Modern post-conflict security sector reform in Africa: patterns of success and failure

Transcription

1 African Security Review ISSN: (Print) (Online) Journal homepage: Modern post-conflict security sector reform in Africa: patterns of success and failure Sarah Detzner To cite this article: Sarah Detzner (2017) Modern post-conflict security sector reform in Africa: patterns of success and failure, African Security Review, 26:2, , DOI: / To link to this article: The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group Published online: 18 Apr Submit your article to this journal Article views: 3814 View Crossmark data Citing articles: 1 View citing articles Full Terms & Conditions of access and use can be found at

2 AFRICAN SECURITY REVIEW, 2017 VOL. 26, NO. 2, FEATURE Modern post-conflict security sector reform in Africa: patterns of success and failure Sarah Detzner ABSTRACT This piece examines recurring patterns in the where and how of security sector reform (SSR) implementation failures in post-conflict African SSR programmes featuring substantial external involvement that have been undertaken since the mid-1990s. It finds, briefly that in these efforts, gaps in SSR implementation have tended to occur at the same points in the SSR process repeatedly. The most common issues include failures to correctly assess the post-conflict security environment, failures to ensure local ownership of reform efforts, failures to devote sufficient resources and attention to disarmament, demobilisation and reintegration (DDR, a process with close ties to SSR and force integration), failures by donors to coordinate goals and resources, and failures to include critical parts of both the de jure and de facto security sector in reforms. Post-conflict African success stories also share common characteristics, most notably a deep and wide level of societal involvement at most stages of the SSR process. These patterns, analysed as a necessary first step to discovering the why of implementation failures in such cases, collectively suggest a focus on the early mobilisation of domestic political demand for SSR, rather than on varying methods of applying external pressure, as a fruitful avenue for future research. KEYWORDS security sector reform; peacebuilding; African Union; disarmament; demobilisation; reintegration; DDR; SSR Introduction Promoting security sector reform (SSR) in countries emerging from war is one of the critical missions that the African Union (AU) following the path laid out by the United Nations (UN), the World Bank, and others has increasingly assumed in recent years. However, despite two decades of implementation experience, as of 2016 there has been no increase in the tiny number of post-conflict SSR efforts generally considered successful. In another field of endeavour, the approach might have been discarded as unworkable in practice. However, in the absence of any alternative path to the same critical ends, i.e. stable, self-governing states in which citizens enjoy basic security and justice services, do not export security problems (refugees, militants, drug-traffickers, etc.), and do not require continual aid and periodic intervention, SSR remains indispensable. It has become a cliché to state that failure to implement is the problem with SSR. The core recommendations for SSR practice have remained fairly constant over time, as has the failure CONTACT Sarah Detzner Sarah.Detzner@tufts.edu 2017 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License ( creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited, and is not altered, transformed, or built upon in any way.

3 AFRICAN SECURITY REVIEW 117 to shape programmes accordingly. While the obvious question regarding failures of SSR implementation is why, it should be equally clear given the vast scope of activity fitted under the heading of SSR that to answer this one must first address the prior questions of where and how, particularly if the answers systematically vary based on region, level of intervention, intensity of conflict, and a host of other variables. This article attempts to examine the where and how of SSR s implementation failures for both the thematic issues and the subset of cases most relevant to the AU instances of postconflict SSR with significant external involvement undertaken since the mid-1990s: Liberia, Sierra Leone, Guinea-Bissau, Somalia, Sudan, South Sudan, Mali, Côte D Ivoire, Burundi, Chad, Libya, the Central African Republic (CAR), and the Democratic Republic of Congo (DRC). To this end, the piece is organised into five sections. The first provides background by reviewing both prevailing definitions of SSR and the conditions that led to its popularisation. The second section reviews and aggregates the experiences of various African states that have undertaken post-conflict SSR efforts (successfully and otherwise). Many thorough case studies of the efforts examined exist; by scrutinising these cases for patterns, this article attempts to draw more generalisable conclusions for this case universe. The third section explores scholars and practitioners evolving and diverging attempts to define and so resolve recurrent issues within SSR, focusing on those most relevant to African post-conflict efforts. The fourth section distils the findings of this examination into a set of recommendations for the AU, and the fifth concludes with thoughts on how these findings suggest further specific directions for research. This examination finds, briefly, that in African post-conflict efforts, gaps in SSR implementation have tended to occur at the same points in the SSR process repeatedly. The most common issues include failures to correctly assess the post-conflict security environment, failures to ensure local ownership of reform efforts, failures to devote sufficient resources and attention to disarmament, demobilisation and reintegration (DDR, a process with close ties to SSR explored in the issues section of this article, and force integration), failures by donors to coordinate goals and resources, and failures to include critical parts of both the de jure and de facto security sector in reforms. Post-conflict African success stories also share common characteristics, most notably a deep and wide level of societal involvement at most stages of the SSR process. Possible directions for further research gleaned from these patterns are explored in the conclusion of this article. Defining SSR Any exploration of SSR successes and failures requires first drawing some boundaries around the concept. Unfortunately, there has never been a consensus definition of SSR, though definitions have tended to expand over time. What consensus exists is largely captured in the definition put forward by the Development Assistance Committee of the Organisation for Economic Co-operation and Development (OECD-DAC) of the security system to be reformed (named to convey the interconnection of institutions) as: [ ] core security actors (e.g. armed forces, police, gendarmerie, border guards, customs and immigration, and intelligence and security services); security management and oversight bodies (e.g. ministries of defense and internal affairs, financial management bodies and public complaints commissions); justice and law enforcement institutions (e.g. the judiciary, prisons, prosecution services, traditional justice systems); and non-statutory security forces (e.g. private security companies, guerrilla armies and private militia). 1

4 118 S. DETZNER Of the rival definitions, the most relevant to the cases examined in this article is offered by Ball et al., who argue that particularly in Africa SSR is not a sufficiently ambitious or comprehensive approach, proposing the following alternative: Security-sector transformation: A holistic change to the security sector, aimed at altering the relations of power within the sector in the direction of civil/constitutional control to transform institutional culture, promote professionalism, improve resource utilization and operational effectiveness (on the side of the security forces, better policy management (on the side 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 [ ]. 2 Ball et al. hold that the term reform has been used too widely in reference to superficial changes meant to legitimise unpopular regimes, while transformation is an accurate description of the task. 3 In practice, however, almost all efforts have fallen short of this expansive definition. The following section explores the particular conditions that gave rise to SSR and subsequent disagreements as to its correct interpretation and implementation. Initial rationale for SSR SSR emerged as a possible solution to pressing problems in both security and development in the 1990s, which witnessed a critical shift in thinking about the military s role in developing countries, sparked by the end of the Cold War. As existential dangers for the great powers receded, experts began to argue that the bilateral, train-and-equip focused assistance that the Union of Soviet Socialist Republics (USSR) and the United States (US) provided to allies/clients could be destabilising. 4 Militaries receiving aid had become dependent, growing larger (and more ambitious) than their economies could sustain and tending to have outsized control over political processes. 5 In Africa, the last colonial/racist regimes and a number of autocracies were faltering and government budgets were shrinking, forcing militaries to adjust to the needs and values of democratic governments as well as finding a way to reintegrate ex-combatants. Concurrently, assessments in developing countries highlighted the lack of security and justice as major concerns for the impoverished. 6 This fed into the evolving acceptance of human security as a guide for operations. Claire Short, who was Secretary of State for International Development in the United Kingdom (UK) at the time, said in her 1999 speech that coined the term security sector reform, that state, regional and even global stability is not simply a question of relations between states. Rather, it is a function of the welfare of the people who constitute these nations. 7 This was compelling, as donors strategised to prevent or resolve the messy New Wars of the 1990s, where increasing danger to civilians was linked to state weakness and failures of critical institutions. 8 It was hoped that change could be encouraged through the main innovation of the SSR model [ and] its focus on governance. The professionalism and effectiveness of the security sector is not just measured by the capacity of the security forces, but how well they are managed, monitored and held accountable. 9 Institutionalisation, disappointment, and new tactics SSR was rapidly accepted as a critical goal of post-conflict state building. However, after more than a decade with few success stories, scholars and practitioners frustration has grown, particularly since, as previously mentioned, basic orthodoxy regarding a sound SSR effort is largely

5 AFRICAN SECURITY REVIEW 119 unchanged, with a fair amount of agreement that these various approaches should address two main problems: the ineffective and inefficient provision of security and, increasingly, justice, in part because the providers may themselves be a source of insecurity; and the inadequacy of accountability and oversight in the security sector. 10 Execution is widely held to be the problem. As Zyck notes, scholars attribute lack of SSR success to donors and recipients failure to live up to principled security sector engagement by, for instance, imposing short timelines, privileging technical capacity-building over statebuilding approaches, bypassing security sector governance and parliaments, marginalizing civil society and outsourcing implementation to private security companies (PSCs). 11 To address this gap, some argue that SSR must remain ambitious when the challenge is to fundamentally alter the relationship between the government and people across interdependent institutions, an incremental approach cannot work. 12 If international actors were serious about devoting political and diplomatic capital, time and resources, they argue, success would follow. On the other side, many emphasise that, especially in Africa, most security and justice for ordinary people is not provided through the state. They have come to believe that SSR s preoccupation with the state and Western liberal principles is unrealistic and counterproductive, and that reformers must seek to shape programs around local dynamics and perceptions of security. 13 Aiming for focused international will and resources that will rarely arrive is futile. This thinking prescribes a more incremental, evolutionary approach to post-conflict SSR and greater non-state actor engagement. Beyond this, Sedra observes three basic approaches. 14 First is the throwback train-andequip, focused on increasing capability but not accountability or governance. Sedra notes that This approach most often comes to the fore in the most difficult reform cases, where insecurity and political instability is acute. Under such pressure donors tend to do two things: instrumentalize SSR to address immediate instability and insecurity, and revert to what is most simple and familiar, training and equipping the security forces. 15 This approach also gives donors cover to pursue more traditional bilateral security priorities; many have attributed its resilience to the requirements of the War on Terror. Ball and Hendrickson describe how, particularly in Africa, many governments supported US counterterrorism efforts in exchange for direct military assistance delivered without pressure to reform. 16 De Goor and Van Veen identify this trend as a major reason why development actors have become less invested in SSR. 17 The second variant orthodox SSR is a recommended order of operations developed in the initial years of practice a security needs assessment, government buy-in, and a holistic and multi-year strategy with clear short-, medium- and long-term goals. The problem was inflexibility and expense; in volatile post-conflict situations, it was easy for newly-arrived interveners to misread initial conditions. A coup, faltering government support, one component not going as well or quickly as expected numerous developments could derail careful sequencing. Additionally, as these programmes tend to target many institutions simultaneously, they require immense resources, intensive multi-actor coordination, and medium- to long-term funding they rarely received. 18 When these programs falter, as in Afghanistan or Haiti, they were sometimes abandoned for the next crisis and sometimes reverted to train-and-equip efforts. Finally, an emerging school of thought holds that, in transitional environments, a better approach is to gradually build reform momentum and buy time for the development of a political consensus and the stabilization of the security environment through ad hoc projects in areas and institutions ripe for reform. 19 This general idea of stabilization, 20 interim

6 120 S. DETZNER stabilization, 21 or similar, acknowledges that in the most volatile situations governments may be completely incapable of meeting basic security needs, thus leaving civilians at greater risk from violent crime, inter-communal reprisal and general opportunistic predation than from the official conflict. 22 Frequently, neither national nor local governments are in a position to locally own any process: Indeed, it is often the case in fragile and failed states that an intervening military force initially operates under the international law governing the occupying powers. 23 These forces often a peacekeeping mission can provide what Downes and Muggah refer to as breathing room, ensuring basic security and suppressing spoilers while encouraging locals (and donors) into dialogue about which SSR activities are most immediately necessary and possible. 24 Some specific stabilisation measures include the establishment of civilian service corps; military or security sector integration arrangements; the creation of transitional security forces; dialogue and sensitization programmes; and differentiated forms of transitional autonomy. 25 The major downside is that this commits an intervening force to filling the security gap indefinitely, as it cannot responsibly withdraw until replaced by local actors. African post-conflict SSR: the record Of course, these divergent approaches developed in response to successes and failures on the ground. This section focuses on the patterns that have recurred most prominently in African post-conflict SSR efforts, first in failed and/or stalled attempts, and later in the smaller universe of successes. Given the current lack of an adequate standard for measuring SSR success or failure (an issue explored later in greater depth), these sections only reference fairly clearcut and generally accepted instances. Failures of security assessment and national ownership It is widely agreed that it is necessary but not sufficient for a proposed SSR programme to have national government support. However, often, the consultations with other parts of society, or even the government itself, that are needed in order to produce a plan that satisfies true security needs do not take place, or are undertaken in a cursory fashion. Notable cases of near-complete failure to assess and consult include:. Guinea-Bissau, where in terms of civil society inclusion the overall impression remains that instead of guaranteeing self-determination, only a minimum degree of local co-determination in the form of selective consultations was granted. 26. Liberia, where in 2006, years after SSR efforts began, the US government (charged with rebuilding the army) commissioned a comprehensive security review from the Rand Corporation in which the Liberian government was so minimally involved that some security agencies were not even aware that it had taken place. 27 Significantly, security was also unaddressed in the otherwise broad national consultations that fed into Liberia s postwar development strategy. 28 In some efforts, these assessments and consultations took place but without follow through, and seemingly to check a box for a donor. In Guinea-Bissau, for example, the process seemed largely driven by international actors with local political power actors using SSR to garner international goodwill. 29 The 2005 Malian National Conference on

7 AFRICAN SECURITY REVIEW 121 Peace and Security and the 2008 CAR Political Inclusion Dialogue were praised when conducted, but produced so little action as to embitter and discourage many of those who participated. 30 The CAR process was relaunched a few years later under a new government, illustrating another common challenge; assessments and consultations often take place in situations (such as Somalia, South Sudan and the CAR) where renewed conflict is imminent. 31 Even if the assessed or assessing government survives, the situation on the ground is likely to have changed enough to justify beginning the assessment process from scratch. Failures of DDR and force integration DDR failures in these cases are numerous and interconnected. In some cases (Liberia, Sierra Leone, Mali, Libya and the DRC), inadequate funding/underestimation of the task left some combatants out of the process, while planned demobilisation periods and reintegration training were shortened beyond usefulness or cut entirely. 32 In others, such as the CAR, assistance was so delayed that cantoned combatants gave up and resumed low-level attacks against rivals. 33 Worryingly, in many states DDR (largely externally funded) seems to have become an expected part of combat cycle. The CAR, the DRC and Guinea-Bissau have all hosted multiple DDR efforts in quick succession, to little or even negative effect. In other states, notably Mali, the promise of an eventual DDR package has been a recruiting tool. 34 Recent Sudanese and South Sudanese efforts were so heavily corrupted as to be largely patronage programmes for various commanders. 35 When government/antagonist forces are combined to form a single military, in many instances Chad, Sudan, Cote D Ivoire, DRC, South Sudan they are combined only on paper and serve under separate, non-integrated chains of command. 36 Payroll funds (to real or fictitious soldiers) function as payment to commanders to buy their temporary loyalty. De Waal diagnoses this process in much greater detail in his writings on the political marketplace. 37 As de Waal (and Debos) also note, what many of the most prolonged and wasteful cases share is the conviction among government elites and the population that, while DDR and force integration may take place, the conflict is not over, but merely paused. 38 This belief is reflected in the lack of care and attention given to long-term considerations in many peace agreements, as such considerations are clearly irrelevant where the average agreement life span is measured in months. Further, when donor timescales are just a few years, there is little incentive among combatants to keep the peace after aid dries up. A few counterexamples help underscore the point Liberia, which ended its war with a widely-accepted political settlement and a long-term security guarantor in the form of the United Nations Mission in Liberia (UNMIL), remains at peace, despite its flawed DDR process and ongoing issues with non-reintegrated combatants. 39 Despite recent instability, Burundi s painstakingly and lengthily-negotiated DDR and force integration arrangements in the peace agreements that ended its civil war kept the security sector functioning, and the military largely out of politics, for many years. 40 Given a true buy-in to a peace by warring parties and the population, DDR and force integration processes can be quite flawed and still not fatally destabilising absent such a buy-in, all technical solutions seem futile. Failures of coordination Donors largely agree that an imperfect but coordinated DDR/SSR strategy is superior to even the best-designed sub-sectorial programmes operating in isolation, given the need for holistic

8 122 S. DETZNER SSR. However, in practice, African efforts have resulted in several prominent coordination failures. In the DRC and Guinea-Bissau, multiple donors have been played off against one another, enabling host governments to postpone reform progress significantly. 41 In Liberia, the major donor (the US) took responsibility for the military, but left the police and other institutions to the UN and other donors, with predictably lopsided results. 42 In Somalia, failures to coordinate within the African Union Mission in Somalia (AMISOM) have led each troop-contributing country to train and equip its own Somali security force according to its own needs and norms, raising serious concerns about the ability of the Somali National Armed Forces to eventually act jointly. 43 Key parts of the security sector left out of reform Human security concerns are theoretically the heart of SSR, due to the destabilising effects of daily insecurity. However, the organs of the security sector most involved in this area the police, the prisons and the judiciary have been persistently neglected or omitted from African post-conflict efforts. As previously noted in Liberia (and similarly in Cote D Ivoire), the police received far less attention than the military, but in an oversight with more immediate consequences, the prisons and the judiciary were initially left out of reform efforts entirely, squandering an opportunity to rebuild the population s trust in the government s ability to prosecute criminals. 44 In Chad, the United Nations Mission in Central African Republic and Chad (MINURCAT) trained a small number of police to effectively police refugee camps, only to face a lack of prisons or courts with which to deal with arrestees. 45 Mali s current troubles have been blamed in part on popular frustration over a corrupt judiciary, and were foreshadowed by the police war of , sparked in part by issues of corruption and nepotism. 46 In Chad, Guinea-Bissau, the DRC and the CAR, the police have been used as a source of jobs for demobilised soldiers with minimal if any attempt at vetting or retraining, worsening already weak institutions. 47 Another less-frequently discussed issue is that of security institutions (usually intelligence services and elite guard units), which are left out of reform efforts by governments for strategic reasons, often as a hedge against coups. Given the extremely sensitive nature of these institutions, external actors have often been unwilling or unable to include them in the reform process. However, as is recognised in frequent calls for holistic SSR when they are excluded, they often undermine the reform process by serving as a home for the types of corrupt and repressive activities that SSR efforts are attempting to eliminate within the regular forces. 48 In Chad, for instance, the presidential guard has been a stronger force than the army, a dynamic echoed in Libya. 49 The presidential guard and intelligence services were similarly left out of reform efforts in the DRC and the CAR. 50 In Sudan, while the comprehensive peace agreement called for the setting up of a non-operational security and intelligence agency[,] this provision was [ ] not implemented [ ]. The National Intelligence and Security Service (NISS) of Sudan, headed since 2009 by Muhammad al-atta Fadl al-mula, remains powerful and feared, and has even gained command of its own paramilitary. 51 Neglected peripheries and non-state actors Few states have productively engaged the non-state actors who provide security services in SSR. This is a particular concern because conflicts originating in neglected peripheries are common, highlighted as a major factor in Liberia, Guinea, Chad, Mali, South Sudan, Sudan,

9 AFRICAN SECURITY REVIEW 123 Libya, the DRC and the CAR. A number of these states are so large and unevenly settled that they are described as unpoliceable by conventional means without an unrealistic expansion of policing budgets; thus, engaging non-state actors may well be the only feasible option. 52 However, it is inaccurate to portray the governments of these states as hapless victims of poverty and geography. In many cases, governments do have a role for non-state security actors in peripheries namely, as militias that are activated when needed (through cash and arms transfers) to forward the government s agenda and suppress rebellions. This dynamic has been observed in the DRC, the CAR, Sudan, Chad and Côte D Ivoire, among others. 53 Debos observes: There is not a zero-sum game between the official and unofficial state: the strengthening of one does not necessarily imply the weakening of the other. 54 Where this pattern appears, governments often resist the inclusion of these non-state actors in SSR efforts, because to do so would be to forfeit a valuable, semi-deniable tool. Success stories? The importance of political demand for SSR Success is a term that must be applied cautiously regarding SSR efforts here, it is used not to imply achievement of a final goal, but rather significant and sustained improvement in day-to-day security conditions for the majority of the people within a state when compared to prior conditions. Thus, cases that are clear SSR successes in terms of improvement may still host high levels of crime, political repression, or other serious threats to human security. Using this standard, three of the clearest examples of success among post-conflict African states are Sierra Leone, South Africa and Ethiopia. Notably, the second and third cases took place before SSR was popularised, and South Africa helped to inspire the approach by example and through the advocacy of its veterans for its process. Examining all the factors that contributed to the relative success in these cases is beyond the scope of this article. However, it is useful to reflect on the most notable factors that the three share: political consensus between the government and the general population that significant SSR must take place, and a process for ensuring that reform priorities were broadly shared. In all three cases, the government s approach to SSR was transformational, not incremental. The Ethiopian and South African governments had been brought to power by popular movements, and thus had a strong mandate for and interest in radically changing then existing institutions. In Sierra Leone, the post-civil-war government was reinstalled after a military coup (following a previous coup d état during the war) and then democratically re-elected, with incentive and mandate to prevent future upheavals. Much of post-apartheid South Africa s leadership was familiar with consultative processes and building consensus from their time in the resistance and established patterns which fuelled the South African defence review process, including hundreds of consultations with a wide range of NGOs [non-governmental organisations], business, academia, other government departments, rural and urban communities and specialist groupings to produce a defence review and security sector approach considered groundbreaking for its focus on internal human security over external threats. 55 Ethiopia s insurgents pursued a more informally consultative process that grew from the necessity of winning and maintaining popular support during the civil war, when insurgent forces depended on winning the loyalty of the population to gain ground. 56 This process culminated in a focus on economic development as the path to security and a commitment to cap defence spending at 2% of the state s budget, far lower than the regional norm. 57

10 124 S. DETZNER In Sierra Leone, Jackson asserts, as the SSR process evolved, the government s main source of information about security concerns outside of the capital was the newly-established Office of National Security (ONS), which in the absence of capacity for other intelligence gathering developed extensive links throughout the countryside with the goal of having locals report threats. The ONS used these links, once formed, to conduct a security review that essentially doubled as a consultation process, in which locals in various regions reported their most pressing security concerns. 58 Several observers link the input from this process to the government s unusually strong focus on the link between poverty reduction/economic development and increased security. 59 This input also fuelled focus on police over military capacity building. The fact that all three successes featured popular consultations resulting in a governmental decision to focus on human security (in the form of a focus on economic development and police capacity) over traditional security goals seems to validate early SSR theorists similar focus, and raise questions about the long-term usefulness of programming focused first on national militaries. Evolving recipes for SSR: issues and debates Having examined both successes and failures of implementation, it is useful to examine how scholars and practitioners have attempted to define and resolve emergent issues over time. Early SSR scholars were optimistic that the details of central norms, etc., would eventually generate an operationalisable consensus. This optimism was misplaced. As Law laments, the lack of a common SSR language [ ] is the IGO [international governmental organisation] equivalent of the non-interoperability of communications systems of security forces. 60 This section explores the discussions and disagreements surrounding key issues. A tailored and holistic approach? Scholars and practitioners have always emphasised that SSR efforts should be tailored to the needs of each case, and that reform programmes created to address these should be as holistic as possible, as reforming one portion of a highly-interconnected set of institutions (for example, police but not prisons) is unsustainable at best. 61 The primary recommendation for mitigating the immense resource demands of a holistic approach is careful sequencing: conducting a needs assessment, carefully considering which goals are interdependent, and finally pursuing programming that lays the groundwork for future efforts while still being politically feasible for example, offering civil society leaders and parliamentarians training in security-sector oversight before legislation/policy changes give them actual oversight. 62 Over time, as these approaches spread but still failed to deliver success, many have argued that assessments needed to consult a much broader range of stakeholders to be accurate. These broader consultations, it was hoped, would both give outsiders critical information about the political and cultural context in which SSR would take place and reassure various critical stakeholders that their concerns were being heard. 63 England and Boucher s list of the expanded universe of actors to be consulted includes: national political stakeholders; local justice networks and informal security and justice providers; local citizens, residents, communities, and neighborhoods; and international actors with their own national interests. 64 Ideally, broader assessments would create a strategic framework guiding all programming. 65 However, some contend that many post-conflict situations are too unstable for SSR; moving too quickly, they argue, risks donors empowering illegitimate and unaccountable

11 AFRICAN SECURITY REVIEW 125 actors. 66 Furthermore, a premature focus on SSR may divert resources from protecting civilians or maintaining peace agreements. 67 Finally, if the principle of local ownership means requiring the host government to set its own security priorities then the host government must first have some capacity to do so and this capacity is often absent immediately post-conflict. 68 Therefore, the first question any assessment should ask is whether or not conditions allow for SSR and, if not, what activities could realistically be pursued to prepare for future efforts. 69 Lack of evaluation mechanisms Disagreement on how to recognise and measure success hindered learning from early SSR experiences. While SSR principles existed, it was not clear how progress toward these general principles could be concretely measured, especially in a way that provided comparability between cases. Expectations that evaluation mechanisms would eventually develop organically proved overly optimistic. 70 However, it eventually became clearer what specific types of indicators were most needed for post-conflict SSR, notably objective criteria to help determine at what point interveners should transfer authority over security institutions back to locals, indicators to measure change in the professionalisation of forces, and tools for assessing whether or not the population is receiving basic security services. Considering paths forward, in 2010 Schroeder proposed an approach prioritising local perceptions, building composite measurements of quality of security sector provision and security sector governance using existing indicator sets, including the Uppsala Conflict Data Program, the Global Peace Index, and the Civil Society Index. 71 Downes and Muggah further suggest that in order to measure the cumulative effects of different programmes, actors should perform joint evaluations. 72 Local ownership: who and how much? 73 From the outset, agreement has been widespread that demand for change must come from within countries. 74 SSR s governance aspect, where security services commit to civilian control and civilian institutions assume oversight, requires fundamental attitude and relationship changes unlikely to be created through donor coercion. Recognition of this has created uncertainty about how to proceed when the demand is not present from official actors or, at least, not present in a form that donors are willing to fulfil. Measuring local ownership is necessarily a combination of different factors given different weights, with the who being the most obvious. It is widely agreed that the host government needs to be active in reform. Most hold that civil society actors should also be involved, both to build support beyond the government reform and to provide vital information about everyday security. 75 While some have argued for direct donor engagement with civil society, others argue that governments need to control such consultations themselves. 76 Adding further complexity, neither government nor civil society are usually strong or unified in post-conflict contexts, as Adedeji Ebo highlights in his 2004 review of the first official SSR efforts in postconflict African states. Government elites may continue jockeying for position and resources. 77 Civil society may divide along sectarian or ethnic lines, or into a donor-friendly contingent of non-governmental organisation (NGO) vs other organisations that are less approachable but more representative. 78 Given the certainty of conflicting interests then, who owns SSR? A related query is how much local input into reform is necessary or desirable, from passive acceptance of externally driven policy reforms to wide societal participation in policy

12 126 S. DETZNER formulation and implementation. 79 Those focused on immediate post-conflict SSR often argue that, given that local actors (however defined) are likely to be unable or unwilling to move forward, international actors may be forced to make sensitive security decisions in place of a host state [and ] should be given the mandate and resources to make decisions. 80 Scheye and Peake suggest that local owners are those who are most effected by an institution and that therefore, in policing, owners should be local government and citizens, while central governments should handle military affairs; this is a less clear-cut division, however, where the military has handled security inside the state. 81 Finally, Bendix and Stanley argue for clarity first, where both donors and recipients should clearly state which actors they have in mind and what precise role they are willing to assign to these actors. 82 Unfortunately, donors have no consistent definition of local ownership [ ] local ownership of externally supported SSR tends to conform to the lowest common denominator. 83 External SSR promoters usually revert to the easy and familiar working with central governments. 84 In a typical scenario, elites wish to maintain an opaque and un-democratic control over the security services to ensure that it meets their interest. Donors allow them to do so partly out of respect for national ownership and state sovereignty and partly because their modes of engagement, being short term and focused upon technical rather than political solutions, are ill-suited to long-term governance challenges [ ]. 85 Regrettably, this dynamic has persisted despite an increasing scholarly consensus that greater inclusion is key to better results; effective SSR must conduct wide-ranging consultations with all affected parties in a society and facilitate these parties dialogue with each other, to generate a rough consensus about the nation s security needs, trade-offs with other priorities, and other key issues. 86 Especially where civil society is weak and interests greatly diverge, many have called for boosting the capacity of civil society actors who call for SSR so as to mobilise existing demand and increase pressure on official actors. 87 Furthermore, in this model, donors must be willing, as they rarely have been, to use political and diplomatic pressure to ensure that this dialogue is conducted in good faith. Non-state security providers In many cases, engaging solely with government security providers neglects the security and justice providers that ordinary citizens rely upon. Informal actors, England and Boucher estimate, make up as much as 80% of the security and justice sector in many post-conflict states, and the public often prefers them to state forces. 88 Williams further argues that, in Africa in particular, guerrilla forces and indigenous military organizations [ ] have played a positive role in contributing to the physical security of communities. 89 This class, not clearly defined, generally includes all those using force inside the state, other than the state, in a manner that is at least somewhat protective guerrillas, liberation movements, community self-defence forces, private security companies, paramilitary groups and political party militias, among others. 90 As frustration with state engagement has grown, so has enthusiasm for working with these providers to meet basic security and justice needs, especially immediately post-conflict and for basic policing functions, which states undergoing reform tend to neglect in favour of the military. 91 There is no consensus about how to engage with these groups. Numerous commentators have warned that such groups have varying levels of legitimacy with different local populations sometimes respected, sometimes tolerated in the absence of better, sometimes feared. 92 They generally do not operate within accepted human rights standards, are likely

13 AFRICAN SECURITY REVIEW 127 to pursue their own communal agendas, and can be corrupt. 93 Engaging with such actors requires detailed contextual knowledge that interveners often lack. Mobekk notes several instances of traditional practices that have been fabricated post-conflict to serve the interests of powerful locals. 94 However, Baker argues that refusing to work with questionable local non-state security providers when official state security providers are equally questionable or unavailable makes little sense: As long as support for such local networks is not strengthening repressive and abusive policing, but moving them towards more democratic policing. 95 Most scholars advocate for a parallel approach, wherein outside support is provided to both formal and informal systems, alongside efforts to regularise irregular forces either gradually folding them into existing structures, creating new ones to formalise them, or encouraging them to form non-violent parties. 96 The challenge may be less to link the two systems than to transform the nature of the linkage; for example, many states habitually activate local protection militias to attack neighbours or out-groups so the government can repress with plausible deniability. Some procedures for engaging non-state actors are emerging. One example is Wulf s subsidiarity principle, which holds that security provision should be as local as possible, and only when ineffective should the next highest level assume responsibility. Conversely, the norms which govern security provision should be established at the top. 97 Peace agreements SSR provisions are increasingly included in peace agreements, usually drafted in negotiations with the most powerful warring parties and those donors willing to fund such efforts. As a result, they often fail to address the security concerns of ordinary citizens or less powerful groups (who may become spoilers if insecure). 98 The desire of mediators to reach agreements that end violence quickly is understandable but has contributed to a succession of weak, unenforced and quickly disregarded agreements. Often, provisions that minimally conform to accepted basic principles civilian control over the military, parliamentary oversight, etc. threaten the powerbases of the negotiating parties if implemented. Further, donors and governments must often plan for funding higher, short-term security costs and scaling back to sustainable levels once stability has been achieved without disrupting re-established security institutions a tricky proposition that can leave donors afraid of having to fund indefinitely lest they re-spark war. 99 Worryingly, Duursma finds that the inclusion of SSR and/or DDR provisions in a given peace agreement does not have a significant impact on the agreement s durability. 100 Hutchful s examination of nine cases of SSR provision in peace agreements provides some evidence that this lack of impact, much like SSR failures generally, can be attributed to systematic failures of provision drafting and implementation; when these failures are avoided, success is more likely than when SSR provisions are either poorly written and implemented or not included in the agreements at all. 101 Specifically, he and other scholars offer evidence that agreements are more durable if more representative, and argue for consultations with as many stakeholders as possible, particularly oft-excluded groups like women. 102 Apart from consultations generating a wider buy-in, these actors may have insights on the drivers of conflicts and possible solutions ranging beyond those at the negotiating table. Finally, some have suggested directly integrating non-state security and justice actors into agreements, particularly DDR and SSR, both as an acknowledgment of their security provision role and as a first step toward legitimising and establishing oversight over them. 103

14 128 S. DETZNER Linking DDR and SSR Best practices literature stresses how the successful reintegration of former combatants creates a permissive environment for SSR, while DDR failures especially of reintegration undermine follow-on SSR efforts by promoting crime, corruption and the persistence of wartime networks. 104 The link between SSR and DDR is ideally iterative; planning for SSR should precede DDR because decisions about the size and composition of forces, but also the funding needs of security sectors, and the roles and objectives of the various institutions should shape DDR programme decisions about demobilisation, training, etc. 105 DDR efforts then shape the environment for SSR implementation by affecting the security situation, particularly with respect to crime and the likelihood of resurgence of armed conflict. 106 In practice, however, DDR often must begin before such planning and immediately postconflict, peace agreements and ceasefires can depend on demobilisation happening too rapidly to allow for an SSR assessment to take place beforehand. 107 However, disarming combatants in such conflicts may prove both difficult and, if the interveners cannot guarantee their security, irresponsible. 108 As England and Boucher argue, [t]he gap between the end of DDR and the beginning of SSR should be minimal to avoid creating a security vacuum [ ]. 109 Unfortunately, changes in practice reflecting understanding of the DDR SSR link have been minimal programming continues to neglect reintegration in favour of easier-to-measure disarmament and demobilisation. This has been amplified by cases in which a gap between the DD and the R (often caused by funding shortfalls) has led to renewed conflict. Common failures include the failure to communicate whom a DDR process will cover and what they can expect to receive to those eligible, and failures to identify and register legitimate DDR beneficiaries. 110 For example, the practice of expecting a combatant to turn in a weapon to participate in DDR both excludes some who are eligible and needy (women and children who served in unarmed roles) and may include some who were not combatants but have acquired a weapon to access benefits. 111 Many suggest addressing these problems by making DDR more participatory (amongst both combatants and receiver communities), hopefully thus shaping programming so that it is appropriate for each situation and generates buy-in. 112 Examples include enlisting communities in regulating weapons ownership as well as tailoring reintegration packages (rural vs urban, etc.). Donor coordination and capacity building Once SSR gained acceptance, donors faced the challenges of building their own capacity and coordinating within their own governments or organisations, as well as with one another. They have seen success towards the first goal and middling progress towards the second; unsurprisingly, the more political rather than logistical and technical the challenge, the greater the difficulty. Early on, the demand (from donors) for SSR programmes outstripped the supply of qualified experts, noted as one explanation for sub-par programming. 113 Eventually, the development of resources such as 2004 s OECD-DAC Guidelines on security system reform and governance makes it easier to train new personnel. 114 Still, in peace operations, the UN has struggled to find qualified recruits, with the Economic Community of West African States (ECOWAS) and the AU even further behind. 115 Other actors tend to use personnel have who security backgrounds but lack SSR-specific training, which contributes to the focus of many efforts on training and equipping rather than governance. Recently, observers have lauded the creation of expert pools of qualified civilians with SSR experience, such as the International Security Sector

15 AFRICAN SECURITY REVIEW 129 Advisory Team (ISSAT) of the Geneva Centre for the Democratic Control of Armed Forces (DCAF), which allow for SSR efforts that are more professionalised. 116 The hybrid security/development nature of SSR had early advocates which urged donor governments and organisations to take a whole-of-government approach, breaking barriers between development and military assistance. Similarly, there was a call for the division of labour between donors, with many arguing that no single entity possesses the necessary expertise to deal with all the interconnected components of any given security sector. 117 Actually achieving adequate coordination in this division of efforts has proved difficult. 118 Schroeder ascribes the lack of cooperation to a number of factors, such as differing definitions of SSR and also resource competition between actors (even intra-organisation and intro-government), different operating procedures, timelines, funding sources, etc., and incompatible mandates. 119 Law characterises much international governmental organisation (IGO) interaction as attempts to either ignore or absorb/merge with other organisations, with true cooperative endeavours, with division of labour by capability, rare. 120 Several years in, Chanaa described situations in which there had been no systematic consideration of which institutions are best suited to which parts of security reform, or which donors are best positioned to influence institutional change, 121 while Hendrickson suggests: Bilateral assistance programmes [ ] have much more latitude to deal with sensitive issues surrounding military reform. As a result, there is more scope for DFID to do the initial diagnostic work on various aspects of security-sector reform, such as defence resource management, as a prelude to greater involvement by the World Bank, for example, in financing a reform programme. 122 To improve coordination, many have called on SSR actors to sign onto a joint strategic framework for each case, derived though a comprehensive security needs assessment. Some propose that regional organisations should further share a common SSR normative framework which provides guidance for all SSR efforts. 123 Operationally, many hold that one organisation or nation should be recognised as the lead in any given effort, tasked with the responsibility of coordinating all SSR-related efforts. 124 With some exceptions, such as Sierra Leone (UK-led), these suggestions have been implemented half-heartedly, if at all. Situations where a host government plays multiple, uncoordinated bilateral partners against one another persist. England and Boucher offer a compelling explanation for this disorder, highlighting that it is unrealistic to view all multilateral coordination issues as communication and logistical problems, ignoring underlying differences in national and organisational interests. 125 For example, a far-off donor may be more concerned with a receiver state military s ability to suppress drug trafficking while a neighbouring state is focused on the reintegration that prevents former combatants from migrating across the border for work. 126 Rather than assuming a pose of disinterested altruism, interveners may be more likely to agree on a common SSR approach considerably enhancing leverage on a receiver state if they openly discuss and negotiate these interests. Similarly, if SSR must be designed based on a deep analysis of the political conditions of the receiver state, donors will benefit from pooling their information regarding those conditions and considering how the kinds of influence they variously wield can be used to shift in a more SSR-friendly direction. Receiver and regional capacity building in Africa Early scholars of SSR have pointed out that, given the emphasis on local ownership, it was essential to build knowledge and capacity in regions likely to embark on SSR