Why do armed groups sometimes switch from voluntary to forced recruitment of childres? : a case study of the Democratic Republic of the Congo

(1)

WHY DO ARMED GROUPS SOMETIMES SWITCH

FROM VOLUNTARY TO FORCED RECRUITMENT

OF CHILDREN?

A CASE STUDY OF THE DEMOCRATIC REPUBLIC OF THE CONGO

Master Thesis Political Science: International Relations

Author: Jessica Besch

Student Number: 11254114

Supervisor: Dr. Abbey Steele

Second Reader: Dr. Jana Krause

(2)

Abstract

Why do armed groups sometimes switch from voluntary to forced recruitment of children? By extending Beber and Blattman’s (2013) theory about the costliness of punishment I argue that an increase in economic resources, being natural resources and foreign government support, leads to a change from voluntary to forced recruitment of children. My within and cross-case analysis about the RCD-Goma and the Mai-Mai, two competing armed groups in the Democratic Republic of the Congo, should test the causal effect between my independent and dependent variable. The observable implications suggest that there is some sort of correlation, but due to a lack of fine-grained data I can only partially confirm my hypotheses. This paper highlights the importance of variation when studying the recruitment of children and calls upon scholars to guide future research on this topic in order to protect children in conflict areas.

(3)

Abbreviations

Alliance des Forces Démocratiques pour la Libération du Congo (AFDL)

Armée Nationale Congolaise (ANC)

Democratic Republic of the Congo (DRC)

Forces Démocratiques Alliées, Uganda (ADF)

Forces pour la Défense de la Démocratie, Burundi (FDD)

Large-N analysis (LNA)

Lord’s Resistance Army, Uganda (LRA)

Mouvement pour la Liberation du Congo (MLC)

Rassemblement Congolais pour la Démocratie (RCD)

Rassemblement Congolais pour la Démocratie – Mouvement de Liberation (RCD-ML)

Rwandese Patriotic Army (RPA)

Small-N analysis (SNA)

Société minière des Grands Lacs (SOMIGL)

Société minière et industrielle du Kivu (SOMINKI)

(4)

I. Introduction

The recruitment and use of children as soldiers during conflict is condemned by the UN Security Council and is in violation of international law (United Nations, 1999). The Optional Protocol to the Convention on the Rights of the Child on the involvement of children in armed conflict (2002) should conduce states not to recruit children under the age of 18 and to prevent such recruitment (United Nations, 2002). However, even if agreements with governments and armed groups are being reached, they do not necessarily provide a safeguard. The “Children, Not Soldiers” campaign is advocated in order to further protect the rights of children in conflict zones (United Nations). Using children in regular and irregular armed groups can prolong negotiated settlements and harm post-conflict reconstruction (Haer and Boehmelt, 2017, p. 5). Studying the mechanisms behind the recruitment of children is therefore an important aspect to peacebuilding. The exact number of child soldiers in the world today is not known, but it is estimated at being around 10.000 and up to more than 100.000 (Child Soldiers International).

The use of children as soldiers has gained more and more attention among scholars. The literature focuses primarily on the reasons of child recruitment and their consequences (Andivg, 2006; Haer and Boehmelt, 2016; McBride, 2014; Singer, 2005). Another wave of literature tries to explain why children voluntarily join an armed group. Possible explanations include escaping poverty, taking revenge and seeking protection (Brett and Specht, 2004; Guichaoua, 2007). The reintegration process of former child soldiers is also widely studied and should push policymakers to address the importance of education and constructing rehabilitation centers (Bracken et al., 1996; Wessells, 2005; Derluyn et al., 2013).

Little however is known about the forced recruitment mechanisms of armed groups. Why do they recruit children by force as opposed to voluntarily? Recruiting tactics

(7)

7

are often referred to as static, producing a gap in the literature about the recruitment of adults and children as soldiers. Eck (2014) points out the importance of variation when analyzing the recruitment of armed groups: “[…] groups are likely to shift recruitment strategies depending on the exigencies of the conflict” (Eck, 2014, p. 374). Consequently, my research question will focus on the change in the recruitment method of armed groups. In this paper I am trying to explain why armed groups sometimes change from voluntary to forced recruitment of children.

I will base my argument on Faulkner (2016) and Beber and Blattman (2013) by claiming that an armed group’s economic endowments influence their recruitment strategies. The latter suggest that a group with initial low level of civilian support and a resource base that is unresponsive to human rights abuses is more likely to recruit children by force. Thus, when an armed group relies primarily on natural resources or foreign donors to finance its organization, it does not have to be dependent on civilian support and can more easily adopt forced recruitment tactics (Beber and Blattman, 2013, p. 95). Faulkner also argues that the decision to recruit children by force relies on whether a group has enough resources to finance warfare. Those who lack economic endowments at the beginning of their mobilization will appeal to the civilian population for support. Groups who were successful in gaining civilian support will consequently refrain from forced recruitment practices. The second part of his theory targets the variation in recruitment practices as a consequence of changes in an armed group’s access to economic resources. Here Faulkner goes a step further than Beber and Blattman by claiming that the level of forcible recruitment of children will increase when groups did not win the support of civilians but increased their resource base (Faulkner, 2016, pp. 217-218). Similarly, I propose that an increase in an armed group’s economic endowments, being natural resources and support from a foreign government, leads to a change from voluntary to forced child recruitment.

(8)

8

My findings on the case study of two armed groups in the Democratic Republic of the Congo (DRC), the RCD-Goma and Mai-Mai, show that economic and social endowments indeed play a crucial role in the variation of child recruitment. The rebel group RCD-Goma changed their recruitment method during a period of increased resource wealth. Failed attempts to gain support from the population caused little voluntary recruits. The Mai-Mai militia on the other hand were initially popular and experienced a high number of recruits. Some groups however sought personal enrichment and endangered their reputation, leading to a decline in voluntary and sometimes also a switch to forced recruitment. A lack in precision in regard to the variations weakens my results but does not undermine the fact that changes in a group’s resources are connected to changes in a group’s child recruitment strategies.

In the remainder of this paper, I will first give an insight into the existing literature that covers the topic of child soldier recruitment. In the second section I will develop my argument and the hypotheses I will test. The third step will be the research design, where I will first explain which methodology I will use in order to analyze my data. I will then move to the case selection and elaborate why I chose two armed groups within the Democratic Republic of the Congo as my cases. In order to proceed, I then need to illustrate the indicators for my measurement. The background of the conflict and the armed groups will serve as an introduction into the section of my data collection. Here I will present all the observable implications that are relevant for my hypotheses, followed by a summary of the results. The evidence partially supports the theoretical expectations of the research and highlights the importance of variation when studying the topic of child recruitment. The last section concludes with a discussion of alternative explanations and opportunities for future research.

(9)

9

II. Literature Review

I will give a brief introduction into the topic of child soldiers in order to situate my paper in the broader universe of theories. How is a child soldier defined and where are children recruited most? More importantly, why do armed groups recruit them as opposed to adults? I will also give an overview of voluntary versus forced recruitment mechanisms and the main scholars involved in the discussion.

UNICEF formulates a standard definition of what is called a child soldier, being any boy or girl under the age of 18 who is part of any kind of regular or irregular armed group and serves as a cook, porter, messenger or combatant (UNICEF, 2011). Child soldiers can also perform tasks such as laying explosives, spying, taking care of the logistics and domestic labor (Coalition to Stop the Use of Child Soldiers, 2010).

To get a better sense of the scope of the recruitment of children, it is important to know where child soldiers are mostly employed. Recruitment of children into the national military is still allowed in 50 countries. Around a dozen armed groups still use child soldiers today, ranging from Colombia, Somalia and Yemen to Myanmar, India and the Philippines. Wealthier states such as Canada and the United Kingdom allow children as young as 16 into their national military with consent of a parent or guardian (Child Soldiers International). All in all, I have barely found a country that has been in a civil war at any given time and where not at least one armed group was recruiting children into their ranks. Given the huge amount of armed groups that have been and still are involved in recruiting children, it is difficult to speculate which one’s are more likely to use child soldiers than others.

Why are children even recruited as opposed to adults? The proliferation of small arms has made rebel groups and militias more inclined in recruiting children (Machel, 1996; Singer, 2005). Such weapons are easy to use and lightweight, making them the perfect combat tools for young soldiers. Their low cost and availability through illegal

(10)

10

markets influence the attractiveness of recruiting children into the ranks of regular and irregular armed groups (McBride, 2014, p. 9).

Another explanation for the rate of child soldier participants is the easy access to refugee camps. Low protection of these camps makes children particularly vulnerable to recruiters (Achvarina and Reich, 2006). Children have also often been taken from camps to third countries for military training (Machel, 1996).

One of the most elaborated theories of child recruitment consists of looking at the demand and supply factors. Demand on the one hand means looking at the decisions made by an armed group when recruiting new soldiers. Which strategies are being implemented and who are the targets? Supply on the other hand is defined by characteristics such as the number of children present in the population or the level of poverty and determines how costly or difficult the recruitment will be. There is also the distinction between forced and voluntary supply (Andvig, 2006, pp. 19-20). Furthermore, armed groups are seen as rational actors who want to maximize their probability of winning with the lowest financial costs. In that sense, children are employed if they are cheap enough to compensate for a group’s low military efficiency (idem, p. 26).

There is a debate whether children are better fighters than adults. Some literature however states that the fact that they are more easily manipulated and indoctrinated ultimately transforms children into effective recruits in the long run (Beber and Blattman, 2013). Compared to adult combatants, they are more adaptable and obedient which in turn makes them easier to control and retain. These characteristics unfortunately facilitate their recruitment (Haer and Boehmelt, 2016, p. 157). Under the influence of drugs, children can be more daring and fearless in combat. Socialization into the armed group makes them acquire the values of their militant environment and thus influences their behavior (idem, p. 154).

Recruiting decisions by armed groups are influenced by cost and benefit calculus. A decline in soldiers motivates group leaders to fill the quota, ideally at a low cost. Since

(11)

11

children are easy to catch due to their low mobility, they become easy targets for groups with low human resources (Lasley and Thyne, 2015, p. 291).

Having summarized the main reasons for armed groups to recruit children, I will now dive into the forced recruitment mechanisms which my paper will focus on. By forced recruitment I mean the opposite of voluntarily joining an armed group. Coercion and abduction thus fall into the category of forced recruitment. Also, when a person is being threatened by members of an armed group and does not have the choice to say no without having to fear consequences, recruitment cannot be counted as voluntary. It is however not easy to completely separate the two categories because even if children claim to have joined at their own free will, indirect but coercive mechanisms could have influenced their decision. Persuasion, intimidation and social pressure can push people, especially children, to become soldiers in a regular or irregular armed group (Honwana, 2006, p. 58). These environmental push factors make a decision seem voluntary, but often children discover too late that “even if the initial act was voluntary, there is no opportunity to exercise further choice” (Brett and Specht, 2004, p. 110). Some were also being misled with false promises, making their reasons for joining an illusion (idem, p. 112).

Why is it important to differentiate between voluntary and forced recruitment in general? The article “Explaining Rape during Civil War” by Dara Kay Cohen (2013) particularly comes to mind, where the author links forced recruitment (of adults) to gang rape in times of war. She argues that low social cohesion is a consequence of forced recruitment. In order to create bonds and increase the morale of the group “through inducing feelings of power and victory”, armed groups commit gang rape (Cohen, 2013, pp. 463-464). Recruitment mechanisms can thus influence a combatant’s tendency to engage in wartime rape. Andvig (2006) points to the importance of voluntary recruits in order to prevent the usage of child soldiers. He demonstrates that economic or ideological incentives of an armed group are necessary in order for people to join them. In case of a declining or initially low success rate, little voluntary recruits are expected. If this stock of

(12)

12

potential adult recruits dries out, we should observe an increase in forced recruitment and child soldier ratio (Andivg, 2006, p. 24).

Cohen’s cross-national dataset reveals that 45% of all insurgencies between 1980 and 2009 recruited their adult fighters by force (Cohen, 2013, pp. 467-468). It was not surprising to see that countries such as the Democratic Republic of the Congo, Uganda and Sudan were in the tables. All three countries are known for using child soldiers and recruiting them by force on a large scale (Coalition to Stop the Use of Child Soldiers, 2000). There may be a correlation between forced recruitment of adults and forced recruitment of children, which I will however not further elaborate on now. My aim was to point out the importance of recruiting mechanisms for further actions of an armed group and vice versa.

Andvig (2006) points out that it is more beneficial to recruit children by force than adults since they are more easily framed and therefore forget more quickly that they were forcibly recruited. Ultimately, the desertion rate will be lower (Andvig, 2006, p. 31). Beber and Blattman (2010) agree with the fact that it is not beneficial for armed groups to recruit adults by force since their incentives to escape will be too high. It is thus easier to coerce children as they have low abilities and reservation utilities (Beber and Blattman, 2010, p. 2).

Some scholars put their emphasis on armed groups’ economic versus social resources (Beber and Blattman, 2013; Salehyan et al., 2014; Faulkner, 2016). They argue that natural resources or support from foreign governments as a resource base can decrease an armed group’s incentive to collaborate with civilians. According to Salehyan et al. (2014), groups that receive external support are more likely to engage in one-sided violence against civilians (Salehyan et al., 2014, p. 640). Faulkner (2016) builds on this theory and adds as a consequence the rise of forced child recruitment. I will now elaborate on these ideas in more detail since I will use them as a foundation for my argument about why armed groups sometimes switch from voluntary to forced recruitment of children.

(13)

13

First, for an armed group to function and achieve its goals, it needs adequate financial support (Salehyan et al., 2014, p. 635). This resource base then further influences its way of recruitment (Faulkner, 2016). In their article “The Logic of Child Soldiering and Coercion”, Beber and Blattman (2013) argue that armed groups with initially low support from civilians and resource bases that neglect human rights, such as natural resources and foreign donors, should be more likely to employ coercion against children (Beber and Blattman, 2013, p. 95). However, in order to be effective, coercion has to be cheap. This means that by recruiting children by force armed groups should have little to lose, especially civilian support or other funding actors who could condemn them for violations against human rights (idem, p. 69).

Rebel groups can become less dependent on civilians when they find alternative means of financing their violent organization, as for example external support from foreign governments or when they get the opportunity to easily extract natural resources (Faulkner, 2016). Easily extractable commodities can be sold on the black market and provide a revenue in order to keep their organization alive. With financial and material support from foreign sponsors, incentives to build or maintain a relationship with the civilian population decrease. Ultimately, this relationship, being a “[…] strategy of ideological persuasion and service provision […]”, is considered costly (Salehyan et al., 2014, pp. 636-637). Unconditional foreign support also reduces rebel groups’ accountability and lets them engage in coercive recruitment and even violence against civilians without having to fear sanctions (Beber and Blattman, 2013, p. 100).

In the next section I will elaborate on my theoretical framework and present two hypotheses.

(14)

14

III. Theory

The last part of the literature review constitutes the foundation of my theory. Combining ideas from Salehyan et al. (2014), Faulkner (2016) and particularly Beber and Blattman (2013), I put an emphasis on economic and social endowments when explaining an armed group’s recruiting strategies. Civilian support on the one hand is especially important in the initial phase of an armed group’s mobilization. With their help, groups can rely on food, shelter and even fighters and thus accumulate their resource wealth. High economic endowments (natural resources and foreign sponsorship) on the other hand give a group little incentive to build up or keep a relationship with civilians. Both these variables, as well as recruitment decisions, can however vary over time.

Faulkner (2016) highlights the importance of variation by saying that “[…] as a group’s access to resources change so too will their level of forcible recruitment” (Faulkner, 2016, p. 218). Moreover, he specifies that rebel leaders make the decision about forced recruitment of children as opposed to adults since they provide an easy access in an environment where local support is non-existent. He argues that groups which manage to gain local support will refrain from forcibly recruiting children (idem, p. 219).

The objective of my thesis is to find out which factors help to explain why armed groups recruit children by force rather than voluntarily. Being aware of the fact that environments and relationships surrounding an armed group are never fixed and vary over time, I want to focus on change rather than only initial conditions. By paying attention to the temporal variation I will argue that a change in a group’s resource base can lead to a change in the way they recruit children. More specifically, a rise in an armed group’s economic resources will consequently lead to a change from voluntary to forced child recruitment. The causal mechanism linking my dependent and independent variable is the decline of civilian support.

(15)

15

My theory puts an emphasis on an armed group’s accumulation of resource wealth. I assume that every group first tries to recruit adults and children voluntarily. This also creates incentives to build a relationship with civilians. If they gained the support of civilians but over the time increased their financial means, a group will gradually lose this relationship. A reason for that is the fact that maintaining a relationship with the local population requires a lot of effort (Salehyan et al., 2014). Furthermore, high economic resources convey the image of greedy and criminal soldiers and consequently weaken an armed group’s reputation. As a result of this, they can no longer depend on voluntary recruitment but will turn to coerced and forced recruitment of children. An increase in economic resources (independent variable) thus leads to a decline in civilian support. Civilian support, being my causal mechanism, then has as an effect the switch from voluntary to forced recruitment of children (dependent variable).

It has to be noted that I am using Beber and Blattman’s (2010) explanation about why it is more beneficial for an armed group to recruit children by force as opposed to adults. Their low abilities and reservation utilities decrease the possibility of escaping (Beber and Blattman, 2010, p. 2). Children thus represent an easy and cheap way to get recruits. Only if the number of needed combatants increased dramatically in a short period of time would a violent organization turn to forced recruitment of adults.

I am aware of the fact that there may be different reasons why armed groups switch their child recruitment strategies. A sudden urge of a high number of troops can be fatal when voluntary recruitment is low and civilian support almost absent (Andvig, 2006; Lasley and Thyne, 2015). This would then rather be observed in times of multiple big attacks between two or more armed groups. I can also think of another reason for a decline in civilian support. A shift in alliances for example can ultimately alter a group’s reputation. Raúl Sánchez de la Sierra (2017) further provides us with an explanation about why some armed groups are more dependent on civilians than others. Accounting for the different ways in which coltan and gold can be taxed, he argues that armed actors in gold

(16)

16

municipalities rely heavily on the civilian population if they want to benefit from the demand shock (de la Sierra, 2017). I will further elaborate on these alternative explanations in the discussion section and indicate why I think that economic resources provide a valuable argument when trying to explain recruitment mechanisms.

Coming back to my theoretical framework, I will first test Beber and Blattman’s theory about initial resource endowments and their correlation with armed groups’ recruitment strategies since I will use their cross-national dataset for my case selection. Testing both their theory and my alternative at the same time has the advantage of finding strengths and weaknesses of the latter. My second hypothesis can be seen as a continuation of Beber and Blattman’s theory by including a temporal variation in child recruitment.

H1: Armed groups with initially low support from civilians and resource bases that neglect

human rights, such as natural resources and foreign government support, should be more likely to recruit children by force.

H2: An increase in armed groups’ economic endowments (natural resources and foreign

government support) leads to a change from voluntary to forced child recruitment.

IV. Research Design

i. Methodology

To test my hypotheses, I will use a nested analysis, “a unified mixed method approach to comparative research” (Lieberman, 2005, p. 435). To begin with, I will look at the quantitative dataset provided by Beber and Blattman (2013). Their conducted cross-national study, which I will elaborate on in the next paragraph, narrows down the universe

(17)

17

of potential cases that could apply to my theory. This existing large-N analysis (LNA) helps me with my case selection strategy (Lieberman, 2005).

Next, I will conduct a small-N analysis (SNA) in which I will test an existing as well as an alternative theory. I will evaluate Beber and Blattman’s argument by using the evidence from my case study as well as test an alternative hypothesis. By comparing two distinct cases, in my context two different armed groups within one country, I allow their distinguishing characteristics to facilitate theoretical reflections (Bryman, 2016, p. 28). I will thus conduct a within and cross-case analysis. More specifically, I will look at two opposing armed groups within the same country and observe if and how their resources and child recruitment strategies have changed over time. I will then analyze how this change differs across the two cases.

The methods used for my data collection consist of analyzing briefing papers and policy documents from non-governmental organizations, as well as examining research articles and qualitative data from former child soldiers and civilians who have experienced the conflict in areas controlled by the armed groups in question.

With this approach I aim to test the causal effect of an armed group’s economic resources on its child recruitment strategies. If there is an association between the relevant variables of my second hypothesis, I will further test whether or not my proposed causal mechanism is correct. The mechanism connecting X to the occurrence of Y is a decrease in civilian support. This support however is difficult to measure which is why I will expect to only partially be able to find evidence on whether the mechanism links my independent to my dependent variable.

ii. Case Selection

For my case selection I will draw on Beber and Blattman’s article “The Logic of Child Soldiering and Coercion” (2013). They first examine the incentives for armed groups to recruit children and suggest that groups who use coercive recruitment strategies are more

(18)

18

likely to recruit children. Their second statement is that low support from the civilian population and natural resources, as well as foreign government sponsorship as a resource base, makes armed groups more likely to recruit children by force (Beber and Blattman, 2013, p. 95).

What is of particular importance to my paper is their cross-national dataset of rebel groups in 12 African countries, with which they want to investigate the variation in the costliness of punishment. The latter implies that forced child recruitment is only useful to armed groups when it is not costly. This means that they should solely apply this tactic when they have little civilian support or important funding actors to lose. In case donors condemn their partners for violating human rights and cease to support them, punishment is considered expensive (Beber and Blattman, 2013). This cross-national survey provides data on forced child recruitment and links them to armed groups’ sources of material support such as civilian and foreign government support, as well as access to natural resources. They were looking at all conflicts in sub-Saharan Africa that have been reported by the Uppsala Conflict Data Program and Peace Research Institute Oslo between 1980 and 2004. They stratified twelve out of forty-two conflicts by the number of conflicts reported in a country. They further collected information on all nonstate actors during that period, from which they excluded coups. Even if this random sample of African rebel groups has to be used with caution due to scarcity and inconsistency of data, it presents evidence on the variables that are relevant for my theory (Beber and Blattman, 2013, pp. 95-96).

I further downloaded their supplement materials in which I found the Excel sheet with the variable coding for the cross-national survey of the African rebel groups (see Appendix A). The coding values however were not included, until I discovered them in the appendix of an earlier article by the same authors called “The Industrial Organization of Rebellion: The Logic of Forced Labor and Child Soldiering” (Beber and Blattman, 2010) (see Appendix B). This document in combination with the Excel sheet got me full access to

(19)

19

quantitative data of 10 African countries, their rebel groups, methods of recruitment and initial endowments (they removed Gambia and Burkina Faso because they were only single military coups rather than lasting insurgencies) (Beber and Blattman, 2010).

First, I removed Uganda from the list of potential cases since Beber and Blattman (2013) conducted their case study on this country, finding out that “this coercive strategy came cheaply; with no civilian support to lose, and material backing coming from one of the regimes least sensitive to human rights (Sudan), the LRA [Lord’s Resistance Army] faced little penalty for its abductions and abuses” (Beber and Blattman, 2013, p. 89). I then selected the first country in the document, which is the Democratic Republic of the Congo. Being aware that child soldiers are a big issue in the conflict and natural resources abundant, I decided to take a closer look at the available literature for this topic. As expected, literature on child soldiers in the DRC, as well as information about armed groups’ access to natural resources and their cooperation with foreign governments, are adequately given. The fact that I am proficient in the French language, which is the official language in the DRC, gave me the opportunity to benefit from an even wider range of literature.

My aim is to compare two armed groups within the DRC, one fighting against and one supporting the government, in order to strengthen my evidence. It is interesting to observe in how far recruitment strategies vary depending on whether a group gets support from its own or a foreign government, or maybe even no support at all. Finding similarities or differences between two groups can help future research in the area of child recruitment. In order to keep selection bias low, I chose two armed groups based on my independent variable, being their resource endowments (King et al., 1994). It is also important to highlight the timeframe of my analysis, considering the fact that the DRC has been in numerous conflicts since the 1990s. I will focus on the Second Congo War (1998-2003) since there were numerous local and international actors at play, some of which were accused of committing human rights abuses. The war is further known for its grave

(20)

20

consequences on the Congolese population, having resulted in at least one million deaths and the internal displacement of about two million people (Nest, 2011, p. 75).

The next step consisted of selecting two groups out of Beber and Blattman’s dataset. It has to be noted that the rebel groups in their dataset were selected randomly (Beber and Blattman, 2013, p. 96). Out of the five armed groups that the authors proposed for the DRC, only three were active throughout the whole Second Congo War: the Mouvement pour la Liberation du Congo (MLC), the Rassemblement Congolais pour la Démocratie (RCD) and the Rassemblement Congolais pour la Démocratie – Mouvement de Liberation (RCD-ML). Given that the sources of each group’s financial and material resources all include natural resources and foreign governments (coded 1 and 2 respectively), I chose the group that has been active for the longest period of time, the RCD. Being involved in the conflict between 1998 and 2006, this group gives me the opportunity to observe its full evolution and hopefully to make conclusive claims about the correlation of my variables. Since Beber and Blattman do not provide me with an armed group that is supporting the Congolese government during its second war, I needed to conduct some further research. I then came across the Mai-Mai militia groups who were actively fighting against the Rwandan backed RCD, in particular the splinter group RCD-Goma, during the Second Congo War (Global Security). The Mai-Mai are particularly known for their social endowments and later on also their activities with third parties in order to accumulate their wealth (Laudati, 2013; Morvan, 2005). Having different resource endowments and being competing actors in the war, I want to see if the accumulation of resource wealth of the RCD and the Mai-Mai during the Second Congo War has had an influence on their child recruitment strategies.

It has to be noted that the Mai-Mai form a variety of groups differing in size and capacity, which is why some of them are difficult to map (Coalition to Stop the Use of Child Soldiers, 2010). Most of the literature therefore, and all of the articles I will use, hardly differentiate between the numerous groups. Additionally, there are a few ways to spell the

(21)

21

name of the armed group and I (randomly) selected the one that I have been using so far: Mai-Mai. When it comes to the RCD, we may not forget that they were officially only active until 1999, which is when the group split into more fractions. What was then left of the initial RCD was renamed RCD-Goma (Richards, 2014). When reporting about incidents after 1999, some documents and articles still refer to the newly named RCD-Goma as RCD. Nevertheless, I decided to separately use their names according to the time period.

Using the Mai-Mai for my analysis will unfortunately increase my measurement error. Since the literature in use barely specifies which group is being talked about, I can only make conclusions about the Mai-Mai in general. For example, it may be that one group is engaged in trading minerals whereas others are not. If I found evidence that some Mai-Mai groups started recruiting children by force, I will not know whether they are the ones who are also exploiting natural resources. I can thus only make conclusions about the Mai-Mai by looking at the overall effects. Without access to more detailed reports about the militia, clear distinctions between the different groups cannot be made.

I will now proceed by explaining how I will measure my variables and propose clear indicators.

iii. Data

Proposing clear indicators for variables that need to be measured has the intention of enhancing reliability. For both hypotheses, the dependent variable is forced child recruitment which I will measure by looking for evidence about any kind of recruitment that has not been voluntary. Threats, coercion and abduction thus fall into this category.

Measuring recruiting strategies is however not always easy since the organizations and authors writing those papers rely on perceptions by former child recruits. As already mentioned in the literature review, children may think they joined voluntarily while in fact coercive measures have been used to convince them. Nonetheless, reports by

(22)

22

organizations as well as research articles all make a clear distinction between voluntary and forced recruitment, which is why I will continue using these two categories.

The independent variable in the analysis is an armed group’s resource endowment. While in the alternative theory I only consider the economic resources, Beber and Blattman (2013) also highlight the importance of social endowments, being civilian support. Testing their existing theory, I will measure civilian support the way Beber and Blattman did, meaning “[…] (non-coerced) support of the population in its base region within [its] home country on which it could draw for resources, cooperation, or harbor” (Beber and Blattman, 2010, Appendix). This support can be seen when an armed group receives voluntary funds or other kind of assistance from the local community. I also try to find written proof in several documents about a group’s popularity and cooperation with the civilian population. The taxation of the population by armed groups is not included in the definition, since it hardly reveals whether this support is voluntary or coerced. Furthermore, taxation is sometimes viewed as a complex relationship between civilians and armed groups, meaning that various factors influence where and when taxation is applied (de la Sierra, 2017).

The indicators for economic resources are in both hypotheses access to natural resources and support from foreign governments. For the former, I will draw on Beber and Blattman’s definition again, being that of “[…] territorial and natural resource control that is convertible into monetary resources” (Beber and Blattman, 2010, Appendix). I will further incorporate the collaboration of armed groups with, as well as taxation of companies that work with natural resources. The United Nations (2001) also gives an elaborate explanation about what they mean by the exploitation of natural resources: “[…] extraction, production, commercialization and exports of natural resources and other services such as transport and financial transactions” as well as “[…] confiscation, extraction, forced monopoly and price-fixing” (United Nations, 2001, pp. 5-6). Support from foreign governments is being measured in financial, material and human resource related reinforcement by mostly

(23)

23

neighboring countries. More specifically, if a foreign government is providing an armed group with money, weapons or soldiers, it is contributing to a group’s resource wealth. I will also need to find evidence on the increase of resources at any point of time.

The indicators for civilian support unfortunately have some weaknesses. Cooperation between communities and armed groups can often only be observed when on site. When interviewing locals and asking about their relationship with a group, fear of retaliation may lead them to refrain from answering the questions. Additionally, support can also be divided into voluntary and coerced. The statement that a group relies on civilian support thus has to be treated with caution. In order to avoid measurement error, I will try to find evidence from numerous sources.

By including some of the same definitions for the variables as Beber and Blattman (2010), I am trying to strengthen the validity of my measurements while also taking into account their limitations. In the next section I will introduce the background of the Congo conflict, including some important background knowledge about its international dimension, natural resources, ethnicity and the problem of child soldiers in the country. I will then turn to a more detailed description about the formation of the RCD and the Mai-Mai.

iv. Historical overview

The next paragraphs should help to understand how multiple factors shaped the conflict and the formation of the armed groups in the DRC.

(24)

24

a. Background of the conflict

Administrative map of the Democratic Republic of the Congo.

Source: http://www.nationsonline.org/oneworld/map/dr_congo_map.htm

The Democratic Republic of the Congo, a country located in the heart of central Africa, is a relatively young country that gained its independence only in 1960 and has been named Zaire by its president Mobutu in 1971. Being rich in natural resources, it has been exploited by businesses, politicians, rebel groups and neighboring states ever since the Belgian colonization in the late 1800’s (BBC, 2016). The conflict is further known to be a mixture of local, national and international actors, all acting upon their own economic interests (Vlassenroot and Raeymaekers, 2004). The consequences of its two wars (1996-2003) are still being felt today, with numerous armed groups constantly forming and dismantling.

Joseph Mobutu became president in November 1965. During his term in office, Mobutu set up a system to control and exploit the country’s natural resource wealth for his own purposes, leaving the Congolese population in bad living conditions (Global

(25)

25

Witness, 2009, p. 5). The heterogeneous composition of the Congolese population in the East, as well as the presence of immigrant communities inhabiting the same territories, has led to regular disputes about the access to and control over land. In the 1990s, a lot of local conflicts could be explained as “the political manipulation by local politicians of the rights of access and control of land” (Vlassenroot and Raeymaekers, 2004, p. 41).

As of 1994, the country has experienced further instability near its boarders with Rwanda. More than one million Hutu refugees fled from the neighboring country into the DRC as a consequence of the genocide. Among them were militias who were accused of being responsible for the horrible act and who started launching attacks from the refugee camps on Rwanda (Vlassenroot and Raeymaekers, 2004, p. 54). This sudden influx of refugees “altered the existing demographic dynamic and ethnic balance in the eastern part of the Congo” (Moyroud and Katunga, 2002, p. 160).

The country experienced its first civil war between 1996 and 1997. The seven months conflict came to an end when the former autocratic president Mobutu was ousted in May 1997 by the Alliance des Forces Démocratiques pour la Libération du Congo (AFDL), an armed coalition led by Laurent Kabila and supported by Rwandan and Ugandan troops, as well as various Mai-Mai groups (Richards, 2014, p. 304). Mobutu fled the country and Kabila declared himself president, renaming the country Democratic Republic of the Congo.

In 1998, Kabila replaced his second Rwandan army commander and turned his back on his former backers in Uganda, Rwanda and Burundi by ordering them to leave the country. Even though he was still dependent on the support of his external allies, Kabila did not want to be portrayed as a Rwandan puppet, which is why he distanced himself from his neighbors (Vlassenroot and Raeymaekers, 2004, p. 49). As a result, Rwanda and Uganda created an anti-Kabila revolt with the goal of overthrowing him. They feared that the tensions along the eastern boarders of the DRC would hinder their goal of security and peace. Neighboring countries like Angola and Zimbabwe fortunately came Kabila to

(26)

26

the rescue. Only a few months later however, different irregular armed groups emerged and created an imbalance in the country, particularly in the East (Kisangani, 2003, pp. 52-57). A second attempt by Rwanda, Uganda and this time also Burundi to launch a coup against the president resulted in a messy armed conflict. Troops from Angola, Chad, Namibia, Sudan and Zimbabwe were once again backing up Kabila, but the economic and infrastructural consequences were devastating (Global Security). Since 1996, the national oriented agenda transformed into a more local and individual agenda, “leading to a total fragmentation of the politico-military landscape”, all of which led to the multiplication of armed actors (Vlassenroot and Raeymaekers, 2004, p. 51).

Kabila was assassinated in January 2001, making his son, Joseph Kabila, the new president. In autumn 2002, he successfully negotiated the withdrawal of Rwandan troops from the eastern part of the country. With the Pretoria Accord signed only two months later, Joseph Kabila was able to set up a transitional government in July 2003, which further resulted in presidency elections in 2005 and provincial legislatures in 2006 (Global Security). This post-transition period has however seen an increase in violence due to the state’s inability to tackle local conflicts. All the “flawed peace agreements” helped former rebels to integrate into the national army, all while preserving their illegal networks (Peace Direct, 2009).

In order to get a better understanding of my results I will give an overview of the main indicators of my variables, only this time they are directly related to my unit of analysis, the Democratic Republic of the Congo. The next paragraphs will briefly introduce the international dimension of the conflict, the situation about natural resources, ethnic diversity in the country in order to understand civilian support and the problem of child soldiers in the DRC.

(27)

27

b. International involvement

The conflict in the Congo, especially during 1996 and 2003, is also known as “Africa’s World War” due to its transnational character. Not only local rebel groups, but also neighboring insurgents and African governments have participated in the conflict for numerous reasons (Williams, 2013). In order to get a better understanding of the international dimension of the First and Second Congo Wars, I will briefly introduce the foreign governments involved and their reasons for participation.

Rwanda was the country that was probably most involved in the DRC. Its stated reasons were the growing insecurity and raids by the Hutu Interahamwe from eastern Congo into the country. Not only was president Kabila not able to control the attacks, but Rwandan officials gained knowledge about him training and arming their enemies. Furthermore, Congolese Tutsis were getting more and more into conflict with native opponents and Hutu refugees from Rwanda. The Rwandan Tutsi government thus saw the need to intervene for humanitarian concerns. Nevertheless, the Second Congo War increasingly demonstrated the country’s hidden motivations, being personal and national enrichment (Williams, 2013, p. 91).

Uganda’s involvement can also be explained by a security dilemma concerning the

attacks by the Ugandan resistance group called Forces Démocratiques Alliées (ADF),

located in eastern Congo. Together with Rwanda, they shared the goal of overthrowing Kabila. Ugandan actors were furthermore driven by economic motivations, being that of benefitting from the Congo’s mineral trade (Williams, 2013, p. 92).

Burundi, similar to Rwanda, experienced a local conflict between the Hutu and Tutsi

population. It saw the formation of the rebel group Forces pour la Défense de la

Démocratie (FDD) who used the DRC as their sanctuary and were armed by Kabila in order to resist Rwandan and Ugandan forces. Burundi thus saw itself forming an alliance with the anti-Kabila states in the interest of repelling the FDD (Williams, 2013, p. 93).

(28)

28

Angola on the other hand was considered the most important state opposing the Rwanda-Uganda alliance. Together with Zimbabwe, Namibia and Sudan, these countries supported Kabila for different personal reasons which I will not further elaborate on now (Williams, 2013).

With this introduction I aim to point out not only the international dimension of the conflict, but also the complexity of interests and thus alliances of neighboring countries.

c. Natural resources

The DRC is rich in a multitude of natural resources, reaching from diamonds, gold, cassiterite (tin ore) and coltan to coffee, timber and palm oil. All of these minerals can be found in the eastern regions of the DRC, making it a perfect battlefield for local and international actors who seek to make easy economic profits (Amnesty International, 2003; United Nations, 2010). These economic interests largely contributed to the nature of the Congolese conflict, where rebel movements, businesses and foreign governments were and still are engaged in the illegal extraction of natural resources (Vlassenroot, 2008). “Thus, the quest to control natural resources have shaped the power strategies pursued by the warring parties that have increasingly been based on the territorialisation of authority around valuable resource areas and trading networks” (Vlassenroot, 2008, p. 3). The practice of resource pillaging, corruption and lack of accountability characterized Mobutu’s and later also Kabila’s regime. It has to be noted that under Mobutu, the mining and forestry sectors had been almost entirely controlled by civilians (even if some of the profits were handed to the military). With the beginning of the first war, these sectors suddenly fell into the arms of the new national army and various armed groups. This militarization of regions for important sectors triggered increasing violence against the Congolese population (United Nations, 2010, pp. 320-321).

The second war then saw an even higher attractiveness for resource exploitation because it helped armed groups to finance their warfare and became a source of personal

(29)

29

enrichment for political and military leaders. The creation of so called “front” companies enabled foreign governments to extract natural resources in collaboration with violent armed organizations without even visibly participating (United Nations, 2010, pp. 321-322).

d. Ethnicity

The DRC is the second most ethnically diverse country in the world, counting 200 diverse groups (Karbo and Mutisi, 2012, p. 382). The North and South Kivu, as well as Ituri regions, are particularly rich in different ethnic groups. During its Belgian colonial administration in the Kivus, different waves of migration highly influenced local society. Between 1937 and 1955, a large number of Hutu-Rwandans were placed in the Kivus as coffee, tea or cotton laborers. These lands were however controlled by different ethnic groups, creating the initial phase of tensions. North and South Kivu further experienced an influx in Rwandan Tutsi refugees due to its Social Revolution. The resulting disputes have been mainly about the access to and control over land (Vlassenroot and Raeymaekers, 2004, pp. 40-41).

These disputes escalated in the early 1990s, when Mobutu encouraged local politicians to mobilize their followings according to their ethnic background. This mobilization and manipulation of entire ethnic communities led to serious levels of violence. “Local politicians and commercial interests mobilized the population on the issue of land by the use of self-serving ethnic discourse” (Vlassenroot and Raeymaekers, 2004, p. 41). Traditional authorities also used these campaigns to regain or increase their control over land and political power. Local trust and cooperation thus became limited to ethnic belonging. This “stirring up of ethnic resentment” is of importance when analyzing the formation of the numerous armed groups. With this discourse, traditional authorities and politicians were able to mobilize young locals into militias. The result was a progressive proliferation of militias, all being supported by their ethnic groups (idem, pp. 42-46).

(30)

30

With this introduction into the Congo’s fractured society, I aim to point out the importance of ethnicity when analyzing civilian support for armed groups like the RCD-Goma and Mai-Mai.

e. Child soldiers

The DRC is known for its large scale recruitment of children. Of all newly recruits in the country, the UN estimates that 15-30% are under the age of 18. Child soldiers in the DRC are also called “kadogos”, meaning little ones in Swahili (Coalition to Stop the Use of Child Soldiers, 2001, p. 144). Armed groups often justify their actions by pointing out the number of soldiers needed to overthrow the government. The lack of military training within rebel groups makes it easy for children to fight among unprofessional combatants. A lot of children in the DRC join voluntarily in order to protect themselves, defend their communities or survive economically. However, children are also frequently recruited against their will. Child recruitment in the DRC has particularly gained national and international attention since 1996 with the AFDL (United Nations, 2010, pp. 307-308).

While during the first war it was mainly the AFDL who recruited large numbers of children, the second war saw a tremendous increase. After the coup against Mobutu in 1997, a large part of these children were integrated into the new government army. In 2003, according to the Report of the Secretary-General on children and armed conflict, 12 parties have been involved in recruiting children into their ranks, including the RCD-Goma and Mai-Mai (United Nations, 2010, p. 309).

v. Background of the armed groups

I will now introduce both armed groups that are relevant for my analysis and give an overview of their formation. The RCD-Goma and Mai-Mai groups constituted competing forces in eastern DRC during the Second Congo War (1998-2003).

(31)

31

a. RCD and RCD-Goma

The Rassemblement Congolais pour la Démocratie (RCD), supported by Rwanda and

Uganda, was organized in Kigali, Rwanda, with the aim of ousting Kabila. “This rebellion, however, was more than a simple repetition of the AFDL campaign, as former allies now became adversaries, new coalitions proved to be fleeting and economic interests became guiding principles” (Vlassenroot and Raeymaekers, 2004, p. 50). In August 1998, Ernst Wamba dia Wamba became its chairman. With Wamba being concerned about the increasing Rwandan influence, he got toppled ten months later and created the RCD-Kisangani in March 1999. After experiencing continuous fighting between Rwanda and Uganda over the control of Kisangani, Wamba relocated his office to Kampala and established the Rassemblement Congolais pour la Démocratie – Mouvement de Liberation (RCD-ML) (idem, pp. 58-60). During the last months of what was still known the RCD, in early 1999, the rebel group together with Rwandan and Ugandan forces controlled over one-third of the DRC (Global Witness, 2004, p. 10).

The remaining RCD became thus known as RCD-Goma and was backed by Rwanda, whereas the RCD-ML was now supported by Uganda (Richards, 2014). RCD-Goma, the Rwandese Patriotic Army (RPA) and the Rwandese Army controlled until late 2002 most of the North and South Kivu provinces, parts of the Maniema province, Kasai Orientale and Oriental province, including Katanga and Kisangani (Amnesty International, 2003, p. 7). The RCD-Goma also had an armed wing called Armée Nationale Congolaise (ANC), which were however known to be undisciplined and badly trained. The combatants were violent, often looting and burning down whole villages in order to get food or acquire property (United Nations, 2002, p. 16).

With the beginning of the Second Congo War, ethnic communities in the Kivu provinces slowly experienced a division: those supporting and those opposing the RCD. One of the main opposing armed forces were the Mai-Mai (Vlassenroot and Raeymaekers, 2004, p. 57).

(32)

32

b. Mai-Mai

Since the 1960s, eastern Congo experienced a mixture of different ethnicities. Workers from Rwanda, known as Rwandophones, a Tutsi Banyarwanda community, later called Banyamulenge, and native communities all lived in the East of the country. Years of tensions erupted into violence in North Kivu in 1993, leading to the proliferation of local defense militias by the native communities, called the Mai-Mai. They constitute a variety of groups differing in capacity and size. The constant emergence of new and the dissolution of former Mai-Mai groups makes them difficult to map (Coalition to Stop the Use of Child Soldiers, 2010, p. 2). Their goal was to “[…] keep land and positions of traditional authority out of the hands of the Rwandan-backed rebellion […]” (Richards, 2014, p. 314). They are known for their war rituals that convey the belief of “mayi”, magic water, to protect the combatants from bullets. Their nationalist ideas are particularly attractive to young people (Jourdan, 2001, p. 90). Stearns (2011) describes the demographic of the Mai-Mai as young, unemployed and disaffected (Stearns, 2011, p. 263).

In 1996, the campaign of the AFDL drove most of the Mai-Mai groups to ally with Kabila and his Rwandan and Ugandan supporters. The second war however saw a change in alliances when the Mai-Mai realized the Tutsi influence within the AFDL. The Mai-Mai soon started opposing the RCD, a rebellion considered to be a Tutsi attempt to regain Kivu, and thus allied with the government of Kabila which gave them military support (Jourdan, 2001, pp. 94-95).

The last few years saw an increase in Mai-Mai groups as a response to the growing insecurity in the rural areas. While some of the groups are still providing protection and regulation in the areas they control, others are guided by violent ethnic ideologies and involved in plunder (Peace Direct, 2009).

(33)

33

vi. Analysis

Before diving into my analysis and discussing the results, I will state which observations I expect to see. By stating my expectations I set clear guidelines concerning the confirmation or rejection of my hypotheses.

First, it is important to mention which kind of variation I included in my hypotheses and therefore also aspire to see in my results. My second hypothesis entails a temporal variation while observing if and when resource bases of an armed group change. If they do, can we also witness a change in a group’s child recruitment strategies? More specifically, if at one stage an armed group experiences an increase in its resource wealth, we should also observe a shift from voluntary to forced recruitment of children. This shift should appear within the next six months. One year would be too wide of a gap and something else could have provoked this change in recruitment. One or two months on the other hand would be too soon to observe such a big change in an armed group’s recruiting tactics. So, if I find evidence for example that one armed group had an increase in economic resources in the month of January, I expect to observe that by July the latest it started recruiting children by force (whereas before it only recruited voluntarily). The time frame of six months can however be subject to change in future studies about the variation in recruitment tactics.

For my first hypothesis I expect to have the following observations: in case the Mai-Mai and RCD-Goma both have initial low civilian support and economic resources at their disposal (access to natural resources and/or support from foreign governments), they will consequently recruit children by force. Here I will focus on the initial resource endowments and recruitment strategies of a group. Forced recruitment should be observed during the same time where civilian support is low and economic resources rather good. I can however also confirm my first hypothesis in case a group has initial high civilian support, no good economic resource base and does not recruit children by force.

(34)

34

My second hypothesis on the other hand is correct if I observe an increase in the economic endowments of both armed groups, which will then as an effect have a switch from voluntary to forced child recruitment. Ideally, I will also observe a decrease of civilian support as causal mechanism. This means that after an increase in resources we should find evidence soon after about the decline of civilian support due to this development. Here again, I set a timeframe of six months. This decline of support then also needs to affect forced recruitment of children. With a higher resource base independent from civilians, armed groups do not feel accountable for their actions anymore and will turn to a more cost-effective way of recruitment.

In order for my findings to be generalizable, I need to be able to confirm my hypotheses for both armed groups. The ultimate goal of this research is to place my case study within the context of a broader universe of units.

It is, on the other hand, also important to mention how I would know that I can falsify my hypotheses. For Beber and Blattman’s theory, I would reject the hypothesis if initial low civilian support and a good economic resource base did not lead to forced recruitment of children. Furthermore, if I only observe low civilian support but no economic resources or vice versa, I would have to refute the theory, no matter the outcome on the recruitment strategy.

The second hypothesis can be falsified if I observe an increase in the economic resources of an armed group but no change in their method of child recruitment (meaning from voluntary to forced). Should I observe a change in recruitment but no increase in economic resources, I know that my theory does not hold and other factors are responsible for the variation.

I will further proceed by summarizing my observations for both cases. I will link the key findings of my causal mechanism, independent and dependent variables to each armed group in the following order: foreign government support, access to natural resources, child recruitment strategies and civilian support.

(35)

35

a. Data collection 1. RCD-Goma

Foreign government support

The reliance of the armed group on foreign governments, particularly on its neighbor Rwanda, has been highlighted in numerous articles and policy documents (Richards, 2014; Amnesty International, 2003). According to reliable sources, Rwanda and Burundi both supplied RCD-Goma combatants with arms and training. The former included “[…] rocket launchers, armoured cars, machine guns, light artillery, mortars and landmines” as well as various helicopters and aircrafts (Amnesty International, 2003, p. 8). Besides military equipment and troops, Rwanda also appointed the rebels with oil and petrol (United Nations, 2001, p. 28).

In 2002, RCD-Goma claimed around 25.000 soldiers, 15.000 of which were well-trained combatants provided by the Rwandan army. These foreign soldiers constituted an essential part of the armed group’s army, without which it could not have survived (Kisangani, 2003, p. 59).

Rwanda even set up a “Congo Desk” within its borders in order to administer the finances needed for the war in the DRC, but also to handle the wealth from the natural resources and other revenues from the neighboring country. The Congo Desk thus contributed to Rwanda’s military expenses, which in turn benefited its ally, the RCD-Goma (Amnesty International, 2003, p. 12).

In exchange for weapons, RCD-Goma repeatedly provided the Rwandan government with direct payments. To finance these arms, the rebel group collected taxes on mineral exports (Kisangani, 2003, p. 67). Since the beginning of the second war in 1998, a huge amount of cassiterite concessions have been sold by RCD and later also RCD-Goma to Rwanda on a daily basis. The allies even set up an efficient network to transport these resources by planes or trucks (Global Witness, 2004, p. 22). This interdependence between demand and supply benefited both Rwanda and the RCD-Goma. A more detailed

(36)

36

explanation about the extraction of natural resources and its taxation by the rebel group will follow in the next section.

The movement is however also often portrayed as being militarily weak if it would not have been for the numerous superior forces of the Rwandese army. “The RCD are the allies of the RPA [Rwandan Patriotic Army], not the other way around” claims a Congolese human rights defender (Amnesty International, 2003, p. 7). Claiming to be the true political as well as military authority in the provinces it controls by undertaking local administrative functions, it has nonetheless constant backup from Rwandese officials. A lot of policies that needed to be implemented were decided upon and approved in Kigali, Rwanda, instead of eastern Congo (ibid.).

As we see, the RCD-Goma, as well as its RCD predecessors, depended highly and continuously on their allies in Rwanda and Burundi. Without their financial, military and material support, the rebel group would have probably either been defeated in a small amount of time or it had to look for different donors.

Access to natural resources

Evidence shows that the Rwandan army and the RCD-Goma gained significant revenue from the trading of diverse natural resources around Kisangani (United Nations, 2001; United Nations, 2002; United Nations, 2010). The rebel group also focused on mineral rich areas in the Kivu-provinces and adopted a strategy called the “archipelago” approach to economic accumulation (Vlassenroot, 2008, p. 5).

The illegal activities with the country’s natural resources can be divided into two phases: during the first year resource stockpiles were looted, while after that period one could observe the direct extraction of minerals as well as taxation of mineral companies (United Nations, 2001, pp. 9-10).

During the first two years of the second war in 1998 and 1999, the RCD took control of the company SOMINKI (Société minière et industrielle du Kivu) and claimed 6 tons of

(37)

37

coltan and 200 tons of cassiterite worth around 722.000 dollars (United Nations, 2001, p. 8). In order to manage the finances and transfers of resources that have been collected, RCD-Goma established SONEX in 1999. The year 2000 saw a change in this administration when the group saw the establishment of the company SOMIGL (Société minière des Grands Lacs), from which it granted monopoly on all coltan exports within the rebel territory control. This monopoly helped to finance the group’s war efforts (Global Witness, 2004, pp. 21-22).

In the year 2000, an RCD-Goma leader reported that the group raises around 200.000 dollars a month from diamonds and some one million dollars monthly from coltan (Kisangani, 2003, p. 69). The extractions were higher in that year than in 1999. In 2000, 100kg of gold and 29 tons of coltan a month instead of 60kg of gold and 27 tons of coltan were extracted. Even though more exact figures are not available, evidence suggests that the financial situation of the RCD-Goma has been improving since late 2000 (United Nations, 2001, p. 32). These increased profits can be explained by a boom in coltan in the year 2000 due to a high demand of cell phones and the Christmas release of the Sony PlayStation console. The price of tantalum, the processed form of coltan, was growing and

numerous comptoirs (mineral trading houses) opened up in Bukavu and Goma. Thus,

between June 2000 and July 2001, the RCD-Goma, the Rwandan government and their business partners experienced record profits (Stearns, 2001, pp. 299-300).

Not only did the armed group exploit and trade with the country’s natural resources, but it also collected taxes on businesses who were engaged in such activities: “[…] for economic agents, connection with the rebels was a necessary condition to continue or expand their activities and to increase the predictability of commercial activities in terms of logistics and revenues” (Vlassenroot, 2008, p. 8).

The relationship between the rebels and businessmen meant money for the armed group in return for military protection and a reduction of taxes (Vlassenroot, 2008, p. 9).

(38)

38

RCD-Goma demanded mining licence fees, non-refundable deposits and various export taxes from companies (Amnesty International, 2003, p. 33).

Another important way of increasing their revenue was by shipping gold, diamonds and timber from the DRC through Dar es Salaam in Tanzania, their financial and logistical network. From there, the minerals were sold on to the international market. In exchange, the armed group received cash, arms, medicines, food or other logistical needs (United Nations, 2001, p. 32).

Rwanda’s increase in its mining profits slowly had consequences for the RCD-Goma in that the rebel group’s share then decreased. Rwanda continuously deprived its ally of significant shares and prerogatives. A 2001 Panel report by the United Nations however reported that the RCD-Goma was gradually becoming autonomous in that it was capable of raising big amounts of cash, purchasing its own equipment and using its own networks (United Nations, 2001, p. 32).

Child recruitment strategies

The RCD and later also the RCD-Goma were among the groups who recruited most children into their ranks. At first, and with the beginning of the second war in summer 1998, the armed group started recruiting children on a voluntary basis by publicly announcing their awareness raising campaign directed at the parents (United Nations, 2010, p. 310). They did this particularly through broadcasts on the local radio station (Human Rights Watch, 2001).

A year later in 1999, forced recruitment strategies already started to be applied. A former RCD instructor explained:

“I was sent to certain zones to recruit people for the RCD army. Because none would volunteer, the RCD forced people into joining its army. In 1999, I was involved in the recruitment of children aged eleven to fifteen in the localities of Kasongo and Kalima in

Why do armed groups sometimes switch from voluntary to forced recruitment of childres? : a case study of the Democratic Republic of the Congo