Judy Smith-Höhn Rebuilding the Security Sector in Post-Conflict Societies: Perceptions from Urban Liberia and Sierra Leone Geneva Centre for the Democratic Control of Armed Forces (DCAF) LIT Judy Smith-Höhn Rebuilding the Security Sector in Post-Conflict Societies: Perceptions from Urban Liberia and Sierra Leone LIT Contents Preface Acknowledgements Tables and Figures Abbreviations Maps vii ix xi xiii xvi 1 Introduction 2 On Security: Definitions and Theories 13 2.1 2.2 2.3 2.4 13 15 17 3 4 17 Methodology 43 3.1 3.2 45 58 The comparative method Tools for comparison: a mixed-methods approach Case Studies 4.1 4.2 4.3 4.4 5 Post-conflict societies Human security: A narrow definition Peacebuilding Security sector reform and reconstruction in the post-conflict setting 1 Liberia Sierra Leone Comparison of cases Multilevel comparison: A synopsis Conclusion 71 71 78 86 142 157 Annex I: Survey Poll 162 i) ii) iii) iv) Questionnaire: Liberia Questionnaire: Sierra Leone Data analysis: Frequencies Liberia Data analysis: Frequencies Sierra Leone 162 169 175 198 Annex II: Guidelines for Focus Group Discussion (excerpt) 224 About the Author About DCAF 227 228 Preface The Geneva Centre for the Democratic Control of Armed Forces (DCAF) develops and disseminates policy-relevant research that contributes new knowledge on issues related to security sector reform (SSR) and security sector governance (SSG). As part of the DCAF Research Division’s strong commitment to promoting innovative, empirically-grounded analysis from different world regions, we are therefore delighted to publish Rebuilding the Security Sector in Post-Conflict Societies: Perceptions from Urban Liberia and Sierra Leone by Dr. Judy Smith-Höhn. Applying a human security perspective, this volume raises important issues relating to gaps between SSR ‘design’ and how individuals and communities experience security in practice. It also constructively identifies policy options that are particularly relevant given the long term nature of the SSR processes in both Liberia and Sierra Leone. An analytical approach that asks questions of individual citizens on how they experience security helps to move the notion of human security from concept to lived experience. Moreover, the prominence of the informal security sector in both settings – from community watch groups to private security companies – provides empirical evidence to support those analysts and practitioners that call for the role such actors play in security sector governance frameworks within many national settings to be properly acknowledged.1 This volume is published in parallel to another contribution in the DCAF LIT series, Security Sector Reform in Sierra Leone 1997-2007: Views from the Front Line edited by Peter Albrecht and Dr. Paul Jackson. While not conceived as companion volumes, they do offer complementary insights and lessons. Albrecht and Jackson consider SSR from the perspectives of influential national and international stakeholders; Smith-Höhn takes the opposite approach in focusing on the perceptions of the ordinary citizens who should be the beneficiaries of reform efforts. We hope that disseminating and discussing the findings contained in these volumes will provide insights that help to bridge top-down and bottom-up perspectives, thus making a concrete impact on the design, implementation and effectiveness of SSR policies and programmes. 1 See: Alan Bryden and ‘Funmi Olonisakin (Eds) Security Sector Transformation in Africa (Münster, LIT Verlag, 2010). Available at www.dcaf.ch/publications/ viii It would not have been possible to successfully complete this volume without the invaluable support of a number of people. I would like to acknowledge the excellent work of the author, Judy-Smith-Höhn, and thank her for the opportunity to review and publish her research. DCAF colleagues Fairlie Chappuis and Lynda Chenaf provided important assistance during the editing process. In particular, I would like to thank Yury Korobovsky, Head of DCAF’s Information Resources Unit, as well as his team for guiding the publication process and liaising with our partners at LIT Verlag. Finally, thanks go to Cherry Ekins for copy editing the draft volume with subtlety and attention to detail. The views expressed in this volume are those of the author alone and do not in any way represent the views of the institutions or their representatives involved in this project. Alan Bryden Deputy Head of Research DCAF Geneva, September 2010 Acknowledgements This dissertation is developed from a research project funded by the German Foundation for Peace Research. Without this financial support and without the assistance and cooperation of a number of people to whom I owe many thanks, I would surely not have been able to complete this study. I am deeply indebted to Prof. Dr. Ulf Engel for always finding time to answer my emails, and for his persistent encouragement and intellectual assistance as well as to Dr. Andreas Mehler for his intellectual guidance, support and constant advice on all project-related matters. I am also very grateful to the Geneva Centre for the Democratic Control of Armed Forces (DCAF) for their generous offer to publish this book as a volume in their DCAF LIT Series and for taking such professional care of publishing matters. In so doing, what was initially perhaps doomed to gather dust on library shelves throughout Germany will now reach a much wider audience than I could have possibly hoped for. I would like to thank my interview partners in Liberia and Sierra Leone, who found the time to answer my questions during my two field visits in November 2005-February 2006 and April-June 2006. The information obtained in the interviews is incorporated throughout this study, but mainly in chapter 4. I also thank Dr. Heinz Jockers, who in his capacity as the director of the Konrad Adenauer Foundation in Liberia, established essential contacts and allowed us to use the foundation’s facilities to conduct research in what would have otherwise been a very difficult environment. Special thanks go to Dr. Mats Utas, who shared not only his in-depth knowledge on Sierra Leone and his valuable contacts, but also his home in Freetown. I also thank Osman Gbla and his Centre for Security and Development Analysis in Freetown, for providing us with competent researchers to conduct the survey poll and for sending some of them off to the hinterland with me on roads that turned out to be rather perilous. I must also thank our moderators, Sunny George, Safie Sesay and Charlie J. Hughes, for skilfully guiding the focus groups in the right direction. I am grateful to my colleagues at the GIGA Institute of African Affairs in Hamburg who always found the time to clarify doubts and give advice but also to ask the right questions that made me think twice. I must especially thank Dr. Matthias Basedau, who played a brief but crucial role as a senior researcher in the research project, for giving me sound advice from the very start. Dr. Daniel Lambach, for providing precisely the type of x support expected of a senior researcher. I thank Alexander Stroh, my office buddy, for being in the same boat as I; several interns at the Institute for their research assistance; and Sven Buth for help with the data analysis. I am further grateful to the participants of several conferences, workshops, and seminars, whose comments proved to be most helpful. I want to express my heartfelt thanks to family and friends. I particularly thank those friends who helped me through the most critical stages in the last two years, especially Elena Meyer-Clement and AmandaLeigh O’Connell. My deepest gratitude goes to my brother, Neil Smith. Thanks for the much needed moral and financial support while I was trying to find my feet in what was then a foreign country to me. Last but certainly not least, I owe love and thanks to my husband Bastian Höhn for his unconditional support and love throughout this entire process. Judy Smith- Höhn Pretoria, September 2010 Tables and Figures Table 2.1 Table 3.1 Table 4.1 Table 4.2 Table 4.3 Table 4.4 Table 4.5 Table 4.6 Table 4.7 Table 4.8 Table 4.9 Table 4.10 Table 4.11 Table 4.12 Table 4.13 Table 4.14 Table 4.15 Table 4.16 Table 4.17 Table 4.18 Table 4.19 Definitions of the ‘security sector’ Four purposes of comparing General and personal security in urban Liberia General and personal security in urban Liberia – Gender General security in urban Liberia – Ethnic group General security in three urban centres of Liberia General security in three urban centres of Liberia – Past and present compared Security actors in urban Liberia – Most important actors Security actors in urban Liberia (%) Security actors in urban Liberia – Biggest threat Changes in general and personal security in urban Liberia – Past versus present General and personal security in urban Sierra Leone General and personal security in urban Sierra Leone – Gender General security in urban Sierra Leone – Religion General security in urban Sierra Leone – Ethnic group General security in three urban centres of Sierra Leone Personal security in three urban centres of Sierra Leone Changes in general security in urban Sierra Leone: Past and present compared Changes in personal security in urban Sierra Leone: Past and present compared Security actors in urban Sierra Leone – Most important actors Security actors in urban Sierra Leone (%) 20 48 87 88 89 90 91 93 94 96 100 105 106 106 107 108 108 109 110 111 112 xii Table 4.20 Table 4.21 Table 4.22 Table 4.23 Table 4.24 Table 4.25 Table 4.26 Figure 2.1 Figure 3.1 Figure 3.2 Figure 4.1 Figure 4.2 Figure 4.3 Figure 4.4 Figure 4.5 Figure 4.6 Figure 4.7 Figure 4.8 Current policing structures in Sierra Leone Mines monitoring officers in three Sierra Leonean cities compared (%) Bike riders in three Sierra Leonean cities compared Sierra Leone – Mapping security actors today Sierra Leone – Security actors past and present General and personal security in urban Liberia and Sierra Leone compared Mapping security actors – Liberia and Sierra Leone today SSR and related activities The case study method – A semantic clarification Comparative case studies – Levels of comparison Liberia today – Security actors and their interactions Liberia 2003 – Security actors and their interactions Sierra Leone today – Security actors and their interactions Sierra Leone 2002–2005 – Security actors and their interactions Sierra Leone today – Security actors and their interactions Sierra Leone in 2001 – Security actors and their interactions Liberia today – Security actors and their interactions Liberia and Sierra Leone today – Security actors and their interactions 114 115 119 120 121 133 134 28 56 58 98 102 116 124 126 128 130 138 Abbreviations ACPP AFL APC AU BCE BICC BIN BTI CCSSP CDF CDID CDU CISU CMR CPA CSO CWT DAC DCAF DDR DDRR DEA DFID DSF ECOMIL ECOMOG ECOWAS EO EU FCO FGD FOHRD GIGA GOL Africa Conflict Prevention Pool Armed Forces of Liberia All People's Congress (Sierra Leone) African Union Bureau of Customs and Excise (Liberia) Bonn International Center for Conversion Bureau of Immigration and Naturalisation (Liberia) Bertelsmann Transformation Index Commonwealth Community Safety and Security Programme (Sierra Leone) civil defence forces Complaint, Discipline and Investigation Department civil defence unit Central Intelligence and Security Unit (Sierra Leone) civil-military relations Comprehensive Peace Agreement (Liberia) civil society organisation community watch team OECD Development Assistance Committee Geneva Centre for the Democratic Control of Armed Forces disarmament, demobilisation and reintegration disarmament, demobilisation, rehabilitation and reintegration Drugs Enforcement Agency (Liberia) Department for International Development (UK) Deutsche Stiftung Friedensforschung (German Foundation for Peace Research) ECOWAS Mission in Liberia ECOWAS Monitoring Group Economic Community of West African States Executive Outcomes European Union Foreign and Commonwealth Office (UK) focus group discussion Foundation for Human Rights and Democracy German Institute for Global and Area Studies Government of Liberia xiv GRC GTZ ICG ICGL IDP IMATT ISU JIU LNP LURD MCP MDSD MNS MoD MODEL MSSD NBI NCDDR NCDDRR NFS NGO NPFL NPRC NSA NSC NTGL NTLA OECD ONS OSD PPB PSC RSLAF RUF SCSL SILSEP SLP Governance Reform Commission Gesellschaft für Technische Zusammenarbeit (Germany) International Crisis Group International Contact Group on Liberia internally displaced person International Military Advisory and Training Team (Sierra Leone) Internal Security Unit (Sierra Leone) joint implementation unit Liberia National Police Liberians United for Reconciliation and Democracy Monrovia City Police most-different systems design Liberian Ministry of National Security Ministry of Defence Movement for Democracy in Liberia most-similar systems design National Bureau of Investigation (Liberia) Sierra Leone’s National Committee for Disarmament, Demobilization and Reintegration Liberia’s National Commission for Disarmament, Demobilization, Rehabilitation and Reintegration National Fire Service (Liberia) non-governmental organisation National Patriotic Front of Liberia National Provisional Ruling Council (Sierra Leone) National Security Agency (Liberia) National Security Council (Sierra Leone) National Transitional Government of Liberia National Transitional Legislative Assembly of Liberia Organisation for Economic Co-operation and Development Office of National Security (Sierra Leone) SLP Operational Support Division Police Partnership Board (Sierra Leone) private security company Republic of Sierra Leone Armed Forces Revolutionary United Front (Sierra Leone) Special Court for Sierra Leone Sierra Leone Security Sector Reform Programme Sierra Leone Police xv SLPP SSD SSR SSS TRC UK ULIMO UN UN CIVPOL UNAMSIL UNDP UNICEF UNIOSIL UNMIL UNP UNSC US WANEP WIPNET Sierra Leone People’s Party Special Security Division security sector reform Special Security Service (Liberia) Truth and Reconciliation Commission United Kingdom United Liberation Movement of Liberia for Democracy United Nations UN Civilian Police UN Mission in Sierra Leone UN Development Programme UN Children’s Fund UN Integrated Office in Sierra Leone UN Mission in Liberia UN Police UN Security Council United States of America West African Network for Peacebuilding Women in Peacebuilding Network The boundaries and names shown and the designations used on this map do not imply official endorsement or acceptance by the United Nations. Mamou N ig G U I N E A te en L ola Ko l SIERRA Faranah LEONE Kindia Médina Dula Falaba Tabili on cie go ba Ka s car M G re at S N O R T H E R N Lit tl Rokupr Kayima Magburaka Lunsar Masingbi li Masiaka a Mile 91 Yonibana Yele WESTERN Songo Waterloo AREA Bradford Rotifunk Bauya Yawri Bay Shenge Plantain Is. Moyamba Sembehun S O U T Sieromco Mokanje rbro Nitti Se w Bo Gerihun a Tikonko Kpetewoma Kenema Tokpombu Blama Koribundu River Turtle Is. Sherbro I. Matru ait Bonthé Str ATLANTIC OCEAN br o Sher SIERRA LEONE Sumbuya e Potoru Pujehun Gorahun Zimmi Lake Mabesi Ma M ano LIBERIA Kongo no Bendaja Lo Sulima Robertsport UNITED NATIONS fa Bopolu Bomi-Hills 0 S ai n tP au l Bong Lake Piso 0 Map No. 3902 Rev. 5 January 2004 o or i Lake Mape National capital Provincial capital City, town Major airport International boundary Provincial boundary Main road Secondary road Railroad Koindu Falla Kailahun Buedu Panguma Bomi Manowa Giehun Boajibu Pendembu Bendu Lago Segbwema Daru oa She Dambara H E R N Bumpe Koundou E A S T E R N Taiama Njala Mano Gbangbatok Koidu-Sefadu Yengema Njaiama NjaiamaTongo Gandorhun Sewafe M Banana Is. Tungie Mongeri Taia York Matotoka M Wellington Hastings e Waan j Freetown Ferry rS an R o k el o mp Pepel Bendou Bodou Yomadu Tefeya Pa Lungi Binkolo Makeni Port Loko Int'l Airport Yombiro Bendugu M eli Mambolo Bumbuna Bambaya Kurubonla Alikalia Pendembu ole Gbinti Mange Kortimaw Is. Ba g b e Karina Mateboi Batkanu Fadugu er Kamalu Madina Jct. M ab Kambia li Se Kukuna cies car eS Banian Bendugu Koinadugu oa Forécariah Kamakwie N ig Konta Fandié Gberia Fotombu Kabala M Coyah Musaia Bafodia Ma ho Dubréka er Kle 20 10 40 20 60 30 40 80 km 50 mi Department for Peacekeeping Operations Cartographic Section 11° Kenema Kongo 12° 20 10 20 40 30 40 60 50 mi 80 km M 11° be ya Gelahun bu z ul i zi R Vi Belefuanai da ian a Zorzor Nzérékoré Yekepa Mt. Gletohn (Ganta) Gahnpa Kahnple Saniquellie Mt. Nuon-Fa GUINEA . Ghapo ss ( Gonglee es C Gboyi Tapeta Zekera RIVER CESS Guata St n Joh e Timbo C Bokoa Seh n k w e River Cess Juazohn Kopo o Sehnkwehn 10° ATLANTIC OCEAN Edina Buchanan Trade Town GRAND BASSA Marshall Hartford Harbel eg Yella Zigida Irié Poabli Shabli v a l l y Kola (Kola Town) Babu gbe Du D Tatuke Tiehnpo Tawake Fish Town RIVER GEE (Kaobli) Yibuke Dube ( Galio (Dubwe) Pelokehn Yakakahn Duabo Drubo (Zwedru) Tchien (Towai Town) Towabli Ca Toulépleu Taï 9° GRAND KRU C. Palmas Harper 8° LIBERIA 7° Tabou 5° 6° 7° 8° Department of Peacekeeping Operatons Cartographic Section 7° CÔTE D'IVOIRE Guiglo MA Kodeke RY Nyaake Nana Kru LA ND Barclayville Sasstown Nemeke Grand Cess Plibo Du SINOE Sinoe Greenville Bay o Tobli Nu l ly Danané 8° GRAND GEDEH Debli Kpeaple Gbarnga Palala Kpein Sagleipie Suakoko Zienzu NIMBA Butlo BONG Yela a t. P Bong Town S on g an a Lo Belle Yella W MARGIBI Careysburg a k a t a BOMI an ol i og LOFA W Mt. Wuteve Wiesua Gbange ffa a Mendekoma Voinjama Kolahun ff 9° a 0 G Bopolu T ub m Monrovia International boundary County boundary National capital County capital Town, village Road Railroad Airport a na GBARPOLU MONTSERRADO LIBERIA 0 Klay Brewerville Piso Bendaja Sulima GRAND CAPE Bo MOUNT Robertsport Lake ano Ma Kailahun Buedu wi go Pendembu Vahun 10° Guéckédou ra Map No. 3775 Rev. 6 UNITED NATIONS January 2004 5° 6° 7° 8° SIERRA LEONE a Mo M ro or M iR a ko Lo n Law g e 12° Timbo rg ba N im e N ni Ma Ra ng e n The boundaries and names shown and the designations used on this map do not imply official endorsement or acceptance by the United Nations. s) to Ca va e ng Ra tu hn Si n be ) Pu uo bo C a v all C ess K Creek nd and G we n ssa Gr Sa Chapter 1 Introduction April 2006, Monrovia. Thick smoke rises from burning tyres. Shop owners have pulled down their shutters and residents have vacated the streets. A group of discontented former soldiers stages protests in the streets, setting up ad hoc roadblocks and effectively paralysing Liberia’s capital. The mayhem is brought to an end when a contingent from the 15,000-strong UN peacekeeping force fires shots in the air to disperse the angry mob. A few thousand kilometres away, the UN-backed Special Court of Sierra Leone, citing an inability to deal with major security threats, awaits a response to its request to relocate the trial of former warlord and Liberian President Charles Taylor to the International Criminal Court in The Hague. Around the corner, a taxi driver who had just been involved in a car accident refuses to leave his damaged vehicle at the police station for fear that it might be stripped of all its parts during the night. Both scenarios attest to the volatility of Liberia and Sierra Leone, and expose the inability of the state security forces – both military and police – to meet the security needs of their citizens effectively. Both countries had only recently emerged from protracted periods of civil war, and the international community has been involved in rebuilding – and in some cases building from scratch – the various state institutions. Following the end of the Cold War, international organisations and coalitions of countries have increasingly engaged in what is commonly labelled post-conflict reconstruction and peacebuilding, particularly in countries recovering from violent conflict that led to state failure.1 Peacebuilding has formed a core aspect of international intervention in postwar societies ever since then UN Secretary-General Boutros Boutros-Ghali introduced the concept as key to successful preventive diplomacy, peacemaking and peacekeeping. In An Agenda for Peace, Boutros-Ghali defined peacebuilding as the medium- to long-term process of rebuilding war-affected communities through identifying and supporting structures to consolidate peace so as to avoid a relapse into conflict.2 Over time, the definition of peacebuilding has gradually expanded to become a multidimensional construct that includes the process of rebuilding the political, security and socio-economic dimensions of societies emerging 2 Judy Smith-Höhn from conflict. Peacebuilding stretches from humanitarian relief through transitional rehabilitation to long-term development efforts such as security sector reform (SSR).3 In recognition of the fact that there can be no development without peace, reconstruction efforts have focused heavily on reforming the security sector in such post-conflict societies.4 Although some authors may not agree with the statement that the (re)installation of an order of violence – that is, freedom from violence and coercion through the provision of security – is an unconditional prerequisite to any effective rebuilding process following a decline in the intensity of armed conflict, most concede that security is a crucial public good in post-conflict societies, though not necessarily the most important one.5 Numerous efforts have been made to reconstruct or build from scratch a security sector that can effectively provide for the physical well-being of its citizens. Successes can be tallied, but there remains significant scope for improvement. A few years following the official cessation of conflict in socalled post-conflict societies – usually once a relatively secure environment has been maintained by external actors through international peacekeeping operations – one can begin to analyse the progress made in the (re)construction of the security sector. While acknowledging that SSR and the demands and challenges faced in such processes are context-specific and ever-changing, lessons that are learnt from such analyses can inform the development of future reform strategies if conducted within a comparative framework. Given that SSR forms part of a wider peacebuilding, conflict prevention and post-conflict reconstruction agenda that obtains its primary impetus from external actors, there is a need for a rigorous academic analysis of this largely policy-dictated concept. In the tradition of Max Weber, where the state is defined as – among other things – ‘a human community that (successfully) claims the monopoly of the legitimate use of physical forces within a given territory’,6 such reform strategies focus on (re)installing the state’s said monopoly and taking the security sector out of private hands and bringing it under the state’s mantle. However, this approach ignores the fact that the post-colonial state has itself often played a central role in causing and promoting the escalation of violence. Strategies implemented by external actors to reform the security sector should therefore include a consideration of whether the state should or can be reconstructed into its pre-conflict shape, since the goal is the establishment of a viable society capable of managing its conflicts without resorting to violence. A notion such as this would then include, for example, Introduction 3 alternative or traditional security models that coexist with the state structures into the reform process. Analogous to the often restricted focus of policy-makers on the role of the state within the security sector, the academic focus has generally been limited to the strategies undertaken by external actors to ensure the survival of the local population through strengthening state security institutions, while the efforts made by locals themselves to survive amid hostile situations and resolve emerging conflicts have been given little attention. Moreover, although current research goes beyond the focus on state structures to include sub-state and transnational actors in the security sector, there is a shortage of empirical evidence assessing the existence and efficiency of all these external, state and non-state actors. Research on SSR is a fairly recent phenomenon, which partly explains the scarcity of empirical evidence from local expert observers, public officials, civil society and members of the local society. Hence the security sector reform debate is characterised by a disproportionate ratio between numerous general recommendations on what could and should be done and specific suggestions that are based on a thorough analysis of a particular post-conflict situation.7 This study aims to overcome this discrepancy by combining the wealth of theoretical insights on security sector reform and reconstruction with empirical evidence gathered on the ground. Thus the primarily deskbased approaches towards defining key elements of post-conflict SSR will be tested against empirical evidence gathered in the two post-conflict societies of Liberia and Sierra Leone. What, then, are the key issues that concern analysts of security sector reform processes in post-conflict settings? Is there general agreement on the most effective strategy, and do the numerous actors that are involved in such processes coordinate their activities? Are current approaches towards reforming post-conflict security sectors successful, and, most importantly, are there tools to measure the effectiveness of these strategies? The key objective of this particular study is to ascertain the suitability of previous strategies for reforming and reconstructing the security sectors in post-conflict societies, thus pointing to key challenges for policy-makers in the development of their reform strategies. Two overarching principles guided the selection of the two country case studies. On the one hand, there was the recognition that an in-depth empirical analysis of all instances of externally driven SSR in societies emerging from conflict would require a large amount of human and financial resources, hence the need to restrict the number of cases. On the other hand, 4 Judy Smith-Höhn there was a need to select cases that were comparable, i.e. the cases needed to share certain characteristics while displaying divergence in others. Liberia and Sierra Leone fulfilled this latter criterion in that they displayed divergence in the different reform approaches that were undertaken by the (mostly international) implementing agencies in each case, with Sierra Leone at a slightly more advanced stage of the reform process. On the other hand, they displayed a number of contextual similarities, i.e. the two countries shared certain background conditions. Both West African countries witnessed protracted periods of violence during the 1990s. In fact, the civil wars in both countries are closely linked: mercenaries surreptitiously crossed the common border to fight in the other’s civil war, and each ruling class has continuously meddled in the affairs of its neighbour. Charles Taylor’s role in instigating violence and supporting rebel groups in Sierra Leone is the most prominent case in point. Yet another contextual similarity between the two is the extensive involvement of international and regional actors such as the United Nations (UN) and the Economic Community of West African States (ECOWAS) during the post-conflict phases. Also, the socio-economic conditions are similar and there appear to be a number of different (types of) security actors in both cases. The overall approach to the analysis of security sector reform involves both deductive and inductive reasoning. Firstly, current approaches adopted by the leading international and regional agencies involved in SSR and reconstruction are scrutinised, particularly those that focus on post-conflict situations. This enables the identification of common criteria that can be utilised to measure the efficacy of SSR strategies through empirical testing. Drawing from the analysis of the empirical data and having tested the validity of the criteria used to measure effective reform of the security sector, the focus returns to the drawing board to formulate or amend existing criteria for the successful implementation of SSR and make some general observations regarding the efficacy of SSR strategies undertaken in postconflict environments. In most cases, the reconstruction of the security sector involves the rebuilding of domestic public security institutions, and particularly the reestablishment of a legitimate monopoly on the use of force.8 This book challenges this normative axiom and suggests a broader understanding of the security sector as an institution to include not only state actors, but also nonstate informal security actors that have proven to play a role in the security arena as stipulated in the OECD DAC9 Guidelines on Security System and Governance Reform. These guidelines define a broader security sector that includes core security actors such as the armed forces and the police, Introduction 5 security management and oversight bodies, and justice and law enforcement institutions as well as non-statutory security forces such as liberation armies and private security companies –donors rarely engage with the latter.10 Though most authors recognise the definition of the security sector or system as stipulated by these guidelines, they tend to focus on the state sector when analysing reform processes. This a priori exclusion of the nonstate players appears to be a key deficit in such studies: in many postconflict societies, non-state actors tend to play a far more significant role in peacebuilding processes than presumed. This limited focus, in part, contributes to some of the shortcomings of SSR strategies. In post-conflict settings, years of civil war have often left state security institutions in shambles. Hence these state actors usually remain heavily dependent on external support even after the externals have officially entrusted local authorities with the primary responsibility of providing security to citizens. In the face of the state’s inability to hold a monopoly on the use of force, the question therefore arises as to whether alternative modes of security provision have emerged to fill this security vacuum. Perhaps non-state actors often found in such environments should be included in the security architecture? If the focus is no longer merely on ensuring that the executive organs of the state are developed into an efficient security sector, but instead on the inclusion of other relevant actors who are bound by the same rules as all others, it might increase the prospects of ensuring the long-term sustainability of security. By exploring the mechanisms applied by the local population to resolve their conflicts and by identifying the actors that are considered to be involved in providing security at all levels – national and local, urban and rural – one counters the policy bias of focusing mainly on external and state actors, albeit that they are important. The issue of local ownership, a commonly accepted prerequisite to effective security provision, plays a key role here. Having identified the salient issues and affirmed the need for an empirical analysis of SSR strategies, what type of empirical data are required to test the hypotheses? Simply put, who should be asked what and why? Country data, such as crime statistics, are one source of empirical information. However, there are two problems associated with the use of such data: the reliability of the source; and the accuracy and currency of the data. This is particularly so for countries that have recently emerged from protracted periods of conflict, where there is a lack of infrastructure and manpower to provide reliable statistical information. In the cases under study, such data are only available for Sierra Leone, and their quality and 6 Judy Smith-Höhn reliability are questionable. Although the UN Mission in Liberia (UNMIL) monitors the security situation and gathers data on security-related matters, the information was not publicly accessible at the time this analysis was performed and could therefore not be used. Primary and secondary literature are a second source of information. Though a comparative desk study may certainly provide new insights, few academic analyses have been conducted on SSR in Liberia and Sierra Leone, and the empirical data, where available, are generally not comparable due to the expected divergence in the research designs of these various studies. There was therefore a need to generate reliable, and more importantly comparable, data on security-related issues in both post-conflict countries. The question is then: what sort of empirical data can facilitate the identification of the strengths and weaknesses of security sector reform strategies in post-conflict societies? Victor Azarya’s work on reordering state-society relations in Africa provides some valuable insights into why local perceptions not only enable an assessment of the performance of security actors, but are also relevant for the future success of reform and reconstruction strategies.11 Azarya suggests an alternative to the state-centred approach to studying state-society relations in contemporary Africa by placing greater emphasis on societal responses to state actions. In his attempt to explain the weakening of the state, Azarya developed the notion of incorporation into and disengagement from the state as ‘societal responses to state actions (or anticipated state actions) which lead to a perceived change in the field of opportunities of given groups or individuals’.12 Applying the concept of incorporation and disengagement to the provision of security within society, with the state ideally dominating and regulating the behaviour of all relevant actors within this realm, acts of association with or distancing from the state by individuals or groups are indicative of that sector’s efficiency. If the weakening of government agencies can lead individuals and groups to devise alternative methods of sustaining themselves economically, then surely the same can be said for the pursuit of interests related to the provision of security. Thus if individuals or groups perceive the state to be incapable of providing or managing security in the country, they will disengage from this arena.13 As other structures gain increased authority, state security structures weaken. Thus, in order to determine whether security bodies exist that are capable of providing security, or whether effective oversight mechanisms are in place, one can consider society’s response to the state in this area. In other words, one must question whether individuals or groups within society perceive the state as Introduction 7 providing effective security, for example, or whether citizens tend to opt for alternative means of security provision. To use Azarya’s terminology: do citizens incorporate into or disengage from the state when it comes to fending for their security needs? An analysis of the efficacy of the security sector that is based on empirical evidence gathered on the ground must therefore take into account the perceptions of the local population as regards security-related matters. Returning to the structure of this study, the approach is analytical, empirical and interpretive. Chapter 2 outlines current concepts of security sector reform and reconstruction as implemented by the leading international agencies, particularly those that focus on post-conflict situations. A basic definition of the terminology and key expressions used precedes the discussion on SSR. The lack of consensus among scholars on the definitions and descriptions of what comprises the security sector necessitates an initial distinction between the various interpretations. A set of key criteria are developed, which are used as guidelines for assessing the current effectiveness of security sector reform efforts. Having outlined the basic guideline for the analysis of SSR, Chapter 3 focuses on the methodology applied. It outlines the benefits of the comparative method as it is – or at least should be – applied in the field of political science. The point of departure for deliberations on the appropriate research design is the recognition that there is no ‘one-size-fits-all’ blueprint. Hence the chapter develops and elaborates on the research design most suited for the analysis at hand, namely the comparative case study approach, and a framework for analysis is developed, outlining the various levels of analysis and identifying the cases to be compared. The chapter concludes with a description of the qualitative and quantitative tools that were used for data gathering. The study takes a mixed-methods approach to data gathering, combining both quantitative and qualitative tools to collect empirical information on the state of the security sector in the case studies. Field research was conducted in both countries, extending over a period of three months in each; and the findings are based on the combination of unique empirical data from three sources: a survey poll of 700 respondents in three urban areas in each country; four in-depth focus group discussions (FGDs), each comprising six to eight participants; and 60 semi-standardised interviews with elites and local and international experts. Where available, non-conventional literature and other key documents such as internal reports, fact sheets and crime statistics were gathered to substantiate the empirical data. The comparison is structured and focused: it is focused in that it considers a selected aspect of each case – i.e. the security sector – and it is Judy Smith-Höhn 8 structured in that the same set of research questions were posed for both cases. Chapter 4 on the case studies comprises four sections that are structured according to the framework developed in Chapters 2 and 3. The first two sections provide a chronological overview of the development of each conflict, as well as the SSR strategy adopted in each case. The next section applies the comparative research design, comparing the security sector reform of both countries. The chapter concludes with a synopsis of the various comparative steps undertaken. Chapter 5 concludes with a summary of the findings from the previous chapters and a reflection on the implications of these findings for further research. 1.1 Background This volume is developed from a research project entitled ‘Legitimate Oligopolies of Violence in Post-Conflict Societies with a Particular Focus on Liberia and Sierra Leone’14 that was implemented by the GIGA Institute of African Affairs in Hamburg, Germany. As part of the research team, the author conducted the fieldwork in Liberia and Sierra Leone and used the project as a framework for the development of a doctoral thesis. The project recognised that little attention had thus far been given to determining which state or pre-state institutions emerge during post-conflict periods, when actors enjoy considerable leeway in the domain of security, and when institutions, concepts of law and order and patterns of action are challenged and renegotiated. It acknowledged the lack of empirical research on the constellations of authority following the cessation of conflicts in Africa, a shortcoming which corresponded to deficiencies on the level of policy-making. It appeared that Western donors based their approaches to post-conflict reconstruction on the wholly unchallenged assumption that only the state could legitimately hold the monopoly on the use of force. However, the project noted that such approaches should, in fact, be preceded by questioning the pre-conflict state’s validity, for it was generally accepted that the post-colonial state had often played an important role in the persistence and escalation of violence. Hence, there should be prior consideration as to whether the state can and should be reconstructed into its pre-conflict shape. The project’s primary goal was to determine which actors (the traditional authorities, the remnants of state security, private entrepreneurs, Introduction 9 international peacekeeping missions and so on) provide security in a situation of fragmented authority and sanctioned violence and crime. Moreover, the project sought to analyse the conditions under which different groups within society considered these actors to be legitimate: some actors might protect specific groups within the population while representing a threat to others. In other words, the project sought to determine whether oligopolies of violence exist in post-conflict societies, and, if so, how they are constituted: whether temporal, homogeneous or heterogeneous, territorial or functional. These questions were addressed in empirical case studies of Liberia and Sierra Leone and a desk study of six other post-conflict societies, namely Afghanistan, Bosnia-Herzegovina, Guatemala, Nigeria, Pakistan and Somaliland. The case studies of oligopolies of violence in post-conflict situations in Sierra Leone and Liberia complemented the general theoretical insights gained from the literature and the subsequent desk study.15 Though not focused on security sector reform, the guiding questions of the research project, in revealing the constellation and perceived legitimacy of security actors in Liberia and Sierra Leone, also highlighted the security sector as a whole and provided insights into the perceived efficacy of this sector, particularly those actors involved in actually providing physical security. Therefore, although the focus of the research project was actor-centred, the data gathered could be adjusted and used effectively to examine the entire reform process. Notes 1 2 3 4 Michael Brzoska and Andreas Heinemann-Grüder (2004) ‘Security sector reform and post-conflict reconstruction under international auspices’, in Alan Bryden and Heiner Hänggi (eds) Reform and Reconstruction of the Security Sector. Münster: LIT, pp. 121– 143; available at www.dcaf.ch/publications/bm_ssr_yearbook2004.cfm (accessed 13 February 2006). Boutros Boutros-Ghali (1992) An Agenda for Peace: Preventive Diplomacy, Peacemaking and Peacekeeping. New York: United Nations; available at www.un.org/ documents/ga/res/47/a47r120.htm (accessed 1 July 2005). J. ’Kayode Fayemi (2004) ‘Governing insecurity in post-conflict states: The case of Sierra Leone and Liberia’, in Alan Bryden and Heiner Hänggi (eds) Reform and Reconstruction of the Security Sector. Münster: LIT, p. 2; available at http://se2.dcaf.ch/serviceengine/ FileContent?serviceID=DCAF&fileid=B0EA7833-62AD-2215-99BF-182ADF9F413A &lng=en (accessed 13 June 2006). Albrecht Schnabel and Hans-Georg Erhart (2005) ‘Post-conflict societies and the military: Challenges and problems of security sector reform’, in Albrecht Schnabel and HansGeorg Erhart (eds) Security Sector Reform and Post-Conflict Peacebuilding. Tokyo: 10 Judy Smith-Höhn 5 6 7 8 9 10 11 12 13 14 United Nations University Press, p. 1; David M. Law (2006) ‘The post-conflict security sector’, DCAF Policy Paper 14; available at www.dcaf.ch/_docs/ PP14_post_conflictss.pdf (accessed 12 December 2006); Heiner Hänggi (2004) ‘Conceptualising security sector reform and reconstruction’, in Alan Bryden and Heiner Hänggi (eds) Reform and Reconstruction of the Security Sector. Münster: LIT, pp. 1–11; available at http://se2.dcaf.ch/serviceengine/FileContent?serviceID=DCAF&fileid= 0CD0E86D-F99B-3B67-0D75-2FFB3FB5CA3D&lng=en (accessed 13 June 2006); Brzoska and Heinemann-Grüder, note 1 above. Cf. Dirk Salomons (2005) ‘Security: An absolute prerequisite’, in G. Junne and W. Verkoren (eds) Postconflict Development: Meeting New Challenges. Boulder, CO: Lynne Rienner, p. 19; Daniel Lambach (2007) ‘Oligopolies of violence in post-conflict societies’, GIGA Working Paper Series, Hamburg, p. 1. Max Weber (1958) From Max Weber: Essays in Sociology, eds H. H. Gerth and C. Wright Mills. New York: Oxford University Press, p. 78. See e.g. Brzoska and Heinemann-Grüder, note 1 above, p. 125. Ibid., p. 121. The Organisation for Economic Co-operation and Development’s Development Assistance Committee (OECD DAC), henceforth referred to as the DAC. OECD (2004) Security System Reform and Governance, DAC Guidelines and Reference Series, pp. 20ff; available at www.oecd.org/dataoecd/8/39/31785288.pdf (accessed 13 January 2007). Victor Azarya (1988) ‘Reordering state-society relations: Incorporation and disengagement’, in Donald Rothchild and Naomi Chazan (eds) The Precarious Balance. State and Society in Africa. Boulder, CO: Westview Press, pp. 3–21. Ibid., p. 6. Incorporation, on the other hand, would entail the integration of traditional, commercial and other alternative security agencies into the state security sector. The term ‘oligopoly of violence’ was coined by Andreas Mehler; it is defined as ‘a fluctuating number of partly competing, partly co-operating actors of violence of different quality’ and has stimulated controversial debate. The term ‘oligopoly’ is derived from economy theory, where it denotes a market or industry that is dominated by a small number of suppliers. Subsequently, an oligopoly with two suppliers is a duopoly, one with many is termed a polypoly. Applied to the security context, an oligopoly of violence signifies a security arena that is occupied by a few security actors providing either the same kind of security (homogeneous oligopoly) or different kinds of security (heterogeneous oligopoly). Mehler further distinguishes between three sub-types of oligopolies: territorial oligopoly, based on an agreement between actors for control of specific territories within states – within these borders, these actors may hold the monopoly on the use of force; functional oligopoly – heterogeneous oligopoly comprising various security actors that provide security to protect from a range of security threats; and temporal oligopoly – displays seasonal variations (e.g. night and day). If understood as an oligopoly where the roles of actors and their relative meaning within the oligopoly can change over time, its temporary nature should be considered as a general characteristic of all types of oligopolies. See Andreas Mehler (2004) ‘Oligopolies of violence in Africa south of the Sahara’, Nord-Süd Aktuell, 18(3): 539–548; available at www.giga-hamburg.de/content/staff/mehler/publications/mehler_nsa_04-3.pdf (accessed 5 August 2005). Introduction 11 15 The cases were subjected to different models of comparison. First, the current situation in both states was compared; second, Liberia as it was at the time of survey was compared to Sierra Leone as it emerged from violent conflict at the turn of the millennium; and, finally, a comparison is made on the timeline between the past and present situation in each country. The conclusions drawn from these comparisons are then collated with the results of a study on the organisation of security in 11 other post-conflict societies, which was developed according to a uniform model of analysis. Finally, the project seeks to establish criteria that permit an assessment of the prospects for goal-oriented cooperation with non-state actors, a result which is of direct relevance for policy-making. Chapter 2 On Security: Definitions and Theories The concept of security is multidimensional, and those analysing the security sector normally define security depending on their analytical purposes and practical problems. The chapter sets out by defining key terms, such as postconflict, peacebuilding, security and security sector. It then provides a brief outline of the various theoretical approaches to security sector reform that will form the basis for the analysis of the efficacy of the security sector in post-conflict societies. This is not exhaustive in its illustration of the various approaches to SSR, but instead outlines the approaches of the lead international agencies involved in these processes (the United Nations, the OECD DAC, the UK Department for International Development (DFID) and the World Bank). 2.1 Post-conflict societies The term post-conflict calls to mind a number of interrelated developmentoriented concepts, such as post-conflict reconstruction, post-conflict recovery and post-conflict peacebuilding. However, there appears to be a lack of consensus in both academia and policy-making circles as regards the definition of a post-conflict situation. Usually, it describes the period when a conflict is either interrupted or (temporarily) halted by means of a peace agreement, the victory or defeat of a conflict party or the intervention of an external actor. Kumar and de Zeeuw, for example, define post-conflict societies as those where the cessation of intra-state conflict is followed by the international community’s recognition of the government as legitimate.1 In other words, a ‘post-conflict society’ is one that has emerged from protracted violent conflict either through externally brokered peace negotiations, international military intervention or the victory of one warring faction over another.2 Like most definitions, this implies that a post-conflict period is one that is characterised by peace. Yet this is frequently not the case. In fact, post-conflict phases are often marked by high levels of violence, and a 14 Judy Smith-Höhn simplified concept of conflict versus post-conflict fails to describe adequately the phenomenon of violent conflict. Similarly, during times of war there are often long periods without any violent outbreaks. As for many authors,3 the inaccuracy of Kumar and de Zeeuw’s definition is a result of preconceived notions implicit in such definitions and is semantic in nature: the prefix post implies that the conflict has ended and the situation is returning to normal,4 hence the term indicates a supposed unidirectional dynamic from a period of war to a period of peace. In other words, the term post-conflict creates a false mental dichotomy between a conflict and a post-conflict phase by attaching the temporal prefix ‘post’ to the ‘conflict’, which, in fact, has no temporal meaning.5 These terms are then used synonymously with ‘war’ and ‘peace’, respectively. In this dichotomy, ‘conflict’ denotes situations that are structured by violence carried out by organised actors according to some dominant conflict narrative, while ‘postconflict’ implicitly signifies the end of such violence and the return to a peaceful, ‘normal’ situation.6 As a result of such misconceptions, most studies on ‘post-conflict’ situations are still informed by the dichotomised understanding of conflict versus post-conflict, with the level of violence acting as a benchmark: There are few truly postconflict situations. Conflicts become more or less violent, more or less manifest or latent, but they seldom stop altogether. ‘Postconflict’… is shorthand for conflict situations, in which open warfare has come to an end.7 This definition upholds the dichotomy by equating post-conflict with the cessation of open warfare. Lambach rightly points to the operational limits of such an approach, since in most cases there is never an identifiable single moment of cessation.8 Conflicts may either intensify or decline towards the end, and research has shown that this is contingent upon the side that wins. When state forces are losing, conflict is likely to intensify, whereas conflict wanes when insurgent groups are defeated and they withdraw into remote locations or hide among the population.9 The developments in Sierra Leone in 2001/2002 following the formal declaration of the end of civil war are a good example of this lack of a definitive ending of the conflict.10 The level of violence is also often used as an indicator of the end of a conflict. Yet experience has shown that so-called post-war periods are often marked by a decline in security.11 On Security: Definitions and Theories 15 As a result, post-conflict realities rarely bear much resemblance to what is implied by their definition. Rather, death and injury rates often remain comparatively high even after an armed conflict has come to an ‘end’.12 Lambach, among others, suggests an improvement to the definition of a post-conflict situation by considering war to be ‘a state of mind shared among participants’.13 By acknowledging that during times of war there are often extended periods of non-violence, and contrarily, times of so-called peace are often tainted by violent outbreaks, one can grasp situations of conflict and post-conflict as social constructs – that is, ‘discursive delimitations of the kind of behavior that is to be expected and allowed in a given set of circumstances’.14 Lambach provides a concise definition of post-conflict as a ‘narrative of peace’.15 In other words, when a violent outbreak is no longer explained in terms of the dominant ‘narrative of conflict’, that conflict can be considered to have come to an end. Violence that occurs in the aftermath of a conflict is often labelled criminal violence, thus effectively assigning a different quality to it than to the violence that transpires during civil war.16 Thus a post-conflict society is one in which the violence that is perpetrated by the former warring factions is no longer carried out in terms of the central narratives of the previous conflict.17 A conflict can therefore end ‘when the narrative of violence changes or when the conflict parties are removed (or remove themselves) from the scene’.18 A further conceptual clarification is needed for the term post-conflict to characterise not a transitional period between war and peace, as is often assumed in academic literature and political practice alike, but a distinctive phase that follows its own logic.19 This would facilitate a better understanding of post-conflict situations and inform and improve policymaking in this arena. By acknowledging, for instance, the depleted absorptive capacity of state institutions during post-conflict periods, development aid could be increased gradually and not disbursed en masse during the first few years and then reduced, as is usually the case.20 2.2 Human security: A narrow definition The 1990s witnessed the emergence of new security concepts such as human security, indicating the paradigmatic shift from the dominance of national (and international) security to the growing importance of transnational, subnational and individual security.21 16 Judy Smith-Höhn Whereas previously security had been defined as national security, understood as the state’s ability to defend itself against external threats, it now came to encompass the broader notion of human security, which moves away from the state-centric approach to focus on the protection of individuals. Despite consensus regarding this primary goal of human security, proponents of the concept disagree on its scope – that is, the types of threats individuals need protection from.22 The narrow understanding of human security confines it to violent threats to individuals. According to former UN Secretary-General Kofi Annan, it is the ‘protection of communities and individuals from internal violence’.23 Others advocate a broader understanding of the concept, moving beyond protection from violence to include protection from hunger, disease and natural disasters as well as from economic insecurity and threats to human dignity.24 In its guidelines on helping prevent violent conflict, the OECD DAC notes that security is: increasingly viewed as an all-encompassing condition in which people and communities live in freedom, peace and safety, participate fully in the governance of their countries, enjoy the protection of fundamental rights, have access to resources and the basic necessities of life, and inhabit an environment which is not detrimental to their health and wellbeing.25 This broader view of human security has many adherents, particularly as it acknowledges the interrelatedness of such societal threats. However, from a methodological and pragmatic perspective, the current study has focused on the narrow understanding of human security. Methodologically, a concept that groups together threats as diverse as genocide and affronts to personal dignity runs the risk of becoming diluted. For, as has been noted elsewhere: ‘once a concept is defined so that it encompasses all the incidents that are members of a given category… it ceases to enhance one’s ability to explain’.26 Moreover, such concepts, though useful for advocacy, do not facilitate policy analysis. This certainly explains, to some extent, why the broad definition of human security developed by the UN Development Programme (UNDP) in its often-cited Human Development Report 1994 has rarely been used to guide research programmes.27 From a pragmatic perspective, the narrow definition of human security as the protection of individuals from physical harm serves a very practical purpose. The broader the concept, the more difficult it becomes to gather data and formulate succinct questions for respondents to answer. In turn, the narrower or more concise a concept, the easier it becomes to On Security: Definitions and Theories 17 pinpoint its distinguishing features and hence pose questions which respondents can answer in a few words. This is particularly crucial when conducting a quantitative survey, where the interviewer poses a number of brief questions which should be self-explanatory, i.e. the questions should not require further elaboration. 2.3 Peacebuilding As noted earlier, peacebuilding became a core feature of international intervention in post-war societies when former UN Secretary-General Boutros Boutros-Ghali introduced the concept in An Agenda for Peace. Over time, the concept has evolved and now applies not only to rebuilding in postconflict settings but also includes conflict prevention, conflict management and post-conflict reconstruction efforts. Broader definitions of peacebuilding incorporate nearly all forms of international assistance to societies that have suffered or are at risk of armed conflict (government decentralisation, rural development, social investment, fiscal reform, human rights, security and justice sector reform, nationbuilding). Others are more specific, yet focus attention on clarifying international mandates rather than conditions for peace in the host country. More recent approaches focus on the comparative value of international peacebuilding efforts in relation to one another and in contrast to regional and domestic efforts. Essentially, post-conflict peacebuilding can be defined as a mammoth experiment in social engineering, with the goal of creating conditions for sustainable peace within countries emerging from civil wars. 2.4 Security sector reform and reconstruction in the post-conflict setting As noted above, SSR forms part of a wider peacebuilding, conflict prevention and post-conflict reconstruction agenda that obtains its primary impetus from external actors. Post-conflict peacebuilding is a multidimensional construct that stretches from humanitarian relief through transitional rehabilitation to long-term development efforts such as justice and security sector reform.28 As part of the peacebuilding agenda, one can identify certain basic principles of SSR as implemented by the leading international agencies involved in such processes (DFID, United Nations, World Bank) that are 18 Judy Smith-Höhn considered relevant and applicable to all post-conflict settings. Both the scholarly and practitioner-oriented work in this field is quite comprehensive and hardly requires further elaboration. What is lacking, however, is an SSR strategy that is context-sensitive – in other words, it takes into account the local and potential regional peculiarities of a particular reform context. This is the point of departure for the current study’s deliberations on security sector reform and reconstruction, which will add to the knowledge base context-specific empirical evidence of the success or failure of current reform efforts. 2.4.1 Defining the security sector The variance in the approaches adopted by external actors when implementing SSR indicates a general divergence in their respective understanding of the security sector. A comprehensive analysis of security sector reform and reconstruction strategies therefore necessitates a prior definition of the sector. Much like the concept of security, there are almost as many definitions as there are scholars and institutional actors on the notion of what comprises the security sector.29 Given the diversity among actors within the development community, definitions are often shaped by the domestic arrangements of these external actors. Such limitations notwithstanding – for scholars cannot escape from the contexts within which they develop their arguments – there appears to be little doubt regarding the basic elements of the security sector among scholars and practitioners alike. Much has been written on the subject, and the following discussion will necessarily be brief. The purpose here is merely to set the context for the description of key SSR strategies that have provided the framework for such intervention in several, particularly African, countries. The core elements that are discernible among the various interpretations of what constitutes this sector will thus be identified. The Geneva Centre for the Democratic Control of Armed Forces (DCAF) is one of the leading international institutions dealing with SSR issues. Established by the Swiss government in October 2000, it collaborates with governments and civil society to promote and support the democratic and civilian control of security sector organisations such as police, intelligence agencies, border security services, paramilitary forces and armed forces.30 DCAF has acted in an advisory capacity to international organisations like the United Nations, the European Union (EU) and the OECD31 as well as governments and civil society actors. In addition to policy advice, it conducts research on issues related to the security sector, On Security: Definitions and Theories 19 going beyond the ‘democratic control of armed forces’. DCAF’s definition of the sector and its approach to SSR can thus be viewed as a showcase for the international donor community’s approach to the subject. In Germany, the Bonn International Center for Conversion (BICC) is quite prominent in this field. Generally, its perspectives on security sector reform and reconstruction are indistinguishable from those of DCAF, and both have collaborated in conducting research and publishing work in this area.32 BICC has also worked extensively with the federally owned Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ), Germany’s international cooperation enterprise for sustainable development.33 In 2000 they co-published a monograph on SSR in developing countries wherein the sector was defined as comprising the: armed forces, paramilitary units, the police and gendarmerie, and the intelligence service. It also includes the judicial and penal systems, as well as civil society in general, which also plays a significant role in democratic control.34 DCAF’s 2004 Yearbook on SSR acknowledges the fact that the definition of what constitutes the security sector is a complex, evolving and therefore unsettled matter.35 It argues for a broad understanding of the sector and provides a useful overview of the key definitions as shaped by the various perspectives adopted (Table 2.1 on page 20). Table 2.1 shows that, but for one, definitions of what comprises the security sector are broad. Definition A, one of the ‘older’ definitions, was formulated by the Informal DAC Task Force on Conflict, Peace and Development Co-operation.36 Like the OECD DAC, Chanaa, in her study on the issues, challenges and prospects for SSR, opted for a narrow definition of the security sector to cover ‘those elements that have been granted a legitimate and exclusive role in the exercise of coercive power in society to deal with external and internal threat to the security sector and its citizens’.37 The security sector thus comprises all those organisations that have the right either to use or order the use or threat of force, as well as the civil structures responsible for their management. This would include military and paramilitary forces, intelligence services, national and local police forces, border guards and custom services, judicial and penal systems and the civil authorities authorised to monitor and supervise these agencies. The key flaw in this definition is that it is confined to the state sector. Non-statutory security actors – such as liberation armies and private military and security Judy Smith-Höhn 20 Table 2.1 Perspectives Narrow Broader Definitions of the ‘security sector’ Definition A Definition B Definition C Definition D Security forces Groups with a mandate to Core wield security instruments actors of violence Organisations authorised to use force Civilian management and oversight bodies Institutions with a role in managing and monitoring Security management and oversight bodies Civil management and oversight bodies Judiciary, penal system, human rights ombudsmen Justice and law enforcement institutions Justice and law enforcement institutions Nonstatutory security forces Nonstatutory security forces Nonstatutory civil society groups Focus Statecentric Humancentric companies – if not considered as part of the de facto security sector, must at least be acknowledged as key actors that influence this sector.38 As for the literature particularly focused on SSR issues in the African context, it appears that definitions of what comprises the sector are in line with the norm. In their handbook on security sector governance in Africa, Ball and Fayemi have a narrow understanding of the sector as such, defining it as: all the institutions of the state responsible for securing the state and its population from fear of violence… the security sector is taken as consisting of the armed forces, the police and intelligence services, the related ministries On Security: Definitions and Theories and departments, organisations.39 paramilitary forces, and governmental 21 oversight The authors do, however, acknowledge that several other actors – both official and non-official – form part of a ‘security community’ that affects the quality of governance of the sector. Hence, an analysis of SSR should focus on this entire ‘security community’ to include non-state security organisations as well as non-statutory civil society bodies.40 Ball and Fayemi distinguish between state and non-state security organisations by describing the former as those groups ‘that can legally use force to protect the state and its population’.41 Again, this is a state-centred approach. Hendrickson, among others, defines the security sector in a slightly broader context to include not only ‘groups with a mandate to wield instruments of violence’ and those institutions that have managing and monitoring responsibilities, but also the judiciary, penal system and human rights ombudsmen.42 The OECD DAC’s earlier attempts to define the security sector, or to use its terminology the security system, encompassed the traditional set of actors associated with the sector, namely the security forces and the relevant civilian bodies and processes needed to manage them: state institutions which have a formal mandate to ensure the safety of the state and its citizens against acts of violence and coercion (e.g. the armed forces, the police and paramilitary forces, the intelligence services and similar bodies; judicial and penal institutions; and the elected and duly appointed civil authorities responsible for control and oversight (e.g. Parliament, the Executive, the Defence Ministry, etc.).43 The OECD DAC has since revised this definition to encompass four sets of actors as stipulated in Definition C.44 The broadest definition of the sector is provided by the UNDP (Definition D).45 In addition to the four sets of actors identified by the OECD DAC, the UNDP includes non-statutory civil society groups – professional groups, the media, research organisations, advocacy organisations, religious organisations, non-governmental organisations, community groups – as an integral part of the security sector.46 Though this definition moves away from the state-centric approach, it tends towards the other extreme by allowing the inclusion of all sorts of actors, thus running the risk of becoming too diffuse. 22 Judy Smith-Höhn Rocklyn Williams provides a second ‘African’ approach to defining the security sector.47 According to Williams, the security sector can be defined in terms of either a maximalist (Definition D), minimalist (Definition A) or pragmatic approach, the author himself favouring the latter. The pragmatic approach views the security sector: as being constituted out of the traditional statutory instruments of statecentred security – the armed forces, the police, paramilitary organisations and the intelligence services as well as such institutions as guerrilla forces (where such forces are eligible for eventual integration into new national security institutions and/or demobilisation), and indigenous military organisations that have played a positive role in contributing to the physical security of communities (civilian defence forces in Sierra Leone, and selfdefence units in South Africa, for example).48 This definition is unique in that it considers both the statutory and nonstatutory actors actively involved in the provision of physical security to comprise the security sector, barring the civilian oversight bodies from inclusion. Although this restriction may prove useful in that it would confine an analysis of the sector to fewer actors, it fails to acknowledge the role of justice and law enforcement institutions, for example. The most commonly cited definition – by scholars and practitioners 49 alike – is that provided by the OECD DAC guidelines on security system reform and governance (Definition C).50 1. 2. 3. Core security actors: armed forces, police, gendarmeries, paramilitary forces, presidential guards, intelligence and security services (both military and civilian), coastguards, border guards, customs authorities, reserve or local security units (civil defence forces, national guards, militias). Security management and oversight bodies: the executive, national security advisory bodies, legislature and legislative select committees, ministries of defence, internal bodies (finance ministries, budget offices, financial audit and planning units) and civil society organisations (civilian review boards and public complaints commissions). Justice and law enforcement institutions: judiciary, justice ministries, prisons, criminal investigation and prosecution services, human rights commissions and ombudsmen, customary and traditional justice systems. On Security: Definitions and Theories 4. 23 Non-statutory security forces, with whom donors rarely engage: liberation and guerrilla armies, private bodyguard units, private security companies, political party militias.51 While most authors recognise the definition of the security system as stipulated by these guidelines, their analyses tend to focus on certain aspects of that sector, mostly the state side.52 Though the limitation of an analysis to a particular set of actors within the security sector is understandable considering the multitude of actors identified as being relevant, the a priori exclusion of non-state actors is, in the author’s opinion, the problem with such studies, as in many post-conflict societies non-state actors tend to play a far more significant role in peacebuilding processes than presumed by (mostly international) observers, as will be shown in Chapter 4. So which of these definitions should be applied in this analysis? Undoubtedly, the security sector cannot be viewed as isolated from other government sectors or as a singular entity detached from the society within which it operates. However, this does not imply that all sorts of actors are included in the equation: this would render an empirical analysis of the sector difficult, if not impossible. A useful distinction can be made between the various approaches by considering the different perspectives from which the sector is defined, namely either from a security or a governance perspective.53 From a security perspective, the sector includes both statutory and non-statutory security forces.54 In other words, in addition to the state institutions with the formal mandate to ensure the safety of the state and its citizens (e.g. armed forces, police, intelligence and secret services, justice and penal institutions), the sector encompasses private and other nonstatutory security actors (e.g. private military and security companies, liberation armies).55 From the governance perspective, the sector also includes those state actors responsible for managing and controlling the security forces (e.g. executive government, ministries of defence and interior, parliament, judiciary) as well as non-statutory civil society groups, since these play a role in the democratic governance of the security sector.56 By adopting a security perspective in his definition of the sector, Williams maintains that it provides a focus within which the complex issue of SSR can be managed.57 Though he rightly argues for inclusion of indigenous military organisations such as civil defence forces, his definition is still limited to those actors that are relevant for the provision of security at a national level. A definition that not only considers the, let’s call it, ‘national-level’ security sector but also encompasses the local or 24 Judy Smith-Höhn community-level security sector would allow for the inclusion of legitimised organisations or groups of actors on the community level and would surely be more appropriate to an analysis of SSR strategies, particularly in a postconflict environment. The development perspectives prove to be more inclusive. However, there is the danger of incorporating far too many actors into the analysis. In sum, notions on what comprises the security sector have undergone radical changes since the end of the Cold War era. Whereas earlier definitions focused on the core security actors only, i.e. those state bodies with the legitimate mandate to use force, more recent approaches have come to consider a very broad range of actors. By and large, there appears to be a consensus that the security sector must be understood in the broader context of governance. For the purposes of this study and in recognition of the widespread application of the OECD DAC guidelines on security system reform and governance, the security sector shall be defined as comprising core security actors, civilian security management and oversight bodies, justice and law enforcement institutions, and non-statutory security forces. 2.4.2 Security sector reform, reconstruction and transformation: A jumble of terminologies or necessary distinctions? A review of the extensive bulk of academic and practitioner-oriented literature reveals that there remains a lack of consensus among researchers and development practitioners on the terminology to be used to describe this process. While some refer to security sector reform, others consider security system reform as more appropriate terminology. Still others speak of security sector reconstruction or transformation. This variety of terms is not necessarily a result of a fundamental difference in perceptions of the security sector per se, as the concepts – much like the definitions of the sector – often overlap. For the sake of uniformity, the term security sector reform will be used to describe such processes. Though it has become a popular area of research for academics, SSR by definition is primarily not an academic exercise but an issue of practical policy-making. By and large, external actors – for it is they who initiate these processes – generally have broad ideas concerning the instruments best suited to a particular reform context. These are often moulded by their perceptions of the respective arrangements in their own countries. Since these ideas differ among major international actors, there is, of course, a lack of policy coherence.58 This lack of consensus on the use of basic concepts On Security: Definitions and Theories 25 and terms notwithstanding, there appears to be an identifiable set of basic principles of security sector reform upon which all agree. Rather than delivering a comprehensive overview of the numerous works on the concept, this section will provide a summary of the literature categorised according to the various approaches adopted. First, it will put the discourse on SSR into historical context. Having shown how the approach to reform has evolved from focusing only on military-related programmes to regarding the process in the wider context of governance, i.e. including ‘civil elements’, it then outlines the various contexts within which SSR takes place, invariably influencing the focus of reform efforts. In a final step, SSR and reconstruction in the post-conflict context are addressed. History of a concept: the multiple origins of security sector reform. Not only is the terminology used to describe the (relatively new) concept of SSR contested, but the literature also reveals a discrepancy as regards its origins. Most authors contend that the concept was originally introduced by development donors.59 Others claim that it stems from two main areas, namely the development community – as an essential requirement for economic development and democratisation –and the field of civil-military relations (CMR), where, following the end of the Cold War and due to the developments in Central and Eastern Europe, analysts began to think more holistically about key aspects of the CMR debate.60 A third, and more accurate, approach is a division into three main categories, grouped according to particular sets of actors and concerns: the United Nations and its role in peace settlements and international administrations; the EU, NATO and the eastward extension of the European security community; and the development agencies and their extensive involvement in influencing policy within developing countries.61 The latter approach elucidates the diversity in priority-setting of SSR approaches. Naturally, the various origins of the concept invariably have an influence on how it is defined and which elements of the reform process receive more emphasis. Leading international agencies’ approaches to security sector reform. The United Nations can probably be considered as the leading international agency in terms of implementing peacebuilding strategies. In the first half of the 1990s global disarmament and regional peace initiatives brought about a number of peace accords that specified the course of action to be taken in reducing troop numbers and reintegrating former combatants into the formal economy.62 Disarmament, demobilisation and reintegration (DDR) programmes, which were usually developed alongside UN peacekeeping 26 Judy Smith-Höhn operations, were widely acknowledged as a key component of peacebuilding strategies. Later, however, in recognition of the fact that there appeared to be no positive correlation between a reduction of forces and military budgets and the success of development, there were calls for more holistic approaches than the DDR programmes undertaken by UN missions.63 One mechanism put in place to avoid cases where underdeveloped SSR programmes could contribute to increased political instability was the UN Civilian Policing Department (UN CIVPOL), established in 1992 to support the reform of police forces in countries emerging from civil war.64 In a recent report on SSR, UN Secretary-General Ban Ki-Moon notes that, despite extensive experience in assisting countries to re-establish security, support for SSR has largely been provided on an ad hoc basis.65 Hence the United Nations lacks a system-wide approach to delivering UN support in the countries where it is currently active.66 Notwithstanding such shortcomings, the United Nations recognises five common features of effective and accountable security sectors. i) ii) iii) iv) v) They have a legal and/or constitutional framework that provides for the legitimate and accountable use of force consistent with universally accepted human rights norms and standards. An institutionalised system of governance and management is in place: these oversight mechanisms can include systems of financial management as well as the protection of human rights. They have the capacity – personnel, equipment and resources – to provide effective security. Mechanisms are in place to facilitate interaction among the various security actors. There is a ‘culture of service’ that promotes unity, integrity, discipline, impartiality and respect for human rights among security actors and affects the manner in which these actors carry out their duties.67 DFID identifies seven key entry points for supporting SSR: building public awareness and engagement; building strategic planning capacity; strengthening legal and constitutional frameworks; strengthening civil oversight mechanisms; strengthening financial management systems; facilitating war-to-peace transitions; and improving human resource management.68 It acknowledges that entry points vary depending on the local context for reform and notes that, in countries emerging from war, ‘root-and- On Security: Definitions and Theories 27 branch’ reforms are needed that involve broad restructuring of the security sector and the transformation of mechanisms for civil control.69 The World Bank’s SSR agenda includes several key elements: promoting judicial independence through improved appointment, financing judicial administration; training judges and court personnel; strengthening good governance (including finding corruption); building capacity in public agencies, supporting bar associations and legal education; and supporting civil society organisations.70 In the case of Sierra Leone, for example, the Bank provided technical, financial and capacity-building support. Financial support was provided mainly through a multi-donor trust fund (of US$31.5 million) and direct financing through emergency recovery credits, budget support and a postconflict fund grant.71 The Bank provided technical advice to both the DDR programme and community-driven recovery efforts. Capacity-building initiatives supported the government in taking a lead in DDR efforts.72 The OECD may have extended its definition of what comprises the ‘security system’ (see 2.3.1), but its approach to its reform has remained unchanged. Accordingly, security system reform still describes the transformation of the ‘security system’ – which includes all the actors, their roles, responsibilities and actions [collaborating] in a manner that is more consistent with democratic norms and sound principles of good governance, and thus contributes to a well-functioning security framework.73 Figure 2.1 on page 28 illustrates how the various areas of SSR are interrelated. More specifically, according to the OECD SSR involves the achievement of four overarching objectives: i) Establishment of effective governance, oversight and accountability in the security system, ii) Improved delivery of security and justice services, iii) Development of local leadership and ownership of the reform process and iv) Sustainability of justice and security service delivery.74 The OECD DAC uses the term to describe a policy agenda, making an effort to distinguish it from security sector reform policies, which, according to the OECD DAC, concern only the armed forces or the system of actors working on security-related issues.75 A review of the work done in this field reveals Judy Smith-Höhn 28 Figure 2.1 SSR and related activities Source: OECD (2005) Security System Reform and Governance, p. 20. general recognition of the wider security agenda; hence this distinction seems unnecessary, particularly in light of the fact that the authors themselves use the terms interchangeably. More important in this context is not the label, but the content of what comprises security sector reform. In their handbook, Ball and Fayemi advocate the use of the alternative term of transformation. The authors argue that security sector reform represents [p]iecemeal, stand-alone changes to the security sector, often limited to doctrinal changes, operational effectiveness and cost-cutting drives, occurring as by-products of other state reform initiatives and often without ‘buy-in’ from critical stakeholders and thus subject to reversal depending on power structures of the state.76 According to Ball and Fayemi, the more holistic approach of security sector transformation is more capable of affecting the power relations within the sector that may inhibit change. This transformation occurs through civil/constitutional control to transform institutional culture, promot[ing] professionalism, improv[ing] resource utilisation and operational effectiveness… of the security forces, better policy management… of civil authorities, in tandem with accountability and respect for human rights and international law and involving inputs from a wide range of stakeholders and role-players.77 On Security: Definitions and Theories 29 As in the case of the OECD DAC, this distinction of terminology does not appear to be necessary, particularly since other approaches also adopt a more holistic agenda. Furthermore, the terminology is unsuitable for the postconflict environment, where a security sector, as such, needs to be built from scratch. The term construction or reconstruction would certainly be more appropriate in the latter case. Clingendael et al. use the terms transformation and reform interchangeably. They describe ‘security sector reform’ as the ‘transformation of security institutions so that they play an effective, legitimate and democratically accountable role in providing external and internal security for their citizens’.78 The objectives of this process are ‘strengthening civilian control and oversight of the security sector; professionalisation of the security forces; demilitarisation and peacebuilding; and strengthening the rule of law’. Of the seven yearbooks published by DCAF since 2003,79 the second volume on the reform and reconstruction of the security sector contains contributions from both academics and practitioners – on the conceptual bases and practical realities, respectively – and provides a sound overview of the complex dynamics of SSR in various regions of the world, focusing particularly on challenges faced in post-conflict environments.80 Since its emergence in the late 1990s, the SSR concept has come to include a plethora of issues and activities concerning the reform of those segments of the public sector assigned with providing both external and internal security. Though there is general consensus that no ‘one-size-fitsall’ model for effective SSR exists, Hänggi provides a useful differentiation into three broad SSR contexts – each reflecting a different rationale for reform – which in turn allows isolation of those SSR theories that are relevant to this thesis:81 the developmental context; the post-authoritarian context; and the post-conflict context.82 Accordingly, SSR can refer to strategies adopted: i) ii) iii) by international development donors as a means to improve the efficiency of development assistance in post-authoritarian states as a tool to enable the practical coordination and conceptual integration of defence and internal security reforms in failed or post-conflict states as an instrument to facilitate postconflict reconstruction.83 30 Judy Smith-Höhn Though the context within which SSR occurs is different in each case, they do have one common feature, namely that external actors are involved to a greater or lesser extent. Though they may not always be the primary actors in reform, they are certainly always the initiators.84 Moreover, regardless of the context, two normative elements are key to SSR efforts: the need for inexpensive security bodies capable of providing security, and the development of effective oversight mechanisms consistent with democratic norms. The post-conflict context is faced with a third and certainly urgent objective of addressing the legacies of the past conflict. DDR of former combatants, issues of transitional justice and proliferation of small arms and light weapons fall under this set of priorities.85 Post-conflict security sector reform. More interesting for this study than the differentiation between security sector reform and transformation (see above) is the differentiation between SSR and security sector (re)construction, as the latter is often used in a post-conflict context. The case studies in this book have both emerged from civil war in the past three to five years. David Law differentiates between post-conflict settings where a security sector has to be reconstructed and one where it needs to be built from scratch, appropriately applying the term security sector (re)construction.86 In his paper on the post-conflict security sector, Law identifies six distinctive features of the post-conflict environment, particularly from the perspective of external actors. Above all, the top priority in a post-conflict setting is the need to focus on security provision, a necessary precondition for the successful implementation of security sector programmes. Secondly, the restructuring of the security sector generally has to proceed before any legitimisation of the process through elections takes place. In non-conflict countries the process is reversed, with the reform effort gaining legitimacy through the preceding democratic elections. Thirdly, security sector programmes in post-conflict settings are often dominated by troop-contributing donor countries; accordingly, they have a strong military element and a weaker civilian one. This echoes the perpetual need to attend to the security situation as (re)construction efforts commence and recognises the ability of the military to develop programmes to create or reorganise armed forces. It may, however, also lead to the neglect of the (re)construction of other security sector actors in favour of defence capacitybuilding projects. On Security: Definitions and Theories 31 The often extensive responsibilities of the intervening forces is a fourth characteristic of the post-conflict environment. The context is generally one of little rule of law and no democracy – ruined state, social and civil infrastructure, compromised or uncooperative local elites – and there may also be fierce resistance to the (re)establishment of a state monopoly on security. A fifth feature of this specific environment is the need for donor agendas to focus on demanding issues rarely addressed in security sector programmes in non-conflict situations – e.g. DDR of combatants – and, consequently, such external actors require special skills and experience that may not be readily available. Finally, Law points to the disparate desired outcomes of security sector programmes in these different settings. Whereas, for example, the post-conflict setting demands a focus less on building national institutions and more on creating the conditions that make this possible, institutionbuilding issues tend to dominate donor programmes in the security sector of non-conflict environments.87 Notwithstanding these differences, Law rightly argues that donor countries ‘need to follow the same basic principles as they do in non-conflict environments’ when analysing what must be done in the security sector.88 One of these three basic principles is that any analysis of the sector should be based on a thorough understanding of its ‘characteristics, procedures and interactions’.89 In addition, they need to ensure that the security forces – be they domestic or foreign – can provide the required security. Finally, they need to ensure ‘that the population has confidence in the actions of security forces, for if it is lacking they will not be able to operate efficiently and will fail in their mission to provide security’.90 Law notes the expeditious engagement of local authorities in this process as crucial. Local ownership has been noted as vital by several authors working in the field. Although Law provides an insightful overview of the uniqueness of the post-conflict environment that external intervening actors face and highlights the basic principles that apply to any situation, his approach remains state-centred. It is the state security actors that are relevant on the local level. In narrowing his focus, Law joins most academic researchers in their analysis of the security sector, limiting themselves to studying merely the state actors in this field. However, despite his narrowed focus, Law’s deliberations can contribute to an overall framework for analysis of post-conflict security sector (re)construction. Generally speaking, the goals of SSR, and particularly so in the post-conflict setting, are very ambitious. To speak of an 32 Judy Smith-Höhn initial focus on activities in this field to include the restoration of the state’s monopoly on the legitimate use of force – as several authors do91 – is farfetched, if not an impossible goal. Should strategies not be more realistic and cater not to what sounds good, but what appears feasible in the respective context? The analysis of empirical data gathered in post-conflict settings can provide answers to this and other related questions. Such an analysis requires a guideline, which can easily be developed by summarising the key elements outlined by leading authors and organisations working in the field and commonly considered to be central to the success of SSR efforts. Perceptions matter: Azarya’s model of incorporation and disengagement. Before moving on to a guideline for analysis, it is necessary to elaborate on the question of the type of empirical data that can facilitate identification of the strengths and weaknesses of SSR strategies in post-conflict societies. Victor Azarya suggests an alternative to the state-centred approach to studying state-society relations in contemporary Africa by placing greater emphasis on societal responses to state actions.92 In other words, instead of explaining deficits in the state by focusing on how the state acts upon society, he proposes an analysis of how society behaves irrespective of or despite certain actions of the state as a means of explaining state failure or weakness. Azarya developed his concept in response to the overwhelmingly state-centred approaches towards explaining state weakness or failure, some of which attribute it to structural weaknesses, or to the colonial legacy of authoritarian rule, or even to international dependency of African states. He points out that the focus has generally been on how the state acts upon the society, with little attention being paid to how societies cope with the state. Azarya’s approach draws greater attention to the response of various groups and sectors within society to the state’s ability or inability to meet a particular need.93 When the focus is shifted from the state to the society, incorporation and disengagement are in a sense analogous to state consolidation and decline.94 According to Azarya, incorporation is ‘the process whereby large segments of the population associate with the state and take part in its activities in order to share its resources’.95 In other words, many individuals and groups perceive an expanded field of opportunities in having close ties with the state. Examples of incorporation include population migration from rural to urban areas, from remote regions to economic and communication centres, greater receptivity to mass media, an increase in the production of goods and services and so on.96 In cases of incorporation, government On Security: Definitions and Theories 33 employment is highly valued, with many seeking public administrative positions as another means of accumulating resources.97 The incentive to engage in such activities is always the same: the state is perceived as the centre of attraction and a superior channel for the distribution of resources, whether economic, symbolic or otherwise.98 Although society’s incorporation into the state does not necessarily create a better social environment – nor is it free from inter-group tensions or social unrest – it does point to the importance of society in enhancing the position of the state as a component part of that society. The state is defined as ‘an organization within the society where it coexists and interacts with other formal and informal organizations, from families to economic enterprises to religious organizations’.99 The state is distinct from other organisations within the society, in that it seeks predominance over them. Its goal is to establish binding rules vis-à-vis the other organisations’ activities, or, as Azarya puts it, ‘to authorize (i.e. to delegate power to) the other organizations to make such rules for themselves’.100 At the other end of the scale there is disengagement, which Azarya describes as the ‘tendency to withdraw from the state’ in response to its increasing instability and dwindling resources.101 As members of a society become increasingly sceptical of the competence and legitimacy of state actions, they respond by means of ‘popular evasion and dissimulation’, thus effectively undermining these actions.102 According to Azarya, the state’s incapacity is the end result of disengagement of social forces from the public sphere.103 Disengagement is a response to the vulnerabilities – both economic and social – that stem from the declining capacities of the state. Society members are focused on contriving ways to guard themselves from the state or somehow to manage if they are unable to disengage.104 Hence, only if people incorporate into the main state policies, inter alia into security sector reform, is there the likelihood of more broad-base support and success of reform efforts. Adapting Azarya’s thinking to the subject of this study may mean that donor-driven and state-centred programmes – including those undertaken in the SSR field – that fail to attain at least a modicum of participation will lead to the alienation of large sections of society, thus fostering disengagement instead of incorporation. A guideline for analysis. Having provided an impression of the plethora of views and theories on the SSR concept, it is apparent that there is hardly a need for another preconceived concept of SSR. What is missing, however, is empirical evidence that these concepts can be implemented and that they are 34 Judy Smith-Höhn effective at all. This is the starting point for this study, which, instead of attempting to redefine the security sector or develop a new approach to reform, builds on those issues that have been identified as key components of effective SSR through an analysis of unique empirical data gathered on the ground. The assessment of the successes and failures within a given postconflict setting will need an actor-centred approach. From the above elaborations, effective SSR should meet five requirements: 1. 2. 3. 4. 5. presence of security bodies – domestic or foreign – that are capable of providing security existence of effective oversight mechanisms sufficient consideration of the specific characteristics, procedures and interactions of the security sector by the international implementing agencies confidence of the population in the ability of the security forces to provide security adequate attention to the legacies of the past conflict in order to avoid a relapse into conflict. The guideline for the analysis of SSR efforts in the empirical section of this study is based on these five requirements. It goes without saying that some questions cannot be answered definitively at an early stage of reform efforts. Whether effective mechanisms have been put in place to oversee the security sector can hardly be judged while international actors are still actively involved in post-conflict peacebuilding efforts on the ground. Such is the case in Liberia and Sierra Leone, which some analysts consider to be UN protectorates.105 In the case of Sierra Leone, one may be able to draw conclusions regarding the presence of oversight mechanisms since the UN mission withdrew the last of its forces in December 2005, a few months prior to the field trip. Such conclusions, however, remain tentative due to the continued presence of the UK, through DFID and IMATT (International Military Advisory and Training Team), which will continue their development partnership with Sierra Leone presumably until 2012.106 Furthermore, one can hardly expect such mechanisms to be in place at all at an early stage of reform efforts. The analysis of the successes of SSR and the challenges that are faced will therefore focus more on the other four items on the checklist. Adapting Azarya’s understanding of states in society to the security sector, one means of assessing the efficacy of security bodies is by On Security: Definitions and Theories 35 determining whether the population integrates into or disengages from the state for the provision of security. Such reactions can be observed by drawing from local perceptions; in other words, by asking the population whether they believe security bodies exist that are capable of providing security or not. Determining whether the population decides to integrate or disengage with the state, in turn, reveals whether they have any confidence in the ability of security forces to provide security, or whether they resort to alternative channels for their security provision. The significance of the need to address the legacies of the past cannot be overstated, as the events and circumstances that led to the conflict in the first place must be dealt with in order to avoid a repetition. However, such issues are complex, and it is often difficult to determine where to begin. Moreover, as with the putting in place of effective oversight mechanisms, addressing and eradicating the causes of conflict are a timely endeavour. That they are crucial for sustainable peace is beyond doubt, but it is doubtful that these can be done adequately and speedily. Furthermore, one needs to question whether this issue can be addressed by the external implementing actor, or whether it is a matter for the local leadership, who have yet to perform their functions without the aid of the externals. This problem will be tackled in more detail in Chapter 4. To sum up, of the five elements outlined as the key criteria for assessing SSR, four can be addressed in this study: 1) 2) 3) 4) whether security actors exist and what their future potential is for providing security whether the context-specific characteristics, procedures and interactions of the security sector have been adequately considered by the international implementing agencies whether the legacies of the past have been adequately addressed so as to avoid a relapse into conflict whether the population has confidence in the ability of security actors to provide security, determined by means of local perceptions. These guidelines will shed light on the effectiveness of current SSR efforts and reveal the particular significance and hitherto lack of approaches that focus on the actors involved and address the challenges faced. Having outlined the basic guideline for the analysis of SSR, the exact methodology to be applied will be dealt with in the next chapter. Judy Smith-Höhn 36 Notes 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Krishna Kumar and Jeroen de Zeeuw (eds) (2006) Promoting Democracy in Postconflict Societies. Boulder, CO: Lynne Rienner, p. 2. Ibid. See for example Albrecht Schnabel and Hans-Georg Erhart (2005) ‘Post-conflict societies and the military: Challenges and problems of security sector reform’, in Albrecht Schnabel and Hans-Georg Erhart (eds) Security Sector Reform and Post-Conflict Peacebuilding. Tokyo: United Nations University Press, pp. 1–16. The authors do not clarify their understanding of the term ‘post-conflict’. In their work on SSR and postconflict peacebuilding with a particular focus on the military, the authors provide a definition of post-conflict peacebuilding as ‘a complex and multidimensional, genuinely political process of transformation from a state of war or violent conflict to one of stability and peace’ (emphasis added). As in many studies of societies with a recent experience of internal war, this lack of clarity leads the reader to assume that it is a selfexplanatory concept. Jeff Crisp (1999) ‘Post-conflict societies: The hidden agenda’, Crosslines, 34. Daniel Lambach (2007) ‘Oligopolies of violence in post-conflict societies’, GIGA Working Paper Series, Hamburg, p. 9. Ibid. See e.g. Gerd Junne and Willemijn Verkoren (eds) (2005) Postconflict Development: Meeting New Challenges. Boulder, CO: Lynne Rienner. Lambach, note 5 above. Ibid., p. 10. Though peace agreements had been signed and the United Nations had already begun sending in peacekeeping troops, the violence continued, and even drew in the external forces when 500 UN peacekeepers were abducted by the insurgent group known as the Revolutionary United Front, culminating in the intervention of British troops. Guatemala is a case in point, where death rates have, in fact, increased since the 1996 peace agreement was signed. See Lambach, note 5 above, p. 10. Small Arms Survey (2005) Small Arms Survey 2005: Weapons at War. Oxford: Oxford University Press, p. 289. See Lambach, note 5 above, p. 10. Ibid. Ibid. Ibid. Ibid., p. 11. Ibid. World Bank (2003) ‘Breaking the conflict trap: Civil war and development policy’, World Bank Policy Research Paper, Washington, DC, p. 83; Nicholas Sambanis (2001) ‘Do ethnic and nonethnic civil wars have the same causes? A theoretical and empirical inquiry (part I)’, Journal of Conflict Resolution, 45(3), p. 847. Quoted in Lambach, note 5 above, p. 5. Lambach, ibid., pp. 5ff. Heiner Hänggi (2003) ‘Making sense of security sector governance’, in Heiner Hänggi and Theodor H. Winkler (eds) Challenges of Security Sector Governance. Münster: LIT, p. 5; available at http://se2.dcaf.ch/serviceengine/FileContent?serviceID=DCAF&fileid= 0CD0E86D-F99B-3B67-0D75-2FFB3FB5CA3D&lng=en (accessed 13 June 2006). On Security: Definitions and Theories 37 22 23 24 25 26 27 28 29 30 31 32 33 34 35 See e.g. Human Security Centre (2005) Human Security Report 2005. War and Peace in the 21st Century, p. vii; available at www.humansecurityreport.info/content/view/28/63/ (accessed 6 January 2006); Jakkie Cilliers (2004) ‘Human security in Africa: A conceptual framework for review’, African Human Security Initiative; available at www.africanreview.org/docs/humsecjun04.pdf (accessed 20 September 2005); Dan Henk (2005) ‘Human security: Relevance and implications’, Parameters, 35(2), pp. 91ff. Kofi Annan (2000) ‘We the Peoples’: The Role of the United Nations in the 21st Century. New York: United Nations, p. 43; available at www.un.org/millennium/sg/report/full.htm (accessed 18 June 2006). Human Security Centre, note 22 above, p. vii. OECD (2001) DAC Guidelines: Helping Prevent Violent Conflict, p. 38; available at www.oecd.org/dataoecd/15/54/1886146.pdf (accessed 13 January 2007). Amitai Etzioni (1988) The Moral Dimension: Towards a New Economics. New York: Free Press, p. 27. Human Security Centre, note 22 above, p vii; UNDP (1994) Human Development Report 1994. New Dimensions of Human Security. New York: UNDP; available at http://hdr.undp.org/reports/global/1994/en/ (accessed 5 October 2007). J. ’Kayode Fayemi (2004) ‘Governing insecurity in post-conflict states: The case of Sierra Leone and Liberia’, in Alan Bryden and Heiner Hänggi (eds) Reform and Reconstruction of the Security Sector. Münster: LIT, p. 2; available at http://se2.dcaf.ch/serviceengine/ FileContent?serviceID=DCAF&fileid=B0EA7833-62AD-2215-99BF182ADF9F413A&lng=en (accessed 13 June 2006). Heiner Hänggi (2004) ‘Conceptualising security sector reform and reconstruction’, in Alan Bryden and Heiner Hänggi (eds) Reform and Reconstruction of the Security Sector. Münster: LIT, pp. 1–11; available at http://se2.dcaf.ch/serviceengine/FileContent? serviceID=DCAF&fileid=0CD0E86D-F99B-3B67-0D75-2FFB3FB5CA3D&lng=en (accessed 13 June 2006). For more on DCAF, visit its website at www.dcaf.ch/. In 2006, for example, DCAF was instrumental in the drafting of the OECD’s Handbook on Security System Reform. See OECD (2007) OECD DAC Handbook on Security System Reform: Supporting Security and Justice. Paris: OECD; available at www.oecd.org/ dataoecd/43/25/38406485.pdf (accessed 20 May 2007). It assisted the EU in shaping and implanting its SSR strategy in South-Eastern Europe along the guidelines of the EU’s overarching policy framework for security sector reform. DCAF also collaborated with the United Nations and its agencies, namely the UN Department of Peacekeeping Operations and the UNDP, in conducting a study on SSR in UN integrated missions. See DCAF (2006) DCAF Annual Report 2006. DCAF: Geneva. For further DCAF activities see the DCAF website. See BICC/DCAF Security Sector Governance and Conversion Studies Series. GTZ (2007) ‘GTZ profile’; available at www.gtz.de/en/unternehmen/1698.htm (accessed 30 June 2007). Herbert Wulf (2000) Security Sector Reform in Developing Countries. Bonn: BICC/GTZ, p. 15; available at www.bicc.de/publications/other/gtz_studien/sicherheitssektor/ securitysector.pdf (accessed 12 July 2007). Alan Bryden and Heiner Hänggi (eds) (2004) Reform and Reconstruction of the Security Sector. Münster: LIT; available at http://se2.dcaf.ch/serviceengine/FileContent?serviceID =DCAF&fileid=0CD0E86D-F99B-3B67-0D75-2FFB3FB5CA3D&lng=en (accessed 25 August 2006). 38 Judy Smith-Höhn 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 Informal DAC Task Force on Conflict, Peace and Development Co-operation (2000) Security Issues and Development Co-operation: A Conceptual Framework for Enhancing Policy Coherence. Paris: OECD, p. 8. Jane Chanaa (2002) Security Sector Reform: Issues, Challenges and Prospects, Adelphi Paper 344. New York: Oxford University Press for International Institute for Strategic Studies, p. 7. Hänggi, note 29 above, p. 6. Nicole Ball and J. ’Kayode Fayemi (2004) Security Sector Governance in Africa: A Handbook, p. vii, emphasis added; available at www.ssrnetwork.net/documents/GFNSSR-SecuritySectorGovernanceInAfrica-AHandbook.pdf (accessed 16 February 2007). Ibid., pp. 15ff. Ibid., p. vii. Dylan Hendrickson (1999) ‘A review of security sector reform’, Working Paper No. 1, Centre for Defence Studies, King’s College, University of London, p. 29; available at http://csdg.kcl.ac.uk/Publications/assets/PDF%20files/Working%20paper%20number%2 01.pdf (accessed 15 July 2007); Owen Greene (2003) ‘Security sector reform, conflict prevention and regional perspectives’, Journal of Security Sector Management, 1(1), p. 2; Clingendael, International Alert and Saferworld (2002) ‘Towards a better practice framework in security sector reform. Broadening the debate’, Occasional SSR Paper No. 1, p. 1; available at www.international-alert.org/publications/getdata.php?doctype= Pdf&id=90 (accessed 15 July 2007). OECD, note 25 above, p. 38. OECD (2005) Security System Reform and Governance, DAC Guidelines and Reference Series, pp. 20ff; available at www.oecd.org/dataoecd/8/39/31785288.pdf (accessed 13 January 2007); Andrzej Karkoszka and Dylan Hendrickson (2002) The Challenges of Security Sector Reform. Oxford: Oxford University Press, p. 179. UNDP (2002) ‘Democratizing security to prevent conflict and build peace’, in Human Development Report 2002: Deepening Democracy in a Fragmented World. Oxford: Oxford University Press, p. 87. Ibid. Rocklyn Williams (2000) ‘Africa and the challenges of security sector reform’, in Jakkie Cilliers and Annika Hilding-Norberg (eds) Building Stability in Africa: Challenges for the New Millennium. Pretoria: ISS. Ibid. For the UN definition of the security sector, see Ban Ki-Moon (2008) ‘Report of the Secretary-General on “Securing peace and development: The role of the United Nations in supporting security sector reform”’, UN Doc. A/62/659–S/2008/39, 28 January, p. 6; available at http://daccess-ods.un.org/access.nsf/Get?Open&DS=S/2008/39&Lang= E&Area=UNDOC (accessed 13 February 2008). DFID also considers these four sets of actors to comprise the security sector. See DFID (2002) ‘Understanding and supporting security sector reform’, p. 7; available at www.dfid.gov.uk/Pubs/files/ supportingsecurity.pdf (accessed 13 January 2007). OECD, note 44 above; OECD, note 31 above. OECD, note 44 above. See e.g. Schnabel and Erhart, note 3 above, p. 7; David M. Law (2006) ‘The post-conflict security sector’, DCAF Policy Paper 14, pp. 12–15; available at www.dcaf.ch/_docs/ PP14_post_conflictss.pdf (accessed 12 December 2006). Hänggi, note 29 above, pp. 5ff. On Security: Definitions and Theories 39 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 Ibid. See e.g. Williams, note 47 above. Hänggi, note 29 above, pp. 5ff; See e.g. OECD, note 44 above; Ball and Fayemi, note 39 above; Clingendael et al., note 43 above. Williams, note 48 above. Michael Brzoska and Andreas Heinemann-Grüder (2004) ‘Security sector reform and post-conflict reconstruction under international auspices’, in Alan Bryden and Heiner Hänggi (eds) Reform and Reconstruction of the Security Sector. Münster: LIT, p. 127; available at www.dcaf.ch/publications/bm_ssr_yearbook2004.cfm (accessed 13 June 2006). Ibid., p. 123; Law, note 52 above. Timothy Edmunds (2003) ‘Security sector reform: Concepts and implementation’, in Wilhelm N. Germann and Timothy Edmunds (eds) Towards Security Sector Reform in Post Cold War Europe: A Framework for Assessment. Baden Baden: Nomos, p. 11; available at www.dcaf.ch/publications/kms/details.cfm?q279=security+sector+reform &ord279=title&lng=en&id=20278&nav1=4 (accessed 15 June 2006). Chanaa, note 37 above, p. 13. Ibid., p. 16. Ibid., p. 17. Ibid. Ban Ki-Moon, note 49 above, p. 1. Ibid. Ibid., p. 6. DFID, note 49 above, p. 19. Ibid. Chanaa, note 37 above, p. 22. World Bank (2002) ‘Sierra Leone: Disarmament, demobilization and reintegration (DDR)’, Findings: Africa Region No. 81, World Bank, Washington, DC, p. 1. Ibid. Compare OECD, note 25 above, p. 38; OECD, note 44 above, p. 20; OECD, note 31 above. OECD, note 31 above, p. 21. OECD, note 44 above, pp. 29ff. Ball and Fayemi, note 39 above, p. vii. Ibid. Clingendael et al., note 42 above, pp. 1ff. Heiner Hänggi and Theodor H. Winkler (eds) (2003) Challenges of Security Sector Governance. Münster: LIT; Alan Bryden and Heiner Hänggi (eds) (2004) Reform and Reconstruction of the Security Sector. Münster: LIT; Alan Bryden and Heiner Hänggi (eds) (2005) Security Governance in Post-Conflict Peacebuilding. Münster: LIT; Alan Bryden and Marini Caparini (eds) (2006) Private Actors and Security Governance. Münster: LIT; David M. Law (ed.) (2007) Intergovernmental Organisations and Security Sector Reform. Münster: LIT; Timothy Donais (ed.) (2008) Local Ownership and Security Sector Reform. Münster: LIT; Hans Born and Albrecht Schnabel (eds) (2009) Security Sector Reform in Challenging Environments. Münster: LIT. Bryden and Hänggi (2004), ibid. There have been other attempts at categorising SSR. Edmunds, for example, drawing from analyses of SSR in post-authoritarian states, distinguishes between first- and second- 40 Judy Smith-Höhn 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 generation SSR. See Edmunds, note 60 above, pp. 16ff. First-generation SSR involves the ‘establishment of new institutions, structures and chains of responsibility for the security sector’, whereas second-generation SSR concerns the ‘consolidation of previous reforms, and the effective and efficient operation of institution and procedures at a sustainable cost for the state and society’. Edmunds notes that in Central and Eastern Europe, SSR occurred concurrent with other transformation processes – e.g. establishment of market economies and wider political reform – which governments and societies considered as more urgent priorities. The consequent lack of political commitment and resources for SSR has compelled governments to prioritise their activities in this area, leading them to emphasise issues of democratic control of the security sector rather than the effective operation of the sector itself. Edmunds’s design is particularly adapted to the postauthoritarian environment of Central and Eastern Europe; his approach has not been implemented in the countries of interest for this particular study, and is therefore not suited as a framework for the analysis of post-conflict SSR. Hänggi, note 29 above, p. 11. Ibid. Ibid., p. 15. Ibid., p. 9. Law, note 52 above, p. 1. Ibid., pp. 2ff. Ibid., p. 3. Ibid. Ibid. See for example Brzoska and Heinemann-Grüder, note 58 above, p. 128; Schnabel and Erhart, note 3 above. Victor Azarya (1988) ‘Reordering state-society relations: Incorporation and disengagement’, in Donald Rothchild and Naomi Chazan (eds) The Precarious Balance. State and Society in Africa. Boulder, CO: Westview Press, pp. 3–21. Ibid., p. 5. Ibid., p. 6. Ibid., p. 7 Ibid., p. 6. Ibid. Identity and legitimacy, for example, can also be understood as cultural and political resources, respectively. Azarya, note 92 above p. 10. Ibid. Ibid., p. 7. Ibid. Standard forms of disengagement include a retreat from the state-cash nexus to either a subsistence economy or alternative channels such as black markets and smuggling, i.e. channels less easily regulated by the state. Society members fail to comply with stateenacted laws and decrees as the judiciary system forfeits its credibility. Azarya, note 92 above, pp. 7ff. Activities aimed at controlling the state and modifying its actions according to an alternative set of values and interests are not acts of disengagement: ‘disengagement does not include active opposition to the regime with the objective to replace either the rulers and/or change government policies’. On Security: Definitions and Theories 41 105 106 See e.g. Wolf-Christian Paes (2005) ‘Eyewitness: The challenges of disarmament, demobilization and reintegration in Liberia’, International Peacekeeping, 12(2), p. 97. DFID (2004) ‘Poverty reduction agreement between the Government of the United Kingdom of Great Britain and Northern Ireland and the Government of the Republic of Sierra Leone’, p. 2; available at http://www2.dfid.gov.uk/search/proxy/cs.html?url= http%3A//www.dfid.gov.uk/pubs/files/sierraleone-prf-arrangement.pdf&qt=hmg&col=&n =172 (accessed 11 April 2007). Chapter 3 Methodology To identify the potential challenges to SSR in post-conflict societies according to the guidelines developed in Chapter 2, two post-conflict societies – Liberia and Sierra Leone – have been selected, and the progress made thus far in reforming their respective security sectors will be tracked by way of a multi-method research design of comparative case studies combining both quantitative and qualitative tools for data gathering. The research design is comparative and case-oriented. Unlike other methods of enquiry – such as the experimental method and statistical analysis – the comparative method is uniquely suited for analysing complex causality, i.e. the fact that a particular social phenomenon is probably affected by several economic, political, cultural and/or socio-economic factors.1 Comparative analysis is qualitative; it considers cases as wholes and compares one with another. In other words, unlike a quantitative statistical or variable-centred analysis, the comparative method can simultaneously consider a broad range of factors within a particular historical context which varies over time.2 The analysis of the efficacy of SSR in post-conflict societies, which depends on the range of factors developed in section 2.4.2, necessitates precisely such an approach. Moreover, comparative analysis facilitates an understanding of the relationship between and among these various influencing factors.3 Case study analysis pays particular attention to the historical context; a single-unit case can be divided into two separate time periods, facilitating an understanding of the change in the phenomenon under scrutiny between these two periods. This is commonly termed within-case or within-unit comparison. Furthermore, and in keeping with Sartori’s assertion that ‘he who knows one country only knows none’,4 comparison across cases (or units of analysis) enables a better understanding of individual cases. To gain an exhaustive understanding of any phenomenon, one would ideally consider all instances where it manifests and, if applicable, also those instances where it does not occur. In this case, one would want to analyse all post-conflict societies that have attempted to reform their security sectors, as well as those post-conflict societies where there have been no such efforts. The latter can be ruled out, since all post-conflict societies have in some way 44 Judy Smith-Höhn or another attempted to implement this process. Further, such an allencompassing analysis is unfeasible for two reasons. Firstly, there is the dichotomy of scope versus depth, since research designs are regularly confronted with a choice between knowing less about more and knowing more about less.5 On the one extreme lies the single case study approach, which has the benefit of depth of analysis to the detriment of the scope of causal arguments. Cross-unit analyses are situated on the other extreme; these provide scope – boundedness of inferences – at the cost of thinness. The research questions of this particular study necessitate an in-depth analysis of SSR efforts, yet at the same time the study requires a degree of universal validity. By opting for a comparative approach studying two cases, the research design positions itself between the two extremes. The limitation to two cases (units of analysis) enables the in-depth analysis that is the virtue of the case study method. The comparison of the units of analysis, on the other hand, allows more broad-based inferences as regards the efficacy of SSR strategies. A second viable argument against an all-encompassing sample of units of analysis (i.e. all post-conflict societies) is of a more pragmatic nature. Such a comprehensive study would require a significant amount of resources, both financial and human, and more than what was available for the study at hand. It would have required the employment of more researchers to collect data during the same period of time, as this concurrence is needed for such data to remain comparable. In the past there has been a good deal of disagreement over appropriate social science research methods, with the basic divide drawn between qualitative case study methods on the one hand, and statistical and formal methods on the other. More often than not, these lines of contention arose because they affected opportunities for funding, teaching positions and publication outlets. It meant that scholars with essentially similar interests had grouped into separate communities along methodological lines. More recently, however, social science research methods are entering a new stage of development that is conducive to cross-method collaboration and multimethod studies.6 In addition to such collaboration, many scholars nowadays are trained in or at least exposed to more than one methodology, allowing easier interpretation between the various methods that are generally restricted by inherent epistemological limits.7 George and Bennett maintain that most cross-method collaborative work has occurred sequentially. Researchers have applied the methods in which they are most adept, but have also drawn on the findings of those using other methods. Cross-method collaboration Methodology 45 has thus not been duly appreciated because, until recently, it rarely involved the application of different methods for one publication.8 In recognition of the benefits of such collaboration, the research design adopted here will make use of both quantitative and qualitative tools for gathering data. In addition to semi-standardised interviews and focus group discussions (qualitative), this study incorporates a survey poll (quantitative). This chapter outlines the benefits of the comparative method as it is, or at least should be, applied in the field of political science. The next section elaborates on the comparative case study approach as the most appropriate research design for the analysis at hand. Having developed the framework of analysis – outlining the various cases to be compared – the chapter concludes with a description of the tools used for data gathering. 3.1 The comparative method The following will illustrate the suitability of the comparative case study method, as opposed to other methods (e.g. statistical, formal, experimental), in determining the efficacy of SSR strategies in post-conflict societies. 3.1.1 The case for comparing It is often held that comparisons can be implicit, or that the scientific approach is in itself intrinsically comparative.9 From this viewpoint it follows that the notion of an independent comparative method is redundant or synonymous with scientific method in the social sciences in general.10 The disadvantage of such viewpoints is that they fail to distinguish between the various methods applied in the social sciences. Of course, a comparison is made at one point or another regardless of the method used (statistical, experimental, etc.) and the comparative strategies applied by comparativists are not, in principle, different from the comparative strategies used by other political scientists, economists or sociologists. However, to claim that all methods are comparative reduces the distinctiveness of the application of each method in research. The true distinctive feature of comparative politics is its explicit and direct focus on the comparative method – as opposed to merely comparing.11 Some authors have attempted to narrow the definition of the comparative method as one that compares a limited number of justified cases and is applied in studies where other methods, namely the statistical and 46 Judy Smith-Höhn experimental, cannot be employed.12 Lijphart defines the comparative method as a qualitative comparison of a few cases, which stands in contrast to the statistical, and hence quantitative, method of comparing a large number of cases (‘cross-national studies’).13 Whatever the case may be, the choice to proceed with a comparison precipitates the need for methodogically sound reasoning.14 This method can be defined as the systematic comparison of cases, mostly employed to generate empirical generalisations and test hypotheses.15 More specifically, the comparative method seeks to investigate causality and attempts to isolate those factors that cause (independent variable) a particular outcome or phenomenon (dependent variable).16 By and large, there is a clear lack of consensus over the reasons for applying the comparative method. While some authors claim the key goal of comparison to be controlling, that is verifying or falsifying whether generalisations hold across the cases,17 others consider control to be one of three, four or even five functions of the comparative method.18 Lim summarises the three purposes of comparing as an answer to the general question as to why one compares, namely to control, to understand and to explain.19 The control function, as mentioned above, is applied as a means of checking whether statements about particular phenomena are valid by controlling for, or holding constant, certain variables.20 Though actual control variables are not used, different types of comparisons allow the researcher to consider a range of similarities or differences as control variables.21 This enables the researcher to dismiss a number of potentially significant factors and focus on those variables he/she considers to be most important.22 Landman and Nohlen label this the hypothesis-testing function23 and it is what Sartori called the ‘control purpose’ of a systematic comparison of two or more cases.24 Comparative checking enables the control of hypotheses by way of eliminating conflicting explanations about particular events, actors, structures, etc. as a means of developing more general theories.25 The second purpose of understanding applies to those comparativists who, instead of testing theories or hypotheses, apply theory in order to interpret the cases compared.26 Such exercises are based on the recognition that context matters; in other words, countries or other macrosocial units all have a unique story to tell and comparative analysis can enable a better understanding of these individual stories.27 Here the comparison serves not to verify or falsify larger arguments, but to facilitate in-depth understanding by observing what other cases reveal about a specific case or country that is Methodology 47 of primary interest. The comparison serves as a heuristic analysis of a particular case in comparison to others.28 Landman describes this function as contextual description in that it interprets a particular case by drawing from the various experiences with other cases through either quantification or qualification.29 A more pragmatic approach to comparison acknowledges the complexity of real-world cases and the difficulties associated with applying the control function. It therefore endeavours to build theoretical generalisations by collecting case-based knowledge; in other words, it wants to explain. Each case or each small-n comparison gives the comparativist another piece to work into a larger puzzle.30 Although the three purposes outlined above are indeed key reasons for comparison, Lim overlooks one other key function of comparison, namely that of prediction. Researchers may use the comparative method as a means to generate hypotheses from the units that have been studied. Such hypotheses enable prediction of the likely outcomes in countries not included in the original comparison, or outcomes in the future given the presence of certain factors.31 Comparisons can therefore serve to generate or modify hypotheses.32 Such hypotheses can and should again be tested in other comparative studies, with the research process having thus gone full circle. Table 3.1 on page 48 summarises the four general purposes outlined above, providing an overview of the emphasis placed by the various authors. Although the various functions have been listed separately here, it is important to note that these functions are not disjunctive and comparisons can indeed have a multifunctional purpose.33 Returning to the research focus of this study, what purpose does the comparative method serve? The key objective is to determine the efficacy of SSR strategies in post-conflict societies by submitting the theory of SSR to empirical scrutiny. The advantages of the comparative method are evident. A single case study would perhaps enable a description of SSR strategies in a given society. Yet the interest is not in a single case, but in post-conflict societies in general, therefore ruling out the function of understanding. The study also does not seek to control, that is verify or falsify, the hypotheses developed by scholars in the field. For this study, comparison serves the two closely related purposes of explanation and prediction. Firstly, a comparison of cases enables the inductive analysis of SSR, which then necessitates strengthening or modification of the existing theory. Secondly, this facilitates the development of hypotheses relating to SSR, possibly in other post-conflict societies. Indeed, predictions about other cases can hardly be Judy Smith-Höhn 48 Table 3.1 Four purposes of comparing General purpose Comparing to control Comparing to understand comparative checking ab hypothesistesting cd interpretation b analytical contextual induction b c description heuristic analysis development of hypotheses cd Logic or approach to comparative analysis Researcher uses a range of cases as a way to ‘test’ (verify or falsify) a specific claim, hypothesis or theory Researcher is primarily interested in a single case and uses different cases or general theories as a way to learn more about the case being studied Researcher uses cases as a way to build stronger theoretical explanations Cases are used in a ‘step-by-step’ manner, with each case contributing to the development of a general theory Researcher uses a range of cases as a way to ‘generate’, (develop) a specific claim, hypothesis or theory Example Begin with a Begin with a claim, then ‘test’ case (and issue), the claim then use existing theories and/or other cases to understand the case better Begin with a general theory, then use cases to develop or strengthen (modify) the theory Begin with a number of cases, then identify similarities or differences between these in order to make claims Basic strategy or purpose Comparing to explain Comparing to predict d a = Sartori 1994; b = Lim 2006; c = Landman 2003; d = Nohlen 2004 Own modification Source: Adapted from Timothy C. Lim (2006) Doing Comparative Politics. An Introduction to Approaches and Issues. Boulder, CO: Lynn Rienner, p. 23 made through the analysis of a single case study. For how is one to know whether a hypothesis generated from the observation of a particular characteristic is not precisely that, namely characteristic for this case only? Methodology 3.1.2 49 What is comparable? Identifying cases Having identified the various functions of the comparative method, the question is now how to determine what, exactly, is comparable. Authors on comparative methods often respond to the common argument that one cannot compare apples and oranges (or apples and pears) with a counterargument, namely: how can one know unless one has compared them?34 Let us, for a moment, take this argument literally. To a certain extent, apples and oranges are indeed comparable: both are fruit, both grow on trees and both are edible. In other words, any two or more entities are comparable with respect to the properties they share, and it follows that they are incomparable with respect to the properties they do not share. Apples and oranges would therefore be incomparable in their colours. If entities have nothing in common, a comparison is of no interest. Logically, if two or more entities share all the same characteristics, then they are a same entity. In sum, a fruitful comparison is both the assimilation and the differentiation of two or more entities to a point. This part of the process is crucial, as it affects the quality and extent of the findings. The comparativist does not choose countries according to personal preference; instead s/he must be guided by relevant criteria.35 Classification provides so-called ‘data containers’ into which empirical evidence can be organised.36 A specific number of items are ordered into categories that are mutually exclusive and altogether comprehensive.37 Classification does not imply real sameness, as Sartori puts it, but groups entities together according to the similarities they share with respect to the selected criteria, and indicates that these entities are more similar among themselves than to the entities that fall into other classes.38 Elsewhere, classification has been categorised as one of the key functions of comparison.39 However, this implies that a researcher may or may not choose to classify the units s/he intends to compare and, as Sartori points out, classification ought to precede any analysis. When classifying cases, three factors can lead the analyst down the wrong path and towards false concepts: parochialism, misclassification and conceptual stretching.40 Parochialism refers to single-country studies conducted in vacuo that disregard existing categories established by either general theories or comparative frameworks of analysis. Such studies tend to invent ad hoc, self-tailored terminology.41 Sundquist’s article on coalition government in the United States is a good illustration of this approach.42 The term ‘coalition 50 Judy Smith-Höhn government’ denotes parliamentary systems – in other words, not Americantype presidential systems – in which governments are elected into office and supported by parliaments, and are not single-party governments. Sundquist’s parochial definition of coalition governments does not include any of these characteristics, his mislabelling thus interfering with what is generally known of coalition governments.43 The present study comprises two country cases and the terminology used is based on existing definitions and established categories. Although various analysts may highlight different elements of security sector reform, for example, there is a general consensus as regards the basic understanding of this process to describe the creation of a secure environment that is conducive to development, poverty reduction and the growth of democratic states and institutions based on the rule of law. The second source of false classification is that of misclassification, that is the creation of pseudo-classes. If we subsume a number of cases into one class that is not derived from a single criterion, then we cannot, for example, test hypotheses based on this classification.44 The study at hand runs no risk of misclassifying cases, since the term post-conflict – albeit delivering a somewhat inaccurate description of such societies, as elaborated in Chapter 2 – denotes a particular type of society, namely one that has emerged from a narrative of conflict, and where the violence that is perpetrated by the former warring factions is no longer carried out in terms of the central narratives of the previous conflict. The third reason for erroneous classification results from conceptual stretching. In other words, when the definitions of particular terms are too broad, hypotheses cannot be tested. For example, if all societies are considered to be pluralistic in some sense, then one can no longer claim that ‘pluralism falls and stands with democracy’.45 This study has taken particular care not to define terms too broadly. Human security is defined in a narrow sense to avoid this problem. Moreover, it was imperative that the understanding of security be operationalised in such a manner as to make it practical to gather empirical data in the field. 3.1.3 Dimensions of comparison: Developing a research design Taking care not to make any of the above mistakes, the comparative scientific enquiry can be conducted according to one or more of four aspects: object, context, space and time.46 The most important criteria are the number of cases and the ratio between context variables (parameters or constants) and operative variables, i.e. dependent and independent variables to be examined. Some authors define operative variables as ‘those allowed to vary Methodology 51 in order to assess their influence upon the dependent variable(s)’.47 However, this causes ambiguity as regards the distinction between dependent variables and context variables, so this distinction will not be made here. Unlike a scientific experiment, the outlying conditions cannot be kept at a constant, and one can only assume that the context variables remain constant to a certain degree. Depending on its aims and objectives, a comparative study of two cases can prove more fruitful than a large-N quantified statistical study. Both types of comparison have their merits and demerits. An increase in the number of cases generally brings a decrease in the qualitative elements of the analysis, i.e. a hypothesis becomes more universally applicable at the expense of losing information. This dichotomy is inherent in the comparative method.48 The four aspects outlined by Nohlen are not mutually exclusive, as they can often overlap and each must be taken into account when developing the appropriate method of comparison: 1) 2) 3) 4) object: can be either large entities (e.g. societies, political systems) or segments (e.g. parties, unions) context: can be either homogeneous or heterogeneous space: is differentiable according to four features – state, intrastate, suprastate and global time: can be either diachronic, synchronic or diachronic-binary.49 A diachronic comparison is one that is longitudinal and occurs over a certain time frame. The number of cases is usually small, mostly applied within one country, and the contextual variables are relatively constant, with the advantage that historical factors can be taken into consideration. The availability of data can pose a problem here. Synchronic comparisons are horizontal and refer to simultaneous occurrences. The comparison can be between (a large number of) countries, or even between regions within one country, but the contextual variables are more difficult to control and the comparison is made mostly between homogeneous cases or by means of an area approach. The diachronic-binary comparison is a hybrid of the former two types of time-sensitive comparisons, employed for example to compare industrialisation and import substitution as they occurred at different times in different countries. The number of cases varies, but is mostly binary, and the context variables are homogeneous. The case studies of Liberia and Sierra Leone are classified as follows. First and foremost, both are post-conflict societies as defined in section 2.1. 52 Judy Smith-Höhn Secondly, Liberia and Sierra Leone are comparable since they share certain contextual similarities,50 but are at different stages of reform. The comparison shall not be made according to the object or space aspects, and while it considers the context aspect to be more or less homogeneous, this research will focus on the time aspect. Furthermore, post-conflict Liberia and Sierra Leone present an opportunity to test the potential for institutional design. The deinstitutionalisation and informalisation of the security sector before, during and immediately after the conflict provide ‘that space for altering the relations of power within the sector in the direction of civil/constitutional control’.51 In other words, the question as to how the security sector should be reconstructed and what issues need to be considered can be addressed with a view to actually being relevant and practicable. Let us now turn to the well-established research designs employed in comparative politics. As noted above, there is no ideal research design for conducting analyses in political science, though some authors claim otherwise.52 Two general comparative strategies in the field are a mostsimilar systems and a most-different systems design (MSSD and MDSD, respectively). Perhaps the commonest strategies of comparative analysis today, these designs are both based on the logic of John Stuart Mill’s classification in his seminal work A System of Logic, where he differentiated between a ‘method of agreement’ and a ‘method of difference’.53 The former compares different instances in which a phenomenon occurs, the latter compares instances that are similar in other respects, but that differ in outcome.54 The terminology used to differentiate between the various research designs has often led to confusion, but most can more or less be categorised under Mill’s two-way classification.55 The MSSD – method of difference – comprises systems that are as similar as possible in as many features as possible in order to single out differences. In other words, the parameters – i.e. context conditions – are similar, but the (dependent and independent) operative variables differ.56 The assumption is that the factors that are common to these homogeneous cases can be ruled out as an explanation for their differences. The MSSD can be considered synonymous with Mill’s method of difference in that it aims to identify difference and compare cases ‘in which the phenomenon does occur, with instances in other respects similar in which it does not’.57 The MSSD allows the researcher to explain a particular social or political phenomenon (the dependent variable) by identifying a significant dissimilarity between two or more systems, which can then be submitted as the causal factor or key independent variable.58 In other words, the Methodology 53 independent variable is causal and affects the dependent variable.59 The key limitation of the MSSD is that no comparison of two or more cases can provide exhaustive proof of a particular causal relationship. Firstly, a comparison of a limited number of cases cannot provide a strong enough empirical basis for verifying wide-ranging claims; secondly, no matter how similar two or more systems appear to be, there are bound to be numerous potentially relevant differences and it would not be possible, in strict methodological terms, to determine precisely which differences possess causal significance and which do not.60 The MDSD – akin to Mill’s ‘method of agreement’ – compares systems that differ as much as possible and yet do not differ on the phenomenon under investigation. The operative variables are similar, while the context varies. Hence, the isolation of another similar variable would serve as a possible explanation for this similarity.61 There are two major and interrelated shortcomings of the MDSD. Firstly, the findings on causal relationships in such a research design are not indisputable. In other words, discovering similarity in one variable across most-different systems does not qualify it as the root cause of a particular phenomenon, as there can be another unidentified factor, i.e. variable, influencing the relationship.62 In fact, and this is the second shortcoming, the MDSD cannot adequately deal with multiple causation.63 More precisely, one cannot assume that a particular phenomenon is the product of one and only one cause.64 It is not the outlined shortcomings that make the MSSD and MDSD unsuitable for the study at hand, however, for any comparative research design has its limitations. Neither design enables an in-depth analysis of a particular phenomenon,65 which is a key objective of this particular study. While both may attempt to control for a particular phenomenon by way of testing hypotheses, the main interest here is not to identify pre-determined causal relationships between SSR and the factors that affect its efficacy, but to illuminate the SSR strategies in each case and then identify other possible factors that play a role in effectively reforming the security sector in postconflict societies. In other words, the goal is to explain and predict. Moreover, the most-similar and most-different systems designs do not allow a precise differentiation between the various ‘cases’ that need to be analysed, as will become more apparent below (see section 3.1.4). The case study method provides a better basis for comparison,66 if defined correctly and properly distinguished from other methods. A brief digression is called for before continuing with a comprehensive outline of the case study method. There is potential for confusion among the terms ‘comparative method’, ‘case study method’ and 54 Judy Smith-Höhn ‘qualitative method’. Some authors consider the comparative and case study methods to be two distinct methods where the former involves the comparison of a number of cases, whereas the latter only focuses on the internal examination of distinct cases.67 This assumption is based on the notion that, when employed, the case study method is usually a single-case investigation. Similarly, earlier definitions of this method characterised it as a small-n study, in contrast to a large-N, statistical (quantitative) study.68 However, given the growing consensus that the most effective means of drawing inferences from case studies is by using a combination of withincase analyses and cross-case comparisons in a single study or research programme, case study methods can indeed be comparative, as they can include both within-case analyses of single cases and comparisons of a small number of cases.69 The elaborations below illustrate the suitability of the comparative case study method for the analysis of SSR strategies in postconflict societies. 3.1.4 The case study method as a method of comparison The term case and the numerous terms linked to the idea of case analysis, despite widespread usage, are ill-defined in social science.70 As a research method, there is no fixed definition as to what exactly a case study is. Though it is widely practised, its status as a method in the discipline of political science is ambiguous: a study which focuses on one example of a broader phenomenon can only be considered a ‘mere’ case study, hence methodologists’ circumspect view of this method.71 This is not to say that no attempts have been made to define the case study as a research approach. In fact, the literature abounds with such definitions. The reference to a work as a case study might indicate that the method applied is qualitative and small-n;72 that the research is ethnographic, clinical or participant-observation;73 that the research is characterised by process tracing;74 that the research analyses a single phenomenon, instance or example – this is the most common usage;75 or that the research examines the properties of a single case.76 The first three items listed above appear erroneous, as each implies a significant ‘shift in meaning relative to established usage’.77 To substitute case study for qualitative, ethnographic or process-tracing research is to limit the case study to one of its subtypes, for each definition describes a certain kind of case study, and not the general phenomenon. The fourth option may not be false, but it is certainly ambiguous. The terms phenomenon, instance or example are simply too broad to define the limits of what a case study is and is not. Finally, equating Methodology 55 a case study with the study of a single case (n = 1) is, as Gerring puts it, ‘simply wrong’.78 Gerring proposes an alternative definition, effectively capturing the uniqueness of the case study as a research method without becoming so specific as to limit its application. He defines the case study as ‘an intensive study of a single unit for the purpose of understanding a larger class of (similar) units’.79 A unit represents a spatially confined phenomenon, such as a nation-state, a political party, an election and so on, observed at a single point in time or over a delimited period of time. Chronological boundaries may not always be explicit, but they are at least implicit. Gerring provides a particularly useful definition of the case study by further clarifying the relationship of the study to other terms in this semantic field. The author defines a population as consisting of a sample (studied cases) as well as unstudied cases. A sample comprises several units, and each unit is analysed at distinct points in time, forming cases. A case contains several relevant dimensions (variables), each of which is developed from an observation or observations.80 This clarification facilitates more accuracy in designing a research approach because the precise definitions of all the terms can be modified depending on the proposition and corresponding research design. In other words, depending on the research focus, a country may be a case, a unit, a population or a case study – thus effectively reducing the risk of methodological confusion. Unlike the MSSD or MDSD, the comparative case study research design permits a clear delineation of the various levels of comparison, particularly on the timeline, as it enables both the within-unit and cross-unit analyses that are required for this study, as will be described in more detail below. The preference for the comparative case study research design, however, does not suggest that context conditions are irrelevant for selecting the units of analysis. On the contrary, analysis of the context is crucial, and the logic that underlies such a two-unit comparison is exactly the same as that of the MSSD or MDSD.81 Having specified what a case study is, Gerring moves on to elaborate on the ways in which case study research can be conducted. To do so, one must examine the various ways in which a covariational relationship can be established within any research design; in other words, observing how X causes Y. Of the seven ways in which covariation can be observed, the single case study research design applies to the first three: i) ii) iii) in a single unit diachronically within a single unit synchronically within a single unit diachronically Judy Smith-Höhn 56 iv) v) vi) vii) across units synchronically across units synchronically and diachronically across and within units synchronically across and within units synchronically and diachronically82 The first type of case study examines variation in a single unit over time (i), the second type divides the primary unit into subunits that are examined synchronically for covariation, while the third type subjects the subunits to a diachronic covariational analysis.83 Of the other cross-unit research designs (iv–vii), the comparative case study is closest to the final method which utilises all forms of covariation in a single research design that is ‘comparative-historical’. Here, a small number of units are subjected to in-depth analysis. Such an intensive study is tantamount to a combined series of case studies with cross-unit analysis.84 Applying Gerring’s definition and bearing in mind the aspects to be considered when using the comparative method as outlined by Nohlen,85 the research design of this study can be visualised as in Figure 3.1. Figure 3.1 The case study method – A semantic clarification POPULATION Post-Conflict Societies CASE 1 before 2003 SAMPLE UNSTUDIED Sierra Leone and Liberia Other Post-Conflict Societies UNIT UNIT Liberia Sierra Leone CASE 2 2003-06 CASE 3 before 2001 CASE 4 2001-05 CASE 5 2006 Methodology 57 This framework reveals two units of analysis that together comprise five cases. This is a crucial distinction, as the division into five time-bound cases enables a more clear-cut comparison within and across cases. The inferences drawn from the comparison are thereby more accurate. Basically, there are three types of comparisons: diachronic, diachronic-binary and synchronic. The diachronic comparison occurs on the timeline between the past and the present situation in each country (cases 1 and 2 for Liberia and cases 3, 4 and 5 for Sierra Leone). This is based on the notion that, to determine the efficacy of the SSR efforts implemented in the two countries since the end of their respective civil wars, one must know the state the security sector – if intact at all – was in before such efforts were undertaken. This enables one to determine the relative progress that has been made in each country. Moreover, within-unit comparisons can potentially reduce the well-known inferential errors that may arise from employing cross-unit comparisons alone.86 The first level of comparison is therefore within-unit, i.e. a diachronic comparison: Liberia just before the end of the civil war in 2003 compared to the period 2003–2006 during the so-called peacebuilding phase, and Sierra Leone before the end of its civil war in 2001 in comparison to both the period 2001–2005 (the peacebuilding phase with heavy international involvement until the withdrawal of the last UN troops) and the period after the withdrawal of the UN peacekeeping force in 2006.87 These will be dealt with in sections 4.3.1 and 4.3.2, respectively. The diachronic-binary comparison is cross-unit, and refers to the comparison of both countries just as they emerged from violent conflict (the immediate post-conflict phases signified by cases 2 and 4). A cursory note on the diachronic-binary comparison of cases 2 and 4: it is necessarily brief, as sufficient data could not be gathered in Sierra Leone. The analysis can only be based on the interviews and one of the four focus group discussions that had looked at this time period, while the survey poll provides no information in this regard. This points to a flaw in the planning and implementation of the field research. The third and final level of comparison is synchronic and compares the current situation, i.e. the SSR efforts, in both countries (cases 2 and 5). The two cross-unit comparisons are dealt with in sections 4.3.3 and 4.3.4. Figure 3.2 illustrates the intended approach, with the arrows indicating the comparisons that will be undertaken within and between the cases. In their work on Case Studies and Theory Development in the Social Sciences,88 George and Bennett dedicate an entire chapter to the comparative Judy Smith-Höhn 58 case study research design.89 Although their definition of case study research90 covers only one of the subtypes of such research, their elaboration on the essential components of a comparison of case studies remains valid – it is both simple and straightforward.91 In fact, the statement applies to any comparison, namely that it requires structure and focus.92 The method should be structured, in that the same set of research questions are posed for each of the case studies to guide and standardise data collection, thus facilitating a systematic comparison and cumulation of the findings of the cases. It should also be focused, in that it considers a selected aspect of each of the cases or units. Figure 3.2 Comparative case studies – Levels of comparison Liberia Sierra Leone CASE 1 CASE 3 before 2003 before 2001 CASE 2 CASE 4 2003–2006 2001–2005 CASE 5 2005 3.2 Tools for comparison: A mixed-methods approach In general, despite most authors now calling for a mix of methods, their designs and elaborations on particular methods either do not permit any integration of methods or their narrow definitions of research strategies limit the potential for a mixed-methods design. George and Bennett are no Methodology 59 exception when defining the case study as a uniquely qualitative method. Though they call for the use of more than one method, their discussion provides no insights as to how such a multi-method approach should or could be structured.93 George and Bennett’s elaborations on the benefits of case studies and their distinction between case studies, statistical studies and formal modelling may obscure the fact that case studies can incorporate both qualitative and quantitative methods for data collection. As Yin noted in his later works, a case study should be defined as a research strategy: an empirical inquiry that investigates a phenomenon within its real-life context… The case study inquiry… relies on multiple sources of evidence, [benefiting] from the prior development of theoretical propositions.94 This all-encompassing method can include both single- and multiple-case studies and the data can include, and even be limited to, quantitative evidence. More specifically, Yin notes that case studies should not be confused with qualitative research, and indicates that they can be based on any mix of quantitative and qualitative evidence.95 Research on mixed methods is still evolving, and there is an ongoing discussion as to how to define such studies. Though most writers concede that a mixed-methods approach is one that includes both a qualitative and a quantitative sub-study, there is a lack of consensus on how the two substudies are related to each other. Basically, one can distinguish between mixed methods as the collection and analysis of two types of data (qualitative and quantitative) and mixed methods as the integration of two approaches to research (quantitative and qualitative). While the former focuses more closely on methods, the latter hones in on methodology. One thus needs to distinguish between studies that employ two types of data without serious integration, and those that integrate findings from qualitative and quantitative elements of a project. This study mixes methods by applying two types of data-collection procedures: a survey poll (quantitative), and focus group discussions and semi-standardised interviews (both qualitative). A comprehensive definition of mixed methods describes it as ‘research in which the investigator collects and analyzes data, integrates the findings, and draws inferences using both qualitative and quantitative approaches or methods in a single study or a program of inquiry’.96 A key element of this definition is integration. 60 Judy Smith-Höhn What, then, are the benefits of mixing methods for the research design developed above? In the field of political science, the research paradigm of mixed methods has proven to be a contentious issue.97 Undoubtedly, merging of quantitative and qualitative findings has the potential to offer insights that could otherwise not be gathered: by drawing from the two sets of findings, the researcher can consider whether they suggest interesting contrasts that require further investigation or whether they help to clarify each other.98 The challenge is to overcome the practical difficulties that may impede such an integration of methods. Bryman published a revealing article on this issue, noting that the development of mixed-methods research was being hindered by the tendency for qualitative and quantitative findings either not to be integrated at all or to be integrated only to a limited extent.99 The author elaborates on the various factors that bring about this lack of integration,100 and concludes by formulating a question that can be used to guide the researcher when conducting mixed-methods research: Has my understanding of my quantitative/qualitative findings been substantially enhanced by virtue of the fact that I also have qualitative/quantitative findings, and have I demonstrated that enrichment?101 Bryman also points to a significant deficit in mixed-methods practice, namely the lack of literature on how to present mixed methods in such a way that the qualitative and quantitative findings are genuinely integrated. Hence there are very few guidelines upon which to draw when writing up findings that employ a mixed-methods research design. Elliott, like Bryman and other leading authors in the field, laments the fact that the qualitative and quantitative elements within a research project are often still ‘compartmentalized into separate chapters’ and not integrated as they might be.102 To overcome this barrier, a fully integrated mixed-methods framework was developed for the current study. Instead of presenting the findings according to the datasets available, i.e. dividing the chapters according to the sets of qualitative and quantitative results drawn from the fieldwork, this study uses the checklist developed in Chapter 2 as a framework. The various comparative steps for the analysis are therefore guided by the four criteria for assessing SSR in post-conflict societies: 1) whether security actors exist and what their future potential is for providing security Methodology 2) 3) 4) 3.2.1 61 whether the context-specific characteristics, procedures and interactions of the security sector have been adequately considered by the international implementing agencies whether the legacies of the past have been adequately addressed so as to avoid a relapse into conflict whether the population has confidence in the ability of security actors to provide security – the existence of these actors can be determined by means of local perceptions. Integrating the data-gathering tools for analysis One can briefly elaborate on the benefits of the methods of data gathering employed, and underline the value of each method in providing answers to the questions posed in the checklist as well as their potential to complement each other. Field research was conducted in both countries for a period of three months in each country.103 Three methods were used for data gathering. i. ii. iii. A survey poll was carried out in three urban areas in both countries. Eight in-depth focus group discussions were conducted, zooming in on security perceptions and including issues such as security actors involved and the relationships between these actors. Approximately 60 focused, semi-structured interviews were conducted with elites and experts – local and international – with the information gathered from experts serving as the more ‘objective’ perception of the security arena in each country. Where available, non-conventional literature and other key documents (also known as grey literature, e.g. internal reports, government documents, newsletters, factsheets, crime statistics) were gathered to substantiate the ‘objective’ performance of security actors. 3.2.2 Survey poll The survey poll is a typical quantitative data-gathering tool which enables the generalisation of findings to the population from which the sample was drawn, with the samples generally being larger than that for qualitative research. A total of 700 respondents were interviewed in three urban areas in each country. In Liberia, the capital Monrovia (500 respondents) and two 62 Judy Smith-Höhn secondary cities, Tubmanburg and Buchanan (200 respondents each), were selected. In Sierra Leone, interviews were conducted in the capital, Freetown (500 respondents), and in Makeni and Koidu Town (200 respondents each). Due to practical considerations, fieldwork was limited to urban areas; thus the figures reflect the perceptions of both countries’ urban population. However, people who had migrated from the hinterland participated in the focus group discussions and interviews were also conducted with a number of key stakeholders from rural areas, thus furnishing some impressions of rural conditions. But given this inherent caveat, results are interpreted bearing in mind the urban bias of the research design. To ensure that the survey sample is representative, four suburbs were selected within each city which together were identified as representing the demographic composition of each city. In addition, interviewers applied a randomised technique for the selection of the interviewer teams, starting points and households. To guarantee a gender balance, every second respondent had to be a woman.104 The survey was intended to give an impression of the perceptions of general and personal security of urban citizens in each country, as well as revealing citizens’ assessment of different security actors as regards their role in public and personal security. The questionnaire included items on the socio-economic background of respondents, allowing a more differentiated analysis of security needs and perceptions in terms of various criteria, such as age, gender, ethnic group, religion and occupation. Data from the survey poll were interpreted using SPSS,105 a software package developed specifically for statistical data analysis.106 3.2.3 Focus group discussions Focus groups are best known for their use in the marketing sector as a means of determining consumer habits. In other areas, focus group discussions have been used as a tool for what is termed social marketing. Here, focus groups facilitate the evaluation of programmes dealing with, for instance, substance abuse or curricular reform, and identify potential targets of such programmes.107 Focus group discussions are defined as a research technique that collects data through group interaction on a topic determined by the researcher.108 They can thus be understood as a data-collecting tool that utilises the interaction in a group discussion as a source of data, with the researcher playing an active role in creating the group discussion for datagathering purposes.109 Methodology 63 Given this broad definition, it is helpful to understand what focus groups are not. The term can sometimes falsely signify groups whose primary purpose is not research, but rather therapy, decision-making, education or even behaviour change. Furthermore, procedures that utilise multiple participants without interactive discussion should not be confused with focus groups.110 Such procedures are more aptly termed group interviews, and offer a time-saving alternative to conducting individual interviews.111 Finally, focus groups should be distinguished from methods that collect data from naturally occurring group discussions where no one functions as an interviewer.112 The latter distinction is crucial to the focus group discussion method, as it points to one of the key strengths of this method, namely that the discussion is guided by the researcher’s interests.113 As a technique of collecting qualitative data, focus group discussion thus has two defining features which render it ideal for this particular analysis: the reliance on the researcher’s focus and the group’s interaction.114 The former enables the researcher to collect concentrated amounts of data on precisely the topic of interest, in this case SSR, while the latter uses the group’s interaction to produce data and insights that would otherwise be less accessible. In comparison to the participant observation method often applied by anthropologists, which involves a variety of techniques including informal interviews, collective discussions and life histories, the focus group method enables the researcher to observe a large amount of interaction on a particular topic in a limited period of time based on his/her capacity to assemble and direct the group sessions.115 The downside of such control is that it inevitably means focus groups occur in unnatural social settings, thus reducing the range of data that can be collected, inhibiting the scope of interactions among participants and preventing an open discussion of the research topic.116 However, as the interest here is not to collect data on social actions but rather on the discussion of these activities, the focus group is the more appropriate method. Moreover, the naturalistic settings required for the participant observation method are often difficult to locate and gain access to.117 The two defining features of the focus group technique point to the two key issues that must be considered during the planning phase: the researcher’s ability to communicate the topic of interest to the participants in such a way as to assemble and direct group sessions (moderation), and the composition of the members of the group. As regards the issue of moderating the focus groups, the distinctiveness of the settings under study – subject sensitivity, language 64 Judy Smith-Höhn barrier (local dialect) – meant that no researcher from the core team occupied the role of moderator. Instead, a local experienced moderator was employed. Careful consideration was also given to the intended composition of the groups. In both countries, four focus group discussions were held, each having six to eight participants. Discussants were chosen according to two criteria: the members of each group had to be homogeneous and virtual. The first criterion denotes similarities among group participants in terms of their social, educational and economic background. Hence one focus group comprised people from the educational sector (teachers and students), group two’s members were all (previously) employed in the health sector (health workers), the third discussion was conducted among market women, while members of the last group were either community or youth leaders. The second criterion – that the groups had to be virtual as opposed to real – merely meant that discussants within each group should not be familiar with each other. The composition of the focus groups was intended to reduce the risk that status-related barriers or entrenched roles would unduly influence participants’ responses. Each round of discussion lasted one day, during which time it was possible to explore the assumed motivation and legitimacy of security actors as well as other issues relevant to the security sector. A systematic comparison with the security situation in the past – more specifically, the period prior to the end of the civil war – was undertaken in order to map the extent of the transformation in the configuration of the various security actors. The key outcome of the focus group discussions was a graphical mapping of security actors which depicts the identity and relative significance of, as well as the interactions among, these various actors. 3.2.4 Focused, semi-structured interviews with elites and experts The semi-structured interview is perhaps the classic form of data collection used in the social sciences. Interviews are conducted with a relatively open framework allowing for focused two-way communication. Unlike a survey containing detailed questions that are formulated prior to the interviews, semi-structured interviews usually comprise a set of general questions or topics to be discussed. These form the basis for more specific questions that arise during the interview, permitting flexibility to probe a variety of issues depending on the knowledge and expertise of the interviewee and, of course, the interests of the interviewer. Methodology 65 Interviews were conducted with local elites as well as local and international experts – these included security actors such as representatives of the national security forces, traditional societies and private security companies. Unlike the specific questions posed in the survey, the interviews were guided by a set of open-ended, semi-structured questions. Conducting interviews with elites, such as local government representatives, traditional authorities and key representatives from civil society, enabled a comparison of perceptions ‘from above’ with those ‘from below’, i.e. the perceptions of the local population gathered by means of the focus group discussions and the quantitative survey. The expert interviews, as mentioned above, add a more objective view on the security sector in each country. Notes 1 2 3 4 5 6 7 8 9 10 Rueschemeyer, quoted in Timothy C. Lim (2006) Doing Comparative Politics. An Introduction to Approaches and Issues. Boulder, CO: Lynne Rienner, p. 26. Lim, ibid. Ibid., p. 27. Giovanni Sartori (1994) ‘Compare why and how. Comparing, miscomparing and the comparative method’, in Mattei Dogan and Ali Kazancigil (eds) Comparing Nations: Concepts, Strategies, Substance. Oxford: Blackwell, p. 16. John Gerring (2004) ‘What is a case study and what is it good for?’, American Political Science Review, 98(2), p. 348. Alexander L. George and Andrew Bennett (2005) Case Studies and Theory Development in the Social Sciences. Cambridge, MA: MIT Press, p. 3. According to Bennett et al., this is generally evident in the mix of methods in leading political science journals since the mid-1980s, with 49 per cent of articles sampled having used statistics, 46 per cent cases studies and 23 per cent formal modelling, and roughly one in five of these having used more than one method. See Andrew Bennett, Aharon Barth and Kenneth R. Rutherford (2003) ‘Do we preach what we practice? A survey of methods in political science journals and curricula’, PS: Political Science and Politics, 36: 373–378, p. 374. Regardless of the accuracy of these figures, and despite its regional bias, with a sample of 1,000 articles in ten journals the survey is considered as representative. George and Bennett, ibid., p. 5. Ibid. Sartori, note 4 above, p. 15; Dieter Nohlen (2004) ‘Vergleichende Methode’, in Dieter Nohlen and Rainer O. Schultze (eds) Lexikon der Politikwissenschaft. Theorien, Methoden, Begriffe, Vol. 2, C. H. Beck: Munich, p. 1042; Alfred Grosser (1974) Politik Erklären: Unter Welchen Voraussetzungen? Mit Welchen Mitteln? Zu Welchen Ergebnissen? Frankfurt/Main: Ullstein, p. 19. Harold D. Laswell (1968) ‘The future of the comparative method’, Comparative Politics, 1(1), p. 3; Gabriel A. Almond (1966) ‘Political theory and political science’, American Political Science Review, 60(4): 877ff, quoted in Nohlen, ibid., p. 1042. 66 Judy Smith-Höhn 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 See e.g. Lim, note 1 above, p. 19; Nohlen, note 9 above, p. 1042; Sartori, note 4 above, pp. 15–34. Arend Lijphart (1971) ‘Comparative politics and the comparative method’, American Political Science Review, 65: 682–693. Ibid. Though this may seem an obvious statement, Giovanni Sartori, speaking of comparative politics in the United States, observed that the field is often defined as the study of ‘other countries’, and in most cases only one. See Sartori, note 4 above, p. 14; Lim, note 1 above, p. 29, n7. In other words, while the field is defined in terms of a particular method – i.e. comparison – several scholars of comparative politics have no interest, notion nor training in comparing. See Sartori, note 4 above, p. 15. Nohlen, note 9 above, p. 1042. Ibid., p. 1043. Sartori, note 4 above, p. 15. Lim, note 1 above; Nohlen, note 9 above; Todd Landman (2003) Issues and Methods in Comparative Politics: An Introduction. London: Routledge. Lim, note 1 above, pp. 20ff. Ibid. Both Lim and Sartori concede that the experimental and statistical methods are better methods of control. In the social sciences, however, the experimental method has limited pertinence and the statistical method requires many cases, which research in this field is generally short of. Hence, researchers often have to opt for the second-best method of control. See Lim, note 1 above, p. 22; Sartori, note 4 above, p. 16. Lim, ibid., p. 21. Nohlen, note 9 above, p. 1044; Landman, note 18 above, p. 4. Sartori, note 4 above, p. 16. Landman, note 18 above, p. 4. Charles C. Ragin (1989) The Comparative Method. Moving Beyond Qualitative and Quantitative Strategies. Berkeley, CA: University of California Press, p. 11. Lim, note 1 above, p. 22. Nohlen, note 9 above, p. 1044. Landman, note 18 above. Lim, note 1 above, p. 22. Landman, note 18 above, p. 4. See e.g. Nohlen, note 9 above, p. 1044; Landman, note 18 above, p. 4. Nohlen, ibid. See e.g. ibid., p. 1043; Sartori, note 4 above, p. 16. Dogan, quoted in Nohlen, ibid., p. 1044. Sartori, note 4 above; Landman, note 18 above, p. 4. Sartori, ibid., p. 17. Ibid. Landman, note 18 above, pp. 4ff. Sartori, note 4 above, p. 19. Ibid. James L. Sundquist (1988) ‘Needed: A political theory for the new era of coalition government in the United States’, Political Science Quarterly, 103(4): 613–635. Sartori, note 4 above, pp. 19ff. Ibid., p. 20. Methodology 67 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 Ibid., p. 21. Nohlen, note 9 above, p. 1044. See e.g. Sartori, note 4 above, p. 22. Nohlen, note 9 above, p. 1045; Lim, note 1 above, p. 55. Nohlen, ibid., p. 1044. Both West African countries have witnessed protracted phases of violence, and the civil wars in both countries are closely linked: mercenaries fought in the civil wars; the ruling class continuously meddled in the affairs of neighbouring countries; and Charles Taylor’s involvement in the Sierra Leonean civil war is now known to most as a result of his ongoing trial in the Hague. Further similarities include the far-reaching involvement of international actors during the post-conflict phases, the prevailing socio-economic conditions and the presence of a number of different (types of) security actors. A detailed description of the similarities between the case studies is provided in Chapter 4. J. ’Kayode Fayemi (2004) ‘Governing insecurity in post-conflict states: The case of Sierra Leone and Liberia’, in Alan Bryden and Heiner Hänggi (eds) Reform and Reconstruction of the Security Sector. Münster: LIT, p. 179; available at http://se2.dcaf.ch/serviceengine/ FileContent?serviceID=DCAF&fileid=B0EA7833-62AD-2215-99BF182ADF9F413A&lng=en (accessed 13 June 2006). Lim, note 1 above, p. 2. John Stuart Mill (1868) A System of Logic: Ratiocinative and Inductive; Being a Connected View of the Principles of Evidence and the Methods of Scientific Investigation, 7th edn. London: Longmans, Green, Reader & Dyer. Ibid,, p. 425. See for example Neil J. Smelser (1976) Comparative Methods in the Social Sciences. Princeton, NJ: Prentice-Hall; Mattei Dogan and Dominique Pelassy (1990) How to Compare Nations. Strategies in Comparative Politics. Chatham, NJ: Chatham House, pp. 111ff; Adam Przeworski and Henry J. Teune (1970) The Logic of Comparative Social Inquiry. New York: John Wiley & Sons; Sartori, note 4 above, p. 22. What these have in common is the agreement that the (similar or dissimilar) dependent variable is known, the independent variable is what is to be determined and neither heterogeneous nor homogeneous contextual factors are to be considered. Lim, note 1 above, pp. 34ff. Mill, note 53 above, p. 425. For a detailed discussion of independent and dependent variables in social science research, see Lim, note 1 above, p. 35. Ibid., pp. 34ff. Ibid., pp. 38ff. See e.g. ibid., pp. 41ff; Nohlen, note 9 above, pp. 1046f. Lim, ibid., pp. 43ff. Ragin, note 26 above. Lim, note 1 above, p. 44. Ibid., pp. 50ff. Lim’s elaboration on the case study as a comparative method, particularly his reference to the binary analysis, i.e. comparison of two units (cases), is synonymous with the MSSD, and hence provides little additional insight into this particular research design. See ibid., pp. 44ff. See e.g. Sartori, note 4 above, p. 23. George and Bennett, note 6 above, p. 17. 68 Judy Smith-Höhn 69 70 71 72 73 74 75 76 77 78 79 80 81 81 82 83 84 85 86 87 88 89 Ibid., p. 18. Charles C. Ragin and Howard S. Becker (eds) (1992) What Is a Case? Exploring the Foundations of Social Inquiry. Cambridge: Cambridge University Press, p. 1. See e.g. Stanley Lieberson (1992) ‘Small n’s and big conclusions: An examination of the reasoning in comparative studies based on a small number of cases’, Social Forces, 70(2): 307–320; Gary King, Robert O. Keohane and Sidney Verba (1994) Designing Social Inquiry. Scientific Inference in Qualitative Research. Princeton, NJ: Princeton University Press; Lijphart, note 12 above. E.g. Robert K. Yin (1994) Case Study Research. Design and Methods. Thousand Oaks, CA: Sage; Joe R. Feagin, Anthony M. Orum and Gideon Sjoberg (eds) (1991) A Case for the Case Study. Chapel Hill, NC: University of North Carolina. Yin, ibid. George and Bennett, note 6 above. Gerring, note 5 above. Harry Eckstein (1975) ‘Case study and theory in political science’, in Fred I. Greenstein and Nelson W. Polsby (eds) Handbook of Political Science,. Reading, MA: AddisonWesley. Gerring, note 5 above, p. 342. Ibid. Ibid. Ibid Lim, note 1 above, p. 53. Gerring, What Is a Case Study, p. 343. Gerring, note 5 above, p. 343. Ibid See James Mahoney and Dietrich Rueschemeyer (eds) (2003) Comparative Historical Analysis in the Social Sciences. Cambridge: Cambridge University Press. This typology has shown that, for case study research design, the comparison is conducted according to the time aspect – time, object, context and space are the four ways in which a comparison can be carried out. See Nohlen, note 9 above, p. 1044. As Lim argues, a single case study can be explicitly comparative when time is taken into account. See Lim, note 1 above. Nohlen, ibid. See e.g. George and Bennett, note 6 above, p. ix. 2006 serves as the cut-off point for analysis as the data were collected during that year. Though there may have been changes in the SSR processes in both countries in 2007, these are believed to be minimal. Institutions are not easily susceptible to change and, more often than not, such change is slow. Moreover, it is believed that the inferences drawn from the analysis retain their relevance for SSR inasmuch as they are generalisable, and particularly they are yet to be considered closely by stakeholders in policy-making. The volume draws from the work of several authors in the field of case study methods, including Alexander George’s structured, focused comparison of cases; Arend Lijphart and Harry Eckstein’s descriptions on the various theory-building types of case studies; Charles Ragin’s analysis of interaction effects and comparative methods of studying them; and Gary King, Robert O. Keohane and Sidney Verba’s work on scientific inferences in qualitative research. See Lijphart, note 12 above; Eckstein, note 76 above; Ragin, note 26 above.; King et al., note 71 above. George and Bennett, note 6 above. Methodology 69 90 91 92 93 94 95 96 97 98 99 100 101 102 103 George and Bennett (ibid., p. ix) define case study research as a qualitative exercise that typically involves a combination of within-case analyses and cross-case comparisons through process tracing and congruence testing, respectively. Since it is not of immediate relevance to the comparative method, here is a brief note on the explanations given by various authors concerning the reasons for and advantages of employing the case study method. Basically, case studies are employed for five reasons: configurative/atheoretical ideographic (Eckstein); interpretive (Lijphart); hypothesisgenerating (Lijphart); crucial (Eckstein), i.e. theory-confirming or disconfirming (Lijphart); and deviant (Lijphart). See Eckstein, note 76 above; Lijphart, note 12 above. According to George and Bennett (ibid., p. 19), case studies have four advantages over statistical methods and formal models in testing hypotheses and developing theory, namely the potential for achieving high conceptual validity; strong procedures for fostering new hypotheses; the value as a useful means to examine closely the hypothesised role of causal mechanisms in the context of individual cases; and the capacity to address causal complexity. Alexander George developed the method of structured, focused comparison of cases as one that ‘outlines process-tracing and other within-case modes of analysis as key complements or alternatives to controlled comparison of cases’. George and Bennett, ibid., p. ix. Ibid. Yin, note 72 above, pp. 13ff. Ibid., p. 14. Abbas Tashakkori and John W. Creswell (2007) ‘Editorial: The new era of mixed methods’, Journal of Mixed Methods Research, 1(1): 3–7, p. 4; available at http://mmr.sagepub.com/cgi/reprint/1/1/3 (accessed 27 November 2006). See for example Henry E. Brady and David Collier (eds) (2004) Rethinking Social Inquiry: Diverse Tools, Shared Standards. Lanham, MD: Rowman & Littlefield. Discourse resulted from an ensuing publication: see Henry E. Brady, David C. Collier and Jason Seawright (2006) ‘Toward a pluralistic vision of methodology’, Political Analysis, 14(3): 353–368; Henry E. Brady, David C. Collier and Jason Seawright (2007) ‘Sources of leverage in causal inferences’, paper presented at American Political Science Association Annual Meeting, Chicago, 29 August–2 September; Nathaniel Beck (2006) ‘Is causal-process observation an oxymoron?’, Political Analysis, 14(3): 347–352; Nathaniel Beck (2007) ‘Causal process “observation”: Oxymoron or old wine’, paper presented at American Political Science Association Annual Meeting, Chicago, 29 August–2 September. See e.g. Alan Bryman (2007) ‘Barriers to integrating quantitative and qualitative research’, Journal of Mixed Methods Research, 1(1): 8–22; available at http://mmr.sagepub.com/cgi/reprint/1/1/8.pdf (accessed 27 November 2007). Ibid. For a detailed outline of possible barriers to integrating qualitative and quantitative data see ibid. Ibid., p. 20. Jane Elliott (2005) Using Narrative in Social Research: Qualitative and Quantitative Approaches. Thousand Oaks, CA: Sage, p. 185. Fieldwork in Liberia was carried out from late November 2005 to mid-February 2006; field research in Sierra Leone followed a few months later, from early April to the end of June 2006. 70 Judy Smith-Höhn 104 105 106 107 108 109 110 111 112 113 114 115 116 117 The guidelines for randomising the survey sample are provided in Annex I. Statistical Products and Service Solutions. For detailed instructions on how to use SPSS see e.g. George A. Morgan, Orlando V. Griego and Gene W. Gloeckner (2001) SPSS for Windows: An Introduction to Use and Interpretation in Research. Mahwah, NJ: Lawrence Erlbaum Associates; George A. Morgan, Nancy J. Leech, Gene W. Gloeckner and Karen C. Barrett (2004) SPSS for Introductory Statistics: Use and Interpretation. Mahwah, NJ: Lawrence Erlbaum Associates. Lengua and Hendershott, cited in David L. Morgan (1996) ‘Focus groups’, Annual Review of Sociology, 22(1): 129–152. Morgan, ibid., p. 129; David L. Morgan (1997) Focus Groups as Qualitative Research. Thousand Oaks, CA: Sage, p. 6. Morgan, note 108 above, p. 129. For a review of other group techniques such as nominal or Delphi groups, see David W. Stewart and Prem N. Shamdasani (1990) Focus Groups: Theory and Practice. Thousand Oaks, CA: Sage, pp. 153ff. Peter Loos and Burkhard Schäffer (2001) Das Gruppendiskussionsverfahren. Theoretische Grundlagen Und Empirische Anwendung. Opladen: Leske + Budrich, p. 12. Such naturally occurring group discussions are often used as a means of data gathering when the researcher is less interested in the content of discussions and more in the dynamics and structure of the conversations. In ethnographic research, such ‘informal interviews’ are used as a means of gathering information about a particular segment of the culture under study. See ibid., pp. 12ff. Morgan, note 108 above, p. 129. Morgan, note 109 above, p. 13. Ibid., p. 8. Ibid. Ibid., p. 9. Chapter 4 Case Studies Before comparing each of the cases according to the research design (section 3.1.4), the case studies will be put into historical context. An analysis of the security sector and the reform efforts thus far undertaken in both countries facilitates an understanding of the setting within which the actors have been implementing their strategies. It will also illuminate the causes of the conflicts, enabling an assessment of whether the pertinent issues have been adequately addressed so as to prevent a relapse into violence or they remain unresolved. The key focus will be on the countries’ post-colonial history, and the chapter will proceed with one caveat: a detailed analysis of the causes of conflict in each country warrants a study all on its own, and no attempt is made here to illustrate the multifaceted nature of conflicts in general.1 A description of the dynamics of a particular conflict that differentiates between various types of causes – triggers, prolonging and accelerating factors and root causes – goes beyond the scope of this study. However, these dynamics shall be kept in mind when elaborating on these legacies of the past and, where necessary, will be referred to in so far as they have a bearing on the extent to which a factor can have a negative effect on SSR efforts. The historical contextualisation is followed by the within-case analyses, the leitmotiv being the four guiding questions formulated in Chapter 2. 4.1 Liberia 4.1.1 Background2 Liberia was founded as a settlement for freed American slaves of African descent, with executive power initially vested in a governor appointed by the American Colonization Society. Although they made up only a fraction of the Liberian population, the American-born Africans, known as AmericoLiberians, and their descendants dominated political life from Liberia's 72 Judy Smith-Höhn independence in 1847 until April 1980, when Samuel Doe overthrew the government in a bloody military coup. The coup was the result of decades of exclusion suffered by the indigenous African population of Liberia. The Doe regime was initially multi-ethnic and participatory in nature, but during the second half of the 1980s this gave way to a new and equally exclusive patronage system, with members of Doe’s own ethnic group, the Krahn, as well as the Mandingo being recruited to key posts in government and the military. From then on, the regime's chief features were the ethnicisation (and brutalisation) of rule and the personalised use of the security apparatus, which was also a common trait of the previous regimes. Liberia has never had a truly national army dedicated to defending the common good. In 1989 a former government official, Charles Taylor, launched an armed uprising. His forces, known as the National Patriotic Front of Liberia (NPFL), soon seized control of 95 per cent of the country. Only the intervention of the military arm of ECOWAS prevented Taylor from capturing the capital, Monrovia. ECOWAS negotiated a peace settlement among the two rebel groups and the government, and installed an interim government led by Amos Sawyer. Yet Taylor's force, with military aid from Libya and Burkina Faso, continued its military campaign, began a siege of Monrovia in 1992 and engaged in fighting with ECOWAS forces. AntiTaylor groups, among them Doe partisans, joined forces as the United Liberation Movement of Liberia for Democracy (ULIMO) and declared war against Taylor. Alongside the three conflict parties (government troops, NPFL and ULIMO), their various factions and local civil defence forces, the ECOWAS troops soon became part of the local constellation of violence in Liberia. The ensuing civil war lasted eight years, claiming more than 200,000 lives and threatening to destabilise the entire West African region. Although ECOWAS had sent troops in 1990, it was only after 14 peace agreements and the establishment of ECOMOG in 1996 that the civil war was brought to an – albeit temporary – end. The elections held under the peace agreement in 1997 resulted in a victory for Charles Taylor, who secured 75 per cent of the vote, largely because he threatened to continue the violence if he lost the election. Like Doe, the warlord-turned-president pursued the tradition of neopatrimonial rule and installed an authoritarian, repressive regime, looting Liberia’s resources. The only difference was that in Taylor's case the apparatus of repression was recruited mainly from the NPFL. The peace agreements were not upheld, and renewed fighting soon erupted between Taylor’s government forces (Armed Forces of Liberia – AFL) and a rebel Case Studies 73 group, Liberians United for Reconciliation and Democracy (LURD), which was backed by Guinea. Two years after the elections, civil war broke out anew and persisted for years, escalating in March 2003 when LURD advanced to within 10 kilometres of Monrovia. At the same time, the newly formed Movement for Democracy in Liberia (MODEL), backed by Côte d'Ivoire, launched an offensive from the south. The fighting claimed more than 10,000 lives. Once again, peace talks were initiated, but the fighting was only brought to a halt when ECOWAS deployed 3,500 peacekeepers in August 2003 and a weakened Taylor was forced to agree to exile in Nigeria. On 18 August 2003, a week after Taylor’s departure, the three warring factions (government forces, LURD and MODEL) signed the ECOWASbrokered Comprehensive Peace Agreement (CPA) in Accra, Ghana. The elaborations above show, that in the case of Liberia, international actors played a considerable role in the developments of its civil war. As fighting escalated in 2003, there were calls for the United States, as a nation with historical ties to Liberia, to send peacekeeping forces. ECOWAS took the initiative and together with the International Contact Group on Liberia (ICGL) – comprising ECOWAS, the United Nations, the African Union (AU), the EU, France, the UK, the USA, Ghana, Morocco, Nigeria and Senegal – organised peace talks with the parties involved. Despite the initial cease-fire agreement of June 2003, LURD continued its march towards the capital. The 3,550 ECOWAS troops – the ECOWAS Mission in Liberia (ECOMIL) – were later relieved by a multinational UN peacekeeping force, UNMIL, which began with a troop strength of 15,000. The presence of US Marines offshore to act as a rapid response force may well have exerted a deterrent effect as well, and helped bring the conflict under control. In August 2007 UNMIL had just over 15,000 uniformed personnel, including 207 military observers and 1,146 police officers. The mission also employs around 500 international civilian staff and more than 900 local UN workers as well as 232 UN volunteers.3 In view of the significant role played by UNMIL in Liberia’s political arena, some observers even claim that Liberia has in effect become a UN protectorate.4 The peace agreement of 2003 established an interim administration that took to power in October 2003. This National Transitional Government of Liberia (NTGL), headed by Gyude Bryant, comprised representatives from all parties present at the peace talks – the warring factions, 18 opposition political parties and five civil society movements. Hence the transitional administration consisted of a number of former fighters, many of whom were less concerned for the welfare of the nation and more interested in carving out rent-seeking opportunities from their allocation of utilities and 74 Judy Smith-Höhn parastatals. The transitional Liberian state was effectively a partnership between UNMIL, which had broad executive and administrative powers, and a transitional government made up of ex-warlords and their proxies.5 Security-related provisions in the CPA included the introduction of a disarmament, demobilisation, rehabilitation and reintegration (DDRR) programme for the combatants of all warring factions, the establishment of a truth and reconciliation commission and transition processes. Part 4 of Liberia’s CPA is devoted entirely to SSR.6 The effects of the 14-year civil war were devastating. By the time the CPA was signed, the state had effectively collapsed, taking the economy and society with it. The war had claimed some 250,000 lives, half of them civilians. In addition to 500,000 internally displaced persons (IDPs), more than 80 per cent of the population were unemployed, with 75 per cent living on less than a dollar a day.7 Even during the transitional phase, after the cessation of war, the national context within which post-conflict reconstruction was being implemented was characterised by deprivation and a lack of opportunity, largely as a result of poor governance.8 In 2005 a democratically elected government under President Ellen Johnson-Sirleaf replaced the interim administration. Beginning in 2006, Liberia’s political institutions were significantly restructured. Shortly after her inauguration, Johnson-Sirleaf selected her new cabinet and allocated other senior government positions. The number of civil servants within all ministries was drastically reduced following personnel audits to determine technical competency. However, since then there have been few changes of staff in central offices and many of those who were involved in the civil war still play a role in the country’s political landscape. Former Taylor associates and other warlords, for example, were elected into parliament in 2005, in the first legislative elections since the end of the war. The most prominent position – House speaker – was at one point held by Taylor's former son-inlaw, Edwin Snowe. Also in a prominent post is former secretary-general of the now-defunct rebel group LURD, Isaac Nyanebo, who was chosen as the leader of the Senate. Liberia therefore remains a difficult environment for post-conflict reconstruction and reform. 4.1.2 Liberia’s security sector reform strategy SSR efforts in Liberia leave much to be desired. After 14 years of civil war and the signing of the CPA in Accra, Liberia was one of the most challenging contexts for post-conflict reconstruction since the end of the Cold War.9 This comes as no surprise: of all aspects of institutional design in Case Studies 75 Africa, reconstruction of the security sector has proven to be the most difficult, complex and delicate task in post-conflict peacebuilding efforts.10 As it is relatively early in the reform process, few authors have written on SSR issues in Liberia, and even fewer have provided a comprehensive overview of the efforts made thus far. This is not an indication of a lack of zeal on the part of authors, but more a reflection of the failure of the country’s transitional government under Guyde Bryant to develop a discernable national security and defence strategy in its period of rule from 2003 to 2005. This, predictably, meant that a coherent and comprehensive framework for the reconstruction of the security sector was lacking.11 Instead, the reform process has been compartmentalised, a situation evident, for example, in the emphasis on reforming the police without developing a corresponding reform strategy for the judiciary and correctional service, with the result that the latter two institutions are overburdened and inundated.12 While the detailed provisions of the CPA have set a high standard, it appears that the practical implementation of these policies has proven to be very difficult. Part 4 of Liberia’s CPA makes three key provisions regarding SSR: the disbandment of all irregular forces; the restructuring of the AFL with a new command structure and forces ‘which may be drawn from the ranks of the present GOL13 forces, the LURD and the MODEL, as well as from civilians with appropriate background and experience’; and the immediate restructuring of the national police force, immigration, the Special Security Service (SSS), customs security guards and other statutory security units.14 The CPA called on ECOWAS, the UN, AU and ICGL to provide advisory staff, equipment, logistics and experienced trainers for the security reform effort. In addition, the agreement stipulated that the USA should play a leading role in organising the military restructuring programme. While the CPA constitutes the major legal reference for the SSR process in Liberia, two other legal frameworks also make mention of security sector reform, namely the 1986 constitution of Liberia and UN Security Council (UNSC) Resolution 1509. These conflicting legal frameworks impede progress towards effective reform of the sector.15 With the inauguration of Johnson-Sirleaf in January 2006, the constitutional provisions that had been suspended during the transitional phase of 2003–2005 were restored.16 Article 54(e) of the 1986 constitution stipulates that the president, as commander-in-chief of the armed forces, ‘appoints members of the military from the rank of lieutenant or its equivalent above; and field marshals, deputy field marshals and sheriffs’.17 The constitution further provides that the national legislature has broad 76 Judy Smith-Höhn responsibilities regarding security issues. According to Article 34(b)–(c), for instance, the legislature has the power: to provide for the security of the Republic and to provide for the common defense, to declare war and authorize the Executive to conclude peace; to raise and support the Armed Forces of the Republic, and to make appropriations therefor provided that no appropriation of money for that use shall be for a longer term than one year; and to make rules for the governance of the Armed Forces of the Republic.18 UNSC Resolution 1509 stipulates that the SSR process shall be supported by UNMIL, which will: assist the transitional government of Liberia in monitoring and restructuring the police force of Liberia, consistent with democratic policing, to develop a civilian police training programme, and to otherwise assist in the training of civilian police, in cooperation with ECOWAS, international organizations, and interested States.19 The resolution further notes that UNMIL, in cooperation with ECOWAS, international organisations and interested states, will assist the NTGL ‘in the formation of a new and restructured Liberian military’.20 Whereas the CPA was specific about the role of the USA in restructuring the army, the UN resolution refers only to ‘interested States’ and is specific only about the role of UNMIL in training the Liberia National Police (LNP). To date, though the legal frameworks vary in focus and emphasis, the question as to which should be the defining document for the SSR process has not been posed. Unlike in the past, when the Taylor leadership, despite the fragile security environment, invoked the Liberian constitution in order to abandon the SSR process prematurely by asking ECOMOG, the recently elected government is not prepared to revert to this old practice.21 The overall understanding of the security sector by those involved in the reform process is that it ‘comprises security agencies and civilian oversight bodies that promote and protect the national security interests of the people and the state’.22 As such, the security sector includes intelligence, paramilitary and military bodies as well as civilian institutions that have oversight responsibilities over these security agencies. Liberia’s security sector comprises 11 institutions: the AFL, LNP, National Security Agency (NSA), National Bureau of Investigation, SSS, Drugs Enforcement Agency, National Fire Service, Ministry of National Security, Bureau of Immigration Case Studies 77 and Naturalisation, Bureau of Customs and Excise and the Monrovia City Police.23 The USA has seconded the programme to rebuild the AFL as a small professional force of 2,000 to the private security company Dyncorp, while UNMIL has taken up the training of the local police through CIVPOL. Dyncorp, a company which has worked with the USA on similar programmes in Iraq and Afghanistan, has already begun implementing the $35 million programme in Liberia. However, several civil society organisations (CSOs) have raised concerns about the efficacy of employing private security companies to train the national army, citing the lack of transparency and accountability in terms of the training procedure as key problems in this area. Unlike the US approach, the UN has employed non-commercial means to retrain the Liberian police force through the UN Police: UNMIL’s training of local police recruits began in 2004, and some are already deployed with international teams of civil police operating throughout the country.24 The training programme was designed for new recruits, existing police officers and those undergoing specialised training, and was targeted to train a total of 3,500 police officers by June 2007. This approach has also met with criticism, with the selective retraining of existing LNP officers causing grievances among those – mostly older LNP staff – who were not selected. These elaborations reveal that the SSR process in Liberia has had a narrowed focus on the AFL, the LNP and the SSS. The other security agencies have not benefited as much; the Bureau of Customs and Excise and the Bureau of Immigration and Naturalisation, for example, have received capacity-building training from BIVAC International and UNMIL, respectively, while the Ministry of National Security and other agencies have merely undergone a vetting procedure.25 In 2003, as part of the CPA, the Governance Reform Commission (GRC) was established to set up a ‘national integrity framework… to promote transparency, accountability and the rule of law’.26 In 2006 the commission initiated nationwide consultations with key players, including district commissioners, chiefs, representatives of security agencies in the counties, UNMIL and CSOs. The problems associated with the GRC are a useful indication of the problems encountered when dealing with issues of local ownership. Whereas the GRC members perceive themselves as an independent body, the UN thinks otherwise; this became apparent in comments made by a senior UN staff member regarding the GRC’s independence.27 Judy Smith-Höhn 78 In recognition of the need to deal with the legacies of the past as a key prerequisite for effective SSR, the issue of transitional justice was addressed in the Liberian peace process. Part 6 of Article XIII of the 2003 CPA stipulates the establishment of the Truth and Reconciliation Commission (TRC), which officially began its work on 22 June 2006. The commission – comprising nine Liberian and three international members – is mandated to investigate human rights violations committed between January 1979 and October 2003. The investigative process is scheduled to take a maximum of two to three years. Several international and local CSOs are involved in this process. Amnesty International, for example, together with a steering group of local organisations, monitors the commission’s work on a regular basis. Established under the auspices of the CPA, Liberia’s National Commission for Disarmament, Demobilization, Rehabilitation and Reintegration falls in line with the broader SSR approach that focuses on the need to deal with the legacies of the past (see Chapter 2).28 The commission was to coordinate all DDRR activities. However, following UNMIL’s deployment, a joint implementation unit – comprising UNMIL’s DDRR Unit, the UNDP and the NTGL – was established to implement the DDRR process and coordinate DDRR activities.29 This unit did most of the actual planning and subcontracted local service providers, such as Monrovia-based offices of the UN Children’s Fund (UNICEF) and the Norwegian Refugee Council, which mostly carried out the technical and educational aspects of the cantonment process.30 In sum, the Liberian peace process is exemplary for the UN modus operandi developed since the end of the Cold War. The approach follows a consecutive series of activities, with the peacebuilding process implemented according to an operational checklist starting with a peace agreement, followed by the deployment of peacekeepers, DDR of former fighters, the repatriation and return of refugees and IDPs, judicial and security sector reform (the latter focusing primarily on the military and the civilian police) and concluding with the organisation of elections.31 4.2 Sierra Leone 4.2.1 Background32 Founded in 1787 as a settlement for freed slaves and recaptives,33 Sierra Leone became a British protectorate34 in 1896 and achieved full Case Studies 79 independence in 1961 under Prime Minister Milton Margai. Since then, six general elections have taken place – as well as five military coups. As in Liberia, here too the descendants of former slaves (Creoles or Krios) formed the elite class during the colonial period, dominating the civil service. After independence, however, the animosity between the indigenous population and the Krios was soon overlaid by political rivalries between various ethnic groups, most notably between the Temne group in the north and the Mende of the south and east. Due to the reinforcement of tribal patriotism through colonial education, for example, appeals to ethnicity became one of the easiest ways to gain support during post-independence elections.35 Moreover, much like Liberia, Sierra Leone is marked by a tradition of patrimonial rule, which has its origins in the pre-colonial period when warrior chiefs who controlled rival trade networks developed large followings of retainers and protected clients, many of whom were first acquired as slaves in war.36 With the British-supervised indirect rule, this form of patronage was perpetuated by the chiefs in the hinterland. Since the chiefs formed the backbone of the Sierra Leone People’s Party (SLPP), which became the ruling party in 1961, the new multi-party system of governance was sustained by patronage networks that extended from the state house and parliament down to the chieftaincies, sections and villages.37 As a system of sponsorship whereby solidarity with a patron was ensured by the big man’s support for his so-called people, this vertically integrated hierarchy often takes on ethnic colourings.38 The increasing clamour for inclusion in this dysfunctional system and the consequent marginalisation and exclusion of large sections of the population from socio-economic advancement were key factors in the outbreak of a civil war that left over 50,000 dead and 300,000–400,000 internally displaced. The desire to benefit from the primary economic resource, namely high-quality alluvial diamonds, was but a perpetuating factor.39 The violent seizure of power by the All People’s Congress (APC) led by Siaka Stevens in 1967–1968 marked the start of a series of military coups. From 1985, Stevens’s self-proclaimed successor, Major-General Joseph Saidu Momoh, governed the country – which was plagued by economic crises and political unrest. The civil war erupted in 1991 when a rebel insurgency group under former army corporal Foday Sankoh, the Revolutionary United Front (RUF), which was supported and armed by Charles Taylor, initiated a campaign against President Momoh, capturing towns on the border with Liberia. In 1992 Momoh was overthrown in a military coup staged by a group of young army officers led by Captain 80 Judy Smith-Höhn Valentine Strasser, who then went on to head the National Provisional Ruling Council (NPRC). The NRPC coup-makers had apparently received support from the RUF in overthrowing the APC’s one-party regime on the understanding that the NPRC and RUF would form a coalition government. Strasser, however, strengthened by a second Nigerian battalion and more Guinean troops, attempted to crush the RUF by re-equipping the army and forcing it to engage with rebel troops in the provinces.40 But many of the poorly trained and underpaid government troops deserted and defected to the RUF as a response to the state neglect of their material and security needs prior to and during the war. Aptly dubbed sobels,41 their atrocities against the civilian population and involvement in the illicit diamond trade prompted Strasser to hire the now-defunct Executive Outcomes (EO), a private military company, to supplement his troops and liberate the most important areas from RUF’s control. Although EO was extremely effective in fulfilling its directive, its involvement revealed weaknesses in government by exposing the incompetence of its own armed forces to manage the situation, consequently arousing their resentment, and also led the latter to assume that the mercenaries were there not only to defeat the RUF but also to protect the government from its own army. In 1996 Strasser was ousted by a military coup led by his former defence minister, Brigadier-General Julius Maada Bio. Bio fulfilled his promise of a return to civilian rule when he handed power over to the newly elected president, Ahmad Tejan Kabbah of the SLPP, in early 1996. Kabbah's government signed a peace accord with Sankoh's rebels later that year, yet rebel attacks continued, again aided by Charles Taylor. During the civil war, so-called civil defence forces (CDF) emerged in response to community-level security problems that were not being addressed by the national security forces. These local self-help measures proved effective in safeguarding security, albeit confined to the community level. Unlike the community-driven efforts in neighbouring Liberia, such measures were carried out in a more organised manner, the CDF having traditionally been managed through various forms of local and customary authority. Civilians organised themselves into civil defence groups that were usually led by traditional hunters. Thus the CDF, commonly referred to under their Mende name kamajors,42 were formed amid fears of threats from both the rebels and government troops. Again, in contrast to Liberia, these traditional institutions – as part of the Poro system of secret societies – were key players during the civil war. Later, however, certain elements within the CDF resorted to extortion and acts of brutality. Case Studies 81 There appears to be some ambiguity as regards the role of the chiefs, depending mostly on the respective chief’s personal ambitions to protect either his people or his own life. Thus while some chiefs deserted their people in the rural areas and fled to other parts of the region, others played an instrumental role in recruiting and funding kamajors. In 1997 a coalition of army officers led by Major-General Johnny Paul Koroma and members of the RUF forced Kabbah to leave office. The UN reacted by imposing sanctions in October that year, impeding the supply of arms and petroleum products. Kabbah was reinstated in March 1998 after the Nigerian-led ECOMOG force successfully ousted the junta, but the fighting continued in many parts of the country and the RUF reclaimed parts of the capital from regional peacekeeping force. Ultimately the rebels were defeated, and the UN intervened and organised a cease-fire agreement between Kabbah and Sankoh’s RUF in 1999. Terms of the agreement included the rebels receiving posts in a new government and a general amnesty from prosecution for war crimes for all forces. The government had largely ceased to function effectively and with at least half of its territory remaining under rebel control, fighting continued despite the arrival of the first UN troops. When in May 1999 nearly all Nigerian forces had left and UN forces were attempting to disarm the RUF in eastern Sierra Leone, Sankoh’s forces clashed with the UN troops, and some 500 peacekeepers were taken hostage as the peace accord effectively collapsed.43 In May 2000 the RUF appeared to be on the verge of capturing Freetown, whereupon the British government decided in favour of military intervention. British troops were initially deployed to evacuate foreign nationals and restore order; later, they also provided support for the UN Mission in Sierra Leone (UNAMSIL) contingents, so that in 2001 the UN troops were able to deploy peacefully in rebel-held territory. Rebels were now being disarmed and the new British-trained Sierra Leone Army began deploying in rebel-held areas. After several failed peace agreements in 1996, 1997, 1999 and 2000,44 Sierra Leone finally emerged from 11 years of civil war in 2002 with a Joint Declaration of End of War between the RUF, the Sierra Leone Army and the quasi-official CDF, with the government and the UN agreeing to set up a special war crimes court. Within a year of signing the peace accord, Kabbah and his party again won landslide victories in the parliamentary and presidential elections, consolidating Kabbah’s position of power by securing a majority in parliament. The country’s progress towards peace was considered so significant that the UNAMSIL troops withdrew in December 82 Judy Smith-Höhn 2005. In January 2006, the UN Security Council authorised that UNAMSIL be succeeded by an assistance mission, the UN Integrated Office in Sierra Leone (UNIOSIL). Upon emerging from war in 2002, the key tasks for the Sierra Leonean government included capacity-building for state structures and public administration. Such goals could only be reached with the help of external actors, particularly the UN and the UK. Like Liberia, the Sierra Leonean government continues to be plagued by rampant corruption and lack of accountability, and although Sierra Leone has succeeded in instituting some democratic structures and holding transparent elections, there are still major shortcomings concerning issues of corruption, mismanagement and low state institutional capacities. The UK, as the former colonial power, is the key benefactor of Sierra Leone’s postconflict reconstruction reform. According to the Bertelsmann Transformation Index (BTI), the government’s fundamental administrative infrastructure now covers the entire country.45 However, the administration remains heavily dependent on external assistance and cannot function effectively without the support of external actors. 4.2.2 Sierra Leone’s security sector reform strategy Before the war, Sierra Leone’s security sector was largely based on ethnic and patrimonial alliances, and stagnated in what Ebo has aptly termed a ‘comatose state’.46 The benchmarks for recruitment and promotion within this sector, as with all other components of the public sector, were party political, ethnic and inter-personal relations, and not objective professional performance. Moreover, inadequate and irregular remuneration exacerbated corruption and abuses of office. Security sector governance was shaped by a focus on state and regime security, and alternative security institutions were established in order to guarantee their loyalty to the respective ruling party, effectively impeding operational coherence and accountability. Moreover, the consequent inter-institutional rivalry precipitated the self-defeat of the security sector within the state.47 After the first coup attempt in 1967–1968, during the one-party rule of Siaka Stevens, the regime consciously started to weaken state security through the creation of an alternative security system; at first through the Internal Security Unit (ISU) and later through an ISU subsidiary, the Special Security Division.48 The overthrow of the APC rule in a military coup in 1992, a year after the onset of rebellion in the diamond region, was precisely a result of this systematic deinstitutionalisation of the Case Studies 83 state, for Momoh was left with little or no capacity to confront the rebellion that removed him from office.49 The considerable influence of colonial history, and in this case British colonial rule, on the security sector in Sierra Leone is typical of many West African states.50 During colonial times, the authorities had established the security forces in order to protect British interests and contain resistance against the colonial administration through the suppression of the local population. The Sierra Leone Frontier Police, from which the Sierra Leone Police (SLP) originate, had been created not to defend the population against threats to their security, but to protect the interests of the colonial administration.51 Thus, despite the country’s independence from its colonial ruler, there continued to be a gap between ordinary citizens and the security institutions that were created for their protection but which, in fact, often served to threaten their well-being. Following the cessation of the civil war, and in keeping with the historical relationship between Sierra Leone and the UK, the former colonial power became the key benefactor of Sierra Leone’s post-conflict reconstruction agenda. More specifically, the UK has been involved in the reform of the country’s security sector through four institutions, namely DFID,52 the Ministry of Defence (MoD), the Foreign and Commonwealth Office (FCO) and the British-led International Military Advisory and Training Team (IMATT).53 Sierra Leone’s perception of the security sector currently encompasses the military, police, correctional, fire-fighting and intelligence services, relevant line ministries, CSOs and international stakeholders.54 In 2002 the National Security and Central Intelligence Act was passed as part of the SSR process, leading to the establishment of the Office of National Security (ONS) and the Central Intelligence and Security Unit (CISU) under section 11.1 of the Act.55 The ONS serves as the secretariat of the National Security Council, which, according to section 4.1 of the Act, is ‘the highest forum for the consideration and determination of matters relating to the security’ of the country.56 As noted above, the UK has been the major contributor towards rebuilding the country’s security sector, and as such, has developed a number of programmes to facilitate this process. The Sierra Leone Security Sector Reform Programme (SILSEP), for instance, was initiated in 1999 as a medium-term programme funded and managed by DFID, which was initially designed with three main objectives: to develop the Ministry of Defence; to support capacity-building within the ONS under the national security co-ordinator; and to rebuild the 84 Judy Smith-Höhn effectiveness of the CISU.57 IMATT, on the other hand, took the lead in training the government’s armed forces.58 Both activities began implementation in 1999. As for the reform and reconstruction of the SLP, the UK and UN have taken a combined role in training senior staff in strategic and operational planning through the Commonwealth Community Safety and Security Programme (CCSSP) and the civilian police section of UNAMSIL (CIVPOL).59 The CCSSP aims to establish the SLP as an ‘effective and accountable civilian police service that focuses on local needs-based policing based on the full involvement of civil society’.60 The objective was to establish a police force which is fully staffed, trained and equipped to maintain law and order without military support after UNAMSIL’s withdrawal.61 The SLP was also supplied with radios, weapons, uniforms and vehicles. Until recently, most police reform efforts were conducted through the CCSSP, with the Commonwealth Police Development Task Force shaping the SLP restructuring programme.62 Britain’s role vis-à-vis the SLP was executive and intrusive; from 1999 to 2003 the SLP was even led by a British national, Keith Biddle, as the inspector-general of police.63 During this period a number of new initiatives were developed, namely the Community Relations Department, the Complaint, Discipline and Investigation Department and family support units. As of June 2006, the police reform efforts were absorbed into SILSEP.64 At the time of writing, SILSEP was in its fourth phase, and had undergone an annual review in April 2007.65 In 2001, in an effort to consolidate the separate efforts by the FCO, DFID and the MoD in conflict prevention activities, the UK created the Africa Conflict Prevention Pool.66 The UN has also played a considerable role in Sierra Leone. At its peak, the UNAMSIL peacekeeping force numbered 17,300 troops; it maintained the peace after the end of the civil war in 2002. In compliance with recommendations not to exit immediately after the holding of elections, UNAMSIL forces gradually withdrew, and disengaged completely by December 2005. Notwithstanding the major role played by the UNAMSIL force in maintaining a stable security environment, British involvement in Sierra Leone has probably had the most significant impact on the country’s security sector. In addition to the training of government armed forces through IMATT, the British government has committed itself to an ‘over-thehorizon’ guarantee to intervene within 48–72 hours in the event that sustained fighting were to break out. Case Studies 85 As an essential tool in addressing the legacies of the past, Sierra Leone’s DDR programme, after nearly five years of operation, was considered by some observers to have been relatively successful.67 Others, such as the former head of the UNDP in Sierra Leone, Alan Doss, were less optimistic. Doss pointed out that the former combatants ‘have joined another, a larger army of young people who are seeking gainful employment… [they] must now compete with all other citizens for any assistance or support provided by the government’.68 In a report on the state of Sierra Leone’s security and governance, the International Crisis Group (ICG) highlighted that many ex-combatants were dissatisfied with the reintegration programme and their prospects for employment in the formal economy remained bleak.69 Nearly four years after the ICG report, the situation remains unchanged, with numbers of former combatants battling with unemployment.70 Upon completion of Sierra Leone’s National Committee for Disarmament, Demobilization and Reintegration programme for more than 70,000 civil war combatants in 2004, UNAMSIL handed primary responsibility for security in the capital over to the local police and armed forces. The reintegration phase comprised a choice of vocational training, formal schooling or tools for various trades as well as farming.71 As another urgent objective towards effective SSR in the post-conflict context, there have been efforts to address the issues of transitional justice and reconciliation in Sierra Leone, such as the establishment of the Truth and Reconciliation Commission as stipulated in the Lomé peace agreement.72 In addition to the TRC, the UN-backed Special Court for Sierra Leone (SCSL) was established to try ‘those who bear the greatest responsibility’ for the country’s civil war.73 The concurrent operation of these two very different transitional justice mechanisms proved to be problematic, particularly as there were no rules regulating their interactions with regard to key concerns such as information-sharing. Although just under 8,000 statements were collected during hearings carried out in 2003–2004, the majority of the population did not participate in this process for a variety of reasons, the most prominent being that they favoured ‘social forgetting’ as opposed to dealing with the past violence – reproducing, denouncing and overcoming – through public speech.74 In addition, civil society and human rights groups criticised the government’s subsequent white paper on the Truth and Reconciliation Project for not committing it to any concrete steps on the issues raised in the report, such as the need to improve governance and tackle corruption.75 86 Judy Smith-Höhn The work of the SCSL has become increasingly controversial. Not only do some local media brand it as ‘white man’s justice’, but the ordinary population also view it as a process that has no impact on their lives, and understandably so.76 Indeed, in its final report in 2004 the TRC questioned the SCSL’s contribution to lasting peace in Sierra Leone: The international community has signalled to combatants in future wars that peace agreements containing amnesty clauses ought not to be trusted and, in so doing, has undermined the legitimacy of such national and regional peace initiatives.77 International actors spearheaded both processes, most prominently among them the UN agencies. In the case of the TRC there was more local involvement through civil society groups, such as the Freetown-based Campaign for Good Governance. 4.3 Comparison of cases 4.3.1 Assessment of SSR: Within-unit comparison of Liberia As elaborated in Chapter 3, one level of comparison will be within-unit, and this section will draw a binary comparison of Liberia’s security sector just before the end of the civil war in 2003 with the period 2003–2006 during the so-called peacebuilding phase, enabling a first assessment of the efficacy of these sector reform efforts. This section begins with a focus on the first of the four criteria for determining effective SSR, namely the existence of security actors that are providing security. The question as to whether the overall security situation has improved will be addressed, as this gives an indication of the performance of the security actors during the reform phase. In other words, is the presence of security actors in fact leading to a positive or improved perception of the general security situation? Undoubtedly, the mere presence of security actors does not necessarily imply that the security situation is favourable and conducive to effective reform. Once a general conception of the security situation and the actors operating within this arena has been established, the focus can be directed towards assessing whether the contextspecific characteristics of the particular post-conflict environment have been taken into consideration by the various stakeholders implementing the Case Studies 87 reforms. The question as to whether the legacies of the past have been addressed concludes this section. Security and security actors in Liberia: Changes over time Has the security situation improved since the end of the civil war, and if so which actors have been key players in that improvement? The survey poll provides some answers. One of the goals of the survey was to determine whether the Liberian population consider the country to be safe at all. Respondents were asked to rate the extent to which they considered it to be safe.78 The rating was made on a four-point scale, thus compelling respondents to make either a negative or positive assessment of the current security situation. Results were generally positive, with more than 60 per cent of respondents rating the situation as okay or very safe. Just over a third felt differently; and 10.5 per cent still thought the country was not safe at all. To ascertain whether there was a variance in perceptions of personal versus general security, respondents were asked to rate their personal security according to the same four-point ranking, as shown in Table 4.1. The results reveal that personal security is rated slightly more positively than general security, yet the overall tendency is similar. Most Liberians have a positive outlook on the overall security situation. Table 4.1 General and personal security in urban Liberia General security (%)* Personal security (%)** Very safe 20.3 25.6 It’s okay 40.1 43.7 Fairly unsafe 26.9 23.5 Not safe at all 10.5 6.2 No answer/don't know 2.2 1.0 N = 698 * Question: ‘In your opinion, how safe is the country today?’ ** Question: ‘How safe do you feel personally in the country today?’ Judy Smith-Höhn 88 The poll also revealed whether perceptions differed among various groups within Liberian society; in other words, whether socio-economic factors influenced the responses, e.g. whether women felt less secure than men, or whether the 10.5 per cent of respondents who did not feel safe at all belonged to a particular ethnic group. The correlation of security perceptions according to gender is shown in Table 4.2.79 Table 4.2 General and personal security in urban Liberia – Gender General security (%)* Personal security (%)** Female Male Female Male Very safe 19.0 22.5 25.4 26.5 It’s okay 39.2 42.8 43.0 44.7 Fairly unsafe 29.8 25.0 24.6 23.3 Not safe at all 11.9 9.7 7.0 5.6 General security: N = 672 (female = 352; male = 320) Personal security: N = 680 (female = 358; male = 322) It appears that, in this case, gender did not influence responses to both questions. This would indicate that, taken from these results alone, security and insecurity do not have a gender dimension in Liberia as both men and women feel equally secure or insecure in their respective environments. This assumption needs qualification, which can be provided by the results from the focus group discussions, particularly that conducted with the group of market women. A comparison of security perceptions and ethnic group produced slightly different results. Members of certain ethnic groups viewed both overall and personal security situations in a more negative light than other groups (Table 4.3). However, the validity of these results is diminished by the fact that in some instances only a handful of respondents were found to belong to one of the 18 ethnic groups listed in the questionnaire. In the case of the Mandingo and Americo-Liberians, where the total numbered less than 20 individuals, the relative percentages can hardly be considered representative. During the planning phase the research team had made particular provisions to enable a Case Studies Table 4.3 89 General security in urban Liberia – Ethnic group Not safe at all Fairly unsafe It’s okay Very safe Total Count % Count % Count % Count % Count % Tribe Kpelle 8 9.3 34 39.5 31 36.0 13 15.1 86 100 Bassa 14 14.4 28 28.9 44 45.4 11 11.3 97 100 Gio 1 3.7 13 48.1 8 29.6 5 18.5 27 100 Kru 14 17.1 25 30.5 31 37.8 12 14.6 82 100 Grebo 5 7.1 21 30 32 45.7 12 17.1 70 100 Mano 7 24.1 6 20.7 8 27.6 8 27.6 29 100 Krahn 3 11.1 8 29.6 9 33.3 7 25.9 27 100 Gola 6 12.2 12 24.5 20 40.8 11 22.4 49 100 Gbandi 4 11.4 4 11.4 20 57.1 7 20.0 35 100 Loma 5 14.7 7 20.6 15 44.1 7 20.6 34 100 Kissi 2 5.1 13 33.3 15 38.5 9 23.1 39 100 Vai 1 2.4 5 12.2 22 53.7 13 31.7 41 100 Mandingo 0 0.0 2 18.2 3 27.3 6 54.5 11 100 Americo0 0.0 0 0.0 3 75.0 1 25.0 4 100 Liberians Congo 2 9.5 5 23.8 4 19.0 10 47.6 21 100 Other 1 3.3 4 13.3 15 50.0 10 33.3 30 100 Total 73 10.7 187 27.4 280 41.1 142 20.8 682 100 N = 682 Question: ‘In your opinion, how safe is the country today?’ focus on the Mandingo ethnic group, who appeared to be generally ostracised from the rest of Liberian society. One of the suburbs pre-selected to implement the survey was known as being populated by this particular group and it was expected that the sample of Mandingo respondents should total approximately 100. Despite these measures, however, only 11 of the altogether 700 respondents were listed as Mandingo. Perhaps the research team underestimated the extent of the isolation of this particular group in that it required interviewers, none of whom was Mandingo, to conduct these interviews. Judy Smith-Höhn 90 Given the fact that during the civil war, the intensity of the war and the extent of the damage it inflicted on each city and its inhabitants varied across all three cities, there was a need to determine whether the experiences and perceptions of security differed across these cities (Table 4.4). Table 4.4 General security in three urban centres of Liberia Public security (%) Buchanan Monrovia Tubmanburg Total Very safe 14.0 20.1 28.0 20.3 It’s okay 44.0 38.4 45.0 40.1 Fairly unsafe 28.0 27.9 21.0 26.9 Not safe at all 14.0 11.2 3.0 10.5 No answer/don’t know 0.0 2.4 3.0 2.2 N = 698 Question: ‘In your opinion, how safe is the country today?’ A significant variance in the results of each city would perhaps give an indication of the relevance of former violence actors (‘Gewaltakteure’), since each was a stronghold for one of the warring factions: the port city of Buchanan was seized by MODEL in July 2003, while Tubmanburg was a stronghold of the rebel faction LURD during the final throes of the civil war in 2003.80 The comparison of the relative frequencies obtained for this question per city revealed that perceptions of general and personal security were much the same for Buchanan and Monrovia, while Tubmanburg residents had a decidedly more positive perception of the general security situation. In order to ascertain more accurately the impact – whether positive or negative – of the presence and activities of these former warring factions on these communities, one must also factor in the extent to which the respective residents consider the general and personal security situation to have improved or worsened since the end of the civil war in 2003. Table 4.5 illustrates the results for this particular question. Case Studies Table 4.5 91 General security in three urban centres of Liberia – Past and present compared Public security (%) Buchanan Monrovia Tubmanburg Total Much better 36.0 32.3 38.0 33.7 Somewhat better 57.0 51.0 47.0 51.3 No change 6.0 5.6 10.0 6.3 Worse 1.0 5.0 5.0 4.4 Much worse 0.0 3.6 0.0 2.6 No answer 0.0 2.4 0.0 1.7 N = 702 Question: ‘In your opinion, compared to before the end of the last war, what is the general state of security like in the country today?’ Two inferences can be drawn from these results. i) ii) Residents of Tubmanburg have a markedly more positive perception of the current security situation than those living in the other two cities (compare results in Table 4.4.). In the comparison of past and present security situations, there was little deviation in the results, illustrating that the withdrawal of the various armed factions perhaps had the same positive effect in all three instances. The positive perception of Tubmanburg respondents (Table 4.4) is partly clarified when one considers the historical context. During the civil war the LURD rebel movement was considered to have established a degree of stability in its stronghold, whereas MODEL was far less organised and more brutal.81 What one may be witnessing is the after-effects of the more positive role played by the LURD insurgents in Tubmanburg; 10 per cent even considered their security situation to have remained the same since the end of the civil war. It is therefore logical that Buchanan respondents considered their security situation to have improved most since the end of the war (93 per cent much/somewhat better) following MODEL’s departure. 92 Judy Smith-Höhn These results are, however, inconclusive. If the survey poll had an added question relating to the various rebel factions specifically, it would have been possible to determine more accurately their respective impact on these three cities. Thus, according to the survey poll results, LURD appears to have had a conspicuously positive impact on the perception of the residents of Tubmanburg. A juxtaposition of these results with those from the four focus group discussions serves two purposes. First, it enables the verification of the survey results with those from the focus groups – in this case, whether discussants generally have a positive perception of the security situation. Second, it facilitates a more in-depth enquiry into the particular reasons for why general and personal security were rated as they were. The focus groups confirmed the overall positive perception of the security situation. Most discussants were of the opinion that Liberia had become a relatively safe country to live in; they were generally more optimistic about their personal security than the overall state of security in the country. Undoubtedly, it was the presence of the large international peacekeeping force, UNMIL, which contributed greatly to this favourable perception, as one discussant stated: ‘UNMIL is here to bring peace’.82 However, discussants all agreed that the national security situation remained unstable, a view that partly reflects the more negative assessments of just over a third of all survey poll respondents (as shown in Table 4.1).83 Both international and local experts interviewed confirmed that despite the large international presence and a by-and-large positive outlook, there remained broad scepticism that there was still potential for large-scale violence. An UNMIL civil affairs officer maintained that the security situation had improved drastically, as evidenced by the reduction of UN highway checkpoints throughout the country: ‘The security situation is very much improved… From Buchanan to Greenville there is no UN checkpoint because things are safe.’84 Yet he conceded that were ‘UNMIL… to leave immediately, the result would be chaos’.85 This points to the significance of the international peacekeeping force in the provision of security in Liberia. As for other security actors and their ability to provide security, the quantitative and qualitative data present a wealth of empirical evidence that gives insights into the situation on the ground and the efficacy of SSR strategies implemented. A first indication of the relevant security actors and their relative importance is shown in Table 4.6. Survey respondents were asked to specify Case Studies Table 4.6 93 Security actors in urban Liberia – Most important actors Actors % UNMIL 75.9 Liberia National Police (LNP) 17.9 Armed Forces of Liberia (AFL) 2.0 Ex-combatants 0.1 ECOWAS 0.1 Poro/Sandee 0.1 Vigilante teams 0.1 Family 0.1 No answer/don’t know 3.4 N = 698 Question: ‘We have spoken to many Liberians and they all have different feelings about which groups in the country protect them and which are a threat to them. Could you please tell us which group is the most important one for your personal safety?’ which actor they considered to be the most important provider for their personal security.86 It comes as no surprise that most respondents afforded UNMIL this key function: 75.9 per cent of respondents consider the international actor to play a crucial role in guaranteeing security. The second most frequently named organisation was the LNP (18.4 per cent), followed – by a considerable margin – by the AFL. Other actors were cited far less frequently (in each case less than 1 per cent of responses). However, the fact that these are not considered the most important actors does not imply that they are insignificant (see below). Because the research was limited to urban areas, these figures reflect the perceptions of Liberia's urban population. However, people who had migrated from the hinterland participated in the focus groups, enabling the research team to gain some impressions of conditions in the rural areas. The general consensus on the importance of the UN troops among participants in the discussions suggests that, in this respect at least, there is no clear rural-urban divide. The figures in Table 4.6 are based on an open-ended question; the survey question was designed in such a way as to encourage respondents to Judy Smith-Höhn 94 name the most important group or set of actors in each case without being influenced or guided towards any particular answer. If respondents had been asked to choose from a list of specific actors, one may have unwittingly excluded certain actors who, from the respondents’ perspective, play a key role in security. These results were tested with a second, closed question: through multiple-choice options, respondents were asked to rate each listed actor according to its level of importance in their security. Table 4.7 thus provides a more detailed illustration of a number of relevant security actors, with each rated according to its impact – both positive and negative – on Table 4.7 Security actors in urban Liberia (%) Very/ somewhat important to personal security Types of actors International State International/ commercial non-state Domestic nonstate Does not affect my personal security Somewhat/ a big threat to personal security UNMIL 94.9 2.0 1.4 LNP 91.9 4.6 2.9 AFL 64.8 23.8 9.3 Private security companies 38.3 52.2 3.8 Community watch teams 56.6 17.9 15.5 Poro/secret societies 15.7 46.4 26.1 Political party militias 7.2 36.3 42.8 Street boys 0.4 6.3 78.6 Ex-combatants 3.3 6.9 86.8 N = 698; percentages of ‘don’t know’ and ‘no answer’ responses are not shown. Question: ‘Now, I’m going to read out a list of groups which are said to affect security in one way or another. For each one, could you tell me whether you feel protected or threatened by them or not?’ Case Studies 95 security. Little difference was found between the two sets of answers as regards the relevance of the international and state security actors. Again, UNMIL received the most positive rating. Yet when one considers the fact that, at the time of survey, it was the only viable, armed security actor around, a rating of around 75 per cent of respondents claiming UNMIL to be the most important security provider (Table 4.6) is no longer as positive. Some results from the focus group discussions underline this point. Though three of the four groups rated UNMIL as being positive, one group (youth and community leaders) noted it as being neutral/disputed, signifying that it was not always considered as contributing towards security, particularly on the community level, where it was said to have little or no presence.87 In the Liberian capital Monrovia, for example, UNMIL troops essentially patrol the main thoroughfares and show little if any presence in outlying areas of the city. The results for UNMIL may be even more dramatic in rural areas, where there is hardly any UN presence. Hence, although there is no doubt that the peacekeepers are perceived to be the main guarantor of security, there is certainly still scope for improving their public image. Certainly one of the more remarkable findings was the relatively favourable perception of the state security actors – the LNP and the AFL – given their current reorganisation and history of violence against citizens in the past. The fact that the AFL was regarded as an important security actor suggests that people’s perceptions are coloured by wishful thinking. The AFL was dissolved after the civil war and, at the time of questioning, was in fact non-existent. Conversely, when asked which group they felt to be the biggest threat to them personally, 1.1 per cent of respondents named the stilldefunct AFL. This may have been a reference to former members of the armed forces or deserters, i.e. ex-combatants, given that 59.5 per cent of respondents regard this latter group as the biggest threat to their personal security (Table 4.8). Private security companies (PSCs) were generally awarded a relatively positive rating of 38.3 per cent (Table 4.7). More importantly, a majority of Liberians (52.2 per cent) felt unaffected by PSCs, which points to the fact that they provide security mainly for wealthier segments of society. The perception that political party militias pose a threat (‘is a big/somewhat a threat to my personal security’) is partly due to the elections which took place in October and November 2005. Still, only 7.2 per cent 96 Table 4.8 Judy Smith-Höhn Security actors in urban Liberia – Biggest threat Actors Ex-combatants Street boys Criminals (armed robbers) LNP Political party militias AFL UNMIL Poro/Sandee (secret societies) Other None No answer/don’t know % 59.5 7.9 3.7 2.0 1.9 1.1 0.9 0.4 2.0 4.6 16.0 N = 698 Question: ‘Which group do you feel is the biggest threat to you personally?’ thought they posed a threat, and only 1.9 per cent considered them to be the biggest threat to security! In addition to highlighting the significance of the international intervening actor UNMIL and the state security actors LNP and AFL, Tables 4.6 and 4.7 show that Liberia’s current security architecture comprises three types of actors: external (such as UNMIL); state (such as the local police forces or the army); and non-state (such as community-based organisations or traditional security institutions). This distinction is crucial, as it is the second type, i.e. state actors, which receive the most attention in the implementation of reform efforts. Non-state actors, be they international or domestic, although acknowledged, are not incorporated into any formal reform framework, leading one to question whether the second criterion for determining the efficacy of SSR strategies has indeed been fulfilled; in other words, whether the context-specific characteristics, procedures and interactions of the security sector have been adequately taken into account by the implementing agencies. In addition to highlighting the significance of the international intervening actor UNMIL and the state security actors LNP and AFL, Tables 4.6 and 4.7 show that Liberia’s current security architecture comprises three Case Studies 97 types of actors: external (such as UNMIL); state (such as the local police forces or the army); and non-state (such as community-based organisations or traditional security institutions). This distinction is crucial, as it is the second type, i.e. state actors, which receive the most attention in the implementation of reform efforts. Non-state actors, be they international or domestic, although acknowledged, are not incorporated into any formal reform framework, leading one to question whether the second criterion for determining the efficacy of SSR strategies has indeed been fulfilled; in other words, whether the context-specific characteristics, procedures and interactions of the security sector have been adequately taken into account by the implementing agencies. Context-specific characteristics of Liberia’s post-conflict security environment Both the focus groups and the expert interviews provide more insights into the perceived roles and significance of each of the non-state actors mentioned above. One of the key findings from the focus groups was that there appeared to be two levels at which particular actors were expected to provide security, namely the national and the local. At the national level, UNMIL, the LNP and other state actors were generally expected to provide for security; at the local level, many agreed that the individual should provide for his or her own security: ‘You yourself provide security… if there is small trouble.’88 The extent to which this varied focus is significant for SSR policies will be dealt with in section 4.3.5. Results from the focus groups also revealed the importance of community-based informal actors in the security arena. Figure 4.1 on page 98 depicts one of the mapping exercises which was conducted in each of the focus groups. As mentioned in Chapter 3, the mapping exercise was one of the key outcomes of the focus groups. Discussants were asked to identify all relevant security actors, determine their relative significance and illustrate the interactions among them within the security arena. Figure 4.1 confirms the survey poll results, namely that, to ordinary Liberians, a variety of actors – beyond the state and external ones – play a role in the security arena in Liberia today. Two informal actors are particularly noteworthy in this regard: the community watch teams (CWTs) and the secret societies, known as the Poro and Sandee. Judy Smith-Höhn 98 Figure 4.1 Liberia today – Security actors and their interactions EXTERNAL STATE NON-STATE Bureau of Immigration and Naturalisation Ministry of National Security UNMIL Community watch teams Poro/Sandee National Security Agency National Bureau of Investigation Ministry of Defence Excombatants Liberia National Police Action agents Special Security Service National Fire Service Private security companies Legend positive cooperation neutral/disputed rivalry negative rivalry within Source: FGD1 Liberia, Teachers and Students. Copyright: Judy Smith-Höhn Case Studies 99 The secret societies tend to play a negligible role in urban areas, and to some extent they are viewed as a substantial threat (as shown in Table 4.8); but above all they are decoupled from the other actors. This was found in all four focus groups conducted in Liberia. Thus, regardless of whether the perception of the secret societies was favourable, uncertain or disapproving, they were consistently isolated from all other actors.89 Interviews confirmed this disconnect, with several local experts acknowledging their importance, yet maintaining their detachment from other actors.90 Admittedly, rural dwellers assign secret societies a central role in active conflict resolution, with Poro hierarchies continuing to provide institutional responses to armed violence in some communities.91 However, the absence of lines of interaction with other security actors indicates that their involvement in the development of SSR strategies would very likely be problematic. Nonetheless, a context-specific SSR strategy must not only identify those actors that need to be incorporated into reform efforts, but should also be aware of those that should be excluded, and for what reason. Interviews with staff from the UN mission in Liberia revealed that they attach little overall significance to local conditions and actors. Social and political mechanisms, capacities and conflicts at local level are largely ignored. Since a temporarily deployed UN mission or the newly trained national police and armed forces cannot be expected to provide security in the medium term, SSR can only work effectively if local conditions are taken into account. An interview conducted with a senior UNMIL staff member who was partly responsible for regularly assessing the security situation in the country points to a cause for concern, as the interviewee had not heard of the community-based secret societies at all.92 Here, clearly, there is an increased risk of developing inappropriate strategies on the basis of ill-informed analyses. The role of CWTs is more ambiguous. According to Table 4.7, more than half of respondents consider the CWTs to be very or somewhat important for their security (56 per cent). Similarly, focus group discussants considered this actor to play a central role in providing security, particularly at the community level, in areas where UNMIL does not patrol. However, perceptions weren’t entirely positive, as shown by the 15.5 per cent who considered CWTs to threaten their security. Again, the focus groups elucidate this negative result. CWTs were associated with the increased practice of mob justice, which most discussants did not tolerate and which called into question the legitimacy of such community-run security provision. However, the objection to mob justice was not unanimous, as several informal discussions and interviews revealed.93 That many Liberians Judy Smith-Höhn 100 considered it the only alternative to dealing with criminal threats illustrates the lack of trust in state institutions, both the police and the judiciary, to punish offenders. The perceived inefficiency of state institutions to perform their given functions and the continued concern over the activities of former combatants inevitably lead one to question whether the legacies of the past have indeed been adequately addressed – the third criterion for measuring effective SSR. Dealing with legacies of the past? Although the same survey was not conducted in 2003 prior to the signing of the CPA, the research team attempted to grasp the dynamics of the security situation over time by asking respondents to compare the current situation with the security conditions before the final peace agreement was signed. The results revealed a significant shift in perceptions of both the country’s security and the level of personal safety (Table 4.9). However, only 28.4 per cent of the respondents felt that their personal security situation had improved significantly since the end of the last war. This certainly falls short of the expectations that some UN administrators may have about the impact of the UNMIL mission. Table 4.9 Changes in general and personal security in urban Liberia – Past versus present Much better Somewhat better No change Worse Much worse No answer Don’t know General security (%)* 33.7 51.3 6.3 4.4 2.6 1.7 0.0 Personal security (%)** 28.4 57.7 5.6 4.6 2.4 1.3 0.0 N = 702 *Question: ‘In your opinion, how safe is the country today?’ **Question: ‘How safe do you feel personally in the country today?’ Case Studies 101 Focus groups substantiated the generally positive assessment of the current security situation, both general and personal – certainly an expected outcome in a country that only recently emerged from a protracted civil war. More interestingly, discussants pointed out that though the security situation had improved dramatically, as there were now hardly any reports of outbreaks of violence, especially in the capital, the basic living conditions had not changed for many ordinary Liberians. Whereas during the war individuals could not gain a livelihood due to the constant fighting, they were now prevented from fending for themselves by to a lack of opportunity. The issue of socio-economic marginalisation was raised repeatedly in all focus groups. The marginalisation of one particular group within society, namely the large group of (mostly male) former combatants, is particularly worrisome. In this regard, the effective implementation of DDR programmes, understood as part of the wider SSR agenda, is a key prerequisite to sustainable reform. Despite the significance of the DDR programme in stabilising Liberia after years of civil war, it had some major flaws. Although it claims to have disarmed 100,000 combatants, less than 28,000 guns were collected.94 Having extended the qualification criteria for admission to the DDR programme to include not only those who could produce a weapon but also those who could present 150 rounds of ammunition, many took advantage of the opportunity. For instance, former commanders provided non-combatants with ammunition in order to receive a percentage of their stipend. At times, women and children were not even required to surrender arms or ammunition, making it easier for rebel commanders to recruit youths from communities. Some sources rated this ‘commercialisation’ of the DDR process to have accounted for as much as 60 per cent of the total caseload of candidates.95 This may mean that officially disbanded factions have retained arms. In addition, many demobilised fighters have not been given adequate and timely reintegration opportunities, the lack of which may be an incentive for some to move to other hotspots in the region (e.g. Côte d'Ivoire) to reenlist. Despite these shortcomings, there has been a significant alteration in the security arena since the end of the civil war. The fact that not one of the former warring factions, for example, was considered a viable security actor or threat shows that the dominant conflict narrative has at least shifted. As with the survey respondents, focus group discussants were asked to compare the current security situation with conditions prior to the end of the civil war in 2003. In contrast to the survey poll, attention was centred less on an Judy Smith-Höhn 102 Figure 4.2 Liberia 2003 – Security actors and their interactions Vigilante Poro/Sandee MODEL LURD Government of Liberia/militia Anti-Terrorist Unit Special Operations Division Legend positive neutral/disputed negative rivalry rivalry all around Source: FGD1 Liberia, Teachers and Students. Copyright: Judy Smith-Höhn Case Studies 103 overall assessment of the security situation and more on identifying the relevant actors, the objective here being clarifying the dynamics within the security arena, particularly between the various actors. In all four focus groups, results were similar. The actors that comprised the so-called security arena prior to the end of the civil war – in this case, the actors that possessed the capacity to use (and abuse) force – were, for the most part, no longer part of the security architecture.96 In Liberia, vigilante is the term generally used for community-based self-help groups, hence its positive connotations.97 During the Taylor regime, the Anti-Terrorist Unit functioned as Taylor’s private security guards. Figure 4.2 on page 102 represents the security arena pre-2003, as identified by one of the focus groups. When compared with Figure 4.1, it demonstrates that, although groups of negative actors are currently still present, the conflict narrative has changed to such an extent that the former warring factions are no longer organised into distinct groups, but have mostly been subsumed under the unified heading of ex-combatants. Yet another legacy of the past – or more precisely, several legacies – is the continued existence of a large number of state security institutions, which poses a challenge. In an assessment of these agencies and the existing legislation that guides their roles and functions, Jaye points to the myriad of problems faced by this sector. In addition to overlapping responsibilities and budgetary constraints, this bloated sector struggles with issues of politicisation, a lack of infrastructure, poor human resource capacity and skewed civilian oversight, not to mention the struggle for local ownership of the SSR process.98 Rather than addressing the issue directly, the current SSR strategy has instead restricted its focus to certain institutions within the state security sector. As mentioned in section 4.1.2, there has been a narrowed focus on the AFL, LNP and Special Security Services, while the other security agencies have to all intents and purposes been sidelined.99 Most of the agencies were mentioned during the focus groups, yet for the most part discussants were vague as to the exact role these various institutions currently played and were expected to play. During one of the focus groups, for example, the Bureau of Immigration and Naturalisation was initially noted as being functional, yet in the course of discussion it was acknowledge that this particular actor was, in fact, not performing its intended role. Although the actor was formally still operational, it was not providing any form of security and was therefore rated negatively, with one discussant justifying the assessment thus: ‘if they were working effectively, the kind of troublesome people that are allowed in this country would not be allowed to enter’.100 104 Judy Smith-Höhn The within-unit comparison of the security arena in Liberia, particularly the actors involved, can be summarised as follows. There is currently only one actor that can assure internal security in Liberia, and that is UNMIL. As this is a key state function, the task will ultimately be handed over to the newly formed AFL. Another characteristic feature of this post-conflict phase is the involvement of international actors in policing, notably in the delivery of training for the LNP. Moreover, since the police service currently remains ill-equipped and short-staffed, the local police forces lack the capacity to provide security without external support. This gap is filled, to a certain extent and only in some districts, by civil defence forces known as CWTs, which take on the responsibility of protecting their local community. In addition, citizens with the necessary financial resources can engage PSCs to protect their property and business premises. The picture is complicated further by the fact that the UN mission hires private security services to protect its premises, and these companies sometimes employ ex-combatants on low wages. On the one hand, this creates employment opportunities for potential troublemakers which would not be available if the mission withdrew altogether; on the other hand, it again reflects the lack of confidence in the state's own security forces. Although the activities of these companies should not be regarded as entirely unproblematic or necessarily positive, they are nonetheless a key actor in providing security in Liberia. 4.3.2 Assessment of SSR: Within-unit comparison of Sierra Leone The within-unit comparison of issues related to SSR Sierra Leone will follow the same course as the previous comparison. Thus the section will draw a binary comparison between Sierra Leone’s security sector just before the end of the civil war in 2001/2002 and the peacebuilding phase from 2003 to 2006. This section will first deliberate on the effective provision of security by the various actors, then focus on the context-specific characteristics of this particular post-conflict environment. Finally, attention is directed to the causes of the conflict and whether these have been taken into account by the various implementing actors. Security and security actors in Sierra Leone: Changes over time As in the case of Liberia, the survey conducted in Sierra Leone illustrated the extent to which the population deemed the general and personal security Case Studies 105 situations to be favourable. To ascertain whether there was a variance in the perceptions of personal versus general security, respondents were asked to rate their security according to the same four-point ranking as used in interviews in Liberia. The results revealed that personal security was rated more positively than general security. In fact, a slight majority of respondents considered the general security situation to be insecure, with many regarding the situation to be ‘fairly unsafe’ (37.9 per cent). Personal security, on the other hand, was rated far more positively, with almost twothirds of respondents answering accordingly (61.6 per cent). Thus, results show that, although both general and personal security have improved substantially in Sierra Leone since the peace agreement was signed in 2002, the security situation remains fragile (Table 4.10). Table 4.10 General and personal security in urban Sierra Leone General security (%)* Personal security (%)** Very safe 14.7 20.7 It’s okay 32.3 40.9 Fairly unsafe 37.9 27.5 Not safe at all 13.4 10.3 No answer/ don't know 1.7 0.7 N = 702 *Question: ‘In your opinion, how safe is the country today?’ **Question: ‘How safe do you feel personally in the country today?’ As in Liberia, there appears to be no significant difference between the assessment made by men and women as regards both overall and personal security situations (Table 4.11). In their overall assessment of the security situation, the focus groups confirmed the general scepticism among Sierra Leoneans, with discussants agreeing that although there had been no major incidents of unrest since the end of the last war, it remained an uneasy peace. The prospect of elections scheduled for April 2007 was noted as particularly worrisome.101 A cross-tabulation of the results for the question on respondents’ views on public security according to religious affiliation is shown in Table 4.12. Judy Smith-Höhn 106 Table 4.11 General and personal security in urban Sierra Leone – Gender General security (%) Personal security (%) Female Male Total* Female Male Total* Very safe 15.2 14.5 14.8 21.4 20.1 20.7 It’s okay 32.1 33.8 33.0 43.7 39.0 41.3 Fairly unsafe 40.5 36.6 38.5 27.3 28.0 27.6 Not safe at all 12.2 15.1 13.7 7.6 13.0 10.4 General security: N = 688 (female 336; male 352) Question: ‘In your opinion, how safe is the country today?’ Personal security: N = 695 (female 342; male 354) Question: ‘How safe do you feel personally in the country today?’ *No answer/don’t know are not included, therefore figures differ from those in Table 4.8. Table 4.12 General security in urban Sierra Leone – Religion General security (%)* Religion Christian Muslim Traditional Other Total* Count % Count % Count % Count % Count % Very safe 40 14.2 58 15.1 1 16.7 3 20.0 102 14.9 It’s okay 89 31.7 133 34.7 1 16.7 4 26.7 227 33.1 Fairly unsafe 119 42.3 138 36.0 2 33.3 5 33.3 264 38.5 Not safe at all 33 11.7 54 14.1 2 33.3 3 20.0 92 13.4 N = 685 Question: ‘In your opinion, how safe is the country today?’ *No answer/don’t know not included. As the total number of respondents who professed to belong to a traditional or other religion is rather small, the results cannot be considered as representative of either of these two groups. In the case of the Christian and Muslim respondents, the figures are substantial enough to be considered Case Studies 107 representative. According to Table 4.12, there is insufficient difference between the ratings given by both groups to presume that there is any correlation between perceptions of security and religious affiliation. According to cross-tabulations of the results on the question of general security in proportion to membership of a particular ethnic group, one of the two largest groups in the country, namely the Mende, gave a generally more positive assessment of the current security situation. Half of all Mende respondents rated public security as okay or very safe, while around 44 per cent of Temne respondents thought likewise and just over 40 per cent considered the country to be still a fairly unsafe place to live. This minor deviation notwithstanding, there appears to be no other significant correlation between perceptions of security and ethnic group; no particular group within society felt more threatened than another. But as in Liberia, this inference cannot be verified as the sampled number of members of most of the groups is simply too small to guarantee that they are representative (Table 4.13). Table 4.13 General security in urban Sierra Leone – Ethnic group Not safe Fairly It’s okay Very safe Total at all unsafe Tribe Count % Count % Count % Count % Count % Temne 31 14.8 85 40.7 70 33.5 23 11.0 209 100 Mende 18 15.8 39 34.2 35 30.7 22 19.3 114 100 Kono 9 13.4 22 32.8 23 34.3 13 19.4 67 100 Creole 6 10.9 18 32.7 22 40.0 9 16.4 55 100 Limba 8 16.0 23 46.0 13 26.0 6 12.0 50 100 Fula 6 12.2 15 30.6 20 40.8 8 16.3 49 100 Mandingo 4 14.3 9 32.1 11 39.3 4 14.3 28 100 Susu 0 0.0 14 50.0 9 32.1 5 17.9 28 100 Sherbro 1 4.5 14 63.6 5 22.7 2 9.1 22 100 Koranko 3 14.3 10 47.6 5 23.8 3 14.3 21 100 Loko 5 25.0 6 30.0 6 30.0 3 16.0 20 100 Other 3 11.1 11 40.7 8 29.6 5 18.6 27 100 Total 94 13.6 266 38.6 227 32.9 103 14.9 690 100 N = 690 Question: ‘In your opinion, how safe is the country today?’ 108 Judy Smith-Höhn When categorising the assessment of the security situation according to the cities in which the survey was conducted, a marked difference is discernible in Koidu, where respondents were decidedly more optimistic (Table 4.14). Table 4.14 General security in three urban centres of Sierra Leone Very safe It’s okay Fairly unsafe Not safe at all No answer Freetown 12.8 32.3 38.3 14.2 2.4 Public security (%) Makeni Koidu 12.0 26.7 29.0 35.6 45.0 28.7 14.0 8.9 0.0 0.0 Total 14.7 32.3 37.9 13.4 1.7 N = 702 (Freetown 501; Makeni 100; Koidu 101) Question: ‘In your opinion, how safe is the country today?’ A look at the ratings for personal security produced similar results (Table 4.15). Table 4.15 Personal security in three urban centres of Sierra Leone Very safe It’s okay Fairly unsafe Not safe at all No answer Freetown 20.4 39.1 27.5 12.0 1.0 Personal security (%) Makeni Koidu 14.0 28.7 50.0 40.6 31.0 23.8 5.0 6.9 0.0 0.0 Total 20.7 40.9 27.5 10.3 0.7 N = 702 (Freetown 501; Makeni 100; Koidu 101) Question: ‘How safe do you feel personally in the country today?’ In all three cities, personal security is rated more positively than general security, thus not refuting the overall assessment of personal versus general security mentioned above (compare Tables 4.15 and 4.16). As with general security, the citizens of Koidu were more approving of their personal Case Studies 109 security situation, with just over two-thirds considering it to be okay or very safe. There may be a number of reasons for this disparity, but a consideration of the context may provide a partial explanation. During the civil war the area around Koidu was initially worst hit by insurgents, particularly as the diamond-rich area is the goose that lays Sierra Leone’s proverbial golden egg. Over the years, however, the conflict moved towards the capital, Freetown, and in the last phases of the conflict Freetown was the key target of the RUF fighters and belligerent sobels. Thus the citizens of Freetown and nearby Makeni have a more recent memory of excessive fighting, whereas Koidu citizens have perhaps experienced a longer period of peace and therefore perceive both general and personal security situations to be more positive than their neighbours. This assumption requires testing. However, neither focus groups nor the interviews shed light on the reasons for this discrepancy, which has highlighted a flaw in the research design: in addition to extending the survey poll to two additional urban areas in each country under study, focus groups should have been conducted in each of these other cities as well in order to obtain a more detailed description of the security arena and perceptions thereof. A comparison of perceptions of the current security situation and that before the end of the civil war provides no further clues as regards this discrepancy between the various cities. Consider Tables 4.16 and 4.17. Respondents were asked to evaluate personal and overall security situations today in comparison to the wartime situations. Table 4.16 Changes in general security in urban Sierra Leone: Past and present compared Much better Somewhat better No change Somewhat worse Much worse No answer Freetown 34.7 56.7 4.6 2.2 0.2 1.6 General security (%) Makeni Koidu 21.0 27.7 71.0 65.3 4.0 5.0 3.0 1.0 1.0 1.0 0.0 0.0 Total 31.8 60.0 4.6 2.1 0.4 1.1 N = 702 (Freetown 501; Makeni 100; Koidu 101) Question: ‘In your opinion, compared to before the end of the last war, what is the general state of security like in the country today?’ 110 Judy Smith-Höhn Table 4.17 Changes in personal security in urban Sierra Leone: Past and present compared Much better Somewhat better No change Somewhat worse Much worse No answer Freetown 28.3 61.1 5.0 2.6 2.0 1.0 Personal security (%) Makeni Koidu 22.0 27.7 72.0 66.3 3.0 3.0 3.0 2.0 0.0 1.0 0.0 0.0 Total 27.4 63.4 4.4 2.6 1.6 0.7 N = 702 (Freetown 501; Makeni 100; Koidu 101) Question: ‘Compared to before the end of the last war, how is your personal safety today?’ The ratio of answers is similar in all three cases, thus providing no further evidence to explain the markedly positive perceptions of security of Koidu citizens. Fortunately, there is a third set of data, namely the semistructured interviews. Interviews conducted with experts and key stakeholders in Koidu provide a plausible explanation for the disparity: the exceedingly efficient work being conducted by the city’s police forces.102 The local SLP unit commander of the security division was a well-respected member of society. The SLP division responsible for Tonkoro district, of which Koidu Town forms a part, appears to be better at fulfilling its stated duties than its counterparts in Freetown and Makeni. Regular night patrols are conducted, for example, and disciplinary measures are usually taken against police officers who break the code of conduct (e.g. participating in extortion or unwarranted assault). A more comprehensive overview of Sierra Leone’s current security sector and the relevant actors involved – or expected to be involved – in the provision of security was gained from other questions posed in the survey. The data gathered from the fieldwork revealed the existence and relevance of several actors – particularly of a non-state, informal nature – in providing (or threatening) security. Since the departure of the last contingent of UNAMSIL troops, which during their deployment had provided the bulk of security in urban Sierra Leone, informal actors have come to play a more significant role in the security arena. Given that the fieldwork was conducted shortly after UNAMSIL’s withdrawal, the sustainability of the current post- Case Studies 111 conflict phase cannot be determined. However, the results reveal the significant vacuum left by the multinational force and expose the inability of the national security forces to fill it. Table 4.18 shows that, here, the external actors – UNAMSIL, ECOMOG, IMATT – do not play as significant a role in the provision of Table 4.18 Security actors in urban Sierra Leone: Most important actors Actors RSLAF Police UNAMSIL ECOMOG Government Former warring faction 1: CDF Husband/family/individual God Rebel/ex-combatant President Youth groups Foreign troops (general) Traditional authorities (paramount chief/village authorities) Private security SLP Operational Support Division IMATT Mende Former warring faction 2: RUF Other None No answer/don’t know % 34.5 33.2 10.8 3.0 2.0 1.6 1.0 1.9 0.9 0.9 0.6 0.4 0.3 0.3 0.3 0.1 0.1 0.1 1.0 0.7 7.2 N = 702 Question: ‘We have spoken to many Sierra Leoneans and they all have different feelings about which groups in the country protect them and which are a threat to them. Could you please tell us which group is the most important one for your personal safety?’ Judy Smith-Höhn 112 security as their Liberian counterparts. The Sierra Leone government and its security forces, on the other hand, scored high in respondents’ expectations for the provision of national security. Surprisingly, it was the fact that the RSLAF stayed out of sight and was confined to barracks that accounts for the discussants’ favourable assessment.103 This to some extent explains why more than two-thirds of survey respondents considered the RSLAF as very important for their personal security, despite the violent, even criminal, history of the military in the country. Table 4.19 Security actors in urban Sierra Leone (%) Types of actors State International/ commercial non-state Domestic non-state Very/ somewhat important to personal security SLP 87.8 RSLAF 81.4 Mines monitoring 22.7 officers Traffic wardens 51.2 Private security 54.6 companies Secret societies 26.6 Youth wings of 30.6 political parties Ghetto Boys 4.4 Bike riders 39.1 West Side Boys 2.9 CDF 20.5 RUF 2.3 Does not affect my personal security Somewhat/ a big threat to personal security 3.4 5.0 55.0 8.0 10.0 5.2 34.0 38.6 7.9 1.6 50.9 33.5 19.7 30.3 16.2 30.1 17.0 20.5 16.1 76.8 26.6 74.6 54.1 78.0 N = 700 (out of total sample of 700); percentages of ‘don’t know’ and ‘no answer’ responses not shown. Question: ‘Now, I’m going to read out a list of groups that are said to affect security in one way or another. For each one, could you tell me whether you feel protected or threatened by them?’ Case Studies 113 Focus group discussants had a relatively negative opinion of the SLP.104 This deviates from the positive ratings given to the SLP in the survey poll, where 87.8 per cent of respondents considered the police to be very or somewhat important for their personal security (Table 4.19). This disparity was clarified in interviews conducted with local and international experts, which revealed that, despite their high expectations concerning the role of national security forces such as the police, Sierra Leoneans are very aware of the shortcomings of the SLP in particular. Despite external support and training received from the Commonwealth Police and the civilian police section of UNAMSIL, the SLP’s ability to perform its duties is hampered, among other things, by a lack of equipment and insufficient remuneration, which partly explains its poor performance. Both Tables 4.18 and 4.19 reveal the state actors – namely the RSLAF and the SLP – to be the most important security providers. In a recent survey of policing agencies in Sierra Leone, Bruce Baker identified three types of policing structures: those operated by the Sierra Leonean state, private and community structures that have state approval and those without state authorisation.105 He found that in contrast to widespread opinion, citizens do not always choose state policing over the non-state alternatives; in fact, many consider non-state policing a viable alternative to the often distant and inefficient state police. Though this may reflect the current situation – and the survey results confirm the importance of non-state actors – Sierra Leoneans do indeed expect the state security actors to be in charge of a properly functioning national security apparatus. Baker’s mapping of the security environment (Table 4.20) is somewhat different to the results of the DSF project. Firstly, Baker identified a number of state actors as being part of the security environment, such as the mine monitoring officers and traffic wardens. Secondly, Baker’s list does not reveal the extent to which particular actors are considered important for the provision of security. The added value of the DSF project’s approach is that it provided actual figures as regards the importance of particular actors through the survey as well as enabling a confirmation of these figures through the focus groups. While Baker’s research certainly provided useful groundwork in preparation for the present fieldwork, with all the actors in Table 4.20 having been included in the survey questionnaire, it is prone to misinterpretation. More precisely, certain actors mentioned by Baker were, in fact, a much less prominent feature in the security sector than the table suggests. The mines monitoring officers, for example, were considered to have no impact on security by the majority of survey respondents. Whereas the Judy Smith-Höhn 114 Table 4.20 Current policing structures in Sierra Leone Policing body SLP Mine monitoring officers, beach and traffic wardens Authoriser Powers Key function State structures Central Standard Serious government police crime powers Ministry of Reduced Supervision Mines, police of mining Tourist powers and Board, Freetown Traffic beaches and Authority traffic Links with other policing bodies Funding Work with partnership boards Work with SLP State Volunteers State State-approved policing Local policing partnership boards Work-based policing associations SLP Citizen Provide intelligence to police Work with SLP Elected leaders of association Citizen Security of workplace Minimal Members’ contact with contributions SLP Customary structures Chiefdoms Judicial Civil and powers in customary civil matters arbitration Minimal Chiefdoms contact with and state SLP Licensed by Guarding police Joint patrols Commercial with armed police in high-risk situations Commercial Private security companies groups Peace monitors Community Citizen groups Civil arbitration Minimal Volunteers contact with police Unauthorised policing Youth groups Mob justice Youth groups Citizen Guarding and rapid response Assaults on alleged criminals Source: Bruce Baker (2005), see note 59, p.373 Minimal Volunteers contact with SLP None Case Studies 115 focus groups and interviews served to strengthen the poll ratings for secret societies, they merely confirmed the inconclusiveness of the ratings for the mines monitoring officers. Not one of the focus groups considered this group of actors to be a relevant security player. Unlike Baker’s list, which does not elaborate on the relative importance of particular actors, a look at the ratings for each of the three cities confirms that the mines monitoring officers are really only visible and relevant in Kono district, where most of the diamond mining is conducted (Table 4.21). The relatively large percentage of ‘don’t know’ responses in Freetown and Makeni confirms this inference. Table 4.21 Mines monitoring officers in three Sierra Leonean cities compared (%) Very/ somewhat important to City personal security 15.2 Freetown 8.0 Makeni 32.7 Koidu Town 12.4 Total Does not affect personal security 6.0 14.0 27.7 10.3 Somewhat/ a big threat to personal security 58.1 64.0 30.7 55.0 Don’t know No answer 18.2 11.0 3.0 15.0 3.0 0.0 1.0 2.3 N = 702 (Freetown 501; Makeni 100; Koidu 101) Question: ‘Now, I’m going to read out a list of groups that are said to affect security in one way or another. For each one, could you tell me whether you feel protected or threatened by them?’ Another interesting observation from the survey poll is the generally positive role of private security companies (PSCs). While a not insignificant total of 38.6 per cent of respondents said that PSCs had no impact on their security, quite a number considered them to be ‘somewhat’ or ‘very’ important for their personal security (54.6 per cent). Secret societies appear to play a significant role in providing security in Sierra Leone, with almost one-third of respondents considering them to have a positive role in security provision. That most respondents (around 50 per cent) considered them to be insignificant for their personal security is certainly partly attributable to the fact that the survey was conducted in three urban areas. Judy Smith-Höhn 116 Figure 4.3 Sierra Leone today – Security actors and their interactions EXTERNAL STATE Customs and Excise UN NON-STATE Al Qaeda connections Smugglers/ business people Prisons Judiciary Chiefs IMATT Secret societies SLP Community Yenga border issue Dogs Private security companies Taylor’s NPFL Political parties RSLAF Charles Taylor Armed robbers Pirates Excombatants Legend positive neutral/disputed negative cooperation rivalry rivalry/cooperation cooperation and avoidance rivalry within cooperation all around Source: FGD2 Sierra Leone, Teachers and Students. Copyright: Judy Smith-Höhn Case Studies 117 Focus groups and interviews confirmed the increased importance of such traditional actors in rural Sierra Leone, which brings us to the second criterion for assessing SSR programmes: the regard for context-specific characteristics in designing implementation strategies. Context-specific characteristics of Sierra Leone’s post-conflict security environment In Sierra Leone, there was an attempt by the government to formalise policecommunity cooperation through the police partnership boards (PPBs). Comprising community members, PPBs are intended to give citizens a voice in security-related matters, allow police officers access to information they would otherwise not have and help to improve the image of the police forces. The PPBs received mixed ratings in both the focus groups and the interviews, pointing to the fact that the success of such initiatives depends primarily on the commitment of individual police divisions to follow up and support community policing efforts. The mapping exercises conducted during the focus groups produced some striking results and led to the identification of additional actors that were considered relevant in the security sector; most prominent among these were the community watch teams (CWTs). They also served to confirm findings from the survey poll. By and large, a number of informal, i.e. non-state, security actors appear to operate in Sierra Leone (Figure 4.3). External actors continue to play a role, despite the withdrawal of UNAMSIL troops in December 2005. However, as in the survey poll, they feature less prominently. Whereas external actors were included in the mapping exercises conducted by the teachers/students and health workers, the market women and youth/community leaders considered them irrelevant in terms of providing security.106 As mentioned above, the traditional actors – secret societies and chiefs – not only feature prominently, but are also integrated into Sierra Leone’s security architecture, as shown by the lines of interaction with other security actors. Of the actors considered to play a positive role in providing security, probably the most striking one mentioned in the focus group is the dogs. Although it may seem odd that a group of animals is considered to play a significant and active role in the security arena, what it in fact indicates is that citizens provide for their own security by acquiring watchdogs to protect themselves and their property. Moreover, it also points to a lack of faith in other security providers. One focus group discussant maintained that he 118 Judy Smith-Höhn could not rely on the police nor on any PSC to provide effective security to him personally.107 The integration of secret societies into the security architecture appears to be a unique feature of Sierra Leone’s post-conflict environment. In addition to secret societies’ acknowledgement as viable security actors in their own right, having points of contact with other actors in the country’s security architecture, they also enjoy hidden representation by other security actors. The paramount chief, for instance, is usually a member of a secret society. Furthermore, membership of these societies is widespread – not least in the police, PSCs and even among youth leaders. As for other traditionally based actors, the CDF no longer play as positive a role in the security arena as in early stages of the civil war. Returning to Table 4.19, more than half of all respondents considered them to be a threat or a big threat to their security. The focus groups confirmed their negative status, which is certainly partly attributable to their changing role during the civil war. The role of the chiefs, on the other hand, has always been ambiguous and depended on the chief’s personal ambitions. While some chiefs in rural areas deserted their people, others played a key role in recruiting and funding kamajors. A number of informal actors were revealed to pose a potential security risk. The Ghetto Boys, for instance, many of whom are ex-combatants, are considered to be a major threat to security. The West Side Boys – a rogue army faction that emerged at the end of the 1990s but which has been officially disbanded – are considered to pose a similar threat. The focus groups confirmed these views, and discussants agreed that the West Side Boys still maintained a modicum of organisational capacity, with several claiming that they were capable of recruiting and reorganising at any time. Regardless of whether this is true or not, the fact that many former combatants – whether Ghetto Boys, West Side Boys or RUF rebels – are unemployed and struggling to make ends meet is cause for concern, for it is from this group that those wishing to destabilise the security sector for their own personal gain, for example, would recruit. Bike riders, many of whom are also ex-combatants, can serve as an example of the potential to work with such actors. All in all, they received the most balanced results of all actors, i.e. an almost equal number of respondents considered them either as important, insignificant or a threat to their security. In contrast to the Ghetto Boys, this group appears to have a more positive impact on security (39.1 per cent, see Table 4.22). A third found that they had no impact on their security, while 26.6 per cent Case Studies 119 considered them to be a threat. These results can be explained if one considers the responses in each of the cities where the survey was conducted. Following conflicts between bike riders and other transport stakeholders in Makeni and Koidu Town, civil society groups and the local police intervened successfully to curb this threat. This led to a rapid decrease in bike-related unrest and explains the favourable results this group obtained in these cities, particularly in Makeni, where 79 per cent of respondents considered them as being important or somewhat important for their security (Table 4.22). In Koidu, 73.3 per cent of respondents considered them as positive, whereas only 24 per cent of respondents in Freetown answered accordingly. In the latter case, most respondents found the bike riders to have no impact on their security (37.1 per cent). Table 4.22 Bike riders in three Sierra Leonean cities compared Very important to personal security Somewhat important to personal security Freetown 9.4 14.6 37.1 19.8 13.4 Makeni 33.0 46.0 15.0 3.0 3.0 Koidu 41.6 32.7 9.9 11.9 3.0 City Does not affect personal security Somewhat a A big threat threat to to personal personal security security N = 702 (Freetown 501; Makeni 100; Koidu 101); percentages of ‘don’t know’ and ‘no answer’ responses are not shown. Question: ‘Now, I’m going to read out a list of groups that are said to affect security in one way or another. For each one, could you tell me whether you feel protected or threatened by them?’ The mapping exercises conducted during the focus groups not only produced colourful results, but also enabled a more in-depth analysis of the actors considered as relevant in the country’s security arena. Table 4.23 on page 120 encapsulates all the security actors identified by the focus groups, categorising them according to their average ranking as well as their status as state, international/commercial non-state or domestic non-state. In most cases the ranking of the various actors confirms the results from the quantitative survey, such as the positive ratings of the RSLAF, UN and IMATT. As noted above, a positive rating does not necessarily imply that these actors play a proactive role in security provision, as in the case of Judy Smith-Höhn 120 Table 4.23 Sierra Leone – Mapping security actors today Types of actors State Positive - Chiefdom police - RSLAF International / commercial non-state Domestic non-state - UN - IMATT - Headmen/ headwomen - Soweis - Media - Dogs Neutral/disputed - SLP - SLP Operational Support Division (OSD) - Customs and Excise - Immigration - Parliament - Government - City council Negative - Prisons - Port security - Judiciary - Secret societies (Poro, Ojeh; hunter militias) - Chiefs - Community watch teams - Private security companies - Ex-combatants Doctors Criminals Ghetto Boys Student politics Political party militias Potential threats - Charles Taylor, NPFL - Yenga border issue - Businessmen - Al-Qaeda connections Source: FGD1–4, Freetown, April–May 2006 the RSLAF, where the positive ranking was attributable to its invisibility rather than its capacity to provide for national security. Similarly, a neutral or disputed ranking requires elaboration. The community watch teams, for instance, were ranked as neutral/disputed not because discussants disagreed on whether they had a positive or negative effect on security; rather, there was unanimous consent that this group had the potential to be a guarantor of security, but the composition of these self-organised squads was key to Case Studies 121 ensuring that they do not pose a threat to security, as was known to have happened with the CDF in the past. Finally, if one compares the current security situation with the period before the end of the war, UNAMSIL’s withdrawal in December 2005 marks a shift in the constellation of security actors. Table 4.24 reveals that Sierra Leone’s current security arena contains far more actors than were present during UNAMSIL’s deployment. Table 4.24 Sierra Leone – Security actors past and present Period Before 2001 External - ECOMOG - Executive Outcomes - 2001 – December 2005 - UNAMSIL - Since January 2006 - State OSD Police Sierra Leone Army City council police Chiefdom police Armed Forces Revolutionary Council Chiefdom police Sierra Leone Police City council police OSD RSLAF OSD RSLAF SLP Council Chiefdom police - Non-state PSCs Civil defence unit (youth) CDF RUF Gbethis - PSCs - Community night watch - Community - Chief - Headman - Market women - Ojeh - Poro - Soweis - Hunter militias Source: FGD 2, Market Women, Freetown, 28 April 2006 Similarly, more actors were involved in the period before 2001, prior to the end of the civil war. This implies that with the presence of UN 122 Judy Smith-Höhn troops the need for security was covered. Moreover, the gradual withdrawal of UNAMSIL troops produced a security vacuum that was filled not by state actors but rather by private non-state actors. Some of them have only been established recently (such as the PPBs and the community night watch), while others have re-emerged (such as the secret societies). Dealing with legacies of the past? The fact that many former combatants – whether Ghetto Boys, West Side Boys or RUF rebels – remain without employment and struggle to make ends meet poses a major security threat and is an indication that the legacies of the past may not have been addressed as effectively as they should. Though officially disbanded, these groups have in the past proven their ability to recruit and reorganise at any given moment. The need to focus on social background conditions, e.g. by creating employment opportunities in order to reintegrate such groups, is therefore crucial in securing a lasting peace. As mentioned earlier, in addition to the measures taken towards SSR and the relevant actors mentioned above, transitional justice mechanisms were also established soon after the cessation of the civil war. Problematic as the concurrent operation of these may have been, Sierra Leone’s TRC as well as the UN-backed SCSL constitute formal efforts to address the legacies of the civil war.108 Sierra Leone has a vibrant civil society, and although the government tends to ignore civil society actors that are out of line with its own political and economic interests, local mechanisms for reconciliation have proven more effective – at least on the community level. Religious groups and particularly women’s movements, as well as other grassroots organisations, have often played a key role in reconciling those who had been adversaries during the war. Furthermore, the present government remains supportive of the chieftaincy system, which is viewed by some as one of the main causes of the war. Colonial indirect rule has a particularly strong legacy in Sierra Leone, and the chiefs continue to be closely involved in almost all aspects of everyday governance in rural areas. This has bred animosity due to the fact that chiefs, as heads of the local judicial system, regularly handed down fines that were grossly disproportionate to the offences committed, as well as obliging their subjects to work for them without payment. During the civil war, however, chieftaincy was also used an instrument of civil-military liaison in northern areas. Case Studies 123 Notwithstanding the major role played by the 17,300-strong UNAMSIL force in maintaining a stable security environment during its deployment from 2001 to 2005, British involvement in Sierra Leone has had a significant, if not the most significant, impact on stabilisation and peacebuilding processes in the country. And although the last UNAMSIL troops withdrew in December 2005, leaving only a small contingent of 250 Mongolian troops to guard the UN-backed SCSL, the Sierra Leonean state still relies heavily on external support for security, conflict management and effective governance of its institutions. Indeed, the massive influx of external funding may have weakened the motivation to deploy the country’s own capacities effectively. Far-reaching structural change can only evolve out of leaders’ and individuals’ sense of political responsibility and ownership; external actors can at best support this process. Moreover, the newly (re-)established security forces have yet to prove their ability to deal with potential security threats. There have been no reports of major incidents of unrest in 2006, with the extradition of Charles Taylor from his exile in Nigeria to the SCSL in March 2006 viewed by many as the major security challenge that year. The court promptly requested the relocation of Taylor’s trial to The Hague, citing security concerns. This is certainly an indication that Sierra Leone’s national security apparatus is still not fully functional and is likely to remain reliant on external support for some time to come. The within-unit comparison of the security arena in Sierra Leone can thus be summarised as follows. State actors are expected to play a leading role in security provision, despite awareness of their shortcomings in terms of capacity, skills and resources. This poses a challenge in that the state institutions – the SLP and the RSLAF, for example – will not be able to fulfil this function exclusively for some time yet. This jeopardises the potential for successful reform, in that there is then a risk of disengagement from the state in this particular arena. If citizens rely increasingly on alternative means of security provision, and these become entrenched in the security architecture not as subordinate and complementary to state institutions but as substitutes for services the latter are expected to provide, the context for effectively reforming the security sector becomes progressively more difficult. The comparison revealed the positive and immediate effect of efforts to tackle security threats on the ground, as was demonstrated by the activities of the local unit commander and his team in Koidu, as well as the Judy Smith-Höhn 124 Figure 4.4 Sierra Leone 2002–2005 – Security actors and their interactions EXTERNAL STATE NON-STATE Chiefdom police UNAMSIL Police Private security companies City council police Operational Security Division Sierra Leone Army Legend positive neutral/disputed cooperation Cooperates with all Source: FGD1, Sierra Leone, Market Women. Copyright: Judy Smith-Höhn Case Studies 125 efforts to curb violence related to the bike riders in both Makeni and Koidu. It also reveals, however, that positive change depends heavily on the integrity and efforts of individuals. A formalised approach to dealing with security threats is not functional. Thus if the police officers of a particular division are not personally committed to performing their duties, and nor are those responsible for ensuring that disciplinary measures be taken against such officers, then the citizens are effectively left to provide for their own security. Finally, the incorporation of traditional, non-state actors in the security arena appears to be a unique feature of Sierra Leone’s post-conflict environment. In terms of SSR, this indicates that it should be easier to acknowledge and integrate such actors in the reform process. 4.3.3 Diachronic-binary comparison This section explores the conditions of the immediate post-war security sector as it was at different times in both countries. As the wider DSF-funded research project necessitated that the same questions be posed for both countries, the comparison across units is unproblematic. However, the field research was not conducted in the immediate post-war period of both countries – research was undertaken within a time frame of six months from November 2005 till June 2006, and in the case of Sierra Leone this was nearly four years after the formal end to its civil war. Therefore, as noted in Chapter 3, the information on Sierra Leone as it emerged from civil war is based on the focus groups and semistructured interviews only. The survey was not designed to explore this dimension of the country’s progress in terms of security and SSR. For the purposes of this study, Liberia is considered to have just emerged from civil war at the time the field research was undertaken. Therefore all findings on the security situation in Liberia today are equivalent to the findings on this situation in Liberia in the immediate postwar phase. Given that the international peacekeeping force, UNMIL, plays such a decisive role in Liberia’s security arena, it is necessary to ask the same about Sierra Leone: what was the significance of the international peacekeeping force during its deployment? In one of the focus groups discussants were asked to compare the security arena across time, not only the present and the period prior to the cessation of the civil war, but also the period of the UNAMSIL deployment between 2002 and 2005. Figure 4.4 on page 124 depicts the status and Judy Smith-Höhn 126 Figure 4.5 Sierra Leone today – Security actors and their interactions STATE NON-STATE Operational Support Division Community night watch Army Police Market women Chief Headman Ojeh Council Market women Poro Chiefdom police Hunter militias Legend positive cooperation neutral/disputed rivalry negative rivalry/cooperation rivalry within Source: FGD1, Sierra Leone, Market Women. Copyright: Judy Smith-Höhn Soweis Case Studies 127 interactions among the various security actors in Sierra Leone in the immediate post-conflict phase. Judging by this scenario, during the period of UNAMSIL’s deployment a decidedly lesser number of security actors were considered relevant for security provision. Moreover, the relationships among the various actors were all cooperative. Although the retrospective viewpoint taken by the focus group may be distorted, it does reveal that the security situation was considered to be more positive during the deployment of the UN peacekeeping force. Certainly, some actors threatened security in the post-conflict phase: the ex-combatants are the first set to come to mind. Yet again, what this particular mapping exercise reveals is the perception that security was provided adequately during the massive deployment of international force. Interviews with local and international experts confirmed that, from 2001 to 2004, the United Nations and the British troops were the most important security providers.109 Interviews generally confirmed the positive impact of the UN force despite its difficult beginnings,110 though all considered the involvement of the UK to have been equally decisive. When one compares Figure 4.4 with the results from the mapping exercises depicting the security situation today and the security arena just before the end of the civil war, both suggest that during the UNAMSIL deployment the security needs of citizens were covered. Figure 4.5 on page 126 demonstrates that a significantly larger number of actors are considered to play a role in the security arena. As noted in section 4.3.2, it appears that the gradual withdrawal of UN troops created a security vacuum that was filled not by state actors but by informal, non-state actors (compare Figures 4.4 and 4.5). Figure 4.4 depicting the security situation during UNAMSIL deployment has a total of only seven actors, most of which were state actors. However, this does not imply that they played the most important role, especially when one considers the rating awarded to them. Most were neutral or disputed, indicating their ambiguous role during this phase. This comes as no surprise, since the state actors were in the early stages of the restructuring and reform process. More importantly, Figure 4.4 shows that one particular actor – UNAMSIL – had cooperative relationships with all actors in the arena. In fact, focus group discussants were clear as to the central role played by UNAMSIL in providing security. In other words, were one to visualise this in the figure, the block of the international peacekeeping force would have to be relatively larger than all others to illustrate the scope of its role. Only one group of non-state actors, the PSCs, was considered in this Judy Smith-Höhn 128 Figure 4.6 Sierra Leone in 2001 – Security actors and their interactions EXTERNAL STATE NON-STATE Civil defence unit (youth) SLP Operational Support Division ECOMOG Executive Outcomes Private security companies Revolutionary United Front Police Civil defence forces Sierra Leone Army City council police Chiefdom police Gbethis Armed Forces Revolutionary Council Legend positive cooperation neutral/disputed rivalry negative rivalry/cooperation Source: FGD1, Sierra Leone, Market Women. Copyright: Judy Smith-Höhn Case Studies 129 retrospective exercise. A review of the literature shows that other actors, particularly traditional, were present throughout this period.111 The fact that these were not included points to the perception that they were not considered to be important in security provision. Security needs of urban citizens were thus considered covered by the actors identified above. A map for market women from rural areas would certainly have produced different results. Figure 4.5, representing the security arena today, is much more diverse. Having omitted the international actor, several non-state actors – community watch teams and particularly some traditional actors (Ojeh, Poro, Soweis, hunter militia, headman, chief) – have now been included in the arena. In this particular focus group (market women), it is noteworthy that the traditional actors were awarded a central role in the security arena following UNAMSIL’s withdrawal. Whereas some actors were considered wholly positive (the headman, the chief and the female secret society known as the Soweis), others played a more ambiguous role (the secret societies of the Poro, Ojeh and hunting militia). State actors’ rankings were even more diverse; the SLP was considered to have deteriorated, while the RSLAF has improved its reputation since UNAMSIL’s withdrawal. Consider for a moment the constellation of security actors in Sierra Leone in the period just before the end of the civil war (Figure 4.6). Again, numbers of non-state security actors were present in the period prior to the end of the civil war – the civil defence units, PSCs, CDF, RUF and the gbethis (a traditional society similar to the kamajors). Aside from the RUF and the CDF, which by that time had forfeited their good reputation due to the actions of some of their men, all other non-state actors were considered to play a positive role in security provision. Many young men who had protected their communities through the civil defence units are now active members of self-help community watch teams.112 The external actor Executive Outcomes was ranked neutral/disputed, to indicate that it had both a positive and a negative role in providing security. As mentioned above, of the state actors, both the police and the army were deemed to play both positive and negative roles. In the case of the army, it was the sobels – soldiers by day, rebels by night – who threatened security. The military has improved its reputation and is now considered to be a positive force – if only because soldiers are confined to their barracks and no longer actively threaten security.113 Judy Smith-Höhn 130 Figure 4.7 Liberia today – Security actors and their interactions EXTERNAL STATE NON-STATE National Bureau of Investigation Bureau of Immigration and Naturalisation Market supervisor National Security Agency UNMIL Special Security Service Liberia National Police Street boys Excombatants Monrovia City Police Political party militias Politicians Warlords Legend positive neutral/disputed cooperation negative rivalry rivalry within Source: FGD1 Liberia, Teachers and Students. Copyright: Judy Smith-Höhn Case Studies 131 The CDF in fact comprised different militia groups previously referred to by ethnically coded titles, such as the gbethis (Temne), kamajors (Mende), tamaboros (Kuranko) and donsos (from Kono).114 The CDF were considered a threat to security while the gbethis were deemed as providing security (Figure 4.6); this is partly because the CDF were not a military organisation, but should be understood as the militarisation of a social network. By the mid-1990s this network had become the symbolic and material focal point for community defence mobilisation after the realisation that the state military was by and large unable or unwilling to defeat the RUF.115 Thus various groups were subsumed under one umbrella term, and later a number of Liberians and mixed-parentage youths with combat experience in Liberia also joined the CDF, changing the dynamic within the network. A roaming population of combatants such as these youths, many of whom began fighting as children during Charles Taylor’s insurgency against the Doe regime, pledge allegiance to local warlords in exchange for payment, logistic support and the prospect of self-enrichment. They do not adhere to a centralised command structure and are usually better trained and more violent than their allies in the various factions.116 The immediate post-conflict period in Liberia indicates that there may be a different dynamic than in Sierra Leone. Figure 4.7 on page 130 illustrates the security arena in Liberia as it is perceived today. The most palpable observation that can be drawn from the comparison of Figures 4.7 and 4.4 (security in Sierra Leone in the immediate postconflict phase) is that in Liberia, aside from the central role played by the international peacekeepers, more actors – both state and non-state – are considered to comprise the security arena. In Sierra Leone, fewer actors were identified as significant during the same time period. This may lead to a number of inferences, e.g. that the UN mission in Sierra Leone was more important in security provision and thus covered a wide range of the security needs of citizens, or that the state security actors in Liberia are more established than those in Sierra Leone, given that several state agencies were identified in the mapping exercise in Figure 4.7. However, two factors play a role here. Since only one mapping exercise was conducted in Sierra Leone that can give an impression of the characteristics of the security arena during the UNAMSIL deployment, it can hardly serve as substantial evidence. Moreover, as noted above, the lack of survey data regarding this particular question makes it even more difficult to draw sound conclusions on the issue. Interviews suggest that non-state actors have played a role in security 132 Judy Smith-Höhn provision for quite some time. And although the presence of UN troops ensured stability during the immediate post-war phase, UNAMSIL was not the only player.117 Secondly, some of the state actors identified in Figure 4.7 were, at the time of questioning, not fully functional. Though this focus group did not explicitly note this during the exercise, the three other focus groups acknowledged the fact that, although formally existent, these actors were effectively not able to provide any security.118 Given the limitations of the data collected, one particular conclusion can be drawn: the significant role played by UNAMSIL during its deployment between 2001 and 2005. Its positive effect is certainly comparable to the equally (though not exclusively) positive role currently being played by UNMIL in Liberia. This raises the question of the possible impact of an UNMIL withdrawal. Would non-state actors (re-)emerge to fill the security vacuum? 4.3.4 Synchronic comparison This final synchronic comparison is the more detailed of the cross-unit comparisons since the data gathered from the survey poll, focus groups and interviews provide extensive information on the conditions within which SSR efforts are currently being undertaken (Cases 2 and 5 in Figure 3.2). Security and security actors in Liberia and Sierra Leone: Changes over time Using the criteria for measuring SSR elaborated in Chapter 2, one would again first enquire into the state of security and the presence of viable security actors. The cross-unit comparison enables a more concise analysis of the conditions in both countries, in that the ratings can be compared systematically. Table 4.25 reveals that overall there has been a predictably dramatic shift in security perceptions of Liberians and Sierra Leoneans compared to the period prior to the end of their respective civil wars. Here too, a significant number of respondents considered their personal security situation to be ‘somewhat’ or ‘much’ better (90.8 per cent in total, 86.1 per cent total in Liberia). Similar results were obtained for public security. A noteworthy result from the survey is that Liberians have a more favourable perception of their country’s general security situation than their neighbours, although Sierra Leone has enjoyed a longer post-conflict phase. Only 47 per cent of Sierra Leonean respondents rated the security situation Case Studies 133 in the country as ‘okay’ or ‘very safe’, compared to 60.4 per cent in Liberia, and they have a much more negative perception of the overall security situation (Table 4.25). A majority of respondents in both countries – 61.6 per cent in Sierra Leone and 69.3 per cent in Liberia – felt personally ‘very safe’ or ‘okay’, although Sierra Leoneans appear to be more sceptical about their personal safety than Liberians (37.8 per cent compared to 29.7 per cent). Table 4.25 General and personal security in urban Liberia and Sierra Leone compared General security (%)* Personal security (%)** Liberia Sierra Leone Liberia Sierra Leone Very safe 20.3 14.7 25.6 20.7 It’s okay 40.1 32.3 43.7 40.9 Fairly unsafe 26.9 37.9 23.5 27.5 Not safe at all 10.5 13.4 6.2 10.3 No answer/ don't know 2.2 1.7 1.0 0.7 N Liberia = 698; N Sierra Leone = 702 *Question: ‘In your opinion, how safe is the country today?’ **Question: ‘How safe do you feel personally in the country today?’ Although this discrepancy may contradict initial expectations, a consideration of the prevailing context in each country at the time of the survey provides some answers. In Liberia, UNMIL was still ever-present; and a fairly free and fair, and most importantly peaceful, election produced the first female African president, Ellen Johnson-Sirleaf. The survey in Sierra Leone was carried out against the backdrop of UNAMSIL’s withdrawal only a few months earlier, as well as the extradition of Charles Taylor to the UN-backed SCSL. Aside from these context conditions, it is precisely the fact that Sierra Leone’s post-conflict phase is older that accounts for the less favourable assessment of the security situation. The initial euphoria that ensued after the end of the civil war had over the years given way to the realisation that socio-economic conditions for the poor had yet to improve. The focus groups confirmed this line of reasoning, as all discussants agreed that the Judy Smith-Höhn 134 Table 4.26 Mapping security actors – Liberia and Sierra Leone today Country Positive Neutral / disputed Liberia - Market superintendent CWTs PSCs National Fire Service LNP Ministry of Defence Agents Against Rape Community chairman - UNMIL - Special Security Service - National Security Agency - National Bureau of Investigation - Ministry of National Security - Bureau of Immigration and Naturalisation - Town chief - Poro/Sandee - Kenewo/Senewo/Bodio - Monrovia City Police - Press union - Drivers union/ car loaders/ wheelbarrow boys - Liberian Marketing Association/ pedicuremanicure boys/ Yanna boys/ shoeshine boys/ money changers - - Headmen/headwomen Soweis Chiefdom police UN IMATT Dogs RSLAF Media - Secret societies (Poro, Ojeh; hunter militias) - Chiefs - City council - CWTs - OSD - PSCs - Customs and Excise - SLP - Immigration - Parliament - Government - Sierra Leone Negative Ex-combatants Street boys Politicians Political party militia Warlords Action agents Prisons Ex-combatants Doctors Port security Ex-combatants Criminals (armed robbers, drug addicts) - Ghetto youths - Judiciary - Student politics - Political party militia - Port security Potential threats - Charles Taylor; NPFL - Yenga border issue - Businessmen - Al Qaeda connections Source: FGD1–4, Monrovia and Freetown, December 2005–May 2006 Case Studies 135 lack of employment opportunities was a potential threat to security and deplored the SLPP government’s failure to address this issue. The government’s performance in other sectors was also criticised, e.g. its failure to improve the working conditions of the SLP, with the result that some police officers are compelled to resort to extortion as a result of long-overdue salary arrears, was identified as a further threat to overall security. Discussants clearly indicated that the primary responsibility for the rebuilding of Sierra Leone’s institutions rested with the government; the external donors were acknowledged as key facilitators for these processes. Consider Table 4.26, which represents a classification of all the actors noted as playing a role in the respective country’s security sector. For each country, the various actors identified as operating in the security arena are grouped into three categories: those considered to have a wholly positive impact on security; those considered to have a negative impact; and finally those that were either considered to be neutral (i.e. to have neither a positive nor a negative impact on security) or where discussants disagreed on the rating to be given (e.g. one discussant may have considered an actor to be positive while another considered it to be negative; or one focus group rated a particular actor as neutral while another rated it as positive). Thus the third category subsumes two types of actors.119 Note that the table does not reflect the relative importance of each actor; in other words, whether a particular actor or set of actors was mentioned in all or merely one of the FGDs is not shown. As far as the security arena is concerned, Liberia and Sierra Leone exhibit the following key similarities. Firstly, and not surprisingly, there is overwhelming evidence that security conditions have improved dramatically in both countries since the end of their civil wars. Secondly, state security actors – both police and military forces – received remarkably positive ratings despite obvious shortcomings in terms of equipment and training as well as their general history of violence and oppression. These perceptions are less a reflection of these actors’ actual performance, but can be interpreted as an indication of the high expectations regarding this. There is little doubt that there is a pronounced preference for state agencies to provide for security in both countries. A third similarity is the significance of non-state informal actors in providing security. This needs some qualification as regards different groups: in both countries non-state groups such as community watch teams were identified which – though not as prominent as state actors – were viewed overwhelmingly positively. Others, particularly secret societies, 136 Judy Smith-Höhn received mixed results, though they were deemed irrelevant for the personal safety of larger parts of the urban population. This is also true for the – generally rather favourably assessed – PSCs. Probably the most clear-cut parallel between the two countries in terms of non-state actors can be drawn with regard to informal groups of young men – such as ex-combatants, street boys and Ghetto Boys – who were undoubtedly considered the most serious threats. This finding points to the urgent need to (re-)integrate them into society by creating job opportunities, and it is nothing short of obvious that Liberia and Sierra Leone face very similar social and economic challenges if they are to eradicate the social roots of violent conflict. Though DDR processes were carried out in both Sierra Leone and Liberia, they were marred by lack of funds to provide sufficient training and difficulties in finding employment pending training. The problem is compounded by the regional dimensions of past and present conflict in this part of the world. In the past, Liberia and its neighbours – Sierra Leone, Guinea, Côte d’Ivoire – have been involved in each other’s conflicts. Not only local but also international ‘violence actors’ (e.g. armed refugees) hired by the warring parties have been involved in the civil wars of the region. With Côte d’Ivoire having only recently signed a peace agreement in Ouagadougou in 2007, the future of Guinea being clouded by rumours of an impending coup and the failing health of Guinea’s Lansana Conté, there is still a formidable level of insecurity in region. If the need arises, this group of disgruntled youths, all with combat skills, can be lured back into militias, particularly in Guinea and Côte d’Ivoire. Notwithstanding many similarities, there are two major differences between the two countries, at least regarding the actors. The first concerns the involvement of international actors, particularly the UN peacekeeping missions. In Sierra Leone, the external actors – UNAMSIL, ECOMOG and IMATT – do not play as significant a role in the provision of security as their Liberian counterpart, UNMIL. With its troop strength of more than 15,000 soldiers, UNMIL is the key provider of security. In Sierra Leone, UN peacekeepers, who during their deployment had most probably played an equally significant role, withdrew in December 2005, and the subsequent security vacuum was only partly filled by state and non-state actors. The second difference relates to the varying importance of secret societies and their ability to have a hand in SSR efforts. The Liberian study revealed that although such groups are acknowledged as security actors, their role is negligible; moreover, they appear to be ‘decoupled’ from the rest of the security arena. In Sierra Leone such traditional actors are much Case Studies 137 more relevant, as they are integrated into the security architecture. Comparing the maps from the focus groups conducted in the two countries, it becomes apparent that in Sierra Leone the traditional structures such as the chiefs are more integrated into the overall system (Figure 4.8 on page 138). While the Poro in Liberia is a security-relevant but nonetheless decoupled actor, the secret societies in Sierra Leone (Poro, Sowei, Ojeh) have points of contact with the other actors. They are also represented by other actors in the security apparatus, for instance through the function of the paramount chief, the police and the youth leaders (as mentioned in section 4.3.2). Figure 4.8 compares the focus groups conducted with health workers in both countries. The three other Sierra Leonean focus groups attributed much more significance to the secret societies. Thus a comparison across countries and focus groups would produce more striking results than are visible here. The survey polls confirmed this notable difference. Whereas in Liberia very few respondents considered secret societies as important security actors (15.7 per cent), almost one-third of Sierra Leonean respondents considered them to play a positive role in security provision. That most respondents (around 50 per cent) considered them to be insignificant for their personal security is certainly partly attributable to the fact that the survey was conducted in urban areas. Interviews confirmed the increased importance of such traditional actors in rural Sierra Leone.120 As indicated earlier, both countries continue to rely heavily on external support for security, conflict management and effective governance of institutions. In both instances, the fact that the police services are illequipped and short-staffed has produced a security vacuum that has been filled, to a certain extent and only in some areas, by CWTs, which have taken on the responsibility of protecting their local community. In addition, citizens with the necessary financial resources engage private security firms to protect their property and business premises. Although the activities of these companies should not be regarded as entirely unproblematic or necessarily positive, they should nonetheless be taken into account when considering strategies to promote security. Finally, civil society movements should not be ignored. During the wars in both countries, women’s movements in particular were instrumental in bringing the conflicting parties to the negotiating table. This points to the significance of such actors and the need to consider them as valuable assets in any peacebuilding strategy. positive neutral/disputed negative Special Security Service Liberia National Police STATE Charles Taylor Guinea Kenewo/ Senewo/ Bodio Private security companies RSLAF influences rivalry within rivalry/cooperation cooperation and avoidance Private security companies Media Community Chiefs Political parties Ghetto youth Armed robbers NON-STATE rivalry Fire service Prisons Judiciary Police Immigration STATE cooperation Car loaders Street peddlers UN IMATT EXTERNAL Community chairman Community watch teams Poro/Sandee NON-STATE Source: FGDs Liberia and Sierra Leone, Health Workers. Copyright: Judy Smith-Höhn Legend UNMIL EXTERNAL Figure 4.8 Liberia and Sierra Leone today – Security actors and their interactions Case Studies 139 The survey polls and focus groups conducted in both countries revealed that the constellation of security actors, and hence the context for SSR, differs in these two post-conflict societies. Thus, from the onset, one should expect that the SSR strategies implemented reflect this difference. Although this is indeed the case, the differences in the approaches adopted by the various implementing actors are less an acknowledgement of the need for context-specific SSR strategies, and more a consequence of the division of labour between the various international actors. Liberia, in particular, has not benefited from the concurrent involvement of international actors in reforming its dilapidated security sector: the US government outsourced the programme to reconstruct the AFL to Dyncorp, a company with somewhat controversial involvement in Iraq and Afghanistan, while the UN focused on DDR and the training of the local police through CIVPOL. In addition, there are the justified concerns voiced by several CSOs as regards the suitability of private military companies to train the national army, such as the lack of transparency in the training procedure and issues of accountability. Although the UN’s non-commercial efforts are preferable, attempts to coordinate the efforts undertaken by the various UN agencies active in the country are still under way. The general lack of coordination is a serious concern. Interviews conducted with UN staff revealed that it was only in January 2006 that the international implementing agencies working to reform the country’s security sector began considering the need for a comprehensive approach that coordinates the activities of the various agencies. Until that point, the external implementing agencies had divided the tasks of reforming the various security institutions among themselves – most notably UNMIL, which had taken on the task of training the Liberian National Police, while the US seconded the task of training, or more precisely building from scratch, the AFL – with the result that neither knew what the other was doing. In comparison, the SSR efforts under way in Sierra Leone appear to be more coordinated. The fact that the reconstruction is being carried out bilaterally through various UK agencies has facilitated such synchronisation. As noted in section 4.2.2, the Africa Conflict Prevention Pool was initiated as early as 2001, in an effort to consolidate the conflict prevention activities of the FCO, DFID and the MoD. A key difference in the reform efforts of both countries is the early establishment in 2002 of the Office of National Security (ONS) in Sierra Leone. The ONS is a post-conflict innovation that receives technical assistance from the UK and is responsible for coordinating the activities of 140 Judy Smith-Höhn Sierra Leone’s security agencies, and in the process for developing joint sectoral assessments on a regular basis. It functions as the coordinating body of Sierra Leone’s security agencies and serves as the secretariat of the National Security Council. Liberia’s Governance Reform Commission (GRC) could either duplicate the role of the ONS or establish a similar body to perform this key function. As noted in section 4.1.2, Liberia’s SSR process has had a narrowed focus on the AFL, the LNP and the SSS, while the other security agencies received little or no attention by the implementing actors. In light of the bloated nature of the country’s security sector, there should be a more concentrated focus on the other state agencies to address the problem of the overlapping responsibilities and duties of certain institutions. Four key lessons can generally be considered when designing security sector reform and reconstruction strategies in other post-conflict societies. First, the success of the reform of the Sierra Leonean army is a good example of the merits of non-commercial forms of external intervention, as opposed to the employment of private military companies. The findings from the fieldwork reveal that the image of the RSLAF has improved substantially, and the training it received through the British-led IMATT team has certainly played a part in this respect. Those in charge of SSR in Liberia should reconsider the continued involvement of the commercial PSC Dyncorp to train the new national army, particularly given the concerns involving transparency and accountability deficits. Second, results from the Sierra Leone case study underscore the importance of non-state actors in this sector. Since the national security apparatus is still not fully functional, citizens will turn to alternative informal actors for security provision for some time to come. Hence this is not a short-term phenomenon and should be taken into account when designing SSR strategies. Third, the concerns voiced by focus group discussants regarding groups of unemployed youth – mostly former combatants – and their negative assessment in the survey polls underline the need to focus on social background conditions. The creation of employment opportunities other than those derived from engaging in combat would serve to reintegrate such groups. This is particularly important in the run-up to elections, when such groups would be vulnerable to manipulation by political leaders. Finally, local ownership of the transformation process in post-conflict societies is crucial if it is to succeed. Two examples from Sierra Leone suggest a direction for Liberia: the ONS and the community-based PPBs. There are obvious benefits in having one overriding national body such as Case Studies 141 the ONS to coordinate all security agencies, as opposed to several autonomously functioning bodies with possibly duplicating roles, particularly in terms of promoting effective security threat assessments. As in Sierra Leone, with the rise in crime and the lack of adequate state-led responses to the problem, community-based neighbourhood watch teams have become more prevalent in Liberia. The fact that such community-based responses are (re-)emerging points to similarities with Sierra Leone’s experiences after the civil war. However, unlike in Sierra Leone, where there was an attempt by the government to formalise policecommunity cooperation through the PPBs, those in charge in Liberia have hitherto limited their strategies to publicly calling for increased community action. Justice Minister Cllr Frances Johnson-Morris called on citizens to organise vigilante groups to protect themselves, citing the police’s inability to deal with all crime-related incidents due to their composition and a lack of equipment. Other public figures criticised this call for self-organisation, among them former speaker of the House of Representatives Edwin Snowe, stating that the government was thereby abdicating its duties.121 Focus groups revealed that, as in Liberia, Sierra Leoneans ideally prefer state actors to non-state actors. Though this is no indication of their actual performance, state security forces are expected to provide security, particularly at the national level. Citizens resort to privately produced security mostly by way of communal self-help organisations, and to a lesser extent from commercial actors, for lack of a state-produced alternative. In addition to the aforementioned PPBs, the data gathered from the fieldwork in Sierra Leone revealed the existence and relevance of several actors – particularly of a non-state, informal nature – in providing or threatening security, with the constellation of security actors being somewhat different than that of Liberia’s security arena. In conclusion, there is no doubt that efforts to improve the performance of the security sector in both countries must continue. Foreign troops can and should not stay forever and, in the long run, there can be no alternative to a locally owned security sector in post-conflict societies. Although I shall refrain from drawing too far-reaching conclusions, the results suggest that in Liberia and Sierra Leone a strong state security sector that incorporates non-state actors as partners will work best. Given the fairly divergent perceptions of the different groups, there can be no simple blueprint for the treatment of these sets of actors. It may be worthwhile considering some form of cooperation with the more positively rated actors – at least from a mid-term perspective – while other groups may warrant a 142 Judy Smith-Höhn more robust approach, in case the improvement of their social background conditions should not suffice to reduce the threat emanating from them. The development of a promising strategy on dealing with non-state security actors will require more positive knowledge about them, particularly as regards the various roles of different groups of actors, their relations and interactions with each other and their impact in rural areas. In short, further research into all aspects of an oligopoly of force is of the essence. 4.4 Multilevel comparison: A synopsis The brief historical review in sections 4.1 and 4.2 illustrated the following key similarities between the two country case studies. In both countries, many different security actors have vied for control and influence over several decades. The national army and police were never able to establish or maintain a state monopoly on the use of force for any significant period of time. In the hinterland, local civil defence forces, rebel groups and secret societies have dominated the security sector. For citizens, these various actors could act as protectors but could also pose a threat to individual security. Actors from neighbouring countries have played a major role during periods of conflict escalation. While the intentions and achievements of the intervention troops (especially ECOMOG) remain open to argument, the peacekeeping troops were ultimately able to impose a much-needed peace, largely through the use of force. The previous sections have shown that the context for reforming security institutions in post-conflict environments is particularly challenging. Having analysed the efficacy of current SSR strategies within and across the two units under study, the fulfilment of the four criteria identified as benchmarks for measuring the success of SSR strategies adopted in the two post-conflict societies of Liberia and Sierra Leone will be summarised under four headings: security actors; context sensitivity; local ownership; and root causes of conflict. The first heading covers the impact and efficacy of all the security actors identified as relevant in both post-conflict environments. Particular attention is given to the impact of external involvement in view of the Case Studies 143 multiplicity of external actors and the overlapping of duties. The second issue is simply an assessment of whether the reform strategies undertaken in both cases were indeed context-specific, which leads to the third issue of formulating strategies that can foster local ownership of reform efforts in order to ensure their sustainability. The final issue summarises whether the root causes of the violence have been adequately addressed by those implementing and overseeing SSR to avoid a relapse into conflict. 4.4.1 A multitude of actors: Many hands make light work or too many cooks spoil the broth? The synchronic comparison proved particularly useful for an assessment of the impact of external involvement. The Sierra Leonean case, for instance, provides some insights into what happens to a post-conflict society following the withdrawal of peacekeeping troops and the cessation of longterm intrusive intervention. Secondly, the comparison of both countries and identification of key similarities between the two post-conflict environments allow a prognosis regarding the possible implications and consequences of international involvement for Liberia. A striking similarity was the insufficient level of coordination of the activities of the numerous agencies involved in SSR. This was evident during the examination of the legal frameworks for reform. The fieldwork not only confirmed this shortcoming, but also enabled an assessment of the impact of uncoordinated activity on the environment. Moreover, it became apparent that reform policies of the lead agencies involved in SSR attached little overall significance to local conditions and actors. Social and political mechanisms, capacities and conflicts at local level were largely ignored. Since it cannot be expected of a temporarily deployed UN mission or the newly trained national police and armed forces to provide security in the medium term, SSR can only work effectively if local conditions are taken into account. Evaluation of the research undertaken in Liberia revealed that the residents of Monrovia regard UNMIL in particular as playing a crucial role in assuring security. The fact that an external actor was named as the key guarantor of security raises a number of serious questions. How long must the UN mission remain on the ground? When and how should responsibility for national security be transferred to local security agencies? How certain can the country’s citizens and donors be that the newly trained police and soldiers genuinely have the capacity and the will to fulfil their functions properly? Interviews with local and international observers revealed that it 144 Judy Smith-Höhn was far too early even to consider a withdrawal of the UN mission at this stage; in many experts’ view, the security situation remains volatile. Dyncorp’s involvement in the Liberian SSR programme through the reconstruction of the AFL and the UN’s training activities with the LNP through its CIVPOL unit are an example of the lack of coordination. Moreover, CSOs have raised concerns about employing PSCs in this way, citing problems of transparency and accountability. Non-commercial efforts may thus be preferable: UNMIL’s training of local police recruits is in full swing and some are already deployed. The work of the GRC, charged with setting up a ‘national integrity framework to promote transparency, accountability and the rule of law’, may successfully foster a dialogue that takes into account security concerns of both state and non-state local actors. The commission recently initiated nationwide consultations with key players, including district commissioners, chiefs, representatives of county security agencies, UNMIL and CSOs. The withdrawal of UNAMSIL international peacekeeping forces from Sierra Leone after five years of deployment had a significant impact on the security environment. The fact that several non-state, informal security actors re-emerged or were established subsequent to UNAMSIL’s withdrawal highlights the fact that a security vacuum was created that could not be, or was not, filled by state actors. That non-state actors such as CWTs became increasingly important points to the need to incorporate them in SSR designs during the planning and implementing phases to ensure that they gain the relevant knowledge and skills to perform a particular function should the need arise. This does not discount the primacy of state actors in their role as the key providers of security, but it does acknowledge the reality of post-conflict environments in which the state actors, due to financial and other restraints, are not in a position to hold the monopoly on the use of force. 4.4.2 Context-sensitive security sector reform: The case for second-best solutions In January–March 2005 Tufts University’s Humanitarianism and War Project conducted a study on security perceptions among three sets of actors – the military contingents of peace support operations, international assistance agencies and local populations – in countries in or recovering from crisis.122 Among the countries studied was Sierra Leone, where key findings were that perceptions differed not only between the three sets of actors but also within each group, and that perceptions evolved over time. Case Studies 145 This also applied in the other two country case studies, Afghanistan and Kosovo. Although the study offered no details about the interactions between the relevant security actors, it highlighted a key issue that guided the DSF project’s research in Liberia and Sierra Leone: policy-makers, international donors and NGOs must take into account local perceptions of security, be it physical, human or other security. This not only increases the acceptability of these actors, but is crucial if any strategy for peacebuilding and development is to be sustainable. However, context-specific conditions can only be identified through accurate and timely groundwork. In the period following the cessation of war in any given post-conflict setting, the immediate objective is maintenance of peace and stability – and well-planned strategies cannot be implemented partly because there is simply no time for preparation. Thus policy-makers resort to blueprint strategies designed without prior knowledge of the context within which such assistance is to take place. Once under way, these processes are not easily altered. Moreover, as mentioned in Chapter 2, reform policies are more often than not shaped by the domestic arrangements of the external actors, rather than by the conditions prevalent in the host country. Since these ideas differ among major international actors, a lack of policy coherence can be expected. During wartime, the images associated with security (or lack thereof) were of child soldiers, warlords, rebels – even the external actors like ECOMOG in Liberia had negative connotations. The results from the fieldwork show a remarkable shift in these perceptions, particularly of the state security actors. In both post-conflict reform contexts, state actors were granted a central role as security providers; that is, citizens expected them ideally to hold the monopoly on the use of force, despite their current reorganisation and history of violence against their own people. Although citizens are well aware of the all-too-blatant shortcomings of this set of actors, they have a remarkable level of confidence in their potential. Regardless of whether this perception is donor-driven, Westernised or inherently African, the fact remains that not only have governments adopted the notion of a democratically elected state with accountable institutions that are there to serve the people, but the population has internalised this notion of a modern state. Whether it is mere rhetoric or reflected in the policies and actions of the state, the change in the population’s perception points to opportunities for change which those implementing SSR programmes can capitalise upon, on the condition that the respective governments are willing and able to take on this responsibility. This development bodes well for the future, as it indicates that trust in state institutions can be developed. Hence 146 Judy Smith-Höhn the state can indeed hold the monopoly on the use of force, or at least be at the top of a hierarchy that delegates roles and duties within a new security architecture. Despite the obvious preference for the state to hold the monopoly on the use of force, there is a marked discrepancy between the expectations of citizens for the state security actors to reassert their monopoly on the use of force and the realities on the ground. This reaffirms the analysis of several observers that, in both contexts, the state will not be able to fulfil this role for some time. Both Liberian and Sierra Leonean governments remain heavily dependent on external support, and the (re-)emergence of non-state security actors such as neighbourhood watch teams and traditional security institutions following the withdrawal of international peacekeeping troops in Sierra Leone, as highlighted above, underscores this fact. A context-sensitive approach calls for a move away from a sole focus on idealised, far-reaching and long-term SSR programmes towards more realistic mid-term strategies. In other words, in light of the improbability of restoration of a state monopoly on the use of force in the near future, a second-best alternative is required. Phrased differently, there is a need for mid-term strategies within SSR frameworks. The findings suggest that in Liberia and Sierra Leone a strong state security sector embedded in sound socio-economic development that takes into consideration the role of non-state actors will work best. Given the fairly divergent perceptions on the ground, there can be no simple blueprint for the treatment of these security actors. Some form of cooperation with the more positively rated actors could be considered – at least from a mid-term perspective – while other groups (e.g. ex-combatants) may warrant a tougher approach in case their fortunes do not improve sufficiently to reduce the threat they pose. Moreover, the incorporation of local mechanisms into SSR measures would enable the implementing actors to address another problematic context-specific characteristic, namely the prevalence of mob justice, which is widely practised in both countries. Research showed that mob justice is widely accepted at all levels of society. As the worrying statement of one Liberia expert reveals: ‘You may call it mob justice, but in real life it’s just getting rid of pests.’123 4.4.3 Whither local ownership: The dilemma of external involvement There is now a widespread recognition of the importance of local ownership of transformation processes in post-conflict societies. Only by building Case Studies 147 reliable and trusted national security institutions through strategies that take into account local needs and perceptions will there be long-term and effective security provision. However, as highlighted by Nathan, for local ownership not to be treated as an idealistic, romantic notion, ‘the reform of security policies, institutions and activities in a given country must be designed, managed and implemented by local actors rather than external actors’.124 The principle of local ownership thus does not entail domestic support for donor activities, but rather donor support for programmes initiated by local actors. Moreover, the range of local owners should not be restricted to the executive, but should include parliament and civil society. If understood as national rather than government ownership, civil society can play a key role in fostering local ownership of reform efforts.125 Although Liberia has a less vibrant civil society than Sierra Leone, a number of NGOs exist that could foster ownership of the reform process in both countries.126 Note that the role of civil society is not to undertake SSR, but to support and lobby for particular reforms. This is the case of Liberia’s TRC process, for example, where several international and local CSOs – Amnesty International plus a steering group of local organisations – are involved in monitoring the commission’s work on a regular basis. In general, current post-conflict transitional arrangements seldom use both formal and informal (traditional) justice mechanisms for sustainable reconciliation. More support should be given to such informal mechanisms and the actors involved in implementing them, as it is they who understand the needs of their communities and can develop adequate strategies for reconciliation. The problem of local ownership is further compounded by the dilemma faced by international actors in such contexts: the longer and more intense their involvement, the more efficient transformation processes appear in the short term, yet the more difficult disengagement becomes in the long term. The incorporation of hitherto-disregarded actors into SSR strategies could contribute towards local ownership. 4.4.4 Addressing the root causes of conflict – Who is responsible? The need to address the root causes of a given conflict in order to ensure effective and sustainable SSR is closely linked to the imperative of local ownership. To prevent any relapse into old patterns of violence, the former conflict parties, which to some extent still enjoy considerable support and 148 Judy Smith-Höhn legitimacy in post-conflict settings, must be integrated into the process to ensure that ownership is not restricted to a new elite which currently cannot guarantee proper representativity or performance. The exclusion of these actors could itself pose a threat to security. It is vital to shift their loyalty away from individuals and towards the democratic state, but this process is likely to be fraught with problems. There is a need to focus on the broader social background conditions in order to address the root causes of the civil war and enable the state and society to resolve conflicts that may arise in the future without resorting to violence. The fact that informal groups of young men – such as excombatants, street boys and Ghetto Boys – were repeatedly mentioned as key threats to security in both country case studies points to the unresolved issues of reintegration and the unemployment of young males. The structural bases of armed conflict will most likely remain intact as long as there are insufficient employment opportunities for this demographic group. The creation of alternative sources of income to that derived from engaging in combat would serve to reintegrate such groups. Efforts to address this problem are under way, e.g. employment of ex-combatants in donor-funded government road construction and other labour-intensive projects. However, overall economic take-off is imperative and will require the rehabilitation of businesses and the effective management of state resources (e.g. those stemming from the export of resources such as timber and diamonds). Undoubtedly, it is essential that the legacies of the past be addressed to ensure the sustainability of reform efforts: there is a need to consider and avoid a repetition of the events and circumstances that led to the conflict in the first place. However, such issues are complex, and it is often difficult to determine where to begin. Moreover, as with putting in place effective oversight mechanisms, addressing and eradicating the causes of conflict are a timely endeavour. That they are crucial for sustainable peace is beyond doubt, but it is doubtful that these issues can be addressed comprehensively and speedily. Furthermore, one needs to question whether the issues can be addressed by the external implementing actor or whether it is a matter for the local leaders, who have yet to perform their functions without the aid of external actors. Case Studies 149 Notes 1 2 3 4 5 6 7 8 9 10 See e.g. Andreas Mehler (2005) ‘Major flaws in conflict prevention policies towards Africa: The conceptual deficits of international actors’ approaches and how to overcome them’, GIGA Working Paper Series No. 4; available at www.giga-hamburg.de/dl/ download.php?d=/content/publikationen/pdf/wp04_mehler.pdf (accessed 12 November 2005). Data in this chapter are taken from the following key sources: Elwood Dunn, Amos J. Beyan and Carl Patrick Burrowes (2001) Historical Dictionary of Liberia, 2nd edn, African Historical Dictionaries. Lanham, MD: Scarecrow Press; Stephen Ellis (2005) ‘Liberia’, in Andreas Mehler, Henning Melber and Klaas van Walraven (eds) Africa Yearbook. Politics, Economy and Society South of the Sahara in 2004. Leiden: Brill, pp. 96–104; Stephen Ellis (2006) ‘Liberia’, in Andreas Mehler, Henning Melber and Klaas van Walraven (eds) Africa Yearbook. Politics, Economy and Society South of the Sahara in 2005. Leiden: Brill, pp. 81-87; Peter Körner (2003) ‘Liberia 2002’, in Rolf Hofmeier and Andreas Mehler (eds) Afrika-Jahrbuch 2002: Politik, Wirtschaft und Gesellschaft in Afrika Südlich der Sahara. Opladen: Leske und Budrich, pp. 138–141; Andreas Mehler and Judy Smith-Höhn (2006) ‘Liberia: Ellen in Wonderland?’, GIGA Focus Afrika No. 5; available at www.giga-hamburg.de/dlcounter/download.php?d=/content/publikationen/ pdf/gf_afrika_0605.pdf; Judy Smith-Höhn and Verena Wiesmann (2003) ‘Krieg Kennt Keine Grenzen: Die Regionale Tragweite Des Konfliktes in Liberia. Eine Hintergrundanalyse’, Afrika im Blickpunkt 3; available at www.giga-hamburg.de/ dlcounter/download.php?d=/content/iaa/archiv/aib/AiB3-03.pdf; Bertelsmann Stiftung (2005) BTI 2006 – Liberia Country Report. Gütersloh: Bertelsmann Stiftung; EIU (various dates) reports and country profiles for Liberia. London: Economist Intelligence Unit. UN Mission in Liberia (undated) ‘UNMIL Facts and Figures’; available at www.un.org/ Depts/dpko/missions/unmil/facts.html (accessed 5 August 2007). See e.g. Wolf-Christian Paes (2005) ‘Eyewitness: The challenges of disarmament, demobilization and reintegration in Liberia’, International Peacekeeping, 12(2): 97. The CPA also stipulated that the executive branch of government be maintained; thus the newly elected National Transitional Legislative Assembly assumed power from the former legislative bodies, the House of Representatives and the Senate, and the 76 seats were divided among the warring factions and other groups. Another key goal of the NTGL was the organisation of the elections, which took place on 11 October 2005. See e.g. EIU (2004) Country Profile 2004: Liberia. London: Economist Intelligence Unit, p. 4; EIU (2006) Country Profile 2006: Liberia. London: Economist Intelligence Unit, p. 6. The finer points of the CPA will be elaborated in more detail in section 4.1.2. See Adedeji Ebo (2005) ‘The challenges and opportunities of security sector reform in post-conflict Liberia’, DCAF Occasional Paper No. 9, p. 6; available at http://se2.dcaf.ch/ serviceengine/FileContent?serviceID=DCAF&fileid=154D8391-8666-9EB5-64DDEC9C9DEA3446&lng=en (accessed 10 January 2006); EIU (2007) Country Report, June 2007: Liberia. London: Economist Intelligence Unit, p. 12. Ebo, ibid., p. 6. Ibid., p. 1. J. ’Kayode Fayemi (2004) ‘Governing insecurity in post-conflict states: The case of Sierra Leone and Liberia’, in Alan Bryden and Heiner Hänggi (eds) Reform and Reconstruction of the Security Sector. Münster: LIT, p. 179. 150 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Judy Smith-Höhn Ebo, note 7 above, p. 1. Ibid., p. ii. Government of Liberia. Articles VII–VIII of the 2003 CPA. Thomas Jaye (2006) An Assessment Report on Security Sector Reform in Liberia. Monrovia: Governance Reform Commission of Liberia, p. 4; available at www.kaiptc.org/_upload/general/Lib_Assess_Rep_on_SSR.pdf (accessed 3 November 2006). CPA 2003, Article XXXV, s. 1(b–e). Constitution of the Republic of Liberia, 1986, Article 54(e). Ibid., Article 34(b)–(c). UN Security Council (2003) Resolution 1509, 19 December, UN Doc. S/RES/1509, para 3(n). A full list of UNSC resolutions is available at www.un.org/documents/scres.htm. Ibid. Jaye, note 15 above, p. 5. Ibid. Aside from those institutions whose name basically indicates their key roles (e.g. Bureau of Customs and Excise, National Fire Service), others such as the NSA require some elaboration. The NSA was created in 1974 as the national body with the ‘sole authority to coordinate the activities of all national intelligence collecting services’. The National Bureau of Investigation was established in 1998, and its duties are to investigate ‘major crimes including homicide… illegal entries into the country, robbery, arson, rape, grand larceny, kidnapping, burglary, embezzlement, forgery, smuggling, violation of the narcotics law, counterfeiting, [and] theft of government property’. Its functions thus overlap with those of the police, Ministry of National Security, NSA and Bureau of Naturalisation and Immigration. The SSS was created in 1966 as a special unit to provide for all the security needs of the incumbent president. These include, for example, the physical protection of the president, his/her immediate family and certain officials and visiting dignitaries as well as protection of the executive mansion. The SSS also has law enforcement powers, as it is authorised to ‘arrest or cause to be arrested any person or persons committing a crime flagrant delicto, or aiding and abetting any crime or violation, or who may reasonably be suspected of the commission of a crime’. The Monrovia City Police was created to enforce municipal and city ordinances and regulations, assist in the collection of real estate tax and provide security to the city mayor and visiting dignitaries. Many of these functions are also provided by the LNP. See Jaye, ibid., pp. 6ff. For more on CIVPOL see the UNMIL website, www.unmil.org/content.asp?ccat=civpol. See e.g. Jaye, note 15 above, p. 12. UNMIL (2006) press release, 13 October; available at www.un.org/Depts/dpko/missions/ unmil/pr119.pdf. Various interviews, Monrovia, November 2005–February 2006. Comprehensive Peace Agreement Between the Government of Liberia and the Liberians United for Reconciliation and Democracy (LURD) and the Movement for Democracy in Liberia (MODEL) and Political Parties, Accra, Ghana, 18 August 2003, Article 6. See e.g. Paes, note 4 above, p. 1; Festus Aboagye and Alhaji M. S. Bah (2004) ‘Liberia at a crossroads: A preliminary look at the United Nations Mission in Liberia (UNMIL) and the protection of civilians’, ISS Occasional Paper No. 95, Pretoria: ISS, p. 7; UNMIL website, note 24 above. Paes, ibid., pp. 1ff. Case Studies 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 151 See e.g. Ebo, note 7 above, p. 6; International Crisis Group (2004) Rebuilding Liberia: Prospects and Perils, Africa Report No. 75. Freetown and Brussels: ICG, p. 9. Data in this section are taken from the following key sources: Krijn Peters (2005) ‘Sierra Leone’, in Andreas Mehler, Henning Melber and Klaas van Walraven (eds) Africa Yearbook. Politics, Economy and Society South of the Sahara in 2004. Leiden: Brill; Krijn Peters (2006) ‘Sierra Leone’, in Andreas Mehler, Henning Melber and Klaas van Walraven (eds) Africa Yearbook. Politics, Economy and Society South of the Sahara in 2005. Leiden: Brill; Bertelsmann Stiftung (2007) BTI 2008 – Sierra Leone Country Report. Gütersloh: Bertelsmann Stiftung; Bertelsmann Stiftung (2005) BTI 2006 – Sierra Leone Country Report. Gütersloh: Bertelsmann Stiftung; Peter Körner (2003) ‘Sierra Leone 2002’, in Rolf Hofmeier and Andreas Mehler (eds) Afrika-Jahrbuch 2002: Politik, Wirtschaft und Gesellschaft in Afrika Südlich der Sahara. Opladen: Leske und Budrich; EIU (various dates) reports and country profiles for Sierra Leone. London: Economist Intelligence Unit. Term used to refer to Africans from regions such as Nigeria, Congo and Angola who were captured as slaves but released before undertaking the Atlantic crossing See e.g. EIU (2007) Country Profile 2007: Sierra Leone. London: Economist Intelligence Unit, p. 4. Also known as a crown colony. The crown colony system was developed in the latter part of the eighteenth century to enable the British government to govern territories with large indigenous populations. At the head of such colonies stood a British governor, who nominated most of the consultative councils; these, in turn, often delegated considerable powers of local government to local rulers. Hong Kong was one of the most enduring examples of a crown colony. See The Columbia Encyclopedia, 6th edn. New York: Columbia University Press, 2007, p. 7167. See e.g. David Keen (2005) Conflict and Collusion in Sierra Leone. Oxford and New York: James Currey/Palgrave, p. 14. Caspar Fithen and Paul Richards (2005) ‘Making war, crafting peace: Militia solidarities and demobilisation in Sierra Leone’, in Paul Richards and Bernhard Helander (eds) No Peace, No War: An Anthropology of Contemporary Armed Conflicts. Oxford and Athens, OH: James Currey/Ohio University Press, p. 119. Amos Sawyer (2004) ‘Violent conflicts and governance challenges in West Africa: The case of the Mano River Basin area’, Journal of Modern African Studies, 42(3): 442. Fithen and Richards, note 36 above, p. 122. See e.g. Keen, note 35 above. Ibid., pp. 94ff. The term is a combination of the words soldier and rebel; Sierra Leoneans described these soldiers as being ‘soldiers by day, rebels by night’. See ibid., p. 109. The term is generally translated as ‘traditional hunter’. Traditionally, the kamajors were responsible for hunting game and protecting the community from danger (both human and animal) with their weapons. In the Koranko-dominated areas of the north the huntermilitiamen were known as tamaboros, while the gbethis were associated with the Temne and the donsos with the Kono. See e.g. Mariane C. Ferme and Danny Hoffman (2004) ‘Hunter militias and the international human rights discourse in Sierra Leone and beyond’, Africa Today, 50(4): 74ff. On 3 May 2000 some 500 UN peacekeepers from Kenya, Zambia and India were taken hostage by loyalists of Sankoh, who at the time held a post in the transitional government. The Abidjan Peace Agreement of 1996, the ECOWAS Peace Plan of 1997, the Lomé Peace Agreement of 1999 and the Abuja Ceasefire Agreement of 2000. Bertelsmann Stiftung (2005), note 32 above, p. 5. 152 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 Judy Smith-Höhn Ebo, note 7 above, p. 483. See e.g. ibid., p. 484; Fayemi, note 10 above, p. 181. Ibid. Fayemi, note 10 above, p. 182. Ebo, note 7 above, p. 484. Osman Gbla (2006) ‘Security sector reform under international tutelage in Sierra Leone’, International Peacekeeping, 13(1): 79. DFID’s country programme focuses on post-conflict reconstruction and involves security and justice sector reform, rebuilding the civil service, putting in place sound public financial management systems and tackling corruption through support to the AntiCorruption Commission. Other elements include support for decentralisation and private sector development. The UK’s strategy for conflict prevention in Sierra Leone has been summarised under three main objectives: ‘effective, affordable and democraticallyaccountable security agencies that are able to counter internal or external threats without significant external military presence; reconciliation and justice and the reintegration of ex-combatants into society; and reducing the external threats to Sierra Leone from the region’. Jeremy Ginifer (2004) ‘Evaluation of the conflict prevention tools: Sierra Leone’, DFID Evaluation Report EV 647, p. 1; available at www.dfid.gov.uk/aboutDFID/ performance/files/ev647sleone.pdf (accessed 2 October 2006). Gbla, note 51 above, p. 82. Peter Albrecht and Mark Malan (2006) ‘Post-conflict peacebuilding and national ownership: Meeting the challenges of Sierra Leone’, paper presented at international seminar of the Center for International Peace Cooperation (ZIF) and the Kofi Annan International Peacekeeping Training Centre, Accra, Ghana, p. 116. Ibid. Ibid. PRIDE (2004) ‘Improving the development response in difficult environments: Lessons from DFID experience’, PRIDE Working Paper 4, pp. 39ff; available at www.dfid.gov.uk/search/proxy/cs.html?url=http%3A//www.dfid.gov.uk/pubs/files/fragile -states/lessons.pdf&qt=SILSEP&col=dfid&n=3 (accessed 10 March 2006). Nicole Ball, Piet Biesheuvel, Tom Hamilton-Baillie and ‘Funmi Olonisakin (2007) ‘Security and justice sector reform programming in Africa’, DFID Evaluation Working Paper 23, p. 32; George Klay Kieh Jr (2005) ‘State-building in post-civil war Sierra Leone’, African and Asian Studies, 4(1/2): 171. See e.g. Ball et al., ibid., p. 32; Bruce Baker (2005) ‘Who do people turn to for policing in Sierra Leone?’, Journal of Contemporary African Studies, 23(3): 377. Ginifer, note 52 above, p. 11. Ibid. Ebo, note 7 above, pp. 485ff. Ibid., p. 486. Ibid. DFID (undated) ‘Terms of reference: Advice on security sector reform – Senior security sector coordinator Sierra Leone’, p. 2; available at www.dfid.gov.uk/search/proxy/ cs.html?url=http%3A//www.dfid.gov.uk/procurement/files/ojec8332tor.pdf&qt=SILSEP &col=dfid&n=1 (accessed 8 March 2008). For a detailed review of the Africa Conflict Prevention Pool see DFID (2004) ‘The African Conflict Prevention Pool: A joint UK government approach to preventing and Case Studies 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 153 reducing conflict in sub-Saharan Africa’; available at www.dfid.gov.uk/pubs/files/ acppinfodoc.pdf (accessed 25 July 2006). At the ceremony marking the official dissolution of the National Committee for Disarmament, Demobilisation and Reintegration, World Bank representative Eileen Murray noted that the country’s programme ‘is considered as the best practice example throughout the world of a successful disarmament demobilization reintegration programme’. See UNOCHA (2004) ‘Sierra Leone: Disarmament and rehabilitation completed after five years’, IRIN News Brief, 4 February. Quotes in UNOCHA, ibid. International Crisis Group (2003) Sierra Leone: The State of Security and Governance, Africa Report No. 67. Freetown and Brussels: ICG, p. 1. Interviews, Freetown, April–June 2006. See e.g. J. Andrew Grant (2005) ‘Diamonds, foreign aid, and the uncertain prospects for post-conflict reconstruction in Sierra Leone’, WIDER Research Paper 2005/49, p. 2; available at www.wider.unu.edu/publications/working-papers/research-papers/2005/ en_GB/rp2005-49/_files/78091760714974671/default/rp2005-49.pdf (accessed 2 October 2006). Article XXVI of Sierra Leone’s Lomé Peace Agreement. Agreement Between the United Nations and the Government of Sierra Leone on the Establishment of a Special Court for Sierra Leone (2002), Article 1. See e.g. Rosalind Shaw (2005) ‘Rethinking truth and reconciliation commissions. Lessons from Sierra Leone’, US Institute for Peace Special Report No. 130, p. 3; available at www.usip.org/pubs/specialreports/sr130.pdf (accessed 2 October 2006). IRIN (2005) ‘Civil society criticises “vague” government plan for post-war reform’, IRIN News Brief, 13 July 2005. Hans Nichols (2005) ‘Truth challenges justice in Freetown’, Washington Times, 5 January. Sierra Leone Truth and Reconciliation Commission (2004) Witness to Truth: Report of the Sierra Leone Truth and Reconciliation Commission, Vol. 1; available at www.trcsierraleone.org/pdf/start.html (accessed 2 October 2006). The question read: ‘In your opinion, how safe is the country today?’ See Table 4.1. Note that the percentages in the cross-tabulations comprise only the valid number of respondents, therefore ‘no answer’ and ‘don’t know’ options are not valued into the relative percentages as was done in Table 4.1. Nicole Itano (2003) ‘Liberating Liberia: Charles Taylor and the rebels who unseated him’, ISS Paper No. 82, November, Pretoria: ISS. Ibid., p. 8. FGD1, Monrovia, 2 December 2005. The author acknowledges that there can be a number of reasons for the negative perceptions of individual respondents, such as an increase in the crime rate, a lack of international peacekeeping presence in their area of residence, experience of domestic violence and so on. Interview with Gray-Johnson, Monrovia, 10 February 2006. Ibid. This was an open-ended question; no multiple-choice options were offered. FGD4, Monrovia, 21 December 2005. FGD3, Monrovia, 19 December 2005. FGD1–4, Monrovia, December 2005. 154 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 Judy Smith-Höhn See e.g. interviews with Tornolah Varpilah, Monrovia, 6 February 2006 and Commany Wesseh, Monrovia, 14 February 2006. See interview with Commany Wesseh, ibid. See also Amos Sawyer (2005) ‘Social Capital, Survival Strategies, and their Potential for Post-Conflict Governance in Liberia’, WIDER Research Paper 2005/15, p. 6; available at http://www.wider.unu.edu/ publications/rps/rps2005/rp2005-15.pdf (accessed 15 October 2005). Interview, Monrovia, 17 February 2006. See interviews, Annex III. Paes, note 4 above, p. 4. Ibid., p. 5. Compare FGD1–4, Monrovia and Freetown, December 2005–May 2006. During the war, the vigilantes were the only actors Liberians could rely upon to protect their lives and property. Having lost their relevance after the end of the war and also being infiltrated by ex-combatants with bad intentions, the vigilantes as a group are no longer viewed so positively. Compare FGD1–4, Monrovia, December 2005. Jaye, note 15 above, p. 9. See e.g. ibid., p. 12. FGD3, Monrovia, 19 December 2005. See FGD1–4, Freetown, April–May 2006. See e.g. interview with Joseph Kabia, Freetown, 21 April 2006. FGD1–4, Freetown, April–May 2006. Ibid. Baker, note 59 above. FGD1–4, Freetown, April–May 2006. ‘I do not trust the police or the private security companies. I have six dogs in my compound. In fact, the way I train them, I give them codes’; FGD2, Sierra Leone, 1 May 2006. See Article XXVI of Sierra Leone’s Lomé Peace Agreement. Interview, Freetown, 22 April 2006. In the early phases of deployment RUF rebels had managed to kidnap 250 UN peacekeepers, who were later released as a result of the intervention of British troops. See e.g. Baker, note 59 above. Interviews, Sierra Leone, April–June 2006. FGD1–4, Sierra Leone, April–June 2006. Danny Hoffman (2007) ‘The meaning of a militia: Understanding the civil defence forces in Sierra Leone’, African Affairs, 106/425: 642. Ibid. Ibid., pp. 657ff. Interviews, Freetown, April–June 2006. Compare FGD1–4, Monrovia, December 2006. For a comprehensive outline of the assessment made by each focus group, see Annex II. Interviews, Freetown, April–June 2006. George D. Kennedy (2006) ‘Speaker Snowe condems Issakaba boys’, Liberian Observer, 12 September. Antonio Donini, Larry Minear, Ian Smillie, Ted van Baarda and Anthony C. Welch (2005) Mapping the Security Environment: Understanding the Perceptions of Local Communities, Peace Support Operations and Assistance Agencies, report commissioned by UK NGO-Military Contact Group. Medford, MA: Feinstein International Famine Center, Tufts University. Case Studies 123 124 125 126 155 Interview with Rufus Kaine, 16 December 2005. Laurie Nathan (2007) No Ownership, No Commitment: A Guide to Local Ownership of Security Sector Reform. Birmingham: University of Birmingham, p. 4. Ibid., p. 5. Such as the West African Network for Peacebuilding, the Women in Peacebuilding Network and the Foundation for Human Rights and Democracy. Chapter 5 Conclusion This study aimed at analysing the efficacy of SSR efforts in post-conflict societies by means of two case studies, with a view to amending previous approaches and contributing to the knowledge base on this issue through the analysis of unique empirical data. It has been shown that, in both Liberia and Sierra Leone, the violent conflicts of the past emerged and persisted because those responsible for preventing or managing conflicts lacked the political will and resources to do so. In fact, they were often part of the problem. Emphasis has been put on the need for appropriate governance strategies devised in a collaborative effort by both government and civil society, with support from external actors, to deliver collective solutions to reforming the security sector. To substantiate the analytical approach, it was necessary to draw attention to the policy-driven nature of SSR programmes in general. Notwithstanding the need for a general understanding of effective SSR, the analysis of approaches adopted by the various donors has shown that a contextual perspective is required. In other words, notions of what comprises the security sector that are restricted to a state-centred focus confine the analysis from the outset and result in the neglect of other possible actors that may have the potential to contribute to (re)construction of a viable security sector in a given environment. Thus a wide-ranging definition of the security sector was devised which could be utilised for a comprehensive and wellstructured analysis of any particular post-conflict reform setting. Chapter 2 illustrated that what was needed was not yet another SSR approach, perhaps fed with new terminology or redefinitions, but empirical evidence of a given security situation on the ground. The study thereby contributes to narrowing the gap between the list of general recommendations for SSR and concrete suggestions based on empirical evidence gathered in specific post-conflict settings. The assessment of the SSR policies of the lead agencies involved led to the identification of four key criteria that can be used as a guideline to determine whether efforts to reform a security sector in societies emerging from civil war are effective: 158 Judy Smith-Höhn the existence of security actors and their future potential for providing security; the adequate consideration of context-specific characteristics, procedures and interactions of the security sector by the international implementing agencies; a focus on legacies of the past in order to avoid a relapse into conflict; and the confidence of the population in the ability of security actors to provide security. The first criterion centres on the competence of security actors; that is, whether actors exist that can provide for the security of citizens and the sustainability of their role as security providers. The fourth criterion, the population’s level of confidence in the ability of these actors to perform this function – in other words, whether citizens perceived actors to be effective – was a key benchmark for testing the validity of the first criterion. The second criterion was more widely focused on the context within which reform efforts take place, questioning whether the context-specific characteristics, procedures and interactions within the security sector of a particular postconflict environment were adequately addressed by the agents implementing reform. The third criterion focused on the legacies of the past conflict and examined whether the causes of a given conflict had been adequately addressed and the lessons learnt from this exercise incorporated into SSR strategies. Having identified the four key criteria for assessing the efficacy of SSR programmes, the methodology for conducting the analysis was elaborated. In recognition of the fact that a study can only be as good as its research design, this was given particular attention. The chapter started by highlighting the advantages of the comparative method vis-à-vis other methods of enquiry, such as the experimental method or statistical analysis. It was proven that the comparative method was uniquely suited for analysing complex causality: it acknowledges that a particular phenomenon – such as the efficacy of SSR – is affected by several factors. In addition, it was shown that because a case study analysis pays particular attention to historical context, and as such facilitates an understanding of the change in a particular phenomenon under scrutiny, a case study approach would be beneficial for the analysis of SSR strategies. The research design was therefore one of comparative case studies that used a mixed-methods approach to data gathering combining both quantitative and qualitative tools. The merging of quantitative and qualitative findings offered insights that could otherwise not be gained. By drawing from the two sets of findings, one could consider whether the results suggested interesting contrasts or helped to clarify or confirm each other. Conclusion 159 Due to their comparability, the post-conflict societies of Liberia and Sierra Leone were chosen as case studies. Both West African countries witnessed protracted phases of violence during the 1990s; in both instances there was extensive involvement of international and regional actors such as the UN and ECOWAS during the post-conflict phases; the socio-economic conditions are similar; and there appeared to be a number of different (types of) security actors in both cases. The case studies complemented the general insights gained from the literature on SSR. A synopsis of the various case comparisons highlighted three overarching issues: the first is directly related to the impact of external involvement and the overlapping responsibilities of the various implementing agencies, the second is the issue of context and, finally, there is the perpetual problem of instilling local ownership. These issues can be formulated into the following three questions. 1. 2. 3. What has been the impact of the international involvement in postconflict peacekeeping and peacebuilding processes? Are the reform strategies of these external actors – for it is they that initiate and guide these processes – effective and, most of all, have they managed to maintain the peace? Considering the fact that the state security actors have, in the past, proven incapable of dealing effectively with threats, in some cases having even been instrumental in perpetuating violence, is there any likelihood that state actors can be incorporated into a security sector institution that is accepted, trusted and deemed legitimate by the general population? Finally, in the wake of calls for ‘African solutions to African problems’, to what extent must local conditions be taken into account? Given relevant actors in a well-functioning security sector, what role can traditional actors – such as the secret societies often found in the African context – play? And to what extent can SSR strategies be owned by local actors? Firstly, the empirical data confirmed the crucial role of the international peacekeepers in maintaining the peace, which is a prerequisite for developing and improving strategies to rebuild security institutions so greatly damaged, if not destroyed, by the conflicts in the past. However, despite positive results regarding their general peacekeeping efforts, evidence gathered on the ground revealed that their presence, and thus impact, was felt mainly in the urban centres. Not only was there an urban- 160 Judy Smith-Höhn rural dichotomy, but there was also a discrepancy in the intensity of peacekeeping efforts depending on which parts of the city one happens to survey. This is understandable considering the limited capacity of such forces with regards to the availability of human resources, logistics and so on. Given the fact that these forces are temporary, the need to focus on strengthening those actors that would later have to provide the public good of security is imperative. Calls for local ownership can be overheard but, in certain cases, may be mere rhetoric. No matter how often it is reiterated in government and research papers as well as UNSC reports, the actual challenge of transferring power and responsibility to a local authority that may have had little or nothing to do with designing programmes, projects or processes that are imposed from ‘above’ has not been met with promising new approaches. Secondly, as far as the potential role of the state security actors is concerned – and here I refer to the a priori notion that the state should hold the monopoly on the use of force – the data reveal that citizens consider the state, ideally, to be the key provider of security. Though citizens are well aware of the all-too-blatant shortcomings of this set of actors, they have a remarkable level of confidence in their potential. This bodes well for the future, in that it indicates that trust in state institutions can be established. Hence the state can, indeed, hold some sort of monopoly on the use of force, or at least be at the top of a hierarchy that delegates roles and duties in a new security architecture. The people of Liberia and Sierra Leone have high expectations of a well-functioning state security apparatus, but these expectations cannot be fulfilled in the foreseeable future. A one-sided approach in which the international actors foster these countries’ newly emerging state security apparatus without consolidating democracy and recognising local self-help mechanisms is almost certainly doomed to failure. Thirdly, the issues of context specificity and local ownership are interrelated. The solutions to African problems may well need to be Africanised, but they are not necessarily exclusively African, and lessons can be learnt from the experiences of other countries in this field. As far as the incorporation of African elements into an African SSR strategy is concerned, those that can be considered African are, in fact, difficult to incorporate. The secret societies in Liberia, for example, cannot be incorporated into a security sector because they are what one could call a parallel society with its own set of rules, norms and objectives. This does not mean, however, that they are irrelevant. They are significant to some members of Liberian society, yet they are insignificant or even pose a threat to quite a number of Conclusion 161 other members of society. Unlike Liberia, Sierra Leoneans afford secret societies much more importance. Solutions must therefore be Africanised to the extent that they need to acknowledge such entities and the role they play in society, but without abandoning so-called modern strategies. In short, solutions and reform strategies must always be context specific; the relative significance of a particular variable cannot simply be inferred, but should be determined through well-grounded empirical analyses. It is a well-known fact that empirical data on Africa are rare and, if available, often unreliable, particularly in the interest areas of political scientists. Finally, the massive influx of external funding may have weakened the motivation to deploy the countries’ own capacities effectively. Farreaching structural change can only evolve out of leaders’ and individuals’ sense of political responsibility and ownership; external actors can at best support this process. In both Sierra Leone and Liberia the operational checklist for international interventions included the deployment of peacekeepers; disarmament, demobilisation and reintegration of former fighters; repatriation of refugees; judicial and security sector reform and, finally, elections. Yet many observers agree that the time frame for these interventions, set at two to five years, was far too short to ensure sustained peace and effective transformation in both countries. A final challenge to reforming the security sector in post-conflict societies such as Liberia and Sierra Leone centres on the need to acknowledge the regional dimensions of conflict and hence the development of regional solutions to SSR. When assessing the potential threat of conflict in Liberia, for example, it is imperative that one considers the fact that Sierra Leone, Guinea and Côte d’Ivoire were involved in the conflict. Hence it is not only the local non-state actors that influence the security sector, but also international ‘violence actors’ such as armed refugees hired by the parties in conflict. As noted above, there is still a high level of insecurity in the region. If the need arises, groups of disgruntled youths, all with combat skills, can be lured back into militias, particularly in Guinea and Côte d’Ivoire. The level of security in both Liberia and Sierra Leone is certainly a far cry from what it was at the turn of the century. In both instances, the international community acknowledged its responsibility to protect, as stipulated in UN Security Council Resolution 1647 on the protection of civilians in armed conflict, by intervening and – with the help of local civil society actors – eventually facilitated resolution of the conflict through the signing of a peace agreement. 162 Judy Smith-Höhn There is no doubt that efforts to improve the performance of the security sector in both countries must continue. Foreign troops cannot and should not stay forever; and in the long run there can be no alternative to a locally owned security sector in post-conflict societies. As noted above, the findings suggest that in Liberia and Sierra Leone the best approach would entail the establishment of a strong state security sector that collaborates with non-state actors within an overall framework of sustainable socioeconomic development. These sets of security actors may need to be handled differently: perhaps some form of cooperation with the more positively rated actors, while others may warrant a tougher approach in case their fortunes do not improve sufficiently to reduce the threat they pose. Developing a promising strategy to deal with non-state security actors requires deeper knowledge, particularly about the roles of different groups, their relations and interactions and their impact in rural areas. Further research is essential to ensure the development of effective and sustainable reform programmes. Much has improved in both countries, where less than a decade ago people lived in fear and uncertainty. If one is to ensure the long-term sustainability of any reform efforts that will institutionalise professionalism in the security sector, those involved on the ground must endeavour to appreciate the context within which they operate. The now commonly accepted postulate of local ownership must go beyond mere rhetoric to acknowledge the heterogeneity of local actors, and the externally-driven reform strategies must be developed through a partnership with the societies in which they operate in order to ensure that institutions are built on a solid foundation that is cognisant of local conditions. Annexes Annex I: Survey Poll i. Questionnaire: Liberia Questionnaire Perceptions of Security in Liberia Oligopolies of Violence in Liberia and Sierra Leone Institute of African Affairs (Hamburg, Germany) Field Number [supervisor use only]: Respondent Number [office use only]: Questionnaire back-checked by: [supervisor use only] …… [Interviewer: Fill in boxes] District Area …………… ………………… Coding checked by: [office use only] Enumeration Area ……………… …….. Street ……………… A. Social background Let us start with a few questions about yourself. Male 1. [Do not read out] Interviewer: What is the respondent’s gender? 2. How old were you at your last birthday? Could not determine age Age [office use only] Female 164 Judy Smith-Höhn 3. What is your marital status? Married Unmarried Divorced Widowed No answer 4. How many children do you have staying with you at home? 5. What is your county of origin? Bomi Margibi Bong Maryland Gbarpolu Montserrado/Monrovia Grand Bassa Nimba Grand Cape Mount River Cess Grand Gedeh River Gee Grand Kru Sinoe Lofa Other: ………………………………………………………………… No answer 6. What is your tribe? Kpelle Loma Bassa Kissi Gio Vai Kru Dei Grebo Bella Mano Mandingo Krahn Mende Gola Americo-Liberians Gbandi Congo Other: ………………………………………………………………….. None No answer …… none Annexes 7. Have you spent more years of your life in a village or in a town? Village Town About equally Don’t know No answer 8. What is your current occupation? Teacher Office clerk Artisan Housewife Student/pupil Fisherman/fish trader Farm worker/tenant Businessman/woman Farmer Trader (street vendor) Government employee Unemployed Employee (private business) Other: ………………………………………………………………… No answer 9. What is your religion/denomination, if any? Roman Catholic African Methodist Episcopal Church African Methodist Episcopal Zion Church Baptist Church Methodist Episcopal Church Presbyterian Church Episcopal Church Pentecostal Church Islam (Sunni) Islam (other) Traditional religion None Other [specify] …………………………………… No answer 10. In what grade did you leave school? No schooling Grade 11–12 Grade 1–2 College/university Grade 3–4 Grade 5–6 [Interviewer: Do not read] Don’t know Grade 7–8 No answer Grade 9–10 165 166 Judy Smith-Höhn B. Perceptions of personal safety/security 11. In your opinion, how safe is the country today? Not safe at all Fairly unsafe It’s okay Very safe [Interviewer: Do not read] Don’t know No answer 12. In your opinion, compared to before the end of the last war, what is the general state of security like in the country today? Much worse Worse No change Somewhat better Much better [Interviewer: Do not read] Don’t know No answer 13. How safe do you feel personally in the country today? Not safe at all Fairly unsafe It’s okay Very safe [Interviewer: Do not read] Don’t know No answer 14. Compared to before the end of the last war, how is your personal safety today? Very much worse Somewhat worse No change Somewhat better Very much better [Interviewer: Do not read] Don’t know No answer Annexes 167 15. We have spoken to many Liberians and they all have different feelings about which groups in the country protect them and which are a threat to them. Could you please tell us which group is the most important one for your personal safety? [Interviewer: Enter respondent’s exact response, in English, in the line below] …………………………….……………………………………… [Interviewer: Do not read] Don’t know No answer 16. Which group do you feel is the biggest threat to you personally? [Interviewer: Enter respondent’s exact response, in English, in the line below] ……………………………………………… [Interviewer: Do not read] Don’t know No answer 168 Judy Smith-Höhn 17. Now, I’m going to read out a list of groups which are said to affect security in one way or another. For each one, could you tell me whether you feel protected or threatened by them or not? Is very Is some- Does Is some- Is a big [Do No import- what what a threat not answer not ant for import- affect threat for my read] my ant for for my personal Don’t my personal my personal personal security know security personal security security security at all A. Liberian National Police B. Armed Forces of Liberia C. UN peacekeepers (UNMIL) D. Poro/Sandee (secret societies) E. Private security companies (e.g. Dyncorp Intl, InterCom Security) F. Ex-combatants (MODEL, LURD, Taylor government) G. Political Party Militias H. Vigilantes/area teams/ neighbourhood watch J. Street boys K. Other (specify) ………….…… Thank you very much! Annexes 169 ii. Questionnaire: Sierra Leone Questionnaire Perceptions of Security in Sierra Leone Oligopolies of Violence in Liberia and Sierra Leone Institute of African Affairs (Hamburg, Germany) Field Number [supervisor use only]: Respondent Number [office use only]: Questionnaire back-checked by: [supervisor use only] …… [Interviewer: Fill in boxes] District Area …………… ………………… Coding checked by: [office use only] Enumeration Area ……………… …….. Street ……………… A. Social background Let us start with a few questions about yourself. Male Female 1. [Do not read out] Interviewer: What is the respondent’s gender? 2. How old did you turn at your last birthday? Could not determine age Age [office use only] 3. What is your marital status? Married Unmarried Divorced Widowed No answer 4. How many children do you have staying with you at home? …… none 170 Judy Smith-Höhn 5. What is your district of origin? Bo Kono Bombali Port Loko Bonthe Pujehun Kailahun Tonkolili Kambia Western Area rural Kenema Western Area urban Koinadugu Other: ………………………………………………………………… No answer 6. Which tribe do you belong to? Creole Mandingo Fula Mende Gola Sape Kissi Sherbro Kono Susu Koranko Temne Krim Vai Limba Yalunka Loko Other: ………………………………………………………………… No answer 7. Have you spent more years of your life in a village or in a town? Village Town About equally Don’t know No answer Annexes 8. What is your current occupation? Teacher Office clerk Artisan Housewife Student/pupil Fisherman/fish trader Farm worker/tenant Businessman/woman Farmer Trader (street vendor) Government employee Unemployed Employee (private business) Other [specify]:………………………………………………………….. No answer 9. What, if any, is your religious denomination? African Methodist Episcopal Church Baptist Church Christ Apostolic Church Evangelical Lutheran Church Freetown Anglican Diocese Islam Methodist Church Pentecostal Church Roman Catholic Traditional religion None Other [specify]………………………………………………………… No answer 10. In what grade did you leave school? No schooling Class 1–2 Class 3–4 Class 5–6 Junior secondary 1–3 Senior secondary 1–3 College/university [Interviewer: Do not read] Don’t know No answer 171 172 Judy Smith-Höhn B. Perceptions of personal safety/security 11. In your opinion, how safe is the country today? Not safe at all Fairly unsafe It’s okay Very safe [Interviewer: Do not read] Don’t know No answer 12. In your opinion, compared to before the end of the war, what is the general state of security like in the country today? Much worse Worse No change Somewhat better Much better [Interviewer: Do not read] Don’t know No answer 13. How safe do you feel personally in the country today? Not safe at all Fairly unsafe It’s okay Very safe [Interviewer: Do not read] Don’t know No answer 14. Compared to before the end of the war, how is your personal safety today? Very much worse Somewhat worse No change Somewhat better Very much better [Interviewer: Do not read] Don’t know No answer Annexes 173 15. We have spoken to many Sierra Leoneans and they all have different feelings about which groups in the country protect them and which are a threat to them. Could you please tell us which group is the most important one for your personal safety? [Interviewer: Enter respondent’s exact response, in English, in the line below] ……………………………………………………………………………………… [Interviewer: Do not read] Don’t know No answer 16. Which group do you feel is the biggest threat to you personally? [Interviewer: Enter respondent’s exact response, in English, in the line below] ………………………………………………………………………………… [Interviewer: Do not read] Don’t know No answer 174 Judy Smith-Höhn 17. Now I’m going to read out a list of groups that are said to affect security in one way or another. For each one, could you tell me whether you feel protected or threatened by them? Is very Is Does not Is Is a big important somewhat affect somewhat threat for for my important my a threat my personal for my personal for my personal security personal security personal security security at all security A. Sierra Leone Police B. Republic of Sierra Leone Armed Forces C. Mines monitoring officers D. Traffic wardens E. Secret societies (Poro/Bondo, Ojeh) F. Private security companies G. National Union of Sierra Leonean Students H. Youth wings of political parties I. Ghetto Boys J. Bike riders Ex-combatants K. West Side Boys L. Civil defence forces (CDF) M. Revolutionary United Front (RUF) N. Other (specify) ………………… Thank you very much! No [Do not answer read] Don’t know Annexes 175 iii. Data analysis – Frequencies Liberia Table A1: District Count Valid % % Buchanan 100 14.33 14.33 Monrovia 498 71.35 71.35 Tubmanburg 100 14.33 14.33 Valid N 698 100.00 100.00 Total 698 100.00 100.00 Table A2: Survey area Count Valid % % Buchanan 100 14.33 14.33 Central Monrovia 100 14.33 14.33 Gardnersville 99 14.18 14.18 Gorblah 34 4.87 4.87 Joseph Street 25 3.58 3.58 New Kru Town 99 14.18 14.18 Paynesville 100 14.33 14.33 Rock Town 25 3.58 3.58 Vai Town 41 5.87 5.87 Water Side 51 7.31 7.31 West Point 24 3.44 3.44 Valid N 698 100.00 100.00 Total 698 100.00 100.00 176 Table A3: Judy Smith-Höhn Q1: Sex Count Valid % % Male 327 47.60 46.85 Female 360 52.40 51.58 Valid N 687 100.00 98.42 Missing 11 0.00 1.58 Total 698 100.00 100.00 Mean Valid N Maximum Minimum 32.7 690 78 16 Table A4: Q2: Age Annexes Table A5: 177 Q2_Kat: Age group Count Valid % % Up to 20 years 88 12.75 12.61 21–30 years 263 38.12 37.68 31–40 years 195 28.26 27.94 41–50 years 90 13.04 12.89 51–60 years 40 5.80 5.73 61 years and older 14 2.03 2.01 Valid N 690 100.00 98.85 No answer 8 0.00 1.15 Don’t know 0 0.00 0.00 698 100.00 100.00 Total Table A6: Q3: Marital status Count Valid % % Married 312 45.81 44.70 Unmarried 311 45.67 44.56 Divorced 27 3.96 3.87 Widowed 31 4.55 4.44 Valid N 681 100.00 97.56 No answer 17 0.00 2.44 Total 698 100.00 100.00 178 Table A7: Judy Smith-Höhn Q4: How many children do you have staying with you at home? Count Valid % % None 176 26.15 25.21 1 75 11.14 10.74 2 111 16.49 15.90 3 76 11.29 10.89 4 64 9.51 9.17 5 61 9.06 8.74 6 35 5.20 5.01 7 22 3.27 3.15 8 15 2.23 2.15 9 7 1.04 1.00 10 14 2.08 2.01 11 5 0.74 0.72 12 5 0.74 0.72 13 2 0.30 0.29 14 1 0.15 0.14 15 1 0.15 0.14 16 1 0.15 0.14 17 1 0.15 0.14 25 1 0.15 0.14 Valid N 673 100.00 96.42 No answer 25 0.00 3.58 Total 698 100.00 100.00 Annexes Table A8: 179 Q5: What is your county of origin? Count Valid % % Bomi 65 9.31 9.31 Bong 56 8.02 8.02 Gbarpolu 24 3.44 3.44 Grand Bassa 86 12.32 12.32 Grand Cape Mount 43 6.16 6.16 Grand Gedeh 28 4.01 4.01 Grand Kru 46 6.59 6.59 Lofa 93 13.32 13.32 Margibi 28 4.01 4.01 Maryland 56 8.02 8.02 Montserrado/Monrovia 54 7.74 7.74 Nimba 39 5.59 5.59 River Cess 27 3.87 3.87 River Gee 9 1.29 1.29 Sinoe 44 6.30 6.30 Other 0 0.00 0.00 698 100.00 100.00 0 0.00 0.00 698 100.00 100.00 Valid N No answer Total Table A9: Total Q5_other: Other county of origin? Count Valid % % 698 100.00 100.00 698 100.00 100.00 180 Judy Smith-Höhn Table A10: Q6: What is your tribe? Count Valid % % Kpelle 86 12.34 12.32 Bassa 98 14.06 14.04 Gio 27 3.87 3.87 Kru 83 11.91 11.89 Grebo 73 10.47 10.46 Mano 29 4.16 4.15 Krahn 27 3.87 3.87 Gola 51 7.32 7.31 Gbandi 37 5.31 5.30 Loma 36 5.16 5.16 Kissi 41 5.88 5.87 Vai 42 6.03 6.02 Dei 5 0.72 0.72 Bella 11 1.58 1.58 Mandingo 11 1.58 1.58 Mende 10 1.43 1.43 Americo-Liberian 4 0.57 0.57 Congo 21 3.01 3.01 Other 5 0.72 0.72 None Valid N No answer Total 0 0.00 0.00 697 100.00 99.86 1 0.00 0.14 698 100.00 100.00 Annexes 181 Table A11: Q6_other: Other tribe? Count Valid % % 693 99.28 99.28 Fula 1 0.14 0.14 Gbarpor 1 0.14 0.14 Gbee 1 0.14 0.14 Naturalised Liberian 1 0.14 0.14 Zaibor 1 0.14 0.14 Total 698 100.00 100.00 Table A12: Q7: Have you spent more years of your life in a village or in a town? Count Valid % % Village 112 16.42 16.05 Town 486 71.26 69.63 About equally 84 12.32 12.03 Don’t know 0 0.00 0.00 Valid N 682 100.00 97.71 No answer 16 0.00 2.29 Total 698 100.00 100.00 182 Judy Smith-Höhn Table A13: Q8_rec: What is your current occupation? Count Valid % % Teacher; student/pupil 218 31.50 31.23 Government employee 53 7.66 7.59 Artisan 8 1.16 1.15 Fisherman; fish trader; farm worker/tenant; farmer 53 7.66 7.59 Housewife 60 8.67 8.60 Businessman/woman; trader (street vendor) 143 20.66 20.49 Office clerk; employee (private business) 82 11.85 11.75 Unemployed 57 8.24 8.17 Other 18 2.60 2.58 Valid N 692 100.00 99.14 6 0.00 0.86 698 100.00 100.00 No answer Total Table A14: Q8_other: Other occupation? Count Valid % % 683 97.85 97.85 Carpenter 2 0.29 0.29 Domestic work 1 0.14 0.14 Driver 1 0.14 0.14 Footballer 1 0.14 0.14 Nurse 1 0.14 0.14 Operator 2 0.29 0.29 Pastor/administrator 1 0.14 0.14 Private security 2 0.29 0.29 Reception 1 0.14 0.14 Social worker 1 0.14 0.14 Tailor 2 0.29 0.29 Total 698 100.00 100.00 Annexes 183 Table A15: Q9_rec: What is your religious denomination, if any? Count Valid % % Christian 532 76.55 76.22 Muslim 91 13.09 13.04 Traditional religion 6 0.86 0.86 Other 55 7.91 7.88 None 11 1.58 1.58 Valid N 695 100.00 99.57 3 0.00 0.43 698 100.00 100.00 No answer Total 184 Judy Smith-Höhn Table A16: Q9_other: Other religious denomination, if any? Count Valid % % 645 92.41 92.41 Church of God 1 0.14 0.14 Dominion Christian Fellowship 1 0.14 0.14 First Assembly of God Church 8 1.15 1.15 Glory of Church 1 0.14 0.14 Greater Grace Church 1 0.14 0.14 Jehovahs Witness 1 0.14 0.14 Kingdom Assembly 2 0.29 0.29 Liberia Open Bible Church 1 0.14 0.14 Lutheran Church 12 1.72 1.72 New Apostolic 1 0.14 0.14 Poutal House Church 1 0.14 0.14 Repentance Baptist Church 1 0.14 0.14 Salvation Army 8 1.14 1.14 Seventh Day Adventist 6 0.86 0.86 The Church of Jesus Christ of Latter Day Saints 1 0.14 0.14 United Liberian Inland Church 5 0.72 0.72 United Methodist Church 1 0.14 0.14 Victory Chapel 1 0.14 0.14 698 100.00 100.00 Total Annexes 185 Table A17: Q10: In what grade did you leave school? Count Valid % % No schooling 75 11.76 10.74 Grade 1–2 3 0.47 0.43 Grade 3–4 16 2.51 2.29 Grade 5–6 37 5.80 5.30 Grade 7–8 72 11.29 10.32 Grade 9–10 95 14.89 13.61 Grade 11–12 265 41.54 37.97 College/university 75 11.76 10.74 Valid N 638 100.00 91.40 No answer 34 0.00 4.87 Don’t know 26 0.00 3.72 Total 698 100.00 100.00 Table A18: Q11: In your opinion, how safe is the country today? Count Valid % % Not safe at all 73 10.69 10.46 Fairly unsafe 188 27.53 26.93 It’s okay 280 41.00 40.11 Very safe 142 20.79 20.34 Valid N 683 100.00 97.85 No answer 15 0.00 2.15 Don’t know Total 0 0.00 0.00 698 100.00 100.00 186 Judy Smith-Höhn Table A19: Q12: In your opinion, compared to before the end of the last war, what is the general state of security like in the country today? Count Valid % % 18 2.62 2.58 Much worse Worse 31 4.52 4.44 No change 44 6.41 6.30 Somewhat better 358 52.19 51.29 Much better 235 34.26 33.67 Valid N 686 100.00 98.28 No answer 12 0.00 1.72 Don’t know 0 0.00 0.00 698 100.00 100.00 Total Table A20: Q13: How safe do you feel personally in the country today? Count Valid % % Not safe at all 43 6.22 6.16 Fairly unsafe 164 23.73 23.50 It’s okay 305 44.14 43.70 Very safe 179 25.90 25.64 Valid N 691 100.00 99.00 No answer 7 0.00 1.00 Don’t know 0 0.00 0.00 698 100.00 100.00 Total Annexes 187 Table A21: Q14: Compared to before the end of the last war, how is your personal safety today? Very much worse Count Valid % % 17 2.47 2.44 Somewhat worse 32 4.64 4.58 No change 39 5.66 5.59 Somewhat better 403 58.49 57.74 Very much better 198 28.74 28.37 Valid N 689 100.00 98.71 No answer 9 0.00 1.29 Don’t know 0 0.00 0.00 698 100.00 100.00 Total Table A22: Q15_rec: Which group is the most important one for your personal safety? UNMIL Count Valid % % 530 78.64 75.93 Armed Forces of Liberia 14 2.08 2.01 Liberia National Police 125 18.55 17.91 1 0.15 0.14 Ex-combatants ECOWAS 1 0.15 0.14 Poro/Sandee 1 0.15 0.14 Vigilante teams 1 0.15 0.14 Family 1 0.15 0.14 Valid N 674 100.00 96.56 No answer 14 0.00 2.01 Don’t know 10 0.00 1.43 Total 698 100.00 100.00 188 Judy Smith-Höhn Table A23: Q16_rec: Which group do you feel is the biggest threat to you personally? Count Valid % % Armed Forces of Liberia 8 1.15 1.15 Armed robbers and criminals 26 3.72 3.72 Congress for Democratic Change 3 0.43 0.43 Demonstrators 1 0.14 0.14 Ex-combatants 415 59.46 59.46 Liberia National Police 14 2.01 2.01 Militias 2 0.29 0.29 Political party militias 10 1.43 1.43 Poro/Sandee (secret societies) 3 0.43 0.43 Private security companies 1 0.14 0.14 Street boys 55 7.88 7.88 UNMIL 6 0.86 0.86 Vigilantes 1 0.14 0.14 None 32 4.58 4.58 Former warring faction 1: Taylor government 2 0.29 0.29 Former warring faction 2: MODEL 3 0.43 0.43 Former warring faction 3: LURD 4 0.57 0.57 Valid N 586 83.95 83.95 No answer 35 5.01 5.01 Don’t know 77 11.03 11.03 Total 698 100.00 100.00 Annexes 189 Table A24: Q17_A: Could you tell me whether you feel protected or threatened by them or not: Liberia National Police? Is a big threat for my personal security Count Valid % % 5 0.72 0.72 Is somewhat a threat for my personal security 15 2.16 2.15 Does not affect my personal security at all 32 4.62 4.58 Is somewhat important for my personal security 223 32.18 31.95 Is very important for my personal security 418 60.32 59.89 Valid N 693 100.00 99.28 No answer 3 0.00 0.43 Don’t know 2 0.00 0.29 698 100.00 100.00 Total Table A25: Q17_B: Could you tell me whether you feel protected or threatened by them or not: Armed Forces of Liberia? Is a big threat for my personal security Count Valid % % 30 4.39 4.30 Is somewhat a threat for my personal security 35 5.12 5.01 Does not affect my personal security at all 166 24.30 23.78 Is somewhat important for my personal security 281 41.14 40.26 Is very important for my personal security 171 25.04 24.50 Valid N 683 100.00 97.85 No answer 11 0.00 1.58 Don’t know 4 0.00 0.57 698 100.00 100.00 Total 190 Judy Smith-Höhn Table A26: Q17_C: Could you tell me whether you feel protected or threatened by them or not: UN peacekeepers (UNMIL)? Is a big threat for my personal security Count Valid % % 5 0.73 0.72 Is somewhat a threat for my personal security 5 0.73 0.72 Does not affect my personal security at all 14 2.04 2.01 Is somewhat important for my personal security 52 7.58 7.45 Is very important for my personal security 610 88.92 87.39 Valid N 686 100.00 98.28 No answer 11 0.00 1.58 Don’t know 1 0.00 0.14 698 100.00 100.00 Total Table A27: Q17_D: Could you tell me whether you feel protected or threatened by them or not: Poro (secret societies)? Is a big threat for my personal security Count Valid % % 129 20.94 18.48 Is somewhat a threat for my personal security 53 8.60 7.59 Does not affect my personal security at all 324 52.60 46.42 Is somewhat important for my personal security 56 9.09 8.02 Is very important for my personal security 54 8.77 7.74 Valid N 616 100.00 88.25 No answer 47 0.00 6.73 Don’t know 35 0.00 5.01 Total 698 100.00 100.00 Annexes 191 Table A28: Q17_E: Could you tell me whether you feel protected or threatened by them or not: Private security companies (e.g. Dyncorp Intl, Inter-Com Security)? Is a big threat for my personal security Count Valid % % 6 0.91 0.86 Is somewhat a threat for my personal security 20 3.04 2.87 Does not affect my personal security at all 364 55.40 52.15 Is somewhat important for my personal security 147 22.37 21.06 Is very important for my personal security 120 18.26 17.19 Valid N 657 100.00 94.13 No answer 23 0.00 3.30 Don’t know 18 0.00 2.58 Total 698 100.00 100.00 Table A29: Q17_F: Could you tell me whether you feel protected or threatened by them or not: Ex-combatants (MODEL, LURD, Taylor government)? Count Valid % % Is a big threat for my personal security 491 72.53 70.34 Is somewhat a threat for my personal security 115 16.99 16.48 Does not affect my personal security at all 48 7.09 6.88 Is somewhat important for my personal security 9 1.33 1.29 Is very important for my personal security 14 2.07 2.01 Valid N 677 100.00 96.99 No answer 10 0.00 1.43 Don’t know 11 0.00 1.58 Total 698 100.00 100.00 192 Judy Smith-Höhn Table A30: Q17_G: Could you tell me whether you feel protected or threatened by them or not: Political party militias? Is a big threat for my personal security Count Valid % % 172 28.57 24.64 Is somewhat a threat for my personal security 127 21.10 18.19 Does not affect my personal security at all 253 42.03 36.25 Is somewhat important for my personal security 32 5.32 4.58 Is very important for my personal security 18 2.99 2.58 Valid N 602 100.00 86.25 No answer 63 0.00 9.03 Don’t know 33 0.00 4.73 Total 698 100.00 100.00 Table A31: Q17_H: Could you tell me whether you feel protected or threatened by them or not: Vigilantes/ area teams/ neighbourhood watch? Count Valid % % Is a big threat for my personal security 39 6.21 5.59 Is somewhat a threat for my personal security 69 10.99 9.89 Does not affect my personal security at all 125 19.90 17.91 Is somewhat important for my personal security 221 35.19 31.66 Is very important for my personal security 174 27.71 24.93 Valid N 628 100.00 89.97 No answer 29 0.00 4.15 Don’t know 41 0.00 5.87 Total 698 100.00 100.00 Annexes 193 Table A32: Q17_J: Could you tell me whether you feel protected or threatened by them or not: Street boys? Count Valid % % Is a big threat for my personal security 438 73.61 62.75 Is somewhat a threat for my personal security 110 18.49 15.76 Does not affect my personal security at all 44 7.39 6.30 Is somewhat important for my personal security 1 0.17 0.14 Is very important for my personal security 2 0.34 0.29 Valid N 595 100.00 85.24 No answer 85 0.00 12.18 Don’t know 18 0.00 2.58 Total 698 100.00 100.00 Table A33: Q17_rA: Could you tell me whether you feel protected or threatened by them or not: Liberia National Police? Is very/somewhat important for my personal security Count Valid % % 641 92.50 92.50 Does not affect my personal security at all 32 4.62 4.62 Is a big/somewhat a threat for my personal security 20 2.89 2.89 Valid N 693 100.00 100.00 No answer 0 0.00 0.00 Don’t know 0 0.00 0.00 693 100.00 100.00 Total 194 Judy Smith-Höhn Table A34: Q17_rB: Could you tell me whether you feel protected or threatened by them or not: Armed Forces of Liberia? Count Valid % % Is very/somewhat important for my personal security 452 66.18 66.18 Does not affect my personal security at all 166 24.30 24.30 Is a big/somewhat a threat for my personal security 65 9.52 9.52 Valid N 683 100.00 100.00 No answer 0 0.00 0.00 Don’t know 0 0.00 0.00 683 100.00 100.00 Total Table A35: Q17_rC: Could you tell me whether you feel protected or threatened by them or not: UN peacekeepers (UNMIL)? Count Valid % % Is very/somewhat important for my personal security 662 96.50 96.50 Does not affect my personal security at all 14 2.04 2.04 Is a big/somewhat a threat for my personal security 10 1.46 1.46 Valid N 686 100.00 100.00 No answer 0 0.00 0.00 Don’t know 0 0.00 0.00 686 100.00 100.00 Total Annexes 195 Table A36: Q17_rD: Could you tell me whether you feel protected or threatened by them or not: Poro (secret societies)? Is very/somewhat important for my personal security Count Valid % % 110 17.86 17.86 Does not affect my personal security at all 324 52.60 52.60 Is a big/somewhat a threat for my personal security 182 29.55 29.55 Valid N 616 100.00 100.00 0 0.00 0.00 No answer Don’t know Total 0 0.00 0.00 616 100.00 100.00 Table A37: Q17_rE: Could you tell me whether you feel protected or threatened by them or not: Private security companies (e.g. Dyncorp Intl, Inter-Com Security)? Count Valid % % Is very/somewhat important for my personal security 267 40.64 40.64 Does not affect my personal security at all 364 55.40 55.40 Is a big/somewhat a threat for my personal security 26 3.96 3.96 Valid N 657 100.00 100.00 No answer 0 0.00 0.00 Don’t know 0 0.00 0.00 657 100.00 100.00 Total 196 Judy Smith-Höhn Table A38: Q17_rF: Could you tell me whether you feel protected or threatened by them or not: Ex-combatants (MODEL, LURD, Taylor government)? Is very/somewhat important for my personal security Count Valid % % 23 3.40 3.40 Does not affect my personal security at all 48 7.09 7.09 Is a big/somewhat a threat for my personal security 606 89.51 89.51 Valid N 677 100.00 100.00 No answer 0 0.00 0.00 Don’t know 0 0.00 0.00 677 100.00 100.00 Total Table A39: Q17_rG: Could you tell me whether you feel protected or threatened by them or not: Political party militias? Count Valid % % Is very/somewhat important for my personal security 50 8.31 8.31 Does not affect my personal security at all 253 42.03 42.03 Is a big/somewhat a threat for my personal security 299 49.67 49.67 Valid N 602 100.00 100.00 No answer 0 0.00 0.00 Don’t know 0 0.00 0.00 602 100.00 100.00 Total Annexes 197 Table A40: Q17_rH: Could you tell me whether you feel protected or threatened by them or not: Vigilantes/area teams/ neighbourhood watch? Is very/somewhat important for my personal security Count Valid % % 395 62.90 62.90 Does not affect my personal security at all 125 19.90 19.90 Is a big/somewhat a threat for my personal security 108 17.20 17.20 Valid N 628 100.00 100.00 No answer 0 0.00 0.00 Don’t know 0 0.00 0.00 628 100.00 100.00 Total Table A41: Q17_rJ: Could you tell me whether you feel protected or threatened by them or not: Street boys? Count Valid % % Is very/somewhat important for my personal security 3 0.50 0.50 Does not affect my personal security at all 44 7.39 7.39 Is a big/somewhat a threat for my personal security 548 92.10 92.10 Valid N 595 100.00 100.00 No answer 0 0.00 0.00 Don’t know 0 0.00 0.00 595 100.00 100.00 Total 198 Judy Smith-Höhn iv. Data analysis – Frequencies Sierra Leone Table A42: District Count Valid % % Freetown 501 71.37 71.37 Makeni 100 14.25 14.25 Koidu 101 14.39 14.39 Valid N 702 100.00 100.00 Total 702 100.00 100.00 Table A43: Survey area Count Valid % % Freetown: Central 100 14.25 14.25 Freetown: East 1 100 14.25 14.25 Freetown: East 2 100 14.25 14.25 Freetown: West 1 100 14.25 14.25 Freetown: West 2 101 14.39 14.39 Makeni: Zone 1 25 3.56 3.56 Makeni: Zone 2 25 3.56 3.56 Makeni: Zone 3 25 3.56 3.56 Makeni: Zone 4 25 3.56 3.56 Koidu: Central 25 3.56 3.56 Koidu: Bungalow 26 3.70 3.70 Koidu: Hill station 25 3.56 3.56 Koidu: Tankoro 25 3.56 3.56 Valid N 702 100.00 100.00 Total 702 100.00 100.00 Annexes 199 Table A44: Q1: Sex Count Valid % % Male 355 50.71 50.57 Female 345 49.29 49.15 Valid N 700 100.00 99.72 Missing 2 0.00 0.28 702 100.00 100.00 Mean Valid N Maximum Minimum 34.1 641 85 18 Total Table A45: Q2: Age 200 Judy Smith-Höhn Table A46: Q2_Kat: Age group Count Valid % % Up to 20 years 74 11.54 10.54 21–30 years 255 39.78 36.32 31–40 years 152 23.71 21.65 41–50 years 77 12.01 10.97 51–60 years 49 7.64 6.98 61 years and older 34 5.30 4.84 Valid N 641 100.00 91.31 No answer 61 0.00 8.69 Don’t know 0 0.00 0.00 702 100.00 100.00 Total Table A47: Q3: Marital status Count Valid % % Married 363 51.86 51.71 Unmarried 264 37.71 37.61 Divorced 36 5.14 5.13 Widowed 37 5.29 5.27 Valid N 700 100.00 99.72 2 0.00 0.28 702 100.00 100.00 No answer Total Annexes 201 Table A48: Q4: How many children do you have staying with you at home? Count Valid % % None 203 29.38 28.92 1 99 14.33 14.10 2 109 15.77 15.53 3 92 13.31 13.11 4 63 9.12 8.97 5 36 5.21 5.13 6 27 3.91 3.85 7 18 2.60 2.56 8 13 1.88 1.85 9 6 0.87 0.85 10 4 0.58 0.57 11 7 1.01 1.00 12 4 0.58 0.57 13 2 0.29 0.28 14 2 0.29 0.28 15 3 0.43 0.43 16 1 0.14 0.14 17 1 0.14 0.14 18 1 0.14 0.14 Valid N 691 100.00 98.43 No answer 11 0.00 1.57 Total 702 100.00 100.00 202 Judy Smith-Höhn Table A49: Q5: What is your county of origin? Count Valid % % Bo 34 4.86 4.84 Bombali 121 17.29 17.24 Bonthe 10 1.43 1.42 Kailahun 19 2.71 2.71 Kambia 47 6.71 6.70 Kenema 30 4.29 4.27 Koinadugu 39 5.57 5.56 Kono 87 12.43 12.39 Port Loko 69 9.86 9.83 Pujehun 20 2.86 2.85 Tonkolili 44 6.29 6.27 Western Area rural 34 4.86 4.84 Western Area urban 104 14.86 14.81 Moyamba 40 5.71 5.70 Other 2 0.29 0.28 700 100.00 99.72 Valid N No answer Total 2 0.00 0.28 702 100.00 100.00 Table A50: Q5_other: Other county of origin? Count Valid % % 700 99.72 99.72 Liberian 1 0.14 0.14 Makeni 1 0.14 0.14 702 100.00 100.00 Total Annexes 203 Table A51: Q6: What is your tribe? Count Valid % % Creole 56 7.98 7.98 Fula 51 7.26 7.26 Gola 1 0.14 0.14 Kissi 14 1.99 1.99 Kono 67 9.54 9.54 Koranko 21 2.99 2.99 Krim 3 0.43 0.43 Limba 53 7.55 7.55 Loko 21 2.99 2.99 Mandingo 28 3.99 3.99 Mende 117 16.67 16.67 Sape 3 0.43 0.43 Sherbro 23 3.28 3.28 Susu 28 3.99 3.99 Temne 210 29.91 29.91 2 0.28 0.28 Vai Yalunka 1 0.14 0.14 Other 3 0.43 0.43 None 0 0.00 0.00 702 100.00 100.00 0 0.00 0.00 702 100.00 100.00 Valid N No answer Total Table A52: Q6_other: Other tribe? Count Valid % % 699 99.57 99.57 Bassa (Liberia) 1 0.14 0.14 Kru 2 0.29 0.29 Total 702 100.00 100.00 204 Judy Smith-Höhn Table A53: Q7: Have you spent more years of your life in a village or in a town? Count Valid % % Village 97 13.90 13.82 Town 556 79.66 79.20 About equally 45 6.45 6.41 Valid N 698 100.00 99.43 No answer 1 0.00 0.14 Don’t know 3 0.00 0.43 702 100.00 100.00 Total Table A54: Q8_rec: What is your current occupation? Teacher; student/pupil Count Valid % % 181 25.86 25.78 Government employee 55 7.86 7.83 Artisan 27 3.86 3.85 Fisherman; fish trader; farm worker/tenant; farmer 37 5.29 5.27 Housewife 45 6.43 6.41 Businessman/woman; trader (street vendor) 159 22.71 22.65 Office clerk; employee (private business) 107 15.29 15.24 Unemployed 62 8.86 8.83 Other 27 3.86 3.85 Valid N 700 100.00 99.72 2 0.00 0.28 702 100.00 100.00 No answer Total Annexes 205 Table A55: Q8_other: Other occupation? Count Valid % % 676 96.30 96.30 Computer instructor 1 0.14 0.14 Diamond digger 4 0.57 0.57 Driver (commercial) 2 0.28 0.28 Footballer 2 0.28 0.28 Hairdresser 5 0.71 0.71 Labourer 1 0.14 0.14 Legal researcher 1 0.14 0.14 Librarian 1 0.14 0.14 Miner 1 0.14 0.14 Photographer 1 0.14 0.14 Retired 1 0.14 0.14 Seamstress 1 0.14 0.14 Security officer 2 0.28 0.28 Soldier 1 0.14 0.14 Solicitor’s secretary 1 0.14 0.14 Traditional leader Total 1 0.14 0.14 702 100.00 100.00 Table A56: Q9_rec: What is your religious denomination, if any? Count Valid % % Christian 287 41.24 40.88 Muslim 388 55.75 55.27 Traditional religion 6 0.86 0.85 Other 15 2.16 2.14 None 0 0.00 0.00 696 100.00 99.15 6 0.00 0.85 702 100.00 100.00 Valid N No answer Total 206 Judy Smith-Höhn Table A57: Q9_other: Other religious denomination, if any? Count Valid %t % 687 97.86 97.86 Assemblies of God 1 0.14 0.14 Christian Ministry Church 1 0.14 0.14 Faith Healing Bible Church 1 0.14 0.14 Grace Assembly of God Church 1 0.14 0.14 Holy Ghost Bible Church 1 0.14 0.14 Holy Trinity Church 1 0.14 0.14 International Church of Christ 1 0.14 0.14 Jehovahs Witness 1 0.14 0.14 Joint Ministry Church 1 0.14 0.14 Rastafarian 1 0.14 0.14 Salvation Spiritual Church 1 0.14 0.14 Seventh Day Adventist 2 0.28 0.28 Waeslan Church 1 0.14 0.14 Winners Chapel International Total 1 0.14 0.14 702 100.00 100.00 Table A58: Q10: In what grade did you leave school? Count Valid % % 148 21.26 21.08 Class 1–2 4 0.57 0.57 Class 3–4 17 2.44 2.42 No schooling Class 5–6 38 5.46 5.41 Junior secondary 1–3 119 17.10 16.95 Senior secondary 1–3 256 36.78 36.47 College/university 114 16.38 16.24 Valid N 696 100.00 99.15 No answer 5 0.00 0.71 Don’t know 1 0.00 0.14 702 100.00 100.00 Total Annexes 207 Table A59: Q11: In your opinion, how safe is the country today? Count Valid % % Not safe at all 94 13.62 13.39 Fairly unsafe 266 38.55 37.89 It’s okay 227 32.90 32.34 Very safe 103 14.93 14.67 Valid N 690 100.00 98.29 No answer 12 0.00 1.71 Don’t know 0 0.00 0.00 702 100.00 100.00 Total Table A60: Q12: In your opinion, compared to before the end of the last war, what is the general state of security like in the country today? Count Valid % % Much worse 3 0.43 0.43 Worse 15 2.16 2.14 No change 32 4.61 4.56 Somewhat better 421 60.66 59.97 Much better 223 32.13 31.77 Valid N 694 100.00 98.86 8 0.00 1.14 No answer Don’t know Total 0 0.00 0.00 702 100.00 100.00 208 Judy Smith-Höhn Table A61: Q13: How safe do you feel personally in the country today? Count Valid % % Not safe at all 72 10.33 10.26 Fairly unsafe 193 27.69 27.49 It’s okay 287 41.18 40.88 Very safe 145 20.80 20.66 Valid N 697 100.00 99.29 No answer 5 0.00 0.71 Don’t know 0 0.00 0.00 702 100.00 100.00 Total Table A62: Q14: Compared to before the end of the last war, how is your personal safety today? Count Valid % % Very much worse 11 1.58 1.57 Somewhat worse 18 2.58 2.56 No change 31 4.45 4.42 Somewhat better 445 63.85 63.39 Very much better 192 27.55 27.35 Valid N 697 100.00 99.29 No answer 5 0.00 0.71 Don’t know 0 0.00 0.00 702 100.00 100.00 Total Annexes 209 Table A63: Q15_rec: Which group is the most important one for your personal safety? Count Valid % % All People’s Congress Armed forces Civil society ECOMOG Ex-combatans Foreign fighters, foreign troops Free Force God Government Husband/my family IMATT Individuals International community Mende National Social Security Information Trust of Sierra Leone Private security Former warring faction 1: CDF Former warring faction 2: RUF Former warring faction 3: West Side Boys None SLP Operational Support Division President Paramount chief Petty Traders Association Sierra Leone Police Rebels SOD UNAMSIL Village authorities Youth groups Valid N No answer Don’t know 1 242 1 21 1 3 1 7 14 6 1 1 1 1 0.15 37.17 0.15 3.23 0.15 0.46 0.15 1.08 2.15 0.92 0.15 0.15 0.15 0.15 0.14 34.47 0.14 2.99 0.14 0.43 0.14 1.00 1.99 0.85 0.14 0.14 0.14 0.14 1 0.15 0.14 2 11 1 0 5 2 6 1 1 233 5 1 76 1 4 651 24 27 0.31 1.69 0.15 0.00 0.77 0.31 0.92 0.15 0.15 35.79 0.77 0.15 11.67 0.15 0.61 100.00 0.00 0.00 0.28 1.57 0.14 0.00 0.71 0.28 0.85 0.14 0.14 33.19 0.71 0.14 10.83 0.14 0.57 92.74 3.42 3.85 Total 702 100.00 100.00 210 Judy Smith-Höhn Table A64: Q16_rec: Which group do you feel is the biggest threat to you personally? Count Valid % % Armed Forces Revolutionary Council Armed forces/army/military/RSLAF Armed persons Armed rebels Armed robbers/bandits/criminal gangs/gangsters/thieves Bike riders Boboh belle* politicians (*Krio term for big belly, meaning greedy) Charles Taylor City council Disgruntled citizens ECOMOG Enemies Ex-combatants/ex-rebels Ex-Sierra Leone Army/retired soldiers Armed men Free-flow police Ghetto Boys Government In-laws Intruders Junta Land thieves Inmates Neighbours Nigerians None Parents Police Political parties/politicans/SLPP Secret societies Special Court for Sierra Leone Temne Traffic wardens Youth Witches Former warring faction 1: CDF Former warring faction 2: RUF Former warring faction 3: West Side Boys Valid N No answer Don’t know 4 90 5 68 89 1 1 3 1 1 3 2 48 5 1 3 7 3 1 2 2 1 1 2 1 30 2 65 13 1 2 2 2 14 2 46 60 9 593 29 70 0.67 15.18 0.84 11.47 15.01 0.17 0.17 0.51 0.17 0.17 0.51 0.34 8.09 0.84 0.17 0.51 1.18 0.51 0.17 0.34 0.34 0.17 0.17 0.34 0.17 5.06 0.34 10.96 2.19 0.17 0.34 0.34 0.34 2.36 0.34 7.76 10.12 1.52 100.00 0.00 0.00 0.58 13.01 0.72 9.83 12.86 0.14 0.14 0.43 0.14 0.14 0.43 0.29 6.94 0.72 0.14 0.43 1.01 0.43 0.14 0.29 0.29 0.14 0.14 0.29 0.14 4.34 0.29 9.39 1.88 0.14 0.29 0.29 0.29 2.02 0.29 6.65 8.67 1.30 85.69 4.19 10.12 Total 692 100.00 100.00 Annexes 211 Table A65: Q17_A: Could you tell me whether you feel protected or threatened by them or not: A: Sierra Leone Police? Is a big threat for my personal security Count Valid % % 23 3.3 3.3 Is somewhat a threat for my personal security 33 4.7 4.7 Does not affect my personal security at all 24 3.4 3.4 Is somewhat important for my personal security 117 16.8 16.7 Is very important for my personal security 499 71.7 71.1 Valid N 696 100.0 99.1 No answer 1 0.0 0.1 Don’t know 5 0.0 0.7 702 100.0 100.0 Total Table A66: Q17_B: Could you tell me whether you feel protected or threatened by them or not: B: Republic of Sierra Leone Armed Forces? Count Valid % % Is a big threat for my personal security 31 4.5 4.4 Is somewhat a threat for my personal security 39 5.6 5.6 Does not affect my personal security at all 35 5.0 5.0 Is somewhat important for my personal security 134 19.3 19.1 Is very important for my personal security 456 65.6 65.0 Valid N 695 100.0 99.0 No answer 1 0.0 0.1 Don’t know 6 0.0 0.9 702 100.0 100.0 Total 212 Judy Smith-Höhn Table A67: Q17_C: Could you tell me whether you feel protected or threatened by them or not: C: Mines monitoring officers? Count Valid % % Is a big threat for my personal security 11 1.9 1.6 Is somewhat a threat for my personal security 25 4.3 3.6 Does not affect my personal security at all 386 66.4 55.0 Is somewhat important for my personal security 72 12.4 10.3 Is very important for my personal security 87 15.0 12.4 Valid N 581 100.0 82.8 No answer 16 0.0 2.3 Don’t know 105 0.0 15.0 Total 702 100.0 100.0 Table A68: Q17_D: Could you tell me whether you feel protected or threatened by them or not: D: Traffic wardens? Count Valid % % Is a big threat for my personal security 20 3.0 2.8 Is somewhat a threat for my personal security 36 5.5 5.1 Does not affect my personal security at all 239 36.3 34.0 Is somewhat important for my personal security 214 32.5 30.5 Is very important for my personal security 149 22.6 21.2 Valid N 658 100.0 93.7 No answer 13 0.0 1.9 Don’t know 31 0.0 4.4 Total 702 100.0 100.0 Annexes 213 Table A69: Q17_E: Could you tell me whether you feel protected or threatened by them or not: E: Secret societies (Poro/Bondo, Ojeh)? Count Valid % % Is a big threat for my personal security 72 10.6 10.3 Is somewhat a threat for my personal security 66 9.7 9.4 Does not affect my personal security at all 357 52.3 50.9 Is somewhat important for my personal security 78 11.4 11.1 Is very important for my personal security 109 16.0 15.5 Valid N 682 100.0 97.2 No answer 4 0.0 0.6 Don’t know 16 0.0 2.3 Total 702 100.0 100.0 Table A70: Q17_F: Could you tell me whether you feel protected or threatened by them or not: F: Private security companies? Is a big threat for my personal security Is somewhat a threat for my personal security Count Valid % % 5 0.8 0.7 6 0.9 0.9 Does not affect my personal security at all 271 40.8 38.6 Is somewhat important for my personal security 244 36.7 34.8 Is very important for my personal security 139 20.9 19.8 Valid N 665 100.0 94.7 No answer 6 0.0 0.9 Don’t know 31 0.0 4.4 Total 702 100.0 100.0 214 Judy Smith-Höhn Table A71: Q17_G: Could you tell me whether you feel protected or threatened by them or not: G: National Union of Sierra Leonean Students? Count Valid % % Is a big threat for my personal security 4 0.6 0.6 Is somewhat a threat for my personal security 28 4.5 4.0 Does not affect my personal security at all 253 40.9 36.0 Is somewhat important for my personal security 133 21.5 18.9 Is very important for my personal security 201 32.5 28.6 Valid N 619 100.0 88.2 No answer 12 0.0 1.7 Don’t know 71 0.0 10.1 Total 702 100.0 100.0 Table A72: Q17_H: Could you tell me whether you feel protected or threatened by them or not: H: Youth wings of political parties? Is a big threat for my personal security Count Valid % % 60 9.0 8.5 Is somewhat a threat for my personal security 153 23.1 21.8 Does not affect my personal security at all 235 35.4 33.5 Is somewhat important for my personal security 118 17.8 16.8 Is very important for my personal security 97 14.6 13.8 Valid N 663 100.0 94.4 No answer 5 0.0 0.7 Don’t know 34 0.0 4.8 Total 702 100.0 100.0 Annexes 215 Table A73: Q17_I: Could you tell me whether you feel protected or threatened by them or not: I: Ghetto Boys? Count Valid % % Is a big threat for my personal security 370 54.1 52.7 Is somewhat a threat for my personal security 169 24.7 24.1 Does not affect my personal security at all 114 16.7 16.2 Is somewhat important for my personal security 19 2.8 2.7 Is very important for my personal security 12 1.8 1.7 Valid N 684 100.0 97.4 No answer 5 0.0 0.7 Don’t know 13 0.0 1.9 Total 702 100.0 100.0 Table A74: Q17_J: Could you tell me whether you feel protected or threatened by them or not: J: Bike riders? Count Valid % % Is a big threat for my personal security 73 10.9 10.4 Is somewhat a threat for my personal security 114 17.0 16.2 Does not affect my personal security at all 211 31.4 30.1 Is somewhat important for my personal security 152 22.6 21.7 Is very important for my personal security 122 18.2 17.4 Valid N 672 100.0 95.7 No answer 13 0.0 1.9 Don’t know 17 0.0 2.4 Total 702 100.0 100.0 216 Judy Smith-Höhn Table A75: Q17_K: Could you tell me whether you feel protected or threatened by them or not: K: West Side Boys? Count Valid % % Is a big threat for my personal security 392 59.1 55.8 Is somewhat a threat for my personal security 132 19.9 18.8 Does not affect my personal security at all 119 17.9 17.0 Is somewhat important for my personal security 7 1.1 1.0 Is very important for my personal security 13 2.0 1.9 Valid N 663 100.0 94.4 No answer 8 0.0 1.1 Don’t know 31 0.0 4.4 Total 702 100.0 100.0 Table A76: Q17_L: Could you tell me whether you feel protected or threatened by them or not: L: Civil defence forces (CDF)? Count Valid % % Is a big threat for my personal security 243 36.4 34.6 Is somewhat a threat for my personal security 137 20.5 19.5 Does not affect my personal security at all 144 21.6 20.5 Is somewhat important for my personal security 90 13.5 12.8 Is very important for my personal security 54 8.1 7.7 Valid N 668 100.0 95.2 No answer 8 0.0 1.1 Don’t know 26 0.0 3.7 Total 702 100.0 100.0 Annexes 217 Table A77: Q17_M: Could you tell me whether you feel protected or threatened by them or not: M: Revolutionary United Front (RUF)? Count Valid % % Is a big threat for my personal security 458 67.7 65.2 Is somewhat a threat for my personal security 90 13.3 12.8 Does not affect my personal security at all 113 16.7 16.1 Is somewhat important for my personal security 9 1.3 1.3 Is very important for my personal security 7 1.0 1.0 Valid N 677 100.0 96.4 No answer 10 0.0 1.4 Don’t know 15 0.0 2.1 Total 702 100.0 100.0 Table A78: Q17_N: Could you tell me whether you feel protected or threatened by them or not: N: Other? Is a big threat for my personal security Count Valid % % 36 75.0 5.1 Is somewhat a threat for my personal security 1 2.1 0.1 Does not affect my personal security at all 0 0.0 0.0 Is somewhat important for my personal security 2 4.2 0.3 Is very important for my personal security 9 18.8 1.3 Valid N 48 100.0 6.8 No answer 654 0.0 93.2 Don’t know 0 0.0 0.0 702 100.0 100.0 Total 218 Judy Smith-Höhn Table A79: Q17_n_ot: Could you tell me whether you feel protected or threatened by them or not: N: Other (strings)? Area groups Count Valid % % 656 93.4 93.4 1 0.1 0.1 Armed robbers 7 1.0 1.0 House burglars 1 0.1 0.1 Council chiefs 1 0.1 0.1 Councillors 1 0.1 0.1 Drivers’ union 1 0.1 0.1 Drug addicts 1 0.1 0.1 Ex-combatants 5 0.7 0.7 Foreign troops 2 0.3 0.3 Gang robbers 1 0.1 0.1 Justice system 1 0.1 0.1 Kamajors 2 0.3 0.3 Kidnappers 1 0.1 0.1 Police SSD 1 0.1 0.1 Political parties’ supporters 1 0.1 0.1 Rebels 1 0.1 0.1 Retired soldiers 2 0.3 0.3 Robbers 4 0.6 0.6 Thieves 5 0.7 0.7 Traditional rulers 1 0.1 0.1 UN mission 2 0.3 0.3 UNAMSIL 1 0.1 0.1 Youths 1 0.1 0.1 Youth groups 1 0.1 0.1 702 100.0 100.0 Total Annexes 219 Table A80: Q17_rA: Could you tell me whether you feel protected or threatened by them or not: A: Sierra Leone Police? Count Valid % % Is very/somewhat important for my personal security 616 88.5 88.5 Does not affect my personal security at all 24 3.4 3.4 Is a big/somewhat a threat for my personal security 56 8.0 8.0 Valid N 696 100.0 100.0 No answer 0 0.0 0.0 Don’t know 0 0.0 0.0 696 100.0 100.0 Total Table A81: Q17_rB: Could you tell me whether you feel protected or threatened by them or not: B: Republic of Sierra Leone Armed Forces? Count Valid % % Is very/somewhat important for my personal security 590 84.9 84.9 Does not affect my personal security at all 35 5.0 5.0 Is a big/somewhat a threat for my personal security 70 10.1 10.1 Valid N 695 100.0 100.0 No answer 0 0.0 0.0 Don’t know 0 0.0 0.0 695 100.0 100.0 Total Table A82: Q17_rC: Could you tell me whether you feel protected or threatened by them or not: C: Mines monitoring officers? Count Valid % % Is very/somewhat important for my personal security 159 27.4 27.4 Does not affect my personal security at all 386 66.4 66.4 Is a big/somewhat a threat for my personal security 36 6.2 6.2 Valid N 581 100.0 100.0 No answer 0 0.0 0.0 Don’t know 0 0.0 0.0 581 100.0 100.0 Total 220 Judy Smith-Höhn Table A83: Q17_rD: Could you tell me whether you feel protected or threatened by them or not: D: Traffic wardens? Count Valid % % Is very/somewhat important for my personal security 363 55.2 55.2 Does not affect my personal security at all 239 36.3 36.3 Is a big/somewhat a threat for my personal security 56 8.5 8.5 Valid N 658 100.0 100.0 0 0.0 0.0 No answer Don’t know Total 0 0.0 0.0 658 100.0 100.0 Table A84: Q17_rE: Could you tell me whether you feel protected or threatened by them or not: E: Secret societies (Poro/Bondo, Ojeh)? Count Valid % % Is very/somewhat important for my personal security 187 27.4 27.4 Does not affect my personal security at all 357 52.3 52.3 Is a big/somewhat a threat for my personal security 138 20.2 20.2 Valid N 682 100.0 100.0 No answer 0 0.0 0.0 Don’t know 0 0.0 0.0 682 100.0 100.0 Total Table A85: Q17_rF: Could you tell me whether you feel protected or threatened by them or not: F: Private security companies? Count Valid % % Is very/somewhat important for my personal security 383 57.6 57.6 Does not affect my personal security at all 271 40.8 40.8 Is a big/somewhat a threat for my personal security 11 1.7 1.7 Valid N 665 100.0 100.0 No answer 0 0.0 0.0 Don't know 0 0.0 0.0 665 100.0 100.0 Total Annexes 221 Table A86: Q17_rG: Could you tell me whether you feel protected or threatened by them or not: G: National Union of Sierra Leonean Students? Count Valid % % Is very/somewhat important for my personal security 334 54.0 54.0 Does not affect my personal security at all 253 40.9 40.9 Is a big/somewhat a threat for my personal security 32 5.2 5.2 Valid N 619 100.0 100.0 No answer 0 0.0 0.0 Don’t know 0 0.0 0.0 619 100.0 100.0 Total Table A87: Q17_rH: Could you tell me whether you feel protected or threatened by them or not: H: Youth wings of political parties? Count Valid % % Is very/somewhat important for my personal security 215 32.4 32.4 Does not affect my personal security at all 235 35.4 35.4 Is a big/somewhat a threat for my personal security 213 32.1 32.1 Valid N 663 100.0 100.0 No answer 0 0.0 0.0 Don’t know 0 0.0 0.0 663 100.0 100.0 Total 222 Judy Smith-Höhn Table A88: Q17_rI: Could you tell me whether you feel protected or threatened by them or not: I: Ghetto Boys? Is very/somewhat important for my personal security Count Valid % % 31 4.5 4.5 Does not affect my personal security at all 114 16.7 16.7 Is a big/somewhat a threat for my personal security 539 78.8 78.8 Valid N 684 100.0 100.0 0 0.0 0.0 No answer Don’t know Total 0 0.0 0.0 684 100.0 100.0 Table A89: Q17_rJ: Could you tell me whether you feel protected or threatened by them or not: J: Bike riders? Count Valid % % Is very/somewhat important for my personal security 274 40.8 40.8 Does not affect my personal security at all 211 31.4 31.4 Is a big/somewhat a threat for my personal security 187 27.8 27.8 Valid N 672 100.0 100.0 No answer 0 0.0 0.0 Don’t know 0 0.0 0.0 672 100.0 100.0 Total Table A90: Q17_rK: Could you tell me whether you feel protected or threatened by them or not: K: West Side Boys? Count Valid % % Is very/somewhat important for my personal security 20 3.0 3.0 Does not affect my personal security at all 119 17.9 17.9 Is a big/somewhat a threat for my personal security 524 79.0 79.0 Valid N 663 100.0 100.0 No answer 0 0.0 0.0 Don’t know 0 0.0 0.0 663 100.0 100.0 Total Annexes 223 Table A91: Q17_rL: Could you tell me whether you feel protected or threatened by them or not: L: Civil defence forces (CDF)? Count Valid % % Is very/somewhat important for my personal security 144 21.6 21.6 Does not affect my personal security at all 144 21.6 21.6 Is a big/somewhat a threat for my personal security 380 56.9 56.9 Valid N 668 100.0 100.0 No answer 0 0.0 0.0 Don’t know 0 0.0 0.0 668 100.0 100.0 Total Table A92: Q17_rM: Could you tell me whether you feel protected or threatened by them or not: M: Revolutionary United Front (RUF)? Count Valid % % Is very/somewhat important for my personal security 16 2.4 2.4 Does not affect my personal security at all 113 16.7 16.7 Is a big/somewhat a threat for my personal security 548 80.9 80.9 Valid N 677 100.0 100.0 No answer 0 0.0 0.0 Don’t know 0 0.0 0.0 677 100.0 100.0 Total Table A93: Q17_rN: Could you tell me whether you feel protected or threatened by them or not: N: Other? Count Valid % % Is very/somewhat important for my personal security 11 22.9 22.9 Does not affect my personal security at all 0 0.0 0.0 Is a big/somewhat a threat for my personal security 37 77.1 77.1 Valid N 48 100.0 100.0 No answer 0 0.0 0.0 Don’t know 0 0.0 0.0 Total 48 100.0 100.0 224 Judy Smith-Höhn Annex II: Guidelines for Focus Group Discussion (excerpt) Questions for opening the discussion What is security all about? Which groups provide security (or consent to violence and crime) in this area? Mapping of various actors that are involved – general. Topics for discussion 1. 2. 3. 4. 5. 6. 7. 8. 9. Understanding of security (protection from harm, use of force). Identification of security providers (names, state or non-state, authorised by the state or not, external). Cooperation and rivalries between these groups, if at all. Expectations from security providers. (Who provides what for whom? Does the situation determine the security provider chosen?) Citizens’ needs met by such groups. (All the time? Some of the time? Never?) Are their actions viewed as justified and righteous (i.e. legitimate) by the discussants? (Would you turn to these people again? Motives for turning to these providers; payment for security?) Need for such groups, i.e. their importance. (Which are indispensible and which should rather be disbanded? Are they a real alternative to state-provided security?) Mapping of various actors – specific: threat; neither/nor; no threat, important or not, relationship among each other, comparison to the past (five years ago). Discussants’ closing remarks (additional points that weren’t raised in the discussion). How to moderate a discussion successfully When moderating a focus group discussion, please comply with the following instructions. 1) 2) Start each discussion with the introduction. Do not cut it short and refrain from interjecting personal comments. Introduce the topic in an honest and fairly general fashion. Annexes 3) 4) 5) 6) 7) 8) 9) 10) 11) 12) 225 State the ground rules for the discussion: mobiles off, only one person speaks at a time, no conversations with your neighbour, etc. Emphasise that you want as many different points of view as possible. Emphasise that there are no right or wrong answers. Begin the discussion by letting each discussant introduce him/herself. Dealing with latecomers: do not include anyone who arrives after the opening statements have ended and the discussion has begun. Often, discussants tend to make general remarks; to avoid them drifting off into generalities, stress that their personal experience is what is of interest. Interview content: maintain the focus and do not explore too many topics. The guideline for the focus group discussions is there to assist the moderator in ensuring that the discussions follow in more or less the same order from group to group. These should not be followed rigorously, i.e. do not adhere to a fixed set of questions. It is therefore best to memorise these topics of discussion; if the moderator constantly refers to a printed list of topics, the discussants will hang back in their discussions, waiting to hear what the next topic for discussion is. The role of the moderator is to guide someone else’s discussion. Do not dominate the discussion. A moderator who appears to be an expert on the topic will shut off many lines of discussion. At the end of the discussion, allow each person to give a final summary statement. Asking each person to give a final statement is not only a useful technique that signifies the end of the discussion, but it also allows participants to make a contribution that they have been holding back from the open discussion. Finally, fill out the questionnaire for each of the participants. Stress that the data will remain anonymous. About the Author Judy Smith-Höhn is a senior researcher at the Institute for Security Studies in Pretoria, South Africa. Prior to her present position she was a research fellow at the GIGA Institute of African Affairs in Hamburg, Germany and later a senior researcher at the Centre for Conflict Resolution in Cape Town in South Africa. Her thematic emphases lie in the areas of violent conflict and its prevention, and security sector reform within a regional focus on Liberia, Sierra Leone, Zambia and Zimbabwe. She has published nationally and internationally on topics ranging from security sector reform and postconflict peacebuilding to democratic transformation in South Africa. She holds a PhD from the University of Leipzig, Germany, and a Diplom (masters degree) in Political Science from the University of Hamburg, Germany. About the Geneva Centre for the Democratic Control of Armed Forces (DCAF) The Geneva Centre for the Democratic Control of Armed Forces (DCAF) is an international foundation whose mission is to assist the international community in pursuing good governance and reform of the security sector. To this end, the Centre develops and promotes appropriate norms at the international and national levels, determines good practices and relevant policy recommendations for effective governance of the security sector, and provides in-country advisory support and practical assistance programmes to all interested actors. Detailed information is available at www.dcaf.ch Geneva Centre for the Democratic Control of Armed Forces (DCAF) Rue de Chantepoulet 11, PO Box 1360, CH-1211 Geneva 1, Switzerland Tel: + 41 22 741 77 00; fax: + 41 22 741 77 05; e-mail: info@dcaf.ch