peace and security in the central african republic: Community perspectives from Bossangoa and the western border zones Report Peace and security in the Central African Republic: community perspectives from Bossangoa and the western border zones May 2020

Acknowledgements and disclaimer: This report was compiled by Ben Shepherd based on reports written by Ben Shepherd for Conciliation Resources and Guy-Florent Ankogui-M’Poko for the Centre for Humanitarian Dialogue. Centre for Humanitarian Dialogue and Conciliation Resources are grateful to the research team who conducted the baseline assessment and the staff members of Association for Humanitarian Action in the Central African Republic (AAHC) who contributed to the development of the final questionnaire, and identification of and contact with community leaders and members of their respective communities.

This publication is funded with UK aid from the UK government. The contents of the publication are the sole responsibility of Centre for Humanitarian Dialogue and Conciliation Resources and the views expressed do not necessarily reflect the UK government’s official policies.

Smart Peace is a four-year UK Department for International Development (DFID)-funded programme (2018-2022) for strategic conflict resolution. Smart Peace is implemented by a specialist consortium, led by Conciliation Resources, in partnership with International Crisis Group, Centre for Humanitarian Dialogue, the Asia Foundation, ETH Zurich, Behavioural Insights Team and Chatham House. The Smart Peace consortium combines expertise in conflict analysis, community dialogue, elite mediation, evaluation, policy influence and behavioural science to deliver targeted and adaptive conflict resolution interventions in CAR, Myanmar and Nigeria. Smart Peace will share learning from practical experience to improve global policy and practice.

In the Central African Republic, the project will be implemented by Centre for Humanitarian Dialogue and Conciliation Resources in Mambéré-Kadéï, Sangha-Mbaéré, Nana-Mambéré and Ouham. International Crisis Group and ETH Zurich will provide research and analytical support.

Published by: Conciliation Resources Burghley Yard, 106 Burghley Road London NW5 1AL

Design & layout: www.revangeldesigns.co.uk Cover photo: Peacekeepers serving with the UN Multidimensional Integrated Stabilization Mission in the Central African Republic (MINUSCA) operate a checkpoint. © UN Photo/Hervé Serefio Executive Summary

• Field research was carried out in late 2019 in two border zone in the last two years, due either to theft locations: in and around Bossangoa town in Ouham or to their large-scale movement to Cameroon for prefecture, in the geographical north west of the protection, had reduced conflict between herders and Central African Republic (CAR) between the capital farmers by about 60-70%, and two thirds of respondents Bangui and the Chadian border; and along CAR’s felt that insecurity was less acute than in 2013. remote western border with Cameroon, across four • In Bossangoa, the former Séléka rebels were no communes in three prefectures, Mambéré-Kadéï, longer present and the Anti-balaka were revealed as Sangha-Mbaéré, Nana-Mambéré. The two research relatively peripheral; the absence of armed groups areas are referred to in this document as ‘Bossangoa’ meant that most security incidents in 2019 were and ‘border zone’ respectively. This report presents low-level crime and concentrated in Bossangoa results from surveys and interviews conducted with town. However, the ex-Séléka nonetheless remained more than 1700 respondents, as well a preliminary the most significant perceived threat for those in comparative analysis. the Bossangoa research area, and just over half of • Both the Bossangoa and border zone areas were respondents felt that security had deteriorated in the found to be relatively peaceful in comparison to past twelve months. the height of the crisis. The absence of cattle in the

central african republic

SUDAN

Vakaga CHAD

Bamingui-Bangoran SOUTH SUDAN

Haute-Kotto Ouham Nana-Grébizi

Ouham- Ouaka Pendé

Kémo Haut-Mbomou Nana-Mambéré Mbomou Ombella-M’Poko Basse- Mambéré-Kadéï Bangui Kotto Lobaye

Sangha-Mbaéré

DEMOCRATIC REPUBLIC OF THE CONGO AMEROON om

C 150km CONGO 100 miles © d-maps.c

Conciliation Resources • Peace and security in the Central African Republic 3 A UN Police officer serving with MINUSCA, conducts a class on gender violence at a school. © UN Photo/Eskinder Debebe

• There were some suggestions of a change in underlying • The extent to which communities still looked to the conflict dynamics, away from organised armed group state, broadly defined, to protect them and their activity towards pervasive low-level insecurity linked communities was also striking, given the very low level to crime and economic interests. Crime, banditry of state capacity in CAR. The United Nations Mission in and young people – especially youth associated CAR (MINUSCA) was the most widely-cited protection with armed groups – were the most acute perceived actor in Bossangoa town, where it is deployed in a threats in the border zone and were also important in significant population centre. Young people consistently Bossangoa town. saw MINUSCA as their biggest protector, while older people were more likely to look to local authorities. • The research found that communities in both areas looked to local actors and traditional methods for • Interviews revealed that the crisis had caused conflict resolution, notably traditional leaders and enormous economic damage, entrenching endemic village chiefs. Respondents were likewise clear that poverty and social divisions (herder-farmer in border these local and traditional actors needed material zones, Muslim-Christian in Bossangoa). Both areas and financial support, as well as appropriate training, are vulnerable to external shocks and cross border in order to work effectively. Further detailed research dynamics from Chad or Cameroon, as well as from is necessary on how these local actors relate to both the return of displaced people – herders in the border the state and communities, and what types of conflicts zone and the wider Muslim population in Bossangoa. they can address. The recognition of non-governmental The structural conditions persist across both areas organisation (NGO)-led mechanisms was low, perhaps for a future resurgence of conflict. because of the limited visibility of NGOs in the border areas and in the rural areas outside Bossangoa town.

4 Conciliation Resources • Peace and security in the Central African Republic background

Significant efforts have been made to pull However, the border zone has seen persistent and the Central African Republic (CAR) out of the endemic conflict between herders and farmers over armed conflict that erupted in 2013. After destruction of crops and access to resources, and is exposed to overspill from the activities of armed several failed peace processes, a peace criminal groups across the Cameroonian border. accord was negotiated in Khartoum and signed in the capital Bangui on 6 February Methodology 2019 by the CAR Government and 14 armed groups, following presidential and local In Ouham prefecture, a team of field researchers elections in December 2015 and early 2016. administered 1010 surveys, conducted 257 structured The deal included provisions to create mixed interviews with a cross-section of survey respondents, and undertook 23 long-form key informant interviews units between armed groups and the national with officials, armed group leaders, and international army, and to bring armed group leaders into observers. government. Implementing this agreement is Due to limitations on time, in part due to security crucial to break the cycle of violence in CAR. concerns limiting time in the field, respondents were selected through convenience sampling. However, care Some leaders of armed groups have already proved to was taken to ensure a representative sample in terms be more interested in consolidating territorial control, of gender, with 54.1% of survey respondents male and profiting from local taxes and control of resources, than 45.4% female.4 The age of survey participants was also in engaging in peaceful politics, and violence continues relatively balanced, though with more young and between non-state armed groups, particularly in north middle-aged respondents compared to those over fifty.5 eastern areas bordering Sudan and Chad. Breaking this down by age and gender, the sample is The country is also now preparing for elections in also relatively balanced, with a slight over-representation December 2020. If successful, the election process of women among younger people, and men among could reinforce the legitimacy and stability of the middle-aged and older respondents. government, but it also brings risks that politicians In the prefectures of Nana-Mambéré, Mambéré-Kadéï will manipulate or deepen existing social divides. and Sangha-Mbaéré, collectively referred to in this Smart Peace project partners – International Crisis report as the border zones, data collection was based Group (ICG), Centre for Humanitarian Dialogue (HD), on a qualitative approach collecting the perceptions ETH Zurich and Conciliation Resources (CR) – have and reflections of key players on the existing conflicting conducted research to understand the views of local dynamics between farmers and herders in cross-border communities in order to inform effective, proactive areas of CAR. Thirty-eight group interviews were policy during this crucial period and beyond. conducted, including eight with community leaders, eight The two research areas presented contrasting aspects with herders, seven with women representatives, six of CAR’s conflict and social dynamics. Bossangoa town with youth representatives and two with all target groups. and its immediate environs were one of the key centres In addition, 29 individual interviews were conducted. of the conflict, a heartland of the Anti-balaka1 militias and the site of fierce fighting with Séléka2 militias. Nearly the entire Muslim population of this area was displaced. By 2019 the immediate area of Bossangoa was reported to be relatively peaceful; tensions nonetheless remain high.3 By contrast, the border zone along CAR’s border with Cameroon was not central to the conflict between Anti-balaka and Séléka.

Conciliation Resources • Peace and security in the Central African Republic 5 findings

Perceptions of security Conflict dynamics Both research areas saw a significant reduction in These perceptions of security conditions may reflect armed violence since the height of the conflict between a shift, indicated in the research, in the predominant 2012 and 2015. In the CAR-Cameroon border zone, dynamics of insecurity from the height of the CAR inter-community conflict between farmers and herders crisis. In the border zone, conflict had been orientated was found to have reduced significantly, and a sharp around long-standing friction between herders and drop in monthly incidents of conflict in 2019 compared to farmers over crop destruction, theft of oxen and access 2012 was reported, down to zero in two cases6. United to resources. Around Bossangoa, the conflict at its Nations (UN) sources reported comparatively few security height had been centred on non-state armed groups incidents in the research area around Bossangoa in associated with different religious communities, broadly 2019; reported incidents were overwhelmingly low-level pitting Muslim against Christian. criminality and concentrated in Bossangoa town.7 This matches the pattern of significantly reduced conflict In both areas, neither of these conflict dynamics was events in and around Bossangoa from 2014 onwards.8 still present. The removal of the majority of cattle from the border zone and the displacement of the Muslim However, this was not unambiguously reflected in population from around Bossangoa meant that the community perceptions of insecurity. In the border perceived nature of the threat to communities had zone, two thirds of respondents felt that insecurity had begun to shift away from the long-standing pattern of become less prevalent since the height of the violence herder-farmer violence or clashes between religious in 2013-14. Only a quarter of respondents felt it was communities represented by armed groups. Instead, worse. However, a small majority of 54% in Bossangoa banditry, crime and young people had emerged as the and its environs felt that security had deteriorated in most significant threats. Of course, ‘young people’ the past twelve months. is a broad category in a context where a significant percentage of the population is under 25; it could include those previously involved in armed groups,

Chart 1: perceptions of security

100

80

60 67%

54% Per cent Per 40 45%

20 25% 8% 1% 0 Since 2013 (border zone) Past year (Bossangoa)

Improved Deteriorated No change

6 Conciliation Resources • Peace and security in the Central African Republic and lingeringfear ofex-Séléka. Despite theircomplete visible inBossangoa, thoughunderneathawidespread Changes inpatterns ofconflict were also to some extent a threat by12% andex-Séléka byjust 6%ofthesample. thieves (17%).Bycontrast, theAnti-balakawere seenas with armedgroups, followed bybandits(23%)andcattle young people (29%),especially young people associated border zone, themost commonly perceived threat was reflected in community perceptions ofthreat. Inthe Changing conflict dynamics were to some extent Perceived threats which CAR state authoritiesare unable to prevent. crossing from acommunity inCameroon, something acute spike inconflict caused bycrop damageby cattle border community ofSabewa, whichhasexperienced an In theborder zonetheserisksare illustrated bythe communities inBossangoa) andcross-border dynamics. populations (herders intheborder zoneandMuslim remained. Triggers could includethereturn ofdisplaced from theconflict. Thus,acute conflict vulnerabilities were exacerbated bydeeppoverty andeconomic damage tensions between herders andfarmers. Thesedivisions reported, albeitto alower level, aswere persistent between religious communities. Ethnicfrictionwasalso that entrenched socialdivisionsremained, most notably However, there was widespread recognition in both areas understood anddisaggregated maybeworthwhile. interrogate how thecategory ofyoung people shouldbe of idleness anddisorderly conduct. Further research to in banditryandillegal taxation, orsimply aperception many ofwhomare now unemployed, orthoseinvolved

Per cent 100 20 40 60 80 Young people BanditsCattle thieves Anti-balakaTaxi moto Ex-SélékaNone 0 29% 23% 17% Border zone 12% Chart 2:perceivedthreats 12% Conciliation Resources •Peace and security inthe CentralAfrican Republic 7 6% [0%]

the overall sample. seen asthemost important threat, namedby37%of absence from theresearch area, ex-Séléka were widely only 6%ofthesample. MINUSCA wasseenasprotecting thecommunity by followed byFACA (23%)and‘local authorities’ (18%). perceived as thelargest single source ofprotection (29%), By contrast, intheborder zonethepolice were deployment intheBossangoa area. sample, whichisunsurprisinggiven thelimited FACA as offering protection byjust 1%oftheBossangoa The Central African ArmedForces (FACA) wasseen by ‘local authorities’ (19%)andthepolice (18%). looked to MINUSCA for protection (39%),followed In Bossangoa, thelargest percentage ofthesample community perceptions ofwhoisable to protect them. Significant differences alsoemerged in relation to of protection Perceived sources likely to perceive acontinued threat from ex-Séléka. experienced very few recent securityincidentswere more By contrast, communities inoutlying areas thathad and Anti-balaka,taken together, thanfrom ex-Séléka. the sample perceived athreat from young people, banditry of recent securityincidentshadbeenrecorded, more of communes. InBossangoa town, where thevast majority variation between Bossangoa town anditsoutlying However, there were indications ofsignificant local identified theAnti-balakaasmost important threat. young people (19%),while only 7%ofrespondents perceived no threat at all (20%) and those who identified 19% 7% [0%] 9 Thiswasfollowed bythosewho Bossangoa 7% [0%] 37% 20%

8 Conciliation R esources •P eace and securityin theCentralA frican Republic resolution mechanisms. community recognition ofexternally-sponsored conflict reported looking to NGOs,whichsuggests limited the populationinborder zoneand8%inBossangoa government (28%)thanto thechurch (8%).Just 11%of much higherproportion ofrespondents looked to the respectively intheborder zone,while inBossangoa a conflict mediators by12%and17%of respondents ‘Government’ andchurches were identifiedaseffective to traditional conflict resolution methods, Both studies found thatthepopulationstill largely looked Conflict resolution and 6%for thepolice. outside thetown thesefigures fell to 27% for MINUSCA source ofprotection, andthepolice cited by30%;while was cited by52%ofrespondents asthemost important of MINUSCA andthepolice: inthetown itself,MINUSCA were differences incommunity perceptions oftherole border zone. However, even withinBossangoa there its absence inthepredominantly rural andremote Bossangoa, animportant local populationcentre, and This disparityreflects MINUSCA’s deployment in chiefs for conflict resolution (43%). pronounced, withmost respondents looking to village mechanism. InBossangoa thistrend waseven more ‘local committees’, anewer structure, asaneffective leaders’ for conflict mediation (26%), while 14% identified largest proportion ofthepopulationlooked to ‘traditional precise distinction difficult.Intheborder zone,the built uponpre-existing ‘traditional’ structures makes a years, althoughthefact thatnew initiatives have often multiplicity ofnew mechanismsestablished inrecent

Per cent 100 20 40 60 80 0 37 Bossangoa 21 Ex-Séléka Youth NoneAnti-balakaBandits/crime Ethnicgroup 12 11 Chart 3:perceptionsofthreat –Bossangoa 12 10

2 Soumbé 30 11 17 19 10 11 despite the 13 7 45 Benzambé

21 26 focus for future research. such asvillagechiefsandthe state, maybeavaluable the relationship between local figures andinstitutions the roles they playinthecommunity, andinvestigating ‘village chiefs’ and‘local committees’, understanding definitions of terms suchas‘traditional leaders’, more thanonecapacity. Exploring andclarifyingthe played bykey individuals,whomaybepresent in These findingsmayrisk over-simplifying the roles economically from thedisplacement ofMuslims. play: engaginginpeace mechanismsbutbenefiting by theambiguousrole thatsomeAnti-balakaleaders of theAnti-balakaheartlands.Thismaybeexplained despite thewidespread perception ofBossangoa asone of respondents inBossangoa looking to Anti-balaka, role, and6%theex-Séléka. Thiscompared to just 2% seeing theAnti-balakaasplayinganimportant mediation mechanisms insomelocations, with14%ofrespondents found thatarmedgroups ‘overshadowed’ traditional in conflict resolution. Intheborder zone,the study emerged between theperceived role ofarmedgroups Significant differences between the research areas can beencouraged. meaningful involvement ofwomen andyoung people understand thebarriers to participationandhow the of thesegroups. More research isneededto better to explicitly work towards supportingtheempowerment peacebuilding. TheSmart Peace programme will need support women’s andyoung people’s engagementin resolution. Social normsandpower relations donot women andyoung people incommunity conflict Under current practices, there is limited participation of 1 1 1 n doromboli Koro-Mpoko 45 3 34 2 3 [0] 32 30 19 2 1 [0]

Per cent Per cent 100 100 20 40 60 80 20 40 60 80 0 0 Police FACA Local authoritiesMINUSCA Anti-balakaTechnical services None 26% 29% Chart 4:perceivedsourcesofcommunityprotection Chart 5:PERCEIVEDCOMMUNITYCONFLICT MEDIATION [0%] Churches NGOsAnti-balaka Ex-Séléka Traditional leaders Villagechief Local communities Government 23% 14% 18% Border zone Border zone 12% 6% 17% 12% 11% Conciliation Resources •Peace and security inthe CentralAfrican Republic 9 12% 14% [0%] 6% 18% 5% 43% 1% [0%] 19% Bossangoa Bossangoa 28% 39% 8% 8% 8% [0%] 2% 7% [0%] Conclusion

The research has three main conclusions: policies across the country, including what it can and cannot achieve, in order to manage expectations. • There is scope and a need for local community leaders, informal and formal peace structures, and • There is a need to address the fears and stigmatisation local government to act as convenors of dialogue of youth as troublemakers and to engage young people, and reconciliation. This is to prepare and accompany including those who have returned from armed groups, their communities in a process that could be termed in peacebuilding efforts within their communities. ‘the cooling of hearts’ from the trauma of the Young people have a valuable and positive role to play 2013-2015 crisis that left communities divided along in their communities, when engaged in productive religious and ethnic lines. There is an urgent need activities and provided with skills in conflict resolution. to support structures and capacities for non-violent means of resolving political and inter-communal Based on these conclusions and consultations amongst conflict and co-creating a shared future together the Smart Peace CAR Hub, the members of the Hub will as Central Africans. Local leaders and peace actors focus on supporting the government’s efforts on the may benefit from support in thinking through and implementation of the peace agreement. Centre for developing ways to address trauma and anger among Humanitarian Dialogue will focus on agro-pastoral conflict members of armed groups and to broker intra- and resolution linked to cross-border transhumance in the inter- community dialogue. border zones of CAR, Cameroon and Chad. Conciliation Resources will support the work of local peace structures • There is a need to consider providing strategic support in Ouham prefecture and link them to the national peace to traditional conflict resolution structures and process. International Crisis Group will lead on advocacy government-led processes such as the Prefectural at the global and national level, including engagement Implementation Committee (CMOP) set up by the CAR Government to implement the peace accord signed on at ministerial and presidential level. This will include 6 February 2019. This is an opportunity to provide the engaging with presidential candidates in the lead-up necessary conflict analysis and peacebuilding skills so to the election set for December 2020, to secure a that CMOP members and other peace structures carry commitment to a peaceful election and acceptance of the out their responsibilities to encourage people to engage result; advocating for an electoral code of conduct; and with local dialogue and reconciliation mechanisms, pushing for a commitment from the current president to convey community perspectives to national to roll out peace programmes that are better able to government, and to communicate government respond to the conflict, especially during elections.

A young man tending his livestock. © UN Photo/Evan Schneider

10 Conciliation Resources • Peace and security in the Central African Republic Endnotes

1. Anti-balaka formed from elements of the national 9. This may be due to the severity of the exactions suffered army and local self-defence militias to resist abuses by by the population in and around Bossangoa, or/and their the Séléka and protect the community. They are often sense of vulnerability to a return of ex-Séléka from the associated with Christian communities. Chadian border areas – as noted, Bossangoa is more 2. Séléka was a coalition of predominantly Muslim rebel directly connected to both national and regional dynamics movements, with strong cross-border links to Chad and than the border zone, which may make the population feel Sudan. more vulnerable. 3. In mid-2019 rumours of the return of Muslim 10. In the border zone, these include mechanisms of representatives as part of government-sponsored talks consultation between herders and community leaders, resulted in significant local resistance. Subsequent and transhumance meetings set up by the state to iterations of the talks did allow Muslim leaders to return prevent conflicts. In and around Bossangoa, this included peacefully, and there were several reports that Muslims accepted processes of community identification of conflict are able to circulate in the town without threat. or risk, in some instances through ‘autodefence’ groups, 4. Overall, 547 respondents were men and 459 women. and the communication and/or escalation to the authority Gender was not recorded for 4 surveys. seen as most appropriate to address it – to village chiefs in the first instance, then to more senior officials and, if 5. Life expectancy at birth in CAR is 52 for men and 54 for necessary, the police and Governor. women. https://www.who.int/countries/caf/en/ 11. Some of these differences are no doubt due to different 6. With the exception of Sabewa, a border region with methodological and survey approaches – respondents may Cameroon where there had been a sharp increase. have understood village chiefs and traditional leaders in 7. Ninety-two incidents were reported by the International different ways. In addition, it is not clear how respondents NGO Safety Organisation (INSO) in Bossangoa itself, understood the differences between ‘government’, compared to eight in Korom-Mpoko, seven in Benzambé, different chiefs and local committees. four in Soumbé and none in Ndorombli. 12. This may be because of the limited visibility of NGOs in 8. The Armed Conflict Location & Event Data Project (ACLED) the border areas and rural areas outside Bossangoa lists just seven ‘conflict events’ in and around Bossangoa town. The distant villages of Soumbé and Ndoromboli, for since the end of 2014. By contrast, they list 151 incidents example, were particularly appreciative that Conciliation for Ouham prefecture as a whole from 2015, with more Resources reached out to them. recent violence concentrated in Batangafo – over 100km distance from Bossangoa. An alternative data source, a CAR-focussed conflict mapping project undertaken by the International Peace Information Service, lists significantly more security incidents in the research area, identifying 27 between early 2015 and the end of 2017. None of these incidents, however, involved conflict between two armed groups, or an armed group against the government.

Conciliation Resources • Peace and security in the Central African Republic 11 Conciliation Resources is an international organisation committed to stopping violent conflict and creating more peaceful societies. We work with people impacted by war and violence, bringing diverse voices together to make change that lasts.

Conciliation Resources, Burghley Yard, 106 Burghley Road, London NW5 1AL UK [email protected] +44 (0)20 7359 7728 www.c-r.org CRbuildpeace ConciliationResources

Conciliation Resources is a charity registered in England and Wales (1055436) and a company limited by guarantee registered in England and Wales (03196482)