Reimagining SSR in Contexts of Security Pluralism

Within the repertoire of international stabilization interventions, security sector reform (SSR) and other conventional efforts to strengthen security and governance institutions remain central. There is increasing recognition that the policies and practices operating under the rubric of SSR are blind to the empirical reality of security pluralism in most stabilization contexts. In these contexts, both security providers directly authorized by the state (police, army) and a multitude of other coercive actors engage in producing and reproducing order, and enjoy varying degrees of public authority and legitimacy. Recognizing this, research was undertaken in three cities (Beirut, Nairobi, and Tunis) to discern the conditions enabling various security providers to forge constructive relations with local populations and governance actors. Drawing on insights generated by these case studies, this article problematizes conventional state-centric approaches and argues for a bold reimagining of SSR. It makes the case for an SSR approach that prioritizes promoting the accountability and responsiveness of all security providers, integrating efforts to strengthen the social determinants of security, and enabling a phased transition from relational to rules-based systems of security provision and governance.


Introduction
For more than two decades, the policies and programs known as security sector reform (SSR) have been a preferred instrument in the stabilization toolkit, aimed at averting the (re-)emergence of armed violence in volatile and post-conflict contexts. Bilateral and multilateral donors and their state counterparts have invested heavily in professionalizing and equipping police, disseminating new methods and strategies of state security provision, and fostering democratic oversight mechanisms. A recent review of SSR programming conducted for DFID casts doubt on the utility of these investments from an end-user perspective, insofar as their effect on the everyday security of citizens is not supported by strong evidence (Denney and Valters 2015).
The report acknowledges that, 'the dominant focus on state providers of security also overlooks the existence of alternative forms of security provision' (Denney and Valters 2015: 11). Compelling evidence on everyday security arrangements in complex contexts indicates that security providers outside state authorization are not necessarily alternatives to state policing. Rather, they stability constitute the main providers of security for many people living in areas where the state is either absent or predatory (Albrecht and Kyed 2014). Likewise, the popular credibility and legitimacy of those state institutions so favored by SSR practitioners is, in many places, doubtful at best.
State-authorized security actors and law enforcement, such as the police or gendarmerie, often serve a narrow portion of the population (Gordon 2014). Recent political economy research provides evidence of the role security arrangements and institutions play in maintaining structural privilege or suppression of certain groups (Price 2016;Van Veen 2015;Luckham and Kirk 2013). Thus, SSR interventions' tendency to exclusively engage a single set of actors, namely state institutions, risks reproducing and entrenching structural inequalities.
The provision and governance of security involves how power is exercised, by whom and for whose benefit. As such, it bears two contrasting yet complementary faces: as a process of political and social ordering, which stabilizes power relations; and as an entitlement of citizens to the protection of their safety and welfare (Luckham and Kirk 2013). Security arrangements are thus infused with and bounded by local repertoires of legitimacy, authority and historical modes of governance. The variety of unique systems this produces has, on many occasions, stymied technocratic attempts to force local security arrangements into alignment with Western-style models of state formation (Debiel and Lambach 2009). SSR policy and practice continues to follow Weberian paradigms, either by imbuing the state with a chimeric monopoly on legitimate violence, or by positioning it as the control center of a coordinated network of security providers. In either case, SSR interventions ignore the complex dynamic of multiple security actors simultaneously coexisting and competing to produce local order (Albrecht and Wiuff Moe 2015). SSR continues to focus on the state as the fulcrum of stability despite increasing recognition that the balance is often determined by what happens on its peripheries.
This article will problematize SSR in contexts of security pluralism, wherein both security providers directly authorized by the state (police, army) and a multitude of other coercive actors (armed vigilantes, political party militias) engage in producing and reproducing order, and enjoy contingent degrees of public authority and legitimacy. Drawing on insights generated by research conducted in Beirut, Nairobi, and Tunis, the article will reimagine an SSR agenda capable of improving the everyday security of citizens. Its primary contribution will be to effectively grapple with the empirical reality of security pluralism and convincingly demonstrate both why and how current SSR strategies must be refitted to address this reality.

Perspectives on Security Pluralism
Security pluralism (or plural security provision) refers to situations in which an array of actors, regardless of their relationship to the state, claim the prerogative to coercive force (Belhadj et al. 2015). 1 In such contexts, a panorama of coercive actors, including statemandated security providers, operate simultaneously and in various configurations to enforce local order. These actors' de facto authority is largely determined by their relations with local communities (Podder 2013). Plural security providers are here defined as: actors characterized by the ability and willingness to deploy coercive force, lack of integration into formal state institutions, and organizational structure that persists over a period of time, that seek to ensure the maintenance of communal order, security and peace through elements of prevention, deterrence, investigation of breaches, and punishment. 2 Numerous studies have documented the pervasive, and sometimes primary, role security actors without state authorization play in areas of limited statehood (Albrecht and Kyed 2014;Baker 2009). Where evidence of public opinion exists, some people profess a preference for these providers (Alemika and Chukwuma 2004). The variety of such actors is described in a number of taxonomies (Bagayoko 2012;Lawrence 2012;Baker and Scheye 2007). Those covered by the research discussed in this article include, inter alia, political parties and their armed retinues, networks built around sectarian identities, occupational communities, voluntary neighborhood watch groups, local youth groups (often scrutinized as 'gangs'), self-organized paid patrolmen, provisional civil defense forces, and collectives of female human rights defenders. However, assigning labels and categories to such actors risks obscuring a key contention of the research: coercive actors are not static entities; their orientation toward public or private interests and their subservience to public control can shift, or be shifted, over time. As such, observing and defining a security actor is only useful insofar as it helps to assemble a longue durée perspective of that actor's trajectory toward or away from public responsiveness.
Plural security actors may acquire popular legitimacy by effectively and efficiently resolving local security issues (Kantor and Persson 2010;Albrecht and Kyed 2014), by proving more relevant to local populations' needs (Willems and Van der Borgh 2016;Ahram 2011), or by offering more accessible services than state alternatives (Scheye 2009). These advantages notwithstanding, coercive actors operating outside the purview of the state are often associated with discriminatory and unaccountable security provision (Bagayoko et al. 2016) and human rights violations (Kirsch and Gratz 2010; Arias 2006), and ultimately prolong and increase insecurity over time, despite periodic contributions to stability (Meagher 2012). Some security providers command the state's recognition, sponsorship, or tolerance, collaborating with its courts and police services to co-produce public goods; some are exploited, with the state co-opting them to project its rule; others are ignored, marginalized, isolated, criminalized, or violently subverted. The relationship security providers maintain with their constituents, and their interactions with state counterparts, may provide key insights into the logic behind providers' use of coercion or conciliation or willingness to submit to constraints on their power.
In recent years, analyses of security pluralism have repeatedly underscored that the multifarious ways in which security provision and governance arrangements take shape demands more empirical inquiry. Specifically, the unique ways in which people access security in various contexts deserves more attention than they currently receive. Encouragingly, academic and policy-oriented literature has begun to engage with the confluence of security provision and statebuilding processes (Raeymaekers et al. 2008). This research has demonstrated that plural security providers have the potential to exert either constitutive or corrosive effects on political order (Menkhaus 2016;Podder 2013), and external engagement can influence the nature of those effects, for better or worse (Derksen 2016). This article seizes the opportunity to offer a more productive approach to security pluralism, exploring ways in which such actors may be proactively engaged to shift local security arrangements toward citizen responsiveness and accountability.

Case Study Analysis
The research initiative carried out three case studies in urban contexts with a recent history of political and/or social instability. By pursuing multiple case studies, the project was able to focus explicitly on contextual conditions unique to each site (Yin 1984), and explore without prejudice which factors appeared locally relevant to explaining how security pluralism may either improve or threaten citizen security. Each case study amassed primary data through semi-structured interviews and focus group discussions with, inter alia, municipal governance actors, members of civil society, community members and the security providers they recognized. In each case, data was collected by multi-disciplinary teams guided by a local research coordinator over the course of 5-7 days in country. To preserve the authenticity and accuracy of the primary source voice, interviews were conducted in the local national language, with translation provided for vernacular dialects. Interviews across the case studies were guided by a common set of research questions, aimed at discerning the conditions that enable security providers to forge constructive relations with local populations and governance actors to promote positive security outcomes.
Data from across the case studies could then be compared and analyzed for nascent theory development and potential generalization (Ragin 2000). Each case study was thus pursued as a singular experiment; however, the findings at the discrete sites could ultimately be tested against one another for patterns of divergence or convergence. In this section, each case is briefly described, distilling the unique evidence offered, before highlighting key findings drawn from across the studies.
Beirut (Boustani et al. 2016) is characterized by a startlingly diverse panorama of actors able to deploy coercive force in different circumstances and with varying degrees of state assent. These range from the security apparatuses of political parties to neighborhood committees to national and municipal police services. Here the boundaries between the public and private identities of agents are blurred, and there exists an unwritten and mercurial, but locally understood, division of labor distributed across a multitude of security actors. These different actors claim to defend elements of the public through their connections to and integration with state security, without ever fully submitting to state authority. For example, local residents often pointed out that, in their neighborhoods of influence, political parties are able to rouse, restrict or oversee interventions by municipal and state police. In some areas, police are seen to enter only after being given permission to do so by the locally preponderant political party. The arrangement, while clouding the supremacy of the state, nonetheless allows de jure authorities to engage de facto providers and negotiate practical arrangements to maintain daily order (Belhadj et al. 2015: 9) Crucially, a high level of social cohesion within neighborhoods and sectarian communities works to mitigate the risks and detrimental effects of an otherwise complex and unstable security system.
Since the beginning of the Syrian civil war, as many as 1.5-million Syrians have fled into Lebanon, with over 300 000 settling in Beirut. Syrians now constitute an acutely vulnerable urban population plagued by constant fear of harassment and detention, lack of protection, and limited mobility. They draw upon diverse strategies to access security, from evasion to reliance on in-group problem-solving to affiliation with sympathetic local security providers. These strategies diverge distinctly from those of Lebanese citizens. For a Beirut native, the ability to assure the safety of oneself or one's family is often determined by wasta (influence). Powerful political, familial, social and business connections can allow an individual with sufficient wasta to organize protection or demand redress, by mobilizing state or other institutions to their advantage (Boustani et al. 2016: 12). While this system has led to a plethora of security arrangements serving divergent needs of different Beirut communities, it remains inaccessible to refugees who lack the social connections required to build wasta. The case study thus concludes that the very nature of security pluralism in Beirut is unlikely to promote the equitable distribution of security as a public good, especially to newcomers.
In Nairobi ) security pluralism emerges in the gap between citizens' need for security and the state's inability to fully meet that demand. Here, plural security takes the shape of self-preservation strategies adopted by inhabitants of the city's poor urban settlements. In these areas, citizens must 'hustle' for security. That is, they must use their wits and personal networks to access protection, all the while remaining highly vulnerable to exploitation and abuse by the state and criminals alike. In many of Nairobi's settlements, state police are regarded as the most dangerous gang (Van Stapele 2015). The state's failure to guarantee security for poor urban residents is but one of the many ways in which it fails to provide for the collective good (Dafe 2009). In this context, people are incentivized and enabled to assert private and in-group interests, while the provision of collective security is contingent on its convergence with powerful private interests. The case highlights a risk of plural security provision not currently addressed under the rubric of SSR: the normalization of security as contingent upon one's individual traits and personal networks. In this way, security reform interventions are not only needed to restore the capacity and accountability of security providers, but must also energetically activate the notion of security as a public good and universal entitlement.
In Tunis (Kahloun et al. 2016), the state's capacity to ensure security has been deeply affected by the disorder stemming from the country's revolution in early 2011, which briefly led citizens to organize themselves through neighborhood protection committees. These provisional citizen collectives, consisting of men connected through personal familiarity or membership in local mosques, patrolled the streets and regulated traffic, reinforcing the local status quo despite a broader context of state withdrawal and disorder. As Tunisia transitioned toward a new democratic dispensation, the citizen collectives implored state actors to resume their hegemonic role in security provision, and, with few exceptions, quickly dissolved. While there have been no overt attempts by informal groups to challenge the state's monopoly on legitimate force, the state has not yet convincingly restored its purported (though contested) pre-revolution efficacy, leaving a noticeable security gap that erodes public confidence in the democratic transition. The revolution's impact also spread to local governance actors, straining the social contract between the state and its citizens. Inadequate urban planning compounds this, leading to social segregation and unregulated occupation of public spaces. The case highlights how the dilemma of security provision in Tunis is part of a crisis of trust between Tunisian society and state institutions.
Across the case studies, the research underlines the extent to which powerful actors are able to deploy force to sustain compliance with security arrangements that benefit them. This phenomenon is certainly not unique to conflict-affected or weak-state contexts. However, in situations where society is fragmented and the state unable to impose a monopoly on legitimate coercion, this situation can prove particularly intractable. In Nairobi's informal settlements, private citizens patrol their neighborhoods at night, acting as an effective deterrent against thieves. Typically, these guardians are chronically unemployed young men, and often begin collecting fees from residents whose homes fall within their self-determined jurisdiction. They soon find the collective threat they pose to bandits proves an equally effective deterrent against rent truancy. Residents unable or unwilling to pay their protection fees can find themselves on the receiving end of the group's intimidation, and are often left without recourse (Price et al. 2016). In Beirut, security providers sponsored by the locally preponderant political party are rarely motivated to serve people outside their constituency. Moreover, the security provided by these actors is informed and activated, not by citizens' needs, but primarily by parties' interests (Belhadj et al. 2015). In both cases the notion of security as a public good is severely undermined.
It should be acknowledged that plural security actors are not necessarily abusive or malicious. Each of the cases found that many people are, in fact, grateful for the protection these actors provide. Nevertheless, fragmented security arrangements complicate and resist the assertion of public oversight. In all three cities, citizens have limited or inconsistent access to opportunities to determine how security is produced and distributed, or hold their designated providers accountable. Notably, the research detected a broad aspiration among citizens for a state monopoly on legitimate coercion, despite contradictory everyday practices across the case studies. This would indicate that, in the three case studies, the role of the state is still valued for its potential to set the parameters of rules-based security provision and provide a mechanism for democratic control and accountability. These findings should encourage those working with state partners to design intervention models in which security providers are subordinated to civilian oversight, operate predictably and impartially, and serve the public good.
Finally, the case studies converge around the finding that, in contexts of security pluralism, security is often accessed through personalized or group-based relations, contingent upon fluctuating negotiations of identity and community. People seek security by building and maintaining relationships, often under precarious or exploitive circumstances-'hustling' in Nairobi, leveraging wasta in Beirut, or, as in Tunis, depending on communal networks for protection during a period of crisis. In each of the three cases, people described their experiences of accessing security through strategic contacts, and cited this as a profound source of anxiety and disempowerment (Price 2016). Building and sustaining more impersonal, rules-based mechanisms for security provision may provide a vital and viable path for improvement of how citizens ensure their everyday security. It may also represent an opportunity for state security actors to become preferred providers, should they furnish credible opportunities to appeal for protection or redress, regardless of identity or social status.
Though it is tempting to pursue models and templates that have 'worked elsewhere', the case studies indicate that arrangements to promote accountability and curtail arbitrary power must be calibrated for context and anchored in cultural and social consensus. This implies the installation of outside models should be eschewed for a gradual approach to encourage rules-based systems that uphold and are upheld by local values and preferences (Price 2016).
To summarize, the three case studies each produced corroborating evidence that in contexts of security pluralism citizens still exhibit a preference for security provided by an impartial authority bound by predictable rules, often imagined as the state. Nonetheless, they are compelled to navigate systems that provide security precariously and contingently, and are organized, not as a public service, but to advance the interests of particular factions. In view of this, efforts to improve security for citizens may be better channeled toward enhancing the accountability and equability of various security providers, with the aim to nudge relational, unchecked systems toward increasingly inclusive, rule-based arrangements.

Implications of Security Pluralism for SSR
A predominant goal of SSR interventions is to expand the state's ability to broadcast security provision across its territory by coopting, supplanting or eradicating those actors operating outside its direct authority. As demonstrated in the previous section, this approach is fundamentally at variance with the empirical reality in contexts like Lebanon, Kenya, and Tunisia, where such actors may be thoroughly entangled with the state apparatus (via political parties in Beirut) or may be serving a client base the state has not proven willing or capable to protect (such as the urban poor in Tunis and Nairobi). While some international donors have sought a more inclusive approach to security provision and governance (for example, OECD 2007, DFID 2004, the majority are reluctant to risk upsetting host government relationships, conferring legitimacy on groups with untenable aims or means, or empowering particular factions over competitors (Derks 2012: 19).
The research outlined in this paper suggests that SSR programs could achieve greater impact by endeavoring to address a broader swathe of the security provision panorama. To cite one example: strengthening oversight and accountability of stateauthorized police is often a key objective of SSR programming, but such work is equally valid for curtailing the power of actors that operate outside state jurisdiction. SSR practice could seek to promote the principle that those vulnerable to transgressions of security providers (of any stripe) should possess swift and effective mechanisms for exercising oversight and accountability (Price and Van Veen 2016: 18). Similarly, activities designed to strengthen the integrity and restraint of state-authorized security providers could be inventively adapted to assist local communities in placing benchmarks and performance standards on their designated providers (Price 2016: 26). In areas where multiple groups operate in parallel, working with civilians to design and broker a functional division of labor amongst security providers could help impose a system of checks and balances and distribute power more evenly across diverse actors.
A key insight of the research is that security entails more than the professional and effective functioning of designated security providers. Across the research sites, people attributed their safety to social cohesion (community solidarity, strong family ties, cultural values, etc.) that reduces vulnerability and militates against disorderly or anti-social behavior; and collective efficacy, or the ability of community members to control the behavior of neighbors through social codes, especially in areas where violence is commonplace (Uchida et al. 2014;Arjona 2016). Community-level networks, trusted dispute resolution mechanisms, and shared norms can empower people to take collective action and demand responsiveness and accountability from authorities, including security providers (Arjona 2016). The research also points toward the importance of reliable public services and infrastructure, not only for the convenience and safety they provide, but also for activating the social contract. Deeper consideration could be given to how citizens' daily security is shaped by mechanisms other than official 'policing'. Bolstering citizens' ability to effectively demand public services can increase their experience of, and claim to, basic entitlements, including-but not limited to-security . The research strongly supports the value of multi-faceted initiatives that complement conventional SSR and community policing approaches with activities to address the wider social determinants of security.

SSR Reimagined: Key Priorities
Drawing on the insights from the case study research, this article encourages international donors and practitioners to boldly and creatively reimagine SSR. By shifting away from a reflexive state-centrism and re-focusing on the functional principles of equitability, accountability, predictability and responsiveness, security provision arrangements (regardless of their form) can be encouraged to produce better outcomes for end-users. At the core of this argument is the observation that those forms of order that have held volatile societies together and offered some degree of protection to citizens should be regarded as at least potentially constitutive of positive development (Meagher 2012). Engagement will demand new strategies and modalities of support, a higher tolerance for risk, and longer time horizons.
Where SSR interventions exclusively engage state-authorized actors, impact will be limited to the beneficiaries of state protection and rarely benefit the most vulnerable in a society. Put differently, assisting one type of actor in contexts of security pluralism inevitably risks privileging some groups over others (Price 2016). Moreover, interventions that focus solely on developing state actors' capacity ignore the potential for innovative and high-impact investments in strengthening social cohesion and collective efficacy.
Intelligent urban design can improve the frequency and quality of interactions between residents and promotes natural surveillance of streets and public spaces (Cassidy et al. 2015). Policies designed to expand access to public services can also reduce citizens' perception and experience of insecurity, and reduce reliance on divisive or unreliable providers (Price 2016). This research indicates that, in order to produce positive effects on citizen security, SSR interventions must be integrated with broader efforts to address these and other social determinants.
Below, two priorities for a reimagined SSR are outlined, with practical recommendations for policymakers and practitioners.

Priority 1: Strengthen civic space and institutionalized mechanisms for asserting public oversight of all security providers
The research identified the lack of responsiveness and accountability among security providers as one of the most pernicious aspects of security pluralism. In contexts of pluralism, formal and informal governance systems do not necessarily feed into each other in predictable ways, and neither is inherently conducive to meaningful public oversight (Bagayoko et al. 2016). It is vital for SSR practitioners to contribute to fostering an environment in which all security actors are subordinated to citizen control and oversight. Given the Janus-faced nature of all coercive actors, practitioners must recognize that these actors' production of security or insecurity is constantly shaped by statutory, political, and economic incentives, and nearly all rely to some extent on local consent or cooperation, or at least the absence of collective resistance. As such, SSR policy and practice could prioritize revealing and testing how the incentive structures for plural security providers might be tilted toward greater citizen accountability and responsiveness. Of equal importance are the processes by which these actors succumb to elite capture or subversion by transnational criminal networks (thereby unmooring them from local populations). Investigations into how such processes might be discouraged or arrested could prove invaluable for effectively insulating security providers from external co-optation (Meagher 2012). Robust political economy analysis to discern sources of power and interests of various providers and gauge the potential to extend protection more inclusively is likely to prove useful if incorporated into program design (Price and Van Veen 2016).
Policies and programs should be tailored to increasing various providers' responsiveness to citizens and reinforcing or expanding their reliance on public approval over time. In each operational context, a starting point should be the identification and analysis of existing local citizen and community strategies for engagement with coercive actors, ranging from co-production of security to resistance and exit from hostile situations. This might be achieved by fostering opportunities to safely articulate citizen and community voices and mobilize collectively for demand-driven reform at the local level. Though variable in form and function, the community policing programs, often deployed under the rubric of SSR, offer a model for providing citizens with the political space to negotiate constraints on security providers' power and make explicit the conditions upon which they may earn public credibility (Denney 2015).
Creative energy should be put behind adapting the strategies and tools of community policing to increase the public responsiveness of the range of security providers that local residents deem relevant (Price 2016). For instance, this could lead to experimentation with novel means to promote dialogue between citizens and plural security providers. Focus could be usefully placed on enabling citizens to: articulate priorities, benchmark and evaluate performance, make actors' legitimacy contingent upon respect for agreed parameters, seek redress for abuses, and circumscribe the ability of coercive actors to rely on external sponsors. Special attention must be paid to understanding the scalability of such initiatives, or the potential for upwards interlacing with state-level mechanisms for responsiveness and accountability.
Encouraging the notion of security as a public good (rather than as a vehicle for group-assertion) might reduce the risk of plural security provision practices reinforcing or exacerbating social cleavages. As a means of encouraging security providers to demonstrate responsiveness to all (or at least an ever-wider constituency) of citizens, SSR initiatives could include offering political and social incentives for security providers to expand their client base, in coordination with other actors. A counter measure may be to invest in exposing the identity politics or financial machinations that underpin unequal security provision, including by state-authorized actors. Stimulating public mobilization in favor of more equitable security, by strengthening civil society and independent media, could raise the standard to which security providers (state or otherwise) adhere or aspire in order to maintain a necessary level of public credibility.
It must be recognized that working to reduce the practice of security as a personalized (especially identity-based) privilege constitutes a social shift, and would therefore require long-term programming that focuses on citizens and providers alike (Price 2016). Careful efforts to consolidate protective communities, through brokering truces or designing divisions of labor, if negotiated patiently and with deep contextual awareness, may not only make security arrangements more inclusive, but could also establish basic performance standards perceived to have universal applicability.

Priority 2: Identify and invest in intermediate steps in the transition from relational to rules-based security systems
Rather than understand the purpose of SSR as the construction or restoration of a state monopoly on legitimate violence, SSR stakeholders should frame and strategize their interventions around stimulating a long-term shift from relational to rulesbased security systems. That is, end-users should experience an incremental change whereby their access to security is eventually conditioned by universal entitlement to security as a public good, not through personal negotiation or political privilege. Such a transition requires changes in the way people and groups relate to one another, and how power is attained, exercised, contested and curtailed. This amounts to a social transformation, a process goal not currently foregrounded in most SSR frameworks.
With such a transformation in mind, SSR policy and practice should endeavor to support context-specific transitional arrangements that represent intermediate steps between relational and rules-based systems of security provision and governance. As a starting point, SSR interventions should address the panorama of security providers, rather than state-authorized actors in isolation. Much as donors invest in the professionalization of state-authorized security providers, assistance could be conditionally offered to plural security providers that meet a minimum set of criteria to reduce (if not eliminate) abusive practices and promote adherence to national and international standards, while improving security outcomes produced for communities (Baker 2012).
An SSR strategy that recognizes the reality of security pluralism might stretch beyond enhancing the quality of security provision to incorporate diverse and empirically grounded models of regulation, seek to promote specific principles of accountability, predictability and equitability therein (for example, Baker 2008, Wulf 2006. Improving cooperation and coordination between state-authorized and other security providers might commence with a mapping of who is providing which security functions, followed by delineation of recognized spheres of influence. Reconciling different types of authority at the various levels of security provision and governance to create a functional distribution of labor would be a central SSR outcome (Wulf 2007). In this regard, practitioners might learn from examples of durable, negotiated, functional divisions of labor between security providers in contexts of pluralism. This is illustrated by earlier research on Beirut (Belhadj et al. 2015) and in relations between the federal police service and Hisbah religious police in Kano, Nigeria (Hills 2014).
There are presently few attractive options for addressing the coordination problems, frictions, and inequalities that emerge from a dispersed and crowded security sector. The principal challenge in navigating any transition from a relational to rules-based paradigm will be evolving a means of articulating and protecting the notion of security as a public good. It remains to be seen if security as a public good can be sustained in the absence of a common institutional framework able to exercise what might be called a 'meta-authority' (Loader and Walker 2004) over the wild profusion of coercive actors, and to what extent intermediate arrangements might congeal into the kinds of political settlements that are constitutive of statebuilding, or entrench dynamics that undermine emergent state legitimacy.

Conclusions
A traditional understanding of the state as the sole legitimate provider of security may be defensible as a normative preference, but is deeply flawed as a description of existing security environments in most countries. This renders it a weak basis for policymaking. This article contributes to the growing body of literature that questions the validity of investing in the establishment of a Weberian state monopoly of legitimate violence in complex contexts where doing so is demonstrably unrealistic in the short-or mediumterm, with deleterious impacts on everyday security for citizens.
As a short-term priority, there is a need to enlarge the scope and deepen the optic of research, to ensure that international and national SSR policymakers and practitioners are able to draw upon nuanced, contextualized evidence as they develop responses to citizen insecurity. In particular, attention is needed for how patterns of security pluralism in contexts of social fragmentation diverge from those in contexts of pervasive social cohesion. There is also an urgent need for detailed empirical inquiry into the impacts such arrangements have on the people who are 'secured', especially in relation to those constituencies that experience inequality, insecurity and violence. Efforts to understand local security contexts should assess whose interests particular security arrangements serve, and gauge the potential for expanding inclusiveness and increasing accountability.
SSR policymakers and practitioners must avoid romanticizing security pluralism, as an undiscerning rush to embrace the 'local' and 'traditional' can obscure regressive and coercive features of governance beyond the state (Meagher 2012). Still, there is a clear need for more innovative, problem-solving approaches that accelerate delivery of security to citizens as an entitlement, rather than a political or economic privilege. The conventional preoccupation of SSR with state-centric institution-building should be shifted toward a more ecumenical approach that puts end-users, and the local arrangements they regard as at least minimally legitimate, at the center of analysis. Donors should experiment with a wide range of new methods and tools aimed at asserting public control over security providers, creating civic space and reliable mechanisms through which people might demand responsiveness and accountability from a panorama of providers, and using experiences from the field of community policing as a guide.
Reimagining SSR summons its proponents to unleash a proliferation of contextuallyrooted models for transitioning toward rules-based systems of security provision and governance, integrated with efforts to strengthen the social determinants