Engaging Non-State Security Providers: Whither the Rule of Law?

The primacy of the rule of law has long been seen as one of the essential principles of security sector reform (SSR) programming, and part of the larger gospel of SSR is that the accountability of security providers is best guaranteed by embedding security governance within a rule of law framework. Acknowledging the reality of non-state security provision, however, presents a challenge to thinking about SSR as merely the extension of the rule of law into the security realm, in large part because whatever legitimacy non-state security providers possess tends to be grounded in extralegal foundations. This paper – more conceptual than empirical in its approach – considers the implications of hybrid forms of security governance for thinking about the relationship between SSR and rule of law promotion, and argues that the rule of law still provides a useful source of strategic direction for SSR programming.


Introduction
The rule of law has long been a core pillar of security sector reform (SSR) programming. To the extent that SSR seeks to ensure that security forces are not only effective but also accountable to both the state and its citizens, the proposition that accountability is best guaranteed by embedding security governance within a rule of law framework has, with a few notable exceptions, gone relatively uncontested despite SSR's uneven track record. In the context of post-conflict transitions in particular, the standard SSR narrative has been that the (re)-consolidation of coercive power within the hands of the state (in a Weberian sense) is both justified and legitimized by the parallel establishment of legal and institutional frameworks that serve to constrain and limit the uses and abuses of this power. Just as SSR lies at the core of the contemporary statebuilding agenda, part and parcel of a larger effort to create capable, accountable and responsive states (Barnes 2009: 3), the rule of law, as a set of principles and practices aimed bringing political and socio-economic relations within a predictable, transparent framework of enforceable rules, remains central to the contemporary SSR agenda.
That this narrative remains compelling both in terms of its internal logic and in terms of how most SSR practitioners differentiate between successful and unsuccessful security sectors goes a long way towards explaining the unease with which the growing emphasis on non-state security provision has been received within the wider SSR community. Non-state security providers, in this context, include those actors -from militias to neighborhood watch groups to traditional stability Wilfrid Laurier University, CA tdonais@wlu.ca chiefs -who command coercive power, and provide a measure of security and protection to particular communities and/or across specific territories, outside of the context of formal state security provision. Hybrid forms of security provision, in which state and nonstate security providers co-exist and overlap, are increasingly acknowledged as the reality in many fragile and conflict-affected states. Typically, however, such security ' orders' are decidedly disorderly, inherently unstable, and sometimes violent, presenting myriad points of friction where the claimed and contested jurisdictions of various security providers overlap. While they may be acknowledged, such hybrid security arrangements have rarely been considered viable alternatives to conventional SSR approaches. At best, they have tended to be viewed as ephemeral features of the transitional landscape, to be tolerated until such point as the state can take up its rightful monopoly over the legitimate use of force.
While SSR is not easily disembedded from these state-centric presumptions, evidence is slowly accumulating that hybrid forms of security governance may be more durable, more effective, and less easily-displaced than previously thought. While it may be premature to declare, as Bruce Baker (2010) has, that 'the future is non-state,' the case for embracing hybridity in SSR programming is gaining strength. It also rests on decidedly pragmatic grounds. As Kate Meagher (2012) has observed, the willingness to re-consider the viability of hybridity as a model of security governance has much to do with the search for less elaborate and less costly forms of governance, and with a growing recognition that existing systems of security governance should be judged for what they are, rather than what outsiders would like them to be. Consistent with broader critiques of liberal peacebuilding, this openness to hybridity is also a reaction to the hubris and ' arrogant managerialism' of most SSR policy interventions, marked by unrealistic and unachievable social engineering ambitions and ongoing efforts to jam the complex realities of weak and conflict-affected states "into a procrustean bed of pre-set rule of law templates" (Meagher 2012(Meagher : 1076Raeymaekers et al 2008: 10).
This paper explores the role of non-state security provision in SSR contexts against the wider backdrop of an ongoing normative and policy commitment on the part of donors to embedding SSR within a rule of law framework. In doing so, it contemplates the possibilities for a 'post-liberal' (if not necessarily post-rule of law) SSR agenda, distinguished from its liberal precursor by a commitment to fashioning SSR strategies on the basis of existing socio-political realities within the society in question rather than on idealized (and possibly unattainable) end-points. This emphasis on starting conditions rather than (or at least in addition to) ultimate outcomes -consistent with Amitai Etzioni's argument that, given the limits of international influence, it makes more sense to build on existing structures and trends "rather than seeking to fashion new ones out of whole cloth" (2009/10: 54) -imposes considerable demands on outside interveners in terms of understanding the local context in all its dynamic complexity. It demands, in short, a systems-analysis approach to SSR, based on a careful reading of the relevant actors, the incentive structures they face, the institutional and relational dynamics that connect them, and the location of potential levers of change. It also leads, almost inevitably, to a form of SSR that is based on the balancing and bridging of existing political forces rather than on Weberian monopolies contained and constrained by a framework of laws capable of regulating political life while simultaneously standing outside of politics. At the same time, while it is unrealistic to expect donors to set aside long-standing commitments to the rule of law in their SSR interventions, it may also be unnecessary. Indeed, the paper makes the case that viewed as one component -albeit one to be progressively expanded over time -of a complex and evolving accountability framework for security provision, the rule of law remains central to the broader SSR enterprise. The remainder of the paper unfolds as follows. The next section unpacks the key foundational premises of conventional SSR, with a particular focus on the prominence of monopoly and accountability and the enormous challenges of achieving either -let alone both -within the standard timeframes of most contemporary international interventions. Next, the paper considers the awkward relationship between non-state security providers and the rule of law, while at the same time outlining why hybrid security arrangements involving a combination of state-based and non-state security provision are likely to be more conflictual than collaborative. The third section outlines a vision for longer-term security sector evolution grounded in a 'rules-based' framework, emphasizing the emerging ability of state institutions to regulate, rather than monopolize, the provision of security. The conclusion re-visits the overall argument, suggesting that to the extent that SSR remains about the systemic transformation of security provision, the rule of law continues to provide an important set of strategic guideposts to guide this process.

Conventional SSR: The Merger of Monopoly and Accountability
Louise Andersen has observed that the socalled monopoly model of SSR -central to the project of establishing liberal peace in fragile states -"involves not merely the taming of the Hobbesian Leviathan but the actual establishing of the Leviathan" (2011: 12). This characterization nicely underlines the intertwined principles of monopoly and accountability upon which conventional SSR is premised, while also hinting at the epic scale of the undertaking. It has been clear for some time that SSR's weak empirical track record has a great deal to do with the gap between lofty principles and on-the-ground realities, and between the broader ambition of the SSR agenda and the time, resources, and political capital required to transform that ambition into reality.
On the monopoly question, perhaps the most important insight to have emerged from the last quarter-century of SSR programming concerns the limited remit of the state -and its security and justice apparatus -across a wide range of fragile and conflict-affected environments. It is now commonly asserted and widely accepted (if difficult to verify) that upwards of 80 per cent of security and justice provision in the states that are the beneficiaries of SSR programming is provided by non-state actors (Denney 2012: 1;Albrecht and Kyed 2011). Indeed, part of the very essence of state weakness or fragility relates to the inability of governments to exercise effective control over territory, while conflict leads to the further fragmentation of security provision. Given these realities, in most cases conventional SSR programming has struggled to meet the challenge of engineering massive transfers of power from non-state actors -most of whom have proven to be reluctant collaborators -to state-level actors.
As Ken Menkhaus (2016: 6) has noted in the case of Somalia, efforts to strengthen the formal security sector in that country are "swimming against powerful currents"; nonstate security providers are not only more capable than state-level actors across most of the country, they have also developed powerful economic interests in the maintenance of the status quo. While Somalia may be an extreme case of fragmented security provision, the failure of the monopoly model of SSR to deliver on its core premise is a common theme across a broad cross-section of post-conflict cases.
Conventional SSR has arguably been no more successful in the achievement of its second core principle: ensuring that those who wield coercive force behave responsibly and can be held accountable for their actions. It is here where the rule of law intersects most directly with security governance; in the typology of Thomas Carothers (1998), this represents 'type three' rule of law reform, aimed at ensuring government compliance with the law and, more generally, putting in place robust mechanisms to constrain the powerful. Particularly in the context of volatile and insecure environments, it should come as no surprise that those in positions of privilege see little self-interest in limiting their power by subjecting it, and themselves, to the rule of law. Indeed, as Agnes Hurwitz (2008: 2) has noted more generally, "programs seeking to strengthen or re-establish the rule of law in peacebuilding contexts have rarely achieved their nominal objectives of delivering human rights, security or development." This is due, in large part, to the reality that the rule of law is about changing norms at least as much as it is about building institutions, and normative change is almost invariably a long-term endeavor (Stromseth et al 2006: 75). For domestic elites especially, respect for and adherence to abstract principles such as justice, accountability, and transparency is a tough sell in cost-benefit terms, particularly when set alongside the more prosaic pursuit of political and economic self-interest. Further, as Alex Berg (2012) has demonstrated, rule of law in conflictaffected contexts rarely emerges as a result of elites ' coming to enlightenment', but is rather the consequence of specific, and relatively uncommon, patterns of state-society relations -notably regimes rooted in broad or fragmented coalitions and lacking easy access to revenue -that alter the incentive structures facing elites in ways that make it more likely for them to accept legal and institutional constraints.
Given the difficulty of realizing the enormous ambition that lies at the heart of the conventional SSR paradigm -vis-à-vis both restoring monopolies over the legitimate use of force and embedding security governance within a robust legal framework -the search for alternative and more realistic models has become increasingly urgent. In this sense, critiques of conventional SSR echo Marina Ottaway's (2003) broader critique of the democratic reconstruction model as attractive in theory but unworkable in practice, given the enormous gulf between ground-level realities and idealized endpoints. Like Ottaway, advocates of second-generation SSR seek more realistic and less hubristic approaches that nevertheless retain a fundamental commitment to improving both human and state security within fragile and conflict-affected contexts. While hybrid approaches promise such realism by eschewing formal templates in favour of strengthening actually-existing mechanisms of security provision, almost by definition hybridity also entails the reconciliation of radically different practices and principles. How to go about reconciling the recognition of non-state security provision with an ongoing commitment to rule of law promotion presents one such paradox.

Non-State Security Providers and the Rule of Law: An Uneasy Relationship
While the flaws of both liberal peacebuilding and conventional SSR have been exposed in recent years, in large part due to the incapacity of each framework to bridge the gap between promise and performance, the rule of law continues -somewhat remarkably -to enjoy deep and near-uncontested legitimacy. This is perhaps even more remarkable given that the rule of law can be considered primus inter pares among all of the core principles underpinning liberal interventions in fragile and conflict-affected states: the rule of law is, in other words, an essential background condition for the achievement of key public goods associated with the modern paradigm of good governance, from economic development to human rights to democratization.
While there is a rich literature debating both its meaning and its substantive content, at its core the rule of law can be defined, in the words of Thomas Carothers (1998: 96), "as a system in which the laws are public knowledge, are clear in meaning, and apply equally to everyone." The rule of law, as Carothers also notes, is fundamentally dependent on the fairness, competency, and efficiency of core legal institutions such as courts, prosecutors, and police, and more generally on the embeddedness of government -and governance -within a comprehensive legal framework (1998: 96). Two aspects of this definition appear especially relevant for the purposes of thinking about the relationship between the rule of law and non-state security provision in transitional contexts. The first is its undeniably statist framing; while most conceptions of the rule of law contain articulations of the rights of citizens to due process and equality before the law, the core puzzle faced by rule of law reformers in fragile and conflict-affected states is ultimately how to both enable and constrain government power, on the wider principle of 'no power without accountability' (Gowlland-Debbas and Pergantis 2009: 321). As Lisa Denney (2012: 1) has suggested, however, the terms 'non-state' and 'informal' remain analytically useful when thinking about hybrid security arrangements precisely because "they denote the broad set of arrangements that, in some way, operate beyond the state's accountability net." Acknowledging the reality of non-state security provision, in other words, remains a challenge to thinking about SSR as merely the extension of the rule of law into the security realm, in large part because the legitimacy of non-state security providers tends to be grounded in extralegal foundations.
The second aspect of the Carothers' definition worth noting in this regard is its seemingly apolitical nature, with laws and their guardians cast as neutral arbiters of political and social life. Framing the rule of law this way, however, conceals as much as it reveals. Given that laws themselves, being little more than words on paper, have no inherent authority, genuine rule of law -as opposed to rule by law -requires a robust and durable intersubjective agreement on the part of the constituent elements of any society, especially those in positions of power, to submit themselves to the authority of abstract law. In this sense, acceptance of the rule of law on the part of both rulers and ruled constitutes -at least in liberal democratic contexts -a central component of the social contract through which state-society relations are governed. Historically, the emergence of social consensus on the centrality of the rule of law as a bedrock of governance has come only through prolonged, and often violent, political struggle (think of England's long journey from the Magna Carta to modern constitutional monarchy), the outcome of which is by no means pre-determined. The fundamental challenge facing those seeking to embed the rule of law within conflict-affected states is, therefore, that few good models exist for how to short-circuit the messy and violent dynamics of political contestation in order to build consensus among differentially-empowered (and mutually-distrustful) social actors on the wisdom, desirability, and legitimacy of the rule of law as an overarching governance principle. Ultimately, as Janice Stromseth et al (2006: 77) have suggested, "few rule of law theorists have grappled with the issue of how rule of law cultures can be created." While holding to the conviction that the rule of law provides the only durable, sustainable framework for responsible, accountable security governance, then, conventional security sector reform models have never really offered a convincing theory of change for how to bring this about. Nor have they fully come to terms with the reality that, in most cases, both state and non-state security provision will continue to co-exist for an indefinite interim, drawing on a wide range of different sources of legitimacy, offering variable levels of security or insecurity, and forcing citizens -as security consumers -to navigate what are often security terrains of exceptional complexity (Menkhaus 2016: 34). For external reformers, such terrains are no less difficult to manage (even if less existentially threatening), in part because of the difficulty of distinguishing good actors from bad ones and in part because of the inherent limits on external leverage. Consequently, donors continue to focus on reforming statelevel security and justice systems, while overlooking the majority of mechanisms through which justice and security are delivered on a daily basis (Denney 2012: 1). Conversely, beginning to think in terms of 'interim security arrangements', even if the interim in this context may be measured not in years but in decades or even generations, necessarily requires a willingness to engage with -rather than attempt to circumvent or transcendthe messy realities of actually-existing security arrangements.
One of the earliest efforts to frame this kind of engagement in policy terms was provided in 2011 by the Development Assistance Committee of the OECD. Highlighting the centrality of legitimacy to larger debates around governance, the OECD-DAC made the case that 'grounded legitimacy' -pursued through "deliberate strategies for supporting the marriage of indigenous, customary and communal institutions of governance with introduced, Western state institutions, with a view to creating constructive interaction and positive mutual accommodation" -should be a key guiding principle in efforts to rebuild fragile or war-torn states (OECD 2011: 38). While the idea of grafting Western norms and institutions onto pre-existing systems that resonate socially and culturally with local populations is compelling, in the particular realm of security provision such 'marriages' between state and non-state actors are likely to be especially fraught. To highlight such tensions is not to deny that collaborative security arrangements across the state/non-state divide can, and do, exist independent of outside intervention; Baker (2016: 4) has described precisely this kind of collaboration between state authorities and customary structures in Somaliland, for example. It is, rather, to note that such arrangements may be the exception rather than the norm precisely because of the breadth and range of actors that comprise security systems in conflict-affected environments, the particular nature of power relations in such contexts, and tensions inherent in the private delivery of public security.
In the first place, the universe of non-state security actors is remarkably varied, ranging from traditional chiefs to secret societies, and from neighborhood watch groups to gangs, militias, and warlords. Such actors may have long-standing bonds of reciprocity with their client communities, or they may have emerged from within the conflict context with little history and few direct connections with particular communities. William Reno, for example, has distinguished between protective and predatory militias, with the former dependent on local communities for resources and connected to them by dense webs of values, beliefs, and identities (cited in Lawrence 2012: 15). Complicating matters, of course, is the reality that particular actors may simultaneously be perceived as both predatory and protective by different segments of the communities with whom they interact, with perceptions varying significantly across time. More generally, Baker and Scheye (2007: 517) have argued that there are no a priori grounds for assuming that non-state actors are less capable than non-state actors of upholding human rights or being held accountable, since they may "more accurately reflect local beliefs and needs and are regarded by local people to be more legitimate." Certainly, the varied experiences of non-state security provision across a range of cases demonstrate that the rule of law is not a pre-requisite for accountability: despite the dramatic power differentials between the providers and recipients of security, there is some evidence of warlords being 'tamed' by links with more traditional forms of organization, and of militias -particularly those that are embedded within specific communities -being ' civilized' by social pressure (Meagher 2012(Meagher : 1080. Social embeddedness, however, offers no durable guarantee that the 'protected' will be able to reliably hold their 'protectors' accountable, given the shifting and unpredictable nature of most informal governance arrangements: in other words, non-state security provision can just as easily erode as uphold the security of particular communities. Perhaps unsurprisingly, then, context remains all-important. Second, as SSR has always been part of a larger project centred around re-arranging the manner in which power is exercised and controlled within particular societies, hybrid security arrangements are as likely to generate competitive power dynamics -both across the state/non-state divide and among non-state security actors -as they are to yield respectful and mutually-reinforcing co-habitation among differentially-situated security providers. There is, on the one hand, the reality that in such contexts the state, given the high stakes involved and the long-standing presumption that security provision is at the very core of what defines contemporary statehood, is unlikely to enthusiastically embrace an emerging norm of hybrid security governance (Lawrence 2012: 18). At the same time, the lingering insecurity of the post-conflict 'moment' and the political economy of private security provision -in contexts of resource scarcity, many security providers find it difficult to resist the temptation to leverage coercive authority for either political advantage or economic gain -point to real risks that in the absence of some form of regulatory framework, ongoing struggles for power and authority could easily turn ugly. Indeed, South Sudan's postindependence descent into civil war can be read precisely through this lens of competitive security dynamics.
Third, as Baker and Scheye (2007: 519) have noted, both justice and security are -at their core -public goods, a reality which sits awkwardly with hybridized security arrangements. While there is of course no guarantee that public security providers will take seriously their responsibilities for public security provision -indeed, post-conflict environments are sadly replete with examples of the exploitation of public office for private gainthere is at the very least a normative expectation that over time, public security forces will act in the name of public security. Hybridity, conversely, implies multi-layered and overlapping security provision, with non-state actors in particular providing security to selected slices of a particular population, while representing agents of insecurity to others. In such contexts, the provision of 'public security' may be uneven and incomplete at best, while the prospects for encouraging a multiplicity of non-state security providers to embrace a public security ethos remain decidedly uncertain.
In light of such concerns, there remain grounds for caution about the long-term capacities of hybrid security arrangements to offer superior outcomes, in human security terms, to the long-suffering populations of conflict-affected states. Indeed, writing about the specific context of Africa, Kate Meagher warns of the dangers of inverting, rather than overcoming, the essentialist tendencies of previous thinking around security governance. As she suggests, to the extent that "the condemnation of non-state order as institutionally destructive has been replaced by its celebration as a vehicle of embedded forms of order and authority," there's a risk of failing to make important distinctions between constructive and corrosive forms of non-state order (2012: 1074). At the same time, in the rush to embrace ' actually-existing' security governance arrangements in lieu of striving for ideal-type outcomes, there is also a danger of losing sight of the reality that SSR is, at its core, about systemic change; indeed, one of the characteristics of the literature on non-state security governance has been an emphasis on tactical improvements to ground-level security arrangements at the expense of a more sustained focus on how wider security systems might be transformed over longer timeframes. I take up this question in the following section, suggesting that even in the context of security hybridity the rule of law, as a set of overarching principles of governance, may continue to offer an important set of guideposts that enable SSR practitioners to go beyond acknowledging the role played by non-state security providers in SSR contexts towards engaging with them in constructive, forward-looking ways.

Squaring the Circle -A Qualified
Defence of the Rule of Law One starting point for reconciling a continued commitment to rule of law promotion with a recognition of the reality of non-state security provision is the realization that most advocates of non-state security strategies are not as radical as they may appear at first glance. Either implicitly or explicitly, most continue to acknowledge a crucial -if somewhat transformed -role for the state within any evolving security governance framework. Likewise, most also continue to acknowledge the imperative of enveloping security governance within an enforceable rules-based framework. Baker and Scheye, for example, posit that regardless of the specific identity of the actors who provide security services: A principle function of the postconflict and fragile state might be to monitor, license, and regulate the activities of non-state service providers. This is no longer a state defined in terms of a monopoly control over violence and coercion, but rather a highly circumscribed and limited state, working in varying unique partnerships and associations with nonstate actors and CSO's (2007: 519).
Michael Lawrence, similarly, in the context of a broader argument around the need to develop non-state SSR strategies, defends the notion of the regulatory state, empowered to set broad parameters for security provision, "particularly standards of human rights, accessibility and accountability" (2012: 10). Even in a context such as Somalia, the classic case of a 'mediated state' where weak central authorities have little choice but to broker deals with powerful non-state actors, Menkhaus concludes that state regulation of private security provision remains a possibility, albeit part of a long, convoluted process "by which state authorities eventually gain primacy over non-state and sub-state security providers" (2016: 38-39).
What emerges from these accounts, therefore, are hints at a long-term, incrementalist theory of security sector transformation aimed at facilitating a gradual shift in the balance of power from non-state to statelevel actors, while at the same time re-positioning the state as a regulator, rather than monopolizer, of security provision. Crucial to this account of change is the state's developing capacity (and legitimacy) to make and implement rules, laws, and regulations. While the state might have little choice but to defer to the capacity/legitimacy of nonstate security providers in the short-term, and share authority with these same actors over the medium term, over the longer term the sovereign state is expected to assert its primacy -through what Menkhaus (2016: 38) terms a combination of negotiation, confrontation, and co-operation -over the non-state in matters of security governance. None of this, of course, is necessarily inconsistent with the shorter-term imperative of making ' actually-existing' security governance work better through ongoing efforts to build partnerships, facilitate collaboration, and ease friction across the broad range of security providers.
Seen in this light, a looser conception of how the rule of law might over time link state and non-state, security provider and security consumer, and different kinds of security providers with each other may still provide a reasonably compelling framework for international engagement with the security sectors of fragile and conflict-affected states. While avoiding the perils of both 'legal orientalism' (Heupel 2012: 168) and externally-driven social engineering, the strength of such a vision may lie in its ability to bridge the gap between the imperative of starting SSR from actually-existing conditions and Alice in Wonderland's dictum that you need to know where you are going if you ever hope to get there. In this sense, then, conceptualizing SSR in terms of the gradual expansion of the state's ability to bring security provision within a common framework of rules provides at least some direction and focus to the generic call to ' engage' non-state security providers, without being overly prescriptive in terms of eventual outcomes.
The importance of holding onto at least a thin vision of how external interveners imagine change in security systems unfolding over time should not be underestimated, especially given the gap between the need to think about change in systemic terms and the chronic inability of international actors -despite ritual nods to the importance of 'holism' as a key SSR principle -to engage with the security sectors of conflictaffected states as complex systems. Indeed, the failure to cooperate, coordinate, and plan strategically remains, in many ways, the Achilles' Heel of the entire SSR enterprise, which in too many cases still unfolds more as a series of discrete, time-bound, and unconnected projects than as a coherent and integrated blueprint for shifting conflict-affected societies along a continuum from insecurity to security. Thus, when Michael Lawrence (2012: 17) -in an otherwise excellent discussion of hybrid security governance -calls for the development and implementation of non-state SSR strategies, it is not entirely clear who, precisely, is being called on to craft, oversee, or operationalize such a strategy (other than a generic reference to 'local civil society', which seems a poor match for the task). Lawrence suggests, similarly, that "a key goal for a non-state SSR strategy is to open new channels of communication and dialogue between on-the-ground security providers, citizens, civil society, international actors and the state" (2012: 22). While it's easy to support such a prescription in principle, the danger of such an approach is that, absent a coherent linkage between means and ends, it adds up to little more than an SSR version of contact theory: bring the key actors together, and assume that good things will result.
While this may represent an overly-minimalist strategy for effective engagement with the interconnected and shifting components of hybrid security governance, marrying an ongoing commitment to loose, contextspecific and flexible forms of rule-based security governance with a renewed commitment on the part of SSR interveners to contribute to what Robert Ricigliano (2003) has called 'networks of effective action' may offer a more promising approach. As Ricigliano has suggested, systems thinking emphasizes iterative approaches, learning by doing, and working with (and within) the system to identify and exploit opportunities for positive change, which may in turn lay the foundation for larger changes down the road (cited in Donais 2013). While this still requires careful, nuanced understanding and analysis of system dynamics, it does not necessarily require sophisticated central planning and coordination. What it does require, rather, are open communication networks, a shared understanding of larger goals and rules of the road, and a willingness on the part of all members to view individual efforts in the context of larger reform dynamics (2003: 446). Within this larger context, a continued commitment to supporting the evolution of rule-based systems may provide a common reference point around which the actions of interveners can converge.
A broad, long-term commitment to the rule of law and to the development of the state's regulatory capacity also, finally, has the potential to mitigate resistance on the part of state-level actors to external engagement with non-state security providers. As the evolution of the discourse on ownership has demonstrated, governments of fragile and conflict-affected states (represented by the so-called g7+) have grown increasingly sensitive to donor infringements -real or perceived -on their sovereign prerogatives. Accordingly, they have attempted in recent years to use international commitments to respecting 'national ownership' as a means of re-asserting control over post-conflict reform agendas. The sensitivities of governing elites, unsurprisingly, are particularly acute in the security governance realm, both because security provision has long been seen as the exclusive preserve of the state and because of the inherent value of security systems as assets through which to maintain control, establish legitimacy, and/or generate political, social, or economic rents (van Veen and Derks 2012: 85). In contexts where governments see non-state actors as being in competition with them for authority or legitimacy, therefore, non-state SSR strategies that are insensitive to such tensions run the risk of alienating the very constituency whose support is an essential prerequisite for enabling SSR in the first place. In the words of Erwin van Veen and Maria Derks (2012: 85), "where justice and security initiatives are perceived by elites as potentially threatening to their interests, they are almost guaranteed to fail." Somewhat paradoxically, therefore, effective engagement with non-state security providers also requires engagement strategies that both acknowledge and align with the incentive structures faced by governing elites. Beyond appeals to pragmatism -that governing elites should support whatever strategies improve security provision, particularly if they can take at least some credit -embedding non-state SSR strategies within a larger framework of state-centric rule of law development may help offset zero-sum calculations on the part of state and non-state actors alike, and provide statelevel elites with some assurances that longterm trends still privilege the state's ability to control and regulate -if not necessarily monopolize -the broader security sector. A self-conscious policy of incrementalism may be perceived as an asset rather than a liability in this context as well, especially given the delicate challenge of ensuring that efforts to bring state-level actors and actions -and not just security providers -within the purview of rule-based frameworks also unfold in ways that are not perceived as an overt threat to elite interests.

Conclusion
The ongoing search for viable second-generation approaches to security sector reform reflects a growing consensus on the prescriptive inadequacies of the monopoly model in the vast majority of reform contexts. In a variation on the theme of 'you can't get there from here', most states undertaking SSR are highly unlikely to be able to monopolize security provision within their territorial boundaries within any realistic timeframe. Yet the alternative notion of hybrid security governance -which recognizes the reality of messy, overlapping, unstable mixes of state and non-state security provision characteristic of a great many fragile and conflict-affected states -appears to suffer from the opposite problem of under-prescription. In other words, while hybridity often accurately describes ' actually-existing security governance', it is decidedly less helpful as a roadmap for charting a coherent course towards the sustainable long-term transformation of security provision in conflict-afflicted states.
With a particular emphasis on the relationship between SSR and rule of law promotion, this paper has made the case that the rule of law, loosely defined, still has a useful role to play as a source of strategic direction for SSR. Crucial to this argument is the conceptual de-linking of monopoly and accountability. While first-generation approaches emphasized accountability within the context of a state monopoly on the legitimate use of force, the argument here -consistent with insights drawn from the literature on nonstate security actors -is two-fold: not only that accountability should matter as much, if not more, in situations of hybrid security governance, but that over the longer term the rule of law may still provide the most stable foundation for ensuring accountability. While emphasizing the gradual expansion of the state's ability to bring security provision within a common (and ultimately enforceable) framework of rules, reformers also need to accept the reality of -and embrace the possibilities for working within -an indefinite interim, understanding that norms underpinning the rule of law evolve slowly and recognizing that relationships between providers and consumers of security will continue for the foreseeable future to be characterized by varied configurations of accountability and legitimacy.
Re-thinking SSR along these lines also necessarily involves re-thinking how external interveners relate to both the security systems and the security actors within reform contexts. In the first place, as Lisa Denney has noted, dealing with non-state security providers is "uncomfortable territory for organizations committed to human rights and good governance principles." Risk-aversion and a distaste for dealing with actors that might otherwise be considered unsavoury represents, therefore, a crucial first obstacle to be overcome in order to create opportunities both for understanding non-state security providers and for beginning to develop "a spectrum of unique partnerships and associations" between state and non-state systems (Scheye and Baker 2007: 525). Along the same lines, external interveners need to increasingly think of themselves as facilitators rather than engineers, with the goal helping to put in place the necessary processes, relationships and dynamics that will enable complex security systems to evolve along constructive channels long after outsiders have gone home. Indeed, Erwin van Veen and Maria Derks (2012: 93) have explicitly called for the donor community to adopt "a process approach to programming," which combines, among other elements, a commitment to short-term results (specifically, supporting existing arrangements that 'work' in a given context), flexible results frameworks supported by sophisticated monitoring and evaluation tools and deeper understandings of the incentive structures facing key actors, and mutual long-term commitments (to be realized over a timespan of decades).
While SSR continues to be, at its core, about the regulation, management and control of coercive power, increasingly the focus of outside intervention needs to shift away from the daunting (and perhaps unachievable) challenge of re-distributing power, towards a project of gradually bringing existing power relations within a broad and predictable regulatory framework. The objective, ultimately, should be to connect short-term initiatives -particularly those that facilitate constructive engagement across the different categories of security actors that constitute hybrid security orders -with a longer-term strategy for systemic change based on the evolution of existing arrangements rather than on the imposition of external ones.