Upload
neretva
View
81
Download
2
Embed Size (px)
Citation preview
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