Whose security? Whose justice? Customary authorities in security and justice interventions in the Horn of Africa

ABSTRACT

An evolving body of literature calls for politically realistic security and justice interventions that reflect the plurality of norms, power and legitimacy in the security and justice arena. In this arena, customary authorities may provide more security and justice than the state, especially in Sub-Saharan Africa. They matter for peace, security, and development, but engaging them requires donors to navigate hidden power structures and illiberal traditions and ideas. In response to this complexity, this article proposes an approach for working with customary authorities – supporting fluid networks of customary, community and state agents and the political bargaining between them. Looking at a specifically conflict-affected area of the Horn, this paper provides an important survey of what we know, and what we don’t know and embraces customary authorities’ agency in change.

KEYWORDS:

• Legal pluralism
• customary justice
• security and justice
• Horn of Africa
• hybrid justice

Introduction

This article takes an important step towards better engaging customary authorities¹ which are often legitimate security and justice² providers, but have historically been poorly understood.³ More recent developments in hybrid security have started to analyse flexible security arrangements around activities such as natural resource extraction.⁴ This article contributes to the evolving literature by proposing an approach for international organisations working with customary authorities in security and justice interventions. The article transcends the narrow position that customary authorities are spoilers and positions them as agents for change acknowledging their position as political actors in their own right. Importantly, it provides a much-needed alternative to the simplistic state/non-state dichotomy, and the constraining perception that liberal and local objectives are mutually exclusive.

Many Western donors have concerns around human rights when dealing with customary authorities in security and justice (S&J). However, ignoring or excluding these institutions may lead to the incidental exclusion of significant elements of the population using these systems. State/non-state interactions provide a way into engaging with and transforming politics, power structures and social norms. We challenge the common problematisation of political will by international actors, recognising that political bargaining is how change happens and ‘the local’ is a politically contested space in its own right.⁵

Working with customary authorities is complicated, takes time and relies on significant trust-building.⁶ Local civil society organisations with a long-term community presence frequently enjoy trusted relationships and are essential partners for the international community. For donors to meaningfully capitalise on these partnerships and avoid harm, interventions must allow local partners space and time to develop healthy relationships and adapt to local actors and sensibilities.⁷ Success requires devolving decision-making and resources to local community actors⁸ and investing in high quality evidence on engagement with customary authorities. It assumes greater donor openness to governance and actors who do not conform to liberal ideals and a recalibration of risk versus return.

This article builds on previous research on security and justice programming that is broader in scope and brings in wider references on different regions, including significant resources looking at customary authorities.⁹ This paper addresses a gap identified within that earlier work, specifically a paucity of information on processes of engaging with customary authorities.¹⁰ The study reviewed academic research, grey literature, programme documentation and media that concern security and justice reforms involving customary authorities.¹¹ The Horn of Africa provided an apt geographical focus given the prominence of customary security and justice systems and the contested nature of authority between state and non-state actors.¹² In Ethiopia, customary authorities are the most utilised security and justice providers in rural areas.¹³ Customary systems are core to Somaliland’s statebuilding project with clan leaders resolving 80–90 per cent of disputes, and governance and politics organised along clan lines.¹⁴

We interrogate the literature for the incentivising factors and enabling processes surrounding customary authorities’ and community members’¹⁵ engagement in security and justice reforms. This acknowledges criticism that ‘what works’ syntheses are too prosaic and miss important questions including what conditions allow interventions to gain traction in different contexts.¹⁶ The paper is primarily a policy-oriented analysis of the role of customary authorities in security and justice provision in the Horn of Africa, and the potential for promoting reforms based on a synthesis of existing literature and reports on security and justice in the Horn of Africa. We build towards a proposal that donors and NGOs working in the Horn might promote reforms of customary authorities by facilitating negotiations between different community-based and state actors, and encouraging bargaining, without predetermining the end point of these reforms. It also provides suggestions of where future research could be usefully developed. The article begins by analysing why customary authorities have engaged in security and justice reforms in the Horn of Africa, outlines several challenges and dilemmas and then goes on to propose an approach to engaging customary authorities in externally supported security and justice interventions.

What does the literature say about customary authorities and international reform?

In efforts to foster peace, security and development, donor-funded interventions often support security and justice reform in conflict-affected places. Although security and justice can be considered sectors, in practice they are really fluid. Security and justice comprise complex networks of state and non-state actors with distinct beliefs and preferences, and which continually exercise different forms of power at the individual, group and national levels.¹⁷ Customary authorities often uphold traditions that contradict liberal human rights frameworks and discriminate against certain groups, while enjoying greater authority and legitimacy than the state.¹⁸

Security and justice research is evolving to better consider customary authorities, but there is no policy consensus regarding effective engagement.¹⁹ Most donor-supported security and justice can be characterised as liberal and Weberian, emphasising the state’s responsibility.²⁰ These interventions aim to bring the state security and justice apparatus under the structures of democratic governance and support the state’s monopoly on the use of violence.²¹

Despite its popularity among liberal donors, several scholars criticised this approach for being technocratic, unsustainable and lacking in local ownership.²² Critics argue that citizens in developing countries often perceive the state as ineffective or illegitimate and rural communities may receive no security and justice from the state at all, turning instead to customary authorities.²³

Recognising that authority and power are contested in most post-conflict and fragile environments, policy reform and approaches to security and justice evolved to consider the interaction between state and non-state actors. However, processes and interaction between state and non-state agents remain relatively unexplored within much of the literature.²⁴ In the early 2000s donor-funded security and justice interventions began to incorporate non-state actors in the reform process, referring to a ‘comprehensive’, ‘holistic’ or ‘multi-layered’ approach.²⁵ This mirrors a shift in the wider development literature towards ‘good enough governance’²⁶ and ‘working with the grain’.²⁷ Whilst these concepts are nuanced, they point to the same thing: rooting interventions in local realities and customs, rather than imposing unfeasible western institutional templates.

Policy interest in non-state actors gradually translated into programming; a 2018 study highlights a significant increase in the evidence on ‘non-state S&J influencing reform and national level processes’.²⁸ However, this ‘holistic’ approach has been problematic to implement. Andersen (2012) even calls this a ‘crisis’, with many scholars agreeing that the approach is too broad, incorporating many actors and institutions, with little guidance on prioritisation or sequencing. An emphasis on holistic approaches, with those who do not conform labelled as spoilers, was reinforced by the OECD DAC’s 2007 Handbook on SSR, which highlights the state’s ‘irreducible’ role in security sector reform relegating local ownership by applying western ideas to local problems of self-governance.²⁹

Richmond cautions against denouncing customary authorities over poor human rights compliance, but Andersen and Schroeder and Chappuis both warn against romanticising non-state actors.³⁰ Albrecht and Kyed point out that state and customary systems are often built on the same social norms, meaning that criticisms around human rights compliance are unlikely to apply to one without the other, challenging an assumption that change is linear and local ownership rests with actors who model Weberian liberal ideas.³¹ This has led to alternative ways of conceptualising the politics of interventions through non-linear approaches transcending the conventional boundaries of state and non-state and embracing the resulting uncertainty in the interest of genuine local ownership.³² In practice this requires international actors to engage proactively with the complexity and fluidity of the S&J arena: reorienting interventions towards processes not end states, systems not sectors, and plurality not homogeneity of ideas, legitimacy and ownership.³³

Research approach

An initial systematic literature review was carried out asking why customary authorities and community members work together to implement security and justice reforms³⁴? In addition, the review also asked how the needs and perceptions of under-served³⁵ groups were incorporated? The study only investigated reforms that explicitly involve community members following Harper’s suggestion that evidence suggests security and justice users’ protection will be enhanced if they are included in reform negotiations or decision-making.³⁶ The search generated 12,882 hits and 28 papers were included: 15 observational case studies, 10 UN- or NGO-produced papers, and three media articles building on Jackson and Bakrania’s security and justice evidence mapping approach, but the study only searched papers in English.³⁷ These papers were then screened for methodological quality and backwards reference snowballing was used to included papers, which ensured all relevant data from the literature was captured.³⁸

Framework synthesis methods were applied to code and analyse the data, drawing on the World Development Report Governance and the Law conceptual framework.³⁹ This conceptual framework highlights that the way decisions are taken on development policies, and by whom, determining whether development outcomes can be realised. Semi-structured interviews with authors and practitioners involved in high quality papers were used to enhance data. These highlighted relevant unpublished information and reduced publication bias.⁴⁰ However, it is likely that data exists that remains not captured and derives from informal decisions, undocumented data and classified data held by governments.

The search identified few papers on Eritrea and Djibouti, and none fitted the inclusion criteria due to lack of detail or quality. Therefore, this analysis only discusses Somalia, Somaliland and Ethiopia. All 28 studied papers focus on the district or village level, reflecting the prominence of customary mechanisms in community-level security and justice provision. Whilst the systematic review findings were nuanced by context, the study identified common trends from contexts with different political economies, cultures and statebuilding histories that could be further explored.

Common preferences

Across Ethiopia, Somalia and Somaliland, customary authorities collaborated on security and justice reforms because they wanted to improve security and development following legacies of conflict, but very few believed that the current system could deliver this vision, thus providing a strong incentive for change.⁴¹ In several Ethiopian districts, customary authorities collaborated on government-led community policing reforms, despite not having requested them.⁴² Initially, some clan leaders resisted this top-down initiative because it challenged their authority but ultimately accepted the reforms because they wanted to support Ethiopia’s development, ‘protect the community’⁴³ and ‘reduce crime rates’.⁴⁴

Denney and Kassaye (2013) find that the police compellingly aggregated customary authorities’ and community members’ nuanced security interests, creating a broad vision that positioned community policing as necessary for Ethiopia’s development. That communities in Dire Dawa district had lived through years of violence likely further strengthened their willingness to reform coinciding with a shared belief that the current system was ‘rotten’.⁴⁵

Somalia’s security and justice arena involves a marked blurring of formality and informality. Somalia uses clan customary law (Xeer), Islamic Sharia law and secular law – with Xeer playing a dominant role in addressing conflict and ensuring citizens’ safety since the state’s collapse.⁴⁶ Somalia’s traditional elders provide clan members with protection and enable them to negotiate power and resources.⁴⁷ However, the clan-based nature of Xeer can pose problems for minority clans, who typically require sponsorship from a majority group to access justice.⁴⁸ As with Somaliland, this politicises the clan leader’s role, putting them at the heart of ongoing bargaining for power.⁴⁹

Mayors and customary elders in Somaliland’s Togdheer region asked the Danish Refugee Council (DRC) and Oxfam Novib for support to Xeer, which they agreed was not coping with shifting patterns of violence.⁵⁰ This catalysed over a decade of support by UNDP and multiple NGOs to customary systems in Somalia and Somaliland. It reflects wider evidence that external security and justice interventions should be timed when communities express the desire for external or state-led support.⁵¹ The bigger picture in Togdheer was an absence of functioning state structures to resolve conflict when required.

Notably, customary authorities in Somalia and Somaliland were motivated by peace rather than improving access to justice, despite the latter being a donor and NGO programme objective. DRC and Oxfam Novib’s support to Togdheer elders’ security goals laid the groundwork for further NGO-supported projects that take a more liberal standpoint. Saferworld, Soyden and Pact are challenging discriminatory justice practices and involving women, youth, ethnic minorities and internally displaced persons in justice processes. Clearly customary and liberal ideals are not inherently incompatible providing external development actors work through local actors’ objectives without imposing external liberal ideas.

Importantly, conflict levels affect reform possibilities. Gundel highlights how large changes to Xeer are traditionally only possible during peacetime.⁵² Similarly, communities in Ethiopia’s Amhara more readily accepted the government’s community policing initiative because tensions between central government and communities had reduced. These tensions had the reverse effect in Dire Dawa; with violence between central government and communities raising interest in some forms of community policing. The specific initiative promised to ease tensions by appointing police officers from the community, training them and then posting them back into that community.

Integrating customary systems

Customary authorities collaborate on security and justice reforms in Ethiopia, Somalia and Somaliland following histories of contestation between the state and clan elders over policing and justice.⁵³ Ongoing contestation has created incentives for customary elites to engage in reform. In Somalia and Somaliland, interviewees highlighted how Somalia’s weak state and fragile security situation motivated customary authorities to drive reform from the bottom up. It was further noted that elders’ have an interest in engaging communities to bolster authority. Post-Conflict reforms both engaged the state and asserted the authority of customary systems, aiming to prevent conflict in situations where citizen-state relationships are fluid and contested rather than a state/non-state dichotomy.⁵⁴

Customary authorities were also motivated by a desire to keep security and justice local. Cooperation on reformed Councils of Elders⁵⁵ and the Community Police Dialogue Cooperation⁵⁶ enabled them to promote customary dispute resolution and maintain ownership of local security issues, rather than relying on the police.⁵⁷ In Somalia’s Southwest State, this culminated in police recognition of the value of customary mechanisms in supporting policing.⁵⁸

In Ethiopia, the government formally recognised customary authorities in community policing, gave them authority to resolve local disputes, and awarded the police the mandate to use customary crime responses.⁵⁹ In turn, customary authorities mostly collaborated in the process.⁶⁰ Customary authorities viewed this as an opportunity to amplify their power and legitimacy, while keeping policing local. Such actions are frequently motivated by supplementing policing with customary mechanisms, including intelligence, and may also represent a trade-off between enhanced police control and less crime through enhanced policing coalitions.⁶¹

Social norms

Social norms around youth and women’s roles, and cultural expectations of what security and justice should look like, have both enabled and impeded reform. This reflects how security and justice interventions are complex and so different papers covering different geographical areas or populations were unlikely to produce homogenous results.⁶² However, it also suggests that social norms can be used strategically to produce different outcomes. Illiberal social norms may not simply make reforms too risky or difficult but provide the basis for dialogue around reform itself.

In Somaliland, clan elders influence security and justice governance at different levels, including through neighbourhood committees.⁶³ This reflects the composite nature of the Somaliland state, built on negotiations between clan elders, businessmen and citizens.⁶⁴ In Ethiopia, community policing gained traction when it met customary authorities’ expectations of how security and justice are provided. According to a member of the Regional Police in Amhara, some customary authorities expressed ‘nonchalance’ towards community policing because it appeared to be a continuation of how things had always been done. Customary authorities’ willingness to reform should be considered alongside their expectations of security and justice provision.⁶⁵ Social norms can also play a disruptive role in reforms. Culturally, Somalian nomadic communities maintain strict lines on who belongs in community structures to protect their livelihoods. This affected nomadic customary authorities’ willingness to allow new participants into reform structures. Conservative social norms around women and youth also impede meaningful participation in reformed security and justice governance in some districts – with elders not taking women and youth perspectives seriously.⁶⁶

Elders have used cases strategically to prevent change around women’s rights. In Somalia’s Warsheikh town, clan elders dismissed a woman’s land dispute, arguing that it was too complicated.⁶⁷ By contrast, in Somalia some elders disrupted new formal court processes on rape cases.⁶⁸ In Ethiopia some elders refused to hand over female genital mutilation and domestic violence cases to the formal system.⁶⁹ Customary authorities may use beliefs and traditional values within individual cases to resist alternative values and threats to those systems. This may manifest itself through withdrawing from some formal systems, but equally it could also result in some cases never entering the system as a result of societal pressure.

Evidence from Somalia and Somaliland illustrates that NGOs can challenge discriminatory social norms through a long-term, sensitive approach.⁷⁰ NGOs have overcome elders’ resistance to supporting women’s rights by taking time to build trust⁷¹ and assisting both men and women on gender issues, for example.⁷² This indirect approach culminated in the acceptance of women in leadership positions in the Community Action Forums and District Policing Committees.⁷³ Bringing communities into the design of these structures from the outset enabled reform negotiations to challenge conservative and exclusionary norms, leading to the participation of minority clans, women, youth and internally displaced persons in reform governance. Contestation within the reform process can also challenge participants’ normative beliefs, although actual incorporation is under-researched.

Rules

The discussion on social norms is mirrored regarding the role of formal and informal rules in shaping customary authorities’ reform motivations. Governments, civil society groups and international NGOs recognise this; they have intelligently deployed rules as an engagement strategy to secure customary authorities’ commitment and change their behaviours. Rules are critical in incentivising actors and holding them to account, while coordinating expectations around how to behave.⁷⁴

Jackson et al. and Belay find that the police in Ethiopia successfully used the constitution – which emphasises cooperation between the police and communities – as a strategy to secure chiefs’ and community members’ commitment to community policing.⁷⁵ Whilst the constitution was not the main reason why customary authorities engaged on community policing, the way the police deployed it as an engagement strategy made it an important factor in their cooperation. Notably, not only did the constitution strengthen commitment to reform, but the reforms also served to assert the constitution. The fact that the constitution legally recognises customary courts may have increased its legitimacy with customary authorities and allowed use of the constitution to change legal practice and norms within that system.⁷⁶ In Dire Dawa, the police successfully appealed to the constitution – which was widely bought into by society and included protections for women and girls – to persuade chiefs to part with discriminatory sexual and gender-based violence and child early and forced marriage practices.

The police framed the constitution as a way for customary authorities to wield power and protect their customs, rather than losing them. However, it should be noted that increasing counterterrorism regulations gave Amhara customary authorities little choice but to engage with community policing and community policing itself should not be taken as being uniformly positive.\For example, the Dire Dawa evidence was partly the result of a specific set of circumstances and much research in this area supports a rather mixed picture.

The Somaliland context supports this. Here, DRC facilitated elders to develop declarations, which contained changes to laws under Xeer that were deemed acceptable to the elders.⁷⁷ Whilst research suggests that the declarations did not improve access to justice for under-served groups, they provided a reference point for subsequent contestation between elders and community members on discriminatory practices.⁷⁸

Ethiopia and Somaliland are interesting examples because of the way these rules were produced (bottom-up versus top-down) and at which level of governance (national versus community). The implicit similarity, however, is that both the constitution and elders’ declaration were legitimate in customary authorities’ eyes. This suggests that legitimate rules offer a powerful tool in the bargaining process around reforms, whether these are developed at the national or community-level. It is conceivable that rules which customary authorities perceived as illegitimate would not be effective. However, whilst this requires further investigation, Ethiopia, for example, had a constitution for a number of years but legitimacy was clearly an issue both in terms of inclusion and exclusion, but also in the way it was enforced. It seems that a more inclusive constitution coupled with a difference enforcement mechanism has gained results. The absence of a legal framework to underpin reforms, by contrast, can hamper progress. Saferworld and Soyden both note the difficult implementation of community security and policing committees in Somalia because these structures have no legal framework.⁷⁹

Dilemmas for international engagement with customary authorities

The previous section shed light on customary authorities’ motivations and incentives for engaging in S&J reform across Ethiopia, Somalia and Somaliland. The analysis supports the position that reform is an ongoing political process and not an end state. This process involves the continual interaction, negotiation and contestation of power and ideas, implying that better understanding of political bargaining may be one key that unlocks change. Although these bargaining processes have secured customary authorities’ voluntary commitment to reforms, and in places reshaped their position on discriminatory practices, their fluidity and complexity create dilemmas for externally supported and highly structured interventions.

The literature on the Horn of Africa unanimously highlights a trend towards reforming security and justice by changing the structures and participation of security and justice governance. The systematic review’s most common finding was that donor-funded and government-led reforms have brought community members, including under-served groups, into security and justice governance to work alongside customary authorities and, sometimes, the police. This reflects the benefits of citizen participation to improve contestation in decision-making, and drive and support change.⁸⁰ Importantly, it supports the position that technically-sound institutional models are insufficient to catalyse development outcomes; what matter are the power structures in the bargaining arena where policy decisions are taken.⁸¹ External development actors too are part of these power structures, and interaction with the bargaining system is an unavoidable part of any development intervention.⁸²

In Ethiopia’s Dire Dawa and Amhara, community policing entails training community members and putting them back into the community as police officers, thereby merging the distinction between the police and community. In Dire Dawa, this model both improved communities’ trust in policing, and community police officers’ knowledge of community members’ needs. This led to better management of sexual and gender-based violence and child, early and force marriage cases for women and girls.⁸³ However, in Amhara community police officers’ consideration of under-served groups depended on their personality and interpretation of their objectives. Community members’ participation in policing does not automatically translate into improved consideration for under-served groups, rather, an approach is needed to address the hidden power structures that discriminate against certain groups in the first place.

In Somalia and Somaliland donor-funded security and justice reforms have led to a proliferation of district-level committees, forums, boards and units. New structures generally bring local elites, such as customary authorities and businessmen, alongside community members including women, youth, human rights activists, internally displaced persons and clan minorities.⁸⁴ On a voluntary basis, they diagnose problems around safety, security and crime, and develop and implement responses. Wiuff Moe and Vargas describe how these structures have enabled community members to ‘become actively part of the process of contestation and negotiation that shape and reshape this architecture’.⁸⁵ Mydlak, however, finds that participation does not automatically translate into influence.⁸⁶

Our research shows that there is a lack of research on how increased participation by under-served groups in security and justice governance has improved outcomes for these groups – or whether the power structures remain the same behind a structural reorganisation. This is exacerbated by a vague approach to incorporating stakeholders in reform. Bennett supports the inclusion of ‘as many local stakeholders as possible’ – although the evidence base for this is unclear – but the World Bank argues for ‘actors who have reason to be included’, but this is open to interpretation.⁸⁷ Andersen argues that external actors should focus on supporting stable power balances, which could come at the expense of inclusion.⁸⁸ Notwithstanding data quality, the literature outlines three factors which appear to matter for whether reform structures challenge discriminatory power structures and better incorporate under-served groups’ security and justice needs: perceptions of performance; perceptions of effectiveness; and the selection process.

There is some evidence that decision-makers may listen to under-served groups whose contribution to security and justice governance is perceived as credible. Elders in Somalia and Somaliland applauded women and youth for their contribution to reform structures when they demonstrated unique skills, knowledge and networks.⁸⁹ This outweighed the reasons why they were previously excluded.⁹⁰ It also created a virtuous cycle whereby women and youth gained social capital through their effective contributions to the elders’ council, which enhanced the council’s traction with the community, therein further enhancing elders’ opinions of these groups.⁹¹ This supports Halloran’s argument for supporting the capacity of local actors so they can actively influence change.⁹²

There is also evidence that the credibility of reform structures also matters for their potential to better incorporate under-served groups’ needs and perspectives. Reform structures in Somalia and Somaliland, which involved under-served groups, were embraced by customary authorities and other community members because they improved security and community cohesion.⁹³ However, further research may measure whether these initiatives genuinely translated into better outcomes for under-served groups. Leite highlights how putting communities in the lead to design and implement community security initiatives in Somalia and Somaliland underpinned these structures’ traction with local communities.⁹⁴

Thirdly, NGOs have developed a range of selection processes to encourage fairness and inclusion in reform governance in Somalia and Somaliland.⁹⁵ However, several selection processes have been captured or dominated by majority groups, local authorities, the educated class,⁹⁶ religious leaders and business owners⁹⁷ and political parties.⁹⁸ This reiterates the importance of considering how real and hidden power are distributed and wielded throughout reforms, how that may be manifest in the local outcomes of these processes and how local coalitions of citizens may be mobilised outside of these core groups.

A way forward: working with customary authorities in S&J reforms

The evidence from the Horn of Africa shows a very mixed picture. Creating structures of inclusion may not result in actual participation in contexts where there is significant structural imbalance of power at the local level. MacGinty and Richmond highlighted a number of issues some time ago with regard to local challenges lack of structural power at the grassroots, the issue of violence, and the reproduction of colonial tendencies by interveners. The key for those who seek to engage in political bargaining with different types of customary authorities needs to acknowledge the differential of bargaining power between the relevant actors. In particular the core issue is how to support those who may lack power in a bargaining situation and those who are seeking justice.⁹⁹

Bargaining processes in this context draw on both Levy’s working with the grain approach and also Leftwich’s emphasis on human agency for reform.¹⁰⁰ This gives an emphasis to fluidity in relationships rather than a static structural approach. The question for external actors, therefore, becomes not which institutions to support, but how they might facilitate networks of customary, state and community actors to build relationships and negotiate reforms including how to balance power differentials in bargaining positions.

Working with customary authorities in this way starts with the factors that influence them. Our evidence reveals that legacies or new patterns of violence, and contestation over authority, can motivate customary authorities to request support and broaden their coalitions. This outweighs their loyalty to old ways of doing things. It also challenges the position that customary authorities’ beliefs are fixed and inflexible. External actors have a role to opportunistically support windows in customary authorities’ reform horizons.

Legacies or new patterns of violence provide external actors an opportunity to coordinate customary authorities around a desire for improved security, and a belief that the current system is inadequate. Although customary, state and community actors’ nuanced security and justice interests are unavoidably pluralistic, external actors should consider how these ideas might aggregate into a broad but common vision for peace. When the context offers sufficient stability, this vision represents an anchor in reform negotiations. As such, the coordination of relevant actors’ ideas and expectations, and a shared vision for success, can work favourably to drive change.¹⁰¹

Central to this approach is engagement with local interests, rather than imposing western ideas of policy success. The evidence from Somalia and Somaliland confirms the feasibility of pursuing liberal goals (such as access to justice) by working through local interests (such as peace). In these contexts, interventions that donors and NGOs call ‘alternative dispute resolution’ and ‘access to justice’ are, from the perspective of customary authorities, about conflict management – resolving disputes before they escalate. By considering what purpose interventions might serve for customary and international actors, external actors can move beyond the constraining perception that liberal and local objectives are mutually exclusive.

Contestation between customary and state systems represents another window of opportunity to work with customary authorities. In this instance, reform serves a function for customary authorities: defending or asserting their authority and keeping security and justice local. Here, the question is whether and how negotiation over power and authority might affect customary authorities’ calculations in relation to the opportunity cost of not reforming. This requires understanding the shifting nature of real and perceived power and authority – including relationships between the local, regional and federal levels.

The relative strength of the state security and justice system compared to customary systems matters here. In contexts with particularly weak or failed states, external actors might support customary systems to fill gaps in state security and justice provision, while using this engagement to foster the relationships and systems needed for scaled-up customary and state security and justice linkages over the long-term. This acknowledges the competing sources of security and justice beyond the state and capitalises on the strengths of customary systems in reforms.¹⁰² Further research could usefully interrogate how customary security and justice systems might engage the state in contexts like Somalia where the state is particularly weak.

When support is requested, external actors’ role lies in facilitating customary authorities and community members to negotiate solutions to their security and justice problems. This does not mean designing model institutional end states or simply consulting civil society at intermittent points.¹⁰³ Rather, it means reorienting intervention objectives to support bargaining processes between customary, state and community actors – without a predetermined view of what reform institutions look like.¹⁰⁴ This could entail coordinating local actors with similar security concerns, building the capacity of under-served groups to participate in negotiations, and creating a platform for customary, state and community actors to build trust and collaborate.¹⁰⁵ At the same time, this must recognise that ‘the local’ is an area of political contestation in its own right and different actors within that space have different levels of power. Local security and justice spaces are not level playing fields and careful consideration needs to be given, for example, to those who lack power of voice, e.g. women and youth.

This approach is supported by the literature on ‘good enough governance’,¹⁰⁶ ‘thinking and working politically’¹⁰⁷ and ‘problem-drive iterative adaptation’.¹⁰⁸ Several scholars advocate inclusive and accountable processes that bring communities into the design, implementation and monitoring of security and justice interventions, particularly those under-served by the status quo.¹⁰⁹ There is some empirical evidence on the effectiveness of facilitating local actors to define solutions to their development problems.¹¹⁰ However, systematic research is required to test the causal links to development outcomes.¹¹¹

The main limitation of this approach concerns the hidden power structures in customary security and justice.¹¹² There is the risk that customary authorities capture reform bargains to entrench a social hierarchy that protects their position and beliefs. Whilst new security and justice governance structures that include community members can be valuable, they are not an automatic solution to transforming discriminatory power structures and empowering under-served groups. They can even be counter-productive if the structures or members are seen as ineffective or illegitimate. At a minimum, external actors should consider how reform structures and their members, particularly representatives of under-served groups, might build legitimacy, credibility and influence. It may be that a ‘good enough’ approach seeks to include under-served groups who might gain backing in the community, have the potential to provide a credible contribution, and who are open to new approaches – recognising that it is probably unrealistic to include all under-served groups at once.

In a promising step forward, the evidence from the Horn of Africa reveals how the strategic and intelligent framing of social norms and rules can re-shape discriminatory attitudes and practices. This may go some way to addressing the predicament of hidden power. The Dire Dawa case shows how conservative attitudes can be re-shaped with astute framing; the police persuaded customary authorities to tackle sexual and gender-based violence and child, early and forced marriage by linking this necessity to the constitution, while appealing to the constitution as a way for them to wield power.¹¹³ The way rules are used as an engagement strategy potentially matters more than their content, providing these rules are perceived as legitimate.

Reform negotiations themselves also offer an avenue for contestation around exclusionary social norms and poor human rights compliance. The Community Action Forums and District Peace Committees in Somalia and Somaliland illustrate this; bringing under-served groups into the design of security and justice reforms helped re-shape discriminatory attitudes by legitimising a role for new groups in decision making. Aligning reform proposals with expectations of how security and justice should be provided (as in Ethiopia’s Amhara) or approaching conversations around discriminatory norms indirectly (as Saferworld’s gender framing in Somalia) can also generate customary actors’ voluntary commitment to a new way of doing things. Herein, social norms are not a barrier to political will; they are a strategy for change.

Overall, the quality of this approach is limited by the strength of the evidence base. If the international community is to improve the evidence base on how change in the security and justice arena really happens, there is a role for practitioners to better document the everyday decision-making and bargains around reforms. Longitudinal research could usefully examine the dynamics of long-term security and justice reform processes, and the contestation and bargaining underpinning this.

A clear priority is research investigating what outcomes security and justice interventions achieve for under-served groups and why. This may require more innovative methodological approaches for studying security and justice outcomes that are hard to measure, particularly over time. It would be beneficial to build experimental evaluations into the design of these projects from the outset. This supports Denney and Valters’ call for research into how and whether security and justice programmes are supporting medium- and long-term outcomes like accountability and stability.¹¹⁴ Such research could provide a starting point to generating quality evidence on how and whether reforms are re-shaping discriminatory practices and engaging with hidden power structures. It could also map interactions between state and non-state actors and the results of this over time.

Conclusion

The international community faces a dilemma in dealing with non-state actors like customary authorities. Overwhelmingly associated with traditional and illiberal social norms, direct interaction with customary authorities has historically been regarded with some trepidation. These social norms and human rights are likely to remain concerns of western donors. The dilemma is exacerbated by a lack of understanding, reflected in the research evidence, of how state and non-state actors interact and influence each other over time. This perpetuates a false dichotomy of state/non-state and the constraining perception that liberal and local are mutually exclusive. It also risks attributing forms of liberal reformism to states that may themselves be problematic. To get beyond this dilemma, future interventions must tune into the fluid interactions between customary, community and state agents. To do this well, international agents must think differently about change, including when designing, developing and measuring their interventions. Although the evidence favours the possibilities of working with customary authorities, donors should not underestimate the length of engagement and trust-building to do this astutely.¹¹⁵ Civil society organisations with a long-term presence and access in rural communities are best placed to understand the nuanced contextual dynamics and build trust. To benefit from these partnerships, donor processes must allow local partners the space, time and authority to build relationships, deliberately experiment, and adapt approaches iteratively based on new learning and evidence.¹¹⁶

Fundamentally, security and justice interventions are usually implemented in places that fall short of a liberal western ideal. Characterising customary authorities as too risky, difficult or separate from the state misses their importance in the webs of interactions at the local level. In reality, the security and justice arenas are not divided by the state/non-state boundaries that the international community assigns to it. This complexity becomes increasingly important as fragmented conflict landscapes leave western and local actors to contend with competing international models of conflict management.¹¹⁷ By reducing customary authorities to spoilers, we leave incomplete any coalition for change.

The paper has analysed evidence generated over a considerable time-period, along with approaches to working with customary authorities including ‘thinking politically’, ‘good enough governance’ and ‘working with the grain’. Far from taking these approaches uncritically, we surmise that none of them have worked adequately because they are based on either fundamental lack of local knowledge, or on a set of misunderstood assumptions in application. Furthermore, there has been a tendency for those intervening to assume that ‘politics’ is what prevents reforms happening rather than local systems and institutions having their own politics. An important element missing in all of these approaches is a comprehensive understanding of how the intervention relates to domestic or local politics and how they change each other to produce unintended consequences or variations from underlying aims. In our view what is required to validate the approaches taken in the longer-term, is a means of empirically demonstrating how sustainable interventions with customary authorities have been over time.

The research suggests that what is required to validate the approaches taken in the longer-term, is a means of empirically demonstrating how sustainable interventions with customary authorities have evolved over time. In purely pragmatic terms part of this might be explained by the expense of long-term research, but also the advent of new techniques and supporting historical case studies could be used to create meaningful narratives of change. The contemporary approaches most frequently used are valuable but could be strengthened by comparative work and also mapping changes in behaviour through experimental methods. None of these approaches is adequate in isolation, however, since the hidden politics of how change comes about are usually hidden for a reason and determined by mapping outcomes over time.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Paul Jackson

Paul Jackson is Professor of African Politics at the University of Birmingham. He is currewntly Director of Research for the Social Sciences and has previously been Head of the School of Government and the International Development Department.

Veronica Stratford-Tuke

Veronica Stratford-Tuke works for the UK Government and is currently Head of the International Centre of Illicit Finance in the UK Foreign, Commonwelath and Development Office

Notes

1. In this article, customary authorities are one group of non-state actors and refer to elders, chiefs and tribal or clan leaders. They exclude religious leaders and members of clan militia or other non-state armed groups. The literature uses several terms interchangeably including community, community leadership, elders, traditional and customary providers. This exacerbates ambiguity on who really owns change processes, and how local ownership in security and justice reform can be enacted in practice.

2. This article takes an expansive definition of security and justice to mean ‘values and goals (e.g. freedom, fairness, personal safety) as well as the various institutions established to deliver them (e.g. defence forces, police, courts)’ (DFID, 2007, 10). This includes both formal (e.g. the police) and informal village courts, community security groups) institutions.

3. Jackson and Bakrania, Is the Future of SSR non-linear? Verweijenaand and Van Bockhaven, Revisiting colonial legacies.

4. Enns et al., Security for whom? Aning and Axelrod, Hybrid Security Provision.

5. Hudson et al., Inside the Black Box of Political Will; Jackson and Bakrania, Is the Future of SSR non-linear?; Andersen, Security Sector Reform; Leftwich, Politics in command; Green, How Change Happens; Acemoglu and Robinson, Why Nations Fail.

6. Mydlak, Engagement of Somali Customary Institutions.

7. Laws et al., LearnAdapt.

8. Midgley et al., Beyond Box-Ticking.

9. For example, Jackson and Bakrania, Is the Future of SSR non-linear?

10. Denney and Valters, Evidence Synthesis.

11. This study defines the Horn of Africa as Ethiopia, Somalia, Somaliland, Eritrea and Djibouti.

12. Bereketeab, State Legitimacy; Hills, Security Sector or Security Arena?

13. Jackson and Bakrania, Is the Future of SSR non-linear? Baker, Where Formal and Informal Justice Meet; Denney and Kassaye, Securing Communities for Development.

14. Abdi Omer and Jibril, Hybrid Security Governance; Wiuff Moe and Vargas Simojoki, Custom, Contestation and Co-operation; Dijkema et al., Rethinking the Foundations of the State; EAJ, Access to Justice Assessment Tool.

15. In this article, community members refer to everyday citizens and civil society organisations.

16. Mallett et al., The Benefits and Challenges of Using Systematic Reviews.

17. Luckham and Kirk, Understanding Security, 10; Lund, Twilight Institutions, 685; Mac Ginty, Indigenous Peace-Making, 140.

18. Denney, Liberal Chiefs or Illiberal Development? Jackson and Bakrania, Is the Future of SSR non-linear?

19. Aning and Axelrod, Hybrid Security Provision.

20. Jackson and Bakrania, Is the Future of SSR non-linear? Jackson, Decentralised authorities; Andersen, The Liberal Dilemmas of a People-Centred Approach.

21. Jackson and Bakrania, Is the Future of SSR non-linear? Sedra, Adapting Security Sector Reform.

22. Jackson and Bakrania, Is the Future of SSR non-linear? Sedra, Adapting Security Sector Reform; Homel and Masson, Partnerships for Human Security; Enfield, Lessons of Working with Informal Security Actors; Harborne, Security and Justice Overview.

23. Jackson and Bakrania, Is the Future of SSR non-linear? Andersen, Security Sector Reform; Bierschenk, Who are the Police in Africa? Richmond, Becoming Liberal.

24. Bell, Jackson and Bakrania, Security and Justice Evidence Mapping.

25. Department For International Development, Safety, Security and Accessible Justice; Organisation for Economic Cooperation and Development, The OECD DAC Handbook.

26. Grindle, Good Enough Governance; Ramalingam et al., From Best Practice to Best Fit.

27. Levy, Working with the Grain.

28. Bell, Jackson and Bakrania, Security and Justice Evidence Mapping, 28; Sedra, Adapting Security Sector Reform.

29. Gordon, Security Sector Reform; Oosterveld and Galand, Justice Reform.

30. Richmond, Becoming Liberal; Andersen, The Liberal Dilemmas of a People-Centred Approach; Schroeder and Chappuis, New Perspectives on Security Sector Reform.

31. Albrecht and Kyed, Introduction.

32. Jackson and Bakrania, Is the Future of SSR non-linear?

33. Enfield, Lessons of Working with Informal Security Actors.

34. The research questions took a broad definition of reforms across S&J institutions and values, rather than studying one intervention type like community policing. Reforms included policies, programmes and projects led by local governments, communities, international organisations or donors. This was appropriate because the study interrogated the bargaining process between actors on reforms, not the type of interventions that emerge.

35. Under-served groups mean those who struggle to access S&J services. They include women, youth, clan minorities, the disabled and IDPs – although members of these groups can also perpetuate discrimination (Harper, Gender Equality and Social Inclusion Analysis, 3).

36. Ibid.

37. Bell, Jackson and Bakrania, Security and Justice Evidence Mapping.

38. Bakrania, Methodological Briefs on Evidence Synthesis, 14.

39. World Bank, World Development Report 2017.

40. Stewart, Changing the World One Systematic Review at a Time, 587; Popay et al., Guidance on the Conduct of Narrative Synthesis, 22.

41. Jackson and Bakrania, Is the Future of SSR non-linear? Belay, Community Policing or Policing the Community? Mydlak, Engagement of Somali Customary Institutions; Wiuff Moe and Vargas Simojoki, Custom, Contestation and Co-operation; Wiuff Moe and Vargas, Supporting Local Peace Building; Gundel, Alternative Dispute Resolution Mechanisms in Somalia; Denney and Kassaye, Securing Communities for Development; Mahmud, Writer Calls for National Dialogue; AllAfrica.com, Somali Leaders Pledge Judicial Reform.

42. Jackson and Bakrania, Is the Future of SSR non-linear? Denney and Kassaye, Securing Communities for Development.

43. Denney and Kassaye, Securing Communities for Development.

44. Jackson and Bakrania, Is the Future of SSR non-linear?

45. Ibid.

46. Beyene, The Security Sector Reform Paradox in Somalia; International Security Sector Advisory Team, Background Note; Vargas Simojoki, Enhancing Legal Empowerment; Burke, Somalia and Legal Pluralism; Leite, Reinvigoration of Somali Traditional Justice.

47. Schroeder and Chappuis, New Perspectives on Security Sector Reform; Beyene, The Security Sector Reform Paradox in Somalia.

48. Vargas Simojoki, Enhancing Legal Empowerment; Harper, Gender Equality and Social Inclusion Analysis.

49. De Waal, Somalia’s Disassembled State.

50. Wiuff Moe and Vargas Simojoki, Custom, Contestation and Co-operation; Gundel, Pathways and Institutions; Mydlak, Engagement of Somali Customary Institutions.

51. Enfield, Lessons of Working with Informal Security Actors.

52. Gundel, Pathways and Institutions.

53. Bell, Jackson and Bakrania, Security and Justice Evidence Mapping; McCullough and Saed, Gatekeepers, Elders and Accountability.

54. Andersen, The Liberal Dilemmas of a People-Centred Approach; Jackson and Bakrania, Is the Future of SSR non-linear? EAJ, Access to Justice Assessment Tool.

55. The Councils of Elders (often called the ‘Guurti’) are traditional forums for elders to mediate and resolve disputes. The Danish Demining Group supports reforms to these councils across Somaliland and Somalia. These reforms bring a wider set of community actors into the council, including women, youth and ethnic minorities, broadening participation in decision making and dispute resolution.

56. The Danish Demining Group supports the Community Police Dialogue Cooperation (CPDC) initiative in several Somalian districts. The CPDC takes some disputes off the Somali police and holds them accountable.

57. PCEF, Perceptions Assessment.

58. UNDP, Programme Quarterly Progress Report.

59. Denney and Kassaye, Securing Communities for Development; Baker, Where Formal and Informal Justice Meet; Belay, Community Policing or Policing the Community?

60. Denney and Kassaye, Securing Communities for Development; Baker, Where Formal and Informal Justice Meet.

61. Belay, Community Policing or Policing the Community? Denney and Kassaye, Securing Communities for Development.

62. Popay et al., Guidance on the Conduct of Narrative Synthesis.

63. Abdi Omer and Jibril, Hybrid Security Governance.

64. Ibid.

65. Denney and Kassaye, Securing Communities for Development.

66. Mydlak, Engagement of Somali Customary Institutions; Tait et al., Community-Police Dialogue; Saferworld, Sustainable Community Approaches; Wiuff Moe and Vargas Simojoki, Custom, Contestation and Co-operation; Gundel, Pathways and Institutions.

67. Mydlak, Engagement of Somali Customary Institutions.

68. UNDP, Programme Quarterly Progress Report.

69. Baker, Where Formal and Informal Justice Meet.

70. Harper, Gender Equality and Social Inclusion Analysis.

71. Gundel, Pathways and Institutions; Leite, Reinvigoration of Somali Traditional Justice.

72. Saferworld, Sustainable Community Approaches.

73. Ibid.

74. World Bank, World Development Report 2017.

75. Jackson and Bakrania, Is the Future of SSR non-linear? Belay, Community Policing or Policing the Community?

76. Baker, Where Formal and Informal Justice Meet.

77. Wiuff Moe and Vargas Simojoki, Custom, Contestation and Co-operation.

78. Ibid.

79. Saferworld, Sustainable Community Approaches.

80. World Bank, World Development Report 2017.

81. Ibid; Denney and Valters, Evidence Synthesis.

82. Jackson and Bakrania, Is the Future of SSR non-linear?

83. Jackson and Bakrania, Is the Future of SSR non-linear?

84. Wiuff Moe and Vargas, Supporting Local Peace Building; Saferworld, Sustainable Community Approaches; Soyden, District Peace.

85. Wiuff Moe and Vargas Supporting Local Peace Building.

86. Mydlak, Engagement of Somali Customary Institutions.

87. World Bank, World Development Report 2017.

88. Andersen, Security Sector Reform.

89. Mydlak, Engagement of Somali Customary Institutions; Wiuff Moe and Vargas Simojoki, Custom, Contestation and Co-operation.

90. Mydlak, Engagement of Somali Customary Institutions.

91. Ibid.

92. Halloran, Thinking and Working Politically.

93. Mydlak, Engagement of Somali Customary Institutions; Wiuff Moe and Vargas Simojoki, Custom, Contestation and Co-operation; Wiuff Moe and Vargas, Supporting Local Peace Building 2012; Soyden, District Peace.

94. Leite, Reinvigoration of Somali Traditional Justice.

95. Ibid.

96. Baker, Where Formal and Informal Justice Meet.

97. Gundel, Pathways and Institutions.

98. Denney and Kassaye, Securing Communities for Development.

99. Jackson, Security Sector Reform.

100. Levy, Working with the Grain.

101. World Bank, World Development Report 2017.

102. Jackson and Bakrania, Is the Future of SSR non-linear?

103. Gordon, Security Sector Reform.

104. Bell, Jackson and Bakrania, Security and Justice Evidence Mapping; Andersen, The Liberal Dilemmas.

105. Enfield, Lessons of Working with Informal Security Actors.

106. Grindle, Good Enough Governance.

107. Leftwich, Thinking and Working Politically.

108. Andrews et al., Escaping Capability Traps.

109. Jackson and Bakrania Is the Future of SSR non-linear? 2018; Albrecht and Jackson, Security Sector Transformation.

110. Booth and Unsworth, Politically Smart.

111. Dasandi et al., Thinking and Working Politically.

112. Jackson and Bakrania, Is the Future of SSR non-linear?

113. Ibid.

114. Denney and Valters’ Evidence Synthesis.

115. Mydlak, Engagement of Somali Customary Institutions.

116. Laws et al., LearnAdapt.

117. Mariani, China’s Engagement.