Traditional authorities as both curse and cure: the politics of coping with violent extremism in Somalia

ABSTRACT This paper explores community perceptions about traditional authorities’ roles during the disarmament, demobilisation and reintegration (DDR) of former combatants. We have selected the case of Somalia, where both government institutions and traditional authorities have partnered with international actors and institutions, as well as non-governmental organisations (NGOs), to prevent and counter violent extremism (P/CVE). International actors have related to traditional authorities based on the assumption that these actors wield a kind of social power that facilitates the reintegration of former members of the violent extremist organisation al-Shabaab. Based on mixed methodology research we explain social reintegration in Somalia from the community perspective, and find that P/CVE programmes are expressive of co-optation of traditional authorities. We make the case that ‘risk coping’ helps explain why a majority of civilians prefer the government-led formal reintegration pathway of ex-combatants to the traditional authorities pathway. We conclude by discussing the implications that this has for NGOs/INGOs active in this P/CVE sector.


Introduction
This article examines what community members' perceptions of the inclusion of traditional authorities into P/CVE tells us about the ways that Somali society copes with violent extremism.Informal rulers such as non-state armed groups have significant military and administrative governance records and perform state-like functions in Somalia as well as in other places such as Colombia, Iraq, Myanmar, and elsewhere. 1he vibrancy of informal order in insurgency contexts presents challenges to the interactions that international peacebuilding institutions attempt to forge with and also beyond the host state; with NGOs, traditional authorities, or community members. 2 This is important because the capacity for physical violence by both formal and informal rulers during war-to-peace transitions informs what strategies peacebuilders will prefer to use.
We show how local, as well as national, power holders contest international peacebuilding.At times, the political strategies of ruling elites and non-state armed groups to co-opt and manipulate community actors may bring statebuilding and peacebuilding logics in conflict with one another.As Menkhaus explains, local political adaptation to state abuse and state collapse in Somalia means that peacebuilding and statebuilding initiatives may work against each other, challenging outside actors and donors to do a better job at disentangling and distinguishing these. 3State fragility makes the logic of working locally very clear to international donors/peacebuilders, and, whilst this makes intuitive sense, researchers and practitioners also need to factor in how local power holders play 'state-like' roles which may instrumentalise international-local partnerships for peace.We illustrate this latter point using a case study of community involvement in P/CVE in Somalia.H _ arakat al-Shabāb al-Mujāhidīn is an al-Qaeda-affiliated Islamist armed group often known by the shorthand al-Shabaab.Due to its governance and service-delivery capacity, semi-territorial control, and military campaigns, al-Shabaab has accrued basic empirical legitimacy. 4Varieties of political order and legal pluralism have emerged in areas where national state institutions have been absent or ineffectual. 5However, it is an organisation that has been terrorist-listed by the US and the proscription regime has made it politically and juridically highly sensitive for international actors to operate locally in Somalia while avoiding recognition of al-Shabaab as a political actor.Since roughly 2007 informal systems of security and rule of law have adapted to and become enmeshed with al-Shabaab's politico-military power.This de facto empirical power of al-Shabaab but de jure international unacceptability raises the question of what scholars and practitioners need to know about local actors' strategies to make international peacebuilding interventions work in their interest.In Somalia, the Somali government has partnered with UNSOM and a number of implementing partners to implement the 'defector rehabilitation programme (DRP)'.Admittedly 'non-typical' and contextual, this programme has overall been implemented as a DDR process with P/CVE components. 6These P/CVE elements were viewed as means to work 'holistically' with broader segments of the Somali society in order to weaken community support for al-Shabaab under conditions of insurgency. 7An important assumption among DDR practitioners has been that traditional authorities in Somalia will play helpful roles during the reintegration phase of former combatants. 8Our objective in this paper is to examine what community members' perceptions of the inclusion of traditional authorities into P/CVE tells us about the ways that Somali society copes with violent extremism.
DDR policymaking has been influenced by the arguments that local community actors need to be consulted and actively involved in the processes that impact on them. 9nternational peacekeeping interventions have been the classical frameworks of DDR processes, and UNSC resolutions have in recent years increasingly incorporated community concerns into the mandates of peace operations.A variety of community-based DDR tools have been applied to both post-conflict and active conflict settings. 10elatedly, emerging best practices and community-based programming on P/CVE approaches is underpinned by the assumption that effectiveness seems to increase when holistic approaches such as working with communities are included. 11e approach traditional authorities in Somalia as 'contemporary actors in various fields, such as politics, economy, law; their acts and roles are perceived as having a link to the past; on this basis they are accepted as legitimate by their followers'. 12Often these actors are caught in a historical dilemma, or a 'conservative element'; they are bestowed with community legitimacy by way of upholding past traditional conventions yet the legitimacy of their rule persists only in as far as it is oriented towards community interests in the present. 13Their authority is continuously assessed by the community, who in turn decide to perform the show of will to obey or not.Collective visions of 'traditional' rules are at the heart of the relations between this kind of authority and its followers. 14Community beliefs about tradition have a relatively stable quality, offering an important source of community order and dispute settlement based on customary justice (xeer), yet there are temporal effects on interpretations of tradition leading to adjustments that reveal the capacity for flexibility which in turn is key for traditional authorities to continuously earn community legitimacy over time. 15Therefore, we need to bear in mind that the interplay of historical and social conditions with individual and collective decision-making in Somalia means that: (i) traditional authorities' positions and powers are dynamic, and vary across different regions, and (ii) social changes (gender relations and urbanisation), transnationalisation of the Somali civil wars, and the war economy's impact on livelihoods are alternative determinants of social identity in Somalia.
In this article, we seek to contextualise and nuance existing knowledge about the locallevel assessments that community members make about traditional authorities' powers when it comes to reintegrating former combatants.Our data shows that community members are keenly aware that traditional authorities' clan biases sometimes undermine (international) statebuilding objectives.Guided by local politics, the traditional authorities mechanism in Somalia is known to sometimes support and sometimes obstruct the efforts to reinvigorate the formal government. 16We highlight contemporary challenges to traditional authorities legitimacy, especially in South-Central Somalia.However, a few caveats deserve mentioning.Firstly, our research design sketches determinants of social reintegration in a particular time and place and does not lend itself to draw out farreaching implications for traditional authorities' positions and powers seen in relation to other pillars of Somali social structure.Such analysis has been an important contribution of Somali studies. 17Secondly, our focus on community perceptions about the main pathways of reintegration led us to examine the assumptions about traditional authorities' functions in the broader international DDR engagements.The purpose of implementing survey techniques in South-Central Somalia is to zoom inside-out from individual and household perceptions to a broader understanding of social reintegration, which encompasses the concept of community-based reintegration more often employed by the UN. 18With this in mind, we are cautious not to inadvertently lend weight to a reductive perspective on socio-political life in Somalia.
Our data, drawing on surveys and focus-group discussions in three Somalian cities in the period 2016-2020, tells us that community members bestow limited trust on traditional authorities on several critical aspects of transnational DDR and P/CVE.At first glance this is a counterintuitive finding.We develop an explanation of this whereby community support for traditional authorities is situational and context-specific.Traditional authority powers in Somalia are fragmented yet vital, and communities depend on their role in managing inter-communal affairs and upholding relative stability. 19Our analysis shows that communities perform an important grassroots discernment of what ends are well served through the informal channel.Perceptions about clan authority are variable and depend on the issue's perceived threat level to vital community interests.We identify the DDR process as a particular and contemporary source of fragility of traditional authority in Somalia.
We have structured the article in the following way.First, we describe national and local level governance logics and the main characteristics of clan politics.After that, we discuss our methodology and survey results.Having explained what is at play behind the variations in public attitudes towards those reintegrating, we proceed to draw out implications for the work by NGOs/INGOs active in P/CVE, DDR and international statebuilding.

Non-state armed groups rule in Somalia is enmeshed with local politics
Since 1991 when the Somali government collapsed, the substate level has been characterised by the politics of coping. 20A mosaic of overlapping formal and informal forms of governance has manifested.Despite messiness and contestation, this has, in some places and under specific conditions, provided basic services, rule of law, and plural forms of justice.We will not repeat the established literature about national versus local governance in Somalia, except to highlight two prevalent characteristics and their relevance to our argument: mobilisation of identity politics and the conditions of insecurity and insurgency.
Clan remains a key characteristic of local politics, both formal and informal ones.Localised politics is significantly identity-based in Somalia, making clan genealogy and xeer into basic social institutions. 21Traditional authorities have historically mediated clan tensions, in order to prevent clan interests and clan alliances from taking on highly destructive forms escalating into communal warfare and ethnic atrocities.This helps explain that clan support or opposition to state-building and state consolidation depends on the perceived advantages that such options afford to local political stability.Clan elders' dispute management function has been susceptible to pressure during the many political crises.Our focus is on how this pressure has manifested since the civil war in Somalia received renewed global attention in the mid-2010s.Local governance by Islamist militants, and in particular al-Shabaab from 2007 onwards, have shown the malleability of the traditional authorities.During the insurgency, the traditional authorities have faced attempts to instrumentalise and manipulate them by co-ethnic political elites and al-Shabaab militants alike.There has been a 'misuse' of the traditional authorities by al-Shabaab serving to embed the movement into the community.There has been a similar 'misuse' of the traditional authorities by government actors who to try to bargain with them over ways to infiltrate the community and collect intelligence about al-Shabaab members. 22As Gaas reminds us, traditional authorities claim power by using their positions in a bi-directional power dynamic. 23o enhance community security under conditions of risk and state weakness, people have devised a multitude of local security and justice practices.These have given rise to polities that have enjoyed public legitimacy in as far as they provided a basic modicum of predictability and security.In South-Central Somalia one of the 'pragmatic' public goods was effectiveness of the al-Shabaab rulings to handle the sheer 'case-load' of violence and criminal acts in the aftermath of the civil war.The death toll had outnumbered the ability to pay blood compensation at mag clan level and the xeer system was overburdened. 24hese unresolved killings had led to an increase in male-on-male revenge killings, and the traditional authorities have since Siad Barre's days been weaker and particularly complex and fragmented in this part of Somalia. 25We can think of the traditional authority actors as local statebuilding 'spoilers' who resist the strengthening of a central state, since such a state could lead to a dismantling of 'what works'.Judging based on contemporary history, the state has tended to be predatory, non-representative, and abusive.Interestingly, these same spoilers could still act in support of international peacebuilding initiatives because these often bring lucrative opportunities, and sometimes bypass Somali government control. 26he jihadist insurgency has nurtured such resistance to state reinvigoration due to al-Shabaab's non-recognition of the Federal Government of Somalia and objectives of applying strict shari'a and ridding the territory of foreign forces.Al-Shabaab has been a major source of political violence directed at AMISOM, the Somali government, and civilian populations.At the same time, the group has also penetrated branches of the Somali government and this 'tactical collusion', although there is limited evidence on the extent to which this is the case, could be understood as a mafia protection racket, constituting a form of loose elite bargain with political actors. 27This hybrid form coexists with the 'conservative' security feature of political and business elites' reliance on the coercive capacity of their clan to mobilise clan hostility for the purpose of advancing their interests.This means that present-day clan leaders can comprise a combination of politicians, clan elders, and businesspeople, meaning that political identities and clan identities can pragmatically blend to pursue overriding interests and during conflict this has often been to ensure that their clan has effective capacity to mobilise to fight, or threaten armed violence against rivals. 28uring al-Shabaab's insurgency, traditional institutions have upheld a local political order, providing dispute settlement and basic services.Their roles are to represent community affairs and to assess the threat of what al-Shabaab might do if community members support state-building efforts and enhanced government presence.Their authority in upholding social order is directly linked to their local sources of legitimacy. 29Traditional authorities' risk assessments can make a big difference for the local protection order.Elders and sheikhs that are deemed favourable of UN and international agencies or nascent state structures have been at risk of death threats, harsh punishments, and assassination. 30As a result, some community members prefer that elders refrain from overtly expressing positive views about the international presence.Therefore, it is probable to assume that people understate their potential support for the government, and internationally supported, DDR programmes. 31Communities also know that some elders have complied with al-Shabaab's demands, making 'survival deals' that have involved terrible ethical dilemmas yet nonetheless serve vital clan-based interests, at least in the short term. 32Furthermore, community members know that some elders have become instrumentalised, they approach their role in the sense of running a political campaign and prioritise self-serving agendas over community affairs.
Islamist militant rulers showed awareness that if stability and predictability of localised governance arrangements can be maintained, then this earns the militants some degree of legitimacy among community members. 33Skjelderup argues that this informed the pragmatic choice by al-Shabaab to sustain and not dismantle the traditional authorities structure in the lower Jubba region of Somalia. 34Assumedly, al-Shabaab viewed these governance arrangements as the most effective through which to pursue their politico-military objectives.The weakly institutionalised elders of South-Central and Southern Somalia faced an 'existential' choice, which involved cooperating with al-Shabaab in return for some means of upholding more of their original authority in managing clan affairs.Communities came to cautiously appreciate the stability that came with the unlikely partnership. 35While considered repressive and brutal, community members could compare the situation with a prior alternative considered worse and up to a point accept the ways in which they performed governance functions, such as basic justice based on shari'a. 36mportantly, the argument is that customary rule was co-opted and not supplanted.The al-Shabaab was to some extent influenced by local institutions and social order, and ultimately did not enforce revolutionary ideology or fully replace national laws with its strict application of shari'a law.During insurgencies, community may be tightly controlled but devises self-protection and resistance strategies. 37Seen in that light, traditional authorities' cooperation with al-Shabaab is indicative of local strategic decisionmaking under repressive conditions more so than a symbol of genuine support for the group's ideology or methods. 38

Methodology
For our work on social reintegration, we have used a mixed methods research design.We carried out our experimental survey between October and November 2020 with 1503 respondents across 75 communities in the three cities of Mogadishu, Kismayo and Baidoa in South-Central Somalia.The experimental design follows recent work on examining citizens' preferences through a forced choice conjoint survey experiment. 39e opted for conjoint analysis since this enabled us to test our hypotheses while allowing us to simultaneously vary different ex-combatant attributes such as recruitment profile, clan identity, combatant profile and the type of DDR programmes they underwent (both formal and informal).Based on a close reading of extant DDR literature, we use a total of nine ex-combatant attributes with each attribute comprising of a few levels making it possible to test the relative explanatory power of our competing hypotheses.Additionally we test our hypotheses with three sub-groups we believe are key in community reintegration namely: the traditional authorities, women leaders, and ordinary civilians.This allows us to identify and compare different attributes within and across these groups in one single experiment.
Site selection was primarily driven by the fact that these three cities have hosted the best known and longest running DDR centres run by the Somali government, with accompaniment from the UN and implementing actors. 40Fielding surveys in these three cities ensured a sample of respondents familiar with reintegration of former combatants.This is imperative for the validity and reliability of the study because it adheres to the principle that experimental survey manipulation must be 'information equivalent' in relation to the background features of the scenarios presented. 41Our survey respondents were randomly selected by our research enumerators from 75 different communities across the three cities.
We have taken several precautions to protect our respondents' privacy and confidentiality, and participation in this study was voluntary. 42Our field research manager had been given an in-person training by one of the authors of this article both on how to recruit respondents and on how to explain and read the conjoint scenarios to the respondents.We held zoom training sessions with the research manager and the team of eight enumerators (all Somali nationals) prior to launching this survey.We checked the quality of the data at the end of each working day and one of the authors carried out a daily debriefing session with the research manager who in turn conducted debriefs with every enumerator during the entire duration of this survey.
We have also used qualitative methods; for instance, we held two focus group discussions (FGDs) with officials having worked in Kismayo and Baidoa DDR centres to discuss the reintegration process, from the time when a former al-Shabaab militant is brought in until the time they are sent back to the community (see Table 1).Additionally, we draw on our previous work on DDR and P/CVE, where one of the authors of this study had conducted semi-structured interviews and FGDs with traditional authorities, women leaders and youth leaders in two of the cities selected, Baidoa and Mogadishu in 2016 and 2017, with a total of 55 individuals. 43here are four reintegration modalities in Somalia: a) The government-led and internationally supported DRP programme for low-risk al-Shabaab defectors.The Somali government, partnering with the UN and implementing actors have been running three DDR centres for persons who choose to leave al-Shabaab; b) The amnesty deal with so-called high-value defectors, who, in exchange for defecting along with their followers, are supposed to receive protection and avoid accountability for their past behaviour; c) The prison programme for defectors who are assessed as high-risk and are sent to military courts which convicted them and handed down prison sentences; and, d) The traditional authorities' channel of reintegration.
Former combatants in Somalia are divided into risk categories, enabling us to study how this categorisation affects community acceptance of former militants.All former al-Shabaab members are first vetted by a dedicated unit within the Somalia National Intelligence and Security Agency (NISA) and sometimes by regional equivalents, after which the internationally funded DDR centres start the rehabilitation and reintegration support of ex-combatants deemed low risk. 44Those vetted as high-risk combatants but who defected may qualify for amnesty while those apprehended and categorised as highrisk undergo trial and are imprisoned.This latter high-risk group rarely goes through the DDR centres' disengagement and deradicalisation programme, although prisons have offered various rehabilitation activities. 45Importantly, the risk category of the ex-

Results
We present a selection of results from our overall social reintegration project in order to analyse and cross-validate our findings relating to the variance in community preferences for the different reintegration pathways by drawing on our qualitative data.Figures 1 and 2  indicate that when it comes to security-related issues, such as ex-combatants involved in killings (Figure 1) and his/her recruitment history (Figure 2), the traditional authorities channel is least preferred.In fact, it is almost 20 per cent less preferred than other types of DDR channels such as the DRP, Government amnesty or the government-run Prison Program. Figure 3 tells us that for duties related to institutional roles and responsibilities regarding rule of law outcomes, the traditional authorities are ranked the highest by as much as 55 per cent more than the AMISOM, 32 per cent more than the UN and 10 per cent more than the Somali government.
Our finding about social reintegration in Somalia tell us that traditional authorities' reintegration pathways enjoy lower legitimacy than expected.Community respondents prefer combatants that undergo the government-led and internationally supported DDR programme, and the government amnesty, over those that go through a communitybased and informal reintegration channel overseen by the traditional authorities.This indicates that ex-combatants and also communities face security dilemmas during their reintegration process. 47Because community generally speaking legitimate the authority of traditional authorities, 48 we unpack this finding further and posit that civilians make situational assessments about when traditional authorities are considered credible or legitimate.
Informal channels driven by traditional authorities constitute an alternative to formal institutions in overseeing reintegration processes.Customary dispute resolution mechanisms assumedly raise community preferences towards ex-militants that reintegrate through this channel. 49Nonetheless, we find that individual security concerns linked to reintegration processes undermine local support for the traditional institutions.This is in line with work showing that heightened threat perceptions can lead to risk aversion that affects political preferences. 50The traditional authorities' channel of reintegration in Somalia involves several security-related dilemmas for community members.This channel does not entail disarmament, whereas the internationally supported DDR programme does require ex-combatants to disarm.Former militants might prefer this informal traditional authorities channel as it does not mean giving up weapons. 51But when ex-militants continue bearing arms, civilians form perceptions about the motivations of these former combatants and whose security they will defend.They assess the likelihood of these ex-combatants exacerbating clan-based disputes or criminality, on the one hand, or participating constructively in community life, on the other.
The UN's and other international actors' continued support for state-building is necessary because the Federal government of Somalia still has limited state capacity.But community actors are understandably cautious in voicing support for this international engagement.This is for at least two reasons; such vocal support could lead to other community members becoming suspicious of this person or organisation, and in the worst case denouncing them to al-Shabaab locally. 52Secondly, state-building projects are locally assessed according to whether they will upset/dismantle the social order that community members depend on for basic services and security.Against this background, we believe that while internationally backed DDR builds some confidence at the community level in the context of active conflict, this is mainly linked to civilian expectations about the opportunities of international peace/statebuilding to local powerholders.Our claim is similar and builds on Menkhaus' point in regard to the UN Operation in Somalia (1993-1995) that despite its many failures according to him had a mixed positive legacy on governance in later years due to how it, 'poured an enormous amount of money as well as sizable employment and contract opportunities into the country and inadvertently helped to stimulate and strengthen legitimate businesses, thereby shifting business activities away from a war economy'. 53There is a similar tendency with the present-day international engagement in that it has created a veritable P/CVE political economy which brings opportunities to those that position themselves well. 54

Micro-politics of risk coping
This section interprets our experimental evidence and cross-validates this using also our qualitative material.In active insurgencies, the pressing need to manage risk under conditions of social interdependence means that community members prioritise security over other considerations. 55The history of violence has produced long-lasting and deeply absorbed adaptations affecting the whole social system, 56 and under these conditions, the 'silver lining' is that civilians develop competence in risk assessment and coping strategies. 57We draw on the recent literature about individuals developing 'expertise'a competence or a self-efficacy in surviving insecurity -as a result of direct exposure to political violence or traumatic events. 58According to this research, individuals tend to develop a heightened capacity to anticipate future threats and when levels are thought to be too high at a specific time/place, they will display risk mitigation or avoidance behaviours. 59Interestingly, the tolerance of risk threshold will be higher among those that have developed this expertise. 60This is a continuous process that informs community members' preferences and undergirds their decision-making. 61e expect based on this account that communities with such heightened threat perceptions also exhibit less trust towards 'outgroups' when these become perceived as predictors of further inter-clan tensions. 62ommunity members are basing their perceptions of former combatants on particular attributes in order to anticipate their likelihood of recidivism, or the risk of revenge attacks on communities that host certain individuals.This functions like a micro-political check on the destructive potential of inter-communal animosity that could be mobilised by violent extremism.As a result, those ex-combatants that fit attributes associated with high risk would be ranked lower across a choice of whom to welcome in a community.We suggest that community members skilfully categorise former militants according to risk attributes based on associating certain attributes, such as (in)voluntary recruitment, association with abusive units, and involvement in killings, with varying risks of leading to the breakdown of social order. 63In the same way, if former militants are deemed less likely to be a threat, then community members are more willing to accept their reintegration.This occurs in a dynamic fashion since beliefs about risk and security are continuously updated. 64ur data shows that the perceived threat posed by a former militant is not significantly rooted in community assessments about the former militants' ideological commitment.In the case of Somalia, becoming a member of al-Shabaab was often gradual, informal and fluid, shaped by a mix of factors other than personal ideological justification of terrorist violence. 65Communities have seen that al-Shabaab leaders found to be high-risk and very radical are wanted by powerful international actors, especially the US and UK, which means that the government amnesty channel or counterterrorism operations will oversee these cases.Some of those categorised as high-risk individuals have held senior ranked positions in the group and bring back considerable numbers of followers, now defectors, with them.Therefore they are considered valuable sources of intelligence for NISA, and this signals to many community members that such a person has crossed a 'point of no return' and can never go back to al-Shabaab, as someone with long experience from DDR work in Baidoa told us. 66ost significant is the group's local capacity for violence and extent of territorial control. 67This could mean, then, that traditional authorities, and P/CVE NGOs, are not primarily assuming roles in transnational P/CVE in an effort to deradicalise al-Shabaab sympathisers, members, or ex-combatants.The more plausible interpretation is that the primary motivating factor for involvement is pro-social, or in other words improving inter-communal stability.NGOs seeking to play roles in P/CVE are acting in the shadow of harder counterterrorism and deradicalisation interventions aimed at the country.While they know that the Somali government receives international assistance for P/ CVE activities, it seems rational for smaller and rural-based NGOs to think about prevention from an extremely localised lens.An implication of this is that community members assessments cannot be understood as a blanket stigmatisation process.The reasonability of this nuanced risk coping also dovetails with a recent study showing how some individuals joined al-Shabaab 'naturally' or gradually in areas it controlled, and that while motives for doing so varied, ideologically-based sympathies were sometimes the next step after training or socialisation occurs inside the organisation. 68ocial capital is also an important part of skilful coping with former insurgents in your community.Rumours in war-time are a part of social capital, and crafting narratives about which individuals are low versus high-risk is an example of how to cope with uncertainty and risk in dangerous environments. 69In regard to our case, rumours and suspicions work as a form of 'public intelligence wing' in a highly interdependent communal society. 70Rumours inform opinions about those that came through the traditional authorities channel. 71The dimension of suspicion and rumour is a wellknown dimension of reintegration. 72Numerous illustrative anecdotes were shared with us in the FGDs, and the bottom line was that in the community's eyes, elders do not on their own have the authority to verify whether someone ceases to support al-Shabaab and has defected for good. 73Serious security incidents have arisen where it has been hard to distinguish whether former members carried out acts of violence due to clan power rivalry or carrying out revenge on order from al-Shabaab. 74There have been cases where the very elder who vouched for an ex-combatant was the target of that individual's lethal assault. 75Government security forces have increased mistrust of former combatants also by enlisting them to infiltrate community and denounce others that may still sympathise with al-Shabaab.Those practitioners that have followed the process for a while can give numerous examples of rumours heightening the anticipation of violence when 'Disengaged have become informants of some kind'. 76he expertise to cope with risks is at work when community members shape their public opinions.The political and ethical implications are part of the strategic and security-centred reasoning, but they have to be ranked according to priority.For certain specific roles, traditional authorities have been delegitimated (and the international and national formal pathways have been accepted) although in regard to their other functions they continue to enjoy support.Capturing how traditional authorities are wedged in between logics of state-building and local insurgency, some respondents expressed that, 'communities are caught between two flames'. 77This strategic element in civilian decision-making came up in a follow-up interview with a traditional leader commenting on the erasure of power, 'The citizens understand how the clan system is undermining the government authority including the justice system'. 78Civilians therefore simultaneously show support for international peacebuilding and scepticism about statebuilding since historically sovereign power has been abusive or woefully absent in such large parts of Somalia, necessitating the emergence of hybrid governance arrangements.While generally local interests go hand in hand with traditional authority, our analysis shows that civilians will adopt other strategies based on their understanding of the issue. 79

Implications for peacebuilding practice and P/CVE
The finding that respondents prefer the formal over the informal reintegration channel tells us that the communities have strong reservations about the involvement of traditional authorities on DDR and especially reintegration in Somalia.Our results show that support for the UN peacebuilding presence and support for the government's DDR activities is high. 80For international peacebuilders, the case of Somalia seems to offer promising signs that community support for former militants coming through the formal DDR channel is much higher than for the informal channel, given the contested applicability of DDR in contexts of active militant armed groups. 81While it seems intuitive for policymakers and donors to consider bottom-up or informal alternatives to DDR instead of quite costly international peacebuilding interventions, our finding suggests that prior conflict analyses need to assess the conditions needed to allow interventions' acceptability and what is at stake for the P/CVE actors and their dependants. 82omalia's DDR experience on the whole has relied on traditional authorities playing important facilitating roles.We believe that in the reintegration phase, the DRP is perceived by community members to complement, rather than replace or contest, local and traditional sources of authority in this specific policy area.Thereby, it poses less concern to those opposed to expansive state-building, who might otherwise act as spoilers.The reason would be that it has taken on a form that does not upset the preexisting, eminently local, forms.Instead, it brings a significant amount of resources and opportunities that community members view as beneficial.
Several national NGOs have benefitted from the international focus on P/CVE and among the informal actors that take up roles and activities linked to DDR and P/CVE we find both 'traditional' and 'modern' ones. 83International peacebuilders need to do continuous learning from the local contexts, because it is not straightforward or automatic when and to whom locally driven peacebuilding is preferable.Depending on what form of DDR is applied, it may fragment traditional authorities' powers further (and fail to address legitimate local grievances).P/CVE programming as part of DDR grew out of the felt need to tailor and contextualise DDR better to insurgencies and 'jihadist' conflict. 84However, our analysis shows that when international actors engage traditional authorities as change agents in their P/CVE then this will have contingent effects, resisting top-down templates.When programming is rolled out, therefore, adjustments must be allowed as part of the process to avoid or minimise the added risks for community members that carry out P/CVE relevant work.
State actors and non-state actors alike have, through omission or commission, played a role in promoting violence.In our FGDs with community members, a common reflection was how extremism had a mutable nature and how it appears 'imperialistic' to them that foreigners can decide what political violence crosses a definitive red line (in turn, triggering P/CVE programming based on certain assumptions that ascribe individuals with extremist or radical characteristics). 85Given the fragmented and co-opted nature of traditional authorities in Somalia in the context of counterinsurgencyinfluenced international engagements, many other NGOs have rebranded their prior peacebuilding or development focus and forged partnerships with P/CVE INGOs.This raises two interrelated concerns.First, the NGOs become vulnerable by affiliating with international actors in the prevailing insurgency context.If their newfound frontline P/ CVE roles lead to backlash against them and they have to close down, they cannot fall back on customary sources of legitimacy in the same way that many traditional authorities do.Second, in order to 'do no harm', INGOs seeking local P/CVE partnerships in Somalia should consult with local authorities and representatives of different segments of society in order to tailor programming to community dynamics.Sometimes programmes may need to be tailored differently to different groups in the same setting in recognition of the rationale of the screening process of ex-combatants, namely that individuals had very different sympathies and roles in an extreme violent group, and this determines duration of rehabilitation. 86n operational terms, funders should consider securing longer project cycles and solid sources of funding in order to help safeguard the high-risk work of NGOs in this domain. 87INGOs need to consider P/CVE programming towards marginalised clans in Somalia separately and set up earmarked forms of support for these groups. 88eintegration options for minor clans are fewer in the Somali case, since these individuals often seek out formal pathways because they can less often count on a powerful elder and clan structure supporting their process. 89This is important because societal norms and socio-political identities do not change easily or quickly.This is a point that also concerns gender norms.For instance, NGOs might continue working for gender equality primarily, but in setting up civic space to consider the connection between gender and VE some of them additionally play a useful early warning and P/CVE role.Alternative ways of imagining gender identities can sometimes work to empower women, as well as men, girls and boys, by charting an alternative path to conservative and restrictive gender norms mobilised by violent extremist groups. 90Worth recalling, though, is that these NGOs run the risk of their gender work becoming regarded as off-sync with society by associating it with international discourses on gender and P/CVE.The meaning-making strategies of women are therefore important; they need to be respected interlocutors in mediating common ground on standards of gender equality. 91t would be well worth safeguarding a terminology that keeps all conflict resolution tools open, and make use of categories that resonate with community needs.It may be that human rights, the rule of law, development or transitional justice more directly resonate with underlying grievances. 92Notions such as P/CVE are strategic and may function as more of an external framing exercise but run the risk of securitising the work of local NGOs.One clear limitation of P/CVE work in Somalia is that due to the high-risk environment, INGOs are operating in urban centres.This makes the dilemma quite apparent, since it means that any NGO candidates motivated to play a role in P/CVE in much of southern Somalia's countryside, where the support base for al-Shabaab is generally stronger than in the major towns, would need to be sufficiently resourcestrong to operate both in urban centres and in the rural areas. 93n attempt to tailor P/CVE approaches to community dynamics would frame the next logical step as one of advancing global south P/CVE perspectives and voices.This entails moving international-local P/CVE partnerships in the direction of mutual exchange and two-way knowledge transfers.It would be important for Somali P/CVE community actors to meet and coordinate initiatives and lessons learned, with one another and with global networks.As our analysis shows, community actors support international involvement in the country's DDR-process.In part because of the many P/CVE networking and funding opportunities that are part of the DDR programming for Somalia and also on offer through other UN agencies in the country.However, it is important to be mindful of the longer-term challenges involved with eroding further the powers of traditional authorities.Our analysis shows that harmful power imbalances in local politics must be managed and that civilians in insurgency-affected areas develop expertise in doing exactly that.
Community members have expectations, and among them are also skills and contributing to community.Those in charge of rehabilitation programming need to know what conditions are valued by communities. 94Taking community conditions seriously is important since in a 'political economy' of P/CVE there are strong incentives to account for volumes of programme beneficiaries to back requests for funding and continued commitment from international implementing partners.Still, the total number of persons that have passed through the centres tells us very little about the process of them contributing to community and political order.As noted by a DDR official with programming insights, 'NISA also has pressure to have beneficiaries, to show that they get people through the existing projects'. 95Importantly, this study was not designed to measure and establish success rates.Echoing Stig Hansen, 'caution is needed when trying to establish success rates of deradicalisation and disengagement programmes, and how studies over longer periods of time perhaps are needed to correctly assess such rates'. 96

Conclusion
Traditional authorities continue to play significant roles in mediating inter-communal affairs and upholding local order. 97However, communities are intimately aware of the limitations of informal reintegration led by traditional authorities.The cooptation of traditional authorities by al-Shabaab result in fear and heightened threat perceptions about new waves of community-level political violence. 98These threat perceptions are also what primarily drive citizens' preferences in accepting ex-al-Shabaab combatants for reintegration.Their lived experiences of insurgency dynamics give the community members reasons to suspect that traditional authorities sometimes obstruct government security forces efforts to apprehend al-Shabaab suspects.The community members have to some extent delegitimated the traditional authorities specifically on reintegration and P/CVE for the reasons analysed in this article.
Support for the traditional authorities is issue and context-specific.It is important to tease out the micro-politics at work in the shaping of these opinions.We suggest that this is due to an important psychological risk coping capacity.From the community risk coping perspective, the assumption of DDR work that excombatants, generally, pose a threat to peace in the aftermath of war needs further nuancing. 99Instead, there may be specific characteristics of former militants vetted as posing a higher risk to the local security concerns.For instance, those thought to have been directly involved in violent extremist acts as opposed to those that had supporting roles such as drivers, informants, cooks, smugglers, etc. 100 P/CVE INGOs will need to keep in mind that P/CVE NGOs in Somalia may welcome and participate in peacebuilding through a wide repertoire of development, rule of law, or transitional justice programmes, yet some of those P/CVE actors could, once well resourced, work against statebuilding ends.Through a Somalia community lens, national state actors have in many areas historically seemed more 'foreign' and 'extremist', in some sense of the term, than locally based rulers.This is significant to our story of how 'the local' may be simultaneously violent and legitimate.Again, this is a sober reminder to INGOs to hold their judgement on which perpetrators of political violence are causing the worst forms of insecurity from the viewpoint of communities.Indeed, al-Shabaab tactical penetration of local governance systems rested on a strategic understanding that this system was held in high esteem by local populations.
International peacebuilders and peacebuilding research would do well to probe deeper into this micro-level 'vetting' process about risks and opportunities, including its emotional and psychosocial dimensions. 101Do determinants of community support mould governing strategies by non-state armed groups such as al-Shabaab?It would also be interesting to conduct further research into the transformation processes of high-risk former combatants, and the probability of their 'violence specialist' 102 coping skills helping or not in them assuming leadership positions.Further research is warranted into understanding traditional authorities' reactions and responses to this delegitimation process.Our data cannot tell us to what extent traditional authorities reasoned that in order to preserve more of their original authority in core roles it may have been useful to have the formal and internationally supported reintegration pathways in place.Additionally, the perceived legitimacy of traditional authorities influences community preferences of the formal versus the informal reintegration pathway.Clearly, there are other factors and drivers involved and the decisions at the individual level are more nuanced still in reality.

Notes
understood the nature of the research, and what participation entailed, and also that participation was voluntary.While precautions to seek active consent in appropriate ways are crucial, we also want to mention that when a data collection has been well planned and grounded in local realities, human subjects understand more than is sometimes thought and take rational decisions about their participation.43.Gelot and Hansen, 'CVE Brokerage'.44.European Union Agency for Asylum, 'Somalia'.45.Felbab-Brown, 'The Limits of Punishment'.46.Confidential interview with NGO practitioner active in P/CVR, April 2017.This person had also observed the disappearance of large numbers of people that have been collected by security forces meaning that they did not reach any collection site for the DRP, which adds to the challenge that unknown numbers of ex-combatants go through some form of informal pathway.

Figure 1 .
Figure 1.Marginal means for type of DDR program by whether ex-al-Shabaab has killed.Acceptance of ex-al-Shabaab combatants by community members based on the type of DDR programme ex-al-Shabaab underwent interacted with if the ex-al-Shabaab killed someone.

Figure 2 .
Figure 2. Marginal means for type of DDR program by ex al-Shabaab recruitment.Acceptance of ex-al-Shabaab combatants by community members based on the type of DDR programme ex-al-Shabaab underwent interacted with ex-al-Shabaab recruitment profile.

Figure 3 .
Figure 3. Citizens trust in institutions to carry out their duties and responsibilities.Response to the question: "How much trust do you have in . . . . . . in their ability to carry out their roles and responsibilities in your community?"Responses were captured in a 4 point scale which were: Strongly Distrust; Distrust; Trust; Strongly Trust.

Table 1 .
Focus group discussions.Table showing the total sample for FGDs carried out before the launch of the survey and after data analyses.CONFLICT, SECURITY & DEVELOPMENT combatants who are accepted through the traditional authority channel is not well known nor is it clear which indicators are used to assess this.This means that traditional authorities facilitate mixed members, varying in risk level and membership status.46