Peace over violence or vice versa? Competition dynamics and conflict behavior

(1)

Peace over violence or vice

versa?

Competition dynamics and conflict behavior

Name: Lorik Rexhepi Student number: 10439722

Lecturer and first reader: Abbey Steele Second reader: Ursula Daxecker Course: Civil War and State Building Wordcount: 8501 (references not included)

(2)

1. Introduction

“It had been evident for many years that Kosovo was a catastrophe waiting to happen” (Judah, 1999). Ethnic tensions between Albanian Kosovars and Serbs had been prevalent for decades, but were intensified after the Kosovo province was stripped from its autonomous status in 1990: thousands of Albanian Kosovars were removed from their public functions (Mulaj, 2008; Bekaj, 2010; IICK, 2000). The domestic political conflict that ensued descended into civil war; in Kosovo, around 12.000 Kosovar Albanians were killed; 120.000 homes were burnt. Moreover, almost 900.000 Albanians were forced to leave Kosovo, while hundreds of thousands were internally displaced (Bekaj, 2010). Prior to the escalation of the conflict, the domestic political conflict was characterized primarily by the nonviolent approach of resistance by the Albanian Kosovars. The Democratic League of Kosovo (LDK) led this peaceful protest, which took the shape of a parallel government and had a widespread support among the Albanian Kosovar constituency (Maliqi, 2001). By remaining pacifist, patient, and approachable, the Albanian Kosovars hoped to attract international attention and support for their civil resistance (Maliqi, 2001). The pacifist resistance, however, did not achieve its goals and lost its magnetism in 1995 (Bekaj, 2010; Kubo, 2010; Maliqi, 2001; IICK, 2000; Mulaj, 2008; Judah, 1999). In 1996, the nonviolent strategy was abandoned and a new faction, the KLA, was publicly exploring the options for armed resistance. In 1998, the KLA had become the dominant force in the resistance and the pacifist movement had perished (Kubo, 2010).

My research question is: What explains the transition from nonviolent to violent resistance in Kosovo? I define Y as ‘the onset and the escalation of violence’.

The scientific relevance of this question lies in the explanation of the onset of violence. Leading theories on violence onset focus primarily on conditions that explain the probability of the occurrence of violence in specific regions (Lawrence, 2010). Offering an explanation of timing of violence will, thus, contribute to the existing literatures. The social relevance is found in humanitarian and economic considerations: was is expensive; it has costs many people their lives and it is always accompanied by economic destruction. Understanding the dynamics of resistance strategies might help policymakers recognize threatening situations and prevent them from escalating.

To answer the research question, I have tested Lawrence’s (2010a) ‘Competitive Violence Theory’ on the case of Kosovo. This theory explains the transition to violence through the concept of

competitive dynamics. This means that co-ethnic factions compete for political relevancy among each other. This has the following observable implications: in the case of a sudden power vacuum, new nationalist factions become active (H1); co-ethnic factions compete with one another in order to gain recognition (H2); factions employ violence to show that they are more involved in the resistance (H3); factions target direct rivals. All four hypotheses are supported by the data.

The remainder of this paper proceeds as follow. The first section will discuss the theoretical arguments in the literature on the onset of nationalist violence. The second section will describe the ‘Competitive Violence Theory’, its causal mechanisms, and the hypotheses that are derived from the theory. In the third section, the case of Kosovo will be examined. Finally, the generalizability of the ‘Competitive Violence Theory’ will be tested when the hypotheses are subjected to an analysis of the turn to violence in Kosovo.

(4)

2. Literature review

The following section will provide an overview of the existing literature on violence onset and the mobilization of a resistance against an established regime. I argue that the arguments they produce fall short of the ability to explain the adoption of violent strategies over nonviolent strategies and the timing of violence. Furthermore, the effectiveness of nonviolent and violent strategies will be

discussed before turning to the main argument of this paper. 2.1 The onset and efficiency of violence

The existing literature on civil war onset has produced a wide spectrum of possible explanations and motivations for violent resistance against an established authority. “The pool of independent variables that have been identified as potential determinants of civil war onset includes the level of economic development, political instability, ethnic heterogeneity, the presence of plentiful natural resources, a history of conflicts, war-prone and undemocratic neighbors, small military

establishments, political regimes that are neither dictatorships nor democracies (‘‘anocracy’’), mountainous terrain, large population, diasporas, and oil production” (Kalyvas, 2007).

One of the most supported findings in quantitative research, regarding this spectrum, is the correlation between poverty and the onset of violence (Lawrence, 2010a ; Fearon & Laitin, 2003; Collier & Hoeffler, 2004). Fearon & Collier (2003) argue that the onset of violence is more likely when conditions favor violent resistance. Poverty, proxied by per capita gross domestic product (GDP), facilitates these conditions; poor states tend to be weak in the sense that they suffer from diminished financial, administrative, and military capabilities. These factors undermine a state’s capacity to prevent violence (Fearon & Laitin, 2003). Collier & Hoeffler (1999; 2004) offer an alternative explanation for the association between poverty (proxied by per capita income) and violent

challenges to the state. The authors have found strong evidence that rebels use violence to acquire wealth.

The poverty determinant, however, suffers from the same lack of explanatory power that the independent variables noted by Kalyvas (2007) suffer from: it cannot account for the actual onset of violence. None of the beforementioned variables, with the exception of ‘political instability’ – if sudden (i.e. state crisis) – are adequate predictors for violence onset since they are stable over time and consequently have no temporal variance. These determinants identify conditions that explain the probability of the occurrence of violence in specific areas, but cannot account for a sudden surge of violence (Lawrence, 2010a).

In this respect, the social movement literature seems more capable of identifying mobilization triggers. This academic stream focuses on political opportunities that arise from social change. Mobilization of state adversaries is dependent on the vulnerability of the political order. Most resistance movements become active after social changes offer the opportunity for confrontational collective action (McAdam, McCarthy & Zald, 1996: 8). The relevance of this viewpoint lies in its ability to account for temporal variation of resistance; social change is a dynamic concept by default; it is a point in time, rather than an occurrence over time. However, literatures on political

opportunities focus on mobilization of movements in general and not on why state adversaries would opt specifically for violent over nonviolent strategies (Lawrence, 2010a).

The dominant viewpoint on the latter, among political scientist, is that that state adversaries opt for violent over nonviolent strategies because they are more effective in realizing political objectives. Moreover, it is widely assumed that the most effective way to achieve desired political goals is through the use of violence. Terrorism would, for example, be particularly effective in coercing governments to make territorial concessions (Stephan & Chenoweth, 2008). Stephan & Chenoweth (2008) & Chenoweth & Stephan (2011) oppose these assumptions: having studied 323 violent and nonviolent resistance campaigns from 1900 to 2006 (all regime change, foreign occupation and

(5)

secessionist movements in this timeframe), the authors find that major nonviolent resistance campaigns have achieved roughly twice as much success than their violent alternative (respectively 53 percent versus 26 percent. “From strikes and protests to sit-ins and boycotts, civil resistance remains the best strategy for social and political change in the face of oppression. Movements that opt for violence often unleash terrible destruction and bloodshed, in both the short and the long term, usually without realizing the goals they set out to achieve” (Chenoweth & Stephan, 2014: 95). Accounting for the success of nonviolent approaches, the authors focus on the advantages nonviolent campaigns have over violent campaigns, regarding mobilization and mass participation: nonviolent action attracts more domestic and international recognition and legitimacy. It therefore encourages more participation and facilitates a broader base (Stephan & Chenoweth, 2008: 9). The strength of nonviolent measures lies in numbers: the costs for targeted governments rise, when more people are mobilized. Acts of violent oppression could prove costly for the regime as well, as violent crackdowns by regimes against nonviolent movements are more likely to backfire (i.e. international repercussions, increased public mobilization, and loyalty shifts) against the regime (idem: 9-12). The increased pressure makes it more likely that the target regime eventually concedes and/or

compromises (Stephan & Chenoweth, 2008: 9; Chenoweth & Stephan, 2014: 97). In the case of violent resistance, governments might easily justify the use of violence against armed separatists, as the repercussions after a counterterrorist or counterinsurgency operations are less severe than after a violent crackdown on a peaceful protest (Stephan & Chenoweth, 2008: 11).

It should be noted that, although nonviolent resistance carries a peaceful connotation, the vast majority of participants view nonviolent resistance as a strategic choice, rather than a principled, moral choice. Nonviolent measures are able to be highly disruptive if employed correctly (Stephan & Chenoweth, 2008: 10). Similar to violent methods, it requires planning, coordination and flexible tactics (Chenoweth & Stephan, 2014: 96).

If nonviolent resistance proves to be more effective and provides a more favorable process and outcome, then why do resistance movements change their strategy in favor of violent measures? The challenge then is identifying the essential opportunity that facilitated the ‘ineffective’ transition to violence in Kosovo.

3. Theory and hypotheses

As stated in the above, the existing literature on violence onset has not been able to account for the timing of the eruption of violence or why resistance groups adopt violent strategies after a nonviolent campaign. Recent literatures suggest that the assumption that conflicts regarding secessions are dyadic in nature has limited our understanding of the developments of conflicts and conflict behavior (Lawrence, 2010a; Cunningham, Bakke & Seymour, 2012; Cunningham, 2013; Pearlman, 2009). The beforementioned theories regard actors as unitary; in other words, they ignore actor fragmentation (Cunningham, Bakke & Seymour, 2012). Adria Lawrence (2010a) and Cunningham, Bakke & Seymour (2012) claim that internal divisions within self-determination movements are essential to our

understanding of conflict behavior. Moreover, Lawrence (2010a) argues that her ‘competitive violence theory’ is able to both establish opportunity and motive and thus account for the timing of violence, as well as the use of violence. The central logic behind this theory is that nationalist violence is the result of a fragmented resistance movement where factions engage in a competition with other co-ethnic factions. Competition among nationalist factions can be incited when the status quo of the movement is unstable. This can be triggered by variety of events (i.e. events that internally divide nationalist movements), such as the death of nationalist leaders or leadership repression (Lawrence, 2010a: 99-101). This paper will argue that in Kosovo the proverbial death of the LDK’s dominant, pacifist ideology served as the main trigger of fragmentation.

(6)

3.1 Dual Contests and Competitive Violence

Conflicts are commonly regarded as dyadic standoffs between self-determination movements and the dominant regime. Self-determination groups, however, often engage in ‘dual contests’ (i.e. versus the state and other co-ethnic factions) dividing them internally. A group is defined as a people with “a shared ethno-nationalist identity” (Cunningham, Bakke & Seymour, 2012: 5). Groups can be subdivided into different factions. Factions claim to represent the same constituent population and are engaged in the same struggle for self-determination. Simultaneously, they have divergent interests beyond self-determination and preferred means to achieve their goals (i.e. nonviolent or violent)(idem: 5-7). The dynamics of conflict behavior are primarily influenced by the latter. As resistance movements (i.e. groups) enter a standoff vis-à-vis the target regime, the individual factions concurrently oppose each other in search of political relevance (idem: 6-7). The former competition (i.e. versus the state) focuses on obtaining public goods that benefit the entire group (i.e. self-determination). The second competition entails a struggle between factions for political power and material goods (i.e. monetary and material gains) that are exclusively beneficial for the individual faction that obtains them. In other words, the spoils of war in the first competition are shared by the group, while exclusively prominent factions are able to stake their claim to private goods at the of the struggle(i.e. material gains, monetary gains and political power)(idem: 6).

The prominence of a faction, and thus their ability to obtain public goods, is determined by its political relevance. The latter refers to the position of the faction “vis-à-vis the state, rival factions, and the population it claims to represent”. A faction is regarded as politically relevant when other actors in the conflict are forced to acknowledge and engage them politically or military (idem, 6-7). Lawrence (2010a: 100) identifies four mechanisms that lead to violence when actors engage in a competition for political relevance. First is the recognition mechanism: to compete with other active parties, a faction needs to be recognized as an prominent nationalist actor. The use of violence is highly visible and a fast way to gain notoriety. This is what Crenshaw (1981: 386) calls “advertisement for the cause”. A second mechanism is outbidding. Actors want to show that they are more involved in the cause against the state. A third mechanism is the use of violence to eliminate local rivals and competitors. A final reason to take up arms is the inability to obtain political relevance through nonviolent strategies. To be visible in the political arena, a nonviolent mobilization requires a mass base. A new player does not have these means but is still able through the employment of violence to carry out highly visible attacks. “For would-be leaders who lack a large base of followers, violence is the only route to claiming an important role in the conflict” (Lawrence, 2010a).

However effective in gaining political relevance short term, violent measures do not guarantee the possibility of maintaining it. Political relevance is, as stated above, dependent on the approval of the constituent population and therefore their strategic preferences. Accordingly, the legitimacy of a faction, and consequently their claim on private goods is directly connected to the popularity they enjoy domestically. Pearlman (2009: 86) suggests that otherwise violent factions would refrain from using violence on the road to secession when the relevant population is optimistic about a peaceful process.

Actors who would want to employ violence might feel inhibited in the shadow of an existing dominant campaign; after the proverbial decapitation of the dominant campaign – and a power vacuum is created - the inhibition drops, clearing the way for their own strategy. Lawrence (2010a) states that “it emboldens other actors who wish to use violence to become the new voice of the movement”. It therefore leads to the following hypotheses1_{based on the mechanisms outlined}

above:

1 Lawrence’s (2010) ‘Competitive Violence Theory’ has implications for the duration of violence as well. The argument suggests that ongoing violence might be observed after the nationalists have achieved their objectives. This, however, will not be tested in this paper, because I do not expect this to occur in the Kosovo case. After the civil war, the UN Security Council passed Resolution 1244, which effectively stated the United Nations Mission in Kosovo (UNMIK) interim administration and provided the legal framework for the NATO-led peacekeeping force KFOR (Bekaj, 2010:45). This leads me to expect that prolonged violence will not occur, which makes H5 irrelevant in this case

(7)

H1. After the emergence of a power vacuum, an increase in the number of ethnic-nationalist factions

should be observed; existing ‘dormant’ parties could use the opportunity to become more involved

and oppose the previously dominant pacifist paradigm.

H2. Struggle for power among local actors should be observed. Specifically, new actors will use violence to gain recognition. However, the self-determination group may contain only one faction that employs violence in search of political relevance.

H3. Actors will employ violence to outbid other actors. H4. Rivals will directly be attacked

4. Data: The case of Kosovo and the adoption of violence

In order to test the beforementioned hypotheses, I use the case of Kosovo and the genesis of the Kosovo Liberation Army (KLA). This is an appropriate case to test the hypotheses, due to its

resemblances with other cases of foreign occupation (as the Albanian Kosovars perceived it) and self-determination; it involves a protracted conflict, a dominant nonviolent resistance movement and a sudden rise of an armed insurgency movement (Bekaj, 2010: 7-8). Furthermore, Lawrence (2010a) argues that leadership repression was the key actor fragmentation trigger. Leadership repression did not occur in Kosovo; instead, I argue that, in Kosovo, similar dynamics resulted from the fact that the effectiveness of the peaceful resistance was disproven in a ‘dramatic’ way. This could contribute to identifying alternatives triggers for violence onset. Moreover, the significant involvement of the international community (i.e. UN, EU & NATO), and the consequent media and scholarly attention for the conflict have brought forth a rich literature and data (IICK, 2000; Maliqi, 2001; Bekaj, 2010; Kebo, 2010; Judah, 1999; Fulton, 2010; Mulaj, 2008; Boyle, 2014). This eases the process of tracing the conflict and establishing an analysis in which nationalist bias is avoided as much as possible. To test the hypothesis, I collected data on the presence of a power vacuum (X) and its effect on possible competitive dynamics between co-ethnic factions. I expect this to result in the emergence/activation of at least one faction that had not been prominent thus far (H1); competition between factions is observable: new factions employ violence to gain political relevance; the former dominant faction might refrain from violent tactics; factions could attempt to undermine each other by denouncing the opposition’s strategies (H2). This could lead new factions to incorporate violence to show that they are more actively involved in the cause. This results in a growing occurrence of violence. An upward trend can be observed. Furthermore, outbidding occurs as well, if the nonviolent factions adopts alternatives strategies (H3). Finally nationalist actors target other co-ethnic nationalists as to eliminate competitors for political relevance (H4). The variation I exploit is over time. Therefore, collected data on X before and after T, and on H1, H2, H3, and H4 during those periods. X is measured as LDK-popularity. The LDK was the dominant faction in the resistance group and enjoyed absolute legitimacy and political power by the grace of the Albanian Kosovar constituency. A power vacuum emerges when their popular support decreases. The sources for the data on X, H1, H2, H3, and H4 are published articles containing detailed descriptions on the Kosovo case and the ICTY.

4.1 The parallel system and nonviolent resistance

Many onlookers had theorized that if a war in Yugoslavia would occur, it would happen in Kosovo. After Milosevic became Serb President in 1989, he seemed to confirm these thoughts in his speech in

(8)

June 1989, on the 600th_{anniversary of the Battle of Kosovo: “Six centuries later, again, we are in}

battles and quarrels. They are not armed battles although such things cannot be excluded” (IICK, 2000: 40-42). The autonomous status of Kosovo was gradually reversed, until the actual abolishment of autonomy in September 1990 (Mulaj, 2008: 1106). Accompanied by an increase in human rights abuses (i.e. arbitrary arrest, imprisonment without trial, and torture) and discriminatory government policies, thousands of Albanian Kosovars were dismissed from public employment: according to the independent Kosovar Albanian Association of Trades Unions, 115,000 people out of a total 170,000 lost their jobs (IICK, 2000: 42). These ingredients seemed to be the perfect mixture for an ethnic-nationalist uprising, but the ones that expected it to happen were, for the time being, disappointed. As the Albanian Kosovars had fought the Serbs before over the past decades, the decision not to fight in the early ‘90s came as relative surprise. Tim Judah (1999) argues that this decision was “born not of passivity but of shrewd logic”. Prior to the complete abolishment of autonomy, the Kosovo Assembly unilaterally proclaimed themselves to be independent when they passed a new constitution. This proclamation was heavily supported by the Albanian constituency; 99 percent voted in favor (87 percent of voters took part)(IICK, 2000: 43-44). In 1992, elections for a new

republican government and assembly were held in Kosovo. One party dominated these elections: the ‘LDK’ won 96 of the 100 constituent seats and was already backed by 700.000 members prior to the elections, confirming the people’s trust in the LDK (Bekaj, 2010). Consequently, the LDK leader, Ibrahim Rugova, became ‘President of the Republic’ (IICK, 2000: 44; Mulaj, 2008: 1107; Judah, 1999). It was clear that Kosovo did not command the military means to consolidate this independence, or at best, to oppose Serbian rule. Although it went against the traditions of Albanians (i.e. the traditional practice of blood feud), Rugova recognized that a violent resistance was simply not possible and even folly. Now being the leading figure in the Albanian Kosovar movement, he claimed: “The Serbs only wait for a pretext to attack the Albanian population and wipe it out. We believe that it is better to do nothing and stay alive than be massacred” (IICK, 2000).

“And so Kosovo became a weird looking-glass world” (Judah, 1999). Under the leadership of Rugova, Kosovo entered a state of neither war nor peace; as most Albanians were removed from public employment, the LDK set up a parallel state apparatus that governed education, health care, and cultural activities (IICK, 2000: 44-46). This shadow government preached the liberating effects of refusing the use of violence, and introduced “the tradition of endurance” and “culture of solidarity” (Maliqi, 2001). This solidarity manifested itself through the success of the “voluntary” tax system; Albanian Kosovars paid a second 3 percent income tax on voluntary basis next to the tax they owed the Serbian government; diaspora could also contribute; non-compliance was low (IICK, 2000). In the meantime, the Albanian Kosovars tried as much as possible to circumnavigate anything that required interaction with the Serbs, public as well as private (IICK, 2000: 49).

The primary objective of this parallel society was independence for Kosovo. The means to achieve this lay in its very own existence: by remaining pacifist, patient, and most importantly, approachable, the Albanians Kosovars hoped to internationalize the Kosovo question. Rugova acted as the primary spokesperson, as the election in 1992 granted legitimacy to the LDK as the representatives of the Albanian Kosovar people (IICK, 2000: 44; Bekaj, 2010: 13). He hoped to convince the international community (i.e. UN and NATO) to establish a temporary UN protectorate, which would facilitate the move to independence (IICK, 2000). The Kosovo media reported that Western leaders appraised the nonviolent resistance and soon Rugova was portrayed in the public as “America’s chosen man” (Mulaj, 2008: 1107; Maliqi, 2001). The Albanian Kosovar community looked up to him as a god, which symbolized the political power and unrivaled position he had received in the movement (Maliqi, 2001). Rugova, however, has been criticized for not having a clear idea how victory could be achieved in the long term and the inability to employ flexible tactics (Maliqi, 2001). He made maximalist political demands (i.e. full independence), while favoring not only a nonviolent strategy but an

(9)

excessively passive strategy void of any risk (IICK, 2000: 48; Maliqi, 2001).

These public ‘transgressions’ did not go unnoticed in Belgrade; the Serbs denounced Rugova as a separatist, but did not seem eager to intervene (Judah, 1999). As time passed, Rugova’s grip on the Albanian Kosovar community got tighter. The LDK used the national fund and Rugova’s charismatic pull to maintain its dominant position in the shadow government and to create a monopoly on political power, and information (Maliqi, 2012). The LDK controlled the public opinion through the media, which had continually reported on the appraisal the Albanian Kosovars had received from Western states for their strategies and persistence. However, the press never revealed that foreign governments were reluctant to recognize the Kosovar independence (Mulaj, 2008: 1107). Their reluctance had a similar origin to the Serbs’ willingness to uphold the status quo and not intervene. “Belgrade may have calculated rightly that if the pacifist Rugova were removed from office, radicals preaching violence would fill the leadership vacuum” (Judah, 1999). This lead to a situation that Tim Judah would later describe as a “grotesque symbiosis between Belgrade and Prishtina “ because the LDK and Rugova ”effectively pacified the province, which was exactly what the Serbian authorities wanted“ (Maliqi, 2001).

Rugova had a tight control of the LDK structure and imposed his nonviolent strategy with an iron fist in a velvet glove (Kubo, 2010: 1145). He did not only represent the collective will, but also articulated it. Peaceful, passive resistance was to be the only form of resistance. Dissidents were regarded as traitors and were to be dealt with accordingly (Maliqi, 2012). However, as Rugova’s policy became more and more passive, and the prospects seemed dim, unrest crept over the Albanian Kosovars. The double taxation and the psychological burden of constant oppression had taken their toll, but progress seemed to be absent (IICK, 2000). The dominance of the LDK and Rugova had thus far been able to stymie any radical voices, due to the following they had mustered. The situation, however, seemed unsustainable as cracks began to appear in the foundations of the Albanian Kosovars’ loyalty to the nonviolent resistance group (Maliqi, 2001).

4.2 The splintering of the nonviolent resistance

Throughout the early ‘90s, the Albanian Kosovar society underwent a transformation in solidarity. It was an artificial solidarity, since the Albanian society was not homogeneous; even within the LDK there was an ideological split between members, with matching ideas on how to resist the oppression effectively (Maliqi, 2001). However, the Serbian threat, and the need of the Albanian community to keep that threat at bay, made sure that the status quo was treated “gently” by all parties. Acts of public defiance towards the dominant movement were, thus, limited. Dissident movement, however, remained operational under the radar. Maliqi (2001) claims that within the LDK, out of the public sight, there were struggles for control going on between different factions. Most prominently the LPK, which was established in the ‘80s, was planning to transform the nonviolent resistance into a violent one (Maliqi, 2001). In 1992 and 1993 the LPK had already begun to establish a guerilla group during secret meeting in Macedonia and Prishtina (Judah, 1999). Simultaneously, the leaders of the LPK were convinced that an early uprising would be foolish, without the support of the Albanian Kosovar population. Since the LDK represented the entire Albanian Kosovar community, it was thought that the community would be convinced most effectively from within the LDK

organization (Maliqi, 2001). The attempts to gain prominent positions within the shadow government, however, proved to be futile, as Rugova managed to ward off the internal takeover (Maliqi, 2001). The LPK made its final split from the nonviolent resistance and now focused solely on their own means, which was rewarded when the Dayton agreements of 1995 came to pass (Kubo, 2010; Maliqi, 2001).

The 1995 Dayton Peace Accords on Bosnia and Herzegovina proved to be “little short of traumatic” for the Albanian Kosovars (Judah, 1999; Kubo, 2010: 1144; Bekaj, 2010: 13). The entire peaceful

(10)

resistance and its widespread support amongst the Albanian Kosovars relied on its ability to

internationalize the Kosovo question. The Albanian leaders had lobbied extensively to ensure that the US government would solve the Kosovo situation at the Dayton agreements (Kubo, 2010: 1143-1144). However, no mentioned was made of Kosovo when a deal was reached for Bosnia and Kroatia. The Kosovo situation did not demand the same sense of urgency as the latter two regions did; Kosovo was relatively stable compared to Bosnia and Kroatia (Maliqi, 2001; Mulaj, 2008: 1109). The dominant Albanian Kosovars perception of these events was that their nonviolent methods and endurance were not rewarded, as opposed to the violent means that the Serbs had incorporated in their conflicts (Judah, 1999). The European Union states and the UN seemed to have rewarded Milosevic for agreeing to the deal by officially recognizing the Federal Republic of Yugoslavia (i.e. Serbia, Montenegro, and Kosovo). The UN embargo on Yugoslavia was lifted, and Kosovo remained an integral part of Republika Srpska (Kubo, 2010: 1144; Judah, 1999). This political exclusion of Kosovo ‘”confirmed to the Albanian Kosovars in a dramatic and humiliating way that Rugova’s policy of passive resistance had failed” (Kubo, 2010: 1144). The key lesson Albanian Kosovars took from this experience is that ‘violence pays’ and that the attention of the international community could only be demanded through war (Mulaj, 2008: 1109; IICK, 2000: 50). Albanian Kosovars began to radicalize rapidly, and the LPK was eager to fill the ideological vacuum left behind by Rugova’s peaceful policy. LPK/KLA spokesman Bardhyl Mahmuti explains the situation in an interview with Shkelzen Maliqi (2001): “Rugova created the illusion that the international community would resolve the crisis and that independence would come a gift. The LDK had controlled people’s imagination and when Kosova was forgotten in the Dayton agreements, the space was created for different and active politics of resistance. Our room to maneuver opened up”.

4.3 The transition to violence

As noted above, slumbering nationalist factions rose up: most prominently, the LPK, who had never been seriously acknowledged due to their extremist views, were now considered serious actors on the playing field (Judah, 1999). The guerilla group (i.e. KLA) the LPK had set up during the nonviolent period now begun to make their presence known in Kosovo. On 22 April 1996, four coordinated assaults on policemen were carried out (Kubo, 2010: 1142-1143) The attacks were accompanied by a written communiqué, in which responsibility was claimed by the KLA (Bekaj, 2010: ). Although the intensity of the rebellion remained low, there was clear increase in the number of attacks(31 in 1996, 55 in 1997 and 66 in the first two months of 1998)(Kubo, 2010: 1142). The KLA, however, did not receive any tangible recognition for these actions. They were mostly perceived as a phantom organization, which was caused primarily by the position Rugova had taken towards them. The KLA was created primarily to oppose Serbian oppression, but simultaneously challenged the LDK’s hold on power, as well as the legitimacy the LDK enjoyed amongst the Albanian Kosovar population and the international community (Mulaj, 2008: 1109). Rugova therefore denounced each attack, claiming that the KLA was a Serbian construct designed to destabilize the pacifist movement (Judah, 1999).

Furthermore, Kubo (2010: 1145) claims that Rugova used his dominant position in the LDK, as well as his tight grip on national media to prevent Albanian Kosovars from joining the KLA, even though almost everybody in the LDK presidency supported the transition of strategies. This ushered in a long-lasting rivalry between the LDK and the KLA (Judah, 1999). The KLA could not count on the LDK to mobilize the masses, and so they were forced to step out of the shadows to advertise the cause themselves.

The KLA made their first public appearance on 28 November 1997 – the Albanian Flag Day - at the funeral of Halit Geci, a local schoolteacher and KLA activist, which was attended by approximately 20.000 people. Three men in KLA uniform appeared before the crowd and proclaimed their resolve to fight the Serbian forces, liberate Kosovo through armed resistance, and protect the civilians in the

(11)

process (Bekaj, 2010: 20; Mulaj, 2008: 1110). This appearance sent “electric waves” through the Albanian population of Kosovo (Bekaj, 2010: 20). The existence of the KLA was now a fact, rather than a rumor.

However, the KLA still was understaffed, underfunded and underarmed. Furthermore, soliciting popular support still proved to be difficult. Mulaj (2008: 1111) argues that this is due to the fact that “expected net returns to membership of the KLA” were still perceived to be low. The people’s loyalty lies where the highest probability of success lies. In other words, the strength of the KLA and the probability of success were dependent on the amount of popular support, but the amount of popular support was dependent on the strength of the KLA and the probability of success. To overcome this ‘opening mobilization dilemma’, the KLA employed tactics that exaggerated the actual strength of their organization and the weakness of their opponent (Mulaj, 2008: 1111-1112). Increasing hit and run tactics (i.e. 10 per month) were combined with the manipulation of violent images, which were spread through national and international media. The media attention amplified the perceived strength of the KLA and effectively popularized them. This caused Albanian diaspora to reroute the voluntary taxes they paid to the LDK to the KLA; between 1997 and 1999 the U.S. branch of Albanian diaspora donated over US$30 million to the KLA (Mulaj, 2008: 1112; Judah, 1999). This surge in popularity coincided with the sudden collapse of the Albanian state in 1977. Hundreds of thousands of Albanians had invested their life savings in fraudulent pyramid banking schemes and consequently saw their money dissolve into thin air. The Albanian army and police force now no longer existed, and the state’s armories were practically up for grabs. The KLA’s funding and arms problems were over (Mulaj, 2008: 1110; Judah, 1999). According to Kubo (2010: 143), however, this did not lead to a direct escalation of the rebellion. The KLA had still not gathered a mass following and did not expect to do so until 1999. So slow but steadily, they made their preparations for war.

The Serbian police did not await them and quickly swept in, launching a ferocious attack in order to eradicate the KLA uprising (Mulaj, 2008: 1111). Between 28 February and 7 March 1998, they targeted the villages of Likoshan, Çirez and Prekaz, which had become associated with KLA activities. Using machine guns, rocket-propelled granades, helicopters and armored vehicles, the Serbian forces killed a total of 85 people – including 12 women and 11 children (Kubo, 2010: 1146). The massacre sparked the outrage of the Albanian Kosovars; revolt spread out and different village militias started to rally under the KLA flag (Kubo, 2010: 1146; Judah, 1999).

The news spread around the world, outraging the international community (Mulaj, 2008: 1110). Albanian Diasporas flooded in to join the rebellion; it is estimated that the KLA membership drastically increased from 1.200 to 25.000 in three months (May – July 1998). The number of KLA attacks increased accordingly: in 1998 there were 1470 attacks, whereas in 1997 there were only 55 attacks (Kubo, 2010: 1148). Instead of planning and executing guerilla tactics, the leadership of the KLA was now burdened with the coordination of many smaller factions without having established a fully operational chain of command yet (Kubo, 2010: 1146-1147; Judah, 1999).

In July 1998, Belgrade deployed the Yugoslav military forces to root out the KLA. The chaos had not yet settled, when the Yugoslav unleashed their artillery, battle tanks and armored vehicles on the Albanian Kosovars. During this period, the LDK had created a military formation of their own: the ‘Armed Forces of the Republic of Kosova’ (FARK) now joined the frontline. FARK temporarily joined forces with the KLA to fight for independence (Boyle, 2014: 190). However, their activities were limited to a few operations. By then the KLA had become such a dominant moral and military force in Kosovo that FARK, and by extension the LDK, was not able to make a significant impact (Bekaj, 2010: 20). operations were actively impeded by the KLA; the ICTY (2012: 98-103) reports that FARK-personnel were told by KLA operatives that they had to comply with the KLA-command or leave Kosovo. FARK-soldiers were repeatedly beaten into submission, and on occasion even shot. According to Boyle (2014: 190), the FARK minister of defense, Ahmet Krasniqi, was shot to death in 1998 after a

(12)

threatening communiqué was released, which proclaimed that “one of these kind of people will pay for the damage they have caused our nation”.

The KLA was now clearly poised to have the lead in the Albanian Kosovar resistance group (Mulaj, 2008: 1110). Although the LDK still believed that they were the legitimate representatives of the Albanian Kosovars since they had won the elections, the Contact Group (comprised of United States, United Kingdom, France, Germany and Russia) seemed to confirm the new faction hierarchy when the most prominent actors of the Kosovo conflict were called to Rambouillet and Paris to talk peace in 1999 (Mulaj, 2008: 1110). Instead of Ibrahim Rugova, Hashim Thaçi – one of the KLA’s political leaders and future Prime Minister of Kosovo – was chosen as head of the Albanian delegation, thus recognizing the KLA as a prominent, and unavoidable actor (Mulaj, 2008: 1110). More importantly, the KLA, who had been formerly viewed as a terrorist organization, were now invited guests at the negotiation table (Fulton, 2010; Bekaj, 2010: 21). NATO and the KLA even joined forces, most notably at the battle of Mount Pashtrik, where the collaboration between KLA ground forces and NATO air forces proved essential for the deterrence of Yugoslav forces (Mulaj, 2008: 1114).

The NATO air strikes, which lasted for 78 days, effectively ended the civil war. On 12 June, NATO ground troops began entering Kosovo, as the Yugoslav forces left for Serbia (Kubo, 2010: 1149; Bekaj, 2010: 45). The UN security council passed Resolution 1244 (UNSCR 1244), which established the UN administration in Kosovo. The KLA, consequently, was expected to disarm after these developments. Bekaj (2010: 45) argues that this demilitarization of the KLA has been achieved and that the

insurrections of the KLA ended in 1999. However, Boyle, (2014: 190) notes that, after the war, both the LDK/FARK and KLA were keen on gaining a dominant position within the provisional government. This set off sporadic acts of KLA-FARK violence; a series of “politically colored executions” of known FARK members followed. Members of the KLA have been trialed for their involvement in cases of torture and murder of ex-FARK officers. However, the trials were not concluded, as the key witnesses were murdered (Boyle, 2014). The LDK managed, nonetheless, to win the first free and internationally recognized elections in 2001 (Bekaj, 2010: 31). The KLA top members, Hashim Thaçi and Ramush Haradinaj, went on to form the biggest political parties besides the LDK: respectively the PDK and AAK, thus establishing themselves as legal, legitimate and relevant political actors (Bekaj, 2010: 31).

5. Analysis

In order to test the generalizability of Lawrence’s (2010a) ‘Competitive Violence Theory’, the

beforementioned five hypotheses are tested through the Kosovo case. If the hypotheses are correct, then I expect to observe that the transition to violence is explained by competitive dynamics. In the occurrence of a sudden power vacuum, more active organizations will emerge (H1). Moreover, co-ethnic factions compete for political relevance; they struggle for legitimacy and recognition (both local and international) by discrediting each other and employing violence to gain visibility and popularity among the target population. Violence will, subsequently, increase over time and

nonviolent actors will adopt violent strategies to stay relevant (H2 and H3). Finally, I expect to observe violence between co-ethnic factions (H4). I do observe this in the data. I am fairly confident of my findings, because the data in different studies overlap regarding key aspects of the case (Bekaj, 2010; Mulaj, 2008; Kubo, 2010; Judah, 1999; IICK, 2000; Maliqi, 2001). However, it should be noted that due to the clandestine nature of actors such as the KLA, inside information is always acquired second hand and thus is inherently biased. By acquiring different sources of information, I have attempted to verify the data. My analysis leads me to qualify the theory.

(13)

Name

Goals

Strategies

LDK

Democratic League of Kosovo Self-determination Pacifist resistance

LPK

People’s Movement of Kosovo Self-determination Armed combat

LKÇ

K

National Movement for the Self-determination Liberation of Kosovo

Active resistance (i.e. public acts of defiance)

KLA

Kosovo Liberation Army Self-determination Armed combat

UPS

UP

Independent Student’s Union More human right, of the University of Prishtina primarily education.

Mass demonstrations

Table 1 - active parties between 1990 and 1999

5.1 The fragmentation trigger, intra-group competition and transition to violence

As the case illustrates, the LDK and Ibrahim Rugova were the dominant actors in the resistance group in the early ‘90s. They were the chosen leaders of the parallel government, and were therefore in charge of the ‘voluntary tax’ funding, its expenses, and consequently the local media (IICK,2000). These assets were employed to create a monopoly on political power and information. Public

dissidence was therefore largely absent or barely visible (Table 1) (Maliqi, 2001). The dynamics of the peaceful resistance movement, however, changed after the Dayton agreements in 1995. Rugova and the LDK lost the unconditional faith the Albanian Kosovars had in them (IICK, 2000: 50; Kubo, 2010: 1144; Mulaj, 2008: 1109; Judah, 1999). The LPK, in the meantime, had secretively established a guerilla group in 1992/1993 and were preparing for an armed struggle (Judah, 1999). Until Dayton, the constituency, however, were backing the peaceful resistance, and an armed insurgency was publicly deemed as foolish and counterproductive. The first attacks of the LDK guerilla group (KLA) followed shortly after these developments, as the KLA saw their chance to jump into the leadership vacuum left behind by the LDK and establish themselves as legitimate and recognized actors (Maliqi, 2001). Lawrence (2010b), however, notes that one must be careful not to identify the triggers of violence by reasoning back from violence (i.e. “doing history backwards”). In this regard, it should be noted that the KLA leadership was, at the time, aware of the implications of the Dayton agreement and their political room for violent maneuvering (i.e. “violence pays”) (Maliqi, 2001; Mulaj, 2008: 1110). Furthermore, the KLA shared the same goals as the pacifist resistance (i.e. independence through internationalization of the conflict), which does not directly cause competition among actors. Mulaj (2008: 1109), however, argues that the KLA was not only created to oppose the Serbian regime, but also explicitly to oppose the civil resistance led by the LDK.

Although the LDK was losing credibility after Dayton, they were still the dominant force in terms of finance, political relevance, and media control. Despite their shared goals, the LDK used these sources to denounce any alleged KLA activity. According to Bekaj (2010: 20), the LDK’s leadership had early on been informed about the KLA’s existence by the KLA themselves. Nevertheless, Rugova had repeatedly tried to discredit the KLA by claiming that they were a Serbian construct designed to destabilize the civil resistance. From then on, a rivalry between the two factions ensued. (Bekaj, 2010: 20; Judah, 1999).

The KLA challenged Rugova’s hold on power in Kosovo, but more importantly, they challenged the legitimacy of his representation of the Albanian Kosovars, both nationally and internationally (Mulaj, 2008: 1109). But to gain political relevancy, the KLA was first required to gain popularity among the constituency. They understood that they needed a mass base to gain legitimacy and consequently

(14)

recognition by the international community. This meant they had to step out of the shadows and advertise their cause. Accordingly, and in response to the LDK’s denouncements, the KLA gradually increased the intensity of their violent efforts (31 attacks in 1996, 55 attacks in 1997 and 66 attacks in the first two months of 1998) (Kubo, 2010: 1142). On 28 November 1997, they made their first public appearance and proclaimed their resolve to fight for the independence of the Albanian Kosovars (Bekaj, 2010: 20). Now the KLA had gained momentum and attracted the attention of national and international media. With their sustained guerilla attacks and manipulation of violent images, combined with the amplifying effect of the media, the KLA seemed far stronger than they actually were and more likely to achieve success at the end of the campaign (Mulaj, 2008: 1111).

Simultaneously, they were attracting attention from the international community. This, however, was a negative blow for the KLA: the United States had marked the KLA attacks as terrorist activities (Bekaj, 2010: 21). In spite of this, the positive image and popularity among the local Albanian

Kosovars and Albanian diaspora soared; many diaspora switched their financial backing for the LDK to the KLA (Mulaj, 2008: 1112; Judah, 1999). Up until this point, the KLA faction was, in spite of their growing popularity, still small and had trouble recruiting men. According to Kubo (2012: 1145) Rugova’s hold of the LDK and his constant public rejection of violence still inhibited KLA recruitment. This drastically changed after the massacres of Lukoshan, Çirez and Prekaz in 1998 . Since the KLA was the only active violent faction that could channel the sparked outrage, Albanian Kosovars joined the KLA’s ranks in numbers (Kubo, 2012: 1146-1147). The KLA thus became the most prominent force in the resistance. The peaceful strategy of the LDK became obsolete in the process, as Kosovo

descended into civil war. This was illustrated by the formation of FARK, the army of the LDK-led government (Boyle, 2014). FARK’s impact, however, proved to be marginal due to the dominance of the KLA (Bekaj, 2010: 20). FARK eventually joined forces with the KLA to secure independence (Boyle, 2014: 190).

The new political power hierarchy (with a dominant KLA) was reflected in the composition of the Albanian delegation during the Rambouillet and Paris peace talks in 1999. As stated by Mulaj (2008: 1110), the fact that the chosen head of the Albanian delegation was KLA political director Hashim Thaçi, clearly expressed the recognition of the KLA as a political relevant actor in the conflict. This recognition is further exemplified by the co-operation of the KLA and NATO during military operations against the Yugoslav Army (Mulaj, 2008: 1113-1114).

In sum, the data has provided the following insights: the seemingly unitary nonviolent resistance fragmented after the Dayton agreements and opened a leadership vacuum. Most notably, the LPK/KLA deliberately made use of this and started to actively pursue violent resistance. Secondly, intra-group competition can be observed between the LDK and KLA: the LDK constantly attempted to delegitimize the new violent movement, which led the KLA to gradually intensify their activities. The KLA started out small and used violence to gain notoriety and momentum for their cause. The LDK eventually took up arms as well, but was, at that stage, not able to compete with the KLA. These observations lead me to accept H1, H2 and H3.

5.2 Intra-group violence

One of the key-aspects of the ‘Competitive Violence Theory’ is its ability to account for the occurrence of violence aimed at co-ethnic nationalists. When factions enter the dual contests of nationalist conflicts, they target co-nationals to eliminate the competition. I do observe these

dynamics in the available Kosovo data as well. Although the factions joined forces after the Lukoshan, Çirez and Prekaz killings, there was a continuation of rivalry and tension between the KLA and the LDK-led FARK from the onset (Boyle, 2014: 190). The LDK were convinced that, because of the elections, they were still the legitimate representatives of the Albanian Kosovars, and therefore expected the KLA to yield (Bekaj, 2010: 20). At this point, the KLA were, however, poised to have the

(15)

lead in the resistance movement (Mulaj, 2008: 1110). The KLA coerced the FARK into compliance with threats and, in some cases, consequent acts of violence. Prior to the merge of KLA and FARK

command, FARK-personnel were told that they had to submit to the KLA command or leave Kosovo, and that their safety could not be guaranteed (ICTY, 2012: 98-103). The ICTY (2012) further reports that FARK-personnel, who were crippled by the fact that the KLA had become such a dominant force, were repeatedly beaten by KLA soldiers. Furthermore, KLA soldiers told members of FARK that they would “eliminate” the FARK and their commanding officers (ICTY, 2012). According to Boyle (2014: 190), the FARK minister of defense, Colonel Ahmet Krasniqi, was shot to death in 1998 by KLA-operatives.

The hostilities between KLA and FARK members decreased in intensity when they faced a common enemy, only to intensify in the immediate post-war period. Both parties (i.e. LDK and KLA) were maneuvering for a prominent place in the provisional government. FARK attempted to back a LDK-led government and consequently came into conflict with the KLA (Boyle, 2014: 190). This caused sporadic violence between the KLA and FARK, and notions of “politically colored” executions of FARK members followed (Boyle, 2014: 190). The occurrence of intra-group violence leads me to accept H4.

5.3 Alternative explanations

An alternative explanation for the transition from nonviolent measures to violent ones would be that the faction that prefers violent strategies might view the nonviolent means as inefficient and

ultimately unsuccessful. It is clear that the LPK/KLA thought that nonviolent resistance would prove unsuccessful, as they had started to prepare for armed battle in 1992/1993 (Judah, 1999). This argument, however, cannot account for the fact that the armed group was not only created to oppose the Serbs, it was explicitly created to oppose the LDK as well (Mulaj, 2008: 1109).

Lawrence (2010a) offers a counterargument to the theory of competitive dynamics as well. She notes that a resistance movement could switch strategies from nonviolent to violent out of anger and frustration. This argument has different observable implications. Violence should be primarily two-sided; factions do not engage in dual contests. It might even unite the population, which causes intra-group competition to be absent. Furthermore, it predicts retaliatory violence; intra-intra-group violence is only aimed at collaborators and the target regime. Lawrence (2010a) further argues that these arguments are not mutually exclusive. The implication of the latter being that both the competitive violence argument and the grievance based argument could both be operating in the same conflict. However, the key empirical challenge of this counterargument is measurement: it is difficult to measure emotions and when the threshold that prompts violence is reached (Lawrence, 2010b). It should be noted that in the Kosovo case, the events in Lukoshan, Çirez and Prekaz mobilized the masses drastically. Therefore, one could make a compelling argument that these events served as a trigger to unite the Albanian Kosovars, which led to a collective resistance campaign of violence. The fact that the competitive dynamics and the counterargument are not mutually exclusive could then account for the fact that there were competitive dynamics from the onset of violence until the mass mobilization trigger (i.e. the Lukoshan, Çirez and Prekaz attacks) and co-operation between factions after the trigger.

My explanation will be better than the alternative, because the data suggests that grievances did not cause the onset of violence post-Dayton in 1996: Lawrence (2010a) argues that actors could adopt violent measures out of anger and frustration. However, the LPK/KLA had already begun their preparations for armed combat in 1992/1993 and were explicitly waiting on an opportunity to rival the LDK and oppose the Serbian regime with violence (Judah, 1999; Mulaj, 2008: 1109). Other actors did not engage in armed combat after Dayton. Furthermore, the competitive violence theory also accounts for the competitive dynamics in the developments of the conflict in Kosovo. Finally, the

(16)

counterargument suggests that other co-ethnic nationalists would not be targeted (Lawrence, 2010a: 101). However, co-operation between FARK and the KLA after the Lukoshan, Çirez and Prekaz attacks only followed after violent coercion by the KLA (ICTY, 2012). The fact that intra-group violence was associated with opportunities to gain political power suggest that competition between the LPK/KLA and LDK/FARK existed.

6. Conclusion

In this paper, an attempt has been made to find an explanation that is able to account for the transition of nonviolent to violent strategies in the resistance movement of Albanian Kosovars throughout the ‘90s. I argue that, in this regard, competitive dynamics among co-ethnic nationalist factions can account for the use, timing and escalation of violence following a nonviolence resistance campaign.

The most prominent findings in existing literature on violence onset focus on independent variables that are stable over time and can thus not account for sudden shifts towards violence (i.e. the timing of violence)(Kalyvas, 2007; Fearon & Laitin, 2003; Collier & Hoeffler, 2004). Moreover, the literatures that focus on social change are promising, but focus on mobilization of movements in general and not on the specific transition from nonviolent resistance to violent resistance (Lawrence, 2010). The findings of Stephan & Chenoweth (2008) show that, contrary to popular belief in academic literature, nonviolent measures are significantly more effective than violent ones. It, therefore, emphasizes the importance of understanding why violent strategies are adopted after a nonviolent campaign. The strength of the ‘Competitive Violence Theory’ lies in its ability to provide a causal mechanism that leads to the use of violence: a leadership vacuum triggers movement fragmentation and subsequent competition (Lawrence, 2010). It is, therefore, able to provide an explanation for temporal variation, by identifying critical opportunities and motives.

I have applied this theory to the case of Kosovo and evaluated the argument at faction level. The dominant viewpoint in conflict literatures, is that nationalist conflicts are dyadic in nature (Lawrence, 2010; Cunningham, Bakke & Seymour, 2012; Cunningham, 2013; Pearlman, 2009). In the case of Kosovo, the early ‘90s were characterized by Albanian Kosovar solidarity in a peaceful resistance campaign. Without the notion of actor fragmentation, one could not explain how the Albanian Kosovar factions (i.e. LDK/FARK and LPK/KLA) used different tactics after the fragmentation trigger (i.e. the Dayton Agreements), and entered in a public escalating contest for credibility, recognition and political power. Furthermore, the competitive dynamics predict violence aimed at other co-ethnic nationalists, which has been observed in Kosovo as well (Bekaj, 2010; Mulaj, 2008; Kebo, 2010; Judah, 1999; IICK, 2000; Maliqi, 2001). Consequently, my analysis has led me to qualify the theory.

The findings in this research carry certain policy implications: first, by adopting violent tactics, the KLA were successful in attracting international attention and expelling the Serbian regime. According to Fulton (2010), this meant a successful resistance campaign carried out by a terrorist organization. The post-war developments culminated in a complete secession from Serbia (Bekaj, 2010). This development opposes the observation noted by Stephan & Chenoweth (2008) that nonviolent measures are more effective than violent strategies in obtaining policy goals. Understanding the dynamic of resistance and the transition from nonviolence to violence in Kosovo thus carries a significance for future policy on secessionist movements and their strategies. Furthermore,

international actors, prior to the Dayton Agreements, seem to have perceived the Albanian Kosovar resistance as a unitary group and believed that the region was effectively pacified (i.e. pacifist without internal dissent) for the time being. In doing so, the international community ignored and neglected the viable and imminent threat of conflict escalation and genocide, taking into consideration the prior events in Bosnia and Croatia. By taking into consideration the possibility of movement fragmentation,

(17)

one could argue that dealing with the demands of a pacifist organization would prove more preferable than dealing with possible violent escalations.

In this paper, I have attempted to test the generalizability of the ‘Competitive Violence Theory’. An important claim of this theory is that it can explain the continuation of violence after the stated aims of the faction (Lawrence, 2010). This, however, could not be tested in the Kosovo case as the UN effectively pacified the region. Further empirical data is required to test its generalizability.

(18)

7. Bibliography

Bekaj, A. R. (2010). “The KLA and the Kosovo War”, Berghof Transitions Series, 8.

Boyle, M. J. (2014). Violence after war: explaining instability in post-conflict states. Baltimore: JHU Press.

Chenoweth, E., & Stephan, M. J. (2011). Why civil resistance works: The strategic logic of nonviolent

conflict. Columbia University Press.

Chenoweth, E., & Stephan, M. J. (2014). “Drop your weapons: When and why civil resistance works”, Foreign Aff., 93, 94.

Collier, P., & Hoeffler, A. (2004). “Greed and grievance in civil war”, Oxford economic papers, 56(4), 563-595.

Crenshaw, M. (1981). “The causes of terrorism”, Comparative politics, 13(4), 379-399.

Cunningham, K. G. (2013). “Actor fragmentation and civil war bargaining: How internal divisions generate civil conflict”, American Journal of Political Science, 57(3), 659-672.

Cunningham, K. G., Bakke, K. M., & Seymour, L. J. (2012). “Shirts today, skins tomorrow: Dual contests and the effects of fragmentation in self-determination disputes”, Journal of Conflict Resolution, 56(1), 67-93.

Fearon, J. D., & Laitin, D. D. (2003). “Ethnicity, insurgency, and civil war” American political science review, 97(01), 75-90.

Fulton, J. R. (2010). “NATO and the KLA: How the West Encouraged Terrorism”, Global Security Studies, 130-41.

Independent International Commission on Kosovo (IICK). (2000). The Kosovo Report: Conflict,

International Response, Lessons Learned, Oxford University Press on Demand.

Judah, T. (1999). “Kosovo's road to war”, Survival, 41(2), 5-18.

Kalyvas, S. N. (2007). “‘Civil Wars’”, Chapter XVIII in Stokes, Susan Carol: The Oxford handbook of

comparative politics. Vol. 4.; Oxford University Press, USA.

Kubo, K. (2010). “Why Kosovar Albanians took up arms against the Serbian regime: The genesis and expansion of the UÇK in Kosovo”. Europe-Asia Studies, 62(7), 1135-1152.

Lawrence, A. (2010a). “Triggering nationalist violence: Competition and conflict in uprisings against colonial rule”. International Security, 35(2), 88-122.

Lawrence, A. (2010b). “Driven to arms? The escalation of violence in national conflicts”. In

Chenoweth, E. & Lawrence, A. (eds) Rethinking violence: states and non-state actors in conflict, 143-171. MIT press

Maliqi, S. (2001). Why peaceful resistance movement in Kosova failed, Chapter IX in UNDP, MDGD,

New States, Old Problems, Innovative Solutions: The Case of the Succesor States of the Former Yugoslavia (rebuilding Fractured Communities serie); New York

McAdam, D., McCarthy, J. D., & Zald, M. N. (1996). Comparative perspectives on social movements:

Political opportunities, mobilizing structures, and cultural framings. Cambridge University Press.

Mulaj, K. (2008). “Resisting an oppressive regime: The case of Kosovo Liberation Army”, Studies in Conflict & Terrorism, 31(12), 1103-1119.

Pearlman, W. (2009). “Spoiling inside and out: Internal political contestation and the Middle East peace process”, International Security, 33(3), 79-109.

Prosecutor v Haradinaj, Balakj & Brahimaj (Judgement) ICTY-2012-IT-04-84bis-T (29 November 2012)

Stephan, M. J., & Chenoweth, E. (2008). “Why civil resistance works: The strategic logic of nonviolent conflict”, International Security, 33(1), 7-44.

(19)

Peace over violence or vice versa? Competition dynamics and conflict behavior