Post-Conflict Economies in Africa Edited by
Augustin Kwasi Fosu and Paul Collier
Post-Conflict Economies in Africa Th...
29 downloads
1272 Views
1MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Post-Conflict Economies in Africa Edited by
Augustin Kwasi Fosu and Paul Collier
Post-Conflict Economies in Africa This is IEA conference volume no. 140
This page intentionally left blank
Post-Conflict Economies in Africa Edited by
Augustin Kwasi Fosu African Economic Research Consortium, Nairobi and Economic Commission for Africa, Addis Ababa
and
Paul Collier St Antony’s College, Oxford, and Centre for the Study of African Economies University of Oxford
in association with the International Economic Association
© International Economic Association 2005 All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication may be reproduced, copied or transmitted save with written permission or in accordance with the provisions of the Copyright, Designs and Patents Act 1988, or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, 90 Tottenham Court Road, London W1T 4LP. Any person who does any unauthorized act in relation to this publication may be liable to criminal prosecution and civil claims for damages. The authors have asserted their rights to be identified as the authors of this work in accordance with the Copyright, Designs and Patents Act 1988. First published 2005 by PALGRAVE MACMILLAN Houndmills, Basingstoke, Hampshire RG21 6XS and 175 Fifth Avenue, New York, N.Y. 10010 Companies and representatives throughout the world PALGRAVE MACMILLAN is the global academic imprint of the Palgrave Macmillan division of St. Martin’s Press, LLC and of Palgrave Macmillan Ltd. Macmillan® is a registered trademark in the United States, United Kingdom and other countries. Palgrave is a registered trademark in the European Union and other countries. ISBN 1–4039–4346–X This book is printed on paper suitable for recycling and made from fully managed and sustained forest sources. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication Data Post-conflict economies in Africa / edited by Augustin Kwasi Fosu and Paul Collier. p. cm. — (International Economic Association series) Originally presented at a workshop in Kampala, Uganda, in July 2002 in colloboration with the World Bank and the International Economic Association. Includes bibliographical references and index. ISBN 1–4039–4346–X (cloth) 1. Africa—Economic policy—Congresses. 2. Political stability—Africa— Congresses. 3. Africa—Politics and government—Congresses. I. Fosu, Augustin Kwasi. II. Collier, Paul. III. World Bank. IV. International Economic Association. V. Series. HC800.P677 2004 330.96—dc22 2004054466 10 9 8 7 6 5 4 3 2 1 14 13 12 11 10 09 08 07 06 05 Printed and bound in Great Britain by Antony Rowe Ltd, Chippenham and Eastbourne
Contents
The International Economic Association
vii
Preface and Acknowledgements
ix
List of Participants and Contributors
x
List of Abbreviations and Acronyms
xii
Introduction: Post-Conflict Economies in Africa Augustin Kwasi Fosu
xv
Part I
Thematic Issues
1 Post-Conflict Reconstruction in Africa: Some Analytical Issues Tony Addison and S. Mansoob Murshed 2 The Economic and Political Consequences of Conflict and Implications for Post-Conflict Recovery in Africa Ibrahim A. Elbadawi and Njuguna S. Ndung’u
3
18
3 Economic Policy in Post-Conflict Societies Paul Collier
45
4 Ethnicity, Institutions of Governance and Conflict Avoidance Mwangi S. Kimenyi
57
Part II
Case Studies: West Africa
5 Liberia and Sierra Leone: Interwoven Civil Wars Victor A. B. Davies
77
6 The Nigerian Civil War: Causes and the Aftermath Abdul-Ganiyu Garba and P. Kassey Garba
91
7 The Economics of Civil Conflict in Africa: The Case of Chad Nadjiounoum Djimtoingar and Djona Atchénémou Avocksouma
Part III
109
Case Studies: The Horn, East and Central Africa
8 Conflict, Post-Conflict and Economic Performance in Ethiopia Alemayehu Geda and Befekadu Degefe v
125
vi
Contents
9 Prospects for Sustainable Peace and Post-Conflict Economic Growth in the Sudan Ali Abdel Gadir Ali and Ibrahim A. Elbadawi 10 Entrenching Peace in Post-Conflict Economies: The Case of Uganda Michael Atingi-Ego and Rachel Kaggwa Sebudde 11 The Challenge of Transition from War to Peace in Burundi Janvier Désiré Nkurunziza and Floribert Ngaruko
Part IV
163 180
Case Studies: Southern Africa
12 The Political Economy of Post-Conflict Economic Recovery: Making Peace Substantial in Mozambique Laudemiro Francisco and Edwin Connerley 13 Transformation for Post-Conflict Angola Fátima Moura Roque
Part V
143
197 213
Conclusion
14 Post-Conflict Economies in Africa: Synthesis and Lessons Augustin Kwasi Fosu
231
Index of Names
241
Subject Index
245
The International Economic Association A non-profit organization with purely scientific aims, the International Economic Association (IEA) was founded in 1950. It is a federation of some sixty national economic associations in all parts of the world. Its basic purpose is the development of economics as an intellectual discipline, recognizing a diversity of problems, systems and values in the world and taking note of methodological diversities. The IEA has, since its creation, sought to fulfil that purpose by promoting mutual understanding among economists through the organization of scientific meetings and common research programmes, and by means of publications on problems of fundamental as well as of current importance. Deriving from its long concern to assure professional contacts between East and West, and North and South, the IEA pays special attention to issues of economies in systemic transition and in the course of development. Since its inception it has organized more than a hundred round-table conferences for specialists on topics ranging from fundamental theories to methods and tools of analysis and major problems of the present-day world. Participation in round tables is at the invitation of a specialist programme committee, but thirteen triennial World Congresses have regularly attracted the participation of individual economists from all over the world. The Association is governed by a Council, composed of representatives of all member associations, and by a fifteen-member Executive Committee elected by the Council. The Executive Committee (2002–2005) at the time of the Lisbon Congress was: President Vice-President Treasurer Past President President-Elect Other Members
Professor János Kornai, Hungary Professor Bina Agarwal, India Professor Jacob Frenkel, Israel Professor Robert Solow, USA Professor Guillermo Calvo, Argentina Professor Maria Augusztinovics, Hungary Professor Eliana Cardoso, World Bank Professor Duardo Engel, Chile Professor Heba Handoussa, Egypt Professor Michael Hoel, Norway Professor Jean-Jacques Laffont, France Professor Andreu Mas-Colell, Spain Professor Kotaro Suzumura, Japan vii
viii
The International Economic Association
Advisers Secretary-General General Editor
Professor Alessandro Vercelli, Italy Professor Fiorella Kostoris Padoa Schioppa, Italy Professor Vitor Constancio, Portugal Professor Jean-Paul Fitoussi, France Professor Michael Kaser, UK
Sir Austin Robinson was an active Adviser on the publication of IEA Conference proceedings from 1954 until his final short illness in 1993. The Association has also been fortunate in having secured many outstanding economists to serve as President: Gottfried Haberler (1950–53), Howard S. Ellis (1953–56), Erik Lindahl (1956–59), E. A. G. Robinson (1959–62), Ugo Papi (1962–65), Paul A. Samuelson (1965–68), Erik Lundberg (1968–71), Fritz Machlup (1971–74), Edmund Malinvaud (1974–77), Shigeto Tsuru (1977–80), Victor L. Urquidi (1980–83), Kenneth J. Arrow (1983–86), Amartya Sen (1986–89), Anthony B. Atkinson (1989–92), Michael Bruno (1992–95), Jacques Drèze (1995–99) and Robert M. Solow (1999–2002). The activities of the Association are mainly funded from the subscriptions of members and grants from a number of organizations. Support from UNESCO since the Association was founded, and from its International Social Science Council, is gratefully acknowledged.
Preface and Acknowledgements In recent years there has been considerable conflict in the form of civil wars across the African continent. To inform policy-makers and analysts, the African Economic Research Consortium (AERC) has focused on post-conflict economies as part of its research portfolio under the Special Workshop (SW) modality, which provides research on a topical policy issue for dissemination in a relatively short period of time. The chapters in this volume were originally presented as papers at a Special Workshop in Kampala in July 2002, the result of collaboration with the World Bank, which has also worked extensively on conflict, and the International Economic Association (IEA), which provides a forum for the analysis and dissemination on important economic issues of international relevance. The volume consists of five parts comprising a set of thematic papers to inform the reader of the various pertinent issues on post-conflict economics; a number of country case studies drawn from all regions of sub-Saharan Africa, in order to shed light on the Africa-specific setting; and a section on conclusions. Thus the volume uniquely and richly provides in a complementary format both theoretical and empirical analyses. Many individuals and organizations have contributed to the success of the project. However, specific mentions are in order. First and foremost, the AERC, through its Advisory Committee, has been instrumental in identifying the topic and executing the research project. Second, the collaboration of the World Bank has been important in bringing the project to fruition. Third, the IEA has served as an important partner through which the quality of the product has been enhanced and the output disseminated appropriately. Much gratitude is due to Jean-Paul Fitoussi of the IEA and the Scientific Program Committee, comprising Eliana Cardoso, Paul Collier, Augustin Fosu (chair) and Mustapha Nabli. Further thanks are due to Heywood Hadfield and Michael Kaser, IEA General Editor, who supervised the editorial process. It is hoped that the present volume will contribute significantly to the salient policy dialogue on how to achieve an enduring post-conflict economy for growth and development, particularly in Africa, where the challenge seems so enormous and the imperative so real.
ix
List of Participants and Contributors
*Professor Tony Addison, UNU/WIDER, Helsinki, Finland *Professor Alemayehu Geda, Addis Ababa University, Ethiopia, and School of Oriental and African Studies, London, UK *Dr Ali Abdel Gadir Ali, Arab planning Institute, Kuwait *Dr Michael Atingi-Ego, Bank of Uganda, Kampala, Uganda *Professor Djona Atchénémou Avocksouma, University of N’Djamena, Chad *Professor Befekadu Degefe, Centre for Economic Studies and Policy Analysis, Ethiopia Professor Eliana A. Cardoso, Georgetown University, Washington, DC, USA *Professor Paul Collier, St Antony’s College, Oxford, UK and the Centre for the Study of African Economies, Oxford, UK *Dr Edwin Connerley, United States Agency for International Development (USAID), Washington, DC, USA *Mr Victor A. B. Davies, Oxford University, UK Mr Nadjiounoum Djimtoingar, Ministère du Plan, du Développement et de la Coopération, N’Djamena, Chad *Dr Ibrahim A. Elbadawi, World Bank, Washington, DC, USA *Professor Augustin Kwasi Fosu, African Economic Research Consortium (AERC), Nairobi, Kenya, and Economic Commission for Africa, Addis Ababa, Ethiopia *Mr Laudemiro Francisco, Howard University, Washington, DC, USA *Dr Abdul-Ganiyu Garba, Ahmadu Bello University, Zaria, Nigeria *Dr P. Kassey Garba, University of Ibadan, Nigeria Professor Mukwanason A. Hyuha, African Economic Research Consortium (AERC), Nairobi, Kenya Ms Darlison Kaija, Makerere University, Kampala, Uganda Ms Rehema Kakonge, Uganda Manufacturers’ Association, Kampala, Uganda *Professor Mwangi S. Kimenyi, Kenya Institute for Public Policy Research and Analysis (KIPPRA), Nairobi, Kenya Mr Rubain Batila Kissita, Université Marien Ngouabi, Congo Brazzaville Professor William Lyakurwa, African Economic Research Consortium (AERC), Nairobi, Kenya *Professor Fátima Moura Roque, Gabinete de Estudos Africano, Lusófona University, Portugal Mr Paul Mpuga, Makerere University, Kampala, Uganda Mr Adam Mugume, Makerere University, Kampala, Uganda, and Oxford University, UK x
List of Participants and Contributors
xi
*Dr Syed Mansoob Murshed, Institute of Social Studies (ISS), The Hague, The Netherlands Dr Ngila Mwase, UNDP, Kampala, Uganda Professor Francis Mwega, African Economic Research Consortium (AERC), Nairobi, Kenya Ms Floribert Ngaruko, Université de Nice-Sophia Antipolis, France *Professor Njuguna S. Ndung’u, International Research Centre (IDRC), Nairobi, Kenya *Mr Janvier Désiré Nkurunziza, Harvard University, USA Dr Marios Obwona, Makerere University, Kampala, Uganda Mr John Okidi, Executive Director, Economic Policy Research, Kampala, Uganda Mr Francis Okurut, Uganda Women’s Finance Trust, Kampala, Uganda Mr Nichodemus Rudaheranwa, Makerere University, Kampala, Uganda The Honourable Dr Abel Rwendeire, Minister of State for Trade and Industry, Government of Uganda *Ms Rachel Kaggwa Sebudde, Bank of Uganda, Kampala, Uganda Dr Germina Ssemogerere, Makerere University, Kampala, Uganda *Authors of chapters
List of Abbreviations and Acronyms
ADF AERC AG AIDS AMODEG ANC APC ATU CBE CCFAN CCM CCSRP CFA cif CNDD CNPC CNR CPC CPI CPIA CSIS CSM CSNPD DDR DfID DP DP DRC DUP ECOMOG ECOWAS EPI EPLF EPRDF
Allied Democratic Forces (Uganda) African Economic Research Consortium Action Group (Nigeria) acquired immunodeficiency syndrome Association of Demobilized War Veterans (Mozambique) African National Congress (South Africa) All People’s Congress (Sierra Leone) Anti-Terrorist Unit (Liberia) Commercial Bank of Ethiopia Northern Armed Forces Command Council (Chad) Mozambican Christian Council Petroleum Revenue Control and Surveillance Assembly (Chad) Communauté Financière d’Afrique cost, insurance and freight Conseil National pour la Défense de la Démocratie (Burundi) China National Petroleum Corporation National Recovery Committee (Chad) conflict and post-conflict country Investment Promotion Centre (Mozambique) Country Policy and Institutional Assessment (World Bank) Center for Strategic and International Studies (United States) Supreme Military Council (Chad) National Committee for Peace and Democracy (Chad) disarmament, demobilization and resettlement Department for International Development (UK) Democratic Party (Kenya) Democratic Party (Uganda) Democratic Republic of the Congo Democratic Unionist Party (Sudan) Economic Community of West Africa Monitoring Group Economic Community of West African States elite political instability Eritrean People’s Liberation Front Ethiopian People’s Revolutionary Democratic Front xii
List of Abbreviations and Acronyms
EPRP ESAF FADM FAR FDC FDD FDI FNLA FNL-PALIPEHUTU FORD FRELIMO FRODEBU FROLINAT FRONASA GDN GNPOC GPA GUNT HIV HDI HIPC IAAC ICG IDA IDP IGAD IMF KANU KIPRA KY LRA LTDS LURD MDD MDJT MPLA NAS NCNC NDP NEPAD NGO NIF NPC
xiii
Ethiopian People’s Revolutionary Party Enhanced Structural Adjustment Fund Mozambican Defence Force Forces Armées Rwandaises (Rwanda) Foundation for Democratic Cooperation Forces pour la Défense de la Démocratie (Burundi) foreign direct investment Frente Nacional de Libertação de Angola Front National de Libération-Parti pour la Libération du Peuple Hutu (Burundi) Forum for the Restoration of Democracy (Kenya) Frente de Libertação de Moçambique Front Démocratique Burundi National Liberation Front (Chad) Front for National Salvation (Uganda) Global Data Network The Great Nile Petroleum and Oil Corporation (Sudan) General Peace Agreement (Mozambique) Transitional National Union Government (Chad) human immunodeficiency virus Human Development Index Highly indebted poor country integrated approach to aid coordination (Sierra Leone) International Crisis Group international development assistance internally displaced person Inter-Governmental Agency for Development International Monetary Fund Kenya Africa National Union Kenya Institute for Public Policy Research and Analysis Kabaka Yekka (Uganda) Lord’s Resistance Army (Uganda) long-term development strategy Liberians United for Reconciliation and Democracy Movement for Democracy and Development (Chad) Movement for Democracy and Justice in Chad Movimento Popular para a Libertação de Angola Native Authority System (Nigeria) National Council of Nigeria and the Cameroons National Development Party (Kenya) New Partnership for African Development non-governmental organization National Islamic Front (Sudan) Northern People’s Congress (Nigeria)
xiv
List of Abbreviations and Acronyms
NPFL NPRC NPV NRA NRM NURP OAU OLF PPT-RDA PRC PRE PRES PRS RENAMO RUF SDR SLPP SOD SP SPLA SPLM SSA SSS SSTP TICAD TPLF UN UNAMSIL UNDP UNITA UNLF UNMIL UNOMOZ UPC USA USAID USSR
National Patriotic Front of Liberia National Provisional Ruling Council (Sierra Leone) net present value National Resistance Army (Uganda) National Resistance Movement (Uganda) Northern Uganda Reconstruction Programme Organization of African Unity Oromo Liberation Front Chad Progressive Party People’s Redemption Council (Liberia) Economic Rehabilitation Programme (Mozambique) Economic and Social Rehabilitation Programme (Mozambique) poverty reduction strategy Resistência Nacional Moçambicana Revolutionary United Front (Sierra Leone) special drawing right Sierra Leone People’s Party Special Operations Division (Liberia) Sudan Policy Sudan People’s Liberation Army Sudan People’s Liberation Movement sub-Saharan Africa Special Security Service (Liberia) structural and systemic transformation programme Tokyo International Conference on African Development Tigray People’s Liberation Front United Nations United Nations Mission in Sierra Leone United Nations Development Programme National Movement for the Total Independence of Angola Uganda National Liberation Front United Nations Mission in Liberia United Nations Operation in Mozambique Uganda People’s Congress United States of America United States Agency for International Development Union of Soviet Socialist Republics
Introduction: Post-Conflict Economies in Africa Augustin Kwasi Fosu
1
Introduction
Conflict has occurred throughout history. In Africa, the partition of the continent and creation of modern states using artificial borders often divided peoples and forced ethnically diverse groups together, creating the conditions for additional conflict. Independence was often contemporaneous with ideologogical conflict, as capitalism, socialism and communism competed for supremacy. Post-independence Africa has been dominated by elite political instability (Fosu, 2003). This form of conflict is observed to have contributed significantly to the dismal economic growth and development observed on the continent (Fosu, 2001, 2002, 2003). Following independence, there remained ongoing rebel movements in several African countries. In many countries, however, the dictatorial leaders appear to have succeeded in suppressing the simmering disputes fuelled primarily by inter-ethnic rivalries. Indeed, in some instances, the forging of nation-states through the use of strong central government control probably contributed to neutralizing the pernicious elements of ethnicity (for example, in Côte d’Ivoire, Ghana, Kenya, Malawi, Tanzania and Zambia). Following global political liberalization, however, open conflict in the form of armed insurrection developed in several African countries (for example, Liberia, Sierra Leone, Burundi, Rwanda and, more recently, Côte d’Ivoire). Between 1990 and 2000 there were some nineteen major conflicts in Africa, mainly in the form of civil wars (Wallensteen and Sollenberg, 2001). In this volume the emphasis is on armed insurrections in the form of civil wars and on the economies following them. It is difficult to define a postconflict economy. Nevertheless, the term is maintained in order to stress the importance of attaining such a state, as a prelude to achieving the political stability required for economic growth and development. This volume has two themes, which complement each other. The first section consists of four chapters on thematic issues pertinent to post-conflict economies. What factors are likely to result in conflict resolution in order to achieve the post-conflict state? How can that state be maintained in terms of avoiding the resurgence of conflict? This is followed by country case studies from the sub-Saharan region: West Africa; the Horn, East and Central Africa; and Southern Africa. These cases were chosen to reflect countries where there xv
xvi
Introduction: Post-Conflict Economies in Africa
have been substantial conflicts since independence. In most instances, the countries are at the time of writing experiencing relative peace. In some cases, however, conflict is ongoing, though with some reasonable hope of resolution and with the expectation that the post-conflict state can soon be attained.
2
Thematic issues
Chapter 1: Post-Conflict Reconstruction in Africa: Some Analytical Issues Tony Addison and Mansoob Murshed model the phenomenon of likely credibility problems associated with peace agreements in post-conflict situations in the immediate aftermath of civil war, and the risks of conflict recurring. Where war provides economic gains to one side, peace is not incentive compatible and peace agreements degenerate as they become time inconsistent. The party that has something to gain from surprise warfare will first agree to peace, but then renege. The levels of conflict chosen by this group are an increasing function of greed and capturable rents, but decrease with the group’s perceived cost of war. Furthermore, in a multiple time framework, uncertainty about the group type that may renege on peace generates greater fighting. In particular, higher rates of time preference associated with the impatience to consume immediately engender greater conflict. In this context, action by external powers could lower the risk of resurgence. For example, sanctions, aid and direct intervention can eliminate conflict, as well as helping in devising ‘commitment technologies’ for peace. This analysis is relevant to the recent cases of Angola, Burundi, Liberia and Sierra Leone. Chapter 2: Economic and Political Consequences of Conflict and Implications for Post-Conflict Recovery in Africa Ibrahim Elbadawi and Njuguna Ndung’u explore empirically the channels through which risk is transmitted to conflict and post-conflict economies. Estimating a system of equations that take into account multi-causal structure in the variables, a number of results emerge. First, these ratify and strengthen two salient results in the aid-effectiveness literature, based on single-equation estimation, that a sound policy environment enhances the effectiveness of aid in augmenting growth and reducing poverty, and that the impact of aid is subject to diminishing returns (Collier and Dollar, 2002). Second, that the non-monotonic impact of variables such as aid effectiveness and social fractionalization on the risk of conflict (Collier and Dollar, 2002; Collier and Hoeffler, 1998), and debt on growth,1 are supportable. Third, that public investment is the main channel through which the deleterious effect of conflict is transmitted to the economy. Fourth, that the absence of political rights in socially fractionalized societies exposes such societies to higher risks of conflict. And, finally, that exports tend to increase the risk of
Fosu
xvii
conflict by augmenting state revenues, thus providing increased rent-seeking opportunities.2 Chapter 3: Economic Policy in Post-Conflict Societies Paul Collier details policies which reduce two sets of risks characterizing post-conflict economies: those leading to the outbreak of the conflict in the first place, and those emanating from the conflict itself. Policies to resolve the first set of risks include: • raising the growth rate; • setting up transparent systems of scrutiny and controlling revenues from primary commodities in the short run; • diversifying the economy away from high primary commodity dependence in the long run; and • attenuating ethnic dominance, such as by guaranteeing the rights of minority groups, which may require the support of international intervention. Policies to reduce the second set of risks include: • providing agreed formulae for sharing government revenues – for example, between regions, and for public-sector jobs – such as between ethnic groups; • minimizing the incentives of groups preferring conflict to revert to this, such as converting rebel groups into political parties;3 and • providing a recognized positive role for the diaspora – for example, in nation rebuilding. Collier considers three specific areas likely to be differentially important in post-conflict settings: demobilization; the restoration of the rule of law; and the rehabilitation of the formal economy. Policies to address these include: • ensuring that demobilized soldiers are engaged meaningfully in the economy (such as by providing them with adequate access to land); • re-establishing property rights, followed by the enhancement of legal and judicial systems and of policing; and • the gradual formalization of the economy, including reducing the tax burden on those in the formal sector. According to Collier, external aid and good policy have special roles to play in post-conflict economies. While aid apparently has no direct impact in terms of reducing the risk of conflict recurrence, it can serve an important indirect role through its ability to increase growth, which can in turn decrease the risk as well as provide reassurance to minority groups that their rights are being protected. Further, aid is increasingly effective when the
xviii
Introduction: Post-Conflict Economies in Africa
policy environment is sound (Burnside and Dollar, 2000). Collier also underscores the importance of social policies within the policy environment, such as the inclusiveness of education and health. Similarly, the marginal product of sound policies also increases with the level of aid. Furthermore, the policy environment of embryonic post-conflict economies starts out as fragile but improves with time, so the optimal policy might then entail increasing aid levels over time for post-conflict economies. Additionally, aid could help to reduce the tax burden on those remaining in the shrunken formal sector in post-conflict economies, so that gradual formalization of the economy can proceed more effectively. Chapter 4: Ethnicity, Institutions of Governance and Conflict Avoidance Mwangi Kimenyi highlights the importance of establishing institutions that harmonize ethnic interests. Unfortunately, recent democratization reforms that have introduced political competition are not sufficient to promote conflict-avoidance. On the contrary, political party competition can intensify conflict-generating ethnic competition in highly centralized unitary states. Kimenyi proposes the establishment of decentralized institutions of governance comprising ethnically homogenous local units of collective choice possessing a high degree of autonomy. He suggests that governments based on ethnicity can be complemented by functional institutions transcending the boundaries of ethnicity within a given country.
3
Country case studies
3.1 West Africa Chapter 5: Liberia and Sierra Leone: Interwoven Civil Wars Victor Davies observes that Sierra Leone’s civil war between 1991 and 2001, and Liberia’s civil war from 1989 to 2003 are highly interwoven. He analyses common and idiosyncratic factors causing and sustaining the civil wars in the two bordering West African countries: natural resource abundance, historical legacies, growth collapse, ethnic rivalries, geography and external instigation. After discussing the social and economic consequences, the author examines Sierra Leone’s seemingly successful peace process, highlighting the domestic and external factors that aided it. He also assesses Sierra Leone’s post-conflict initiatives and challenges, especially natural resource management. He then asks why Liberia was at war for so long. Davies attributes much of the blame for the looting and quasi-criminal behaviour to the head of state, concluding that there is no simple solution to the Liberia problem, and that any hope for transition to an enduring, post-conflict peace requires continuous international attention.
Fosu xix
Chapter 6: The Nigerian Civil War: Causes and the Aftermath Based on a concept of conflict as a struggle for supremacy and strategic advantage by antithetical forces, Abdul-Ganiyu Garba and Kassey Garba address three sets of issues: the primary or root causes of the Nigerian civil war; the duration and cessation of the war; and the aftermath and risks of recurrence. Analysing the emergence and evolution of antithetical forces before and after the formation of Nigeria, the authors conclude that the seeds of civil war were sown in Nigeria’s amalgamation, the way it was governed when it was colonized, and the way its independence was arranged. Their analysis reveals that the civil war was extended by the interplay between conflicting foreign interests and actions, as well as differential domestic interests, capacities and actions, while the landlocked nature of Biafra reduced the duration of the conflict. The risk of conflict recurring has been increased by the exploitation of asymmetric institutional, political, economic and social relations by postconflict governments, as well as by the consequent adverse effects on governance, macroeconomic management and economic progress. While the 2003 elections presented both dangers and opportunities, the landmark judgements by the Nigerian Supreme Court in 2001–3 offer a basis for building a modern state on the colonial foundations if Nigeria overcomes the current challenges. Chapter 7: The Economics of Civil Conflict in Africa: The Case of Chad Nadjiounoum Djimtoingar and Djona Avocksouma argue that the Chad conflict is inherently internal, since it involved Chadian belligerents and took place within the country. Although the civil war broke out in 1979 and lasted until 1982, the genesis of the conflict can be traced back to the 1960s. The establishment in 1966 of the first politico-military movement, the National Liberation Front (FROLINAT), marked the first armed opposition to the government since independence in 1960. However, the change to a FROLINAT government worsened the hostilities. Lacking sound public management strategies, succeeding governments – from President Goukouni’s National Union Government to Presidents Habré and Déby – became embroiled in side issues far from the concerns of the northern and southern communities. Chad has not enjoyed widespread peace since President Déby seized power in 1990: pockets of resistance still operate in both the northern and southern regions, casting doubts on the credibility of the north–south, Muslim–Christian and animist, and pastoralist–farmer cleavages as the causes of the conflict. The costs of the ongoing conflict have been enormous. Mineclearing operations cost an estimated US$88 million, while demobilization and integration of combatants cost even more. The most important factor remains a lack of self-confidence among Chadians. Governments have been
xx
Introduction: Post-Conflict Economies in Africa
unable to mobilize the nation because of poor governance, institutionalized insecurity and ethnicization of public management. Several attempts have been made to resolve the conflict with the help of third parties, but these efforts have been only minimally successful. Indeed, successive governments have not accepted democratic succession, and negotiations have only succeeded in absorbing opponents into the government. Unfortunately, promises made are binding only to the signatories and conditions for transition to a peaceful, post-conflict phase are far from satisfied. Despite possible ethnic or religious undertones, bad governance is the key cause of the conflict. 3.2 The Horn, East and Central Africa Chapter 8: Conflict, Post-conflict and Economic Performance in Ethiopia Ethiopia has had a long history of relative cultural homogeneity, notwithstanding the large number of ethno-linguistic groups. Nevertheless, that long history has been punctuated by a number of conflicts, including the last religious war fought between the Christians and Muslims in the sixteenth century, and the era of the princes in the nineteenth century. According to Alemayehu Geda and Befekadu Degefe, the more recent periods of conflict have been the imperial era, which lasted until the fall of Haile Selassie in 1974; the 1974–91 Dergue period, during which a civil war took place; and the 1998–2000 Ethiopia–Eritrea war. While the conflict of the imperial period involved rivalries among the aristocracy over the benefits of control of the peasantry, the 1974–91 conflict mainly entailed rivalries among the educated elites. Similarly, the 1998 Ethiopian–Eritrean war most probably resulted from a miscalculation as each side threatened the other with military power in search of a more favourable bargaining position following the agreement on Eritrea’s secession from Ethiopia in 1993. In Ethiopia, the post-conflict periods have so far been relatively brief as there has been a pattern that the means of resolving one conflict sows the seeds of the next. This phenomenon deserves special attention, if an enduring post-conflict period is to be fostered. International intervention is also recognized as having played a salient role, either as a catalyst, or a restraint, of conflict. With the end of the Cold War the geopolitical importance of Ethiopia has diminished, thus attenuating the catalytic role of international interventionism. In the final analysis, however, Alemayehu Geda and Befekadu Degefe believe that the major cause of conflict in Ethiopia lies in the political economy and is underscored by competition for power. If so, then the apparent solution for achieving a more stable post-conflict economy is a strategy to minimize such competition. One such potential mechanism might involve the establishment of a democratic system that makes leaders more accountable to the people.
Fosu xxi
Chapter 9: Prospects for Sustainable Peace and Post-Conflict Economic Growth in the Sudan The Sudanese conflict, at the time of writing, spanning more than thirtyeight years of civil war since independence in 1956, is characterized by religious and cultural divides between north and south, according to Ali Ali and Ibrahim Elbadawi. These fundamental divides dominate the country’s tribal and regional interaction between north and south and within the regions. While the north–south divide explains external interventions, culture and religion account for the relatively high risk of conflict predicted for Sudan. The conclusion, according to the authors, is that these two realities about the Sudanese conflict are the key to understanding why the war started and why it has persisted for so long. Furthermore, these realities define several preconditions for sustainable peace, including the centrality of a genuine democratic transition in both the north and the south. The failure of the 1972 Addis Ababa peace accord, which allowed eleven years of fragile peace, shows the importance of democracy for sustainable peace and viable economic management in a divided Sudanese society. The accord finally broke down as a result of the autocratic rule of President Nimeiri, who, inter alia, attempted to impose Islamic law on the south. The authors argue that party-based democracy is the most viable instrument for promoting stable and developmental bargaining processes among potentially conflicting social groups. Unfortunately, the peace initiative under the auspices of the Intergovernmental Authority for Development (IGAD), being confined to the two main military protagonists, does not seem to accommodate this basic but crucial lesson. With the emergence of Sudan as an oil-producing country, transparent and democratic governance becomes even more critical. International evidence suggests that dictatorships (as well as unstable factional democracies) have almost always failed to harness oil resources for development. Instead, under such regimes, oil resources have been associated with deepened poverty, social cleavages and conflict. Even if peace cannot be achieved without partitioning the country, the transition to democracy in both future states would be critical. Given current conditions in the Sudan, partition without democratic transition risks replacing a civil war with an international war between two autocratic and non-representative regimes, both of which may also risk internal conflict. Chapter 10: Entrenching Peace in Post-conflict Economies: The Case of Uganda Michael Atingi-Ego and Rachel Sebudde show that, since independence in 1962, there has been at least one conflict in Uganda during each decade. Between 1964–6 the crisis was an ethnic power struggle which resulted in suspension of the Constitution, as well as detentions and killings of Bugandan civilians. In 1971, Idi Amin replaced President Milton Obote via a coup d’etat,
xxii
Introduction: Post-Conflict Economies in Africa
and then the liberation struggle between 1978 and 1980 culminated in the overthrow of President Amin in 1979. Conflict between 1981 and 1986 brought President Museveni to power, since which there have been continuing insurgencies, mainly in the north and west of the country, although Uganda has enjoyed relative peace in the post-conflict period starting from 1986. The authors argue that grievance, exacerbated by weak institutions, has been the main source of conflict. Various factors have led to improvements of the post-conflict economy since 1986: political reconciliation, including the return of some of the former leaders; the restoration of kingdoms; and improved economic conditions, such as per capita income growth and poverty reduction. Relations with Uganda’s neighbours have improved. Institution building, with international support, has been important, and has included: • • • • •
grassroots democratization; enhancement of press freedom; decentralization of power to local government; public service, market and legal reforms; a legal framework strengthening commercial jurisdiction, property rights and the settlement of business disputes; • support for the development of effective civil society groups; • the establishment of a properly functioning financial system as well as safety nets; and • demobilization of the military and the reintegration of rebels. The problem of uneven distribution of development remains, however. While the southern part of the country has benefited from the above reforms, the north has continued to lag behind. Meanwhile, the insurgency in the north continues. It is hoped that solutions will be found to achieve both economic progress and abate the insurgency in the north. Chapter 11: The Challenge of Transition from War to Peace in Burundi Since independence in 1962, Burundi has had five civil wars: the first was in 1965 and the most recent began in 1993. This latest conflict has been the longest and costliest, and at the time of writing, the situation is barely postconflict. The cycle of violence has been played out between two ethnic groups: the Hutus, and the minority (but politically and economically dominant) Tutsis. According to Janvier Nkurunziza and Floribert Ngaruko, powerful, self-interested domestic groups have manipulated the apparent ethnic rivalry to their advantage. The authors observe that Hutus and Tutsis lived together peacefully for at least four centuries before the outbreak of hostilities in the early 1960s, and shared a normative system of governance and a political system that allowed divergent interests to be reconciled, while social and economic activities allowed Hutus and Tutsis to identify themselves as members of one nation.
Fosu
xxiii
Thus, post-conflict Burundi can learn from its distant past. Justice is a prerequisite if the wounds of the recent past are to be healed. This requires an impartial administration to punish those inciting violence for their own selfish purposes. One way to proceed, the authors argue, is to use a prisoner’s dilemma-based system of amnesty and justice that induces perpetrators to tell the truth. Meanwhile, the role of the international community is crucial, not only in terms of engendering confidence for any agreement among the warring parties, but also in keeping neighbouring countries at bay. 3.3 Southern Africa Chapter 12: The Political Economy of Post-Conflict Economic Recovery: Making Peace Substantial in Mozambique Examining the political economy of post-conflict Mozambique, Laudemiro Francisco and Edwin Connerley argue that the durability of the peace and the domestic viability of current reforms are dependent on making peace substantial. This would require a comprehensive approach to reconstruction and development, including self-rule, proper macroeconomic policies, and building real foundations for social and political integration. Following the 1992 General Peace Agreement (GPA), Mozambique has experienced an impressive economic recovery. However, the GPA failed to address the grievances that motivated the combatants, and the risk of renewed conflict has therefore remained high. Thus an approach that would realistically foster economic, social and political development is still required. How the main actors continue to perceive the ability of the GPA to address their concerns is politically crucial for the viability of reconstruction and development efforts. Chapter 13: Transformation for Post-Conflict Angola Fátima Moura Roque writes that Angola was impoverished by twenty-six years of civil war. Political interference in economic management resulted in perverse policies, financial sector weakness, opaque public accounts, and corruption. The death of UNITA leader, Jonas Savimbi, in 2002 was followed by a Memorandum of Understanding which consolidated the ceasefire and the terms of peace. Though too early at the time of writing to be completely confident, the peace seems to be holding. Post-conflict, Angola must now confront widespread malnutrition; provide accelerated basic training for 80,000 demobilized soldiers and create safety nets for about 300,000 dependents; reintegrate 4.2 million internally displaced people and 500,000 refugees; and address extreme poverty. Following such a protracted conflict, a strong commitment to national reconciliation is required – that is, if the parties are to overcome so many years of resentment and social exclusion. Meanwhile, democratic governance and good economic fundamentals are essential for the post-conflict transformation.
xxiv Introduction: Post-Conflict Economies in Africa
4
Conclusion
Chapter 14 concludes the volume by synthesizing and drawing lessons from the earlier chapters. The synthesis comprises the following ten themes: colonial history; the role of neighbours; the relevance of elite political instability (EPI); the importance of rent-seeking opportunities; the role of child combatants; the proliferation of cheap weapons; justice: inter-ethnic and class inequities; justice and impartial, effective legal systems; the role of the international community; and the importance of democracy.
Notes 1 See, for example, Fosu (1996); Elbadawi, Ndulu and Ndung’u (1997). While the latter finds a non-monotonic impact of debt on growth based on a quadratic specification of the debt variable, the former employs an analysis-of-covariance model and finds that the debt effect is positive when the level of investment is low, but becomes negative at a sufficiently high level of investment, suggesting that the marginal product of capital decreases with the level of debt. 2 This result supports the finding in the literature that the prevalence of lootable commodities, especially primary exports, tends to raise the risk of civil war (Collier and Hoeffler, 1998). It also provides support for the result based on a more general concept of conflict in terms of ‘elite political instability’ (EPI). Fosu (2003), for example, finds that a higher exchange rate misalignment tends to raise the likelihood of EPI in African countries. 3 Not only might such a strategy convert, to use Mancur Olson’s words, ‘roving bandits’ into ‘stationary bandits’, but possibly to responsible non-bandits.
References Addison, T. (ed.) (2003) From Conflict to Recovery in Africa, WIDER Studies in Development Economics (Oxford: Oxford University Press). Burnside, C. and D. Dollar (2000) ‘Aid, Policies and Growth’, American Economic Review, vol. 90, no. 4, pp. 847–68. Collier, P. and D. Dollar (2002) ‘Aid Allocation and Poverty Reduction’, European Economic Review, vol. 46, issue 8, pp. 1475–500. Collier, P. and A. Hoeffler (1998) ‘On Economic Causes of Civil War’, Oxford Economic Papers, vol. 50, no. 4, pp. 563–73. Collier, P. and A. Hoeffler (2001) ‘Greed and Grievance in Civil War’, working paper no. 2355, World Bank, Washington, DC. Elbadawi, I., B. Ndulu and N. Ndung’u (1997) ‘Debt Overhang and Economic Growth in Sub-Saharan Africa’, in Z. Iqbal and R. Kanbur (eds), External Finance for Low-Income Countries (Washington, DC: International Monetary Fund). Fosu, A. K. (1996) ‘The Impact of External Debt on Economic Growth in Sub-Saharan Africa’, Journal of Economic Development, vol. 21, no. 1, pp. 93–118. Fosu, A. K. (2001) ‘Political Instability and Economic Growth in Developing Economies: Some Specification Empirics’, Economics Letters, vol. 70, no. 2, pp. 289–94.
Fosu xxv Fosu, A. K. (2002) ‘Transforming Growth to Human Development in Sub-Saharan Africa: The Role of Elite Political Instability’, Oxford Development Studies, vol. 30, no. 1, pp. 9–19. Fosu, A. K. (2003) ‘Political Instability and Export Performance in Sub-Saharan Africa’, Journal of Development Studies, vol. 39, no. 4, pp. 68–82. Wallensteen, P. and M. Sollenberg (2001) ‘Armed Conflict 1989–2000’, Journal of Peace Research, vol. 38, no. 5, pp. 629–44.
This page intentionally left blank
Part I Thematic Issues
This page intentionally left blank
1 Post-Conflict Reconstruction in Africa: Some Analytical Issues Tony Addison and S. Mansoob Murshed
1
Introduction
In contrast to much of the twentieth century, when warfare between rich states was the norm, contemporary conflict now occurs almost exclusively in poor developing countries and is mainly internal in nature (although external parties and neighbouring countries may support internal belligerents). The period 1990–2000 saw nineteen major armed conflicts in Africa, ranging from civil wars to the 1998–2000 war between Eritrea and Ethiopia (Wallensteen and Sollenberg, 2001). Although the picture is grim, there has recently been some good news: a peace agreement has been signed in Angola; the governments of the Democratic Republic of the Congo (DRC) and Rwanda have signed a peace agreement; talks have resumed between the government and the rebels in Sudan. Sadly, these positive developments need to be counterbalanced with the empirical regularity with which peace agreements break down (Walter, 2001). Ensuring that the benefits of reconstruction are broad-based rather than narrow in their benefits, it is important to minimize the chances of conflict reigniting, since grievances will otherwise fester. Accordingly, the containment and reduction of inequality, and not just a reduction in absolute poverty, may be central to broad-based reconstruction. This means that transition from war to peace may not be seen as only a political process, or one in which attention being paid to the economic dimensions of conflict should take a back seat to the political settlement (as is sometimes argued). Concomitantly, economists may not ignore the implications for social conflict of economic decisions: conflict and the potential for conflict play a strong role in retarding growth, development and poverty reduction (Murshed, 2002). Indeed, civil wars and conflict are among the major causes of development and growth failure in today’s developing world, a point increasingly emphasized by aid donors such as the UK’s Department for International Development, the World Bank, and the United Nations (DfID, 2001; World Bank, 2001; United Nations, 2000). 3
4
Thematic Issues
As societies slide into war, competing groups are formed to overcome collective action problems (in the sense of Olson, 1965). Ethnicity, whether based on language, religion or other distinctions, is often a superior basis for collective action in contemporary conflicts in poorer countries than other social divisions such as class that featured in the internal conflicts of earlytwentieth-century Europe.1 Moreover, as Azam (2001) indicates, organization or action along ethnic lines is often a product of the failure of the state to provide public goods. Thus conflicts rooted in grievances over one group’s standard of living relative to another, can take an increasingly ethnic dimension as leaders mobilize followers by appealing to ethnic differences (for example, the appeal to religious affiliation in Nigeria’s local-level conflicts). So too with greed over valuable natural resources; in the 1990s, Angola’s rebel movement – the National Movement for the Total Independence of Angola (UNITA) – increasingly emphasized ethnicity to organize support among the Ovimbundu people so as to win control over the country’s diamonds and oil. The greed versus grievance dichotomy is important to the debate on the nature of conflict (see Murshed, 2002). But for either of these forces to cause large-scale violence there must be other factors at work, specifically a weakening of dispute resolution mechanisms, which some refer to as state failure (Marshall, 1999, for example). We may view today’s African conflicts as being rooted in a partial or complete breakdown of the social contract – the agreed rules that govern the distribution of resources and obligations across society – and the concomitant mechanisms for settling differences. Furthermore, the social contract has economic and redistributive dimensions that are, in turn, embedded in fiscal institutions. The social contract, its collapse, and the centrality of its reconstruction post-conflict therefore provide the organizing themes of this chapter. Section 2 is concerned with conflict and the breakdown of the social contract, including possible reasons why a peaceful social compact or agreement might become infeasible in the pre-conflict phase. Section 3 analyses how commitments to peace, and the broad-based fiscal transfers it entails, can be sustained and enhanced. Section 4 continues on the theme of designing commitment technologies to peace, this time in the context of capturable natural resource rents. Section 5 discusses post-conflict reconstruction, and how its design can rebuild a peaceful and broad-based social contract. Finally, Section 6 concludes with a discussion of areas of major concern in post-conflict societies.
2
The collapse of the pre-conflict social contract
Within nation-states, the fiscal system will secure a workable social contract if the allocation of public expenditure and the apportionment of taxes are judged to be fair, or at least not so unfair that some groups regard taking resources by force the better option. There are many examples of conflicts
Addison and Murshed: Reconstruction
5
emerging out of fiscal disputes. Côte d’Ivoire, for example, has become increasingly unstable with the collapse of the social contract engineered by the late President Houphouët-Boigny, in which he allocated public spending across the regions to buy the loyalty of the country’s ethnic groups. Similarly, conflict in Burundi and Rwanda has had strong fiscal dimensions (Ndikumana, 2001). Disputes over the apportionment of revenues from natural resources are especially common and, as in Nigeria, they often take on ethnic and regional dimensions. Azam and Mesnard (2001) analyse the difficulties in sustaining social contracts that involve revenue sharing via fiscal transfers.2 Contemporary civil wars are more often related to the breakdown of explicit or implicit arrangements to share resources or revenues, rather than their absence. One reason that a contract to share revenues encounters difficulties is the imperfect credibility with which the side that controls the resources honours its commitment. Let there be two parties to the potential armed conflict, the government and a rebel group. The government party has access to revenues and royalties, but is threatened by the excluded rebel group, which may violently overthrow the government with probability π, related to a contest success function. On the other hand, it may choose not to fight if it receives a fiscal transfer from the government. Similarly, the government has a choice between fighting the rebels and offering them a fiscal transfer. The utility function of the government (UG) may be denoted as:
U
G
R
μF G G -⎞ ( Y – F – λG ) = ( 1 – π ) ⎛ --------------------⎝ F G + μF R⎠
(1)
where YG is the income of the government, FG represents government military expenditure, G the transfer to the rebels, λ < 1 captures the imperfect credibility with which the government transfer is made, and the probability of the government staying in power is 1 − π, which in turn is a function of the contest success function, as well as the relative fighting efficiency of the rebels vis-à-vis the government μ. The contest success function (see Hirshleifer (1995) for example) is given by: R
μF π = --------------------G R F + μF It is defined in terms of the rebels, and says that the probability of their success in the military contest with the state depends on their military expenditure, FR relative to the total military expenditure made by both sides to the conflict FG + μFR, weighted by the rebel side’s relative military efficiency, μ. The rebel utility function (UR) is given by: U R = Y R − F R + λG + π(·)(Y G − F G − λG)
(2)
6
Thematic Issues
The income of the rebel group is described by YR, and the cost of waging a war represented by FR. A great deal can be said about the sources of income and war finance for the rebel side. The rebels may obtain an income from natural resources under their control (illegal diamond mining in Sierra Leone, for example), by raising external finance via promising commercial concessions once they seize the state (the strategy of Laurent Kabila when he overthrew Mobutu Sese Seko in DRC in 1997), criminal activities such as drug trafficking, or sympathetic contributions from diasporas (Eritrea). A feasible social contract favouring peace must give the rebels as much utility via a credible transfer as they would get in the event of a (probability weighted) overthrow of the state. A social contract favouring peace in return for a transfer is infeasible if, at a critical level of π = ϕ, the probability of toppling the government by war is greater than the chances of it credibly making the transfer, λ.3 Also the social contract is less likely with regimes that prefer military expenditure F G over making a fiscal transfer to the rebels, a common outcome in countries with powerful militaries, such as Sudan. Note that, for the government, no transfer is made in the case of a civil war, and in the case of a credible transfer being possible there is a saving in military expenditure. There is therefore a trade-off between military expenditure and a credible transfer. For the rebels, there is the transfer and no fighting expenditure if there is peace. When the transfer is highly improbable, and the potential spoils are great, warfare is more likely. This captures the pre-conflict situation. We now turn to problems of enhancing commitment to peaceful behaviour, post-conflict.
3 Sustaining the post-conflict commitment to peace by the government We now consider policies that may enhance the credibility of the government’s commitment to the broad-based transfer to the rest of the society, which will increase the chances of an enduring peace. Let us postulate that peace has been established, usually through external intervention, and for the moment there is no war, such that π = 0 in Equation (1), and the government is in no imminent danger of losing power to the rebels. The price of peace is the transfer to the rest of society, including the erstwhile rebels. Will the government stick to this commitment? There are two types of government: the first type (labelled as 1) will honour its commitment, and the second type (labelled 2) may not honour its commitment. Let us consider the utility function of the first or good type of government: UG1 = f(Y G − F G − λG)
(3)
Addison and Murshed: Reconstruction
7
Its disutility from fighting (differentiating Equation 3 with respect to F G) leads to: −f1 = 0
(4)
This type of government obtains no utility from fighting the rebels and makes the transfer, hence λ = 1 in Equation (3). Turning to the second type of government, which is perhaps more realistic in relation to post-conflict situations: UG1 = f(Y G − F G − λG) + h(F G)
(5)
This type of government derives positive utility from reneging on its commitment to the transfer, captured by the parameter h, implying that it is willing to fight in order to make a smaller transfer or no transfer at all. Maximization with respect to fighting the rebels yields: f1 = h1
(6)
Thus this type of government will equate the marginal cost and benefit of reneging on pre-commitments to peace via the transfer, so future peace is not guaranteed and the problem of imperfect credibility of the commitment to the broad-based transfer remains λ < 1. The presence of imperfect credibility requires policy-makers, particularly external powers and donors (which can be crucial to ending civil war) to devise commitment technologies to peace. Consider an example, leading to an innovative term in the utility function of the ‘bad’ (type 2) government: UG1 = f(Y G − FG − λG) + h(F G − m(F G))
(7)
Here, m refers to the cost of fighting or reneging on strategic pre-commitments imposed by donors. This takes the form of a variety of sanctions, military as well as through the suspension of aid. Maximization with respect to fighting will now yield: f1 = h1(1 − m1)Λm1 ≥ 1
(8)
When compared to Equation (6), Equation (8) states that the incentive to deviate from peaceful commitments is diminished, as the marginal utility of fighting is reduced. Furthermore, as m1 → 1, the temptation to deviate from agreements is zero, as with the first type of government in Equation (4).
8
Thematic Issues
4 Natural resource rent contests and imperfect commitment among rebels Many conflicts in Africa occur amid natural resource wealth. However, it is not natural resource endowments per se but rather the type of natural resource that matters. Addison, Le Billon and Murshed (2002) distinguish between point resources, which are non-renewable and geographically concentrated (for example, diamonds and oil), and diffuse resources, which are renewable and geographically spread, such as soil and water. Conflict occurs in countries with diffuse-resource endowments when access and ownership are highly unequal, but its probability falls when asset redistribution is undertaken and is certainly less evident than in countries richly endowed in point resources such as Angola, Congo-Brazzaville, DRC and Sierra Leone. Exactly how point-resource endowments increase the chances of conflict is now debated vigorously. One factor might be the difficulty of managing the macroeconomic consequences of resource windfalls, resulting in boom and bust cycles that ultimately lower the rate of economic growth, where low or declining growth in employment and living standards is a factor in generating conflict (as found by Nafziger, Stewart and Väyrynen, 2000). Another is the incentive of political insiders to undermine state institutions (such as budgetary systems) deliberately, in order to conceal the siphoning-off of resource revenues. This leads to the collapse of the state from within, together with its ability to exercise social control and defend itself against usurpers (as in Mobutu’s Zaire). Support for the government also declines as its commitments take the form of highly imperfect contracts. Additionally, opposition groups attract the finance of unregulated international trading networks interested in capturing valuable natural resources (giving rise to merchant-capital wars, such as those in West Africa, which receive arms and finance from entrepreneurs based in the Middle East and the former Soviet Union). In this section we examine credibility problems related to the temptation to renege on peace agreements when capturable (point) natural resource rents are at stake. Again, we have two sides, whom we refer to as government and rebels.4 One side is either tired of fighting or it has an interest in peace. In our example this group is the government. The other side, the rebels, may have something to gain from the resumption of fighting, such as rents from resources (as in Angola). The roles played by the government and rebels in the games that follow can be reversed without altering the results. Both sides have entered into a peace deal. The government side derives no benefit from breaking this agreement. Consider the utility function of the rebel group (UR): UR = −(1/2)c1w2 + Bc2(w − we)
(9)
Addison and Murshed: Reconstruction
9
The first term on the right-hand side of Equation (9) is the pure cost of conflict in quadratic form, where w represents warfare or belligerent behaviour and c1 is the parameter measuring the direct cost of warfare. The negative sign before it is to indicate the cost or disutility from fighting. The quadratic form of the cost function indicates that the costs of war rise more than proportionately as the level of w rises. The second term on the right-hand side of Equation (9) indicates the gains to the rebels from reneging on a peace agreement, or the benefit from a ‘surprise’ war, where the level of actual conflict (w) exceeds the level of conflict expected in advance (we). In other words, the spoils of war can only be wrested via the ruse of peaceful intentions. The parameter c2 captures the magnitude of this effect: the higher is c2, the greater is the gain from feigning peacemaking first and looting later. It may also be viewed as a subjective measure of greed. In addition to this, the greater the abundance of lootable resources (B), or rents to be extracted, the higher is the gain from a surprise war. The rebels maximize their utility in Equation (9) subject to w, which leads to: w = Bc2/c1
(10)
This result can be interpreted in the following manner: the equilibrium choice of warfare is greater the higher is the element of pure avarice, c2, the higher the availability of lootable resources, B, and the smaller the direct cost of fighting, c1. As far as the government is concerned, its utility function could take the form: UG = −(w − we)2∀w ≥ we and = (w − we)2∀w < we
(11)
We shall focus our attention on cases where the government’s utility declines in surprise warfare when actual war is greater than expected. The opposite, when actual levels of belligerency fall below expectations (pleasant surprises) are intellectual curiosa. The disutility from surprise war arises either because the government has to engage in unforeseen military expenditure that diverts income from other types of public expenditure, or because war increases its need for foreign loans or aid. For example, following UNITA’s return to war in 1998, Angola’s government had to raise military expenditure sharply, much of it financed by oil revenues and foreign borrowing. Now assume that the rebels enjoy a first-mover advantage and can announce total peace, and then engage in surprise warfare. In this case, the actual and
10 Thematic Issues
expected levels of warfare would diverge — w = Bc2/c1 and we = 0 in Equation (9). This involves cheating on a pre-announced commitment. At this juncture, reputation becomes important. The government forms a view of the rebel announcements based on the past behaviour of the rebels. This implies that, for the rebel group, there exists a future cost of cheating in the context of a low-intensity conflict. The cost is equal to the loss of reputation for honesty, but this cost comes in the future. In summary, the socially optimal policy of zero warfare (w = 0) is time inconsistent or incentive incompatible, and thus will not be a possible outcome. The optimal policy of no conflict is infeasible, as it is inconsistent with the incentives and expectations of the parties to the game. More particularly, the government knows it is in the interests of the rebels to renege on a preannounced policy of total peace, and thus will not find any peace-offering credible. Furthermore, there will be a range of possible conflict intensities that are feasible equilibrium outcomes; thus multiple equilibria are possible. A rise in war booty raises the minimum conflict threshold. Also situations where the future is heavily discounted are likely to raise the fighting threshold associated with the best enforceable outcome. We now consider policies to reduce conflict. Much of this implies manipulating the attitudes of the rebel leadership via sanctions, arms controls, trade restrictions and foreign aid. Consider a reformulated version of the rebel utility function where we embed external conflict prevention policy parameters, and an additional cost component associated with an implicit or explicit international agreement: UR = −(1/2)c1(M1)w2 + Bc2(T,M2)(w − we) − c3(M3)(w − we)
(12)
Aid (T) may be utilized by a foreign power to reduce greedy attitudes, c2. Similarly, trade sanctions (M2) on items such as ‘conflict’ diamonds, money laundering and the activities of foreign entrepreneurs (in supplying arms and finance) might have the same effect. International controls on arms transfers (M1) and/or sympathetic assistance from non-residents could be utilized to raise the direct cost of war, c1. This lowers the fighting efficiency of the rebels, μ in Equations (1) and (2). The last term in Equation (12) represents another commitment technology by delegation, and c3 measures the costs of reneging on peace agreements as a function of sanctions (M3) imposed by other external signatories or parties to the agreement. The presence of commitment technologies, sanctions, conditional aid and controls on international trade (from looted resources) lowers the optimal level of belligerency among rebels. This can be seen by maximizing Equation (12) with respect to w, which yields:
Addison and Murshed: Reconstruction 11
w = (Bc2 − c3)/c1
(13)
This leads to a lower level of warfare when compared to Equation (10). There is a direct effect of the external sanction or outside commitment technology, c3. Then we have the indirect effects emanating from the manipulation of the behavioural parameters of the rebels, c2 and c1. Our discussion indicates that establishing peace is a tough job, as the recent experiences of Angola, the Democratic Republic of the Congo (DRC), and Guinea-Bissau show. But even if a peace agreement can be put in place, the difficulties do not end there: the form that reconstruction takes also affects the sustainability of peace. It is to these issues that we now turn.
5
Broad versus narrow reconstruction
The 1990s saw substantial donor-supported reconstruction: Angola, Ethiopia, Mozambique and Sierra Leone are African examples. To be successful, reconstruction must rebuild the social contract, otherwise conflict can reignite. Fundamental political and economic reform is therefore necessary to change the rules of the game, such innovation being akin to the deep interventions described in Dixit (2001). In principle, the temptation to renew conflict could be eliminated completely if the ‘prize’ (oil revenue, for example) is absorbed into national income (then w = 0, in Equations (10) and (13)). The gains from capture and surprise warfare will vanish in rebel utility functions, as B = 0. Input by external actors, such as linking peaceful behaviour to aid flows, will then be unnecessary (c3 = 0 in Equation 12). But promises of action are not enough. As we saw in Section 2, the social contract became unviable because the credibility of a transfer (λ) is low relative to the probability of rebel success in armed conflict. Also, as in Section 4, the commitment to peace may be subject to temptations to renege on the agreement. The credibility of the fiscal transfer may be restored through the creation of democratic institutions. But such institutional reform takes considerable time. A badly functioning democracy also implies incomplete contracts and imperfectly credible institutions of commitment. For example, the quality of fiscal institutions may be so low (as a result of destruction or neglect) that they are unable to deliver the necessary transfer even when the government is committed to making the transfer to its former opponents. The size of the transfer (G) in Equation (2) may be too small in per capita terms if the income of the excluded rebel group is low. Our discussion so far has been quite pessimistic, in part because peace is viewed as being dependent on the distribution of a ‘fixed pie’. But peace is more likely to endure if the ‘pie’ can be made to grow. Growth, by raising the tax base, increases the possibilities of redressing grievances through
12 Thematic Issues
fiscal transfers. Growth itself provides additional income to further reduce grievance (and perhaps mitigate greed) and tightens the labour market, thereby reducing the attractiveness of joining a warlord or rebel group (a point emphasized in Collier and Hoeffler, 2001). The initial years of peace may offer considerable scope for growth as the economy re-achieves its pre-war production frontier (Mozambique, whose war ended in 1992, has achieved annual real GDP growth of 6.2 per cent over the period 1990–9, with growth of over 11 per cent in the late 1990s). Just as economic growth in general can distribute its benefits broadly or narrowly across society (depending on the initial distribution of assets and skills), so too does reconstruction-led growth. Pre-war asset and skill distributions may have been highly unequal (with the resulting grievances contributing to conflict), and can worsen dramatically during wartime. Those who are already poor often lose the few assets they have, and looting adds to the number of poor (Nafziger, Stewart and Väyrynen, 2000). In contrast, warlords and their followers accumulate assets, and so while the early years of peace may see quite rapid gross domestic product (GDP) growth, it can be very narrow in its benefits, unless policies are put in place to restore the productive assets and human capital of the poor. The immediate post-conflict situation may also offer a golden opportunity for pro-poor asset redistribution (something that cannot be done easily during peace), although this can be impeded when rich ‘winners’ from war block the necessary measures. However, the ‘post-conflict’ economy will be highly distorted, and this can impose an unfortunate path-dependence on reconstruction and growth. One source of distortion is the sharp increase in transaction costs (σ) resulting from war (including the destruction of transport, the planting of land mines, and institutional collapse) that drive a wedge between producer and consumer prices. Typically, production (especially agriculture) is more vulnerable, leading to a sharper increase in its transactions costs compared to other sectors such as urban-based trade and services (as was the case in Mozambique’s sixteenyear civil war). To get a handle on this problem, consider a model with two sectors, agriculture and services, whose relative prices are PA and PS respectively, with transaction costs σA and σS. The transaction costs lower the supply price of the two goods because producers have to incur this extra cost. Given σA > σS, relative prices shift, resulting in a switch in activity from production (agriculture) to services: PS – σS PS ------------------- > -----PA – σA PA
(14)
if σA > σS This output change is additional to the increase in production costs resulting from the direct destruction of institutions, infrastructure and skills in the
Addison and Murshed: Reconstruction 13
two sectors (which usually affect production disproportionately, especially agriculture). Services and trade, including smuggling and other rent-seeking, dominate wartime economies and, because of the relative price effect, the collapse in production usually exceeds that resulting from destruction alone. Smallholder agriculture, the basic livelihood of Africa’s poor, is hit especially hard. For example, at the time of writing, Angola’s agricultural output is now less than 5 per cent of its pre-war level. The result is shown in Figure 1.1. Three production frontiers are indicated: PP (peace), WW (war) and RR (recovery). These have the usual properties. Assume a peace equilibrium of Ep given by the relative price (with σA , σS = 0).5 Growth will be along a ray such as Gp, as aggregate investment pushes the production frontier outwards over time. Contemporary wars generally reduce aggregate output. The wartime production frontier must therefore lie within the peace-time frontier. Agriculture (A) is hit harder than services (S), and the production frontier changes shape (to form WW reflecting the greater difficulties of A production). And the relative price shift in favour of services causes a structural shift to the new equilibrium Ew. There may be periods of growth during wartime, from a point below the peacetime production frontier, but it will be S-biased (along the ray Gw) and is therefore unlikely to reduce poverty sufficiently.
S Gw P w p Gp R Ep
W Ew
Eb
Gb
b
W Figure 1.1
War, peace and recovery
R
P
A
14 Thematic Issues
If a peace agreement is signed, then transaction costs start to fall. Accordingly, agricultural output recovers as relative prices shift back in its favour and as infrastructure is rebuilt; the economy moves through a series of reconstruction frontiers such as RR. Although this reconstruction is characterized by a more than proportionate rise in A, the main livelihood of the poor, benefits of recovery may still be distributed too narrowly because of the loss of assets and skills among the poor (see de Sousa, 2003 on Mozambique). Moreover, the economy’s pre-war structure may have been far from optimal for growth, poverty reduction or peace because of policy distortions — a history of policy bias that depresses PA relative to PS as in many African economies. Hence significant economic reform must take place alongside reconstruction if broad, rather than narrow, recovery is to take place. In Figure1.1 reform is shown as raising the return further on A relative to S activities (that is, to the relative price line b so that the economy recovers along Gb). There is wide agreement on the need for fiscal reform to raise public spending on pro-poor basic health care and primary education, although implementation is often fraught with difficulties. If peace is uneasy then there is an additional constraint on rebuilding the social contract via broad-based reconstruction and growth. To see why, consider the net present value (NPV) of a private investment project in sector i, producing an output Qi (Equation 15). In addition to raising transaction costs (σ) and production costs (C), conflict raises uncertainty about the future, and therefore affects the private discount rates of potential investors (r):6
∑ Qi ( Pi – σi ) – Ci t NPV i = --------------------------------------------t (1 + r)
(15)
The rise in r has differential sectoral effects depending on the time-profile of private benefits and costs associated with different types of activity. Thus production sectors such as agriculture require more fixed capital than trade and services (which may only need working capital), and profits from trade can be realized more quickly than from production, whose profits may only start to flow after several periods (particularly in agriculture, where tree crops have a lengthy gestation). Any rise in the private discount rate will tend to lower the NPV of investments in production relative to trade and services. This amplifies further the war-time collapse in the production sectors, and the expansion of trade, services and associated rent-seeking. Restarting activity in pro-poor production activities such as agriculture can therefore be especially hard after war when the peace is uneasy, since while transaction costs can be reduced (for example, de-mining rural roads lowers σ) and infrastructure rebuilt (lowering C), an uneasy peace keeps private discount rates (r) high. Consequently, the reconstruction-growth path may create insufficient income and employment, and its narrowness lead to frustration among large sections of the excluded population. Hence the
Addison and Murshed: Reconstruction 15
effectiveness of donor support to reconstruction, while lowering σ and C, will be reduced if parallel political initiatives (both domestic and international) are not undertaken to secure peace, and lower r. In summary, if grievance is the main source of conflict, then broad-based reconstruction (along Gb) is necessary to re-create a sustainable social contract. In particular, growth raises wage levels and reduces the attractiveness of predation as a livelihood (Collier and Hoeffler, 2001).
6
Conclusions
We have stressed the importance of the breakdown of the social contract, in particular its fiscal component, as contributing to civil war in Africa. We went on to examine the problems of sustaining a commitment to peace both by a government whose commitment may be tenuous, and by a group that has something to gain from the resumption of hostilities in terms of controlling resource rents. Finally, we considered the difficulties in engendering broad-based and pro-poor growth and reconstruction in post-conflict societies. This chapter has emphasized the importance of economics and rational-choice political science. This is not to underestimate the complexity of contemporary war, nor its deep roots in cultures and history (especially colonial legacies), nor to neglect the importance of other social science methodologies in addressing it. Nevertheless, there is a danger in seeing every conflict situation as being in some way culturally or historically unique, and thus perhaps intractable. The approaches discussed in this chapter do provide a framework in which to identify the fundamental constraints that must be overcome to end war, as well as the trade-offs involved. To conclude, we highlight two areas of pressing concern that need attention. The first is the issue of grievance. For a social group that has experienced years of decline in its absolute living standards (whether measured by income, human development indicators or both), a peace deal that delivers an absolute rise in its living standard may be sufficient to secure its withdrawal from violent conflict. In this case, the current donor focus on absolute poverty reduction will also be conducive to reviving the social contract. But grievances may also be expressed in relative terms; the rebel group may seek a closing of the gap in its standard of living with wealthier groups, not just a rise in its absolute well-being. In this case, governments (and donors) will need to be much more concerned with reducing inequality, especially across regions. For example, Mozambique has seen some (modest) improvements in living standards in areas dominated by the rebel group, RENAMO, since the end of the civil war in 1992. But, continuing discontent among RENAMO supporters, which flared into violence in 2001, reflects their perception that the fruits of reconstruction have so far been concentrated in regions controlled by the FRELIMO government, so that the regional gap shows no sign of narrowing, and may indeed have widened.
16 Thematic Issues
Additionally, relative differences in living standards across social groups may play a greater part in fermenting grievances in societies rich in point resources, since the monetary value of the associated flows of income is more evident than in the case of societies dependent on diffuse, renewable resources, where much of the associated output is not marketed. The expected return from controlling natural-resource-rich countries is therefore considerably higher than in controlling countries which may have rich diffuse resources, but in which there are considerable costs in organizing and profiting from production (such as in agriculture) even if a belligerent secures the territory. Indeed, in natural-resource-rich countries it is sufficient to control only that part of the territory in which the resources are mined, which may exacerbate state collapse (DRC, for example). Second, the peace dividend. Development economists typically classify military spending as unproductive. But donor efforts to reduce military spending, and free public money for development spending, are typically thwarted, by moving the spending off-budget, even in countries such as Uganda that have demonstrated commitment to poverty-reduction. Countries can have legitimate security interests, especially when they find themselves in bad neighbourhoods, and poor people place security high on their list of priorities. Defence spending will remain high, with large opportunity costs for development spending, until the international community can offer a credible regional peace, thereby preventing harm to countries that contract their defence capabilities unilaterally. Yet UN peacekeeping operations remain grossly underfunded; their cost in the year 2000 amounted to less than half of 1 per cent of the approximately US$800 billion that member states spend on national defence (United Nations, 2000). If peace were credible, then the developmental impact of releasing those resources from military spending could be immense, and the resulting broad-based development thereby in turn solidifying the social contract upon which peace is ultimately sustained. Notes 1 On the complexity of the relationship between ethnicity and violence see Bates (2000), among others. 2 The term fiscal transfer also covers broad-based social expenditure across all groups, and the award of jobs and other benefits from government contracts. It could even mean power sharing. 3 Alternatively, the ‘contract’ governing the transfer from the government to the rebels may be highly incomplete. This is likely to arise in the context of state failure, and where government actions are not transparent. 4 A more detailed version of the analysis in this section can be found in Addison and Murshed (2002). 5 The pre-war equilibrium will depend on policies that in turn affect the relative price, and thus the structure of output. As drawn, agriculture takes a larger share of pre-war output relative to services, which is typical of a low-income country when policy does not have a large bias against farmers.
Addison and Murshed: Reconstruction 17 6 Moreover, conflict increases the cost of capital (as risk premiums on loan rates rise and as the monetary base shrinks), acting as a further disincentive to invest.
References Addison, T. and S. M. Murshed (2002) ‘Credibility and Reputation in Peacemaking’, Journal of Peace Research, vol. 39, no. 4, pp. 487–501. Addison T., P. Le Billon and S. M. Murshed (2002) ‘Conflict in Africa: the Cost of Peaceful Behaviour’, Journal of African Economies, vol. 11, no. 3, pp. 365–86. Azam, J.-P. (2001) ‘The Redistributive State and Conflicts in Africa’, Journal of Peace Research, vol. 38, no. 4, pp. 429–44. Azam, J.-P. and A. Mesnard (2001) ‘Civil War and the Social Contract’, paper presented at the Royal Economic Society Annual Conference, Durham, 9–11 April. Bates, R. H. (2000) ‘Ethnicity and Development in Africa: A Reappraisal’, American Economic Review, vol. 90, no. 2, pp. 131–4. Collier, P. and A. Hoeffler (2001) ‘Aid Policy and Peace: Reducing the Risks of Civil Conflict’, paper presented at the Royal Economic Society Annual Conference, Durham, 9–11 April. de Sousa, C. (2003) ‘Rebuilding Rural Livelihoods and Social Capital in Mozambique’, in T. Addison (ed.), From Conflict to Recovery in Africa, WIDER Studies in Development Economics (Oxford University Press). DfID (2001) The Causes of Conflict in Africa, Department for International Development, London (mimeo). Dixit, A. K. (2001) ‘Some Lessons from Transaction-Cost Politics for Less-Developed Countries’, Princeton University, USA (mimeo). Hirshleifer, J. (1995) ‘Anarchy and its Breakdown’, Journal of Political Economy, vol. 103, no. 1, pp. 26–52. Marshall, M. G. (1999) Third World War: System, Process, and Conflict Dynamics (Lanham, Md.: Rowman & Littlefield). Murshed, S. M. (2002) ‘Conflict, Civil War and Underdevelopment’, Journal of Peace Research, vol. 39, no. 4, pp. 387–93. Nafziger, E. W., F. Stewart and R. Väyrynen (eds) (2000) War, Hunger and Displacement: The Origins of Humanitarian Emergencies, Volume I, WIDER Studies in Development Economics (Oxford: Oxford University Press). Ndikumana, L. (2001) ‘Fiscal Policy, Conflict, and Reconstruction in Burundi and Rwanda’, WIDER discussion paper 2001/62 (Helsinki: WIDER). See www.wider. unu.edu. Olson, M. (1965) The Logic of Collective Action (Cambridge Mass.: Harvard University Press). United Nations (2000) ‘Report of the Secretary-General on the Work of the Organization’, United Nations General Assembly, Official Records, Fifty-Fifth Session, Supplement No. 1 (A/55/1) (New York: United Nations). Wallensteen, P. and M. Sollenberg (2001) ‘Armed Conflict 1989–2000’, Journal of Peace Research, vol. 38, no. 5, pp. 629–44. Walter, B. F. (2001) Committing to Peace: The Successful Settlement of Civil Wars (Princeton, NJ: Princeton University Press). World Bank (2001) ‘Development Co-operation and Conflict’, Operational Manual OP 2.30 (Washington, DC: World Bank).
2 The Economic and Political Consequences of Conflict and Implications for Post-Conflict Recovery in Africa Ibrahim A. Elbadawi and Njuguna S. Ndung’u
1
Introduction
In sub-Saharan Africa (SSA) one in every five people is directly affected by civil war. So it is not surprising that a recent multi-agency report on the prospects of Africa’s development identifies the task of overcoming the conflict–poverty trap as the most basic and most serious challenge facing sub-Saharan Africa at the time of writing (World Bank, 2000).1 The essence of this trap is not only reflected by the bi-causal relationship between growth (and hence poverty) and the risk of conflict, but also has pivotal linkages with other fundamental variables necessary for post-conflict economic recovery. Recent studies on the causes and consequences of conflicts, as pioneered by Collier and his research associates2 have focused on economic growth and other indicators of economic performance prevailing in periods before the war (Collier and Hoeffler, 1998, 2001) or assumed that the two phenomena of conflict and growth are driven by the same process (Elbadawi and Sambanis, 2001). Elbadawi and Ndung’u (2001) developed the models to ascertain multiple causal relationships to include additional intermediate development indicators such as private investment, public investment, and monetary and fiscal policies. This chapter utilizes the results of their structural model, composed of six behavioural equations: risk of conflict; per capita growth; private investment; public investment; money demand; and government revenue.3 Based on a panel of 130 countries over eight 5-year periods between 1970 and 1999, the results, when combined with a fiscal consistency equation, allow for the solution of the reduced-form equations for the six endogenous variables, as well as government consumption. The results inform further discussion of the economic and political agenda for post-conflict economic recovery in SSA countries. 18
Elbadawi and Ndung’u: Consequences of Conflict
19
The results of the structural model complement the endogenous growth analyses in three ways. First, the replication of empirical regularities, especially the causes and consequences of conflicts and civil wars in a multi-causal structure key, illuminates key indicators, policies and outcomes. Second, the empirical regularities may have underestimated these effects by failing to take into account the bi-causal and multi-causal structure of the key indicators, policies and other intermediate development goals. Finally, in incorporating a multi-causal structure, the analysis allows us to draw conclusions regarding appropriate policies for post-conflict recovery. The results corroborate the basic predictions of the literature on the determinants of the risk of civil wars. They also lend support to two key conclusions from studies on aid: that it is effective only in the context of good overall policy and institutional environment; and that its impact is subject to diminishing returns (Collier and Dollar, 1999; Burnside and Dollar, 2000). The results also lend support to key predictions on debt overhang; our estimations are stronger than the evidence found in the literature, where single-equation models are used. Five conclusions from the structural model results underpin this chapter: • by exploiting the multiple channels of influences allowed by accounting for mutual causation among the seven pivotal variables, our results show that some popular empirical findings (such as the non-monotonic effect of social fractionalization on risk of conflicts) as well as the aid effectiveness and debt overhang predictions (in the cases of growth and private investment) in fact apply to a much wider class of performance indicators; • in addition to its direct influence on the risk of conflict, the degree of political rights also influences economic performance through the public revenue channel, suggesting that countries with high levels of political rights have a lower risk of conflict and superior economic performance, because the legitimacy of their governments allows mobilization of resources for financing larger public investments; • public investment, rather than overall economic growth, appears to be the main economic variable associated with risk of conflict, implying that specific economic policies can reduce risk of conflict; • the association between public investment and risk of conflict on the one hand and public revenue and public investment on the other, provide multiple channels for the effects on the risks of conflicts of political rights, overall economic and institutional policy environment as well as primary exports. Similar effects on private investments and growth also obtain; and • the structural model also incorporates a money-demand equation allowing fiscal and monetary policy to interact and produce a policy mix required for economic management in post-conflict economic recovery.
20 Thematic Issues
Section 2 of this chapter provides an overview of the development conditions in conflict and post-conflict countries (CPCs), in particular those in SSA, highlighting the severity of the development trap facing CPCs, including the adverse external debt burden and policy environment they inherit in the immediate post-conflict period. The empirical results from the reduced-form analysis are discussed in Section 3, providing evidence on the extent of the development trap and the potential for reversing it through more aid and debt relief (under appropriate policy and institutional environments). Section 4 concludes.
2
The development tragedy of conflict
Conflict and post-conflict countries face a development tragedy. Conflict generates further conflict and poverty; countries emerging from conflict continue to suffer poverty and lingering risks of renewed conflict.4 The Human Development Index (HDI) for CPC countries averages only 0.362, much lower than other low-income, highly-indebted poor countries (HIPCs) at 0.457 (see Table 2.1). CPCs have lower levels of well-being; they also have higher infant mortality rates, an average of 195 per 1,000 births (using 1996 figures) compared to the HIPC average of 152. Even CPCs that have benefited from the HIPC initiative (such as Uganda and Mozambique) are still precarious. Uganda has an NPV of debt to exports of 239 per cent, while Mozambique’s is 588 per cent. This shows that reconstruction has not supported or influenced export activities. Furthermore, Mozambique has Table 2.1
Indicators of debt and wellbeing of the CPCs
Angola Burundi CAR Chad Ethiopia Eritrea Liberia Mozambique Rwanda Sierra Leone Somalia Sudan Uganda HIPC average HIPC II proposal CPC Average
HDI 1997
Infant mortality 1996 (per 1,000 births)
NPV of debt to GNP (%)
NPV of debt to exports (%)
0.398 0.324 0.378 0.393 0.298 0.346 na 0.341 0.379 0.254 na 0.475 0.404 0.457
209 176 164 189 177 na na 214 205 284 na na 141 152
291 59 48 34 127 56 218 137 437 95 73 192 31 107
0.362
195
138
165 548 244 196 791 na na 588 373 778 na 1358 239 372 150 528
Source: World Bank (2001), World Development Indicators.
Elbadawi and Ndung’u: Consequences of Conflict
21
one of the highest infant mortality rates in the group, and in both cases their HDI is lower than average HIPC levels. Furthermore, most CPCs face unsustainable debt ratios. The thirteen countries considered to be CPC countries in Africa have NPV debt ratios to export of goods and services higher than those prescribed in HIPC II (150 per cent) and higher than the HIPC average (see Table 2.1). The average NPV of debt to exports is 528 per cent for CPCs, while the HIPC average stands at 372 per cent, against a benchmark of 150 per cent. These indicators imply that CPCs are a special type of HIPC. However, CPC countries also face the risk of further conflict arising out of the underlying economic constraints that differ from other HIPCs. Understanding the causes of poverty and conflict requires a better understanding of the mutual causation between growth and conflict. A structural model of growth and risk of conflict should endogenize investment and indicators of fiscal and monetary policies. Figure 2.1 provides comparisons of these indicators for conflict and non-conflict countries. Average annual per capita growth rates between 1970 and 1999 were less than −0.2 per cent compared to almost 2 per cent for non-conflict countries. Other intermediate economic performance indicators show a similar pattern. The average shares of public investment, private investment and public revenue were about
35 30
Percentage
25 20 15 10 5 0 –5 gcapGDP
Figure 2.1
priv_inv
pub_inv
pub_rev
Conflict
5 years following the end of conflict
Non-conflict
10 years following the end of conflict
Economic performance in conflict and non-conflict countries, 1970–99
Notes: gcap GDP: GDP per capita growth; priv_inv: share of private investment; pub_inv: share of public investment; pub_rev: share of public revenue. Sources: World Bank Data Base; and Small and Singer (1994).
22 Thematic Issues
7 per cent, 10 per cent and 19 per cent, respectively, compared to 10 per cent, 14 per cent, and 29 per cent in non-conflict countries. Evidence suggests that post-conflict countries achieve significant recovery during the five years immediately following the end of the conflict. However, sustaining postconflict recovery for a longer period is problematic. An enriched analysis can furnish a long-term perspective of post-conflict recovery in order for these countries to escape the cycle of conflict and poverty. 2.1 Macroeconomic policies CPCs face dilemmas establishing macroeconomic stability, promoting growth, reducing poverty and establishing credible fiscal, monetary and exchange rate policies. Given the low growth rates registered during conflicts, there is usually pressure on the government to improve the economy. Considering general government consumption, total tax revenue and overall surplus or deficit as shares of GDP, it is evident that the CPCs gain large deficits during the conflict. The majority had healthy overall surplus/deficit ratios in the early 1980s. However, in the 1990s, the deficit situation worsened. Deficit levels were high even for countries that had entered a relatively peaceful period, such as Uganda.5 High deficits during conflict and immediately afterwards can be explained by high military expenditure. General government consumption as a share of GDP is not large relative to non-conflict countries, but this is because of low revenues, as shown by the total tax revenues to GDP ratio. Government consumption is higher during conflict, yet this is when total tax revenues as a proportion of GDP are poorest. Angola, for example, had government consumption in excess of 40 per cent, even during the UN peace accord period. But for Mozambique, the share of government consumption fell by 15.9 per cent at the cessation of conflict, to 10.4 per cent in 1997. However, as noted above, most CPCs have not exhibited dramatic declines in the share of government consumption in GDP during peace. This can be attributed partly to the neglect of fiscal discipline during conflict having a lagged impact on the following period of peace. Total tax revenues, hence general government consumption, for CPCs are low, except for Angola and to some extent Eritrea. The low ratios, particularly of total tax revenues, may be attributed to the collapse or breakdown of the revenue-collection mechanisms. It is clear that revenue growth is not commensurate with GDP growth rates witnessed at corresponding periods. Clearly, CPCs take a long time to rebuild institutions such as tax-collection agencies, but also take time to reduce war-related expenditures. Consequently, there is scope during peace to improve these ratios, even though this might be counter-productive in terms of business development, since businesses would also be trying to rebuild themselves. Debt relief assistance and fiscal support, such as International Monetary Fund (IMF) Enhanced Structural Adjustment Fund (ESAF) programmes, would enhance
Elbadawi and Ndung’u: Consequences of Conflict
23
reconstruction, easing pressure on the government to levy taxes on businesses in its attempt to restore social services. To avoid stifling growth while enabling CPC governments to provide social services, there is need to consider the special circumstances of post-conflict economies. Comparisons of policy and growth performance between CPCs and SSA countries are shown in Table 2.2. Real growth improved between 1995 and 1998 for a few successful CPCs such as Uganda. For those countries still in conflict, real growth remained low or negative. The outcome of policies, as reflected by inflation, shows high inflation regimes for the CPCs compared to SSA countries. CPC inflation apparently increased in the last period, rather than stabilizing, as it did in SSA countries. CPC fiscal policy stance is consistent with expectations: high fiscal deficits, more than the SSA average. But there were dramatic improvements during 1995–8: the average deficit to GDP fell to −4.7 per cent for the CPCs compared to the −2.5 per cent SSA average. This dramatic improvement indicates that peace settlements improve macro policy formulation. CPCs have four basic characteristics. First, regarding investment and savings, evidence shows they attract little foreign direct investment (FDI), and have low rates of savings and investment. FDI flows decline during conflict. Where flows have been reversed, as in Uganda, it is the peace dividend that has helped. FDI flows and domestic investments have a dimension indicating credible domestic policies and stability. Second, GDP has been precarious but population growth (hence dependency rates) has increased Table 2.2
Growth and macroeconomic policy stance (means)
Indicator Real GDP growth 1970–9 1980–4 1985–94 1995–8
SSA Countries
CPC Countries
4.40 1.80 2.70 2.60
0.63 0.91 0.35 4.20
Inflation 1970–9 1980–4 1985–94 1995–8
10.40 12.14 9.86 9.25
5.70 9.40 16.70 18.60
Overall fiscal deficit (incl. grants) 1970–9 1980–4 1985–94 1995–8
−2.70 −7.10 −4.30 −2.50
−5.90 −9.00 −8.90 −4.70
Sources: African Economic Indicators; World Bank, various years.
24 Thematic Issues
tremendously, leading to increased poverty. Third, regarding external debt and the ability to pay, CPCs form the bulk of countries considered to be HIPC. Ability to pay is minimal because of the destruction of economic activity and a shift of available resources to military activity. In order to support postconflict reconstruction and economic policies, debt stock reduction would be an appropriate starting point to allow flexibility for economic reform policies. Finally, fiscal (monetary) policies, exchange rate management and inflation management policies are required in support of growth. Once normality has been restored, strong growth-friendly policies to support and sustain economic recovery and minimize the risk of conflict can be implemented, strengthening private-sector activities in rebuilding markets, domestic competitive production and ensuring resource flows into profitable sectors. Collier and Gunning (1995) show that resolving the conflict-uncertainty problem takes a long time, but it is necessary to stabilize markets and induce a portfolio substitution towards fixed irreversible investments.
3
Empirical analysis6
We utilize a structural model of growth and conflict anchored around two pivotal variables: the rate of growth and the risk of civil war, where the latter is defined as the probability of observing a civil war at any given point in time (Elbadawi and Sambanis, 2001). The model also includes three other endogenous variables associated with growth and risk of conflict: private and public investments; a money demand equation; and public-sector revenue (expressed as ratios to GDP). Success in the area of public investments is likely to be the trigger for the virtuous circle of growth, poverty reduction and conflict prevention. On the other hand, failure can cause countries to slip into the cycle of rising poverty and conflict. Moreover, since public investment is usually the first casualty in conflicts, post-conflict reconstruction must begin by rebuilding the physical infrastructure, as well as public institutions. In addition to foreign aid, public investments will, in turn, depend on government revenue, mainly mobilized through the taxation of primary exports. Finally, in post-conflict countries, private investment must help to sustain growth, so that a meaningful reduction in poverty can be achieved and risks of future conflict reduced sufficiently. In addition, renewed private investment is a signal for a stable polity and for the credibility of domestic economic policies, which should translate into further inflows of foreign direct investment. Collier and Gunning (1995) show that fears of insecurity encourage private agents to hold their portfolios in liquid form, resulting in unpredictable returns because of uncertainty in the economy. This unpredictability induces erratic changes in money demand, complicating monetary policy. So rates of return should be adjusted in relation to the levels of risk of conflict prevailing in the economy.
Elbadawi and Ndung’u: Consequences of Conflict
25
Following the work of Collier and Hoeffler (2001), and Elbadawi and Sambanis (2001), several explanatory variables are included in the structural model. Controlling for social characteristics (such as ethnic and religious diversity), risk of civil war is explicable in terms of fundamental economic variables (per capita GDP; economic growth; public revenues; money demand; and dependence on primary exports, especially minerals) as well as indicators of political and human rights. Aid effectiveness and debt overhang literature are the most relevant to the empirical endogenous growth model for the analysis of economic performance in conflict and post-conflict countries. Consequently, the additional explanatory variables we consider include the stock of external debt to GDP; external debt service ratio; ratio of aid to GDP; the interaction of aid with indicators of overall policy environment; and traditional growth and investment explanatory variables (initial income, educational attainment, population growth, average inflation and real exchange rates). The structural model can be expressed as follows: F[gy, Pw, Ip, Ig, Ry, DEBTY, DEBTY2, INTPYMT, RINT, AIDY, AIDY2, AIDPOL, CPIA, GIY, Y0, School0, lagged gy, lagged Ip, FINDEP, FBY, AVGINFLN, RER, POPGRWTH, LEGACY, AIDYLGCY, POLITY, PRIMEXP, ETHNIC, ETHNIC2, ETHPOLITY] = 0 where gy, Pw, Ip, Ig, Ry are the five endogenous variables (rate of real GDP growth, risk of civil wars, private and public investments, and government revenue), with the last three variables expressed as ratios to GDP. The explanatory variables include the following: DEBTY, which stands for external debt stock to GDP; DEBTY2 is external debt stock to GDP lagged one period and squared to capture the overhang effect; INTPYMT is external debt service payment as a share of GDP; RINT is real interest rate; AIDY is foreign aid to GDP; AIDY2 is foreign aid to GDP squared; AIDPOL is foreign aid to GDP interacted with an overall policy index, CPIA, which ranges from zero (bad policy) to 6 (for good policy environment). Both AIDY2 and AIDPOL reflect two of the key findings of the aid effectiveness literature, where aid was found to be associated positively and robustly with economic performance in the context of good policy environment; however, the effect of aid is usually subject to diminishing returns. GIY is aggregate investment and POPGRWTH is the rate of population growth. The initial level of development is reflected by initial income and schooling (Y0 and School0, respectively). Traditional variables reflecting overall macroeconomic stability and competitiveness include: FINDEP, FBY, AVGINFLN and RER, which are, respectively, financial depth (ratio of stock of money to GDP), fiscal balance to GDP ratio, average inflation, and real exchange rate. LEGACY is a positive indicator for post-conflict countries (0 elsewhere), which is equal to the number of months during the first decade following cessation of conflict
26 Thematic Issues
multiplied by the total length (in months) of the previous war (Collier, 1999a). Collier regards this variable as a proxy for the ‘peace dividend’ effect. Both LEGACY and AIDYLGCY (AIDY × LEGACY) are expected to be associated positively with growth. The remaining variables (POLITY, PRIMEXP, ETHNIC, ETHNIC2, ETHPOLITY) are motivated by the literature on the risk of civil war. The variables, respectively, stand for the extent of political rights, share of primary exports to GDP, degree of ethnolinguistic fractionalization, its square and its multiple with (1-POLITY). POLITY is predicted to be associated negatively with the hazard of conflict, although Gleditsch and Hegre (1997) find the effect of POLITY to be positive and non-monotonic, suggesting that the risk of conflict can rise with democratization before reducing as the democratic process matures. Primary exports (PRIMEXP) is expected to be associated positively with conflict. Social diversity (ETHNIC) has been found to be associated either positively and non-monotonically with the risk of conflict (Elbadawi and Sambanis, 2001) or associated negatively with conflict when controlling for ethnic dominance. The latter effect is found to be associated positively with the risk of conflict (Collier and Hoeffler, 2001). In this model we also add an interaction term between ETHNIC × (1-POLITY), which should be associated positively with conflicts. Such an effect can be explained by the ‘grievance’ motive of civil war, which suggests that an absence of political rights renders socially diverse societies particularly vulnerable to the risk of civil war.
4
The estimation results
The structural model of the above equation was estimated for 133 countries over eight 5-year periods covering 1960–99, giving rise to a sample space of 1,064 data points. We first estimated fixed and random-effects panel regressions, and then re-estimated the models in non-linear, multivariate and full information maximum likelihood (see Tables 2.3–2.7). The last three methods utilize the minimum distance estimators. These three estimators are consistent with the two fundamental assumptions of the structural model of Equation (1): that the five development and intermediate target variables are characterized by mutual causality, and that the relationships between the variables may be non-linear. Because of this interdependence, the minimum distance estimators dropped the variables that were not significant even though they might have been significant in the panel regressions. The empirical procedure was to utilize various estimation methods in order to get robust estimates to help us to map the channels through which these five endogenous variables affect each other. In the multivariate regression, we utilize the fact that the disturbances from the equations may be correlated, since we assume that simultaneity exits within the endogenous variables. If this is the case, the estimation method would provide us with more efficient
Elbadawi and Ndung’u: Consequences of Conflict
27
estimates compared to the panel regressions, where each equation is estimated separately. In addition, since these equations share explanatory variables, the estimation method imposes cross-equation constraints. This method uses the generalized least squares estimator and, if the residuals are multivariate normal, the estimator will converge to the maximum likelihood estimator. In order to check for the robustness of the results we have finally estimated the model in full information maximum likelihood (FIML). This is an asymptotically efficient estimator under our assumption that the disturbance terms are multivariate normal. It has been shown that even when this assumption does not hold, FIML will still give us asymptotically efficient coefficients. If there are no simultaneity and non-linearity problems among the endogenous variables, the FIML estimator will coincide with the multivariate results. The results have no marked differences between the multivariate and FIML results. They also suggest that non-linearity may not be a problem in these results, but that the endogenous variables are interdependent. Therefore, we chose the multivariate regressions for further analysis and for the subsequent simulation exercise of the next section. The structural model can be elaborated in diagram form to show the causal structure of the three pivotal variables: growth, private investment and risk of war (see Elbadawi and Ndung’u, 2001). Other variables can be added to the comparative statics to show the richness of the mutual causality among the variables and the key indicators. 4.1 The growth model The results of the growth model suggest that the risk of conflict is associated negatively with growth, but that the length of peace following long wars (LEGACY) is associated with high growth. Both results are consistent with the received literature (Collier, 1999). Also our results strongly corroborate the theoretical predictions of the aid effectiveness and debt overhang strands of the literature. The results suggest that the growth–debt relationship is located on the ‘bad’ side of the Laffer curve, where only DEBTY2 is significantly (and negatively) associated with growth. In addition, the crowding effect of interest payments (INTPYMT) is also associated negatively with growth. Our results also corroborate a modified version of the aid effectiveness literature, where – regardless of the policy environment – aid contributes to growth but is subject to diminishing returns. However, our results also show that good policy environment further adds to the effectiveness of aid. In terms of macroeconomic policy, high inflation was found to be associated negatively with growth, while more competitive real exchange rates and high aggregate investments were positively associated with growth. The rate of population growth was positively, and perhaps surprisingly, associated with growth, but the full effect on growth for this variable will turn out to be negative (see the analysis of the reduced form effects in the following section). Finally, the multivariate estimation results suggest that
28 Thematic Issues Table 2.3
Growth model results
Explanatory variable
Fixed effects model
−0.224 (−3.970) −0.148 (−5.220) 0.077 −1.700 0.525 −11.120
−9.460 (−2.830) −0.261 (−5.750) −0.157 (−6.580) 0.084 −2.190 0.583 −19.470
0.023 −2.820 0.029 −0.895
0.017 −2.270 0.005 −0.190
CONSTANT RISK OF WAR DEBTY2 LEGACY AIDY
Random effect Non-linear model estimation
AIDY2 AIDPOL GIY Lagged growth AVGINFLN INTPYMT RER POPGRWTH R2 Hausman test
−0.030 (−1.330) −0.016 (−.802) 0.017 (−0.890) 0.654 (−26.350) 0.830
−0.024 (−1.160) −0.014 (−0.750) 0.011 (−0.640) 0.634 (−32.670) 0.800 4.980
−11.480 (−3.800) −0.330 (−6.100) −1.440 (−47.200) 0.051 −0.940 0.607 22.370 −0.090 (−1.150) −0.012 (−1.470) 0.015 −0.550 0.162 (−8.060) −0.027 (−1.340) −0.019 (−1.030) 0.014 (−0.780) 0.607 (−29.090) 0.740
Multivariate FIML estimation estimation −38.880 (−39.340) −1.133 (−26.130) −1.014 (−75.700) 0.203 −4.670 0.625 −24.480 −0.114 (−2.040) 0.074 −11.690 0.065 −4.440 0.043 (−3.280) −0.041 (−3.790) −0.075 (−6.720) 0.041 (−4.270) 0.641 (−30.030) 0.570
−28.690 (−8.290) −0.784 (−3.550) −1.060 (−42.790) 0.190 −3.660 0.621 −24.380 −0.014 (−0.060) 0.042 −3.590 0.068 −2.750 0.101 (−4.420) −0.044 (−2.370) −0.051 (−2.640) 0.032 (−1.490) 0.621 (−16.410) 0.670
Note: Figures in brackets are t-ratios. Source: Computed from estimated equations.
‘growth begets growth’; that is, past growth is likely to generate more growth in the future. This is important in explaining the episodes in CPC countries where risk of conflict depresses growth not only in the current period but also in subsequent years. 4.2 Private investment Table 2.4 shows the results from estimating the private investment equation in the same way. Public investment is associated positively with private investment, while the risk of conflict is inversely related to private investment. This is a channel through which the risk of civil war further reinforces a negative association with growth, which suggests that single equation
Elbadawi and Ndung’u: Consequences of Conflict
29
models are likely to understate the extent of this association. Public investment and risk of conflict are by far the most important determinants of private investments. The debt overhang does not have a direct effect, although, as expected, the real interest rate (RINT) has a negative and significant effect. On the other hand, aid is associated positively with private investment. Moreover, this direct effect of aid is not subject to diminishing returns. However, the net reduced-form effect would restore the predictions of the aid effectiveness literature (see section 5). Finally, the results suggest that Table 2.4
Private investment model results
Explanatory variable
Fixed effects Random effect Non-linear Multivariate FIML estimation model model estimation estimation
CONSTANT ATWAR
−0.002 (−0.405)
4.587 −4.686 −0.003 (−0.300)
DEBTY DEBTY 2 RER FINDEP AIDY AIDY2 AIDYLGCY PUINV RINT Lagged private Investment CPIA FBY AVGINFLN R2 Hausman test
−0.009 (−1.820) −0.006 (−0.500) 0.007 −0.454 0.002 −0.490 0.0005 (−0.170) 0.00008 −1.350 0.989 −24.300 −0.023 (−1.640) 0.067 −1.530 −0.0005 (−0.160) 0.0336 −1.310 0.011 −1.740 0.970
−0.008 (−1.960)
0.002 −0.310 0.003 −0.310 0.00004 −0.130 0.974 −93.830 −0.027 (−2.150) 0.070 −7.290 −0.0003 (−0.100) 0.027 −2.760 0.0097 −2.400 0.970 4.350
Note: Figures in brackets are t-ratios. Source: Computed from estimated equations.
1.880 −2.510 −0.0001 (−0.010) −0.001 (−0.050) 0.001 (−0.050) −0.011 (−2.780) 0.024 −3.650
−38.880 (−39.340) −1.050 (−121.860) 0.032 −1.540 −0.033 −1.560 −0.007 (−2.190) 0.018 −3.270
−28.690 (−8.290) −1.110 (−36.060) −0.016 (−0.400) 0.017 −0.040 −0.009 (−2.180) 0.019 −2.490
0.007 −0.640
0.201 −11.190
0.179 −9.100
0.954 −88.740 −0.024 (−2.700) 0.078 −7.740 −0.113 (−0.340)
0.480 −50.170 −0.026 (−3.840) 0.026 −3.170 0.341 −29.720
0.549 −17.360 −0.028 (−3.120) 0.021 −2.580 0.413 −11.170
0.960
0.110
0.120
30 Thematic Issues
lagged investment, financial depth (FINDEP) and overall policy stance (CPIA) were all associated positively with private investment, while the latter is affected negatively by the real exchange rate. 4.3 The risk of civil war Table 2.5 shows the results from the estimated model of the probability of war, ATWAR. The model corroborates findings from previous literature in several aspects. For example, social diversity has a positive and non-monotonic effect, while primary exports are associated with a high risk of conflicts. A new result on the effect of social diversity, so far not obtained in the literature, is that absence of political rights makes socially diverse societies particularly Table 2.5
Risk of war model results
Explanatory variable
Fixed effects model
Random effect model
Non-linear estimation
Multivariate estimation
FIML estimation
−0.018 (−1.920) 0.283 −6.510 −0.018 (−0.850) 0.081 −1.090 −0.001 (−2.090) 0.431 −3.490 0.029 −3.250
5.620 −2.180 −0.019 (−1.820) 0.242 −11.740 −0.005 (−0.230) 0.068 −1.880 −0.0009 (−3.630) 0.500 −13.920 0.031 −3.430
2.110 −0.920 0.019 −1.770 0.170 −8.950 0.009 −0.410 0.024 −0.750 −0.0002 (−0.900) 0.756 −19.990 0.013 −0.123 0.0003 −0.780
−38.880 (−39.340) 0.005 −1.890 0.010 −2.280 −0.501 (−91.870) 0.017 −2.820 −0.00014 (−3.090) 0.165 −12.380 0.320 −30.360 0.0001 −2.070
−28.69 (−8.29) 0.006 −1.25 0.006 −0.800 −0.431 (−14.780) 0.018 −1.460 −0.00012 (−1.380) 0.145 −9.280 0.370 −12.600 0.0001 −0.130
CONSTANT Lagged growth POLITY PUINV ETHINIC ETHNIC2 PRIMEEXP CPIA ETHPOLITY AVGINFLN DEBTY DEBTY2 ETHPOL R2 Hausman test
0.043 −3.950 0.174 −1.610 −0.176 (−1.640) 0.144 −2.440 0.890
0.047 −4.09 0.158 −2.33 −0.158 (−2.33) 0.001 −3.67 0.87 43.03
Note: Figures in brackets are t-ratios. Source: Computed from estimated equations.
0.780
0.210
0.190
Elbadawi and Ndung’u: Consequences of Conflict
31
vulnerable to the risk of conflict. Another new, and potentially very important, result is that public investment, rather than economic growth, was the economic variable associated most strongly and robustly with low risk of conflict. Finally, both POLITY and CPIA have had a perverse association with high risk of conflict. However, again the net reduced form effects restore the theoretically consistent predictions of the literature. Therefore, unlike single-equation analysis, this structural model allows for multiple channels of influence, which should prove valuable for developing policy instruments for post-conflict reconstruction as well as conflict prevention. 4.4 Public investment Table 2.6 shows the results for the public investment equation. The evidence suggests that growth and government revenues are associated with high public investments. Two important results that strongly corroborate the aid effectiveness literature are that a high level of aid tends to reduce public investment, but if delivered in a good policy environment it can augment such investment. LEGACY and population growth tend to reduce public investment. One explanation for the LEGACY effect, which also obtains in Table 2.6 Public investment model results Explanatory variable
Fixed effects Random model effect model
0.183 −3.650 0.028 −1.230
−75.490 (−24.950) 0.208 −9.370 0.029 −1.380
−0.008 (−1.440)
−0.006 (−1.140)
−0.039 (−2.600) 0.047 −1.640 0.025 −1.050 −0.0002 (−0.700)
−0.041 (−2.820) 0.038 −1.690 0.013 −0.450 −0.00008 (−0.300)
CONSTANT Revenue to GDP AIDY Growth AIDPOL LEGACY POPGROW POLITY ETHNIC ETHNIC2 Lagged per capita income
0.019 −1.210
0.019 −1.320
Non-linear Multivariate FIML estimation estimation estimation −66.160 (−23.820) 0.299 −13.020 0.051 −2.230 0.013 −0.510 0.003 −0.460 0.017 −0.690 −0.043 (−1.870)
−38.880 (−39.340) 0.743 −79.910 −0.307 (−21.090) 0.477 −50.360 0.070 −13.620 −0.122 (−5.760) −0.331 (−24.760)
−28.690 −8.290 0.899 −9.290 −0.170 −4.580 0.266 −5.990 0.082 −7.600 0.060 (−1.320) −0.178 (−4.670)
32 Thematic Issues Table 2.6
(continued)
Explanatory variable ATWAR DEBTY2 AVGINFLN AIDYLGCY R2 Hausman test
Fixed effects Random Non-linear model effect model estimation −0.002 (−0.090) 0.003 −0.152 −0.030 (−1.640) −0.003 (−0.970) 0.610
0.022 −0.580 0.010 −0.540 −0.040 (−2.570) −0.0005 (−0.368) 0.550 12.930
0.160
Multivariate FIML estimation estimation
0.970
0.130
Note: Figures in brackets are t-ratios. Source: Computed from estimated equations.
the reduced-form case, is that, following long wars, government revenues are shifted increasingly from physical investment towards other social sectors as peace becomes established. 4.5 Government revenue Finally, we present the estimation results of the government revenue model (see Table 2.7). The results are important as they show that the risk of conflict has a strong and negative effect on government revenues. The other variable that affects government revenue negatively is external debt service payment. Two other important results that may be linked to aid effectiveness literature are that good policy (CPIA) and aid in a good policy environment (AIDPOL) are both associated with high public revenues, as are growth and primary exports. The latter provide for multiple channels for the effect of primary exports on the risk of conflict. Operating through public investment, the effect of primary exports on revenue would have a negative influence on the risk of conflict. This effect, however, is more than outweighed by its direct impact on the risk of conflict. Finally, the positive effect for POLITY suggests that political legitimacy is important for attaining high public revenues. This result has profound policy importance, especially in terms of highlighting the interdependence between economic and political development. Our results corroborate the basic predictions of the literature on the determinants of the risk of civil wars, and the aid effectiveness and debt overhang literature (the growth and private investment models). The results of our system estimations are much stronger than the evidence of either of the individual strands of the literature, which estimate single equation models.
Elbadawi and Ndung’u: Consequences of Conflict Table 2.7
33
Government revenue model results
Explanatory variable
Fixed effects model
Random effect model
Non-linear estimation
Multivariate estimation
FIML estimation
−0.079 (−1.380) −0.024 (−3.040) 0.013 −0.550 0.039 −2.080 0.016 −0.540 −13.950 (−5.360)
−83.350 (−31.990) −0.074 (−1.280) −0.023 (−3.350) 0.031 −1.110 0.050 −2.490 0.016 −0.190 −14.890 (−12.060)
−75.290 (−14.950) −0.083 (−1.600) −0.010 (−0.580) 0.056 −2.260 0.067 −3.730 0.035 −0.650 −17.350 (−13.130) 0.046 −1.250
−38.880 (−39.340) −1.590 (−56.030) 0.117 −24.170 0.042 −3.630 0.015 −2.280 0.414 −11.260 −3.640 (−6.580) 0.718 −30.630
−28.690 (−8.290) −2.260 (−8.730) 0.121 −8.130 0.042 −1.440 0.012 −0.640 0.435 −3.540 −1.460 (−1.950) 0.871 −9.890
CONSTANT ATWAR AIDPOL POLITY Growth PRIMEXP INTRY CPIA DEBTY DEBTY2 AIDY2 AVGINFLN LEGACY AIDY R2 Hausman test
0.074 −1.770 −0.065 (−1.580) 0.453 −4.020 0.003 −0.140 −0.360 (−3.620) −0.030 (−1.100) 0.580
0.047 −0.400 −0.035 (−0.300) 0.440 −4.150 −0.014 (−0.690) −0.360 (−6.950) −0.029 (−1.110) 0.520 0.630
0.17
0.210
0.250
Note: Figures in brackets are t-ratios. Source: Computed from estimated equations.
Moreover, our estimation provides some further insight. For example, public investment, rather than growth, appears to be a main economic variable associated strongly (and negatively) with the risk of conflict. Moreover, the association between public investment and risk of conflict on the one hand, and public revenue and public investment on the other, provide multiple channels for the effects on the risk of conflict of POLITY and CPIA, as well as PRIMEXP. Similar effects on private investments and growth also obtain. A more detailed discussion of these reduced-form effects follows.
34 Thematic Issues
5
Reduced-form analysis
The empirical estimation started with panel estimation to assess the validity of the arguments. But, from the argument of mutual causality, it was felt that simultaneity would bias the results as well as the possibility of non-linear relationships in the model. So the next stage in the estimation was to re-estimate the model in non-linear form, and non-linearity was not a problem. The end result was to compare FIML estimation and multivariate estimation, which was preferred, since multivariate regression accommodates non-linear relationships in the data as well as simultaneity, otherwise addressed individually by non-linear and FIML estimations. But there arose a problem of identification, which was solved by reducing the model and retaining only the statistically significant variables in the multivariate regressions. This helped in identifying the reduced-form equations used in the discussion. Thus the results from the multivariate regressions have been used to construct a system of equations for growth; risk of war; private investment; public investment; money demand; and government revenue. In addition, we add a fiscal account identity, which allows derivation of an expression for government consumption (Gy). The reduced-form results for the six endogenous variables, shown in Table 2.8 were analysed for seven categories of exogenous variable: debt overhang; aid effectiveness; macroeconomic environment; structure of exports; levels of political rights; social diversity; and the peace dividend following end of civil war (legacy) and monetary policy.
Table 2.8
DEBTY DEBTY2 AIDY AIDY2 LEGAY AIDPOL GIY INF INTRY RER POP FINDEP RINT CPIA
Reduced-form results8 gy
Ip
Pw
Ig
Ry
Gy
0 −3.158 0.6167 −0.355 0.1038 0.9102 0.2024 −0.1277 −11.9483 0.1277 0.5626 0 0 −0.4559
0.032 −3.833 −0.0147 −0.2262 −0.1759 1.4821 0.2436 −0.1536 −21.0438 0.1466 −0.1389 0.018 −0.026 −0.1604
0 1.8923 0.0073 0.2127 0.0876 −0.7381 −0.1213 0.0765 10.4795 −0.0765 0.0692 0 0 0.4024
0 −3.7771 −0.0146 −0.4246 −0.1748 1.4732 0.2421 −0.1527 −20.9173 0.1527 −0.1381 0 0 −0.1644
0 −3.0562 −0.0024 −0.3436 −0.1377 1.3042 0.1959 −0.1236 −20.4817 0.1236 −0.1016 0 0 0.0714
1.000 0.7209 1.0122 0.0811 0.0371 −0.169 −0.0462 0.0292 −0.5644 −0.0292 0.0365 0 0 0.2358
Elbadawi and Ndung’u: Consequences of Conflict
Table 2.8
(Continued)
POLITY POLITY2 ETH ETH2 PEXP ETHPOL LGY PIYL DEBTYt−1 CONST.
0.0487 0 −0.147 0.0012 −0.0942 −0.0009 0.0907 0 0 −78.4795
0.0959 0 −0.2442 0.002 −0.0088 −0.0014 0.0893 0.026 0 −146.3149
−0.043 0 0.1297 −0.0011 0.0831 0.0008 −0.0421 0 0 34.951
0.1058 0 −0.225 0.0019 0.1635 −0.0013 0.094 0 0 −147.3673
0.1111 0 −0.2085 0.0017 0.2804 −0.0012 0.0683 0 0 −95.6293
35
0.0053 0 0.0165 −0.0001 0.117 0.0001 −0.0257 0 −1.000 51.738
Source: Computed from estimated equations.
5.1 Debt overhang Increases in external debt lead to a rise in government consumption, while foreign debt has a negative and a quadratic effect on growth, and private and public investment as well as money demand, but has no effect on public revenues (see Figures 2.2 and 2.3). Debt flows reduce the risk of conflict but the stock of debt increases the risk (see Figure 2.4). External debt seems to find its way into financing government consumption (see Figure 2.5) but, interestingly, it affects public revenues in the same way. However, the stock of external debt does not affect public revenues (see Figure 2.6) but has a negative effect on public investments. 5.2 Aid effectiveness Abstracting from the quality of the policy environment, the effect of aid (AIDY, AIDY2) has the usual inverted U-shape in the case of growth, which corroborates the previous evidence on diminishing returns to aid. However, aid seems to increase private and public investments but raises the risk of conflicts and government consumption quite substantially. Taking into consideration the effect of the overall policy environment (CPIA),7 the monotonic effect on growth disappears and the positive effects on private and public investment are strengthened, the effects on public revenues and government consumption are significantly negative, and the risk of conflict is reduced significantly. Therefore, accounting for policy, the overall net effects of aid on growth, private savings, public savings and public revenues produce a family of non-monotonic, inverted U-shaped (and, at times, mixed) relationships, reflected generally by Figure 2.7. Moreover, accounting for policy, aid is associated negatively and non-monotonically with the risk of conflicts and with government consumption (see Figure 2.8).
Growth
36
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.1
0.12
0.14
0.1
0.12
0.14
External debt
Growth and external debt
Private investment
Figure 2.2
0
0.02
0.04
0.06
0.08
External debt
Private investment and external debt
Public investment
Figure 2.3
0
Figure 2.4
0.02
0.04
0.06 0.08 External debt
Public investment and external debt
37
Revenue
Elbadawi and Ndung’u: Consequences of Conflict
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
External debt Government consumption and external debt
Government consumption
Figure 2.5
0
Figure 2.6
0.02
0.04
0.06 0.08 External debt
0.1
0.12
0.14
Revenue and external debt
These are important results. First, they corroborate the two main theses of the aid-effectiveness literature: that aid is effective in a good policy environment, but is subject to diminishing returns. Second, whereas previous analyses focused on growth, investments and savings, these results corroborate the aid-effectiveness theses for a wider class of economic performance indicators as well as for the risk of conflict. Third, they show the dire consequences of provision of aid in the absence of a good policy and institutional environment (in terms of high risk of conflict, low public and private investment, low public revenue, high government consumption, and possibly an inherently inefficient patronage system). In the absence of a good policy environment aid flows increase the risk of conflict, but in a positive policy environment they reduce the risk of conflict and also have a positive effect on private and public investment.
38
80
Gy, Iy, Ig, Ry (%)
60 40 20 0 –20 –40
0
2
4
6
8
10
12
14
16
18
Ratio of aid to GDP PPP Gy Figure 2.7
Ig
Iy
Ry
Aid effectiveness
Risk of war, government revenue (%)
70 60 50 40 30 20 10 0
0
5
10
15
Aid as a ratio of GDP PPP Pw Figure 2.8
Aid effectiveness
G
20
25
Elbadawi and Ndung’u: Consequences of Conflict
39
5.3 Macroeconomic environment Three macroeconomic policy variables (INF, DNER, FITRINT) are considered. Real interest rate (FITRINT), adjusted for risk of conflict, is associated negatively with private investment and money demand, but has no effect on the other variables. Inflation is represented by its level and a crisis dummy (INFL40) when inflation level is consistently above 40 per cent. The results seem to show that inflation reduces growth, private investment, public investment and money demand, but is associated with a higher risk of conflict, public revenue and government consumption. The change in nominal exchange rate (DNER) has no effect on the other variables except a weak negative effect on money demand. Exports The results suggest that a high share of primary exports (XGDP) is associated with lower growth and private investment. The literature suggests that economies dominated by primary exports tend to have higher public investment, public revenues and government consumption, as well as a higher risk of conflict. It is only the risk of conflict that appears to be significant in the results we report here, but, as the literature suggests, in economies dependent on a precious non-renewable natural resource base (such as oil), these effects tend to be much stronger. However, this version of the model fails to pick up the non-monotonic effect discussed in the economics of conflicts literature (for example, Collier and Hoeffler, 2001), implying that countries very rich in such resources are likely to be safe from the risk of civil war because they have the capacity to build sufficient deterrence, such as high public investment. Political rights The results suggest that high standards of political rights are associated with lower risks of conflicts, as predicted in the literature (see Elbadawi and Sambanis, 2001), but also with high levels of growth, private investment, public investments and public revenue as well as government consumption. The implication of this is that countries with high standards of political rights have superior economic performance and a lower risk of conflicts because the legitimacy of their governments allows the mobilization of resources for financing larger public investments.9 However, these results fail to pick up the effects of POLITY that would run from public revenues to growth, private investment and public investments that would strengthen the indirect channels. Social fractionalization Consistent with some findings in the literature on the determinants of civil conflicts,10 these results find a strong positive but non-monotonic relationship
40 Thematic Issues
between social fractionalization and risk of conflict. Moreover, there is a family of non-monotonic relationships between social fractionalization and economic performance. For example, growth, private investment and public investment, as well as public revenue, all have a U-shaped relationship with social fractionalization (see Figure 2.9). On the other hand, as in the case of the risk of conflicts there exists an inverted U-shaped relationship between government consumption and social fractionalization (see Figure 2.10). The figures reflect the net effect of social fractionalization, which also account for the incremental influence of the combination of low standards of political rights (1-POLITY) and social diversity. This latter effect brings home the fact that, in the absence of political rights, socially fractionalized societies are particularly vulnerable to the higher risk of conflict (as can also be seen with risk of conflict increasing with the number of ethnic groups) and economic development failures. The results suggest that socially diverse societies with high standards of political rights and other favourable initial conditions do well, but socially polarized societies are predicted to have a much harder time. The peace dividend The extent of the peace dividend following civil wars is modelled by Collier (1999) through the effect of LEGACY on economic growth, where he defines
25
Ig, Gy, Iy, Ry
20
15
10
5
0
0
20
40
60
80
Ethnic fractionalization Gy Figure 2.9
Iy
The consequences of social diversity
Ig
Ry
100
120
Elbadawi and Ndung’u: Consequences of Conflict
41
Risk of conflict, government revenue
6 5 4 3 2 1 0
0
20
40
60
80
100
120
Ethnic fractionalization Pw Figure 2.10
G
Consequences of social diversity
the former as the number of months during the first decade following the cessation of conflict multiplied by the total length (in months) of the previous war. Contrary to Collier (1999), we find LEGACY to be associated positively with private investment. However, when LEGACY is crossed with AIDY, it is found to be associated negatively with private investments and has no effect on other variables. This suggests that, following long wars, growth is driven primarily by capital accumulation efficiency, as one would expect. We have included other variables which define the risk of conflict for CPC countries; for example, we have POSTC, which depicts countries considered to be post conflict, and WLENGTH to capture the effects of prolonged conflict. The variables, as expected, have the effect of increasing conflict and have negative effects on growth and private investment. Money demand and fiscal policy implications Money demand is driven by the standard variables, but in addition, we have all the other variables in the system entering the model. The results show that aid increases money demand but, in a poor policy environment, money demand is affected negatively. External debt and debt stock reduce money demand. What is interesting is that the variables that seem to increase the risk of conflict enter the money demand equation and seem to support Collier and Gunning’s (1995) thesis that portfolio substitution,
42 Thematic Issues
high liquidity and fiscal demand after the end of conflict make money demand erratic. On the fiscal side, the success of government revenue generation is important in financing government consumption and predicting future public revenues. Thus lagged effects of public revenues have significant consequences for current revenues, risk of conflict, private investment, government consumption and money demand, reflecting the complicated web of policies required for post-conflict economic recovery.
6
Conclusions
This chapter has attempted to replicate and strengthen the results of the recent literature on growth and poverty eradication in conflict and post-conflict countries, and to try to show new direct and indirect transmission channels through which risk of conflict is mediated in CPC countries. This strengthens the discussion of the recovery agenda in CPCs. There are three sets of conclusions. First, we have corroborated the evidence of previous studies even when we have calculated sets of equations that take into account multi-causal structures in the variables. The results strongly support the conclusions of the recent literature – for example, that a good policy environment will make aid more effective, and that the effect of aid is subject to diminishing returns. But the results in this chapter provide stronger evidence, suggesting that single-equation estimation may have underestimated these effects. Second, we have shown that single-equation models estimated in the recent literature may have underestimated the strength of the transmission channels from growth to conflict via other intermediate variables and policies. This may also imply that the policy discourse from these sets of results can include a variety of options that will lead to an effective postconflict recovery. Finally, there are new channels of transmission that are likely to be important in economic policies and the development path of post-conflict recovery: • The non-monotonic effect of, for example social fractionalization, on risk of conflict, aid effectiveness and debt overhang on growth and private investment, found in the recent analyses, apply not only to these variables but can be generalized to a wider class of performance indicators. • The degree of political rights has a stronger effect on economic performance through its effect on public revenues. • Public investment, rather than overall growth, is the main channel through which the impact of conflict is mediated to the rest of the economy. This insight provides a useful handle for reducing risk of conflict and revamping growth and investment.
Elbadawi and Ndung’u: Consequences of Conflict
43
• Public investment, risk of conflict and public revenues jointly provide multiple channels through which the effects of risk of conflict on political rights, on the overall economic and institutional policy environment and on primary export performance, are transmitted. • Absence of political rights in socially fractionalized societies makes them particularly vulnerable to a higher risk of conflict, compounds the development failures of poverty, and the collapse of law and order. The impact is further compounded when an absence of political rights and the number of ethnic groups have a combined effect on increasing the risk of conflict. • Exports are associated with high public revenues. The results show that exports seem to increase the risk of conflict: exports of primary extraction owned by the state increase public revenues, providing multiple channels where the effect of primary exports on risk of conflict is compounded by its effect on public revenues and public investment. Notes 1 The World Bank publishes the book on behalf of five development institutions, which include in addition to the Bank, the African Development Bank, the Economic Commission for Africa, the African Economic Research Consortium and the Global Coalition for Africa. 2 See, among others, Collier and Hoeffler (1998, 2001), Collier (1999) and Collier, Hoeffler and Pattillo (1999), to mention just a few. 3 This model allows a choice of policy mix in the development agenda of post-conflict recovery. 4 In the period 1960–99, sixty-six countries experienced at least one episode of civil war. 5 Data for CPC countries in the SSA region until 1997 are available on request from the authors. These include indicators of total external debt to GDP ratio; overall surplus/deficit as a share of GDP; general government consumption as a share of GDP; FDI as a percentage of GDP; gross domestic investment as a percentage of GDP; gross domestic savings as a share of GDP; and real growth of GDP at market prices. 6 An earlier draft of this chapter presented a model consisting of six behavioural equations and one identity, and the reduced-form equations derived from these. The model and the reduced-form equations, together with further details of the endogenous and exogenous variables, are available from the authors. 7 On the other hand, the influence of CPIA through aid dominates its independent effect. At least in the cases of low income countries, this result appears plausible. 8 t-ratios are available from the authors. 9 Azam (2001) notes that, under President Houphouët-Boigny, the Ivorian government bought peace by engaging in an active policy of visible redistribution across regions via public investment. For example, he notes that the taxation of cocoa and coffee was used to fund, among other public investments, the development of the San Pedro harbour in the south-west region, the area most hostile to national unity. 10 Collier and Hoeffler (1998) provide the arguments for the positive but nonmonotonic association between social fractionalization and the risk of civil wars (see also Elbadawi and Sambanis, 2001).
44 Thematic Issues
References Azam, J.-P. (2001) ‘The Redistributive State and Conflicts in Africa’, Journal of Peace Research, vol. 38, no. 4, pp. 429–44. Burnside, C. and D. Dollar (2000) ‘Aid, Policy and Growth’, American Economic Review, vol. 90, no. 4, pp. 847–68. Collier, P. (1999) ‘On the Economic Consequences of Civil War’, Oxford Economic Papers, vol. 51, no. 1, pp. 168–83. Collier, P. and D. Dollar (1999) ‘Aid Allocation and Poverty Reduction’, World Bank Policy Research working paper no. 2041, World Bank, Washington, DC. Collier, P. and J. W. Gunning (1995) ‘War, Peace and Private Portfolios’, World Development, vol. 23, no. 2, pp. 233–41. Collier, P. and A. Hoeffler (1998) ‘On the Economic Causes of Civil War’; Oxford Economic Papers, vol. 50, no. 4, pp. 563–73. Collier, P. and A. Hoeffler (2001) ‘Greed and Grievance in Civil war’, working paper no. 2355, World Bank, Washington, DC. Collier, P., A. Hoeffler and C. Pattillo (1999) ‘Flight Capital as a Portfolio Choice’, World Bank Policy Research, paper no. 2066, World Bank, Washington, DC. Collier, P., A. Hoeffler and M. Söderbom (2002) ‘On the Duration of Civil War’, World Bank, Washington, DC (mimeo). Elbadawi, I. A. and N. Ndung’u (2000) ‘External Indebtedness, Growth and Investment in Conflict and Post-Conflict Countries’, paper presented to World Bank Conference ‘The Economics of Political Violence’. Elbadawi, I. A. and N. Ndung’u (2001) ‘The Economics of Civil Wars and PostConflict Recovery’, paper presented at World Bank Conference ‘Civil Wars and Post-Conflict Transitions’. Elbadawi, I. A. and N. Sambanis (2001) ‘How Much War Will We See? Explaining the Prevalence of Civil War’, Journal of Conflict Resolution, vol. 46, no. 3, pp. 307–34. Elbadawi, I. A., B. Ndulu and N. Ndung’u (1997) ‘Debt Overhang and Economic Growth in Sub-Saharan Africa’ in Z. Iqbal and R. Kanbur (eds), External Finance for Low-Income Countries (Washington DC: International Monetary Fund). Gleditsch, N. P. and H. Hegre (1997) ‘Peace and Democracy: Three Levels of Analysis’, Journal of Conflict Resolution, vol. 41, no. 2, pp. 283–310. Small, M and J. D. Singer (1994) ‘Correlates of War Project: International and Civil War Data 1816–1992, Inter-University Consortium for Political and Social Research, Ann Arbor, Ml. World Bank (2000) Can Africa Claim the 21st Century? (Washington, DC: World Bank). World Bank (2001) World Development Indicators (Washington, DC: World Bank).
3 Economic Policy in Post-Conflict Societies Paul Collier
1
Introduction
Once a civil war has ended, what economic policies should a society adopt? Should institutions such as the World Bank, which provide policy advice, recommend policies that are distinctive because the society is post-conflict? Should donors modify their own policies – notably their rules of aid allocation – so as to make special provision for countries that are post-conflict? At some general level, the objectives of policy should be common across societies. Regardless of whether societies have recently come out of conflict, governments, and the donors that support them, should presumably seek to maintain peace and to promote growth. Although objectives may be common, policies will legitimately differ. The policies that are successful in one country are likely to bear a family resemblance to those that are successful in others, but they will differ somewhat according to country circumstances. Hence, the policies appropriate for a particular post-conflict society will inevitably depend in part on characteristics specific to that country at that time. The question posed here is not whether policy needs to be tailored to the circumstances of an individual country, but whether policy should differ systematically for a country which has recently come out of civil conflict. Post-conflict policies may need to be distinctive, for three reasons. First, peace may need to be given an unusually high priority in post-conflict societies, and so policy will need to be shaped disproportionately by the needs of peace-building. Second, post-conflict societies may have systematically distinctive economic problems that change the detail of which policy configurations are most effective for growth. Third, the political economy of post-conflict societies may be systematically distinctive, so that the attainment of policy reform in general, and of some reforms in particular, is made easier or more difficult than in other societies. I deal with each of these in turn. Section 2 discusses which government and donor policies are likely to be important for sustaining post-conflict peace, Section 3 discusses which government and donor policies are likely to be important in increasing growth 45
46 Thematic Issues
rates in the post-conflict phase, and Section 4 considers the political economy of policy reform in post-conflict situations. Section 5 concludes.
2
Prioritizing peace
The typical post-conflict society has a high risk of reverting to conflict: a typical society emerging from conflict has a 50 per cent chance of reverting to conflict within five years. This alarmingly high risk is made up of two approximately equal components. One is the risks that generated the previous conflict; usually, these risks will have remained. The other component is made up of those risks that were generated during the previous conflict. If peace is maintained, this second group of risks gradually recede, so that after around twenty years the society recovers to the level of risk it had prior to the conflict. The risks that caused the original conflict will differ, society by society. However, Hoeffler and I found that four risk factors are usually important: (i)
countries with a low per capita income face a much higher risk of conflict – and a doubling of per capita income approximately halves the risk. This is why conflict is common in low-income countries; (ii) countries with slow or negative growth face increased risks. In any five-year period, an additional percentage point on the growth rate reduces the risk of conflict by around one percentage point. For example, the catastrophic economic decline in Zaire during the 1990s may have contributed substantially to the civil war of the late 1990s; (iii) high rents from primary commodity exports increase the risk of conflict substantially. For example, the rents from oil and diamonds in Angola evidently contributed to its civil war; (iv) the structure of the society; countries in which the largest ethnic group constitutes a majority of the population (somewhere in the range 45–90 per cent) have around double the risk of conflict compared to countries with either more ethnic diversity or ethnic homogeneity. Evidently, in such circumstances the largest ethnic group can, in a democracy, be permanently dominant. Such ethnic dominance is less common in Africa than in other developing regions, but some countries, such as Rwanda and Burundi, are characterized in this way. Hence, part of post-conflict policy for peace is to reduce these risks that were inherent prior to the conflict. Since three of these risks concern economic performance, economic policies are directly important for risk reduction. Raising the growth rate will reduce risk directly, and further, will reduce risk cumulatively by raising the level of income. The policies appropriate for rapid growth in post-conflict societies are discussed below.
Collier: Economic Policy in Post-Conflict Societies
47
Additionally, policy may need to be aimed at managing the risks from high rents from primary commodities. In the short term, the best strategy is to reduce popular suspicions that such rents are misappropriated by setting up transparent systems of scrutiny and control. The involvement of civil society in the scrutiny of the revenues from the Chad–Cameroon pipeline is a pioneering example of such an attempt. In the long term, it is desirable to diversify the economy away from high primary commodity dependence. Research has shown that the same economic policies that promote growth also facilitate diversification (Collier and Hoeffler, 2002a). The remaining risk factor, ethnic dominance, cannot be dealt with simply by economic policies. Evidently, in such situations minorities require credible guarantees that their rights will be respected. In post-conflict situations, constitutional guarantees or other means of government commitment may have limited credibility to minority groups, and there may be a need for international intervention to provide reassurance. I now turn to the risks that are generated during the conflict. These arise partly because of the build-up of hatreds, and partly because of the build-up of interests. It is unsurprising that civil war polarizes society, making cooperative behaviour more difficult. Various strategies can be used to make coexistence easier, recognizing that the starting position is usually an extreme lack of trust. One approach is to agree on sharing formulas for government revenue between regions, and for public-sector jobs between ethnic groups. For example, following the Biafra war, Nigeria adopted a revenue-sharing formula that has been successful in avoiding further secessionist conflict. Even Switzerland has an explicit formula for sharing civil service jobs among the three main ethno-linguistic groups in Swiss society. The other way in which risks are generated during conflict is that some participants in the conflict come to have an interest in continuing that conflict and so may make active attempts to undermine any peace settlement. Although the divided society as a whole loses from conflict, some participants do well out of war. Further, they accumulate skills and armaments that have few uses other than in the context of conflict. There is often only a fine line between the activities of rebel groups and of organized criminal extortion rackets. One of the central tasks of a post-conflict government is to provide ‘carrot and stick’ incentives for such groups to integrate into the political and economic opportunities of a peaceful society, rather than to revert to violence. A good example of political integration was the change in RENAMO from being a rebel military organization in Mozambique into a democratic political party. A common source of post-conflict risk comes from diasporas, sometimes in neighbouring countries and sometimes in developed countries. Diasporas may harbour grievances and help to finance extremist organizations. A useful approach to reducing this risk is to provide the diaspora, and in particular diaspora organizations, with a recognized and positive role in
48 Thematic Issues
rebuilding the society. For example, in the current rebuilding of Afghanistan, the Internet has been used to create a network of the Afghani diaspora, classified by skills. The diaspora is often a major resource for building business networks. By giving it a formal role in the peace process it is more likely to be harnessed for economic recovery as opposed to subversion. In addition to policies that are targeted at a specific component of risk, such as ethnic dominance or diasporas, some policies may have overall effects on risk. Here I consider military expenditure and aid. Military expenditure is initially high following a civil war. Globally, during a civil war military expenditure rises by around 1.9 percentage points of GDP, compared to its usual level of around 3.3 per cent. Following conflict this is a considerable fiscal burden, and an important question is whether this level is appropriate, given the high risk of further conflict. It might be that continued high military expenditure is an effective way of maintaining peace. Governments not only increase military expenditure during a civil war, but also in response to a risk of civil war. However, such deterrent expenditure appears to be ineffective. A high level of military expenditure does not, in general, reduce the risk that a civil war will be initiated (Collier and Hoeffler, 2002b). There are various reasons why this might be the case. Much military expenditure is often highly inefficient: the military is the least scrutinized part of the public sector. Further, there may simply be insufficient intelligence information to make good use of military forces in the inception stages of a rebel organization. Finally, counter-insurgency operations are extremely difficult. The most common military error is to over-react, harming the civilian population and creating resentments that assist rebel recruitment. Most armies lack the tight discipline necessary to avoid such behaviour. Hence, even when government forces engage with incipient rebel forces they may be as likely to increase the problem as to reduce it. Evidently, if rebel forces escalate, then the government has no alternative but to counter the threat with military force. However, military deterrence appears to be an ineffective strategy for securing post-conflict peace. Economic and political strategies are likely to be more effective. Aid can affect the risk of conflict through various routes. It has effects on growth (discussed below) and thus has indirect effects on the risk of conflict. Here I consider whether aid has a direct effect on the risk of conflict, over and above these growth effects. Some analysts have suggested that aid might increase the risk of conflict directly by increasing the resources to be contested (Grossman, 1992). Others suggest that aid can enable the government to placate the opposition (Azam, 1995). Collier and Hoeffler (2002a) introduce aid directly into their model of conflict risk, controlling for economic outcomes, and find that there is no significant effect one way or the other. Hence, on the evidence, aid does not have strong, direct peacebuilding effects other than via its effect on economic performance, but equally, nor does it increase the risk of conflict.
Collier: Economic Policy in Post-Conflict Societies
49
Concern has also been expressed as to whether aid inadvertently finances military expenditure. Collier and Hoeffler (2002b) investigate this by introducing aid into a military expenditure function. Military expenditure varies considerably between countries, and these differences are well-explained by factors such as the previous history of conflict, and the military expenditure of neighbours. However, aid does not influence military expenditure significantly. This is surprising, since it might be expected that aid would be sufficiently ‘fungible’ so that some of it would finance indirectly the military component of the budget. This is especially the case since military expenditure is income-elastic, rising more rapidly than national per capita income as a whole. Evidently, donor pressure to moderate military spending is normally sufficient to counter this tendency. The most important conclusion from this discussion of post-conflict risks to peace is that rapid economic recovery is likely to be important in sustaining peace. Although in some situations there may be trade-offs between growth and peace, in most cases, policies that augment growth will tend to build peace. Aid does not generally have adverse consequences for peace and so can be determined sensibly by the ability of the economy to absorb it productively in a growth process, a matter discussed in Section 3. Military expenditure retards growth without deterring rebellion and so the high levels found in post-conflict societies are likely to be dysfunctional.
3
Rebuilding the economy
I now turn from the objective of maintaining peace to that of rebuilding the economy. During a civil war the economy typically contracts, either absolutely or at least relative to its potential. It also changes its structure: commonly there is a retreat into subsistence. All forms of capital are reduced: private investment collapses and is replaced by capital flight; public capital, such as infrastructure, is destroyed; and ‘social capital’, such as trust and reputation, is lost. The post-conflict society is thus short of both private and public capital, and is characterized by opportunistic behaviour. Post-conflict economies have wide variations in performance. Some have recovered very rapidly, while others have continued to decline virtually as if the war were continuing. A critical aspect of performance is whether people come to believe that peace will be maintained. Confidence in peace is selffulfilling, since it encourages private investment, and the resulting growth tends to consolidate the peace. A useful indicator of success in restoring confidence is whether flight capital returns. Typically, in post-conflict societies a large part of private wealth has been taken out of the economy. For example, in Uganda, by the end of the period of social disruption (1986) over 60 per cent of private wealth was held outside the country. By taking strong action to restore confidence in property rights and other reforms, such as restoring a competitive exchange rate, by the early 1990s this flight
50 Thematic Issues
capital was starting to return to the economy. In some years, capital repatriation was larger than exports. Globally, both policy reform and aid are important motors of economic growth. They are complementary – aid is more effective in a better policy environment, and policy reform is more effective in conjunction with large aid inflows (Collier and Dollar, 2002). In post-conflict situations, the complementary role of aid and policy reform is even more important (Collier and Hoeffler, 2002c). Other than in post-conflict situations, the relationship between aid and growth is estimated to be: dG/dA = 0.127P − 0.056A
(1)
where G = the growth rate of GDP; A = aid as a share of GDP, measured at purchasing power parity exchange rates; and P = the World Bank index of policy, the ‘Country Policy and Institutional Assessment’, with 1 as very poor policy and 5 as very good policy. The relationship described in Equation (1) tells us that there are diminishing returns to aid, but that aid is more effective in a better policy environment. For example, a fairly typical policy score is 3, and a fairly typical aid inflow is 3 per cent of GDP at purchasing power parity prices (equivalent to around 9 per cent of GDP at market exchange rates). For such a country, the marginal percentage point of aid raises the growth rate by around 0.2 percentage points. Hence, the marginal rate of return upon aid is around 20 per cent. For a country during the first decade of post-conflict, the relationship between aid, policy and growth is much more favourable, being: dG/dA = 0.313P − 0.056A
(2)
Hence, for the same values of the policy score and aid, the marginal contribution of aid to the growth rate would be around 0.77 percentage points. Therefore the marginal return on this volume of aid in a post-conflict situation would be around 77 per cent. An implication of this is that donors should allocate aid disproportionately to post-conflict situations. In practice, historically, this has not been the case: donors have given about the same amount of aid to post-conflict countries as to other countries. The appropriate amount of aid a postconflict country should receive, however, is a difficult issue. But there is a reasonable presumption that the lower bound for an allocation is that in which the marginal contribution to growth would be approximately equal to that of aid in non-conflict countries. This is a lower bound because, in post-conflict settings, aid may have a peace-stabilizing effect over and
Collier: Economic Policy in Post-Conflict Societies
51
above its effect on growth, although, as discussed above, we did not find evidence for such an effect. On the criterion of growth alone, aid as a share of GDP in post-conflict situations should be approximately double that in non-conflict situations. We can derive this from Equations (1) and (2) by introducing the concept of the saturation point where the society has so much aid that diminishing returns have driven the marginal contribution of aid to growth to zero. The saturation point is determined by setting dG/dA = 0, and solving for A in terms of P. For non-conflict situations, the saturation point is around 2.5P, whereas for post-conflict situations it is around 5.6P. Thus, for a common level of policy, the maximum volume of aid that is useful in a post-conflict setting is over twice the normal level. In the typical post-conflict setting, policy is initially considerably worse than normal, and this qualifies the amount of aid that can usefully be absorbed. However, even once we introduce a discount to correct for worse policy, aid should still be approximately double its normal level. This suggests that the donor community could improve its performance by allocating more aid to post-conflict countries. The same analysis also provides some guidance as to how these additional flows of aid should be timed. For the first three or four years after the end of conflict, the economy does not have any above-normal capacity to absorb aid. The extra scope for aid is approximately from the fourth or fifth year through to the end of the first decade. This pattern may be a result of the general disarray of public institutions in the initial post-conflict years which makes it difficult to absorb large volumes of aid productively. Hence, post-conflict aid should taper in gradually over a decade. In practice, aid flows have had precisely the opposite time profile. There is typically a surge of aid during the immediate post-conflict period, followed by a rapid decline after the first three years: aid tapers out just as it should be tapering in. Of course, there may be good reasons other than growth for aid to be high in the early post-conflict years. However, on the evidence, the present temporal pattern of aid is inappropriate to the problem of restoring post-conflict societies. The appropriate response is probably not to forgo the initial surge of aid. Politically, this is the best opportunity to get higher levels of aid commitment. Rather, donors should couple these large commitments with an understanding that the lag between commitment and disbursement will be much longer than in normal situations. Just as the relationship between aid, policy and growth can be investigated to determine the marginal impact of aid on growth, so it can be used to investigate the contribution of policy improvement to growth. Other than in post-conflict situations, the relationship between policy and growth is estimated to be: dG/dP = 1.017P + 0.127A
(3)
52 Thematic Issues
In post-conflict situations this becomes: dG/dP = 1.017P + 0.313A
(4)
An implication is that good policy is even more important for growth in post-conflict situations than in other contexts, and that the extent of this difference depends on the amount of aid received. Suppose that donors were to get aid inflows ‘right’ in the sense discussed above, so that they were approximately double in post-conflict situations to what they were in non-conflict situations – say 6 per cent of GDP (at purchasing power parity exchange rates) versus 3 per cent. In a non-conflict country a one-point improvement in the five-point policy scale used by the World Bank would raise the growth rate by around 1.5 percentage points. By contrast, in a post-conflict society, the same one-point improvement would raise the growth rate by around 3 percentage points. Thus, if donors get aid policy right in post-conflict societies, policy improvement will have double the effect on growth that it has in normal environments. The conjunction of both effective aid and policy in post-conflict situations usually leads to substantially above-average rates of growth during the first decade of peace, typically by a little over one percentage point. All this extra growth occurs from around the fourth year of peace through to the end of the first decade, with a peak around the sixth or seventh year. The policy scoring system of the World Bank is an average of four components of policy: macro, structural, social and governance. At the time of writing, twenty different aspects of these policies are considered. On the whole, good policies in one area tend to be associated with good policies in another, although analysis of the components suggests that many of these policies matter for the growth process. I now turn to the question of whether any of these policies are particularly important for growth in a post-conflict situation. That is, how might policy priorities differ if a society is post-conflict and growth is the objective? The World Bank policy index can be used to investigate this question. However, the governance component is too recent an addition to the index for it to be used in the analysis of growth. For the other three components – macro, structural and social – there are consistent data through the decade of the 1990s that can be used to investigate whether any component is differentially important in post-conflict situations. Collier and Hoeffler (2002c) find that of the three, social policies are differentially important in post-conflict societies, whereas macroeconomic policies are differentially less important. This does not mean that macroeconomic policies do not matter, or even that in some absolute sense they are less important than social policies. What the result shows is that, using as a benchmark the relative importance of the three policies in normal situations, in post-conflict societies social policies are more important and macro policies less important. Thus,
Collier: Economic Policy in Post-Conflict Societies
53
for example, it may be the case that in all situations macro policies are in some sense more important than other policies. The World Bank’s main indicators of social policies are the inclusiveness of education and health policies, and the provision of social protection. Superficially, it is surprising that these policies are important for growth in post-conflict situations. Evidently, the effect of education on growth is only long term, whereas the results suggest that social policies have an effect on growth within four years. What might be the channel by which social expenditures affect the growth rate? The most likely route is, perhaps, that inclusive social policies act as a signal to the population regarding the government’s intentions. If the government signals inclusion, then the population recognizes that peace is more likely to be maintained and is therefore more willing to invest. Recall that the restoration of confidence is likely to be extremely important in post-conflict situations. During a successful restoration, like that in Uganda, the investor risk rating has improved from only 5 on a scale of 0–100 in 1992 (at the time the lowest in all Africa) to 23 by 2002, at which time it was well above the African average. The move to universal primary education during the first peaceful and democratic presidency in the country’s history may have played a part in this remarkable recovery.
4
The political economy of policy reform
Post-conflict societies come out of conflict with significantly worse policies than the average for low-income countries. Using the World Bank’s policy scale, the average for all low-income countries during the 1990s was around 3.00, while the average for post-conflict countries during the first three years of peace was only 2.41. A one-point difference in the policy scale indicates a substantial difference in policies, so this 0.59 difference matters, although the fact that countries coming out of conflict have poor policies is not surprising. Should governments attempt rapid reform, or is the political situation so delicate that reform risks a resumption of conflict? Or should they accelerate reform because of its atypically important role in the growth process? Obviously, such delicate matters can only be decided by a careful analysis of the particular situation. However, the global pattern of policy reform post-conflict provides some useful guidance. First, I consider whether policy reform appears to increase the risk of conflict. Introducing both policy and the rate of change of policy into the previously discussed model of conflict risk, we find that neither the level of policy nor an improvement in policy have significant direct effects on the risk of conflict. All their important effects work through their impact on economic performance. As discussed above, policy improvement raises growth, and growth reduces the risk of conflict. Hence, policy reform should not, in general, be delayed from fear that it will lead to renewed conflict.
54 Thematic Issues
The actual behaviour of post-conflict governments bears this out. By around the end of the first decade of peace, policy improves dramatically to 3.05. Hence, during the first decade, those post-conflict governments that had managed to sustain peace had also managed to improve policies from the very poor initial level of 2.41 to being a little better than the average for low-income countries. This is a far more rapid rate of policy improvement than is found on average in non-conflict situations. Hence we can conclude that the first decade after conflict is politically a relatively easy time for a government to introduce substantial policy reforms. This is indeed consistent with theoretical analysis in political science, which suggests that the vested interests that normally block reform are weakened by periods of civil war. Further, since growth is typically more rapid in this period, this in itself may make reform easier. However, while overall policy improvement is impressive, the composition of that improvement shows no signs of being skewed towards social policies, which our previous analysis suggests should be differentially a priority. Rather, all three groups of policies tend to improve in step with each other. Finally, I turn from the level of the three broad policy groups to consideration of three specific policies that are likely to be differentially important in post-conflict settings. These are demobilization, the restoration of the rule of law, and the restoration of a formal economy. Demobilization is desirable in that, typically, the government and the rebel army will be merged, producing a very large army at a time when the need for a large army is reduced by the return to peace. However, demobilization can potentially pose a threat to both the security of ordinary households, if demobilized soldiers resort to crime, and to the security of the state if they become a recruitment pool for subsequent rebellions. With regard to the impact of demobilization on crime, there is some evidence from the effect of the Ugandan demobilization of the early 1990s (Collier, 1994). Despite widespread fears that this demobilization would lead to an escalation in crime, the detailed evidence suggested that it had no effect, with one exception: where soldiers were demobilized into districts where there was a severe land shortage, so that many of the soldiers lacked access to land, then crime in those districts escalated. Hence, it is probably important to ensure that demobilized soldiers have adequate access to land. On the threat to the security of the state there is no systematic evidence. While there may be some risk of recruitment into rebel organizations, the latter more typically target youth. Further, there may be more of a threat to the state from an excessively large regular army that is consequently badly paid and underemployed. Given the lack of evidence for a deterrence effect from military expenditure, there is on balance a case for early and substantial demobilization. The restoration of the rule of law typically involves early action on property rights, followed by a gradual improvement in the functioning of the courts and the police. Civil wars generally leave a legacy of confused property
Collier: Economic Policy in Post-Conflict Societies
55
rights as different vintages of powerful groups lay claim to the same property without a chain of legal transfer of title. In this environment there are often also opportunistic claims to property. While property rights are confused, investment in physical assets will be curtailed because investors cannot be sure that their titles will be upheld. A major success of the Ugandan recovery was the handling of these disputed claims. The key step was to establish a credible process for the adjudication of claims and the restitution of property, but one that was strictly time-limited. All claims had to be filed with the court by a set date to be considered, and decisions were rapid. By this means the Ugandan government restored the market in property quite rapidly while at the same time building a good image with the foreign investor community. The highly opportunistic environment typically following civil wars makes it both important and difficult to restore the courts. For example, until the banks are able to foreclose reliably on assets pledged as collateral, default rates are likely to be too high for the banking system to be able to lend profitably. Commonly, the police force will have been run down during the conflict. The rebuilding of an honest police force is important, both for the functioning of the courts and for the curtailment of opportunism. However, it is itself made difficult by the prevalence of opportunism. Police work offers many opportunities for corrupt income, and a police force can easily become part of the problem that it is designed to reduce. During a civil war the economy becomes less formal. This has consequences for the taxable base – most economic activity is not taxable. This is one reason why aid is so important in the early post-conflict years. Without aid, the government is faced with the choice between extremely low levels of expenditure and extremely high rates of taxation on the small part of the economy that is taxable. High tax rates are likely both to limit the growth of formal sector enterprises and to retard the repositioning of enterprises from the informal sector into the formal sector. Thus, in post-conflict societies, aid can legitimately be used to reduce the burden of taxation. During this phase the taxable base of the economy recovers so that when taxes are eventually imposed revenue will be sufficient to meet the public expenditure needs of the society. Premature attempts to tax the fragile and small formal economy risk delaying fiscal sustainability.
5
Conclusion
Post-conflict societies are distinctive. They require different policies for the maintenance of peace, which needs to be a much higher priority than in countries without a recent history of conflict. They require different donor policies and different government policies for the restoration of growth. To date, there is evidence that both donor and government policies have often been inefficient. Military expenditure has been too high, whereas aid has been too low. If this donor behaviour is based on a fear that aid will leak into
56 Thematic Issues
military expenditure, such fears seem on average to have been misplaced. Aid has tapered out too early in the post-conflict process. Governments have been correct to implement rapid reform, but these reforms may not have been sufficiently tailored to the special needs of post-conflict societies. References Azam, J.-P. (1995) ‘How to Pay for the Peace? A Theoretical Framework with References to African Countries’, Public Choice, vol. 83, issue 1–2, pp. 173–84. Collier, P. (1994) ‘Demobilization and Insecurity: A Study in the Economics of the Transition from War to Peace’, Journal of International Development, vol. 6, no. 3. Collier, P. and D. Dollar (2002) ‘Aid Allocation and Poverty Reduction’, European Economic Review, vol. 46, no. 8, pp. 1475–500. Collier, P. and A. Hoeffler (2002a) ‘Aid, Policy and Peace: Reducing the Risk of Conflict’, Defence and Peace Economics, vol. 13, no. 6, pp. 435–50. Collier, P. and A. Hoeffler (2002b) ‘Military Expenditure: Threats, Aid and Arms Races’, Development Research Group, World Bank, Washington, DC (mimeo). Collier, P. and A. Hoeffler (2002c) ‘Aid, Policy and Growth in Post-Conflict Societies’, Development Research Group, World Bank DC (mimeo). Grossman, H. I. (1992) ‘Foreign Aid and Insurrection’, Defence Economics, vol. 3, pp. 275–88.
4 Ethnicity, Institutions of Governance and Conflict Avoidance Mwangi S. Kimenyi
1
Introduction
Of the many challenges facing Africans, the establishment of durable peace is probably the most critical, yet the most difficult to achieve. According to some estimates, one-fifth of the people on the continent live in countries involved in conflict. It is now well established that conflicts in Africa have been a major hindrance to the improvement of the wellbeing of Africans (World Bank, 2000). In addition to costs such as loss of life, injuries and the general disruption of economic and social activities (Stewart, 2000; and Elbadawi and Ndungu, 2001), investment flows are damaged as a result of the increased risk premium. They have an impact on human development by disrupting investment in human capital and the provision of basic services, and increasing mortality rates. Civil wars are also associated with a prevalence of child labour and the disruption of social networks, and they weaken governments’ capacity to deliver crucial services. Conflicts are also associated with disease outbreaks that have a negative impact on the quality of life. In addition, conflicts involve collateral damage to neighbouring countries as a result of the inflow of refugees and guns. To the extent that conflicts disrupt investment in human capital, their impact on development tends to be long term because generations of young people miss education opportunities, as has been the case in both Angola and Sudan. Furthermore, conflicts have a major impact on investment flows because of increased risk premiums. The various development targets, such as those articulated in the Tokyo International Conference on African Development (TICAD-II), or the more recent Millennium African Renaissance Programme, are unlikely to be achieved unless there is a marked reduction in the incidence, severity and duration of civil wars. Post-independence conflicts have been relatively more intense and have entailed enormous costs; it is estimated that conflict lowers economic growth by two percentage points every year. Additionally, conflicts divert huge resources from development purposes. In Central and West Africa, such 57
58 Thematic Issues
resources have been estimated at US$1 billion and more than US$800 million a year, respectively (World Bank, 2000). These estimates exclude the indirect costs of conflict, such as costs for neighbouring countries generated by refugee flows, and costs arising from impeded communication routes and disruption in economic activity. Post-independence conflicts have generally been between different ethnic groups within a country. Virtually every African conflict has an ethno-regional dimension to it. Even conflicts that are apparently free of ethnic considerations involve factions and alliances built around ethnic lines (Deng, 1997). One of the first post-independence conflicts was the civil war in Nigeria. The conflict arose when the Ibo in the eastern region, fearing domination by other ethnic groups in the country, wanted to secede and form their own Biafran nation. This was prevented by the use of military force. In Rwanda and Burundi, where the Hutu constitute a numerical majority while the minority Tutsi dominate the countries’ economic and political life, interethnic conflicts have been reported as some of the most intense in the developing world; each ethnic group has attempted genocide aimed at the eradication of the other. In Ethiopia, ethnic conflict lasted for hundreds of years as the Eritrean people sought independence from Ethiopia. Similarly, members of the Somali ethnic group sought independence from the Amhara who have for years dominated Ethiopian politics and the economy. Other examples of interethnic conflict deteriorating into wars that have claimed the lives of hundred of thousands of people include Sudan, Mozambique, Somalia, Angola and the Democratic Republic of the Congo (DRC). Clearly, to understand conflicts in Africa necessarily requires an appreciation of ethnic dynamics. In particular, it is important to understand why ethnic groups engage in conflict, and which features of ethnicity make Africa prone to conflict. A number of causes have been advanced for the occurrence and persistence of conflicts. These include natural resource endowment, ethnic and religious hatreds, economic inequalities, lack of political rights and economic mismanagement of the state, among many others (Collier and Hoeffler, 1998 and 2001; Collier, Hoeffler and Söderbom, 1999). While it is a fact that ethnicity does play a part in triggering conflict, it has been established that it is not the mere existence of ethnic diversity that causes conflict but rather that much depends on the institutions of governance. Institutional arrangements play a critical part, not only in determining economic outcomes but also in establishing peace within nations. It is, for example, evident that the economic and political crises that characterize post-independent Africa are often an outcome of poorly-designed institutional frameworks. Specifically, political instability, civil strife and economic stagnation are to a large extent the products of institutions that are not capable of dealing with heterogeneous populations. The failure of political institutions to accommodate diverse interests (ethnic, religious, linguistic) has generated conflict situations that have an adverse
Kimenyi: Ethnicity and Conflict Avoidance 59
effect on political and economic outcomes. It is becoming increasingly clear that African countries are unlikely to achieve stability and economic growth in the absence of major institutional reform. In particular, the greatest challenge facing African countries is the establishment of institutional arrangements that can accommodate ethnic diversity effectively and facilitate the peaceful coexistence of population groups (Kimenyi and Mbaku, 1995). If institutions determine outcomes, then Africa’s problems should not be attributed merely to the existence of extremely heterogeneous population groups but must be seen as the result of institutions that are not well placed to harmonize ethnic claims and result in various undesirable outcomes. Thus the problem is not diversity but rather the existing institutional arrangements. This chapter suggests that the centralized unitary states of Africa have failed to deal effectively with diversity because the institutions have been unable to achieve cooperative agreements among the various groups. Rather, unitary centralized states tended to encourage ethnic competition for resources and power, which in turn led to conflict. The basic argument advanced in this chapter is that ethnic groups are important institutions that organize for the production of ‘ethnic goods’ in order to generate a surplus for their members. This surplus can be increased either through the creation of value (production and specialization) or by seeking transfers from other ethnic groups. Since there are substantial benefits that accrue from associating and identifying with an ethnic group, there is an interest among members in identifying with their group. The ability of an ethnic group to ‘force’ transfers from other groups is largely determined by the institutions of governance. Ethnic conflict is more likely to arise in institutions that make it easy for one group to improve its welfare by transferring surplus from other groups. This chapter contends that the optimal institutional arrangement in nations characterized by strong ethnic identification is one that allows for the establishment of governments based on local ethnic jurisdictions, each possessing a fair degree of autonomy, permitting groups to retain their surplus. In such institutions, groups can increase their wellbeing by improving efficiency of production or by engaging in mutually beneficial exchange with other groups. Section 2 discusses post-independence African institutions and policies that have tended to trigger conflicts. Section 3 outlines various features of ethnic groups in Africa and stresses that they are important institutions that can be used to organize collective activity. Section 4 introduces the idea of functional governments to complement ethnic governments, and Section 5 concludes.
2
Post-independence African institutions and ethnic conflicts
Probably the most important event explaining contemporary African institutions is European colonialism. Colonialists partitioned Africa among
60 Thematic Issues
themselves in a manner that reflected their spheres of influence, paying little regard to ethnic heterogeneity or hitherto existing institutions of governance. As a result, different ethnic groups, each with their own language, culture, religion, traditions, customs and political and economic systems, were brought together to form a single political and administrative unit. For example, in Nigeria, 400 ethnic groups were combined to form one country. The partition of Africa also resulted in situations whereby members of the same ethnic group were separated and placed in different countries. Members of the Somali community, for example, found their members dispersed between Somalia, Ethiopia and Kenya. The Yoruba and the Aja were divided between Nigeria, Benin and Togo; the Wolof and the Serers between Senegal and the Gambia. The political units that were created were not only arbitrary but also lacked any strong unifying factors. Partition disrupted long-established kingdoms and tribal governments, and disrupted existing trade networks and other inter-ethnic linkages. Furthermore, colonial rule did little to develop democratic institutions in Africa, the colonialists being interested mainly in maximizing economic gain. Members of certain tribes were favoured over others – a policy of divide and rule that weakened the opposition to colonial rule. The ethnic divisions thus created were manifested during the struggles for independence and continue to be important today. In fact, ethnic differences in levels of education and income that exist today are a legacy, in large part, of the colonial experience. The post-colonial African leaders sought to retain the unitary governments bequeathed by colonialists. They argued that for the various ethnic groups to coexist peacefully, it was necessary to homogenize otherwise heterogeneous groups. Unifying the groups into one national state was considered to be crucial for political stability. Additionally, the centralization of power allowed for the adoption of uniform policies and the balancing of economic resources. Similar arguments were advanced for the preference for singleparty political systems where all ethnic groups would be joined together in one party. Leaders warned that unity would be compromised if political party competition were to be introduced, because different political parties would be dominated by particular tribal and linguistic groups, essentially promoting tribalism. The leaders therefore sought ways for the diverse ethnic groups within each country to come together in a unitary state. The approaches to coexistence that were adopted involved suppressing expressions of ethnic identity, ostensibly to promote national unity and integration. Thus the common response to diversity in many African countries has been the adoption of institutional arrangements that seek to unify heterogeneous populations by limiting expressions of group differences (Kimenyi, 1997a). African leaders have therefore been eager to weaken the tribe as a unit of collective choice, as the tribe is considered to be harmful to peaceful coexistence. Determined to preserve national unity, leaders sought to establish unitary states with power concentrated at the centre. The result was the
Kimenyi: Ethnicity and Conflict Avoidance 61
creation of institutions where the head of state held the power over all public policies without constitutional constraints on the central authority in its exercise of power. As a result, in most African countries, the state wields immense power over the economy and the allocation of resources. Thus, for the most part, post-independence African institutions were characterized by one-party rule and an excessive centralization of power. The experiment with one-party centralized government has shown that such institutional arrangements are not suited to, or capable of, harmonizing the interests of heterogeneous groups. Since the state is highly centralized, the leader of the ethnic group that captures the state has control of enormous resources. To retain control of the instruments of wealth transfers, those in power have an interest in erecting barriers to entry into political and economic markets. There is ample evidence that many African governments have maintained their control over governance structures by redistributing resources in favour of groups such as the military, urban dwellers and specific ethnic groups. Instead of unifying the various ethnic groups, centralized unitary states have provided leaders with opportunities to extend benefits to some groups at the expense of others (Brough and Kimenyi, 1986; Kimenyi, 1989; Kimenyi and Shughart, 1989). Discriminatory allocation of resources creates dissatisfaction among groups that are not in the ruling coalition, which then becomes a primary cause of ethnic conflict. Thus ethnic conflicts on the continent can be viewed as a reflection of sub-optimal institutional arrangements that make it possible for governments to use discriminatory policies to transfer benefits from one group to another. The ethnic groups that are not part of the ruling coalition are excluded from full and effective participation in economic and political markets. Many turn to violence as a way to minimize further marginalization and to improve their participation in these markets. Given that control of the government has potentially large benefits for members of the ruling ethnic group, they have an interest in taking control of the leadership. Since there are many barriers that make it difficult to remove leaders through democratic processes, one common approach to attaining power has been through military coups. Conclusive evidence of the ethnic orientation of most military coups is revealed by changes in the composition of members of the cabinet and senior civil servants before and after the coup. Between 1960 and 1982 some forty of the forty-five independent countries in sub-Saharan Africa experienced a military coup an attempted coup or a plot to overthrow the government. In almost all cases military leaders awarded top government positions to members of their own ethnic groups. For example, before the 1966 coup in Central African Republic that deposed President Dacko (a member of the Baya group), no member of the Mbaka ethnic group was in the cabinet. After the coup, the composition of the cabinet changed so that 23 per cent of cabinet members were Mbaka.
62 Thematic Issues
Similarly, when Kwame Nkrumah was president of Ghana, 71 per cent of cabinet members were from Nkrumah’s Akan ethnic group. Following the coup organized by members of the Ga and Ewe ethnic groups, representation of Akan people in the cabinet dropped to 25 per cent, and that of the Ga increased from 7.7 per cent to 38 per cent (Kimenyi, 1997a). Such changes in the ethnic composition of the cabinet have also been evident after peaceful transitions of government (Kimenyi and Shughart, 1989). These examples demonstrate the notoriety of ethnic groups of imposing costs on nonmembers. The same ethnic unit that advances the wellbeing of its members can also undermine the liberty of others. In a centralized state, ethnic groups engage in conflict primarily because they are in a struggle to capture the state; they mobilize to control the instruments of government as part of the competition for resources. Control of governance structures allows the controlling group to redistribute income and wealth in favour of its members and supporters (Collier, 1999). Recently, there have been political and economic reforms in most African countries. These reforms have generally taken the form of opening up political markets to allow opposing political parties, thus replacing single-party and military dictatorships. There is little doubt that political party competition is likely to make political leadership more responsive to a wider cross-section of the population. However, the problems that plague the continent are unlikely to be resolved by such reforms. It is possible that political party competition might widen ethnic division unless it is accompanied by other far-reaching institutional reforms. This is because, under a multi-party system, issues take on an increasing ethnic character as power brokers prey on people’s ethnic background to drum up political support. The limited experience of political party competition in Africa provides an example. Political parties have more often than not turned out to be tribal factions that have no clear ideology or plan of action other than transferring benefits to members of their coalition. Following the introduction of multi-party politics in Kenya, political parties tended to form along ethnic lines: the ruling Kenya Africa National Union (KANU) became the party of the Rift Valley; the National Development Party (NDP) evolved as the party of the Luo/ Nyanza tribes; the FORD-Kenya became the party dominated by Luhyas in Western Kenya; while the Democratic Party (DP) and FORD Asili were largely dominated by the Kikuyus of Central Province (Murungi, 1995). The legalization of opposition parties has not resulted in truly competitive political markets that would otherwise serve as effective vehicles in the design of viable constitutional contracts guaranteeing protection of individual and group rights. In almost all countries that have recently entered the era of multi-partyism, the structures of governance have remained much as they were during the single-party era. As a result, African states continue to suffer from the single-party legacy of over-centralization, weakly developed civil societies and a lack of the independent institutions necessary for proper
Kimenyi: Ethnicity and Conflict Avoidance 63
governance. Clearly, conflicts among groups are not likely to be eliminated by political party competition alone and it is conceivable that such conflicts may be intensified by its introduction. In short, democratization as represented by political party competition is an important institutional reform but not sufficient for harmonizing ethnic interests and not necessarily conflictreducing. Thus alternative forms of institutional design are required.
3
Ethnic units of collective choice
The centralized African state has clearly been a failure. Our argument above is that marginal reforms, such as the introduction of political party competition, are unlikely to result in significant changes in terms of the peaceful coexistence of ethnic groups. Instead, we suggest that more radical reforms providing for greater decentralization of political power are necessary. Decentralization must necessarily involve the creation of autonomous ethnic governments. To appreciate the thrust of the argument advanced here, particularly why the ethnic unit is a suitable unit of collective choice, it is important first to appreciate the importance of the ethnic group as an institution as well as the complexity of ethnicity in Africa. One of the most important features of sub-Saharan Africa is the high degree of ethnic heterogeneity. Countries in this region are home to over 2,000 distinct ethnic groups, each having its own language, culture and traditions. Sudan has over 170 distinct linguistic groups, while in Zaire there are over 250 ethnic groups and in Nigeria between 200 and 400 distinct linguistic groups. The size of each tribe varies, some having as few as 100,000 members, while others have millions. One important feature of African societies that makes ethnic units suitable as units of collective choice is geoethnicity: in many cases each group associates with a particular territory. The various ethnic groups have what they refer to as their territory and in many respects members of other ethnic groups are considered to be outsiders. Although the boundaries that demarcate ethnic territories are not always clearly specified, they nevertheless exist and to a large extent are respected by the groups that live next to each other. For example, although the Democratic Republic of Congo (formerly Zaire) consists of over 250 ethnic groups, they can be grouped into 23 distinct units that are closely related in terms of culture and language. As is evident, each of the groups is confined to a clearly defined ethnic territory. Cobbah (1988, p. 73) observes that: In Africa, this ethnic identity is above all other things a territorial identity. Nothing defines the ethnic group better than its ‘standing place’. Thus the term geoethnicity has been used to describe the African ethnic phenomenon. Geoethnicity as opposed to non-territorial ethnic identification involves the historic identification of an ethnic group with a given
64 Thematic Issues
territory, an attachment to a particular place, a sense of place as a symbol of being and identity. Thus, by and large, ethnic units make up what could be called ‘ethnic nations’, which have organized and served their members for generations. Because they occupy a specific territory does not necessarily qualify ethnic groups as appropriate units of collective choice. However, associated with diversity is the fact that ethnic groups have through the years continued to remain distinct, with members identifying strongly with their own tribes. Unfortunately, tribal identification is usually viewed negatively, and African leaders and intellectuals have engaged in highly publicized efforts to denounce this identification as tribalism. Identifying with a particular tribe is considered to be regressive. Tribalism, many claim, has no place in the modernizing nations of Africa. At the national level, identifying with a tribal group is considered to be inconsistent with development. The truth, however, is that tribal attachment is alive and well throughout Africa. By and large, identifying with one’s tribe is highly valued, and resources are devoted to make certain that members of a group continue to identify with that group. Children are taught to identify and be proud of their tribe, ethnic identity is strongly felt, and behaviour based on ethnicity is normatively sanctioned (Horowitz, 1985). The result of strong ethnic identification is that most voluntary and business organizations form along ethnic lines. The majority of urban social welfare organizations, urban soccer leagues and even trade unions are organized along tribal lines. It has been observed that tribally-orientated associations rate among the most important voluntary associations in West Africa (Little, 1957). More telling evidence of the degree of ethnic identification in politics is revealed by voting patterns. No matter which country one looks at, it is clear that people vote very much along ethnic lines; political ideology in the Western sense rarely plays a significant role. The fact that we observe high degrees of tribal identification implies that the expected benefits of identifying with a tribal group exceed the costs of such identification. Because there are benefits associated with identifying with a particular tribe, the tribal leadership has an interest in investing in tribal identification. Kimenyi (1998) presents a simple model of tribal identification based on costs and benefits, as shown in Figure 4.1. The horizontal axis measures the degree to which individuals identify with their tribe, and the vertical axis measures the marginal costs and benefits of ethnic identification. While there are benefits from identifying with one’s tribe, they tend to decrease the more one identifies with it. This is represented by the demand (TI) for tribal identification. Notice that even if the costs of tribal identification were zero, it would not pay to identify completely with the tribe. MC represents the cost of identifying with one’s tribe. Point A is the optimal level of tribal identification (50 per cent in this case) and TAM is the tribal surplus.
Kimenyi: Ethnicity and Conflict Avoidance 65
T
Marginal cost and marginal benefit
C
A
M
0
50 Tribal identification %
80
100 I
Figure 4.1 The market for tribal identification Source: Kimenyi (1998).
This point is optimal because identifying more with one’s ethnic group beyond this point implies that the additional benefits are less than the costs associated with such identification. In essence, beyond point A, the benefits of associating with one’s tribe are less than those that would accrue from identifying with members of other tribes. The idea of maximizing tribal surplus assumes that the main focus is on group welfare and thus benefits are shared by all members. This does not imply that members share the benefits equally, but rather that the group is better off in general. Strong ties within ethnic groups make them ideally suited as units for organizing collective provision. This is because ethnicity economizes on organizational costs. This is consistent with the suggestion by Addison and Murshed in Chapter 1 of this volume, that the primary reason why many civil wars have an ethnic dimension is precisely because ethnicity is efficient at overcoming collective action problems. In particular, ethnic-based institutions have a comparative advantage in solving prisoner’s dilemma problems. It can be argued that the outcomes observed in centralized African states are the result of the fact that the costs of arriving at cooperative outcomes are very high, so that countries ultimately suffer Pareto-inferior non-cooperative
66 Thematic Issues
outcomes. Since members of ethnic groups have a long-term attachment to their own group (for example, through blood lines or past memories), they are more likely to have continuous dealings with members of their ethnic group than with members of other ethnic groups. Continuous dealing reduces cheating in prisoner’s dilemma situations and as a result ethnic groups may be more efficient than the state in the provision of public goods (Roback, 1991). The free-rider problem is likely to be less prevalent when the group is made up of a single ethnic group than when several ethnic groups are involved. If we define an optimal unit of collective choice as one that minimizes the costs of achieving cooperative agreements, we would then expect jurisdictions to vary in both size and composition. Since transaction costs associated with achieving cooperative agreements depend on the composition and size of the decision-making group, those groups composed of homogeneous members face lower transaction costs of achieving agreements than groups whose members are heterogeneous, ceteris paribus. The other important characteristic is that ethnic groups act as interest groups in competing for transfers, maximizing the wellbeing of their members by competing for wealth transfers from other groups. However, they are distinct from other interest groups such as labour unions or producer groups in that membership of an ethnic group is permanent, since entry and exit are not possible (Kimenyi, 1989). The interest group theory of government teaches that public policy outcomes reflect the interplay between demanders and suppliers of wealth transfers. Transfers may involve taxes, subsidies, regulations, public projects or even government jobs. In Africa, though many interest groups compete for wealth transfers, ethnicity represents the most important group. Kimenyi and Mbaku (1995) identify various African interest groups competing for wealth transfers, including farmers and urban dwellers, but they note that the most important characteristic of these groups in Africa is ethnicity. Permanent interest groups attempt to maximize their share of wealth transfers through the political process. To ensure that it maximizes its share of wealth transfers, a group needs to control the instruments of wealth transfers, primarily political leadership. It follows that it is more profitable for groups in African countries to invest to influence government policy than it is for them to invest their scarce resources in the market, where the returns are, in comparison, lower. Not surprisingly, the competition to control the instruments of wealth transfer has been intense in most African countries, and has been one of the underlying causes of ethnic conflict. Since governments do not create wealth, such transfers must be taken from other groups, which essentially implies that the transfer-seeking behaviour by permanent interest groups is at best a zero sum game. This analysis helps to explain the continued state of conflict and political instability in Africa, where there are so many tribal groups attempting to control governments, and why the interest-group behaviour of ethnic groups is more pronounced in centralized unitary states.
Kimenyi: Ethnicity and Conflict Avoidance 67
Efficient production increases the producer and consumer surplus, essentially a measure of an ethnic group’s wellbeing. Different ethnic groups within the same country are likely to be vastly different in terms of their ability to generate surplus, but each can increase its ‘ethnic’ or ‘tribal’ surplus by entering into mutually beneficial exchanges with other ethnic groups. Thus, under conditions whereby ethnic groups are able to organize production freely and to enter into market exchange with other groups if they are able to retain their tribal surplus, then each ethnic group would seek to be efficient in production in order to increase its surplus. However, efficiency in production is not the only way that an ethnic group can increase its surplus. One attractive alternative is to use the political process to capture the surplus generated by other groups through rent-seeking. Under a system of government where taxing and spending decisions are centralized, rulers are able to redistribute surplus from some ethnic groups to others. Given that ethnic groups occupy different regions, governments are able to direct the resources to specific groups. Such ethnic rent-seeking can be minimized under a system of decentralization. The discussion has so far emphasized that ethnic groups possess specific advantages in organizing for the provision of ‘ethnic’ goods. However, in unitary centralized states, ethnic groups can use their organizational capacity to undermine the wellbeing of others. As such, the first reforms that can help to reduce ethnic conflicts are those providing ethnic units with a fair degree of autonomy. Ideally, decentralized federal states would be best suited to deal with diversity. Of the various strategies aimed at resolving conflicts in societies characterized by diversity, federalism is the most prescribed system of governance. Decentralization has the potential to resolve the institutional problems affecting African countries, by increasing the efficiency and responsiveness of the government in two ways. First, compared to central government officials in distant offices, local leaders have superior information about local needs. Devolving resource allocation decisions to locally elected leaders therefore helps to improve the match between the preferences of local people and the government’s output. Second, local leaders are more accessible to local people and can be held more accountable. This results in improved management of resources through improvement in performance. Decentralization also encourages innovation in the delivery of services, as individuals have a greater incentive to participate in all aspects of community life and to seek solutions to individual and collective choices. When considering the institutional problems in Africa, however, one can argue that the primary advantage of decentralization is that it is probably the only system of organizing collective activities that protects groups from oppression by others as well as accommodating diversity. It provides mechanisms for bringing rival groups into a formal bargaining process (Burki, Perry and Dillinger, 1999), thus reducing conflicts. Decentralization
68 Thematic Issues
reduces the points of inter-ethnic contacts that result in conflict, and instead increases the contacts that are mutually beneficial. In such a decentralized system, each autonomous political jurisdiction has greater latitude in policy design and implementation, and can also determine the nature and scope of interaction with other groups. Under constitutional federalism, individuals can migrate freely and cheaply to competing political jurisdictions. This reduces significantly the ability of the state to coerce. Moreover, this ability of citizens to vote with their feet enhances competition among units and results in improved service delivery. While there is consensus about the advantages of decentralized states, the use of ethnic units of collective choice is controversial. We suggest that ethnic units have distinct efficiency characteristics that should be exploited in designing the units of collective choice. Given geoethnicity, effective decentralization naturally implies that the local governments will be composed primarily of ethnically homogenous groups. The use of ethnic groups as units of collective choice has been advanced by Kimenyi (1998). Kimenyi compares ethnic groups to voluntary associations or clubs whose members have fairly homogenous preferences. Using the model of clubs as developed by Buchanan (1965), the optimal size of the club is determined by the membership at which marginal cost is equal to marginal benefit (see Figure 4.2). In this model, costs increase and benefits decrease steadily as membership increases. Figures 4.3 and 4.4 extend the model to cases of ethnic units of collective choice. Based on previous discussion, the organization of collective activities
MB, MC MC
MB
N Number of club members Figure 4.2
Optimal size of clubs
Kimenyi: Ethnicity and Conflict Avoidance 69
MB, MC MC
MB
Members of same ethnic groups Figure 4.3
Members of other ethnic groups
Optimal ethnic club size: kinked marginal cost
by one ethnic group is expected to resemble the club models: costs increase and benefits decrease steadily as membership increases. This is because the ethnic group possesses various advantages that economize on organizational costs. However, increasing the membership beyond a certain level results in a situation where the benefits of organizing the activity are lower than the costs. This would imply that the benefits of ethnic provision are less than the costs (Landa, 1994). The benefits and costs change more dramatically when more than one ethnic group is involved. Organizing across different ethnic groups is more complicated, and adding members from other ethnic groups creates a discontinuity in the costs and benefits. These discontinuities result from difficulties in communicating across ethnic and linguistic boundaries as well as differences in preferences, taste and reduced trust across ethnic groups rather than within the same ethnic group. The implication of the foregoing analysis is that the optimal units of collective choice are ones that are ethnically homogeneous. Establishment of autonomous ethnic governments does not preclude collaboration across ethnic groups. However, this collaboration will only be acceptable if all
70 Thematic Issues
MB, MC MC
MB
Members of same ethnic groups Figure 4.4
Members of other ethnic groups
Optimal ethnic club size: kinked marginal benefit
groups benefit from the association. In essence, a system of ethnic governments would help to reduce ethnic rent-seeking that often translates into ethnic conflicts.
4
Overlapping functional and ethnic jurisdictions
The central theme of this chapter is that ethnic governments have desirable features that potentially can reduce conflict. The proposal, then, is for the establishment of decentralized federalism with autonomous ethnically homogenous local governments. However, this should not be seen as totally severing linkages between ethnic groups. Ethnic governments, while retaining a significant degree of autonomy, must necessarily associate with other ethnic governments. In particular, it is important to note that such governments will only be able to provide a limited number of ‘local’ public goods. Thus each unit has a desire to cooperate with other units in the provision of public goods that cannot be provided efficiently within the confines of the individual units. Such joint provision would promote cooperation across ethnic units based on mutual interest, discouraging conflicts and encouraging peaceful coexistence. In addition to the ethnic governments, various other complementary functional governments for the provision of various services are proposed (see Frey and Eichenberger, 1999). These governments would cover different regions, some transcending the boundaries of the ethnic governments. In
Kimenyi: Ethnicity and Conflict Avoidance 71
addition, in some cases, the functional governments would overlap because there are many goods that are best provided by more than one ethnic community. For example, the provision of water may require communities to join together to build a dam or to distribute water to their communities. Note that the communities cooperate out of self-interest. In essence, the establishment of ethnic governments can be an important cornerstone for defining the nature of cooperation among ethnic groups. Ethnic governments can be inefficient, particularly in cases where individuals do not have any mobility option and thus are not able to ‘vote with their feet’. This is quite often the case in Africa, where migration from one ethnic community to another is quite difficult. However, the establishment of overlapping functional governments such as for water government acts as a good substitute for ‘voting with one’s feet’ and thus enhances the efficient provision of services. The point to emphasize is that the establishment of ethnic governments does not preclude cooperation between ethnic groups. Instead, such institutions form the basis for groups to enter into mutually beneficial arrangements in the provision of public goods.
5
Conclusion
This chapter highlights the importance of establishing institutions that harmonize ethnic interests in Africa. It is argued that the recent democratization reforms involving the introduction of political competition are not enough to deal with the issue of conflict avoidance. In fact, it is argued that political party competition can intensify conflict-generating ethnic competition in the current highly centralized unitary states. In this chapter I have proposed the establishment of decentralized institutions of governance that involve ethnically homogeneous local units of collective choice possessing a fair degree of autonomy. In addition, the chapter suggests that ethnic governments might be complemented by functional governments that transcend the boundaries of ethnic governments within a given country. The chapter does not attempt to discuss institutional issues relating to conflicts between countries or wars of secession. These topics are discussed in Kimenyi 1997b, 1997c and 1999. References Borner, S. and M. Paldam (1998) The Political Dimension of Economic Growth, IEA Conference vol. 119 (London: Macmillan). Brough, W. T. and M. S. Kimenyi (1986) ‘On the Inefficient Extraction of Rents by Dictators’, Public Choice, vol. 48, pp. 37–48. Buchanan, J. M. (1965) ‘An Economic Theory of Clubs’, Economica, vol. 32, pp. S1–4. Burki, S., G. Perry and W. Dillinger (1999) Beyond the Centre: Decentralizing the State (Washington, DC: World Bank).
72 Thematic Issues Cobbah, J. A. M. (1988) ‘Toward a Geography of Peace in Africa: Redefining Sub State Self-Determination Rights’ in R. J. Johnson, D. B. Knight and E. Kofman (eds), Nationalism, Self-Determination, and Political Geography (London: Croom Helm). Collier, P. (1999) ‘Doing Well Out of War’, Paper prepared for the Conference on Economic Agendas in Civil Wars, London, 26–27 April. Collier, P. and A. Hoeffler (1998) ‘On Economic Causes of Civil War’, Oxford Economic Papers, vol. 50, pp. 563–73. Collier, P. and A. Hoeffler (2001) Greed and Grievance in Civil War (Washington, DC: World Bank). Collier, P., A. Hoeffler and M. Söderbom (1999) On the Duration of Civil War (Washington, DC: World Bank). Deng, F. M. (1997) ‘Ethnicity – An African Predicament’, Brookings Review, vol. 15, no. 3, pp. 28–31 (Washington, DC: The Brookings Institution). Elbadawi, I. A. and N. S. Ndungu (2001) ‘The Economics of Civil Wars and PostConflict Recovery’, Paper presented at the Development Economic Research Group Conference, University of California, May (mimeo). Frey, B. and R. Eichenberger (1999) ‘A New Proposal for Federalism and Democracy in Developing Countries’, in M. S. Kimenyi and J. M. Mbaku (eds), Institutions and Collective Choice in Developing Countries (Aldershot: Ashgate). Horowitz, D. (1985) Ethnic Groups in Conflict (Berkeley, Calif.: University of California Press). Kimenyi, M. S. (1989) ‘Interest Groups, Transfer Seeking and Democratization: Competition for the Benefits of Government Power May Explain African Political Instability’, American Journal of Economics and Sociology, vol. 48, pp. 339–49. Kimenyi, M. S. (1997a) Ethnic Diversity, Liberty and the State: The African Dilemma, The Shaftesbury Papers No. 12 (Cheltenham: Edward Elgar). Kimenyi, M. S. (1997b) ‘Redrawing Africa’s Borders: Logical Foundations for Consensual Units of Collective Choice’, Department of Economics, University of Connecticut. (mimeo). Kimenyi, M. S. (1997c) ‘Ethnic Rent-Seeking and Optimal Integration’, Department of Economics, University of Connecticut (mimeo). Kimenyi, M. S. (1998) ‘Harmonizing Ethnic Claims in Africa. A Proposal for Ethnic Based Federalism’, Cato Journal, vol. 18, no.1, pp. 43–63. Kimenyi, M. S. (1999) ‘Spatial Competition, Ethnicity, and the Optimal Size and Composition of Units of Collective Choice’, in M. S. Kimenyi and J. M. Mbaku (eds) Institutions and Collective Choice in Developing Countries (Aldershot: Ashgate). Kimenyi, M. S. (2001) ‘Institution of Governance and Ethnic Conflict in Africa: A Positive View of Ethnic Governments’, Development Economic Research Group Conference, University of California, May 2001 (mimeo). Kimenyi, M. S. and J. M. Mbaku (1995) ‘Democratization in Africa: The Continuing Struggle’, Coexistence, vol. 32, pp. 119–36. Kimenyi, M. S. and W. F. Shughart (1989) ‘Political Successions and the Growth of Government’, Public Choice, vol. 62, pp. 173–9. Landa, J. T. (1994) Trust, Ethnicity and Identity (Ann Arbor, Mich.: University of Michigan Press). Little, K. (1957) ‘The Role of Voluntary Associations in West African Urbanization’, in P. L. Van den Berghe (ed.), Africa: Social Problems of Change and Conflict (San Francisco: Chandler).
Kimenyi: Ethnicity and Conflict Avoidance 73 Murungi, K. (1995) Ethnicity and Multi-Partysim in Kenya (Nairobi: Kenya Human Rights Commission). Roback, J. (1991) ‘Plural but Equal: Group Identity and Voluntary Integration’, Social Philosophy and Policy, vol. 8, no. 2, pp. 60–80. Stewart, F. (2000) ‘Civil Wars in Sub-Saharan Africa: Counting the Economic and Social Cost’ in D. Ghai (ed.) Renewing Social and Economic Progress in Africa (London: Macmillan). World Bank (2000) Can Africa Claim the 21st Century? (Washington, DC: World Bank).
This page intentionally left blank
Part II Case Studies: West Africa
This page intentionally left blank
5 Liberia and Sierra Leone: Interwoven Civil Wars Victor A. B. Davies
1
Introduction
Liberia and Sierra Leone have common borders and interwoven civil war experiences. Rebels from the ongoing Liberian civil war ignited the 1991–2001 Sierra Leone war. Combatants subsequently moved between them with ease. This chapter addresses the following questions. What factors have caused and sustained war in the two countries? Why in 2002 was Liberia still at war? What lessons can be learned from Sierra Leone’s peace process? What are the prospects for peace for the two countries?
2
Context and background
2.1 Liberia The Liberian civil war began in December 1989 when Charles Taylor’s National Patriotic Front of Liberia (NPFL) invaded Liberia from Côte d’Ivoire (See Map 5.1). Within months, the rebels were on the outskirts of Monrovia, the national capital. In August 1990, the Economic Community of West African States (ECOWAS) despatched to Liberia a Nigerian-led intervention force known as the Economic Community of West Africa Monitoring Group (ECOMOG). Using bases in Sierra Leone, ECOMOG prevented the NPFL’s imminent capture of Monrovia, marking the beginning of a long military stand-off between ECOMOG, which held Monrovia and the surrounding regions, and the NPFL, which held the remaining 90 per cent of Liberia. In 1997, after nearly ten failed peace accords and cease-fire agreements, and several interim governments, elections were conducted under the 1996 Abuja Peace Accord. Charles Taylor won and was inaugurated in August 1997. However, once in office, Taylor flagrantly violated the Abuja Peace Accord by refusing to allow ECOMOG to restructure the national army to absorb combatants from the warring factions and promote ethnic balance. This caused ECOMOG’s premature withdrawal in 1998. Taylor became increasingly 77
78 Case Studies: West Africa
15° w
5° w
10° w
15° N SENEGAL
Dakar
GAMBIA MALI
Banjul
Bamako
GUINEA BISSAU Bissau GUINEA 10° N Conakry SIERRA LEONE
Freetown Atlantic Ocean
CÔTE D’IVOIRE LIBERIA Monrovia
5° N
0 0
15° w
Map 5.1
100
200 miles
Abidjan
100 200 300 km
10° w
5° w
Liberia and Sierra Leone in West Africa
repressive, also supporting rebels in Sierra Leone. By early 1998, small rebel groups were again active in Liberia. The Liberians United for Reconciliation and Democracy (LURD) appeared in July 2000, and, although participating in the peace agreement of August 2003, failed to disarm as required. The conflict has claimed over 150,000 lives and displaced more than 1 million people out of a pre-war 1998 population of 2.4 million. It has been factious, spawning ten ethnically-based warring factions in addition to the dominant NPFL. 2.2 Sierra Leone In March 1991, Sierra Leonean and Liberian rebels fighting for the NPFL in Liberia attacked south-eastern Sierra Leone, triggering the Sierra Leone civil
Davies: Liberia and Sierra Leone
79
war. Their leader, Foday Sankoh, announced the intention of the Revolutionary United Front (RUF) to overthrow the kleptocratic All People’s Congress (APC) government. ECOMOG, based in Sierra Leone and Liberia, supported the government in the war against the RUF. In 1992, young soldiers from the war front overthrew the APC, setting up the National Provisional Ruling Council (NPRC). The war escalated under the NPRC, coming to within twenty miles of Freetown, the national capital, in 1995. The NPRC organized elections in 1996 and handed over to the victorious Tejan Kabbah and his Sierra Leone People’s Party (SLPP). Kabbah signed the Abidjan Peace Accord with the RUF in 1996 but was overthrown by the army in collaboration with the RUF in May 1997. ECOMOG intervened to reinstate Kabbah in February 1998, driving the rebels to the countryside. Kabbah was almost overthrown again in January 1999 when the rebels captured parts of the capital before ECOMOG repulsed them again. The Lomé Peace Accord was signed in July 1999 which, though frequently jeopardized, provided a framework for ending the war, with the completion of the disarmament and demobilization of combatants in January 2002. In May 2002, general elections were held, and won by the incumbent President Kabbah and his SLPP. The RUF failed to win a single parliamentary seat but, together with other contesting parties, accepted the election results. The conflict claimed over 25,000 lives and displaced 2.1 million people, nearly half of the country’s population. The RUF had about 20,000 combatants, and pro-government forces of over 25,000, in addition to ECOMOG. Both sides used mercenaries. The rebels committed widespread atrocities, including mutilations, rape, arson and looting. This was partly an RUF military strategy, otherwise lacking an ideology to motivate and guide its campaign. The atrocities bonded combatants to the rebel organization and subjugated civilians. Recruits were drawn largely from among socially-deprived youths and children, often forcibly, and made to commit acts of terror against family and community members, ensuring exclusion from their community so they would subsequently not desert. The rebels were also drugged to induce fearlessness and to dehumanize them. Civilians and new recruits were used as human shields, positioning them in the firing line, with hard-core rebels at the rear. Both sides used child soldiers, totalling about 10,000.
3
Factors causing and sustaining civil war
3.1 Causes Modern Liberia emerged when freed slaves from the USA established settlements around the coast of Monrovia from the early 1820s. These AmericoLiberians declared independence in 1847. Constituting less than 5 per cent of the population, the settlers monopolized power, disenfranchising indigenous people. In 1931, the League of Nations established that Americo-Liberians
80 Case Studies: West Africa
were using indigenous Africans as forced labour, an act of slavery. Enduring ethnic aristocracy foreclosed all peaceful means for the majority indigenous population to access power, leading to Samuel Doe’s coup of 1980. During the 1980s, Doe’s policies resulted in the most immediate causes of the civil war. Doe used ethnicity as a survival strategy, practising nepotism for his minority Krahn ethnic group, which constituted only 2 per cent of the population. He executed powerful non-Krahn fellow coup-plotters and members of the ruling People’s Redemption Council, notably Thomas Weh Sien, his second in command, and Thomas Quiwonkpa. Consequently, Doe alienated his executed colleagues’ ethnic groups, particularly Quiwonkpa’s Gios. In 1985, Doe’s security forces massacred hundreds of Gios in Nimba County. That same year, Doe stage-managed presidential elections to become the elected president. Corruption and mismanagement led to declining economic performance. Per capita income growth fell from 0.5 per cent during 1965–80 to minus 5.2 per cent during 1980–7. Thus, when Charles Taylor launched his rebellion in Nimba County in 1989, he gained widespread support among Liberians, particularly Gios and Manos from Nimba County, enabling the rebellion to spread rapidly. Diasporas, mostly USA-based, featured prominently in the conflict, some allegedly financing the NPFL, and expecting power in return. Like Liberia, Sierra Leone’s modern period began when its present capital, Freetown, was chosen as a home for freed slaves from Europe and America. However, unlike the situation in Liberia, the settlers never ruled Sierra Leone. Although ethnic differences emerged between the settlers and indigenous ethnic groups, after independence in 1961 these rivalries were superseded by ethnic rivalries among the indigenous population. The domestic causes of war in Sierra Leone can be traced to the APC government of 1968–92. The APC practised kleptocracy and political repression as a survival strategy through a patrimonial system of asset-stripping, including the allocation of access rights to natural resources; rents from pricing interventions; income redistribution measures such as exchange rate overvaluation; and taxation of rural agriculture to subsidize urban consumption. These measures were inimical to economic growth and development. The APC suppressed or bought out any opposition. Consequently, by the late 1980s, under a one-party dictatorship, Sierra Leone was heading for economic and political implosion. Living standards were among the lowest in the world, as reflected in a rank of 174th out of 174 on the UNDP Human Development Index for 1991. Basic amenities such as electricity and running water had virtually collapsed even in the national capital. Per capita income growth was negative. Youths, hardest hit by unemployment, bore the brunt of the economic and political atrophy of the 1980s and became the instruments of choice for political violence under the APC. Drugged and used as thugs during elections, and abandoned subsequently, they became increasingly rebellious, retreating
Davies: Liberia and Sierra Leone
81
into crime and drug-taking. By the late 1980s, Sierra Leone’s unemployed youths constituted a large pool of potential rebels. University students spearheading the anti-APC campaign recruited volunteers from this pool for revolutionary training in Libya. External support helps to explain why war broke out in Sierra Leone and Liberia rather than other African countries with similar risk factors. Taylor and Foday Sankoh received training and military support from Libya as part of Colonel Gaddafi’s grand plan to install satellite regimes in West Africa. Côte d’Ivoire and Burkina Faso also supported the rebellions. Côte d’Ivoire hosted Foday Sankoh and offered Charles Taylor a base; Taylor, in turn, offered Sankoh a base and troops to launch civil war in Sierra Leone. Presidents Houphouët-Boigny of Côte d’Ivoire and Blaise Campoaré of Burkina Faso apparently supported Charles Taylor for personal reasons: in the coup of 1980, Doe had executed Liberian President Tolbert’s son who, like Blaise Compoaré, was Houphouët-Boigny’s son-in-law (Reno, 1998, p. 81). While there is no obvious reason why Côte d’Ivoire and Burkina Faso supported the RUF, such support may have been obtained through Charles Taylor. 3.2 Factors sustaining civil war The conjunction of abundant natural resources and the porosity of national borders has been a major factor sustaining rebellion in the two countries. Pre-war legal and illicit exports of natural resources were equivalent to more than 30 per cent of GDP in Liberia, and some 25–40 per cent in Sierra Leone. Timber and rubber resources helped to finance Liberia’s civil war; diamonds financed the rebellion in Sierra Leone and produced a warprolonging congruence of interests (Davies, 2000). Exports of conflict diamonds for 1999 were estimated at US$68.5 million, while government defence expenditure during the war averaged less than US$25 million a year. Pro-government combatants also mined diamonds, sometimes in collaboration with the rebels. The illicit trade of conflict diamonds to Europe via Liberia partly explains why Taylor, at the centre of a lucrative arms-for-diamonds trade, chose the path of warlordism, even after winning the 1997 presidential elections. Table 5.1 shows that Liberia ‘exported’ diamonds far in excess of domestic output. Maximum annual production capacity is approximately 150,000 carats, while recorded imports of ‘Liberian’ diamonds in Belgium exceeded Table 5.1
Illicit diamond traffic 1990–8 (in 000s carats)
Year
1990
1991
1992
1993
1994
1995
1996
1997
1998
Production Exports
100 5,523
100 658
150 1,909
150 5,006
100 3,268
150 10,677
150 12,320
150 5,803
150 2,558
Source:
Smillie, Gberie and Hazleton (2000).
82 Case Studies: West Africa
12 million carats in 1996. A report by a UN panel of experts states that the value of Belgian imports of ‘Liberian’ diamonds was US$217 million for 1998, while official exports recorded in Liberia were valued at only US$800,000 (United Nations, 2000, p. 10). Ethnic rivalries sustain civil war in both countries, more strongly in Liberia because of the social and political divide between Americo-Liberians and the indigenous population, and Samuel Doe’s ethnically-divisive rule in the 1980s. Gios and Manos from Nimba County whose ethnic groups were massacred by Doe’s army harbour deep hatred for the Krahns, Doe’s ethnic group. Such divides tend to break down national unity and make it easy to recruit within ethnic groups by appealing to ethnic emotions. There has been no enduring social divide among the population in Sierra Leone. Intermarriage, migration and other forms of socialization have gradually blurred the Creole-Protectorate divide. Although the APC did not engage in systematic ethnic massacres, ethnic undercurrents did affect the war. The APC, the first government that fought the rebellion, was often accused of using the war to disenfranchise the south-eastern regions where the war started. People from these regions initially joined the rebel movement voluntarily out of disillusionment with the APC. Northerners distrusted the war effort of the south-eastern based SLPP, especially with regard to the prominent role of the ethnic Mende Kamajor civil defence militias. Both conflicts created a substantial supply of small arms and a generation of young men knowledgeable in the use of them. The risks of cross-border contagion were high; and combatants redeployed across countries with relative ease. The RUF field commander, General Mosquito, allegedly commanded some of Taylor’s troops. Worse still, in 2000, an insurgency in Guinea, which shares borders with both Liberia and Sierra Leone, was aided by RUF rebels and, allegedly, the Liberian government. Also, it is relatively easy to acquire small arms: West Africa is awash with weapons despite an ECOWAS moratorium on shipments to the region (United Nations, 2000, p. 1). In Liberia, deep mistrust among the protagonists makes it difficult for them to cooperate, which is one reason for the failure of past peace accords and interim governments. Peace is constrained by a severe time-consistency problem, but a major factor for disruption was removed by Taylor’s flight into exile in August 2003. The two countries’ geography favours warlordism. Much of the hinterland is heavily forested, allowing rebels to operate with ease. Unlike Sierra Leone, with a relatively high population density of 65 people per square kilometre, Liberia has a low population density of 26 people per square kilometre, which makes it more difficult to defend. The rural population density is much lower in both countries. Continued external support of the rebels, particularly from Libya, and Taylor’s support for the RUF, helped to sustain the two civil wars. On the other hand, apathy by the international community has been detrimental.
Davies: Liberia and Sierra Leone
83
The belated international efforts in Sierra Leone demonstrate the potency of such support. Apart from ECOWAS, it was not until February 2004 that a UN donors’ conference put together a US$ 500 million aid package for Liberia.
4
The consequences of civil war
4.1 Liberia Liberia’s pre-war per capita income of US$485 in 1987 was higher than the SSA average of US$440 and US$237 in Sierra Leone (see Table 5.2). The 1990 Human Development Index rank of 26/130 on an ascending scale of human achievement put Liberia above twenty other African states, with 70 per cent of the labour force engaged in agriculture and forestry. Leading exports included iron ore, rubber and timber. Diamonds and gold were exported in small quantities. The effects of war have been devastating: total official annual exports averaging over US$400 million before the war plummeted to less than US$60 million in the period 1997–2000. Much of Liberia’s infrastructure was destroyed; and at the time of writing there is still no running water or electricity supply. The war has claimed over 150,000 lives (equivalent to more than 6 per cent of the population) and experienced massive emigration, especially of skilled human capital. Table 5.2 shows that per capita income fell from US$485 in 1987 to US$177 in 2000. In the 1990s employment decreased by 50 per cent in the formal sector, which accounted for about 30 per cent of total employment in the mid-1980s. Alluvial diamond mining, which employed some 60,000 Liberians prior to the war, employed only 6,000 in 2001, following UN sanctions on Liberian diamonds (United Nations, 2001). Support for Sierra Leone’s rebel RUF led to Liberia’s international isolation. After May 2001 UN sanctions banned President Taylor and his government from international travel. The export of rough diamonds was
Table 5.2
Socio-economic indicators in Liberia and Sierra Leone Period
Per capita income (US$)
Per capita income growth (%)
Incidence of poverty (%) (
HDI rank1
Liberia
Pre-war War
4852 1773
−5.206 na
– 803
26/1304 na
Sierra Leone
Pre-war War
2374 1423
−1.006 −8.007
82 90
4/1304 174/1745
Notes: 1 1990: increasing scale of human development; all other years, decreasing scale; 2 1987; 3 2000; 4 1990 (war had already started in Liberia); 5 2001; 6 1980–90; 7 1991–2000. Sources: Liberia: UNDP Human Development Report 1990 and United Nations 2001; Sierra Leone: Government of Sierra Leone (2001).
84 Case Studies: West Africa
banned and an arms embargo enforced. Non-humanitarian assistance became minimal and aid flows to Liberia fell from US$138 million in 1998 to US$30 million in 2001. Economic management, particularly the capacity to collect data, was severely affected: balance of payments statistics have not been published since 1989. UN sanctions worsened matters. However, following Taylor’s abdication international aid began to flow by 2004. Liberia has the world’s second-largest shipping register in tonnage terms. Lax registration and control procedures offer a ‘flag of convenience’. Registration revenue, totalling US$18 million (or 20 per cent of total revenue) in 2000, has helped to sustain the economy. Remittances totalling US$23 million in 2000, along with humanitarian relief, have also been key sources of support. Strong family and kinship bonds have also served as a safety net, with income earners having to support as many as twenty dependants. 4.2 Sierra Leone In 1990, Sierra Leone was a low-income economy with a per capita income of US$237 and GDP of US$800 million. Agriculture was the dominant economic activity, accounting for over 70 per cent of employment, 35 per cent of GDP and about 15 per cent of export earnings. The country is richly endowed with mineral resources. In the early 1970s, diamonds accounted for about 60 per cent of export revenue and 15 per cent of GDP. However, increased smuggling contributed to a decline in official diamond exports from an annual average of US$78 million in 1972–5 to US$15 million by 1990. Between 1997 and 2000 official exports all but collapsed. During the 1990s, rutile displaced diamonds as the leading export, accounting for 40 per cent of export earnings and about 20 per cent of fiscal revenues. Gold and bauxite were also mined, although the mining of rutile and bauxite ceased after 1995. Table 5.2 shows that the percentage of the population living below the poverty line of US$1 per day increased from 82 per cent in 1989 to 90 per cent in 2000, and that per capita income fell from an already low US$237 in 1990 to US$142 in 2000. Government control and the provision of social services were often limited to Freetown. A structural adjustment programme launched in 1989 liberalized domestic markets and international trade, reformed public-sector management and stabilized the economy somewhat, so that inflation declined from a peak of 180 per cent in 1987 to less than 30 per cent for most years after 1992. However, during the adjustment period (which after 1991 coincided with the war) per capita income declined further. Gross national savings were negative; and the ratio of investment to GDP (at international prices) was less than 2 per cent, see Davies (2000). The social costs of the war have also been high: 25,000 people died and about 2.1 million (about 40 per cent of the population) were displaced. The conflict occurred primarily in rural areas, causing rural–urban displacement. The population of Freetown increased from 500,000 pre-war to 2 million
Davies: Liberia and Sierra Leone
85
during the war. Overcrowding in urban areas increased the incidence of communicable diseases. Many displaced persons are now reluctant to return to rural areas which lack modern amenities despite dire urban living conditions. Consequently, urban unemployment is high. Many continue to be displaced to neighbouring Guinea and Liberia, where they have been further affected by insecurity. Human rights abuses, murder, mutilation, rape and arson, sometimes carried out by family or community members, have generated grievances against the rebels and their perceived collaborators. Many former rebels are terrified of returning to their communities. Youths and children have been particularly hard hit. Out of a total of 48,000 combatants, some 10,000 were child soldiers; they were exposed to drugs and have been victims or perpetrators of atrocities. Commercial sex was a coping strategy during the crisis. This, together with population movements, low contraceptive use, physical and sexual abuse, and ignorance about HIV/AIDS, has exacerbated the growth of HIV/AIDS.
5
Peace in Sierra Leone
The initial framework for peace was the July 1999 Lomé Peace Accord, which promoted power sharing and granted RUF leader, Foday Sankoh, the status of vice-president and de facto powers of mines minister. The Accord was threatened in May 2000 when the rebels kidnapped 500 UN peacekeepers and attempted to seize power, leading to the arrest of Foday Sankoh and other RUF ministers, and Sankoh’s subsequent trial for murder. However, renewed peace efforts culminated in the Abuja Ceasefire Agreement in November 2000, and the completion of the disarmament and demobilization process by January 2002. A weakening of RUF military capability aided the peace process. British troops intervened in May 2000. They thwarted an RUF attempt to seize power, wiped out the West Side Boys rebel faction and strengthened government military capacity by revamping the Sierra Leone army. In September 2000, the Guinean army inflicted heavy casualties on RUF rebels aiding an insurgency in Guinea. Also, renewed civil war in Liberia in the year 2000 led to redeployment of pro-RUF combatants from Sierra Leone to Liberia. Other factors also aided the peace process. A UN embargo on arms to and diamonds from Liberia since May 2001 forced President Taylor to reduce his arms-for-diamonds support for the RUF. Also, following the kidnapping of some of its troops in March 2001, the UN peacekeeping force UNAMSIL was expanded to 17,500, becoming the world’s largest UN peacekeeping operation. UNAMSIL robustly countered RUF ceasefire violations. In 2001 a more cooperative General Issa replaced the incarcerated and recalcitrant RUF leader, Foday Sankoh. By then, many of the rebels were reportedly tired of fighting and living in dire conditions in the bush, and some had reportedly amassed wealth and were looking forward to returning to civil society to enjoy it.
86 Case Studies: West Africa
While the unexpected British military intervention has been critical for peace, it has prompted considerable speculation. Some have expressed far-fetched fears of, and others a desire for, British recolonization. The motivation for the military intervention and high-level support can be attributed to a combination of debatable factors including historical links and the desire to prevent a former colony from becoming another Somalia. Whatever the reason, the military intervention was feasible and successful because the risks were very low. The rebels were disorganized and poorly trained. Geographically, the country is manageable, with a total area of only 72,000 square miles, in sharp contrast to other conflict countries such as the Democratic Republic of Congo. Domestic agencies also played a major role in the peace process. Widespread disillusionment with the military for collaborating with the rebels led to the formation of civil defence militias, notably the Kamajors in the south-eastern regions, the Gbethis in the north, and the Donsos in eastern Kono District. The Kamajors, numbering about 20,000, became a powerful military force, keeping much of the south-eastern region rebel-free. Civil opposition to the military coup of May 1997, which toppled the elected President Kabbah, was instrumental in the eventual restoration of democracy. The populace refused to recognize the Armed Forces Revolutionary Council. A campaign of non-cooperation, coupled with international sanctions, ostracized the junta and led to their being flushed out by ECOMOG in February 1998. Civil society groups, particularly the InterReligious Council, grouping Christians and Muslims, participated actively in the peace process.
6
Post-conflict initiatives in Sierra Leone
6.1 Peace-building The disarmament and demobilization of 48,000 ex-combatants, including 10,000 children, was completed in January 2002. Some, from both sides, have been absorbed into the army which, together with the police, is being restructured. A National Reconstruction, Resettlement and Rehabilitation Programme is ongoing. A war crimes court and a truth and reconciliation commission have been established. The peaceful May 2002 presidential and general elections were also a critical part of the peace-building process. 6.2 Poverty reduction The government has made poverty reduction its principal objective and adopted a two-phase strategy. The first, transitional, 2001–2 phase focuses on consolidating peace and security; the resettlement and rehabilitation of displaced persons and refugees; and the rehabilitation of a basic economic and social infrastructure. Other measures include:
Davies: Liberia and Sierra Leone
• • • •
87
improving governance through decentralization and judicial reform; stabilizing the economy to facilitate private-sector-led economic recovery; privatizing inefficient state enterprises; and debt-reduction efforts under the highly indebted poor country (HIPC) initiative.
Some transitional-phase interventions continue into the second, mediumterm, 2003–5 phase, addressing longer-term development challenges formally articulated in a Poverty Reduction Strategy agreed between the government and the IMF in April 2003. 6.3 Natural resource management Sierra Leone’s diamond deposits pose enduring risks: they can be mined illegally using simple equipment, and policing is difficult. A pre-war liberal exploitation policy induced severe economic and social dislocation. Migration from the agricultural to the diamond-mining sector constrained agriculture while fostering unemployment, crime and violence in diamond-mining areas. Smuggling was widespread. The diamonds fuelled the war by providing opportunities for illicit miners and criminals who joined the rebellion to gain access to the diamonds. Exploitation policy remains liberal, driven by entrenched private interests. To percolate rents, 0.75 percentage points of a 3 per cent tax on diamond exports now goes to the diamond-producing community. Fees and licences also generate government revenue. However, a Lebanese cartel, possessing capital and external marketing contacts, dominates the industry and enjoys much of the rents. Sierra Leoneans are employed mainly as diggers. The presence in Kono District of tens of thousands of diamond diggers, mainly unemployed former combatants, poses a threat to peace and security in Sierra Leone. Furthermore, the people of Kono District are deeply aggrieved, feeling that they were heavily exploited before the war, and suffered disproportionately during the war as the RUF tenaciously held and virtually completely destroyed the district. It is therefore important to target diamond diggers and the Kono District in post-conflict initiatives. Key international initiatives have been undertaken to curb the trade in conflict diamonds. A UN diamond certification scheme for Sierra Leone was established in July 2000. This led to increased official diamond exports, from US$1.5 million in 1999 to US$10 million by October–December 2000. However, the end of the war made the scheme redundant. 6.4 An assessment Despite weak institutional capacity, the government has made considerable efforts in consolidating peace and security, partly because it started planning long before the end of the conflict. Donors have played a key role by providing technical and financial support. Institutional reforms have improved
88 Case Studies: West Africa
economic management capacity, and technocrats employed in key ministries such as finance have been attracted by special remuneration packages. Markets and trade have been liberalized and the economy stabilized as part of the adjustment programme, while debt-relief efforts are ongoing. Key services such as electricity and telecommunications are to be privatized. However, tremendous challenges lie ahead. Corruption, partly a legacy of past kleptocracy, continues. There is a risk that ethno-regional rivalries might escalate when President Kabbah, generally regarded as a symbol of ethnic harmony, completes his current (at the time of writing) and final term of office. There is a need to develop rural Sierra Leone, home to 90 per cent of the pre-war population but lacking electricity, running water, telecommunications and other modern facilities. The government’s decentralization programme requires speedy implementation. Another challenge is to reduce donor dependence, accounting for highly donor-driven policies. The country is saddled with a plethora of policies, programmes and processes, some designed merely to attract foreign aid. Aid coordination remains problematic despite the establishment of the Integrated Approach to Aid Coordination (IAAC) in 2000. The war spawned over seventy international, and 400 local, nongovernmental organizations (NGOs). Some receive funds directly from donors without informing the government.
7
Why was Liberia so long at war?
The Abuja Peace Accord of 1996, leading to Charles Taylor’s presidential election and inauguration in 1997, did not bring peace to Liberia. Undeniably, there were formidable obstacles at the outset, notably the adverse economic consequences of the war and Taylor’s questionable legitimacy resulting from suspicions that ECOMOG had rigged his election. Furthermore, the pre-election disarmament process, which ECOMOG had declared successfully completed, was in reality far from successful. Factional combatants retained their weapons and remained in the bush. After the elections, many people fled Liberia, fearing persecution from Taylor. However, for some, the elections were a referendum for peace (Toure, 1999). They voted for Taylor fearing that he would continue the war if he lost, and expecting Taylor’s victory to lead to peace. Some saw this as the reason for ECOMOG’s alleged involvement in favour of Taylor. Peace depended very much on Taylor’s desire and strategy for reconciliation. Taylor made some reconciliatory gestures, appointing former opposition or faction leaders such as Tom Woewiyu, Jenkins Scott and Roosevelt Johnson as ministers. These gestures did not lead to peace, however. In 1998, Roosevelt Johnson and some of his supporters had to be evacuated from Liberia after clashes with Taylor’s security forces. In 1999, Taylor charged Ellen Johnson Sirleaf, his main opponent in the 1997 general elections, and Alhaji Kroma, a faction leader, with treason.
Davies: Liberia and Sierra Leone
89
In the main, Taylor chose the path of ‘warlordism’ – personalized rule through a network of security forces designed to terrorize the opposition, and to facilitate extortion and shady business operations, rather than to promote national reconciliation, democracy and respect for the rule of law. He ran the economy as a personal estate, allocating business contracts and access to state resources to family, friends and cronies, receiving unrecorded extra-budgetary income from the exploitation of these facilities. Taylor became linked with the trade in illicit diamonds and encourages the use of the Liberian flag as a flag of convenience by operators involved in illicit maritime and aviation activities. Thus many aircraft flying the Liberian flag are apparently unknown to the Liberian authorities (United Nations, 2000). Within months of Taylor’s inauguration, key critics of him were assassinated. Taylor vitiated the Abuja Peace Accord by refusing to allow ECOMOG to recruit and train a new army, as stipulated in the accord. This led to ECOMOG’s premature withdrawal from Liberia in 1998, enabling Taylor to consolidate power by establishing security units loyal to him personally, such as the Anti-Terrorist Unit (ATU), the Special Security Service (SSS) and the Special Operations Division (SOD). After winning the 1997 elections, Taylor had two paths to choose from: either transform himself into a statesman, as Museveni did in Uganda, or continue as a warlord. Why did he choose the latter? Greed – rents from criminal and quasi-criminal activities – is part of the answer. However, after winning the 1997 elections, Taylor forwent statesmanship and himself succumbed to force. LURD and other armed groups compelled his resignation and exile in August 2003 and joined an interim administration protected by peacekeepers from the UN Mission in Liberia (UNMIL) until elections planned for 2006.
8
Conclusions
The interwoven civil wars in Liberia and Sierra Leone may be blamed on similar risk factors: high dependence on natural resources; poverty and economic collapse; and histories of political repression. The wars demonstrate the risks of regional contagion once war starts in one country, and the importance of external actors in triggering and sustaining civil war and in leading to the eventual attainment of peace. Much of the blame for the continuation of the civil conflict in Liberia after Charles Taylor’s 1997 inauguration lay with Taylor himself. The prolonged transition to peace also highlights the dangers of incomplete disarmament and demobilization prior to elections following a peace settlement. Arguably, in Sierra Leone, British military and diplomatic intervention was a critical factor as was the US ‘quick reaction force’ that was flown to Monrovia in August 2003 in support of ECOWAS peacekeepers who had arrived ten days previously. Any resumption of war in either Liberia or Sierra
90 Case Studies: West Africa
Leone is a threat to peace and security in the other country, and indeed to other countries in the West Africa sub-region. References Davies, V. A. B. (2000) ‘Sierra Leone: Ironic Tragedy’, Journal of African Economies, vol. 9, no. 3, pp. 349–69. Government of Sierra Leone (2001) Interim Poverty Reduction Strategy Paper (Freetown: Government of Sierra Leone). Reno, W. (1998) Warlord Politics and African States (Boulder, Col.: Lynne Rienner). Smillie, I., L. Gberie and R. Hazleton (2000) The Heart of the Matter: Sierra Leone Diamonds and Human Security (Ottawa: Partnership Africa Canada). Toure, A. T. (1999) ‘Instability in Liberia and Sierra Leone and the Growing Menace of Political Violence’, Liberian Journal of Democracy, vol. 1, no. 1. UNDP (1990) UNDP Human Development Report 1990 (Oxford: Oxford University Press). United Nations (2000) Report of the Panel of Experts Appointed Pursuant to UN Security Council Resolution 1306 (2000), Paragraph 19, in Relation to Sierra Leone, S/2000/1195 (New York: United Nations). United Nations (2001) First Report of the Secretary-General in Pursuance of Paragraph 13(a) of Resolution 1343 (2001) Concerning Liberia, S/2001/424 (New York: United Nations).
6 The Nigerian Civil War: Causes and the Aftermath Abdul-Ganiyu Garba and P. Kassey Garba*
1
Introduction
This chapter addresses three sets of issues: the primary causes of the Nigerian civil war; its duration and the aftermath; and the risk of recurrence. In Section 2 we address the formation of the Nigerian state, focusing on the origins of dominant antithetical forces, struggles and their consequences over four distinct periods leading up to the civil war. Section 3 addresses its duration: why did the war last thirty months, and why did it end when it did? Although Nigeria has not relapsed into civil war, since 1983 peace has become increasingly threatened. In Section 4 we therefore analyse the post-civil war threats and risks. This discussion emphasizes post-conflict governance, nation-building, economic management and development, allowing recommendations to be made to prevent a recurrence of conflict.
2
Pre-civil war: seeds of crisis
There have been several explanations of civil war.1 One type focuses on motive, such as greed and grievance, as causative forces (see Collier and Hoeffler, 1998; Elbadawi and Sambanis, 2000a, 2000b and 2001; Collier, Hoeffler and Söderbom, 2001) and is essentially an equilibrium or neoclassical analysis. A second type emphasizes ethnicity and population pressures as causes (Hildyard, 1999), tending to ignore social, economic and political struggles that develop into open ethnic or shortage-induced conflicts (Hildyard, 1999). Both provide only a limited understanding of open conflicts.2 A third type accepts that greed, grievances, tribal hatred and population pressures are powerful explanatory forces but attempts to locate the roots of conflict in the ‘complex web of politics, economics, history, psychology and (struggles) for identity’ (Hildyard, 1999; see also Adedeji, 1999 and Suliman, 1999). Garba and Garba (2002) represent a variant of this third type, putting forward five propositions. The first is that conflict is inherent in humans, in relations between humans and states, and between humans and the physical 91
92 Case Studies: West Africa
environment.3 Second, inherent and open conflicts in relations between humans and states result from the process of forming, consolidating and dissolving nations. Third, inherent conflict exists prior to open conflict; however, open conflict does not necessarily follow all inherent conflict. Fourth, when a state, its institutions and market system approach the threshold of failure, incidences of inherent conflict that become open tend to increase: increasing incidences of open conflict are indicative of a failing national social order.4 Fifth, in considering national conflicts (especially in periods of global shifts, turbulence and insecurity), there are links between national and global levels. In applying this analysis to Nigeria we first review the origins and evolution of the Nigerian nation state between 1900 and 1967, which we have divided into four periods: formation (1900–14); foundation (1914–46); decolonization (1947–60); and transition to civil war (1961–7). 2.1 The formation of the Nigerian nation state, 1900–14 The formation of a colonial Nigerian state occurred in three phases: the formation of three separate colonial governments; the amalgamation of the two governments in the south in 1906; and the amalgamation of the southern and northern governments in 1914. Lawal (1998) and Ikime (1980) show that formation was inherently conflictual. A struggle between imperial forces and local resistance dominated the first phase and in many parts of Nigeria this was open. In essence, the three colonial states were products of open conflict and, specifically, the defeat of local resistance by imperial force. Once colonial rule was imposed, the historic conflict between social cohesion and individual liberty was resolved using the threat of force. The colonial administrators had no direct ‘social contracts’ with the civil society, and taxation was without representation. As a result, those paying tax had no influence on how it was spent. The absence of a social contract had implications for the inherent conflict between the three colonial governments, the resolution of their conflicts and the amalgamations of 1906 and 1914, and hence the formation of the colonial Nigerian state and the evolution of the Nigerian nation-state. Lawal (1998) identified two sources of conflict between the three governments. First, conflict in colonial orientation; and second, different fiscal systems and revenue generation: the fiscal problems of the northern protectorate were the main source of conflict. Conflict between colonial authorities did not become open for two interrelated reasons. First, because all three governments were agents of, and subject to, the authority of the British Colonial Office – which, for example, resolved the fiscal problems of the northern protectorate by enforcing an arrangement whereby, against the wishes of southern administrative officers, the two southern protectorates made mandatory annual transfer payments to the north to supplement annual voluntary grants-in-aid from the British Treasury (Lawal, 1998). These payments from the south to the north rose
Garba and Garba: Nigerian Civil War 93
steadily from £10,000 in 1899 to £70,000 in 1907. Had the use of revenue been accountable to those taxed then this south–north flow of funds would have been unlikely. Nor would surplus revenue of the south have been transferred to the Treasury or used to grant loans to the north.5 The second reason is that Britain used amalgamation as a strategy to resolve conflicts between the various colonial authorities and in the short term to reduce these annual transfers to the north, which ended in 1918. The interests of Nigerians were not considered during the formation of the Nigerian state, which began as a colonial state in conflict with its people. The threat of force by the colonial administration was the means of resolving this conflict. 2.2 The foundation phase, 1914–46 On amalgamation, fiscal problems ceased to be an issue, and conflicts between the northern and southern authorities ended. However, the creation of a unified colonial government in Nigeria in 1914 did not bring to an end all problems: some conflicts persisted, others took new forms and new ones emerged. The conflict between the colonial government and civil society remained. The decision-making process remained closed to, and disconnected from, civil society, as exemplified by the imposition of the Native Authority System (NAS), the Law of Nigeria (1914), the Clifford Constitution (1922), the split of the southern protectorate into east and west districts (1939), and the Richards Constitution (1946). In each case the colonial state prevailed over local preferences and aspirations and, in the case of the Native Authority System, provoked riots. Each of these institutional changes generated new points of conflict. The imposition of the NAS throughout Nigeria transformed independent native states into provinces, divisions and districts, and created local agents where centralized authority did not exist. The centralized Native Authority, with its executive, legislative and judicial powers, generated tensions between the chosen local autocrats on the one hand, and the emergent unions, professional groups, educated elite and political parties on the other.6 These tensions were particularly strong in the south, which was modernizing faster. Attempts at direct taxation, an integral part of the Native Authority Ordinance of 1916, generated armed revolts, such as the anti-tax riots in Warri in 1927–8 and the Women’s Riot in Aba in 1929 (Asiwaju, 1980). Riots and agitation by the educated elite in the south induced reforms that prepared the way for democratically elected councils in 1952. The colonial administration played off one group against another by granting partial concessions and favours. In Nigeria, the aggregation of hitherto independent people to meet the demands of colonial administration, with no regard to their pre-colonial context, created ethnic asymmetry. The majority–minority divide, the major manifestation of ethnic asymmetry, came to be defined by struggles for identity and autonomy, rendering ideas of
94 Case Studies: West Africa
nationhood, citizenship and patriotism problematic. Asiwaju (1980) and Abubakar (1980) showed that ethnic asymmetries prevailed in the more heterogeneous and non-Muslim central provinces (for example, Ogoja, Ogoni, Andoni, Ijaw) and the middle belt (Tiv, Bura, Margi, Gwari, Borgu). Ethnic asymmetries led to incidences of open conflict which the state pacified by force. Even at the time of writing, the erstwhile central provinces and middle belt remain highly volatile, tension is still caused by issues of identity and autonomy, and the current government still uses force to quell the violence. Tension was strongest in the eastern provinces, where local agents were appointed without regard to traditional political and social structures, in a culture that was historically republican. Thus, when Lord Lugard, head of the colonial administration, experimented with native authorities based on emirates, administered through paramount chiefs and permanent presidents, this approach failed in the eastern provinces (Afigbo, 1980) and generated tension in many other parts of the new nation. In addition, according to Afigbo (1980), the failure of the NAS, particularly in the east, split the colonial service into pro- and anti-Lugard officers, with the former being in the north and the latter in the south. Therefore, while tensions between the northern and southern colonial authorities ended with amalgamation, the colonial state did not attain internal harmony. The heterogeneity of the amalgamated units made uniform administration difficult. Lord Lugard, for example, retained a truncated Legislative Council for Lagos and created a Nigerian Council for the rest of the country, with four important chiefs of southern and northern Nigeria as members (Olusanya, 1980). When, in 1922, the Clifford Constitution was enacted, it institutionalized administrative segregation and segregated political and development paths along the pro- and anti-Lugardian split. First, a Legislative Council with jurisdiction over the entire south replaced the Nigerian Council, whose jurisdiction was limited to Lagos. Second, the Clifford Constitution provided for rule by proclamation for the North.7 The north–south split within the colonial administration and its institutionalization by the Clifford constitution, propelled northern and southern Nigeria towards divergent political, economic and social development paths.8 According to Abubakar (1980), the British isolated the northern provinces from the south and the Islamic world to prevent the possible penetration by subversive influences. In fact, the attempt to segregate Nigerians was extended to communal levels where, for example, the concept of Sabon Gari (sections of communities in northern Nigeria whose residents were not indigenous to northern Nigeria) was used by colonial administrators in north-western Nigeria to isolate the local community from southerners and middle belters, who tended to be more exposed to Western education and culture. Segregation was so entrenched that when the central administration in the 1930s sought to narrow the gaps between the north and the south through integration, British officials in the north resisted successfully.9
Garba and Garba: Nigerian Civil War 95
The division of the southern region into east and west regions created a three-region administrative structure that created further antithetical forces, struggles and conflicts. The Richards Constitution of 1946 ossified this by creating ethno-regional institutional structures that divided the elite of Nigeria. The new ethnically based regional structures created a northern region greater than the east and west combined. Such a polity was inherently unstable because it lacked counterbalancing forces. The premise that a regional consciousness was a prerequisite for national consciousness was advanced in support of ethno-regionalization. However, while the deliberate promotion of ethno-regional consciousness was successful, this countered the development of a national consciousness, the development of national institutions, common citizenship, a cohesive approach to decolonization, and, ultimately, the building of a modern state. The colonial authorities were also unwilling or unable to provide public goods efficiently or equitably. Present-day inequities in the spatial allocation of public goods can be traced back to colonial policy and the differential pace of economic development in Nigeria could be considered to be a consequence of colonial policy. The colonial state failed to perform its function as the provider of public goods, offsetting externalities and market failure by taxes, subsidies and social equity. In fact, the colonial state tended to subvert the development of a competitive market system by creating an environment conducive to the development of British enterprises.10 The resolution of conflict between the colonial government and civil society by force therefore built the foundations for political and economic processes that were monopolistic. 2.3 The decolonization phase, 1947–60 The Second World War altered global power structures and gave an impetus to decolonization globally. The difficulties of building a modern state became manifest at the first All Nigeria Constitutional Conference in 1950. The three regions failed to agree on any issue, and ethno-regional consciousness dominated the positions of each region. The 1951 Constitution generally reflected the position of the north. The conference was the first time that Nigeria faced the threat of secession, when northern demands for revenue allocation based on population were not accepted.11 However, the 1951 Constitution incorporated most of the positions canvassed by the delegates from the north, thus provoking criticism from the leading politicians and elites of both east and west. Following the conference, two political parties, the western Action Group (AG) and the Northern People’s Congress (NPC), were formed. Both were ethnically based regional parties. In the 1951 regional and 1952 federal elections, AG won in the west and the NPC won in the north, while the National Council for Nigeria and the Cameroons (NCNC) won in the east. The defeat of Dr Azikiwe in the west further deepened the east–west divide, and when he shifted his political base to the east, the majority–minority divide there also deepened, later undermining the survival of Biafra.
96 Case Studies: West Africa
The demand for independence in 1953 has often been explained as nationalism. However, closer analysis suggests that it was motivated by ethno-regional consciousness rather than nationalism. The north was not yet as prepared as the east or west for independence. The fears of the northern elite were warranted, and had eastern and western political elites been more understanding, the Kano riots, the first politically motivated ethnic crisis in Nigeria, might have been avoided.12 The threat of secession by the northern elite prior to the Kano riots is also an indication that they subordinated national consciousness to ethno-regional consciousness. The 1953 crisis revealed the fragility of the federation. The 1951 Constitution had indicated that independence would not produce a balanced, equitable and stable polity. Furthermore, the constitutional process 1950–9 leading up to independence was not accountable or accessible to the public. Although the constitution was agreed by elected politicians, the Nigerian people did not ratify the 1959 Constitution and in this sense, it was similar to those that went before it. 2.4 Transition to civil war, 1961–7 At independence on 1 October 1960, Nigeria was in disequilibrium: the minority problem was not resolved; and nationhood, citizenship and patriotism were abstract and ambiguous concepts. Furthermore, the emergent state, like the colonial state before it, was neither accountable nor transparent, as well as lacking the capacity to fulfil its role in a modern state. Within a month of Nigeria gaining independence, Britain and the new government signed an Anglo-Nigerian Defence Pact, a secret agreement. Although, it was later terminated after protests by the political opposition and students, the signing of the agreement exemplified the new government’s disconnection from civil society. The conflicting relations between native authority and traditional authority diminished in the south following reforms after the Aba Women’s Riot. However, in the north, it was manifested in intensified political activity. The Native Authority System (NAS) became a focus of mobilization against the dominant northern party (the NPC). The new form of asymmetrical relations that developed tended to be volatile and posed great danger to lives, property and civil liberties. The majority–minority problem remained unresolved at independence, although the Willinks Commission of Inquiry into Minority Fears had been set up in 1957. The minority east and west supported the creation of more states, which was opposed by the north. The Colonial Secretary claimed that no case had been made for the immediate creation of states, and the 1959 Constitution imposed conditions for the creation of new states that would deepen the potential instability of an already unbalanced and unstable polity.13 The east, west and north divisions had dominated the constitutional debate and the result was not fair, just or balanced. The NPC, through the NAS, controlled the north, the NCNC the east and the AG the west. An NPC–NCNC
Garba and Garba: Nigerian Civil War 97
alliance formed the government, and what followed was a political struggle for power. Taking advantage of the 1962 constitutional crisis, the alliance took control of administration in the west and created the mid-west in 1963. The western crisis, its exploitation by the alliance and the collapse of the alliance preceded the 1964 and 1965 elections, both of which were followed by violence in the west and north (the Tiv Riots). A military coup on 15 January 1966 caused a shift in ethno-regional power balance, with the eastern military elite replacing the northern elite. The new military government proposed a Unification Decree in 1966 that would have reconfigured Nigeria politically to reflect the position eastern delegates advocated at the 1950 Constitutional Conference. The Decree elicited at least the third clamour for secession by the north. Unlike the previous attempts (1950 and 1953) this one produced widespread demonstrations of support in the north, and a second coup, in July 1966, shifted the regional balance back to the north. However, the new government failed to bring security, and when war broke out on 6 July 1967 Nigeria was no longer a unified nation, nor did it have a national government. Although analyses such as Collier and Hoeffler’s (1998, 2002) are relevant, the Nigerian civil war is best understood as a historical event, the consequence of historical forces (colonization) and its ramifications. These were the set of asymmetrical relations and institutions built and nurtured between amalgamation in 1914 and the Richards Constitution. The asymmetrical relations and institutions excluded the people and gave opportunities for self-interest and those acting against the public good, and created an imbalanced and inequitable ethno-regional nation. Britain’s bias towards the north during the constitutional process frustrated the resolution of the minority problem as well as the resolution of what is today, commonly referred to as the ‘nationality question’. The discovery of oil in 1958 in the Niger Delta (a southern minority area), had serious consequences for stability and nation-building. The Constitution vested property rights in the central government, but its lack of accountability and transparency created a moral hazard problem that produced negative net effects on the local population.14 The economics and politics of oil involve the most powerful local and international players, and control of major economic resources was at stake: sooner or later, Nigeria’s dysfunctional state would generate open conflict at any of the major points of antithetical forces. The most likely points in terms of degree of volatility and scope were the divisions between east, west and north, the majority–minority relationship, and between state and civil society.
3
The civil war: duration and cessation
On 27 May 1967 the military government of Lt. Col. Gowon created twelve states. Three days later, Biafra declared itself an independent republic and
98 Case Studies: West Africa
on 6 July 1967 war began. According to Obasanjo (1981), the creation of new states was designed to exploit the majority–minority divides in the east, suggesting that the declaration of Biafra was a response to retain control over the eastern region. It is estimated that, when the war ended on 12 January 1970, more than a million Biafrans had died. Why did the war last as long as it did, and why did it end when it did? 3.1 Why the war lasted so long According to Obasanjo (1981), it was commonly believed that it would take just one month to capture Enugu and end the rebellion. Instead, the war lasted thirty months. There was a lag of a year between the July 1966 coup and the start of the war, a period that both sides used to prepare for war. The rebel side was able to mobilize its forces because after the ad hoc Regional Representatives Conference of 9 August 1966, all Igbo and eastern officers relocated to the east as recommended. By the beginning of the war, the rebel side had succeeded in mobilizing soldiers, population and resources as well as winning the support of influential international players through its relatively more effective use of the news media. The federal government also mobilized its troops and resources, although it was less successful in mobilizing the population and gaining international support. Biafra was better prepared for war. The federal side had planned to use only soldiers based in the north (Obasanjo, 1981). Such a plan indicated a lack of unity and purpose. Furthermore, the federal command and control system was neither cohesive nor disciplined; field commanders acted unilaterally and many disagreed with their commanders – such as Colonel Murtala, who deserted his command after accusing his headquarters of undermining the war effort (Obasanjo, 1981). Obasanjo also described weak morale among officers and men, as well as disagreements between officers. This lack of motivation diminished the capacity of the federal army. The mid-west and west lacked enthusiasm for the federal cause, which also contributed to low morale. As Obasanjo (1981) pointed out, it was only the Biafran leader Chukwuemeka Odumegwu Ojukwu’s invasion of the mid-west and his brief venture into parts of the west that united them with the federal cause. An important factor was the role of international players. The discovery of oil in 1958 had already drawn Nigeria into global economics and politics. Cold War politics ensured a strategic interest in Nigeria and its oil. Britain and France also continued to compete for influence in West Africa. Bertrand Russell maintained that all nations face the threat of ossification and foreign threats in the struggle to survive. Colonialism had already ossified, and was too fragile to cope with the more sophisticated predation of the Cold War era. Prior to the Nigerian war, the OAU, supported by General Ankrah, Emperor Haile Selassie and Martin Luther King had sought a peaceful resolution. However, the civil war presented opportunities to interested powers and, in particular, arms dealers. British political and economic interests, notably in the oil and
Garba and Garba: Nigerian Civil War 99
gas resources of the Niger Delta, were best served by a short war that quelled the rebellion. However, a short war was in conflict with the interests of arms dealers, many of whom were British. France would also have preferred a short war, but a rebel victory. Former French colonies such as Côte d’Ivoire and Gabon supported French interests. However, British and French military support (including French mercenaries) rendered a short war unlikely. Other European countries sold arms to one or both sides, and we can infer that a long war would have satisfied their immediate commercial interests. Others were more altruistic, such as humanitarian and religious groups, as were Zambia and Tanzania. The altruists tended to sympathize with Biafra, since its civilian population suffered greatly and the rebel cause resonated more deeply with the international media and non-governmental organizations in the First World, leading to support for the underdog. Islamic interests tended to support the federal government, while Christian organizations favoured Biafra. Israel, engaged in its own conflict with the Arabs, sided with the rebels, as their struggle for identity and autonomy echoed Israel’s own. To summarize, interplay between conflicting foreign interests and domestic interests prolonged the war. 3.2 Why the war ended Numerous civil wars in Africa defy a military solution. With Biafra’s surrender on 12 January 1970, the Nigerian war was an exception. Biafra was essentially an enclave of Nigeria. Once the federal army entered and captured key river routes, supply routes to the Igbo heartlands were cut. Ojukwu left Biafra on 11 January 1970 before his only escape route (Uli-Ihiala airstrip) was captured. Several hours later, the federal forces were in control of Biafra. Biafra’s geography therefore played an important part in bringing a swift end to the war once key positions had been captured. It is also important to note that neighbouring Cameroon, did not support the rebels.15 If it had, Biafra could have established bases for guerilla attacks, allowing a more flexible resistance strategy.
4
The aftermath and risks of recurrence
4.1 Post-conflict governance and nation-building The civil war ended with the restoration of federal military government control in eastern Nigeria, military rule and a return to northern dominance over the east. Formal surrender by the rebels was not followed by a national dialogue. Issues such as national integration, constitutional government, the minority problem, the nationality question and ethno-regional imbalances remained unresolved. Indeed, the government demanded and secured from the rebels a renunciation of secession, affirmations of loyalty, and acceptance of the existing twelve-state federation. After that, the government offered
100
Case Studies: West Africa
a general and, arguably, a non-binding commitment to reconciliation, rehabilitation and reconstruction entirely on its own terms. Since the end of the civil war, Nigeria has had ten governments, with an average tenure of 3 years 4 months. Only three were led by civilians. Peaceful transfer of power through elections has occurred only twice, but there have been five coups. The army has governed, by decree, usually without the support and consent of the civilian population. The first elected civilian administration ruled on the basis of a constitution decreed by the erstwhile military government. Similarly, the current civilian administration is legitimized by a Constitution, the preamble of which begins ‘Now the Federal Military Government hereby decrees as follows . . . ’. Nigeria has yet to develop a liberal democracy or democratically accountable institutions. Indeed, government and governance remain largely asymmetrical, elitist and disconnected from civil society. Military rule retarded the development of institutions to enhance the capacity of government and, by extension, of economic mechanisms that would allow non-state agents to compete globally. Six issues shed additional light on post-conflict governance in Nigeria and its consequences on institutional, political and economic development after 1970: the creation of states; petroleum industry laws; the Land Use Act (1978); implementation of a structural adjustment programme (1986); the annulment of the 12 June 1993 election; and the hanging of Ken Saro Wiwa (1996). From the 1950 conference to 1963, the northern political elite opposed the creation of new states. However, in 1963 and 1967, the northern-dominated central government created more states. Neither exercise was aimed at promoting effective governance but rather was motivated by ethno-regionalism and the extension of the tenures of military dictators. Since 1967, the twelve states have been expanded to a fragmented thirty-six through four statecreation exercises in 1976 (to 19), 1988 (to 21), 1991 (to 30) and 1996 (to 36). Each restructuring of the federation fragments and weakens the capacity of state and local governments, while strengthening central government. As oil revenue became Nigeria’s principal economic resource and its control became progressively centralized through military decree, the states have become dependent on transfers. In fact, these transfers have motivated demands for more states and local government councils. With limited manpower, the fragmentation of states has also reduced the quality of personnel, while administrative overheads have raised the costs of governance and reduced expenditure on social and economic goods and services, especially after 1988. State creation has also deepened ethno-regional political imbalances. For example, had the 1967 state structure prevailed in 1979, the north would have had 50 per cent of Senate members and the east and west 25 per cent each. However, after 1976, the north was allocated 53 per cent, the west 26 per cent and the east only 21 per cent. After the 1996 exercise, the west had only 22 per cent and the east 25 per cent. However, the north-west still
Garba and Garba: Nigerian Civil War
101
lags behind the west and east in access to electricity, safe drinking water, health and education services (Garba, 1999). Most of the changes in the laws regulating the petroleum industry have favoured either the central government or foreign oil companies. Those favouring central government include the 1969 Petroleum Laws, which vested onshore and offshore property rights in the federal government; the 1973 Uniform Pricing Decree, which, inter alia, transferred sales tax on petroleum products to the federal government; and the 1974 Petroleum Refining (Regulations) Acts, which created a federal monopoly in the refining and distribution of petroleum products. All undermined efficiency in the industry and an effective federal fiscal system. Laws favouring oil companies include the Oil Pipelines Acts (1958, 1965 and 1985) and Memoranda of Understanding (1988 and 1991). The former subordinated the rights of Nigerian citizens to private property to those of oil companies to lay and maintain oil pipelines subject to compensation determined by the government. The Memoranda of Understanding guaranteed oil companies risk-free investments in oil and gas by setting a minimum profit margin. The consequence was to shift the burden of fluctuations in global oil prices to the government and non-state agents through an inflation tax. As Garba and Garba (2001) show, collusion between the government and oil companies, and the former’s dependence on oil revenue, acted as a disincentive to control gas flares. As a consequence, Garba and Garba estimated that about 109,000 metric tonnes of carbon dioxide was emitted into the environment of the oil-producing communities between 1986 and 1996, equivalent to 380,000 barrels per day. As well as degrading the environment, US$28.14 billion was lost between 1986 and 1996. The Land Use Decree (Act, 1978) also appropriated property rights for the federal state. In a low-income agrarian society, competition for land is fierce since survival depends on it. The potential for abuse, especially by government officials, is high, increasing risks of tension and conflict, especially when the economy dips into recession. In 1986, despite the overwhelming opposition of Nigerians, the military government imposed a structural adjustment programme. Nigeria’s economic decline after 1986 is well documented. Following the structural adjustment programme there was an increase in conflict and acts of violence. The annulment of the June 1993 election and the hanging of Ken Saro Wiwa in 1996 both exemplify the lack of political development since the civil war ended. In addition, the hanging of Saro Wiwa and the state-sponsored violence in Ogoni shows the failure of the state and the social costs of a complementary relationship between state failure and market failure. The two events contributed to the voluntary and peaceful transfer of power in 1999, but nation-building in Nigeria is still undercut by the ossified and asymmetric political, economic, social and institutional forces that post-colonial governments inherited and nurtured. This is why divisive issues such as the Sovereign National Conference, sharia’a law and religion, control of resources, ethnic
102
Case Studies: West Africa
militia, citizenship and the national ID card scheme dominate public discourse, generating deep division and widespread apprehension. 4.2 Post-conflict economic management and development The emergence of oil revenue as the principal resource and major source of foreign exchange is probably the single most important economic change in the post-conflict era. Oil has dominated the economy since 1971, following a 206 per cent growth in oil export revenue. By 1980, oil accounted for 89 per cent of Nigeria’s revenue. A positive effect was the removal of budgetary constraints on post-conflict reconstruction. Oil revenue underwrote the National Development Plans for the reconstruction and rehabilitation of war-damaged infrastructure. In terms of growth, capacity utilization, employment, investment, external balance and strength of the naira, the decade after the civil war was the best for Nigeria. Growth estimates using Central Bank data show that between 1971 and 1980 the economy grew at an annual average rate of 6.2 per cent, compared to an annual decline of 5.7 per cent during 1981–5, and 4.1 per cent growth during 1986–96. Capacity utilization averaged 73.7 per cent (1971–80), 53 per cent (1981–5) and 37.5 per cent (1986–96). Economic growth between 1971 and 1985 mirrors that of oil revenue. The reversal of the economy after 1980 shows that oil-induced growth was unsustainable. The conventional wisdom is to blame political and economic policy. Expenditure levels were unsustainable, investment choices were inefficient and unproductive, and in many cases (steel, telecommunications, road construction) public investment was merely a mechanism for creating, sharing and exporting rent, as the many commissions of enquiry have subsequently revealed. From an institutional perspective, the critical factors include the lack of accountability and transparency of the public policy process, fiscal centralization, subordination of monetary policy to fiscal policy, and short-sighted macroeconomic management. In 1978, workers and students unsuccessfully opposed the so-called ‘jumbo loan’, which laid the foundation of the 1982 external debt crisis and created the conditions that precipitated the choice between surrendering to external control by accepting structural adjustment or maintaining policy autonomy and national sovereignty. Again, against the opposition of Nigerians, the government chose to accept structural adjustment. Government expenditure, both current and capital, favoured expenditure on administration rather than on the provision of social services. The available evidence indicates that the number and intensity of conflicts escalated (Egwu, 1998). Successive governments centralized fiscal policy, enlarging the public sector and imposing public monopolies in key areas such as oil refining and the supply and distribution of petroleum products. The creation of new states, the imposition of the petroleum laws, and revenue allocation decrees enabled central government to increase its control of resources and to expand its
Garba and Garba: Nigerian Civil War
103
control of economic, social and political processes. A former governor of the Central Bank of Nigeria stated that: The military abolished the ‘fair shares’ principle and the country’s political, social and economic development have been the worse for it. The military centralized and concentrated the country’s revenues in the hands of the Federal Military Government and it has now been revealed since the military left governance that the concentration of revenues in their hands facilitated the plundering of the national coffers by the departed military rulers. (Vincent, 2001) Fiscal centralization also had an adverse effect on macroeconomic stability, the effectiveness of monetary policy and the provision of public goods. Oil revenue is the basis of Nigeria’s centralized fiscal federalism. Failure to meet revenue targets compromised monetary policy, as in 1999 and 2001. Government borrowing overshot targets, causing domestic credit and money supply also to overshoot, which crowded out private borrowing (especially in 2001). This type of macroeconomic management undercuts growth objectives, sustainable employment and private investment, especially as the Central Bank tends to fight inflationary pressures by raising interest rates. Post-conflict mismanagement of the economy has shifted Nigeria from being a low-to-middle-income and low-debt country to a low-income, heavilyindebted country. Although the growth data indicate an average annual growth rate of 4.1 per cent between 1986 and 1996, such a rate of growth is not consistent with the decline in capacity utilization and decline in average annual investment (from US$12,110 million in 1981–5, to US$2,364 in 1986–8 and US$3,429 million between 1989 and 1993), the shrinking of real activity, rising unemployment, poverty and conflict. 4.3 Risks of further conflict Despite military coups, the 1970s were relatively peaceful and the economy performed well. From 1980, the incidence, scope and spread of open conflict increased progressively. Major points of conflict include religion, communality, oil, the annulment of the 12 June 1993 election, and the growth of militiapolice. The three fundamental divides (government–civil society; east– west–north; and majority–minority) remain at the root of both rural and urban communal conflicts. In the case of oil communities, wider issues are at stake. The Ogoni crisis, which cost 331 lives between 1990 and 1994, was partly related to local causes but also linked to external manipulation relating to oil, and compounded by a failure of the government to protect the lives and property of minorities. This, again, exemplifies the dysfunction of the Nigerian nation-state. Since the 1993 crisis, and particularly under the 1993–8 government, the violent role of ethnic militias has increased, and ethnic militias such as the
104
Case Studies: West Africa
Odua People’s Congress and Bakassi Boys have often usurped the police in preventing crime and administering justice.16 This loss of trust in the state has deepened tensions between the militia and state security organs, especially the police. Militia justice is swift, violent and arbitrary, lacking the principles of crime control and dispensation of justice expected of a modern state. Militias are therefore more likely to intensify violence. The government–civil society divide, dating back to the colonial period, was at the root of religious crises that swept the north between 1980 and 1985.17 However, subsequent religious crises have been essentially ethnic rather than religious in nature. Those involved in religious conflicts tend to be poor and their manifestation is usually looting and killing (sins in all religions). Causal factors are economic and political in addition to religious. The cumulative effect of post-conflict governance and economic mismanagement has produced conflicts in diverse locations: Zangon Kataf, Wukari and Takum in the north; Ife/Modakeke (west); Aguleri and Umuleri (east); Ijaw, Urhobo and Ishekiri (Niger Delta); and Hausa-Yoruba (Lagos, Sagamu and Kano). As economic conditions worsened, and people depended more on their land, shortages produced tensions and conflicts in rural communities (see Egwu, 1998). The frequency and intensity of communal conflict is linked to the scarcity of land caused by the 1978 Land Use Decree, which enabled the misappropriation of land by politicians and foreign companies. The Ogoni crisis of 1990–4 claimed at least 2,300 lives, while during the 1990s at least 1,056 people died as a result of ethnic conflict, which also injured many, displaced many others and caused considerable damage to property. Communal crisis over land resources has become more frequent, widespread and deadly. That Nigeria has not relapsed into a civil war is partly because of the economic boom of the 1970s, which allowed for reconstruction. The election of an Igbo as vice-president in 1979 and the granting of a pardon to the rebel leader were also positive signs. However, economic crises after 1981 exposed the fragility of the Nigerian economy, while the progressive frequency, scope and growing intensity of open conflict reveals the fragility of the Nigerian nation. Citizenship, for example, remains amorphous. A typical Nigerian has multiple layers of citizenship and allegiance: national, state, ethnic, religious and local, and differential rights in different parts of Nigeria. Multiple citizenship breeds confusion and controversy, frustrates the development of national unity and a harmonious political community, and generates labels such as ‘strangers’, ‘settlers’ and ‘aliens’ applied to Nigerians within Nigeria. Multiple citizenship also causes discrimination in jobs, land purchase, access to housing, admission to educational institutions, marriage, business transactions and the distribution of social services (Nnoli, 1998). The idea of a national conference to resolve these issues does not yet enjoy a consensus among the Nigerian elite. However, Nigeria in its present form is unsustainable. The spate of bomb blasts and acts of terrorism, and the rise
Garba and Garba: Nigerian Civil War
105
in intensity and spread of other forms of conflict, necessitate urgent reform of political and economic institutions. The ongoing political process, while fraught with dangers, also presents great opportunities for reform. The judiciary, especially the Supreme Court, has begun the process of institutionbuilding through landmark judgements on the boundaries of littoral states, the 2001 Electoral Bill, the registration of political parties, and the participation of civil servants in politics. The judgement on the boundaries of littoral states has led to the passage of the landmark Onshore–Offshore Bill. Supreme Court guidelines covering the registration of political parties liberalized the political arena, reduced tension and the need for violence, and allowed nonprofessional politicians and civil rights activists to participate. The judgment on the participation of civil servants in politics should increase levels of participation of the working class in the political process. These institutional changes should increase Nigeria’s chances of overcoming its incapacitating colonial and post-colonial heritage.
5
Conclusions
This chapter has shown that the seeds of civil war were sown during Nigeria’s amalgamation, the way it was governed as a colony and the way its independence was arranged, producing an unequal and inherently unstable polity with an elite preoccupied with power struggles instead of nation-building. The civil war was the climax of a phase of this power struggle, prolonged by the interplay between conflicting foreign and domestic interests. Several lessons can be learnt from Nigeria’s post-civil-war experiences. First, recurrence of conflict is less likely if the war ends in military victory. Second, rebellion in an enclave region is more likely to end in defeat. Third, settlements that exclude foreign interests are more likely to endure. Fourth, a positive external trade shock and the absence of a debt overhang have positive effects on reconciliation, rehabilitation and reconstruction. Fifth, lack of democracy and poor governance increase risks of recurrence. Finally, the risks tend to rise as economic conditions worsen. At the end of the war, the principle of ‘no victor, no vanquished’ informed the reconciliation, rehabilitation, reconstruction and reintegration programmes. National planning was made easier because of the oil boom and the absence of a debt problem, which rendered foreign support unnecessary. However, the exploitation of ossified and asymmetric institutional, political, economic and social conditions by post-conflict governments undermined governance and efficient macroeconomic management. Conflict along key ethnic/communal, religious, state–civil society and majority–minority divides increased and intensified after 1980. Before the 2003 elections, discussions focused on divisive issues such as a sovereign national conference, sharia’a law, control of resources, the offshoreonshore dichotomy, religion, ethnic militias, social tensions, citizenship, the national ID Card scheme and political assassination. After the 2003 elections,
106
Case Studies: West Africa
additional divisive issues emerged and they include mass action, state of emergency, deregulation, National Economic Empowerment and Development Strategy (NEEDS), etc. If before the 2003 elections, Nigeria was at a crossroads, facing dangers as well as opportunities, the political events of 2003 and 2004 and the aftermath suggest that Nigeria has not begun exploiting the opportunities provided by the judgments of the Supreme Court to begin the long and difficult process of nation-building. The rising tension and conflicts seem to suggest that Nigeria must introduce political and economic reform or transformation, although its ruling elite opposes this, in order to build a functional modern state. The great challenge for Nigeria thus remains: How can it overcome the ossified asymmetrical relations, consciousness and vested interests that have prevented Nigeria from achieving its potential? If Nigeria does not succeed, it could not show the way for other African countries that have also failed to build modern states on colonial foundations.
Notes * This chapter is an abbreviated version of Garba and Garba (2002). 1 Garba and Garba (2002) conceptualized conflict as a struggle for supremacy or strategic advantage by antithetical forces and can be either open (when antithetical forces fight) or inherent (an inner struggle). The civil war is thus one example of open conflict. 2 Elbadawi and Sambanis (2001) limit their analysis to the correlation between risks of war, on the one hand, and political liberalization, standards of living and economic diversification on the other. Relations between political liberalization, standards of living and economic diversification were under-emphasized, as were the relations between national and international process. Reliance on cross-country regression analysis is, in our opinion, ill-suited to historical phenomena intrinsically linked to processes forming, consolidating and dissolving nation-states. 3 Individuals continually confront antithetical forces, such as: passion/pleasure; prudence/forethought; mind/matter; duty to God/duty to the state; good/evil. Examples of antithetical relationships in society might include: state/citizens; social cohesion/individual liberty; order/anarchy; slaves/masters; aristocrats/commoners; colonizers/colonized; Western values/African values; tradition/modernity; science/ religion; rich/poor; military/civil; socialism/communism; capitalist/worker; debtor/ creditor; and unlimited wants/limited resources. 4 Increased incidences of open conflict preceded the failure of Yugoslavia, the Soviet Union and other former communist states. 5 In 1907, the Crown Agents overruled objections from the southern government when granting a loan of £1,230,000 to the southern government, to build a railway in the north out of surplus southern funds. Similarly, in 1910, the Colonial Office enforced a loan of £200,000 from the south to the north. 6 The colonial state adopted a strategy of ‘divide and rule’. Intrinsically asymmetrical, its system of indirect rule used local collaborators as a bridge between it and the people. 7 Sir Hugh Clifford justified this on the grounds of practical difficulties caused by the size of the country, its poor communications infrastructure, Nigeria’s ethnic diversity and the legal distinction between a colony and a protectorate (Olusanya, 1980). However, as Olusanya showed, Clifford’s arguments were weak.
Garba and Garba: Nigerian Civil War
107
8 For example, while southern Nigeria ‘led the way in the achievement of the elective principle in Tropical Africa’ in 1922, it was not until 1951 that northern Nigerians gained similar rights. It is worth noting, however, that even in the south the elective principle was limited only to Lagos and Calabar and was not based on universal adult suffrage. 9 According to Abubakar (1980), integration was resisted by colonial officials in the north, who preferred adherence to tradition and a north that developed along its own lines. The officials played on differences between north and south, using scare tactics, arguing that integration would lead to the enslavement of Muslims by semi-Europeanized southerners. 10 Chapter 130 of the Laws of Nigeria: ‘The powers conferred upon the Governor General to grant licences and leases for mineral oils shall be exercised subject to the following conditions: No lease or licence shall be granted except to a British subject or to a British company and having its principal base of business within Her Majesty’s dominions, the chairman and the managing director (if any) and the majority of the other directors of which are British subjects.’ 11 After the civil war, population and land mass did become criteria for resource allocation. 12 The motion called for Britain to grant Nigeria independence by 1956. Northern politicians argued that the south would dominate; southern newspapers were highly critical of the northern politicians, who were booed in Lagos. The Kano riot was triggered by an AG visit to Kano to explain to the people why they should support independence in 1956. After four days of riots in Kano, fifty lives were lost, more than 200 were injured and property was damaged. 13 There were two methods of creating a new state: a two-thirds majority in the Federal Legislature and a simple majority in two regions, one of which must be the region to be split; or a simple majority of the Federal Legislature and a referendum in the area to be transferred. See Olusanya (1980) for further details. 14 The moral hazard problem is, in essence, about a lack of incentive for central government to ensure that the Niger Delta derives benefits equitably from its endowments. Ongoing discussion of the Onshore–Offshore Bill exemplifies the problem. 15 It has been speculated that Nigeria may have made territorial concessions to Cameroon to secure its support. If this is true, then there may yet be a high cost, given the ongoing conflict between Nigeria and Cameroon over the Bakassi Peninsula. 16 The Odua People’s Congress was formed after 1995 to oppose the perceived persecution of the Yoruba. From 1999, the Bakassi Boys and the Movement for Actualization of the Sovereign State of Biafra also developed in the east. The political crisis that followed the annulment of the 11 June 1993 elections deepened the ethnic divide in Nigeria, and for the first time since the end of the civil war, Nigeria faced threats of dissolution. 17 This was the first time an organized military group opposed the government. Religion had motivated mobilization. The political objectives were not as clearly defined as those of the militias or those of other non-violent groups such as the Movement for the Survival of the Ogoni People.
References Abubakar, S. (1980) ‘The Northern Provinces Under Colonial Rule’ in O. Ikime (ed.), Groundwork of Nigerian History (Ibadan: Heinemann). Adedeji, A. (ed.) (1999) Comprehending and Mastering African Conflicts: The Search for Sustainable Peace and Good Governance (London: Zed Books).
108
Case Studies: West Africa
Afigbo, A. E. (1980) ‘The Eastern Provinces Under Colonial Rule’ in O. Ikime (ed.), Groundwork of Nigerian History (Ibadan: Heinemann). Asiwaju, A. I. (1980) ‘The Western Provinces Under Colonial Rule’ in O. Ikime (ed.), Groundwork of Nigerian History (Ibadan: Heinemann). Collier, P. and A. Hoeffler (1998) ‘On Economic Causes of Civil War’, Oxford Economic Papers, vol. 50, no. 4, pp. 563–73. Collier, P. and A. Hoeffler (2002) ‘Greed and Grievance in Civil War’, Centre for the Study of African Economies Working paper No. 128 (Oxford: CSAE). Collier, P., A. Hoeffler and M. Söderbom (2001) ‘On the Duration of Civil Wars’, working paper (Washington, DC: World Bank). Egwu, S. G. (1998) ‘Structural Adjustment, Agrarian Change and Rural Ethnicity in Nigeria’, Nordic Africa Institute Research Report 103 (Uppsala: Nordiska Afrikainstitutet). Elbadawi, I. and N. Sambanis (2000a) ‘External Interventions and the Duration of Civil Wars’, paper presented to the conference on ‘The Economic and Politics of Civil Conflicts’, World Bank and Princeton University, 18–19 March 2000. Elbadawi, I. and N. Sambanis (2000b) ‘Why Are There So Many Civil Wars in Africa? Understanding and Preventing Violent Conflict’, Journal of African Economies, vol. 9, no. 3, pp. 244–69. Elbadawi, I. and N. Sambanis (2001) ‘How Much War Will We See? Estimating the Incidence of Civil War in 161 Countries’, Working Paper No. 2533 (Washington, DC: World Bank). Garba, A.-G. (1999) ‘Are African Countries Configured for Good Governance?’. Prepared for the International Conference on Good Governance and Sustainable Development in Africa held in Abidjan on 22–24, November 1999 (mimeo). Garba, A.-G. (2002) ‘Macro-institutional Framework for Harmonizing Fiscal and Monetary Policy’. Presented at CEAR Staff Seminar, Department of Economics, University of lbadan, 31 January 2002 (mimeo). Garba, A.-G. and P. K. Garba (2001) ‘Market Failure, State Failure and Air Pollution in Nigeria: A Theoretical Investigation of Two Cases’ in A. G. Garba (ed.), Natural Resource Use, The Environment and Sustainable Development (Ibadan: Nigerian Economic Society). Garba, A.-G. and P. K. Garba (2002) ‘Open Conflicts when the State and the Market are Failing: The Case of Nigeria’, presented at Post-Conflict Economies World Bank, AERC, IEA Special Workshop, Kampala, 5–6 July 2002. Hildyard, N. (1999) ‘The Political Economy of Violent Conflict’, in M. Suliman (ed.), Ecology, Politics and Violent Conflict (London: Zed Books). Ikime, O. (ed.) (1980) Groundwork of Nigerian History (Ibadan: Heinemann). Lawal, A. A. (1998) ‘British Financial Administration in Nigeria: 1900–1960’, The Nigerian Journal of Economic History, vol. 1, pp.101–28. Nigerian Economic Society (1996) Beyond Adjustment: Towards Better Management of the Nigerian Economy, Selected Papers for the 1996 Annual Conference (lbadan: Nigerian Economic Society). Nnoli, O. (1998) Ethnic Conflicts in Africa (Dakar: ODESRIA). Obasanjo, O. (1981) My Command: An Account of the Nigerian Civil War 1967–70 (Ibadan: Heinemann). Olusanya, G. O. (1980) ‘Constitutional Development 1861–1960’, in O. Ikime (ed.), Groundwork of Nigerian History (Ibadan: Heinemann). Suliman, M (ed.) (1999) Ecology, Politics and Violent Conflict (London: Zed Books). Vincent, O. (2001) ‘Fiscal Federalism: The Nigerian Experience’, Nigerian Economic Society 2001 Public Lecture, Abuja, 11 December.
7 The Economics of Civil Conflict in Africa: The Case of Chad Nadjiounoum Djimtoingar and Djona Atchénémou Avocksouma
1
Introduction
As several authors have pointed out (Collier and Hoeffler, 1998, 1999, 2002; Collier, Elbadawi and Sambanis 2000; De Soysa, 2002; Sambanis, 2002) political and economic shortcomings constitute the basis for civil war. This chapter applies these ideas to Chad, leading up to the outbreak of civil war in 1979, and shows how the conflict had an impact on the subsequent development of Chad’s economy. At the time of writing, petroleum is beginning to flow but conflict continues, the causes of which endure. Chad is a large, land-locked country with a population of approximately 7.5 million, almost half of whom are below 15 years old. Over 300 dialects are spoken, but French and Arabic are the two official languages. Just under half the population are Christian or animist farmers, living in the southern regions where cotton, the principal export commodity, is produced. Islamic pastoralists and traders live in the northern and central regions, the larger portion of the country. Petroleum, found in the south of the country, is the most important natural resource and exports began in October 2003. Other resources include salt, uranium and gold; however, the exploitation of these has been handicapped by the conflict. Statistics for 1995 indicate that about 60 per cent of the population live below the poverty line. In 1997, the infant mortality rate was 103 per 1,000. Maternal mortality for recent years is estimated at 827 out of 100,000. According to UNDP (2000), the human development index for 1999 was estimated at 0.403, while the human poverty index was estimated at 57 per cent.1 This means that, in 1999, about six out of ten Chadians suffered from the effects of poverty. About 36 per cent of Chadians are likely to die before the age of 40; only a third of the population had access to basic health services, and eight people out of ten did not have access to potable water. The illiteracy rate is about 80 per cent. This gloomy picture is a result of decades of poor economic growth and resource distribution. Available data indicate an average annual growth rate 109
110
Case Studies: West Africa
of less than 1.5 per cent since 1980. The Chad economy is vulnerable to climatic risks, given the dominance of the primary sector (80 per cent of the active population). Insecurity also helps to explain why Chadian entrepreneurs prefer the less risky commercial sector to production activities. The economy is currently dependent on two primary exports – cotton and cattle. National wealth is monopolized by a minority. Development projects have been mismanaged, leading to the failure of almost all political and economic projects. In the final analysis, the conflict is inextricably linked to the way that major issues of political, economic and social development are managed.
2
The historical dimension
2.1 The pre-colonial period Prior to the colonial period, Chad comprised a multitude of territories and diverse communities. Influenced by the Kanem, Baguirmi and Ouaddaï empires, Islamic pastoralists settled on the right bank of the Chari river. Animist farmers were settled on the south bank – ‘useful Chad’, as the French were later to describe it. Interaction, defined by tradition and custom, was confrontational but symbiotic, characterized by competition for land and resources but tempered by the need for trade and exchange. Gali (1985, p. 32) shows that the better-organized and better-equipped northern Muslims waged war incessantly among themselves and with their southern neighbours. The north–south dualism of economic activity was glaring. Their limited defence capacity made southern communities the preferred target for northern raids2 to acquire horses, slaves, women and provisions. Their organization into decentralized communities handicapped their mobilization against invaders. Thus from the beginning of the nineteenth century, the south was a slave reservoir for the Sahelian sultans. At this time, north–south conflict took the form of military dominance, economic in nature, by the north over the south, only becoming political conquest with the later success of the National Liberation Front (FROLINAT). 2.2 The colonial period Modern Chad is a product of the Berlin Treaty (1885) in which France, Germany and Britain divided Africa. According to Bouquet (1982) Chad was created by the courage of soldiers pursuing exalted ideals, and by treaties signed, modified and ratified in the corridors of European power. The reality was somewhat less romantic and much more difficult. Colonial troops took fifteen years (1900–15) to pacify Chad, then forty years (1920–60) to organize and administer it (see Gali, 1985, p. 31). The colonialists created a state which ignored the boundaries of existing communities, bisecting those now also
Djimtoingar and Atchénémou Avocksouma: Chad 111
found in Sudan,3 Niger and Libya. From the outset, Chad suffered from poor institutional arrangements (Ziegler, 1978; Michalon, 1979; Bouquet, 1982). Just as the economy remained dualistic, political authority was duplicated. Chad comprises two distinct parts, differentiated by geography and culture. Consequently, Chad began as a collection of communities having nothing in common apart from a conflict of interests and unpleasant memories of raids and slavery. In asserting their authority, the colonialists superimposed a system that was not adapted to local values: ‘the policy of the colonialists and governments that followed practically sought to destroy the subsistence economy so as to promote commercial agriculture and thereby obtain a commercial surplus’ (Grellet, 1982, p. 25). The French regarded the south as ‘useful Chad’. In comparison, in the north there was nothing much to develop, at least from the standpoint of the colonial power primarily interested in primary products and export crops. In 1920, cotton cultivation was introduced to the south. This had the effect of exposing the settled southern communities gradually to capitalist production, including waged labour and modern infrastructure such as roads, factories, schools and health posts. In contrast, these benefits were inimical to the northern pastoralists, who rejected them, with the result that Chad’s economic duality widened. The south effectively became the economically useful zone for France. Initially, the southern communities were also uneasy over the introduction of cotton, since its cultivation required production structures that were inconsistent with their traditional socio-economic organization. For many authors, the southern socio-economic systems, far from being a handicap, proved to be beneficial in the long run. According to Teisserenc (1972, p. 26) ‘the colonial administration gradually marginalized the traditional administration. The weakening of traditional authority was accompanied by the weakening of its underlying institutions and beliefs’. This dismantling process occurred with the integration of Western values. Magnant (1986, p. 268) shows that ‘following the introduction of cotton, the state . . . intervened directly in the land tenure system. In the face of peasant revolt, force was used to impose the new crop and new working methods, and also to determine what piece of land to allocate to cotton cultivation and ensure compliance’. Added to cotton cultivation was mandatory taxation. Diguimbaye (1969) justified this policy on the grounds that ‘cotton would generate enough revenues to enable tax collection and develop a market economy’. Consequently, according to Cabot (1982), at independence the south was economically better placed than the north: more trained cadres, cash crops accounting for 70 per cent to 80 per cent of the country’s exports, modest but discernible industrialization. Furthermore, by adapting to the needs of the colonizers, the southern population, especially its elite, gained economic and political advantage.
112
Case Studies: West Africa
It is therefore understandable that, by independence in 1960, the northern population had begun to feel neglected and resentful. FROLINAT, the National Liberation Front, later based its policies on the southern-bias argument: ‘the entire Chad economy is based on enhancing production in the south . . . Gradually the north was drained of its lively force by the degradation of the standard of living, the cessation of all economic activity, the accelerated pauperisation of rural communities overwhelmed by the burden of all sorts of taxes.’
3
The civil war
Civil war broke out in 1979, but the beginning of the conflict can be traced back to the early 1960s. A characteristic of post-independence politics is that those in power base their strategy on fomenting discord between communities. In Chad at the time of writing, regime change is driven by the desire for revenge and to exclude others from the national wealth. No government has truly sought to develop and implement a policy of unity and equity, or embrace mediation and compromise. One consequence is a lack of unanimous support: decisions are always challenged by a particular interest group. Thus, rightly or wrongly, many northerners did not see a role for themselves in the government of Tombalbaye, Chad’s first president and a southerner. This does not mean that he had unanimous support in the south; indeed, the first uprisings against Tombalbaye’s rule began in the south, following the death of the member of parliament, Jacques Nadingar, who had initiated a motion of censure against the government in October 1960. Following the introduction of one-party rule under the Chad Progressive Party (PPT-RDA) clashes occurred between Ngarta (François) Tombalbaye and leaders of political parties disbanded in 1962. The ruling party’s decision, taken at its Fort Archambault congress, to impose one-party rule led to demonstrations by opposition parties in September 1963. These were violently suppressed and were followed by the arrest of the leaders (who were mainly northerners). This marked a historical turning point, engendering sentiments of frustration among northerners, who felt excluded from national politics, and reinforcing sentiments of discrimination felt by northern Muslims following the accession to power of Christian elites in 1960. In 1966, FROLINAT was established. This was Chad’s first politico-military movement and marked the birth of armed opposition to the government. FROLINAT intensified its armed struggle despite the overthrow of the Tombalbaye dictatorship by the military on 13 April 1975. Although the new president and head of the Superior Military Council (CSM), General Félix Malloum, was a northerner, the situation did not change. FROLINAT gained access to power in 1978 with the nomination of Hussein Habré as prime minister. However, differences soon surfaced between the president and the prime minister, leading to civil war in February 1979.
Djimtoingar and Atchénémou Avocksouma: Chad 113
The conflict lasted until 1982. It was characterized by the division of the country into zones of influence managed by militias and warlords (1979–80), and by a coalition in the exercise of power (1981–2) under the Transitional National Union Government (GUNT). Although Hussein Habré’s accession to power on 7 June 1982 marked the end of the civil war, conflict has continued to underpin politics in Chad, particularly since Idriss Déby seized power in 1990. Pockets of resistance persist in both the north and south, casting doubts on the north–south, Muslim–Christian–animist, pastoralist–farmer divides generally considered as causes of the conflict. According to Bétoudji (2003, p. 6) social injustice, especially nepotism among members of the ruling ethnic group, the poor management of public resources and repression sustain the conflict.
4
The negative impact of the civil war
Lack of any rigorous study of the economic costs of the Chad conflict make it difficult to furnish quantitative data and an objective analysis. However, while the value of a human life or a burnt-out lorry full of cotton or a bridge damaged by rebels might be estimated, it is difficult to put a value on the hopes, plans and ideas stifled by the conflict. Another difficulty is that the civil war’s negative impact occurred before its outbreak and will be felt long after the final reconciliation. Given these problems, this chapter does not attempt to detail either the direct or indirect costs. Rather, Chad’s economic backwardness is ample evidence of the scale of the damage caused by the war. The war destroyed much of Chad’s heritage and infrastructure – public and private buildings, roads and cattle, and had negative environmental impact. More than half the buildings in the capital were destroyed. Many pastoralists fled to neighbouring countries. Wildlife in the national parks suffered. Whole production sectors disappeared. High military expenditure deprived spending on health, education and social development. Furthermore, much of the country continues to be covered in mines, rendering agriculture and cattle-breeding problematic and dangerous. In addition to the human costs, de-mining operations are estimated at US$88 million, while the costs of demobilizing and reintegrating former combatants would be much higher. For several years, thousands of men were away from productive activities and mobilized for war. Besides the combatants and civilians killed in the war, some 40,000 people, including trained staff, large-scale farmers and traders, disappeared between 1982 and 1990, especially during the Habré regime. At the time of writing, both the civil service and the private sector face a shortage of trained staff. The cost of rebuilding this human capital reflects the scale of lost income. The excesses committed during the civil war led to further mistrust between north and south Chadians. Southern civil servants working in the
114
Case Studies: West Africa
north were massacred by northerners, as were northerners working in the south. Added to this was the cruel treatment of war prisoners in the north. These acts have sown doubt and mistrust among communities. Abuses committed by different ethnic groups under the different regimes in power over the years have added to this mistrust, which no government has attempted to alleviate with any success. This mistrust is damaging to the economy: according to authors such Fukuyama (1995) and Charreaux (1998) trust plays an important role in promoting economic performance and is a factor in explaining differences in performances across national economic systems. Lack of trust in Chad helps to explain the imbalances between the regions of the country, the overall poor economic development and the resulting frustrations of the population. Dishonesty and corruption have become a way of life. Regarding the management of public resources, the 1979 civil war marked the beginning of a period of corruption and inefficiency. The state was unable to assert its authority throughout the country, and fraud and crime went unpunished. Appointments to senior public positions, often guided by considerations other than competence, and the association of political power with a particular group of individuals or ethnic group, engendered grievances, with disastrous consequences for Chad’s economic prospects. The multiplicity of taxes and barriers was aimed at private gain for relatives, military leaders or other allies. Lack of transparency in public management, which characterized the different regimes, eroded their credibility. The high turnover rate in top public posts, notably at ministerial level, encouraged the embezzlement of public funds. The climate of insecurity and the vagueness of property rights discouraged investors, in particular foreign investors. Access to public resources came to be based on clan membership, to the exclusion of non-members, thus sapping the state’s social fabric. The concept of collective responsibility introduced under Hussein Habré meant that a person’s family and ethnic group became liable for an illegal act committed by that person. This led to abuses: whole families were dispossessed of their property, driven from their homes, imprisoned or simply eliminated.
5
The aftermath of civil war: mismanagement of the economy
Redressing injustice felt by northerners was the basis of FROLINAT’s cause. It was expected that, once in power, FROLINAT’s political leaders would administer public affairs differently. Unfortunately, though, the situation changed from bad to worse. Whereas the first post-independence government appeared to have an economic agenda, succeeding governments failed to come up with any economic programme at all. This shortcoming has been more pronounced since FROLINAT came to power in 1979 and is all the more worrying as the level of national output has hardly changed while the population has been growing. Competition for the economy’s
Djimtoingar and Atchénémou Avocksouma: Chad 115
resources, unchanged since independence, became fierce. It may be argued, therefore, that one of the root causes of conflict in Chad has been the redistribution of resources through confiscation by the political authorities. The primary objective of political power in Chad is control of resources, whether emanating from public or private enterprises, the civil service or international aid. According to Dadi (1987, pp. 137–8), Hussein Habré’s appointment as prime minister in 1978 was an incarnation of the aspirations of the Muslim population and ‘implied a sharing of resources between the north and the south. The Muslim population expected this, more so given their conviction that central government authority held by the south systematically denied Muslims access to high-level offices: participation in public administration was a fundamental demand made by northerners’. Developing the reasoning further, ‘geopolitics’ became institutionalized. Since 1979, when the president comes from the north, the vice-president, prime minister or speaker of Parliament must come from the south; if the army chief of staff is from the south, this must be balanced by an appointment from the north, and so on. However, in reality, power was in the hands of northerners. As Dadi acknowledges, the exigencies of justice are not always compatible with the basic principle of the civil service, notably competence. This discriminatory practice has a tribal undertone. As Beyem (2000, p. 238) argues, tribalism poses great dangers for socio-political stability, and fosters the erosion of national integration. The current practice of tribalism in the management of state affairs is based on the fear of actual or potential dangers facing the group in power. From this perspective, power becomes an instrument for political, economic and social promotion of members of the ethnic group of the head of state: those in power are concerned only with enriching themselves and rewarding their allies, exploiting the state while they can, and flouting constitutional authority in the process. From independence to the present day, nothing has changed fundamentally. Members of the ruling ethnic group grab everything, be they the Mbayes under the regime of Malloum and Kamougué, the Goranes under Habré or the Zaghawa under the current Déby regime. Knowing this, other tribes or tribal groups are waiting their turn, thus triggering the recurrence of conflict. A general lawlessnes has developed in Chad. The lack of an independent judiciary, delays in pronouncing sentences, political interference in the judicial process, and the impunity enjoyed by people who embezzle state funds, are major factors leading to a crisis of confidence between the public and the authorities. Insecurity, including assault, crime and armed robbery, enables particular groups to enrich themselves at the expense of others. These problems may be a consequence of the widespread availability of weapons, but the failure to tackle them suggests that they are condoned as an inherent tool of the regime. ‘Road cutters’, who extort money from
116
Case Studies: West Africa
people at night, are often members of the security forces – a consequence of the absence of a professionally organized and managed national army. Beyond the ecological and environmental causes that force pastoralists to clash with sedentary farmers in the south, the conflict between these two groups appears to be encouraged by administrative and military leaders who own cattle herds and employ armed pastoralists. Possessing firearms, the latter do not respect transhumance corridors, destroying all harvests in their wake. The complacency with which this conflict has been handled constitutes implicit protection for the cattle owners and their cohorts. The north– south divide has led to double standards in the economic life of the country and is likely to be a source of further conflict. Under Habré’s presidency, Chadians realized that they are not all created equal. Based on the Islamic customary law practice of paying ‘dia’ compensation, the penalty for killing a member of the president’s tribe (Gorane) in a fight or an accident rose to 4 million CFA francs, but only 200,000 CFA francs if the victim belonged to another tribe. This is the price of a human life, since the culprit is not prosecuted separately through the formal courts. Furthermore, the crime is no longer the responsibility of the individual, but a collective one, since members of the culprit’s community are mobilized to pay the compensation. Under Idriss Déby, it is the president’s clan that manages public affairs. Decisions are made at clan meetings. Entire economic sectors are monopolized by the ethnic group in power. Professionalism has been degraded to such an extent that even illiterates, fraudsters and embezzlers have been awarded senior public appointments. As mentioned earlier, insecurity is encouraged even though a priority obligation of a responsible government is to ensure security. The secularity of the state, as enshrined in the constitution, is made meaningless by the proliferation of mosques constructed in traditionally Christian zones, often provoking the indigenous communities to anger.
6
The search for peace
Governments before, during and after the civil war have sought continuously to find solutions to the different conflicts that Chad has experienced. In addition to internal initiatives, many international efforts have taken place. These include the Sebha–Benghazi (Libya) accords agreed by the Supreme Military Council (CSM) and FROLINAT in 1976; the Khartoum accords of September 1977, which led to the establishment of the Northern Armed Forces Command Council (CCFAN) by Hussein Habré and the Chad government; the Kano I accords (March 1979); Kano II (April 1979) and the Lagos accord (August 1979) concluded between the CSM and the twelve politico-military groups. All these accords, mainly brokered by the host countries, had the backing of the Organization of African Unity (OAU), but failed through lack of commitment.
Djimtoingar and Atchénémou Avocksouma: Chad 117
The problem of reaching agreement is exacerbated by the multiplicity of protagonists and stakeholders, each with divergent interests and personal ambitions. The Kano and Lagos conferences took place at a time when the central state did not exist, putting the armed opposition groups in a position of strength. The accords were based on balance-of-power considerations; the losers were resentful, cracks appeared easily and the cycle of violence restarted. Several peace initiatives took place after the re-establishment of state authority under Hussein Habré and since Idriss Déby gained power. These include the 1993 accord between the government and Abbas Koty’s National Recovery Committee (CNR); the 1994 accord between the transitional government and the National Committee for Peace and Democracy (CSNPD); and the 1998 accord between the government and the Movement for Democracy and Development (MDD). However, these accords have generally had tragic outcomes, often with the assassination of the principal opposition leaders involved in the agreement. Most recently, the Tripoli 2002 accords between the Movement for Democracy and Justice in Chad (MDJT) and the government of Chad could be threatened following the disappearance of the MDJT leader Youssouf Togoïmi. Generally, the accords have focused on the distribution of ministerial posts and support to former rebels, such as their absorption into the national army, or demobilization and reintegration. The ineffectiveness of these reconciliation accords as a whole, and in particular those involving the international community, reflects a lack of binding instruments of coercion on the one hand and, on the other, government intransigence and inflexibility, their position being that rebels must accept the conditions they lay down. Notably absent is agreement concerning the distribution of national resources, commitment to good governance and respect for human rights, and any discussion about issues of social stability and cohesion. Many wonderful documents have simply never been implemented. The appointment of a national mediator in 1997, charged with the resolution of major conflicts, was a positive initiative, although he was not given the means to carry out his mission effectively. International effort has also been made to help bring peace to Chad, although this is generally influenced by donor interest, and in certain respects appears in its totality to run contrary to Chad’s interests. Multilateral aid to Chad has been substantial and enabled a kick-start of the economy. The need for reconstruction and the restoration of state authority after 1982 led to the implementation of the 1984–6 emergency plan, which addressed the economic and social needs caused by the civil war. The emergency plan was followed by an interim 1986–8 plan, to rehabilitate infrastructure. In addition, Chad has been implementing a stabilization and adjustment programme under the aegis of the IMF and World Bank since 1987, and a medium-term economic recovery strategy since July 1995, intended to
118
Case Studies: West Africa
establish the basis for durable economic and social development. The programme seeks to achieve a high growth rate, enhanced competitiveness and a reduction in dependence on external assistance. The IMF and World Bank supported this strategy under the aegis of a triennial accord, the Enhanced Structural Adjustment Fund, covering the period 1995–9. Since 2000, the government has been operating, with the support of most donors, a national strategy for poverty reduction, implementation of which is expected to be completed by 2015. However, while at first glance the objectives of these plans are to secure positive benefits for the people of Chad, nevertheless they involve the government in committing itself to reducing its provision of social services and its involvement in the productive sector. Given the chaos of the private sector and the strength of interest groups aiming to secure their own benefit at the expense of others there is a risk that the current policies militate against harmonious economic development.
7
The future
Petroleum, due to flow commercially in 2003, dominates Chad’s future. Its revenue offers hope to the people of Chad, but it is a potential cause of conflict, especially if the benefits do not trickle down to the majority of the population. Initial planning and preparations look promising: structures and measures were enacted to ensure that the revenues benefit the people of Chad. The statute of 11 January 1999 (later modified by law 016/PR/2000 of 18 August 2000) safeguards the management of financial resources and encompasses experience from other countries in setting up mechanisms to avoid over-consumption and misuse of revenue on ill-conceived projects. The statute is based on five economic management principles. The first is that revenue from petroleum is to be allocated to poverty reduction as a matter of priority. The second is to protect the resources in the interests of future generations. Third, there should be a participatory approach to policy formulation and implementation. Fourth, utilization of revenue should be transparent, efficient and accountable. And finally, re-allocation of revenue should be guided by regional balance and equity. The legal framework is significant in that the government has established an explicit link between petroleum revenue and economic development and social objectives. As poverty reduction is the principal objective, the revenues will finance expenditure on education, health, rural development and basic infrastructure. The framework also protects revenue from being used on non-productive activities and sets out important guidelines for monitoring and overseeing the management of petroleum resources. It is planned that 70 per cent of revenue will be allocated to the priority poverty-reduction sectors; 15 per cent for state operations and current state investment, for a period of five years from the initial date of production; and 5 per cent to the producing region’s local government, as stipulated in Article 212 of the Constitution. The balance of 10 per cent is to be banked for the benefit of
Djimtoingar and Atchénémou Avocksouma: Chad 119
future generations. The Petroleum Revenue Control and Surveillance Assembly (CCSRP) has been created to oversee the management of Chad’s petroleum resources. Furthermore, in support of the guiding principles, a process of decentralization will grant legal status to local governmental authorities comprising elected members, which will have their own autonomous resources and manage the benefits of petroleum at a local level. However, while initial planning and implementation by the government appears to indicate its good intentions, there continues to be an undercurrent of concern. This became manifest as a result of the government’s hostile response to criticisms about its management of petroleum profits from the Chad Consortium in 1999, during which it emerged that the government used profits to purchase, among other items, military equipment and weapons (which were not listed in the priority items as determined by the legislature). The disparity between the stated intentions of the government and its actual actions called into question both the integrity of the government and, more generally, its interpretation of the guiding principles. This issue has raised doubts as to whether the checks and balances that have been put in place will be allowed to function effectively. Ultimately, this is an issue of political will: if the Chad authorities genuinely intend to use the petroleum resources for the well-being of all Chadians, then the law suffices. If they do not, then past experience in Chad indicates that there is a high risk that the resources will be redirected towards suppressing opposition and in support of pro-government groups. A programme of good governance is the best way to end hostilities decisively in Chad. Transparent government, a participatory approach to the involvement of civil society, and transparent bilateral and multilateral accords, would all be proof of political goodwill and provide a basis for building democracy. Strengthening the judicial apparatus, ensuring the impartiality of judges, is required to combat corruption and fraud, and to challenge the impunity currently enjoyed by those favoured by the government. Decentralization and opening up the remote regions, enhanced by sub-regional integration, are also measures to fight poverty and enable Chad to manage better the effects of globalization. There is therefore a need for Chad to formulate and implement a participatory and coherent socioeconomic development plan, the absence of which may jeopardize the benefits of petroleum.
8
Conclusion
The civil conflicts in Chad are essentially internal, even if exacerbated by foreign powers. They assume different dimensions: tribal, ethnic, regional and religious. They can be explained partly by economic conditions, aggravated by poverty, and the fact that sound economic plans have never been formulated and implemented efficiently. However, poor governance remains the common denominator. Historically, poor colonial institutional arrangements
120
Case Studies: West Africa
in what was conveniently referred to as the state of Chad brought together different peoples, with diverse social systems and values, without a common basis for establishing a united nation. Following independence, the new leaders, as well as those who followed them, exacerbated differences within Chad. The accession to power of all regimes has always been rushed, lacking any sound economic or political programme, and above all, without any serious preparation for the management of public affairs. It is not surprising, therefore, that the causes of conflict that were discernible in the 1960s are still discernible in the 2000s. Each government, whether from the south (the Tombalbaye and Malloum governments) or from the north (the Habré and Déby governments) bears responsibility for the consequences. The north–south divide has served as a pretext to enable a handful of elites to monopolize power. The issue underlying conflict has never been an issue of harmonious economic development for the different regions of the country. It is, rather, a problem of how to share the available national resources and to escape the cycle of blame for historical wrongs and injustices. Ultimately, Chad looks like a country impoverished by its own elite. The Chad conflict is far from over, given lack of national consensus on development, lack of trust among the citizens, and a deliberate policy of unfair distribution of the fruits of public resources. Recent problems relating to the exploitation of Chad’s petroleum serve to highlight this concern.
Notes 1 In addition to the monetary dimension, this index incorporates other dimensions of poverty, notably life expectancy and education level. 2 According to René Beyem (2000, p. 224), the claim that the sultans of the north raided southern zones to procure slaves was more fictional than real, since from 1860 to 1895 slaves from the south constituted only about 13 per cent of the total number of captives in the Baguirmi Kingdom. The argument advanced focuses on an intra-north phenomenon, at least from 1900. On the other hand, the period considered as the heyday of slavery in southern Chad is between 1900 and 1912, with the intervention on the one hand of the Baguirmiens, and on the other, the Fellatas (Peulhs) from the west of Lake Chad and northern Cameroon. 3 For example, ramifications of this can be seen in 2004 by the influx of Zaghawa and other refugees from neighbouring Darfur.
References Bétoudji, D. A. (2003) ‘Les Causes de la Périnnisation de la Guerre Civil au Tchad’, Le Temps, no. 337, 12–18 March. Beyem, R. (2000) Tchad: l’Ambivalence Culturelle et l’Intégration National (Paris: L’Harmattan). Bouquet, C. (ed.) (1982) Tchad: Genèse d’un Conflit (Paris: L’Harmattan). Cabot, J. (1982) ‘Postface’, in C. Bouquet (ed.) Tchad: Genèse d’un Conflit (Paris: l’Harmattan). Centre Culturel Al-Mouna (1996) Tchad: Conflit Nord-Sud Mythe ou Réalité? (Paris: Editions Sépia).
Djimtoingar and Atchénémou Avocksouma: Chad 121 Charreaux, G. (1998) ‘Le Rôle de la Confiance dans le Système de Gouvernance d’Entreprise’, in Économies et Sociétés, Sciences de Gestion, série S.G. no. 25, no. 8–9/ 1998, pp. 47–65. Collier, P. and A. Hoeffler (1998) ‘On Economic Causes of Civil War’, Oxford Economic Papers, no. 50, pp. 563–73. Collier, P. and A. Hoeffler (1999) ‘Loot-Seeking and Justice-Seeking in Civil War’, World Bank, Washington, DC (mimeo). Collier, P. and A. Hoeffler (2002) ‘On the Incidence of Civil War in Africa’, Journal of Conflict Resolution, vol. 46, no. 1, pp. 13–28. Collier, P., I. Elbadawi and N. Sambanis (2000) ‘How Much War Will We See? Estimating the Likelihood and Amount of War in 161 Countries, 1960–1998’, World Bank, Washington, DC (mimeo). Dadi, A. (1987) Tchad: L’É tat Retrouvé (Paris: L’Harmattan). De Soysa, I. (2002) ‘Paradise is a Bazaar? Greed, Creed and Governance in Civil War, 1989–99’, Journal of Peace Research, vol. 39, no. 4, pp. 395–416. Diguimbaye, G. (1969) L’Essor du Tchad (Paris: Presse Universitaires France). Frémeaux, P. (1973) La Rébellion Tchadienne (Paris: Université de Paris). Fukuyama, F. (1995) Trust: The Social Virtues and Creation of Prosperity (New York: The Free Press). Gali, N. G. (1985) Tchad: Guerre Civile et Désagrégation de l’Etat (Paris: Présence Africaine). Gomez, P. Y. (1996) Le Gouvernement de l’Entreprise (Paris: Interéditions). Goual, N. (1995) 300 Dates de l’Histoire du Tchad (N’Djamena: Centre d’Étude et de Formation pour le Développement). Grellet, G. (1982) Les Structures Economiques de l’Afrique Noire (Paris: Presse Universitaires France). Khayar, H. I. (1976) Le Refus de l’Ecole (Paris: Librairie d’Amérique et d’Orient). Magnant, J. P. (1986) La Terre Sara, Terre Tchadienne (Paris: L’Harmattan). Michalon, T. (1979) ‘L’Echec de la Greffe Jacobine’, Le Monde Diplomatique, April. Ndikumana, L. (1998) ‘Institutional Failure and Ethnic Conflicts in Burundi’, African Studies Review, vol. 41, no. 1, pp. 29–47. Ndikumana, L. (2000) ‘Towards a Solution to Violence in Burundi: A Case for Political and Economic Liberalization’, Journal of Modern African Studies, vol. 38, no. 3, pp. 431–59. Sambanis, N. (2002) ‘A Review of Recent Advances and Future Directions in the Quantitative Literature on Civil War’, Defence and Peace Economics, vol. 13, no. 3, pp. 215–43. Teisserenc, P. (1972) De la Communauté Ethnique à la Communauté Nationale: Les Lycées d’Abéché et de Sarh (Paris: École Pratique des Hautes Études). UNDP (1999) Gouvernance et Gestion des Conflits au Tchad (New York: United Nations Development Programme). UNDP (2000) National Human Development Report for Chad 2000 (N’Djamena: United Nations Development Programme). World Bank (1997) Tchad: Evaluation de la Pauvreté: Les Obstacles au Développement Rural, report no. 16567-cd (Washington, DC: World Bank). Ziegler, J. (1978) Main Basse sur l’Afrique (Paris: Editions du Seuil).
This page intentionally left blank
Part III Case Studies: The Horn, East and Central Africa
This page intentionally left blank
8 Conflict, Post-Conflict and Economic Performance in Ethiopia Alemayehu Geda and Befekadu Degefe
1
Introduction
Ethiopia is a conflict-prone country, characterized by ideological, religious, regional, ethno-linguistic and sociological divisions (Gebre-Hiwot, 1924; Pankhurst, 1963; Alemayehu, 2002a; Alemayehu and Befekadu, 2003). The consequences have been devastating: loss of life and the destruction of infrastructure, and a negative impact on growth and development. At times even the very existence of the Ethiopian polity has been threatened. As this chapter will show, conflict is the major explanatory factor for the country’s poverty. There is a growing literature focusing on the economics of conflict, and its causes, consequences and policies to address related problems. Neoclassicalbased analyses have focused on global data using logit-regression (for example, Collier and Hoeffler, 1998 and 2002; Collier, 1999, 2000a and 2000b; Elbadawi, 1999; Elbadawi and Sambanis, 2000a and 2000b; Reynal-Querol, 2002). A second approach, contrasting with the neoclassicists, focuses on political economy (Gebre-Hiwot, 1924; Nafziger and Auvinen, 1997; Pastor and Boyce, 1997; Stewart, 1998, Cramer, 1999, 2001, 2002 and Addison, 2001), while a third might be described as neoclassical game theory (Grossman, 1991; Noh, 1999; Azam, 2001, among others). Despite differences, each is helpful in explaining conflicts such as those in Ethiopia.1 This chapter draws on different aspects of these approaches in explaining conflict in Ethiopia. Three categories of conflict can be identified in Ethiopia. Type 1 is characterized by competition for positions of power; Type 2 as popular revolt against those in power; and Type 3 as conflict among people over resources. The major analytical basis for such a distinct set of conflicts lies in the characterization of (a) the nature of agents involved in the conflict; (b) the subject of their conflict; and (c) the specific form in which (a) and (b) are combined. In Type 1 conflicts, the agents of conflict are either regional aristocrats (before 1974) or the ‘educated elite’ (after the 1974 revolution). The subject of conflict is power in its generic form (both political and economic) which can be exercised through capturing the state machinery. In Type 2 conflicts the agents in 125
126
Case Studies: The Horn, East and Central Africa
conflict are the elite in power and the masses (usually the underclass). The subject of conflict is usually a violation of basic economic and political rights. In Type 3 conflicts the agents are peasants, usually organized across regions or ethnicity. The subject of conflict here is usually competition over scarce resources, such as grazing land, among the pastoralist communities. This chapter analyses conflicts in Ethiopia with this framework in mind.
2
Background to conflict in Ethiopia
Ethiopia has an estimated population of 67 million. According to the 1994 Population and Housing Census of Ethiopia (CSA, 1999), there are no fewer than eighty ethnic groups. However, two major ethnic groups, the Oromo and Amhara, account for 62 per cent of the population and, together with another four groups (the Tigrawi, Somali, Guragi and Wolaita), make up 81 per cent. More than 60 per cent of the population are Christian (mainly Ethiopian Orthodox) and 33 per cent are Muslim. Amharic is spoken as a first or second language by 42 per cent of the total population and 81 per cent of the urban population. The existence of a large number of ethno-linguistic groups notwithstanding, Ethiopians are considered to be a culturally homogeneous people because of continuous interaction and intermarriage, trade and migration creating unity in diversity, which has helped to resist the pressure from internal disintegration and external aggression (see, for example, Pankhurst, 1966; Hess, 1970; Levine, 1974, 1992). In the long history of the country, neither ethnicity nor religion has been the overriding motive behind conflict. One of the bloodiest periods of conflict in the nineteenth century, called ‘the era of the princes’, is characterized by regional, as opposed to ethnic, civil war. The last religious war was fought between Christians and Muslims in the sixteenth century.2 Three distinct recent periods can be identified during which conflict occurred: the imperial period, prior to the 1974 revolution; the Dergue (an Amharic word equating to a ‘committee’ of soldiers) of 1974–91, during which the civil war took place; and the government current at the time of writing, during which the 1998–2000 Ethiopia–Eritrea war occurred. External intervention has had an important role in the initiation and continuation of conflict in Ethiopia. External intervention can be classified as: historic (religious and colonial); Cold-War related; and conflict with neighbouring countries. The country has had a number of historic conflicts: between 1868–96 there were three with the Egyptians, four involving the Dervishes, five with Italy, and one with the British. These external conflicts were protracted and shaped the Ethiopian state, making it militaristic, and they aggravated and exploited internal conflict. The economic impact of such intervention was devastating and contributed to the backwardness of the country.
Alemayehu and Befekadu: Ethiopia 127
Ethiopia’s strategic position in the Horn of Africa also attracted superpower interest. Pre-1974 support from the USA was reversed following Ethiopia’s socialist revolution. Subsequently, the USA switched its support to Somalia, and the USSR switched from Somalia to Ethiopia. This disrupted the socio-economic organization of society (with a subsequent detrimental impact on economic performance) and accentuated conflict between Ethiopia and Somalia. Similarly, conflict with Eritrea had its roots in Italian and British colonial history but also led to the involvement of China and Middle Eastern countries. Their finance and training of the warring parties prolonged the conflict. The conflict also led to migration, creating a large diaspora who further prolonged hostilities through financing both them and the economy (see Alemayehu and Befekadu, 2003). Although the effect of the re-engagement of the USA in the region following 11 September 2001 is hard to assess, it is hoped that the end of the Cold War has minimized the strategic importance of Ethiopia and risks of further occurrences of this type of conflict.
3
The imperial period
Ethiopia was ruled by a monarchy (although not necessarily hereditary – see Clapham, 1988) for around a millennium. The socio-economic system was essentially ‘feudal’, with two distinct classes, the aristocracy and the peasantry, with the monarchy at the apex. Until the early twentieth century, before the establishment of a salaried bureaucracy and military, the peasantry carried the burden of supporting government functionaries. An official, along with his retinue was assigned a given part of the country from whose residents he collected rent. Accountability was to the chief who gave him this right, and not to the peasants. The official’s sole responsibility was to ensure that his patron remained in power. Chiefs derived their power from military strength; the tenure of officials depended on the will of the patron, so pillaging was rampant, with the grievances arising engendering another cycle of violence. Despite a century of a salaried civil and military bureaucracy, at the time of writing, former traditions continue, resulting in instability and continued grievance. In 1924, B. Gebre-Hiwot identified competition for resources through banditry as a major source of conflict in Ethiopia, the fundamental problem of twentieth-century Ethiopia (Alemayehu, 2002a). He defined banditry as the capacity to exercise violence to ensure a regular flow of income without engaging directly in the production process. The dispersed settlement of peasants limited their capacity to organize and defend themselves and their resources. Gebre-Hiwot also recognized the link between this fundamental constraint and inter- and intra-ethnic/linguistic/cultural group (what he calls neged) conflict. He noted that banditry operates to a degree in each neged. Peasants were squeezed through tax levies and surplus extraction by their own groups’ elites. This process pushed peasants away from their earlier settlements
128
Case Studies: The Horn, East and Central Africa
(usually near fertile land) onto marginal land, leading to conflict with members of other neged (Alemayehu, 2002a). Gebre-Hiwot is clear about the class dimension of this ethnic conflict. He emphasized that the ruling class of each ethnic group waged war using peasants. Control of resources was achieved through power (Gebre-Hiwot, 1924). Each aspirant revolted against the government in power, mobilizing supporters by appealing against perceived oppression by those in power. The new leaders then treated the people no better than had those they had unseated. This cycle of revolt and conflict, seen in the context of Ethiopia’s history, was regional and class-based rather than ethnic. This is supported by the ‘king of kings’ system, where the strongest regional-based king became the king of all regional kings and occupied the central power. This king usually belonged to one regional group but maintained power by recruiting officials from different regions (usually through a system of systematic intermarriage with his own family). Furthermore, all peasants from each ethnic group were subjugated by the ruling elite.3 As noted by Clapham (1988), the root causes of the fall of Haile Selassie and the imperial regime, and hence the birth of the 1974 revolution, are structural rather than personal. Factors included the oppression of the majority of the rural population (who were basically in a condition of serfdom) and the government’s alienation from the urban intelligentsia, military and bureaucracy. The immediate causes were linked to famine in northern Ethiopia (Wallo), which the government attempted to hide; the reaction of taxi drivers following the 1974 oil crisis; and a revised education curriculum, which was opposed by the intelligentsia. The military staged a coup in 1974, declaring a socialist state. The revolutionaries believed that the grievances against the system could only be resolved through its abolition by popular uprising or military coup. The importance of the peasants’ grievances can be inferred from the first proclamation of the military government, which was the nationalization of land and its distribution to the peasants.
4
The Dergue, 1974–91
There were two major conflicts in this period: with Somalia (July 1977–March 1978), and the civil war that began immediately after the coup in 1974 and resulted in the overthrow of the Dergue in 1991. The war with Somalia was related to Somalia’s attempt to create a greater Somalia, one part of which, the ethnically Somali Ogaden, lies within the state of Ethiopia. This was also a period of the Cold War during which the USA and the former USSR were realigning their strategic allies in the Horn of Africa. The USSR had previously supported socialist Somalia and the USA imperial Ethiopia. Following the coup in Ethiopia, the two powers switched their allegiances. After a brief but intense war, the Ethiopian forces repelled the Somalian forces in March 1978.
Alemayehu and Befekadu: Ethiopia 129
4.1 The civil war, 1974–91 During this period the dominant ideology of the Ethiopian elite and educated populace was primarily socialist or left-leaning, so the basis of opposition analysis tended to be either a critique of socialism with a view to addressing the Ethiopian context, or a rejection of socialism in favour of historical grievances such as ethnicity. The opposition included Meison (the All-Ethiopian Socialist Movement), the Ethiopian People’s Revolutionary Party (EPRP), the Eritrean People’s Liberation Front (EPLF), and the Tigray People’s Liberation Front (TPLF). Meison and EPRF were multi-ethnic, with the objective of forming a democratic Ethiopia. However, TPLF and EPLF analysis focused on ethnic dominance or difference (whether perceived or real), resulting in a regional nationalist ideology with the objective of secession. The EPLF had been established following the abolition of a UN-sponsored federation of the Italian colony of Eritrea with Ethiopia by Haile Selassie in 1961. The TPLF was a Marxist guerilla opposition established with the help of the EPLF, and it flirted initially with secession but in the 1980s changed its objective to overthrowing the Dergue and creating a democratic Ethiopia. The years following the 1974 coup were unstable. Power struggles within the Dergue led to the assassination of three consecutive chairmen (General Aman Amdo, Brigadier-General Teferi Banti and Major Atnafu Abate) by Major Mengistu Haile Mariam, who eventually controlled the helm of state power until 1991. The opposition was also fragmented, with different origins, leadership, ethnic balance, objectives and attitudes from the Dergue (see Clapham, 1988). For example, the EPLF and TPLF were active in their home region in the north and followed the strategy of rural insurgency. On the other hand, Meison and EPRP were multi-ethnic and operated nationwide. Meison was prepared to cooperate with the Dergue, with the objective of capturing power. All drew support from young people, especially of university students, both at home and abroad, as well as from trade unions. By 1976, all the major opposition groups, with the exception of Meison, had opted to take up armed struggle against the Dergue. The EPRP launched its urban struggle and the Dergue (together with Meison) responded with the ‘red terror’, in which thousands of the educated elite perished. By 1977, the EPRP had been defeated in the urban areas and began a guerilla struggle in the north. The Dergue’s ‘red terror’ now turned towards Meison, at that time its ally, with the aim of dislodging it from the power structure. This left the TPLF, EPLF and to a lesser extent the EPRP, as opposition, based mainly in the north. The TPLF began to engage with other ethno-linguistic based organizations (including the Amhara Nation Democratic Movement, the Oromo People’s Democratic Organization, and the South Nations and Nationalists Organization) to form what is called the Ethiopian People’s Revolutionary Democratic Front (EPRDF), whose chairman, Meles Zenawi, is, at the time of writing, the
130
Case Studies: The Horn, East and Central Africa
prime minster of Ethiopia. The EPRDF, together with the EPLF, overthrew the Dergue in 1991 and established a transitional government which included other ethno-linguistic organizations such as the secessionist Oromo Liberation Front (OLF). The transitional government reorganized the country as a federation structured along ethno-linguistic lines. It also conducted two elections to form the current federal government, which claimed to win more than 90 per cent of the seats. It is interesting to reflect on current sources of conflict. It is not ideology, since almost all these political organizations were Marxist. Although grievances over cultural/linguistic domination by Amharic speakers might have a grain of truth, the fundamental conflict is principally power among the educated elite. Ethnicity is used as a convenient ideology in the struggle for power. 4.2 The costs of conflict: expenditure on defence and the impact on growth The costs of conflict are multifaceted and complex, ranging from the destruction of human and physical resources to the disruption of economic activity and social disorder. In this section we shall focus on the measurable dimensions of conflict in Ethiopia in recent years.4 Defence expenditure has been a major proportion of government expenditure since the mid-1970s. In the last year of the imperial regime it accounted for 13 per cent of government expenditure (1.9 per cent of GDP). Under the Dergue regime it increased to 17 per cent (3 per cent of GDP). In the period 1977–9, during the Somali invasion, it rose to 26 per cent (6 per cent of GDP). This high share continued through the era of the Dergue until its overthrow in 1991 by the current government (See Figure 8.1). From 1991, defence expenditure at first declined, but then increased again towards the end of the decade. This pattern had a detrimental effect on growth. Although growth variability in Ethiopia often results from drought, Table 8.1 shows that GDP growth declined during years of conflict and recovered during post-conflict years. For example, the table shows that GDP growth averaged 3.8 per cent during 1969–74, and fell to 2.6 per cent during 1975–80 (the civil war period and conflict with Somalia). Between 1980–5 there was a drought, then 1991–2 was the peak period of the 1974–91 civil war and instability associated with the regime shift. Post-conflict, growth improved, with the exception of 1993–4. Looking at sectoral contributions to GDP from pre-conflict to postconflict, one observes that although no sector is immune from the effects of conflict, agriculture is relatively less affected because it has a structure of a large number of small-holders. During conflict, assuming that weather conditions are propitious, its share decreases, but then rises after the cessation of conflict. The industrial sector, especially manufacturing, suffers a
Alemayehu and Befekadu: Ethiopia 131
40 35 30
% of GDP
25 20 15 10 5
/9
/7
98 19
/6
96 19
/4
95 19
/3
93 19
/1
92 19
19
90
/9
/7
88 19
/5
86 19
/3
84 19
/1
82 19
/9
80 19
/7
78 19
/5
76 19
74 19
19
72
/3
0
Year Total govt expenditure.
Figure 8.1 of GDP)
Defence expenditure
Total government and defence expenditure, 1972–2000 (as a percentage
Source: Macroeconomic Data, Ministry of Finance and Economic Development (2004).
greater negative impact, because of a reliance on imported inputs, which during periods of conflict are affected by the government’s increased need for foreign exchange. On the other hand, the service sector has an inflated contribution because of increased government expenditure, and the need to finance the combatants (Alemayehu and Befekadu, 2003).
Table 8.1 Conflict and growth performance in Ethiopia: first conflict context, 1969–97 Growth performance Pre-conflict Year 1969/70 1970/1 1971/2 1972/3 1973/4 Source:
During conflict
Post conflict
Growth (%)
Year
Growth (%)
Year
Growth (%)
2.52 4.53 4.14 3.63 4.03
1975–80 1980–5 1985–90 1990/1 1991/2
2.60 −0.80 5.70 −4.33 −3.69
1992/3 1993/4 1994/5 1995/6 1996/7
6.06 −3.65 3.39 14.68 3.44
Authors’ computations based on Ministry of Finance and Economic Development data.
132
Case Studies: The Horn, East and Central Africa
Increases in defence expenditure had an impact on investment and external debt. The current account deficit doubled from around US$340 million during 1995–8 to about US$700 million in 1998–9. Following the conflict, defence expenditure declined significantly, from an average of 34 per cent of government expenditure in 1989–91 to an average of 9 per cent between 1992 and 1997. An additional cost to the EPRDF government was the demobilization of combatants, numbering about 509,000, which cost US$10 million. It is estimated that 40 per cent of the target group was not reached (Daniel, Dercon and Krishnan, 1999), thus shifting the burden of adjustment onto the rural population. However, demobilization was reversed at the onset of the conflict with Eritrea, between 1998 and 2000; huge mobilization took place, and defence expenditure escalated again. 4.3 The modalities of resolving the civil war Conflicts arising out of economic or resource factors where one group invades another, such as among pastoralists, are usually easier to resolve. They are usually short-lived and are settled easily by the intervention of elders and/or the government. But conflicts that are motivated by political interests are difficult to resolve amicably. For example, the most recent civil war in Ethiopia was resolved only when the Dergue lost to the rebel forces after the failure of international mediation to resolve the issue peacefully. During the civil war, the elites did not suffer opportunity costs in continuing to achieve their ambitions through conflict. Under the Dergue the promised benefits of socialism failed to materialize. The population became frustrated, especially the educated elites, whose exclusion from power by the Dergue aggravated the problem. As noted above, the dominant ideology in Ethiopia during the Dergue period was left-leaning, and opposition to the Dergue implied either refinement of socialism to meet Ethiopia’s needs or an appeal to alternative analyses, such as ethnicity. Of the opposition groups, those that were successful in toppling the Dergue – the EPRDF and EPLF – used ethnicity as an ideology. With the reinstatement of peace, the modality of resolving conflict and political discourse came to be based on ethnicity. This is exemplified by the policy priority of the EPRDF to reorganize the country using a federal structure based on ethno-linguistic states. Ethnicity therefore replaced socialism in informing the economic, political and social arena of the Ethiopian polity. However, this sowed the seeds for further potential conflict, because those elites excluded from power began to oppose the EPRDF government, either for its focus on ethnicity at the expense of undermining national unity and/or its bias towards its own ethnic group. The opposition groups invariably used similar ethnic analysis and ideology as that used by those they opposed. Risks of further conflict increased because, during the civil war, the EPRDF and EPLF had accumulated wealth (both established companies). In peacetime, the state machinery further
Alemayehu and Befekadu: Ethiopia 133
supported their commercial activities (see Tekeste and Tronvoll, 2000), triggering both greed (at the elite level) and grievance (at the ordinary citizen level) which served as a catalyst for the next round of conflict. An intriguing question is why there was a failure to design peaceful power-sharing mechanisms. The historical and social specificity of Ethiopia helps to answer this. A broader answer is greed for power; the negligible (relative to those in power) opportunity costs of pursuing conflict to those excluded from power; the role of external intervention and the fear of reprisals for past wrongs when the incumbent plans (if at all) to transfer power peacefully – most rebels come to power with blood on their hands and it is hard to imagine them leading a peaceful life in their own country once they relinquish that power. This is accentuated by the existence of institutions that are informed by the historically militaristic nature of the Ethiopian state (see Alemayehu and Befekadu, 2003, for details).
5
The Ethiopia–Eritrea war, 1998–20005
Historically, Eritrea was part of Ethiopia but became an Italian colony following the partition of Africa in the late nineteenth century. It remained a colony for about sixty years, composed of about nine ethno-linguistic groups and a population of about 3 million divided equally between Moslems and Christians. Following the defeat of Italy in 1945, it became a British protectorate. In 1952, a UN-sponsored referendum giving a choice of federation with Ethiopia or independence resulted in the former option. However, annexation by Ethiopia in 1960 led to a secession struggle that lasted over three decades. The former province finally seceded in 1993 under the leadership of the EPLF. This was achieved with the help of the government in Addis, current at the time of writing, with whom the EPLF had fought against the Dergue. Given the historic, cultural and political ties between Ethiopia and Eritrea, and in particular the civil war period 1974–91, the Ethiopia–Eritera war can be seen as a continuation of the civil war, since prior to this they acted as a single country with two states allied by ethnic ideology and configuration. 5.1 Characterization of the causes and duration of the conflict Ethiopia and Eritrea signed cooperation agreements following Eritrea’s independence in 1993. Eighty per cent of Ethiopia’s foreign trade passed through Eritrean ports (which are free ports for Ethiopia). Ethiopia was also a market for about 80 per cent of Eritrea’s exports, and the two countries commonly used the Ethiopian currency, the birr, as well as sharing the oil refinery at Assab port. There were obstacles to the implementation of the agreements, illustrated by two conflicting views (Tekeste and Tronvoll, 2000; Socio-Economic Study Team (1992) cited in EEA, 2002). One holds that economic relations between
134
Case Studies: The Horn, East and Central Africa
the two countries were based on mutual benefit. The other view asserts that, whether deliberately or because of negligence by the Ethiopian government, Eritrea took unfair advantage of Ethiopia, or that the political decision-making processes and structure of the Ethiopian government were under the control of Eritrea. The trade agreement allowed the tax-free mutual importation of products from each country, although commodities that were sources of foreign exchange earnings were excluded from the agreement. When imported commodities passed through each other’s country, they were considered to be in transit and free from customs duties. The agreement on transport stipulated that Ethiopian Airlines flew to Asmara. Eritrean nationals residing in Ethiopia were allowed to live and work in Ethiopia with a similar status to Ethiopians and vice versa (Economic Analysis Team (2000) cited in EEA, 2002). Ethiopia complained that the agreement was abused by Eritrea. For example, when a US dollar was exchanged at 6.25 birr in Ethiopia, its value in Eritrea was 7.2 birr, contrary to the trade agreement. Eritrea bought coffee and oil seeds for re-export (EEA, 2002). In addition, the Eritreans allegedly sold untaxed commodities (imported as goods in transit) in Ethiopia, excluding Ethiopian traders from the market. Furthermore, grievance was generated because the Ethiopian government had committed itself to the reconstruction and rehabilitation of the Eritrean economy, before Eritrean independence, signing a US$18.1 million Special Drawing Right (SDR) loan agreement with the International Development Assistance (IDA) group of the World Bank, without passing on the debt to Eritrea (Negarit Gazeta (1993) cited in EEA, 2002). On the other hand, Eritrea argued that Ethiopia was being protectionist and discriminating against Eritreans engaging in the Ethiopian economy (Tekeste and Tronvoll, 2000). There was also a problem coordinating macroeconomic policy. The two countries adopted different development strategies: Eritrea was outwardorientated while Ethiopia was inward-looking. Monetary union in such a context strained macro and fiscal policy coordination, which was apparent just before the onset of the war. When Eritrea issued its currency, the nakfa, in November 1997, the Ethiopian government took the position that economic relations with Eritrea should be identical to those with its other neighbouring countries, using the US dollar as the medium of exchange through letters of credit. This became one of the major causes of the outbreak of the conflict. According to this analysis, the Eritrean military drive to the border might have been motivated by an attempt to threaten the Ethiopian government so as to obtain favourable policy, on the assumption that the Ethiopian government was weak and minority-based, fragmented along ethno-linguistic lines (which was not an unreasonable assumption at that time). An impasse on the border, coupled with the arrogance and stubbornness of the leadership in the two countries,6 led to war.
Alemayehu and Befekadu: Ethiopia 135
However, Eritrea has blamed the northern Ethiopia region of Tigray (where the core of the current ruling elite in Ethiopia, including the prime minister, come from) for blocking competition from Eritrea. Ethiopia’s protectionist stance and its refusal to accept the Eritrean currency with a value on a par with the birr are some of the economic reasons for the conflict (Tekeste and Tronvoll, 2000) 5.2 Cost of the conflict to Ethiopia Destruction of infrastructure The estimated cost of destroyed public and social infrastructure is over US$200 million (World Bank (2000) cited in EEA, 2002). The number of people displaced was about 674,000. The number of people killed is not yet known, but media sources estimate up to 70,000 lives were lost. According to the Maritime and Transit Service Enterprise, the total volume of property in transit to Ethiopia but looted at Eritrean ports was 137,000 tons with a cif value of US$133 million. Since Eritrean consumption habits are similar to those of Ethiopia, commodities that cannot be exported to other countries, such as teff (a staple grain) and berberie (red pepper), used to be exported to Eritrea. Eritrea’s products also used to be imported into Ethiopia at competitive prices. Following the conflict, trade and economic relations with Eritrea were embargoed. This also affected the cross-border petty trade, which used to support the livelihood of many Ethiopians and Eritreans living in the border areas. Development costs Defence expenditure increased from 9 per cent of total expenditure in the 1990s to about 40 per cent at the height of the conflict, thus having a negative effect on social and economic development. For example, capital expenditure on transport construction decreased by 6.1 per cent from the 1997/8 level. Capital expenditure in the industrial sector declined from 93 million birr in 1997/8 to 40 million birr in 1998/9. Using foreign exchange for military expenditure, apart from the budgetary implication noted above, limits the purchase of raw materials for the industrial and service sectors. This increased domestic borrowing, risking a reduction in private-sector investment. According to the Commercial Bank of Ethiopia (CBE) figures in 1998, outstanding loans to 393 Eritreans who had been deported from Ethiopia totalled 376 million birr. Of this amount, 22 per cent had yet to be recouped. Furthermore, Ethiopia had a huge transfer account balance over Eritrean claims (EEA, 2002). Ethiopian Airlines has also been affected. During the conflict, the airline cancelled flights to the north and north-western parts of the country and, for a short period, international flights were in fact based in Nairobi, Kenya. Since the airline could not overfly Eritrean airspace, there were
136
Case Studies: The Horn, East and Central Africa
probably re-routing costs. Flights to Asmara and Assab were cancelled and, because of a decrease in tourist inflow, there was a loss of income from tourism as well. Ethiopian Airlines reported the first-ever annual loss in its long history. Changes in the magnitude and composition of public expenditure Before the conflict, recurrent budget expenditure was around 60 per cent of total expenditure. During the conflict, this share rose to 70 per cent in 1998/9 and 80 per cent in 1999/2000. This is attributed to the increase in defence expenditure, which was about 9 per cent of the total, or 14.6 per cent of recurrent expenditure (4.9 per cent of GDP) in 1996/7, and 40 per cent of the total or 49.8 per cent of recurrent expenditure (13.2 per cent of GDP) in 1999/2000. As defence expenditure increased, the overall government deficit increased nearly ninefold (from 635 million birr in 1996/7 to 5,365 million birr in 1999/2000). Similarly, the import–export gap widened and the foreign exchange reserve was reduced from six months of imports in 1996/7 to two months by the end of 1999/2000. The conflict also reversed the demobilization of Dergue solders. Before the conflict, Ethiopia had an army of about 60,000. At the end of the Ethiopia– Eritrea conflict its size had expanded to about 350,000. Although it was then reduced by about half following the peace agreement (EEA, 2002), the conflict (and absence of normalization with Eritrea) left Ethiopia with a large army and a drain on its resources (the 2003/4 defence budget being about 37 per cent and 16 per cent of current and total planned public spending, respectively). Re-routing of freight via Djibouti At the outset of the conflict, Ethiopia’s import and export activities shifted to Djibouti port. It is estimated that this shift resulted in a 31 per cent increase in transportation costs and, according to a World Bank estimate, a cost to GDP of nearly 1 per cent. There was also a tariff of US$1 per ton at Djibouti port, which also increased the prices of commodities to consumers (World Bank (2000) cited in EEA, 2002). The effects of conflict on growth As can be seen from Table 8.2, it is less clear that conflict had a negative impact on growth when compared with trends during the civil war. Indeed, during the two years of conflict with Eritrea, GDP growth was 6.31 per cent in 1998/9 then reduced to 5.33 per cent. However, there was a shift in the source of this growth. Of the 6.3 per cent growth in 1998/9, 27 per cent came from agriculture, while the contribution of industry amounted to 20 per cent. The major growth impetus came from the service sector, with a contribution to GDP growth of 52 per cent. Government consumption contributed the major share to the service sector, accounting for 35 per cent of the GDP growth for 1998/9.
Alemayehu and Befekadu: Ethiopia 137 Table 8.2 Conflict and growth performance in Ethiopia: second conflict context, 1992–2002 Growth performance Pre-conflict Year 1992/3 1993/4 1994/5 1995/6 1996/7
During conflict
Post conflict
Growth (%)
Year
Growth (%)
Year
Growth (%)
6.06 −3.65 3.39 14.68 3.44
1997/8 1998/9 1999/2000
−1.16 6.31 5.33
2000/1 2001/2
7.20 9.30
Source: Authors’ computation, based on Ministry of Finance and Economic Development data.
Growth declined relatively between 1999 and 2000, which had an impact on poverty reduction. A growth rate of about 5.7 per cent per annum is needed to reduce poverty by half to meet international development objectives (Alemayehu, Shimless and Weeks, 2002). Not only did the growth rate decelerate to 5.4 per cent, but its quality was further eroded as the major part of the growth was the result of inflated government expenditure to finance the war. Increased government consumption accounted for 52 per cent of the growth rate, while the contribution of agriculture and industry plummeted to 15 per cent and 9 per cent, respectively. Considering all the above, and without taking into account human loss and suffering, the total measurable cost of the war sustained by Ethiopia is estimated by the Ethiopian Economic Association (EEA) at 25.4 billion birr (about US$3 billion), excluding the unrecovered loans given by banks to deported Eritreans. This is equivalent to eighteen months of the national budget, or nearly half of the annual GDP. 5.3 Post-conflict issues Reconstruction, infrastructure and resettlement of displaced people The two-year war ended following a peace agreement signed in Algiers in December 2000. The agreement outlined the establishment of a commission in The Hague, to rule on the border issue, and this ruling, based on relevant treaties and applicable international law, is final and binding. Prior to this, diplomatic efforts by the USA and Rwanda, and by the Organization of African Unity, had failed to bring an end to the war. The Algiers peace accord was a consequence of Ethiopia’s successful military offensive deep into Eritrea and its military success in controlling the disputed areas. This explains Eritrea’s acceptance of the ‘OAU Framework’ (which it had previously rejected) and the establishment of the 25-kilometre buffer zone inside Eritrea’s territory.
138
Case Studies: The Horn, East and Central Africa
A UN peacekeeping force currently guards the buffer zone, awaiting implementation of The Hague commission’s ruling. The economic crises brought on by the war, together with drought, encouraged both countries to come to the negotiating table. The result has been a state of neither war nor peace, with both Ethiopia and Eritrea rejecting the normalization of relations. It is interesting to observe the role of external forces as agencies of restraint in bringing the warring parties towards a peace settlement. The arms embargo imposed on both countries (despite smuggling from the former USSR and Eastern Europe), and most importantly the freezing of aid to both countries, were of critical importance. Sustainable peace, however, requires going beyond such aid-related coercive measures and addressing the root causes of the problem. In this endeavour, amicable solutions to the economic co-existence problem are central for durable peace. Furthermore, examining the historical, political and sociological as well as the regional dimension of the conflict helps in resolving it and achieving durable peace. In each country, the first step is to address the damages caused by the war.7 Following the end of the conflict, Ethiopia began the reconstruction of its damaged and partly destroyed social and physical infrastructure; the resettlement of IDPs and the deportees from Eritrea; and the rehabilitation of disabled people. The World Bank estimates that this emergency recovery programme will cost US$555 million (World Bank (2000) cited in EEA, 2002).
6
Conclusions
Conflicts and their violent resolution seem to be an enduring characteristic of the Ethiopian polity. Recently, periods of peace have been short-lived; new regimes have failed to learn the lessons of their predecessors. In Ethiopia, power is absolute, despite the reforms of each incoming administration.8 The problem is that the laws are not applied, and it is the whim of the leaders that rules the country. This is the substance from which recurring instabilities are made. The only way the country can end the vicious circle of conflict and instability is for the leaders to become accountable to their people. External intervention has also had repercussions, either as a restraint or a catalyst of conflict. Which way the external forces react depends on their long-term national interests. During the Cold War, US and Russian interests were a catalyst of the Ethiopia–Somalia conflict. On the other hand, to some degree, the international community was a constraint on the Ethiopia–Eritrea conflict. Post-conflict years experience high rates of economic revival. Economic growth normally picks up and life becomes more orderly. Such is the case in Ethiopia under the current government. Between 1992 and 1997, GDP growth averaged 6 per cent, the longest and the highest rate in the recent past.
Alemayehu and Befekadu: Ethiopia 139
Years of conflict show either low or negative growth rates, or where they are high they are of the wrong type and incapable of reducing poverty. This chapter has shown that recent research on conflict sheds important light on the nature and driving force of conflict in Ethiopia. An attempt to characterize conflicts and their economic ramifications has also been made. Analysis of conflict, post-conflict and economic conditions in Ethiopia stressed the class nature of major hostilities and the significance of the political economy approach to offer adequate explanation. The major cause of conflict in Ethiopia is grounded in the political economy of the country, reinforced by competition for power. History, institutions, the path-dependent nature of state formation and external intervention are important. Ethiopian governments have never been accountable to the people. This created a fertile ground for insurrection and offered opportunities for people to express their grievances or greed only through violence, often at negligible opportunity cost to the opposition but relative high cost to the incumbents. Inevitably, the incumbents’ attempt to defend and maintain their position led to increased militarism: just as good institutions enhance growth, so bad institutions are detrimental to long-term development and durable peace. This seems to be the case in Ethiopia. Considering the three recent regimes – the imperial, the Dergue and the EPRDF – one discerns a certain pattern: the means of resolving one conflict sowed the seeds of the next. The downfall of the imperial regime can best be explained by the misery it caused to the majority of the rural population, who were essentially serfs. Resolution was through popular uprising and a military coup, replacing the existing system with a socialist system. This grievance was expressed in the first important proclamation of the Dergue – nationalization of land and its redistribution to the peasants. The failure of the Dergue to meet the expectations of the population, and the educated elite, led to grievances which fuelled the next round of conflict, in 1974–91, the basis of which became ethnicity. And the resolution of the civil war on an ethnic basis precipitated its next phase, the Ethiopia–Eritrea war. Conflict is the major and fundamental explanatory factor for the poverty and backwardness of Ethiopia. It not only creates human misery but also suppresses economic activity. Resources shift from productive to destructive activities, reducing growth or having a negative effect on its quality and composition. Economic revival intensifies following conflict. Such is the case under the current government. Between 1992 and 1997, annual GDP growth averaged a healthy 6 per cent, the longest period of growth at such a high rate. As noted above, during periods of conflict Ethiopia generally experienced either low or negative growth rates or, where they were high, an inherently unstable growth which exacerbated rather than reduced poverty. However, in Ethiopia, power remains absolute. Despite institutional and legal reforms implemented by incoming administrations, laws are not applied equitably, transparently or accountably but according to the whims of the
140
Case Studies: The Horn, East and Central Africa
leaders who rule the country. This is the substance of recurring instability. Peace and stability depend on the extent to which political leaders are made accountable to the people. In essence, the only way that the country can end the vicious cycle of conflict and instability is to make the leaders accountable to their people. Will there be another conflict in Ethiopia? Unless and until there is real political reform bringing government accountability to the people, further conflict is probable. Notes 1 See Alemayehu and Befekadu (2003) for a detailed theoretical and historical discussion of the conflict literature and its implication for the Ethiopian context. 2 The facts correspond neither to the Collier and Hoeffler (2002) quantitative definition of ethnic dominance or the Reynal-Querol (2002) concern of religious fractionalization, both of which are associated with high risk of civil conflict. 3 This argument in no way denies the historic domination of the northern highlanders’ language and culture over the others, but takes it as secondary (compared to the power-grabbing motive). 4 See Addison, Chowdhury and Murshed (2002) and Addison and Alemayehu (2003) regarding the impact of conflict on financial development. Issues of financial reconstruction are also discussed in the studies. 5 Information used in this section is largely based on EEA (2002). 6 Asked about the cause of the war, one prominent Ethiopian politician answered: ‘The war is the result of the psychological problem of the highland Eritrean and the Tigrian elite.’ Given the history and sociology of that part of the country, this observation is important in partly explaining the causes of the conflict. 7 At the time of writing, Ethiopia has rejected the border ruling by the Hague-based commission and asked the UN Security Council to intervene. 8 For example, the current government promised devolution of power to the regions, yet the centre still retains control of about 80 per cent of public revenue and expenditure. There have been democratic elections, but their results have been questioned. A clear distinction between the mandates of the party and civil servants (technocrats) has been made, yet party cadres are primarily in control. A free and competitive economic space has been promised to the private sector, yet a focus on party-owned businesses is the rule, not the exception.
References Addison, T. (2001) ‘Reconstruction from War in Africa: Communities, Entrepreneurs, and States’, WIDER Discussion Paper 2001/18 (Helsinki: WIDER). Addison, T. and G. Alemayehu (2003) ‘Ethiopia’s New Financial Sector and Its Regulation’, in Tony Addison (ed.), From Conflict to Recovery in Africa (Oxford University Press). Addison, T., A. R. Chowdhury and S. M. Murshed (2002) ‘By How Much Does Conflict Reduce Financial Development?’, WIDER Discussion Paper 2002/48 (Helsinki: WIDER). Addison, T., G. Alemayehu, P. Le Billon and S. M. Murshed (2001) ‘Financial Reconstruction in Conflict and Post-conflict Socities’, WIDER Discussion Paper 2001/90 (Helsinki: WIDER). Alemayehu, G. (2002a) ‘The Gebre-Hiwot Model: A Pioneer African Development Economist’, Ethiopian Journal of Development Research, vol. 24, no. 1, pp. 1–28.
Alemayehu and Befekadu: Ethiopia 141 Alemayehu, G. (2002b) Finance and Trade in Africa: Macroeconomic Response in the World Economy Context (London: Palgrave). Alemayehu, G. and D. Befekadu (2003) ‘Conflict and Economic Performance in Ethiopia’, Proceedings of the 1st International Conference on Ethiopian Economy (Addis Ababa: EEA/Addis Ababa University). Alemayehu, G., A. Shimless and J. Weeks (2002) ‘The Pattern of Growth, Poverty and Inequality in Ethiopia: Which Way for a Pro-poor Growth?’, Research report for the Ministry of Finance and Economic Development (Addis Ababa: MoFaED). Azam, J.-P. (2001) ‘Looting and Conflict between Ethno-Regional Groups: Lessons for State Formation in Africa’, Institut Universitaire de France, Toulouse (mimeo). Clapham, C. (1988) Transformation and Continuity in Revolutionary Ethiopia (Cambridge: Cambridge University Press). Collier, P. (1999) ‘On the Economic Consequence of Civil War’, Oxford Economic Papers, vol. 51, no. 1, pp. 168–83. Collier, P. (2000a) ‘Policy for Post-conflict Societies: Reducing the Risk of Renewed Conflict’, paper presented to symposium on ‘The Economics of Political Violence’, Princeton University, Princeton, NJ. Collier, P. (2000b) ‘Economic Causes of Civil Conflict and their Implication for Policy’, World Bank Research paper (Washington, DC: World Bank). Collier, P. and A. Hoeffler (1998) ‘On the Economic Causes of Civil War’, Oxford Economic Papers, vol. 50, no. 4, pp. 563–73. Collier, P. and A. Hoeffler (2002) ‘Greed and Grievance in Civil War’, World Bank Research paper (Washington, DC: World Bank). Cramer, C. (1999) ‘The Economicus and Political Economy of Conflict in Sub-Saharan Africa’, CDPR Discussion Paper 1399 (London: Centre for Development, Policy and Research). Cramer, C. (2001) ‘Economic Inequality and Civil Conflict’, CDPR Discussion Paper 1501 (London: Centre for Development, Policy and Research). Cramer, C. (2002) ‘Homo Economicus Goes to War: Methodological Individualism, Rational Choice and the Political Economy of War’, World Development, vol. 30, no. 11, pp. 1845–62. CSA (1999) The 1999 Population and Housing Census of Ethiopia: Results at Country Level, Vol. II, Analytical Report (Addis Ababa: Central Statistical Authority). Daniel, A., S. Dercon and P. Krishnan (1999) ‘Demobilization, Land and Household Livelihoods: Lessons from Ethiopia’, WIDER Research in Progress, No. 24 (Helsinki: WIDER). EEA (Ethiopian Economic Association) (2002) Second Annual Report on the Ethiopian Economy (Addis Ababa: EEA). Elbadawi, I. (1999) ‘Civil War and Poverty: The Role of External Intervention, Political Rights and Economic Growth’, paper presented to World Bank conference ‘Civil Conflicts, Crime and Violence’, Washington, DC. Elbadawi, I. and N. Sambanis (2000a) ‘External Interventions and the Duration of Civil Wars’, paper presented to World Bank conference ‘The Economics and Politics of Civil Conflicts’, Princeton University, Princeton, NJ. Elbadawi, I. and N. Sambanis (2000b) ‘Why Are There So Many Civil Wars in Africa? Understanding and Preventing Violent Conflict’, Journal of African Economies, vol. 9, no. 3, pp. 244–69. Gebre-Hiwot, B. (1924) The State and Public Administration (in Amharic) (Addis Ababa: Teferi Mekonnen).
142
Case Studies: The Horn, East and Central Africa
Gebre-Hiwot, B. (1995) The State and Economy in Early 20th Century Ethiopia, trans. by T. Bonger (London: Karnak House). Grossman, H. I. (1991) ‘A General Equilibrium Model of Insurrections’, American Economic Review, vol. 81, no. 4, pp. 912–21. Hess, R. (1970) Ethiopia: The Modernization of Autocracy (New York: Cornell University Press). Levine, D. (1974) Greater Ethiopia: The Evolution of Multi-Ethnic Society (Chicago: University of Chicago Press). Levine, D. (1992) ‘Greater Ethiopia Reconsidered’, Ethiopia Review, vol. 2, no. 8, pp. 14–16. Nafziger, E. W. and J. Auvinen (1997) ‘War, Hunger, and Displacement: an Econometric Investigation into the Sources of Humanitarian Emergencies’, WIDER Working Paper No. 142 (Helsinki: WIDER). Noh, S. J. (1999) ‘A General Equilibrium Model of Two-group Conflict with Endogenous Intra-group Sharing Rules’, Public Choice, vol. 98, no. 3–4, pp. 251–67. Pankhurst, R. (1963) ‘The Ethiopian Army of Former Times’, Ethiopian Observer, vol. 2, no. 2, pp. 118–42. Pankhurst, R. (1966) Economic History of Ethiopia (Addis Ababa: Haile Selassie University Press). Pastor, M. and J. K. Boyce (1997) ‘The Political Economy of Complex Humanitarian Emergencies: Lessons from El Salvador’, WIDER Working Paper no. 131 (Helsinki: WIDER). Reynal-Querol, M. (2002) ‘Ethnicity, Political Systems and Civil Wars’, Journal of Conflict Resolution, vol. 46, no. 7, pp. 29–54. Stewart, F. (1998) ‘The Root Causes of Conflict: Evidence and Policy Implication’, paper prepared for WIDER Conference on ‘War, Hunger and Displacement: The Economics and Politics of the Prevention of Humanitarian Emergencies’, Stockholm. Tekeste, N. and K. Tronvoll (2000) Brothers at War: Making Sense of the Eritrean–Ethiopian War (Oxford: James Currey).
9 Prospects for Sustainable Peace and Post-Conflict Economic Growth in the Sudan Ali Abdel Gadir Ali and Ibrahim A. Elbadawi
1
Introduction
The Sudan is defined by conflict, having suffered more than thirty-seven years of civil war since independence in 1956, interrupted only by an uneasy peace from 1972 to 1983. Historical, social and geographical factors have all contributed to the nature and duration of the conflict. Socially diverse, Sudan is characterized by the religious and cultural dominance of the Arab and Arabized Muslim majority from the north of the country. Ali, Elbadawi and El-Batahani (2002) argue that this divide was further polarized by colonial policy insulating the south from the north, pending its proposed integration into East Africa. This policy was reversed less than ten years before independence, and at the time of writing, remains marginalized from the political, economic and social developments of the north. A meaningful contribution to the resolution of this conflict must address three fundamental questions. First, what are the major determinants of the occurrence and duration of the civil war in Sudan? Second, given an understanding of the causes of the civil war, how has peace been brokered in the past, and what is necessary for successful peace initiatives in the future? Finally, how does the developing oil industry impact on the war and the requirements for peace? Social polarization 1 and external intervention have been the defining factors of Sudan’s civil war. Social polarization allows for sustained coalitions between fighting forces, while external intervention in support of the weaker party prevents one side from overcoming the other. These factors set the experience in Sudan apart from that of many other countries, even those that share Sudan’s other high-risk factors associated with civil war, such as low levels of economic development and lack of democracy. Peace, whether in a unified or partitioned Sudan, must address the deeprooted social polarization that has previously marginalized relevant groups. 143
144
26
28
30
32
34
36
38
UD SA
The boundaries and names shown and the designations used on this map do not imply official endorsement or acceptance by the United Nations.
N Jeb et
Oo zA b
tA be
Je
Milk
KHARTOUM
WARAB
Wau
Fathai
ite
Town
0
100
200
24
Map 9.1 Source:
26
28
4
Wodi Od ih
UGANDA
V i c t o ri a
L. Albert 30
6
Ch’ew Bahir
Lalagipi Swamp
le
300 Km 200 ml
100
Kabowen Swamp
o
Administrative boundary
L. Turkana (L. Rudolf)
e Nil
0 22
Om
Kapoeta Juba BAHR AL Torit Nagishot JEBEL
Airport
2
Ukwaa Towot
Ni
Regional capital
Yei
A lb ert
DEMOCRATIC REPUBLIC OF THE CONGO
National capital
8
EASTERN EQUATORIA
e
Track Railroad
Adis Abeba (Addis Ababa)
Kenamuke Swamp
Nil
Amadi
EQUATORIA Yambio Maridi
Road
10
ETHIOPIA
Akobo
r
JONGLI d LAKES d Bor
Li Yubu
State boundary (approximate)
ay
Kigille
Wh
International boundary
4
i
Rumbek
SUDAN 6
Ab
N IL E ba
Pib o
AL GHAZAL
h
12
u
Po ng
o
Bentiu
S
BAHR
J ur
Bahr et Z Jande i Ca a a f nal
L o l NORTHERN
WESTERN Aweil BAHR Raga AL GHAZAL
8
UNITY
T’ana Hayk’ Aba y (B lu
Renk Ed Damazin BLUE Famaka NILE
Paloich UPPER Malakal So
CENTRAL AFRICAN REPUBLIC
Kafia Kingi
ite
W
Abyei
Gonder
e Nile)
e t ' A rab
SINNAR
r
Talodi
Ba
10
D
WHITE NILE
de
Kadugli
Sinnar
Kosti
Nuba Mts. Kologi
Muglad
14
Te kez
GEDAREF
WESTERN Abu Zabad KORDOFAN SOUTHERN Al Fula Ed Da'ein KORDOFAN
Tullus Buram SOUTHERN DARFUR hr Radom
Asmara
G EZ IR A Wad Medani
in
Nya l a
Kassala
Gedarei
En Nahud
12
E R I T R E A16
KASSALA Halfa al Gadida
Ha
El Obeid
WESTERN DARFUR
Meroe
Shendi
Omdurman
ar
18
Karora Gadamai
N ile
D
Haiya Tokar
ra
Sodiri
Al Fasher
Suakin
Atbara
D
u
KO R D O R A N m i d
Geneina
20
Port Sudan
RIVER
Ed Damer wuNILE
et
Umm Badr 14
l
e
te a u dP la
by a
Ni
Khartoum NORTHERN
Miski
Muhammad Qol
RED SEA
Merowe
sh hu as
di Wa
Salala
ba
W
ag
n
Abu Hamed
Old Dongola
Abu 'Uruq
NORTHERN DARFUR
a
At
16
r owa iH ad
i
D e s e r t
Karima
IA
t s e r D e
El’Atrun
CHAD
b
Kerma Dongola
AB
u
Laquya Arba' in
Nukheila
18
A
N
24
R IA
Kumma
Lake Nubia
SE
Halaib Wadi Halfa
Semna West
Selima Oasis
N O R T H E R N 20
D
y a n L i b
Administrative boundary
E
EG YPT
SUDAN
R
Lake Nasser
32
L. Kyoga
L. Salisbury
34
Sudan
http://www.un.org/Depts/Cartographic/map/profile/sudan.pdf.
KENYA 36
38
2
Ali and Elbadawi: Sudan 145
Specifically, the principles of neutrality, inclusion and proportionality must be integrated into the peace process, alongside a transition to democracy.
2
Background to the conflict
Violence erupted in 1955 during the closing days of the Anglo-Egyptian colonial administration (1898–1956). Southern ranks garrisoned at Torit in the southerly Equatoria Province mutinied and were joined by civilians, police and prison guards. For two weeks, northern Sudanese civilians, including women and children, were massacred. Government punishment was brutal, and many mutineers fled into the bush or to neighbouring countries. This massacre, and what was perceived to be a miscarriage of justice by the northern-dominated administration following the event, would set the stage for a civil war that has devastated the country for all but eleven years (1972–83), of its independence since January 1956. The colonial administration’s policy on southern Sudan (SP) was particularly important in shaping the initial conditions for the conflict. The policy contained an ensemble of implicit and explicit measures insulating southern Sudan from the north, pending its proposed integration into East Africa on independence. Education and development policies augmented the fundamental objective of keeping the two entities apart while both were under a single colonial administration, to facilitate eventual partitioning. These initial conditions, we would argue, made Sudan ripe for a civil war. However, just eight years before independence, the SP was reversed in a dramatic decision by the colonial administration to maintain a united Sudan. Unfortunately, this reversal came after the SP had been in effect, implicitly or explicitly for more than forty years, during which time the two entities had already diverged. Culturally, politically and economically, the south was already substantially marginalized in relation to the north. Nevertheless, the new SP was institutionalized in two major events which reverberated throughout the Sudanese conflict. In 1947, a formal conference to discuss the future of Sudan, attended by representatives of both north and south, was held in the southern capital of Juba under the auspices of the colonial government. Its outcome was then ratified as the ‘Independence Motion and Resolution’ during Sudan’s first Parliament in 1955.
3
Why has the civil war lasted so long?
Why has the civil war in Sudan been the longest war in Africa, encompassing almost the entire history of post-colonial Sudan? One possible explanation looks at the determinants of the risk of war in terms of the likelihood of prevalence of war at any given period. Elbadawi and Sambanis (2001) estimate the probability of a civil war occurring based on the probabilities of war duration and war initiation. This model predicts that increased economic
146
Case Studies: The Horn, East and Central Africa
External intervention (% of years with external interventions)
development and democracy lower the probability of war, whereas ethnic polarization2 rather than ethnic fractionalization raises it. The prediction of this model is strongly corroborated by the data. The situation in the Sudan, however, has an additional significantly important determinant – external interventions by third parties. Such interventions are associated closely with the fundamental cultural–religious divide that characterizes the conflict (Ali, Elbadawi and El-Batahani, 2002). Common in civil wars, this factor has been found to be an important determinant of civil war duration, since the weaker party is often supported by outside interventions (Elbadawi and Sambanis, 2000; Regan, 2000). According to a global index on the extent of external intervention in civil wars by third parties (developed by Regan, 2000), Sudan has experienced high levels of external intervention, particularly during the second insurgency (1983–present), consistent with the apparent regionalization and internationalization of the conflict (see Figure 9.1). Evidence suggests that external intervention in favour of the southern rebels has been critical for both the onset and the sustainability of the war. External support to the southern rebels by some of Sudan’s neighbours has been motivated partly by sympathy
120
100
80 Sudan Mean SSA
60
40
20
0 1955–72 Sudan: 1st civil war
1973–82 Sudan: peace period
1983–present Sudan: current civil war
Figure 9.1 External interventions in Sudan and sub-Saharan African civil wars by regional and global third parties Note: A country in a civil war is coded as experiencing intervention only if the extent of intervention surpasses a certain threshold. Source: Based on Regan (2000) data on external intervention, including economic, military and mixed interventions.
Ali and Elbadawi: Sudan 147
for their cause and partly because of the wider tendency of neighbouring countries in the region to involve themselves in each other’s civil wars (see Alier, 1990). In addition, the rebels have also received support from Christian missionaries and non-neighbouring countries, motivated mainly by religious or strategic considerations. On the other hand, Arab and Islamic countries have provided support to various Sudanese governments.3 We have therefore extended the Elbadawi and Sambanis (2000) model to account for external invention. The analysis now assumes that while the northern and southern parts of the country are very diverse, the dominant factor affecting the prevalence of war is the social polarization between the two regions. Including this aspect of the Sudanese conflict greatly enhances the predictive power of the model, as shown in Figure 9.2. The simulation is based on an extended regression of probability of war prevalence (Elbadawi and Sambanis, 2001), with ‘anticipated’ external intervention added, generated for three periods using average data for Sudan during the periods and for two scenarios. The regression is given by: Pw = 13.05 isxp − 26.40 isxp2 − 0.03 poll 1 − 0.38 rgdplag + 0.03 elf − 0.0003 elf 2 + 0.57 log pop + 0.74 expint − 0.0086 elf . expint − 9.49
(1)
where isxp is primary exports to GDP ratio; poll 1 is the polity index lagged by one period; elf is the degree of ethno-linguistic fractionalization, given by 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Short-lived peace (1972–83)
First civil war (1955–72)
Uniform diversity (elf = 73%)
Figure 9.2
Polarization between north and south (elf = 38%)
Risk of conflict in Sudan (on a scale of 0 to 0.8)
Source: Elbadawi and Sambanis (2001).
Second civil war (1983–2002 ongoing)
148
Case Studies: The Horn, East and Central Africa
the probability that two people selected at random in a given society would come from two different ethnic or linguistic groups; logpop is the log of population size; expint is expected intervention and elf.expint is the interaction between intervention and elf. Anticipated external intervention is derived from the predicted models below (Elbadawi and Sambanis, 2000): P (extintervention)= −4.8 e−0.7 death and displacement −1.67 ethnicwar −0.001 size of military (−1) −0.48 average regional democracy (−1) −0.16 regional dummy + 1.51 war (−1) + 4.3. The uniform diversity scenario treats ethno-linguistic diversity uniformly across the country and hence it does not account for the polarization between north and south Sudan. Hence the standard elf index (available in global databases) is used, which assigns an elf of 73. The polarization scenario obtains when we transpose religious fractionalization into cultural and linguistic diversity. This scenario assumes that, despite the overall diversity of the Sudan, the indicator relevant to the war in the Sudan is the one that recognizes the fundamental divide between the two parts of the country. According to 1983 population data, southern Sudanese and other small minorities in the north account for about 25 per cent of the total population of Sudan. The elf index, according to this polarization scenario, is given by: [1 − (0.25)2 − (0.75)2] × 100% = 38%
(2)
The fundamental divide between north and south, which manifests itself in terms of social polarization, is therefore significant in explaining the Sudanese experience. To understand better the contribution of ethnic polarization, external intervention, and other economic and political factors to the civil strife in Sudan, we compare its experience with that of Egypt (see Figure 9.3) and Tanzania (see Figure 9.4). Neither of these countries has experienced a civil war, and their predicted probabilities for conflict are correspondingly lower than that of the Sudan. As is seen clearly in Figure 9.3, the predicted probability of conflict in Sudan is higher than that for Egypt by 0.55 for the first Sudanese civil war, 0.41 for the second civil war, and 0.29 during the peaceful period in Sudan; and for Tanzania the probability differentials are 0.49, 0.48 and 0.35. Sudan is comparable to many countries with regard to other factors associated with a risk of conflict, such as a lack of democracy and low levels of economic development. While Sudan scores badly in both areas, its situation is not significantly different from that of Tanzania and median sub-Saharan Africa. However, since the civil war has largely been confined to the south, economic factors are likely to be more significant than the model suggests. Dr John Garang de Mabior, the leader of the main post-1983 rebel movement, argued that, in the south it paid to rebel. The factors in which Sudan does stand out are, as expected, social polarization and external intervention. Social polarization sets Sudan apart
149
1 First civil war
Current civil war
0.8 0.6 0.4 0.2 0 –0.2 –0.4 Pw(Sudan)–Pw(Egypt) Share of ethnic Share of economics Share of population (1) polarization & political factors size & external intervention (3) (4) (2)
Figure 9.3 Risk of civil wars in polarized and homogeneous societies: Sudan compared to Egypt (on a scale of −0.4 to 1) Source: See Figure 9.4.
1.2 First civil War
1
Current Civil War
0.8 0.6 0.4 0.2 0 -0.2 -0.4 -0.6 -0.8 Pw(Sudan)–Pw(Tanzania) (1)
Share of ethnic polarization & externalintervention (2)
Share of economic & political factors (3)
Share of population size (4)
Figure 9.4 Risk of civil wars in polarized and diverse societies: Sudan compared with Tanzania (on a scale of −0.8 to 1.2) Notes: The simulations are based on the same model as in Figure 9.1, where period averages are used and the polarization scenario is assumed for the Sudan (elf = 38 per cent); first civil war relates to 1955–72 and current civil war to 1983–2002. Source: Ali, Elbadawi and El-Batahani (2002), table 17.
150
Case Studies: The Horn, East and Central Africa
from both diverse Tanzania and homogenous Egypt. Political violence is likely to be expressed differently in homogenous societies (for example, through coups and popular uprisings), while in very diverse societies, like that of Tanzania, civil wars are likely to be short if they occur, since rebel movements have difficulty maintaining cohesion (Collier, Hoeffler and Söderbom, 2001). External intervention is the most significant factor in explaining the higher risk of civil war prevalence. In Sudan it has been substantial, as supported by this analysis and by anecdotal evidence. The more relevant question is why Sudan attracts so much external intervention. A fundamental divide between groups with international associations, rather than localized ethnic conflict, attracts external intervention. It is also linked to regional conflicts. Sudan, for example, supported the Eritrean rebels against Ethiopia, while the latter lent support to the southern Sudanese rebels during the first civil war. Interventions have important implications for the dynamics of war and peace, and for democratic transformation and accountability. In all likelihood, in Sudan they have reduced the opportunity cost of peace, by reducing the cost of warfare for the rebels and the government; and undermined accountability to local constituencies, thus prolonging the war and complicating transformation towards more accountable and democratic governance. This should, of course, be distinguished from the role of ‘neutral’ multinational external interventions promoting peace, such as the current efforts by the Inter-Governmental Agency for Development (IGAD). This understanding allows for the design of appropriate strategies for conflict resolution and peace-building, and points to the necessity of democratic transformation in both north and south for a sustainable peace in Sudan.
4
The making and collapse of peace, 1972–83
After several serious attempts at conflict resolution, peace was achieved in 1972 with the Addis Ababa peace agreement signed by the main insurgents (Anya-Nya) and the military government of General Nimeiri. Mediated by the World Council of Churches and the Ethiopian Emperor Haile Selassie, the Addis Ababa peace agreement brought peace to the Sudan for more than eleven years (1972–83), allowing for partial southern autonomy with a regional parliament and regional government. These democratic institutions were confined to the south, but were embedded in the autocratic regime of General Nimeiri. We shall argue that this precarious system led to the re-emergence in 1983 of the civil war which continues at the time of writing. 4.1 Economic mismanagement: a missed opportunity Peace benefited Sudan and the Nimeiri regime. For most of the eleven peaceful years, Sudan grew by almost 5 per cent a year, and substantial investment entered Sudan, including some to the south to finance post-conflict reconstruction. However, as Ali and Elbadawi (2002) show, growth was not
Ali and Elbadawi: Sudan 151
sufficient to alter economic conditions radically, especially in the south. The results are summarized in Table 9.1, reporting average growth rate calculations for sub-periods corresponding to the various stages of the civil war. While real per capita GDP increased during the peace period, from US$705 in 1973 to US$886 in 1983, the pattern of growth during the whole post-conflict period continued to show relatively high volatility, with a coefficient of variation of 6.71, only slightly lower than that of the conflict period (7.91). Thus growth was not stable, especially in the southern part of the country. Over the period 1960–83, the highest GDP per capita of US$992 is recorded for 1977, followed by US$962 for 1978. Thus, judging the early 1980s by the ‘living conditions’ of 1977 and 1978 as reflected by real per capita GDP, there was indeed some deterioration. Given the initial development gap between the two geographical regions of the country, the deterioration of ‘living conditions’ could have been magnified in the south. The economic opportunity provided by the peace dividend was squandered, and growth gave way to deep economic crisis requiring economic adjustment under the auspices of the World Bank and the IMF, which, if anything, further deepened the crisis. This economic mismanagement, linked to the autocratic nature of the government, contributed to the collapse of peace and recurrence of the civil war. Furthermore, prior to the renewal of conflict, the south remained marginalized, both in terms of income and access to social services. Of 5,912 public primary schools in the country in 1980–2, only 10 per cent were in the south; and shares of the south in public intermediate and secondary schools were even lower. The south had just 20 per cent of the nation’s hospitals, and only 2 per cent of the health clinics. Finally, by 1984, out of 181 bank branches nationwide, only 15 were located in the south. The opportunity Table 9.1 Years
GDP per capita growth rates during conflict and post-conflict periods Status of conflict
1961–72 Low intensity war 1973–83 Peace on Addis Ababa Accords 1973–7 Sub-period: first five years of peace 1978–83 Sub-period: last six years of peace Source:
Average Standard deviation Coefficient Number of years annual (percentage points) of variation with negative percentage growth growth rate −0.53
4.19
7.91
8 out of 12
1.25
8.39
6.71
3 out of 11
4.82
10.66
2.21
1 out of 5
−1.73
5.15
2.98
2 out of 6
Authors’ calculations based on GDN database.4
152
Case Studies: The Horn, East and Central Africa
cost of recurrence of civil war remained very low in the south, despite more than a decade of peace. 4.2 Political miscalculations: another missed opportunity The unpredictable nature of autocracy contributed in two significant ways to the collapse of peace, and to the eventual demise of the regime itself. First, the regime used the achievement of peace with the south as a substitute for democracy throughout the entire country. The failure to submit to popular demands for a return to constitutional democracy increased political tensions between the regime and popular forces in the north, undermining the political ‘peace dividend’. Sudan’s regional relations were similarly precarious, as the regime was associated with the Egypt–USA axis, in opposition to the Sovietbacked Aden alliance that included South Yemen, Ethiopia and Libya. Regional tensions deepened domestic ones; Ethiopia, for example, provided considerable support to the newly formed Sudan People’s Liberation Army/ Movement (SPLA/SPLM). Second, as a dictator with absolute and arbitrary powers, General Nimeiri had resisted devolution of power, risking violation of the peace agreement to strengthen his grip. The sequence of events that led to the resurgence of civil war included repeated interventions by General Nimeiri in the Southern Executive Council. In the most flagrant violation of the peace agreement, on the advice of General Joseph Lagu, he further divided the south into three regions. Finally, imposition of the September laws (Islamic sharia’a laws), in 1983, effectively buried the Addis Ababa agreement. General Nimeiri’s calculations proved incorrect and his regime was eventually overthrown by a popular uprising that led to a military coup in April 1985, two years after the onset of the new civil war. With the benefit of hindsight it is clear that the collapse of the Sudanese peace was largely the result of an attempt to make a transition from war to peace without making a parallel transition to democracy including a broad-based dialogue of all the popular political forces in the country. The democratically elected government (1986–9) that followed the overthrow of General Nimeiri inherited a civil war. The insurgency was no longer content with a return to the old paradigm and vowed to continue fighting for a ‘new’, radically restructured, Sudan. 4.3 Why peace collapsed Could the low probability of successful peace-building in Sudan under the Addis Ababa agreement have been predicted? Based on initial conditions and design and, controlling for other factors, have peace-building experiences associated with democratic transition been more successful than those that were not? To address these questions, we employed a peace-building model designed by Doyle and Sambanis (2000), which models the probability of successful peace-building as a monotonic function of what they termed the ‘peace
Ali and Elbadawi: Sudan 153
triangle’. The size of this triangle depends on three elements: the extent of hostility; the local capacity for change; and the degree of external commitment to assist change. Using a version that identifies successful peace-building with both cessation of war and transition to democracy, in Table 9.2 and Figure 9.5 we simulate probabilities for successful peacebuilding for the Sudan in 1972 and for Mozambique in 1992. Mozambique’s experience, unlike that of the Sudan two decades earlier, was associated with a transition to democracy. The model predicts a low probability of successful peace in Sudan (0.180) compared to Mozambique (0.490). In both cases, the extent of hostility was high, and local competency low; the differentiating factor between the two experiences was the degree of external commitment to change. Mozambique’s success was greatly facilitated by the presence of a UN peace-keeping operation. This assertion is consistent with the global evidence of Doyle and Sambanis (2000), which suggests that, controlling for local capacities and hostilities, the highest probability of successful peacebuilding occurs when there is both a treaty and a UN operation.
5
Oil and post-conflict recovery
5.1 The influence of oil on Sudan’s economy The major oilfield currently being exploited straddles the border between north and south Sudan, falling largely within the historical borders of the south. Prospects for a sustainable peace and economic growth are closely linked to the effective management of oil resources, which have become an integral part of the conflict, and must therefore be part of the resolution. Democratic transition in both north and south alongside coherent criteria for revenue sharing is necessary for a sustainable peace.
2.00
1.00
0 Prob (Moz – Sudan)
Local Hostility (Moz – Sudan)
Local competency (Moz – Sudan)
External competency
Figure 9.5 External competency and successful peace-building: Sudan (1972) compared to Mozambique (1992) (on a scale of 0 to 2.00) Source: Ali, Elbadawi and El-Batahani (2002), table 24.
Probability of successful peace-building comparing Sudan (1972) with Mozambique (1992)
Variable
Parameter estimate
Sudan
154
Table 9.2
Mozambique
Value in Contribution to Value in Contribution to 1972 linear probability 1992 linear probability Extent of hostility War type (whether identity-based war, i.e. ethnic /religious) Logcost (natural log of death death and displacement) Wardur (duration of war in months) Factnum (number of major factions) Factnum2 (square of number of factions) Total External competency Trnsfcap (net current account transfers %, current US$) un2cint (type of UN operations, by mandate type) Treaty (parties signed treaty to end the war) Total Local competency Develop (per capita electricity consumption; pre-war) Exp (primary resource exports as % of GDP) Decade (0–6: decade war started) Constant Linear components Logit probability
Linear probability difference Sudan – Mozambique
−2.080 −0.540 0.009 −1.760 0.110
1.000 13.700 108.000 3.000 9.000
−2.080 −7.320 0.980 −5.290 1.030 −12.680
0.000 15.200 156.000 5.000 25.000
0.000 −8.130 1.420 −8.820 2.860 −12.670
−2.080 0.810 −0.440 3.350 −1.830 −0.010
0.004 0.710 2.080
29.970 0.000 1.000
0.110 0.000 2.080 2.190
8.300 3.000 1.000
0.030 2.120 2.080 4.230
0.080 −2.120 0.000 −2.040
0.000 −5.340
42.230 0.130
369.660 0.140
−0.350 10.320
2.000 1.000
0.020 −0.700 −0.680 −0.700 10.320 −1.540 0.180
0.210 −0.730 −0.520 −1.400 10.320 −0.030 0.490
−0.190 0.030 −0.160 0.730 0.000 −1.480 −0.310
4.000 1.000
Note: Dependent variable is coded 1 if war stopped and a transition to at least ‘low’ level democracy was achieved during the first two years of peace. Source: Doyle and Sambanis (2000), table 2, model A.
Ali and Elbadawi: Sudan 155
Oil is changing the profile of Sudan’s economy significantly since the country exported its first barrels in August 1999.5 Reserves are estimated at between one billion and five billion barrels.6 Recent estimates show a total annual production of about 52.8 million barrels, where profits constitute 30.62 million barrels – 58 per cent of production. Of this, the government received 22.05 million barrels, its net revenue from oil, after deducting oil-related loan repayments, amounting to US$292 million. In terms of the production structure of the economy, preliminary projections by the IMF show that the share of the oil sector in GDP should increase from 4 per cent in 2000 to 6.4 per cent in 2005. Oil has already begun to dominate exports, at 35.4 per cent in 1999. The commercial exploitation of oil has had a salutary impact on the growth performance of the economy. Ali (2002a) shows that sustained positive per capita growth is a feature of the second half of the 1990s. Following past trends, growth fluctuated greatly during the period 1990–4, with the real GDP growth rate ranging from 0.89 per cent to 7.0 per cent. In the second half of the 1990s, however, relatively high and sustained GDP growth emerged, rising to about 4.5 per cent, then exceeding 6 per cent for four years in a row. As a result, real per capita GDP amounted to about US$976 by the year 2000, recovering almost to the 1977 level of US$992. However, there is concern that this impressive growth did not have any impact on poverty reduction. As noted earlier, poverty is a major risk factor in civil wars. Notwithstanding difficulties in obtaining data, Ali (2002a) reported results of poverty studies during the 1990s, where poverty is defined in terms of the head-count ratio.7 This analysis estimated a rate of increase in the Gini coefficient of 4.51 per cent for the 1990s. The results show that the spread of poverty increased during that decade despite the overall growth in per capita GDP, as a result of the relatively low growth rates in per capita expenditure and because of the deterioration in the distribution of expenditure. For the first half of the 1990s, poverty increased marginally at an annual rate of 0.24 per cent, but for the second half the increase in the head-count ratio was significant at 2.4 per cent. This appears strange, given the impressive growth in per capita GDP, yet it can be explained by the allocation of the increased resources to the war effort. Despite progress on the economic and political fronts, there is no evidence to suggest that the National Salvation regime current at the time of writing has abandoned its strategic political, economic and social objectives. These include making radical changes to the civil service, police and the military; and creating a new entrepreneurial class through redistribution, fiscal and monetary policy. Outright military defeat of the rebels is no longer the regime’s prime objective, since it became seriously engaged in the peace process when oil production started to pay dividends. However, predation on state resources continues. The lack of transparency in the budget process, especially with regard to the allocation of the oil resources and the continued control of
156
Case Studies: The Horn, East and Central Africa
the state apparatus by the ruling party, seem to suggest that the state is still, in essence, ‘privatized’. This is consistent with global evidence of mismanagement and regressive development outcomes in oil-exporting countries ruled by unstable military regimes or factional democracies (Eifert, Geib and Tallroth, 2002). 5.2 Generating economic growth While oil revenues can greatly aid economic development, oil will not lead to sustainable growth without a sound macroeconomic and institutional environment, accumulation of human and physical capital, and due attention to factors such as geography and ecology, ‘demographic transition’ and, most importantly, peace and political stability. When compared to the best practices found in the East Asian economies, and even to those without these economies’, success, Sudan lags behind on almost every indicator. Economic growth will not arise without a comprehensive strategy addressing these factors. Table 9.3 shows the growth differential between the Sudan and East Asia. Much of Sudan’s weakness in these areas can be directly attributed to the long civil war. Human capital investment has been low; for example, as of 2000 the population over the age of 15 averaged just 2.14 years of schooling (Ali, 2002a; and Barro and Lee, 2000). Furthermore, civil war has been shown to be more intensely damaging than other types of conflicts, affecting growth through multiple channels (Collier and Hoeffler, 1998). If Sudan is to generate economic growth it must end not just the civil war, but also reverse its effects on institutions and on human and physical capital investment. The model includes a consideration of geographical factors. First, temperate climates have been shown to be associated more robustly with growth (Gallup and Sachs, 1998). Second, southern Sudan is isolated and landlocked, with access to ports only via the north or neighbouring countries. Even on Sudan’s Red Sea coast economic density is very low, at just a sixteenth of the East African average. Economic revival will need a deeper economic integration of the south (and other marginal regions of Sudan) with the Sudanese coast and the development of economic links with neighbouring African countries. For Sudan to benefit from oil revenues, resources should be used to rehabilitate the south and other marginal areas, build strong institutions, fund social services such as health and education, develop the infrastructure needed to diversify the economy, and be coordinated with an effort at regional cooperation to minimize the effects of isolation. To attain global competitiveness, Sudan should also exploit its geographic advantage as an exporter of agricultural and labour-intensive processed goods to the Middle East and Gulf markets (Ali, 2002b).
6
Building peace: past initiatives and future prospects
Numerous peace initiatives have been proposed since 1983. During the third democracy period between 1985 and 1988 there were three purely Sudanese
Ali and Elbadawi: Sudan 157 Table 9.3 Growth differential between Sudan and East Asia, 1975–99 (growth in East Asia minus growth in Sudan; difference in averages for period 1975–84, East Asia and Sudan 1975–99) Variable
1975–84
1985–9
1990–4
1995–9
−9.80 2.09
−12.49 2.01
−14.35 2.06
−16.70 1.83
Demographic shift Proportion of labour force to population (log)
2.58
3.56
3.69
3.57
Policy and institutional environment CPIA
1.45
2.05
2.65
3.03
Geography Coastal/navigable river population density Proportion of land in non-temperate zone (log)
6.72 0.17
6.72 0.17
6.72 0.17
6.72 0.17
0.15 3.35
1.21 3.24
1.51 2.44
1.21 −0.18
1.90 −1.45
−0.50 −3.74
0.30 −2.74
2.90 −3.08
Initial conditions/human capital Initial income Primary school enrolment
Conflict Civil war Explained difference Actual difference = difference in average growth rate Unexplained difference
Notes: The simulation is based on the difference in magnitude of the variable in question between the two regions multiplied by the coefficient taken from Elbadawi (2002), model 2a. CPIA: Country Policy and Institutional Assessment (from the World Bank). The CPIA is composed of twenty components covering four categories: macroeconomic management and sustainability of reforms; policies for sustainable and equitable growth; policies for reducing inequality; and public sector management and service delivery.
initiatives. In March 1986, all the northern and southern Sudanese political parties apart from the Democratic Unionist Party (DUP)8 and the National Islamic Front (NIF) had a meeting with the SPLA/SPLM and agreed the Kokadam Declaration. This provided for, inter alia, the lifting of the state of emergency, the abrogation of the September laws,9 the revocation of foreign pacts affecting the sovereignty of the Sudan, and the convening of a national conference to lay down the basic principles of the constitution of a reconciled Sudan. Subsequently, the DUP (the second major northern political party) and the SPLM reached agreement, leaving the NIF isolated. On 30 June 1989, just days before the government was due to approve the agreement formally the NIF staged a coup. The new military regime conducted peace talks with the SPLA/SPLM through IGAD, 10 meanwhile intensifying the war effort and increasing political repression. The regime also signed agreements with a number of southern
158
Case Studies: The Horn, East and Central Africa
splinter groups in a ‘divide-and-rule’ strategy that has come to be known as the ‘peace from within initiative’. Perhaps the most important achievement of IGAD has been the 1994 declaration of principles, stating support for a ‘peace agreement structured around a democratic and secular Sudan and call for the sharing of national wealth and resources. Failing that, they support negotiations on modalities of a self-determination referendum for the south’ (International Crisis Group, 2002b). This declaration formed the basis of subsequent negotiations between the government and the SPLA/SPLM. The international community has become involved increasingly in efforts to secure an end to the war. In September 2001, Senator John Danforth was appointed US special peace envoy to Sudan. Previously, the Center for Strategic and International Studies (CSIS) had issued a task force report on US–Sudan policy which stated that a peace process should be based on IGAD’s declaration of principles. The report (International Crisis Group, 2002a) recognizes that ‘oil is changing Sudan’s war. It is shifting the balance of military power in favor of Khartoum’, yet underlines the incentive of both parties to negotiate. Given ‘the single overriding objective of ending Sudan’s war’, the CSIS report proposed the following strategy for building peace: • An ‘inclusive’ external agency for ending the war and for providing a ‘commitment technology’ for subsequent peace and nation building, led by the USA. • The external agency should transparently and fairly deploy economic, diplomatic and political instruments to advance the cause of conflict resolution and peace-building.11 • Establish a transitional governance structure (one Sudan, two systems) that would lead to two self-governing entities, one in the north and one in the south. • Enable the exercise of the right to self-determination by the people of the south, in the hope of maintaining Sudan’s unity. • Design revenue-sharing criteria for oil, minerals and water. • Design a multi-agency, post-conflict reconstruction programme for the south and other affected areas to augment local competencies with expertise and resources from multilateral and bilateral development institutions. A modified version of the CSIS proposal could form the basis for a resolution of the conflict. Defining the details, however, requires a sustained, systematic and credible peace process. Major consideration should be given to the centrality of democratic transition (see Bigombe, Collier and Sambanis, 2000). Analytical evidence suggests that a peaceful but nondemocratic transition would be unstable and might result in Sudan’s partition, potentially leading to conflict between the partitioned states, and internal conflict within the states once the divide between north and south is removed.12 The three principles of neutrality, inclusiveness and
Ali and Elbadawi: Sudan 159
proportionality should therefore be integral to the design of the peace process. The proposed external agency and the peace-building process should be neutral relative to all relevant parties, inclusive of more than just the armed forces, and the influence of different groups should be proportional to the relative weights across factions and between north and south. Criteria for the allocation of oil revenues across regions, ensuring a fair share for the south and other marginalized regions might include the following considerations: the intensity and spread of poverty; the financing requirements of post-conflict reconstruction; and specific shares for the oil-producing regions. Success would necessitate transparent and broad-based governance at both federal and regional levels. This would suggest that any proposal similar to that of Danforth’s report calling for an equal share of the oil revenues must be based on genuine democratic transition in both north and south, whether in a united Sudan or in two independent countries. There are two further issues requiring careful and timely attention. First, the displaced southern Sudanese in the north, and second, tribal disputes over the natural resources straddling the north–south border. About 4 million southern Sudanese currently live in the north as internally displaced refugees. A post-conflict priority is to relocate these refugees to neighbourhoods equipped with social services and a physical infrastructure conducive to a decent and productive livelihood. This will require massive rehabilitation of internally and externally displaced southern Sudanese around productive economic and social projects in both the north and the south. In particular, the rings of poverty and destitution around the capital and other major cities must be eliminated by relocation either to planned residential areas within these cities, or to rural schemes. Regarding north–south cross-border relations, about 8 million Sudanese, mainly cattle-rearing nomads, cross the borders between the two regions and have traditionally struggled over natural resources. During the second insurgency (1983–present) these tribes have been dragged into the conflict, a sharp reversal of the historically generally cooperative relationships between these tribes. Traditional tribal leaders, whose effective leadership underpinned such cooperation in the past, have lost much of their influence. Therefore, as an immediate instrument for managing cross-border relations, the traditional systems for conflict resolution across the north–south borders should be revitalized. In the longer term there should be concerted efforts aimed at diversifying and modernizing their sources of economic livelihood to minimize their mutual competition for the natural resources on which they collectively depend.
7
Conclusion
With over thirty-seven years of civil war since independence in 1956, the Sudan is a country in conflict, characterized by religious and cultural divisions
160
Case Studies: The Horn, East and Central Africa
between the north and south, and tribal and regional cleavages throughout the country. This north–south divide also explains external interventions in the civil war. It is the main conclusion of this chapter that these two factors account for the higher risk of conflict in Sudan predicted in formal models of the hazards of civil wars, and are crucial to understanding why the war began and why it has persisted for so long. These factors should be addressed by the conflict resolution process, bearing in mind five preconditions for sustainable peace. First, appropriate constitutional measures and system of government to ensure equal rights and positive cooperation in a socially polarized and diverse society. Second, the right of self-determination for southern Sudanese, following a transitional period of peace and normality. Third, the peace process must be anchored by external global and regional stakeholders, including those who have actively taken sides in the conflict. Fourth, there must be explicit mechanisms in the peace process for a genuine democratic transition in both the north and the south. Fifth, there should be coherent plans for post-conflict reconstruction, including explicit criteria for the allocation and utilization of oil resources. It is widely agreed that the collapse of the Addis Ababa peace agreement in 1983, marking the start of the current phase of the civil war, was caused by its failure to recognize the centrality of democratic transition for sustainable peace and viable economic management in a divided society also characterized by second-order cleavages. The key lesson is that meaningful involvement of political parties and other stakeholders in the peace process is critical for ensuring an eventual transition to democracy. Unfortunately, the current peace initiatives, most notably the IGAD, being confined to the two main military protagonists, do not seem to be taking this basic but crucial lesson into account. Moreover, with the emergence of Sudan as an oil-producing country, transparent and democratic governance becomes even more crucial. Substantial evidence suggests that dictatorships or factional and unstable democracies have almost always failed to harness oil resources for development. Instead, under such regimes, oil resources have been associated with deepened poverty, social cleavage and conflict. Even if peace in Sudan is only attainable through partition, transition to democracy in both future ‘states’ will be critical. Given that the oilfields straddle the north–south divide, and the large numbers of southern Sudanese displaced to the north, partitioning the country without democratic transition risks replacing a civil war with an international war between two autocratic and non-representative regimes. Notes 1 In this chapter we describe a society or a country as being socially polarized when a plethora of ethnic, cultural and religious factors create a major divide between
Ali and Elbadawi: Sudan 161
2
3 4 5
6
7
8 9
10 11 12
two or more communities within a country. This divide is manifest in terms of major differences about identity, cultural aspirations, human and political rights. ‘Ethnic polarization’ can be thought of as a subset of ‘social polarization’, where tensions stem from a combination of issues such as ethnicity, religion, culture, history and language. For details about specific incidences of external intervention in the Sudanese civil wars, see Ali, Elbadawi and El-Batahani (2002). GDN at http://www.worldbank.org/research/growth/GDNdata.htm. According to an Amnesty report, the major companies currently active in the oil sector include: the Great Nile Petroleum and Oil Corporation (GNPOC), who hold the concessions for the two-main oil-producing areas; the China National Petroleum Corporation (CNPC), holding a 40 per cent share in the project; Petronas Bhd of Malaysia, holding a 30 per cent share; Talisman Energy of Canada, holding a 25 per cent share; and Sudapet of Sudan, holding a 5 per cent share. Estimated reserves were presented in the context of the recent report by the US Special Envoy, J. Danforth, to President Bush on the civil war in Sudan. The figures were reported in the St. Louis Post-Dispatch (http:// home.post-dispatch.com) on 29 April 2002. The head-count ratio, H, is defined as the number of those whose consumption expenditure falls below a given poverty line, say q, as a ratio of total population, say n. Thus H = q/n. The most widely used poverty measures can be defined in terms of a general functional form such that P = P (μ/z, θ), where μ is the mean consumption expenditure, z is the poverty line and θ is a measure of the inequality in the distribution of consumption expenditure usually taken as the Gini coefficient. Differentiation with respect to time will show that changes in poverty will depend on the elasticity of the poverty measure with respect to consumption expenditure and the Gini coefficient as well as the rates of change of these two determinants over time. This is the method used in Ali (2002a) to look at the changes in poverty during the 1990s in Sudan. The DUP is the second largest northern Sudanese political party after the Umma party. These penal codes and other laws were enacted in September 1983 by General Nimeiri. Claimed by his regime to be Islamic laws, they were viewed widely in the Sudan as nothing more than political opportunism. Nevertheless, they increased political tensions and complicated the initial conditions for the third democracy period, which followed the Nimeiri regime. IGAD is a regional organization that includes Sudan’s neighbours: Kenya, Uganda, Ethiopia, Eritrea and Djibouti. This was implied but not stated explicitly in the report. It has always been emphasized that once the fundamental divide is removed by partitioning the country, tribal differences are likely to plunge the south into inter-tribal war. However, the difference between north and south on this point could just be a matter of degree. For example, under the current regime, the absence of genuine multi-party bargaining has led to acrimonious regional political conflicts and the intensification of tribal conflict in western Sudan. However, recent conflict in Darfur, predicted by the analysis, falls outside the scope of this chapter.
References Ali, A. A. G. (2002a) ‘Economic Growth and Growth Prospects in Sudan’, background paper prepared for a World Bank country economic memorandum on Sudan (mimeo).
162
Case Studies: The Horn, East and Central Africa
Ali, A. A. G. (2002b) ‘Diversification and Structural Transformation of an Agrarian Economy: The Case of Sudan’, in Arab Planning Institute and the United Nations Economic and Social Commission for Western Asia (ESCWA), Diversification Experience in the Arab World (Beirut: ESCWA). Ali, A. A. G. and I. Elbadawi (2002) Explaining Sudan’s Economic Growth Performance (Nairobi: African Economic Research Consortium). Ali, A. A. G., I. Elbadawi and A. El-Batahani (2002) ‘On the Causes, Consequences and Resolution of the Civil War in Sudan’, paper presented at the Yale University– World Bank Workshop on ‘The Economics of Political and Criminal Violence’. Alier, A. (1990) Southern Sudan: Too Many Agreements Dishonoured (Exeter: Ithaca Press). Barro, R. and J.-W. Lee (2000) ‘International Data on Educational Attainment: Updates and Implications’, Center for International Development working paper no. 42 (Cambridge, Mass.: Harvard University). Bigombe, B., P. Collier and N. Sambanis (2000) ‘Policies for Building Post-Conflict Peace’, Journal of African Economies, vol. 9, no. 3, pp. 323–48. Collier, P. and A. Hoeffler (1998) ‘On the Economic Consequences of Civil War’, Oxford Economic Papers, vol. 50, no. 4, pp. 563–73. Collier, P., A. Hoeffler and M. Söderbom (1999) On the Duration of Civil War (Washington, DC: World Bank). Doyle, M. and N. Sambanis (2000) ‘International Peacebuilding: A Theoretical and Quantitative Analysis’, American Political Science Review, vol. 94, no. 4, pp. 779–802. Eifert, B., A. Gelb and N. Tallroth (2002) ‘The Political Economy of Fiscal Policy and Economic Management in Oil Exporting Countries’, World Bank, Washington, DC (mimeo). Elbadawi, I. (1999) ‘The Tragedy of the Civil War in the Sudan and Its Economic Implications’, in K. Wohlmuth, A. Gutowski, E. Grawert and M. Wauschkuhn (eds), African Development Perspectives Yearbook 1999: Empowerment and Economic Development in Africa (London: LIT Verlag). Elbadawi, I. (2002) ‘Reviving Growth in the Arab World’, World Bank, Washington, DC (mimeo). Elbadawi, I. and N. Sambanis (2000) ‘External Interventions and the Duration of Civil Wars’, World Bank Working Paper (Washington, DC: World Bank). Elbadawi, I. and N. Sambanis (2001) ‘How Much War Will We See? Explaining the Prevalence of Civil War’, Journal of Conflict Resolution, vol. 46, no. 3, pp. 307–34. Gallup, J. L. and J. Sachs (1998) ‘Geography and Economic Growth’, in B. Pleskovic and J. E. Stiglitz (eds), Annual World Bank Conference on Development Economics 1998 (Washington, DC: The World Bank). International Crisis Group (2002a) Capturing the Moment: Sudan’s Peace Process in the Balance, Africa Report No. 42 (Brussels: International Crisis Group). International Crisis Group (2002b) God, Oil and Country: Changing the Logic of War in Sudan, Africa Report No. 39 (Brussels: International Crisis Group). Regan, P. (2000) Civil Wars and Foreign Powers: Interventions and Intrastate Conflict (Ann Arbor, Mich.: University of Michigan Press).
10 Entrenching Peace in Post-Conflict Economies: The Case of Uganda Michael Atingi-Ego and Rachel Kaggwa Sebudde
1
Introduction
There has been at least one conflict in Uganda in each of the four decades since independence in 1962. However, since 1986 when Yoweri Museveni gained power, relative peace and security have been restored, which have in turn supported sustained economic growth. Collier and Reinikka (2001) point out that civil wars create intense inter-group hatreds and grievances, which persist in post-conflict society, increasing the probability of the re-emergence of civil conflict, which could explain the recurrence of conflict in Uganda. Given the profound gaps between popular perceptions on the causes of conflict, economic theory and empirical results, the choice of policy responses for post-conflict economies remains an important debate. Models using global data to predict the risk of conflict occurring are limited by country-specific factors. In this chapter we consider the experience of Uganda from the perspective of global models, and review the particular factors relevant to Uganda.
2
Conflict in post-independence Uganda
Using common definitions of conflict such as those adopted by Licklider (1993, 1995) Singer and Small (1994) and Elbadawi and Sambanis (2001), there have been three major incidents of conflict in Uganda since independence: the 1964–6 crisis; the 1979–80 liberation war; and the 1981–6 National Resistance Movement (NRM) guerrilla war. Since 1986, during the NRM government, the greater part of the country has been peaceful, although some peripheral areas have continued to be affected by armed violence and other forms of insecurity. 2.1 The 1964–6 crisis This crisis was rooted in Uganda’s colonial history and the process of independence. In 1962, prior to independence, a Constitution was promulgated to provide for democratic leadership, the observance and preservation of 163
164
Case Studies: The Horn, East and Central Africa
human and property rights, the authority and functions of administration, and the structures and institutions of government. However, political analysts argue that the Constitution was biased in favour of the powerful kingdom of Buganda, which raised negative sentiments from other tribes. The Mengo1 establishment contributed significantly to the fragility of the postindependence political system because it favoured secession from the rest of Uganda. Although Mengo abhorred party politics and direct elections, to remain active in the government, it formed a political party, the Kabaka Yekka (KY) headed by the Kabaka, the King of Buganda, to compete with the Democratic Party (DP) that drew its support largely from Catholics, and the Uganda People’s Congress (UPC), which was mainly Protestant. Opposed to direct elections, KY won a concession that its representatives to the national assembly would be elected from the Lukiiko (Buganda Parliament), hence through indirect election. The DP’s strong opposition to the differential treatment of Mengo affected its rivalry with the KY. Despite resistance from some Baganda (disapproving the active participation of the Kabaka and the involvement of Buganda with national politics), the UPC formed an alliance with KY. The alliance gave the KY exclusive rights to field all candidates in Buganda in order to compete against DP, which had a strong base in Buganda, while UPC fielded candidates elsewhere. On winning the elections, the UPC–KY coalition government installed the Kabaka as president and Milton Obote as prime minister. The foundation of the alliance had been opportunistic and it lacked any mechanism for solving internal conflicts, so no sooner had the alliance government been set up than conflicting interests began to strain it (Museveni, 1997). These included the Mengo versus central government, Baganda versus non-Baganda interests, leftists’ versus rightists’ views and, later, Bantu versus the Nilotics’ interests. Tensions relating to Buganda versus non-Buganda interests can be identified in the 1964 anti-government riots. As had been stipulated in the preindependence conference in Britain, the Minister of Justice proposed 4 November 1964 as the date for a referendum to determine whether six ’lost counties’2 would be returned to the Bunyoro Kingdom. Although the bill was passed by Parliament, the President (who was also the Kabaka of Buganda) declined to sign it, as he was reluctant to give away his land to Bunyoro. However, provision for such an eventuality had already been made by Parliament, allowing the prime minister to sign the assent, which he did. The result of the referendum was in favour of the ‘lost counties’ being returned to Bunyoro. Deep divisions within the UPC–KY alliance emerged, the Katikiro (prime minister) of Buganda resigned, and riots in protest at the results of the referendum spread across Buganda, forcing the government to impose a state of emergency. Similarly, divisions based on leftist versus rightist ideologies led to the arrest of four ministers who were considered extreme rightists. Analysts have argued that the KY wanted to continue working with the UPC as a party, but
Atingi-Ego and Sebudde: Uganda 165
the leadership of the latter was a stumbling block. Consequently, the right extremists within the UPC wanted to change their leadership by any means, for which they received the support of the KY. This factor, rather than ideology, led in 1966 to the arrest and detention of the four ministers including the Secretary General of the UPC. Misunderstandings within the executive continued, but the relationship between the executive and the legislature was also deteriorating. A central issue was Uganda’s involvement in Congo in 1966, during which President Obote and the army deputy-commander, Idi Amin, were alleged to have looted gold. A motion was moved in Parliament accusing the two of the offence, but a near-unanimous vote to set up a commission of inquiry and suspend Amin was reversed by Obote.3 No action was taken on Amin. In 1966, the 1962 independence constitution was suspended. The Buganda Parliament resisted this by expelling the central government from its lands. The political crisis in Uganda deepened when the UPC-led government attempted to maintain power unconstitutionally, declaring perpetual states of emergency in Buganda. Finally, in May 1966, the Kabaka’s palace was attacked and the Kabaka escaped into exile and Obote declared himself the executive president. Arrests, detentions without trial and killings of innocent civilians in many parts of Buganda ensued. The Kabaka’s palace was overrun and later converted into a military barracks, while the Lukiiko (Buganda Parliament) building became the Ministry of Defence headquarters, an action that provoked hatred for the Obote regime among the Baganda. 2.2 The liberation struggle, 1978–80 The Idi Amin years which followed the military coup in 1971 arose out of a number of political and economic factors (see Kasekende, Atingi-Ego and Sebudde, 2002), which further intensified the political crisis in Uganda. Even the initial support that this change of government received was shortlived as repression and military control escalated. Within a year, Ugandan exiles in Tanzania had made an abortive attempted to overthrow the Amin government. The resultant bombing by Uganda of Bukoba and Mwanza in Tanzania and the subsequent death of innocent civilians in these towns caused tension in Tanzania, with President Nyerere being accused by fellow Tanzanians of inviting war on their country. The Mogadishu peace accord of 7 October 1972, brokered by Somalia’s President Siad Barre supported by the Organization of African Unity (OAU), resulted in the Tanzania government finally accepting the Amin regime as the legitimate government of Uganda. The peace accord stipulated that propaganda and subversive activities against the Amin regime should cease, thus creating hostility towards Amin’s opponents, most of whom, such as the Front for National Salvation (FRONASA), led by Yoweri Museveni, were based outside Uganda. Analysts have suggested that the 1978–9 conflict arose mainly from the grievances within Uganda and Uganda’s annexation of the Kagera Salient in
166
Case Studies: The Horn, East and Central Africa
1978 (Mamdani, 1984; Mudola, 1993; Museveni, 1997). In 1979, a combined force of Tanzanian and Ugandan exiles overthrew Idi Amin and installed the Uganda National Liberation Front (UNLF) as the government. However, despite this, political instability increased, characterized by the short-lived regimes following the Amin era. The instability largely resulted from the ad hoc manner in which the UNLF was assembled during the 1979 Moshi conference held in Tanzania. Museveni (1997) argues that the main concern of the delegates was to form a viable vanguard against Amin. The Tanzanians, on the other hand, wanted to have a Ugandan Front to popularize the fight against Amin. In summary, because of the lack of a serious post-Amin political agenda, political strife, suspicion and the resulting tensions tore the UNLF apart two years after the fall of Amin. 2.3 The civil conflict, 1981–6 The 1980 elections reintroduced an opposition in Parliament but it was an ineffective restraint on the abuse of power. Community participation in the political process equally was insufficient as the elected district councils were generally constrained by the central government. Although Obote was declared the victor in the elections, not all parties were satisfied that the poll had been free and fair. Some of the dissatisfied parties chose to take up arms and launched an armed struggle in February 1981, marking the beginning of five years of conflict. Led by Museveni, the National Resistance Movement Army (NRA) gradually expanded its campaign through politicizing communities and the voluntary recruitment of civilians into its ranks. Weapons, artillery and communications equipment were acquired through attacks on the military and police. Given its guerrilla nature, the war affected large areas of the central region, the ‘Luwero Triangle’. The guerrillas relied on the resources of communities, such as loans of cattle (Museveni 1997). For example, to sustain the force between September 1983 and June 1984, close to 21,000 head of cattle were borrowed from the Bahiima community in the region. In some instances, money looted from commercial banks enabled the purchase of food. Raids on government health facilities provided medical supplies and personnel. Factions within the ruling alliance wanted to remove Obote so that they could reach a peace agreement with the military wing of the NRM, the NRA. In July 1985, government soldiers, mainly Acholi, supported by some civilian opportunists, mutinied as a result of war losses. However, no sooner had they captured state power than they lost the contact they had established with the NRA. Their plan was had been to co-ordinate with foreign powers to reinforce the regime, isolate the NRA and later defeat it. Contacts were established subsequently with former Amin soldiers, giving concern to President Nyerere of Tanzania, who then gave his support to the NRM/A.
Atingi-Ego and Sebudde: Uganda 167
Attempts to foster peace by President Moi of Kenya through the Nairobi peace accord could not therefore be fully implemented under such circumstances, and it is against this background that the NRA decided to launch a final assault on Kampala. On 25 January 1986, it captured the capital and formed a new administration. Analysts argue that this five-year protracted war widened Uganda’s north–south divide and resulted in major economic losses because of the displacement of people, loss of property and the contraction of agricultural production. 2.4 The post-1986 insurgencies The new government had a ten-point-programme aimed at restoring democracy at the grassroots level of the population, ensuring security and restoring economic growth through prudent policies and institutional reform. Some of the measures taken included a constitutional review, which resulted in the promulgation of the 1995 Constitution, the introduction of a decentralization policy and the re-institution of cultural leaders, all of which were geared towards empowering the people through democratization, accountability and transparency. However, over the past 15 years challenges have arisen in the form of insurgencies in the northern and western parts of the country. While some of these incidences were only short-lived, others have persisted. The Lord’s Resistance Army (LRA), led by Joseph Kony, has paralyzed economic activity and strained development in the north of the country. In addition, during the late 1980s, the north and some parts of eastern Uganda also suffered from cattle rustling by the Karamojong, which according to United Nations (UN) estimates, had by 2000 caused the displacement of up to 90,000 people. The Karamojong had acquired a substantial number of weapons on the fall of Amin in 1979 and continued to acquire more from nomadic tribes among Uganda’s neighbours. The ineffective 2002 voluntary disarmament programme was followed by forceful disarmament to reduce the number of weapons held by the Karamojong. Furthermore, in the 1990s, the activities of the Allied Democratic Forces (ADF) in western Uganda intensified, forcing government to send its troops into the Democratic Republic of the Congo (DRC) to cut off its supplies. Although this resulted in the elimination of the ADF, Uganda became involved in the ongoing regional conflict centred on the DRC. The increasing civil conflict in this period is also evident in household survey data, which shows that 10 per cent of the surveyed households had been affected by civil conflict in the year 2000, compared to 7 per cent in 1992. UN data also shows that the recorded number of people affected by war increased to a peak of 937,431 in May 2000, before the trend reversed to 830,175 in December of that year (see Table 10.1).
168
Case Studies: The Horn, East and Central Africa
Table 10.1
Estimated effect of insurgencies, 1998–2000 By end of November 1998
September 1999
September 2000
December 2000
Refugees IDPs Abducted children
183,695 469,525 na
196,251 529,215 na
201,427 736,004 4,804
213,872 610,240 6,063
Total
653,490
730,270
937,431
830,175
Source: UN database.
3
Explaining Uganda’s recurring conflict
Global studies on the onset, duration and incidence of war (see, for example, Collier and Hoeffler, 1998, 2001; Collier, 1999; Elbadawi and Sambanis, 2001), have generally argued that conflict is caused by two major factors. First, the presence of aggravated parties who turn to war to settle their differences or grievances, and second, the desire of the rebels to satisfy some form of economic interest. In this section we test the extent to which some of the existing models can explain conflicts in Uganda. Specifically, we use the Collier and Hoeffler (2001) and Elbadawi and Sambanis (2001) models.4 3.1 The Collier–Hoeffler model Collier and Hoeffler (2001) argue that the risk of war increases as the resource base increases, and reduces as the opportunity cost of rebellion increases. Educational attainment prevents the onset of conflict significantly, as it is also an opportunity cost. The availability of cheap and illiterate manpower in Uganda provided a pool for recruitment into armed rebellion. Using a set of six observations (1966, 1978, 1980, 1988, 1998 and 2001), the model predicts low risks of the onset of civil war, even though conflict did break out. This notwithstanding, the model’s predictions support the hypothesis that grievance accounts for most Ugandan conflicts, and that the contribution of these factors to the risk of war evolves with the change in the structure of the economy. In addition, reliance on a few primary commodities, together with the stagnation of the economy amid a growing population, increased the risk of conflict in Uganda. That the highest risk of conflict was recorded in 1980, and was mainly caused by of the economic dysfunction and dilapidation of the country during Amin’s regime, was therefore not surprising. Other factors relating to justice-seeking or grievance-induced civil conflict seem to arise from the population factor rather than social fractionalization and ethnic dominance. On the demand side, Collier and Hoeffler (2001) argue that grievance increases with social fractionalization, political repression
Atingi-Ego and Sebudde: Uganda 169
and economic dysfunction. These factors, though predicted to be well below the global average, have been increasing because of the rising population. We also found the risk of conflict caused by social fractionalization to be low. This is because, in spite of significant ethnic and religious fractionalization, Ugandan society is highly heterogeneous, which reduces its contribution to the risk of conflict. 3.2 The Elbadawi–Sambanis model estimation of risk of war in Uganda Developing the Collier, Hoeffler and Söderbom (2001) proposition that onset and duration are quantitatively different processes, Elbadawi and Sambanis (2001) test empirically for war incidence, combining both onset and duration factors. They conclude that all the economic and political factors included in the model, together with social fractionalization, are associated robustly not only with the onset of war, but also with the incidence of civil war. Social fractionalization reduces the incidence of war through its effect on the onset of war. Ethnic fractionalization contributes positively to the incidence of war through its effects on duration. Unlike Collier and Hoeffler (2001), who fail to find a robust association between political rights and war initiation, Elbadawi and Sambanis (2001) find that low levels of political rights are correlated negatively with the incidence of war, but do not significantly support a non-monotonic effect of political rights on the incidence of war as predicted in the political science literature. Furthermore, their results show that the risk of incidence of war is related more robustly and negatively to initial levels of economic development and less robustly to economic growth. Primary exports to GDP have a positive and significant (though non-monotonic) correlation with the incidence of war. This is because, while they can be used to fund loot seekers and support justice seekers, beyond a certain range they are a formidable source of funding for government armies, buy popular support and compensate external allies. Their results also suggest that population size is correlated positively with the incidence of war: it appears that countries with smaller populations have a lower risk. But in fact small countries face a comparatively higher risk of war, since risk does not increase proportionatally with population size. The authors argue that, while ethnic diversity is positively and non-monotonically associated with the incidence of war, social fractionalization becomes associated with the higher incidence of war as it moves towards polarization. Applying the Elbadawi and Sambanis (2001) model to the data set used earlier indicates that grievance was the main factor contributing to the increased risk of conflict in Uganda. Because of the different weighting of the explanatory factors used in the two models, actual levels of risk resulting from each model vary at any given time in history, although the pattern is the same. Both models, however, predicted low risks of conflict at times when conflict in fact occurred, but that the risk of conflict in the period 1998–2001 was greatly reduced.
170
4
Case Studies: The Horn, East and Central Africa
Identifying additional factors
Given the limitations of the existing models in predicting conflict in Uganda, we now review other factors that may have contributed, directly or indirectly, to conflicts and their recurrence in Uganda, and which can be incorporated into existing models. These include the formation of the state of Uganda, social fractionalization, external factors and institutional development. 4.1 Formation of the state of Uganda Uganda was formed as a conglomerate of several tribal groupings and kingdoms that had been in conflict with one another for a number of centuries. From the beginning, divisions were exploited by British colonists in pursuit of their interests and by Arabs trading slaves and commodities. Mamdani (1984) summarizes the political impact of the British rule in Uganda as having capitalized on divisions and differences among the people to create a comprador class, usually from southern Uganda, which acted as agents for the colonial government. Oppression by the agents of their subjects promoted class conflict under the policy of indirect rule, and was particularly pronounced in the traditional kingdoms in the south of Uganda, and to a lesser extent in the east and north. In the conquest of Bunyoro, for example, its traditional rival, Buganda, was used and later rewarded with six counties (‘the lost counties’), which later became one of the major causes of the 1966 crisis. Similarly, conquest of the east by Kakungulu’s forces was rewarded with chieftainship in the conquered areas. Acts of brutality and extortion did not endear these chiefs to their subjects and, until recently, it was easy to arouse anti-Baganda feeling in other parts of the country through these memories. The colonial government often favoured southerners for office jobs, while northerners were used in the army and police and as plantation workers in the south. Since the army and police were dominated by the northern tribes that were often used to settle political differences and protect the interests of the colonialists, this tended to create tensions between southerners and northerners rather than between Ugandans and the colonialists. In the postindependence period, these divisions were manifested in a politically fluid environment that needed only a small trigger factor to spark off violence. 4.2 Fractionalization Uganda’s society is inherently fractionalized. While this in itself may not promote conflict, Mudoola (1993) argues that groups organized on the basis of religion and ethnicity have been involved in political activity since the establishment of colonialism. Ethnic fractionalization Ethnic groups were the most powerful historical socio-political force and have thus dominated the political arena, thus increasing the risk of polarization
Atingi-Ego and Sebudde: Uganda 171
and the incidence of war in Uganda. The constitutional arrangements accommodated their needs, but engendered problems that promoted conflict. First, in the pre-independent state, the elites (who later took over leadership roles) that colonialism bequeathed to Uganda, attached greater sentimental value to their tribes than to nationalism, thus straining the effectiveness of post-independence institutions. Ethnic groups and politicians used resources to promote their own interests. As a result, Uganda’s political arena lacked effective institutions, and groups with political resources easily seized power, which further provoked grievances. The negative effects of ethnicity were exacerbated by uneven socio– economic development. Given the low level of economic development and high illiteracy, and tribal and ethnic polarization, the promotion of narrow interests such as tribalism and ethnicity was a lot easier, as agents would support others because they belonged to their group and not for the policies they put forward. Museveni (1997) argues that such polarization induced a false consciousness, which kept Uganda underdeveloped. Religious fractionalization Uganda has a history of religious conflict, reflecting doctrinal differences between Catholicism, Protestantism and Islam. Rivalries were perpetuated by the Arabs, and British and French missionaries, in their bid to dominate the Kabaka. The controversial takeover of Catholic schools by the government at the time of independence was an attempt to weaken this religion. The Catholic Church lobbied for control of Catholic schools and its moral right to educate its followers. Its marginalization is evidenced by the positive attitude towards the 1966 crisis and the reception of the 1971 Amin coup (Mudoola, 1993). Protestantism, on the other hand, could have been considered the state religion during the colonial period. Kasekende, Atingi-Ego and Sebudde (2002) observe that the 1900 Land Agreement won over most chiefs by awarding land, but also by entrenching Protestantism over other religions. This polarization was maintained until ethnic politics spilled out into the Church of Uganda, resulting in struggles for succession and confrontations between Buganda and central government. Obote manipulated this cleavage during the 1966 crisis. Up to this time, the Ugandan Islamic community was marginalized; Mudoola (1993) argues that they then shifted from a patron– client relationship during colonial days to become mediators of the balance of power between the two larger Christian groups. The north–south divide The tribal and regional allocation of economic resources during the colonial period created divisions which translated subsequently into a conflict-prone environment. Cash crops were grown in the south, which also benefited from basic education services, providing recruitment into the central
172
Case Studies: The Horn, East and Central Africa
administration. The north, on the other hand, was a labour reserve for the security services, factories and the plantations in the south. The resulting unequal distribution of incomes and economic opportunity created grievances, thus exacerbating the polarization of ethnic and tribal tensions. Indeed, one of the reasons advanced by the LRA rebellion against the Museveni regime is the marginalization of the north. Following the loss of political and military status in 1986, the north became a fertile ground for insurgency. While micro factors may have triggered the rebellion, its sustainability has been affected by perceptions in the north of their persecution, isolation and victimization, while institutions which had employed substantial numbers of disadvantaged and less educated northern youth became dominated by southerners. 4.3 External factors Apart from colonialism, several other external factors are important in explaining conflicts in Uganda. As noted earlier, the presence of Ugandan troops in Congo was a trigger factor in the 1966 crisis. Poor relations with Tanzania in the period 1972–8 contributed to the 1979 war. The support Congo and Sudan allegedly gave to rebel groups in the 1990s helps to explain the sustained insurgencies in the post-1986 period. Other reasons for the insurgency notwithstanding, analysts have argued that an alliance between the LRA and the Sudanese government in retaliation for Uganda’s alleged support to the Sudanese People’s Liberation Movement has sustained the rebellion. These external factors introduced new and complex regional dynamics to Uganda’s history of conflict. In March 2002, the government launched ‘Operation Iron Fist’ to end rebellion in the north. Following the events in America of 11 September 2001, Uganda’s relationship with Sudan has improved. The USA’s recognition of the LRA as a terrorist organization, and accusations by Washington of Sudanese government links with Osama bin Laden, have resulted in the Sudan regime seeking to improve its image and end its diplomatic isolation. In an important political investment, regionally and internationally, Sudan has provided intelligence information on Al-Qaeda activities and allowed Uganda’s military to operate within its country in pursuit of the LRA. Operation Iron Fist overran several LRA camps and captured substantial quantities of military equipment. That the LRA had sophisticated resources valued at almost US$2 million suggests that they received considerable assistance within Sudan. 4.4 The role of institutions While institutions to resolve internal conflicts ought to be independent, institutional development in Uganda has been distorted by the unbalanced allocation of resources along ethnic, regional and religious lines. Mudoola (1993) notes that friction between the well-endowed and the marginalized
Atingi-Ego and Sebudde: Uganda 173
groups has complicated the process of institution-building. Given the conflict of interests, the colonial agenda did not prioritize the establishment of institutions compatible with the Ugandan political situation and its needs. Indeed, the formula for integrating the different political and cultural traditions to provide a basis for sustaining a new political order was also ill-equipped to handle the complex post-colonial situation in Uganda. Later, during the early post-independence period, although the attempt by the army to mutiny in 1964 indicated weaknesses within this military institution, no effort was made to rebuild it as a national institution. Rather, it was used for political survival by subsequent governments. This is in contrast to the Tanzanian experience, where, because of the one-party system, Nyerere was able to reform the army after an attempted mutiny, thereby transforming it into a national institution. Despite its shortcomings, the 1962 Constitution empowered different interest groups by making explicit the roles of different stakeholders. However, during the 1966 crisis, the army was used to support the concentration of power in the presidency and dismantle the political system enshrined in the 1962 Constitution. The unpopular interim ‘pigeonhole’5 Constitution of 1966 reduced the powers of central government and increased those of the presidency, resulting in authoritarianism and the decline of democracy. All levels of administration, including the judiciary, became subjected politically and their decisions were circumvented by retroactive legislation. This threatened human and property rights, cast doubt on transparency and accountability, and eliminated community participation. Although structures of governance continued to exist from parish6 up to district levels, this local government was now powerless. The use of repression – in particular by the military – to solve political problems is an indication that Uganda’s political inheritance at independence lacked resilient democratic institutions (Reinikka and Collier, 2001). Political leaders instead reinforced inherent institutional weaknesses in the army to exploit then for their own benefit. The army was not only biased in favour of the northern region, but also tribally skewed in favour of the Acholi, Langi and West Nilers. Museveni (1997) observed that, by 1971, the 8,500-strong force comprised 4,300 Acholis and 1,300 Langis, with the balance mainly from West Nile. Used against the Parliament, the Constitution and the people of Uganda, the army became dominant, eventually overthrowing the Obote regime in 1971. The 1971 coup elevated the army as an institution into the supreme organ of state (Mamdani, 1984). It not only used terror to defend the ruling class, but was also used by individuals to eliminate obstacles in their search for wealth. The rule of law and order collapsed completely in 1972 when the military tribunal took over from the civilian courts, contrary to the claim often made by the regime that the independence of the courts was absolute (Kasekende, Atingi-Ego and Sebudde, 2002). As Amin assumed legislative and executive
174
Case Studies: The Horn, East and Central Africa
powers, most legislation took the form of presidential decrees, while military tribunals judged major violations of the law. For safety, civil servants sought connections with the agents of the regime. Public service delivery thus collapsed as civil servants took orders from the military or became corrupt. In spite of the political will to promote good governance and accountability during the immediate post-Amin era, institutions that had been run down over the years were ineffective. Even as the Museveni government embarked on the task of rebuilding institutions, many analysts have noted that this will be a long process, especially when the consequences of prolonged conflict are a legacy of suspicion and bitterness (Collier and Reinikka, 2001). The cultural values, attitudes and beliefs that could promote a system of good governance, which were eroded under the rule of the army, were replaced with opportunism, and civil and political strife, constraining the proper functioning of government authority.
5
Measures to address conflicts
5.1 Political measures Measures to diffuse conflict included peace pacts, often negotiated by external mediators. Key among these were the 1972 Mogadishu peace accord, which diffused tensions between Uganda and Tanzania, and the 1985 Nairobi peace accord which, negotiated at the height of the protracted 1981–6 war, failed because of lack of commitment. Since 1986, the Museveni government has made several attempts to end insurgency within the country through negotiation, sometimes accompanied by offers of incorporation into a broad-based national government. Such strategies yielded a positive response from the rebels who were operating in the east and west, but placating northern rebels has proved to be more problematic. In the belief that ‘the cost of peace is nothing compared to the price of war’, the 2000 Amnesty Act offered blanket forgiveness to those who would renounce rebellion, lay down their arms and reintegrate into normal life. In addition, the government encouraged organized groups to pursue talks with the LRA to achieve peace. Furthermore, programmes such as the Northern Uganda Reconstruction Programme (NURP), supported by the World Bank and other donors, were introduced in the 1990s to promote economic activity in the north, intended to address this regional economic imbalance. Although NURP operations were constrained by the continued insurgency, support from the World Bank, valued at US$ 150 million, continues to aid reconstruction in the Northern region. 5.2 Institutional measures During the first three decades after independence, Uganda’s political institutions remained weak and were manipulated by the political elite to satisfy
Atingi-Ego and Sebudde: Uganda 175
their own interests. The post-1986 government began rebuilding institutions to allocate resources and resolve disputes. That democratization was high on the new government’s agenda was reflected in the creation of Resistance Councils to facilitate grassroots democratization, press freedom and decentralization of power, through the 1993 Local Government Statute. In the early 1990s, a commission surveyed the views of the public for inclusion in the formulation of a new constitution. These views were later debated by the Constituent Assembly7 and those agreed upon enacted into law as the Constitution of 1995. To accommodate the provisions of the Constitution (1995), the Local Government Statute was revised in 1997 and became the Local Government Act of 1997, which is the legal document underlying the decentralization programme. This specifies the structure of local government units, their power and function, level of central government control, sources of revenue, and the basis for the staffing and management of these units. In addition, the role of local communities and other stakeholders in the decision-making process, planning and implementation, and co-ordination procedures, are specified. This empowerment and community participation improved service delivery. Institutional changes to improve productive capacity have included public service, market and law reforms, with the aim of improving the effectiveness and efficiency in the allocation of resources in both public and private sectors. These have included the provision of a legal framework to strengthen commercial jurisdiction, property rights and the settlement of business disputes; support for the development of effective interest groups representing the needs of the private sector; and the establishment of a properly functioning financial system and safety net, such as deposit insurance schemes, in order to meet liquidity needs and minimize liquidity crises. Finally, with the return to the rule of law and order, it is envisaged that the army will become a national institution. Indeed, demobilization of army personnel to accommodate rebels being reintegrated into the army has to a large extent been implemented successfully (Collier and Reinikka, 2001). 5.3 Increasing economic opportunities Many of the existing models8 that estimate the risk of conflict underscore the importance of economic opportunities and income levels. Although incomes in Uganda were relatively high by African standards at the time of independence, the subsequent slow growth of the economy and reduced economic opportunities for a fast-growing population resulted in the erosion of per capita income. That social conflict and political disorder affected both the level and composition of economic activity shows how conflict and economic dysfunction can interact in a vicious circle. Collier (1999) argues that conflict reduces economic activity through destruction, disruption, reduced savings and investment, and capital flight, diminishing
176
Case Studies: The Horn, East and Central Africa Table 10.2
Impact of political instability on GDP per capita growth Political instability contribution (pp)
1960–4 1965–9 1970–9 1980–5 1986–9 1990–7
0.02 −0.11 −0.39 −0.29 −0.32 −0.08
Estimated GDP per capita growth 4.2 4.3 −4.3 4.1 2.2 2.5
Source: Kasekende, Atingi-Ego and Sebudde (2002).
the human and capital stock, reducing productivity and diverting expenditure to military activities. The impact on economic activity of social conflict and political breakdown is highlighted in Kasekende, Atingi-Ego and Sebudde (2002), where the analysis of growth predicted that the contribution of political stability to overall growth in GDP per capita, though marginal, was positive, at 0.02 percentage points in the period 1960–4, but becames negative with the deterioration of Uganda’s political economy. As shown in Table 10.2, the greatest negative effect of political instability on GDP growth was in the 1970s. As noted by Reinikka and Collier (2001), by 1999, the risk of large-scale internal conflict was considerably reduced compared to independence and the beginning of Museveni’s government. This is partly because of economic recovery during the 1990s. Policy measures contributing to this recovery, which in turn generated a virtuous circle reducing risks of conflict in most of Uganda, include, inter alia, economic liberalization, privatization, the prioritization of public spending, institution-building and investment policy. Liberalizing economic activities in Uganda reduced rent-seeking behaviour among the politically well-connected Ugandans, which had led previously to the misallocation of resources in favour of the ruling class, thus increasing the risk of conflict. Privatization, apart from increasing efficiency, promoted the role of private agents in production, and hence the accumulation of wealth, thus raising the opportunity cost of war. Rapid growth and a corresponding increase in economic activity has benefited greatly from the support of development partners, both financially and through technical assistance. By closing the savings–investment gap, official development assistance in the form of loans and grants (see Table 10.3) has supported government expenditure, despite a low tax base. Opening up the capital account has encouraged investment by facilitating the direct transfer of resources from external sources to domestic private entities. Supported by the international donor community, government poverty reduction strategies have improved living standards. Household survey data
Atingi-Ego and Sebudde: Uganda 177 Table 10.3
Official development assistance to Uganda (US$m) Financial year
Budget support Project support Total inflows Inflows/GDP (%)
1992/3
1993/4
1994/5
1995/6
1996/7
1997/8
1998/9
165.1 307.8 472.9 15.7
154.8 331.8 486.6 13.1
178.9 419.0 597.9 11.3
86.2 446.2 532.4 9.7
110.8 419.1 529.9 9.3
166.4 475.8 642.2 10.4
127.8 405.9 533.7 9.2
Source: Bank of Uganda (various years).
show that, while 56 per cent of Ugandans were poor in 1992, this had declined to 44 per cent in 1997/8, and 34 per cent in 1999/2000, a significant reduction over the relatively short period of seven years.
6
Lessons and the future
While grievance or justice-seeking causes have been the main drivers of conflict in Uganda, weak institutions have exacerbated the problem. Without well-functioning institutions, Uganda’s conflicts can be explained as a positive function of socio-political forces where groups dictated political terms acceptable to themselves. Marginalized groups, in turn, sought to use political resources and force to overthrow the established political order. This, coupled with declining economic opportunities, created the perfect conditions for conflict to occur. Since 1986, political reconciliation, including the return of some former leaders, the restoration of kingdoms, and the creation of economic opportunities which increased per capita income and reduced poverty, have combined to reduce the risk of the re-emergence of large-scale conflict and secured peace over most of the country. Reconciliation efforts and amnesties to rebels reduced the scale and impact of any conflicts that did occur. Rebuilding institutions and assistance from development partners has been critical in achieving this. While political and economic liberalization helps to reduce the incidence of war (irrespective of the ethno-linguistic fractionalization of society), strengthening institutions is an important priority for polarized societies, such as Uganda, to provide conflict resolution mechanisms, particularly for the marginalized. A regional analysis of the promotion of peace and security within the country would be beneficial, and would lead to improved relations with Uganda’s neighbours. Empowering people economically in post-conflict zones and a greater sense of political inclusion in order to give the marginalized a stake in the national economy and political system would also add to the securing of peace in post-conflict Uganda.
178
Case Studies: The Horn, East and Central Africa
Finally, economic opportunities, poverty reduction strategies and the support of the international community are important in creating peace and stability through sustained growth. These measures increase the opportunity costs of opting for conflict and reduce economic marginalization. However, although growth has resulted in poverty reduction, it remains unevenly distributed. The south, mainly engaged in cash crop agriculture, manufacturing and trade, has benefited in particular, whereas the northern region has continued to lag behind as a result of historical factors and the prolonged insurgency. The challenge continues to be that of ensuring sustainable growth and poverty reduction across all sectors of society and all regions, which in turn should further reduce the risk of conflict. Notes 1 Mengo was the seat of power for the Buganda kingdom. 2 The six counties (normally referred to as the ‘lost counties’) had earlier been given to the Buganda kingdom as a gift for their help to the British in taking over the Bunyoro kingdom that had resisted British rule. 3 Buganda, for example, largely supported the coup because of its hatred of Obote and the dismantling of the Buganda establishment. 4 The empirical results from the two models can be obtained from the authors on request. 5 ‘Pigeonhole’ Constitution refers to an interim Constitution that was prepared hurriedly by a small section of people to provide an interim guide for running the country. It was given the name ‘pigeonhole’ because it was deposited in the Parliament’s pigeonholes for circulation. 6 The lowest form of government organizational structure. 7 The constituent assembly was made up of elected representatives from all parts of Uganda. 8 For example, Collier and Hoeffler (2001) and Elbadawi and Sambanis (2002).
References Collier, P. (1999) ‘The Coming Anarchy? The Global and Regional Incidence of Civil War’, Working Paper, Centre for the Study of African Economies, Oxford. Collier, P. (2000) ‘Rebellion as a Quasi-Criminal Activity’, Journal of Conflict Resolution, no. 44, pp. 839–53. Collier, P. and A. Hoeffler (1998) ‘On Economic Causes of Civil War’, Oxford Economic Papers, vol. 50, no. 4, pp. 534–62. Collier, P. and A. Hoeffler (2001) ‘Greed and Grievance in Civil War’, Working Paper no. 2355, World Bank, Washington, DC. Collier, P., A. Hoeffler and M. Söderbom (2001) ‘On the Duration of Civil War’, World Bank, Washington, DC (mimeo). Collier, P. and R. Reinikka (2001) ‘Reconstruction and Liberalization: An Overview’ in R. Reinikka and P. Collier (eds), Uganda’s Recovery: The Role of Farms, Firms and Government (Washington, DC: World Bank). Elbadawi, I. and N. Sambanis (2001) ‘How Much War Will We See? Explaining the Prevalence of Civil War’, Journal of Conflict Resolution, vol. 46, no. 3, pp. 307–34.
Atingi-Ego and Sebudde: Uganda 179 Kasekende, L. and M. Atingi-Ego (1996) Implications of Financial Liberalization on the Domestic Financial System: The Case of Uganda, final research paper submitted to the AERC, Nairobi. Kasekende, L., M. Atingi-Ego and R. Sebudde (2002) ‘The African Growth Experience: Uganda Country Case Study’, Global Development Network, www.gdnet.org (mimeo). Licklider, R. (1993) Stopping the Killing: How Civil Wars End (New York: New York University Press). Licklider, R. (1995) ‘The Consequences of Negotiated Settlements in Civil Wars, 1945–1993’, American Political Science Review, vol. 89, no. 3, pp. 681–90. Mamdani, M. (1984) Imperialism and Fascism in Uganda (Lawrenceville, NJ: Africa World Press). Mudoola, D. (1993) Ethnicity, Religion and Politics in Uganda (Kampala: Fountain Publishers). Museveni, Y. (1997) Sowing the Mustard Seed: The Struggle for Freedom and Democracy in Uganda (London: Macmillan). Reinikka, R. and P. Collier (eds) (2001) Uganda’s Recovery: The Role of Farms, Firms and Government (Washington, DC: World Bank). Sambanis, N. (2001), ‘Do Ethnic and Non-ethnic Civil Wars Have the Same Causes?’, Journal of Conflict Resolution, vol. 45, no. 4, pp. 259–82. Singer, J. and M. Small (1994) Correlates of War Project: International and Civil War Data 1816–1992, data set held at University of Michigan, Ann Arbor.
11 The Challenge of Transition from War to Peace in Burundi Janvier Désiré Nkurunziza and Floribert Ngaruko
1
Introduction
Since independence in 1962, Burundi has had, up to the time of writing, five civil wars: in 1965, 1972, 1988, 1991 and 1993. Given that the root causes of these wars have never been properly addressed, we consider these as five episodes of one conflict that has claimed the lives of about 10 per cent of the population. The ongoing 1993 conflict has been the longest, costliest and most complex. It has claimed 250,000 victims, has created up to 700,000 refugees and about 900,000 internally displaced people (Nkurunziza and Ngaruko, 2002b). In terms of direct economic costs, the first four episodes had little impact and were relatively short. It is the fifth episode that has had dramatic consequences. Nkurunziza (2001) and Nkurunziza and Ngaruko (2002b) cover the social and economic consequences in greater detail. The Burundi conflict has been depicted as ethnic, the majority Hutu in opposition to the minority Tutsi,1 leading to violent clashes between the two groups. Why should these groups hate each other so much? These communities lived together without ethnic clashes for several centuries. The interesting question is why is it that violent conflicts only erupted from the 1960s onwards? While ethnicity is important in explaining conflict in Burundi, it is only one among several factors: economic factors; bad governance; the application of external ‘role models’; and ‘regionalism’. Beyond the use many analysts have made of ethnicity to explain ‘seemingly bizarre or incomprehensible phenomena’ in Africa (Monga, 1998, p. 33), we argue that it is the combination of these factors that explains war in Burundi (Ngaruko and Nkurunziza, 2000). As war has not fully ceased in Burundi, this chapter focuses on the transition from war to peace, and the implication of the process for long-term stability. For the first time in Burundi’s post-independence history, the international community has engaged in devising ways of finding a lasting solution. Under the leadership of Julius Nyerere, former president of Tanzania, and 180
Nkurunziza and Ngaruko: Burundi
181
Nelson Mandela, former president of South Africa, Burundi politicians have negotiated a new political dispensation to end this cycle of violence. What, then, are the requirements for the restoration of long-term stability? We attempt to answer these questions after a presentation of the history and the root causes of a complex conflict.
2
A brief history of the conflict
2.1 Origin and root causes The origin of the Burundi conflict lies in two major historical events during the colonial period and one development four years after independence.2 Before colonization, being Tutsi or Hutu constituted one of many facets on a malleable stratification grid by which one’s social position was defined (Gaffney, 2000). However, when Belgium colonized Burundi, the colonial power saw Tutsis as being superior to Hutus. Guided by this preconception, they introduced administrative reforms marginalizing Hutus from leadership positions. In 1929, Hutu and Tutsi chiefs each represented about 20 per cent of the number of chiefs, with the Ganwa accounting for the balance, but by 1945, there were about 28 per cent Tutsi chiefs and no Hutu (Lemarchand, 1994). This policy ended a relative ethnic balance in leadership that had been a cornerstone of Burundi’s stability throughout its history. Understandably, this divisive policy polarized Hutus and Tutsis. The Hutus’ feeling of marginalization and victimization would be used later to justify their recourse to violence as a mode of political expression. The second event was the so-called ‘social revolution’ in neighbouring Rwanda in 1959. With an ethnic mix and social and institutional structure similar to that in Burundi, Rwanda experienced a bloody coup in 1959 during which the Hutus overthrew the Tutsi monarchy.3 Scores of Tutsis were killed and many more were forced into exile in Burundi and other neighbouring countries. Following this experience, many among the Hutu elite in Burundi embraced the view that their group’s numerical majority should guarantee them de facto control of state institutions, a view shared by those in power in Rwanda. On the other hand, many Tutsis felt it was legitimate to prevent a similar ‘revolution’ in Burundi by keeping a grip on state institutions, especially security institutions. The many Rwandan refugees in Burundi and elsewhere in the region shared this view. Up to the time of writing, these divergent perceptions and agendas have produced continuing tensions between the two ethnic groups. The third factor relates to the failure of the successive governments to deal with these problems. Instead, governments fuelled more conflicts through their catastrophic leadership. On 28 September 1966, Captain Michel Micombero led a group of young Tutsi officers and civilians, mainly from the southern province of Burur, in a coup that deposed the monarchy.
182
Case Studies: The Horn, East and Central Africa
This capture of power by the Bururi Tutsis spelt the end of a governance system based on a balance of power between the two main groups (Manirakiza, 1992), and power became the prerogative of the new Bururi elite. Coups became the mode of regime change thereafter, even while power remained within the same Bururi group. Presidents from the same commune of Rutovu in Bururi have ruled the country for all but three years since November 1966. There are many ways in which Bururi leadership has fuelled conflict. First, by overthrowing the monarchy in 1966, they plunged the country into political instability by replacing a system that had clear rules governing regime change with one where change was decided through the barrel of a gun. This uncertainty enabled leadership conflicts between the Hutus and Tutsis that had previously been impossible. Second, the replacement of the Ganwa leadership by Tutsis destabilized a governance system that had maintained harmony for centuries. Third, most Hutus and non-Bururi Tutsis were shocked by the irresponsible behaviour of the new masters. They enjoyed every benefit while excluding most Burundians outside their circle. These tensions resulted in the 1972 Hutu genocide, while the assassination of a Hutu president by members of the Bururi-dominated army in 1993 triggered the Tutsi genocide.4 While Hutu and Tutsi extremists may each believe strongly in their right to rule, recent history in both Burundi and Rwanda has proved that neither extreme has been able to deliver long-term stability. In Rwanda, retention of power by the Hutu majority since 1959 led to the 1994 genocide. In Burundi, power in the hands of the Tutsi minority failed to prevent the killings of Tutsis by Hutus and led to two genocides: the 1972 Hutu genocide and the 1993 Tutsi genocide. The lesson is that it is not the identity of the leader that matters but rather the quality of leadership. The leaders manipulated ethnicity to safeguard their personal interests. Holding power in Burundi implies controlling economic spoils through predatory taxation, the creation and looting of state corporations, skewed resource and contract allocations, and appointments to strategic positions (Nkurunziza and Ngaruko, 2002a). The strategy has been to make Hutus, who have never identified with Bururi governments, wrongly believe that power is in the hands of Tutsis rather than the small Bururi group. To paraphrase Monga (1998, p. 52) the ideological mystification of ethnicity has allowed the Bururi elite to deflect anger from their abysmal thirty-five-year record of rule. 2.2 Intensity and duration All five episodes of war followed an identical pattern.5 Believing that their numerical majority gave them a right to leadership, and exasperated by the manoeuvres of extremist Tutsis to exclude them, some Hutus organized themselves to take power by force, and killed thousands of Tutsis before the army counter-attacked. The army engaged in indiscriminate and brutal repression of the Hutus as a group, using the opportunity also to eliminate
Nkurunziza and Ngaruko: Burundi
183
non-Bururi Tutsis who were perceived as threats. What remains in people’s memory about the conflict is the image of a brutal army killing Hutus indiscriminately, but it has been forgotten that the army responded, albeit in a brutal and excessive way, to Hutu killings of innocent Tutsis. The absence of an impartial legal system that identifies and punishes the culprits from both groups perpetuates the problem. In addition to needing to keep power to control its spoils, some influential members of the ruling elite have blood on their hands and fear that losing power will be tantamount to their being condemned. Table 11.1 provides some characteristics of the different episodes. We shall now focus our analysis on the 1993 episode because, strictly speaking, the first four episodes of the conflict were mass killings and repressions rather than civil wars (Ngaruko and Nkurunziza, 2000). By assassinating the first democratically-elected Hutu president on 21 October 1993, the army provided Hutus with an excuse to engage in a horrifying killing spree. Tens of thousands of civilian Tutsis lost their lives in what the United Nations classified as genocide against the Tutsis (United Nations, 1996). Following the usual pattern, the army moved in to put down the revolt by targeting Hutus. From October 1993 to the time of writing, it has been estimated that the total death toll is 250,000. There are many reasons why the 1993 episode lasted longer. First, having learned from past rebellions, the rebels slowed army advances by blocking roads and destroying bridges. It took the army more than a week to reach some areas in the countryside, a period during which almost all Tutsis had been killed and their killers had had time to retreat. This prevented the army from stopping the killings as rapidly as they had in 1965 and 1988. Second, the rebels secured safe havens to which to retreat, mainly in the Democratic Republic of Congo (DRC), and in Tanzania and Rwanda until Table 11.1 Years
1965 1972 1988 1991 1993 on Total
Duration and incidence of civil war in Burundi Duration (months)
Deaths (000s)
Refugees (000s)*
Years from Provinces affected previous period
2 4 2 1
5 200 15 1–3
0 300 50 38
na 6 16 3
107
250
687
2
24 per cent of time
10 per cent of population
–
Muramvya Whole country Ngozi, Kirundo Cibitoke, Bubanza, Bujumbura Whole country
–
Source: Nkurunziza and Ngaruko (2002b). Note: * Number of Burundi refugees in DRC, Rwanda and Tanzania related to specific conflict (UNHCR data). It is defined as the difference between total refugees and number of refugees a year before the crisis.
184
Case Studies: The Horn, East and Central Africa
the first half of 1994. Third, after the president was killed, the Hutus remained in power, creating an armed rebel movement to protect themselves. The Forces pour la Défense de la Démocratie (FDD), the largest rebel group, was created in December 1993, followed by the Conseil National pour la Défense de la Démocratie (CNDD), the political wing of the armed movement, in February 1994 (International Crisis Group, 2003). As Hutu parties controlled the state and security institutions (apart from the army), they were able to mobilize state resources to assist their movement. Consequently, FDD had access to resources no other Burundi rebel group had had previously, thus giving it more fighting capacity. Fourth, by mid-1994, Burundi rebels were joined in Eastern DRC by the ex-Forces Armées Rwandaises (FAR), who had fled with their country’s arsenal. By joining forces with a well-armed professional army, Burundi rebels secured access to training and to relatively sophisticated weaponry. Later, the government in DRC used these rebels to fight its own war against Burundi, Rwanda and Uganda. The war became regional.
3
The Arusha peace negotiations
Section 3.1 discusses the context within which the Arusha peace talks (named after the Tanzanian town where they were launched in 1998) started, and how they were mediated. Section 3.2 discusses the challenges of implementing the accord. 3.1 Context The Arusha peace negotiations were the first comprehensive initiative to end the cycle of violence. When Pierre Buyoya overthrew the elected government on 25 July 1996, the international community condemned him and within one week neighbouring countries had imposed an economic blockade, followed by a suspension of international aid, designed to force Buyoya into negotiations with the opposition, including the armed rebels. After almost two years, the government agreed to participate in the Arusha negotiations in June 1998. Regional heads of state appointed first Julius Nyerere, former president of Tanzania, as mediator, and then former South African president, Nelson Mandela. In addition to the complexity of the conflict, there were major problems with the way the negotiations were conducted. First, there was divergence between the objectives of the different participants. The mediators were aiming implicitly at the restoration of a legal government on the basis of the 1993 elections. In contrast, the government sought the removal of the economic sanctions and the condemnation of rebels. As they were associated with the Front Démocratique Burundi (FRODEBU) and the 1993 Hutu government, the rebels’ objectives were relatively close to those of the mediators. The Tutsis expressed concern about allowing FRODEBU’s return
Nkurunziza and Ngaruko: Burundi
185
to power, given its presumed implication in the 1993 Tutsi genocide. Thus, from the Tutsi perspective, the facilitators and the international community backing the negotiations started with a bias against the new government, a feeling reinforced by the negotiations being imposed following the coup d’é tat. As there were only two blocks, namely the Hutu and Tutsi, though each was highly fragmented,6 any perceived bias against one side was interpreted as support for the other. The second problem was one of representation. The main rebel factions, the FDD led by Jean-Bosco Ndayikengurukiye, and the Front National de Libération-Parti pour la Libération du Peuple Hutu (FNL-PALIPEHUTU), currently led by Agathon Rwasa,7 did not participate in the negotiations; they were first banned by Nyerere, then, when Mandela invited them to join the negotiations, they imposed unacceptable preconditions: the release of political prisoners, the disbanding of cantonment areas, the confinement of the army to barracks, the political recognition of rebel groups, and direct negotiation with representatives of the army (International Crisis Group, 2003). The third problem concerned the venue and the facilitators. Burundi’s government did not consider Tanzania to be a neutral venue, or Nyerere a neutral facilitator as the country sheltered hundreds of thousands of Burundi Hutu refugees – the recruiting pool of rebel movements. Moreover, Tanzania was a safe haven and a base for Burundi rebels (International Crisis Group, 2001; Balancie and de la Grange, 1996). In addition, the composition of the mediation team and their anti-government declarations eroded the Burundi government’s trust in the Tanzanian team. When Mandela took over, there were no fundamental changes to the approach apart from his unsuccessful attempts to include the armed rebel groups.8 Finally, language was a problem. The mediators spoke neither Kirundi nor French, the country’s second language. Although Burundian negotiators understood Kirundi, their facilitators had to rely on back-and-forth translations. This slowed the negotiations, increasing their costs and hence the impatience of those funding them. It is within this context that on 28 August 2000 the negotiators were forced to sign a ‘peace agreement without peace’ (Van Eck, 2002). As the government had always insisted it would not honour any agreement without negotiating a ceasefire first (Sindayigaya, 2002), this was seen as a victory of Hutu over Tutsi parties.9 In reality, it was FRODEBU that made the most concessions in the last hours of negotiation (International Crisis Group, 2001). In view of the conditions under which the accord was signed, there are at least two reasons why its implementation will be difficult. First, the agreement is not incentive-compatible, because the government party will benefit from reneging (Addison and Murshed, 2002). The problem is that the Tutsi ‘losers’ are powerful people in leadership positions who will not relinquish their privileges without a fight. Centralization of economic and
186
Case Studies: The Horn, East and Central Africa
political activity means that a number of those leaving the state sector would suffer disbenefits. Those with blood on their hands have additional reasons to cling to the status quo. The second reason is reputation. The government in Burundi has a poor reputation with regard to conflict resolution. Using Azam’s (1995) framework, it is difficult to trust a government that has traditionally opted for force over cooperation and power sharing, but then becomes willing to do so after failing to defeat the rebels. As for the rebels’ lack of reputation,10 no party is able to commit credibly to peaceful behaviour, so peace cannot be maintained for long. In such a context, action by an external agent is needed to enforce both parties’ commitment, although external agents tend to prolong wars (Elbadawi and Sambanis, 2000). 3.2 Implementation of the peace accord When nineteen parties signed a peace agreement on 28 August 2000 in a high-profile ceremony attended by heads of state, including US president, Bill Clinton, most people knew the signatories had not really agreed to peace. Tired of the two-year-long wrangling among the negotiators, the mediators decided to force them to sign a text that, because of last minute modifications, most of them had not even read (Sindayigaya, 2002). However, while the peace accord did not deliver a ceasefire, it did provide a minimum political agreement, particularly with regard to the origins of the conflict, the path to reconstruction, and the role and composition of the army and its relationship to civil power (International Crisis Group, 2001). The document recognized that acts of genocide, war crimes and other crimes against humanity had been perpetrated in Burundi for decades, and it provided for an inquiry by an international judicial commission on human rights abuses since independence. The accord also recognized the need to establish a new political, economic, social, cultural and judicial order based on the principles of democracy and the rule of law. It included the commitment to fight all forms of discrimination regarding access to education, the administration of justice and the management of public services, and the adoption of a multi-party political system. It provided for the establishment of an independent electoral commission, a senate comprising two delegates from each province, the elimination of ethnic dominance in the army, and the need to reconsider the cases of political prisoners, and it allowed for the involvement of an international peacekeeping force to help in stabilizing the country. Not surprisingly, the implementation of the accord raised many difficulties. First, its Implementation and Monitoring Committee was ineffective. Second, violence increased as rebels sought to show that no agreement was viable without them, which undermined the very essence of the peace agreement. Third, the resurgence of violence hardened political positions, and put immense pressure on the government. As a sign of this general political volatility, the government faced two attempted coups – in April and July
Nkurunziza and Ngaruko: Burundi
187
2001. As a result, the government made the implementation of the agreement conditional on the prior signature of a ceasefire. Realizing that the agreement could not be implemented without the cooperation of those in power, Mandela allowed the incumbent president to lead the first half of a three-year transitional period, with a Hutu deputy president. During the second period, the main Hutu opposition party would take over, with a Tutsi as deputy. This decision led to the first steps in the implementation of the accord on 1 November 2001, some fourteen months after its signature. The government guaranteed the safety of Hutu leaders coming back from exile by giving them provisional immunity and accepting South African troops to protect them. Continued conflict and the rebels’ uncompromising attitude have, however, derailed the implementation of the accord. Despite speculation that President Buyoya might refuse to leave the presidency at the end of his term, he handed over in 2003 as agreed (International Crisis Group; 2003).
4
Restoring sustainable peace: the role of the international community
This section argues that the first step towards the restoration of peace should be the adoption of a ceasefire agreement, and the second step should be a long-term commitment by Burundian leaders and the international community to creating a just and equitable society. 4.1 Ceasefire negotiations Despite threats of sanctions by regional states, and fruitless meetings with government representatives, at the time of writing, the two main rebel groups – the FDD led by Peter Nkurunziza and the FNL-PALIPEHUTU led by Agathon Rwasa, have failed to agree on a ceasefire. They have pushed for a different negotiating forum to discuss the reform of the army, which is their main demand. Only two minor groups – the CNDD-FDD led by Jean-Bosco Ndayikengurukiye, and the FNL-PALIPEHUTU wing led by Alain Mugabarabona, both of which are at the time of writing almost inactive within Burundi – have signed a ceasefire with the government. Their low profiles suggest that the impact of this ceasefire will be limited. As we write, the mediators are putting pressure on the major rebel groups to arrive at a ceasefire agreement with all the rebels. The government’s approach is clear: it is ready to negotiate separately with anyone fighting before reaching a general ceasefire agreement involving all armed groups. This strategy raises certain questions. First, focusing on rebel groups while ignoring unarmed opposition within the country neglects a potential source of future problems. So far, the mediators and the government have conducted negotiations as if the army and Tutsi militias will de facto comply with the agreement, but the coup attempts and the aggressive opposition to
188
Case Studies: The Horn, East and Central Africa
negotiations by radical Tutsi groups suggest the reverse. These groups are comforted by recent developments in the region. With the withdrawal of the Rwandan and Burundian armies from DRC in late 2002, the Congolese government pledged to withdraw its support to the Hutu rebels. Many in Burundi assume that the move will weaken the rebellion, forcing it to accept negotiations without preconditions. The FDD branch controlled by Peter Nkurunziza agreed to negotiations in December 2002. In November 2003 an agreement was signed to end the civil war, and Nkurunziza was included in the government as Minister for Good Governance. The outcome, however, remains uncertain, and Tutsi hardliners believe that the government should not make any concessions to the rebels. Second, the government of transition current at the time of writing unites artificially two groups with divergent objectives. The Hutu-dominated ‘G7’ seeks to maximize institutional reform, especially security institutions, while the Tutsi-dominated ‘G10’ seeks to minimize it. Nevertheless, the government has endured, to date, because of two factors: the presence of South African troops in charge of the security of the Hutu politicians (illustrating Addison and Murshed’s (2002) claim that incentive compatibility problems can be overcome only through the external enforcement of agreements) and the explicit commitment made by the army to Mandela to accept the government as long as he appointed President Buyoya, a former army major, to lead the first half of the transition period. Each of these groups has its own alliances in the region. This could be a threat to peace, since each of these governments pushes for its own interests in backing these groups. Rwanda would prefer a Tutsi government, given ethnic affinities and the recent history of ethnic conflict in the region. For similar reasons, Tanzania would favour a Hutu-dominated government (International Crisis Group, 2002b). There is a risk of escalation: the reported involvement of Rwandan troops alongside the Burundian army and the use of Hutu rebel groups by DRC and Tanzania to put pressure on the army gives cause for concern (International Crisis Group, 2002). In this context, to prevent the exploitation of the conflict by Burundi’s neighbours, the international community needs to be more involved. So far, US$830 million has been committed to reconstruction. Structural aid cannot be disbursed, because instability has resulted in an absence of projects. If used adequately, however, external aid may be a decisive lever to force concessions. Nevertheless, the international community should recognize that none of its decisions is neutral. Each camp seeks to exploit any flaw in the mediation process to promote its cause. 4.2 Questions of justice Once a viable and balanced agreement is signed, justice should take its course. The trial of individuals responsible for ethnic violence will be required to break the cycle of impunity. Recently, the government has created a commission of enquiry into genocide and crimes against humanity similar to
Nkurunziza and Ngaruko: Burundi
189
the Truth and Reconciliation Commission in South Africa. However, the work of this commission will face tremendous problems: • the problem of credibility and independence: the commission was appointed by the president, who himself has a tainted reputation, which casts doubt over its credibility; • difficulties probing senior members of the army in the absence of army reforms: the army still hosts many of the authors of repressions committed over the years; • the commission will need to surmount the negative solidarity among the members of each group, both the army and the rebels: solidarity has crossed political and ethnic lines, and opinion in Burundi holds that both Hutus and Tutsis sharing power now constitutes an alliance between criminals to protect each other’s interests;11 and • unlike South Africa, where the truth and reconciliation model was used, killers and surviving victims live together in Burundi, so day-to-day encounters are more likely to degenerate into violent conflict more frequently. In South Africa, the members of the most powerful party from which concessions were logically expected had accumulated private wealth and did not need to cling to the state as a source of income. The high opportunity cost of war and the benefits from the removal of economic sanctions by the international community were strong incentives to accept a compromise. Such incentives are weak in the case of the Burundi. On the other hand, proof of guilt for these crimes through more traditional enquiries by judicial authorities would be costly to establish, as the example of Rwanda has shown (Smith, 1999). Moreover, punishing all those found guilty is not feasible, and may be harmful for the survivors (Minow, 1998), and a sustainable solution to Burundi’s violent conflicts needs to weaken the belligerents. However, the possibility of amnesty is highly controversial in Burundi. In August 1999, several individuals testified in order ‘to purge souls and to heal the nation’12 but granting a blanket amnesty to unknown criminals would be counter-productive (Human Rights Watch, 2000). It seems more appropriate to clarify responsibility for the crimes committed over the years before talking of amnesty. The ‘generalized prisoner’s dilemma’ approach proposed by Ngaruko and Nkurunziza (2000) analyses incentives that could meet these requirements in terms of selective pardon and punishment, as well as a cheap, efficient and quick system of enquiries. The generalized prisoner’s dilemma is based on the assumption that, as the crimes are political and often collective, the questioning of one testifier can shed light on the responsibility of accomplices or silent partners. In contrast to common crime, political crime is committed in the name of a group. Instead of using prosecutors to investigate crimes, the international community should assist in creating chambers to record the testimonies of each social corps (such as the army, civil society, politicians), assuming that
190
Case Studies: The Horn, East and Central Africa
credible international officers guarantee the confidentiality of testimonies. A criminal who agrees to testify is guaranteed a partial amnesty calculated as a proportion of the penalty if judged in the classic way, or if refusing to testify. For those testifying, death and whole-life sentences are excluded, and the same criminal can testify many times, each testimony being registered and scored separately. Punishment is then calculated, factoring in the total number of testimonies registered at the end of the period of registration and the position of the criminal’s testimony on the chronological list of testifiers. By making coordination between accomplices difficult and even counterproductive, the procedure puts criminals and their accomplices in the situation known as the prisoner’s dilemma. In deciding whether or not to testify, a criminal ignores what other testifiers have said about past collective actions of the group to lessen their own penalties, while being encouraged to denounce accomplices to lessen his own penalty. Thus, suspicion spreads rapidly throughout the groups of criminals, penetrating otherwise hermetic criminal groups. At the end of the period of testimony, the process of enquiry by an international commission and judgment by an international court can begin. The principle underlying this procedure can serve both as a framework for collecting information on past crimes and dismantling subversive organizations, and can also help to prosecute criminals and prevent their coordination in the future.
5
Conclusion
On 28 August 2000, Burundi political parties signed a peace accord with the government, expected to end the long conflict spanning four decades. Numerous further accords followed, leading to the formation of a transitional government in October 2001. However, despite subsequent accords and ceasefires, violence has continued, including an upsurge in July 2002 which delayed peace talks, and a rebel assault on Bujumbura in July 2003. Rebel groups have attempted to show that no peace accord is viable without their participation, using violence not only as a military tactic, but also for political ends. While the political accords are important in the quest for peace, to be viable they must be complemented by an agreement on security. The continuing transition to peace will be difficult for many reasons. First, the attitude of the army, the real power holder in Burundi, will be crucial but remains unpredictable. Second, the problems of security in Burundi overlap with the interests of powerful domestic groups whose cooperation will be essential for peace. In the medium term, a credible and impartial system of justice to support the political accords and the peace programme is essential. The international community has a central role to play in each of these stages. It should prevent the governments of neighbouring countries from intervening, thus destabilizing and aggravating the conflict. During the period of justice and reconciliation, the international community would be
Nkurunziza and Ngaruko: Burundi
191
required to help the country put in place the prisoner’s-dilemma-based system of amnesty and justice. It could use the disbursement of nonhumanitarian aid to encourage politicians to accept this scheme. What is at stake is not just the support of a ceasefire. As noted earlier, most of Burundi’s history has been a long period of peaceful coexistence by different groups. From the foundation of the kingdom of Burundi at least four centuries ago to the early 1960s, Hutus and Tutsis lived together peacefully. They shared a normative system of governance and a political system that allowed divergent interests to be reconciled, while social and economic activities allowed Hutus and Tutsis to identify themselves as members of one nation. There is no need to reinvent the wheel. Burundians should draw on their distant past for inspiration. However, only if justice heals the wounds of the recent past will Burundians be able to refocus on their common future. The impartial administration of justice should identify those few Burundians who have compromised the future of the country in order to satisfy their selfish interests, and disqualify them from participating in any future political activity. This would ensure that Hutus and Tutsis return to living peacefully together, as they did for centuries. Notes 1 There are two more minor groups, the Twa and the Ganwa. The Twa are thought to represent 1 per cent of the population, while the Ganwa, the group representing those of princely lineage, are fewer. Unless there is a specific reference to either of these two groups, the analysis focuses on the Hutu and Tutsi groups. Ngaruko and Nkurunziza (2000) discuss why the concept of ethnicity in the case of Burundi needs to be taken with caution. 2 For a detailed historical overview, see Ngaruko and Nkurunziza (2000). 3 Unlike in Burundi, where political power was in the hands of the king and the Ganwa group, the Tutsi group controlled power in Rwanda. 4 In April 1972, a group of armed Hutus killed thousands of civilian Tutsis, mainly in the south. Instead of arresting and judging them, the Bururi Tutsi-dominated army opted to execute summarily 200,000 Hutus in high office, businessmen, civil servants, and university and secondary school students. Many non-Bururi Tutsis were also eliminated if they were perceived as threatening the Bururi grip on power. As no one has ever been prosecuted from either group for the 1972 events, deep scars remain and have coloured subsequent events. 5 The 1993 war is an exception as it was triggered by the killing of the first democratically elected president, a Hutu, by members of the Bururi Tutsi-dominated army. 6 The two sides later became known as Group of Ten (G10), representing mainly Tutsi parties, and Group of Seven (G7), representing mainly Hutu parties. In reality, there are some disagreements within these blocks, making the negotiating landscape highly fragmented. 7 These groups later had internal splits. Within the FDD, one wing was controlled by Jean-Bosco Ndayikengurukiye, and the other – the most active inside Burundi – by Peter Nkurunziza. The FNL-PALIPEHUTU group split into the more radical Agathon Rwasa faction and a less radical faction led by Dr Alain Mugabarabona.
192
Case Studies: The Horn, East and Central Africa
8 There has always been controversy regarding who would have been the most appropriate mediator. Informally, some critics argue that Nyerere’s and Mandela’s personal experiences greatly influenced their views. The critics contend that the two facilitators identified themselves with the Hutus, whom they saw as freedom fighters in a system paralleling South Africa’s apartheid. Knowing Nyerere’s historical contribution to freedom movements in east and southern Africa, including Mandela’s ANC, and Mandela’s life under apartheid, some believe that understanding Burundi’s conflict from this matrix greatly shaped the outcome of the negotiations. 9 Sindayigaya (2002) has an interesting account of the signing of the accord. During the night preceding the signing day, a secret closed-door meeting was held between representatives of the government and FRODEBU, who spent the night changing the document to be signed without the knowledge of the other negotiating parties. 10 Note the difference between bad reputation and lack of reputation. The government has a bad reputation given the way it handled past crises. Rebellion leaders have no reputation because they have not been in power to reveal their leadership type, apart from the three months between July and October 1993 during which they may have organized the genocide against the Tutsi. Even without considering this short spell at the country’s helm, critics refer to the Rwandan case to argue that the (Hutu) rebellion also had a bad leadership record. 11 The Hutus in power are accused of masterminding the 1993 genocide against the Tutsis, while the Tutsis in power (the army in particular) are accused of army repressions and the killing of the democratically elected (Hutu) president in 1993. 12 See Vick (1999).
References Addison, T. and M. Murshed (2002) ‘Credibility and Reputation in Peacemaking’, Journal of Peace Research, vol. 39, no. 4, pp. 487–501. Azam, J.-P. (1995) ‘How to Pay for Peace? A Theoretical Framework with Reference to African Countries’, Public Choice, vol. 83, pp. 173–84. Balancie, J.-M. and A. de la Grange (1996) Mondes Rebelles: Acteurs, Conflits et Violences Politiques (Paris: Michalon). Elbadawi, I. and N. Sambanis (2000) ‘External Interventions and the Duration of Civil Wars’, Paper presented at the conference on ‘The Economics and Politics of Civil Conflicts’, March 18–19 Princeton, Princeton University, NJ. Gaffney, P. D. (2000) ‘Burundi: The Long Sombre Shadow on Ethnic Instability’, in E. W. Nafziger, F. Stewart, R. Väyrynen (eds), War, Hunger and Displacement, Vol. 2 (Oxford University Press). Human Rights Watch (2000) World Report (New York, NY: Human Rights Watch). International Crisis Group (2001) ‘Burundi: Breaking the Deadlock: The Urgent Need for a New Negotiating Framework’, Africa Report, No. 29 (Brussels: International Crisis Group). International Crisis Group (2002) ‘Après Six Mois de Transition au Burundi: Poursuivre la Guerre ou Gagner la Paix?’, Africa Report, no. 46 (Brussels: International Crisis Group). International Crisis Group (2003) ‘The Burundi Rebellion and the Ceasefire Negotiations’, Africa Briefing (Brussels: International Crisis Group). Lemarchand, R. (1994) Burundi: Ethnic Conflict and Genocide (Cambridge: Cambridge University Press).
Nkurunziza and Ngaruko: Burundi
193
Manirakiza, M. (1992) Burundi: de la Révolution au Régionalisme 1966–1976 (Brussels: Le Mât de Misaine) Minow, M. (1998) Between Vengeance and Forgiveness: Facing History After Genocide and Mass Violence (Boston, Mass.: Beacon Press). Monga, C. (1998) The Anthropology of Anger: Civil Society and Democracy in Africa (Boulder, Col.: Lynne Rienner ). Ngaruko, F. and J. Nkurunziza (2000) ‘An Economic Interpretation of Conflict in Burundi’, Journal of African Economies, vol. 9, no. 3, pp. 370–409. Nkurunziza, J. (2001) ‘Exchange Rate Policy and the Parallel Market for Foreign Currency in Burundi’, African Journal of Economic Policy, vol. 8, no. 1, pp. 69–121. Nkurunziza, J. and F. Ngaruko (2002a) ‘Explaining Economic Growth in Burundi: 1960–2000’, Working Paper WPS/2002–03, Centre for the Study of African Economies, University of Oxford. Nkurunziza, J. and F. Ngaruko (2002b) ‘Civil War and its Duration in Burundi’, paper presented at the conference on ‘The Economics of Political and Criminal Violence’ Yale University, New Haven, Conn. Sindayigaya, J.-M. (2002) La Saga d’Arusha (Brussels: Association de Réflexion et d’Information sur le Burundi). Smith, S. (1999) ‘La Faute du TPR’, Libération, 11 November. United Nations (1996) Rapport de la Commission d’Enquête Internationale chargée d’établir les faits concernant l’assassinat du président du Burundi, le 21 octobre 1993, ainsi que les massacres qui ont suivi, Document S/1996/682 (New York: United Nations). Van Eck, J. (2001) ‘Polarisation of Parties Into “Win Power” and “Keep Power” Camps Threatens the Collapse of Peace Process’, Burundi Report no. 2001/1 (Cape Town: Centre for Conflict Resolution). Van Eck, J. (2002) ‘The “Dar” Talks: No Chance of Progress Without Inclusivity’, Burundi Update (Cape Town: Centre for Conflict Resolution). Vick, K. (1999) ‘In Burundi, Confessions of Genocide’, Washington Post, 30 August.
This page intentionally left blank
Part IV Case Studies: Southern Africa
This page intentionally left blank
12 The Political Economy of Post-Conflict Economic Recovery: Making Peace Substantial in Mozambique* Laudemiro Francisco and Edwin Connerley
1
Introduction
Following a sixteen-year civil war, Mozambique, at the time of writing, is enjoying a period of relative peace and stability. The General Peace Agreement (GPA), signed by the government (Frente de Libertação de Moçambique – Frelimo) and the rebels (Resistência Nacional Moçambicana – Renamo) on 4 October 1992 was sponsored by several international actors whose involvement was crucial, making possible the demobilization and restoration of ex-combatants to civilian life; the formation of a new army; and the transformation of Renamo into a political party. National elections followed in October 1994 and December 1999, both won by President Joaquim Chissano and his party, Frelimo – which had been in power since independence. Renamo remains the major opposition to the government. Internationally, Mozambique stands out as an example of a successful transition from a cycle of war and destruction to one of peace and progress. It not only avoided sliding back into war, but also established democracy and impressive economic growth rates. Since the end of the war, the economy has grown at a relatively steady pace approaching 10 per cent a year. Income per capita increased from US$90 at the end of the war to US$238 in 1999 (Instituto Nacional de Estatistica, 1999). However, Mozambique still faces serious challenges. Official and popular discourse on the state of the nation diverges, and many believe that the pace and content of the reforms is controlled by the governing elite for its own benefit. Are these potential causes of the resurgence of war? What do they portend for future social and political stability? What helps to keep the peace? This chapter examines the main lessons from the Mozambican experience. 197
198
2
Case Studies: Southern Africa
Understanding the civil war
2.1 Risk factors Recent studies of civil war identify a number of crucial risk factors – incentives for conflict onset, duration and termination. The most widely known, Collier and Hoeffler (2000), identify the following factors associated with the onset and duration of civil wars: a low level and rate of growth of income; dependence on primary commodity exports; and a low level of male secondary school enrolment. Other relevant factors include ethnic dominance and fragmentation; a history of previous conflict; and the existence of large geographic areas (including mountains and forests) that are not incorporated effectively into the national polity and economy (Collier and Hoeffler, 2000). Table 12.1 illustrates these factors. According to Collier and Hoeffler’s model, between 1995 and 1999, Mozambique still presented relatively high levels of a substantial number of key conflict risk factors. Secondary school enrolment for school-age males averaged only 9 per cent between 1995–9. This resulted in a predicted risk of renewed conflict of 28.8 per cent. Between 1995 and 1999, although there were substantial improvements in Mozambique’s economic outlook, the estimated probability of war remained very high at 23.5 per cent. These probabilities far exceed the Collier and Hoeffler sample mean of countries without civil wars (6.7 per cent). The model assigns relatively high political openness to the Mozambican regime. A recent public opinion survey indicates Table 12.1
Mozambique risk factors Population Male gross (millions) secondary school enrolment ratio
1960–4 1965–9 1970–4 1975–9 1980–4 1985–9 1990–4 1995–9 Sample mean (with war) Sample mean (no war)
7.46 8.34 9.40 10.50 12.10 13.50 14.20 17.40
Average annual per capita GDP growth (%)
Ratio of Ethnic Political primary domination openness commodity exports/GDP
30.00 3.00 5.00 4.00 8.00 10.00 10.00 9.00
na 1.85 3.37 −4.73 −4.95 −4.18 0.29 0.77
0.099 0.099 0.099 0.099 0.099 0.099 0.099 0.102
1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
na na na 0.25 0.00 0.00 1.20 6.00
30.30
−0.23
0.149
0.45
1.82
44.39
1.74
0.169
0.47
4.07
na = not available Source: Collier and Hoeffler (2000).
Francisco and Connerley: Mozambique
199
growing political intolerance (CEP, 2001). Combined with an extensive know-how for waging war, these probabilities deserve attention. Although several measurements of key risk factors for conflict were relatively high at the time civil war broke out in Mozambique, external geo-political actors exploited local grievances to influence the onset and duration of the war. A second widely applied analytic model (Elbadawi and Sambanis, 2000) incorporates geo-political factors, particularly external intervention, which reduces the costs of waging war for small rebel groups. 2.2 Causes and nature of the conflict The causes and nature of the civil war in Mozambique are a matter of dispute. Both Frelimo and Renamo had strong external supporters, tending to confound analyses of the nature and dynamics of the conflict, and resulting in opposing interpretations of the causes of the war and ensuing attempts to resolve it. Analysts such as Minter (1994) emphasize destabilization by external forces, first by the ailing Rhodesian regime and then by South Africa’s apartheid regime, backed by conservative Western forces opposing communism, seeking to exact retribution from Mozambique for its active role within the Frontline States regional diplomatic coalition, and to weaken FRELIMO’s ability to export socialism. The destabilization perspective has been criticized increasingly for ignoring internal trigger factors, namely FRELIMO’s legitimacy deficit,1 exacerbated by its unpopular policies. Others have focused on internal factors explaining the war’s initiation and duration, with RENAMO enjoying an apparently high degree of popular acceptance, particularly in rural areas. This analysis holds that RENAMO is a product of FRELIMO’s politics of intolerance, both in the definition of a socialist orientation for the liberation movement before independence and in the project of building a modern socialist state after 1975 (Geffray, 1991; Cabrita, 2001). Based on its assumed historical right to rule, FRELIMO imposed development strategies that caused disruption to the peasant economy. These strategies, based on socialist modernization tenets, aimed to create a Mozambican nation by suppressing traditional and religious practices, perceived as impediments to progress, and replacing them with a new and progressive mentality. The strategy was two-pronged: widespread and indiscriminate nationalization of private property; and the forced ‘villagization’ and collectivization of peasant livelihoods (Geffray, 1991; Cabrita, 2001). FRELIMO’s policies created opposition, which the new state met with repression, imprisoning people or sending them to re-education camps, thus breeding large pools of domestic discontent and a politicized diaspora of both African and Portuguese descent. Most became the bearers of RENAMO’s cause abroad (interviews, Maputo, March 2001; Luis Serapião, Washington, DC, January 2002).2 These grievances defined the context contributing to the duration and intensity of the conflict. However, Renamo seemed unable
200
Case Studies: Southern Africa
to develop a rhetoric capitalizing on these grievances until the late 1980s. This influenced the dynamics of the war, which became highly disruptive and predatory, including indiscriminate violence against civilians and destruction of the infrastructure. Although external intervention was very high, it is short-sighted to argue that the primary causes of Mozambique’s conflict were exogenous. Domestic and external factors were equally important to a more convincing explanation of the war, including RENAMO’s enduring control of sizeable rural areas. 2.3 Consequences of the civil war The war had a destructive impact on Mozambique and its people, disrupting basic social values and wrecking the infrastructure: 150 villages and localities were destroyed, half the country’s roads, half the health posts, 1,800 schools and 1,500 rural shops. It also caused an estimated 4.5 million people to be displaced within the country, and more than 1.5 million refugees. Teachers and health care personnel were killed, kidnapped or forced to flee. The total material damage to Mozambique’s economy was estimated at over US$20 billion. By the end of the war, 60 per cent of the population lived in extreme destitution (FDC, 1997, p. 8; Hanlon, 1996, p. 15). The impact on the education system was enormous. Progress achieved in basic education through an expansion programme and compulsory education in the first years of independence was nullified. The situation deteriorated further with the subsequent impact of the economic restructuring programme, decreasing basic education spending from 12 per cent to 4 per cent of the state budget. The end of the war allowed the rebuilding and expansion of the education system, leading to gross admission rates of 82 per cent in 1999, up from 63 per cent in 1994 (UNDP, 2000a, pp. 29–64; Diogo, 2002, p. 13). The war enabled the military to dominate the Mozambican political process, and several ministerial positions were held by generals. The end of the war brought about the formal disengagement of the military from politics, and the creation of a multi-party political system, including elected local government structures for a limited number of areas. However, the Mozambican political system remains highly centralized. Another consequence was the entrenched antagonism between the warring parties, which continues to cause polarization into the twenty-first century. While FRELIMO tries to monopolize power, RENAMO shows a continued interest in power-sharing, clinging to its assertion of a ‘right to rule’ in the areas under its control during the war, and where it won the majority of the vote. If the core issues of the current Mozambican political process concern power-sharing and other unfinished business of the peace process, why, then, did the parties settle? What were their expectations and anticipated approaches to post-war political interaction?
Francisco and Connerley: Mozambique
3
201
The peace process: domestic and external factors
Domestic and external social, economic and political factors combined to force the parties to settle. A deepening economic crisis, drought and growing anti-war sentiment affected the parties’ will and ability to fight. These were among the main factors drawing the parties to the negotiating table. Other strategic considerations also forced the parties to negotiate. RENAMO was keen because it needed a genuine opportunity to voice its grievances to a larger and previously inaccessible international audience – thus materializing its long-sought recognition as a legitimate political force. This was important given the changing regional and international conditions associated with the end of the Cold War. It became clear to Renamo that times had changed: South Africa’s domestic and foreign policies were demilitarized and the agenda of RENAMO’s main source of external support, the international conservative movement, changed dramatically (interview with L. Serapião, Washington, DC, January 2002). FRELIMO’s bid for peace was a strategic move geared towards regime survival. Voices demanding the end of the war increased during debates over a new constitution. More pressing was the large toll the war was taking on an economy already in crisis. Measures introduced to address the economic crisis3 forced changes in the political system, resulting in incentives for peace as the governing elite recognized the personal advantage of a peaceful Mozambique. A reduction in public spending and an economic revival via a marketoriented economy were the basis of the reform programme. This strategy forced the government to make substantial cuts in spending in subsequent years (see Table 12.2). The war had made it impossible for the government to generate domestically the resources it needed to operate and meet the increased costs of a widening war. Foreign aid had at times met more than 60 per cent of the government’s budget requirements. By the war’s end, the government’s accumulated external debt totalled over US$5 billion. These factors put FRELIMO under pressure to accept non-military solutions to end the war (Landau, 1998; Gonçalves, 1998). Table 12.2 shows that, on average, between 1985 and 1987, the government spent about 26 per cent of its budget on defence. These years coincide with major military offensives against RENAMO. With the introduction of IMFmonitored reforms, defence spending declined as a share of GDP. Drought and military fatigue were other important factors in the decision of both parties to end the war in the early 1990s. The failure of peasant agriculture affected both sides. FRELIMO forces had to confine their operations to areas that could be re-supplied conveniently from external sources, while RENAMO combatants had to cover longer distances to find food. Morale was low and both sides faced increasing rates of desertion.
202
Case Studies: Southern Africa
Table 12.2
Government expenditure: military, health and education, 1985–951
Year
Total government expenditure
Total expenditure
Defence as percentage of total government expenditure
Total expenditure on education and health2
Education and health as percentage of total government expenditure
1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995
1.69 2.20 6.85 12.28 20.13 29.48 40.78 61.77 96.60 174.34 219.66
0.47 0.53 1.77 2.48 4.36 5.79 7.57 11.03 17.74 32.43 22.21
27.70 24.00 25.90 20.10 21.60 19.60 17.80 18.50 18.30 18.50 10.10
0.37 0.42 1.01 1.77 2.74 4.09 na 13.40 19.57 39.62 48.85
21.90 18.90 14.50 14.80 13.60 13.80 na 21.60 na 22.70 22.20
Notes: 1Approximate figures, in US$m, calculated from the original figures in Mozambican currency: exchange rate US$1 = MZM 23,500 (May 2002); 2 Education and health includes salaries, goods and services; na = not available. Source: Adapted from Lundin et al. (2000, p. 179).
These factors provided incentives for extra-regional actors to invest in a peace process, achieved after a fierce battle for recognition between the parties was settled with skilful help from the Christian churches. The Catholic Church issued numerous public pastoral letters and appeals for peace and reconciliation at a time when the government refused to recognize RENAMO. The government tried to end the war under its own terms, luring RENAMO members to defect in exchange for an amnesty. The adoption of a new constitution in 1990 allowed the churches to capitalize on popular sentiment against the continuation of the war, forcing the parties to seek a political settlement (Sengulane and Gonçalves, 1998). Official diplomacy was crucial to achieving peace, just as external support had been in sustaining war. With changing global and regional contexts, the major powers and regional states disengaged from the conflict, allowing private actors to play a larger role as peace-brokers.4 The Rome-based Catholic Community of Sant’Egidio, supported by the Italian government, played a major mediation role, leading to the GPA in 1992. Yet only official diplomacy could provide guarantees for the peace process. The Mozambican case corroborates the analytical assumption that, in the presence of deep, conflict-related hostility and lack of domestic capacity, the success of the peace process depends on full support and effective monitoring by the international community (Doyle and Sambanis, 2000). The UN established
Francisco and Connerley: Mozambique
203
its Operation in Mozambique (UNOMOZ)5 to supervise and monitor the implementation of the GPA and provide resources and expertise to enable the parties to meet their obligations to organize elections, create a new and unified army and, above all, to demobilize and reintegrate former combatants. The government had anticipated the end of the war and prepared a reconstruction plan, presented to the donors in late 1992 immediately after the war. A donors’ conference, to raise resources for the peace process and post-conflict reconstruction, resulted in donor pledges of US$318.68 million, considerably less than the US$402 million that the government had requested initially for the first phase of post-conflict recovery. Donors pressed for the inclusion of rebel combatants in the post-demobilization reintegration programme (Ball and Barnes, 2000, pp. 176–7). The initial stages of the transition were very delicate. Both parties still exhibited distrust and animosity, which slowed the demobilization process, as they tried to hold on to their best troops in case the process derailed. RENAMO’s leaders were also dissatisfied with governmental control of funding. They perceived that the terms of peace were forced upon Renamo by an international community interested in ending the war, but not necessarily in addressing the internal causes of conflict (interview with L. Serapião, Washington, DC, January 2002). Hence, RENAMO leaders set out to realize their peace dividends, demanding funds promised for RENAMO’s transformation into a political party.6 The UN Special Representative recognized promptly how crucial this was for the success of the peace process, and advocated the provision of ‘democracy aid’ by the international community. Two trust funds were established: one to accommodate RENAMO’s demands and the other for registered political parties participating in the pending elections (Ball and Barnes, 2000). Issues concerning compensation continued, because only a few people, such as party leaders, had access to the trust funds, and soldiers perceived that their leaders, in Maputo, were benefiting from peace dividends at their expense.7 The demobilization process was complex and delicate, as many former combatants on both sides refused their leaders’ plans to retain them on standby and demanded rapid demobilization, refusing requests to join the new army.8 UNOMOZ managed this situation skilfully by linking the electoral to the demobilization process, ensuring that elections would take place only after the demobilization process was completed. International assistance was invaluable in supporting the political transition and ensuring economic recovery. Many high-priority activities were funded, particularly during the most delicate years of the peace process, thus reducing any risk of further conflict. Former combatants were demobilized and reintegrated into civilian life; refugees were repatriated; and mines were cleared. There was a transition to an elected government, including RENAMO’s transformation. The majority of the funds, roughly two-thirds, were allocated
204
Case Studies: Southern Africa
to economic revitalization activities promoting incentives for peace (Ball and Barnes, 2000, p. 171). The emphasis on economic revival constrained the reintegration of ex-combatants, limiting their re-socialization. It was assumed that a sixteenmonth subsistence allowance, a tool kit and short training programmes would result in their self-sufficiency (World Bank, 1997, p. 5). However reintegration projects were discontinued as demobilized soldiers became a lesser priority to the government and donors. Pension payments to some of the demobilized remained unpaid.9 Many felt abandoned, resented the perceived lack of appreciation for their sacrifices, and eventually expressed their frustration vigorously.10 Nevertheless, given the extent of the destruction and disruption produced by the war, reintegration was in the main successful (Ball and Barnes, 2000).
4
Post-conflict political and economic processes
4.1 National and regional views The GPA left Mozambique with unfinished business that may be a long-term obstacle to political stability and economic progress. Both FRELIMO and RENAMO were pressured by their international sponsors to end the war. The GPA accomplished its objectives but failed to address the grievances that had motivated the war. Although, at the time of writing peace continues, war is now being conducted via political means. RENAMO’s reluctance to accept the results of the 1994 and 1999 elections indicate that several sources of conflict remain unresolved. Largely relating to issues of governance in the context of national development and the distribution of economic opportunities across the country’s three regions, tensions are exacerbated by the availability of weapons.11 There is an increasing perception that the governing elite uses the state to reward political supporters and marginalize opponents, and that the content and pace of the post-conflict political process is controlled and manipulated by the governing elite, which lacks the political will to liberalize the political system. The Mozambican political system is highly centralized. Coupled with a ticket-voting, winner-takes-all electoral system, checks on political authority are limited, breeding political intolerance. With the exception of the parliament, the opposition has little access to power; RENAMO’s poor organization allows FRELIMO to exclude them from national and local government, including areas (particularly central Mozambique) where RENAMO won more than 90 per cent of the vote in the national elections. This exclusion engenders perceptions of government favouritism towards its supporters (the south), prompting the belief of bias, fuelled by unbalanced regional development. The centre and north comprise, territorially, 79 per cent of Mozambique and about 75 per cent of the population but, as Table 12.3
Francisco and Connerley: Mozambique
205
Table 12.3 GDP (1998), average consumption and poverty estimates, by region and sub-region Region
Population (%) GDP (%) Average real Poverty Depth of Distribution monthly incidence poverty (%) of poverty (%) consumption (%) ($US)
Rural Urban
79.70 20.30
na na
6.40 8.60
71.25 62.01
29.92 26.67
81.80 18.20
North Centre South (inc. Maputo) South (exc. Maputo)
32.50 42.60
21.00 31.40
7.10 6.00
66.28 73.81
26.62 32.71
31.00 45.30
24.90
47.60
7.80
65.80
26.80
23.60
18.80
13.30
6.90
71.67
30.17
19.40
100.00
100.00
6.80
69.37
29.26
100.00
National
Notes: Approximate figures calculated from the original figures in Mozambican currency: exchange rate US$1 = MZM23,500 (May 2002); na = not available. Sources: Compiled from Government of Mozambique (2001); and UNDP (2000b, p. 39).
shows, produce only about 52 per cent of national wealth, and exhibit higher poverty levels. The south, with 25 per cent of the population, accounts for approximately 48 per cent of the GDP. However, these discrepancies are, in some respects, more apparent than real. There is a tendency to engage in a ‘politics of amalgamation’ and treat the south as a single homogeneous unit. The economic performance of the country’s capital, Maputo City, increases the perception of favouritism towards the south. Six per cent of the population live in Maputo, producing 34 per cent of the country’s GDP, a per capita GDP six times higher than the national average (UNDP, 2000b, pp. 39–42). Excluding Maputo City, the averages for the three regions are very similar, particularly regarding poverty (Government of Mozambique, 2001, p. 23). Although recent data are unavailable, a possible explanation of the perception of southern favouritism is because the south is the main destination of foreign and domestic investment. Additionally, Maputo City has a relatively developed infrastructure unscathed by the civil war. Decentralization was perceived to be a mechanism both to address developmental issues of local social and economic organization, and redress perceived regional imbalances. While opposition and civil society organizations sought to devolve decision-making, the government sought to keep its political monopoly through the privatization of the economy. The 1998 local elections installed thirty-three urban local governments, which remain subordinate to the central government and lack institutionalized mechanisms for popular
206
Case Studies: Southern Africa
participation (Weimer, 2000). The government emphasizes privatization as being crucial to the country’s ability to develop. In theory, privatization allows market forces to allocate national resources more effectively and efficiently. However, the government’s privatization strategy is believed to favour the ‘southern’ elites,12 thus creating political and social dissatisfaction. 4.2 The political economy of economic reforms War and flawed economic policies caused economic collapse and dependence on foreign aid. With peace, Mozambique became a fast-growing economy and a favoured destination for investors. A steady influx of funds (see Figure 12.1) from both bilateral and multilateral aid sources and strong macroeconomic discipline led to high growth rates and promising development.13 Peace has also encouraged domestic entrepreneurs to invest. Out of more than US$6 billion of investment, the national share amounts to 11 per cent (US$657 million), which is substantial given the short period of full economic activity. The economic activity and monetary inflow presented here concerns the formal sector only.14 The 1990s also witnessed a growing government propensity to privatize state assets and contract-out state services to increase revenue. More than 1,000 firms were privatized (Diogo, 2002, p. 10). These reforms required ‘adjustments’ to the political system and a redistribution of power among the governing elite. To avoid major political opposition to these reforms from within its own ranks, the government devised mechanisms to create a national bourgeoisie and facilitate domestic capital accumulation while maintaining control of the economy, and implementing an investment code which introduced foreigner investors to the benefits of joint ventures with local partners by limiting property rights to nationals (Africa Confidential, 1993a, p. 8). The 1995 Investment Promotion Centre (CPI) facilitated investment procedures while making sure that some shares in privatized companies were allocated to nationals. The main issue is not what the government privatizes, but how it privatizes. There is a consistent pattern of privatization by proxy, with the state running the economy by privatizing without selling.15 Landau (1998) finds that the government’s regulations on privatization, the relative scarcity of domestic capital, and privileged access to state and foreign funds by members of the governing elite combine to produce de facto control of the most attractive investment opportunities. Small businesses outside the patronage system face enormous challenges, exacerbated by the government’s preference for large-scale projects. Hence the perception that privatization aims to perpetuate the governing elite’s hold on power harbours the potential for instability. Mozambique’s peace rests on making economic opportunities inclusive. RENAMO’s refusal to recognize the government formally, claiming control over the areas in which it won a majority vote in the national elections, is
2,000 1,750 1,500
US$ (m)
1,250 1,000 750 500 250
0
1985
1987
1989
1991
1993
1995
1997
1999
Total
Years Foreign direct investment
National investment
Figure 12.1 Mozambique investment outlook, 1985–2000 Source: Compiled from Centre for the Promotion of Investments (2001, pp. 9–10).
207
208
Case Studies: Southern Africa
a crucial reflection of the non-governing elite’s perceptions of exclusion from economic opportunities. To reap its peace dividends, RENAMO tried to create de facto dual administration, suggesting that, when providing loans to Mozambique, the World Bank and partners should give RENAMO control of at least 30 per cent of the funds ‘because the projects only benefit Frelimo or people linked to it’ (interview, Maputo, March 2001). This is permitted by the highly centralized nature of the political system, which limits avenues for the opposition to participate in government, and prevents a broader distribution of the benefits of economic growth.
5
Conclusion
Continued peace is dependent on the responsiveness of current political institutions to popular aspirations and needs. The system of governance in place at the time of writing does not enable diverse social and economic groups to resolve the fundamental problems of development. Perceived as the preserve of the political and economic governing elites, the current political system breeds discontent and fosters political intolerance. Mozambique still presents a high risk of renewed conflict as measured by Collier and Hoeffler. These risks are exacerbated by a post-war political economy that channels the benefits of economic growth to a restricted elite. Although Collier and Hoeffler’s main risk factor, greed, was relatively low at the onset of the civil war, current developments seem to indicate that greed informs a potential for future conflict in Mozambique. Comparatively, geopolitical factors seem not to favour the possibility of further conflict, yet international actors are urged to play a more positive role. Their disengagement from Mozambique’s post-war political process may undermine the country’s stable development and create perverse incentives for political opportunism. The primary conclusion of this chapter relates to the lessons learned for postconflict recovery: the Mozambique case suggests a positive correlation between conflict resolution and the need for political and economic decentralization. Concomitant to demobilization, reintegration, national elections and macroeconomic adjustments, the peace process should include a component addressing issues of local self-rule. There is a need to ensure the emergence of a system of multiple checks on political authority at various levels. A functioning political opposition, decentralization of power to local authorities, and the creation of effective civic organizations would, in the authors’ opinion, enhance governance and diffuse tensions in Mozambique. Excessive centralization of the political system hinders the opposition’s access to government and limits local inputs in post-conflict recovery initiatives. Elected provincial and local authorities with defined spheres of autonomous action would be better advocates for local interests, and would influence national priorities and the allocation of government investment.
Francisco and Connerley: Mozambique
209
Experience in Mozambique suggests that international assistance should emphasize economic and political liberalization during, and for several years after, a peace settlement. Neither FRELIMO nor RENAMO has understood or appreciated the benefits of genuine political and economic liberalization. Both parties prefer to maintain a system in which politically bestowed privileges heavily influence the allocation of economic benefits. In post-conflict Mozambique, international assistance efforts have supported political and economic liberalization, but have not changed in a fundamental way the ‘zero-sum, winner take all’ thinking of the ex-combatants. A final lesson that may be deduced from the Mozambican experience is that peace settlements should include specific consideration of grievance. This somewhat obvious point may be under-appreciated, as we believe it was in Mozambique, by the international ‘peace bureaucracy’, when external interests strongly pursue peace. Notes * This chapter draws on work undertaken by Laudemiro Francisco in collaboration with Jeremy Weinstein, ‘External Actors as a Source of War and Peace in Mozambique’, for the World Bank’s project on ‘The Economics of Civil and Common Violence’. 1 In this chapter, legitimacy is understood as popular acceptance of a leader’s or regime’s claimed right to rule. At independence, FRELIMO assumed an ‘historical’ right to rule, based on its role in the struggle for independence from the Portuguese. However, other groups also advanced their own claims and denied FRELIMO’s. 2 RENAMO’s external representatives were members of the Mozambican diaspora. These individuals either left Mozambique during the independence struggle and decided not to return, given the new government’s policies, or were forced into exile immediately after independence, leaving behind all their possessions. For interviews where interviewees requested anonymity, we provide only the place and date of interview. 3 In the mid-1980s the government tried, but failed, to address the economic crisis. As a condition of maintaining access to World Bank funds, the government accepted the IMF prescriptions, initiating a structural adjustment programme in 1987. Initially named the Economic Rehabilitation Programme (PRE), it was renamed Economic and Social Rehabilitation Programme (PRES) later in the decade. 4 In addition to the churches, another private actor was Lonrho’s ‘Tiny’ Rowland, who became involved to protect his economic interests in Mozambique. Rowland made use of Lonhro’s financial muscle in Africa to push for a platform of direct dialogue between the warring parties, opening the way for official diplomatic involvement. 5 UNOMOZ was established under UN Security Council Resolution 797 (1992), 16 December 1992. 6 RENAMO’s chief negotiator, Raul Domingos, demanded on two occasions in June 1993 that the donor community provide RENAMO with US$100 million for its conversion into a political party as well as demanding the provision of new uniforms and professional training centres for RENAMO soldiers as a condition of demobilization (Africa Confidential, 1993b). 7 There was growing animosity within RENAMO ranks, resulting in a physical assault on Brigadier Dick, a commander in an assembly area in Mocubela. UNOMOZ had
210
8
9
10
11
12
13
14
15
Case Studies: Southern Africa called him in to calm a riot. At this time a number of riots were taking place in cantonment areas of both sides around the country, with the rank and file demanding proper treatment and rapid demobilization (Africa Confidential, 1994). A total of 92,881 soldiers were demobilized; 30,000 of them were supposed to join the new Mozambican Defence Force (FADM). By December 1994, when the UNOMOZ mandate expired, only 11,579 had joined the new army (United Nations, 1996, p. 42). Scepticism about serving in the armed forces continues to preclude meaningful voluntary recruitment, and the government was eventually forced to reintroduce conscription. A new law on conscription was passed by the Parliament in 1997. According to ex-combatants, most payment centres were difficult to access; they were usually located in district or provincial capitals, far away from ex-combatants’ resettlement homes (interviews in Iapala, Ribáuè District, Nampula Province, April 2001; and Maringue District, Sofala Province, May 2001). In April 1995, ex-combatants from FRELIMO and RENAMO, gathered under their umbrella organization, AMODEG, rioted in central Maputo, claiming that they had been abandoned by the authorities. The police dispersed them violently (Africa Confidential, 1995). Impact evaluations of demobilization and re-insertion programmes show that problems of reintegration were widespread. Such problems included an inability to lead a stable life and perceptions of social exclusion. The government sought to reduce the sources of immediate insecurity and ensure control of the national territory via a disarmament campaign in cooperation with the South African police to uncover RENAMO’s arms caches. The Mozambican Christian Council (CCM) was also involved in the collection of arms and ammunition left from the war in exchange for equipment such as sewing machines and hoes. The concept of ‘southerner’ includes individuals from the central and northern part of the country who migrated south and became ‘southernized’; that is, part of the ‘system’. Figure 12.1 shows a positive evolution of investment after the war, as compared to the negligible levels of the war years. Some large-scale projects such as Mozambique Aluminum, a US$1.3 bn venture where the government holds 4 per cent of the shares, were approved in 1997 (Centre for the Promotion of Investments, 2001, p. 10). Economic activities financed through the informal sector are not shown. One study estimates that real GDP is underestimated by approximately 79 per cent if the informal sector is not considered (Economist Intelligence Unit, 1998, p. 13). Some reports doubt that the state has collected the money due to it from the privatization of companies which collapsed shortly after privatization. Recent cases include the People’s Development Bank and the brewery, Cervejas Reunidas de Moçambique (Mozambiquefile, 2001a, 2001b, 2001c).
References Africa Confidential (1993a) vol. 34, no. 8, 16 April. Africa Confidential (1993b) ‘Mozambique: Renamo Plays for Time’, vol. 34, no. 15, 30 July. Africa Confidential (1994) ‘Mozambique: The Freelance Warriors’, vol. 35, no. 19, 23 September. Africa Confidential (1995) ‘Mozambique: Money and Votes’, vol. 36, no. 20, 6 October. Ball, N. and S. Barnes (2000) ‘Mozambique’, in S. Forman and S. Patrick (eds), Good Intentions: Pledges of Aid for Postconflict Recovery (Boulder, Col.: Lynne Rienner).
Francisco and Connerley: Mozambique
211
Cabrita, J. M. (2001) Mozambique: The Tortuous Road to Democracy (Basingstoke: Palgrave). Centre for the Promotion of Investments (2001) Tendências, Clima e Oportunidades do Investimento em Moçambique (Maputo: CPI). CEP (2001) Inquérito Nacional de Opinião Pública, Agosto de 2001 (Maputo: Universidade Eduardo Mondlane). Collier, P. and A. Hoeffler (2000) ‘Greed and Grievance in Civil War’, World Bank Policy Research Paper no. 2355 (Washington, DC: World Bank). Diogo, L. (2002) ‘Experiência e Desafios de Boa Governaçã o em Moçambique’, in C. Rolim, A. Franco, B. Bolnick and P. Andersson (eds), A Economia Moçambicana Contemporánea: Ensaios Gabinete de Estudos (Maputo: Ministry of Planning, Government of Mozambique). Doyle, M. and N. Sambanis (2000) ‘International Peacebuilding: A Theoretical and Quantitative Analysis’, American Political Science Review, vol. 94, no. 4, pp. 779–801. Economist Intelligence Unit (1998) Country Profile: Mozambique 1998–9 (London: EIU). Elbadawi, I. and N. Sambanis (2000) ‘External Intervention and the Duration of Civil Wars’, World Bank Working paper (Washington, DC: World Bank). Fundaçã o para o Desenvolvimento da Comunidade (Foundation for Community Development, FDC) (1997) Relatório do Seminário sobre a Criança Afectada pela Guerra (Maputo, FDC). Geffray, C. (1991) A Causa das Armas: Antropologia Social da Guerra Contemporánea em Moçambique (Porto: Ediçõ es Afrontamento). Gonçalves, F. (1998) ‘Ideological Shifts, Economic Imperatives: Southern African States and the Mozambican Peace Process’, in The Mozambican Peace Process in Perspective (London: Accord). Government of Mozambique (2001) PARPA – Action Plan for the Reduction of Absolute Poverty (2001–2005) – Strategy Document for the Reduction of Poverty and Promotion of Economic Growth (Maputo: Government of Mozambique). Hanlon, J. (1996) Peace Without Profit: How the IMF Blocks Rebuilding in Mozambique (Oxford: James Currey). Instituto Nacional de Estatistica (National Institute of Statistics) (1999) Mozambique – Statistical Yearbook 1999 (Maputo: INE). Landau, L. (1998) ‘Rebuilding the Mozambique Economy: Assessment of a Development Partnership’, World Bank Operations Evaluation Department Country Assistance Review no. 18589 (Washington, DC: World Bank). Lundin, I., M. Chachiua, A. Gaspar, H. Guebuza and G. Mbilana (2000) ‘Reducing Costs through an Expensive Exercise: the Impact of Demobilization in Mozambique’, in K. Kingma (ed.), Demobilization in Sub-Saharan Africa: the Development and Security Impacts (New York: St Martin’s Press). Minter, W. (1994) Apartheid’s Contras: An Inquiry into the Roots of War in Angola and Mozambique (London: Zed Books). Mozambiquefile (2001a) ‘Shareholders Pull Out of Austral Bank’, no. 298, May, p. 4. Mozambiquefile (2001b) ‘BCM Fraud: The Thief’s Version’, no. 299, June, p. 9. Mozambiquefile (2001c) ‘Merger of Two Largest Banks Agreed’, no. 303, October, p. 13. Sengulane, D. S. and J. P. Gonçalves (1998) ‘A Calling for Peace: Christian Leaders and the Quest for Reconciliation in Mozambique’, in The Mozambican Peace Process in Perspective (London: Accord). United Nations (1996) The United Nations and Mozambique, 1992–1995, United Nations Blue Books Series, vol. 5 (New York: UN Department of Public Information). United Nations Development Programme (2000a) Mozambique – Education and Human Development: Trajectory, Lessons and Challenges for the 21st Century (Maputo: UNDP).
212
Case Studies: Southern Africa
United Nations Development Programme (2000b) Mozambique – Economic Growth and Human Development: Progress, Obstacles and Challenges (Maputo: UNDP). Weimer, B. (2000) ‘Local Government/Decentralization’, Democratic Governance in Mozambique: Priorities for the Second Generation, 2002–2006, Report prepared for the United Nations Development Programme in Mozambique, Occasional Papers no. 2 (Maputo, UNDP). World Bank (1997) ‘War-to-Peace Transition in Mozambique: The Provincial Reintegration Support Programme’, Findings, Africa Region, no. 90, July (Washington, DC: The World Bank).
13 Transformation for Post-Conflict Angola Fátima Moura Roque
1
Introduction
Angola has extensive reserves of oil, diamonds and other minerals, and a burgeoning oil industry. At independence in 1975 it had a well-developed energy, transportation and communications infrastructure. However, the potential for balanced growth has been blighted by civil war, a critical shortage of skills, a centrally-planned economy subordinated to a military agenda, economic mismanagement, endemic corruption, and dependence on oil for foreign exchange and revenue. The country has been impoverished by war. Political interference in economic management resulted in perverse policies, financial sector weakness, opaque public accounts, and corruption. Welfare and personal security declined; Angola lacks qualified public managers and technicians. The public sector is inefficient and a weak private sector is engaged largely in rent-seeking. The death in battle of the leader of the National Movement for the Total Independence of Angola (UNITA), Jonas Malheiro Savimbi, on 22 February 2002, after the progressive destruction of UNITA’s military force and rural support base over the previous thirty months, was followed on 4 April by the signature of a Memorandum of Understanding consolidating the ceasefire and peace terms. Angola now faces extensive poverty and widespread malnutrition; there is a need to provide accelerated basic training for 80,000 demobilized soldiers, to create a safety net for about 300,000 of their dependants and to reintegrate 4.2 million people displaced within the country and 500,000 refugees. A strong commitment to national reconciliation, democratic governance and good economic fundamentals is essential if Angola is to overcome continuing resentment, social exclusion and corruption. To enable sustainable development there is a need to consolidate peace through national reconciliation, integration of the armed forces and the creation of a government of national unity. Institutional and constitutional 213
214
Case Studies: Southern Africa
reform are required. A technically qualified, independent team needs to be appointed to implement an effective emergency plan for economic stabilization and structural transformation; to promote democratization and build consensus in support of transformation; to enable transparent financial institutions; and to reschedule external debt. In 2002 none of these actions had been started.
2
Historical overview
Angola, with about 13 million inhabitants in 2001, has eight ethnic groups: the Bakongo, the Ambundu, the Lunda-Chokwe, the Ovimbundu, the Nganguela, the Nhaneka-Humbi, the Herero and the Chindonga. The Portuguese began colonization in the fifteenth and sixteenth centuries, when Diogo Cão arrived in the kingdom of Kongo in 1482, and Paulo Dias de Novais landed in Luanda in 1575. Subjugation was effected by conquest and by settling metropolitan Portuguese on tribal lands. Makidi (1983) describes how the peasantry was dispossessed and forced to work as farm labourers or unskilled industrial workers. According to Kambwa (1999, p. 57) ‘Portuguese ambitions . . . met with fierce opposition . . . Portuguese traders, missionaries, soldiers and even criminals were arriving and settling, and practising slavery.’ McCotter and Eddy (1999, p. 58) assert that four million were enslaved, with no other African nation suffering as severely. Portugal controlled little of the interior until after the wars of occupation (1845–1917). Administration was extended during the governorships of Paiva Couceiro and Norton de Matos, during which traditional authorities were subordinated and kingdoms abolished. Only the Kongo kingdom survived – until independence in 1975. Angola’s borders were defined in treaties between 1880 and the 1920s. Military penetration intensified after the Berlin Conference of 1884–5. In 1930, to facilitate natural resource exploitation, the Portuguese government passed the Colonial Act ‘which would submit the Angolan people to . . . worsened labour conditions . . . prohibition of cultural manifestations, assimilation and forced labour’ (Kambwa, 1999, p. 58). The abolition of the slave trade in 1878, rising coffee production and the development of the rubber industry led the authorities to establish control over the interior. ‘One of the highland kingdoms, Bailundu, was a particularly oppressed zone of labour extraction throughout the twentieth century’ (Birmingham, 2002, p. 139). Lacking capital, Portugal failed to invest, and key ventures passed into British, American, German and Japanese ownership (Makidi, 1983). Portuguese entrepreneurs established coffee, sisal, sugar, cotton, rubber, tobacco and fruit production. The manufacturing sector focused on food processing, textiles and clothing. Most surplus production was exported to
Moura Roque: Angola 215
Portugal. Meanwhile, ‘unskilled Portuguese immigrants compete[d] with indigenous workers, typically winning employment and contracts based on “racial superiority” . . . [This] led to heightened prejudice and increased hostility’ (McCotter and Eddy, 1999, p. 59). Indigenous Angolans were denied education and employment, leading to unrest and conflict. Even in 1955, 380,000 worked as forced labourers. According to Tvedten (1997, p. 28) ‘important cleavages developed between the small (mostly urban) elite of Africans and mestiços and the large (mostly rural) majority of African peasants’. Although missionaries and private landowners provided limited education for rural people, only in the final decade of colonial occupation did the government deliver education and health services outside the main cities. Colonialism led to the emergence of three classes: a petite bourgeoisie of intellectuals and commercial traders (white, mestiços and assimilados); an unskilled, industrial and agrarian near-proletariat, employed as forced labourers; and the residual peasantry (indígenas), engaged in subsistence agriculture and providing labour for the bourgeoisie. Language was power in post-colonial African politics and it was the assimilados who spoke Portuguese, the language of command (Birmingham, 2002, p. 149). The emergence of liberation movements during the 1950s and 1960s reflected this pattern. The Movimento Popular para a Libertaçã o de Angola (MPLA) was founded by left-wing, urban elites in Luanda. Its background was (and is) Luso-African; its leading cadres were Catholic seminarians or other assimilados and mestiços with a limited rural base in Kimbundu. The Frente Nacional de Liberte de Angola (FNLA) was a Bakongo movement, rural in character, although many members later gravitated to the northern coastal cities, where they were known as Zairenses, as their leader, Holden Roberto, was the brother-in-law of Mobutu Sese Seko, then President of Zaire. The National Movement for the Total Independence of Angola (UNITA) was founded in 1966 by the Ovimbundu, although it included Cabindans, Bakongo, Lunda-Chokwe and others. Its institutional culture is ruraltraditional and it included few assimilados. Angola gained independence on 11 November 1975 during a struggle between the MPLA, FNLA and UNITA. The MPLA, after defeating the others, formed the ruling government. Angola’s urban elites, MPLA supporters, were forged in the Portuguese administrative culture and seen by the colonialists as their natural successors when they withdrew in 1975. ‘The colonial class of 300,000 privileged and semi-privileged expatriates had been replaced by a similar number of black Portuguese-speaking Angolans who retained many of the old colonial attitudes of social and moral superiority’ (Birmingham, 2002, p. 184). The sense of superiority of these Luandenses to the Ovimbundu of the central plateau, whose economic roles were menial both before and after independence, is palpable even into the twenty-first century.
216
3
Case Studies: Southern Africa
Causes and effects of the conflict
3.1 Causes of the conflict Historical, political, socio-economic and deep-rooted cultural factors precipitated the war in Angola.1 Analysis of the conflict indicates causes that must be addressed if reconciliation is to be achieved: incorrect definition of the cause of a serious conflict leads to . . . procedures inconsistent with the realities of that conflict . . . [thus] likely to be unsuccessful. If a conflict is caused by . . . [a] need for identity and cultural security, but is defined as one stemming from aggressiveness, the likely outcome will be protracted and escalating conflict. (Burton, 1987, p. 21) Ethnic divisions, class divisions, and competition for resources, cannot alone account for the Angolan conflict. Ethno-cultural and class divisions lead to conflict only when dysfunctional political systems entrench politicoeconomic disparities and exclude groups from access to power, leading to widespread resentment, human rights abuses and fear, giving opportunities for political mobilization on ethnic or class grounds which can lead to conflict. The causes of the Angolan conflict are clear. The country’s historical legacy includes different socio-cultural identities and values, and class distinction between assimilated and indigenous people, overlaid on the ethno-linguistic groups. Political causes include the concentration of power in one social group that excluded the legitimacy of others. There was a lack of an institutional mechanism for those excluded to have their needs addressed, and an absence of a common normative system to reconcile divergent interests. This was manifested in systematic violation of fundamental rights, pervasive corruption and perverse external influences.2 Economic causes included the inequitable distribution of income, resources and national wealth, and imbalances between regions and urban and rural areas. Economic mismanagement and poor governance promoted a large and inefficient state sector, extreme poverty and weak institutions, despite revenues from oil and diamond exports. These distortions prevented the development of the non-oil sector and the accumulation of human capital. Social and cultural causes included poor education, health and social services, and the manipulation of ethnic and regional sentiments by politicians. Economic and political power was concentrated in the MPLA elites, the heirs of the Portuguese colonialists. Their refusal to accept the legitimacy of other interests prevented factions excluded from power from having their needs addressed. The war led to the systematic violation of the rights of ordinary citizens. The consequences are evident in the extremes of elite wealth and abject poverty, the rejection of their African identity by the elite, and the political manipulation of ethnic sentiments.3
Moura Roque: Angola 217
3.2 Effects Collier (2000, p. 4) suggests that ‘what matters is whether [a rebel] organization can sustain itself financially . . . it does not really matter whether rebels are motivated by greed, by a lust for power, or by grievance, as long as what causes conflict is the feasibility of predation.’ Oil and diamond revenues financed the Angolan war. In 1999, almost US$4bn (65 per cent of Angola’s GDP) was allocated to defence and internal security, unclassified items and (accounting) discrepancies. This is in sharp contrast to allocations for health and education, which according to the IMF (2002) were only 2.7 per cent of GDP. Angola’s social indicators are well below the sub-Saharan African (SSA) average. In 1999, infant mortality was 172 per 1,000 live births in Angola, compared to 107 in SSA. Life expectancy was 45, against an average of 49 years. At US$270, Angola’s per capita GNP was just over half the SSA average of US$492 (IMF, 2002). In 2000, Angola ranked 160th out of 174 countries in the UNDP’s Human Development Report. More than 70 per cent of Angolans live in absolute poverty, less than 25 per cent have access to health care and less than a third have access to safe water. About half the population is illiterate. Children especially suffer in Angola, and about 42 per cent are malnourished: About half of the 4 million internally displaced people are children . . . with 70,000 amputees due to landmines . . . one child in three never reaches the age of five . . . more than two thirds . . . that survive do not attend school . . . more than half of those that should have gone to school are street children . . . Angolan children are the most vulnerable . . . [in] Sub-Sarahan Africa. (Olara Otunnu, Público, 12 May 2002) A child dies of a preventable disease or malnutrition every three minutes. Income distribution is skewed. In 1998, the richest 10 per cent enjoyed a 44 per cent increase in wealth, while the poorest 10 per cent suffered from a 59 per cent decrease (World Bank, 2000). There are no public services in the rural areas and the urban shanty towns have no infrastructure. More than 90 per cent of rural people live off subsistence agriculture, despite more than 10 million landmines that prevent planting.
4
The post-conflict transformation strategy
4.1 Transformation from ceasefire to durable peace The Angolan civil war continued until 2002, despite three peace agreements (the Alvor accord in 1975, the 1991 Bicesse accords and the Lusaka protocol in 1994) because of a failure to address the root causes of the conflict.4 Crocker (1992, p. 468) emphasizes the need to address the sources of conflict,
218
Case Studies: Southern Africa
not just the symptoms. Similarly, Annan (UNSC, 12 July 2000) observes that ‘prevention strategies must address the root causes of conflicts, not simply their violent symptoms’ and ‘prevention has to address the structural faults that predispose a society to conflict’. Kambwa (1999, p. 56) stresses that Angolans ‘must be fully aware of the historical, political, economic and social contradictions in relation to the peace process, and inculcate a social climate of democracy, justice, tolerance and co-existence’. The conflict has left deep social antagonisms, and ethnic, regional and urban–rural imbalances, exceptional poverty and income gaps; it has destroyed infrastructure and greatly distorted the economy. Achieving durable peace means addressing these distortions and their causes. Angola’s former belligerents reflect different value systems with different needs. An inclusive political, constitutional and socio-economic framework is needed for a sustainable peace. This will only be achieved through reconciliation and the acceptance of diverse identities, values and interests. Durable peace, reconciliation and sustainable development require a regionally decentralized state, democratic governance and a market economy combined with a strong social development focus (see Moura Roque, 2000). Decentralization needs institutional change to bring government to the people. Democratic governance requires political and institutional transformation, while a social-market economy implies a long-term development strategy, incorporating structural and systemic transformation and a strategy of poverty reduction. Civil society must help divided communities to reconcile and reconstruct. Success will demand an extraordinary act of national will and community-wide participation, while failure will condemn another generation to poverty. Nation-building requires a transitional government representing all Angolans, since premature elections risk reigniting the conflict. A priority is to establish mutual acceptance as the foundation of national reconciliation, promoting political dialogue to define an inclusive national vision, strengthen civil society and negotiate a constitution that guarantees basic rights, and accommodates the values and needs of all. Political negotiations must transcend bargaining and address underlying causes. Civil society can play a mediating role, although Angola does not have the resilient civil institutions forged in South Africa in the struggle against apartheid. Interim emergency humanitarian assistance is essential for the most vulnerable and dispossessed; a transitional government should agree on a reconstruction strategy with the relevant international agencies. National reconciliation requires each party to distinguish between their interests (which are negotiable), and needs and core values – identity, security, political and economic participation – which may not allow for compromise. There is a need for collaboration in defining the requisite institutions, laws and policies, and open communication between the parties and their constituencies by guaranteeing media freedom. Violence must therefore be
Moura Roque: Angola 219
decoupled from the efforts of groups to protect their security and identities, building a political culture that accepts the diversity of identities, values and needs, and establishes popular accountability and an inclusive economic system.5 Attempts by elites to impose their values through a monopolistic and authoritarian political system prompted marginalized Angolans to turn to violence. Building sustainable peace means effecting inclusion through new political and economic systems. Party-based bureaucracy, war and economic collapse led to corruption and the withdrawal of social services. Reconstruction and development require a transparent and democratically accountable system of government, which can be promoted through devolution of power and resources to decentralized but inclusive local and provincial authorities. A transitional national unity government priority should be to promote reconciliation, encouraging an inclusive national identity with common values and norms, but enabling the expression of diverse needs and interests. Demilitarization poses a special challenge. Creating a neutral, competent administration, defence force and police service, and building trust between them and the society they serve, is an enormous task. A priority is to build a unified professional defence force under civilian authority while at the same time implementing progressive demobilization, and ensuring skills training for former soldiers not needed for defence. Reintegrating demobilized troops, displaced people and refugees into a fractured society, and clearing millions of land mines, will take a decade. In the economic sphere, a long-term development strategy should aim to promote balanced regional and national development and an equitable allocation of resources. Rebuilding institutional capacity is required to ensure sound economic planning and management and economic stability, while emergency aid can be used initially to assist the most vulnerable and provide a platform for rehabilitation. 4.2 Political instruments for national reconciliation Angola faces two great challenges: one politico-institutional; and the other socio-economic. Political change is needed to build accountable government and effective institutions that respect basic rights, empower the least advantaged and promote lasting peace. The socio-economic challenge lies in promoting equitable, sustained socio-economic development while reducing poverty and social exclusion. Adedeji (1999, p. 17) emphasizes the need for a post-conflict governance system inspired by traditional African values of solidarity, cooperative spirit, self-help development and humanism, in order to ensure sustainable peace and development. It will take time to rebuild confidence between people who have fought bitterly. The major elements of national reconciliation are as follows. First, each party should accept the legitimacy of the other. To achieve this, each party defines itself and selects its own representatives, agreeing to put the past
220
Case Studies: Southern Africa
behind it and move to address the future together. Second, there is a need for recognition and acceptance of the differences between the parties. These might include different identities, constituencies, cultural values, needs and interests. Third, diversity should be accommodated within a constitutional framework. Provision should be included to ensure universal protections (a bill of rights, a constitution, courts); systematic checks and balances (such as separation of powers); inclusive legislative representation (such as multimember constituencies); and geographical, community-based representation and autonomy (such as decentralization). Fourth, human rights should be guaranteed by ensuring the independence of the judiciary; providing civic training for armed services and police; ensuring freedom of the media; and the involvement of civil society, including churches and traditional authorities. Finally, sound economic practice should be promoted by ensuring good macroeconomic management; reducing the incidence of poverty; investing in social development; improving access to, and the quality of, public services; facilitation of grassroots participation in the development process; and the promotion of the development of provinces suffering from economic neglect and devastation from war. Mutual hostility between UNITA and the MPLA reflects the humiliation and resentment rural Angolans feel about their inferior status and exclusion from public life, and the Luanda elite’s fear of reprisals if UNITA were to gain power. Class and social identities deriving from colonial times continue to define each Angolan’s social role. The urban administrative class does not take the rural and labouring classes seriously, while the dispossessed resent the privileged. The fact that the urban elites assumed Portuguese names and customs but the rural areas take pride in their African names, languages and customs, exacerbates the deep fissure in Angolan society. How are all to become Angolans? Kofele-Kale (1981) addressed the problem of exclusive national and subnational identities by arguing that each person can have more than one identity. Adopting a common national identity does not require denial of another ethnic, religious or cultural identity. The two can coexist, transcending narrow identities, in a democracy that values national diversity. To enable this, the government must promote national unity and purpose; consolidate democracy; protect fundamental rights with disciplined, non-partisan security forces; combat corruption and administrative abuse; and create space for civil society to make constructive contributions. It must also ensure macroeconomic and monetary stability; facilitate the operation of market forces; promote private investment; redistribute income; and promote development by investing strategically in key sectors and provinces. Finally, it must provide public goods that the market cannot offer. Green (1999, p. 34) concludes that ‘a post-war state needs to restore security, non-repressive governance, basic infrastructure and delivery of basic services, or it will not be perceived as legitimate for long and new conflicts will arise’.
Moura Roque: Angola 221
As Elbadawi and Sambanis (2000) show, strategies for recovery should be based on promoting political freedom and formulating a governance framework that can accommodate Africa’s social diversity. 4.3 The role of civil society Civil society can help to build diversity, encouraging patriotism and pride in African identity so as to sustain the quest for national unity. It can build commitment to democracy and encourage participation in global affairs. National unity will only emerge if the supporters of all political parties feel secure. Inclusive institutions ‘build confidence in divided societies, they facilitate a better distribution of resources from the centre to the country as a whole, while democracies with exclusionary institutions create an incentive for ruling parties to pamper their home region or ethnic group to the detriment of other . . . groups’ (Reynolds and Sisk, 1998, p. 30). Similarly, Olukoshi (1995, p. 27) observes that ‘no democratic system can flower which is not anchored in domestic social forces’. Angola must develop its civil society, a community of independent organizations dissociated from the interests of the two main political parties, to facilitate reconciliation and reconstruction. Although the formal institutions of constitutional democracies – an independent press, a system of law, regular elections, a code of ethics for public servants, a full-time Parliament, an accountable executive, and well-organized political parties – are essential, Angolans themselves must build consensus and solidarity. In turn, the state must recognize the essential role of civil society in building national unity, and must prove its respect for this role by guaranteeing the civic and political rights of Angolan people. This entails: • • • • •
separating the judicial power from the executive; guaranteeing the protection of the courts to the citizens; ensuring media freedom; encouraging civic activism to resolve common social problems; and creating institutions to address complaints by citizens against government officials.
4.4 The economic dimension Transparent governance, prudent monetary policy, fiscal consolidation and a flexible foreign exchange-rate regime are needed for macroeconomic stabilization. A long-term development strategy (LTDS),6 incorporating a structural and systemic transformation programme (SSTP) and a povertyreduction strategy (PRS) are essential economic tools to promote reconciliation and development with social justice, in the context of the post-conflict reconstruction transformation. This requires carefully designed economic policies to achieve long-term economic benefits, eliminate inherent distortions
222
Case Studies: Southern Africa
and mitigate political costs. As Lipumba (1995, p. 56) shows, designing and implementing a poverty-eradicating development strategy is a political issue requiring a government capable of building its legitimacy through economic growth and structural transformation. Reconstruction must focus on the transformation to a market economy based on social development. This can be achieved through the implementation of a comprehensive SSTP, including emergency assistance to the most vulnerable and a programme of vocational skills training to promote demilitarization and demobilization, and encourage income generation. The re-establishment of properly trained professional police and armed forces is also a priority. State institutions must be strengthened, reorientated and staffed with competent officials. Effective institutions, staffed by qualified civil servants and accountable politicians will help to ensure effective policy management and implementation, and oversee the development of a diversified economy7 based on macro-economic stability. Such a vision would also include redirecting government spending towards key physical infrastructure and social welfare programmes such as housing, primary health care, education and skills training. Policies should be implemented to increase the accessibility of property ownership and ensure that its distribution is more equitable. In the business sphere, public–private partnerships should be encouraged so as to rebuild the production and industrial sectors, opening them up to foreign direct investment (FDI), skills and technology; and regional and global integration. Reconstruction will take between two and ten years, depending on the government’s commitment, the engagement of civil society and the fit between social programmes, stabilization measures and institutional reforms. It is likely that a period of austerity will be required, including a drastic reduction of military and non-productive consumption spending, together with rigorous anti-corruption measures, diversification of the productive sector, better revenue collection, and international aid for community rehabilitation. Additional support will be needed for infrastructure and macroeconomic restructuring, including debt relief and enhanced market access to support trade and price reform. A priority should be to set up a programme of national and provincial peace and reconciliation accords (outside the Lusaka protocol) with the aim of extending government administration peacefully and disarming civilian militias. Such a programme should work hand in hand with the formation of a unified, professional defence force and provide a framework for the provision of vocational skills training to demobilized soldiers. Similarly, a programme of consultative workshops can secure popular input into policy, the formation of a representative, technocratic transitional government and the formulation of a new constitution. The most urgent measures pertaining to public finances include the reform of the oil and diamond sectors and the central bank. Extra-budgetary
Moura Roque: Angola 223
expenditure and fiscal discrepancies must be eliminated and transparent controls implemented.8 All external loans must be documented, and information on public borrowing made public.9 Similarly, oil-related transactions must be recorded and offshore accounts eliminated. Information on oil-backed loans must be provided, and their terms and conditions compared with external debt stock and flow data. The transparent allocation of diamond concessions is needed, and questionable purchasing and marketing arrangements must be ended. Procurement procedures must be revamped and public finances made transparent. A strong, credible, independent central bank is essential. Efficient public administration, the rehabilitation of physical infrastructure and human capital development are vital. Low-inflation policies should be adopted and a realistic exchange rate imposed so as to develop an efficient financial system, support private-sector development, encourage FDI and promote competitive markets to facilitate long-term development. Moura Roque (2000) outlines the main foundation-building measures that should be implemented in Angola. These comprise agreement on a social contract on reforms,10 macroeconomic stabilization, microeconomic liberalization, the restructuring and privatization of state-owned enterprises in a sound legal and institutional framework, investment in skills and market development, and agreement with foreign creditors on debt restructuring. Furthermore, addressing famine and reducing poverty require an immediate response, while the reintegration of displaced persons and skills training and income generation activities targeted at demobilized soldiers are essential for a durable peace. These foundation-building and stabilization measures must be implemented immediately, with funding assistance from the international community on a long-term basis, aiming for sustainability. A social-market economy requires a legal framework, an effective fiscal system, an enabling environment for investment, human resource development and integration into the external economy. The linkages between poverty alleviation, job creation, spatial development, human capital development, income distribution and economic growth demand the prioritization of social and sectoral investment.
5
Poverty reduction strategy (PRS)
Poverty and huge disparities in income, wealth and access to education and health care have been exacerbated by war, bad governance and endemic corruption. Poverty has worsened in recent years: preliminary results from a household survey conducted in 2000–1 indicate that the percentage of households living in extreme poverty had increased since 1995 and that income inequality increased during the same period (IMF, 2002). The same report raises concern regarding the profile of government expenditure: compared with the 2001 budget, the 2002 budget shows both a reduction in
224
Case Studies: Southern Africa
social expenditure in real terms and a large increase in the wages of government officials and the day-to-day costs of running and maintaining central government and ministerial departments (IMF, 2002). Decisive measures to combat corruption are needed. Poverty reduction requires growth and the equitable allocation of welfare. This can be achieved through measures to improve productivity and promote private-sector development and capacity building. There is a need for investment in infrastructure, especially rural infrastructure, and social services – such as improving water delivery, housing, health and education services. Human resource development priorities should include women’s education and vocational training. Social safety nets can be provided for the most vulnerable through labour-intensive public works and social funds, restoring agriculture and extension services, and rebuilding rural markets. Access to finance and loans is required, in addition to creating rural co-operatives and micro-enterprises for farmers. These policies, many of which would be essential for sustainable development, even if poverty were less acute, must be integrated into the transformation programme. To ensure proper targeting of the poorest communities, the PRS should be designed and implemented on a decentralized, though integrated, basis. The objective in each province should be to contribute towards the empowerment of the poor by offering them better access to social services and infrastructure, and facilitating their consultation and participation in regional resource allocation decisions. There should be safety nets for targeted households. Economic measures might include a focus on labour-intensive agriculture, export-orientated light manufacturing and tourism. Improved training facilities and technical support are required. These objectives can only be achieved through regional fiscal efficiency and cooperation between efficient, transparent and accountable central and regional governmental authorities. The impact of these programmes on key poverty indicators must be monitored carefully.
6
Decentralization
Because of the divisions exacerbated by the civil war, central government cannot work effectively in the areas where people suffered as a result of its military activities. Both armies, moreover, forced reprisals on their opponents when control of towns changed during the fighting. These cleavages can be addressed through decentralization, to accommodate Angola’s diversity, and encourage both democratization and balanced regional development. Delegation of authority to properly constituted provincial authorities will facilitate effective governance, the protection of basic rights and freedoms, the growth of civil society, and national harmony. Decentralization will also help to address the need for balanced development, poverty reduction
Moura Roque: Angola 225
and grass-roots participation in development, since the decentralization process can include democratic political reform. Effective local governments can better develop remote areas whose needs the capital does not understand. By encouraging civic education and participation, they can also check authoritarian instincts at the centre. Regional demarcation is a sensitive issue. In determining boundaries one must plan for administrative capability, infrastructure for administration and an economic base allowing a measure of autonomy. All parts must be accessible from the administrative capital, so proximity is relevant, though the optimal size of each region depends on the available infrastructure. Ethnic diversity or coherence is less important unless past ethnic conflict requires division for community rehabilitation. The criteria – social, historical and geographical – and economic functionality and administrative efficiency must be considered carefully. Extreme asymmetry in economic potential, population size and social composition should be avoided. Decisions on regional demarcation, the composition, functions, resources and accountability of regional institutions, and the nature of regional representation in central government must be constitutionally entrenched.
7
Conclusions
Durable peace in Angola depends on constant dialogue and true national reconciliation, and both must address honestly the underlying historical, economic, political, social and cultural factors that drove the conflict. Demilitarization of the administration, society and the economy is, moreover, essential, while mutual respect between the military and civil society must be established. To facilitate this, the media should use their freedom in a manner calculated to promote reconciliation and reconstruction. The economic and social reintegration of demobilized soldiers, displaced people and refugees must be undertaken. Moreover, good governance based on democratic principles, popular accountability, respect for fundamental human rights, and effective application of the rule of law and public order are essential to promote reconstruction. Post-conflict reconstruction poses exceptional challenges in Angola. In essence, it is the reconstruction of a whole society. If the people have not lived and worked together and experienced a collective and externally-defined political identity as nationals of one state in the context of a common vision and sense of destiny, there is little by way of social capital to reconstruct. I have therefore preferred the phrase ‘post-conflict transformation’. Economic reconstruction and political reform will not, in themselves, achieve social harmony, nor will they lay the foundations for sustainable development. The task is far greater: a sense of transcendent national identity will only emerge if Angolans – especially those who were excluded in the past – feel
226
Case Studies: Southern Africa
that their political and economic prospects are attractive, while institutions and capacities that make this possible will only be forged if the political elites are determined to build an inclusive society. The social, economic and political transformation of Angola is thus part of an organic whole. Transformation of political, economic and judicial institutions and systems must be buttressed by the creation of a common social identity. This requires psychological rehabilitation, a task that can only be performed by the churches and other civil society bodies. International institutions that choose to engage in the task of post-conflict reconstruction in Angola will need to bear all this in mind when considering what contributions they are able to make.
Notes 1 Toure (1999, p. 23) states that ‘conflicts arise from human relations in two principal ways: first, individuals or groups of individuals have different values, needs and interests; and, second, most resources are not available in unlimited quantities and so access to them must be controlled and fought for’. 2 See Hodges (2001), Global Witness (2002) and IMF (2002), on the effects of corruption on the development of Angola, and how corruption fed the conflict. 3 Elbadawi and Sambanis (2000, p. 1) show that ‘deep political and economic development failures – not tribalism or ethnic hatred – are the root causes of Africa’s problems.’ Collier (2000, p. 8) emphasizes that ‘whereas natural resources, a lack of economic opportunities and ethnic dominance are causes of initial conflict, once conflict has arisen it generates grievances which themselves increase the risk of subsequent conflict’. 4 There was war from 1975 to 1991, from October 1992 to November 1994 and from December 1998 to March 2002. Peace agreements applied between June 1991 and September 1992, and from November 1994 to November 1998. 5 World Bank president, James Wolfensohn, wrote after 11 September 2001 that ‘central to conflict prevention and peace-building must be strategies for promoting social cohesion and inclusion, ensuring that all have opportunities for gainful employment, that societies avoid wide income inequalities that can threaten social stability and that the poor have access to education, health care, and basic services such as clean water, sanitation and power’ (in Global Witness, 2002, p. 4). 6 See Moura Roque (1991, 1994, 1997 and 2000) concerning the sequencing, intensity and duration of LTDS comprising a SSTP with a PRS. 7 The Angolan state depends on oil for 80–90 per cent of its fiscal revenue and over 90 per cent of its exports (Hodges, 2001, p. 132; Global Witness, 2002, p. 33). The situation before independence was very different (see Moura Roque, 1991, ch. 3). 8 In 2002 the IMF reported that in the preceding five years, ‘extra-budgetary expenditures and residual unexplained discrepancies in the fiscal accounts have averaged 11 per cent and 12 per cent of GDP per year, respectively . . . substantial funds received as signature bonuses for oil contracts and oil royalties have been outside the control of the treasury; non-transparent external debt transactions have been made . . . oil companies are pressed to donate sums to funds and foundations’ (IMF, 2002, box 3, p. 13).
Moura Roque: Angola 227 9 Publication of government revenue and foreign debt flows is essential for transparency and has been recommended by the IMF for many years. 10 The ‘social contract’ must be based on the society Angola seeks to create, and an understanding of how the programmes are to be financed.
References Adedeji, A. (1999) ‘Comprehending African Conflicts’, in A. Adedeji (ed.), Comprehending and Mastering African Conflicts: The Search for Sustainable Peace and Good Governance (London: Zed Books). Birmingham, D. (2002) ‘Angola’, in P. Chabal (ed.), The History of Postcolonial Lusophone Africa (London: Hurst). Burton, J. W. (1987) Resolving Deep-Rooted Conflict: A Handbook (Lanham, Md.: University Press of America). Collier, P. (2000) Economic Causes of Civil Conflict and their Implications for Policy (Washington, DC: World Bank). Crocker, C. A. (1992) High Noon in Southern Africa: Making Peace in a Rough Neighbourhood (New York: W. W. Norton). Elbadawi, I. and N. Sambanis (2000) ‘Why Are There So Many Civil Wars in Africa? Understanding and Preventing Violent Conflict’, Journal of African Economies, vol. 9, no. 3, pp. 244–69. Global Witness (2002) All the Presidents’ Men: The Devastating Story of Oil and Banking in Angola’s Privatised War (London: Global Witness). Green, R. H. (1999) ‘Angola: Seeking to Remedy the Limitations and Bias in Media and Scholarly Coverage’, in A. Adedeji (ed.), Comprehending and Mastering African Conflicts: The Search for Sustainable Peace and Good Governance (London: Zed Books). Hodges, A. (2001) Angola from Afro-Stalinism to Petro-Diamond Capitalism (Bloomington, Ind.: Indiana University Press). IMF (2002) Angola: Staff Country Report (Washington, DC: International Monetary Fund). Kambwa, A. E. (1999) ‘Angola’, in A. Adedeji (ed.), Comprehending and Mastering African Conflicts: The Search for Sustainable Peace and Good Governance (London: Zed Books). Kofele-Kale, N. (1981) Tribesmen and Patriots: Political Culture in a Poly-ethnic African State (Washington, DC: University Press of America). Lipumba, N. H. (1995) ‘Long-Term Development Strategies for Africa as We Enter the 21st Century’, in A. O. Olukoshi and L. Wohlgemuth (eds), A Road to Development: Africa in the 21st Century (Uppsala: Scandinavian Institute of African Studies). Makidi-Ku-Ntima (1983) ‘Class Struggle and the Making of the Revolution in Angola’, Contemporary Marxism, vol. 6. McCotter, B. C. and M. Eddy (1999) ‘Angola: Political Development and Local Government’, in P. S. Reddy (ed.), Local Government Democratisation and Decentralisation: A Review of the Southern African Region (Johannesburg: Juta). Moura Roque, F. (1991) Economia de Angola (Lisbon: Bertrand Editora). Moura Roque, F. (1994) ‘Economic Transformation in Angola’, The South African Journal of Economics, vol. 62, no. 2, pp. 71–91. Moura Roque, F. (1997) Building the Future in Angola: A Vision for a Sustainable Development (Oeiras: Celta Editora). Moura Roque, F. (2000) Building Peace in Angola: A Political and Economic Vision (Lisbon: Ediçõ es Universitárias Lusófonas).
228
Case Studies: Southern Africa
Olukoshi, A. (1995) ‘Africa and the “New” Political Conditionality: What Prospect for Sustainable Political Change?’, in A. O. Olukoshi and L. Wohlgemuth (eds), A Road to Development: Africa in the 21st Century (Uppsala: Scandinavian Institute of African Studies). Otunnu, O. (2002) Público, 12 May, Angola. Reynolds, A. and T. D. Sisk (1998) ‘Elections and Electoral Systems: Implications for Conflict Management’ in T. D. Sisk and A. Reynolds (eds), Elections and Conflict Management in Africa (Washington, DC: United States Institute of Peace Press). Tvedten, I. (1997) Angola: Struggle for Peace and Reconstruction (Boulder, Col.: Westview Press). Toure, A. T. (1999) ‘Mastering African Conflicts’, in A. Adedeji (ed.), Comprehending and Mastering African Conflicts: The Search for Sustainable Peace and Good Governance (London: Zed Books). UNDP (2000) Human Development Report (New York, NY: UNDP). UNSC (2000) Address to the United Nations Security Council, Open Session on Conflict Prevention, 12 July, SG/SM/749/SC/6893, New York. World Bank (2000) Angola (Washington, DC: World Bank).
Part V Conclusion
This page intentionally left blank
14 Post-Conflict Economies in Africa: Synthesis and Lessons Augustin Kwasi Fosu
It is impossible to do justice to all the studies presented in this volume, but this chapter attempts to synthesize the material and draw lessons from the case studies. The synthesis has ten thematic issues: colonial history; neighbours; elite political instability; rent-seeking opportunities; child combatants; the proliferation of cheap weapons; inter-ethnic and class injustice and grievance; the need for impartial and effective judicial and legal systems; the role of the international community; and the role of democratic institutions.
1
Colonial history
Colonial history played a major part in fostering conflicts in Africa. International boundaries determined the ethnic composition of nations and, in many instances, borders cut through existing ethnic groups, sowing the seeds of diluted loyalty and the likelihood that a disenchanted minority would seek refuge across the border. Furthermore, in many countries, the colonial power empowered certain ethnic groups over others. For example, the Belgians, as the colonial power in Burundi, promoted Tutsis and marginalized Hutus, thus contributing to the subsequent ethnic imbalance and the conflict (Nkurunziza and Ngaruko). Similarly, the original constitution at the time of independence in Uganda favoured the powerful kingdom of Buganda, raising resentment among other ethnic groups and contributing to the power interplay that followed (Atingi-Ego and Subudde). In the case of Sudan, the colonial administration’s policy towards southern Sudan contributed to the initial conditions for the conflict, as the policy contained measures that insulated the south from the north in anticipation of the former being integrated into East Africa (Ali and Elbadawi). Although the specifics may differ from country to country, the role of colonial history in sowing the seeds of conflict tends to permeate the case studies. The saliency of colonial history in influencing the incidence of conflicts suggests that former colonial powers have a stake in promoting post-conflict states in Africa. For example, the British intervention in Sierra Leone, an 231
232
Conclusion
erstwhile British colony, is generally credited as a critical turning point in ending the Sierra Leone conflict (Davies). Similarly, French troops and diplomacy have intervened in the recent Côte d’Ivoire conflict. Although, at the time of writing, the post-conflict state is fragile, French intervention appears to have helped attenuate the conflict and has contributed towards some semblance of peace in the country. These two and other cases clearly demonstrate the potential part to be played by erstwhile colonial powers, not only in ending African conflicts, but also in fostering enduring postconflict economies.
2
Neighbours
Neighbours can play a major role in the spread of conflict as well as in its resolution. The existence of a sympathetic ethnic group across the border might provide protection for a rebel force and thus prolong the conflict. Such a situation would also help to spread an existing conflict across borders. This need not be restricted to cross-border ethnic groups: the interwoven conflicts involving Liberia, Sierra Leone and Guinea and, to some extent, Côte d’Ivoire and Burkina Faso, indicate a wide range of rationales. In many instances, rebels receive sanctuary in neighbouring countries: Sierra Leone’s rebels in Liberia; Liberian rebels in Guinea; Rwandan rebels in the DRC; Ugandan rebels in Tanzania during the 1970s and, at the time of writing, in the Sudan (as claimed by the Museveni government, though denied by the Sudanese authorities); and Côte d’Ivoire’s rebels in Liberia. Although cross-border operations usually contribute to the spread of conflict, in some cases they have been useful in ending them. Rebels operating from and supported by Tanzania were instrumental in ending Idi Amin’s regime in 1979 (Atingi-Ego and Sebudde). Similarly, in September 2000, the Guinean army inflicted heavy casualties on the Revolutionary United Front (RUF) of Sierra Leone, which helped to weaken the RUF and pave the way for the resolution of the conflict (Davies). Realistically, though, such interventions usually occur for reasons of national self-interest, because a conflict in a neighbouring country can have adverse spillover effects. Thus an affirmative action across the border is a way to capture that externality. None the less, achieving and maintaining a stable post-conflict economy requires intercountry and intra-regional cooperation.
3
Elite political instability
The role of coups d’é tat cut across many of the case studies and these are usually associated with civil wars. Such elite political instability could constitute an efficient method of unseating an unpopular and dictatorial authority (Fosu, 2002), though it might also be used by a self-interested group to overthrow a democratically elected president, to enable that group to gain access to
Fosu: Conclusion
233
certain rent opportunities (Fosu, 2003). Perceiving their chances of a successful coup as effectively being hampered, opposing coalitions may resort to rebellion to unseat the entrenched government. For example, the Bururi Tutsis overthrew Burundi’s monarchy in 1966 via a coup. This was followed by a series of coups to effect regime change even within that same Bururi group, which led to a very bloody civil war in 1972 (Nkurunziza and Ngarako). The Burundi case also shows that control of the military by the Tutsis probably gave rise to the Hutu rebel action, which was apparently perceived by the Hutus as the only real way of resisting Tutsi domination. Similarly, in Liberia, the seizure of power in 1980 by Samuel Doe via a coup, and his subsequent efforts to suppress opposition, finally resulted in rebel action led by Charles Taylor in 1985 (Davies). The Nigerian situation is another case in point. The first military coup in January 1966 changed the balance of power from the north to the east, with the eastern military elite replacing the northern elite (Garba and Garba). In July 1966, however, there was a counter-coup, resulting in a shift of regional balance back to the north. The Biafra war, which was primarily against the apparent political control of the north, broke out a year later, in July 1967. Hence elite political instability seems to be associated closely with civil wars, albeit in a complicated way. On the one hand, the occurrence of coups could obviate the need to resort to open warfare; but on the other, the prevalence of coups appears to signal the existence of deep-seated problems of governance that might eventually result in civil war.
4
Rent-seeking opportunities
Rent-seeking opportunities act as an incentive for initiating and prolonging conflict, whether as elite political instability (Fosu, 2003) or civil war (Collier and Hoeffler, 1998). As such, it is important to ensure that these opportunities are curtailed. One way to accomplish this is through liberalization and deregulation, though that may only be a starting point (Collier, 2000). For example, the prevalence of elite political instability is observed to be associated positively with the level of exchange-rate overvaluation (Fosu, 2003). For civil wars, however, it is the rent opportunities afforded by primary exports that provide the attraction (Collier and Hoeffler, 1998), as in the cases of Sierra Leone, Nigeria and Angola. Thus there is need to set up transparent systems of scrutiny and control of revenues from primary commodities in the short run, diversify the economy away from high primary commodity dependence in the long run, and provide agreed-upon formulas for sharing government revenues and public-sector jobs between regions and ethnic groups, for example (Collier). Additional measures to rationalize the spending of government revenues from primary commodities are also required, for there is a tendency for governments to be profligate, given revenue windfalls. By exaggerating and
234
Conclusion
overspending the revenues from such sources, governments create further expectations for additional spending. Furthermore, when the revenues contract, the economic downturn can be precipitous and calamitous. Corrupt behaviour may also result (Collier and Hoeffler, 2001). To forestall the natural resource curse, and under pressure from the World Bank, Chad has set up a watchdog agency, the Revenue Oversight Committee, which scrutinizes the use of oil revenues (The Economist, 2003). If constituted and empowered appropriately, such a mechanism should go a long way towards fostering the rationalization required to allocate Chad’s oil resources prudently. The success of such an experiment would surely bode well for replication in many African countries.
5
Child combatants
The use of child combatants is common to nearly all the case studies. This has been a particular problem in Algeria, Angola, Burundi, Congo, DRC, Liberia, Mozambique, Rwanda, Sierra Leone, Sudan and Uganda. It is estimated that more than 120,000 children under the age of 18 years have been engaged in African conflicts (Taylor, 2000). For example, more than 10 per cent of the fighters in the Liberia conflict are children; in Mozambique, the RENAMO rebel group had an estimated 8,000–10,000 children fighting in its force against the government; in Sudan, northern soldiers are reported to have been captured from the south to fight a ‘holy’ war against the south; and in Uganda, it is estimated that around 90 per cent of the soldiers in the LRA rebel force are abducted children. The role of children in the Sierra Leone conflict has been particularly poignant. Some 10,000 children or youths are estimated to have been involved as combatants on all sides in this conflict, committing atrocities beyond belief (Davies). Contributing to this participation was the characteristically low opportunity costs of these young people, as most were hardest hit by unemployment following the economic woes of the 1980s precipitated by the All People’s Congress (APC) government (1968–92). Thus the youths became the recruiting pool for revolutionary activities by rebels, who began increasingly to oppose the one-party dictatorship of the APC. In all these cases children were relatively easy recruits, perhaps given their low opportunity costs and high pliability, as well as non-stringent institutional protection. Thus recent and current efforts by UNICEF and other international agencies to shield children from militaristic exploitation would appear to be in the right direction.
6
The proliferation of cheap weapons
Conflict has become much deadlier. More than 5 million deaths are attributed to conflict in West and Central Africa since 1990. (Woudenberg, 2002).
Fosu: Conclusion
235
These deaths, as well as rape, torture and maiming, which are particularly evident in Sierra Leone and Liberia (Davies), can be blamed in great part on the proliferation of cheap weapons. According to Salopek (2001), over 100 million small arms and light weapons are circulating in Africa today, many having been acquired in the 1990s following the collapse of the Soviet Union. Hence the elimination of the surplus weapons would be a necessary condition for attenuating insecurity in Central Africa (Atanga, 2003) and, by obvious extrapolation, elsewhere on the continent. Although the issue of the proliferation of cheap weapons is not accorded much attention in the foregoing chapters, except in the areas of disarmament, demobilization and resettlement (DDR), these weapons have played a very important role in conflict and post-conflict states of Africa. By lowering the marginal cost of waging war, the availability of cheap weapons is likely not only to ignite wars, but also to prolong them (Collier, 1998; Collier and Hoeffler, 1998). When combined with the existence of easily exploitable natural resources such as rubber, timber and alluvial diamonds, as in the case of Liberia and Sierra Leone (Davies), such availability becomes quite catalytic. Therefore, if appropriate DDR is to be attained in the post-conflict economy, a solution to this problem of a surplus of cheap weapons must first be achieved, as in the recent efforts by ECOWAS, as evidenced by its moratorium of 1998.
7
Justice: inter-ethnic and class equity
Threading through the various case studies is the issue of ethnic rivalry. It is important, therefore, to understand its role in conflict initiation and duration, as well as how the pernicious role of ethnicity can be attenuated through the implementation of appropriate measures. For example, the ethnic polarization problem, exacerbated by the colonial administration, is believed to have played an important role in Nigeria, culminating in the thirty-month Biafran civil war (Garba and Garba). Conversely, the election of an Ibo as a vice-president in 1979 and the pardon of Odumegwu Ojukwu, the former Biafran war leader, helped to preserve peace during the post-conflict era. In the case of Sudan, culture (arguably based on ethnicity) and religion are major determinants of the conflict (Ali and Elbadawi). To help address the problem of ethnicity, Kimenyi proposes the establishment of decentralized institutions of governance involving ethnically homogenous local units of collective choice possessing a fair degree of autonomy. While the proposal seems reasonable theoretically, the proposed system may prove impractical. Though Kimenyi also suggests that ethnically-based governments can be complemented with functional governments that transcend the boundaries of ethnicity, there are bound to be major problems in making such governments functional. For example, what happens to the minorities in localities where a given ethnic group is in the majority? Might
236
Conclusion
such a system not also breed further ethnic rivalry and polarization if there are perceptions of ethnic-based domination? Might it not be better to encourage cross-ethnic socialization, including inter-marriage, in order to blunt the ugly projection of ethnicity? Claude Ake (1996) advanced a similar proposal to Kimenyi’s, and going even further, to include representation by gender and other attributes in a form of proportional representation. Again, there would be major hurdles in implementing this, mainly because of the problem of ‘boundedness’ – that is, how far one must go to be inclusive, given the multiplicity of attributes (Fosu, 1998). A more feasible approach to the ethnicity problem might entail simply ensuring that the rights of ethnic minorities are reasonably guaranteed (Collier). Class politics has also played an important role in the initiation and sustenance of conflict. Alemayehu Geda and Befekadu Degefe, for example, argue that most of the conflicts in Ethiopia can be attributed to rivalry among the aristocracy in their attempts to dominate the peasantry. Indeed, ‘socialistic’ rebellions are usually based on class inequities, with numerous cases, such as Ethiopia (the Dergue), Mozambique (FRELIMO) and Angola (MPLA). Ironically, in the case of Angola, the left-wing MPLA was in fact dominated by the elites, the heirs of the Portuguese colonialists, while its UNITA opposition represented the rural–traditional (Roque). Realistically, class and ethnicity interact in a pernicious way to foster conflicts. Such a situation runs through nearly all the case studies presented in this volume. For example, in Liberia, the Americo-Liberian upper class dominated the political scene right from the inception of the country until the overthrow of the government in a military coup in 1980 led by Samuel Doe, a non-Americo-Liberian (Davies). In Sudan, the relatively well-off northerners have sought for many years to subjugate the south, thus greatly contributing to the conflict; this source has been characterized as ‘culture and religion’ (Ali and Elbadawi). Similarly, in Burundi, the Tutsis have been the dominant class over the Hutus, a situation that has constituted the primary rationale for conflict there (Nkurunziza and Ngaruko).
8
Justice: an impartial and effective legal system
Just as taking up arms may result from the lack of a more appropriate forum to address grievances (Collier and Hoeffler, 2001), the search for justice is a key element in ensuring durable peace (Minow, 1998). The search may range from the milder ‘truth and reconciliation’ model used in South Africa to the court-type commission of enquiry into the Rwanda genocide, and crimes against humanity established in Arusha, Tanzania. In Sierra Leone, and elsewhere, both mechanisms have been used to provide justice. Whether handled internally or externally, the most important aspect of the quest for justice is the availability of an impartial and effective legal system. What remains, then, is the mechanism to ensure a fair outcome of such
Fosu: Conclusion
237
proceedings. As in the case of Burundi, where killers and victims continue to live together and integration among perpetrators seems solid, a special device may be required to induce the truth. Nkurunziza and Ngaruko, for example, recommend a prisoner’s dilemma system of amnesty and justice as the appropriate solution to this problem. That is, a process should be put in place such that suspected perpetrators would perceive truthfulness as a superior option, with the belief that their collaborators would also tell the truth, in which case their failure to do so would invite severe sanctions.
9
The international community
The international community has a central role to play in ensuring an enduring peace in a post-conflict state. First, external agents are required to enforce agreements in a post-conflict state (Addison and Murshed, 2002), given the depreciation of trust among the antagonists during major conflicts. This concept is applicable in most of the cases presented in this volume. Second, to the extent that neighbouring countries are involved, specific suasion, including the possibility of sanctions, from the international community may be called for in keeping potential combatants apart – as in the case of Burundi, Rwanda and DRC, for example. Nor should the role of the international community be limited to neighbouring countries. There is, for example, ample evidence to support the view that Libya provided training and arms to Charles Taylor and others, and that this support fuelled the Liberian and Sierra Leonean conflicts (Davies). More generally, given their respective self-interest, ranging from Cold War rivalry to the security of access to natural resources, external actors have been involved in practically all the conflicts chronicled in this volume. Thus it will require international efforts to deal effectively with such realities. Third, increased external aid may be required as an incentive to encourage peace. Such measures include the infusion of additional revenues to facilitate the process of re-formalizing the economy by reducing the tax incidence on the few who remain in the shrunken formal sector, as well as the use of increased aid to raise the growth rate and thus reduce the risk of the recurrence of conflict (Collier). Finally, in all cases, institutional weakness has been a major basis for engendering war. Conversely, institutional strengthening, supported substantially by the international community, has helped to prolong the peace, as has been the case with Uganda, for example, following Museveni’s assumption of power in 1986 (Atingi-Ego and Sebudde), with Sierra Leone (Davies), and with Mozambique (Francisco and Connerley).
10
Democracy
Finally, the recurrent theme throughout this volume is the need for democracy as a key factor for forestalling conflict and for keeping the peace. Most of the
238
Conclusion
above proposals, including transparency in the use of government revenues, guarantees of the rights of ethnic minorities, and the availability of a system of an impartial and effective legal system, prosper best under democracy. For example, Ali and Elbadawi argue that full, non-compromised democraticparty politics represent the best solution for the Sudanese conflict problem. The many years of peace following the Addis Ababa accord of 1972 broke down, according to the authors, because the democratic institutions in southern Sudan were supplanted by the dictatorial authority of President Nimeiri, who attempted, inter alia, to impose the Islamic laws of the north on the south. In a potentially fractious society, a democratic system that respects the rights of the various groups is necessary to preserve the peace. Similarly, in the Ethiopian case, Alemayehu and Befekadu write: ‘In essence, the only way that the country can end the vicious circle of conflict and instability is when the leaders become accountable to their people.’ According to Djimtoingar and Avocksouma, the competition for resources for redistribution through confiscation by the political authority was one of the root causes of the Chadian conflict. They conclude that ‘a programme of good governance is the best way to end hostilities decisively in Chad’. A similar argument is made for Angola by Moura Roque. Similarly, Francisco and Connerley express the concern that the highly centralized and politically intolerant government of Mozambique continues to mar the state of the post-conflict economy attained in that country. Thus, democracy would appear to be the most appropriate mechanism for attenuating conflict. Ake (1996) proposes that a democratic setting where there is representation by as many attributes as possible (for example, gender, youth, profession) might be a way to achieve peace and development in Africa. There are potential problems, however, associated with implementing Ake’s proposal, as argued above. Realistically, the circumstances are likely to differ across African countries. The real challenge, then, is to define the most appropriate form of democracy to fit a particular set of national conditions.
11
Conclusion
The six decades or so since the Second World War have witnessed many struggles on the African continent. First, there was the struggle for national independence. In many cases, conflict was required to achieve that outcome. Second, there was the struggle to maintain the nation-state following independence. Unfortunately, the seeds of conflict were already sown by the manner in which many of these nations were created in the first place, with national borders incorporating manifold (ethnic) groupings, or simply cutting through ethnic entities. As a mechanism for achieving national unity, a number of these newly formed countries adopted strong central governments. Unfortunately, such governments resulted in a dictatorial entrenchment of power that required forceful removal. Consequently, elite political instability,
Fosu: Conclusion
239
in the form of military coups d’é tat, littered the post-independence terrain of the continent. More recently, open armed insurrections such as civil wars have been particularly dominant (Wallensteen and Sollenberg, 2001). Although civil wars raged in several countries on the continent before the 1990s, they were considerably less frequent. What seems certain, though, is that these conflicts have been particularly protracted and especially deadly thanks to the availability of cheap weapons and huge reservoirs of youths and others with relatively low opportunity costs. The current struggle, then, is to end at least this pernicious form of conflict and attain a state of post-conflict economy. This volume has served to provide a forum for understanding the various issues involved in such an effort. If the final struggle for economic development is to be achieved, conflicts must give way to enduring peace. As Simon Kuznets (1996, p. 451) states: ‘Clearly some minimum political stability is necessary if members of the economic society are to plan ahead and be assured of a relatively stable relation between their contribution to economic activity and their reward.’ A message that emanates clearly from this volume is that the root causes of conflict must be addressed if an enduring state of post-conflict peace is to be attained. That would require judicial reform to provide justice for all: a rule of law that is not arbitrary, and legal systems that respect the rights of minorities as well as the majority. Political liberalization and democratic accountability are further prerequisites. At the time of writing, several African countries have already embarked on such a path, and it is hoped that others will follow suit. While as yet it is too early to tell, it is hoped that these experiments will mature into standard political systems. African countries must, therefore, increasingly demonstrate their commitment towards this course of political transformation. This is arguably the optimal domestic political agenda. Meanwhile, the role of the international community is equally important. Attaining political maturity is a complicated and costly process that requires the nurturing of institutions, and the international community can and should assist. It is hoped that the new initiative, New Partnership for African Development (NEPAD), will help. Finally, as is clear from this volume, many of the conflicts exhibit salient multilateral dimensions and thus require multilateral solutions. Regional groups in Africa could be useful, as exemplified by particularly recent efforts undertaken by ECOWAS/ECOMOG, the OAU and others to resolve disputes and, inter alia, stem the inflow and sale of arms. The role of the external community has been, and remains, just as essential. References Addison, T. and M. Murshed (2002) ‘Credibility and Reputation in Peacemaking’, Journal of Peace Research’, vol. 39, no. 4, pp. 487–501. Ake, C. (1996) Democracy and Development in Africa (Washington, DC: Brookings Institution).
240
Conclusion
Atanga, L. M. (2003) ‘Tackling Small Arms in Central Africa’, BICC paper no. 29, Bonn International Centre for Conversion. Collier, P. (1998) ‘The Role of the State in Economic Development: Cross-Regional Experiences’, Journal of African Economies, vol. 7, supp. 2, pp. 38–76. Collier, P. (2000) ‘How to Reduce Corruption’, African Development Review, vol. 12, no. 2, pp. 191–205. Collier, P. and A. Hoeffler (1998) ‘On Economic Causes of Civil War’, Oxford Economic Papers, vol. 50, no. 4, pp. 563–75. Collier, P. and A. Hoeffler (2001) ‘Greed and Grievance in Civil War’, working paper no. 2355, World Bank, Washington, DC. Economist, The (2003) ‘Can Oil Ever Help the Poor?’, 6 December, pp. 37–8. Fosu, A. K. (1998) ‘Review of Claude Ake, Democracy and Development in Africa’, Journal of Development Studies, vol. 34, no. 4, pp. 171–2. Fosu, A. K. (2002) ‘Political Instability and Economic Growth: Implications of Coup Events in Sub-Saharan Africa’, American Journal of Economics and Sociology, vol. 61, no. 1, pp. 329–48. Fosu, A. K. (2003) ‘Political Instability and Export Performance in Sub-Saharan Africa’, Journal of Development Studies, vol. 39, no. 4, pp. 68–82. Kuznets, S. (1996) Modern Economic Growth: Rate, Structure, and Spread (New Haven, Conn.: Yale University Press). Minow, M. (1998) Between Vengeance and Forgiveness: Facing History After Genocide and Mass Violence (Boston, Mass.: Beacon Press). Salopek, P. (2001) ‘Leftover Arms Fuel Continent’s Ruinous Wars: Cold War Surplus Wreaks Havoc’, Chicago Tribune, 23 December. Taylor, K. (2000) ‘Child Soldiers in Africa: The Problem and the Solutions Fact Sheet’ (Washington, DC: TransAfrica Forum). Wallensteen, P. and M. Sollenberg (2001) ‘Armed Conflict, 1989–2000’, Journal of Peace Research, vol. 38, no. 5, pp. 629–44. Woudenberg, A. V. (2002) ‘Africa at the Crossroads’, briefing paper no. 19, Oxfam GB.
Index of Names Abate, Major A. 129 Abubakar, S. 94, 107(n9), 107 Adedeji, A. 91, 107, 219, 227 Addison, T. x, xvi, 8, 16(n4), 17, 65, 125, 140(n4), 140, 185, 188, 192, 237, 239 Afigbo, A. E. 94, 108 Agarwal, B. vii Ake, C. 236, 238, 239, 240 Alemayehu, G. x, xxi, 125, 127, 128, 131, 133, 140(n1, n4), 140, 141, 236, 238 Ali, A. A. G. x, xxi, 143, 146, 149, 150–1, 153, 155, 156, 161(n3), 161(n7), 162, 231, 235, 238 Alier, A. 147, 162 Amdo, General A. 129 Amin, I. xxi, xxii, 165, 166, 167, 168, 171, 173–4, 232 Ankrah, General 98 Andersson, P. 211 Annan, K. 218 Arrow, K. J. viii Asiwaju, A. I. 94, 108 Atanga, L. M. 235, 240 Atingi-Ego, M. x, xxi–xxii, 163, 165, 171, 173, 176, 179, 231, 232 Atkinson, A. B. viii Augusztinovics, M. vii Auvinen, J. 125, 142 Avocksouma, D. A. x, xix–xx, 109, 238 Azam, J.-P. 4, 5, 17, 43(n9), 44, 48, 56, 125, 141, 186, 192 Azikwe, Dr 95 Balancie, J.-M. 185, 192 Ball, N. 203, 204, 210 Banti, Brigadier-General T. 129 Barnes, S. 203, 204, 210 Barre, President S. 165 Barro, R. 156, 162 Bates, R. H. 16(n1), 17 Befekadu, D. x, xx, 125, 127, 131, 133, 140(n1), 141, 236, 238 Beyem, R. 115, 120(n2), 120 Birmingham, D. 214, 227
Bolnick, B. 211 Bouquet, C. 110, 111, 120 Boyce, J. K. 125, 142 Brough, T. 61, 71 Bruno, M. viii Buchanan, J. M. 68, 71 Buganda, Kabaka of 164, 165, 171 Burki, S. 67, 71 Burnside, C. xviii, xxiv, 19, 44 Burton, J. W. 216, 227 Bush, G. W. 161(n6) Buyoya, President 184, 187, 188 Cabot, J. 111, 120 Calvo, G. vii Campoaré, B. 81 Cão, D. 214 Cardoso, E. A. vii, ix, x Chachiua, N. 200, 211 Charreau, G. 114, 121 Chissano, J. 197 Chowdhury, A. R. 140(n4), 140 Clapham, C. 127, 128, 129, 141 Clifford, Sir Hugh 94, 106(n7) Clinton, W. J. 186 Cobbah, J. A. M. 63–4, 72 Collier, P. ix, x, xvii–xviii, xxiv, 12, 15, 17, 18, 19, 24, 25, 26, 27, 39, 41–2, 43(n2, 10), 44, 45, 47, 48, 49, 50, 52, 54, 56, 58, 62, 71, 72, 91, 97, 108, 109, 121, 125, 140(n2), 141, 150, 156, 162, 163, 168–9, 173, 174, 175, 176, 178, 179, 198, 208, 211, 217, 226(n3), 227, 233, 234, 235, 236, 237, 240 Connerley, E. x, xxiii, 197, 237, 238 Constancio, V. viii Couceiro, P. 214 Cramer, C. 125, 141 Crocker, C. A. 217, 227 Dacko, President 61 Dadi, A. 115, 121 Daniel, A. 132, 141 Danforth, J. 158, 159, 161(n6) 241
242
Index of Names
Davies, V. A. B. x, xviii, 77, 81, 84, 90, 232, 233, 235, 237 de la Grange, A. 185, 192 de Matos, N. 214 de Novais, P. D. 214 de Sousa, C. 14, 17 De Soysa, I. 109 Déby, I. xix, 113, 115–17, 120 Deng, F. 58, 72 Dercon, S. 132, 141 DflD 3, 17 Dick, Brigadier 209(n7) Diguimbaye, G. 111, 121 Dillinger, W. 67, 71 Diogo, I. 206, 211 Dixit, A. K. 11, 17 Djimtoingar, N. x, xix–xx, 109, 238 Doe, S. 80–2, 233, 236 Dollar, D. xvi, xvii, xxiv, 19, 44, 50, 56 Domingos, R. 209(n6) Doyle, M. 152–3, 154, 202, 211 Drèze, J. viii Eddy, M. 214, 227 Egwu, S. G. 104, 108 Eichenberger, R. 70, 72 El-Batahani, I. A. 143, 146, 149t, 153t, 161(n3), 162 Elbadawi, I. A. x, xvi–xvii, xxiv(n1), xxiv, 18, 24, 25, 26, 27, 39, 43, 44, 106(n2), 108, 109, 121, 125, 141, 143, 145, 146, 147, 149t, 150–1, 153t, 157(9n), 161(n3), 162, 163, 168, 169, 186, 211, 221, 226(n3), 227, 231, 235, 238 Ellis, H. viii Engel, D. vii Fitoussi, J.-P. viii, ix Forman, S. 210 Fosu, A. K. ix, x, xv, xxiii, xxiv(n1–2), xxiv, 71, 231, 233, 240 Francisco, L. x, xxiii, 197, 209(n*), 237, 238 Franco, A. 211 Frenkel, J. vii Frey, B. 70, 72 Fukuyama, F. 114, 121 Gaddafi, Colonel 81 Gaffney, P. D. 181, 192 Gali, N. G. 110, 121
Gallup, J. L. 156, 162 Garang de Mabior, J. 148 Garba, A. G. x, xix, 91, 101, 106(n1), 198, 233, 235 Garba, P. K. x, xix, 91, 101, 106(n1), 198, 233, 235 Gaspar, A. 200, 211 Geberie, L. 90 Gebre-Hiwot, B. 125, 127–8, 141, 142 Gleditsch, N. P. 26, 44 Global Witness 226(n2, n7) 227 Gonçalves, F. 202, 211 Goukouni, President xix Gowon, Lieutenant-Colonel 97 Green, F. H. 220, 227 Grellet, G. 111, 121 Grossman, H. I. 48, 56, 125, 142 Guebuza, H. 211 Habré, H. xix, 112–17, 120 Hadfield, H. ix Haile Selassie, Emperor xx, 98, 126, 129, 150 Handoussa, H. vii Hanlon, J. 200, 211 Harberler, G. viii Hazleton, R. 90 Hegre, H. 26, 44 Hess, R. 12, 142 Hilldyard, N. 91, 108 Hirschleifer, J. 5, 17 Hodges, A. 226(n2), 227 Hoeffler, A. 12, 15, 17, 18, 21, 25, 39, 43(n2, n10), 44, 46, 47, 49, 50, 52, 56, 58, 72, 91, 97, 108, 109, 125, 140(n2), 141, 150, 156, 162, 168–9, 198, 211, 233, 234, 235, 236, 240 Hoel, M. vii Horowitz, D. 64, 72 Houphouët-Boigny, F. 5, 43(n9), 81 Human Rights Watch 189, 192 Hyuha, M. A. x Ikime, O. 92, 108 International Crisis Group 158, 162, 186, 187, 188, 192 International Monetary Fund 217, 226(n8), 227 Issa, General 85 Johnson, R.
88
Index of Names 243 Kabbah, T. 79, 86, 88, 89–90 Kabila, L. 6 Kaija, D. x Kakonge, R. x Kakungulu, 170 Kambwa, A. E. 214, 218, 227 Kasekende, L. 165, 171, 173, 176(t), 176, 179 Kaser, M. viii, ix Kimenyi, M. S. x, xviii, 57, 60, 61, 62, 64, 65–6, 66, 68, 71, 72, 235–6 King, M. L. 98 Kingma, K. 211 Kissita, R. B. x Kofele-Kale, N. 220, 227 Kony, J. 167 Kornai, J. vii Koty, A. 117 Kroma, A. 88 Krishnan, P. 132, 141 Kuznets, S. 239, 240 Laffont, J.-J. vii Lagu, General 152 Landa, J. T. 69, 72 Landau, I. 206, 211 Lawal, A. A. 92, 108 Le Billon, P. 8, 17 Lee, J.-W. 156, 162 Lemarchand, R. 181, 192 Levine, D. 126, 142 Licklider, R. 163 Lindahl, E. viii Lipumba, N. H. 222, 227 Little, K. 64, 72 Lugard, Lord 94 Lundberg, E. viii Lyakurwa, W. x Machlup, F. viii Makidi-Ku-Ntima 214, 217 Magnant, J. P. 111, 121 Malinvaud, E. viii Malloum, F. 112, 115, 120 Mamdani, M. 170, 179 Mandela, N. 181, 184–8, 191–2(n8) Manirakiza, M. 182, 193 Marshall, M. G. 4, 17 Mas-Colell, A. vii Mbaku, J. M. 66, 72 Mbilana, G. 200, 211 McCotter, B. C. 214, 227
Mengistu Haile Mariam, Major 128 Mesnard, A. 5, 17 Michalon, T. 111, 121 Micombero, M. 181 Minter, W. 199, 211 Minow, M. 189, 193, 236, 240 Mobutu Sese Seko 6, 8, 215 Moi, President 167 Monga, C. 182, 193 Mosquito, General 82 Moura Roque, F. x, xxiii, 213, 218, 223, 226(n6–7), 227, 238 Mpuga, P. x Mudoola, D. 170, 171, 172–3, 179 Mugabarabona, A. 187, 191(n7) Mugume, A. x Murshed, S. M. xi, xvi, 3, 4, 8, 16(n4), 17, 65, 71, 140, 185, 188, 192, 237, 239 Murtala, Colonel 98 Murungi, K. 62, 73 Museveni, Y. xxii, 89, 165, 166, 171, 172, 173, 174, 176, 179, 232, 237 Mwase, N. xi Mwega, K. 62, 73 Nabli, M. ix Nadingar, J. 112 Nafziger, E. W. 8, 12, 17, 125, 142, 192 Ndayikengurukiye, J.-B. 185, 187, 188, 191(n7) Ndikumana, L. 5, 17 Ndulu, B. xxiv, 44 Ndung’u, N. S. xi, xvi–xvii, xxiv, 27, 44 Ngaruko, F. xi, xxii–xxiii, 180, 182, 183, 189, 191(n1–2), 193, 231, 233, 236, 237 Nimeiri, General xxi, 150, 152, 161(n9), 238 Nkrumah, K. 62 Nkurunziza, J. D. xi, xxii–xxiii, 180, 182, 183, 189, 191(n1–2), 193, 231, 233, 236, 237 Nkurunziza, P. 187, 188, 191(n7) Nnoli, O. 104, 108 Noh, S. J. 125, 142 Nyerere, J. 165, 166, 173, 180, 184–6, 191–2(n8) Obasanjo, O. 98, 108 Obote, M. xxi, 164, 165, 166, 171, 173, 178(n3)
244
Index of Names
Obwona, M. xi Ojukwu, M. xi Okidi, J. xi Okurut, F. xi Olson, M. xxiv(n3) Olukoshi, A. 221, 228 Olusanya, G. O. 94, 106(n7), 107(n13), 108 Osama bin Laden 172 Pankhurst, R. 125, 126, 142 Papi, U. viii Pastor, M. 125, 142 Patrick, S. 210 Pattillo, C. 43(n2), 43, 44 Perry, G. 67, 71 Quiwonkpa, T. 80 Regan, P. 146, 146f, 162 Reinikka, R. 163, 173, 174, 176, 178, 179 Reno, W. 81, 90 Reynal-Querol, M. 125, 140(n2), 142 Reynolds, A. 221, 228 Roback, J. 66, 73 Roberto, H. 215 Robinson, Sir Austin viii Rolin, C. 211 Rowland, R. W. (‘Tiny’) 209(n4) Rudaheranwa, N. xi Russell, B. A. W. 98 Rwasa, A. 185, 187, 191(n7) Rwendeire, A. xi Sachs, J. 156, 162 Salopek, P. 235, 240 Sambanis, N. 24, 25, 26, 39, 43(n10), 44, 46, 91, 106(n2), 108, 109, 121, 125, 141, 145, 146, 147, 152–3, 154, 162, 163, 168, 186, 192, 211, 221, 226(n3), 227 Samuelson, P. A. viii Sankoh, F. 79, 81, 85 Saro Wiwa, K. 100, 101 Savimbi, J. M. xxiii, 213 Schippa, F. K. P. viii Scott, J. 88 Sebudde, R. K. xi, xxi–xxii, 163, 165, 171, 173, 176, 231, 232 Sen, A. viii Sengalane, D. S. 202, 211 Serapião, L. 203
Shimless, A. 137, 141 Shughart, W. F. 61, 62, 72 Sien, T. W. 80 Sindayigaya, J.-M. 192(n9), 193 Singer, J. 163 Sirleaf, E. J. 88 Sisk, T. D. 221, 228 Small, M. 163 Smillie, I. 81, 90 Smith, S. 189, 193 Söderbom, M. 58, 72, 91, 108, 150, 162 Solow, R. vii, viii Ssemogere, G. xi Stewart, F. 8, 12, 17, 125, 142, 192 Suliman, M. 91, 108 Suzumura, K. vii Taylor, C. 77, 80–4, 88–90, 233, 237 Teisserenc, P. 117, 121 Tekeste, N. 133, 134, 142 Togoimi, Y. 117 Tolbert, President 81 Tomballaye, N. 112, 120 Toure, A. T. 88, 90, 226(n1), 228 Tronvoll, K. 133, 134, 142 Tsuru, S. viii Tvedten, I. 215, 228 UNDP 83, 90, 205, 217, 228 United Nations (UN) 16, 17, 82, 90, 211 Urquidi, V. I. viii Väyrynen, R. 8, 12, 17, 125, 142, 192 Vercelli, A. viii Vick, K. 192(n12), 193 Vincent, O. 103, 108 Wallensteen, P. xv, xxv, 3, 17, 239, 240 Walter, B. F. 3, 17 Weeks, J. 137, 141 Weimer, B. 206, 212 Weinstein, J. 209(n*) Woewiyu, T. 88 Wolfensohn, J. 226(n5) World Bank (author) 3, 17, 18, 20, 23, 43(n1), 44, 57, 136, 138, 157n, 217, 228 Woudenberg, A. V. 234, 240 Zenawi, M. 129–30 Ziegler, J. 111, 121
Subject Index
Abidjan Peace Accord (1996) 79 Abuja Ceasefire Agreement (2000) 85 Abuja Peace Accord (1996) 77, 88, 89 accountability xx, 68, 96, 97, 100, 102, 118, 127, 139, 140, 150, 167, 173, 219, 224, 238, 239 Acholi (tribe) 166, 173 Action Group (AG, Nigeria) 95, 96, 107(n12) Addis Ababa peace agreement (re Sudan, 1972) xxi, 150, 151t, 160, 238 Aden 152 Afghanistan 48 Africa institutional problems 68 partition 133 post-colonial leaders 60–1 post-independence institutions and ethnic conflict 59–63 ‘root cause’ of problems 226(n3) see also Central Africa African Development Bank 43(n1) African Economic Research Consortium (AERC) ix, x, xi, 43(n1) African National Congress (ANC) 192(n8) agents 125–6, 171, 186, 237 agricultural goods 156 agriculture 12, 14, 16, 16(n5), 80, 83, 84, 87, 113, 137, 178, 201 commercial 111 smallholder 13, 130 subsistence 217 Aguleri 104 aid xvii, 9, 10, 11, 33, 38f, 43(n7), 88, 201, 206, 209, 219, 222, 237 diminishing returns xvi, 19, 32, 37, 42, 50 effect on growth 37 effective in good policy environment 37 frozen (Eritrea–Ethiopia) 138 Liberia 84 Mozambique 203
multilateral (to Chad) 117 non-conflict countries 50 peace-stabilizing effect 50 post-conflict societies 55 relationship with economic growth 50–1 risk of conflict xvii, 48 saturation point 51 structural 188 aid donors 7, 45, 87, 117, 118, 174, 176, 204, 209(n6) post-conflict societies 55–6 pressure 39 ‘should allocate aid disproportionately to post-conflict situations’ 50–1 timing of aid to post-conflict countries 51 aid effectiveness xvi, xvii–xviii, 15, 19, 25, 31, 35, 36–9, 42 Aja 60 Akan ethnic group (Ghana) 62 Al-Qaeda 172 Algeria 234 Algiers peace agreement (Eritrea-Ethiopia, 2000) 137 All Nigeria Constitutional Conference (1950) 95 All People’s Congress (APC, Sierra Leone) 79, 80, 81, 82, 234 Allied Democratic Forces (ADF, Uganda) 167 Alvor accord (1975) 217 Ambundu (Angolan ethnic group) 214 Americo-Liberians 79–80, 82, 236 Amhara (ethnic group) 58, 126 Amhara Nation Democratic Movement 129 amnesty 189–90, 191, 237 Amnesty Act (Uganda, 2000) 174 AMODEC (Association of Demobilized War Veterans, Mozambique) 210(n10) Andoni province (Nigeria) 94 Anglo–Nigerian Defence Pact 96 245
246
Subject Index
Angola xvi, xxiii, 4, 9, 11, 46, 57, 58, 233, 236, 238 agricultural output 13 causes/root causes of conflict 216, 217–18, 226(n1–4) child combatants 234 decentralization 224–5 donor-supported reconstruction 11 economic dimension 221–3, 226–7(n6–10) effects of conflict 217 ethnic groups 214 fiscal revenue 226(n7) fissure in society 220 government consumption 22 historical overview 214–15 indicators of debt and well-being 20t peace agreements 217, 226(n4) point resources 8 political instruments for national reconciliation 219–21 post-conflict transformation 213–28 poverty-reduction strategy (PRS) 223–4 rents from resources 8 role of civil society 221 slavery 214 social indicators 217 tax revenues 22 transformation from ceasefire to durable peace 217–19 see also independence animists xix, 109, 110, 113 Anti-Terrorist Unit (ATU, Liberia) 89 Anya-Nya 150 apartheid 192(n8), 199, 218 Arabs 99, 143, 147, 170, 171 aristocracy xx, 125, 127, 236 armed forces/armies 54, 61, 127, 128, 155, 159, 173, 220, 222, 224 Burundi 182–3, 184, 192(n4–5), 187–8, 189, 190, 233 Eritrea 134 Ethiopia 136 Guinea 232 Mozambique 200, 203 Rwanda 188
Sierra Leone 85 see also coups d’état Armed Forces Revolutionary Council (Sierra Leone) 86 armed revolt (Nigeria) 93 arms supplies 237 arms-for-diamonds trade 81, 85 Arusha (Tanzania) 184, 236 Arusha peace negotiations (Burundi civil war) 184–7 conduct of negotiations 184 context 184–6 implementation 186–7 ‘peace agreement without peace’ (28 August 2000) 185, 186, 190 representation 185 venue and facilitators 185, 191–2(n8) Asmara 134, 136 Assab 133, 136 Association of Demobilized War Veterans (AMODEC, Mozambique) 210(n10) ‘autonomous ethnic governments’ 63 autonomy 93, 94, 99, 225, 235 Baganda 164 Baguirmi Kingdom 120(n2) Bahiima community 166 Bailundu kingdom (Angola) 214 Bakassi Boys (1999–) 104, 107(n16) Bakassi Peninsula 107(n15) Bakongo (Angolan ethnic group) 214, 215 balance of payments 84 banks 151, 166 Bantu 164 Baya ethnic group (CAR) 61 Belgium 81–2, 181, 231 Benin 60 berberie (red pepper) 135 Berlin Conference (1884–5) 110, 214 Biafra xix, 47, 95, 97–9, 107(n16), 233, 235 see also Nigerian Civil War Bicesse accords (1991) 217 birr (Ethiopian currency) 133, 134, 135 blood on hands 133, 182 Borgu province (Nigeria) 94 ‘boundedness’ 236
Subject Index 247 bourgeoisie national 206 petite bourgeoisie (Angola) 215 Britain see United Kingdom Bubanza province (Burundi) 183t Buganda kingdom 164, 165, 171, 179(n1, 3), 231 Bunyoro’s six ‘lost counties’ 164, 170, 179(n2) Bujumbura province (Burundi) 183t Bukoba (Tanzania) 165 Bunyoro kingdom resistance to British rule 179(n2) six ‘lost counties’ 164, 170, 179(n2) Bura province (Nigeria) 94 Burkina Faso 81, 232 Burundi xv, xvi, xxii–xxiii, 5, 46, 58, 231, 236, 237 assassination of democratically-elected Hutu president (1993) 182, 183, 184, 191(n5), 192(n11) child combatants 234 generalized prisoner’s dilemma approach xxiii, 189–90, 190–1, 237 government 187, 192(n9) Hutu government (1993) 184 indicators of debt and well-being 20t lesson from its own distant past xxii–xxiii, 180, 191 questions of justice 188–90, 192(n11–12) role of international community 187–90, 192(n11–12) state and security institutions (Hutu-controlled) 184 three-year transitional period 187, 188 transition from war to peace 180–93 war with DRC 184 Burundi civil war Arusha peace negotiations 184–7 ceasefire problem 185, 186, 187–8, 191 explanation for war 180 history xxii, 181–4, 191(n2) intensity and duration 182–4 origins and root causes 181–2 polarization 81 requirements for restoration of stability 181
Burur province (Burundi) 181 Bururi Tutsi 182, 183, 192(n4–5) Cabindans 215 cabinets: ethnic composition 61–2 Calabar (Nigeria) 106(n8) Cameroon 99, 107(n15), 120(n2) ‘capacity utilization’ 102, 103 capital 14, 49, 214 public 49 domestic 206 marginal product xxiv(n1) physical 156 capital accumulation domestic 206 efficiency 41 capital expenditure 135 capital flight 49–50 cash crops 111, 172, 178 Catholic Community of Saint’Egidio 202 Catholicism/Catholics 164, 171 Central Africa 57, 235 see also East Africa Central African Republic (CAR) 20t, 61 Central Bank of Nigeria 102, 103 central banks 222, 223 central government xv, 107(n14), 140(n8), 164, 175, 204, 224 Mozambique 205 Nigeria 100, 101 Uganda 172 central planning 213 centralization 60–1, 62, 103 economic and political activity 185–6 fiscal 102, 103 Mozambique 200, 204 Centre for Strategic and International Studies (CSIS) 158, 161(n13) Cervejas Reunidas de Moçambique 210(n15) Chad 109, 234, 238 causes of conflict 120 civil war (1979–82) xix, 109, 112–13 colonial period 110–12 divides xix, 113, 120 economic mismanagement 114–16 economic recovery strategy 117–18 economics of civil conflict xix–xx, 109–21
248
Subject Index
Chad – continued emergency plan (1984–6) 117 historical dimension 110–12 ‘impoverished by its own elite’ 120 indicators of debt and well-being 20t interim plan (1986–8) 117 ‘key cause of conflict’ xx negative impact of civil war 113–14 north–south dualism xix, 110 northern raids 110, 120(n2) oil revenues 234 petroleum legislation 118 prospects 118–19 search for peace 116–18 southern-bias argument 112 Chad Consortium 119 Chad Progressive Party (PPT-RDA) 112 Chad-Cameroon pipeline 48 Chari River 110 children 57, 86, 145 combatants/soldiers 79, 85, 234 one death every three minutes (Angola) 217 China 127 China National Petroleum Corporation (CNPC) 161(n6) Chindonga (Angolan ethnic group) 214 Christians xix, xx, 86, 99, 109, 112, 113, 116, 126, 133, 147, 202, 210(n11) Church of Uganda 171 churches 226 Cibitoke province (Burundi) 183t citizenship 104, 105 civil servants 61, 105, 113–14, 140(n8), 174, 192(n4) civil service/public service xxii, 113, 115, 155, 175 civil society xxii, 61, 62, 85, 86, 96, 97, 103, 104, 105, 119, 205, 218, 220, 221, 222, 224, 226 Nigeria 93, 95 civil wars/civil conflict ix, xv, 3, 156, 157t, 233, 239 Angola 213, 216–17, 226(n1–3) basis 109 breakdown of social contract 15 breakdown of social contract (revenue-sharing) 5 Burundi 180
cause (Ethiopia) 139 cause of development and growth failure 3 causes and consequences 19 ‘complex web of politics, economics, history, psychology, and identity’ 91 cost in lives and displacement of population 79, 83, 84 determinants of risk 19 economics (Chad) 109–21 equilibrium or neoclassical analysis 91 Ethiopia (1974–91) 128, 129–30 ethnic dimension 66 ethnicity and population pressure as causes 91 example of ‘open’ conflict 106(n1) explanations 91, 106(n1) external intervention 146f, 146–7, 161(n3) five propositions (Garba and Garba) 91–2 ‘grievance motive’ 26 Liberia and Sierra Leone (interwoven) 77–90 major explanatory factor for poverty 125 modalities of resolving (Ethiopia) 132–3 motive 91 risk factors 140(n2), 198–9 risk in polarized and diverse societies (Sudan versus Tanzania) 148, 149f, 150 risk in polarized and homogeneous societies (Sudan versus Egypt) 148, 149f, 150 Sub-Saharan Africa 18, 43(n4) vested interests in continuance 47 see also conflict Clifford Constitution (Nigeria, 1922) 93, 94, 106(n7) coffee 134, 214 Cold War xx, 98, 126–7, 128, 138, 201, 237 collaborators (with colonial rule) 106(n6) collective responsibility 114 Collier–Hoeffler model 168–9, 178(n9)
Subject Index 249 Colonial Act (Portugal, 1930) 214 colonial history/colonialism 59–60, 98, 143, 231–2, 238 Angola 214, 215, 216 Burundi 181 Horn of Africa 127 Nigeria 92, 235 Uganda 163, 170, 171, 178(n2) see also history Colonial Office (UK) 92, 106(n5) Commercial Bank of Ethiopia (CBE) 135 commitment technologies (to peace) xvi, 7, 10–11, 158 communality 103, 104, 105 communism xv, 106(n4), 199 comprador class 170 conflict behavioural equations 18 cause of development and growth failure 3 causes 18–19, 21, 43(n2) Colonial Office (British) 92 consequences 18–19, 43(n2) countries with diffuse–resource endowments 8 definitions 163 development tragedy 20–4, 43(n4–5) economic and political consequences 18–44 ethno-regional dimension 58 explanations advanced 58 greed versus grievance dichotomy 4 implications for post-conflict recovery 18–44 ‘inherent’ 92, 106(n1) ‘inherent in humans’ 91–2, 106(n3) national (links between national and global levels) 92 negative effect on economic growth 57–8 ‘open’ 91, 92, 94, 103, 106(n1, n4) perfect conditions 177 policies to reduce 10–11 seeds of 238 spill-over effects 57, 58 temptation to renew 11 see also civil wars; risk of conflict conflict avoidance 57–73 central theme 70
‘conflict oil’ 143 conflict and post-conflict countries (CPCs) 20–4, 41, 42, 235 characteristics 23–4 data 43(n5) economic performance (1970–99) 21f growth and macroeconomic policy stance (1970–98) 23, 23t indicators of debt and well-being 20t risk of conflict 32 ‘special type of HIPC’ 21 structural model of growth and conflict (empirical results) 24–31 tax revenues 22 conflict and post-conflict recovery aid effectiveness 19, 25, 35, 36–9 behavioural equations 18 debt overhang 25, 35–6 economic growth 27t, 31–2, 33, 34, 35f, 36 economic growth, lagged 28t economic and political consequences xvi–xvii, 18–44 empirical analysis 20, 24–6, 43(n6) endogenous growth model 25 estimation results 26–34 five conclusions 19 fixed effects model 26, 27t–31t full information maximum likelihood (FIML) estimation 26, 27t–31t, 31, 34 Hausman test 27t, 28t, 29t, 30t, 31t initial income 25 lagged growth 27t, 28t macroeconomic policies/ environment 22–4, 35, 39–42, 43–4(n9–10) multi-causal structure 19 multivariate estimate 26, 27t–31t, 31, 32, 34 non-linear estimation 26, 27t–31t, 34 overall policy index 25 per capita economic growth 18, 21, 21f, 24, 25, 27t policy environment 25, 28t, 29t, 30t, 36, 43(n7) political and human rights 25, 26, 28t, 29t, 30t, 32, 33, 35 population growth 25, 27t, 29t
250
Subject Index
conflict and post-conflict recovery – continued private investment 18, 21, 21f, 24, 25, 27–8t, 31, 32, 33, 34, 35, 36f, 39 public investment 18, 19, 21, 21f, 24, 25, 28t, 29–30t, 32, 33–4, 35, 36f, 39 random effect model 26, 27t–31t reduced form equations 18, 20, 34–42 risk of conflict/civil war 18, 20, 21, 24, 25, 26, 27t, 28–9t, 31, 32–3, 34, 37, 39 simultaneity and non-linearity problems 31 social diversity 26, 28t, 29t, 35, 40f, 41f structural model (Elbadawi and Ndung’u, 2001) 18–19, 25, 43(n3) sustainability 21 variables 25 conflict prevention 226(n5) conflict re-ignition/reversion 3, 11, 163, 177, 235 avoided (Mozambique) 197 Ethiopia 139 rendered less likely by a decisive military victory 105 risk xxiii, 132, 203, 208, 226(n3), 237 risks generated during previous conflict 46, 47 conflict resolution 158, 159, 160, 186, 208 sowing seeds of the next (Ethiopia) 139 conflict threshold 10 conflict-poverty trap 18 conflicts origins (Toure) 226(n1) regional (multi-country) 150 root causes must be addressed 239 ‘Congo’ [unspecified] 165, 172 Congo-Brazzaville 8, 234 Congo-Kinshasa (Democratic Republic of Congo, DRC; ex-Zaire) 3, 6, 8, 11, 16, 17, 46, 58, 63, 64(map), 167, 184, 188, 232 Burundi refugees 183, 183n, 184 child combatants 234 conscription 210(n8)
Conseil National pour la Défense de la Démocratie (CNDD, Burundi, 1993-) 184, 187 Constituent Assembly (Uganda) 175, 178(n8) Constitutional Conference (Nigeria, 1950) 97, 100 constitutions Chad 116, 118 Mozambique 201 Mozambique (1990) 202 Nigeria 100 Nigeria (1922) 93, 94, 106(n7) Nigeria (1946) 93, 95, 97 Nigeria (1951) 95, 96 Nigeria (1959) 96 Uganda (1962) xxi, 163–4, 165, 173 Uganda (1966, ‘pigeonhole’) 173, 178(n6) Uganda (1995) 167 contracts 215 corruption xxiii, 55, 80, 88, 114, 119, 174, 213, 219, 220, 222, 234 Angola, 216, 226(n2) cost of war xvi Côte d’Ivoire/Ivory Coast xv, 5, 43–4(n9), 77, 81, 99 French intervention 232 national unity 44(n9) cotton 110, 111, 214 councils (Nigeria, 1952) 93 counter-insurgency 48 ‘Country Policy and Institutional Assessment’ (CPIA, World Bank) 25, 28t, 29t, 30t, 32, 33, 34t, 36, 43(n7), 50, 52–3, 157t, 157n country size (risk of war) 169 coups d’é tat 61, 139, 150, 232–3, 239 Burundi (1966) 181 Burundi (1996) 184, 185, 233 Burundi (attempted, April and July 2001) 186–7 Chad (1975) 112 Ghana (1966) 62 Liberia (1980) 80, 233 Nigeria 100, 103 Nigeria (January 1966) 97, 233 Nigeria (July 1966) 97, 98, 233 Rwanda (1959) 181 Sierra Leone (1997) 79, 86
Subject Index 251 Sudan (April 1985) 152 Sudan (1989) 157 Uganda (1971) xxi, 165, (1971) 171, 173, 178(n4) courts (of law) 54, 55, 116, 173 international 190 CPCs see conflict and post-conflict countries CPIA see ‘Country Policy and Institutional Assessment’ credibility xvi, 6–7, 8, 11, 189 domestic economic policies 24 crime 6, 54, 81, 87, 104, 114, 115–16 organized 47 crimes against humanity 186, 188, 236 generalized prisoner’s dilemma approach xxiii, 189–90, 190–1 cross-country regression analysis 106(n2) Crown Agents 106(n5) culture xxi, 63, 111, 140(n3), 161(n1–2), 236 cycles conflict and poverty 22, 24 growth → poverty-reduction → conflict-prevention 24 dams 71 Darfur 120(n3), 161(n12) debt/external debt xvi, 20, 35f, 36f, 37f, 42, 131, 214, 223, 226(n8) ability to pay 24 impact on economic growth xvi, xxiv(n1), 35f debt crisis (Nigeria, 1982) 102 debt overhang 19, 25, 31, 32, 33, 35–6, 42, 105 debt relief/restructuring 22, 87, 88, 222, 223 decentralization xxii, 119, 167, 220 Angola 224–5 Ethiopia 140(n8) Mozambique 205 political and economic, need for 208 political power 63, 67–8, 70, 72, 87, 88 primary advantage 68 decolonization (Nigeria) 95–6 decree, presidential 174 delegation 10 demilitarization 219, 222, 225
demobilization xvii, xix, xxii, xxiii, 54, 79, 85, 86, 175, 197, 203, 208, 209(n6), 210(n7–8), 213, 219, 222, 223 dangers of incomplete 89 Dergue soldiers 136 reintegration of ex-combatants 204, 210(n9–10) democracy xx, xxi, 46, 53, 89, 90, 100, 119, 146, 148, 152, 173, 186, 197, 220, 221, 237–8 lack of (risk factor) 148 multi-party (‘unlikely to resolve Africa’s problems’) 62–3 Sierra Leone 86 southern Sudan (1972 agreement) 150 ‘democracy aid’ 203 Democratic Party (Kenya) 62 Democratic Party (DP, Uganda) 164 Democratic Republic of Congo (DRC) see Congo-Kinshasa democratic transition 145, 153, 158, 159, 160 essential for transition from war to peace 152 Democratic Unionist Party (DUP) 157, 161(n9) democratization xxii, 26, 63, 150, 175, 214, 224 demonstrations 97, 112 Dergue (Ethiopia, 1974–91) xx, 126, 128–33, 136, 139, 236 civil war 129–30 expenditure on defence (impact on growth) 130–2 modalities of resolving civil war 132–3 development xxi, xxiii, 43(n3), 64, 80, 156, 219, 226(n3) issues of economics vii sustainable 213, 224 tragedy of conflict 20–4 see also economic development development costs (Ethiopia) 135–6 development trap 20 ‘dia’ compensation 116 diamonds 4, 8, 46, 82, 83–4, 87, 213, 216, 217, 222, 223, 235 illicit 6, 81, 81t, 89 Lebanese cartel 87
252
Subject Index
diamonds – continued UN certification scheme (Sierra Leone), 87 diasporas xvii, 47–8, 127 Liberian 80 Mozambican 199, 209(n2) dictators/dictatorship 80, 100, 160 diffuse resources 8 diplomacy 202, 209(n4) disabled people 138 disarmament 79, 85, 86, 167, 210(n11), 222 incomplete 89 disarmament, demobilization and resettlement (DDR) 235 discrimination 104, 186 disease 57, 85, 217 dispute resolution xxii, 4, 175, 239 district government (Uganda) 166, 173 divide and rule policies 60, 106(n6), 158 Djibouti 136, 161(n11) Donsos 86 drought 130, 138, 201 drugs 6, 85 East Africa 143 see also ‘Horn, East, and Central Africa’ ‘East Asia’ 156, 157t economic collapse xviii, 89 Economic Commission for Africa 43(n1) Economic Community of West Africa Monitoring Group (ECOMOG) 77, 79, 86, 88, 89, 239 Economic Community of West African States (ECOWAS) 77, 82, 83, 235, 239 economic crisis (Mozambique) 201, 209(n3) economic decline (risk of conflict) 46 economic development 118, 145–6, 239 initial levels (risk of war) 169 interdependence with political development, 33 low levels (risk factor) 148 see also development economic diversification xvii, 48, 106(n2), 156, 159, 222, 226(n7) economic growth xvi, xvii, 11–12, 14, 15, 18, 19, 21f, 22, 23, 31–2, 33, 37,
39, 40, 41, 42, 43, 48, 125, 139, 157n, 169, 178, 222, 237 benefits of peace 138–9 Chad 109–10 civil war risk factor (Mozambique case) 198, 198t contribution of policy improvement 51–2 effects of conflict (Ethiopia) 136–7, 137t, 139 and external debt 35f impact of debt xvi, xxiv(n1) impact of military spending 130–2 impact of oil (Sudan) 155 impact of political instability 176t, 176 institutional reform a prerequisite 59 measures inimical (Sierra Leone) 80 Mozambique 197, 206 negative effect of conflict 57 Nigeria 102 post-conflict prospects (Sudan) 143–62 post-conflict societies 45–6, 49–53 prospects for sustainable peace 153 raised by policy reform 53–4 reconstruction-led 12 reduces risk of conflict 53–4 relationship with aid 50–1 risk of conflict reduced 46 Sudan 150–1, 151t economic mismanagement 58, 150–2, 213, 216 economic opportunities 177, 178, 204, 226(n3) inclusive 206, 208 economic performance 18, 19, 25, 43, 53 conflict and non-conflict countries (1970–99) 21f Ethiopia 125–42 Liberia 80 risk of conflict 46 economic policy 102 economic recovery xxiii, 49, 203–4 economic reform 14, 24 Economic Rehabilitation Programme (PRE, Mozambique) 209(n3) economic restructuring 200, 201
Subject Index 253 Economic and Social Rehabilitation Programme (PRES, Mozambique) 209(n3) economics vii, 15, 91, 109–21 ECOWAS see Economic Community of West African States education xviii, 101, 113, 118, 120(n1), 145, 156, 172, 186, 200, 215, 217, 226(n5) effect on economic growth (long term) 53 government spending (Mozambique 1985–95) 202t lost opportunities 57 primary 14, 53 educational attainment 25 civil war risk factor (Mozambique case) 198, 198t risk factor (outbreak of conflict) 168 Egypt 126, 152 avoidance of civil war 148, 149f, 150 Elbadawi–Ndung’u model 18–19, 43(n3) Elbadawi–Sambanis model 169, 178(n9), 199 elections 221 Burundi (1993) 184 Ethiopia 130, 140(n8) indirect (Uganda) 164 Liberia (1985) 80 Liberia (1997) 77, 82, 88, 89 local (Mozambique, 1998) 205 Mozambique 203 Mozambique (1994, 1999) 197, 204 national 208 national (Mozambique) 206 Nigeria 100 Nigeria (1951, 1952) 95 Nigeria (1964, 1965) 97 Nigeria (June 1993) 100, 101, 103, 107(n16) Nigeria (2003 forthcoming) xix, 105 Sierra Leone 80 Sierra Leone (1996) 79 Sierra Leone (2002) 86 Uganda 164 Uganda (1980) 166 Electoral Bill (Nigeria, 2001) 105 electoral commission 186 electricity 80, 83, 88, 101
Eleventh of September (2001) 127, 172, 226(n5) elite political instability (EPI) xv, xxiv(n2), 232–3, 238 elites Angola 215, 216, 219, 220, 226 Bururi Tutsi 182 Chad 111, 112, 120 eastern Nigeria 233 educated xx, 93, 125, 129, 130, 132, 139 elites 127, 128, 236 Ethiopia 129, 130, 132, 133, 139, 141 governing 197, 201, 204, 206 highland Eritrean and Tigrian 140(n6) Hutu (Burundi) 181 Mozambique 197, 201, 204, 206, 208, 210(n12) MPLA 216 Nigeria 93, 95, 96, 97, 104, 105 northern Nigeria 233 political 174–5, 226 in power 126 ruling 141, 183 Uganda 171, 174–5 urban 215, 220 embezzlement 115 emigration 83 empiricism xvi, 19, 43(n6) breakdown of peace agreements 3 post-conflict recovery 20, 24–31 employment 8, 16(n2), 47, 67, 84, 215, 233 empowerment 177, 224 enclave regions 105 endogenous growth 19 Enhanced Structural Adjustment Fund (ESAF), IMF 22–3, 118 entrepreneurial class 155 Enugu 98 environment (natural) 91–2, 101, 116 see also policy environment Equatoria Province (Sudan) 144(map), 145 ‘era of princes’ (Ethiopia) 126 Eritrea 6, 129, 132 arms embargo 138 economic relations with Ethiopia 133–4
254
Subject Index
Eritrea – continued ethno-linguistic groups 133 historical background 133 IGAD 161(n11) ‘neither peace nor war’ (2000-) 138 rebels against Ethiopia 150 secession from Ethiopia (1993) 133 tax revenues 22 Eritrea–Ethiopia war (1998–2000), see Ethiopia–Eritrea War (1998–2000) Eritrean People’s Liberation Front (EPLF) 129, 130, 132, 133 Eritreans 58 Ethiopia 58, 60, 152, 236, 238 arms embargo 138 background to conflict 126–7 case study xx, 125–42 conflict ‘major explanatory factor for poverty and backwardness’ 139 current account deficit 132 cycle of revolt and conflict 128 Dergue (1974–91) xx, 126, 128–33, 139 divisions 125 economic relations with Eritrea 133–4 emergency recovery programme 138 external intervention 126 further conflict ‘probable’ 140 geopolitical importance xx, 127, 128, 138 IGAD 161(n11) imperial period (pre-1974) xx, 126, 127–8, 139 import-export gap 136 indicators of debt and well-being, 20t ‘king of kings’ system 128 ‘militaristic’ 126, 133 ‘neither peace nor war’ (2000-) 138 north 129 political reform still needed 140 resolution of one conflict sows seeds of next xx revolution (1974) 125, 128 transitional government 130 unity in diversity 126 Ethiopia–Eritrea War (1998–2000) xx, 3, 126, 127, 133–8, 139, 140(n5–7) border issue 137, 140(n7) buffer zone 137–8
changes in magnitude and composition of public expenditure 136 characterization of causes and duration of conflict 133–5 cost to Ethiopia 135–7 development costs 135–6 effects on economic growth 136–7 peace agreement (Algiers, 2000) 137 post-conflict issues 137–8 re-routing of freight via Djibouti 136 reconstruction, infrastructure, resettlement of displaced people 137–8 root causes (need to address) 138 UN peacekeeping force 138 Ethiopia-Somalia conflict 127, 128, 130, 138 Ethiopian Airlines 134, 135–6 Ethiopian Orthodox (Christians) 126 Ethiopian People’s Revolutionary Democratic Front (EPRDF) 129–30, 132, 139 Ethiopian People’s Revolutionary Party (EPRP) 129 ethnic asymmetry (Nigeria) 93–4, 97, 106(n6) ethnic conflict discriminatory allocation of resources 61 post-independence institutions 59–63 ethnic diversity 68, 225 ethnic dominance/domination xvii, 47, 48, 129, 140(n2), 168, 226(n3), 236 civil war risk factor (Mozambique case) 198, 198t ‘ethnic’ goods’ 59, 68 ethnic governments 70–1, 72 ethnic groups 43, 113, 114, 115, 128, 233 government favouritism 61 ‘optimal size of club’ model 68–9, 69f, 70f, 71f ethnic jurisdictions 70–1 ethnic minority rights xvii, 236, 238 ‘ethnic nations 64 ethnic polarization 146, 161(n2), 236
Subject Index 255 ethnicity 4, 5, 104, 126, 161(n2), 231, 235–6, 238 central theme 70 concept ‘needs to be taken with caution’ (Burundi) 180, 191(n1) conflict avoidance xviii, 57–73 ‘convenient ideology in struggle for power’ 130, 132 ‘efficient at overcoming collective action problems’ 66 equity 235–6 ‘ideological mystification’ 182 manipulation 216 negative effects 171 overlapping jurisdictions 70–1 risk of conflict 46 units of collective choice 63–70, 71f, 72, 235–6 ‘used as survival strategy’ by Doe (1980s) 80 and violence 16(n1) voluntary and business organizations 65 ethno-linguistic diversity/states 132, 148 ethno-regionalism 95, 96, 100 Ewe ethnic group (Ghana) 62 ex-combatants xix, 204, 209, 210(n9–10) exchange rate management 24 exchange rates 49, 223 misalignment xxiv(n2) nominal 39 over-valuation 80, 233 real 25, 27t, 28t, 32 exiles 165, 181 exogenous variables 35 exports 111, 216 Angola 214 Eritrea 133 primary 24, 25, 26, 29t, 30t, 34, 35, 39, 43 risk of conflict xvi–xvii external agency (peace-building) 158, 159 external geo-political actors 199 external intervention 139, 148, 150 expected 148 global index (Regan) 146, 146f Sudan 143, 146–7, 161(n3)
extortion 89, 115–16, 170 extra-budgetary expenditures 226(n8)
222–3,
‘fair shares’ principle 103 family 84, 114 famine 128, 223 Federal Military Government (Nigeria) 103 federalism 68, 70 Fellatas (Peulhs) 120(n2) financial depth 25, 28t, 32 first-mover advantages 9 fiscal accounts 226(n8) consistency 18 deficit 23, 23t discipline 22 disputes 5 policy 18, 19, 24, 42, 155 reform 14 support 22 system 4, 92 transfer 5, 11–12, 16(n2) force 94, 111, 186 Forces Armées Rwandaises (FAR) 184 Forces pour la Défense de la Démocratie (FDD, Burundi, 1993-) 184, 185, 187, 188, 191(n7) FORD Asili (Kenya) 62 FORD-Kenya 62 foreign direct investment (FDI) 23, 24, 206, 207f, 210(n13), 222, 223 see also investment foreign exchange 102, 131, 135 formal sector xvii, xviii, 55, 83, 237 Fort Archambault 112 fractionalization ethnic 146, 169, 170–1 ethno-linguistic 26, 28t, 29t, 40f, 41f, 147–8, 149f, 177 north-south divide (Uganda) 170, 171–2, 178 religious 140(n2), 148, 169, 171 social xvi, 19, 40–1, 42, 43, 168–9 see also polarization France 99, 110 fraud 114, 119 free-rider problem 66 Freetown 79, 80, 84–5
256
Subject Index
Frente de Libertaçao de Moçambique (Frelimo) 15, 197, 199, 200, 201, 204, 208, 209, 210(n10), 236 Frente Nacional de Liberte de Angola (FNLA) 215 FROLINAT see National Liberation Front (Chad) Front pour Démocratie au Burundi (FRODEBU) 184, 185, 192(n9) Front National de Libération – Parti pour la Libération du Peuple Hutu (FNL-PALIPEHUTU) 185, 187, 191(n7) Front for National Salvation (FRONASA, Uganda) 165 Frontline States 199 full information maximum likelihood (FIML) estimation 26, 27t–31t, 31, 34 functional jurisdictions 70–1, 72, 235 Ga ethnic group (Ghana) 62 Gabon 99 Gambia 60 game theory (neoclassical) 125 Ganwa (ethnic group, Burundi) 181, 182, 191(n1) gas 99, 101 Gbethis 86 GDN (Global Data Network) database 151n, 161(n5) gender 238 General Peace Agreement (GPA, Mozambique, 1992) xxiii, 197, 202, 203, 204 generalized least squares 31 genocide 58, 186, 188 of Hutus by Tutsis (Burundi, 1972) 182, 191(n4), 233 of Tutsis by Hutus (Burundi, 1993) 182, 183, 185, 192(n10–11) Rwanda (1994) 182, 236 geoethnicity 63–4, 68 geography 111, 156, 157t Germany 110 Ghana xv, 62 Gini coefficient 155, 161(n8) Gios (ethnic group in Liberia) 80, 82 Global Coalition for Africa 43(n1) globalization 119
gold 83, 84, 109, 165 Gorane tribe 115, 116 governance xix, 52, 87, 117, 119, 159, 173, 174, 208, 216, 219, 221, 233, 238 Burundi xxii–xxiii, 182 conflict avoidance 57–73 critical role 58 decentralized (Kimenyi) 235 democratic xxiii, xxi, 160, 213, 218 local ethnic jurisdiction 59 Mozambique 204 post-conflict (Nigeria) 99–102, 105 Sudan 158 government consumption 18, 22, 34, 35, 37f, 37, 39, 40, 137 government contracts 16(n2) government expenditure 9, 176, 201, 223, 233–4 apportionment 4–5 early post-conflict years 55 magnitude and composition (Ethiopia) 136 Mozambique (1985–95) 201, 202t off-budget 16 pro-poor 14 war-related 22 government overthrow probability (versus chances of it making fiscal transfer) 6 government revenue 18, 21, 21f, 24, 25, 30–1t, 33–4, 35–7, 39, 40–3, 87, 140(n8), 238 control xvii generation (Nigeria) 92 scrutiny 233 sharing xvii, 47, 153, 158, 233 transparency 47 windfalls 233 see also oil revenue government/s 45 chance of enduring peace 6–7 choice between fighting rebels and offering them fiscal transfer 5 constitutional 99 Ethiopia 139 local 204, 208 Mozambique 210(n13) national 204
Subject Index 257 post-conflict commitment to peace 6–7 provincial 208 Great Nile Petroleum and Oil Corporation (GNPOC) 161(n6) grievances xxii, xxiii, 3, 4, 15–16, 26, 47, 91, 127, 128, 139, 165, 168, 172, 177, 199, 201, 204, 226(n3) consideration in peace settlements 209 cultural/linguistic 130 exploited by external geo-political actors 199 persisting after end of civil wars 163 source of conflict 15 gross domestic product (GDP) 22, 43(n4) Mozambique 205t, 205 Mozambique (informal sector) 210(n14) overall surplus or deficit as shares of 22 sectoral contributions (Ethiopia) 130–1 gross national product (GNP) Angola/Sub-Saharan Africa 217 Group of Seven 188, 191(n6) Group of Ten 188, 191(n6) growth see economic growth Guinea 85, 232 Guinea-Bissau 11 Guragi (ethnic group) 126 Gwari province (Nigeria) 94 Hague, The: Eritrea–Ethiopia commission 137–8, 140(n7) Hausa 104 head-count ratio 155, 161(n8) health/health care xviii, 14, 53, 101, 109, 113, 118, 156, 215, 217, 226(n5) government spending (Mozambique 1985–95) 202t health care personnel 200 Herero (Angolan ethnic group) 214 highly-indebted poor countries (HIPCs), 20, 20t, 21, 24, 87 history xviii, 91, 97, 106(n2), 133, 139, 161(n2, n4) see also colonial history HIV/AIDS 85 Horn, East, and Central Africa: case studies xx–xxiii, 123–93
Burundi xxii–xxiii, 180–93 Ethiopia xx, 125–42 Sudan xxi, 143–62 Uganda xxi–xxii, 163–79 see also Southern Africa household survey data 167, 176–7 human beings ‘conflict inherent in’ 91–2, 106(n3) human capital 12, 57, 83, 113, 156, 157t, 216, 223 Human Development Index (HDI) 20, 21, 80, 83t Angola 217 Chad 109 CPCs (1997) 20t human poverty index (Chad) 109, 120(n1) human resource development 223 human rights 85, 117, 161(n1), 164, 173, 220 Hutus xxii, 58, 180–93, 231, 236 Ibo/Igbo 58, 98, 99, 104, 235 identity 64–6, 91, 93, 94, 99, 161(n1), 218, 220, 221, 225, 226 identity cards 103, 106 ideology 65, 129, 130, 132, 133, 165 Ife/Modakeke (Nigeria) 104 IGAD see Inter-Governmental Agency for Development Ijaw province (Nigeria) 94, 104 illiteracy 109, 168, 171, 217 IMF see International Monetary Fund Implementation and Monitoring Committee (Burundi) 186 incentive 10 inclusiveness principle 158 income levels 46, 175 independence xv, 189, 238 Angola (1975) 213, 215 Burundi (1962) xxii, 180 Chad (1960) xix, 112 Eritrea (1993) xx, 133 Liberia (1847) 79 Nigeria (1960) xix, 96 Sierra Leone (1961) 80 Sudan (1956) xxi Uganda (1962) 163, 171, 175 ‘Independence Motion and Resolution’ (Sudan, 1955) 145
258
Subject Index
indígenas (Angola) 215 indirect rule 106(n6), 170 industrial sector 130–1, 135 industry 137 inequality 3, 15, 58, 157n, 161(n8), 216, 217, 223, 226(n5) infant mortality 20, 20t, 21, 217 inflation 24, 25, 27t, 28t, 29t, 30t, 32, 39, 103, 223 CPCs and SSA (1970–98), 23, 23t inflation tax 101 informal economy (Mozambique) 210(n14) infrastructure 14, 24, 49, 83, 117, 118, 125, 138, 156, 159, 200, 217, 220, 223, 224, 225 communications 106(n7) Ethiopia 135 Maputo City 205 institution-building/rebuilding 173, 176, 177 institutional capacity 87 environment 19 reform 11, 87–8, 188 weakness 237 institutions 92, 133, 139, 226 Chad 111 collapse 12 conflict avoidance 57–73 democratic 11, 60 democratic resilience 173 fiscal 11 public/state 24, 181 transformation 218 Uganda 172–4, 174–5 weak 177, 216 insurgency (insurrection/revolt/ uprisings) xv, xxii, 93, 111, 125–6, 129, 139, 150, 152, 172, 174, 178, 239 Integrated Approach to Aid Coordination (IAAC) 88 intellectuals/intelligentsia 128, 215 Inter-Governmental Agency for Development (IGAD) xxi, 150, 157, 158, 160, 161(n11) declaration of principles (1994) 158 inter-marriage 126, 128, 236
Inter-Religious Council (Sierra Leone), 86 interest groups 67, 118, 173, 175, 190 interest rates 103 real 25, 28t, 32, 39 internal security 217 internally-displaced persons (IDPs) xxiii, 223 Angola 213, 219 Burundi 180 Ethiopia 138 Mozambique 200 Sudan 159, 160 see also refugees international community xxiii, 138, 180, 184, 185, 188, 190, 201, 208, 223, 237, 239 monitoring Mozambique peace process 202–3 International Development Assistance (IDA) 134 International Economic Association (IEA, 1950–) vii–viii, ix international intervention xx international judicial commission (Burundi human rights abuses) 186 International Monetary Fund (IMF) 22, 117–18, 151, 155, 201, 209(n3) 227(n9) International Monetary Fund: Enhanced Structural Adjustment Fund (ESAF) 22–3 internet 48 interviews 201, 203, 209(n2), 207, 210(n9) investment 23, 37, 53, 55, 150, 156, 176 aggregate 25, 27t domestic 205, 206 impact of military spending 132 lagged 32 private 14, 17(n6), 18, 19, 32, 39, 40, 41, 42, 49 public xvi, 18, 32, 33, 39, 40, 43, 44(n9) risk-free 101 see also foreign direct investment Investment Promotion Centre (CPI, Mozambique) 206
Subject Index 259 Ishekiri (Niger Delta) 104 Islam/Muslims xix, xx, 86, 99, 109, 110, 112, 113, 115, 126, 133, 143, 147, 171 customary law 116 mediators (Uganda) 171 Nigeria 94, 106(n9) ‘proliferation of mosques’ (Chad) 116 sharia law xxi, 101, 105, 152, 157, 161(n10), 238 Israel 99 Italy 126, 127, 133 joint ventures 206 Juba (Sudan) 145 judiciary/legal system xvii, 115, 119, 173, 183, 190 impartial and effective 236–7, 238 independence 220 reform 87, 239 justice 104, 188, 191, 237 justice-seekers/seeking 169, 177 Kabaka Yekka (KY) 164–5 Kagera Salient 165 Kamajor 82, 86 Kampala ix, 167 Kano 96, 104, 107(n12) Kano accords (1979) 116, 117 Karamojong 167 Katikiro (Prime Minister) of Buganda 164 Kenya xv, 60, 62, 161(n11) Kenya Africa National Union (KANU) 62 Khartoum 159 Khartoum accords (1977) 116 Kikuyus (Kenya) 62 kingdoms (African) 60 Kirundo province (Burundi) 183t ‘kleptocracy’ 79, 80, 88 Kokadam Declaration (1986) 157 Kongo kingdom (Angola) 214 Kono District (Sierra Leone) 86, 87 Krahn ethnic group (Liberia) 80, 82 labour 12, 83, 111, 172, 214, 215 labour-intensive processed goods 156 Lagos 94, 104, 106(n8), 107(n12)
Lagos accord (1979) 116, 117 land xvii, 104, 110, 126, 128, 139 Land Agreement (Uganda, 1900) 171 land redistribution 128, 139 land shortage 54 land tenure 111 Land Use Act/Decree (Nigeria, 1978) 100, 101, 104 landmines xix, 12, 113, 203, 217, 219 Langi 173 language 63, 126, 161(n2) Amharic 126, 130 French 185 Kirundi 185 Portuguese 215 ‘power’ (Angola) 215 law 137, 221 Law of Nigeria (1914) 93 Laws of Nigeria (Chapter 130) 107(n10) law and order 43 law reform xxii, 175 lawlessness 115–16 leadership 67, 181, 185, 192(n10) Bururi 182 dictatorial xv tribal (Sudan) 159 League of Nations 79 Legislative Council (Clifford Constitution, 1922) 94 Legislative Council (Lagos) 94 legitimacy 199 definition 209(n1) liberalization 209, 233 economic 176 political 106(n2), 204, 239 Liberia xv, xvi, 232, 236, 237 causes of civil war 79–81 child combatants 234 civil war (1989-) xviii context and background 77–8 ethnic rivalries 82 ‘exported’ diamonds in excess of output 81, 81t, 82 factors sustaining civil war 81–3, 88–9 indicators of debt and wellbeing 20t national army 77 proliferation of cheap weapons 235 and Sierra Leone (interwoven civil wars) 77–90
260
Subject Index
Liberians United for Reconciliation and Democracy (LURD) 78 Libya 81, 82, 111, 152, 237 life expectancy 120(n1), 217 literature aid 19 aid effectiveness xvi, 25, 31, 32, 33, 37, 39 debt overhang 25, 31, 33 determinants of social conflicts 40, 44(n10) economics of conflict 39, 125, 140(n1) length of peace following long wars 31 political rights 39 political science 169 risk of civil war 19, 26, 32, 33 single equation models 42 living standards/standard of living 8, 15–16, 80, 106(n2), 112, 176 loans 9, 92, 106(n5), 137, 173, 208, 223 local government 200, 205, 224–5 Local Government Statute (Uganda, 1993) 175 local interests 208 logit-regression 125 Lomé Peace Accord (1999) 79, 85 long-term development strategy (LTDS) 221, 226(n6) Lonrho 209(n4) looting xviii, 9, 10, 12, 169 Lord’s Resistance Army (LRA) 167, 172, 174, 234 loss of life 125, 234 Angola (children) 217 Biafra 98 Burundi 180, 183, 183t Ethiopia 135 see also genocide; massacres low-income countries 43(n7), 46, 53, 54 Luanda 214, 220 Luhyas (of Western Kenya) 62 Lukiiko (Buganda Parliament) 164, 165 Lunda-Chokwe (Angolan ethnic group) 214, 215 Luo tribe (Kenya) 62 Lusaka protocol (1994) 217, 222 ‘Luwero Triangle’ 166
macroeconomic adjustment 208 environment 39–42 management xix, 102, 103, 105, 157n, 220 policy xxiii, 22–4, 32, 52–3, 134 stability 25, 222 stabilization 221, 223 majority–minority divide (Nigeria) 93, 96, 97, 98, 99, 103, 105 Malawi xv malnutrition xxiii, 213, 217 Manos (ethnic group in Liberia) 80, 82 manufacturing 131, 178, 214, 224 Maputo 199, 203, 205t, 205, 210(n10) Margi province (Nigeria) 94 marginalization 177, 181 Maritime and Transit Service Enterprise 135 market economy/forces xxii, 92, 201, 206 Marxists 129, 130 massacres 82 Burundi 183 Chad 114 Liberia 80 Sudan 145 of Tutsis by Hutus (April 1972) 191(n4) see also genocide; loss of life Mbaka ethnic group (CAR) 61 Mbayes 115 media 98, 99, 225 xxii, 175, 218, 221 Meison (All-Ethiopian Socialist Movement) 129 Memoranda of Understanding (Nigeria, 1988, 1991) 101 Memorandum of Understanding (Angola peace terms) xxiii, 213 Mende 82 Mengo (Buganda) 164, 179(n1) mestiços (Angola) 215 Middle East 8, 127, 156 military expenditure 9, 22, 48, 55, 56, 81, 113, 119, 135, 136, 217 Ethiopia (impact on growth) 130–2 Mozambique (1985–95) 202t ‘not influenced significantly’ by aid 49 opportunity costs 16
Subject Index 261 ‘retards growth without deterring rebellion’ 49 trade-off with transfers to rebels 6 military tribunals 173–4 militias 82, 86, 101–2, 103–4, 105, 107(n17), 112, 187, 222 Millennium African Renaissance Programme 57 minerals 25, 84, 158, 213 minority rights xvii, 48, 239 missionaries 147, 171, 214, 215 Mocubela (Mozambique) 209(n7) modernization 64, 159 socialist tenets 199 Mogadishu peace accord (re Uganda-Tanzania, 1972) 165, 174 monarchy 127, 181–2, 191(n3) monetary policy 18, 19, 24, 35, 103, 155 monetary union (Eritrea-Ethiopia) 133, 134 money demand 18, 19, 24, 25, 34, 39, 42 monopolies 101, 102 Monrovia 77, 79 moral hazard 97, 107(n14) mortality rates 57 Moshi Conference (Tanzania, 1979) 166 Movement for Actualization of Sovereign State of Biafra (1999–) 107(n16) Movement for Democracy and Development (MDD, Chad) 117 Movement for Democracy and Justice in Chad (MDJT) 117 Movement for Survival of Ogoni People 107(n17) Movimento Popular para a Libertaçao de Angola (MPLA) 215, 216, 220, 236 Mozambican Christian Council (CCM) 210(n11) Mozambican Defence Force (FADM) 210(n8) Mozambique xxiii, 14, 20–1, 47, 58, 197–212, 236, 237, 238 agriculture and services/trade (model) 12–13 challenges 197 child combatants 234 consumption 205t donor-supported reconstruction 11
economic growth 12 electoral system 204 GDP (1998) 205t indicators of debt and well-being 20t investment outlook (1985–2000) 207f lessons 208–9 national and regional views 204–6, 210(n11–12) peace process: domestic and external factors 201–4 peace triangle 152–3, 154t post-conflict political and economic processes 204–8, 210(n11–15) poverty 205t, 205 prospects for peace (1992) compared with Sudan (1972) 153, 153f, 154t regional (north–centre–south) 205t, 205 regional poverty 15 rural–urban 205t Mozambique Aluminium 210(n13) Mozambique civil war 198–200 causes and nature 199–200 consequences 200 risk factors 198–9 multi-party system 186, 200 Muramvya province (Burundi) 183t Mwanza (Tanzania) 165 Nairobi 135 Nairobi peace accord (re Uganda, 1985) 167, 174 nakfa (Eritrean currency, 1997-) 134 Nampula Province 210(n9) nation-building 97, 99–102, 105, 106, 218 nation-states xv, 4, 238 formation, consolidation, dissolution 92, 106(n2) Nigeria 91–108 National Committee for Peace and Democracy (CSNPD, Chad) 117 national consciousness 95 identity 220, 225 integration 99, 115 reconciliation (Angola) 219–21 resources 117, 120
262
Subject Index
national – continued sovereignty 102 unity 60–1, 82, 132, 213, 220, 221, 238 National Development Party (NDP, Kenya) 62 National Islamic Front (NIF, Sudan) 157 National Liberation Front (FROLINAT, Chad) xix, 110, 112, 116 redressing injustices felt by northerners 114 National Movement for Total Independence of Angola (UNITA, 1966–) xxiii, 4, 9, 213, 215, 220, 236 National Patriotic Front of Liberia (NPFL) 77, 78, 80 National Provisional Ruling Council (NPRC, Sierra Leone) 79 National Reconstruction, Resettlement and Rehabilitation Programme (Sierra Leone) 86 National Recovery Committee (CNR, Chad) 117 National Resistance Army (NRA, Uganda) 166, 167 National Resistance Movement (NRM, Uganda) 163, 166 National Salvation regime (Sudan) 155 National Union Government (Chad) xix nationalism 96, 171 nationalization 128, 139, 199, 234 Native Authority System (NAS, Nigeria) 93–4, 96 natural resources xviii, 4, 5, 16, 39, 58, 80, 81, 226(n3), 235, 237 Angola 214 ‘curse’ 234 dependence on, 89 management (Sierra Leone) 87 north–south border (Sudan) 159 rebel income 6 rent contests 8–11, 16(n4) type 8 windfalls (macroeconomic consequences) 8 NCNC (National Council of Nigeria and Cameroons) xiii, 95, 96–7
neged group 127–8 neighbours (international) 232 nepotism 80, 113 neutrality 145, 158 New Partnership for African Development (NEPAD) 239 Nganguela (Angolan ethnic group) 214 Ngozi province (Burundi) 183t Nhaneka–Humbi (Angolan ethnic group) 214 Niger 111 Niger Delta 97, 99, 104, 107(n14) Nigeria 4, 5, 47, 58, 233 amalgamations (1906, 1914) xix, 92, 97, 105 central provinces 94 colonial legacy xix constitutional crisis (1962), 97 creation of new states 96, 97–8, 99, 100, 102, 107(n13) decolonization phase (1947–60) 95–6, 107(n12) dysfunctional state 96, 97, 100, 103, 104–5 east 93, 94, 97, 98, 100–1, 103, 104 east–west divide 95 economic crisis 104 ‘elective principle’ 106(n8) ethnic groups 60, 63 ethnic diversity 106(n7) ethno-regional imbalances 99 federal government 98, 99 fiscal problems 92 foreign interests xix foundation phase (1914–46) 93–5 inter-ethnic and class equity 235 legal distinction between ‘colony’ and ‘protectorate’ 106(n7) ‘local resistance’ versus ‘imperial force’ 92 nation-state formation 92–7 north 92–5, 97, 100–1, 103, 104, 106(n8), 107(n12, n17) post-conflict economic management and development 102–3 post-conflict governance (six issues) 100–2 revenue allocation 95 risk of further conflict xix, 103–5 secessionism 95
Subject Index 263 sources of conflict (colonial era) 92, 93–5 south 92–4, 96, 106(n5, n8) south (division into east and west districts) 93, 94–5, 96 three fundamental divides 103 transfer payments 92–3, 100 transition to civil war (1961–7) 96–7 west, 93, 97, 100–1, 103, 104 Nigeria: Federal Legislature 107(n13) Nigerian Civil War xix aftermath and risks of recurrence xix, 99–105 ‘best understood as historical event’ 97 causes and aftermath xix, 91–108 duration and cessation xix, 97–9 lessons 105 seeds of crisis 91–7 Nigerian Council 94 Nigerian Supreme Court xix Nilotics 164 Nimba County (Liberia) 80, 82 nomads 167 non-governmental organizations (NGOs) 88, 99 non-oil sector 216 Northern Armed Forces Command Council (CCFAN) 116 Northern People Congress (NPC, Nigeria), 95, 96–7 Northern Uganda Reconstruction Programme (NURP), 174 Nuba Mountains (Sudan) 161(n14) Nyanza tribe (Kenya) 62
oil companies 101 oil shock (1974) 128 Oil Pipelines Acts (Nigeria) 101 oil refineries/refining 102, 133 oil revenue xxi, 9, 11, 103, 217, 226(n7–8) allocation 155–6, 159, 160 Nigeria 100, 102, 105 potential to cause conflict 160 oil seeds 134 oil-exporting countries 156 one-party (single-party) systems 60, 61, 112, 173, 234 Onshore–Offshore Bill (Nigeria), 105, 107(n14) Operation ‘Iron Fist’ 172 opportunism 49, 55, 161(n10), 164, 174, 208 opportunity costs 132, 133, 139, 150, 176, 189, 239 child combatants 234 low 148 opting for conflict 178 recurrence of civil war 151–2 risk of war 168 Organization of African Unity (OAU) 98, 116, 165, 239 ‘OAU Framework’ 137–8 Oromo (ethnic group) 126 Oromo Liberation Front (OLF) 130 Oromo People’s Democratic Organization 129 ossification 98 Ovimbundu (ethnic group) 4, 214, 215
Odua People’s Congress (1995-) 104, 107(n16) Ogaden (Ethiopia) 128 Ogoja province (Nigeria) 94 Ogoni province (Nigeria) 94, 101, 103, 104 oil/petroleum 4, 8, 39, 46, 98, 102, 103, 107(n10), 158, 213, 216, 222, 223 Angola 226(n7) Chad 109, 118–19, 120 Nigeria 97 Sudan 153, 155–6, 157t, 161(n6–8) reserves (Sudan) 155, 161(n7)
Parliaments Buganda 164, 165 Mozambique 204, 210(n8) Nigeria: Federal Legislature 107(n13) Nigeria: Senate 100 Uganda 164, 165, 166, 173 party politics 164, 238 pastoralists xix, 109, 110, 111, 113, 116, 126, 132, 159 path-dependence 12, 139 patriotism 221 patronage/patron–client relationships 39, 127, 171
264
Subject Index
peace 33, 50, 52, 53, 57, 140, 239 broad versus narrow reconstruction 11–15, 16–17(n5–6) five preconditions 160 imperfect commitment (by rebels) 8–11 ‘must address social polarization’ 143 opportunity cost 150 post-conflict commitment by governments 6–7 post-conflict economies (Uganda) 163–79 post-conflict prospects (Sudan) 143–62 regional (more than one country) 16 Sierra Leone 85–6, 87 social contract favouring 6 peace agreements xvi Angola (civil war) 3 breakdown 3 DRC–Rwanda 3 importance for subsequent maintenance of peace 153 ‘improve macroeconomic policy formulation’ 23 temptation to renege 8–9 ‘peace bureaucracy’, international 209 peace dividend 16, 23, 26, 35, 41, 151, 152, 203, 208 peace triangle 152–3, 154t external commitment to change 153, 153f, 154t extent of hostility 153, 153f, 154t local capacity for change 153, 153f, 154t peace-building 45, 46–9, 226(n5) prospects (Sudan) 156–9, 161(n9–14) peace-building model (Doyle and Sambanis, 2000) 152 peace-keeping force, international 186, 187 peaceful coexistence (ethnic) 63 peasantry/rural population xx, 111, 126, 127, 128, 139, 199, 201, 214, 215, 220, 236 pension payments 204 People’s Development Bank 210(n15) People’s Redemption Council 80
petroleum laws (Nigeria) 100, 101 petroleum products 101, 102 Petroleum Revenue Control and Surveillance Assembly (CCSRP, Chad) 119 Petronas Berhad (Malaysia) 161(n6) point resources (Addison, 2002) 8, 16 ‘increase chances of conflict’ 8 polarization ethnic 149f social 143, 147, 148, 150, 160–1(n1–2) see also fractionalization police/policing xvii, 54, 55, 104, 155, 210(n10–11), 219, 220, 222 policy environment xviii, 19, 20, 25, 28t, 29t, 30t, 31–2, 33, 36, 39, 42, 43, 43(n7), 50 contribution to economic growth 51–2 policy reform political economy 45, 53–5, 56 raises economic growth 53–4 risk of conflict resumption 53 political economy xxiii, 125, 139 economic reforms (Mozambique) 206–8, 205(n13–15) policy reform 45, 53–5, 56 post-conflict economic recovery (Mozambique) 197–212 political legitimacy 33, 40 political openness civil war risk factor (Mozambique case) 198, 198t political opposition 205, 206, 208 political parties xxi, 62–3, 93, 105, 160, 203, 221 ‘can intensify ethnic competition/ conflict’ 63, 72 conversion of rebel groups into xvii, xxiv(n3) political prisoners 186 political rights xvi, 39–40, 43, 43–4(n9), 58, 161(n1), 169 extent 19, 26 political science 54, 169 political stability 60, 176 political transformation 239 political will 119, 174, 204
Subject Index 265 population/demography 95, 107(n11), 114, 156, 157t, 168, 169 civil war risk factor (Mozambique case) 198t population growth 23–4, 32, 33 Population and Housing Census of Ethiopia (1994) 126 population size 148, 149f, 169 ports 133, 156 Portugal/Portuguese 209(n1), 214, 216, 236 post-conflict economic recovery (Mozambique) 197–212 post-conflict economies background xv–xvi case studies: Horn, East and Central Africa, ix, xv, xx–xxiii, 123–93 case studies: Southern Africa, ix, xv, xxiii, 195–228 case studies: West Africa, ix, xv, xviii–xx, 75–121 difficulty of definition xv thematic issues ix, xv, xvi–xviii, 1–73, 229–40 post-conflict economies: synthesis and lessons xxiv, 231–40 child combatants 234 colonial history 231–2 democracy 237–8 elite political instability 232–3 impartial legal system 236–7 inter-ethnic and class equity 235–6 international community 237 neighbours 232 proliferation of cheap weapons 234–5 rent-seeking opportunities 233–4 post-conflict era first five years 22 prospects for sustainable peace and economic growth (Sudan) 143–62 post-conflict reconstruction: analytical issues 3–17 broad versus narrow reconstruction 11–15 collapse of pre-conflict social contract 4–6 natural resource rent contests and imperfect commitment among rebels 8–11
sustaining post-conflict commitment to peace by government 6–7, 15 themes 4 war, peace, and recovery 13f post-conflict societies ‘distinctive economic problems’ 45, 55 economic policy xvii–xviii, 45–56 peace-building 45, 46–9, 55 ‘policies may need to be distinctive’ (reasons) 45 political economy of policy reform 45, 53–5, 56 rebuilding the economy 49–53 two sets of risks (Collier) xvii post-conflict transformation (Angola) 213–28 poverty xxi, xxiii, 20, 24, 43, 89, 103, 115, 119, 120(n1), 155, 159, 205t, 205, 213, 216, 219 absolute 3, 15, 217 causes 21 civil war risk factor 155 head-count ratio 155, 161(n8) regional (within one country) 15 relative 15–16 poverty line 84, 109, 161(n8) poverty-eradication 42, 222 poverty-reduction xvi, 3, 13, 14, 16, 118, 137, 155, 176–7, 178, 220 Sierra Leone 86–7 poverty-reduction strategy (PRS) 221, 223–4, 226(n6) Poverty-Reduction Strategy Paper 87 power 130, 192(n10) absolute 138, 139 Angola 216 competition for xx, 125, 139 decentralization 175 peaceful transfer 133 political 114, 115, 128, 129, 133, 138, 140(n3, n8), 183 power-sharing 16(n2), 133, 189, 200 prices 12–14, 16(n5), 135 primary commodities/exports xvii, 19, 32, 33, 111, 169, 233 risk of conflict 46, 47, 168, 198, 198t primary sector 110 prisoner’s dilemma approach xxiii, 189–90, 190–1
266
Subject Index
prisoners of war 114 private sector 24, 87, 113, 118, 140(n8), 175, 213, 223, 224 privatization 87, 88, 176, 205, 206, 210(n15), 223 pro-poor asset redistribution 12 production 13, 14, 59, 111 projects, large-scale 206 promises (commercial concessions) 6 propaganda 165 property rights xvii, xxii, 49, 54–5, 97, 101, 114, 164, 173, 175, 206 onshore and offshore 101 proportionality principle 159 protectionism 134, 135 Protestantism/Protestants 164, 171 psychology 91, 140(n6) public administration 223 public appointments (Chad) 116 public expenditure/public spending see government expenditure public goods 4, 66, 70–1, 95, 97 ‘local’ 70 public resources, access to 114, 115 public sector/state sector 48, 102, 213, 216 public sector management and service delivery 157n quality of life
57
race 215 raids 111 railways 106(n5) rational choice 15 ratios 24 aid to GDP 25, 26, 27t, 28t, 29t, 31t, 38f aid to GDP interacted with overall policy index 25, 27t, 29t, 30t, 32, 33 aid to GDP squared 25, 27t, 28t, 30t debt to GDP 20t, 25, 27t, 29t, 30t, 43(n4) debt overhang to GDP, 25 debt to exports 20t, 20–1 domestic investment to GDP, 43(n4) external debt service 25 FDI to GDP 43(n4) fiscal balance to GDP 25, 28t
government consumption to GDP 22, 43(n4) primary exports to GDP 26, 29t, 30t, 147, 169 revenue to GDP 29t savings to GDP 43(n4) stock of money to GDP, 25 surplus/deficit to GDP 43(n4) raw materials 135 re-education camps 199 rebel utility function 5–6, 8–9 rebels xv, 5–6, 48, 132, 150, 187, 190, 203, 232 Burundi 186, 188 chance of enduring peace 6–7 conversion into political parties, xvii, xxiv(n3) finance 47, 217 Hutu (Burundi) 188 imperfect commitment to peace 8–11, 16(n4) propensity to join 12, 15 reintegration xxii social contract favouring peace 6 war finance 6 reconciliation xxii, xxiii, 100, 105, 177, 190, 213, 218, 221 national 89, 225 reconstruction xxiii, 23, 100, 105, 134, 174, 188, 219, 221, 222, 225 benefits 3 broad versus narrow reconstruction 11–15, 16(n5–6) donor-supported 11 economic 49–53 Ethiopia 138 financial issues 140(n4) post-conflict 24, 158, 159, 160 post-conflict policy instruments 33 redistribution 43–4(n9), 67, 155, 238 unfair measures 80 ‘red terror’ 129 redress of grievances 11–12 reduced form equations 18, 20, 34–42 referenda 107(n13), 158, 164 refugees xxiii, 57, 58, 180, 200, 213, 219 Burundi (1972-) 183t repatriation (Mozambique) 203 see also internally-displaced persons (IDPs)
Subject Index 267 regional balance (Chad) 118 development 204, 224 imbalance (Angola) 216 Regional Representatives Conference (1966), 98 regions (intra-country) xvii, 233 rehabilitation 100, 105, 134, 156 formal economy xvii, xviii, 237 religion xx, xxi, 4, 103, 104, 105, 107(n17), 161(n2), 235, 236 remittances 84 Renamo 15, 199–204, 206, 209, 209(n2, n7), 210(n10–11), 234 conversion to political party 47, 209(n6) rent (economic) 89, 102 capturable xvi from pricing interventions 80 from primary exports 46, 47 rent contests 8–11, 16(n4) rent-seeking xvii, xxiv(n2), 13, 14, 67, 70, 176, 213, 233–4 repression 89, 157, 173 reputation/image 10, 49, 55, 186 ‘bad’ distinguished from ‘lack of’ 186, 192(n10) Resistance Councils (Uganda), 175 Resistência Nacional Moçambicana (Renamo) 197, 199 resource allocation 68, 155, 171, 172, 175, 208–9, 219, 224 ethnic favouritism 61–2 Nigeria 95, 107(n11) resource misallocation 176 resources competition for 127–8 conflict over 125 Revenue Oversight Committee (Chad), 234 Revolutionary United Front (RUF, Sierra Leone) 79, 81, 82, 85, 87, 232 Rhodesia 199 Richards Constitution (Nigeria, 1946) 93, 95, 97 Rift Valley 62 riots 93, 164, 210(n7, n10)
risk of conflict xvi–xvii, xxiv(n2), 19, 39, 41f, 42, 43, 46, 106(n2), 148, 176, 178 determinants 145 economic factors 148, 149f four factors (Collier and Hoeffler), 46 generated during conflict, xvii leading to outbreak of original conflict, xvii policy reform and, 53 political factors 149f social diversity and, 40–1, 44(n10) see also conflict re-ignition risk premiums 17(n6) road construction 102 rubber 81, 83, 214, 235 rule of law xvii, 89, 173, 175, 186, 239 absence 139–40 post-conflict restoration 54–5 ruling class 128, 173, 176 rural areas 84, 85, 88, 104, 112, 118, 132, 199, 200, 213, 217 rural–urban displacement 84–5 rutile 84 Rutovu commune (Bururi) 182 Rwanda xv, 3, 5, 46, 58, 137, 182, 192(n10) Burundi refugees 183, 183n child combatants 234 indicators of debt and well-being 20t ‘social revolution’ (1959) 181 war with DRC 184 ‘would prefer Tutsi government in Burundi’ 188 Sabon Gari 94 safe havens/sanctuary 183, 185, 232 safety nets xxii, xxiii, 175, 213, 224 Sagamu (Nigeria) 104 Sahelian sultans 110 San Pedro harbour 44(n9) sanctions xvi, 10, 83–4, 85, 86, 184, 189 Sao Tomé 99 savings 23, 37 savings-investment gap 176 scarce resources 126, 226(n1), 238 schooling 151, 156, 157t
268
Subject Index
Sebha–Benghazi accords 116 secessionism 47, 58, 72, 95, 96, 97, 99, 129, 130, 164 Second World War 95 post-war era (1945-) 238 secularism: ‘made meaningless’ (Chad) 116 security institutions 181, 188 segregation (Nigeria) 94, 106(n9) self-determination 158, 160 self-interest 71, 237 Senegal, 60 separation of powers 221 September laws (Sudan, 1983) 152, 157, 161(n10) Serapiao, L. 199, 201, 203 Serers (people) 60 serfs/serfdom 128, 139 service sector 12, 13, 14, 16(n5), 135 shipping 84, 89 Sierra Leone xv, xvi, 6, 233, 236, 237 British intervention (2000) 85–6, 89, 231–2 causes of civil war 80–1 child combatants 234 civil war (1991–2001) xviii consequences of civil war 84–5 context and background 78–9 Creole–Protectorate divide ‘blurred’ 82 donor-supported reconstruction, 11 ethnic rivalries 80, 82 export earnings 84 factors sustaining civil war 81–3 indicators of debt and well-being, 20t and Liberia (interwoven civil wars) 77–90 natural resource management 87 peace xviii, 85–6, 87 per capita income 83, 83t, 84 point resources 8 post-conflict initiatives 86–8 poverty-reduction 86–7 proliferation of cheap weapons 235 socio-economic indicators 83t Sierra Leone People’s Party (SLPP) 79, 82 single-equation estimation xvi skills 48, 213 slavery 80, 110, 111, 120(n2), 170, 214 freed slaves 79, 80
smuggling 13, 84, 87, 138 social capital 49, 225 social class 4, 128, 139, 220, 235–6 Angola 215 class conflict 170 social contract 16, 16(n3), 92, 223, 227(n10) breakdown 4–6, 11, 15 fiscal component 15 rebuilding 11–15 time of uneasy peace 14–15 see also fiscal transfer social diversity 26, 32, 220–1 social exclusion xxiii, 14, 210(n10), 213, 219, 225–6 social inclusion xviii, 53, 219 social policy 54 post-conflict societies 52 social services 23, 84, 102, 118, 151, 159 social structure: risk of conflict 46 social-market economy 223 socialism xv, 127, 128, 129, 132, 139, 199 socialization (cross-ethnic) 236 Sofala Province 210(n9) Somalia 20t, 58, 60, 86, 127, 128, 130, 165 Somalis 58, 126, 128 dispersed between Somalia, Ethiopia, Kenya 60 South Africa 199, 201, 218, 236 peace-keeping in Burundi 187, 188 South Nations and Nationalists Organization 129 Southern Africa: case studies xxiii, 195–228 Angola xxiii, 213–28 Mozambique xxiii, 197–212 see also Sub-Saharan Africa Southern Executive Council (Sudan) 152 Southern Yemen 152 Sovereign National Conference (Nigeria), 101 Soviet Union see Union of Soviet Socialist Republics Special Operations Division (SOD, Liberia) 89 Special Security Service (SSS, Liberia) 89 SSA see Sub-Saharan Africa
Subject Index 269 St. Louis Post-Dispatch 161(n6) stability 24, 140, 239 stabilization (economic) 87, 88 state corporations 182 state failure 16(n3) state machinery 125 state-formation 139 states 91, 92 centralized unitary 59, 63, 66, 72 states of emergency 164, 165 structural adjustment programmes Chad 117 Mozambique 209(n3) Nigeria (1986) 100, 101, 102 structural and systemic transformation programme (SSTP) 221, 222, 226(n6) students 81, 96, 129, 192(n4) Sub-Saharan Africa (SSA) civil war 18, 43(n4) economic recovery agenda (post-conflict) 18 ethnic heterogeneity 63 external intervention in civil wars 146f growth and macroeconomic policy stance (1970–98) 23, 23t risk factors (civil war) 148 see also West Africa subsidies 67 subsistence 49, 111 subversion 165 Sudan xxi, 3, 6, 57, 58, 111, 172, 231, 236, 238 background to conflict 145 child combatants 234 colonial administration 145 conflict resolution (three fundamental questions) 143 economic growth 156, 157t economic mismanagement and a missed opportunity 150–2 external intervention xxi, 146, 146f, 149f, 160, 161(n3) first civil war (1955–72) 146f, 147f, 148, 149f, 150, 151t indicators of debt and wellbeing 20t inter-ethnic and class equity 235 length of civil war (reasons) 145–50 linguistic groups 63
making and collapse of peace (1972–83) 150–3, 154t, 161(n4–5) map 144 north 145, 157, 158, 159, 160, 161(n9, n14) north–south divide xxi, 238 oil and post conflict recovery xxi, 153, 155–6, 157t, 161(n6–8) oilfield straddling north and south 153 partition (mooted) xxi, 158, 159, 160 peace (reasons for collapse) 152–3, 154t peace period (1973–82) 146f, 147f, 151t peace triangle 152–3, 154t ‘peace from within initiative’ 158 peace-building 156–9, 161(n9–14) polarization between north and south 147f, 148 population data (1983) 148 prospects for peace (1972) compared with Mozambique (1992) 153, 153f, 154t prospects for sustainable peace 143–62 regional (multi-country) economic integration 156 risk of conflict xxi, 147f, 148, 149 second civil war (1983-present) 146–8, 149f, 151t, 159, 160 south 151, 152, 156–60, 161(n14), 231 south (proposed integration into East Africa) 143, 145, 231 state of emergency 157 ‘third democracy’ (1985–8) 156, 161(n10) tribal disputes over natural resources straddling north-south border 159 ‘uniform diversity’ 147f western 161(n14) Sudan: Anglo-Egyptian (1898–1956) 145 Sudan People’s Liberation Army/ Movement (SPLA/SPLM) 152, 157, 158 Sudanese People’s Liberation Movement 172
270
Subject Index
Sudapet (Sudan) 161(n6) suffrage 106(n8) Supreme Court (Nigeria) 105, 106 ‘surprise’ warfare xvi, 9–10, 11 Switzerland 47 Takum (Nigeria) 104 Talisman Energy (Canada) 161(n6) Tanzania xv, 99, 172, 174, 232 avoidance of civil war 148, 149f, 150 Burundi refugees 183, 183n, 185 one-party system 173 Ugandan exiles 165, 166 ‘would favour Hutu government in Burundi’ 188 tariffs 136 tax base 11, 55, 176 tax burden xvii, xviii tax revenues 22 taxation 24, 43(n9), 67, 93, 111, 112, 114, 127, 182, 237 agriculture 80 business 23 early post-conflict years 55 taxation without representation 92, 93 teachers 200 technocrats 88, 140(n8) teff (a staple grain) 135 telecommunications 88, 102 textiles 214 thematic issues ix, xv, xvi–xviii, 1–73, 229–40 economic policy in post-conflict societies xvii–xviii, 45–56 economic and political consequences of conflict and post-conflict recovery xvi–xvii, 18–44 ethnicity, institutions of governance, and conflict avoidance xviii, 57–73 post-conflict reconstruction xvi, 3–17 Tigrawi (ethnic group) 126 Tigray (Ethiopia) 141 Tigray People’s Liberation Front (TPLF) 129 timber 81, 83, 235 time xvi, xviii, 10, 13, 14, 22, 51, 55, 161(n8) Tiv province (Nigeria) 94
Tiv Riots 97 Togo 60 Tokyo International Conference on African Development (TICAD-II), 57 Torit (Sudan) 145 tourism 136, 224 trade 10, 12, 14, 60, 84, 110, 133–4, 178 trade shocks 105 trade unions 65, 67, 129 training xxiii, 127, 204, 209(n6), 213, 220, 222, 237 transaction costs 12, 14, 66 transition from war to peace Burundi 180–93 economic dimensions 3 Mozambique 197 transparency xvii, xxiii, 16(n3), 47, 96, 97, 102, 114, 118, 119, 139, 155, 158, 159, 160, 167, 173, 214, 219, 221, 223, 224, 227(n9), 233, 238 transport 12, 135, 136 Treasury (British) 92 ‘tribal’ surplus 67 tribalism 60, 64–5, 115, 161(n14), 171, 226(n3) tribes 63, 64, 171 Tripoli accords (2002) 117 truth and reconciliation 86, 189, 236 Tutsi monarchy (Rwanda) 181, 191(n3) Tutsis (Burundi) xxii, 58, 180–93, 231, 236 Twa (ethnic group, Burundi) 191(n1) Uganda 16, 20, 89, 231, 232, 237 capital flight 49–50 child combatants 234 civil conflict (1981–6) 163, 166–7 civil war (onset and duration factors) 168, 169 Collier-Hoeffler risk-of-war model 168–9, 178(n5) conflict (additional factors) 170–4 crisis (1964–6) xxi, 163–5, 170, 171, 172 deficit levels 22 demobilization 54 economic growth 23 economic imbalance (regional) 174
Subject Index 271 economic opportunities 175–7, 178(n9) effect of insurgencies (1998–2000) 168t Elbadawi-Sambanis risk-of-war model 169, 178(n5) entrenching peace in post-conflict economy 163–79 explaining recurrent conflict 168–9, 170–4 external factors 172 formation of state 170 fractionalization 170–2 heterogenous society 169 indicators of debt and wellbeing 20t institution-building xxii institutional measures 174–5 investor risk rating 53 IGAD 161(n11) leftists versus rightists 164 lessons 177–8 liberation struggle (1978–80) xxii, 163, 165–6, 172 limitations of conflict-prediction models 169, 170 ‘main source of conflict’ xxii measures to address conflicts 174–7 north-south divide xxii, 167 official development assistance (1992–9) 175, 177t peace dividend 23 political instability (impact on economic growth, 1960–97) 176t property rights 55 relationship between executive and legislature 165 role of institutions 172–4 ten-point programme (1986) 167 UPC–KY coalition government 164–5 war with DRC 184 see also independence Uganda National Liberation Front (UNLF) 166 Uganda People’s Congress (UPC) 164–5 Umma Party (Sudan) 161(n9) Umuleri (Nigeria) 104 UN Mission in Sierra Leone (UNAMSIL) 85
UN Operation in Mozambique (UNOMOZ, 1992-) 203, 209–10(n5, n7–8) UN peace-keeping operations Ethiopia 138 Mozambique 153 Sierra Leone 85 underfunded 16 UN Security Council 140(n7), 209(n5) UN Special Representative (in Mozambique) 203 uncertainty 14, 24 unemployment 80–1, 85, 87, 103, 234 UNESCO viii UNICEF 234 Unification Decree (Nigeria, 1966) 97 Uniform Pricing Decree (Nigeria, 1973) 101 Union of Soviet Socialist Republics (USSR) 106(n4), 127, 128, 152, 235 former 8, 138 UNITA see National Movement for Total Independence of Angola United Kingdom 110, 126, 127 intervention in Sierra Leone (2000) 85–6, 89 political and economic interests (Nigeria) 98–9 Uganda conference 164 United Nations 83, 87, 129, 167 United States of America 128, 137, 138, 152, 172 freed slaves 79 Horn of Africa policy 127 Liberian diaspora 80 Sudan policy 158 unlikely to intervene in Liberia 89–90 urban areas/cities 12, 61, 67, 159 urban–rural divide (Angola) 216 Urhobo (Nigeria) 104 Western 111 vested interests 47, 54, 106 ‘villagization’ 199 violence 87, 186, 188, 190, 200, 218 political (in homogeneous societies) 150 voluntary associations 65, 68–9, 70f, 71f voting patterns (ethnic identification) 65
272
Subject Index
wages 15, 224 war, marginal cost of waging 235 war crimes 86, 186 warlords/warlordism 12, 81, 82, 112 ‘survival strategy’ 89 Warri riots (1927–8) 93 wealth transfers 61, 66–7 weapons/arms 57, 82, 115, 116, 167, 204 cheap 239 proliferation 234–5 welfare 65, 224 West Africa xviii-xx, 8, 57, 65, 75–121 Chad xix–xx, 109–21 Liberia and Sierra Leone xviii, 77–90 Nigeria xix, 91–108 see also Africa West Nilers (tribe) 173 West Side Boys (rebel faction) 85 Willinks Commission of Inquiry (1957) 96 Wolaita (ethnic group) 126 Wolof (people) 60 women 110, 145, 224 Women’s Riot in Aba (1928) 93
working class 105 ‘near-proletariat’ (Angola) 215 World Bank ix, 45, 52, 134, 151, 174, 208, 226(n5), 234 Chad 117–18 Mozambique 209(n3) project on ‘Economics of Civil and Common Violence’ 209(n*) World Bank index of policy (‘Country Policy and Institutional Assessment’) 50, 52–3 World Council of Churches 150 Wukari (Nigeria) 104 Yoruba 60, 104, 107(n16) youth 54, 57, 80–1, 85, 129, 172, 238, 239 Yugoslavia 106(n4) Zaghawa (ethnic group, Chad and Sudan) 115, 120(n3) Zaire see Congo-Kinshasa Zambia xv, 99 Zangon Kataf (Nigeria) 104 zero-sum game 67, 209