Disarmament, Demobilization, and Re-integration (DDR) programs are today often implemented in contexts of multi-party civil war, where more than two armed groups are engaged in armed violence (Cockayne and O’Neil 2015; Muggah and O’Donnell 2015). The fragmentation of non-state armed groups is a common feature of these contexts, sometimes occurring as hardline or opportunistic factions break away from those willing to keep faith with a peace process (Christia 2012: 9, Greenhill and Major 2006/7; Stedman 1997). The DDR process in the Democratic Republic of Congo (DRC) exemplifies this pattern, as numerous non-state armed groups in the DRC have fragmented, with some recalcitrant factions forming new groups and fighting on even as others demobilized their troops or integrated into the national army.
Although the fragmentation of armed groups has occurred in almost 45 per cent of civil wars taking place since 1989 (Findley and Rudloff 2012), little operational guidance has been given to practitioners on the specificities of implementing DDR programs in settings where some groups opt in to an ongoing peace process and others opt out. Indeed, neither the United Nations Integrated DDR Standards (IDDRS) nor the African Union’s Operational Guidance Notes on DDR provide instruction on this issue (UN 2014; AU 2014). Ongoing conflict associated with the fragmentation of armed groups, however, creates specific challenges for DDR. As this Practice Note indicates, fragmentation may prompt desertion among combatants who do not wish to fight on within new breakaway groups. These deserters may subsequently report to DDR programs that, in settings of ongoing conflict, are likely to be monitored by groups outside of the peace process. The remobilization of ex-combatants may follow, either as former fighters are targeted for forced re-recruitment by active armed groups, or as they try to escape this fate by choosing a new armed group to join for protection.
The Practice Note begins with definitions of relevant concepts and a brief explanation of the research methodology. The following section introduces two non-state armed groups previously active in eastern DRC – the Rally for Congolese Democracy-Goma (RCD-Goma) and the National Congress for the Defense of the People (CNDP). These groups were selected for analysis due to the fact that CNDP formed and remained active even as RCD-Goma’s fighters integrated into the national army of the DRC and then demobilized and returned to civilian life. Drawing on ex-combatant interviews, the Practice Note describes the military organization of these two groups and parses the linkages between armed group fracturing, desertion, and demobilization. The concluding section provides policy guidance for practitioners tasked with implementing DDR in settings of partial peace.
In its traditional sense, DDR refers to a process in which fighters disarm by handing over their weapons and/or ammunition. Demobilization follows and occurs when combatants delink themselves from military command structures through a process of “formal and controlled discharge from armed forces or other armed groups” (UN 2014, 4.20: 25). Unlike desertion, which occurs when fighters run away from their armed groups without the permission of a superior military officer, demobilization is authorized by senior armed group members. It is also followed by reintegration, which is the “process by which ex-combatants acquire civilian status and gain sustainable employment and income” (UN 2014, 4.30: 2). Although a setting of partial peace is likely to affect all three components of a DDR program, this Practice Note concentrates on the impact of ongoing conflict on demobilization and reintegration, and does not fully address disarmament.
Information is taken from interviews with 32 former members of RCD-Goma and 14 former members of CNDP.1 A further 42 interviews were also conducted with former members of the Armed Forces of the DRC (FARDC), many of whom had previously served in various non-state armed groups. The interviews were conducted in 2011 and 2012, and interviewees were accessed using a chain-referral “snowball” sampling technique (Cohen and Arieli 2011). Snowball sampling can sometimes generate a collection of interview responses that are relatively alike, particularly when interviewees refer the interviewer to friends and relatives with similar experiences. To counter this, a diverse array of individuals were sought out to begin different chains of referrals. These individuals were accessed via ex-combatant associations, reintegration projects, and referrals from civilians across the city of Goma in North Kivu Province. The interviews were then transcribed and different responses to similar interview questions – concerning group fragmentation, demobilization and desertion – were compared.2
The origins of RCD-Goma begin with its predecessor group, the Rally for Congolese Democracy (RCD). RCD participated in the “Second Congo War” beginning in 1998 and aimed to remove the DRC’s President Laurent Kabila from power. Kabila had assumed the presidency earlier in 1997 after leading a Rwandan-backed rebellion which toppled President Mobutu. However, once in charge, Kabila soon turned against his sponsors, ordering Rwandan soldiers to leave the DRC. Drawn together by the common aim of ousting Kabila, RCD was a coalition of otherwise disparate interests, and cracks in the alliance began to emerge in May 1999 when key RCD officials announced they were removing Ernest Wamba dia Wamba, the group’s leader. Wamba subsequently began a new group, RCD-Kisangani, while the remainder of RCD became known as RCD-Goma. Although RCD-Goma failed to swiftly remove Kabila, the group went on to control a large zone of territory in eastern DRC, including both North and South Kivu.
Following a series of negotiations and peace agreements throughout the Second Congo War, the Global and Inclusive Peace Agreement was signed in 2002. Subsequent agreements to implement this peace accord committed RCD-Goma and other signatory armed groups to integrate into a reformed national army, henceforth known as the Armed Forces of the DRC (FARDC). Combatants who integrated into the army were to be given the choice to stay or, alternatively, to demobilize and return to civilian life. However, in September 2003, as the integration of FARDC’s new chain of command got underway, RCD-Goma’s divisional commander for North Kivu, Laurent Nkunda, refused to cooperate. He subsequently formed a new armed group, later known as CNDP, from three RCD-Goma brigades under his command, the 81st, 82nd and 83rd brigades in Masisi and Walikale (Stearns 2012: 20). Nkunda’s defection was allegedly ordered by Rwanda, although he also stressed how RCD-Goma had done poorly out of the peace deal (ibid: 19).
In terms of basic military organization, RCD-Goma and CNDP had much in common. Although technically non-state armed groups, both were conventionally organized due, in part, to the influence of Rwanda’s national Patriotic Army (RPA). RPA soldiers formed and participated in RCD-Goma, and a large number of RPA-trained RCD-Goma members also went on to form CNDP’s senior command (Stearns 2012: 11). As a result, both groups had similar conventionally-inspired mechanisms in place to prevent desertion among their fighters, including a military intelligence officer for each brigade and battalion. These intelligence officers were part of a broader executive staff section for intelligence (the S2 department), and could sometimes go out into the community to gather information on a deserter’s whereabouts. Brigade and battalion level S2 officers also appointed intelligence agents at the lower levels of the company, platoon, and section in both RCD-Goma and CNDP. These individuals, known as “Intelligence Security” (I.S) agents, were often unknown among the troops but were expected to watch over the rank-and-file and report signs of potential desertion to higher level intelligence officers. As former members of CNDP recalled, the role of I.S agents was to “know military intelligence, and the state of military units, including desertions.”3 In the case of a low-level desertion in CNDP, “The S2 would then meet with his collaborators, the I.S agents at the company level, and study the case with them.”4
Desertion was a punishable offense in both RCD-Goma and CNDP and those who made a run for it could be tracked down and picked up. As a former adjutant in RCD-Goma explained,
Not only were deserters pursued, but once the runaway was caught he would be executed in front of everyone in order to scare the rest of the fighters.5
Another former member of RCD-Goma also reported how,
I ran away [from RCD-Goma] and went home. When RCD-Goma found out they sent someone to assassinate my parents. They forced me to go back and I spent one month in prison. I accepted to stay in RCD-Goma, instead of death.6
Others stated that punishment for desertion could vary, particularly if you were caught during a first-time offense. As a former CNDP member stated,
If you are caught and you are lucky, you could just get a heavy punishment, but if you have a bad commander, without scruples, you risk being condemned to death and executed.7
A former S2 officer from the group corroborated this, stating that,
I have known some deserters to be executed and some to be placed in prisons underground known as ‘Andac.’ Sometimes these prisons collapsed in landslides which killed the prisoners.8
As these quotations indicate, if low-level combatants in RCD-Goma and CNDP wanted to leave their group then they had to desert or, alternatively, wait for a formal program of DDR in which their official demobilization would be authorized by their military hierarchy. As will be seen in the remainder of this Practice Note, however, the fact that CNDP remained active and outside of the peace process meant that both options (of desertion and/or demobilization), left some ex-combatants at risk of recapture.
During the army integration process following the 2002 Global and Inclusive Agreement, many low-level members of RCD-Goma entered FARDC (the new national army of the DRC) and then subsequently demobilized and returned to civilian life. However, low-level combatants within brigades under Laurent Nkunda’s control had little opportunity to choose whether or not they wanted to continue fighting within the new group, CNDP, or whether they wanted to join the program of integration and/or demobilization with the rest of their (former) group, RCD-Goma. For example, ex-combatants who served directly under Nkunda’s command in RCD-Goma stated that they simply continued to follow orders and consequently found themselves in CNDP. As a former RCD-Goma captain remembered,
There was integration in the national army but we followed the movement of Mutebutsi and Laurent Nkunda, first to Bukavu in South Kivu, and then we came here to Goma. We went to Rutshuru with Colonel Makenga and then we had to go to Kitchanga, but without knowing the aim of the situation. We found ourselves away from the government. The government remained in Goma and we were elsewhere. We learned from our hierarchy that we were CNDP. They were the ones who knew the reasons for the change.9
However, not all low-level combatants were content to join CNDP once they realized a new group was being formed in the midst of an ongoing army integration and demobilization process. As a former member of Nkunda’s guard in RCD-Goma explained,
Nkunda took refuge in Kitchanga and made his fief. Certain of his close aides fled with their arms during the night. They presented themselves to MONUC [the United Nations Mission in the DRC] at Mubambiro for demobilization.10
A soldier who served in the Congolese Armed Forces (FAC) prior to the creation of FARDC and throughout the post-2002 army integration process also explained,
When FAC and RCD were mixed [to form FARDC], we hadn’t understood that among us there were individuals who wanted to join CNDP. The kadago (child soldiers) from RCD-Goma told us that they wanted to reinforce the troops of Laurent Nkunda. When President Joseph Kabila wanted to visit Goma, he asked all troops to go to Tongo, on the side of Rutshuru, which was in the immediate vicinity of CNDP, because he didn’t want to see any soldiers in Goma. When MONUC came to transport us to Tongo, I fled on a motorcycle to a DDR office. I demobilized because I did not want to become a CNDP rebel.11
While CNDP was initially open to Hutu, Tutsi, and other communities, 80 per cent of CNDP’s senior commanders were Tutsi and the group became increasingly ‘Tutsified’ over time, particularly as Hutu began to desert to the ongoing army integration and DDR program (Stearns 2008: 260; Stearns 2012: 11). The first senior Hutu commander to leave CNDP was Colonel David Rugayi who, in 2004, was appointed major of FARDC’s 83rd brigade – a former RCD-Goma unit that had yet to report for army integration. Rugayi left CNDP with over 1,400 of his troops in 2005 while Colonel Smith Gihanga, the 81st brigade commander and a fellow Hutu, followed several months later in 2006 (Stearns 2012: 24). A former battalion-level S2 officer in CNDP also spoke of these developments,
Actually we recruited mainly among two tribes – the Hutu and the Tutsi – and indiscriminately among children and adults. At one point the Hutu started to desert, some went to join PARECO [an active non-state armed group with a large Hutu faction] and they took their arms and munitions. Finally CNDP stopped accepting PARECO combatants and Hutu combatants, and only recruited among the Tutsi.12
Similar dynamics were also observed much later in the DRC, after CNDP eventually integrated into FARDC in early 2009. In this case ex-CNDP combatants, now inside the army but disgruntled with the terms of their integration, formed a new armed group known as the March 23 Movement (M23). Just as had occurred when Nkunda formed CNDP, low-level soldiers who did not share their commander’s support for the new group began to desert. As a former member of FARDC explained,
After integration, I was not happy with our new ex-CNDP commander … He created divisions between us and rumours started to circulate about the creation of a new movement, in this case, M23. We were in full combat operations around Mushake when the brigade commander gave orders to eliminate some of us. So I decided to desert during the night. During our meetings with the brigade commander, certain ex-CNDP combatants split off to discuss the establishment of a new movement. Even though we were not included, some of our colleagues, who were part of the discussions, told us and asked us to adhere to the new movement. We found this suspect and we started to become scared, because they were all our former enemies. So we planned to desert. One day, our brigade commander who was Tutsi was on a trip. His second, a Congolese commander, informed his men to adhere to the new movement. Those who were not for this movement started to flee.13
As the above quotations indicate, low-level combatants who did not want to be incorporated into new groups outside of the ongoing peace process in DRC sometimes opted to desert and report to DDR. The fact that CNDP also served a much narrower ethnic constituency than the broad-based RCD-Goma also contributed to desertion among those who did not fit the narrower profile, and who subsequently went over to the army integration/DDR process or joined other non-state armed groups whose goals were closer to their own. As seen with the formation of CNDP, when groups are organized similarly to national armies, the decision to fragment is likely to be made by high-level leadership without consultation of the rank-and-file. Group fragmentation may therefore serve as a critical juncture at which point it may be possible to encourage the desertion and subsequent DDR of low-level combatants, not least because some of those who find themselves in spoiler groups will do so purely because of a decision made by their superiors.
Those who deserted and/or demobilized from RCD-Goma and CNDP did not always stay in civilian life. Indeed, as CNDP remained active, many ex-combatants feared and experienced recapture by CNDP combatants. As a demobilized former RCD-Goma combatant explained,
We [ex-combatants] returned to our villages in Masisi [a CNDP stronghold] but we were pursued by soldiers from CNDP and [so] we came to live here [in government controlled Goma]…CNDP followed the demobilization process closely. They knew where the demobilized went. We were constantly denounced.14
Another demobilized RCD-Goma combatant shared these concerns:
When I received the USD 100 [from the DDR program] it was at the same time that CNDP became autonomous. CNDP said that ‘we will arrest and kill the demobilized that are going out and abandoning us on the field of combat, because they are giving away our arms.’ It was then that I was afraid to go back to Masisi.15
A former member of CNDP stated similarly,
At first I supported the group [CNDP], until I realized that there were many deaths, victims of the war and this bad life we led. One night I decided to desert and since my desertion I have never returned to my home village of Mweso [a predominantly CNDP-controlled zone]. I’m scared of being captured or denounced.16
While the above interviewees were able to evade recapture by staying in government-controlled areas away from CNDP, others who returned home to CNDP strongholds were often not so fortunate. As a demobilized former member of RCD-Goma, later recaptured and taken into CNDP reported,
We quit RCD-Goma with the help of MONUC which demobilized us. I was based at Ngungu when MONUC came. We were six soldiers and we decided to desert [from RCD-Goma] together. My comrades were caught by our pursuers and we exchanged fire. Me, I saved myself and hid in the forest until a MONUC truck picked me up. The others were caught. The soldiers loyal to CNDP came to the MONUC camp to get me but MONUC took me to Goma to be demobilized. As I knew that there was peace after the war in 2006 I returned to my village [in CNDP’s zone of influence in Masisi]. The FARDC soldiers there were deployed to [a different area], Bukavu, and then, [when they were gone], the CNDP soldiers came to get me. They took me to Kitchanga where I stayed for five months.17
The above quotations highlight that active armed groups may follow ongoing demobilization processes and attempt to forcibly recapture ex-combatants who return to areas under the group’s influence or control. Some ex-combatant interviewees also indicated that they were particularly attractive targets for CNDP because they were known to have been provided with DDR assistance, often in the form of cash.18 As also outlined above however, ex-combatants could reduce their odds of recapture by settling in areas away from armed groups and their informants. Cantonment sites protected by UN peacekeepers also provided safe sites from which to evade recapture, as did both state and non-state armed groups willing to accept deserters from other groups. As a former intelligence officer in CNDP explained,
Some deserters took refuge with MONUC, and we occasionally fought with the troops of the United Nations [to get these deserters back]. The other deserters took refuge in enemy armed groups, and there were those that we reintegrated back into our group.19
A deserter from the rebel group, RCD-Goma, also similarly explained that because of his fear of recapture,
If MONUC had not helped me it would have been necessary to join another armed group, different to RCD-Goma.20
Another former RCD-Goma combatant also described how he joined the group after deserting from a rival militia, Mai Mai Mongol,
As soon as life became very hard in the bush, we quit [Mongol] to go to RCD. As my mother was a civilian in the service of RCD I followed her situation closely and decided to join. When I arrived, my mother told me, as you will be pursued [by Mongol], it is better to go straight into RCD.21
As stated above, joining a new armed group helped low-level combatants to evade recapture because it was unlikely that their former comrades would confront another armed group and risk combat for the sake of one low-level individual. Just as UN peacekeepers protected individuals within their care, armed groups willing to accept deserters from other groups also offered protection to the individuals who decided to join their ranks.
This Practice Note has outlined the organization of two non-state armed groups, RCD-Goma and CNDP, and shown how the conventionally inspired organization of these groups impacted upon the ability of low-level combatants to desert and/or demobilize. Owing to their position in the military hierarchy, certain low-level combatants had little choice but to follow recalcitrant commanders into a non-state armed group (CNDP) outside of an ongoing peace process. CNDP used its organized structures of military intelligence not only to monitor its own active fighters, but also to track those going through the ongoing DDR process. Deserters and demobilized ex-combatants from both RCD-Goma and CNDP were consequently targeted for recapture.
These findings suggest that, in settings where certain armed groups comply with a peace process and others opt out, it is imperative to ensure that deserters and the demobilized have safe places in which to resettle. More specifically, ex-combatants should not begin the reintegration process in areas frequented by armed groups that aim to recapture former fighters and which practice more generalized forced recruitment. This is especially true if the ex-combatants in question are recently demobilized and are known to be carrying DDR benefits (including cash), which are likely to make them even more attractive targets. As seen in eastern DRC, when demobilized former fighters are faced with the prospect of forced recruitment by one particular group, then voluntary remobilization in a different armed group in the vicinity may be the “least worst” option available. Deserters face similar choices, and may also resort to joining new armed groups when their ability to evade former comrades in pursuit is challenged.
In order to reduce the likelihood that ex-combatants are re-recruited into either their former groups or into new armed groups, practitioners in settings of partial peace may wish to consider the use of mobile patrols able to pick up deserters and safely transport them to DDR. Mobile disarmament sites may also be useful, particularly since active armed groups have been known to set up watch around stationary DDR sites.22 Furthermore, in instances where the dangers of forced re-recruitment are particularly acute, practitioners may wish to consider temporarily relocating and reintegrating ex-combatants in areas where non-state armed groups are absent. Relocation, however, generates issues which do not arise when ex-combatants simply return home. At the very least, ex-combatants will have to find housing and employment while removed from family, friends, and community networks. Ex-combatants might therefore turn to extended family or, as seen in many cases, other ex-combatants for support.
While the UN IDDRS (2006, 2014) previously suggested that practitioners should work to dismantle ex-combatant ties, more recently these guidelines have cautiously recognized that the maintenance of such networks may be beneficial and somewhat inevitable. Contingent on the size of the caseload, practitioners may therefore find it appropriate to support the formation of ex-combatant associations in areas of relocation and resettlement. These ex-combatant associations should also include members of the host civilian community in order to support social as well as economic reintegration (Lemasle 2012). Another, more non-traditional option, would be the establishment of transitional assistance camps where ex-combatants are housed together in a secure compound and provided with basic reintegration services. Examples of these camps can currently be found in Somalia, where support and protection are provided to former members of the still active armed group, Al Shabaab. Transitional assistance camps of this type can be controversial, however, due to the fine line that exists between confinement in a camp for an indefinite period of time (albeit for ex-combatants’ own protection), and detention (AU 2014a). To counter this, practitioners should ensure that such facilities have clear Standard Operating Procedures (SOPs) concerning the rights of ex-combatants to leave.
Finally, it is important to note that not all settings of partial peace are alike. The aforementioned recommendations are likely to be most applicable to settings where armed groups practice forced recruitment and track and pursue deserters. However, outside of the DRC, different armed groups may have different approaches to the entry and exit of their members, and may even allow fighters to resign.23 In these latter contexts, group fragmentation would not prompt the desertion of low-level combatants as this kind of unauthorized exit would be unnecessary. Practitioners therefore need to understand how armed groups approach desertion, demobilization, and forced recruitment in order to formulate appropriate policy responses. Given that much scholarship has already focused on entry into armed groups, future research should examine how different armed groups organize themselves to manage desertion, and why different policies towards armed group exit occur across different groups and over time.
2For more on this, see Richards, forthcoming 2016/2017.
12Interview 2, October 12th, 2012; Also see ICG 2007: 10.
18Some interviewees complained that their DDR assistance was stolen by CNDP, although this could not be confirmed. Similar instances have occurred among other groups, see Vogel and Musambe 2016: 2–3.
22This tactic was employed by M23, see IPIS 2013.
The author declares that they have no competing interests.
African Union (2014). Operational Guidance Notes on DDR, Available at: ddrpeaceau.org/en/ [Last accessed 23 August 2016].
African Union (2014a). AU guidelines on detention and DDR, Available at: ddrpeaceau.org/en/ [Last accessed 23 August 2016].
Christia, F (2012). Alliance formation in civil wars. Cambridge University Press, DOI: https://doi.org/10.1017/CBO9781139149426
Cockayne, J and O’Neil, S (2015). UN DDR in an era of violent extremism: Is it fit for purpose. United Nations University. Available at: https://collections.unu.edu/eserv/UNU:5532/UNDDR.pdf [Last accessed 23 August 2016].
Cohen, N and Arieli, T (2011). ‘Field research in conflict environments: Methodological challenges and snowball sampling,’. Journal of Peace Research 48: 423–435, DOI: https://doi.org/10.1177/0022343311405698
Findley, M and Rudloff, P (2012). ‘Combatant fragmentation and the dynamics of civil wars,’. British Journal of Political Science 424: 879–901, DOI: https://doi.org/10.1017/S0007123412000099
Greenhill, K M and Major, S (2006/7). ‘The perils of profiling: Civil war spoilers and the collapse of intrastate peace accords,’. International Security 31(3): 7–40, DOI: https://doi.org/10.1162/isec.2007.31.3.7
ICG (2007). International crisis group, Congo: Bringing peace to North Kivu, Africa Report No. 133. Available at: http://www.crisisgroup.org/africa/central-africa/democratic-republic-congo/congo-bringing-peace-north-kivu [Last accessed 23 August 2016].
IPIS (2013). International peace information service weekly briefing, April 25th – May 1st, Available at: http://ipisresearch.be/weekly-briefing/25-april-1-may-2013/ [Last accessed 23 August 2016].
Lemasle, N (2012). From conflict to resilience: Ex-combatant trade associations in post-conflict. lessons learnt from the Republic of Congo and the Democratic Republic of Congo. World Bank. Available at: http://www.tdrp.net/PDFs/DRC%20ROC%20Excombatants%20trade%20associations%20Jan%202012.pdf [Last accessed 27 August 2016].
Muggah, R and O’Donnell, C (2015). Next generation disarmament, demobilization and reintegration. Stability: International Journal of Security and Development 4(1): Art. 30.DOI: https://doi.org/10.5334/sta.fs.
Richards, J (2016/2017). ‘Troop retention in civil wars: Desertion, denunciation, and military organization in the Democratic Republic of Congo,’. forthcoming in the Journal of Global Security Studies. ,
Stearns, J (2008). ‘Laurent Nkunda and the national congress for the defense of the people (CNDP),’ In: Maysse, S., Reyntjens, F. and Vandegiste, S. eds. L’Afrique des Grands Lacs: Annuaire 2007–8. Paris: L’Hartmattan, pp. 245–267.
Stearns, J (2012). From CNDP to M23: The Evolution of an armed movement in Eastern Congo. Rift Valley Institute. Available at: http://riftvalley.net/publication/cndp-m23#.VzXeDqHfrIU [Last accessed 23 August 2016].
Stedman, S J (1997). ‘Spoiler problems in peace processes,’. International Security 22: 5–53, DOI: https://doi.org/10.2307/2539366
United Nations (2014). Operational guide to the integrated DDR standards, (2006). Available at: unddr/org/iddrs.aspx# [Last accessed 23 August 2016].
Vogel, C and Musamba, J (2016). Recycling rebels: Demobilization in the Congo. Rift Valley Institute. Available at: http://riftvalley.net/publication/recycling-rebels-demobilization-congo#.VzXeNqHfrIU [Last accessed 23 August 2016].