Challenges of Reintegrating Self-Demobilised Child Soldiers in North Kivu Province: Prospects for Accountability and Reconciliation via Restorative Justice Peacemaking Circles

Abstract

Social reintegration of self-demobilised child combatants can be seriously imperilled by the lack of accountability for human rights violations allegedly carried out during their soldiering life and the failure to pursue reconciliation with their respective communities. This paper examines the circumstances leading young soldiers to voluntarily exit armed groups and militias and the extent to which resettling in the community can be facilitated by restorative justice mechanisms. The findings suggest a large support by war-affected communities for restorative justice peacemaking circles as potential accountability and reconciliation measures to help reintegrate self-demobilised soldiers into society. These results were obtained from interviews, focus group discussions and descriptive statistics in which 1447 respondents participated. These included young ex-soldiers, students, educators, government official, members on NGOs, traditional leaders and councils, Security Services, and the Police. The inquiry was conducted between 3 May and 17 December 2014 in North Kivu Province, Democratic Republic of Congo (DRC).

Introduction

The use of children as soldiers, commonly known as the Kadogo (little ones) in the Democratic Republic of Congo (DRC), or children associated with armed forces and groups (EAFGA), has been ongoing since 1996, during the campaign led by the Alliance of Democratic Forces for the Liberation of Congo/Zaire (AFDL) that ousted president Mobutu Sese Seko.

When children are recruited and used as soldiers, they are paradoxically regarded as victims, perpetrators, and heroes. They are victims of violence, experienced when forced into armed groups, and moreover, they are exposed to situations that, unfortunately, lead to atrocities being committed against innocent, civilian populations (Drumbl 2012; Popovski and Arts 2006; Wessells 2006). Unfortunately, war-affected communities do not always accept overemphasising the victim image, where former young soldiers are called to account for their past crimes. In some cases, the lack of acceptance of child soldiers by recipient communities and non-compliance by child soldiers, with established, traditional practices that condition reintegration, cause them to not return home or not to reintegrate. This was experienced in Sierra Leone, following this country’s 11 years of civil war (Schotsmans 2012). ‘Non-reintegration’ may also be sustained by failure to break ties of friendship developed during war, through the ‘command and control’ chains and the changes that war-stricken societies undergo (De Vries and Wiegink 2011). The reverse side of ‘non-reintegration’ or ‘no return home’ may see ex-combatants end up in criminal activities (Schotsmans 2012, p. 239) or resort to violence, especially those leaving in unsafe environment together, while also marked by a lack of opportunities and livelihoods (De Vries and Wiegink 2011, p. 47). Transition is not easy, but it is possible, by allowing children to account for their past antisocial conduct. However, in the context of North Kivu, there are no traditional transitional justice mechanisms¹ in place to facilitate this process, as many child soldiers auto-demobilise.

The study details the culmination of the researcher’s experience over 7 months of fieldwork, in North Kivu Province’s war-stricken communities, which resulted in interactions with the movements of several children exiting armed groups voluntarily, some switching from one militia to another, many voluntarily enrolling with militias, while many others continue to be abducted. Another issue in point that was experienced is the shaky nature of reintegration by self-demobilised child soldiers, due to the lack of a proper reintegration process and sustainable programmes to accompany this process. Lack of reconciliation and accountability threatens the safety of demobilised soldiers and their reintegration, besides various other constraints.

The objective of this paper is to investigate the circumstances that lead young soldiers to voluntarily exit armed groups and militias and the extent to which their reintegration into society is imperilled by the lack of accountability, adequate justice processes in the environment where they attempt to return home, and embark on reconciliation. Furthermore, it explores restorative justice (RJ) mechanisms, as an approach that could overcome judicial challenges, in order to ease the reintegration of child ex-combatant perpetrators. These will be drawn from the existing literature and empirical data from the field, in five geographical areas most affected by armed conflict and interethnic violence and where child recruitment is ongoing in North Kivu, namely, Goma, Rutshuru, Kiwanja, Masisi centre and Kitshanga.

The paper is structured as follows: an introduction to the problem and objectives of the study, a background on child ex-combatants’ self-demobilisation, an argument on RJ used as the theoretical framework, research designs and methods, data gathering, analysis and findings, discussion of findings and conclusion.

Background of Self-Demobilisation: Concept and Context

In order to understand the concept of self-demobilisation, Zena (2013) discusses three patterns. These are, ‘self-demobilised or voluntary exit’, ‘combatants with deep ties and vested interests in militancy’ and ‘combatants between self-demobilised and the hardcore’. Child soldiers who decide to voluntarily exit armed groups in non-conventional way and without proper demobilisation support fall into the category of self-demobilised or auto-demobilised child soldiers (Juma 2012; Rouw and Willems 2010; Zena 2013).

The phenomenon of self-demobilisation in the DRC has been observed since 2009, at the end of official, nationally and internationally sponsored phases of the disarmament, demobilization and reintegration (DDR) that followed the first and second Congo War (1996–1997; 1998–2002). According to a 2010 report by the DRC Department of Defense and Ex-Combatants, in partnership with CARITAS Congo, a total of 120,000 adult ex-combatants had to be demobilised and receive incentives for social and economic reinsertion, while 30,000 EAFGA had to be reunified with their families and sent back to school. According to the same report, the official DDR ended in December 2009, including all structures planned by the Military Structure of Reintegration (SMI). The responsibility for any combatants who had not been demobilised was shifted to the United Nations Security and Stabilization Support Strategy (UNSSSS) for the eastern DRC and the Congolese Programme of Stabilisation and Reconstruction of Areas Exiting Armed Conflicts (STAREC) (Boshoff 2010, p. 1).

The conventional demobilisation consisted of structuring demobilised and local communities in productive entities; reinforcing the quality of benefits granted to local associations (skills training, professional kit, suivi [follow up], and technical monitoring); and reinforcing technical capacities, human and material resources for executing agencies and their subalterns for an effective and durable support of those associations (Democratic Republic of Congo 2010, p. 14). This official DDR programme ended prematurely because of resumed fighting and the lack of ownership for this process by the DRC Government because the overall programmes were funded by the World Bank (Boshoff 2010, pp. 2–3). In 2010, 2816 children were freed from armed forces and groups (Democratic Republic of Congo 2010). They were oriented towards agents that assure children protection under UNICEF. Of the 8000 children associated with armed forces and groups targeted, 1940 were girls and 4704 were boys (Democratic Republic of Congo 2010, p. 21).

Narratives of child soldiers and young revolutionary combatants identify socio-economic constraints, the lack of livelihood opportunities and socio-economic marginalisation as motives for engaging in soldiering activities (De Vries and Wiegink 2011, pp. 41–42; Jourdan 2011, pp. 90–97; Vlassenroot 2001, p. 124). These conditions often persist during and after hostilities even in the context where children have participated in DDR programmes. Seymour confirms this as she notes: ‘Children’s DDR programmes have often been little more than expensive logistical exercises, which ignore the conditions of entrenched poverty and extremely limited choices that lay the foundations of the child recruitment phenomenon’ (2011, p. 59). The limitations of the official DDR and its non-existence in zones where armed groups are entrenched in North Kivu explain the massive self-demobilisation of child soldiers. A study by Rouw and Willems found approximately 3000 auto-demobilised children in the south of Butembo only (2010, p. 22).² According to Juma, approximately 11,000 young ex-combatants auto-demobilised in the DRC (2012, p. 377)

After evading fighting forces, these young ex-combatants remain vulnerable to re-recruitments due to the lack of amelioration of their socio-economic conditions, and they have also to deal with an environment that accused them of perpetrating atrocities against civilian innocent populations. This is a serious conundrum to reintegration. But while socio-economic constraints are critical issues, the lack of accountability could also be an immediate threat to reintegration of child ex-combatants into the community. And as the situation stands in North Kivu, there is no transitional accountability measure in place to help reintegrate self-demobilised child soldiers into society. For this reason, it is capital to envision adequate accountability mechanisms that are mindful of child soldiers’ various circumstances or applying ‘alternative diversionary measures away from the judicial system…’³ An analysis of RJ was also included in the inquiry a follow-up on a study by Juma which had identified this model of justice as one of the useful tools for the reintegration of self-demobilised child soldiers in this environment (2012, p. 388). However, the author observes a lack of skilled personnel in the field of RJ in the DRC, who could be able to address the challenges of reintegration. In order to deal with this serious gap, this study in its entirety involved training local people in RJ. The theory and praxis of RJ are therefore discussed, while its application to address the problem of accountability and reconciliation is also shown, in reintegrating self-demobilised child soldiers.

Theoretical Background

According to Zehr, ‘[r]estorative justice is a process to involve, to the extent possible, those who have a stake in a specific offense and to collectively identify and address harms, needs, and obligations, in order to heal and put things as right as possible’ (2002, p. 37). It is ‘[…] a process whereby parties with a stake in a specific offence, collectively resolve how to deal with the aftermath of the offence and its implications for the future’ (Marshall and Britain 1999, p. 5). The United Nations Office on Drugs and Crime (UNODC 2006, p. 11) describes RJ as ‘a way of responding to criminal behaviour by balancing the needs of the community, the victims and the offenders’. These various understandings of RJ show that it is an approach that is adaptable, flexible and participatory in searching for solutions in a way that satisfies all parties involved in a stake, namely victim, offender and community.

The RJ programme concerned with this paper is peacemaking circles that are also used in various levels of criminal justice, for young and adult offenders, from lowly rated to severe crimes, in as much as they facilitate reintegration into the community, in assisting to break chronic violent behaviour by breaking the circle deleterious behaviour (Stuart and Pranis 2008, p. 125). Circles are used to create a respectful space where an offender and victim meet to address the hurt caused by a crime and look into ways that various communities can act in response to crimes or wrongdoings. This implies considering the needs and interests of affected parties, namely victims, offenders, their acquaintances, their families and the community in large (Pranis et al. 2003). In circles, parties affected by a crime develop an understanding and empathy that lead to complying with agreements reached in a non-coercive manner, according to Stuart and Pranis (2008, p. 125). Four kinds of reparation are commonly used, namely apology, restitution, changed behaviour and generosity (Centre for Justice and Reconciliation 2008). RJ circles contribute to limiting the recurrence of anti-social behaviour and inhumane conduct and also encourage change in individual offenders (Van Ness 2005).

Notwithstanding some criticisms levelled against RJ (Ashworth and McLaughlin 2003, pp. 170–171), this model of justice is gaining ground as an alternative to justice, when one considers the number of diversionary and referral cases, in contexts where RJ is adopted in the legislation (Van Ness 2005). It has come forth as a preferred alternative to justice because punitive sanctions destroy empathy in offenders who may resort to further violent behaviour. In addition to this, it has been noted that, in a number of instances where youths go through conferencing programmes, they actually improve their conduct (Umbreit et al. 2002, p. 8). When justice for child soldiers is considered from the perspective of RJ, dialogue, reconciliation, forgiveness and healing are consistently prioritised, which contrasts with retributive approaches that focus on punishment, isolation and exclusion under harsh conditions and instead of making a person better, turn him/her against his/her own.

Research Design and Methodology

This paper employed sequential mixed methods, using qualitative and quantitative methods in a sequence with findings from one method feeding into the other, and a convergent design that merges analysis of statistical findings with qualitative themes emerging from fieldwork. Data was obtained through structured and semi-structured interviews, focus group discussions and face-to-face surveys using questionnaires. Child soldiers in centres of transit and orientation (CTOs) and students were assisted in responding by legitimate guardians (caretakers and teachers). This was necessary to ensure questionnaires were understood and accurately responded.

Sampling Techniques

The research proceeds from a purposive sampling of individuals who have been identified as representative of the larger target population being studied and considered to have experienced child soldiering practices or having been directly or indirectly touched by systematic conflicts in North Kivu. All child soldiers found in interim centres at Kiwanja, Nyahanga and Masisi and in 12 selected schools, including members of NGOs and public servants and officials, were selected through snowball sampling and by convenience.⁴ A purposive random sampling was subsequently used, drawing a random sample of a small number of entities from a larger target population. This sampling method is employed to increase the credibility of the findings of a larger study, as sustained by Teddlie and Tashakkori (2009, p. 187). The target population included the youths from two parish communities in Goma and students from 12 schools (two primary and ten secondary), accessed through the parish and school registers, totalling 1900 subjects, from which a random sample was drawn. Interval or systematic sampling was applied ‘based on the selection of elements at equal intervals, starting with a randomly selected element on the population’ (Bless et al. 2006, p. 102). The total number of participants equalled 1447, which was assigned to the sample by both probability and non-probability sampling.

By randomly assigning units of analysis, the research intended to avert the effects of history and sensitivity of participants to the instruments, in addition to which it was also necessary for internal validity of the study (Bless et al. 2006, p. 91), by limiting data collection to only two rounds with the same subjects, within a period of 7 months. To ensure the reliability of the study, the sampling was based on the variability, heterogeneity of groups and diversity of the study population and by examining variables simultaneously during the data analysis. This practice is suggested by Bless et al. (2006, p. 108). Different categories of participants included self-demobilised child soldiers, and members of security services, as well as the Armed Forces of the Democratic Republic of Congo (FARDC), members of NGOs, war victims, students, educators, local and traditional leaders and members of war-affected communities. Research sites were selected by convenience because of their exposure to systematic wars, interethnic violence and ongoing child soldiering practices.

The study was longitudinal, involving two stages of data collection. The aim of the first phase was to appraise the conductivity of justice, effects of DDR programmes and the role of the community of post-war context in North Kivu, to hold child soldiers accountable and deter child soldiering practice, with various circumstances leading to self-demobilisation that were examined. The second phase of the study had a threefold objective: (1) to involve more participants in RJ training through workshops (role plays of peacemaking circles involving victims, child soldiers and community members), in order to allow a good assessment of its potential so as to assist with the reintegration of child soldiers; (2) to evaluate the immediate effects of the training on participants; and (3) to conduct focus group discussions to deepen the themes emerging from this data. For this reason, the sample was extended from 282 to 1447, over the course of the two phases of data collection. The training consisted of 2 days and half-day training by the research team in concerned institutions, mainly schools.

Quantitative Analysis

Survey questionnaires were based on nominal and ordinal scales. Respondents were provided with a scale, where their opinions were to be selected from 1 to 5, 1 representing strongly disagree and 5 strongly agree, while 3 suggested neutrality/no opinion. Data was analysed with SPSS version 22.0. Reliability statistics were computed by taking several measurements on the same subjects. Normally a reliability coefficient of 0.70 or higher is considered as ‘acceptable’, and the overall reliability score for survey questionnaires approximated the recommended value of a 0.70 Cronbach’s alpha score, for all the items that constituted the questionnaire. This indicates an overall high degree of acceptable, consistent scoring for the research. SPSS analyses were mainly descriptive statistics, frequency distributions, factor analysis, cross-tabulation and chi-square tests of significance. For factor analysis, each matrix table is preceded by a table that reflects the results of KMO and Bartlett’s test. The requirement is that the Kaiser-Meyer-Olkin measure of sampling adequacy should be greater than 0.50 and Bartlett's test of sphericity less than 0.05. In all instances, the conditions are satisfied, which allow for the factor analysis procedure (Table 1).

Table 1 KMO and Bartlett’s test

Qualitative Analysis

Discourse analysis and content analysis were applied to qualitative data, with both falling under the interpretive paradigm. Discourse analysis employs the constructivist worldview, by interpreting and handling interview talks with various expectations that are based on our way of learning, through analysis and interpretation of people’s daily, verbal language, as Talja explains: ‘Participants’ accounts, or verbal expressions, are not treated as descriptions of actual processes, behaviour, or mental events. Interview talk is by nature a cultural and collective phenomenon’ (1999, p. 2). And content analysis is utilised if data are gathered primarily by means of interviews, open-ended questions and probing questions to deepen understating of issues from participants such ‘Can you tell me more about that?’ (Hsieh and Shannon 2005, p. 1278). These techniques were employed during interviews and focus group discussions.

Ethical Considerations

Voluntary participation, informed consent, risk of harm and the privacy of participants were preserved by observing two standards of ethics, namely the guarantee of confidentiality and anonymity. Particular consideration was given to children used as source of information. Guardians’ and parents’ consent were given for participation of children below the age of 18. Sensitivity to their child vulnerability, incompetence and powerlessness were taken into account when using them as relevant ‘social actors in their rights’ and ‘sources of valid data’ (Morrow and Richards 1996, p. 95). The pluralist methodology and cross-examining both methods and data ensured the validity and reliability of the findings.

While public opinion shows reservation about associating children in research either as co-researcher or source of data, some scholars have sustained that children should not be excluded in search of viable solutions to issues of their interests. They can provide feasable contribution to improve their condition if they are treated with fairness and dignity. Scott ‘argues that the best people to provide information on the child’s perspective, actions and attitudes are children themselves. Children provide reliable responses if questioned about events that are meaningful to their lives’ (2000, p. 99). She has found unquestionably that only by interviewing children directly that their social worlds can be understood. She further argues that general population surveys that do not include children’s own narratives would come up with ‘biased estimate of many important variables’ (Scott 2000, p. 115). It is however important, she observes, that the methodological approach should be cognisant of the environment, the social position of the researcher and the techniques of data collection (structured questions or face-to-face interviews) because it may impact negatively the outcomes. While researching using children as sources of empirical data is research justice, focus groups are perceived as the most advantageous strategy (Morgan et al. 2002; Scott 2000, p. 115). The inclusion of children in this study was done with fairness and observing all ethical requirements as explained above. Taking this into consideration, more interactive discussions in groups were considered to gather information from young participants.

Research Findings: Data Gathering and Analysis

The results presented here are from descriptive statistics in the form of graphs, cross tabulations and qualitative data collected. Inferential techniques included the use of factor analysis, correlations and chi-square test values that are interpreted using the p values.⁵ This section is subdivided into two subsections (phase one and phase two). Each phase is analysed separately and followed by a comprehensive discussion and conclusion.

Phase 1: Assessment of Post-war Context and Impact on Child Soldiers

Although this paper focuses on accountability and reintegration of child soldiers, the overall study from which it is drawn sought to prevent youngsters from joining armed groups and the subsequent necessity of opening them up to RJ approaches. For this reason, the highest rate of participants was that of learners and students (55.7 percent) and RJ implies involving victims, perpetrators and community members. Former child soldiers represent offenders in the process (25.9 %) (Table 2). Other categories of participants provided insights that assist in dealing with the issues discussed in the study. Within the age category of 20 to 29 years, 79.6 % were male. Within the category of males (only), 20.8 % were between the ages of 20 to 29 years. This category of males, between the ages of 20 to 29 years, formed 15.2 % of the total sample.

Table 2 Participants’ occupation and status in phase one

Factors Analysis: Factors Surrounding the Environment Where Child Soldiers Auto-demobilise

The survey questionnaire established that, since 1996, children continue to enrol because of a variety of reasons, as set out below. The average level of agreement in this section is 73.2 %, ranging from a high of 88.7 % to a low of 63.5 %. Nearly two thirds of the respondents (63.5 %) favour prosecution and punishment for children involved in war crimes. It is noticeable that very few respondents (for most statements) have ‘no opinion’. This reflects the effect that this situation has on respondents (as they are all directly affected).

• (1) Belligerent children do not account for their past, brutal actions: 22.5 % of the respondents disagreed, 12.6 % had no opinion, while 64.9 % agreed.

• (2) The DDR was failing to prevent the recruitment of child soldiers; the local criminal justice system was failing to curb involvement of children in hostilities: 69.8 % agreed, against 25.7 % who disagreed and 4.5 % remained neutral.

• (3) Inefficiency of the local justice system to prevent children’s involvement in armed conflict: 21.6 % of the respondents were in agreement, as opposed to 70.7 % who were in disagreement and 7.7 % reflected no opinion.

• (4) The traditional justice system was not effective to deter child soldier practices: 19.4 % of the respondents disagreed, whereas 76.6 % agreed and 4.1 % had no opinion.

• (5) Impunity and lack of accountability for previous atrocities encourage greater involvement of children in hostilities: 6.8 % agreed, with 4.5 % who expressed no opinion, while 88.7 % agreed.

• (6) The communities in Northern Kivu Province do have sufficient and efficient means to intervene against participation of children in armed conflict: 17.1 % of the respondents disagreed, 5.0 % remained neutral, whereas 77.9 % agreed.

These factors are symptomatic of a hostile environment for the reintegration of young soldiers who have exited conflicts and are attempting to find their feet back in the community. The average level of agreement in this section is 73.2 %, ranging from a high of 88.7 % to a low of 63.5 %. Nearly two thirds of the respondents (63.5 %) favour prosecution and punishment for children involved in war crimes. It is noticeable that very few respondents (for most statements) have ‘no opinion’. These figures reflect the effect that child soldiering practice has on respondents (as they are all directly affected). It is noted that 38.3 % of the respondents had no opinion, while 21.6 % indicated that it was a combination of all of the factors to influence the decision to join armed groups. These factors include instability, insecurity caused by war, ethnic tensions and poverty.

With reference to Table 3:

• (1) The principle component analysis was used as the extraction method, and the rotation method was Varimax, with Kaiser Normalization. This is an orthogonal rotation method that minimises the number of variables that have high loadings on each factor. It simplifies the interpretation of the factors.

• (2) Factor analysis/loading shows intercorrelations between variables.

• (3) Items of questions that loaded similarly imply measurements along a similar factor. An examination of the content of items loading at or above 0.5 (and using the higher or highest loading in instances where items cross-loaded at greater than this value) effectively measured along the various components.

• (4) All conditions, illustrated in Table 3, satisfy the requirement of the factor analysis procedure that the Kaiser-Meyer-Olkin measure of sampling adequacy should be greater than 0.50 and Bartlett’s test of sphericity less than 0.05. These conditions satisfactorily allowed factor analysis procedure.

Table 3 Rotated component matrix

Factor analysis was adopted in order to identify different factors that underpin the variables, so as to establish the correlation between them. It was clearly demonstrated that these sets of independent variables are strongly related and all influence the conditions that do neither sustain the reintegration of child soldiers nor prevent new recruitment and enrolments. This procedure assisted to determine the causality effects of independent variables (accountability, justice system, impunity, role of the community) and the dependent variable (child soldiering practice). Several variables are linked to justice, and some are correlated to social and economic security (poverty and interethnic tensions). These sets of factors strongly cause the existing child soldiering phenomenon and its complexities.

Status of Juvenile Justice

It is noted that only 5.4 % of the respondents believed that juvenile justice was being done efficiently. Similar numbers of respondents rated the service as being inefficient or non-existent (40.3 and 39.4 %, respectively) (Fig. 1). This explains why child soldiers struggle to settle in a community that cannot protect them and help them deal constructively with their past.

Fig. 1

Participants’ views on the state of juvenile justice

War Adversities Inside Armed Groups and Militias

Underage soldiers confirm that extreme suffering, beating and being on guard for several days without sleep push them to desert. Refusal or resistance to take a commanders’ orders, delaying the execution of orders, such as killing innocent people, executing those who loot without permission, cruel training by burying them alive and suffocating them to death with their head wrapped in plastic bags are among the motivations of these young soldiers to exit armed groups and militia.⁶ Immediately, they find themselves out of armed groups and they search for a temporary home that would provide them with security and facilitate their return home.

Akiza (an 11-year-old boy from Bambo, now under care at a CTO) was abducted by the Democratic Liberation Forces of Rwanda (FDLR) when he was nine and later joined the M23 (Insurrectional Movement of March 23). He reports: ‘I was taken by the FRDLR from our house… I ran away from the FDLR and joined again the M23… … the M23 threatened to kill me because I refused to kill… I decided to run for my life and my brother PTK [pseudonym] brought me to the CTO’. Debandes (a 12-year-old boy from Chengerero, now at a CTO) narrates: ‘I was taken by the M23 in our house while my parents had gone to Uganda… they killed those [who] were with us… I ran to Uganda… afterwards I came back home… my parents welcomed me and took me to CTO’. Sertwe-Kahanda (a 17-year-old boy from Bambu village, now at CTO) recounts his tale:

I was taken at home by the FDLR who use to come in the neighbourhood to loot in Kibirizi… they send us to kill someone who attempted to escape… they ask us to kill a person alive… they show us how to bury a person alive… they show us all sorts of evil things… I escaped… I can’t return in armed group… my mother brought me to CTO.

Extreme tortures were also reported as a reason for self-disarmament and self-demobilisation, as reported by Kadogo (a 17-year-old from Chanza, now at a CTO) who was abducted by the M23 when he was 12 on his way back from school. He narrates: ’Beating us all the time… once you are sent if you waste a minute, they tie your hands and legs and throw you in a trench of 1,5 m / 200 m about 1 m deep… You could spend two to three nights in it… with food or without… it depended on those who went to loot food whether they succeeded… they were putting us in front line to fight the Government forces (DRC)… and I escaped and ran away’.

In spite of having access to good food and drink from looting, there is no joy, no peace and no freedom in armed groups for these young combatants. Suffering and the horror of killing are unbearable. Several prefer poverty and the lack of necessities to life in the armed groups and militias. This is again supported by Patient (a 32-year-old NGO agent working with child protection in a CTO): ‘Children attempt to escape when they realise that the sufferings are unbearable, they come back to their conscience, they listen to news in the camps… they hear that there are CTOs and they will be safe and well-treated’. The CTO has played a key role in encouraging the decision to self-disarm and self-demobilise. When they come out of the ‘defence’ (rebel headquarters), they immediately ask where there is a CTO.

Sartwe Kahuna (a 17-year-old boy from Mondo) was abducted by the FDLR; he reports: ‘[… ] I ran away and my mother brought me to CTO which is she said there is a fence of the Caritas where I can hide’. Kabudogo (a 17-year-old boy from Mutiri village) joined the Nyatura at the age of 15; he narrates: ‘My parents paid a lump sum of money to our commander and he released me. When I got home I was very happy’. Joel Hirondele (an 18-year-old young man from Mutiri village) was in the UPC of Thomas Lubanga; he reports: ‘I escaped… one day my friend wanted to adjust his arm, but he didn’t know how to do it. I offered to help him and suddenly a bullet went… they scourged me severely and put me in jail in the defence. At night the guards were sleeping and I managed to run away’.

Rudugiri (a 13-year-old boy from Mondo village) was abducted by the FDLR; he reports: ‘They sent me to fetch water and I ran away because they kill people and they cut his head… I was afraid to see that…’ Blaise (16 years old, from Nyamilima village) was abducted by the FDP; he tells: ‘When we were fighting the M23, they sent us to loot… I threw away my arm and uniform and ran away’. Kadogo_1 (a 13-year-old from Bambu village) was abducted by the FDLR in the field when he was nine. He recounts: ‘I used to stay at the defence… they sent me to wash their clothes… I ran away tracing and following the path they used to take me in their camp. I met my friend Patrick who was also escaping… and we ran together till we reached Bambu’. Kiwede (a 16-year-old from Chengerere village) was abducted by the M23; he reports:

The M23 were beating us, killing people, raping women and children… One day they gave a gun and sent me to kill my father and my mother. Then they will give a 1000USD and I would be going to Uganda to buy supplies for them… I left and went home… because of fear that they will want me, I decided to join the Nyatura. When I arrived in their camp they took my gun and they gave me another one.

This is quite an extreme case, where rebels intend to assassinate the innocent parents of the abducted children, yet this kind of incident happens often in areas controlled by armed groups. Rebel commanders use children after they have been paid to kill certain people. This is a way of settling scores. Sometimes, they abduct members of the targeted families and threaten to kill them if they refuse to take orders to carry out that odious mission. People in conflict use armed groups to wipe out old scores, and child soldiers are used to fulfil that role. Along the same vein, Neri (a 15-year-old boy from Cheya) joined the CNDP (National Congress for the People’s Defense) and was then abducted by the M23. He narrates:

I was a friend of one armed group commander. Then he took me there. Life was difficult. One day they sent me to buy maize and I escaped… But the M23 abducted me again when I was returning from school. Life was tough. I killed a guy who went to loot without permission… There was another guy who stole a cow in the village… they brought him to our commander by co-villagers. He asked me to shoot him and I did. One day my friend told me to run away. We threw our ammunitions and we ran away. My friend has gone to join the Armed Forces of the Democratic Republic of Congo before they were threatening to kill us (FARDC).

Kadogo_G (a 13-year-old girl from Mondo village, who stayed with the Nyatura for 6 years) reports: ‘after killing my grand-father, the commander of the Nyatura, I left and went back home’. Rushati (a 16-year-old from Chengerere) was abducted by the M23 when he was 14. He reports: ‘we escaped… one younger boy whose uncle was a Colonel [high ranking army officer]… he told him to go home when we were about to get to the front line… the colonel told his nephew to remain behind… This young boy also warned me and we remained… this is how we escaped’.

The young soldiers’ strategies vary from opportunities to initiatives. Opportunities come their way, such as the death of the commander that precipitates exit from the jungle and at times, the end of the group. Sometimes the group is defeated, such as the M23, and the end of hostilities offer opportunities to leave. There are, nonetheless, also other occasions that offer opportunities for escape, such as when one boy is asked to take another wounded friend to the hospital, when they are sent to loot or to fetch water, and even when they go on a mission to recruit other young boys or to spy or to kill targeted persons. This is because armed groups hold lists of people they target for execution.

Non-existent or Dysfunctional Mechanisms of Juvenile Justice for Child Soldiers’ Offences

Kamundala (a 15-year-old boy from Masisi) reports:

I joined the APCLS at the age of 12 to fight the Government soldiers. Because of suffering ‘fimbo kila siku ukikosa’ [being scourged everyday if you make a mistake]; I left and went back to the village. I had a miserable life… one day I raped a girl. I was arrested and detained. Because of this situation, I escaped and joined the Nyatura. Life was again tough and I managed to escape and was brought to the CTO.

Several children, who enrol, do so because of an inadequate justice system that does not look into the vulnerability of young offenders. Incarceration conditions, very often with adults, revolts them, and this becomes a stimulus for them to return to the bush.

Mudimbe (a 16-year-old boy from Masisi) reports: ‘I was in the army of the APCLS. Due to unbearable sufferings I left the armed group and returned to school as grade four in primary school. When playing soccer, I broke a friend’s leg. They wanted to arrest me. I preferred going back and re-joined the APCLS’. This is one of the rare cases that a child soldier re-enrols in the same, armed group. Often, they switch between different armed groups because their commanders become suspicious, as many children are used as spies against rival armed groups. The less fortunate are killed; survivors will continue until they escape again. These two incidents demonstrate the extent to which the non-existence of adequate juvenile justice contributes to children re-joining armed groups. Executions and severe punishment of returnee child soldiers are widely used by armed groups’ commanders to dissuade the remaining child soldiers from escaping. These facts were reported by several child soldiers who joined or/and were abducted by the FDLR and the M23.

Kamukongoto (a 14-year-old boy from Masisi) voluntarily joined the APCLS, and he tells: ‘What can lead us to return to armed group is… for example, when we stole somebody’s chicken. When the owner sees us she/he wants our death as punishment. And we are afraid of such severe punishment’. The same story is also told by Buffalo (a 13-year-old boy who voluntarily joined the APCLS): ‘there are mothers in the village with whom we live who do not like us… they look badly at child soldiers’. Francois (an 18-year-old young man from Goma) explains: ‘Communities create armed groups to defend themselves because the Government is not doing it… this facilitates the child soldiering phenomenon… there is no justice… people are jailed… there is no trial… They are enrolled assuming they will be heroes to fight for their communities… they are victims’. Unfortunately, this dream turns into nightmare in armed groups and militias.

Return into Unsafe and Hostile Environment

Child soldiers return to a hostile environment after escaping soldiering life. They are subject to verbal abuse and criminalisation of their actions. Katamasoko (a 17-year-old boy from Lushebere village) joined the APCLS and moved to the Nyatura. He narrates:

I was in the APCLS and I ran away… But my fellow-child soldiers in the APCLS were coming to look for me to return in the bush but I was refusing. I felt I was wanted and my life was in danger. I decided to go and join the Nyatura to protect myself… We prefer going to other groups because there is no justice… and they can kill you.

The same view was shared by Kasole (a 13-year-old from Miandja who was abducted by the FDLR), who narrates: ‘what pushes me to go to the bush is suffering… when I returned to the community, I was hiding because they wanted to kill me… our community cannot protect children who were in the bush… it is good to have the ‘barza’ [customary justice council] to educate children’.

Without a doubt, insecurity and a lack of welcoming attitudes from some families and antipathy from certain members of the community become setbacks to reemmergence in their peaceful life once back in the community.

Kiriboli (a 16-year-old boy from Masisi) joined the PARECO (Coalition of Congolese Patriotic Resistance) at the age of 15 and moved to the Nyatura. He reports: ‘There are so many issues that cause demobilised children [to] return to army… incriminations, tortures by the militaries’. Eh-Litien (a 14-year-old from Masisi) joined the PARECO and enrolled again with the Nyatura. He tells: ‘After leaving the bush… there are rebels who come nightly knocking on the door… you think it is a neighbour… after opening the door… you realise it is a rebel… he can kill you… better stay in the bush’. The whole climate of insecurity is a nightmare for former child soldiers, especially around Masisi, in Rutshuru, Walikale and along the road to Kanyabayonga. Former child soldiers are always targeted by armed groups because of their expertise and are used to obtain information about rival armed groups.

When asked what attracts children to return to the bush, they point to four main reasons. These constitute the pull factors for re-recruitment: a free meal, money, drinks and women. Life of the armed groups, who are entrenched on the mountains, the jungle and national parks of the North Kivu Province, is sustained by looting crops, livestock, shops, vehicles and passengers. Child soldiers are mostly those involved in ‘road-blockings’. Kavuno (a 16-year-old from Rubona) narrates:

I was in the Group of Manoti,⁷ the chief of an armed group called IDIPC [unidentified armed group]… I had a panga. We didn’t have ammunitions. When we were looting, whoever resisted I hit him with the panga at his… when we looting if I got 200USD, I hid 5USD. And you hide all… they find you… they kill you. Any amount I confiscated I slid it in my boots.

Debandes (a 14-year-old, abducted by the M23 in Kiwanja) reports: ‘what I found good…cow meat… drinking alcohol, sleeping with women… they gave us women to sleep with them… if you refuse they beat you severely…. No I didn’t do this act (raping)’. Blaise (a 26-year-old who joined the FPD) narrates: ‘What was good! … stealing… eating… looting’. Kadogo G (a 13-year-old, the only girl associated with armed groups found in the CTOs) joined the Nyatura as a cook and housekeeper of her grandfather NDJ, who was a Nyatura commander. She reports: ‘What was good… to eat… and get money freely without efforts… my grandfather used to send me to buy drinks for him… I was eating the change… he was killed and I went back home’. Actually, Kadogo G’s grandfather was the leader of the Nyatura. He was shot during the operations against the M23 for also being accused of practising witchcraft. They tracked him down to their headquarters and shot him. The entire movement of the Nyatura was dispersed, and the combatants returned to the communities. Kadogo_1 (a 16-year-old from Chanza) was abducted on his way back from school by the M23 at the age of 14. He reports: ‘what was good… I was eating well, and looting, you find money without a problem, you rape, I took a woman by force… she was afraid of me and I slept with her. We looted hospital and shops’.

Socio-economic Constraints: Ill-Treatment, Poverty and Starvation

The majority of returnee child soldiers identify suffering as the main reason for their desertion because of the threat posed by returning to the same fighting force one has just left. Nevertheless, defection is the result of insufficient measures to protect former child soldiers after their reinsertion into the community. They are caught in between hardships in armed groups and rejection by society. This opens the way to several defections and switching between different armed groups. Funding of post-demobilisation programmes has been a serious challenge for children entangled in armed groups, after the primary efforts that provided demobilised combatants with skills and toolkits to allow their socio-economic reinsertion.

Kinige (a 14-year-old boy from Masisi) reports: ‘I was in the APCLS. I was ill-treated, I decided to reintegrate and joined the GIDE of the Nyanga and Hunde communities. But children of other communities were also welcome’. The same problem was confirmed by Kasongo (a 16-year-old boy from Katoyi village who was abducted by the FDLR and joined the Nyatura). He explains:

Our people do not want us to live in this world when we leave the bush [armed groups] after looting and stealing chickens, goats… They want our death. This threat leads us to return to the bush… say that I deserve death penalty as punishment for looting… this is what drives me back to the bush because people bother me and I have no peace in the community (Personal interview with Kasongo, Masisi, 10 June 2014).

Child soldiers need to be exposed to accountability processes in order to protect them and to minimise their re-enrolment and re-recruitment. Shetani (an 18-year-old young man from Mutiri village) joined first the CNDP at the age of 12, then moved to the APCLS, and finally joined the Forces of the DRC, before he decided to leave the armed forces and military groups. He is now in grade 11 at a local, secondary school. He reports:

I was jobless when I joined. I fought several times in different armed groups, including in the Government forces. I had to return in armed groups and forces because there was too much suffering in the armed groups and in the community. I was treated like an animal… the Congolese government does not like people who have stayed in armed groups in the forest (Personal interview with Shetani, Masisi, 10 June 2014).

When persisting insecurity hinders their resettling process, armed groups remain the last refuge for child soldiers. Armed groups are perceived as a safe haven for perpetrators of crimes, looters, thieves and those involved in antisocial acts (impregnating a girl, causing injury to a person, escaping justice, etc.).⁸ There are circumstances that are so complex that they become even more problematic, especially when child soldiers are caught between violence in both the armed groups and in the community. Suffering in the bush and violence by the security services and the police worsen the prospect of living in the community. This is a very dangerous situation that can become suicidal. In this context, it is extremely difficult for child soldiers to escape from soldiering life.

Phase 2: Follow-Up Study of Restorative Justice

Participants’ Rate According to Research Sites

Although questionnaires were dispatched to be completed and returned the next day, some participants preferred answering and handing them back on the same day. A total of 1165 respondents involved came from Goma representing 69.2 %, while those from Masisi were at 1.5 %, Rutshuru 12.8 %, Kitshanga 5.3 %, Walikale 0.2 % and Kiwanja 11.1 %. Of the total of the sample, 43.1 % of participants were below the age of 18, while 56.9 % were 18 years old and above, with ages on average ranging between 10 years old (the youngest) and 78 years old (the oldest). Male respondents represented 51.6 % and females 48.3 % of the respondents. A total of 121 self-demobilised child soldiers were interviewed, and six focus groups were conducted with them, in four different sites in the districts of Masisi and Rutshuru.

As explained earlier, the predominant number of learners (79.1 %) was intended to involve as many juveniles as possible in RJ programmes, in order to obtain their views regarding their perception of ex-combatant soldiers, their accountability and reintegration. The pedagogic aim of the programme was also to discourage young people from joining armed groups and militias. Victims were represented quite well, at 22.1 % of the other categories of participants. It was difficult to separate child soldier victims and perpetrators from the community, though the majority were found in interim centres. They identify themselves as victims in Table 4. Over the course of both phases of the study, 80 self-demobilised child soldiers were below the age of 18 years old, while 41 were older than 18.

Table 4 Participants’ features in phase two (evaluation and follow-up)

During the first phase of data collection, search conferences suggested that Baraza⁹—a traditionally based council of elders for prevention and resolution and mechanisms of justice—and justice that is restorative can be viable approaches to ensure accountability and reconciliation of child soldiers. These two mechanisms were examined in the second phase, although restorative peacemaking circles were already conducted through participatory action research, in the initial phase of this inquiry. This paper’s only focus is on statistical indicators on the potential of RJ to overcome address the accountability challenges pertaining to child soldiers’ reintegration.

Acceptance of Self-Demobilised Child Soldiers in the Community, Without Accountability

Respondents shared their views as to whether alleged charges of criminal acts by former child soldiers should be dropped or ignored. A total of 69.7 % of the participants disagreed that former child soldiers should not be accountable for their past crimes. This was opposed to 23.6 % who expressed that no charge should be brought against child soldiers. Among former child soldiers, 2.3 % joined voluntarily, 3.1 % were forced into hostilities, while 94.6 % never participated. The high number of non-soldiers in this phase was used to gain an in-depth perception from various populations of child soldiering accountability via RJ while at the same time instructing younger participants in ways of building peace and addressing the rights of victims and communities who are affected by wars and violence related to ethnicity, and other causes of conflicts.

Attaining Justice Through RJ for Child Soldiers and Community Members

• (1) It appears that 68.5 % of the respondents agree that RJ attains justice, whereas 16.8 % disagree and 14.9 % have no opinion.

• (2) Participants’ views on the impact of RJPCs on reconciliation between war victims and child soldiers was rated as follows: 83.4 % agreement, 8.2 % disagreement and 8.4 % no opinion.

• (3) The level of agreement, regarding the potentials of RJ to prevent children from joining armed groups, was expressed as follows: 74.9 % agreement, 16.7 % disagreement and 8.4 % neutral.

• (4) The overall level of participation in peacemaking circle programmes and subsequent focus group discussions indicated that 73.8 % of the respondents were in agreement, against 8.4 % who disagreed, while 17.7 % had no opinion, with 0.1 % missing data.

Respondents were asked to express their views regarding their preferred model of accountability for child soldiers. There was no significant difference in the respondents’ view on applying conventional criminal justice as a model of accountability for self-demobilised child soldiers. This split in opinion shows that holding former child soldiers accountable is a great desire of war-affected communities in eastern Congo.

Discussion of Findings

The findings of this research describe the interplay of factors that affect the reintegration process for young soldiers, namely the lack of accountability of child soldiers, failure of the DDR programmes and the insufficiency of the justice system. For instance, the correlation value between ‘demobilisation, disarmament and reintegration/reinsertion programmes, are failing to prevent new enrolment of children in armed forces and groups’ and ‘the local criminal justice system that is in place, is not preventing the involvement of children in armed conflict’ is 0.421. This is a directly related proportionality. Respondents agree that the more effective the criminal justice system, the more likely it is that children would be prevented from joining the armed forces, and vice versa.¹⁰ It is argued that, if self-demobilised child soldiers are accountable through RJ and other adolescents participate in this process, RJ mechanismscan inhibit the feeling of marginalisation and rejection and create a welcoming environment for children. Then they would not have to return to armed groups. The other advantage will be that of dissuading other children from venturing to soldiering life.

In crosstabulating variables, negativevalues imply an inverse relationship. That meansthe variables have an opposite effect on each other. For example, the correlation value between ‘Do you intend to join armed groups or become a soldier?’ and ‘Traditional justice is failing to stop children to participate in armed conflicts’ was –0.296. Respondents indicate that the failure of traditional justice leads to a greater likelihood of children joining the armed forces, and vice versa. These analyses show that the failure to address accountability impedes reintegration and sustains remobilisation of minor soldiers after exiting armed groups and militias.

Challenges of the Environment Where Child Soldiers Exit Armed Groups and Militias

Socio-economic constraints (poverty and lack of employment and education opportunity) are viewed as an additional obstacle to reintegration. It should, however, be noted that socio-economic challenges, faced by the communities of North Kivu, is an arguable issue because all young people in the DRC experience the same lack of education opportunities, poverty and precarious life conditions. The eastern DRC is better off than the western, central and northern DRC, in terms of infrastructure and opportunities, though these are neither sufficient nor according to standard. Notwithstanding the critical effects of these challenges on reintegration, the real factors that impede the process of reintegration are the lack of accountability by child soldiers for their past crimes, the failure of DDR programmes, the lack of an adequate justice system, impunity, insufficient structures for juveniles and the incapacity of the community to protect children. There are close links between these factors and the presence of active, armed groups and runaway insecurity in North Kivu Province. The opinion held by numerous scholars combines several indicators of the disastrous situation of the DRC, characterised by the environment where child soldiers exit armed groups and forces. Looking closely into this situation, one may ask the following question: why then is child soldiering practice so entrenched in North Kivu and South Kivu and not in other provinces? The answer then would preclude social and economic constraints as main factors that push minors to soldiering, but rather ongoing conflict, insecurity, and lack of accountability for both child and adults associated with armed groups and militias. Judicial challenges are part of the difficult environment where justice systems cannot provide protection of communities and perpetrators. Moreover, reconciliation is often ignored.

Avoiding Stereotypes: Mixed Reactions of Recipient Communities to Returnee Child Ex-soldiers

Another important issue to debate is the mixed feeling reaction that is shown by communities into which children need to be reintegrated, namely sympathy, hostility and rivalry. The hostility of victims and rivalry of the Security services and the FARDC members, who have engaged with child soldiers when these minors fight in non-States’ forces, leave self-demobilised child soldiers vulnerable. The unconventional and informal ways through which they exit hostilities explain some of the challenges they face. It is necessary to isolate the circumstances of each child soldier and deal with each one separately, in order to better understand their individual circumstances. RJ was thus examined, and the findings indicate that this justice paradigm is a better alternative for accountability and reconciliation of child soldiers, in order to assist them with reintegration into the community.

Argument on Restorative and Conventional Justice

Firstly, the fact that the majority of respondents called for accountability of child soldiers and expressed concerns in accepting them into the community without justice suggests the critical role accountability can play, in the process of reintegration of ex-combatants. Secondly, the inadequacy of conventional criminal justice to hold former child soldiers accountable has provided the ground for exploring informal and non-judicial mechanisms of accountability. The results of the study show that war-affected communities support RJ as a model of accountability for child ex-combatants. Both victims and child soldiers who participated in RJ training programmes expressed satisfaction for the programmes.

The viability of RJ to ensure reintegration of child soldiers is sustained by two main facts:

• (1) It is important to note that, in the aftermath of bloody conflict and civil war, the pursuit of criminal justice is illusive because of the devastating and protracted consequences of hostilities. These include the judicial system structure, killing of judges and members of the judicial system, dilapidated judicial infrastructures and logistics able to prosecute criminals, anarchy and chaos, as also argued by some scholars. In this environment, reconciliation and accountability cannot be solely achieved through a prosecutorial justice.

• (2) Taking the route of criminal accountability then requires that prosecutorial policies for child ex-combatants assess the intentionality and materiality of the crime, which is a difficult exercise, in order to establish child soldiers’ culpability. This makes justice even more intricate. In addition, the findings showed that there is impunity and a lack of justice and accountability for human rights, particularly as perpetrated by child soldiers.

In order to surmount these two predicaments, RJ tests yielded positive perceptions regarding accountability, reconciliation and reintegration. Without reconciliation, accountability does not guarantee harmonious relations between child perpetrators and affected communities. Additionally, it is noted that global peace and healing of relations cannot be attained by the mainstream justice system with its prosecutorial apparatus. By advocating RJ, as the preferred model of accountability and social reintegration of child soldiers, the focusis on the attainment of justice that enhances societal harmony, while allowing an offender to reintegrate into his/her community and to break with his/her past. This gives hope to society and offenders to move forward.

Conclusions

This paper examined various circumstances that push young soldiers to voluntarily exit armed groups and militias, and the extent to which the lack of accountability imperils their stay in the community. While their challenges are enormous and multi-levelled, this paper focused on accountability and reconciliation. The results suggest that RJ peacemaking circles embed potential to avert the adversities of child soldiers’ reintegration into society without proper accountability measures. This argument was sustained by voices of ex-child soldiers in the field.

The established premise is that war-affected participants disagreed strongly that child soldiers who have committed atrocities should be returned to the community without any form of accountability. This strong objection was persistently found over the course of the two rounds of data collection. The assumption, according to which all ex-combatants are inclined to return home, has been mitigated by previous studies, due to sociological and cultural factors causing the ‘no return home’ or the ‘no reintegration’ of child soldiers, should be placed in context. It was observed that the proportion of the ‘no return home’ factor was very negligible among self-demobilised child soldiers in North Kivu. These young soldiers exit hostilities wounded, vulnerable and traumatised. There is no criminal justice system that can justify prosecution of this category of minors, in this context. Through RJ peacemaking circles, they can obtain a chance for peaceful reunification with their families, reconnection with their communities and finding a new identity in society. While socio-economic constraints and collateral challenges undoubtedly affect child soldiers’ reintegration, these conditions do affect their relationship with other community members to the extent of immediately threatening their stay in the community and making them more vulnerable to hostility of war survivors. But the lack of accountability does as it has been demonstrated in this paper. It is sustained that taking the route of RJ is a genuine path that can lessen the traumas of rejection, marginalisation and exclusion that can only worsen child soldiers’ vulnerability and shatter their hope for the future. This implies empowering more community members in the use of restorative justice mechanisms. This is a challenge that the DRC Government and local NGOs can take.

This study was limited by a short-term engagement with self-demobilised child soldiers and war-affected communities. Yet, it provides grounds for prolonged RJ programmes, in order to ascertain assertively the effectiveness of this justice paradigm to ease the reintegration of child soldiers into the community. This is an area for further study. Finally, sustainable RJ cannot be a panacea to the global issue of child soldiers. There is a need to also address their livelihoods and opportunities for those who return home.