Who returns home? : explanation of the regional variation in internally displaced returnees with the case of Eastern region in the Democratic Republic of the Congo

(1)

Merel Hendriks | 10823247

University of Amsterdam | 24 June 2016

Master Thesis Political Science | International Relations Number of words | 14.711

Explanation of the Regional Variation in Internally Displaced Returnees with the Case of Eastern Region in the Democratic Republic of the Congo

WHO RETURNS HOME?

(2)

Abstract

Internal displacement has been an ongoing issue for years in most of Africa and for the particular case studied in this thesis, the Democratic Republic of the Congo (DR Congo). In 2015, 1.6 million people were internally displaced in DR Congo, yet the most surprising figure is the low number of returnees. Compared to 2014, eight times less internally displaced people have returned to their place of origin in 2015. This is a striking decrease and varies in each region, which this thesis aims to explore by answering - What explains the regional variation in internally displaced returnees? Several possible explanations were identified, and by using qualitative- and quantitative methods these different explanations were explored. The results of this thesis suggest that, in Africa, the presence of mineral mines leads to a significant decrease in returnees. This is also the case in DR Congo. There is an inverse relation between mines and returnees, in which more mines are associated with fewer returnees in a region. More violent incidents also reduce the number of returnees. There was no significant relation between the quantity of armed groups and the number of returnees. By testing the different types of mines individually, I found that gold and wolframite have had the greatest impact on the number of returnees. To reduce the number of internally displaced people and to increase the number of returnees, I would advise the development of a

transparent supply chain for the different natural resources. People may be more willing to purchase merchandise made from products originating from sustainable mines, reducing the negative impact on the citizens of DR Congo.

COVER IMAGE BY DOMINIC NAHR - NATIONAL GEOGRAPHIC

Some of the tens of thousands of Congolese displaced from the area around sake by clashes between M23 and Congolese soldiers in 2012 flee on the road linking Bukavu and Goma.

(3)

Preface

From the start of my MA thesis project, I preferred writing on a topic concerning Africa. After delving into several potential topics, I discovered the large degree of variation in the number of returnees after internal displacement. It was motivating to choose a topic that had not yet been comprehensively explored, and that could contribute to resolving issues in several African countries that are poorly understood. When I visited Africa for the first time, approximately eight years ago, I was a teenager. At the time, I had never reflected on the difficult issues that a country such as the Democratic Republic of the Congo has had to, and continues to, deal with. After many years of reading, learning, and writing, I have come to know a lot more and that is why this topic appealed to me. Although I have learned a lot throughout the writing of this thesis, it has also reminded me that there is much more to learn. I am very motivated to do so, which is why I am looking forward to continuing working in a field concerning Africa, and especially natural resources, in my future career. Please do not forget to take a look at my short video (1.23 minutes), on the USB-stick, where I summarize my findings visually.

Acknowledgements

Many thanks to my supervisor dr. Seiki Tanaka, for his precious advice and guidance, and to dr. Ursula Daxecker for taking the time to be the second reader of my thesis. I would also like to thank my interview respondents, who provided me with information and extended my network: Fleur Scheele (Stichting Onderzoek Multinationale Ondernemingen Amsterdam), Han van Dijk (African Studies Centre Leiden), Dirk-Jan Koch (The Dutch Ministry of

Foreign Affairs), and Clens Santa (former refugee of the Democratic Republic of the Congo). My final thanks goes to Alexandre Jaillon of the International Peace Information Service. Although his research has not yet been published, he was willing to share his data for my thesis. Thank you all for supporting me through this process, and I wish everyone the best with their further work.

(4)

List of tables, figures and models

Table 1: Hypothesis argument 1 11

Table 2: Hypothesis argument 2 12

Figure 1: The regions of DR Congo highlighted in this thesis 21 Figure 2: The regions of DR Congo highlighted in this thesis including IDP, 25

returnees, mines and armed groups

Table 3: Includes model 1 and 2 26

Model 1: Results of regression analysis Africa exclusion of armed conflicts 26 Model 2: Results of regression analysis Africa inclusion of armed conflicts 26 Table 4: Number of returnees across the different regions in 2013, 2014 and 2015 27 Table 5: Number of IDP across the different regions in 2013, 2014 and 2015 28

Figure 3: Scatterplot returnees in DR Congo 28

Figure 4: Scatterplot IDP in DR Congo 29

Model 3: Results of regression analysis DR Congo 30

(6)

Who Returns Home?

Explanation of the Regional Variation in Internally Displaced Returnees with the Case of Eastern Region in the Democratic Republic of the Congo

1. Introduction

“Internal displacement is the great tragedy of our times, internally displaced people are among the most vulnerable of the human family”, Kofi Annan (2004), former Secretary General of the United Nations, noted. Internally displaced people (IDP) are often the most forgotten people in global emergencies (Annan, 2004). The global crisis of internal displacement has already been a challenging issue since the late 1980s, and the issue has grown continuously since then (Weiss & Korn, 2006). Historically, the number of IDP has varied substantially, on occasion thousands have fled in a single day, with other periods characterized by little

displacement (Schon, 2014, p. 437-438).

Overall, the total number of IDP globally has been increasing, whereas the number of internally displaced returnees has been decreasing (Schon, 2014). People have been forced to leave their homes due to several concerns, such as armed conflict and the risk of having their human rights violated (Cohen & Deng, 1998, p. 1). However, even when the violence has subsided, most do not return. People end up relying on host families or living in camps that provide temporary shelter on the way to more permanent homes (International Committee of the Red Cross, 2009, p. 12). This is why the return of IDP is especially important and relevant to research, as it is important to look into factors beyond violence to explain why most

individuals do not return.

One of the countries with high levels of displacement and a decrease in returnees is the Democratic Republic of the Congo (DR Congo). In DR Congo, there was a striking decrease in the amount of returnees between 2014 and 2015, with eight times less people returning to their place of origin in 2015 (OCHA, 2015). There are several reasons for citizens to (not) return to their place of origin. The most commonly specified reason is the presence of violence, yet there may be other reasons. By focussing on a country with a high amount of potential returnees, patterns can be discovered that may explain this phenomenon. Therefore, the case study of DR Congo is adopted in this thesis, as to obtain further in-depth insight into the mechanisms underlying displacement. This research involves five regions of DR Congo: Nord-Kivu, Sud-Kivu, Tanganyika & Haut-Katanga, Maniema and Ituri/Orientale. These regions are located in the Eastern part of DR Congo and are characterized by varying degrees

(7)

of conflict.

Several studies have documented the causes of internal displacement in a variety of ways however no previous study has investigated returnees. The work of the different scholars about IDP is taken into account by formulating a new research question about returnees. Preliminary work on displacement, undertaken by Van Dijk et al. (2001), Moore and Shellman (2004), Vlassenroot and Huggins (2005), and Schon (2014) reported causes of displacement. They provided several arguments for internal displacement, such as changes in the balance of power, threats to personal integrity and transitions in political, economic and social structures. Although there is an abundance of literature on migration and mobility in Africa and increasing displacement, a systematic understanding of why returnees decide (not) to go back to their places of origin was still lacking. This thesis addressed this gap with the following question:

What explains the regional variation in internally displaced returnees?

When answering this question, I focus on conflict related variables. Many other variables could play a role in internally displaced returnees, although in a conflict setting violence usually explains internally displaced returnees. Drawing on existing literature, I considered the following argument: The presence of mineral mines in an area attracts more armed groups, in turn leading to more violent incidents between the different groups. This process results in more displacement and less returnees. If this argument was found to be true, there could be various explanations.

This research aimed to make several contributions to research and policy; little is known about internally displaced returnees, internally displacement is a global problem occurring on all continents and results of this thesis could be valuable for other countries that experience similar patterns. By investigation DR Congo, a country with so much IDP, I could develop new insight for a global problem. I will do this by using quantitative- and qualitative methods.

This thesis proceeds according to the follow structure. In chapter 2, a detailed theoretical overview on the existing literature about internal displacement and returnees is provided. In chapter 3, I describe how I formed my argument based on this literature. Chapter 4 addresses the methods of this study. Chapter 5 provides information about internal

displacement and gives an overview of the conflict in DR Congo, different regions and several factors will be explained. Chapter 6 shows the results of the regression analyses and

(8)

the expert interviews. The thesis will end with a conclusion of the findings.

2. Theoretical framework 2.1 Definitions

IDP is defined as “people or groups of persons who have been forced or obliged to flee or to leave their homes or places of habitual residence, in particular as a result of or in order to avoid the effects of armed conflict, situations of generalized violence, violations of human rights or natural or human-made disasters, and who have not crossed an internationally recognized state border” (United Nations Guiding Principles on Internal Displacement, 2004). There is a clear difference between refugees and IDP. Unlike refugees, IDP are still citizens within their country. Their government is legally responsible for their protection and welfare (OCHA, 2010). The number of IDP has differed from time to time. In 2004, the number of IDP peaked during the Darfur crisis and around 2011 during the Arab Spring. The Arab Spring has led to twice as many IDP as refugees. During the Arab Spring, about half a million people had to flee within Libya, Syria, Yemen and Iraq (UNHCR, 2012, p. 73). However, whereas refugees are often attended to be international support, IDP are neglected.

In the conceptualization of internal displacement, a differentiation can be made

between forced displacement and willing displacement, although both involve citizens having to leave their homes. The World Bank defines forced displacement as the situation in which people are forced to flee their homes due to conflict, violence and human rights violations (The World Bank, 2015). The DR Congo has had to deal with such armed conflict and generalised violence, forcing citizens to flee, due to the peril of staying in their homes (Norwegian Refugee Council & International Displacement Monitoring Centre, 2015, p. 3). Van Dijk, Foeken and Van Til (2001) distinguished between people who chose to migrate (willing displacement) and people who were given no other option (Van Dijk, Foeken & Van Til, 2001, p. 13). They mention that forced displacement is widespread in contemporary Africa (Van Dijk, Foeken & Van Til, 2001, p. 20). Forced displacement does not only occur when there are violence, both oppressing economic conditions and natural disasters could also lead to forced displacement (Van Dijk, Foeken & Van Til, 2001, p. 20).

Following OHCHR’s (n.d.) definition of ‘returning refugees’, I conceptualized the return of IDP as: “a person who was an IDP but who has returned to his/her region of origin.” Although, the original definition pertained to refugees that return to their homeland, I used this similar definition for IDP returnees because no functional definition had been

(9)

2.2 Theories

A number of authors have examined the causes of internal displacement (e.g Zolberg 1989; Cohen & Deng 1998; Weiss & Korn 2006; Terminski 2012). These authors found several reasons for internal displacement and returning after being displaced. A first reason for internal displacement is safety. According to Schon (2014), this is the basis of an individual’s decision to leave their place of origin. People are expected to make strategic choices when they decide to leave their homes, based information about local violence. The greatest incentive of leaving their homes is that they can escape from physical danger to themselves (Schon, 2014, p. 439). Besides safety, moving to a non-conflict location can improve access to social services, healthcare, education, or clean water and food (Schon, 2014, p. 439).

A second potential reason for internal displacement is the presence of mineral mines. Collier (2004) described that natural resources contributed to the conflict in Eastern-Congo. A predominant theory is the resource curse, which might be defined as the adverse effects of a country’s natural resources on economic, social and political institutions (Ross, 2014, p. 240). The theory has influenced many debates within the political sciences. One of the debates is whether natural resources could promote civil war (Ross, 2014, p. 240).

Throughout the 1990s, many armed groups have relied on revenues from natural resources. Resources not only financed, but in some cases also motivated conflicts (Le Billion, 2001, p. 65). Zeender and Rothing (2010) noted that members of the army, as well as military groups, have been involved in the illegal exploitation of minerals, contributing to violence and displacement. Within the context of DR Congo, a report examining this

phenomenon in the provinces of Kasai Oriental and North-Kivu describes that Hutu militants forced civilians to act as slave porters for mining activities, which led to many people fleeing their homes (Zeender & Rothing, 2010, p. 10). Other cases reveal similar dynamics, for example in India more than 22.5 million people were moved because of mining-induced displacement between 1950 and 1990 (Terminski, 2012). Mines have been developed in various regions across DR Congo, and when such developments lead to conflict, citizens may not be able to return home.

As Vlassenroot and Huggins (2005, p. 115) described, the interplay between local, national and regional dynamics in the conflict has drawn much attention from the internal community. As a part of these dynamics, they mention natural resources as a key issue in shaping warfare in DR Congo. Although different scholars have found that natural resources

(10)

can lead to displacement, I aimed to discover if resources might also be associated with the number of returnees.

According to Ross (2003) different types of resources may influence conflicts uniquely. In his research, data from 1990 until 2000 revealed that diamonds and drugs were more strongly associated with civil wars. A possible explanation for this is that some resources are more easy to exploit and thereby more popular with armed groups than other resources, leading to more violent competition for these resources (Ross, 2003).

The presence of mines and violence could directly explain the number of internally displaced returnees, or could affect other factors that may be related to the number of

returnees. An example of a direct relation would be that IDP recognize areas containing high levels of violence as being unsafe and do not return because of this. An indirect relation could be that the presence of violence degrades infrastructure, such as social services, making an area less attractive to potential returnees. For mines, an example could be that the

development of mines leads to water pollution, which could be an important consideration for returning.

A third reason for internal displacement is strategic displacement. Steele (2011) defined strategic displacement as “the expulsion of civilians from a territory by an armed group”. In some cases, the armed groups do not explicitly force people to leave, but the situation becomes so unstable that people flee. Strategic displacement is more likely when there is competition amongst armed groups for territorial control or when an armed group attempts to gain control over a territory (Steele, 2011, p. 423-424). Armed groups can use different strategies to achieve their goals.

According to Sinno (2011), strategic choices of armed groups are affected by whether the group developed by accident or by design (Sinno, 2011, p. 311). This implies that if the group was developed by accident, they often do not make strategic choices and thereby rarely intentionally displace people. Vlassenroot and Huggins (2005) argue that most scholars focus on ‘greed’ as a dominant military strategy, they also see structures and patterns of economic control and exploitation. For example, competition for land was one of the causes of the conflict pertaining to Ituri and the Kivus, as the unequal access to land has lead into competition between different groups (Vlassenroot & Huggins, 2005, p. 115-116). In this study, the origin of the armed groups will be taken into account, as the traits of these groups may affect whether people are able or willing to go home in a particular region.

After being displaced, some people make the decision to return home or they are forced to go back home, while others start a new life elsewhere in the country. There are

(11)

substantial differences in reasons for displacing. In an unstable country, such as DR Congo, some people decide to stay in a situation while others leave. According to Van Dijk, Foeken and Van Til (2001) reasons for migration differ, as people are different in age, gender and education. Literature about IDP may be extended to returnees because, in many cases, the reason for leaving may also be the reason they decide not to return.

3. Argument

The argument of this thesis draws upon previous work on the relation between natural resources and conflicts (e.g. Le Billion, 2001; Collier, 2004; Vlassenroot, 2008; Zeender & Rothing 2010; Batware, 2011; Steele, 2011; Sinno, 2011; Terminski, 2012; Schon, 2014; Ross, 2014). Based on this theoretical framework, the argument underlying this thesis was that the combined presence of mineral mines and multiple armed groups, in a country with ongoing conflict, fosters competition between the groups and thereby an increase in violent incidents. This would be associated with greater displacement and less returnees.

Support for the proposed mechanism comes from Collier’s (2004) description of how natural resources have contributed to the conflict in Eastern-Congo. The resource-driven conflict is sustained by the presence of rich mines, leading to continued tension and

instability. As such, people continue to perceive the region as being unsafe. If this argument was accurate, I expected to see that fewer people returned to their places of origin in an area where there are many mines, multiple armed groups and a large amount of violent incidents after people are internally displaced. If there was just a single armed group in an area with many mines, it would be a monopoly and there would be no competition and thereby less violent incidents. This would result in less displacement and more returnees compared to an area with more armed groups. These expectations are illustrated in Table 1.

Table 1

Expectations for the effect of mines and armed group competition on IDP and returnees Regional condition Displacement

Large amount of mines Number of armed groups (monopoly or competition?) Number of internally

displaced people Number of returnees

No 0 Few Many

No 1 (monopoly) Few Many

(12)

Yes >1 (competition) Many Very few

In this thesis, I focused on areas with mines and armed groups because this makes the regions more comparable, therefore the quantitative analyses did not consider the condition in which there were no armed groups. I also looked into the impact of different natural

resources. Based on these findings, my expectation was that some natural resources would have a larger influence on the number returnees than others. This could lead to varying amounts of returnees for different mines and thereby regions. The natural resources obtained from mines in DR Congo include gold, cassiterite, coltan, wolframite, diamond, tourmaline, amethyst and manganese.

For my argument, I specified one model without the number of violent incidents and compared it to a saturated model with all the variables mentioned in my mechanism: The amount of mines, number of armed groups, number of violent incidents, number of IDP and number of returnees. By doing this, I was able to consider whether the presence of natural resources also predicted the number of returnees while controlling for levels of violence. In this way the influences of natural resources could be investigated independently which relates to my main research goal. As explained in Table 1, the focus of this thesis was on areas with armed groups, implying that not all the expectation as illustrated in the model will be tested. The option where there is no armed group around will not be tested. These expectations are illustrated in Table 2.

Table 2

Expectations of the argument including violent incidents More than one

armed group (monopoly or not?)

Competition what leads into more violent incidents

Many internally displaced people

People that return to their places of origin

No (no armed groups)

No No Yes

No (monopoly) No No Many

Yes Yes Yes Less

Yes No (monopoly) No No

(13)

The argument underlying this thesis produced several observable expectations. First, if my argument was correct, fewer people would return to their places of origin in an area where there are many mines. Second, by including violent incidents as a predictor and the result is significant, I could say that natural resources had an impact on returnees beyond its relation with the number of violent incidents. Third, in considering the impact of different natural resources, I expected that some natural resources would have a greater impact on the number of returnees. The next chapter will provide the methods of this thesis.

4. Research design

This study used a mixed-method data collection strategy, combining quantitative data with qualitative expert interviews. I used explanatory sequential mixed methods because I first conducted quantitative research. After analysing the data, I built on the results using qualitative research to explain them in more detail. Because the initial quantitative data results are explained further with the qualitative data it is considered explanatory (Creswell, 2014). This combination complements and facilitates triangulation of the findings. The core

assumption of this form is that this combination provides a more complete understanding of a research problem than either approach alone (Creswell, 2014). The units of analysis in this study were the different regions. Five regions were selected based on availability and

importance. These regions are located in the Eastern part of DR Congo and are characterized by varying degrees of conflict. For these regions the locations of the mines, violent incidents and armed groups were used.

Regression analyses were useful for answering the research question as they allowed for examining how well the variables of interest (i.e., violent incidents, armed groups, and natural resources) predicted the number of returnees while controlling for other covariates. Given the continuous nature of the dependent variable, I used a multiple linear regression analysis using Ordinary Least Squares estimation to examine the relations. It was used to test the potential predictors and assess the relative contribution of each individual variable

(Pallant, 2013, p. 126). 4.2 Operationalization

I started by conducting cross-sectional analyses with information on all African

countries. These analyses were aimed at identifying overall trends and patterns underlying the overall amount of returnees after displacement across Africa. As there was no time-series data available, I only focused on data from 2013. In these analyses, 22 countries were included.

(14)

These are not all the countries in Africa, nevertheless the detected patterns were still reliable because the main countries of the continent, with reliable data, were included. Once the overall trends were identified, I conducted in-depth analyses on DR Congo, this involved time-series analyses with lagged independent variables. A dummy for time was included to control for temporal variation. Concerns of reverse causality were mitigated by lagging the independent variables. To investigate other potential factors relating to the amount of returnees, other regression analyses were run in an exploratory manner. These in-depth analyses considered various variables and were executed using time-series regression analyses. Because Multiple Regression Analyses are sensitive to outliers, and some of the variables’ distributions were skewed, I used a logarithm of these variables.

In the analyses, I assessed several variables on their ability to predict the number of returnees. This was done in two steps. The first step involved investigating an African cross-national dataset. For the cross-sectional analyses of the Africa data, I only had to use violent incidents and mines as predictors, nevertheless I decided to explore the relation between these extra variables and the number of returnees because they could provide additional context, as they have been shown to demonstrate the development of a country and could all have been related to the number of returnees. For the African cross-national analyses, I used a dataset from the University of Gothenburg1 (QoG) and The UN Refugee Agency2 (UNHCR). The next section describes how these variables were measured.

Due to the limited data available, I could not match data on the smallest unit of analysis (e.g. villages). Therefore, I aggregated the datasets at the largest unit of analysis, which was the province level. In this dataset, the number of violent incidents, the amount of mines and the amount of returnees per village were summarized within regions. Although, aggregating the data had disadvantages as well as advantages, as some important spatial information was lost, this solution allowed for the use of all datasets and was thus deemed most suitable for testing my argument. Furthermore, aggregating the data was also deemed advantageous for surfacing trends that would otherwise have been undetectable. This is because combining data from different sources provided a clean yet comprehensive overview. In sum, the methods used in this thesis were suitable for illustrating the broader association between returnees and potential predictors. This method was aimed at providing insight into coarse patterns in the data, which could be disaggregated into more detailed units of analysis in future research.

1_{http://qog.pol.gu.se/data/datadownloads}

(15)

The amount of internal armed conflicts was measured on an annual basis. Internal armed conflict was said to occur between the government of a state and one or more internal opposition group(s) without intervention from other states (Quality of Government, 2013). As a way to assess democracy, I used a combined score of polity and Freedom House. The polity scores range from 0-10 where zero indicated low levels of democracy and ten indicated high levels. Freedom House scores were transformed to an identical scale and both variables were averaged to assess polity (Quality of Government, 2013). Gross Domestic Product (GDP) refers to the sum of gross value added by all resident producers in the economy plus any product taxes and minus any subsidies not included in the value of the products (World Bank). GDP per capita is the GDP of a country divided by midyear population. It was calculated without making deductions for depreciation of fabricated assets or for depletion and degradation of natural resources (World Bank). In this data, GDP and GDP per capita were expressed in constant 2005 U.S. dollars (Quality of Government, 2013). The last variable included was total natural resource rent. Natural resource rent is the sum of oil rent, natural gas rent, coal rent (hard and soft), mineral rent, and forest rent, as calculated by the Quality of Government for each country.

Following this, the sub-national data of DR Congo was investigated. For the in-depth analysis on DR Congo, I used the variables specified in the proposed argument, namely the number of violent incidents, natural resources, and number of armed groups. I used data provided by the United Nations Office for Coordination of Humanitarian Affairs3 (OHCD), Armed Conflict Location & Event Data Project4 (ACLED) and International Peace

Information Service5 (IPIS). These variables were measured in the following way. In the DR Congo dataset, violence against civilians was defined as the number of deliberate violent acts perpetrated by an organized political group, such as rebels, militias, or government forces, against unarmed non-combatants (ACLED). IPIS collected coordinates of the most important mining areas and information on the security. Their data includes nearly 1100 mining sites (IPIS). A variable for the year was used to compare different periods to see if the amount of IDP in 2013 could be used to predict the amount of returnees in 2014. The numbers of armed groups included all groups active in the mining areas of the region (IPIS). 4.3 Methods

3_{http://reliefweb.int/map/democratic-‐republic-‐congo/democratic-‐republic-‐congo-‐internally-‐displaced-‐} persons-‐and-‐returnees

4_{http://www.crisis.acleddata.com/category/democratic-‐republic-‐of-‐congo/} 5_{Data
is
not
publically
available,
available
upon
request.}

(16)

The five databases that I used all publish independent reports on several themes. Independence is important because this ensured their reliability. The organizations work in several countries on humanitarian challenges, collect data annually and are well known. These five databases were all developed and maintained by prominent research institutes and have a long tradition in conducting research in Africa and DR Congo. The data is publicly available and frequently updated.

For the qualitative component of the analysis, I interviewed four experts. The first interview was with Fleur Scheele, a researcher at a Dutch non-profit research institute, SOMO (Stichting Onderzoek Multinationale Ondernemingen). In April 2016, she published a report on the effects of cobalt mines on human rights in Katanga, also covering topic of internal displacement. The second interview was with Han van Dijk, a researcher at the African Studies Centre in Leiden. He is a specialist on political decentralization, land tenure, natural resource management and development policy. The third interview was with Dirk-Jan Koch, he is the Special Envoy of Natural Resources at The Dutch Ministry of Foreign Affairs. Before, he worked as a diplomat at the Dutch Embassy in DR Congo. The fourth expert interview was with Clens Santa, who fled from DR Congo in 1993. These four experts were chosen to understand how professionals with different experiences gave input on the subject.

Two methodological concerns remained. First, the data concerning Africa was limited. This implied that some areas did not have specific IDP and returnee numbers. Second, the sample size for the subnational analysis was quite small (5 regions over 3 time periods). Although, only five different regions were used, this analysis was still expected to be powerful enough to provide significant results. The reason that the results were reliable was that the data on all variables were comprehensive. This means that all the mines, all the violent incidents and all the armed groups of these regions are taken into account. Data from 2013, 2014, and 2015 was used. These periods were used because of the enormous decrease in returnees in 2015. Another reason was that violence erupted between The March 23

Movement (M23) rebels and the government in 2012. Many people were forced to leave their homes. This conflict ended in 2013, allowing many people to return home afterwards

(Cordell, 2014). The next chapter will provide contextual information relating to this thesis, including the history and background of DR Congo

(17)

5. Case study: The Democratic Republic of the Congo

In order to explain the regional variation in the amount of IDP returnees worldwide and in the case of DR Congo, thus answering the central research question of this thesis, it is important to provide an understanding of the background of DR Congo, internal displacement and returning after being internally displaced. The aim of this chapter is to provide a better understanding of the importance of this case study, and to develop the required understanding of the country and the variation of internal displaced returnees across regions within the country.

The first section gives information about internal displacement worldwide and in DR Congo. The second section covers the country’s history. It is important to note that there are multiple versions of this history, depending on the sources that are consulted. Due to the numerous incidents and many involved factors, there are large variations in how this history has been perceived and described. I choose to describe the history based on Vlassenroot and Huggins (2005), Van Reybrouck (2010), Koch (2014) and Cordell (2014) because they provided coherent explanations, and especially Van Reybrouck is experienced on the topic of DR Congo. In the third section I will explain the differences between the five regions that I took into account in this research, as the regions differ substantially. These differences were deemed non-negligible for answering the research question. The fourth section describes the consequences of internal displacement and returning, highlighting the necessity of

understanding factors that serve as barriers to returnees. I conclude this chapter by providing a brief overview of the different actors that play a role, relevant to this thesis, in DR Congo. 5.2 Internal displacement during armed conflict

It is unknown how many IDP there are, since many are unaccounted for. According to the ICRC, some governments deny the existence of any IDP. However, by one measure, approximately 26 million IDP exist across the world (ICRC, 2009, 2). Many of them are displaced due to armed conflict and may be able to go home after some time (ICRC, 2009, p. 2). However, many people in DR Congo never return and the amount of people returning has been decreasing significantly, possibly because of persistent insecurity (OCHA, 2015).

In DR Congo, there are approximately 1.6 million IDP (in 2015) (OCHA, 2015). Millions of citizens of DR Congo were repeatedly forced to leave their places of origin to flee from violence and commotion, with devastating consequences. This has been going on for years, and is still part of reality in DR Congo. According to the figures of the ICRC, from the total of 1.6 million IDP, 90.6% fled because of armed attacks and 75.1% ended up in host

(18)

families. In the second semester of 2014, 162.100 people returned home and 47.949 people in the second semester of 2015, which is a significant difference and means that people are less likely to go back to their places of origin (ICRC, 2009, p. 2). Although ICRC mentions that people leave from violence, this could also be an indirect case. My argument suggests that when there are mines, this leads to more armed groups and more violence. In that case, the actual cause are these natural resources.

Land access plays a big role in the conflict in DR Congo. Especially because of the valuable natural resources the country has. Customary land rights, which are about 97 percent of the country, are not perfectly defined or secured in land law. This means that there are many inequalities around land and according to Vlassenroot and Huggins (2005) this

contributes to the conflicts in DR Congo; this is one of the reasons why the conflict cannot be solved only through legislative reforms (Vlassenroot & Huggins, 2005, p. 116). This leads into interest for the soil of many companies and countries and results in the fact that citizens of DR Congo do not get any of the profits of their rich soil.

Most IDP are forced to live with host families that provide temporary shelter on the way to more permanent homes (ICRC, 2009, p. 12). When people flee their homes in DR Congo and do not end up in a host family, they can stay in camps. The United Nations (UN) states that this is a last resort, people only go to such camps if no other options exist.

However, some camps have well-developed infrastructure in which children have access to education and clean drinking water is provided, whereas this was not always the case in their own villages. An example of such a camp is Gereida in Darfur. This camp was erected as part of an emergency operation but is now used as a regular service. People travel to the camp to benefit from the services they offer (ICRC, 2009, p. 12). According to Kellenberger, president of the ICRC, these camps do not provide any incentive for people to return home and may partly contribute to the lack of returnees, however I am not aware of any such camps in DR Congo.

5.3 DR Congo and its history

From 1870 until 1960, DR Congo was a colony of Belgium. In this period, two factors became important to the country’s development. First, DR Congo has a rich endowment of natural resources. Western powers saw the country as a bottomless source of copper, gold, diamonds, cobalt, rubber and timber, resulting in structural conflict (Collier, 2004). That the presence of mines fosters strife became evident several years ago. As explained in the

(19)

and have also contributed to the on-going conflict in Eastern-Congo (Collier, 2004). A second factor that became more salient under the Belgian governance was that the colonizers made distinctions between different ethnicities, which also led to civil wars. This implies that violence is not a historical anomaly; violence and the manipulation of identity groups has been a characteristic of Congolese life since the idea of ‘the Congo’ first egressed

(Vlassenroot & Huggins 2005, p. 117). According to several ethnographers, the country has about four hundred different ethnical groups and the citizens are all well known within their own group of origin (Van Reybrouck, 2010, p. 22-26).

The violence of the European power was meted out through proxies, which resulted in tensions between different identity groups. There have been structural clashes in history between different groups, mostly in Nord-Kivu. One example is the clashes between the local Luala and the Luba people. The Luba people migrated to the area to work the diamond mines, which led to the first civil war. At the time of writing this thesis, many similar conflicts are still ongoing. Although these conflicts are mostly local, the impact of colonialism on the local balance of power cannot be neglected. The effects of the colonial era in DR Congo are still felt in the conflicts between ‘local communities’ and ‘immigrant communities’ (Vlassenroot & Huggins, 2005, p. 117).

In 1960, DR Congo gained independence from Belgium. From 1971 until 1997, the country was renamed Zaïre by the ruler, Mobutu. The first years of independence were a continuation of existing norms and structure, which was marked by unrest. During this time, independence leader Patrice Lumumba was assassinated and the United Nations intervened in the country. Natural resources were an important currency for political leaders and the elite (Vlassenroot & Huggins, 2005, p. 117-119). Exploiting these resources remains common among political leaders and the elite. Access and ownership of land and human right violations eventually paved the way to violent conflict.

In 1994, DR Congo plunged into civil war. This was partly caused by tensions from coping with large amounts of refugees from Rwanda and Burundi. In 1997, Kabila became president of the country, ending the dictatorship of Mobutu. Yet the civil war continued and intensified in 1998, when Rwandan and Ugandan rebels tried to expel Kabila. When he died in 2001, Kabila’s son became president of the country. This is the same period in which the United Nations entered the country as part of the peace operation Mission de l’Organisation des Nations Unies pour la stabilisation en RD Congo (MUNUSCO). The civil war ended officially in 2003 and in 2006 DR Congo had its first democratic election. Kabila won although the elections caused unrest. In 2012, violence erupted between the The March 23

(20)

Movement (M23) rebels and the government of North-Kivu. Thousands of people were forced to leave their homes. This conflict ended in 2013, allowing many people to return home afterwards. Although the country is now officially at peace, the violence continues in the Eastern part of the country (Cordell, 2014).

According to the Refugee Studies Centre (2010a), there are some particular events that resulted in IDP. The first event that led to many IDP was the political crisis of the military operation Force Publique Mutiny in July 1960. The second crisis was the Katanga secession from 1960 until 1963 and the third crisis was the Kasai secession in 1960. After this, several wars were fought between the rebel groups, followed by two Shaba wars in 1977 and 1978. These events resulted in even more IDP. The first liberation war was in 1996 and the most recent wars in DR Congo were from 1998 until 2003. This war, between several rebel groups, caused the displacement of approximately 3.4 million people. Although, soldiers were

fighting to protect the citizens of DR Congo, they were not able to protect the people against the rebels. The conflict resulted in (sexual) violence, killing and cruelty towards civilians (Refugee Studies Centre, 2010a, p. 9). Several of these events occurred in the presence of mineral mines.

In 2010, the United States of America introduced the Dodd-Frank act (Dodd-Frank Wall Street Reform and consumer Protection Act) focusing on regulating conflict minerals around the Great Lakes (Rwanda, Burundi, Uganda, DR Congo, Tanzania and Kenya). Koch (2015, p. 1-4) argued that this law is incomplete because the focus is limited to the Great Lakes, however it serves as a legislative foundation and companies are required to report the origin of their minerals. This has supposedly fostered many conflict free chains. For example, the Conflict Free Tin Initiative. Although the Netherlands was involved in implementing this initiative, the European Union still does not have any legislation concerning conflict minerals. Finally, there have been no investigations evaluating the effectiveness of the Dodd-Frank act. Koch (2015) positively claimed that around 60% of the tin, tantalum and wolframite used to be in the hands of rebels, which is now probably around 30%. According to Koch (2015, p. 1-4), this can be attributed to the Dodd-Frank act.

5.4 Different regions

This paragraph shortly explains the situation in every region that is part of this thesis: Nord-Kivu, Sud-Kivu, Tanganyika & Haut-Katanga, Maniema and Ituri/Orientale. Figure 1 illustrates the locations of these regions.

(21)

Figure 1

The regions of DR Congo that are included in this thesis

In Nord- and Sud-Kivu, two of the main provinces of the country, people have had to frequently endure being caught in the crossfire during armed confrontations (Refugee Studies Centre, 2010a, p. 4). These confrontations were between several rebel groups and the

Congolese army. After a peace agreement was signed in 2009, many IDP returned to their place of origin. However, there are still frequent confrontations between the different groups. These confrontations have led to violations of human rights and the destruction of villages. During these attacks, especially women have been the victims of (sexual) violence (Refugee Studies Centre, 2010a, p. 4). In Nord-Kivu, several armed groups have been active around mines, namely Nyatura, the Armed Forces of the Democratic Republic of the Congo (FARDC), The Democratic Forces for the Liberation of Rwanda (FDLR), Mai-Mai and UPCP. In this region, armed groups are active around 79% of the mines. The products exploited in Nord-Kivu are gold, cassiterite, coltan, wolframite, tourmaline, amethyst and diamonds (IPIS, 2014, p. 14-15).

According to IPIS (2014), armed conflicts in Sud-Kivu have been of relatively low-intensity in the last few years. However, there are still many armed groups active in the area and also in some mining areas. The groups that are active in Sud-Kivu are Raïa Mutomboki and FARDC. In this area, mainly coltan and gold is mined (IPIS, 2014, p. 16-17). The

products that are exploited in Sud-Kivu are gold, cassiterite, coltan, wolframite, diamonds and manganese (IPIS, 2014, p. 14-15). Refugee Studies Centre (2010a) stated that, as long as armed groups prey on the population and the Congolese government does not establish basic security, displacement would continue in Nord- and Sud Kivu (IPIS, 2014, p. 6).

(22)

Tanganyika & Haut-Katanga are registered as one province in the maps of Office for the Coordination of Humanitarian Affairs (OCHA) and are both located in the south-east side of DR Congo. Tanganyika has a lake, the Tanganyika Lake, which supports a large fishing industry. The water of this lake is also used as drinking water, which has led to the spread of cholera (UNOCHA, 2010). Katanga is renowned for the mining of tin and tantalum, although there are also mines with gold, casserite, coltan, wolframite and tourmaline. According to IPIS (2014), armed groups no longer affect the exploitation of these resources, as the FARDC is in control (IPIS, 2014, p. 21).

Buchanan (2015) has stated that the mines around Maniema are mostly free from militias. However, according to IPIS, there is still illegal mining going on by the Mai-Mai rebels. In Maniema, gold is exploited, but also casserite and diamond is found in this area (IPIS, 2014, p. 19).

Mines in the Orientale province mostly contain gold, with a yearly production of two tonnes of gold. According to IPIS, the mining operations are free of military involvement. They mentioned that the armed group Front for Patriotic Resistance in Ituri (FPRI) was historically involved in gold mining, but have been driven away by the FARDC since 2013. The Mai-Mai had also been active in this area before 2013 (IPIS, 2014, p. 19). Historically, this lack of conflict in uncharacteristic of the region, Vlassenroot and Huggins (2005) explained that the district has been the location of DR Congo’s bloodiest civil wars, which have led to the displacement of half a million people. They illustrate that some areas – land between different armed factions – have become desolate while other cities observe the return of people from different ethnical backgrounds due to the process of ethnic cleansing. In general, some groups have lost substantial land while others have gained exclusive access to territory, trade routes and other advantages (Vlassenroot & Huggins, 2005, p. 160).

5.5 Consequences of internal displacement and returning

The humanitarian challenges caused by internal displacement are immense (ICRC, 2009, p. 2). While fleeing, people are vulnerable and especially women and children are often victim of (sexual) violence. These problems persist once people arrive at camps. IDP thus struggle with increased danger to their physical and psychological wellbeing. Furthermore, children are susceptible to being recruited by armed groups in camps. Armed groups use these children to work as soldiers or as (sex) slaves. These children are separated from their

families and lose access to education and health services (ICRC, 2009, p. 2). In 2015, International Alert, Climate Interactive, Norwegian Refugee Council, and International

(23)

Displacement Monitoring centre found that repeated displacement has an impact on family composition, relationships and roles. They interviewed displaced people in the Eastern part of DR Congo to find out what consequences they had experienced. The biggest issue was that they had to leave family behind (Kesmaecker-Wissing & Pagot, 2015, p. 8-9).

5.6 Different actors

Several actors have been involved in the conflicts in DR Congo, and over time the state has been challenged by the introduction of new types of actors (Vlassenroot, 2008, p. 2). In DR Congo, many human right abuses have taken place by both armed groups and some FARDC militants (Refugee Studies Centre, 2010a). The main actors that can be distinguished are the government, the international community and armed groups.

The government has set up two national groups with the responsibility of keeping the country safe. First, FARDC, which is the state institution responsible for defending the country (Stearns & Vogel, 2015, p. 5). This group consists of Congolese Tutsis and Hutus. Many of them have deep resentment against their former allies in Kigali (Stearns & Vogel, 2015, p. 5). In protecting national security, the group receives assistance from the second group, the stabilisation mission of the United Nations, the MONUSCO (Global Centre for the Responsibility to Protect, 2016). Vlassenroot (2008) claimed that being reliant on the

FARDC, implies that the country has a weak centre of power, because it also has to rely on international mediation in order to maintain a semblance of authority (Vlassenroot, 2008, p. 15).

The second national group responsible for safety is the Police Nationale Congolaise (PNC), the national police force. This is a government agency, responsible for several tasks. For instance, during elections they are responsible for supporting communication and protecting civil order. The PNC does not have the capacity to safely store weapons, so these are stored with the FARDC. This has led to a lack of regulation, and can lead to the

proliferation of weapons trade with illegal armed groups (Buchanan, 2015).

The Congo Research Group (in Stearns & Vogel, 2015) stated that there are seventy armed groups active in the region, whereas the Global Centre for the Responsibility to Protect estimated that there are thirty armed groups. According to Vlassenroot (2008), several armed groups have only aim to control local natural resources, whereas others have attempted to forge local and transborder alliances in order to increase their economic and political power (Vlassenroot, 2008, p. 2). For armed groups in DR Congo, rather than growth, the most common trend has been fragmentation. According to Stearns and Vogel (2015, p. 7), this has

(24)

led to a rapid increase in smaller armed groups. Most of the armed groups in DR Congo have fewer than 200 soldiers and recruit largely on the basis of ethnicity (Stearns & Vogel, 2015, p. 5). As this study only involved five regions, only the groups that are active in these specific regions will be explained in more detail. This includes the FPRI, Raïa Mutomboki, Nduma Defense of Congo (NDC), Militaires indisciplinés and different Mai-Mai groups (Nyatura, Simba, Mangaribi Sadala, Kifuafua) (IPIS, 2015).

The first armed group is the FPRI, this armed group has mostly operated in the Ituri district. The FPRI is the result of an inter-ethnic clash between two groups, the Lendu and Hema. The group has existed since 1999 and supports the Congolese army. The former commander of the group (Germain Katanga) has been found guilty of war crimes and has spent twelve years in prison. The FPRI has been found guilty of murder, rape and employing child soldiers (Buchanan, 2015).

The second armed group is Raïa Mutomboki. This group has existed since 2011, and has been predominantly active in mining areas across several provinces (IPIS, 2015). They started operating out of self-defence as a response to rampant insecurity against attacks from the FDLR (Buchanan, 2015). As the group had success against the FDLR, they gained popularity and eventually support from local populace. Eventually, families even decided to send their children to help them. Raïa Mutomboki has claimed that they should be recognised as a legitimate security force because the Congolese army is incapable of offering protection to the people (Buchanan, 2015). A publication of the United Nations claimed that the group is responsible for 246 civilian deaths, of which most were women, children and elderly people (Buchanan, 2015).

There is very little documentation on the third and fourth groups in the regions. The NDC is said to have control over many mining areas, but this is not known for the Militaires Indisciplinés. The second last group is the Mai-Mai. The Mai-Mai consists of several groups, but is has relatively few members and, according to Stearns and Vogel (2015), not very active. This group of Hutu militia has existed since 2010 and has collaborated with FDLR rebels and the FARDC in protecting Hutu interests against other armed groups. The Nyatura has initiated ethnic conflict, mainly targeting people of Tembo ethnicity. The International Business Times (Buchanan, 2015) published that this group has integrated into the FARDC since 2012. "Nyatura is engaged in confrontations both as a standalone group but also with some subgroups having been employed as proxy forces by the Congolese army", said Christoph Vogel, political analyst (Buchanan, 2015). The Nyatura have opposed the Raia Mutomboki for targeting Hutu communities. Although the Nyatura have been working with

(25)

the FARDC, they have been accused of rape, violence, recruiting child soldiers and illegal executions. Over a five-month period in 2012, they are suspected of being involved in the execution of 264 civilians, including 83 children (Buchanan, 2015).

Figure 2

The regions of DR Congo that are included in this thesis including IDP, returnees, mines and armed groups

5.7 Summary

In conclusion, internal displacement is a global phenomenon, but is especially prominent in DR Congo, possibly due to poorly defined land rights. Two factors became increasingly important in DR Congo since the colonial period, the presence of natural resources and distinctions between different ethnicities. Both factors have led to increased violence and a clear distinction between communities. The role of these two factors provided some initial support for my argument that the presence of mines and multiple armed groups has led to IDP and less returnees, plausible. Furthermore, that this mechanism works

differently in different regions was supported by the differences in armed groups, intensity of incidents, natural resources, and the presence of absence of the FARDC, PNC and

(26)

6. Results

This section presents the results of the quantitative- and qualitative analyses to seek an explanation for the regional variation in internally displaced returnees. First, regression analyses were run, after which I extended upon the quantitative research with expert interviews.

6.2 Cross-sectional analysis with African data

Model 1 included polity, GDP per capita, total population, IDP, total natural resource rent as predictors. In this model, I looked into the relation between the number of returnees and polity, GDP per capita (growth), total population, IDP and total natural resource rent in Africa. As described in my mechanism, my expectation is that natural resource rent will predict less returnees.

Model 2 included the same variables as Model 1, but with the addition of internal armed conflicts. By doing so, I could test whether the number of violent incidents would mediate the relation between natural resources and the number of returnees. As explained in the Research section, some of the variables were skewed and were therefore log-transformed. These variables were IDP, returnees, GDP per capita (growth) and total population. Using the Africa dataset, I conducted several regression analyses. The first regression models (Models 1 & 2) focused on natural resources. The results of these analyses are presented in Table 3.

Table 3

Models 1 and 2 testing the relation between returnees in Africa and GDP per capita (growth), total population, IDP and total natural resources rents and armed conflicts (N = 22)

Model 1 (exclusion of armed conflicts variable)

Model 2 (inclusion armed conflicts)

B (se) Beta B (se) Beta

Polity -.031*** (.048) -.040 3.252 (.062) -.037 IDP .803*** (.049) .976 8.112*** (.061) .943 Total natural resource rent -.023*** (.006) -.215 2.224** (.008) -.196 GDP per capita (growth) .649* (.269) .125 84.043 (.307) .117 Total population .163 (.144) .066 20.925 (.226) .056 Internal armed conflicts 43.476 (.469) .086

Notes: R2_{= .942, Adjusted R}2_{= .930, N = 22, * p < .05, ** p < .01, *** p < .001. Parameters are presented in}

their original units.

In Model 1, I found that polity, GDP per capita (growth), IDP and natural resources predicted lower numbers of returnees. This implies that the more natural resources a country has access to, the less returnees in a country. Although Model 1 suggests a significant relation

(27)

between natural resources and returnees, this association could potentially be attributed to the mediating effect of armed conflicts. To examine the possibility, Model 2 was specified with internal armed conflict included in the regression analysis.

Model 2 revealed that when conflict was added to the previous model, the negative relation between the number of returnees and natural resources remained significant. This confirmed that natural resources led to fewer returnees. Aside from natural resources, IDP was also statistically-significantly related to the number of returnees. The variables internal armed groups, polity, GDP per capita (growth) and the total population were not significant at a .10 significance level. The next regression analyses were centered on the sub-national level in DR Congo. By looking into sub-national data, I could use the locations of mines and violent incidents per region, and examine if the relation would still hold.

6.3 Analyses with the DR Congo data

To look into detailed information of DR Congo, I compared five regions. This includes all regions that are part of Eastern-Congo: Nord-Kivu, Sud-Kivu, Tanganyika & Haut-Katanga, Maniema and Ituri/Orientale. These regions have to deal with conflict

regularly. Data on the number of returnees in 2013, 2014, and 2015 was used. See Table 4 for an overview of returnees across regions and years.

Table 4

The number of returnees across five regions in eastern DR Congo in 2013, 2014 and 2015.

2013 2014 2015

Nord-Kivu 635,738 664,251 29,000

Sud-Kivu 498,198 513,284 17,000

Tanganyika & Haut-Katanga 232,841 217,380 46,000

Maniema 140,725 140,725 6,000

Ituri/Orientale 156,356 274,108 1,000

*These moments of measurements in the year itself sometimes vary6

Before presenting the results of the regression analyses, I provide descriptive statistics on the number of IDP and returnees to show the trends over time. Figure 2 illustrates how the amount of returnees decreased in all regions, however Figure 3 shows that the amount of IDP decreased as well. The amounts of IDP in 2013, 2014 and 2015 in the different regions can be

6_{http://reliefweb.int/report/democratic-‐republic-‐congo/democratic-‐republic-‐congo-‐internally-‐} displaced-‐people-‐and-‐returne-‐2

(28)

seen in Table 5.

Table 5

The number of IDP across five regions in eastern DR Congo in 2013, 2014 and 2015.

2013 2014 2015

Nord-Kivu 1,123,446 1,076,745 166,000

Sud-Kivu 579,607 518,201 27,000

Tanganyika & Haut-Katanga 402,220 500,284 19,000

Maniema 293,510 172,840 14,000

Ituri/Orientale 549,921 366,802 51,000

*These moments of measurements in the year itself sometimes vary6

Figure 3

(29)

Figure 4

The decrease in the number of IDP between 2013 and 2015 in DR Congo.

To develop a better understanding of the different causes of the decrease of returnees in DR Congo, I did a time-series regression analysis with lagged independent variables (number of IDP, number of returnees and number of violent incidents). Temporal variation was accounted for using a dummy variable. The results can be found in Model 3.

Using Model 3, I looked into the relation between the number of returnees in DR Congo and the number of violent incidents, mines, year, IDP and armed groups. These variables were included in this analysis as they were all components of my proposed

mechanism and could have been related to the number of returnees. My expectations were to find results similar to the Africa data, with mines predicting the number of returnees.

(30)

Model 3

The number of returnees in DR Congo regressed on violence against civilians, total amount of mines, year, IDP and number of armed groups.

B (se) Beta

Violence against civilians -1.207** (.002) -.539 Mines total (2013/2015) -.200* (.001) -.335

Year -1.04*** (.120) -.605

IDP 3.344** (.710) .896

Number of armed groups 2.429 (.123) -.036

Notes: R2 = .986, Adjusted R2 = .969, N = 15, * p < .05, ** p < .01, *** p < .001. Parameters are presented in their original units.

Model 3 revealed that the number of violent incidents, year and the amount of mines all have a negative significant impact on the number of returnees. This implies that the more mines a region has, the less returnees. In conclusion, this is the same result as the result of the cross-section analyses of Africa as a continent, suggesting an overall trend, generalizable for other African countries with many mines and returnees. The variables year and violence against civilians had a smaller unique contribution.

The number of armed groups did not predict the number of returnees in DR Congo. IDP made the strongest unique contribution to explaining the dependent variable.

Furthermore, fewer IDP was associated with less returnees. Although I only presented one model, several regression analyses were run using the original and the lagged variables. As the results led to identical conclusions, I only presented this model.

In the last regression analysis, the effects of the presence of specific resources was investigated. By testing these variables individually, we could see which variables had the most influence on returnees. Model 4 shows the results for the different natural resources included. In this analyses (see Appendix A-I), the variables IDP, violence against incidents, year and number of groups were also included.

Model 4

The relation between the presence of various types of resources and the number of returnees in DR Congo B (se) Beta Amethyst -.236 (.452) -.109 Casserite .003 (.005) .197 Coltan .008 (.007) .184 Copper -.092 (.100) -.213 Diamond -.012 (-.007) -.319 Gold -.002* (.001) -.246 Monazite .745 (.517) -.346

(31)

Tourmaline .200 (.487) .208 Wolframite -.112** (.021) -.566

Notes: R2 _{= .948. * p < .05, ** p < .01, *** p < .001. Data are in their original units.}

Model 4 shows that only the presence of gold and wolframite were significant predictors of the number of returnees. There could be several explanations for this relation, which will be investigated in the second part of this thesis.

6.4 Summary quantitative research

Although the statistical analyses revealed a strong association between the number of returnees and the amount of violence and presence of mines, three different explanations may explain why people do not want to return when there are natural resources in their places of origin.

One possible explanation is that people could be afraid to go back to their homes close to mines because of anticipated or perceived violence surrounding the mines. They left

because of the violence or the consequences of mines and may have expected that these issues remained. Although the previous regression analysis controlled for past conflicts,

recollections of conflict may have biased perceptions regarding the likelihood of future conflicts associated with natural resources. If people perceive that violent incidents are associated are with natural resources, this may result in a decrease of returnees.

A second explanation could be that people do not come back to their place of origin because villages are already destroyed by previous conflict or mining operations, and as such their houses are uninhabitable.

A third explanation could be that mining companies may have forced people to leave, and people do not return because of these land-grabbing practices. The presence and operating of multinational mining companies were not considered in the quantitative part of this thesis, but it does offer a potential explanation.

The final part of the quantitative analyses showed that gold and wolframite were significantly related to the number of returnees. This could be because gold and wolframite are easier to exploit than other minerals. Alternatively, the mining of these natural resources could be more harmful to civilians, for example because of increased water pollution, and therefore people are less likely to come back after displacement. A third explanation could be that gold and wolframite are more interesting for armed groups, promoting violence and competition around these mines. Companies often have access to advanced equipment and are less dependent on resources that are easy to exploit.

(32)

In the qualitative part of the research, these interpretations will be evaluated through interviews with several experts. Taken all together, I will examine if the interpretation(s) are all plausible or which interpretation is the most plausible to explain the impact of (different) natural resources on the number of returnees.

6.5 Qualitative research

In this section, I incorporated the results of the interviews into the structure of my mechanism: The presence of mineral mines in an area attracts more armed groups, in turn leading to more violent incidents between the different groups. This process results in more displacement and less returnees.

I conducted four expert interviews. The first interview was with Fleur Scheele7_{, a}

researcher at SOMO. Scheele did research on cobalt mines in Katanga and published the results in a report, Cobalt Blues. In Cobalt Blues, Scheele shows the role of the region Katanga, on cobalt’s world market. The natural resource cobalt is used in many electronics and industrial applications but the presence of cobalt mines has led to human rights violations and environmental negligence, causing internal displacement and a decrease in returnees. The second expert interview was with Han van Dijk8_{, a researcher at the African Studies Centre}

Leiden. The third expert interviewee was Dirk-Jan Koch9_{, the Special Envoy of Natural}

Resources at The Dutch Ministry of Foreign Affairs. He worked as a diplomat at the Dutch embassy in DR Congo for five years. The fourth expert interview was with Clens Santa10, who fled from DR Congo twenty years ago with his family during the regime of Mubutu. His father opposed the regime, and the family became political refugees. Santa and his family left the country, so they are refugees rather than IDP. Nevertheless, since his family remained in DR Congo, he is well informed about the current and historical situation.

First, I will outline the descriptions of the respondents on the influence of mines, armed groups and violence in DR Congo. Second, I will outline the descriptions of the respondents on reasons for internal displacement and returning home after being displaced. Third, I will summarize this information into the most accurate answer to my argument. Finally, I will discuss alternative reasons for the decrease of returnees in DR Congo.

7_{Translated from Dutch (Fleur Scheele, personal interview, April 26, 2016, Amsterdam), complete interview}

available upon request.

8_{Translated from Dutch (H. van Dijk, personal interview, May 18, 2016, Leiden), complete interview available}

upon request.

9_{Translated from Dutch (D. Koch, personal interview, May 24, 2016, Utrecht), complete interview available}