Competition, patronage and fragmentation

People are affected by different kinds of insecurity in the Ituri Province in the northeastern region of the Democratic Republic of Congo (DRC). This article investigates donor-driven attempts to improve security governance there. More specifically, it investigates bottom-up approaches to security governance in Ituri’s capital of Bunia and in Irumu territory. Whereas in Bunia people are faced with high levels of violent crime, Irumu is the site of a violent conflict between the Ituri Patriotic Resistance Force (FRPI), an armed group connected to the Ngiti community, and the Congolese army. Involving local non-state security actors in security governance is perceived by international and national actors as a pragmatic way to improve security conditions. However, we show that these bottom-up security governance initiatives have not succeeded in resolving the issues that generate insecurity. We argue that this is because the drivers of insecurity in northeastern Congo are translocal and too complex for localised bottom-up approaches to significantly change the status quo.


Introduction
Classic donor approaches to state-building and security governance in conflict and post-conflict contexts usually adhere to a Weberian understanding of the modern western state, characterized by its monopoly on the legitimate use of coercion, its effective assertion of public authority, and its legal-rational bureaucracy. Internationally supported state-building policies tend to aim at rebuilding the institutions of socalled fragile or failed states in the global south in this Weberian image of the modern state. This policy has been driven by the assumption that only by building westernstyle public institutions, would these states be able deliver security, development and other public goods to citizens.
The policies usually entail comprehensive security sector reform (SSR). In most cases, SSR concentrates on reforming the formal arrangements of the state and its security and justice institutions, focusing on clear objectives such as stronger mechanisms of civilian control, parliamentary accountability, budgetary management, and the training and professionalization of police, army and the judiciary (Bagayoko et al. 2016;Schroeder et al. 2014). It is anticipated that these reforms will create security services that both respect stability * University of Copenhagen, DK † Ghent University, BE basic human rights and effectively combat security threats posed by rebel movements, terrorist groups or criminal gangs.

The Turn to the Local in Peacebuilding
Disillusionment with the results of the formalistic and state-centric approach to SSR has provoked criticism and introspection. The failures of top-down state and peacebuilding interventions led to an acknowledgment among policy-makers that the exclusion of local arenas by the central state inhibits state legitimacy. As a result, policy-makers have pushed for exploring the conflict-mitigating potential of local governance (Leonardsson and Rudd 2015: 828). While the UN emphasizes the vital link between local capacity and ownership in peacebuilding processes (UN 2010), DFID and the United States Institute for Peace (USIP) suggest building on already local formal and informal institutions, and supporting systems that strengthen civil society and link traditional authorities with local governance structures (DFID 2010;USIP 2012). Along these lines, it has been argued that non-state security actors could be included as viable partners in the governance of security in fragile contexts. As a result of such narratives, references to local non-state security actors have crept into state-building and SSR toolkits (Bagayoko et. 2016: 2). For instance, while the OECD has called for a 'multi-layered' approach to SSR, which includes local non-state security actors (Scheye and McLean 2006), DFID has argued that 'non-state systems' may play a critical role in restoring security in the aftermath of war (DFID 2004).
Some scholars and NGO activists advocate for an even more profound change of peacebuilding based on local agency (e.g. Autesserre 2010;Boege et al. 2008Boege et al. , 2009Jarstad and Belloni 2012;Mac Ginty and Richmond 2015;Richmond 2010). They criticize mainstream international peacebuilding agendas either for ignoring local agency beyond its rhetorical inclusion in policy papers or seeing it as an instrument for the implementation of liberal peace. SSR, in particular, has been criticized for reinforcing elite interests. This implies that donors may be complicit in deepening structural inequalities, creating insecurity and deepening social divisions (Jackson 2015). Critics argue that attempts to export western-style institutions to conflict and post-conflict situations are doomed to fail because they are perceived as illegitimate locally and differ dramatically from how security is governed on the ground (e.g. Boege et al. 2008;Mac Ginty 2010Richmond and Franks 2009;Scheye 2009). These critiques have paved the way for the 'local turn' in the peacebuilding literature (Paffenholz 2015). It is argued that by ignoring local agency, in spite of rhetoric to the contrary, peacebuilding will necessarily fail because its design rests on externally imposed legitimacy and norms (Richmond 2011: 119). Only by tapping into local forms of peacebuilding, can more self-sustaining forms of peace emerge (Autesserre 2007;Mac Ginty 2010;Richmond and Mitchell 2011;Richmond 2012;Roberts 2011). In the case of the DRC, Séverine Autesserre has recently argued that local matters such as poor access to land, justice, and education, are at the root of Congo's longstanding violence and that these factors should be dealt with by local actors and mechanisms promoting peace and reconciliation (Autesserre 2017).
In recent years, however, some scholars have claimed that the notion of the 'local' as espoused by the peacebuilding literature is problematic. It is argued that whatever is considered local is always made within the dynamics of larger, even global encounters. Furthermore, the notion of 'the local' remains under conceptualised and is often taken for granted, as something ' out there', typically represented as 'indigenous ', ' customary', or 'traditional', to be discovered, understood, or empowered (Autesserre 2010;Richmond 2010;Roberts 2008;Boege et al. 2009;MacGinty 2010;MacGinty 2008;Chopra and Hohe 2004). This is leading to a depoliticisation and reification of the local in peacebuilding literature (Hirblinger and Simons 2015;Hoffmann and Kirk 2013;Sabaratnam 2013: 267).

The Argument
In this article, we contribute to this debate by investigating attempts by international organizations to govern security through local actors in the Ituri province in northeastern DRC. In doing so, it provides an analysis and critique of current international efforts to harness local agency in security governance. More specifically, it investigates 'bottomup' security governance in Ituri's capital of Bunia (section 2), which is facing a high level of violent crime, and in the Irumu territory (section 3), which is the site of a violent conflict between Forces de Résistance Patriotique d'Ituri (Front for Patriotic Resistance of Ituri -FRPI), a militia connected to the Ngiti community, and the Forces Armées de la République Démocratique du Congo (Armed Forces of the Democratic Republic of the Congo -FARDC), which receives support from the UN peacekeeping mission in the Congo (Mission de l'Organisation des Nations Unies pour la Stabilisation en République Démocratique du Congo -United Nations Organization Stabilization Mission in the DR Congo, MONUSCO). 1 Our analysis shows that attempts to harness local agency in security governance is not merely rhetoric by external actors. Rather local non-state actors have increasingly been engaged in efforts to improve the security situation. However, we argue that this engagement with local non-state actors has not led to a durable solution to people's daily security problems, largely because the dynamics fuelling insecurity are not limited to the local level and are too complex for bottom-up approaches to significantly change the status quo. One main argument, therefore, is that while local security actors were empowered by international support, such support challenged existing security actors, which led to intensified competition in the field of security governance, affecting security conditions themselves. Based on our findings, we argue that security governance should not only be understood as a universal public good to be maximised, but rather as a highly contested political issue, which, ultimately, is about who has the right to enforce a certain political order. Security governance is, therefore, always for someone, and not for everyone.

Method
Data for this paper was gathered during a 10-day field trip in September 2015. We make no claim that our study constitutes an in-depth ethnography; yet the authors have extensive research experience in eastern Congo, including in Ituri (see Vlassenroot and Raeymaekers 2004). In total, 26 interviews were conducted with UN agencies, NGOs representatives, and government officials as well as non-state security actors and key informants (civil society activists and university professors). 2 Furthermore, data gathered during the field research was triangulated with relevant academic literature, news articles, official statistics and reports of NGOs and the UN.

Urban Security Governance: The Case of Bunia
Between 1999 and 2003, Ituri was the scene of one of the most horrendous episodes of the Congo Wars that led to the death of more than 55,000 people and displaced hundreds of thousands more (HRW 2003;Vlassenroot and Raeymaekers 2004). The mass-scale violence in 2003 prompted the EU to intervene in support of the UN peacekeeping mission, which was no longer able to manage the situation. One year later, the International Criminal Court (ICC) indicted four of the leaders of Ituri's armed groups. However, violent conflict and insecurity persisted, especially in the rural areas. In 2007, after significant military pressure from the peacekeeping forces and the Congolese army, leaders of the remaining armed groups joined the Congolese army (Fahey, 2013;Tamm 2013). However, in Irumu territory, where the FRPI armed groups continued to operate, everyday security conditions remained precarious. Furthermore, a new series of armed attacks too place between December 2017 and March 2018 in Djugu territory revealing how fragile the security situation is. Djugu was one of the hotbeds of violence during the Ituri conflict, but it has been considered a stable since 2007. Yet in December 2017 Lendu youth started attacking Lendu villages again. More than 100.000 people fled the area in search of security. Among these 40.000 fled into Uganda. The attacks shows that the underlying drivers of conflict have not been resolved and undermine the legitimacy of the Congolese authorities and security services due to their inability protect people. At the same time, it further weakens MONUSCO for its lack of capacity to protect civilians, despite that being a key pillar of its mandate.
Furthermore, in Bunia, where security conditions have considerably improved since 2007, its residents continue to deal with a wide range of interconnected political, military, social and economic issues, which combine to produce an insecure and unpredictable environment (Hoffmann et al. 2016: 5;Pottier 2010: 30). Especially in the peri-urban areas of Bunia, there is a high rate of violence, a rise in drug and alcohol abuse, abusive and extractive authorities, illegal roadblocks, kidnapping and armed robberies (CDJP 2011;CDJP 2010;CDJP 2011b;CDJP 2011c;CDJP 2012a;CDJP 2014;Rapport annuel 2014;UNDP 2012;UNDP 2015b). 3 Another issue that is consistently referred to as a significant source of insecurity is the easy access to small arms, despite different disarmament efforts (UNDP 2012; UNDP 2015b). 4 Additionally, access to security is very unequally distributed. Whereas the wealthy can afford to pay expensive private security companies or privately hire policemen or soldiers for their own protection, 5 the vast majority of the population has to rely on the inefficient and resource-starved Congolese security services or self-organised local security actors, or techniques of self-protection (Hoffmann et al. 2016: 5).

Engaging the Local in Security Governance
The precarious security situation in Bunia in the aftermath of the war prompted concerned international and local organisations to try to improve the security conditions in the city. One of these initiatives was the creation of platforms that engage both local state and non-state security actors in exchanging information and coordinating security governance efforts. At the municipal level, a Local Proximity Security Committee (Comité de Sécurité Locale de Proximité) is organised on a weekly basis. The mayor of Bunia, who is the highest administrative authority in town, heads it. Participants include the neighbourhood chiefs, the Proximity Police (Police de Proximité), elected members of civil society, and international donors. At the neighbourhood level, Neighbourhood Forums (Forums des Quartiers) are regularly held, during which inhabitants can present and discuss their security problems with local authorities and the Proximity Police. Every week, the mayor reports the security situation to the security committee of the Ituri province, in which all the major formal security services are represented (intelligence services, immigration services, army, police, the head of Ituri province, MONUSCO, etc.). There is also the Urban Assembly (Assemblée Urbaine), created by international and local NGOs, which is held every trimester. It convenes state security services (intelligence services, immigration services, army, police, the head of Ituri province) and MONUSCO and local non-state security actors.
All of these platforms are meant to ensure that the different actor's engaged with security governance collaborate in providing security to the residents of Bunia. However, the provision of security is severely hampered not only by problems related to coordination, resources and communication, but also by competition between the different actors involved and the structural factors fuelling insecurity, not least of which are the clientilistic norms prevalent in the Congolese security services. This can be illustrated by the difficulties and challenges faced by two projects intended to harness local agency for the improvement of the security situation in Bunia: The Local Participative Governance Committees (Comités locaux de gouvernance participative), which are funded by two international NGOs, the Diocesan Justice and Peace Commission (Caritas) and IKV Pax Christi; and the Proximity Police units, which are funded by the UNDP in Bunia.

The Local Participative Governance Committees
The Local Participative Governance Committees were created in 2010 as an initiative of Diocesan Justice and Peace Commission and IKV Pax Christi, in partnership with Congolese NGO network Reseau Haki na Amani (Reconciliation and Peace Network). The primary objectives of the Local Participative Governance Committees were to engage youth in crime prevention in Bunia's 12 neighbourhoods, 6 and to create a forum through which its youth could regularly meet with the formal security services (army, police, immigration and intelligence services) and politico-administrative authorities (Hoffmann et al. 2016: 7).
By engaging the youth in crime prevention, the involved NGOs sought to simultaneously harness their potential as security providers and transform vigilante groups into non-violent security actors (CDJP 2012b). 7 Additionally, they hoped to create synergies between existing youth organisations, including confessional youth groups, the association of taxi drivers (many of whom are former militia members), art groups, street kids (called shegue and maibobo), and political party youth groups. 8 This was seen as a way to avoid conflict and competition between these groups. One of the important roles of the Local Participative Governance Committees was to produce improved documentation on the local security situation, as the government system was not working due to a lack of means and unmotivated local officials. 9 However, the project created suspicion among the security services, who thought that the Local Participative Governance Committees had stepped into their domain and were concerned that they would expose and denounce their illegal revenue-generating activities. 10 These concerns must be seen in light of the clientilistic norms prevalent in the Congolese security services. Authority does not simply follow the formal hierarchy in the Congolese security services. Rather, the latter consist of a collection of competing patron-client networks. These networks can be based on different relationships such as shared ethnic or geographical origins, past army unit affiliation, or former armed group membership. Forced to serve their patrons and living in precarious conditions themselves, security agents constantly try to exploit the benefits of their position to illegally collect resources from civilians (a practice that usually involves an element of coercion), or through forging profitable links with criminal actors. This creates a context of constant unpredictability about the conduct of these agents and tends to result in increased insecurity for Congolese civilians (Eriksson Baaz and Verweijen 2013). A striking example of how security forces' illegal revenue-generating activities fuel insecurity in Bunia took place when Colonel Bonane Habarugira was deployed as deputy commander between 2011 and 2013. 11 During this period, urban crime levels increased (UNDP 2015b) with Bonane being reported as sponsoring criminal networks in Bunia, including by providing weapons (UNSC 2014).
Partly as result of the spike in insecurity during Bonane's deployment, the Local Participative Governance Committees began to reinforce and support existing local community alert systems in 2012. 12 These committees provided neighbourhood inhabitants with megaphones, whistles, cans, and alarm bells so they could alert the youth and security services when incidents occurred. As mentioned, the project also sought to improve the strenuous relationship between the youth and their neighbourhoods, on the one hand, and the Congolese security forces and politico-administrative authorities, on the other. This was done through the creation of an Urban Assembly. During its meetings, members of the Local Participative Governance Committees would try to persuade the police to intensify their patrols in the most problematic areas, as indicated by collected statistics. However, youth have expressed fear of reporting security incidents to local authorities, especially in cases where people believed that the police or army were involved, because 'you never know who is invited to the Urban Assembly meetings'. 13 This project has produced some positive effects on local security governance: it has led to an improved documentation of security incidents; created a platform through which local communities can approach the Congolese security forces and authorities; and helped to support an existing alert system. After some time, Congolese state authorities also began to appreciate the collaboration with the Local Participative Governance Committees, which provided them with valuable information.
Despite these effects, the relationship between local youth and the police remains tense. Distrust in the police continues to be high, not only because they often do not show up when a violent crime has been committed, but even more so because they are believed to be complicit in acts of crime. According to different sources, police officers rent their guns to bandits and take part in armed robberies. 14 At the same time, security forces continue to be engaged in illegal revenue-generating activities. Moreover, the police can easily be bribed, which means that while criminals are often let go, innocent people are arrested, or worse, and police officers impose self-invented infractions, fees and fines on people. As a result, some youth groups have unsuccessfully requested that the police no longer enter their neighbourhoods after sunset, unless they receive permission from their superiors, and then also only if the youth are allowed to participate in the patrols.
Such dynamics create a general sense of distrust and suspicion by urban inhabitants towards the security services. 15 According to a recent survey, 36 per cent of people in the area believed that the police protected the population in Bunia, 24 per cent thought that they did nothing, 33 per cent believed that they were involved in crime, and 7 per cent were not sure (Hoffmann et al. 2016: 9). 16 The alleged passivity and complicity of the Congolese security forces has also encouraged some local youth to revert to their former roles as vigilantes. In such cases, revenge or deterrence could lead them to hurting or even killing people. In rare cases, youth have even attacked police officers or soldiers, who they suspect of involvement with other violent incidents. However, vigilante justice is risky as people from the victims' social network may take revenge and trigger vicious cycles of tit-for-tat violence. 17 Furthermore, the Local Participative Governance Committees do not represent all youth in the city. Many youth do not acknowledge them as legitimate security actors and sometimes confuse them with state security actors or see them as complicit with the latter. 18

The Proximity Police
An important new actor in urban security provision in Bunia is the Proximity Police (Police de proximité). It was initiated in 2009 as part of a larger GDP 60 million police reform programme (the Security Sector Accountability and Police Reform, SSAPR), funded principally by the UK and the European Commission (Boshoff et al. 2010;Thill et al 2018). The objective of the intervention was to transform the Congolese police into a Police de Proximité force capable of providing protection to the Congolese population through regular popular consultation and more direct contact with civilians. It aimed at strengthening police-community partnerships and turn police forces into a more accountable institution. However, from the outset senior police chiefs and members of the government showed little interest in reforming the police institutions that are considered important instruments of power and of resource generation, and thus kept under the firm control of those profiting from them (Boshoff et al. 2010: 14). 19 For donors supporting police reform, the Proximity Police is meant to be a guiding philosophy to be applied for the entire Congolese National Police (Police Nationale Congolaise). With its creation, donors sought to increase trust between the population and the police and improve communal security by harnessing local energies. 20 Pilot projects were launched in four major cities around the DRC: Matadi (Bas-Congo), Bukavu (South Kivu), Kinshasa, Kananga (Western Kasai) and Bunia (Ituri). For Bunia, a total of 260 civilians have been recruited and trained since September 2013 (PNUD 2015). Following a sensitisation campaign carried out by local NGOs, these recruits were deployed in five of the city's most insecure neighbourhoods in 2014 (PNUD 2015). 21 To harness local energies in improving security provision, the UNDP and the Congolese authorities also created Neighbourhood Forums (Forum des Quartiers). In these forums, civilians are provided the opportunity to discuss their security problems with local authorities and the Proximity Police. The Proximity Police can then relay these problems to the municipal level security committee.
However, the Proximity Police's capacity to improve security in Bunia is hampered for several reasons. Firstly, these newly trained police officers lack sufficient resources to cover their operating costs, despite donor support. Secondly, when problems arose with the timely payment of salaries, Proximity Police officers started quitting their jobs. Thirdly, since the Proximity Police are unarmed, they are faced with considerable constraints when intervening against armed bandits (UNDP 2015a). Fourthly, the creation of the Proximity Police has produced new competition within the already deeply divided Congolese National Police. Having received training, new equipment and buildings, Proximity Police officers faced resentment from some of their colleagues in the police force. 22 Finally, concerns rose about the sustainability of the positive conduct of the Proximity Police after donor funding was no longer secured in 2015. It was feared that the Proximity Police would in the end become an additional source of coercion, developing similar practices to other security services.
Soon after its deployment, the Bunia population became disappointed with the Proximity Police, accusing them of gradually taking over the 'bad habits' (corruption, extortion, passivity etc.) of their colleagues in the police force. 23 The critique of the Proximity Police in Bunia also affects people's attitudes towards donors, who, along with the government, are accused of being responsible for increased competition within the police and for the lack of improved security conditions (Hoffmann et al. 2016: 11-12). These conclusions are in line with dynamics observed in Bukavu, where police reform prematurely ended following donor reactions against the brutal repression of youth gangs by police forces in Kinshasa as part of operation Likofi in 2013 and 2014. Thill et al (2018) argue that patronage structures guiding revenue generation activities and loyalty structures within the police, as well as the contradiction between the democratic and preventative principles of this reform process and the repressive force needed for regime security, are among the main reasons why security reform has made limited progress. As they state, ' despite some successes, ambitious donor-funded police reform programmes have not sufficiently taken into account the political economy of the police and how their intended reform initiatives may impact institutional incentives and hidden interests' (Thill et al 2018).
The above shows that international donors' engagement with local actors in security governance is not just rhetoric. They have tried alternative ways to provide security to Bunia's inhabitants, have funded existing non-state actors through the Local Participative Governance Committees and have funded and pushed for the creation of the Proximity Police to respond to local security needs. Yet, our research indicates that the effects of these projects have been rather limited, as they do not address the complex of interconnected social, political and economic factors that combine to produce urban insecurity. The Local Participative Governance Committees and the Proximity Police do not act in a vacuum, but are part of a larger political and social context, conditioned by a political economy that itself is characterised by patronage politics and high levels of structural violence, and that informs and guides the conduct of security actors. Even if they had been fully resourced and received the best training available, these police forces would still be subjected to the pressures and constraints of the larger context marked by poverty, violent crime, political tension and fierce competition over power and (limited) resources. Similarly, Eriksson Baaz and Ohlsen (2011), who have studied unofficial economic activities within the Congolese police, conclude that an increase in logistical capacity and better salaries are no guarantee that security services will be transformed into less violent and predatory actors.
Even more, supporting such actors challenges the coercive authority of existing state security actors, who benefit from the status quo. In the end, this generates further competition and even resistance in the field of security governance and hampers the provision of security services. In other words, the potential of such 'bottom-up' approaches to improve security provision in urban contexts can only be understood when taking into account the broader context. It explains why such strategies have unpredictable political effects. This is particularly the case in areas such as northeastern DRC, where political and coercive authority is deeply fragmented and contested (Hoffmann et al. 2016: 12).

Bottom-Up Security Governance in the Context of Violent Conflict in Irumu Territory
Different initiatives to improve security conditions by mobilising local non-state actors have also been developed in Ituri's rural areas.
In this section, we outline how MONUSCO has engaged with local non-state actors to improve security for civilians. MONUSCO has a broad mandate to simultaneously protect the civilian population, neutralise armed groups, stabilise the country and restore state authority. Aligning these mandates has proven difficult. MONUSCO has provided critical lessons on peacekeeping for the UN Security Council and the Department of Peacekeeping Operations, which in recent publications recognise the importance of local conflict drivers and the reinforcement of local capacities to tackle insecurity (UN 2015;UNSC 2015b). As a result of these lessons and critiques, MONUSCO has gradually transformed into a vehicle for the development of new approaches to peacekeeping, including supporting and working with local non-state actors (Stearns 2015).
The FRPI is the last remaining armed group from the Ituri war. The reasons for its existence and its modes of operation are similar to a number of armed groups operating in eastern Congo. Its authority is drawn from the historical grievances of the Ngiti community, which it claims to protect against neighbouring communities and the government (cf. Hoffmann and Vlassenroot 2014). However, at the same time, it is also deeply involved in acts of extortion and abuse and contributes to the creation of insecurity. In other words, it is an armed group, which gradually evolved from a community protection force to a source of insecurity.

The Failure of Carrot and Stick Solutions
A combination of negotiation and force has been used to counter the FRPI. Negotiations between the group and Congolese authorities took place between November 2014 and January 2015 and again in May-June 2015. As in most negotiations with armed groups, the talks with the FRPI focused on issues such as the recognition of military ranks, amnesty, 24 and one-off payments (UNSC, 2015a). Talks in January 2015 failed because of the arrest of the FRPI leader, Cobra Matata. In June 2015, additional demands from the FRPI and its refusal to deliver military equipment put an end to the negotiations. This led to new clashes between the FRPI and the Congolese armed forces around Aveba in the Walendu-Bindi chiefdom. In June 2015, the FARDC and MONUSCO changed strategies and started carrying out joint operations against the FRPI (Hoffmann et al. 2016: 14).
Despite the military operations and negotiations towards demobilisation, the FRPI remains a key actor in the politico-military landscape of Ituri and the Irumu territory. Today, however, it is highly fragmented and no longer able to control significant territory. Previously, the group had its own taxation system for generating revenue, but now it is increasingly relying on robberies, looting and extortion, particularly during market-days and along main roads (UNSC 2015a).
The group's relation with local society is ambiguous. On the one hand, it tries to coerce local authorities to obey them while also abusing and extorting ordinary people. 25 On the other hand, the group is still considered by residents as a protection force of the Ngiti community, which has historically faced marginalisation and exclusion. They are perceived as ' children of the community' and local residents are hesitant to collaborate with local authorities in combatting them. A UN community liaison assistant told us that the local population often tell them that the FRPI fighters are ' children of the village, (so) we cannot ask the soldiers to go and shoot them'. 26 Furthermore, it is difficult to distinguish combatants from civilians and local civilian specialists of traditional medicine, called 'lenga na kisi' are among the group's leaders. 27 Also some customary chiefs are in direct and constant contact with the armed group. 28 Sources claim that neither Kinshasa nor the FRPI leadership ever really intended to reach an agreement. It is argued that for the fragmented FRPI leadership, these talks were considered as an opportunity to get access to food and other resources. 29 Others point to the provision of support to the FRPI by political and community leaders from the area, including members of parliament, and attempts to prevent a further demobilization of the group because of its strategic importance as a reserve force. 30 A local reserve force can potentially be utilised by political actors as leverage during negotiations and to mobilise support for the Irumu territory. 31 Sources told the United Nations Group of Experts that the Congolese government does not want to integrate the FRPI into the Congolese army (UNSC, 2015a). Kinshasa is believed to want to dismantle the armed group to demonstrate to the international community that maintaining stability is mainly a law-and-order issue (UNSC, 2015a). 32 At the same time, rumours are circulating that certain officers in the Congolese army are not interested in finding a solution to the problem as the operations against the FRPI increase their budgets and provide access to income derived from informal taxation of the local population (Hoffmann et al. 2016: 14). These allegations have a history. Already in 2010, unspecified Congolese authorities, told a UN Group of Experts that the FARDC was involved in taxing gold mining in Geti in the FRPI heartland (UNSC 2010: 67).
The deployment of units from the Congolese armed forces has aggravated insecurity at the local level. The Congolese army is accused of extorting and abusing populations in the areas it cleared in South Irumu, which is a pattern that goes back to earlier operations against the FRPI (Justice Plus 2007). 33 Subjected to abuse and extortion by the army as well as by the FRPI, the population in Irumu is trapped between a rock and a hard place (see also Suarez 2017). Human rights organisations have developed a number of protection strategies and tried to sensitize the Congolese army, but they have had a limited effect on the conduct of the FARDC. 34 This underscores MONUSCO's difficulties in aligning the different objectives of its mandate; particularly, the mandates to protect civilians, restore the authority of the state and neutralise armed groups. For their part, FRPI members who want to demobilise are caught in a dilemma. While they fear what the Congolese soldiers might do to them if they surrender to the Congolese armed forces, they also fear being killed by their own commanders if they would be caught trying to leave the group. 35 This all said, the violence in Irumu, and the stakes of those involved, are further evidence of how local and supra-local dynamics are closely interlinked and how security issues cover a wide variety of dynamics and processes dividing elites and populations.
A similar conclusion can be drawn from the 2018 upsurge of violence in the Djugu territory, where Lendu youth started attacking Hema villages. Statements by local and national authorities and rumours about hidden support have prompted local sources to suggest that the attacks have been carried out to exploit existing tensions between local communities due to intensified political competition linked inter alia to the presence of oil deposits in the province and the prospects of elections (Mahamba and Sengenya 2018). While no firm evidence has been brought forth in support of these claims, they fuel popular perceptions and point to the lack of popular trust in political elites and state authorities.

MONUSCOs Bottom-Up Approaches and their Limits
Besides increasing the military capacity of the FARDC in Irumu territory, MONUSCO has deployed initiatives to strengthen local capacities for security governance. To fulfil its mandate to protect the civilian population, MONUSCO has created several non-military approaches to mobilize local actors and strengthen local conflict prevention capacities. In 2009, it launched Joint Protection Teams involving several sections of the mission. These teams gathered information about security threats and socio-economic conditions on the ground to produce recommendations for MONUSCO and the Congolese authorities. Even though military protection was provided, it proved difficult to collect information in remote areas because of a lack of contact points and poor communication infrastructure. Another strategy was the creation and support of Community Alert Networks. This project aimed to improve communication and relations between MONUSCO and local populations in Irumu through a network of Community Liaison Assistants, which are in contact with community leaders. In case of an emergency, community leaders contact the liaison assistant, who, in turn, inform MONUSCO of the security situation. Mobile phones are used to pass on the alerts. However, even with the information on immediate security threats, MONUSCO has struggled to mobilise the necessary (military) responses, particularly in remote areas. Moreover, local populations and leaders were afraid to be associated with the Community Liaison Assistants for fear of reprisals by the FRPI. According to one Community Liaison Assistant: 'People are afraid of talking. It is only through telephone. Even the village chiefs are afraid. If you pass by, they are immediately suspected by the FRPI. 36 Another protection mechanism initiated and supported by MONUSCO were the local protection committees, which were being created in Irumu in September 2015, but were already operational elsewhere in Ituri. These committees consisted of local authorities, including customary leaders, who were tasked with preparing community protection plans. The main objective of this project was to create local capacity and ownership of security governance and to transfer competences to local actors so they could help to improve security conditions. 37 However, the effects of these initiatives were limited by security conditions and conflict dynamics in Irumu as well as by Kinshasa's security policies.
The Congolese regime considers the presence of armed groups a matter of national security and sovereignty. As a result, MONUSCO has been politically sidelined on this issue (De Vries 2015;Boshoff et al. 2010). Yet, as is the case with several other community-based militias, the FRPI militia is not only a security problem, but also a political issue related to the political and socio-economic marginalisation of the Ngiti community. In the context of violent conflict in south Irumu, it is doubtful that MONUSCO's bottom-up initiatives will prove to be a solution to people's complex security problems. The extractive and abusive behaviour of the Congolese armed forces, the strong ties between the FRPI and the populations in Walendu-Bindi, and the persistence of political grievances among the Ngiti render it unlikely that basic security will be restored in south Irumu in the short-term. Military operations have only further escalated the conflict and have had an adverse impact on the security situation. For its part, MONUSCO's effectiveness is crippled by restrictive operational and security rules. Armed convoys are required to escort staff to areas where security incidents have already occurred, limiting their flexibility and response speed. This tends to alienate and anger local populations in Irumu. 38 People on the ground are unaware of the security restrictions and, in some cases, perceive the mission as complicit with armed groups. 39 This not only raises fundamental questions about the use of violence to deal with armed groups, but also MONSUCO's support to the Congolese army more broadly. These conditions also point to the limits of the current development of locally-rooted protection mechanisms, which seem to be disconnected from the larger politico-military power struggles of the region and potentially unfit for solving the deeper causes of insecurity.

Conclusion
Insecurity and violence remains part of everyday life in Ituri. To improve security governance, international actors have tried to engage with local state and non-state security actors. In doing so, international interventions have tried to strike a balance between supporting a regime and its security services with a well-documented history of human rights abuses, and empowering non-state security actors based on the idea that these actors are considered more legitimate by the population. However, there is reason to be sceptical that this kind of bottom-up security governance is a solution to people's immediate physical security needs. In Ituri, existing security actors tend to see the involvement of new actors in security governance as a challenge to their authority and in certain cases as a potential threat to their income-generating activities. As international support to non-state security actors tend to directly impact the distribution of power and resources, they are likely to create reluctance, suspicion and resistance from state security actors benefitting from the status quo. This suggests that security governance is concerned with questions about who has the right to enforce a certain political order and to whom the benefits of security provision accrue.
The suspicious attitude of the Congolese security forces towards the Local Participative Governance Committees and even their colleagues in the Proximity Police should be understood in this light. The context of violent conflict in Irumu implied that local populations and leaders were afraid to be seen with the Community Liaison Assistants. International support to local security actors should, therefore, not be understood as the provision of a universal public good. Engaging with local security actors creates competition and suspicion as well as collaboration in the search for common solutions to shared security problems.
Our research also suggests that bottom-up approaches to security are unlikely to fundamentally improve people's security conditions in a sustainable manner if the systemic causes of insecurity -including ethnic tensions, conflicts over natural resources, lack of state authority and capacity and the clientilistic modus operandi of the Congolese security forces -continue to undermine the creation and support of local security governance mechanisms. It also underscores MONUSCO's difficulties in aligning the different objectives of its mandate; particularly, the mandates to protect civilians, restore the authority of the state and neutralise armed groups.
Our research findings also align with other emergent critical investigations of the local turn in peacebuilding showing that it should not be assumed a priori that bottom-up security governance is better able to meet the security needs of local populations (e.g. Bagayako et al. 2016;Hoffmann and Kirk 2013;Meagher 2012;Wiuff 2016). Instead, these initiatives become embroiled with complex and deeply-rooted translocal security issues, which they are ill-equipped to tackle. We, therefore, suggest that future research endeavours to study security governance systematically as a social space within which actors vie to establish authority over resources and people. This social space is not defined by scale (i.e. the local vs. nonlocal), but by social relations between different actors and networks connecting them. When international actors support one or more actors in security governance they make a political choice, rather than simply providing a public good.  Interview, university assistant, Bunia, 24 September 2015. 18 For instance, during a burial in Saïo neighbourhood in October 2010, the young gravediggers, who were allegedly drunk, dropped the body in to the grave. The neighbourhood's state based Youth Committee (Comité des jeunes) and the family of the deceased eventually sued the gravediggers. In solidarity with the gravediggers many youth from the neighbourhood refused to recognise any youth organisations including