Civilians in the borderlands
A case study of rebel organization in the war in Donbas
Tugba Gunes
Master’s thesis, Department of Political Science, Faculty of Social Sciences
UNIVERSITY OF OSLO
Fall 2017Word count: 40 583
Civilians in the borderlands
A case study on the empirical applicability of Jeremy Weinstein’s (2006) theory on external patronage, rebel organization and indiscriminate violence against civilians. Tested through pattern matching of theory on the pro-Russian rebel movement in Donbas for the period between March 2014- April 2015.
Tugba Gunes
© Tugba Gunes 2017
Civilians in the borderlands Tugba Gunes
http://www.duo.uio.no/
Press: Reprosentralen, Universitetet i Oslo
Abstract
Russia has since the disintegration of the Soviet Union supported numerous separatist movements in its near abroad. Yet, no other rebel movement in Russia’s near abroad receives, or has received, as much Russian funding and weapons as the pro-Russian rebels in the war in Donbas in eastern Ukraine. At the same time, this is the conflict in Russia’s near abroad with the highest number of civilian deaths, with indiscriminate attacks accounting for close to 90 per cent of all civilian deaths (OHCHR, 2016, p. 11). This thesis seeks to shed light on the relationship between funders, rebels and violence through opening the “black box” of rebel organization in Donbas. It will do so by testing theoretical predictions outlined by Jeremy Weinstein in his book Inside rebellion (Weinstein, 2006).
Weinstein (2006) studies the ethnography of rebel organization in rebel movements in Uganda, Mozambique and Peru, and finds support for the initial endowments in rebel groups restraining the universe of organizational strategies rebel leaders can choose to employ.
Based on these restraints, Weinstein (2006, p. 7-16) formulates two paths of expected organizational strategies for rebel groups, wherein the first induces indiscriminate violence, and the second induces capacity for selective violence against civilians.
Through examining the degree of (non) congruence of empirical observations to Weinstein’s theoretical predictions, I find that the case of Donbas provides largely corroborating evidence for Weinstein’s hypothesis (H1):
H1: External patronage for rebel movements induces favourable organizational conditions for indiscriminate violence against civilians
The thesis covers the period of war from March 2014 to April 2015. Its reach applies to the whole territory of rebel-held Donbas, the line of contact and to the zones of conflict
controlled by the Ukrainian government. Through the analytical technique of pattern matching, the thesis finds that Russian external patronage for the pro-Russian rebel
movement induces favourable organizational conditions for indiscriminate violence against civilians. This is because the pro-Russian rebel movement employs largely an opportunistic recruitment and control strategy, which attracts few high-quality recruits and provides rebel leaders with few tools of exerting control and discipline over unit commanders and
combatants.
Acknowledgements
This master’s thesis is the result of a process in which I have received excellent support from my family, friends, colleagues and supervisors. I am grateful to many. My utmost gratitude goes to my two supervisors, Tore Nyhamar and Tor Bukkvoll. Your advice and perseverance in helping me move forward in the face of challenges and setbacks has been truly remarkable.
Thus, all errors in this thesis are mine alone.
Tugba Gunes
Warsaw, October 2017
Maps
Map 1 (Sakwa, 2014): The etymology of the name “Ukraine” is widely believed to mean
“borderlands” (Yekelchyk, 2007). This map depicts the regions (oblasts) of Ukraine. The Crimean Autonomous Republic is controlled by The Russian Federation, but is
internationally recognized as part of Ukraine.
Map 2 (Sakwa, 2014): Map of Luhansk and Donetsk regions (oblasts). The “Donbas” in Ukrainian and “Donbass” in Russian is an abbreviation for “the Donetsk [River] Coal Basin”.
The most common definition of the Donbas refers to Luhansk and Donetsk regions (Yekelchyk, 2015, p. 134). In map 2, the rebel-held areas are coloured in dark grey. It is nonetheless important to note that the dark grey area which represents the rebel-held areas has not remained fixed for the period of 2014-2015.
Table of content
1 Introduction ... 1
1.1 Findings ... 7
1.2 The relevance of the research ... 8
1.3 Overview ... 9
2 Research methodology ... 10
2.1 Case-specific delimitations ... 10
2.2 Case study ... 11
2.3 Why case study research design? ... 11
2.4 Analytical technique- Pattern matching ... 14
2.5 Research method- Content analysis ... 15
2.6 Selection criteria and selection considerations ... 16
3 Theoretical framework ... 18
3.1 Inside rebellion ... 18
3.2 The “black box” of rebel organization ... 19
3.3 Endowments, recruitment and control ... 20
3.4 Paramilitaries in conflicts ... 23
3.5 Defining the scope of the thesis ... 24
4 General background to the war in Donbas ... 26
4.1 A brief outline of main events in Donbas ... 26
4.2 How is the rebel movement structured? ... 27
4.3 Battalions, commanders and leadership according to geography ... 28
4.3.1 Donetsk city and Donetsk oblast ... 28
4.3.2 Donetsk airport ... 29
4.3.3 Sloviansk ... 30
4.3.4 Ilovaisk, Horlivka, Debaltseve ... 30
4.3.5 Donetsk, Makiivka and southern parts of Donetsk oblast ... 30
4.3.6 Luhansk Oblast ... 30
5 Initial endowments ... 32
5.1 Economic endowments- Russian external patronage ... 33
5.1.1 Russian funding ... 36
5.2 Natural resources ... 37
5.3 Social endowments ... 37
6 The structure of rebel organization ... 39
6.1 Recruitment ... 39
6.1.1 How can rebel leaders reduce information asymmetries? ... 41
6.1.2 Predictions on rebel recruitment ... 42
6.2 Recruitment in practice ... 42
6.2.1 How were recruits recruited? ... 43
6.2.2 Propaganda as a recruitment strategy ... 44
6.2.3 Promises of monetary benefits as a recruitment strategy ... 47
6.2.4 What types of rebels did the recruitment strategies attract? ... 49
6.2.5 Information asymmetry ... 55
6.3 Control ... 58
6.3.1 How can rebel leaders create an environment of cooperation among combatants
and commanders? ... 60
6.4 Control in practice ... 60
6.4.1 Have rebel leaders been able to create an environment of cooperation among combatants and commanders? ... 61
6.4.2 Disobedience, defection and operational mistakes ... 62
7 Indiscriminate violence ... 68
7.1 What is indiscriminate violence against civilians? ... 69
7.2 Civilian killings and violence against civilians in Donbas ... 69
7.2.1 Violence against pro-Ukrainian unity protest marches ... 70
7.2.2 Abuses against civilians ... 71
7.2.3 Shelling of populated areas ... 72
7.2.4 Unexploded bombs and shells, and mines ... 73
8 Analysis- Inside the “black box” of rebel organization ... 75
8.1 Russian external patronage, rebel organization and violence ... 75
8.1.1 Russian external patronage and recruitment strategy ... 76
8.1.2 Recruitment strategy and control strategy ... 77
8.1.3 Recruitment and control strategies, and favourable conditions for indiscriminate violence ... 78
9 Discussion and concluding remarks ... 81
9.1 Omitted variables ... 81
9.2 Validity of the findings ... 82
9.3 What do my findings suggest? ... 84
Sources ... 87
Attachment ... 94
List of figures
Figure 1: Weinstein’s (2006) theory on the relationship between endowments, recruits, control, and employment of violence……….…...….….4 Figure 2: The challenges of recruitment and control.………....….…….….…..6 Figure 3: Gering’s (2006) four-point list on trade-offs in research goals……..………12 Figure 4: The structure of the rebel movement in Donbas………27-28 Figure 5: Empirical findings on external patronage, recruitment strategies, control strategies, and the profile of recruits in the war in Donbas………54 Figure 6: The structure of the rebel movement in Donbas and its organizational features...67
1 Introduction
Since the beginning of the 1990s, Russia has either directly supported or in other ways contributed to ethnic conflicts in four breakaway regions in eastern Europe and
Transcaucasia; in Nagorno-Karabakh, in south-western Azerbaijan; in Transdnistria, a self- declared autonomous republic in Moldova situated between the Dniester River and the Ukrainian border; in Abkhazia, a strip of land on Georgia’s north-western flank; and in South Ossetia, a small territory located north of the Georgian capital Tbilisi (Mankoff, 2014;
Tolstrup, 2009).
Many minority groups which had previously regarded the Soviet Union as their protector against larger ethnic groups, protested in the 1990s against the nation-building projects in the newly independent states. Moldova’s independence in 1991 is one such example. The
disintegration of the Soviet Union sparked tension in the eastern region of Transdnistria, which preferred to declare independence from Moldova or become a part of Russia. In 1992 when fighting between the Moldavian government and the de-facto authorities in
Transdnistria started, Russia’s 14th Army, already stationed in Transdnistria, provided support to the Transdnistrian separatists (Mankoff, 2014, p. 63). The European Court of Human Rights concluded in 2004 that the Transdnistrian Republic “remained under the effective authority, or at the very least under the decisive influence, of Russia, and in any event it survived by virtue of the military, economic, financial and political support Russia gave it” (Popescu, 2010, p. 40). Similar to Russia’s involvement in Transdnistria, the Soviet and Russian military also intervened on the side of the separatists in Georgia, first in South Ossetia in 1990, and thereafter in Abkhazia in 1992. After the cease-fires in 1992 in South Ossetia and 1994 in Abkhazia, Russian troops remained on the territories as peacekeepers. In 2008, when Georgia’s military responded with the use of military force to separatist activities by South Ossetia, Russia was again back to militarily supporting the separatists in South Ossetia (Asmus, 2010;19-52). While the case of the conflict in Nagorno-Karabakh pertains different, in that Moscow does not directly support the entity or the separatists, it nonetheless indirectly supports the Republic of Nagorno-Karabakh, by supporting Armenia, which is a party to the conflict (Panossian 2010).
All these conflicts have experienced Russian meddling and support for separatists. One explanation is that Russia regards these territories as a part of its sphere of influence- described by Russian ministers and state officials as its near abroad (in Russian: blizhneye zarubezhye). This is a term for the territories of the Soviet successor states (Moss, 2004, p.
566). Mankoff (2014, p. 67) argues that the idea behind it, namely that former Soviet states in Eastern Europe, Transcaucasia and Central Asia are not fully sovereign and that Moscow can demand special rights to assert influence in these areas, still resonates amongst Russian elites.
Such attitudes can partly explain repeated Russian involvement in support for separatist movements in Azerbaijan, Georgia and Moldova (Manea, 2015).
One of Russia´s modus operandi in conflicts in its near abroad is to support separatist groups; it remains one of the strongest, bluntest and most frequently used tools at Russia’s disposal. Nevertheless, while Russia has deployed its regular armed forces and financed separatist movements in Georgia and Moldova, it is the ongoing war in Donbas (Donetsk and Luhansk regions) in eastern Ukraine that has the highest numbers of civilian deaths and displaced people.
Since the start of the war in Donbas in the end of March 2014, from a population of 4.8 million, approximately 2500 civilians have been killed, 7000 to 9000 are inflicted with conflict-related physical injuries, about a million have fled to Russia, and another 1.8 million people have fled to other parts of Ukraine (OHCHR, 2017, p. 12). Of all conflicts in Russia’s near abroad, the war in Donbas is with its three years of battle 1.) the conflict with deepest Russian involvement in financing rebel fighters and the “state” budgets in the People’s Republics of Donetsk and Luhansk, and in recruiting and training volunteers and mercenaries (Röpcke, 2016). It is also 2.) the conflict with the highest number of civilian deaths, with up to 90 per cent of civilian deaths caused by shelling of populated areas with mortars, canons, howitzers, tanks and multiple launch rocket systems, mostly in indiscriminate attacks (OHCHR, 2016, p. 11). This thesis seeks to shed light on these conflict-specific characteristics. More specifically, in the context of the near aboard, it asks why
indiscriminate violence has been particularly high in the war in Donbas? Can high levels of Russian funding to the pro-Russian rebel movement to some extent explain the high levels of indiscriminate violence against civilians? The thesis approaches this question through
opening the theoretical “black box” of rebel organization, as suggested by Jeremy Weinstein
(Weinstein, 2006), and testing his theoretical propositions on a new empirical case (see figure 1)
Jeremy Weinstein (Weinstein, 2006) studies in his book Inside rebellion the ethnography of rebel organizations in conflicts in Uganda, Mozambique and Peru and finds that rebel organizations that emerge in environments with external support of an outside patron will tend to commit high levels of indiscriminate violence against civilians. Rebel organizations that arise in resource-poor contexts on the other hand employ violence selectively and strategically.
Figure 1 is my own adaptation from the book Inside rebellion. It illustrates expected outcomes upon endowment of economic endowment and social endowment. Rebel groups that arise in resource-poor contexts, and are endowed with social endowments, are expected to follow an activist organization strategy, while rebel groups with external patronage are expected to rely on an opportunistic organization strategy (see figure 2.). The former organizational strategy will, according to Weinstein (2006), be able to ensure targeted and selective violence, because it attracts a pool of high-quality recruits and endow rebel leaders with the capacity to control their unit commanders and combatants. The latter, however, will tend to attract low-quality recruits and restrain rebel leader’s capacity for control, therein inducing favourable conditions for indiscriminate use of violence.
Figure 1. Weinstein’s (2006) theory on the relationship between endowments, recruits, control, and employment of violence
The case of Donbas exhibits both high levels of Russian financing (external patronage) for the pro-Russian rebel movement and high levels of indiscriminate violence against civilians.
Nonetheless, prevalence of high levels of Russian financing and high levels of indiscriminate violence does not prove causality. This thesis seeks therefore to test Weinstein’s theory, on the causality between external patronage, organizational strategies and indiscriminate violence, on the case of the pro-Russian rebel movement in Donbas. This will be done by operationalizing Weinstein’s theory in terms of propositions to be matched with empirical
data. The pattern that constitutes the theory will be evaluated through the analytical technique of pattern matching.
The aim of this thesis will therefore be to evaluate whether Weinstein’s theory, as outlined in figure 1, can be applied as an explanatory theory for the high levels of indiscriminate
violence against civilians in the Donbas. Moreover, also in the face of empirical nuances to the original theory, the aim will be to narrow or extend the scope conditions of the original theory.
As the Donbas case is endowed with external patronage, the expected hypothesis drawn from Weinstein’s theory (2006), will be the following:
H1 X: External patronage Y: favourable organizational conditions for indiscriminate violence against civilians External patronage for rebel movements induces favourable organizational conditions for indiscriminate violence against civilians
In order to theoretically operationalize how X is expected to lead to Y, we need to, in the words of Gerring (2010, p. 1500): “get inside the box of causation”. Weinstein (2006) argues that inside the box of causation, the so-called “black box” of rebel movements, hides
organizational challenges (see figure 1 and figure 2) that either creates favourable or unfavourable organizational conditions for indiscriminate violence against civilians. In his own words, Weinstein emphasises that “understanding differences in the strategies pursued by groups require an examination of the organizations themselves- their membership, their structure, and the ties that hold them together” because “recognition of the diversity of membership and the differing structures of rebel groups offers significant insight into the mechanisms through which indiscriminate violence (…) are observed in conflict” (Weinstein, 2006, p. 258-259). Here Weinstein underlines the importance of rebel organization.
Weinstein proposes that initial endowment of either economic or social endowments plays a part in the organizational capacities of a rebel movement, because it restrains the universe of organizational strategies rebel leaders can employ, and thus the organization’s capacity for targeted and selective violence against combatants.
Weinstein (2006, p. 7-16) finds that the initial endowments of rebel movements constrain rebel organization and shape the violence they employ against civilians. He concludes that high levels of indiscriminate violence are committed by rebel movements that are unable to police defection within their ranks, because of the constraints presence of economic
endowments, and absence of social endowments, imposes on strategies towards challenges of 1.) recruitment and 2.) control (Weinstein 2006, p. 12, 50-53, see figure 2 for detailed
summary of propositions on rebel recruitment and control).
Figure 2. Illustration of the challenges of recruitment and control
In order to evaluate Weinstein’s explanatory theory, operationalized in figure 1, this thesis will first account for the presence of external patronage in the case of Donbas. It will then zoom into the organization of the rebel movement. Lastly, it will detail the level of violence employed by the pro-Russian rebel movement against civilians. Of interest here is squarely the congruence/ non-congruence of theoretical predictions to empirical observations. As the rebel movement receives assistance from Russia (economic endowments through external patronage), for Weinstein’s theory to be true for the case of Donbas, we would expect empirical observations to be in coherence with the opportunistic strategy. This strategy is expected to induce favourable organizational conditions for indiscriminate violence against civilians.
Findings of congruence/ non-congruence will contribute either to strengthen or weaken Weinstein’s theory, and more specifically the theoretical causal hypothesis, which expects Russian support for the pro-Russian rebel movement in Donbas to induce favourable organizational conditions for indiscriminate violence against civilians.
1.1 Findings
This thesis finds that there is largely congruence of empirical observations found in the pro- Russian rebel movement in Donbas to expected theoretical predictions on organizational strategies. The organizational strategy is mostly opportunistic and follows the causal predictions for rebel movements with external patronage.
One nuance to Weinstein theory is the finding that external patronage is not a temporally fixed one-dimensional phenomenon. I find Russian external patronage taking a pragmatic stance to aiding the pro- Russian rebel movement. First, offering leadership, training and supplies of inferior weapons, then providing salaries and better equipment, later deploying brigades from the Russian regular armed forces, and finally financing the budgets of the Luhansk People’s Republic and Donetsk People’s Republic.
Furthermore, I find that variation in assistance shapes the recruitment strategies rebel leaders choose to employ. While, the pro- Russian rebel movement has been endowed with external patronage, it chose nonetheless an activist recruitment strategy in the first months of the war.
When that largely failed, the external patron started providing salaries to prospective recruits, and the recruitment strategy changed to an opportunistic recruitment strategy. I find that rebel leaders were vocal about their demands for more assistance after the activist
recruitment strategy failed, and were thereafter successful in activating the external patron to provide more assistance. Thus, I detect a nuance to Weinstein’s theory. I conclude that in the case of Donbas, the character and level of external assistance, not merely the binary
distinction of being endowed or not endowed with external patronage, shape rebel leaders’
choice of recruitment strategy.
Both the activist recruitment strategy and the opportunistic recruitment strategy mostly failed in recruiting high-quality recruits to the rebel movement. The former attracted militarily
inexperienced volunteers with limited military value to the rebel movement, and the latter attracted mercenaries with former battle experience and military specialties from the Soviet and Russian armed forces, but disadvantaged, and motivated by salaries and payment. The senior Russian leadership was only able to follow an opportunistic control strategy, meaning that military strategy was controlled from the top, and indiscipline down the chain of
command was tolerated. Only when indiscipline, mistakes and defections were particularly grave, did the senior leadership intervene.
The lack of high-quality recruits and the inability of the senior leadership to exert control over battalions contributed to high levels of defections, mistakes and indiscipline, resulting in indiscriminate violence against civilians. These findings support the hypothesis based on Weinstein’s theory, namely that external patronage to rebel movements induces favourable organizational conditions for indiscriminate violence against civilians.
1.2 The relevance of the research
Opening the “black box” of rebel organization for the case of Donbas is relevant for two main reasons. First because, on a policy level civilians in civil wars matter. Civilians are significant actors for belligerent parties- both as means for reaching political outcomes and as ends for employing military force (Smith, 2008). Understanding the conditions that induce rebel movements to employ violence against civilians is henceforth a relevant aim towards the end aim of protecting civilians against predatory violence in conflicts (Beadle, 2014;
Weinstein, 2006, p. 258).
Second because, on the theoretical level, theory testing in case studies are useful for
evaluating the external validity of the original theory (Yin 2014). This is particularly the case for theories inducted from qualitative research, where small N-samples and case studies are frequently used (LeCompte & Goetz, 1982). Weinstein builds his theory mainly on “thick data” from case studies on rebel movements in Uganda, Mozambique and Peru, which has to my knowledge not been tested on other cases. This, despite being a state of the art
contribution to the civil war literature. Thus, a study of a representative case, such as the case of Donbas, ought to give weight to testing the strength of Weinstein’s explanatory theory.
Although confirmatory or dis-confirmatory evidence in the case of Donbas cannot on its own confirm or disconfirm Weinstein’s hypothesis for the whole population of cases, it can
nonetheless give a better basis for assessing its value in terms of its strengths, weaknesses and limitations.
1.3 Overview
The red line for the outline of this thesis is detailed in figure 1. Thus, in order to orderly test the strength of Jeremy Weinstein’s (2006) theory on the empirical case of Donbas, the thesis will follow each causal step accounted for in figure 1. Each step in the process of theory testing will be circled in red, in the end the summary of findings will be illustrated in chapter 8. In chapter 8, the points of congruence and non-congruence will be summarized through illustrative red circles added to figure 1.
Before testing my theory, I will first in chapter 2 detail the research methodology employed in this thesis. Thereafter in chapter 3, I will outline the theoretical framework of Jeremy Weinstein (2006) as presented in his book Inside rebellion. Chapter 4 will provide a background to the war in Donbas and describes the organizational structure of the rebel movement. Chapter 5 will account for the first step in figure 1, namely the distribution of initial endowments. Chapter 6 moves on to the second and third step in figure 1 and matches the theoretical predictions on rebel organization to empirical findings. Chapter 7 details the final step in figure 1, that is the employment of indiscriminate violence by the pro-Russian rebel movement. Finally, chapter 8 accounts for the degree of congruence between theoretical predictions and empirical findings. Chapter 9 discusses the findings in chapter 8 in light of methodological concerns and provides concluding remarks.
2 Research methodology
In this chapter, I will account for delimitations in conducting my research, my research design, analytical technique and research method. The thesis employs a case study research design, which pattern matches theory to empirical observations collected through content analysis of interviews, reports, diaries and other documentary sources.
2.1 Case-specific delimitations
The pro-Russian rebel movement comprises of a plurality of rebel battalions, many of which with little interaction and coordination with each other, but all profess they fight against the Ukrainian armed forces and the pro-Ukrainian paramilitary groups. Estimates vary, but is likely that 10 000- 20 000 rebel combatants were active in the first year of battle (TASS, 2014). Notwithstanding high numbers of rebel fighters, there exists little reliable open-source data on the organization of the rebel movement in Donbas, primarily because few rebel fighters are willing or able to talk to independent news outlets. The main source of data in this thesis is interviews with former rebel fighters conducted outside the rebel-held territories by independent news outlets (Western, Russian and Ukrainian). These interviews are
conducted in the Russian language. I have therefore translated, transcribed and compiled available in-depth interviews conducted by independent news outlets with former volunteer fighters, mercenaries, journalists and rebel leaders. As these are few in number, the scope of the thesis will also be limited by the availability of data. Nonetheless, as biased interviews with civilians, rebel fighters and rebel leaders are frequently used by the parties in the war (OSCE, 2015b, p. 27), upholding stern selection criteria for the sample of interviews, will here be considered important for the validity of the findings. This concern will be further discussed in section 2.6.
The unit of analysis for this thesis is “the pro- Russian rebel movement” and not any
particular rebel battalion. The time frame is limited to the first year of the war (March 2014- April 2015). This delimitation is chosen, because the recruitment of fighters was at its peak in the first year. Besides, interviews are also more available for this period.
The violence against civilians comprises mostly of indiscriminate shelling. While the thesis seeks to explain indiscriminate violence by the pro-Russian rebel movement, it is important to note that observations made by the Office of the UN High Commissioner for Human Rights (OHCHR) and The Organization for Security and Co-operation in Europe (OSCE) finds that both the Ukrainian armed forces and the pro-Ukrainian paramilitary groups, and the pro-Russian rebels and the Russian regular armed forces, are responsible for indiscriminate attacks against civilians (OHCHR, 2016, OSCE, 2015). I will return to this point in chapter 7.
2.2 Case study
A case can be defined as “a spatially delimited phenomenon (a unit) observed at a single point in time or over some period of time” (Gerring, 2006, p. 19). Each case consists of either one or multiple observations that compiles together to a sample (Gerring, 2006, p. 21). A sample of cases belongs to a population, which is characterized by the equal extension of inferences to all cases within the population. The unit of study, or the case, in this study is the “pro-Russian rebel movement in Donbas”. The sample consists herein of one single case.
The case “the pro-Russian rebel movement in Donbas” can be defined as a representative or typical case, because it is not an extreme or unusual case, but instead exemplifies the broader category of cases (population), where it is itself a member (Gerring 2006).
2.3 Why case study research design?
A case study is a study that “entails detailed and intensive analysis of a single case” (Bryman, 2012, p. 66) for the “purpose of understanding a larger class of similar units” (a population of cases) (Gerring, 2006, p. 37). This thesis employs a case study research design. The main research goal for the thesis is to account for (non) congruence of expected strategies to empirical observations, and therein find corroborating or non-corroborating evidence for hypothesis (H1). Towards this end, I will investigate the single case of the “pro-Russian rebel movement in Donbas”. I will collect and compile unique observations on rebel organization made primarily by former rebels, which may or may not provide corroborating evidence for the causal hypothesis. In case of corroborating evidence, the single case will support the hypothesis (H1) that “external patronage induces favourable organizational conditions for rebel movements’ use of indiscriminate violence against civilians”.
The thesis seeks to strengthen or weaken the external validity of a causa pattern through a deep study of a single case. This research goal matches both methodological goals found in a case study research design and a cross-case study research design. I will therefore follow Gerring’s (2006, p. 37) four-point list on trade-offs in research goals to account for the choice of case study research design.
Gerring’s (ibid) four-point list on trade-offs in research goals covers (1) hypothesis, (2) validity, (3) causal insight and (4) scope of proposition, and we can here detect clear
oppositions in the general tendencies in research goals in the case study research design and the large-N cross- case studies (see figure).
Figure 3. Gering’s (2006, p. 37) four-point list on trade-offs in research goals
Gerring (2007: 39) divides social science research into categories of (1) hypothesis generating research and (2) hypothesis testing research. He makes the argument that case studies in general are better suited for theory generating research, while large- N cross-case studies with greater degree of confidence can test hypotheses (Gerring, 2006, p. 41). A similar argument is propounded by Cavaye (1996). Nevertheless, Gerring (2007) does not completely exclude the viability of theory testing in case studies, he merely notes that while
“case studies are often useful for the purpose of elucidating causal mechanisms”, few general theories “offer the kind of detailed and determinate predictions on within-case variations that would allow one to reject a hypothesis through pattern matching” (2006, p. 42). My research goal is to test predictions on initial endowments and organizational strategies. The hypothesis holds that the resources groups have at their disposal in organizing violence restrain their universe of organizational strategies, inducing either targeted or indiscriminate violence against civilians. This hypothesis (H1) is based on a detailed specification of patterns of causal mechanisms found in Weinstein’s theory on insurgency violence. Detailed predictions are a prerequisite for rejecting or confirming, or weakening or strengthening, of a hypothesis.
In this thesis, the purpose is to test the strength of Weinstein’s theory in elucidating causal mechanisms in the case of Donbas, I therefore consider case study research design fitting for my research goal.
The validity- that is the integrity of the conclusions generated from research (Bryman, 2012, p. 47)- in a study can be distinguished according to two criteria (1) internal validity, covering the sample of study and (2) external validity, applies to the whole population of unstudied cases (Gerring, 2006, p. 43). External validity poses particularly a problem for case study research, because it per definition draws inferences from observations of a single case or a small number of cases, thus creating problems of representativeness between the sample and the theoretically defined population. Weinstein (2006, pp. 53-55) employs a mixed-method, where he draws on case studies of four rebel movements as well as cross-case studies.
Notwithstanding the mixed –methods approach, it is primarily qualitative data collected from over 200 interviews with rebels, rebel leaders and civilians, which form the core of
Weinstein’s inductive evidence (2006, p. 58). Weinstein’s theory will benefit from theory testing on new cases, because its evidence relies mainly on “thick data” from a small-N sample. This thesis will test parts of Weinstein’s theory on the case of the pro-Russian rebel movement. By studying a case from his population of unstudied cases, I hope to contribute to evaluating the external validity of Weinstein’s theory.
The third trade-off concerns insight into causality, either through the study of causal mechanisms or the estimation of causal effects (Gerring, 2006, p. 43). The study of causal mechanisms maps how X leads to Y, while causal effects estimate statistical associations (ibid). Gerring (2006, p. 45) finds preference for the study of causal mechanisms in the case study research design, because case studies may through its intense study of one or few cases, open the box of causality, and therein locate intermediate factors between X and Y. As this thesis seeks to study the role of rebel organization, it will test the intermediate factors between X-external patronage- and Y- favourable organizational conditions for
indiscriminate violence, and thus study causal mechanisms. This thesis follows Gerring (2006), Echstein (1975, p. 80), and Stake (1978, p. 7) in considering the case study framework suitable for the purpose of theory testing. The thesis will test the pattern of causality in Weinstein’s theory through the study of one single case. Such an analytical technique is often called “pattern matching”. The aim of this analytical technique is to determine the degree to which observations from the empirical case correspond with the
original theory. This thesis will examine how well observations, mainly from interviews, fit the theoretical pattern drawn from the original theory.
Several notable figures in the field of pattern-matching (Donald T Campbell & Stanley, 1971;
Thomas D Cook, Campbell, & Day, 1979; Cronbach & Meehl, 1955) have contributed to associating pattern matching with construct validity. Construct validity evaluates the degree to which observations fit their theoretical constructs (Trochim, 1985, p. 575), and is per definition dependent on pattern matching. Henceforth, as this thesis will employ pattern matching, it will at the same time shed light on the construct validity between the theoretical constructs in Weinstein’s theory and empirical observation in Donbas.
The forth, and the final trade-off in research goals, is that between a deep versus broad study of cases. Gerring states that “a proposition with a narrow scope is more conducive to case study analysis than a proposition with a broad purview” (2006, p. 49), herein concluding that deep studies are better suited for the case study research design. The purpose of this thesis is to test theoretical predictions on a case; the chosen case (pro-Russian rebel movement) is here considered to be representative of the phenomenon (rebel movements that employ violence against civilians), a so-called representative case (Gerring, 2006, p. 89). So, while a deep study of a representative case may on its own not provide sufficient evidence for a causal proposition, it may nonetheless contribute to evaluating the strength, weaknesses, boundaries and other dimensions of the theoretical predictions in the context of a case study of a new representative case (Løkke & Sørensen, 2014, p. 67).
2.4 Analytical technique- Pattern matching
Pattern matching is an analytical technique for theory testing with cases. A pattern can be defined as “any arrangement of objects or entities”, where the pattern is non-random in arrangement (Hak & Dul, 2010; Trochim, 1985). The first step in conducting pattern
matching is to establish the content of the expected pattern (a hypothesis). Next, observations must be collected and compiled to an observed pattern. Finally, the theoretically expected pattern is matched with the empirically observed pattern, resulting in either a match, supporting the hypothesis, or not a match, rejecting the hypothesis (Almutairi, Gardner, &
McCarthy, 2014).
Contrary to Gerring (2006), Donald Campbell (1966) argues that single case study research designs provides a solid test of a theory, if the expected pattern can be shown to be true in a case. For the expected pattern to be true, each separate prediction in the pattern will have to be in congruence with the empirical pattern of observations. In alignment with Campbell (1966), Robert Yin (1984)- also a central figure in the study of pattern matching- voices support for the suitability of pattern matching in case study research, and even goes further in describing it as the most desirable analytical strategy in case study research design.
In sum, there is largely agreement in the literature (D.T. Campbell, 1966; T.D. Cook &
Campbell, 1969; Yin, 1984) on the suitability of the case study framework for pattern matching, and on it providing a strong test of existing theory. As the aim of this thesis is to exactly test the strength of a theory through a case study framework, the analytical technique of choice for this study will be pattern matching. In the case of points of non-congruence, the aim will be to account for the case-specific empirical nuances, as to establish the strengths, weaknesses and boundaries of the original theory.
2.5 Research method- Content analysis
Content analysis is used in qualitative research when analysing a wide range of documentary sources, such as letters, diaries, newspapers and more. One main characteristic of this type of data is that it is not produced at the request of the researcher. The task of the researcher is herein confined to collecting and analysing available data (Bryman, 2012, p. 543).
I will employ content analysis in my research on the “pro- Russian rebel movement in Donbas”. I have compiled 7 full-length interviews, in addition to drawing on a TV-reportage and a TV round-table discussion, and supplementary news articles, reports, peer-reviewed articles and diaries. My target group consists of individuals with eyewitness observations related to recruitment and organization in the pro-Russian rebel movement in Donbas. This includes former rebels, rebel leaders, regular Russian conscript soldiers, journalists and civilians. All 7 full-length interviews are in the Russian language, they have therefore been translated and transcribed to English, either by me or the news agencies themselves. I have chosen to not translate and transcribe in full a 1, 5 h long oral interview with the first prime minister of Donetsk People’s Republic, Alexandr Borodai. Only a few parts of this interview touches upon themes related to recruitment and organization, I have therefore chosen to
translate and transcribe only the parts of the interview relevant for this thesis. The sampled interviews vary in length, with an average length of approximately 1000 words.
2.6 Selection criteria and selection considerations
I have been sensitive to two selection criteria proposed by Scott (1990, p. 6) in selecting my sample of interviews. The first being the criteria of authenticity-is the evidence and its origin genuine? The second is that of credibility- is the evidence free of error or distortions?
(Bryman, 2012, p. 545; Scott, 1990).
In considering the first criteria and the second criteria of authenticity and credibility, I have chosen to exclude full-length interviews with rebels conducted inside the rebel-held areas of Donbas. This is because many of these interviews are staged, or at best suffer from major distortions, due to the security situation inside the conflict zone. For this reason, only
interviews conducted outside the rebel-held territories are included in the interview sample.
Both considerations on authenticity and credibility are relevant selection consideration for the case of Donbas. For the evidence and its origin to be genuine, the producers of the evidence must be unbiased in selecting interview objects and in conducting interviews. In Donbas there are several reports of “fake news”, many of which draw on supposed eyewitness accounts on violence (Latgalets, 2015; Feldman, 2014; OSCE, 2015b). This thesis will thus not only exclude interviews from the areas under rebel control, it will also only sample interviews conducted by independent news outlets, acknowledged for their critical and investigative journalism. These include mainly Russian news channel Dozhd, Russian radio channel Radio Svoboda, and Russian newspaper Novaya Gazeta. While there are other critical news agencies that fit the criteria set for this thesis, including for instance the St.
Petersburg-based newspaper Fontanka- few of these news agencies publish full-length interviews with former volunteer fighters and mercenaries. For this reason, these will not be included in the interview sample, but are nonetheless included in other parts of the thesis.
My sample of interviews is small relative to the total number of existing interviews with my target groups. This is because I have chosen to adhere to strict selection criteria intended to reduce errors, distortions, and falsity, and sample solely interviews conducted by independent news agencies outside the rebel-held areas. In so doing, the unintended result is that local rebels are omitted from the interview sample. This is unfortunate, but I have been unable to
find full-length interviews with local rebels conducted outside the rebel-held areas.
Information about local rebels will therefore be drawn from other sources, including peer- reviewed articles, non-local volunteers and mercenaries, interviews with journalists, and diaries.
In sum, my sample of interviews comprises of interviews with foreign (Russian, Armenian, Latvian) volunteers and mercenaries fighting for the pro-Russian rebels and an interview with a British journalist who spent time inside one of the main rebel battalions. In addition to these interviews that are translated and transcribed in full, the thesis will also draw on an interview with Alexandr Borodai, the first prime minster of Donetsk People’s Republic, a press
conference with rebel leader Igor Girkin (nom de guerre: Strelkov), and finally a round table with Donbas-stationed Russian journalists from news agencies in my target group.
All my full-length interviews are conducted by Russian or Ukrainian news agencies, but I also draw on information from Western news agencies such as VICE, the BBC, the Guardian, and reports compiled by human rights and security organizations. Finally, I use two diaries, one by a Ukrainian novelist, and the second by a Russian journalist, for contextual
descriptions on the war and quotes from conversations with civilians, rebels and other journalists. Worth mentioning is that I have had meetings and correspondence with
Norwegian journalists and former colleagues working for the Organization for Security and Co-operation in Europe (OSCE) and the International Committee of the Red Cross (ICRC), all with experience from covering the war in Donbas. These sources have not been directly used in the thesis, but they have nonetheless provided me with useful points of departure in initiating my research.
3 Theoretical framework
This chapter outlines the theoretical framework that will be applied to the case of the pro- Russian rebel movement in the rebel- held territories of Donbas, the line of contact and the zones of conflict controlled by Ukraine. In the first section, I will detail the theoretical framework of Jeremy Weinstein (2006) outlined in his book Inside rebellion. Chapter 6, will be divided into two parts, each part starting with detailed predictions on the constituent parts of the theoretical pattern. I will therefore return to the theoretical framework in chapter 6.
3.1 Inside rebellion
Since 1945 more than 16 million combatants and civilian have perished because of
belligerent fighting in civil wars (Weinstein, 2006, p. 4). While the aggregated death toll in civil wars are high, on closer inspection we can find significant variations in-between cases of civil wars in the spectre of violence perpetrated against civilians. For instance, “the war in Congo claimed nearly 100, 000 lives directly in battle, while Colombia’s civil war, which lasted for more than four times as long, is responsible for only one-fifth of the killing”
(Weinstein, 2006, p. 5). These variations in the use of violence raise the puzzle Weinstein seeks to explain. If there are significant variations in the employment of violence in civil wars; if some rebel groups kill their victims indiscriminately, while other groups exhibit restraint and discipline; if some rebel groups abuse the civilian population, loot and destroy their property, while others seek cooperative relations with civilians and engage in the process of political change, then how can these divergent patterns of violence be explained?
Weinstein’s theory accounts for the strategies pursued by rebel leaders in civil wars to explain varying patterns of insurgency violence. To explain the variations in patterns of violence employed by rebel groups in civil wars, Weinstein turns to causal mechanisms caused by the initial conditions rebel leaders face in organizing rebellion.
Starting from an original position, rebel leaders are confronted with factors that either raise or lower their barriers to organizing armed rebellion (Weinstein, 2006, p. 7). The argument is as follows; rebel groups with abundant material resources, either from natural resource
extraction, or from external support of an outside patron, will tend to commit high levels of indiscriminate violence, whilst groups without these material supplies, will tend to directly
depend on the civilian population, and in consequence perpetrate far less abuses against civilians and use violence selectively and strategically (ibid).
3.2 The “black box” of rebel organization
Rebel leaders face several challenges in organizing rebel groups capable of challenging regular armed forces. The main challenge lies in acquiring capital to finance military campaigns and participation of recruits. To face this challenge, rebel leaders can draw on what Weinstein terms initial endowments. These initial endowments can be categorized into two sub-group: social endowments and economic endowments.
Economic endowments refer to resources used to finance rebel groups. These sources of funding can either be supplied by extraction of natural resources, such as minerals, timber, and gemstones, or by donations from external supporters, hereunder supporters from a diaspora or an external patron (Weinstein, 2006, p. 47). According to Weinstein, both these forms of funding are a function of geography, in that natural resources are fixed to territory, and an external patron’s willingness to support rebellion often is a product of external factors, outside the control of the rebel groups (Weinstein, 2006, p. 50). Another common
characteristic is that neither requires voluntary contributions from civilians to initiate and sustain rebellion (Weinstein, 2006, p. 52). Rebel groups with economic endowments are therefore in theory not dependent on civilian support to sustain continued armed rebellion against the government.
Social endowments refer to “distinct identities and dense inter-personal network that can be mobilized for collective action” (Weinstein, 2006, p. 48). Weinstein draws on the work of Robert Putnam (1994), and finds similarities in the role of social capital- networks of civic engagement and norms of reciprocity in a population- in creating efficient institutions, and the common beliefs and norms deriving from ethnicity, ideology, kinship for recruiting and controlling recruits in rebel groups. Like economic endowments, also social endowments are largely considered to be fixed, because although the geographic distribution of ethnic, religious and cultural groups can evolve, it requires time and investment, and is difficult to alter (Weinstein, 2006, p. 50). Weinstein finds that material resources can be mobilized externally or generated from striking bargains with non-combatants. These different environments of endowments shape according to Weinstein the distinct strategies rebel
leaders can employ to solve organizational challenges. Examining how rebel leaders choose to solve organizational challenges, thus requires recognition of the resource environment of rebel groups.
Figure 1. Weinstein’s (2006) theory on the relationship between endowments, recruitment, control, and employment of violence
According to Weinstein’s theory, the distribution of economic and social endowments in rebel groups are decisive for the paths rebel groups pursue, and their employment of violence against civilians (see figure 1). The next section will detail the restraints induced by the initial endowments on rebel organization.
3.3 Endowments, recruitment and control
In Weinstein’s (2006) book Inside rebellion, a rebel group’s distribution of endowments will have consequences for the recruitment-process, the composition of the membership pool, internal control, and finally the violence employed against civilians. Departing from the distinction in the civil war literature on “greed and grievance”-motivated participants in
rebellion (Collier & Hoeffler, 2004), Weinstein finds that the distribution of endowments creates divergent incentives for participation in rebellion, and has consequences for the composition of rebel fighters in rebel groups and the violence they employ against civilians (Weinstein, 2006, p. 259).
Weinstein’s starting point is that all rebel groups are restrained by their endowments. From this original position, rebel leaders need to motivate potential recruits to participate in high- risk collective action against a government (Weinstein, 2006, p. 8). Calculating that potential recruits regard participation in terms of costs and benefits, rebel leaders will need to offer incentives to mobilize participation. Rebel groups with economic endowments, either from natural resources or external patronage, are able to translate economic endowments into economic incentives for participation, by providing salaries, equipment, ammunition and more. Here lies the distinguishable effect of endowments on rebel leaders’ recruitment of prospective recruits. Whereas rebel leaders with economic endowments are able to offer material incentives to recruits, rebel groups with little resources must develop other
alternative strategies to attract recruits. Instead of offering instant benefits for participation, rebel leaders can only offer promises of future individual and collective benefits that are to be reaped when the group-goals are attained. For these promises to be perceived to be credible, rebel leaders will have to draw on social endowments that “tie them to potential followers by means of ethnic, religious, or ideological ties (…) and activate norms and expectations that promote or reinforce cooperation” (Weinstein, 2006, p. 9). As the initial distribution of endowments restrains the incentives rebel leaders can offer, and so too the recruitment strategies rebel leaders can pursue, it will in the next phase determine the composition of rebels in rebel groups. Different initial endowments, attracts different groups of potential recruits, depending on their perceived costs and benefits of participation.
Rebel groups with economic endowments, where individuals can expect immediate reward for participation, tends to attract consumers, partaking in opportunistic rebellions. In environments where participation is filled with risks, and short-term rewards are unlikely, rebel groups will tend to attract only the most committed investors, partaking in activist rebellions (Weinstein, 2006, pp. 9,12).
The composition of consumers and investors in rebel groups will according to Weinstein affects the internal organization of rebel groups and their civilian-belligerent strategies in
war. Weinstein highlights two choices facing rebel leaders once the group is operational;
first, how to ensure members follow orders in conducting rebel-activity, and second, how to extract resources from the civilian population, without disrupting their base of support (Weinstein, 2006, p. 10). The options available to rebel leaders are restrained by the distribution of initial endowments and the profile of recruits these endowments are able to attract. Activist movements with investors can draw on established norms and networks embedded in social endowments to ensure internal discipline, whilst opportunistic rebellions with consumers, lacking in equivalent social relations of reciprocity, must permit indiscipline not to lose members (Weinstein, 2006, p. 9-13). Rebel leaders’ ability to control their troops is relevant in explaining the character and level of violence against civilians. This is because Weinstein finds that structures of internal control shape the capacities of rebel groups to discipline rebel behaviour (Weinstein, 2006, p. 158-159).
Rebel groups drawing on social endowments can use social and political ties to develop effective organizations and therein obtain the capacity to employ violence selectively. Such control is possible, because rebel leaders have “clear guidelines about how combatants should behave and strong mechanisms for enforcing discipline”, and are therefore better equipped to identify targets and discipline the use of force (Weinstein, 2006, p. 9-13)). In contrast,
opportunistic rebellions tend to be without local ties, having greater difficulty in identifying defectors and more frequently making mistakes. Moreover, with a pool of consumers, and without social and political ties to discipline members, opportunistic groups will seek short- term rewards in war, more frequently leading to looting, destruction of property and
indiscriminate attacks (Weinstein, 2006, p. 7-13).
In sum, the distribution of endowments, the types of recruits these endowments attracts, and therefrom rebel leaders’ ability to instil discipline, are all inter-linked to explaining variations in how rebel groups employ violence. Weinstein’s theoretical framework links the initial distribution of endowments to variations in membership and group structure in order to provide perspectives on variations in rebel violence against civilians.
Opportunistic rebel groups with rebel-consumers tend to employ significant levels of indiscriminate violence against civilians, firstly because their source of funding does not dependent on maintaining cooperative relations with the civilian population, and secondly
motivated by short-term material gain. Activist rebel groups on the other hand, often depend on the civilian population for survival, and sustain internal structures of control embedded in social endowments, making them able to discipline the behaviour of their members and commit to selective use of violence (ibid).
3.4 Paramilitaries in conflicts
Weinstein’s unit of analysis is delimited to “rebel movements in civil wars”. The focus of this thesis is the “pro-Russian rebel movement in Donbas”. The pro-Russian separatist movement can be distinguished from other violent groups in conflicts, through its dependence on an outside government (Russia). In their state of the art article Jentzsch. Kalyvas, and Schubiger (2015) study militias in civil wars and note that there are several terms for such armed
groups; militias, civil defence forces, and paramilitaries. What these groups, nonetheless, have in common is that, while there might be a relationship between a paramilitary group and a government, they can operate both “alongside state security forces and independently of the state” (ibid). This feature is present in the pro-Russian separatist movement. It is a
paramilitary group, which depends on Russia for survival, but at the same time does not always comply with it commands. This relationship can be defined as a principal-agent relationship, and will be further elaborated on in chapter 6.
The presence of a principal-agent relationship between a government and a paramilitary group (here the pro-Russian separatist movement), can pose challenges when it comes to violence against civilians. The principal often cannot observe and control the actions of the agents. Thus, the relationship between the principal government and the local communities where the acts of violence find place, can be both protective and predatory. Predatory violence against civilians is often not a deliberate strategy, but rather a result of a failed principal-agent relationship. This is because paramilitary groups, which engage in irregular warfare, contrary to conventional warfare, often depend not only on their relationship to a government, but also on community support and local knowledge for military successes (Kalyvas, 2006).
Predatory and indiscriminate violence has been widespread in Donbas. So, if the literature suggests that paramilitary groups ought to refrain from acts of indiscriminate violence against civilians in order to maintain communal support, then why has indiscriminate violence been
particularly high in Donbas? This thesis will establish, through the theoretical framework of Weinstein (2006), if Russian assistance and the expected organizational restrains it imposes, can provide an explanation to this question.
3.5 Defining the scope of the thesis
The aim of this thesis is to test congruence of Jeremy Weinstein’s (2006) predictions to empirical observations found in the rebel movement in Donbas. To precisely test these predictions, I find it necessary to structure the thesis around concepts and define these concepts in equivalence to Weinstein’s study (2006).
Civil war will be treated as political violence with two belligerent parties; a state and an organized non-state group that seeks to capture control over the government, a region, or influence the policies of the government through violence. In addition, the killings must amount to at least 1000 people in course of the conflict and on average 100 yearly. Finally, there must be 100 dead on each belligerent side (Weinstein, 2006, p. 16).
Violence will encompass both the character of insurgent attacks (the extent to which groups use force selectively) and its aggregated level (number of killings, abductions, and so on) (Weinstein, 2006, p. 6). In chapter 7, I will focus on the temporal developments in the use of force against civilians and account for the degree of selective and indiscriminate employment of violence.
Weinstein distinguishes violence against civilians into two categories; selective and indiscriminate violence. Violence is selective when it targets “individuals or groups that threaten to undermine a rebel organization” and “has a tactical purpose for the group”
(Weinstein, 2006, p. 18). Violence that is applied selectively gives civilians reason to believe that cooperation can be exchanged for the right to survive. Indiscriminate violence on the other hand, makes “no distinction among potential victims, neither protecting supporters nor punish defectors” (ibid).
Organization will be defined as “the internal characteristics of a movement: its membership, policies, structures, and culture” (Weinstein, 2006, p. 19). Organization is crucial for
Weinstein’s explanation for why some rebel groups commit high levels of indiscriminate violence, and is therefore the prime focus of this thesis.
The task of correctly counting the number of deaths caused by violence of belligerent parties in conflict is not an easy task. Nonetheless, Weinstein sets himself the task of striving for more than counting deaths in civil war (Weinstein, 2006, p. 18). He seeks to capture the dynamics of belligerent-civilian interaction. To this end, he adopts multiple methods, and combines the methods of counting deaths, coding human right violations, and qualitative analysis of in-depth interviews. In this thesis, I will attempt to follow Weinstein’s approach.
First, in chapter 7 I will account for civilian deaths and patterns of violence. I will primarily employ human rights reports to establish the presence of high levels of indiscriminate
violence. Then in chapter 6 and 8, I will present and analyse in-depth interviews and accounts of journalists, civilians, volunteers and mercenaries, to determine whether economic
endowments, and the expected retrains it imposes on rebel organization, can have had consequences for employment of indiscriminate violence against civilians by rebels.
The thesis does not cover all aspects of Weinstein’s theory outlined in his book Inside rebellion. Weinstein presents five challenges in organizing rebel movements (Weinstein, 2006, p. 12). This thesis will test only the first two challenges of rebel organization. The choice of delimitation is made primarily due to the limited scope of this thesis. Weinstein based his study of four rebel movements on field work consisting of over 200 interviews (Weinstein, 2006, p. 58), such a wide scope has not been possible in this thesis, both due to time and space restrains, but also because of the challenge of accessing key rebel leaders, unit commanders and combatants. Besides, the war in Donbas is an ongoing war, which means that there are severe security risks associated with entry to the conflict-zone. Due to these considerations, my data will comprise of interviews and diaries from the war, collected and authored by journalists, writers, and human rights and security organizations.
4 General background to the war in Donbas
In this chapter, I will give a brief outline of the chain of events leading up to the escalation of war in Donbas and the main events that followed it. I will thereafter detail and map the structure of the rebel movement in Donbas, accounting for both the hierarchical structure and the specific geographic distribution of rebel leaders and battalions.
4.1 A brief outline of main events in Donbas
The first military operations in Donetsk and Luhansk regions began in end of March 2014.
By 7 April, Donetsk People’s Republic (DPR) was proclaimed, and soon after on 27 April Luhansk People’s Republic (LPR) followed suit. The war in Donbas ought to be seen in the context of the 2014 Ukrainian Euromaidan and the annexation of the Crimea by the Russian Federation. These two events directly contributed to pro-Russian, or anti-Euromaidan, demonstrations finding place in southern and eastern parts of Ukraine in the first months of 2014- either as a response to widespread Euromaidan demonstration or as support for the Russian annexation of Crimea (BBC 2014c).
In April, the separatists in the two regions declared autonomy to the DPR and LPR. By this time, the pro-Russian demonstrations had already escalated into armed conflict between the pro-Russian separatists and the Ukrainian armed forces. As a response to this escalation, the Ukrainian government initiated in the same month the Anti-Terrorism Operation (ATO) against the pro-Russian rebels in the rebel-held territories of Donetsk and Luhansk regions.
The government counter-offensive (ATO) was largely successful in pushing back the rebels, until a new offensive was initiated by the rebels in August. This rebel offensive was largely made possible by the influx of Russian regular armed forces, Russian financing of
mercenaries and Russian shipments of military equipment. As consequence of the pro- Russian offensive, the parties agreed to convened to discuss conditions of peace. In September the same year, representatives of the Russian Federation, Ukraine and the DPR and the LPR signed the Minsk Protocol, which was an agreement meant to end further acts of war. The agreement collapsed however a few months after in January 2015 (BBC, 2014b;
BBC, 2014c).
4.2 How is the rebel movement structured?
In April and parts of May 2014 the chain of command comprised mainly of local unit
commanders, in various battalions and their combatants, without obvious coordination from a senior leadership (Franchetti, 2014; Dozhd 2014; Isayev, 2015). From end of May however, with the Prime Ministry of Alexandr Borodai, the mostly Russian trained military leaders established themselves as the senior leadership in the two people’s republics (Borodai, 2014;
Dozhd, 2014). This induced the chain of command illustrated below. While the civilian leadership changed after the Minsk Protocol in September 2014 to comprise of local leaders, the mostly Russian senior military leadership nonetheless remained throughout the timeframe of this study (Artem, 2015; Borodai, 2014).
Figure 3. The structure of the rebel movement in Donbas
External patronage- The Russian Federation and agents closely tied to it (Röpcke, 2016;
Wilfried Martens Centre for Europan Studies, 2015)
Senior leadership - Russian trained leaders, mostly (former) officers, colonels and generals, some from the Federal Security Service (FSB), others brought in as “expert consultants”
(Borodai, 2014; Kurkov, 2014, p. 212; Sapozhinkov, 2015; Zhetenov, 2015). These leaders plan military operations for unit commanders to execute (Sapozhinkov, 2015). Till the Minsk Protocol, the front-figures were Alexandr Borodai and Igor Girkin (nom de guerre: Strelkov), whilst after the Minsk Protocol, Alexandr Khodakovsky and Aleksandr Zakharchenko
became the new front-figures.
Unit commanders –There are mostly local commanders in the main battalions (Kenyon, 2014), such as Alexander Khodakovsky in Vostok battalion, Alexander Zakharchenko in Oplot battalion, and Mikhail Tolstykh (nom de guerre: Givi) in Somalia battalion
(Polukhina, 2016; RFE/RL, 2017; Vzglyad, 2014). Foreign commanders mainly lead specific battalions, such as Ataman Nikolai Kozitsyn in The Cossack National Guard (Socor, 2015).