) to

Afghanistan (in 2001). For different reasons – decolonization, region-

ally destabilizing civil wars, and global terrorism – the world body has

become engaged and thus resorted to dealing with the situation by

promoting democratic solutions. It can thus be suggested that the

way an international organization such as the UN enters a conflict is

fundamental for understanding which democratic challenges are likely

to result in action by that organization. Once committed, electoral

irregularities are likely to be interpreted as issues for preventive action

and the world body will remain engaged.

Uganda’s situation was not on the agenda of the world body. Also

the country had a record that generated international support for its

policies, internally as well as externally, thus, in effect, giving more

autonomy to its leader. The democracy crisis of 2005 did not connect

to a previous history of Uganda as an agenda item. It may now be

valuable to turn to another example. An entirely different case is the

international involvement in Bosnia-Herzegovina, where the Office of

the High Representative (OHR) often reacts to events which are seen to

threaten the 1995 Dayton Peace Agreement and prevent progress

toward a multiethnic democratic society. Thus, OHR has removed

candidates from electoral rolls, restricted parliamentarians, and even

replaced members of the cabinet. This has made the OHR a central

source of power in the country, highly contested by all the groups. In

reality, OHR is a body appointed from abroad, albeit by democratic

states, and has not emerged through a democratic process in the

country in which it is operating. It makes Bosnia-Herzegovina a demo-

cracy with an ‘‘enlightened’’ overlord. As the heavy international

International responses to crises of democratization 221

peacekeeping presence is seen as necessary more than a decade after the

peace agreement, this case also illustrates the protracted nature of a

‘‘robust’’ democracy promoting effort.

The issue of democracy promotion thus relates to the horizontal

dilemma by asking which body is to react to a democratization crisis.

An increasing number of organizations have democracy explicitly on

their agenda. This includes the EU, OSCE, and NATO, all active in

conflicts and democracy issues in the larger European space. The

Organization of American States (OAS) takes this role in the Americas.

The African Union (AU) has an expressed objective of promoting democ-

racy in Africa. Also international bodies less directly concerned with

international peace and security do have democracy on their agenda: this

is true for the United Nations Development Program (UNDP), the

Organization for Economic Cooperation and Development/Develop-

ment Assistance Committee (OECD/DAC), as well as the different

commissioners on Human Rights (UN, EU, Council of Europe). The

reports of these bodies are significant in highlighting challenges to

democratic development. However, the non-governmental organiza-

tions are those most unconstrained to deal with these issues, at least as

a way of alerting the governmental communities on what goes on in

particular countries (Amnesty International, Human Rights Watch,

International Alert, International Crisis Group).

The large and increasing number of actors engaged in promoting

democracy and human rights means that threats to democratic devel-

opments in post-war societies are increasingly unlikely to take place

without considerable attention. This, furthermore, may in itself work

as a deterrent to governments contemplating, for instance, changes in

the constitution favoring the incumbent. The record in Africa of peace-

ful transitions from one president to the next is improving (Southall,

Simutanyi, and Daniel 2005). However, often issues of candidacy for

re-election are seen as sensitive matters, and neighboring or even extra-

continental governments are careful not to involve themselves openly

in a legislative domestic process. It will require particular conditions

for explicit government reaction. The fact that a country finds itself in a

post-war transition with international commitment is likely to be one

such event. In this situation the international community has invested

considerably in future stability and will react more legitimately than

otherwise would be the case. Such an investment, it is argued here,

provides the general background for international concern and even

222 Peter Wallensteen

legitimacy for international actions in a democratization crisis. The

systemic dilemma does at the same time involve a horizontal dilemma.

If there is a record of commitment, an international body is more likely

to get involved also in a democratization crisis and thus be highly

inclusive. This means more international consensus on action.

However, if the democratization crisis takes place without such a

prior organizational commitment, it is less likely to have universal

support. Instead, action will be pursued by regional organizations

(building on regional consensus), by individual donor countries (with

a commitment to democracy), or by civil society organizations. Still,

there is a general trend of democratization crises generating more

international concern. The question whether this results in a tilting

toward more international rather than local ownership of democrati-

zation will, however, require further analysis.

International response: when?

The temporal dilemma involves two aspects: when will action take

place and when will it end, i.e., the short-term vs. the long-term

dimension? Let us begin by raising the issue of when international

actors enter a democratization crisis. This requires us to ask which

challenging events are expected to lead to international reaction: It is

here suggested the events in electoral processes are more probable than

others to trigger an international response. This is so – as we observed

earlier – because the election process is regarded as central to the

conduct of democracy. According to conventional democracy theory,

the electoral process and elections are the occasions when the general

public (the electorate) can pass judgment on the achievements and

responsibilities of the incumbent government. The public’s main sanc-

tion is to vote for an alternative to the present rulers. Peaceful change is

accomplished this way. Free and fair elections are important elements

of democratic participation. Thus, in democracy building the first and

second elections after a war are those most closely watched. For inter-

national media, elections also incorporate the drama that provides for

readers’ interest. It is a peaceful battle that can be followed and under-

stood globally. It reduces the complexities of ‘‘distant’’ societies into

manageable alternatives. With this in mind, it can be concluded that

there are several reasons for elections to become central indicators of

the progress of democracy promotion. Challenges to the electoral

International responses to crises of democratization 223

process, then, would most typically involve events that may trigger

international responses. There are, however, different elements that

together constitute ‘‘the electoral process’’ and they may solicit differ-

ent reactions. The example of Uganda illustrates a first type of trigger-

ing event: the patent arrest of a leading opponent. This could be

generalized: physical and direct threats to opponents are events that

generate considerable attention internally as well as internationally.

The attempt to poison a leading opposition candidate in Ukraine in

2004 generated international concern and galvanized domestic opposi-

tion, preventing the incumbent government from manipulating the

elections to its advantage. This, of course, was not a case of a post-

war situation, but one which was close to both the EU and Russia,

which had a strategic interest. As a consequence, it became a concern to

the EU, rather than to the UN (Maksymenko 2005; Solana 2005). In a

post-war situation, the physical security of the leaders is likely to be a

major interest, and probably an issue to which they themselves pay

considerable attention. Particularly if there is a peace agreement, such

threats are closely watched by international observers. This means,

paradoxically, that the safety issue may become more urgent, the

further the process moves along. When a democratic transition see-

mingly has been stabilized, the security provisions may be relaxed.

After all, the war in Uganda that led to the new Constitution ended in

1986. By 2005, matters should have been more stable. In fact, they

were not.

This is further illustrated by the modern experience of Lebanon. The

civil war ended in an agreement in 1990. A policy of reconstruction was

initiated and, by many standards, a return to a new ‘‘normalcy’’ seemed

to take place. For instance, in 2000, Israel decided to withdraw from

the south of the country, in the expectation of stable conditions as well

as in the implementation of a gentlemen’s agreement with the domi-

nant force in the south, Hizbollah. However, in the run-up to the 2005

elections, the leader associated with the recovery – as well as with a

desire for independence from Syria’s influence – Rafik Hariri, was

assassinated together with twenty others in a huge bomb blast in

Beirut on February 14, 2005. The electoral process had therefore lost

one of its main contenders, and this led to an international as well as a

national crisis. Indeed, the democratic gains in the country seemed to

be threatened. In fact, they may have been strengthened as people

turned out in large demonstrations to demand Syria’s withdrawal.

224 Peter Wallensteen

Interestingly, the issue was also brought to the UN Security Council

and an international investigation was ordered, as well as the with-

drawal of the Syrian troops. In April 2005, the Syrian military had

pulled out and the elections in May and June gave strong support to

Lebanese factions that were critical of Syria, including giving a promi-

nent role to Saad Hariri, the son of Rafik Hariri. In this case, we see that

the elimination of a democratic leader, who was also regarded as

successful in maintaining economic development and balancing

between different factions, immediately sparked an international reac-

tion. Certainly, Lebanon’s geopolitical location helps explain this, but

it is also significant that the stability of the country has been of con-

tinuous concern to the international and regional community (includ-

ing peacekeeping and peacemaking efforts). There was considerable

visibility coupled with a willingness to act.

The examples of threats to leaders may be extreme cases, but there is

also an issue of widespread electoral violence. Often, the fate of leaders

may be only the most easily observable event of systematic practices.

For the international governmental community, it may be easier to act

if leaders (who may be acquaintances to other leaders) are exposed to

dangers. For the international non-governmental community, not least

those concerned with fair electoral policies, events such as intimidation

of voters, obstacles to campaign meetings, killings of local party work-

ers, as well as interference at polling stations are equally or more

worrying. Reports on such events help form the picture an interna-

tional audience acquires on a particular election campaign. For a more

direct international reaction, this may not be enough. Elimination of

one leader may trigger more media coverage and result in more outrage

internationally than the fate of dozens or hundreds of ‘‘anonymous’’

election workers.

A second set of events that might result in international action relates

to the election outcome. If there is international monitoring of the

process, there will be reports to work from. However, if there is such

a presence it is more likely that the outcomes are correct. The non-

observed cases are probably those with the most irregularities, parti-

cularly in post-war conditions. A paradox is that such elections

generate less international concern, and those outcomes may be more

readily accepted, although they are domestically contested. The result,

as a consequence, is that internationally observed elections are also

those that are the most likely to trigger international reactions. If the

International responses to crises of democratization 225

electoral conduct can be criticized, and if the outcome is also affected

this is likely to spur international responses. Attempts by incumbents to

‘‘steal an election’’ in a process of great international significance is

likely to be resisted.

A now almost classical case in point is the outcome of the presiden-

tial elections in the Federal Republic of Yugoslavia in September 2000.

Yugoslavia (today Serbia) found itself in a post-war situation, following

its involvement in the wars of Slovenia, Croatia, and Bosnia-

Herzegovina (1991–1995), as well as in Kosovo in 1999. The incum-

bent, Slobodan Milosevic, was running for re-election. The electoral

announcement of his victory was met with disbelief, a general strike

was announced, and demonstrators assembled around the Parliament

building in Belgrade in early October 2000. The international reactions

were sharp. It was clear, for instance, that international sanctions

would not be lifted unless the results were correctly reported. In the

end, Milosevic had to admit that he had lost the elections and he

stepped down. This development was actually not part of the peace

agreements or other post-war arrangements. The events of 2000 were a

matter of following the constitutional set-up arranged by the domestic

political process. Still, the outcome had repercussions for the entire

region, and a legitimate outcome was important for the possibilities of

peacemaking in the post-war societies of the Western Balkans. The

elections received intense international coverage. The manipulation of

the results triggered sufficient action to make clear to the incumbent

government that it would not enjoy international legitimacy if it per-

sisted. The domestic opposition may in the end have been more impor-

tant in Milosevic’s decision to step down. His power base was eroded,

but also his main international allies, notably Russia, abandoned him

during the first days of October 2000. The triggering event was elec-

toral fraud: Milosevic was a key regional actor, the legitimacy issue (an

element in the vertical dilemma) was no longer in his favor (but rather

to the advantage of the international actors), and there was sufficient

consensus among key actors to unite against him (an aspect of the

horizontal dilemma). By his actions, Milosevic orchestrated his own

downfall.

An additional aspect is the aftermath of the elections; will the winner

and loser have a chance to play the role the results demand? As the

previous examples suggest, it is difficult for a national government to

withstand the pressures resulting from an electoral process. It may face

226 Peter Wallensteen

the inevitable choice of giving up power. From the point of view of the

power holders this may be a difficult choice. What will happen to the

outgoing leader, the losing administration, the party, his/her family?

There are many elements of an internal security dilemma posing ago-

nizing choices on the loser (Wallensteen 2007). A number of leaders

have had to face this, and chosen to abide by the result. The list includes

Daniel Ortega in Nicaragua, Kenneth Kaunda in Zambia, Daniel arap

Moi in Kenya. In all these cases, the former leaders have been able to

make a living in their own countries following the electoral defeat.1

The recent advent of war crime tribunals may suggest one more disin-

centive for leaders in post-war societies to step down. Milosevic found

himself at the International Tribunal in The Hague less than a year after

his electoral defeat. Other leaders, such as Moi, may face charges of

corruption (as happened to Frederick Chiluba, who overturned

Kaunda and himself lost a later election). In these instances, the inter-

national community is recorded to have acted in favor of the loser

yielding power.

In some instances, obeying by the outcome has been met with resis-

tance, however, resulting in protracted stalemates. In 1988, war-torn

Burma/Myanmar had a democratic election to appoint a constitutional

conference. The democratic opposition clearly had the upper hand. The

National League for Democracy led by Ms. Aung San Suu-kyi won

overwhelming support and as a consequence should have had the

dominant role in Burmese politics. The military regime, however,

refused to give up its power. A stalemate resulted, still remaining.

The pressure on the military regime has been strong, including indus-

trial sanctions by the United States and targeted action by the European

Union. The Nobel Peace Prize was awarded to Ms. Aung San Suu-kyi in

1991. The UN Secretary-General has appointed a special representa-

tive for this issue, as has the Association of Southeast Asian Nations

(ASEAN). The pressures are obvious, and the neighboring countries

have found themselves in a squeeze. Lately, ASEAN has become more

vocal in demanding democratic reform in Burma. Burma’s closest ally,

1 Most remarkably, Daniel Ortega was again elected president in Nicaragua in2006, thus, being at the same time the first president in the country’s history tostep down peacefully and be re-elected. In 2007, the Sudanese businessman MoIbrahim announced a substantial prize for African leaders who ‘‘have taken officethrough proper elections and left having served the constitutional term stipulatedwhen taking office,’’ in effect rewarding constitutional behavior.

International responses to crises of democratization 227

China, however, has not imposed any action, preferring to build up its

own investment in the country. The Burmese issue has been blocked

from entering the formal agenda of the UN Security Council, but an

informal discussion has been allowed. The military government has

defended itself by pointing to its ability to conclude agreements with a

number of the ethnically based insurgencies. This, it has seen as a

development toward a more multinational state. However, given this,

international actors have argued that now may be the time to deal with

the issue of national governance and democracy, as part of a post-war

process and as a way of preventing a return to serious conflict.

The observations so far point to the electoral process, particularly

the election campaign, as a triggering event for international action in a

democratization crisis. Clearly, also the outcome will be closely mon-

itored by international actors. Possibly, there is a sequence of attention:

if the campaign has been followed, it is likely that also the outcome is of

concern. In other ways, if the issue is on the agenda then it will not

easily go away, be it an international or regional organization or a

coalition of donors that are concerned. It still asks the question raised

by the temporal dilemma: is there a preference for short-term commit-

ment, or will it be sustained over a period of time? To this we now turn.

International response: for how long?

The duration of commitment is a second aspect of the temporal

dilemma. It will partly, of course, depend on what type of action is

taken (to which we will turn in the next section). Normally, action

would end when a clear-cut success or failure can be established. The

promotion of democracy, particularly in post-war conditions, seldom

offers such easy points. In reality, for instance, peacekeeping opera-

tions will be sustained as long as the troop-contributing countries are

willing to put up forces, or funds are enough for continuing. As there

are always some possible threats to refer to, the operations can become

protracted. There is a similar tendency for economic sanctions to go on

for a considerable period of time. Even when sanctions are targeted, it

can be politically difficult for the initiators to terminate them, as that

might be seen as yielding to the targeted country or its regime. If the

demands are linked to decolonization, however, independence might

be a clear event for major disengagement (e.g., East Timor in 2002).

Even so, the concern for democracy may extend far beyond such points.

228 Peter Wallensteen

Some of our examples – Lebanon, Uganda, Bosnia-Herzegovina,

Burma/Myanmar, and, more recently Liberia – point to the long-term

nature of democracy building in post-war conditions. Formative crises

might come late in the process, even after a series of elections. As we

have already mentioned repeatedly, once an issue is on the interna-

tional agenda, it is likely to remain there for a considerable period. The

attention may vary across time, however.

One might even suggest that there are different phases of such inter-

est. In the war period, the focus is likely to be on issues of negotiations

and peace agreements. This involves international diplomacy, special

missions, mediation efforts, strategic political considerations, etc.

Once a war has ended, there might be a requirement for peace opera-

tions, humanitarian efforts, refugee assistance, demobilization, and

technical support for the implementation of agreements. When this is

seemingly managed, it will be followed by international development

assistance aiming at recovering from the war effects and the creation of

self-generating economic conditions. Supposedly, from this will follow

the building of democratic conditions in all its aspects (for instance,

media, civil society, rule of law, academic freedom, security sector

reforms). After this, aid will be replaced with trade and promotion of

democracy with the indigenous solidification of good governance. This

is an ideal model. In fact, the international commitment may wane very

quickly. For instance, once refugees have returned, armies have been

reduced, and the first elections have been conducted, global concerns

will shift to other crises. ‘‘Real’’ attention will not return until a new

crisis unfolds. Lebanon might be a case in point. The international

commitment between 1990 and 2005 was limited to particular issues.

Similarly, Haiti was of great concern in 1993–1994, with sanctions and

threat of military actions unless the deposed government was allowed

to return. It did so in 1994, and Haiti dropped from the immediate

concern of international organizations. In 2004 a new crisis took place.

What had been achieved in a period of ten years did not provide for a

sustained democratic form of government.

In the case of Uganda, Museveni managed to win the elections, and

the donors faced a new dilemma. Were they going to sustain their

actions against the regime or were there other means available to deal

with the government? Donors gradually reestablished relations. For

instance, half a year later, Sweden restarted 90 percent of its budgetary

support, arguing on the basis of an EU evaluation that there was now

International responses to crises of democratization 229

multi-party representation in the parliament. Ten percent was withheld

as an indication of continued concern. The conduct of the elections

process was not fully acceptable, but the outcome was likely to have

been the same, a victory for the incumbent. However, to work with

Uganda and the new parliament would increase the chances of further-

ing democracy in the long run. ‘‘A democratic state is not shaped

overnight,’’ said the head of Sida’s Africa Division (Sida 2006).

Thus, democracy promotion in post-war conditions is a long-term

commitment, but international high-level attention is impossible to

maintain for such a period of time. New issues will emerge and right-

fully deserve their shares of attention. The realpolitik conditions may

also dictate against a sustained and coherent democracy approach:

provision of aid to exposed populations requires a functioning state.

Thus, Uganda could find a way back to donor funding. If Uganda,

Afghanistan, or Bosnia-Herzegovina were to relapse into war, consid-

erable efforts and resources would have been wasted.

Thus, there is a need to develop a sustained international interest in

post-war conditions. In 2005, the UN General Assembly and the

Security Council jointly agreed to set up a Peacebuilding Commission.

In the words of Assembly President Jan Eliasson (Sweden) this would

mean that ‘‘post-conflict does not mean post-engagement of the inter-

national community.’’ The task of this commission was to keep coun-

tries on the ‘‘verge of lapsing or relapsing into conflict’’ on the agenda of

UN action (UN DPI, December 27, 2005). This is the first organiza-

tional expression of the need for sustained efforts in peacebuilding,

including the promotion of democracy. It testifies to a general recogni-

tion of the exit problem, as a problem of premature, or too early,

departure by the international community. During its first year of

operation, the new commission selected two situations to work with,

in agreement with the governments: Burundi and Sierra Leone. Both

countries are operating within a democratic framework, and refining

this is one of the ambitions of this new cooperation.

The temporal dilemma points to the issue of commitment. Clearly,

democracy building requires long-term efforts, but management of

crises requires short-term action. What might be emerging is an aware-

ness of the need for long-term international commitment. This, how-

ever, will also require the creation of new institutions or other durable

arrangements (bilateral agreements, regional organs, development

cooperation, etc). It cannot solely rest on the type of early reactions

230 Peter Wallensteen

that we see when triggering events take place. Short-term actions, often

negative in nature (sanctions of different sorts), ultimately yield in

favor of more lasting measures. Still, when facing a crisis, the dilemma

is activated. There is a need for short-term reaction, without at the

same time jeopardizing long-term investment. Skillful political leaders

can exploit this dilemma to their advantage. After all, Museveni was

re-elected and the outcome accepted. The donors’ choice between

demonstrating dissatisfaction and jeopardizing long-term gains was

balanced in the direction of a sustained commitment. The dilemma

remains, however.

International reactions: by which means?

There is an additional dilemma to consider: the vertical dilemma,

which posits legitimacy versus efficacy: Which are the measures used

by the international community to react to a democratization crisis?

Are they legitimate, in the eyes of that community as well as to the

inhabitants of the country exposed to the actions, and are they also

effective in bringing about the changes expected?

Reacting to a democratization crisis involves delicate issues of legiti-

macy, as it concerns the sovereignty of the targeted country. For the

countries contemplating or carrying out action it has to do with their

investment in the situation. There may have been a long-term commit-

ment, for instance, through a peace process during the war period.

Donors have supported the post-war developments. Thus, to them it is

legitimate to take action. It might be different on the other side of the

divide: government and population in the targeted country. The gov-

ernment is likely to be suspicious; the population may be more divided.

This affects the means chosen. The issues of governance (including a

democratization crisis) are within the sovereignty of a member state of

the international community (be it the UN or a regional body). This

may make other member states hesitant to act. The post-war conditions

add to this. Humanitarian concerns will enter into the decision making

as well: aid might be of critical significance for exposed groups. This

differs from the typical Cold War era or to reactions to military coups.

A typical reaction to military coups has been not to allow the particular

regime a seat in the UN. The suspension of diplomatic recognition (or

not extending relations to a new regime) was a common measure.

There were cases of open or clandestine military interventions to

International responses to crises of democratization 231

impact on the domestic dynamics, either favoring democratization or

protecting democratic gains (Western support to armed opposition in

Afghanistan against the Soviet Union may be an example of the former,

the US intervention in Panama in 1989 an example of the latter). These

measures are still available, as seen in the case of Iraq in 2003, but the

resort to actions which aim at the removal of a regime or the exclusion

of an entire country have been limited to a few situations. The post-

Cold War era has resulted in a need for refined ways of reacting. This

points to ways in which the vertical dilemma is handled: finding ways

which are at the same time legitimate and effective. A solution has been

to find targeted sanctions, that focus on the responsible actors rather

than on the population; or affect the revenue of the government in

selected areas (Wallensteen and Staibano 2005).

Some examples may help us to pursue the discussion. After about

fifteen years of post-war transition Lebanon could be regarded as

having recovered as a functioning society, albeit under heavy Syrian

tutelage. There were international contributions to reconstruct the

country’s democratic power-sharing system. Financial institutions

were again returning, trade relations were beginning to reestablish

themselves. There are many lessons to be learned from this, in terms

of long-term peacebuilding with a democratic framework in mind.

With the assassinations of the former Lebanese Prime Minister Rafik

Hariri and twenty-two others on February 14, 2005, however, the

country was again thrown back into severe crisis. This time, the inter-

national reaction was immediate, and the murder entered the agenda of

the UN Security Council. A special prosecutor was appointed and a

report was delivered, pointing to the involvement of Syria in the plot.

The newly elected parliament was decidedly more critical of Syria’s

role in the country. The conditions did not remain calm, however.

Throughout 2005 there were car bombs and political leaders critical

of Syria were killed. Once the prosecutor’s report was delivered, a new

crisis threatened. By the end of October the Security Council sharpened

its attitude, imposing sanctions on those individuals that were men-

tioned in the report (e.g., travel bans and freezing of assets) in order to

ensure that the legal process would be maintained. In Syria this was

perceived to be part of a strategy of isolating the country. The issue of

the democratic process in Lebanon began to take on new dimensions: it

threatened to become a matter of dealing with state-sponsored terror-

ism and the regime in Syria itself, thus connecting to the entire web of

232 Peter Wallensteen

Middle East politics. Interestingly, the Security Council maintained its

uniform position.2 Targeted sanctions as well as the UN investigation

that focused on individuals rather than on Syria’s involvement were

instrumental in this process. The approach, in other words, was to

target particular actors that could potentially undermine the demo-

cratic gains in Lebanon. The government welcomed the international

engagement.

Uganda provides a different story. In this case wars and repression

also lasted for about fifteen years and the reconstruction took an

equally long time. As we already observed, the international commu-

nity embraced Museveni and for many years applauded his regime. The

democratization crisis in 2005, thus, was a novel experience for a

leader that had so far enjoyed good reviews internationally. Also for

that community it was difficult to reverse its position. The arrest of the

opposition leader was, indeed, unprovoked and became a formative

event for international reactions. This was, nevertheless, less severe

than the killings that were witnessed in Lebanon. However, there was

no expectation of violence or human rights abuses. The fear of the

elections of 2006 not meeting international standards may have trig-

gered reaction more than anything else. The responses were different:

criticism by some states at the Commonwealth summit meeting in

Malta in November, the reduction of direct budgetary aid by signifi-

cant donors. The resulting reaction of Museveni was not one of imme-

diately releasing the leading opponent but instead he argued that these

responses showed that ‘‘aid is arrogantly mixed up with an effort to

interfere with our sovereignty’’ (Museveni 2005). Museveni, who had

long been a benefactor of international support, now tried to define

such support as colonial, and himself as a man of independence. The

donors that reacted thus had made the arrest of the opposition leaders a

major public event. The government resisted international engage-

ment, but in the end the opponent was released while other forms of

harassment continued. Through international attention and limited

actions, the international community may thus have achieved some of

its goals, and certainly made the elections more closely watched.

2 Obviously the war in 2006 between Israel and Hizbollah challenged many ofthese democracy gains, but the Lebanese government still remained in office inearly 2007.

International responses to crises of democratization 233

The sanctions approach in the case of Lebanon–Syria was targeted at

particular individuals, thus seemingly freeing the Syrian regime itself.

The regime could presumably deal with this by forcing these indivi-

duals to cooperate. By doing this, the regime might prevent itself from

being the next target. The implicit threat of escalating actions by the

international community may have had an impact. The actions on

Uganda, however, covered budgetary support that went straight into

the treasury of the government. The impact would affect Uganda’s

ability to conduct national policy. Whether the sums blocked actually

were large enough for such an effect can be debated. The demands built

on previous bilateral agreements but may have been somewhat diffuse.

It held out the possibility that the release of the arrested opponents

would restore budgetary aid, at some point. At the same time there was

an implicit threat of escalation of donor measures.

The demands of the international community and the actions taken

are, thus, two considerations which are intrinsically linked. The mea-

sures chosen are likely to be proportional, however difficult it may be

to determine that balance. In addition, there will be a need for inter-

national agreement on the actions. In the Syria–Lebanon case measures

has been taken by the Security Council, where such coordination is

routine. The actions on Uganda lacked such a forum. The options

available, furthermore, would not be global in reach. For instance,

donor countries coordinate through OECD/DAC or through EU,

both organizations covering about one eighth of all states. Uganda

might claim an ability to replace aid lost from some donors with aid

from other countries (although China showed no such record). The

sums involved might not have been prohibitive either. The concerted

actions of the Security Council, in other words, would have advantages

of more precise demands, targeted action, and international agreement.

Actions by other actors would suffer from having too wide demands,

difficulties in targeting, and difficulties in achieving international

(near-) consensus. One might say that the choice of action in both

cases, targeted sanctions, was a solution to the vertical dilemma.

They had legitimacy (being part of the UN, or being undertaken by

democratic states), they avoided affecting ‘‘ordinary’’ citizens (thus not

being punitive), and they were focused on particular individuals or

capacities of the government (travel, financial assets, budgets, respec-

tively). A course of action was chosen trying to combine legitimacy

with efficacy, thus suggesting a way to manage the vertical dilemma. In

234 Peter Wallensteen

the case of Syria–Lebanon some cooperation was achieved; in the case

of Uganda, the actions did possibly modify the incumbent’s policy.

These two cases demonstrate traditional versions of sanctions: by

breaking off certain types of relations, and inflicting some pain, there is

an expectation of change. There are some new options developing in

international practice, however, pointing to a second category: positive

sanctions (Wallensteen and Staibano 2005: ch. 15). Such sanctions

have also been discussed in terms of incentives or rewards (Dorussen

2001; Rothchild 1997). A case not mentioned so far in this contribu-

tion is instructive: Liberia. The most recent civil war ended in 2003, by

the removal of the incumbent, Charles Taylor (he fled to Nigeria, from

where he was extradited to the court in Sierra Leone in 2006, although

that court was based in the Netherlands). An interim government was

installed. UN peacekeepers were sent to the country. UN sanctions had

a role in achieving this change. By preventing the exports of diamonds

and timber from Liberia, one of the main sources for pursuing the civil

war was blocked, particularly for Taylor. In addition, Taylor and some

associates were exposed to financial and travel sanctions. When the

war ended there was a general expectation that the sanctions would be

lifted. The Security Council did not do so, however. Instead sanctions

were maintained and were regarded as a way of supporting post-war

reconstruction and democracy. By maintaining sanctions, no faction in

Liberia could restart war using these sources of income. The Security

Council wanted to make sure that a responsible government would be

in place and that these resources would be used for the benefit of the

entire economy of the country. Thus, Liberia sustained itself on huma-

nitarian support. The presidential elections of 2005 resulted in a run-

off of two candidates. The Security Council held out the possibility that

the winner, who in the end was Mrs. Johnson Sirleaf, would benefit

from the lifting of sanctions, once the government was able to control

the resources in question. A carrot was in place. The approach, never-

theless, suggested a development in the practice of international reac-

tions. During 2006 sanctions were eased, contributing to increased

governmental revenue by early 2007.

Sanctions thus have legitimacy (being a measure in the UN Charter),

and the efficacy of this measure has gradually improved, no longer

being comprehensive and general but targeted and with positive

options. There has been some learning from the negative humanitarian

effects in the case of Iraq during the 1990s. Practice has improved. The

International responses to crises of democratization 235

vertical dilemma expresses the problem well and the international

community has seemingly understood how to deal with this for further-

ing democracy. Still, the record of actually achieving democratizing

effects is in doubt. Lebanon remains in a fragile state, Uganda may see

increasing authoritarianism, while Liberia may have a historical

chance. International responses to democratization crises are often

too late to be effective.

Conclusions

In this chapter the experiences in international post-war democratiza-

tion crisis management have been exposed. There has been a strong,

global tendency for the international community to engage in the

internal affairs of member states more often. This seems like a logical

consequence of the fact that most of the wars since the end of the Cold

War have concerned intrastate matters. In such situations, the creation

or re-creation of democratic forms of governments have become cen-

tral. As this chapter has illustrated, this situation generates a number of

delicate dilemmas, and I have suggested ways in which they have been

dealt with by the international community.

First, there is a systemic dilemma: whether to react or not. A hypoth-

esis has been suggested: local threats to the electoral process in situations

where the international community has already been involved seem to

trigger reaction without involving serious sovereignty problems. Also,

there are particular moments in the process that draw attention: physical

threats or actions against leading opponents (democratic proponents),

followed by the manipulation of outcomes of such elections. In many

other instances of threats to the electoral process, the international

community is not likely to act strongly. Instead, threats to constitutional

processes, the undermining of the rule of law and freedom of speech are

often referred to non-public diplomatic dialogues. These issues may, on

occasion, enter into bilateral cooperation agreements.

Second, the international organizations face an obvious temporal

dilemma: if one is to react, when is that to be done (early or late?) and

for how long is action to be maintained? There is a tendency to crisis

reaction, meaning that it might be too late to be effective (as earlier

signs have been ignored), at the same time that there is an urge to

maintain long-term relations (thus potentially reducing the impact of

the short-term measures).

236 Peter Wallensteen

Third, there are also key elements of the vertical dilemma: by what

means, that is, which are the most effective ones that can legitimately

be used? This chapter points to the resort to targeted measures (differ-

ent forms of sanctions) as a way of mitigating the dilemma. These are

measures supported by international agreements; at the same time they

are not harming the population at large or the fragile democratization

process. The targeting in itself is a problem that has to be pursued: one

may suggest that if it is the opponents to democratization that are

targeted, there may be more effectiveness.

Fourth, the chapter has illustrated the horizontal dilemma through-

out the text. Central for international bodies and for international

actions is the ability to muster international support (for instance,

from the UN Security Council). This might be highly generic to inter-

national action. When it is not possible to get broad-based interna-

tional backing, actors have pursued other ways, for instance regional

organizations or donor meetings. In such, more limited settings, the

agreement on democracy promotion may be higher. At the same time

actions may be undermined by other actors, who step in to prop up the

target. Thus, actions will be limited and, possibly, short-lived (as illu-

strated in the case of Uganda). Sustained, effective, legitimate action, in

other words, will have to be based on international support (e.g.,

through the UN, as in Lebanon and Liberia).

Democratization crises are likely to activate all these dilemmas, more

or less at the same time. This may often make it difficult to predict

which actions will be taken when the international community faces a

problem. Policymakers looking for an appropriate mixture of action

will more or less instantaneously consider whether there is a suffi-

ciently triggering event, what options are available, what international

support can be marshaled, what would be effective, and for how long

action can be maintained.

The options available are key. Recent experience has stimulated new

thinking. One is the use of targeted sanctions. There is a range of

options in the mandates and composition of peacekeeping operations.

There is also the creation of the UN Peacebuilding Commission, con-

centrating on the post-war period of situations on the UN agenda.

There is the formation of more positions as Special Representatives

(by the UN, EU, ASEAN, AU, etc.) with varying assignments and

resources. An example is the appointment in 2004 of a special advisor

of the UN Secretary-General on the prevention of genocide, which also

International responses to crises of democratization 237

has to deal specifically with post-war situations as they have a higher

likelihood of genocidal actions. There is now a similar position on

conflict prevention. Remarkably, however, there is not yet a Special

UN Envoy for democracy building.

Similarly, the re-creation of the UN’s work on human rights, with the

new Human Rights Council in 2006, indicates a wider support of

concerted action, where also democratization crises may receive earlier

attention. A Democracy Fund has been established within the UN, and

international organizations demonstrate an increased willingness to

entertain inputs from civil society organizations. Together, these inno-

vations may signal new instruments for quicker action as well as a long-

term readiness for the international community to remain engaged, for

the benefit of democracy and peace.

238 Peter Wallensteen

9 Peacebuilding as democratization:findings and recommendations

T I M O T H Y D . S I S K

Peacebuilding, as a leitmotif concept guiding policy strategy for inter-

national intervention in war-torn societies that have negotiated a set-

tlement, has matured considerably since the term first entered the

lexicon of the international community in the immediate years after

the Cold War’s end. Peacebuilding evolved dramatically from its first

articulation in United Nations Secretary-General Boutros-Boutros

Ghali’s Agenda for Peace in 1992 to the landmark creation of a

permanent institution at the UN, the Peacebuilding Commission in

2006, to advance the aim of preventing the recurrence of war.

Likewise, policy practice has equally evolved from the initial experi-

ences of ‘‘complex, multidimensional peace operations’’ in Namibia

and Cambodia, for example, to the challenge of building states in

intractable conflicts such as Afghanistan or Liberia. Common to the

peacebuilding experiences of this era and today is the reliance on a

principal formula for durable war termination: the introduction of

democracy as a way to move the theatre of conflict off the battlefield

and into the institutions and processes of politics.

As the authors in this book forcefully show, this formula for durable

peace is premised on a fundamental contradiction that presents policy

practitioners with a serious set of dilemmas: pursuit of democracy can

undermine efforts to secure peace, and efforts to secure peace can

undermine the meaning and quality of democracy. Thus, in practice,

the promotion of democracy and the pursuit of peace can work at cross

purposes.1

1 Moreover, the conceptualization of linkages between peacebuilding anddemocratization have also been clouded by the intervention of the US-ledcoalition in Iraq in 2003, which sought to unilaterally impose a process ofdemocratization after the fall of Saddam Hussein’s regime; this democratizationprocess unfolded with escalating violence such that the military occupation anddemocratization efforts have in effect precipitated a failed state environment in

239

In transitions to democracy following deadly wars, international

mediators and protagonists alike face trade-offs that cannot be easily

reconciled or avoided. Indeed, at times choices must be made between

competing priorities at different moments in difficult processes;

usually, and understandably, international policymakers choose

peace over democratization and human rights because stopping the

killing is a higher-order priority than seeking justice or promoting

multi-party electoral competition (as Anna K. Jarstad points out in her

conceptual chapter in this volume). It is a common mantra among

policymakers that democratization is unlikely to proceed until peace is

achieved (an assumption only partially supported by this research).

The wide-ranging challenges of war-to-democracy transitions

explored in this volume reinforce the realization that the international

community must have sober expectations about the partial and

restricted nature of outcomes in post-war transitions: neither the peace-

making nor the democratization aims are likely to be fully realized in

transitional processes.2 The finding that policymakers often choose

pursuing peace over democracy and justice should not lead to pessi-

mism, for the research in this volume also shows that there are situa-

tions when innovation and careful planning can surmount the

dilemmas such that peace and democratization can progress more or

less simultaneously. The central task, then, is not whether democrati-

zation should or should not occur, whether democratization should

wait for peace or vice versa, or even how long democracy processes

such as elections should be delayed (a decision that will in any event

vary widely according to circumstance). Confronting the deep dilemma

between conflict management and democratization involves designing

ways in which the conflict-inducing nature of transitional processes can

be mitigated such that the initial constraints upon democratization that

arise from peace imperatives can, over time, fall away as trust and

legitimacy ostensibly build in the post-war period.

This concluding chapter evaluates the summary findings of systematic

evaluation of dilemmas in war-to-democracy transitions and explores

Iraq and levels of internecine conflict that have the characteristics of civil war. Foran analysis, see Diamond (2005).

2 Thomas Carothers does a good job of highlighting the partial nature of transitionaloutcomes in his piece, ‘‘The End of the Transition Paradigm,’’ reproduced in hisbook (2004), together with rejoinders from other prominent analysts.

240 Timothy D. Sisk

the implications of these findings for international policymakers. The

chapter synthesizes the ways in which each type of dilemma – horizontal,

vertical, systemic, and temporal – is experienced in the issue areas

explored by the authors. The chapter also seeks to make the link between

research findings and policy implications. The need for more effective,

theory-informed strategic directions to policy is especially acute at a time

when there is significant reform of international institutions mandated

to realize international peace, such as the Peacebuilding Commission

and a United Nations Democracy Fund (also launched in 2006).3 For

policy formulation and programming by these potentially promising

new entities, and in the decision-making processes of the Security

Council, understanding democratization dilemmas is the first step to

improving the capacity to confront or avoid them.

The policy implications of this research are organized into four

categories: designing transitions from war to democracy, managing

political violence, engaging civil society, and improving international

stewardship. The chapter ends with a call for further evolution in the

conceptualization of peacebuilding today and in particular ways to

redress the myopic approach that plagues present efforts to create

sustainable peace through democratization. As Peter Wallensteen

observes in his chapter, crisis management by international actors

such as the UN needs to have a clearer appreciation of the fundamental

dilemma between external supply of, and internal demand for, democ-

racy and greater ability to respond earlier to crises of democratization

in war-torn societies when they occur . . . often years after a war-ending

settlement has been reached.

Findings

The contributions of this volume reveal how the four types of dilemmas

articulated in Jarstad’s theory chapter are experienced by international

conciliators and protagonists in contemporary efforts to settle civil

wars through the democratization formula in a critical set of issue

areas. What synthesis findings emerge from the analysis?

The first finding is that often the conundrums of war-to-democracy

transitions are addressed not once a peace agreement is signed, but

3 Analysis of the UN’s role in democracy promotion is found in Newman andRich (2004).

Peacebuilding as democratization 241

in the course of negotiating settlements themselves. As I. William

Zartman has noted (1991), in the evaluation of a phase of peace process –

the post-agreement phase in this instance – one must look to antecedent

decisions made in prior moments of bargaining. Thus, tough choices are

often made in the negotiation of peace agreements that affect the post-

settlement phase of implementation. Among the critical issues that are

addressed in early phases of bargaining are the following:

� The nature and relative balance of power of various protagonists in

the war, to include the nature of elites who organize and wield power

(e.g., nationalist, ideological, or economic elites), the nature and

capabilities of military factions (e.g., government forces, rebel

groups, private militias);

� The possible types of sequences for setting up a transition – to

include protagonist perspectives on interim, transitional regimes,

on constitutional or final status questions, and in the imagination

of how various choices among political institutions such as elections

(such as electoral system choice) or territorial matters (federalism,

autonomy, division, or unitary states) affect the interests of the

protagonists in terms of their primary objectives;

� The economic base of the society and in particular the political

economy of the war – over issues of absolute scarcity as in Somalia

or over ‘‘lootable goods’’ in Sierra Leone or Angola – to include the

international dimensions of economic support for protagonists (as in

diaspora support for Sri Lanka’s rebels) or in economic dependencies

in terms of trade and investment (such as Angola’s oil revenues for

the government);

� The social divisions that exist to include the structure and nature of

civil society (e.g., to change the regime, to be included in power, to

advocate for minority interests, or to divide the state through seces-

sion); and

� The nature, extent, and resources of external involvement by inter-

national organizations such as the UN, regional organizations, pivo-

tal global or regional states, non-governmental organizations, and

eminent persons.4

4 The contributors to this volume are indebted to Pauline Baker, who made theseobservations at the 2006 Vail Symposium, ‘‘Dilemmas of Democratization inWar-Torn Societies,’’ at which the book’s chapters were presented, together withTom Farer who emphasized the economic dimension.

242 Timothy D. Sisk

Understanding and addressing the trade-offs of the post-settlement

phase thus find their origins in the difficult issues of negotiating peace

settlements, where some of the early choices – such as who is included

and who is excluded in peace processes or democratization pacts – is

decided.

Second, the trade-offs evaluated here are not dichotomous choices in

most instances: there are shades of gray in ways in which the problems

have been handled. For example, the evaluation of post-Dayton Bosnia

is often cited as an instance in which the cause of an ideal, socially

integrative democracy – not based on the ethnic divisions over which

the war was fought – was sacrificed in the interests of peace. Yet post-

Dayton Bosnia has experienced elements of democracy, to include

reasonably free and fair elections, greater protection of human rights,

and – as Roberto Belloni shows – resuscitation of an integrated civil

society. Similarly, the presence of peace in some post-settlement situa-

tions is also not an ideal that has been fully achieved, as Kristine

Hoglund’s analysis of post-settlement violence vividly demonstrates.

Thus, most war-torn societies find themselves somewhere in between

war and peace and war and democracy.

The answer to this conceptual gray zone – which afflicts scholarship

and policymaking alike – is to conceptualize war-to-democracy transi-

tions not just in terms of static regime types, but as dynamic situations

that can experience progress and regression in the consolidation of

post-war peace through democratization processes.

Horizontal dilemmas

The most pervasive of these dilemmas, readily seen in the foregoing

analysis in this volume, is one of inclusion and exclusion: horizontal

dilemmas. This dilemma is very much about inclusion or exclusion of

‘‘spoilers,’’ but it goes well beyond that principal concern, for example,

on how electoral processes affect the inclusion or exclusion in govern-

ing coalitions, or the involvement of civil society in negotiation pro-

cesses. In horizontal dilemmas, the critical question revolves around

the relationships among the parties in civil war: the roles of states

(political elites and military forces), rebel forces, and other factions

with the capacity to engage in violence such as militias. Much has to do

with these parties’ ideological or cause-of-conflict position (e.g., eth-

nonationalist, religious fundamentalist, or perhaps Marxist/Maoist

Peacebuilding as democratization 243

orientation), their bargaining dynamics, and their worthiness as parties

to agreements of the social contract type reflected in their conduct

during the war and ongoing into the peace. It may also occur in

considerations of their behavior during the war: should amnesty be

granted to all former combatants, including those who committed

gross violations of human rights?

Likewise, much depends on parties’ resource endowments and the

networks of political economy they use to fund their capacity for

violence. In the chapter by Hoglund on violence, for example, we

learn that efforts to create strict preconditions for inclusion (such as

linking Sinn Fein participation in Northern Ireland’s political process

to abandoning the use of violence by the IRA) may induce parties to

evade the criteria by shifting tactics from one form of violence (punish-

ment killings) to another form (punishment beatings). Efforts to

minimize violence by bringing in those who bear responsibility for

killing – such as rebel forces, splinter factions, or ‘‘rogue’’ elements of

the security forces – raises concerns of appeasement, of the shelving of

human rights concerns, or in some instances (such as Afghanistan

or Liberia) the creation of a ‘‘warlord’’ democracy (Wantchekon 2004).

Benjamin Reilly’s contribution carefully shows how electoral pro-

cesses not only define the nature of governing coalitions and the valida-

tion and legitimacy of negotiated settlements, but that electoral

processes fundamentally shape the nature of representation in the

war-torn societies context. Whether underlying social forces are

included or excluded – and how that representation actually takes

place – is often a function of the critical electoral system choice. In

this regard, electoral processes present clear trade-offs between desir-

able features of democracy that elections can provide: proportionality

of representation by group or political party and accountability of

individual candidates, between inclusion of representatives of a wide

swath of society and the creation of cohesive, effective governing

coalitions, and between the realities of identity-based voting behavior

and policies designed to induce moderation in societies emerging from

civil war. Innovative engineering to ameliorate these tensions, while

worthy to pursue, is limited by the realities of context (Reilly 2001).

It is for this reason, in part, that Jarstad’s observations about the

need in some instances to include representation quotas for some

groups in power-sharing pacts needs to be heeded. At the same time,

there is the equally important consideration that inclusion and

244 Timothy D. Sisk

exclusion is not just about warring parties: in most situations of war-

torn societies, there are large numbers of internally displaced or refu-

gees whose interests may not be well represented by the organized

factions at the negotiating table or in an interim government; devising

ways for participation of these excluded interests in electoral processes,

for example, is an urgent task (Lacy 2004).

Vertical dilemmas

Vertical dilemmas, these authors show, arise in relation to what can be

accurately described as the elite–mass nexus. Here, in issues of voter

participation or in engagement of civil society, dilemmas arise between

efficacy and efficiency and the equally desirable pursuit of legitimacy.

Elites make the deals, but peace and democratization – to be sustain-

able – need to involve the people. Whether in constitution-making

processes that yield power-sharing agreements, or in an electoral pro-

cess in which people formally participate in the war-to-democracy

transition, a reference to the people in peace processes is costly and

potentially subjects the carefully balanced elite pacts to ‘‘disruption

from below.’’ Particularly, the research here shows that mid-level elites,

such as rank-and-file of state military or rebel forces, or mid-level

political entrepreneurs, are the under-appreciated actors in vertical

dilemmas. The broad public is often too devastated from war –

displaced or dejected – to have the capacity for extensive participation

in the war-to-democracy transition process. Indeed, their attitudes

during the war may well have hardened, leading to maximalist claims

that cannot be reconciled with the imperatives of peace.

Vertical dilemmas are all about the practical difficulties in pursuing a

strategy of conflict transformation, in which all elements of society –

from elites to grass roots – are involved in the peacemaking process;

conflict transformation is so difficult to implement precisely because

the causes of conflict have deep structural roots that feature ‘‘horizon-

tal inequalities.’’ When wealth and control over resources overlaps

with other identities, for conflict transformation social forces will

experience a realignment that can be highly conflict-inducing.5 Thus,

the project of state-building is equally about creating the means for

5 John Paul Lederach (1997), argues that three levels need to be involved to achieveethnic reconciliation. At the top level, political and military leaders are involved in

Peacebuilding as democratization 245

fairer distribution of resources and creating the capacity of the state to

deliver services equitably, provide educational, health, and public sec-

tor opportunities, improve policing and public safety, and create the

conditions for economic prosperity.6

Persistent human rights abuses and recourse to the use of force by

governments and rebel factions undermine the impetus to peace gener-

ated by a negotiated settlement to civil wars. Electoral violence, assas-

sinations, and public riots are all indicative that societies torn by war

are prone to renewed and/or ongoing violence, the origins of which are

found deep in society. Likewise, when such instances of transitional

violence do occur, often it is the UN peace operation that must respond:

generating the perception, and often the reality – as in the Democratic

Republic of Congo – in which the UN becomes another party to a

complex, ongoing violent encounter.

The transformation of rebel forces is an essential part of peacebuild-

ing but with potential negative consequences for democratization

unless the parties fully give up the armed struggle and wage solely

peaceful campaigns for power. As Mimmi Soderberg Kovacs points

out in her analysis, the warring groups must shed the identities of

mobilization used during the war and reinvent themselves along lines

that cross-cut divisive identities. This is illustrated by the more glorified

cases of the transition, such as South Africa’s ANC conversion from

guerilla force to political party, or the transformation of the republican

cause in Northern Ireland from the Irish Republican Army to the

political party Sinn Fein, and today’s cases of the makeover of Aceh’s

Free Aceh Movement (GAM) to a political entity. Likewise, there is a

clear political economy of security sector reform and rebel movement

demobilization: unless the resource-capture dynamics that allow for

the mobilization and operational effectiveness of warring militaries

high-level negotiations to achieve settlements between rivaling ethnic groups.At the middle level, economic, religious, and humanitarian leaders organizeproblem-solving workshops and peace committees, and provide training forconflict resolution. At the grass-roots level, local leaders and officials organizeneighborhood committees and workshops for prejudice reduction. For discussionof the political economy of horizontal inequalities, see Frances Stewart (2000).

6 See, the report of the Making States Work project (Chesterman, Ignatieff, andThakur 2004), a collaboration of the International Peace Academy, the CarrCenter for Human Rights Policy at Harvard University, and the United NationsUniversity; also, see the work of the Research Partnership on Post-WarStatebuilding at www.statebuilding.org.

246 Timothy D. Sisk

and rebels are addressed, it is unlikely that peace will be sustainable. At

the same time, demobilizing a security force without an appreciation of

the deep, personal, economic effects such a move will have on indivi-

duals – and thus their incentives for political or criminal violence – will

undermine the possibility of peace.

With regard to linking politics with the people, power-sharing pacts

constrain democracy horizontally, but they are also limited in the

vertical vein as well. As Jarstad notes in this volume, ‘‘In divided

societies power sharing freezes the conflict lines, and the parties do

not need to compete for votes among their former foes . . . Instead,

nationalist parties [in Bosnia and Herzegovina] prevail in elections.’’

She later writes that despite the finding that elections are ‘‘meaningless’’

in power-sharing situations, at times elections can make a difference in

power-sharing outcomes, citing the cases of Burundi and Cyprus. At

least one option for addressing the problems of elite domination of

power-sharing agreements is through evaluation of the ways in which

political leaders are linked to the elements of society they purport to

represent, either through political parties or civil society. Reilly’s

admonition in his chapter that large, integrative political parties are

best able to contribute to peace and to the sustainability of long-term

democratization is a critical finding on the ways in which the challenges

of popular participation can be potentially resolved, allowing transi-

tional power-sharing pacts to be less critical for peace and allowing for

elections to be more meaningfully fought without endangering the

peace.

As the work of Roberto Belloni shows, after war civil society’s

capacity to participate is deeply weakened at the same time that asso-

ciations and organizations are needed to transcend lines of conflict.

The key to sustainable peace in the long term is the integration of civil

society across lines of conflict, requiring, over time, a different config-

uration of civil society, social structures, and political institutions:

identity must cross-cut lines of conflict, leading to bargaining based

on class concerns rather than those that revolve around the negotiation

of competing identity.7 Belloni concludes that ‘‘An effective civil

society development strategy capable of overcoming the dilemmas of

7 South Africa emerges as a case in which civil society did manage to congeal acrosslines of historical conflict, greatly facilitating progress in the war-to-democracytransition; see Stremlau (1998).

Peacebuilding as democratization 247

democratization and peacebuilding requires . . . the ability to identify

and support inclusive, bridging, and pro-democracy organizations.’’

Systemic dilemmas

Systemic dilemmas relate to the very nature of contemporary internal,

civil wars that cause spillovers and contagions for the entire world:

invariably, these conflicts gain international attention and intervention

by external forces, from neighboring states that feel the direct and

tangible effects to globally powerful states that see ‘‘soft’’ threats ema-

nating from weak state environments. The international community is

compelled to intervene, wherever possible, to manage the international

effects of internal civil wars (Sisk 2001).

When external parties engage, as in a peace negotiation, they tacitly or

explicitly are involved in recognition of parties – or exclusion of others –

as recognized, legitimate parties to the conflict. When international

peacebuilders include a group at the negotiating table, they implicitly

provide legitimacy for these factions (Rothchild 1997). Likewise, in

including rebels at the table as legitimate actors for bargaining, they

also endorse the difficult transformations that such factions are intended

to make as political entities. Sometimes the international community

makes grave mistakes: as Soderberg Kovacs observes in her chapter,

recognition of the notoriously brutal RUF in Sierra Leone’s 1999 Lome

peace pact had the effect of legitimating the party – and providing them

with considerable power in a transitional pact – when in elections the

former bandits garnered only 1.7 percent of the popular vote.

Virginia Page Fortna’s contribution demonstrates that international

peace operations do provide the essential credible commitment sorely

needed in the volatile, post-war context. She writes that ‘‘Because

peacekeeping is clearly and unambiguously good for peace, countries

emerging from civil war are better off when peacekeepers deploy than

when they are left to their own devices.’’ But she also notes that in some

instances peacekeeping operations may actively inhibit some elements

of democratization. The finding is that more than transitional, short-

term peace operations are required: to reduce violence for example, it is

equally salutary to deploy large international electoral observation

missions as a way to redress the uncertainty dilemmas that plague

domestic protagonists as they contemplate loss, or envisage aggrand-

izement, in electoral processes (Bjornlund 2004).

248 Timothy D. Sisk

Temporal dilemmas

The research in this volume also points to one of the most engaging

temporal dilemmas that occur in war-to-democracy transitions. Deals

needed to achieve war termination and bring an end to violence may

require compromises that constrain democratization later. Jarstad’s

analysis of power sharing reveals that both protagonists and interna-

tional mediators face powerful temporal dilemmas when negotiating

peace agreements: she observes that the legitimacy of power sharing

can be questioned when such arrangements are imposed by third

parties. This suggests that, over time, they are prone to collapse unless

a way is found to revisit the terms of settlement gradually without

inducing some of the uncertainty dilemmas that have led to such

pacts in the first place. Pacts that guarantee representation for

warring parties run against the subsequent need in flourishing democ-

racy to have uncertainty about who wins and loses in electoral

competition.

Because violence often accompanies the transition from war to peace

into the post-settlement period, there are calls for the bloodletting to

end before democratization can proceed (especially, elections).

Security must come first. At the same time, there is often the reality

that the violence is instrumental to the uncertainty of the moment, and

that violence will only subside once the turbulence of the transition has

passed. Thus, the sequencing puzzle – the planning and timing of

peacebuilding (e.g., demobilization) and democratization (e.g., elec-

tions) is critical. As Hoglund concludes in her essay, ‘‘Political violence

is frequently a response to too little democracy.’’

Perhaps the most appreciated, and yet continually most poignant,

finding of this research is that often external agendas – often at the level

of the Security Council – push electoral timetables even when the

implication may be deleterious to the long-term needs of democratiza-

tion. Reilly notes in his contribution to this volume, that ‘‘Hasty or

rushed ‘instant elections’ have become common for several reasons.’’

‘‘Democracy, as the mantra goes, is a long-term process, but the

domestic political pressures that weigh on the Western states that

usually fund and implement peacekeeping missions are almost all

short-term . . . such ‘premature elections’ can also create multiple,

ongoing problems for the development of peacetime politics in deeply

divided societies even years after the war has ended.’’

Peacebuilding as democratization 249

Recommendations for peacebuilders-as-democratizers

The principal finding of this book is that, indeed, the goals of demo-

cratization and the demands of conflict management often work at

cross-purposes: dilemmas of democratization and peacebuilding are

ubiquitous in war-torn societies. Dilemmas of democratization arise

when political violence continues into the post-war era and indeed

escalates as elections approach; when rebels face the need to transform

their identities from military organizations to political parties; when

imperatives of justice and an end to impunity for human rights abuses

collide with peacemaking and conciliation; when power sharing is

accepted as an interim and, sometimes, permanent solution to con-

tested governance but impedes accountability and popular participa-

tion; and in strategies of international democracy building, with

trade-offs of short-term needs for transitional success and long-term

demands of state building.

Designing war-to-democracy transitions

The findings have policy implications in two important areas. First is in

the analysis or context-specific assessment of war-to-democracy transi-

tions, where there is the need for theoretically informed understanding of

a given particular context. Research on dilemmas of democratization in

relation to peacebuilding can help those evaluating a particular context

to identify the right variables, to ask the right questions, and to take

findings from comparative research into new settings. Here, there is a

need to fuse two common methodologies of analysis employed by the

international community to determine context appropriateness: democ-

racy assessments and conflict vulnerability assessments. The former

seeks to identify strengths and weaknesses in political processes, and

the latter seeks to identify vulnerability to violence.8

8 For such a comprehensive conflict analysis model, see the World Bank’s ConflictAnalysis Framework (available at lnweb18.worldbank.org/ESSD/sdvext.nsf/67ByDocName/ConflictAnalysis). Similarly, the World Bank’s governanceindicators are an accepted set of measures for evaluating democratic practice:www.worldbank.org/wbi/governance/govdata/. International IDEA’s Stateof Democracy project offers a comprehensive assessment guide, and bilateraldonors such as US Agency for International Development also have developedoff-the-shelf democracy and governance assessment tools.

250 Timothy D. Sisk

Assessment approaches can help address some of the most vexing

questions that the war-to-democracy process raises in specific contexts.

For example, the question of whether a faction or party can be

excluded or included involves a decision made – often by international

mediators such as a Special Representative of the Secretary-General –

that a party is redeemable as a partner for peace. Sometimes, as in

Uganda and the rebel Lord’s Resistance Army (LRA), the decision is

made that some factions simply cannot be included either because of

what they have done (in this case, gross violations of human rights,

especially against children) or what they believe (espousing an extre-

mist and indeed bizarre evangelical ideology in this instance). Even

there, despite an indictment against LRA leaders by the International

Criminal Court, the domestic imperatives for peace have led to a

negotiation process between the government and the rebel forces.

Context-based assessments are also critical for determining how a

transition could be designed to more effectively minimize trade-offs

between democracy and peace. The example of sequencing helps illus-

trate how assessments can improve policy strategies. A constant con-

cern of researchers and policymakers alike is how to evaluate a strategy

of sequenced change in outlining war-to-democracy transitions. The

sequence of events – interim agreements and arrangements, local and

national elections, constitution-making processes, referenda, and the

like – is critical to resolving some of the challenges that peacebuilders-

as-democratizers face. For instance, in some situations it may make

sense to have local elections first, as in Kosovo, and then follow with

national-level polls to reflect a gradual, bottom-up approach to demo-

cratization that does not beg the question of national-level political

authority but does provide some sense of legitimacy for governance. In

other situations, however, the need is for consolidation of national-

level authority such that relatively early national elections – as in

Afghanistan – can begin to create an authoritative state to deal with

local power centers that are injurious to democratization aims.9

Likewise, there is also good analysis available on how to sequence

concerns regarding transitional justice on the route to peace.10 The

variety of sequences addressed in this book – Northern Ireland, Bosnia,

9 For evaluation of local-level processes in United Nations peacebuilding, seeRisley and Sisk (2005).

10 For a thorough evaluation, see Mani (2002).

Peacebuilding as democratization 251

South Africa, Sri Lanka, Kosovo, and others reinforces the principal

finding of sequencing: there is no single or ideal sequence for a war-

to-democracy transition, and in this domain the primacy of analysis

and decision making rests on the uniqueness of context.

Managing political violence

The challenge of managing political violence is acute precisely because

democratization processes can stimulate new bloody struggles for

power: elections lead to mobilization, media openness can lead to the

dissemination of hate speech, security sector reform is often inade-

quate, and rebels often are insufficiently transformed in their meta-

morphosis to becoming a political party. Democratization can,

however, resolve problems of violence through the inclusion of the

formerly excluded into political power, suggesting that with sufficient

commitment, time, and resources the problem of violence can be

managed so that peace can be self-sustaining.

The challenge for external peacebuilders is to try to turn degenera-

tive cycles of violence into generative cycles of trust, tolerance, and a

willingness to play the game of democracy non-violently. Critical to

this aim is the creation of credible commitment to peace agreements,

primarily by international peacekeeping forces but also through the

well-considered, long-term engagement of the international commu-

nity. Likewise, when parties in such situations do use violence to pursue

politics by non-democratic means, the international community must

be willing and able to respond quickly and effectively, both in terms

of military responses and in the willingness to punish (through de-

legitimation, sanctions, or other ‘‘negative inducements’’) violators of

agreements.

As Fortna demonstrates, too often peacekeeping forces are an insuf-

ficiently powerful tool to dissuade parties from turning to violence or

to address spoiling when it occurs. This finding suggests that successful

peace operations need two important preconditions: first is the capa-

city to act, and to act forcefully and convincingly, to provide a mod-

icum of security on the ground. The second is to be deployed for much

longer periods of time, and not to have six-month mandates (a common

time frame for Security Council authorization of missions) for a multi-

year job. Unless the international community is seriously committed

to peace operations for the long term – and the creation of the Peace

252 Timothy D. Sisk

Support Fund in 2006 for this purpose is a hopeful first step – efforts to

manage transitional violence in peace operations will likely fall short,

jeopardizing peace-through-democratization aims. Other proposed

reforms, such as allowing the World Bank to fund peace support

operations or to support programs such as police force training

and reform (with a Security Council mandate and coordination with

the UN force on the ground) are promising and deserve serious

consideration.

Another critical challenge in this domain is the question of anticipat-

ing and predicting electoral violence. The research here points to key

indicators for understanding when a situation is predisposed toward

electoral violence occurring, and the measures can be put in place in a

preventive and post-electoral dispute resolution process to manage the

likelihood of violence. An especially robust set of indicators are the

residual coercive power of armed factions and the capacities of agen-

cies of restraint – civilian authorities, police and military forces, inter-

national peacekeepers – to limit these capacities. Another key indicator

to analyze is the possible outcome dimensions of electoral processes

themselves: if significant factions are likely to lose the power they

already have, or formerly excluded factions have no chance of mean-

ingful representation or influence, the likelihood of ‘‘spoiler violence’’

is high.

Designing institutions: beyond ‘‘once-and-for-all’’settlements

The second implication is for those who are actively engaged in

peacebuilding-through-democratization. A clearer understanding of

the dilemmas that arise for the parties with which they may be working,

and for themselves, can lead to more informed decision making and,

over time, more successful outcomes in war-to-democracy processes.

Peace agreements may contain features of a social contract, but it is not

a self-enforcing one: the key to the long-term success of the war-to-

democracy transition is the perpetuation of the bargaining process that

led to the settlement long into the democratic future.11 Such a future,

11 This is not a new finding. For an earlier analysis on the importance of theproliferation of bargaining institutions for the management of social conflict, seeRothchild (1973).

Peacebuilding as democratization 253

however, is not a quick one. Those who seek to promote war-to-

democracy transitions should expect that the outcome of their efforts

will be inherently partial, superficial, and in some ways deeply disap-

pointing as the dilemmas of democratization present themselves

forcefully.12

At least one solution to this problem is to build into peace settlements

a longer-term transition for war-torn societies, one that envisages not

just a single process of negotiation ending in a set of once-and-for-all

institutional arrangements. Instead, clever inclusion of future negotiat-

ing processes, of longer-term moments of decision, can help address

short-term needs for certainty on final status or on the immediate

question of territorial sovereignty with longer-term needs for flexibil-

ity. At least two practical measures can accomplish this aim. One is to

feature in the course of peace settlements ‘‘sunset’’ (or expiry) clauses

that restrict some of the more democracy-constraining elements of

peace agreements such as power sharing. Sunset clauses may sound

like a good way to resolve the temporal dilemma, but often parties in

conflict – particularly relative weaker parties politically (but not neces-

sarily militarily) – demand a permanent share of power. Likewise, some

parties such as governments are unwilling to agree to ‘‘down the road’’

provisions such as referendums on independence (even though the

January 2005 peace agreement for Sudan included a provision for a

referendum on independence for the South six years following).

The second is to encourage national dialogue processes on democ-

racy that can allow for supplementary consensus building to occur

outside of formal institutions; such dialogues have the benefit of creat-

ing consensus first on possible institutional or procedural reforms,

following which implementation of reforms can be less controversial.

Examples of such dialogues are found both at the national level (as in

South Africa’s highly participatory constitution-making process) and

at the local level, for example those involving mayors in Colombia.

Opening up political space for broader participation has been useful in

a number of transitional arrangements, for example in the appointment

of interim governments (Afghanistan, Lloya Jirga, and in Liberia,

12 Indeed, some scholars have argued that well beyond the initial peace agreementthere is the need for ‘‘post-settlement settlements’’ that review the terms of peacelong after the wounds of war have begun to heal and societies can withstandanew a period of political uncertainty that reopens issues over which the war wasinitially fought (du Toit 2003).

254 Timothy D. Sisk

Benin, Togo, and other West African national conferences such as in

Congo–Brazzaville and in Niger), in the drafting of constitutions

(Eritrea, Rwanda, Afghanistan, and more recently Nepal), and in

various elements of implementation of peace agreements as in

Guatemala (Papagianni 2006: 23–27).

Engaging civil society

One of the problems of international bilateral assistance in war-torn

societies has been that of ‘‘stove-piping’’ in which various elements of

donor agencies, or across different types of donors, agencies, and

elements of a coalition that intervenes, narrowly focus on a single

aspect of engagement of civil society without a more holistic picture

of how such assistance can be better provided to support peacebuilding

efforts more coherently. Assistance for health, education, or human

rights entities often occurs without consideration of the implications of

such assistance for conflict dynamics or for democratization objectives.

Conflict mitigation needs to be more fully mainstreamed into all

aspects of civil society promotion and governance processes such that

a more consistent, coherent approach to addressing the causes of

violence is employed.

Likewise, there needs to be greater appreciation of what civil society

can, and cannot, accomplish and more careful planning for how assis-

tance to non-governmental organizations, trade unions, business asso-

ciations, and religious entities can reinforce peacebuilding objectives.

Belloni points out in his chapter that not all civil society groups are the

same: some groups are virtuous and contribute to peace, but some civil

society groups can in fact frustrate durable peace. Strategically, the

promotion of civil society cannot occur in a platitudinous fashion that

sees all civil society as an inherent good for peace and democratization.

Quite the contrary, there needs to be a sharp strategy of differentiation

in civil society promotion by which international donors are quite

discriminating in identifying three types of non-state actors to support:

those that cross-cut identity lines or fissures of conflict (as highlighted

above), those that are moderate but reflecting primarily one perspective

or protagonist social group, and those that are more extreme but

which, through coaxing and inclusion, can become moderate (such

as ex-prisoners, as was the case in Northern Ireland). Likewise, effec-

tive international action requires identifying and working diligently

Peacebuilding as democratization 255

against those civil society groups that are deemed not constructive to

peacebuilding aims, either because of their irredeemably extreme nat-

ure and positions or because they have other interests or activities

(such as engagement in crime, resource exploitation, or activities such

as human trafficking) that work against progress toward peace or

democracy.

Rethinking international stewardship

The present approach to international stewardship of war-to-democracy

transitions aimed at bringing about lasting peace through demo-

cratization in societies shattered by war needs to be fundamentally

reevaluated. As noted in the introduction to this chapter, the present

approach has evolved in recent years but the fundamental formula

remains the fusion of peacekeeping, peacebuilding, and democratiza-

tion functions in a single, complex, multidimensional peace support

operation. This approach does make sense in the immediate response of

the international community to war-torn states that have managed to

clinch a cease-fire. Indeed, where such negotiation-backstopping

deployments of peacekeepers are not forthcoming for whatever reason,

as in Azerbaijan (Karabagh) or Sri Lanka (2002–2006), moving beyond

cease-fire to comprehensive settlement appears to be stifled, not

assisted.

Tough measures are needed to keep the peace and prevent violence,

development and economic restructuring require early intervention,

civil society (generally) needs to be included, and democratization –

to include electoral processes and constitution making – cannot be

indefinitely delayed. As the imbroglio in Iraq would reaffirm, success

in this realm appears to require the United Nations or at least its

imprimatur in the form of a Security Council mandate that enjoys a

broad consensus among leading states; as Fortna shows, the actual

composition of the force is less important that its response capacities,

and its legitimacy. However, despite years of experience and no dearth

of research, the current approach toward peacebuilding remains myo-

pic; the causes of such myopia are found in the limits of resources

(personnel, funds, and political will by leading states) for risky, long-

term engagements; this is especially true when new emergencies clamor

for attention and require new or diverted resources. Democratization

is the long-term solution for war-torn states, but it cannot be

256 Timothy D. Sisk

accomplished with a one-off, peacekeeping mission that ‘‘exits’’ when

the political will for the mission wanes (usually, as violence begins to

subside and transitional elections are over). Events in Angola, DRC,

Haiti, Liberia, and Timor Leste (formerly East Timor), which have

required repeated UN-mandated military interventions, reinforce this

now well-appreciated observation.

The fallacy of the regrettably still-current approach is the perception

that a deployed, bounded mission lasting at least a year but not more

than a decade can facilitate the process of war to democracy.13 The

present formula: intervene, establish order through blue-helmet force,

fund some civil society, facilitate a new constitution, hold an election . . .

and then leave – may suffice for short-term peace but does not suffice for

democratization.

To redress the fallacy of myopia, and the need to sustain support for

the long transition from war to democracy, international stewardship

of peacebuilding will need to evolve institutionally and in practice

much more significantly than it has so far. The war-to-democracy

transition demands a series of successive missions to address the

longer-term phases of consolidating peace, of building democratic

institutions and processes, of reacting to subsequent political crises

and power-grabs, of addressing the socioeconomic conditions that

give rise to violence, and, ultimately, the creation of a self-sustaining

social contract. Successive missions will require a seamless set of phases

of international involvement that may begin with heavy-handed

deployments but that changes over time into a deeper, generations-

long commitment to democratization. An example of where this has

been somewhat effective is the continued commitment by the interna-

tional community to Mozambique, which benefited from an initial

peace operation but has also been subsequently subject to extensive

monitoring, electoral assistance, donor support, and other governance-

improving engagement by the international community. Such extended

commitments to war-torn societies needs to be the norm, not the

exception.

13 The latter is quite unusual and found only in the post-Cold War era in theBalkans where European resources and NATO troop commitments account forthe longevity of commitment. Even there, initially, there was an ill-consideredeffort to impose a one-year constraint on NATO deployment.

Peacebuilding as democratization 257

Commensurate with such a longer-term view is the need for constant

monitoring and a willingness to intervene again early when progress

toward peace or democratization is threatened. As Wallensteen notes,

many of the crises in the cases he has evaluated – Uganda, Lebanon,

and Liberia – have come a decade or more after the peace agree-

ment was reached. Likewise, Zimbabwe is an example of a situation

in which the legacies of war and social conflict threatened peace and

democratization more than a decade after the war was ostensibly

settled, with devastating social, economic, and regional effects of the

slide into authoritarianism and state failure in that country.

International reaction in this case has been too partial, too restricted,

and too late.

This need for a rethinking of the present approach will require the

reinvigoration of instruments that exist but are insufficiently used –

such as extensive United Nations civilian observer missions and of

long-term UN civilian police deployments to be in place once military

security seems to be achieved. Indeed, there is the need to consider the

deployment of civilian missions who can readily observe when back-

sliding is occurring; increasingly, it is regional organizations (in

Europe, Africa, and the Americas, especially) that are effectively play-

ing this role. A new evolution of peacebuilding will also require new

instruments that have been considered but not realized, especially the

creation of a military rapid deployment force (such as an international

gendarme or rapid reaction force) to respond quickly when new crises

arise,14 new funding mechanisms for providing the resources required

for long-term engagements contexts far away from prosperous Western

states, and the will to react earlier when there are indicators – such as

political crises or the removal of independent judicial officials – that

peace or democratization is threatened anew.

Measures such as these are ideal but, as the challenges of responding

early to the Zimbabwe crisis or empowering Afghanistan’s nascent

democratic institutions, for example, show, are hampered by the

stark realities that international intervention remains limited by inter-

national consensus and domestic contexts. As a result, the outcomes of

14 This recommendation was included as a key finding of the Carnegie Commissionon Preventing Deadly Conflict but has not been embraced by the permanentmembers of the United Nations Security Council. See the Commission’s FinalReport for the rationale and specifics of this recommendation (CarnegieCommission on Preventing Deadly Conflict 1998).

258 Timothy D. Sisk

war-to-democracy transitions are likely to remain, for the foreseeable

future, inherently limited despite the best efforts of the international

community to confront and address innovatively the dilemmas such

transitions pose. More optimistically, however, some of the ways in

which the dilemmas are addressed are amenable to innovation and

implementation by the international community – as the deployment

of an all-female UN peacekeeping unit in Liberia in 2006 following the

election there of Africa’s first female president, Ellen Johnson Sirleaf.

This deployment pairs the need for a longer-term security commitment

with an implicit appreciation for human rights and conflict transfor-

mation. Likewise, the deployment of a significant security force and

large-scale election-assistance mission in the Democratic Republic of

Congo in July 2006 is also indicative of learning and adaptation in

facilitating war-to-democracy transitions, even as post-election vio-

lence has created new insecurities in 2007. Small and incremental

steps such as these do not vitiate altogether the dilemmas in war-torn

societies that the international community and internal protagonists

experience; however, even such small steps do indicate that innovation,

a stronger will and more forceful deployments, and an abiding commit-

ment to a longer time horizon can go some way toward minimizing the

deleterious impact that dilemmas of democratization impose on the

pursuit of peace.

Peacebuilding as democratization 259

References

Abbink, Jon, and Gerti Hesseling, eds. 1999. Election Observation and

Democratization in Africa. Basingstoke: Macmillan Press.

Andersson, Andreas. 2000. ‘‘Democracies and UN Peacekeeping Operations,

1990–1996.’’ International Peacekeeping 7 (2): 1–22.

Arato, Andrew. 1981. ‘‘Civil Society vs. the State: Poland 1980–81.’’ Telos

47: 23–47.

Archer, Dane, and Rosemary Gartner. 1976. ‘‘Violent Acts and Violent

Times: A Comparative Approach to Postwar Homicide Rates.’’

American Sociological Review 41 (6): 937–963.

Arzt, Donna E. 2006. ‘‘Views on the Ground: The Local Perception of

International Criminal Tribunals in the Former Yugoslavia and Sierra

Leone.’’ Annals of the American Academy of Political and Social

Sciences 603 (1): 226–239.

Atlas, Pierre M., and Roy Licklider. 1999. ‘‘Conflict Among Former Allies

after Civil War Settlement: Sudan, Zimbabwe, Chad and Lebanon.’’

Journal of Peace Research 36 (1): 35–54.

Baker, Pauline. 2001. ‘‘Conflict Resolution versus Democratic Governance:

Divergent Paths to Peace?’’ In Turbulent Peace: The Challenges of

Managing International Conflict, ed. Fen O. Hampson, Chester A.

Crocker and Pamela Aall. Washington, DC: United States Institute of

Peace Press (753–764).

Baldwin, Clive. 2006. Minority Rights in Kosovo under International Rule.

London: Minority Rights Group International.

Ballentine, Karen, and Jake Sherman, eds. 2003. The Political Economy of

Armed Conflict: Beyond Greed and Grievance. Boulder, Colo.: Lynne

Rienner Publishers.

Bangura, Yusuf. 2000. ‘‘Strategic Policy Failure and Governance in Sierra

Leone.’’ Journal of Modern African Studies 38 (4): 551–577.

Barnes, Catherine. 2002. Owning the Process: Mechanisms for Political

Participation of the Public in Peacemaking. London: Conciliation

Resources.

Barnett, Michael. 2006. ‘‘Building a Republican Peace.’’ International

Security 30 (4) (Spring): 87–112.

260

Barro, Robert J. 1999. ‘‘Determinants of Democracy.’’ Journal of Political

Economy 107: 158–183.

Bartolini, Stefano, and Peter Mair. 2001. ‘‘Challenges to Contemporary

Political Parties.’’ In Political Parties and Democracy, ed. Larry

Diamond and Richard Gunther. Baltimore: Johns Hopkins University

Press (327–343).

Belloni, Roberto. 2007. ‘‘Shades of Orange and Green: Civil Society and

the Peace Process in Northern Ireland.’’ Paper presented at the 48th

Annual Convention of the International Studies Association, Chicago,

February 27.

2004. ‘‘Peacebuilding and Consociational Electoral Engineering in Bosnia

and Herzegovina.’’ International Peacekeeping 11 (2): 334–353.

2001. ‘‘Civil Society and Peacebuilding in Bosnia and Herzegovina.’’

Journal of Peace Research 38 (2): 163–180.

Belloni, Roberto, and Shelley Deane. 2005. ‘‘From Belfast to Bosnia:

Piecemeal Peacemaking and the Role of Institutional Learning.’’ Civil

Wars 7 (3): 219–243.

Bentley, Kristina A., and Roger Southall. 2005. An African Peace Process:

Mandela, South Africa and Burundi. Cape Town: Human Sciences

Research Council.

Berdal, Mats, and David M. Malone, eds. 2000. Greed and Grievance:

Economic Agendas in Civil Wars. Boulder: Lynne Rienner Publishers.

Berman, Sheri. 1997. ‘‘Civil Society and the Collapse of the Weimar

Republic.’’ World Politics 49 (3): 401–429.

Bertram, Eva. 1995. ‘‘Reinventing Governments: The Promise and Perils of

United Nations Peace Building.’’ Journal of Conflict Resolution 39 (5):

387–418.

Bieber, Florian. 2003. ‘‘The Other Civil Society in Serbia: Non-

Governmental Nationalism – The Case of the Serbian Resistance

Movement.’’ In Uncivil Society? Contentious Politics in Post-

Communist Europe, ed. Petr Kopecky and Cass Muddes. London and

New York: Routledge (19–36).

Bjornlund, Eric. 2004. Beyond Free and Fair: Monitoring Elections and

Building Democracy. Baltimore, Md.: Johns Hopkins University

Press.

Bojicic-Dzelilovic, Vesna. 2006. ‘‘War Veterans’ Associations in Bosnia and

Herzegovina.’’ In Challenges to Peacebuilding: Managing Spoilers

During Conflict Resolution, ed. Edward Newman and Oliver

Richmond. Tokyo, New York, and Paris: United Nations University

Press (200–218).

Brass, Paul R. 1991. Ethnicity and Nationalism: Theory and Comparison.

New Delhi, Newsbury Park and London: Sage Publications.

References 261

Brown, Frederick Z. 1998. ‘‘Cambodia’s Rocky Venture in Democracy.’’ In

Postconflict Elections, Democratization & International Assistance, ed.

Krishna Kumar. Boulder, Colo. and London: Lynne Rienner Publishers

(87–110).

Brown, Michael E., Sean M. Lynn-Jones, and Steven E. Miller, eds. 1996.

Debating the Democratic Peace. Cambridge, Mass.: MIT Press.

Bueno de Mesquita, Bruce and George W. Downs. 2006. ‘‘Intervention and

Democracy.’’ International Organization 60 (3): 627–650.

Burkhart, Ross E., and Michael S. Lewis-Beck. 1994. ‘‘Comparative

Democracy: The Economic Development Thesis.’’ American Political

Science Review 88: 903–910.

Burnell, Peter. 2005. ‘‘Political Strategies of External Support for

Democratization.’’ Foreign Policy Analysis 1: 361–384.

Call, Charles T. 2003. ‘‘Democratisation, War and State-Building:

Constructing the Rule of Law in El Salvador.’’ Journal of Latin

American Studies 35: 827–862.

Call, Charles T., and Susan E. Cook. 2003. ‘‘On Democratization and

Peacebuilding.’’ Global Governance 9: 233–246.

Call, Charles T., and William Stanley. 2003. ‘‘Military and Police Reform

after Civil War.’’ In Contemporary Peacemaking: Conflict Violence and

Peace Processes, ed. J. Darby and R. Mac Ginty. Basingstoke: Palgrave

Macmillan (212–223).

Carnegie Commission on Preventing Deadly Conflict. 1998. Preventing

Deadly Conflict: Final Report of the Carnegie Commission on

Preventing Deadly Conflict. Washington, DC: Carnegie Commission

on Preventing Deadly Conflict. Available at wwics.si.edu/subsites/

ccpdc/pubs/rept97/finfr.htm.

Carothers, Thomas. 2004. Critical Mission: Essays on Democracy Promotion.

Washington, DC: Carnegie Endowment for International Peace.

1999. Aiding Democracy Abroad: The Learning Curve. Washington, DC:

Carnegie Endowment for International Peace.

Cawthra, Gavin, and Robin Luckham. 2003. Governing Insecurity:

Democratic Control of Military and Security Establishments in

Transitional Democracies. London: Zed Books.

Cerdas Cruz, Rodolfo. 1998. ‘‘Political Parties and Party Systems.’’ In

Guatemala after the Peace Accords, ed. Rachel Sieder. London:

Institute of Latin American Studies (15–54).

Chandler, David. 2004. ‘‘The Responsibility to Protect? Imposing the

‘Liberal Peace’.’’ International Peacekeeping 11 (1): 59–81.

1999. Bosnia. Faking Democracy after Dayton. London: Pluto Press.

Chandra, Kanchan. 2004. Why Ethnic Parties Succeed: Patronage and

Ethnic Headcounts in India. Cambridge: Cambridge University Press.

262 References

Chesterman, Simon. 2004. You, The People: The United Nations, Transition

Administration, and State-Building. Oxford: Oxford University Press.

Chesterman, Simon, Michael Ignatieff, and Ramesh Thakur. 2004. Making

States Work: From State Failure to State Building. Tokyo: United

Nations University Press.

Chesterman, Simon, Tom Farer, and Timothy Sisk. 2003. ‘‘Competing Claims:

Self-Determination, Security, and the United Nations.’’ A report of the

project on ‘‘Self-Determination, Security, and the United Nations,’’

Graduate School of International Studies, University of Denver, and the

International Peace Academy. Available at www.ipacademy.org/

Publications/Reports/Research/PublRepoReseCompClaims_body.htm.

Clapham, Christopher, and John A. Wiseman. 1995. ‘‘Conclusions:

Assessing the Prospects for the Consolidation of Democracy.’’ In

Democracy and Political Change in Sub-Saharan Africa, ed. J. A.

Wiseman. London: Routledge (220–232).

Colletta, Nat J., Markus Kostner, and Ingo Wiederhofer. 2004. ‘‘Disarmament,

Demobilization, and Reintegration: Lessons and Liabilities in

Reconstruction.’’ In When States Fail: Causes and Consequences, ed.

Robert Rotberg. Princeton: Princeton University Press (170–181).

Colletta, Nat J. and Michelle L. Cullen. 2000. Violent Conflict and the

Transformation of Social Capital: Lessons from Cambodia, Rwanda,

Guatemala, and Somalia. Washington, DC: World Bank.

Collier, Paul, V. L. Elliot, Havard Hegre, Anke Hoeffler, Marta Reynal-

Querol, and Nicholas Sambanis, eds. 2003. Breaking the Conflict

Trap: Civil War and Development Policy. Washington, DC: World

Bank and Oxford University Press.

Collin Marks, Susan. 2000. Watching the Wind: Conflict Resolution during

South Africa’s Transition to Democracy. Washington, DC: United

States Institute of Peace Press.

Collins, Kathleen. 2006. The Logic of Clan Politics in Central Asia: The

Impact on Regime Transformation. New York: Cambridge University

Press.

Commission on Global Governance. 1995. Our Global Neighbourhood. The

Report of the Commission on Global Governance. Oxford: Oxford

University Press.

Cousens, Elizabeth M. 2001a. ‘‘Building Peace in Bosnia.’’ In Peacebuilding

as Politics: Cultivating Peace in Fragile Societies, ed. Elizabeth M.

Cousens and Chetan Kumar. Boulder, Colo. and London: Lynne

Rienner Publishers (113–152).

2001b. ‘‘Introduction.’’ In Peacebuilding as Politics, ed. Elizabeth M.

Cousens and Chetan Kumar. Boulder Colo. and London: Lynne

Rienner Publishers (1–20).

References 263

Cowen, Michael, and Liisa Laakso. 1997. ‘‘An Overview of Election Studies

in Africa.’’ Journal of Modern African Studies 35 (4): 717–744.

Dahl, Robert A. 1971. Polyarchy: Participation and Opposition. New

Haven: Yale University Press.

Damm, Darlene. 2003. Managing Election Violence in Cambodia: New

Strategies for the International Donor Community. Draft Report

1 December 2003. Johns Hopkins University: International Donor

Community Cambodia Steering Committee. Available at www.sais-

jhu.edu/bwelsh/DarlenePolicyPaper.pdf.

Darby, John. 2006. Violence and Reconstruction. South Bend, Ind.:

University of Notre Dame Press.

2001. The Effect of Violence on Peace Processes. Washington DC: USIP

Press.

1986. Intimidation and Control of Conflict in Northern Ireland. Dublin:

Gill and Macmillan.

Darby, John, and Roger Mac Ginty, eds. 2000. The Management of Peace

Processes. Basingstoke: Macmillan Press.

de Nevers, Renee. 1993. ‘‘Democratization and Ethnic Conflict.’’ In Ethnic

Conflict and International Security, ed. Michael Brown. Princeton:

Princeton University Press (61–78).

DeVotta, Neil. 2004. Blowback: Linguistic Nationalism, Institutional Decay,

and Ethnic Conflict in Sri Lanka. Stanford: Stanford University Press.

de Zeeuw, Jeroen. 2005. ‘‘Projects Do Not Create Institutions: The Record of

Democracy Assistance in Post-Conflict Societies.’’ Democratization

12 (4): 481–504.

du Toit, Pierre. 2003. ‘‘Why Post-settlement Settlements?’’ Journal of

Democracy 14 (3): 104–118.

Diamond, Larry. 2005. Squandered Victory: The American Occupation

and the Bungled Effort to Bring Democracy to Iraq. New York:

Time Books.

1999a. ‘‘Toward Democratic Consolidation.’’ In The Global Resurgence

of Democracy, ed. Larry Diamond and Marc F. Platter. Baltimore and

London: Johns Hopkins University Press (227–240).

1999b. Developing Democracy: Towards Consolidation. Baltimore: Johns

Hopkins University Press.

1997a. ‘‘Promoting Democracy in the 1990s: Actors, Instruments, and

Issues.’’ In Democracy’s Victory and Crisis, ed. Axel Hadenius.

Cambridge: Cambridge University Press (311–370).

1997b. ‘‘Introduction: In Search of Consolidation.’’ In Consolidating the

Third Wave Democracies: Themes and Perspectives, ed. Larry

Diamond, Marc F. Plattner, Yun-han Chu, and Hung-mao Tien.

Baltimore and London: Johns Hopkins University Press (xiii–xlvii).

264 References

Diamond, Larry, Juan J. Linz, and Seymour Martin Lipset, eds. 1988.

Democracy in Developing Countries. Boulder, Colo.: Lynne Rienner

Publishers.

Diamond, Larry, Marc F. Plattner, Yun-han Chu, and Hung-mao Tien, eds.

1997. Consolidating the Third Wave Democracies: Themes and

Perspectives. Baltimore and London: Johns Hopkins University Press.

Dorussen, Han. 2001. ‘‘Mixing Carrots with Sticks: Evaluating the Effectiveness

of Positive Incentives.’’ Journal of Peace Research 38 (2): 251–262.

Doyle, Michael W., and Nicholas Sambanis. 2006. Making War and

Building Peace: United Nations Peace Operations. Princeton:

Princeton University Press.

2000. ‘‘International Peacebuilding: A Theoretical and Quantitative

Analysis.’’ American Political Science Review 94 (4): 779–802.

Druckman, Daniel. 1986. ‘‘Stages, Turning Points, and Crises.’’ Journal of

Conflict Resolution 30: 327–360.

Edwards, Michael. 2004. Civil Society. Cambridge: Polity Press.

Eide, Espen Barth, Anja Therese Kaspersen, Randolph Kent, and Karen von

Hippel. 2005. Report on Integrated Missions: Practical Perspectives and

Recommendations. Independent Study for the Expanded UN ECHA

Core Group.

Elklit, Jørgen. 1999. ‘‘Electoral Institutional Change and Democratisation:

You Can Lead a Horse to Water, but You Can’t Make it Drink.’’

Democratisation 6 (4): 28–51.

Elklit, Jørgen, and Palle Svensson. 1997. ‘‘What Makes Elections Free and

Fair?’’ Journal of Democracy 8 (3): 32–46.

Eriksson, Mikael, and Peter Wallensteen. 2004. ‘‘Armed Conflict

1989–2003.’’ Journal of Peace Research 41 (5) (September): 625–636.

Eubank, William Lee, and Leonard Weinberg. 1994. ‘‘Does Democracy

Encourage Terrorism?’’ Terrorism and Political Violence 6 (4):

417–443.

Farer, Tom J. 2004. ‘‘The Promotion of Democracy: International Law and

Norms.’’ In The UN Role in Promoting Democracy. Between Ideals and

Reality, ed. Edward Newman and Roland Rich. Tokyo: UN University

Press (32–61).

Fearon, James D., and David D. Laitin. 2004. ‘‘Neotrusteeship and the

Problem of Weak States.’’ International Security 28 (4): 5–43.

2003. ‘‘Ethnicity, Insurgency, and Civil War.’’ American Political Science

Review 97 (1): 75–90.

Fischer, Jeff. 2002. Electoral Conflict and Violence: A Strategy for Study and

Prevention. IFES White Paper.

Fitzduff, Mari. 2004. Civil Society and Peacebuilding: The New Fifth Estate?

European Center for Conflict Prevention. Available at www.gppac.net.

References 265

Foley, Michael W. 1996. ‘‘Laying the Groundwork: The Struggle for Civil

Society in El Salvador.’’ Journal of Interamerican Studies and World

Affairs 38 (1): 67–104.

Fortna, Virginia Page. 2008. Does Peacekeeping Work? Shaping Belligerents’

Choices After Civil Wars. Princeton: Princeton University Press.

2005. ‘‘Where Have all the Victories Gone? War Outcomes in Historical

Perspective.’’ Paper presented at International Studies Association,

March, Honolulu, HI.

2004. ‘‘Does Peacekeeping Keep Peace? International Intervention and the

Duration of Peace after Civil War.’’ International Studies Quarterly

48 (2): 269–292.

Francis, David J. 2000. ‘‘Tortuous Path to Peace: The Lome Accord and

Postwar Peacebuilding in Sierra Leone.’’ Security Dialogue 31 (3):

357–373.

Franck, Thomas M. 1992. ‘‘The Emerging Right to Democratic

Governance.’’ American Journal of International Law 86 (1): 46–91.

Freedom House. 2006. Freedom in the World Comparative Rankings:

1975–2005. Available at www.freedomhouse.org.

2002–2005a. Country Report: Cambodia: Freedom House. Available at

www.freedomhouse.org.

2002–2005b. Country Report: East Timor: Freedom House. Available at

www.freedomhouse.org.

Fukuyama, Francis. 1995. Trust: The Social Virtues and the Creation of

Prosperity. New York: Free Press.

Gamba, Virginia. 2003. ‘‘Managing Violence: Disarmament and

Demobilization.’’ In Contemporary Peacemaking: Conflict, Violence

and Peace Processes, ed. J. Darby and R. Mac Ginty. Basingstoke:

Palgrave Macmillan (125–136).

Gidron, Benjamin, Stanley N. Katz, and Yeheskel Hasenfel. 2002.

Mobilizing for Peace: Conflict Resolution in Northern Ireland, Israel/

Palestine, and South Africa. Oxford: Oxford University Press.

Gilligan, Michael, and Stephen John Stedman. 2003. ‘‘Where Do the

Peacekeepers Go?’’ International Studies Review 5 (4): 37–54.

Gleditsch, Nils Petter, and Havard Hegre. 1997. ‘‘Peace and Democracy:

Three Levels of Analysis.’’ Journal of Conflict Resolution 41 (2):

283–310.

Greif, Avner. 1998. ‘‘Self-Enforcing Political Systems and Economic Growth:

Late Medieval Genoa.’’ In Analytic Narratives, ed. R. H. Bates, A. Greif,

M. Levi, J.-L. Rosenthal, and B. Weingast. Princeton: Princeton

University Press (23–63).

Grugel, Jean. 2002. Democratization. A Critical Introduction. Houndmills:

Palgrave Macmillan.

266 References

Guelke, Adrian. 2003. ‘‘Civil Society and the Northern Ireland Peace

Process.’’ Voluntas: International Journal of Voluntary and Nonprofit

Organizations 14 (1): 61–78.

Gunther, Richard, and Larry Diamond. 2001. ‘‘Types and Functions of

Parties.’’ In Political Parties and Democracy, ed. Richard Gunther and

Larry Diamond. Baltimore and London: Johns Hopkins University Press

(3–39).

Gurr, Ted Robert. 2000a. Peoples Versus States: Minorities at Risk in

the New Century. Washington, DC: United States Institute of Peace

Press.

2000b. ‘‘Ethnic warfare on the wane.’’ Foreign Affairs 79 (3): 52–64.

Hadenius, Axel, and Fredrik Uggla. 1996. ‘‘Making Civil Society Work,

Promoting Democratic Development: What Can States and Donors

Do?’’ World Development 24 (10): 1621–1639.

Haggard, Stephan. 1997. ‘‘Democratic Institutions, Economic Policy, and

Development.’’ In Institutions and Economic Development, ed.

Christopher Clague. Baltimore and London: Johns Hopkins University

Press (120–149).

Hammer, Peter J., and Tara Urs. 2005. ‘‘The Elusive Face of Cambodian

Justice.’’ In Bringing the Khmer Rouge to Justice: Prosecuting Mass

Violence before the Cambodian Courts, ed. Jaya Ramji and Beth Van

Schaack. Ontario: Edwin Mellen Press (13–58).

Hampson, Dave. 2004. ‘‘It’s Time to Open Up:’’ Ten Years After the Genocide

in Rwanda. Christian Aid, March. Available at www.discover.ac.uk/

socialsciences/cgi-bin/fullrecord.pl?handle=sosig1.

Hampson, Fen Osler. 1996. Nurturing Peace: Why Peace Settlements

Succeed or Fail. Washington, DC: United States Institute of Peace Press.

Hann, Chris, and Elizabeth Dunn, eds. 1996. Civil Society: Challenging

Western Models. London: Routledge.

Haraszti, Miklos. 2004. The Role of the Media in the March 2004 Events in

Kosovo. Vienna: OSCE.

Harbom, Lotta, Stina Hogbladh, and Peter Wallensteen. 2006. ‘‘Armed

Conflicts and Peace Agreements.’’ Journal of Peace Research 43 (5):

617–631.

Harbom, Lotta, and Peter Wallensteen. 2005. ‘‘Armed Conflict and Its

International Dimensions, 1946–2004.’’ Journal of Peace Research

42 (5): 623–635.

Harris, Peter, and Benjamin Reilly, eds. 1998. Democracy and Deep-Rooted

Conflict: Options for Negotiators. Stockholm: International Institute

for Democracy and Electoral Assistance.

Hartlyn, Jonathan. 1998. The Politics of Coalition Rule in Colombia.

Cambridge: Cambridge University Press.

References 267

Hartzell, Caroline A., and Matthew Hoddie. 2003. ‘‘Institutionalizing Peace:

Power Sharing and Post-Civil War Conflict Management.’’ American

Journal of Political Science 47 (2): 318–332.

Hearn, Julie. 2000. ‘‘Aiding Democracy? Donors and Civil Society in South

Africa.’’ Third World Quarterly 21 (5): 815–830.

Hegre, Havard, Tanja Ellingsen, Scott Gates, and Nils Petter Gleditsch.

2001. ‘‘Towards a Democratic Civil Peace? Democracy, Political

Change and Civil War, 1816–1992.’’ American Political Science

Review 95 (1): 33–48.

Heldt, Birger. 2007. ‘‘Peacekeeping Operations and Post-conflict Transitions

to Democracy?’’ Paper presented at the 2007 Annual Conference by the

Swedish Network of Peace, Conflict and Development Research: ‘‘The

Democratization Project: Challenges and Opportunities,’’ Uppsala

October 23–24, 2007.

Hoddie, Matthew, and Caroline A. Hartzell. 2005. ‘‘Power Sharing in Peace

Settlements: Initiating the Transition from Civil War.’’ In Sustainable

Peace: Power and Democracy after Civil War, ed. Philip G. Roeder and

Donald Rothchild. Ithaca and London: Cornell University Press

(83–106).

Hodges, Tony. 2001. Angola From Afro-Stalinism to Petro-Diamond

Capitalism. Bloomington and Indianapolis: Indiana University Press.

Hoglund, Kristine. 2008. Peace Negotiations in the Shadow of Violence.

Leiden: Martinus Nijhoff Publishers.

2005. ‘‘Violence and the Peace Process in Sri Lanka.’’ Civil Wars 7 (2):

156–170.

Hoglund, Kristine, and I. William Zartman. 2006. ‘‘Violence by the State:

Official Spoilers and their Allies.’’ In Violence and Reconstruction, ed.

J. Darby. South Bend, Ind.: University of Notre Dame Press (11–31).

Hohe, Tanja. 2002. ‘‘Totem Polls: Indigenous Concepts and ‘Free and Fair’

Elections in East Timor.’’ International Peacekeeping 9 (4): 69–88.

Holbrooke, Richard. 1998. To End a War. New York: Random House.

Holm, Tor Tanke, and Espen Barth Eide. 2000. Peacebuilding and Police

Reform. London: Frank Cass.

Holohan, Anne. 2005. Networks of Democracy: Lessons from Kosovo

for Afghanistan, Iraq, and Beyond. Stanford: Stanford University

Press.

Holsti, Kalevi J. 1996. The State, War, and the State of War. Cambridge:

Cambridge University Press.

Horowitz, Donald L. 2001. The Deadly Ethnic Riot. Berkeley: University of

California Press.

1985. Ethnic Groups in Conflict. Berkeley, Los Angeles, and London:

University of California Press.

268 References

Howard, Ross. 2005. ‘‘The Media’s Role in War and Peacebuilding.’’ In

Postconflict Development: Meeting New Challenges, ed. G. Junne and

W. Verkoren. Boulder, Colo. and London: Lynne Rienner Publishers

(117–128).

Human Rights Watch. 2004. Failure to Protect: Anti-Minority Violence in

Kosovo, March 2004. New York: Human Rights Watch.

Human Security Report 2005: War and Peace in the 21st Century. 2005.

New York and Oxford: Oxford University Press.

Huntington, Samuel P. 1991. The Third Wave. Democratization in the Late

Twentieth Century. Norman, Okla.: University of Oklahoma Press.

1957. The Soldier and the State: The Theory and Politics of Civil-Military

Relations. Cambridge, Mass.: Harvard University Press.

ICG. 2007a. Nepal’s Constitutional Process. Asia Report No. 128.

Kathmandu and Brussels: International Crisis Group.

2007b. After Mecca: Engaging Hamas. Middle East Report No. 62.

Amman, Jerusalem, and Brussels: International Crisis Group.

2006a. Sudan’s Comprehensive Peace Agreement: The Long Road Ahead.

Africa Report No. 106. Nairobi and Brussels: International Crisis Group.

2006b. Lebanon at a Tripwire. Middle East Briefing No. 20. Beirut and

Brussels: International Crisis Group.

2005a. Elections in Burundi: A Radical Shake-up of the Political

Landscape. Africa Briefing No. 31. Nairobi and Brussels: International

Crisis Group.

2005b. Kosovo after Haradinaj. Europe Report No. 163. Pristina and

Brussels: International Crisis Group.

2005c. Aceh: A New Chance for Peace. Asia Briefing No. 40. Jakarta and

Brussels: International Crisis Group.

2005d. Liberia’s Elections: Necessary but Not Sufficient. Africa Report

No. 107. Dakar and Brussels: International Crisis Group.

2004a. Collapse in Kosovo. Europe Report No. 155. Pristina, Belgrade,

and Brussels: International Crisis Group.

2004b. Tajikistan’s Politics: Confrontation or Consolidation? Asia

Briefing No. 33. Dushanbe and Brussels: International Crisis Group.

2004c. Elections in Burundi: The Peace Wager. Africa Briefing No. 20.

Nairobi and Brussels: International Crisis Group.

2003. The Special Court for Sierra Leone: Promises and Pitfalls of a ‘‘New

Model.’’ Africa Briefing No. 16. Freetown and Brussels: International

Crisis Group.

2002. Sierra Leone after Elections: Politics as Usual? Africa Report

No. 49. Freetown and Brussels: International Crisis Group.

Inglehart, Ronald. 1997. Modernization and Postmodernization. Princeton:

Princeton University Press.

References 269

Jarstad, Anna K. 2007. ‘‘To Share or to Divide? Negotiating the Future of

Kosovo.’’ Civil Wars 9 (3): 227–242.

2001. ‘‘Changing the Game: Consociational Theory and Ethnic Quotas in

Cyprus and New Zealand.’’ Doctoral Dissertation, Department of Peace

and Conflict Research, Uppsala University.

Jarstad, Anna K., and Ralph Sundberg. 2007. ‘‘Peace by Pact: Data on the

Implementation of Peace Agreements.’’ In Globalization and Challenges

to Building Peace, ed. Ashok Swain, Ramses Amer, and Joakim Ojendal.

London: Anthem Press (73–89).

Jeong, Ho-Won. 2005. Peace building in Postconflict Societies. Strategy and

Process. Boulder, Colo.: Lynne Rienner Publishers.

Jung, Dietrich, ed. 2003. Shadow Globalization, Ethnic Conflicts and New

Wars: A Political Economy of Intra-State War. London and New York:

Routledge.

Kandeh, Jimmy D. 2003. ‘‘Sierra Leone’s Post-Conflict Elections of 2002.’’

Journal of Modern African Studies 41 (2): 189–216.

Kaldor, Mary. 2003. Global Civil Society: An Answer to War. Cambridge:

Polity Press.

Kasfir, Nelson. 2004. ‘‘Civil Society, the State and Democracy in Africa.’’ In

Civil Society in Democratization, ed. Peter Burnell and Peter Calvert.

London and Portland, reg.: Frank Cass (117–142).

Katana, Gordana. 1999. ‘‘NGOs in Republika Srpska: Bashful Support of the

Regime.’’ AIM Banja Luka, September 20. Available at www.

aimpress.ch/dyn/trae/archive/data/199909/90926-002-trae-sar.htm.

Kaufmann, Chaim. 1997. ‘‘Possible and Impossible Solutions to Ethnic Civil

Wars.’’ In Nationalism and Ethnic Conflict, ed. Michael Brown et al.

(265–304).

Keane, John. 2004. Violence and Democracy. Cambridge: Cambridge

University Press.

1998. Civil Society: Old Images, New Perspectives. Cambridge: Polity Press.

Keck, Margaret, and Kathryn Sikkink. 1998. Activists Beyond Borders:

Advocacy Networks in International Politics. Ithaca: Cornell

University Press.

Knaus, Gerald, and Felix Martin. 2003. ‘‘Travails of the European Raj.’’

Journal of Democracy 14 (3): 60–74.

Kohn, Richard H. 1997. ‘‘How Democracies Control the Military.’’ Journal

of Democracy 8 (4): 140–153.

Kritz, Neil J. 2001. ‘‘The Rule of Law in the Postconflict Phase: Building a

Stable Peace.’’ In Turbulent Peace. The Challenges of Managing

International Conflict, ed. Fen O. Hampson, Chester A. Crocker, and

Pamela Aall. Washington, DC: United States Institute of Peace Press

(801–820).

270 References

Krznaric, Roman. 1999. ‘‘Civil and Uncivil Actors in the Guatemalan Peace

Process.’’ Bulletin of Latin American Research 18 (1): 1–16.

Kumar, Krishna. 1998a. ‘‘Postconflict Elections and International

Assistance.’’ In Postconflict Elections, Democratization, and

International Assistance, ed. Krishna Kumar. Boulder, Colo.: Lynne

Rienner Publishers (5–14).

1998b. ‘‘After the Elections: Consequences for Democratization.’’ In

Postconflict Elections, Democratization, and International Assistance,

ed. Krishna Kumar. Boulder, Colo.: Lynne Rienner Publishers

(215–227).

ed. 1998c. Postconflict Elections, Democratization, and International

Assistance. Boulder, Colo.: Lynne Rienner Publishers.

ed. 1997. Rebuilding Societies after Civil War: Critical Roles for International

Assistance. Boulder, Colo. and London: Lynne Rienner Publishers.

Kydd, Andrew, and Barbara F. Walter. 2002. ‘‘Sabotaging the Peace:

The Politics of Extremist Violence.’’ International Organization

56 (2): 263–296.

Lacy, Brett. 2004. ‘‘Building Accountability, Legitimacy, and Peace:

Refugees, Internally Displaced Persons, and the Right to Political

Participation.’’ IFES Occasional Paper, Washington, DC.

Lanegran, Kimberly. 2001. ‘‘South Africa’s 1999 Election: Consolidating a

Dominant Party System.’’ Africa Today 48 (2): 81–102.

Lederach, John Paul. 1997. Building Peace. Sustainable Reconciliation in

Divided Societies. Washington, DC: United States Institute of Peace.

Levi, Margaret. 1996. Consent, Dissent, and Patriotism. Cambridge:

Cambridge University Press.

Licklider, Roy. 2005. ‘‘Comparative Studies of Long Wars.’’ In Grasping the

Nettle: Analyzing Cases of Intractable Conflict, ed. Chester A. Crocker,

Fen Osler Hampson, and Pamela Aall. Washington, DC: United States

Institute of Peace (33–46).

Liddle, R. William. 1997. ‘‘Coercion, Co-optation, and the Management of

Ethnic Relations in Indonesia.’’ In Government Policies and Ethnic

Relations in the Asia-Pacific, ed. Michael E. Brown and Sumit

Ganguly. Cambridge and London: MIT Press (273–319).

Lijphart, Arend. 1999. Patterns of Democracy. Government Forms and

Performances in Thirty-Six Countries. New Haven and London: Yale

University Press.

1994. ‘‘Democracies: Forms, Performance, and Constitutional Engineering.’’

European Journal of Political Research 25: 1–17.

1993. ‘‘Consociational Democracy.’’ In The Oxford Companion to

Politics of the World, ed. Joel Krieger. New York and Oxford: Oxford

University Press (188–189).

References 271

1985. Power-Sharing in South Africa. Berkeley: Institute of International

Studies, University of California.

1977. Democracy in Plural Societies. New Haven and London: Yale

University Press.

1968. ‘‘Typologies of Democratic Systems.’’ Comparative Political Studies

1 (April): 3–44.

Linz, Juan, and Alfred Stepan. 1996. Problems of Democratic Transition and

Consolidation: Southern Europe, South America, and Post-Communist

Europe. Baltimore: Johns Hopkins University Press.

1992. ‘‘Political Identities and Electoral Sequences: Spain, the Soviet

Union, and Yugoslavia.’’ Daedalus 121 (2): 123–139.

Lipset, Seymour Martin. 1959. ‘‘Some Social Requisites of Democracy:

Economic Development and Political Legitimacy.’’ American Political

Science Review 53: 69–105.

Lithander, Anna. 2000. Engendering the Peace Process: A Gender Approach to

Dayton – and Beyond. Stockholm: Kvinna till Kvinna Foundation.

Londregan, John B., and Keith T. Poole. 1996. ‘‘Does High Income Promote

Democracy?’’ World Politics 49: 1–30.

Lopez-Pintor, Rafael. 2000. Electoral Management Bodies as Institutions of

Governance. New York: United Nations Development Programme.

1998. ‘‘Nicaragua’s Measured Move to Democracy.’’ In Postconflict

Elections, Democratization, and International Assistance, ed. Krishna

Kumar. Boulder, Colo.: Lynne Rienner Publishers (39–55).

1997. ‘‘Reconciliation Elections: A Post-Cold War Experience.’’ In

Rebuilding Societies after Civil War: Critical Roles for International

Assistance, ed. K. Kumar. Boulder, Colo. and London: Lynne Rienner

Publishers (43–61).

Louw, Antoniette. 1997. ‘‘Surviving the Transition: Trends and Perceptions

in Crime.’’ Social Indicators Research 41: 137–168.

Ludwig, Robin. 2004. ‘‘The UN’s Electoral Assistance: Challenges,

Accomplishments, Prospects.’’ In The UN Role in Promoting

Democracy: Between Ideals and Reality, ed. Edward Newman and

Roland Rich. Tokyo: UN University Press (169–187).

Lyons, Terrence. 2005. Demilitarizing Politics: Elections on the Uncertain

Road to Peace. Boulder, Colo.: Lynne Rienner Publishers.

2004. ‘‘Transforming the Institutions of War: Post-Conflict Elections and the

Reconstruction of Failed States.’’ In When States Fail, ed. Robert I.

Rotberg. Princeton: Princeton University Press (269–301).

2002. ‘‘The Role of Postsettlement Elections.’’ In Ending Civil Wars: The

Implementation of Peace Agreements, ed. Stephen John Stedman,

Donald Rothchild, and Elisabeth M. Cousens. Boulder, Colo.: Lynne

Rienner Publishers (215–235).

272 References

Mac Ginty, Roger. 2006. ‘‘Post-Accord Crime.’’ In Violence and

Reconstruction, ed. J. Darby. South Bend, Ind.: University of Notre

Dame Press (101–120).

Mac Ginty, Roger, and John Darby. 2002. Guns and Government: The

Management of the Northern Ireland Peace Process. Basingstoke:

Palgrave.

MacGarry, John, and Brendan O’Leary, eds. 2004. The Northern Ireland

Conflict: Consociational Engagements. Oxford: Oxford University

Press.

MacMillan, John. 2003. ‘‘Beyond the Separate Democratic Peace.’’ Journal

of Peace Research 40 (2): 233–243.

Maksymenko, Serhiy. 2005. ‘‘European Neighbourhood Policy and Post-

Orange Revolution Ukraine.’’ In Development, Security and Conflict

Prevention, ed. Anders Mellbourn. Brussels: Madariaga European

Foundation Bank of Sweden Tercentenary Foundation and Gidlunds

(164–174).

Malone, David, ed. 2004. The Security Council: From the Cold War to the

21st Century. Boulder, Colo. and London: International Peace Academy

and Lynne Rienner Publishers.

Mani, Rama. 2002. Beyond Retribution: Seeking Justice in the Shadows of

War. Cambridge, UK: Polity Press.

Mann, Michael. 2005. The Dark Side of Democracy: Explaining Ethnic

Cleansing. Cambridge: Cambridge University Press.

Manning, Carrie. 2004. ‘‘Armed Opposition Groups into Political Parties:

Comparing Bosnia, Kosovo, and Mozambique.’’ Studies in Comparative

International Development 39 (1): 54–76.

Mansfield, Edward D., and Jack Snyder. 2005. Electing to Fight: Why

Emerging Democracies Go to War. Cambridge: MIT Press.

2002a. ‘‘Democratic Transitions, Institutional Strength, and War.’’

International Organization 56 (2): 297–337.

2002b. ‘‘Incomplete Democratic Transitions and the Outbreak of Military

Disputes.’’ International Studies Quarterly 46 (4): 529–549.

2001. ‘‘Democratic Transitions and War: From Napoleon to the

Millennium’s End.’’ In Turbulent Peace: The Challenges of Managing

International Conflict, ed. C. A. Crocker, F. O. Hampson, and P. Aall.

Washington, DC: United States Institute of Peace Press (113–126).

1995a. ‘‘Democratization and the Danger of War.’’ International Security

20 (1): 5–38.

1995b. ‘‘Democratization and War.’’ Foreign Affairs 74 (3): 79–97.

Maoz, Zeev, and Bruce M. Russett. 1993. ‘‘Normative and Structural Causes

of Democratic Peace 1946–1986.’’ American Political Science Review

87 (3): 624–638.

References 273

Marshall, Monty G., and Keith Jaggers. 2002. Polity IV Project: Political

Regime Characteristics and Transitions, 1800–2002. College Park,

Md.: Center for International Development and Conflict Management

(CIDCM), University of Maryland.

Marten, Kimberly Zisk. 2004. Enforcing the Peace: Learning from the

Imperial Past. New York: Columbia University Press.

Mathie, Alison, and Gord Cunningham. 2002. From Clients to Citizens:

Asset-Based Community Development as a Strategy for Community-

Driven Development. The Coady International Institute, St. Francis

Xavier University, Occasional Paper.

McCoy, Jennifer, Larry Gerber, and Robert Pastor. 1991. ‘‘Pollwatching and

Peacemaking.’’ Journal of Democracy 2 (4): 102–114.

Melber, Henning. 2002. ‘‘From Liberation Movements to Governments: On

Political Culture in Southern Africa.’’ African Sociological Review 6 (1):

161–172.

Mendelson, Sarah E. 2005. Barracks and Brothels: Peacekeepers and Human

Trafficking in the Balkans. Washington, DC: Center for Strategic and

International Studies.

Metzl, Jamie F. 1997. ‘‘Information Intervention: When Switching Channel

Isn’t Enough.’’ Foreign Affairs 76 (6): 15–20.

Morrow, James D. 1994. Game Theory for Political Scientists. Princeton:

Princeton University Press.

Mukherjee, Bumba. 2006. ‘‘Why Political Power-Sharing Agreements Lead

to Enduring Peaceful Resolution of Some Civil Wars, But Not Others?’’

International Studies Quarterly 50 (2): 479–504.

Mullojanov, Perviz. 2001. ‘‘Civil Society and Peacebuilding. Politics of

Compromise: The Tajikistan Peace Process.’’ Conciliation Resources,

Accord No. 10, March. Available at www.c-r.org/accord10/

peacebuilding.shtml.

Museveni, Yoweri. 2005. Quoted on BBC December 21, 2005. Available at

http://news.bbc.co.uk/1/hi/world/africa/4549220.stm.

National Democratic Institute. 1996. Bosnia and Hercegovina (BiH):

Strengthening Political Parties. Washington DC: NDI Reports.

Nelson, Sue. 1998. ‘‘Haitian Elections and the Aftermath.’’ In Post-

conflict Elections, Democratization, and International Assis-tance,

ed. Krishna Kumar. Boulder, Colo.: Lynne Rienner Publishers

(71–86).

Newman, Edward. 2004. ‘‘UN Democracy Promotion: Comparative

Advantages and Constraints.’’ In The UN Role in Promoting

Democracy: Between Ideals and Reality, ed. Edward Newman

and Roland Rich. Tokyo: United Nations University Press

(188–207).

274 References

Newman, Edward, and Roland Rich, eds. 2004. The UN Role in Promoting

Democracy: Between Ideals and Reality. Tokyo: United Nations

University Press.

Nilsson, Desiree. 2006. In the Shadow of Settlement: Multiple Rebel Groups

and Precarious Peace. Uppsala: Department of Peace and Conflict

Research, Uppsala University.

2005. ‘‘Inclusive Peace Agreements: A Path to Durable Peace?’’ Paper

presented at a review conference October 24–26, 2005 in Uppsala on

the book project From Intra-State War to Durable Peace: Conflict and

Conflict Resolution in Africa 1989–2004.

Nilsson, Desiree, and Mimmi Soderberg Kovacs. 2005. ‘‘Breaking the Cycle

of Violence? Promises and Pitfalls of the Liberian Peace Process.’’ Civil

Wars 7 (4): 396–414.

Oakley, Robert B., Michael J. Dziedzic, and Eliot M. Goldberg, eds. 2002.

Policing the New World Order: Peace Operations and Public Security.

Washington DC: National Defence University Press.

O’Donnell, Guillermo, Philippe C. Schmitter, and Lawrence Whitehead, eds.

1986. Transitions from Authoritarian Rule: Prospects for Democracy.

Baltimore: Johns Hopkins University Press.

Ohlson, Thomas. 1998. Power Politics and Peace Policies: Intra-State

Conflict Resolution in Southern Africa, Uppsala: Department of Peace

and Conflict Research, Uppsala University.

Oneal, John R., Bruce M. Russett, and Michael L. Berbaum. 2003. ‘‘Causes of

Peace: Democracy, Interdependence, and International Organizations,

1885–1992.’’ International Studies Quarterly 47 (3): 371–393.

Orjuela, Camilla. 2005. ‘‘Civil Society in Civil War: The Case of Sri Lanka.’’

Civil Wars 7 (2): 120–137.

OSCE. 2006. ‘‘International Election Observation Mission, Bosnia and

Herzegovina – General Elections, 1 October 2006.’’ OSCE/ODIHR

Election Observation Mission Final Report. Warsaw: OSCE.

Ottaway, Marina. 1997. ‘‘From Political Opening to Democratization?’’ In

Democracy in Africa. The Hard Road Ahead, ed. Marina Ottaway.

Boulder: Lynne Rienner Publishers (1–14).

1991. ‘‘Liberation Movements and Transition to Democracy: The Case of

the ANC.’’ Journal of Modern African Studies 29 (1): 61–82.

Paffenholz, Thania. 2001. ‘‘Designing Transformation and Intervention

Processes.’’ In Berghof Handbook for Conflict Transformation, ed.

David Bloomfield, Martina Fischer, and Beatrix Schmelzle. Berlin:

Berghof Center for Constructive Conflict Management. Available at

www.berghof-handbook.net

Paffenholz, Thania, and Christoph Spurk. 2006. Civil Society, Civic

Engagement, and Peacebuilding. Washington, DC: Social

References 275

Development Papers, Paper No. 36, commissioned by the Social

Development Department of the World Bank, October.

Papagianni, Katia. 2006. ‘‘Transitional Politics in Post-War Countries.’’

Paper presented at the International Symposium ‘‘State, Conflict and

Democracy,’’ Lund, May 12–13, 2006.

Paris, Roland 2004. At War’s End: Building Peace after Civil Conflicts.

Cambridge: Cambridge University Press.

Patrick, Ian. 2001. ‘‘East Timor Emerging From Conflict: The Role of Local

NGOs and International Assistance.’’ Disasters 25 (1): 48–66.

Pausewang, Siegfried, Kjetil Trondvall, and Lovise Aalen, eds. 2002.

Ethiopia since the Derg: A Decade of Democratic Pretension and

Performance. London: Zed Books.

Perito, Robert M. 2004. Where Is the Lone Ranger When We Need Him?

America’s Search for a Postconflict Stability Force. Washington, DC:

United States Institute of Peace Press.

Phillips, Anne. 1998. The Politics of Presence. Oxford: Oxford University Press.

Pickering, Jeffrey, and Mark Peceny. 2006. ‘‘Forging Democracy at

Gunpoint.’’ International Studies Quarterly 50 (3): 539–560.

Plunkett, Mark. 2005. ‘‘Reestablishing the Rule of Law.’’ In Postconflict

Development: Meeting New Challenges, ed. G. Junne and W. Verkoren.

Boulder and London: Lynn Rienner Publishers (73–98).

Posen, Barry R. 1993. ‘‘The Security Dilemma and Ethnic Conflict.’’ Survival

35 (1): 27–47.

Posner, Daniel N. 2004. ‘‘Civil Society and the Reconstruction of Failed

States.’’ In When States Fail: Causes and Consequences, ed. Robert

Rotberg. Princeton: Princeton University Press (237–255).

Pouligny, Beatrice. 2005. ‘‘Civil Society and Post-Conflict Peacebuilding:

Ambiguities of International Programmes Aimed at Building ‘New’

Societies.’’ Security Dialogue 46 (4): 495–510.

Przeworski, Adam. 1991. Democracy and the Market: Political and

Economic Reforms in Eastern Europe and Latin America. Cambridge:

Cambridge University Press.

Przeworski, Adam, Michael Alvarez, Jose Antonio Cheibub, and Fernando

Limongi. 1996. ‘‘What Makes Democracy Endure.’’ Journal of

Democracy 7 (1): 39–55.

Przeworski, Adam, and Fernando Limongi. 1997. ‘‘Modernization: Theories

and Facts.’’ World Politics 49 (2): 155–183.

Pugh, Michael. 2000. ‘‘The Civil/Social Dimension.’’ In The Regeneration

of War-Torn Societies, ed. Michael Pugh. London: Macmillan

(112–133).

Putnam, Robert. 2000. Bowling Alone: The Collapse and Revival of

American Community. New York: Simon & Schuster.

276 References

1993. Making Democracy Work: Civic Traditions in Modern Italy.

Princeton: Princeton University Press.

Rabushka, Alvin, and Kenneth Shepsle. 1972. Politics in Plural Societies: A

Theory of Democratic Instability. Colombus, OH: Merrill.

Randall, Vicky, and Lars Svasand. 2002. ‘‘Political Parties and Democratic

Consolidation in Africa.’’ Democratization 9 (3): 30–52.

Rapoport, David C., and Leonard Weinberg, eds. 2001a. The Democratic

Experience and Political Violence. London: Frank Cass.

2001b. ‘‘Elections and Violence.’’ In The Democratic Experience and

Political Violence, ed. D. C. Rapoport and L. Weinberg. London:

Frank Cass (15–20).

Rawls, John. 1971. A Theory of Justice. Cambridge, Mass.: Harvard

University Press.

Reilly, Benjamin. 2007. Democracy and Diversity: Political Engineering in

the Asia-Pacific. Oxford: Oxford University Press.

2006. ‘‘Political Engineering and Party Politics in Conflict-Prone

Societies.’’ Democratization 13 (5): 811–827.

2004. ‘‘Elections in Post-Conflict Societies.’’ In The UN Role in Promoting

Democracy: Between Ideals and Reality, ed. Edward Newman and

Roland Richs. Tokyo: United Nations University Press (113–134).

2003a. ‘‘Democratic Validation.’’ In Contemporary Peacemaking:

Conflict, Violence and Peace Processes, ed. John Darby and Roger

MacGinty. London: Palgrave (174–183).

2003b. International Electoral Assistance: A Review of Donor Activities

and Lessons Learned. The Hague: Netherlands Institute of International

Relations ‘‘Clingendael.’’

2002. ‘‘Post-Conflict Elections: Constraints and Dangers.’’ International

Peacekeeping 9 (2): 118–139.

2001. Democracy in Divided Societies: Electoral Engineering for Conflict

Management. Cambridge: Cambridge University Press.

Reinicke, Wolfgang H., and Francis Deng. 2000. Critical Choices: The

United Nations, Networks, and the Future of Global Governance.

Ottawa: International Development Research Centre.

Reynolds, Andrew, ed. 2002. The Architecture of Democracy: Constitutional

Design, Conflict Management, and Democracy. Oxford: Oxford

University Press.

Reynolds, Andrew, Benjamin Reilly, and Andrew Ellis. 2005, Electoral

System Design: The New International IDEA Handbook. Stockholm:

International Institute for Democracy and Electoral Assistance.

Rich, Roland. 2004. ‘‘Crafting Security Council Mandates.’’ In The UN Role in

Promoting Democracy: Between Ideals and Reality, ed. Edward Newman

and Roland Rich. Tokyo: United Nations University Press (62–85).

References 277

Risley, Paul, and D. Timothy Sisk. 2005. Democracy and United Nations

Peacebuilding at the Local Level: Lessons Learned. Stockholm: Interna-

tional IDEA. Available at www.idea.int/publications/democracy_un/

index.cfm.

Roeder, Philip G., and Donald Rothchild, eds. 2005. Sustainable Peace:

Power and Democracy after Civil Wars. Ithaca and London: Cornell

University Press.

Rosato, Sebastian. 2003. ‘‘The Flawed Logic of Democratic Peace Theory.’’

American Political Science Review 97 (4): 585–602.

Ross, Michael L. 2001. ‘‘Does Oil Hinder Democracy?’’ World Politics 53

(3): 325–361.

Rothchild, Donald. 2005a. ‘‘Problems in Applying Executive Power Sharing

to Africa: The Impact of Sequencing on Political Consolidation.’’ Paper

presented at 46th Annual International Studies Association Convention,

1–5 March, at Honolulu, HI.

2005b. ‘‘Reassuring Weaker Parties after Civil Wars: The Benefits and

Costs of Executive Power-sharing Systems in Africa.’’ Ethnopolitics 4

(3): 247–267.

1997. Managing Ethnic Conflict in Africa: Pressures and Incentives for

Cooperation. Washington, DC: Brookings Institute.

1973. Racial Bargaining in Independent Kenya: A Study of Minorities and

Decolonization. Berkeley: University of California Press.

Russett, Bruce. 1993. Grasping the Democratic Peace. Princeton: Princeton

University Press.

Russett, Bruce M., and John R. Oneal. 2001. Triangulating Peace:

Democracy, Interdependence, and International Organizations.

New York: W. W. Norton.

Rustow, Dankwart A. 1970. ‘‘Transitions to Democracy: Toward a Dynamic

Model.’’ Comparative Politics 2 (3): 337–363.

Salih, M. A. Mohamed. 2003. ‘‘Introduction: The Evolution of African

Political Parties.’’ In African Political Parties: Evolution,

Institutionalisation and Governance, ed. M. A. Mohamed Salih.

London: Pluto Press (1–33).

Salomons, Dirk. 2005. ‘‘Security: An Absolute Prerequisite.’’ In Postconflict

Development: Meeting New Challenges, ed. G. Junne and W. Verkoren.

Boulder Colo. and London: Lynne Rienner Publishers (19–42).

Shaw, Mark. 2002. Crime and Policing in Post-Apartheid South Africa:

Transforming Under Fire. Cape Town and Johannesburg: David Philip

Publisher.

Shoemaker, Jolynn. 2005. Conflict Prevention and Transformation:

Women’s Vital Contribution. Cambridge and Washington, DC:

Women Waging Peace and the United Nations Foundation.

278 References

Sida. 2006. Press Statement, August 9, 2006. Stockholm: Swedish

International Development Cooperation Agency.

2005. Promoting Peace and Security through Development Cooperation.

Stockholm: Swedish International Development Cooperation Agency.

Simonsen, Sven Gunnar. 2004. ‘‘Nationbuilding as Peacebuilding: Racing to

Define the Kosovar.’’ International Peacekeeping 11 (2): 289–311.

Sisk, Timothy D. 2001. ‘‘Intrastate Violence.’’ In Managing Global Issues:

Lessons Learned, ed. P. J. Simmons and Chantal de Jonge Oudraat.

Washington, DC: Carnegie Endowment for International Peace

(534–563).

1996. Power Sharing and International Mediation in Ethnic Conflict.

Washington, DC: Carnegie Corporation of New York.

1995. Democratization in South Africa: The Elusive Social Contract.

Princeton: Princeton University Press.

Sisk, Timothy D., and Christoph Stefes. 2005. ‘‘Power Sharing as an Interim

Step in Peace Building: Lessons from South Africa.’’ In Sustainable Peace.

Power and Democracy after Civil Wars, ed. Philip G. Roeder and Donald

Rothchild. Ithaca and London: Cornell University Press (293–319).

Smith, Anthony L. 2004. ‘‘East Timor: Elections in the World’s Newest

Nation.’’ Journal of Democracy 15 (2): 145–159.

Snyder, Jack. 2000. From Voting to Violence: Democratization and

Nationalist Conflict. New York and London: W.W. Norton.

Snyder, Jack, and Karen Ballentine. 1996. ‘‘Nationalism and the

Marketplace of Ideas.’’ International Security 21 (2): 5–40.

Soderberg Kovacs, Mimmi. 2007. From Rebellion to Politics: The

Transformation of Rebel Groups to Political Parties in Civil War

Peace Processes. Uppsala: Department of Peace and Conflict Research,

Uppsala University.

Solana, Javier. 2005. ‘‘Prevention Must Be Concrete.’’ In Development,

Security and Conflict Prevention, ed. Anders Mellbourn. Brussels:

Madariaga European Foundation (9–14).

Southall, Roger, Neo Simutanyi, and John Daniel. 2005. ‘‘Former Presidents

in African Politics.’’ In Legacies of Power, ed. Roger Southall and

Henning Melber. Uppsala: Nordic Africa Institute (1–25).

Spear, Joanna. 2002. ‘‘Disarmament and Demobilization.’’ In Ending Civil

Wars: The Implementation of Peace Agreements, ed. John Stephen

Stedman, Donald Rothchild, and Elizabeth M. Cousens. Boulder,

Colo.: Lynne Rienner Publishers (141–182).

Spears, Ian S. 2000. ‘‘Understanding Inclusive Peace Agreements in Africa: The

Problems of Sharing Power.’’ Third World Quarterly 21 (1): 105–118.

Stedman, Stephen John. 1997. ‘‘Spoiler Problems in Peace Processes.’’

International Security 22 (2): 5–53.

References 279

Stedman, Stephen John, Donald Rothchild, and Elisabeth M. Cousens, eds.

2002. Ending Civil Wars: The Implementation of Peace Agreements.

Boulder, Colo.: Lynne Rienner Publishers.

Stewart, Frances. 2000. ‘‘Crisis Prevention: Tacking Horizontal

Inequalities.’’ Oxford Development Studies 28 (3): 245–262.

Stremlau, John. 1998. A House No Longer Divided: Civil Society in South

Africa. Washington, DC: Carnegie Commission on Preventing Deadly

Conflict.

Stubbs, Paul. 1999. Displaced Promises: Forced Migration, Refugees and

Return in Croatia and Bosnia-Herzegovina. Uppsala: Life & Peace

Institute.

Tangri, Roger. 2005. ‘‘Politics and Presidential Term Limits in Uganda.’’ In

Legacies of Power, ed. Roger Southall and Henning Melber. Uppsala:

Nordic Africa Institute (175–196).

Temporary Media Commissioner. 2004. An Inquiry into the Performance of

Kosovo-Wide Television (16–17 March 2004). Pristina: Temporary

Media Commissioner.

Thompson, Mark. 1994. Forging War: The Media in Serbia, Croatia and

Bosnia-Herzegovina. London: Article 19.

Thompson, William R., and Tucker Richard. 1997. ‘‘A Tale of Two Democratic

Peace Critiques.’’ Journal of Conflict Resolution 41 (3): 428–454.

Tilly, Charles. 1990. Coercion, Capital, and European States, AD 990–1992.

Cambridge Mass.: Basil Blackwell.

1985. ‘‘War-Making and State-Making as Organized Crime.’’ In Bringing

the State Back In, ed. P. Evans, D. Rueschmeyer, and T. Skocpol.

Cambridge: Cambridge University Press (169–191).

Toft, Monica Duffy. 2003. ‘‘Peace Through Victory?’’ Paper presented at

American Political Science Association, August 27–31, Philadelphia.

UCDP. 2007. Uppsala Conflict Data Program. [cited 24 March 2007].

Available at www.pcr.uu.se/database/.

2006. ‘‘Sudan Conflict Summary’’. Uppsala Conflict Data Program.

Department of Peace and Conflict Research, Uppsala University [cited

24 May 2006]. Available at www.pcr.uu.se/database/.

2005. Uppsala Conflict Data Program. Department of Peace and Conflict

Research, Uppsala University [cited June 2, 2005]. Available at

www.pcr.uu.se/database/.

United Nations. 2001. Letter dated 12 February 2001 from the Secretary-

General addressed to the President of the Security Council.

Uvin, Peter. 1998. Aiding Violence: The Development Enterprise in Rwanda.

West Hartford, Conn.: Kumarian Press.

Uvin Peter, and Charles Mironko. 2003. ‘‘Western and Local Approaches to

Justice in Rwanda.’’ Global Governance 9 (2): 129–231.

280 References

van Biezen, Ingrid. 2003. Political Parties in New Democracies: Party

Organization in Southern and East-Central Europe. Houndmills:

Palgrave Macmillan.

Van Tongeren, Paul, Malin Brenk, Marte Hellema, and Juliette Verhoeven,

eds. 2005. People Building Peace II: Successful Stories of Civil Society.

Boulder, Colo. and London: Lynne Rienner Publishers.

Vanberg, Victor, and James Buchanan. 1989. ‘‘Interests and Theories in

Constitutional Choice.’’ Journal of Theoretical Politics 1: 49–62.

Varshney, Ashutosh. 2002. Ethnic Conflict and Civic Life: Hindus and

Muslims in India. New Haven: Yale University Press.

2001. ‘‘Ethnic Conflict and Civil Society.’’ World Politics 53 (April):

362–398.

Viggo Jakobsen, Peter. 2003. ‘‘Military Forces and Public Security

Challenges.’’ In The United Nations and Regional Security: Europe

and Beyond, ed. M. Pugh and W. P. Singh Sidhu. Boulder, Colo. and

London: Lynne Rienner Publishers (137–153).

Vinegard, Anna. 1998. ‘‘From Guerrillas to Politicians: The Transition

of the Guatemalan Revolutionary Movement in Historical and

Comparative Perspective.’’ In Guatemala after the Peace Accords,

ed. Rachel Sieder. London: Institute of Latin American Studies

(207–227).

Vines, Alex. 1996. Renamo: From Terrorism to Democracy in Mozambique?

London: James Currey.

Wade, Christine. 2007. ‘‘El Salvador: The Electoral Success of the FMLN.’’ In

From Soldiers to Politicians: Transforming Rebel Movements after War,

ed. Jeroen de Zeeuw. Boulder, Colo.: Lynne Rienner Publishers (33–54).

Wallensteen, Peter 2007. Understanding Conflict Resolution, 2nd edn.

London: Sage; 1st edn 2002.

2002. Understanding Conflict Resolution. War, Peace and the Global

System. London: Sage Publications.

Wallensteen, Peter, and Carina Staibano, eds. 2005. International Sanctions:

Between Words and Wars in the Global System. London and New York:

Frank Cass.

Walter, Barbara F. 2002. Committing to Peace: The Successful Settlement of

Civil Wars. Princeton: Princeton University Press.

1999. ‘‘Designing Transitions from Civil War: Demobilization,

Democratization, and Commitments to Peace.’’ International Security

24 (1): 127–155.

Wanis-St. John, Anthony, and Darren Kew. 2006. ‘‘The Missing Link? Civil

Society, Peace Negotiations: Contributions to Sustained Peace.’’ Paper

presented at the 47th Annual Convention of the International Studies

Association, San Diego, March 23.

References 281

Wantchekon, Leonard. 2004. ‘‘The Paradox of ‘Warlord’ Democracy: A

Theoretical Investigation.’’ American Political Science Review 98 (1)

(February): 17–33.

Wantchekon, Leonard, and Zvika Neeman. 2002. ‘‘A Theory of Post-Civil

War Democratization.’’ Journal of Theoretical Politics (14) 439–464.

Ward, Michael D., and Kristian S. Gleditsch. 1998. ‘‘Democratizing for

Peace.’’ American Political Science Review 92 (1): 51–61.

Waterman, Harvey. 1993. ‘‘Political Order and the ‘Settlement’ of Civil

Wars.’’ In Stopping the Killing: How Civil Wars End, ed. Roy

Licklider. New York: New York University Press (292–302).

Weinstein, Jeremy M. 2005. ‘‘Autonomous Recovery and International

Intervention in Comparative Perspective.’’ Working Paper, Washington,

DC: Center for Global Development.

Whitehead, Laurence. 2002. Democratization: Theory and Practice. Oxford:

Oxford University Press.

Wilkinson, Steven I. 2004. Votes and Violence: Electoral Competition and

Ethnic Riots in India. Cambridge: Cambridge University Press.

Williamson, Arthur, Duncan Scott, and Peter Halfpenny. 2000.

‘‘Rebuilding Civil Society in Northern Ireland: The Community and

Voluntary Sector’s Contribution to the European Union’s Peace and

Reconciliation District Partnership Programme.’’ Policy & Politics

28 (1): 49–66.

Wolfsfeld, Gadi. 2004. Media and the Path to Peace. Cambridge: Cambridge

University Press.

2003. ‘‘The Role of the News Media in Peace Negotiations: Variations

over Time and Circumstance.’’ In Contemporary Peacemaking:

Conflict, Violence and Peace Processes, ed. J. Darby and R. Mac

Ginty. Basingstoke: Palgrave Macmillan (87–100).

2001. ‘‘The News Media and Peace Processes: The Middle East and

Northern Ireland.’’ Peaceworks (37). Washington, DC: United States

Institute of Peace (1–55).

Woodward, Susan L. 2002. ‘‘Economic Priorities for Successful Peace

Implementation.’’ In Ending Civil Wars. The Implementation of Peace

Agreement, ed. John Stephen Stedman, Donald Rothchild and Elisabeth

M. Cousens. Boulder, Colo.: Lynne Rienner Publishers (183–214).

World Bank 2003a. Breaking the Conflict Trap: Civil War and Development

Policy. Washington DC: World Bank and Oxford University Press.

2003b. Working Together: World Bank-Civil Society Relations.

Washington, DC: IBRD/World Bank. Available at: http://siteresources.

worldbank.org/CSO/Resources?WorkingTogetherBrochure.pdf.

2002a. The Right to Tell: The Role of Mass Media. Washington DC:

World Bank.

282 References

2002b. Local Level Institutions and Social Capital Study, vol. I, June,

Washington, DC: IBRD/World Bank.

Zahar, Marie-Joelle. 2005a. ‘‘Violence and War-to-Peace Transitions:

A Second Look at the ‘Spoiler’ Debate in Light of African Case

Studies.’’ Paper presented at 101st Annual Meeting of the American

Political Science Association, September 1–4, 2005, Washington, DC.

2005b. ‘‘Power Sharing in Lebanon: Foreign Protectors, Domestic Peace,

and Democratic Failure.’’ In Sustainable Peace. Power and Democracy

After Civil Wars, ed. Philip G. Roeder and Donald Rothchild. Ithaca and

London: Cornell University Press (219–240).

2003. ‘‘Reframing the Spoiler Debate in Peace Processes.’’ In

Contemporary Peacemaking: Conflict, Violence and Peace Processes,

ed. John Darby and Roger Mac Ginty. Basingstoke: Palgrave Macmillan

(114–124).

Zartman, I. William, ed. 1995. Elusive Peace: Negotiating an End to Civil

Wars. Washington, DC: Brookings Institute.

1991. ‘‘Common Elements in the Analysis of a Negotiation Process.’’ In

Negotiation Theory and Practice, ed. J. William Breslin and Jeffrey Z.

Rubin. Cambridge, Mass.: Program on Negotiation, Harvard Law

School (147–159).

References 283

Index

Page numbers followed by n indicate footnotes.

Abuja accords, 152–153accountability

effect of foreign peacekeeping on,45–46

electoral systems and, 172Aceh, 180Afghanistan, 31–32, 171, 174African National Congress (ANC),

115–116amnesty for war crimes, 142–145,

154–155Angola, 25–26, 42, 114, 118–119Annan, Kofi, 5, 27armed conflict

defined, 4[n2 ]armed forces, demobilization, 27armed groups. See rebels and

insurgentsAung San Suu-kyi, 227–228authority issues. See legitimacy;

systemic dilemma

Bosnia and Herzegovina, 24, 31–32, 42,91, 105, 111–112, 125, 127,151, 160, 167, 201, 221–222

Bougainville, 158, 169–170Boutros-Ghali, Boutros, 27, 239Burma, 227–228Burundi, 22, 111–112, 116–117,

128–129Bush administration, 162

Cambodia, 22, 25–26, 69–70, 122, 145,165–166

cease-fires, 96–97Central America, 149

civil liberties, 64–65, 75civil society

blurring between governmentand, 201

dilemmas of democratization, 23,183–184, 194–207

exclusion from power sharing, 23international encouragement of,

201–207, 209–210in peace process, 189, 191–199as pro-democracy ideal, 184–186,

189strategy for promoting, 208–210uncivil and violent forms, 186–188,

189–190, 191–194in war-to-democracy transitions,

186, 187–190, 199–200,255–256

civil wardefined, 4[n 2]effect of victory on democratization,

62legacy of violence, 31–32partition of country following, 2–3

Clinton, Bill, 162Clinton administration, 167Colombia, 121[n20]colonies, 63–64Congo. See Democratic Republic of the

Congo (DRC)consociational democracy, 110–111,

122–123constitutions, 10, 105coordination problems, 6–7credible commitments

to democratic rules, 10

284

peace agreements, 108, 252–253,256–259

crime, 83, 89–90, 93–94, 101See also amnesty for war crimes;

human rights abusesCroatia, 25–26, 201Cyprus, 23–24, 119–120, 126, 128

Darfur agreement, 22, 117Dayton Peace Agreement, 91, 105,

111–112, 195–196, 221–222demobilization, 27‘‘democratic peace,’’ 5, 29, 162–163, 214Democratic Republic of the Congo

(DRC), 22, 23, 98, 124–125,127, 198, 259

democratic transition theory, 1democratization

assessment approaches, 250–252civil war protagonists’ views, 7–10as conducive to conflict and violence,

12, 28–31, 80–84, 101consociational democracy, 110–111,

122–123country rankings, 25–26defined, 17, 32impact of peacekeeping

case studies, 69–71overview, 40, 44–48, 74–77short- versus long-term

effects, 51–52, 64–74, 75–76,256–259

statistical analysis: approaches andmethodology, 48–54, 55–64,75–76

statistical analysis: findings,54–55, 64–74

international actors’ support for, 4–7,256–259

international versus local ownership,24–25, 220

peace as precondition, 3–4, 5, 26, 46,256–259

peace through, 3–5, 7–10, 181,239–241

political parties as agents, 139–141power sharing and, 122–129, 161See also democratization crises;

democratization versuspeacebuilding dilemmas

democratization crisesBosnia and Herzegovina, 221–222electoral violence and irregularities,

223–228forms of international intervention,

229, 256–259international versus domestic

control over, 6, 219–223,231–236

Lebanon, 224–225, 232–233timing and duration of intervention,

223–231, 236, 256–259triggers for international

engagement, 217–219,222–223, 236

types, 217–219Uganda, 216–217, 220–221, 224,

229–230, 233–235United Nations role in, 214–215,

219–221, 237–238democratization versus peacebuilding

dilemmasinternational actors’ perspectives,

4–7need for trade-offs, 17–18, 35–36,

243overview, 1–2, 4–7, 10–13, 17–20,

239–241, 250peacebuilding given priority, 18, 26,

35–36policy recommendations, 250–259protagonists’ perspectives, 7–10See also horizontal dilemma; systemic

dilemma; temporal dilemma;vertical dilemma

disarmament, 27

East Timor, 3, 70–71, 86, 171, 174efficacy versus legitimacy. See vertical

dilemmaEl Salvador, 25–26, 42, 90,

93, 149elections

contradictory goals and outcomes,157–159, 161–163, 180–181

democratic dilemmas, 159–161,165, 172

and democratization crises, 223–228electoral administration, 166,

175–177

Index 285

elections (cont.)electoral systems, 172–175electoral violence, 84–87, 225as indicators of democratization, 223international support for, 161–163,

165–166, 181, 223–228minimal security as precondition,

35–36outbidding in, 31–32, 160–161sequencing of local and national, 170timing, 25, 91, 154, 167–168in war-to-democracy transitions,

84–87See also referendums

elite-mass dilemmas. See verticaldilemma (legitimacy versusefficacy dilemma)

Eritrea, 3Ethiopia, 87ethnic conflict

Burundi, 128–129and later democratization, 56managed through power sharing,

110, 111–112ethnic representation, 127–129

See also consociational democracyexclusion. See horizontal dilemma;

power sharing

Fiji, 215force

controlling violence with, 94–96legitimacy, 97peacekeepers’ use, 6, 95–96

GDP per capita, and democratization,57–61

genocide, 88, 145Germany, 213great powers, 63Guatemala, 42, 90, 95, 149, 198–199guerillas. See rebels and insurgents

Hamas, 22, 151, 160Haradinaj, Ramush, 143–144Hariri, Rafik, 224hate propaganda, 88Herzegovina. See Bosnia and

HerzegovinaHizbollah, 120–121

horizontal dilemma (inclusion versusexclusion dilemma)

civil society, 23, 183–184definition and overview, 11, 21–23,

243–245power sharing, 107, 115–117,

124–125, 131rebel groups in post-war politics, 22,

81, 91, 98, 99, 141–145human rights abuses, 92–93, 94,

142–145, 188Human Rights Council, 238human rights networks, 204–205

illiteracy, 61inclusion. See horizontal dilemma;

power sharing; rebels andinsurgents

India, 189Indonesia, 177, 180insurgents. See rebels and insurgentsinternational actors

coordination problems among, 6–7democratization agendas, 3–5, 222,

256–259and the democratization vs.

peacebuilding dilemma, 3–5,24–25, 201–205

versus local control, 7, 24–25, 44–45myopic time frame, 205–207,

256–259in the peacebuilding process, 4–7,

24–25, 76, 95–96promotion of civil society, 201–205,

209–210, 255–256promotion of parties and elections,

149–151, 161–163, 165–166,177–180

responses to democratizationcrises

duration of intervention, 228–231legitimacy of intervention, 6,

219–236range of bodies involved, 222timing of intervention, 223–228,

230–231, 236triggering events, 217–219,

222–223, 236types of intervention measures,

229, 231–236

286 Index

role in power-sharing agreements,107, 119 – 120, 126 –127 ,131– 132

role in violence management, 81,98, 99

international criminal courts, 143Iraq, 23– 24, 171 , 174, 239[n 1]Irish Republican Army (IRA),

96– 97, 116Israel, 93Ivory Coast, 124 –125 , 198

justice, 142– 145

Khmer Rouge, 145Kosovo, 25–26, 71, 88, 95 –96, 126–12 7,

143–144, 148, 155, 171

Lebanon, 110 , 120– 121, 224 –225 ,232– 233

legitimacyformer armed groups as parties,

149– 151international criminal tribunals, 143international intervention, 6 ,

219– 223 , 231– 236peace process, 22, 23–24, 152 –153power-sharing arrangements,

118– 119 , 125– 126via elections and referendums, 22,

23– 24, 25, 164violence and crime as challenge,

93– 94Liberia, 23, 91, 124 – 125, 152– 153 ,

154, 158 , 173, 235 , 259literacy, and democratization, 61Lome accord, 121– 122, 150 –151Lusaka accord, 118– 119

Maori, 118 [n16], 129media, and violence, 87–88, 89, 90– 91military coups, UN reaction, 215Milosevic, Slobodan, 182, 196 , 226Mozambique, 42, 148 , 179Museveni, Yoweri, 216– 217, 229 –230 ,

231, 233Myanmar, 227– 228

Namibia, 25– 26, 40, 42Nepal, 125

New Zealand, 118 [n16], 129NGOs, 206– 207

See also civil societyNorthern Ireland, 94, 95, 96– 97, 116,

126 , 189, 190 , 192, 196

oil resources, 63

Palestinian territories, 151, 160Papua New Guinea, 42, 158, 169– 170Paris agreement, 145partition, 2– 3, 110peace

as precondition for democracy, 46through democratization, 1 – 2, 3 –5 ,

7 – 10, 27, 161– 163, 181 , 214,239 – 241

peace agreementscredible commitments, 108 ,

252 – 253, 256– 259domestic legitimacy, 5effect on democratization, 62failure to end conflict, 8– 10list of agreements/accord s, 32– 35,

111 – 113negotiations

conditions for entering, 136– 138critical issues, 241–242inclusion of warring parties,

21–23, 96–97, 105–106participation of civil society,

194–199protagonists’ perspectives, 7–10provision for democratization, 3–4uncertainty for participants, 8–10

peacebuildingconcept, 1, 17, 17[n1 ], 32, 239– 241versus democratization (See

democratization versuspeacebuilding dilemmas;peacekeeping: impact ondemocratization)

given precedence over justice,142–145

negative effects on democratization,26–28

policy recommendations, 250–259recent cases, 32–35risks for protagonists, 8role of civil society, 189

Index 287

Peacebuilding Commission, 27, 237 –238peacekeeping

defined, 41effectiveness, 39, 52–54, 74, 76– 77,

256 –259impact on democratization

case studies, 69– 71overview, 40, 44– 48, 74– 77short- versus long-term effects,

51– 52, 64– 74, 75– 76, 256 –259statistical analysis: approach and

methodology, 48– 54, 55– 64,75– 76

statistical analysis: findings,54– 55, 64– 74

legitimacy and accountability issues,44– 46, 74

types of missions, 41– 43, 65– 69,74–75

UN vs. non-UN, 43[n7 ], 71– 74, 75use of force in, 6, 94–96

police reform, 89–90political parties

as agents for democratization,139–141, 177–178, 179

domestic versus internationallegitimization, 149–151

ethnic versus multiethnic,177–178, 179

former armed rebels as, 145–149,179–180

international promotion of, 177–180one-party tendencies, 145–147role in electoral administration,

175–177sedimentation, 141–142, 178–179weaknesses in new democracies,

140–141political rights, 64–65, 75power sharing

arguments for, 105–106, 161changing power relations under,

118–119, 120–122defining the included/excluded

groups, 115–117, 124–125,127–129

democratization versuspeacebuilding dilemmas, 21–23,106–107, 115–117, 122–129,131–132

guidelines, 132–133international mediation, 119–120,

126–127list of power-sharing accords,

111–113meaning, 108–113partial inclusion, 116–117popular support for, 125–126protagonists’ views, 8recent agreements, 111–113as source of conflict and violence,

106–107, 114–117, 119,120–122, 130–132

sunset clauses, 121–122, 254‘‘projectism,’’ 207proportional representation,

172–174, 175

rebels and insurgentsamnesty for, 154–155commitment to democratic politics,

145–149contraband financing, 63inclusion in new regime

arguments for, 96, 136–138, 153authoritarian tendencies, 145–147continuation of wartime

allegiances, 141–142,145–149, 155

democratization dilemmas, 21–23,81, 91, 98, 99, 141–153

domestic versus internationalauthorization, 149–151,155–156

horizontal dilemma, 21–23, 81, 91,98, 99, 115–117, 141–145

impact on rule of law, 142–145in police and security forces,

91–92as political parties, 145–149short- versus long-term effects,

152–153, 156reconciliation commissions, 143referendums, 23–24, 125–126,

168–169rule of law, 142–145, 188Rwanda, 25–26, 88, 201

sanctions, international, 232,234–235

288 Index

security issues: in war-to-democracytransition, 10, 11– 12, 35

security sector reform (SSR), 89– 90,91– 92

sedimentation of wartime divisions,141– 142 , 154, 204

Serb Republic of Bosnia, 201Serbia, 182 , 226short- versus long-term dilemma. See

temporal dilemmaSierra Leone, 42, 114, 121 – 122,

144– 145 , 150– 151single non-transferable vote, 172 –174 ,

175Sinn Fein, 96– 97, 116social capital, 185 , 190– 191South Africa, 115 –116 , 121, 127 , 189,

196– 197Soviet Union, 3spoilers, 23, 84, 96, 98Sri Lanka, 97, 101 , 197– 198Sudan, 22, 117, 124 , 125– 126sunset clauses, 121– 122 , 254systemic dilemma (local versus

international control)civil society building, 183– 184 ,

201– 205definition and overview, 11,

24– 25, 248democratization crises, 219– 223, 236former armed rebels in politics,

149– 151peacekeeping, 44– 46, 74power sharing, 107, 126 – 127,

131– 132violence management, 81, 98, 99

Tajikistan, 95Taylor, Charles, 152– 153temporal dilemma (short- versus long-

term dilemma)civil society building, 183– 184 ,

205– 207definition and overview, 11, 25, 249democratization crises, 223– 226,

228– 231 , 236election timing, 90– 91, 166– 171former armed rebels in politics,

152– 153media reform, 90–91, 99

peacekeeping, 48, 64–74power sharing, 120 , 132security sector reform (SSR),

91– 92, 99violence management, 81

terrorist organizations, 22Thailand, 215Timor Leste . See East Timortolerance, 185transitional arrangements, 121– 122,

154 , 177transnational democracy networks,

204 – 205trust, 186, 190truth and reconciliation

commissions , 143

Uganda, 216– 217, 220 –221 , 224,229 – 230, 233– 235

Ukraine, 224uncertainty dilemmas, 8– 10, 165United Nations

Cambodia, 69– 70, 165– 166East Timor, 70–71election supervision, 173Kosovo, 71, 126– 127peacebuilding agendas, 27, 230,

237 – 238, 239peacekeeping missions, 43[n7 ]reaction to armed conflicts, 214– 215reaction to democratization crises,

214 – 215, 219– 221 , 237– 238short-term project orientation, 207Uganda, 221UN vs. non-UN peacekeeping,

43[n 7], 71– 74, 75use of force by peacekeepers, 6

vertical dilemma (legitimacy versusefficacy dilemma)

civil society building, 194–199crises of democratization, 237definition and overview, 11, 23–24,

245–248election timing, 91–92inclusion of former armed groups,

145–149power sharing, 107, 125–126, 131security sector reform, 91–92violence management, 81, 97, 99

Index 289

victory, effect on democratization,61–62

violencedemocracy as conducive to, 29–31,

80–81, 101and dilemmas of democratization,

81, 90–92, 97–101electoral, 84–87, 225legacy of, 31–32management of

conditions for success, 100,252–253

dilemmas, 97–101negative effects on

democratization, 80, 92–96through inclusion of spoilers, 96use of force to curb, 94–96

media as instigator, 87–88, 89, 90–91political versus criminal, 83, 89–90under power-sharing arrangements,

106–107, 114–117, 119,120–122, 130–132

security sector reform (SSR) and,89–90, 91–92

in war-to-democracy transitions, 12,31–32, 80–84, 92–101

war as spur to democratization, 46,61–62

war-to-democracy transitionscivil society in, 186, 187–190,

199–200, 255–256concept and processes, 20–21international engagement, 2–3, 4–7,

95–96, 256–259protagonists’ perspectives, 7–10security issues, 10, 11–12, 35uncertainty dilemmas, 8–10violence in, 12, 31–32, 80–84,

92–101See also democratization;

democratization versuspeacebuilding dilemmas; peaceagreements; peacebuilding

warring parties. See rebels andinsurgents

Western Sahara, 42women’s groups, 189

Yugoslavia, 3, 143–144, 155, 182,226

Zimbabwe, 258

290 Index