1
Natural resources characteristics and
type of conflict
By Natasja Branković
June, 2014
Master thesis Political Science: The Political Economy of Conflict
Supervisors: Geoffrey Underhill Brian Burgoon
2 Contents
1. Introduction 4
2. Natural Resources and Conflict; a review of the perspectives 7 2.1 How are natural resources connected to conflict?
2.2 Does the type of resource matter? Oil versus Diamonds
2.3 Lootability
The extraction process The Location of resources The Geographical concentration 2.4 Legality
2.5 Typology of resource characteristics 2.6 Operationalization and Methodology
3. Angola 21
3.1 Colonial Angola
3.2 the Independence Struggle 3.3 Civil war
The post independence and post cold war period The post cold war period: talking and fighting The war comes to an end
4. Oil in Angola 33
4.1 The Oil industry 4.2 Oil Characteristics
Where is the oil? Concentration of oil 4.3 Legality of oil
5. Diamonds in Angola 44
5.1 The diamond industry 5.2 Diamonds Characteristics
Where are the diamonds? Concentration of diamonds 5.3 The Legality of Diamonds
3
6. Conclusion 53
7. References 55
Maps
Oil fields in Angola 38
Concentration of diamond fields 47
Figures
Angolian Oil Production 32
Offshore and onshore oil production 38
Tables
Lootability 13
Location 14
Concentration 15
Typology of characteristics and type of conflict 17
Composition of GDP by Sector, 1966–2004 25
Government expenditures 36
Government expenditure by function 40
Breakdown of Angolan GDP by sector 44
Diamond production in Angola (carats) 45
4 1 Introduction
Natural resources have become of more importance in civil wars. With the demise of the Cold War and the withdrawal of foreign sponsorship in the 1990s, natural resources have gained strategic importance for rebels and governments in conflict situations. From Liberia to Sierra Leone, both rebels and states have sought alternatives of revenues and the availability of local resources created new opportunities for them. The trade in natural resources enabled governments and rebels to finance the war and sustain it. With few exceptions, it had become more evident that natural resources were more of a curse than a blessing to the stability to developing countries (bron). The strategic importance of natural resources in civil wars has been recognized within academic and policy circles and resulted in a substantial body of literature.
Within this context, most studies have focused on the economic dimensions of civil war arguing greedy rebellion is the major cause of conflict and not social, ethnic or political grievances (Collier & Hoeffler, 2003) and offered important insights into the relationship between natural resources and conflict. However, as a result of studies using different measures, different units of analysis and different interpretations others have emphasized “that we still know little about the processes that tie natural resources” (Ross, 2004a; Humphreys, 2003). In addition to the relationship being complex, most studies do not reflect on the role played by the nature of natural resource and their particular characteristics in relation to the type of conflict that might emerge. The problem is unless we understand how the mechanisms work and how natural resources matter for conflict, it will be difficult for scholars and policy makers to suggest policy development and action on this subject.
Therefore this research argues that for a more comprehensive understanding of the relationship between natural resources and conflict, it is necessary to explore the role of the nature of natural resources and their characteristics. The central question in this research revolves around the particular characteristics of natural resources and their implications on the nature and duration of conflict. Answers to this question can provide a more comprehensive and systematic understanding on how these mechanisms in relation to natural resources and conflict work and may contribute to more complete theories. Furthermore, viewing conflict from a natural resource perspective can lead to different suggesting concerning conflict resolution and peace building processes. This said, viewing conflict from this perspective should be seen as an important contribution to the contemporary debate.
5 This research was inspired by Le Billon and will draw upon his theoretical framework. Le Billon argues that “the nature and geography of resources play a crucial role in shaping conflicts” (2001:570). His theoretical proposition that resources have particular characteristics that might shape patterns of conflicts have not have not been carefully tested, but should be because they can have powerful policy implications. By building upon Le Billons arguments this research will produce two broad arguments. Firstly it will argue that a natural resource is likely to shape more favourable opportunities for rebellion and for the conflict to take the form of a war economy (“warlordism”), when the nature of a natural resource requires less capital intensive skills and equipment for its exploitation, it is located in remote areas distant from the capital, it’s more spatially spread and when its access to the market is legalised. Secondly, it will argue that in the case of a natural resource requiring capital and technical expertise, located in areas proximate to the capital, more concentrated and access to the international market is legalised, it is more likely to produce more favourable conditions for the state and it is more likely to be associated with coup d’états.
Therefore this research will offer an important basis establishing the role of the nature of natural resources and their characteristics. To do so it will develop a framework based on resources characteristics in relation to nature of conflict for future scholars. Furthermore, it will offer an important signal to countries that possess different resources with particular characteristics. This however does not mean other non resource specific factors should be neglected, such as social and ethnic relations, they should be incorporated as well but this is beyond the scope of this research.
To illustrate the above arguments, this research will use an in-depth case approach drawing on the experiences of Angola’s natural resources industries. One of the advantages of using a cast study approach is it will allow us to identify the mechanisms more carefully, which is more difficult when using large n studies. Furthermore, Angola has an abundance of different types of natural resources with particular characteristics, such as oil and diamonds. Because this research is interested in finding how different types of resources interact in relation to conflict, it will disaggregate these resources and rely on the oil and diamonds sources derived from the PRIO databases.
This research falls into 5 parts. The first part (chapter 2) will discuss the different perspectives linking natural resources, their characteristics to conflict, their empirical findings and their limits and contributions. Furthermore it will introduce a typology of particular
6 characteristics of natural resources in relation to the nature of conflict and discuss the generated hypotheses. Chapter 3will introduce the case of Angola and provide a historical analysis arguing that underlying grievances may have triggered the Angolan conflict, but that in the late 1980s natural resources became the driving factor in determining the course of the civil war. Chapter 4 will introduce the case of oil and argue that the nature of oil and its characteristics are crucial variables in explaining the seizure of Luanda by the MPLA movement. Chapter 5 will investigate the characteristics of diamonds and argue that resources, such as alluvial diamonds that are accessible in terms of location, more scattered and easier to extract and market are more likely to be associated with warlordism conflicts. Finally chapter 6 will summarize the findings of this research and draw conclusions for future research and policy implications.
7 2 Natural Resources, their characteristics and Conflict; a review of the perspectives
Do particular characteristics of natural resources matter for the nature and duration of conflict? To be able to answer the research question of this research this theoretical chapter will draw on the theoretical framework of Le Billion, which studies different types of natural resources and its relation to type of conflict. According to Le Billon, resource specifics such as location and geographical concentration of a resource matter for the type of conflict. These factors remain largely untested and this research therefore will build upon Le Billon’s theoretical framework so as to contribute to its empirical underpinnings. It will argue that resource specifics are crucial in shaping the opportunity structure for both rebels and the state and they matter for the type of conflict that might emerge. This research will furthermore add another dimension to Le Billions theoretical framework and argue resource specifics do not only matter for type of conflict, but also for the duration civil conflict.
This chapter will look at what we know about the various factors affecting conflict and falls into three parts. Firstly it will discuss the different strands concerning the natural resource – conflict nexus. Then it will explore different types of resources and introduce Le Billons’ theoretical framework. And lastly it will develop a typology of different characteristics in relation to nature of conflict.
2.1 How are natural resources connected to violent conflict?
Several strands of literature offer a range of explanations as to how natural resources are connected to violent conflict. This section will elaborate on the different theoretical approaches which fall within the opportunity versus grievances debate, the empirical evidence found and their merits and limits. Although these studies have contributed to the study of civil war, it will be argued that most scholars have neglected an important dimension in explaining the complex relationship between natural resources and conflict, namely the nature of natural resources and their particular characteristics in relation to conflict. Firstly, the grievances debate will be discussed, followed by two approaches within the opportunity debates.
One strand of approaches explaining how natural resources wealth interacts with violent conflict is the grievances debate. The idea behind this debate suggests that underlying political, ethnic or social motives are most important in explaining the outbreak of conflict. This argument is derived from the relative deprivation theory, which emphasizes the
8 perceived deprivation of social groups. Smith argues groups and individuals compare their situation with that of another that impacts on how beneficial or worse their situation is perceived (2001). The negative consequences of natural resource wealth and particularly the uneven distribution of this wealth can create grievance among groups and or individuals and trigger violence. The mismanagement of natural resource wealth or the perception of this mismanagement may fuel local resentments and feed violence as occurred in Bougainville Papua Guinea and Sudan (Regan, 2003). Humphreys et al argue that the extraction process itself may result in grievances, such as forced-out migration, environmental pollution and degradation (2007:13).
Having established that the grievance argument insists that political, ethnic and social motives lie at the heart of the conflict, another approach claims that not underlying grievances are the most important factor in explaining the outbreak of conflict, but the most important factor is the immediate economic gains and opportunities available to potential rebels. On the one hand the availability of natural resources provide the motive to capture natural resource rents and on the other hand natural resource rents provide an opportunity to finance violence. The opportunity argument (or greed) is thus about motives and opportunities to start violence. It suggest rebels are rational actors who have an economic motive based on greed or private gain and therefore can do “well out of war” (Collier, 1999). The economic opportunity position suggests the natural resource rents provide the financial means to start violence. The notion is the existence of natural resource rents increases the risk of onset of violent conflict and is more important than underlying grievances as an explanation of the occurrence of civil war.
The most influential empirical work supporting the opportunity position is that of Collier & Hoeffer (2004) who argue that a less developed economy based on primary commodities exports increases the likelihood of onset of civil war. In their study they attempt to predict for the onset of civil war during a 5-year episode by testing two models where one is based on the opportunity model and other on grievance. Their findings suggest the opportunity model has more explanatory power than the grievance model and their model suggests that when primary commodity exports are equivalent to 32 per cent of the GDP, the risk of civil war is about 22 per cent while a country with no such exports has a risk of only 1
9 per cent (580)1. However beyond this point the likelihood of civil war diminishes; the larger the amount of natural resources countries export, the less the likelihood of civil war. This they argue means that states beyond a certain point will be able to outspend rebels and the motive and in particular the opportunity for rebels to “do well out of war” decreases. Furthermore Collier & Hoeffler found that for one particular resource, oil, the risk of civil war is most likely.2
Their results however should be treated with caution as there several methodological and analytical limitations to their findings and interpretations. Firstly, other factors are excluded in their study and Maloney & Nitzschke find that other natural resources, such as diamonds or narcotics are also crucial in financing rebels (2004:5). Furthermore country case studies have found a greater level of complexity: it is not the economic motives and opportunities as claimed by Collier & Hoeffler that cause the onset of violent conflict, but rather the interaction of economic motives with various social and political grievances (Ballentine & Nitzscke 2009:5). Moreover, Collier & Hoeffler’s study has been ‘rebel centric’, neglecting the role of other actors, such as the state.
Another approach supporting the opportunity position explaining the link between natural resource wealth and violent conflict is the weak state approach. It focuses on state institutions in contrast to Collier & Hoeffler and provides an understanding as to how natural resources affect conflict via state capacity. This proposition is twofold. The first factor underlying this proposition relates to the state and the second factor to the society. Natural resource wealth, it is argued may weaken the state because this wealth allows states to become rentier states. As a result of the resources rents, rentier states do not have to rely on domestic taxation and therefore do not have the need to develop a strong bureaucratic apparatus to levy taxes on its citizens. It is argued that much of the state’s strength comes from its ability to establish the type of bureaucracy that can provide effective public goods and regulate social conflicts (Ross, 2003:10). From the perspective of the society, not raising tax revenues on citizens might lessen the power the state has over them. Furthermore the state might be less able to respond to citizens’ demands and engage them with society, which could make conflict more likely. These two factors combined, are likely to lead weak
1 Collier & Hoeffler measure the state’s dependence on natural resources as the ratio of primary commodity
exports to GDP.
2
10 organizational structures and the weaker the state’s capacity, the greater the likelihood of civil war onset.
Fearon & Laitin (2003, Fearon, 2005) find support for the weak state model and argue that oil dependent countries are more prone to civil conflict than non-oil countries. They challenge Collier & Hoeffler’s study and find no connection between primary commodities export and civil war3. By contrast they argue “that economic variables such as per capita income matter primarily because they proxy for state administrative, military, and police capabilities” (ibid, 4). Furthermore they argue oil exports may trigger conflict because oil exports have negative effects on state institutions, such as weak political control. In turn a state without strong institutions renders more favourable conditions for insurgency because it may be more difficult for such a state to deter rebellion.
To sum up, the above debates indicate that the natural resource wealth-conflict nexus in general is contested among scholars and that the empirical findings vary. The best we can say is that the relationship remains complex. Furthermore, the available literature does not address all aspects of the problem and downplays the significance of resource characteristics. If natural resources appear to play a role in conflict situations, do the particular characteristics of specific types of resource matter, and if so how and why? The next section will introduce another framework for explaining the complex relationship between natural resources and conflict and argue how different types of resources, particularly their characteristics are likely to influence conflict dynamics.
2.2 Does the type of resource matter?
The previous section has pointed out the literature discussed does not address all aspects of the natural resources and conflict nexus. In order to better understand the complex relationship between natural resources and the nature and duration of conflict, this section will introduce a framework that analyzes the contrasting characteristics of the particular types of natural resources and the revenues that they produce in relation to type and duration of conflict (Le Billon 2001; Auty 2001; Ross; 200). The banana curse may be different from the oil curse (Basedau & Wegenast, 2009:39) because it is likely to shape the opportunity structure for rebels differently. Very few scholars have taken these differences into account and this section will introduce two resource type dimensions potentially influencing the
3 They eventually cover the dame years as Collier & Hoeffler (1960-99) and argue the opposite, namely that a
11 natural resource and conflict relationship (Le Billon, 2009). These two dimensions of lootability and legality, it will be argued are crucial in understanding how natural resources are being exploited and how they may come to benefit different actors in the type of conflict they produce. Firstly, the findings of oil and diamonds in relation to conflict will be discussed followed by a discussion concerning the lootability and legality aspects in relation to the type of conflict that might emerge.
2.2.1 Oil & Diamonds
Several empirical studies using different datasets and varying definitions of oil dependence, suggest oil is connected to civil conflict. De Soysa (2002) and Fearon & Laitin (2003) (they measure oil exports as a percentage of GDP) find that oil exporting countries are more prone to conflict than non-oil exporting countries. A more recent study by de Soysa & Neumayer who measure resource wealth in terms of resource rent (energy and mineral rents) as a share of gross national income confirm the Fearon and Laitin model that energy wealth (fossil fuels), not mineral wealth increases the risk of civil war (2007:202).4 However the coding Fearon & Laitin use (a threshold of a 1000 battle deaths) is not significant in their model and their results are only significant using a lower threshold of 25 battle deaths. Ross (2004) also concludes that oil matters for conflict after examining 13 cases. Furthermore, according to Lujala et al. (2007) the location of oil apparently matters because they are more likely to be associated with longer governmental conflicts (2007:239). Besides oil, a previous study by also Lujala et al (2005) find that countries possessing secondary diamonds (alluvial diamonds) are more likely to experience ethnic conflicts. Ross finds that the civil war rate among diamond producers is high (five wars in eighteen states), but among the alluvial diamond producers is exceptionally high (four wars in eight states) (2003:52).
The picture emerging from these studies is that there is a relationship between the different sort of resources, oil versus diamonds, and the nature of conflict that might arise. is association with oil and type of conflict and with the production of alluvial diamonds and type of conflict. What accounts for this pattern? In this context two important dimensions are of importance that might influence the natural resource and conflict relationship. Firstly, the most known dimension of lootability distinghuises lootable and unlootable resources. Ross
4
Their measure of resource wealth is aggregated according to mineral and energy rents and they claim that this measure instead of primary commodity exports like Collier & Hoeffler use, this measure of natural resource wealth is more capable to capture the value of natural resources more precisely and therefore is more valuable to test the link between natural resources and conflict.
12 argues lootable resources, such as alluvial diamonds, timber and narcotics are more likely to be associated with non-seperatist conflicts and may prolong the conflict. In contrast, unlootable resources such as kimberlite diamonds, oil and gas are more likely to be associated with separatist conflicts. The second dimension refers to legality distinguishing legal and illegal resources where legal resources are (Le Billon, 2009:17)
The following sections will analyze both lootability and legality as characteristics of natural resources and argue how they become two important explanatory factors behind determining the nature and duration of conflict related to particular resources. Furthermore, there a many factors influencing lootability and legality and these will be discussed in the following sections
2.3 Lootability
The most well known dimension possibly influencing the relationship between natural resources and type of conflict is the idea of lootability. Lootability refers to “The resource’s ease with which a rebel group could access revenue from this resource” (Le Billon, 2009:17)5 . Drugs, alluvial diamonds and timber are categorized as lootable resources while oil and gas are considered to be unlootable. The proposition is that in the case of available lootable resources, the risk of civil war is likely to be higher. There are several components possibly influencing the lootability of natural resources. The next section will discuss these components, the location and geographical concentration and of natural resources and how they might influence the lootability of a resource.
2.3.1 The extraction process
One of the components influencing the lootability is related to the extraction process of the different types of resources. There are two types of resources in this context, those requiring less skills, capital and equipment to exploit them, such as alluvial diamonds. These type of resources are highly lootable because they are easy accessible, their rewards are high and they can be extracted with minimal effort.6 And those resources that require technological know-how and capital to explore and exploit them, such as offshore oil. These resources are highly unlootable because they are more difficult to access and require capital intensive skills
5
Michael Ross defines lootability as the ease with which a resource can be extracted and transported by individuals or small teams of unskilled workers (in Karen Ballentine and Jake Sherman, 2003: 54)
6 Onshore oil is considered to be lootable, but to a lesser extent than alluvial diamonds because in terms of the
13 and equipment. These characteristics might be more attractive to rebels in the case of highly lootables and in this case conflict is to be more likely. On the other hand conflict will be less likely in the case of unlootables because they are less attractive to rebels. Lootability thus will likely affect the opportunity structure for rebels to finance their activities. Table 2.1 summarizes the different types of resources in terms of lootability. Offshore oil and primary diamonds are considered to be unlootable while onshore oil and secondary diamonds are considered to be lootable. Although onshore oil is a capital intensive activity, it is still a lootable. In terms of accessibility the conditions for capturing onshore oil are more likely to be favourable for rebels than for capturing offshore oil. Rebels can extort oil companies or tap pipelines or other onshore infrastructure. In Nigeria for example rebels tapped pipelines and in Colombia guerrillas extorted the oil sector. So, the extraction process influences the lootability of a resource; the easier extractable a resource is the more lootable and the more difficult a resource is extracted the less lootable. The following section will discuss how another factor, location influences the lootability of a resource.
Table 2.1 Lootability
Resource Unlootable Lootable
Oil Offshore Onshore
Diamonds Primary Secondary
2.3.2 Location of resources
Another factor suggested by Le Billon that shapes the nature of conflict relates to the location of natural resources (2007:8-9). The assumption is that a resource proximate to the capital is less likely to be captured by rebels than a resource more distant (and nearer the border) from the capital. The political representation is often centred in the capital and the greater the distance from the capital, the greater the difficulty and accessibility for the state to control remote parts. In this case the conditions are likely to be more favourable for rebels than the state. Fearon & Laitin argue
Resources can thus be categorized as proximate and distant resources. Table 2.2 adds the location component to the previous table. On the on hand, there are proximate resources, such as offshore and onshore oil that proximate to the capital and therefore easier defendable for the state. On the other hand, there are diffuse resources, such as primary and secondary
14 diamonds that are more distant from the capital and therefore more difficult for the state to control these parts, but most likely to be easier to capture by rebels. The following section will discuss the last factor of geographical concentration and how it shapes the nature of conflict.
Table 2.2 Location
Resource Unlootable Lootable
Oil Offshore (proximate) Onshore (proximate)
Diamonds Primary (distant) Secondary (distant)
2.3.3 The geographical concentration
In addition to the location of resources, the geographical concentration of resources affects the lootability of resources. Auty identified two categories of resources in this context: these are point and diffuse resources (2001). Point resources are concentrated in small areas and often require capital intensive skills and equipment for their exploitation. As a result point resources are unlootables because it is more difficult for rebels to access them in terms of location and exploitation. However, point resources are more likely to benefit the government because they are easier to access in terms of location exploitation. Resource rents derived from the exploitation of point resources can be an incentive for the state to defend these resources and as they are located proximate to the capital this is likely to be easier. Oil, in general is concentrated in small areas, and for the exploitation capital and large infrastructure is required. Although onshore oil is considered to be a point resource, as mentioned before it is still a lootable resource and therefore the conditions are more likely to be favourable for rebels than for capturing offshore oil. On the other hand diffuse resources are more spread over wide areas and more likely to be looted because it is more difficult for the state to control these remote areas. Furthermore they are easier to extract as minimal skilled labour and equipment is required. Alluvial diamonds, timber and fish are examples of diffuse resources. Diamonds in general are diffuse resources as they more scattered over wide ares, however primary diamonds are unlootable because they are more capital and technological intensive than secondary diamonds are. In Burundi, access to kimberlite diamonds has been more complicated than the case in Sierra Leone where alluvial diamonds
15 provided tens of millions of dollars for the RUF (The Washington Post, 2001). Table 2.3 adds the concentration dimension to the natural resources.
2.3 Concentration of resources
Unlootable Lootable
Oil Offshore (point) Onshore (point)
Diamonds Primary (diffuse) Secondary (diffuse)
The previous sections have illuminated how different factors, such as the extraction process, location and geographical concentration might shape the nature of conflict. In relation to the extraction process, resources that require minimal effort in terms of skills and equipment are more likely to be captured by rebels than resources that require capital intensive skills and equipment. Furthermore, where resources are located is likely to affect the opportunity structure for rebels and the state. Resources located proximate tot the capital are more likely to benefit the state as it will be easier to defend these parts. Resources that are located more distant from the capital however, are more likely to benefit the rebels as it will be more difficult for the state to control these remote areas. Finally, point resources may be more easily monopolised than diffuse resources because they are concentrated in smaller areas (Le Billon, 2001:572). The next section will discuss another factor that of legality and argue how this factor might affect the opportunity structure for rebels and/or the state.
2.4 Legality
The second dimension influencing the relationship between natural resources and conflict relates to legality. Legality refers to “(i)legal character of a resource along the commodity chain” and it is defined in judicial and moral terms (Le Billon, 2009:17). In judicial terms this refers to the status of a resource in domestic and international markets. Most resources can be legally traded on markets and are subjected to international trading rules. However, in the case of illegal commodities such as narcotics there are international sanctions against it. Morally, this means that what is illegal by law can be considered legal by the population. The assumption is that an illegal resource will be more favourable for rebels than for the state because the state risks losing its legitimacy on the domestic and international market if it engages in illegal trafficking. Furthermore, states are less likely to profit from illegal natural resources because of the risk being subjected to international
16 sanctions (Ross, 2003:63). The legality of a natural resource in judicial and moral terms is likely to affect the opportunity structure for rebels, rebels being advantaged in the case of illegal resources and the government advantage in the case of legal resources. In the case of oil and diamonds, onshore oil and alluvial diamonds are considered as legal lootables and offshore oil and kimberlite diamonds as legal unlootables. It is important to stress however that trading in kimberlite diamonds became ‘illegal’ in 2003 when the The Kimberley Process Certification Scheme was established in response to the role of ‘blood diamonds’ in financing armed conflict in Sierra Leone, Liberia, Angola and the DRC (Global Witness, 1998). These sanctions imposed by the international community attempted to regulate the illegal diamond trade which has played a role in financing the conflict in Angola, Sierra Leone, Liberia and the DRC. Furthermore in this context the creation of KPCS was able to label these resources as illegal.
So far this research has discussed particular characteristics of oil and diamonds and how these characteristics (lootability and legality) might shape different opportunities for rebels and/or the state. It is however interested in how these different characteristics shape the nature of conflict. The following section will therefore discuss the nature of conflict in relation to the different characteristics of natural resources. Also the hypotheses set up! 2.5 Typology of characteristics and conflict
The previous section has established that particular characteristics, such as the extraction process, location, concentration and (i)legal status of natural resources shape the opportunity structure in favour of rebels and/or the state. This section will build upon these specific characteristics and relate these to the nature of conflict. What are the implications of these characteristics on the nature of conflict that might emerge? Firstly it will be argued that the social construction of natural resources play a role in how we understand the nature and duration of civil conflict. Since the demise of the Cold War, natural resources have become of more importance to civil wars. Then, Furthermore, particular resources are more likely to shape different opportunities for prolonging wars. Then it will discuss the hypotheses set up. Time and the meaning society give to natural resources as well as the function of them are crucial in understanding the role of natural resources. As Zimmerman argued “resources are not, they become” (1951). Resources are neutral, humans recognize their presence, give them a function that they are desired or needed. It is about the importance attached to them and how groups construct their notions of natural resources. This resonates
17 with Kaldors “new wars” logic who argues since the demise of the Cold War, it about who has access to the resources (2003:1). Empirical evidence suggests a strong relation for diamonds, particularly alluvials, on conflict in the post Cold War period (Lujala et al. 2005). Thus, the meaning society gives to natural resources needs to be kept in mind. Based on the different characteristics discussed, we can yield a typology of these characteristics in relation to the nature of conflict. To summarize, the extraction process, the location, the geographical concentration and the (i)legal status of a natural resources have been discussed as characteristics of natural resources. Within this context two types of resources have been identified, point and diffuse resources. These characteristics are likely to have implications on the opportunity structure of both rebels and states and therefore might shape different types of conflict. Table 2.4 shows the typology developed and categorizes these two types of resources in relation to their other characteristics.
Table 2.4 Typology of characteristics and type of conflict
Characteristics
Type of Conflict
Coup d’état War economy
(warlordism)
Lootability Capital intensive Less capital intensive
Proximate Distant
Concentrated Spread
Legality Legal Illegal
Having established this typology of the characteristics of natural resources in relation to conflict, the following propositions have been generated. Firstly, the typology suggests that point resources that are capital intensive, proximate to the capital and legal are more likely to be associated with coup d’états. Resources that are concentrated in smaller areas, closer to the capital and are capital intensive, such as offshore and onshore oil are more likely to be defended by the state. Rebels or opposition groups hoping to capture the oil resources will have to capture the state. The first proposition is:
H1resources that are capital intensive, proximate to the capital, concentrated in small areas and legal are more likely to be associated with coup d’états.
18 Secondly the typology suggests that diffuse resources, not capital intensive, distant from the capital and illegal are more likely to be associated with war economies (“warlordism”). According to Ballentine & Nitzschke war economies have several distinctive features, such as “they involve the destruction of the formal economy and the growth of the informal economy, they rely on the licit or illicit exploitation of / trade in lucrative natural resources and they thrive on cross-border trading networks” (2005:2) These types of resources, such as alluvial diamonds produce high rewards for rebels because they easier accessible and marketable. They do not require capital intensive skills and can easily be sold on the international markets. Furthermore, the distance from the capital shapes more favourable conditions for warlords/rebels and makes it harder for the state to control these remote areas. Also, remote areas near the border enable cross-border trading networks. This leads to the second proposition:
H2 resources that are less capital intensive, distant from the capital, geographically scattered and illegal are more likely to be associated with conflicts taking the form of “warlordism”.
These two propositions will be tested in the empirical part of this research. Firstly the different variables and the methodology will be discussed in the following section.
2.6 Operationalization and Methodology
The dependent variable of this study is conflict, in particular the specific nature and the duration of conflict. Violent or armed conflict is defined as “a contested incompatibility, which concerns government and/or territory where the use of armed force between two parties, of which at least one is the government for a state, results in at least 25 battle-deaths per year”. This definition is derived from the Uppsala Conflict Data Program and distinguishes incompatibilities, concerning the government and concerning territory. To be consistent with previous studies on the relation of natural resources and conflict this study will use this definition of conflict. There are different datasets available; however being consistent this will benefit further research as well. The UCDP data related to conflict sites will be used as it provides data on the location the conflict has taken place and the nature of the conflict.
The independent variables of this study are related to the characteristics of natural resources. The first independent variable explaining the relationship between natural
19 resources and the nature of conflict, relates to lootability. This variable is influenced by the extraction process, the location and concentration of natural resources. This research will disaggregate data on the location and concentration of diamonds and oil. The PRIO diamond dataset (diadata) will be used because it produces data on the exact diamond production sites. Furthermore, descriptive statistics, such as production figures concerning diamonds and oil will be sampled. Not only will the formal economy be taken into consideration, figures of the informal economy will be crucial as well. This is important data as it is expected rebels have raised revenues through informal networks. Concerning oil, the petroleum dataset (petrodata) of PRIO will be used to locate the exact on and offshore deposits. It is crucial to obtain the exact production sites in order to connect these to distance to the capital.
The second independent variable explaining the relationship between natural resources and the nature of conflict related to legality. This variable is influenced by the domestic and international status of oil and diamonds.
In order to test the generated hypotheses, this research will use an in-depth case study approach drawing on Angola’s natural resource experiences. It is an attempt to understand how two different types of resources and particular their characteristics shape the nature of conflict. This research will examine two cases from Angola, one based on oil deposits and the other on diamonds deposits. Why Angola? Firstly, it is a representative case because it offers variation between different types of resources and their characteristics. Angola has oil and diamond deposits with particular characteristics. As discussed in section alluvial diamonds are lootables and oil is less lootable. The presence of different resources enables this research to compare and contrast and we will see that resources with particular characteristics, such as lootability are likely to shape the nature of conflict.
Furthermore, variation regarding the dependent variable is present, although Angola experienced a civil war the nature of the conflicts and groups involved differed. UNITA under the leadership of Jonas Savimbi, challenging the MPLA operated in central Angola. The MPLA was like UNITA a political movement that engaged in the civil war. These were however not the only parties involved as international powers, such as Cuba, Russia and the US have played a role in the Angolan Civil War as well. Therefore this research has chosen to split the Angolan civil war period in 3 periods to trace the process of the conflict. Besides tracing the process of the conflict, it will also allow to process the meaning and function of natural resources as well.
20 This chapter has developed a typology based on the literature discussed. It has argued that while other perspectives have offered important insights regarding the natural resources and conflict nexus, they do not reflect on the role of the particular characteristics in relation to the nature of conflict. By viewing conflict from a characteristics perspective, this research attempts to improve our understanding on these complex dynamics. Furthermore this chapter has set up two propositions to be tested using an in-depth case study approach within Angola. The following chapter will provide an in-depth background analysis of Angola and argue that the Angolan Civil War may have been driven by underlying political, ethnic and social grievances, but that the most important factor determining the course of the civil war in the late 1980 has been the availability of natural resources for both the government and rebels.
21 3 Angola
Several factors have played a role in shaping contemporary Angola. Firstly its colonial heritage cannot be dismissed; it has been a Portuguese colony for almost five centuries. Its abundance in natural resources is relevant to its economy and has affected the course of the civil war. Furthermore the colonial struggle against Portugal marked the way for the civil war that broke out in 1976. In order to understand the role natural resources have played in the course of the civil war it is necessary to start telling the story from the beginning. It will be argued that although the parties may have been driven by underlying political, ethnic and social grievances, the most important factor that affected the course of the civil war in the late 1980s and beginning of 1990s were the available economic gains and opportunities to potential rebels and the state.
This chapter falls out into 3 sections. Firstly Angola under the Portuguese will be analysed and it will point out that long standing grievances can be traced back to this period and had been present even before the civil war started. Then, the second section will show these grievances became politicized by the movements during the independence struggle and how they contributed to the conflict that followed. The last section will discuss the civil in 3 periods where the first period is referred to the cold war period. This period was mostly characterized by the cold war ideological struggle where foreign assistance sustained the war. Then the following period will demonstrate once foreign assistance declined, economic imperatives became of more importance, particularly how the war had become more of a resource war. Lastly the last period will be analysed and discuss how the war came to an and how the sanctions imposed by the UN played a role in decreasing UNITA revenues generated from diamonds.
3.1 Colonial Angola
In 1576 the Portuguese established the colony of Angola and was ruled by the Portuguese for nearly five centuries. The Portuguese paid relatively little attention to Angola than to Brazil, which had a growing sugar economy. Brazil’s growing economy was in need of labor and Angola became the main source for slavery trade. It is estimated that approximately 2 million slaves that went to Portugal’s overseas colonies, such as Brazil and the Caribbean may have come from Angola (Warner, 1991: 15 in Collelo). The trade in slaves established a pattern of exploitation and advantage among Angolans. The Mbundu
22 peoples living mostly in Luanda in their hinterlands came to be regarded of the collaborators with the Portuguese engaging in the trade in human beings (Cornwell, 2000:1).
In the twentieth century Angola’s future prospects looked brighter when Antonio Salazar came to power in Portugal. Salazar introduced the Colonial Act, which brought economic, political and social policies implemented in Angola in line with the policies at home in Portugal. The measures taken however, did not bring real prosperity to Angola until the Second World War when higher coffee prices led to an explosion of economic growth brought economic growth. The coffee boom created a greater demand for forced labour and Salazar induced a great number of Portuguese immigrants to settle in Angola. In reality however, Salazar’s policies had created a more divided society in two ways. It firstly produced a divide among Angolans. On the hand mestiços/asimilados could easily become citizens, but on the other hand, indigenas could not and were defined as a separate element in the Angolan population. Mestiços or asimilados were multiracial, mostly educated and living in urban areas and were regarded as the collaborators of the Portuguese. Indigenas were black, uneducated and mostly living in rural areas. They felt mestiços/asimilidos were privileged by the Portuguese and resented them for it. Secondly the Portuguese immigration created a divide between the Portuguese and mestiços as they began to compete for employment. Mestiços until then had identified with the Portuguese. Angolans could also not qualify for certain roles because they had to have a certain level of education that most Angolans did not have.
To wrap up, this section has shown that long standing grievances, embedded in Angola’s historical memories about the trade in human beings and in cultural attitudes was present in Angolan society. Salazar’s policies had led to an intensification of economic and social tensions. His policies created a divided society in racial, material and political terms in which the centre was privileged and separated (mestiços) from the other rural parts of the country (indigenas). The economic and social tensions gave rise to African nationalism and anti-colonial sentiment. The next section, in turn, will analyse these grievances and how they became politicised during the independence struggle.
3.2 The independence struggle
The measures taken during the colonial system created a divided society and shaped the domestic political arena in Angola. This section will discuss the emergence of the 3
23 political movements in Angola and how this division was reflected in their struggle for independence. Furthermore it will discuss how different factors, such as ethnicity and territory have contributed to new grievances and potential conflict. It will also be argued that at the end of the independence a shift in the economy structure occurred; Angola had become an oil exporter.
The first political associations appeared in the 1950s and began activating the anti-colonial sentiment in Angola. In 1956 the MPLA ((Movimento Popular or Libertacao de Angol) was formed with the aim to overthrow Portuguese rule and establish an independent Angola. The MPLA was based on Mbundu nationalism and its peoples were mostly located around the Kwaza river from Luanda to the Cassange highlands in the north east of Angola. The Mbundu peoples were the first to be subjected to Portuguese rule and collaborated with the Portuguese. They became to be regarded as collaborators engaging in the exploitation of human capital in the interior of Angola. As a result of their collaboration they enjoyed more privileges, such as studying in Portugal and having access to certain position the indigenous Angolans did not have. They tended to identify with the Portuguese culture, because they spoke Portuguese and acted like the Portuguese; they became known as mestiços and assimilidos.
Another association, the FNLA (Frente Nacional de Libertaçio de Angola) emerged from the UPA (União das Populações de Angola) organized by Holden Roberto in 1958. It was centred on the old Kingdom of Kongo in northern Angola, an area less influenced by the Portuguese than the Mbunda area of Luanda. They were mostly concentrated in Uige, Zaire and the Cabinda province. Nationalism among the Bakongo peoples emerged from the colonial practice recruiting forced labour for the cotton plantations. The harsh and inhuman conditions during the colonial era resulted in many Bakongo fleeing across the border to Belgian Congo. The large refugee population in Congo enabled the FNLA make its headquarters in Leopoldville.
Besides the FNLA and the MPLA being established in the 1950s, another association was formed by Jonas Savimbi, UNITA (União Nacional para a Independencia Total de Angola). UNITA was to represent the Ovimbundu peoples mostly located in the central highlands of Angola (in Bailundo, Bie and Moxico provinces). The Ovimbundu, mostly peasants, (indigenas) made up 40% of the Angolan population. They were distant from the urban centres of Luanda and Leopoldville and less influenced by their activities (Cornwell, 48).
24 It is important to stress that although these three political associations emerged as a response to colonialization, that underlying ethnic and social grievances were present. The three movements politicised these grievances which gradually contributed to a greater division between them. Furthermore, the movements sowed the seeds for new grievances based on ethnicity and political rivalry.
The growing nationalism and economic recession led to the 1961 uprisings in Angola. Angolans attacked a prison in Luanda and killed Portuguese policemen. The Portuguese took repressive measures, such as executing civilians, to regain control over Angola. This did not however put a hold on the uprisings as they spread quickly across the country. The associations’ activities also became more organized in the 1960s and the FNLA in collaboration with the PDA created a provisional government in Leopoldville, the Revolutionary Government of Angola in Exile (GRAE). As a result of Roberto living in Leopoldville and as just discussed as a result of the Angolan refugee population, it enabled the FNLA to base its headquarters in Leopoldville. Locating the GRAE was strategic because it was more favourable for insurgency activities. Furthermore the Organisation for African Unity (OAU) recognized GRAE in 1963 as the official Angolan government. To be able to undertake insurgency activities, GRAE received training and arms from Algeria and funds from the OAU (30 in Collelo). In 1973 GRAE fully merged in the FNLA.
These favourable conditions marginalized MPLA’s efforts. By the mid 1960’s however this changed when the MPLA improved its military operations and became a greater threat to Portugal. This was a result of Chinese and Soviet arms deliveries to the MPLA. Furthermore, the MPLA had political advantages over the FNLA because of its Marxist-Leninist orientation. It needs to be stressed that this happened during the cold war. Therefore the MPLA received support from Cuba and the Soviet Union.
At that stage there were two political movements active, but in 1966 Savimbi returned to Angola and established UNITA. Inspired by Mao who he had met when travelling to China, he concentrated on the black peasants in contrast to the MPLA dominated by mestiços and assimilados. UNITA focused on the anti-colonial and nationalist struggle emphasizing on the peoples from the south-east who have resisted the Portuguese the longest in contrast to the mestiços and asimilados who collaborated with the Portuguese. By the mid 1970s UNITA controlled many of the central and southern provinces which were rich food producing regions.
25 Thus, the GRAE/FNLA was operating from Leopoldville and supported by other African countries. The MPLA which was Marxist orientated was located in the Luanda area and supported by the Soviet Union and Cuba. UNITA inspired by Maoism focused on the black peasants in the southern provinces received limited support from China.
Ironically, according to the World Bank during the independence struggle the GDP rose at an annual growth rate of almost 7 per cent in Angola (12). As mentioned in the previous section, it was not until the end of the Second World War that economic growth exploded in Angola. As a result of the high coffee prices, coffee had become the main export product (36 per cent of the total value) by 1961. At that stage coffee together with other agricultural goods (such as cotton, sugar and wood) accounted to 56 per cent of total export value. Diamonds only accounted 17 per cent. In 1955 the first oil deposits were discovered and in 1969 the Cabinda Gulf American oil company started to exporting crude oil. This development led to a shift in Angola’s export structure from agricultural goods being Angola’s main export product to raw materials becoming its main export product. In 1973 raw materials (oil, diamonds and iron) accounted 47 per cent of total exports and agricultural goods 33 per cent. As an illustration, table 3.1 shows that after the 1970s the mining sector increased its share in GDP drastically.
26 In 1974 general Spinola together with other officers suddenly overthrew the Lisbon government. The coup d’état brought along decolonization Portugal granted independence to its territories in Africa and thus in Angola. As a result of the triangle struggle between the movements Portugal played an active role in forming an independent Angola including the 3 major parties. In 1975, the MPLA, FNLA and UNITA agreed on a transition period before becoming officially independent. A transitional government consisting members of the 3 major parties was achieved during the Alvord Accord. The coalition formed soon failed as a result of the leaders competing for personal power. Heavy fighting broke out, spread throughout the country and the parties declared war to one another. As the civil war began, Portugal withdrew from the situation.
This section has shown how ethnic grievances became embedded in the political struggle for independence. The Mbundu, Ovimbundu and Bakongo had different experiences with the Portuguese during the colonial era and this pattern of intensified during the independence struggle. Furthermore, ethniticy had divided the country in demographic terms, the Mbundu based in the Luanda area, the Ovimbundu in the south east and the
Despite the independence struggle, the economy grew and by the end of the independence experienced a shift from being an agricultural producer to an oil exporter. Within the context of the cold war, the independence movements were supported by different parties and the role of natural resources had been minimal. The next section will analyze the civil war in three periods and show how natural resources gradually began to play a more prominent role.
4.3 Civil war
Soon after Angola became independent, the civil war broke out. Once the colonial threat disappeared the underlying ethnic and social tensions returned to the foreground. The ethnic and social tensions were present during the colonial regime (mesticos and or asimilados versus indigenas) and during the colonial struggle (MPLA versus FNLA versus UNITA), but during the latter they had a common goal; independence. Once independence was achieved and they realised they had to give up personal power, the parties decided to go to war. The 27 year civil war can be divided into three periods. This is necessary because it will show how the different parties involved changed strategies as a result of the changing domestic and international environment. The first period is the period after independence or the cold war period from 1976-1990. This period particularly points out the relevance of the international environment. It was the Cold War where East and West were in competition.
27 This East-West competition is of importance because foreign parties eventually intervened in Angola choosing sides and maintained the war through offering military and financial assistance. The following period of 1991-1997 marks the end of the Cold War and an end to direct foreign intervention in Angola and a shift from the parties receiving external support to sustaining themselves. It is a period of talking and fighting and where the role of Angola’s natural resources begins to play a more evident role in sustaining the conflict. For the government this meant oil revenues and for UNITA diamonds revenues to stay in control. The last period from 1998-2002 is characterized by an intensification of the war; more civilians were killed compared to the previous the periods. It also shows the diminishing influence of UNITA at t domestic and international level. Domestically, they were forced out several areas and internationally the UN sanctions ended the sale of conflict diamonds.
4.3.1 The post independence and cold war period (1975-1990)
Three factors characterized the first period of the civil war that followed the independence struggle. Firstly, once independence was achieved, political and ethnic tensions returned to the foreground. Secondly it was deeply embedded in the Cold War ideological struggle where East-West rivalry became to dominate African politics. Lastly, the political choices the movements made during this period have been strategic and crucial in prolonging the civil war, particularly after foreign intervention ended. This section will discuss the first period of the Angolan Civil War and analyse the political choices the movements made and how these affected the opportunity structure for the MPLA and UNITA in the long run. Furthermore it will be argued the movements internationalised their domestic struggle and created a framework for foreign intervention which also shaped more favourable opportunities for the MPLA and UNITA to continue the war.
After the independence struggle, the political conflict between the movements began. Although the ideas from the movements varied, they had the same objective and that was capturing the state. The MPLA inspired by the Soviet model soon seized control of Luanda and announced the People’s Republic of Angola. Under President Agostinho Neto it installed a socialist government and shifted the Angolan economy from a market to a command one. As the civil war spread and the domestic production fell, the MPLA opened Angola to western capital and formed alliances with oil companies. Gulf Oil (now Chevron) started operating in the Cabinda region in 1979. As shown in the previous section the country’s as a result of the oil development, Angola’s economy changed from being an agriculture producer
28 to an oil exporter. In the 1980s Angola’s reliance on oil revenues doubled since the independence and according to a study of Ferreira sometimes more than 40% of oil revenues was taken from the budget to finance the war (2006:25). In addition to the MPLA financing its military projects through oil revenues, it also sought militarily aid from the Soviet Union and the Soviet Union became MPLA’s major supplier in arms (Marcum, 1978:253). This resulted in more favourable opportunities to sustain insurgency activities against UNITA. Furthermore as a result of the Soviet Union’s involvement in Angola the MPLA had transformed the domestic political struggle into an ideological one. In addition the MPLA’s long relationship with Cuba resulted in the Cubans sending military personnel. The militarily and financial assistance the MPLA received from the Soviet Union and Cuba led to a shift in the balance of power in favour of the MPLA. By 1975 the MPLA was able to arm a force of 5000-7000 men in comparison with 1973 when they did not have more than 1500 men (ibid). The FNLA portrayed itself as a nationalist movement opposing the MPLA. UNITA believed it represented the largest ethnic group in Angola and compared to the other movements insisted on sticking with the transitional government. This changed after MPLA killed 206 UNITA recruits in 1975 and the FNLA and UNITA formed an alliance because separately they were not military strong enough to fight the MPLA. The same year Washington authorized the CIA to commit $14 million worth of weapons and supplies to the FNLA and UNITA (Weigert, 2011:59). Furthermore, as South Africa feared a Marxist Angola, it set up training camps for the FNLA – UNITA alliance. China indirectly intervened by supplying arms. The military and financial assistance from different parties enabled the movements to conduct military operations against each other. The figures from Marcum (from 1500 to 5000-7000 MPLA armed forces) indicate at the end of the independence struggle neither of the movements had professional fighting machines, but this rapidly had changed after foreign intervention.
The South African involvement ensured a fighting force for UNITA. Furthermore, its interference caused regional tension and several African countries reversed their support towards the MPLA while they initially had opposed the MPLA. The MPLA as a result of Soviet and Cuban support became military dominant and defeated FNLA guerrillas. The US who financially supported UNITA introduced the Clark amendments, which prohibited any form of assistance to Angola. The withdrawal of the US, the Soviet-Cuban intervention and the recognition of the MPLA regime as the official Angolan government of the Organisation of African Unity, strengthened MPLA’s position.
29 With FNLA being defeated, UNITA retreated and sought assistance from South Africa. At that time, the end of the 1970s meant the détente period had come to an end. The end of the détente period meant the end of arms control talks and conventional rearming in Europe (Beer & Gamba, 2000: 73). In response to the Soviet – Cuban involvement increasing in Angola, Reagan repealed the Clark Amendment and the US resumed its military assistance to UNITA. It is reported by 1982 UNITA had armed forces numbering 30 000 men, by 1983 almost 35 000 and by 1987-88 this even doubled (ibid, 76)
By the 1980s UNITA had brought larger areas of south-east Angola under its control and pushed the war further into western and central Angola. UNITA’s move to the north east was strategically crucial because it enabled UNITA insurgents to smuggle supplies across Zaire. Furthermore it is important to stress that the UNITA doctrine was inspired by Maoism which emphasised on self-reliance. This self-reliance strategy was important to UNITA and reveals how they operated. Therefore, in the mid 1980s UNITA attacked diamonds fields in the Cuango valley and Andrada (Nzaji area) because Savimbi did not want to rely only on external support. Unfortunately evidence of diamond revenues is not available in the 1970s and 1980s, but UNITA’s decision to attack certain diamond fields at the end of the 1980s indicates UNITA rebels began to pay closer attention to these significant economic areas. As this research will discuss at a later stage having access to these diamonds fields would become more crucial in the funding of UNITA’s military activities.
As the war intensified, the battle of Mavinga in 1987-88 was a turning point for the foreign parties involved. It was disastrous because in Luanda alone 4,700 people were killed. The Mavinga battle led to a redefinition of the Angolan Civil War and to a decline of foreign intervention. While Moscow reassessed its political options in Angola, Cuba by 1988 had agreed to discuss the withdrawal of Cuban forces (Weigert, 2011:90). In the same year the peace talks between US, Cuba, Angola and South Africa had begun. They agreed on South African withdrawal of troops gradually from Angola, on the withdrawal of Cuban troops and on Namibian independence. The UN was called to verify these commitments. The changing international environment directly affected the military position of UNITA and MPLA. With the disintegration of the Soviet Union, the MPLA lost its major supplier of arms. For UNITA it lost South African withdrawal and US assistance. With the peace talks in place, Savimbi and dos Santos agreed on peace and national reconciliation, a cease fire and on creating a UNITA-MPLA negotiating commission (ibid, 92)
30 In conclusion the first period of the Angolan civil war was deeply embedded in the cold war ideological struggle and Angola had become a cold war battlefield. Through financial assistance, training and supplies, the foreign parties sustained the war. By the end of 1980s the movement had professional fighting machines. Even though during this period foreign intervention largely shaped the opportunity structure for the movements, the political decisions made by MPLA and UNITA leaders should not be underestimated. The decicion of dos Santos asking oil companies to manage Angola’s oil production drastically changed the structure of the economy. Becoming an oil exporter in the late 1980s, it enabled the government to finance the war through oil revenues in the long run. Furthermore, UNITA’s move to the north east of Angola enabled UNITA to smuggle supplies and its focus on diamond fields affected its opportunity structure to prolong the war, particularly during the second period of the Angolan civil war, which will be discussed in the next section. With the demise of the Cold War and an end to foreign intervention the following period is more dominated by economic imperatives.
4.3.2 The Post cold war period; talking and fighting (1991-1998)
This section will discuss the post cold war period which was different in several manners. Firstly, the decline of foreign assistance changed the focus of the MPLA and UNITA more on economic imperatives, particularly on oil and diamond revenues. Furthermore, it was different from the cold war period because the tactics applied were very different and more devastating than during the Cold War period. Global Witness reports an estimated 300,000 people died as a result of the fighting between the period of 1992-1994 (1998:7).
In 1991 during the peace talks, the civil war seemed to have ended at the negotiating table. The aura of peace led the formal diamond sector to flourish and to further development, particularly in the lower Cuango valley which had been captured by UNITA rebels in the late 1980s. According to Reuters, by 1992 Endiama, the diamond state company, exported US $250 million in diamonds (2003). In addition the government during this period legalised the sale of rough diamonds. This legislation encouraged illicit mining and by 1992 US $ 5-6 million was smuggled (Reuters, 1991) Then in 1992 Savimbi rejected the elections won by Dos Santos and returned to war. Without US and and South African assistance, UNITA was forced to develop other strategies to survive. This led to UNITA overrunning the diamonds fields in the Lundas and allowed UNITA to survive after the cold war. Savimbi’s move to capture the diamonds fields in the Lundas became crucial in financing UNITA military
