The risk of domino secessions: interdependent secessions and lessons from the Western Balkans

ABSTRACT

The resurgence of nationalist tendencies around the world has brought new attention to the problem of smaller regions seceding from larger states. While research exists regarding the contagious nature of self-determination, ethnic conflict and civil wars, the interrelationships among various secessionist conflicts have not been systematically studied. We show that the circumstances of secession conflicts in different countries are interconnected by the perceptions of elites and that outcomes in one case may, therefore, influence expectations, motivations, strategies and outcomes in others. We also identified several indicators of interconnectivity among secessionist conflicts in the Western Balkans and visualised the strength of the links among them using social network analysis tools (Pajek – Spider). After the outbreak of a secession conflict, several inter-secession effects will be directly transferred to the most interconnected cases, and the greatest domino risk results from the combined impacts of direct and indirect nth-order effects.

KEYWORDS:

• Secession
• independence
• interdependency
• Western Balkans
• Balkanisation
• security

Introduction

The secession of smaller territorial areas from larger states has been an issue for international order and stability ever since the state-based system was established. Borders have frequently been disputed by internal or external actors and states tend to see secessionism as a threat to their security, stability or existence. Many states naturally wish the existing international order to remain unchanged, but several leaders of secessionist movements (e.g. political parties, NGOs, terrorist groups, paramilitary groups, etc.) and the people who support them hold different ambitions. Disputes over borders and territory continue to be a key cause of intra- and inter-state conflicts with no country being immune to the problem. We can find secession challenges not only in dictatorships, communist and poor countries but in democratic and rich countries. The list of secession conflicts or cases is long and includes Catalonia and the Basque country in Spain, Corsica in France, Northern Ireland and Scotland in the United Kingdom, Chechnya in Russia, the Eastern regions of Ukraine, Tibet in China, Kashmir in India, Kurdistan in Iraq and Turkey, etc.

Secession as a “major contemporary socio-political problem” and “worldwide phenomenon” (Beran 1984, p. 21) has emerged in the Western Balkans in multiple subtle ways. The secession of Slovenia and Croatia in 1991 led to a further declaration of independence by the Serbs living in Croatia. The ensuing declaration of independence by Bosnia and Herzegovina immediately led to another secession by the Serbs (Republika Srpska) from the newly established Bosnia. This was followed by Kosovo’s secession from Serbia in 1999 (physically due to NATO’s intervention) and in 2008 (by a unilateral declaration), the Albanians’ attempt to secede from FYROM (North Macedonia) in 2001, the rebellion of the Albanians in the south of Serbia. It is thus clear that this region may offer important lessons about the complex interconnections among various secessionist conflicts or situations.

The goal of this paper is to explore interconnections among secessionist conflicts theoretically and empirically in the context of the Western Balkans. The question concerns how different secessionist situations are linked and what indicates such links. While several secession processes are unfolding in the Western Balkans, it remains unclear how they influence each other. We argue that secessionist conflicts in the different Western Balkan countries are inter-connected in complex ways through the perceptions of elites and that outcomes in one may influence the expectations, motivations, strategies and outcomes in others. These connections may be observed using various indicators. Moreover, we expect to find a complex cognitive network of interdependent secessionist conflicts. The complexity of the interconnections among secessions not only refers to the multiplexity of connections (as expressed by indicators) but also to the existence of a network made up of both direct (1st-order connections) and indirect (2nd- or n-order) connections among secessions. Many of these interconnections exist in the minds of elites and decision-makers. Wood (1981, p. 124) stressed that secession, like revolution, begins in the minds of men. This means that the perceptions held by political leaders, experts and the general population are key to each individual secession process and key to their mutual relations and influences. If those participating believe that the outcome of another secessionist conflict has influenced their secession and think and act accordingly, these secession processes are interdependent.

The problem of interconnected secessions has not been given sufficient attention in the literature and we aim to address this deficiency. The dominant focus is on the definition of secession, theory (explanatory and normative theories), legal aspects of secession, secession movements and state response, the role of external actors (states and international organisations), self-determination, counter-secession, recognition and cases of secession. In the first part of this paper, we review certain relatively rare arguments about interconnected secessionisms from the secession and international relations literature. In part two, we create an indicator-based model for exploring the connections among secessionist conflicts in different countries and test it in the Western Balkans. The last section presents a quantitative network visualisation of interconnections found in this region enabled by the use of the social network analysis tool Pajek or Spider. The empirical part is based on our analysis of regional media assessments,¹ policy reports and interviews with a sample of 33 experts and informed observers (academic researchers, professors, policy analysts from think-tanks, journalists and representatives of international organisations) from countries facing a secession conflict – Bosnia and Herzegovina (5), Serbia (8), Kosovo (9) and North Macedonia (11) – from 2012 to 2018. Interviewing was used as one of the most common ways of understanding human behaviour (Fontana and Frey 1994, p. 361) in a situation when we are unable to study the totality of a population and try to find representative opinions (Tashakkori and Teddlie 1998, pp. 62–64). Internal validity of the results was achieved by combining media assessments, policy assessments and qualitative and quantitative assessments by interviewees, while external validity or transferability was achieved by connecting empirical results to theory. We selected experts based on nonprobability purposive and sequential sampling (see Tashakkori and Teddlie 1998, p. 76) by contacting and interviewing experts from different fields and with high levels of professional knowledge of secession conflicts in their countries until their opinions became saturated (they did not bring anything new to the views of others). Although we are aware that the use of media and policy assessments and interviews all have certain limitations, we combined them to reduce the risk of interest-based misinterpretation.

Interdependent secessionist conflicts in the literature

Secession is a contested concept definable as “a process of withdrawal of a territory and its population from an existing state and the creation of a new state on that territory” (Pavkovic and Radan 2016a, pp. 1–2, 2016b, pp. 1–2). The dissolution of states (e.g. the former USSR and SFRY) is technically a different phenomenon, yet it is almost always preceded and caused by secessions. When a movement wants to incorporate its territory into the already existing neighbouring country, we talk about irredentism (Pavkovic and Radan 2016a, p. 9, Coggins 2016, p. 23).

The term secession is used in this paper similarly to Pavkovic and Radan (2016a, p. 5) not only to represent completed cases of secession (where a declaration of independence was followed by the recognition by a significant number of other states and international organisations) but primarily to denote a stage before the completion of the process (unfinished secessions). Unsuccessful secession attempts are quite frequent (see Coggins 2014, p. 5, Mayall 2016, p. 11, Krause 2022, p. 788) and do not lead to a new state being established, despite all efforts and even the possible declaration of independence. Attempts at secession usually lead to conflict with the host state, often violent but occasionally not so (in the cases of Norway from Sweden, Slovakia from Czechoslovakia while Czechoslovakia dissolved, and Iceland from Denmark). In addition, the situation is often complicated by gross human rights violations, ethnic cleansing, terrorism, counterinsurgency campaigns, etc. (see Coggins 2014, p. 3). Secession often poses a threat to the host state (see Roeder 2018, p. 2, Butt 2017, p. 8) as it loses its territory, population, sovereignty and jurisdiction, including associated resources (taxes) and power. Furthermore, secession is also an emotional process involving the betrayal of the central state. This “transferability of emotional effects” (Butt 2017, p. 34, 222) is likely to occur among secession cases as well.

A secession crisis occurs when the leaders of a territorially concentrated and distinct community within a larger state translate their discontent into demands for secession and hold sufficient power, either through strong internal mobilisation or the use of force, to compel the central government to react to these demands (Bartkus 1999, p. 10). The universe of secession cases (or our research units), therefore, consists of cases of antagonistic relations between the central government and secessionist movement(s) in more or less autonomous regions. Such crises also affect other emerging or existing crises, including secessionist conflicts. In general, secessionist crises/conflicts can be negatively linked through shared risk and spillover violence and positively linked through policy emulation and convergence.

The existing literature refers to interconnected secessions in a limited way. Most authors depict secessions and their interconnections negatively, as threats to the international order. For example, it is stressed that a “never-ending fissuring of states” would be accompanied by continuous civil war and international anarchy (Glaser 2003, p. 384) or also “international chaos” (see Ker-Lindsay 2012, p. 26). Duerr (2015, pp. 185–186) warned that a central issue with secessionism is that it might encourage other minority groups across the world to make similar claims for independence. In Europe alone, there are numerous regions where support for independence is low but could change quite quickly. For example, several regions might decide to copy the Scottish referendum for independence, such as in Basque country, Padania, Veneto, Brittany, Corsica and Szeklerland (Hungarians in central Romania). Moreover, other regions that already have some autonomy might alter their approaches, such as Wales, Greenland, South Tyrol, Northern Ireland, Sicily, Sardinia, Galicia and the Faroe Islands. Several regions in Eastern and South-Eastern Europe could also face disruptive changes: the problems in Bosnia with Republika Srpska, in Kosovo with its north, in Transnistria in Eastern Moldova, in Abkhazia and South Ossetia in northern Georgia, in Nagorno Karabakh in Azerbaijan, in North Macedonia, in Chechnya, etc.

Kempton and Clark (2002, pp. 1–4) stressed that the Soviet Union’s dissolution was an epic event that opened Pandora’s box by unleashing centrifugal forces with no predetermined or natural endpoint. The process of “Matryoshka nationalism” emerged because it was feared that the new successor states would gradually shatter into smaller and smaller states. Just like a Russian Matryoshka doll, a wide variety of ethnonationalist groups have sought to carve out homelands within the 15 successor states. Many of these new successor states today have a centre–periphery relationship problem, requiring endless negotiation/renegotiation. History, justifiably or not, offered the term Balkanisation to depict the problem of recursive and mutually connected secessionist conflicts, unreasonable fragmentation and related violence and other inter- and intra-state problems. Wood (1981, p. 121) mentioned that the urgent desire to preserve a poly-ethnic union may stem from the recognition that one secession may lead to another, possibly resulting in Balkanisation. Indeed, the Balkans has been the least stable area in Europe and its wars (the two Balkan wars, two World Wars and armed conflicts in Slovenia, Croatia, Bosnia, Kosovo and Macedonia) add to its negative reputation. The post-war situation over the last 20 years reflects a delicate mix of positive and negative trends. Observers have noted that the region has been captured between the forces of integration and disintegration (Bildt 2004, pp. 10–13), using terms like “unstable security” (Vukadinovic 2003, p. 65) and “relative stability” (Gazdag et al. 2007, p. 17). We believe that, when applied to secessions, Burr’s notion of the prevalence of demonstration effects in the mobilisation of ethnic grievances (see Hale 2000, pp. 33–36) suggests that domino secessions are more likely in smaller geographic contexts where demonstration effects in one secessionist case are easily transferred to a neighbouring case.

The explanatory theory of secessionism additionally mentions demonstration effects whereby the success of one nearby or similar secessionist group necessarily impacts the demands of other states or groups (see Wood in Pavkovic and Radan 2016a, p. 177). It should also be noted that strategically-minded secessionist leaders will always take a close look at successful secessions (see Roeder 2018, p. 20). For a better understanding of the positive interconnections among secessions, we can draw on the literature on policy convergence and policy diffusion. The policy convergence literature defines convergence as any increase in similarity between the characteristics of policies across space and time (see Dyson 2008), while the spatial diffusion literature assumes that policies are likely to spread in space and time due to communication links, political factors, the political and cultural similarity of proximal states (neighbours), mutual familiarity, etc. Neighbour-to-neighbour diffusion or contagion is considered a regularity in such cases. The result of policy diffusion across space and time leads to wave-like diffusion patterns (Jensen 2004). Such policy isomorphism stems from a “constraining process that forces one unit in a population to resemble other units that face the same set of environmental conditions” (DiMaggio and Powel in Jensen 2004, p. 112). The literature on norm diffusion also suggests that policy emulation is motivated primarily by the success and legitimacy of attractive policies (see Farrell and Terriff 2010, p. 7), in our case, the attractive other secession policies.

Interestingly, Lenin took a broader view of secessions in the October Revolution of 1917. His concept of revolution was established as a platform for the mobilisation and programmatic coordination of several secession campaigns against the host state (Roeder 2018, pp. 7–9). The political environment provides good conditions for coordination in larger states with several simultaneous national secession campaigns. Such “multiple movement states” (see Butt 2017, p. 13) are statistically more prone to declarations of independence due to the coexistence of several secessionist movements (Griffiths 2018, p. 65). In the period 1945–2010, 3 states had 10 or more secession campaigns, 5 states had 5–9 campaigns and 21 states had 2–4 campaigns. States with a large number of simultaneous campaigns were the Soviet Union, India, Indonesia, France, Yugoslavia, etc. (Roeder 2018, p. 28). In her dynamic model, Walter (2006a) pointed to the importance of complex strategic interactions between the government and a larger number of ethnic groups over time. Ethnic groups decide to challenge the government based on the government's past accommodation policies towards other secessionist groups and associated expectations about it in the future. Empirical results have shown that the odds of initiating conflict with the government are six times higher if an ethnic group has observed the government granting concessions to others in the past than if no accommodation has ever been observed. On the other hand, governments know this and worry about their reputation in terms of their own readiness for compromise and accommodation. They are much more likely to fight a secessionist if the number of similar challengers in the future is high (Walter 2006b) and if they fear that giving up part of their territory will lead other groups to demand independence (e.g. risk of setting a precedent) (Toft 2002). Permitting secession for one region may set precedents for other potential secessionists if they exist within a state (Griffiths 2015, p. 371). Cognitive links between secessions are, therefore, more or less directly incorporated into the strategic game between governments and ethnic groups.

The concept of Balkanisation also raises the question of the involvement of foreign powers that may want to influence secession and other internal affairs. The international community’s attitude to secession can generally increase or decrease motivations for and costs of secession. The prevalent conservative attitude shown by states and the international community to secession (see Walter and Ungern-Sternberg 2014, p. 3) shows that the chances of domino secessions and the related instability and chaos are actually quite real. It is these fears that led states to limit the justifiable application of self-determination. As Bartkus (1999, p. 68, 111) wrote, the UN Charter recognises the right to self-determination but only in a very restricted fashion when applied to the context of de-colonisation (only for the existing administrative units of colonies) and not for others. In most secession cases, international organisations remain silent and the governments affected tend to preserve the status quo in the international structure.

The fact that self-determination is to some extent contagious has also been theoretically and empirically demonstrated using global data for the period from 1960 to 2005. A diffusion of self-determination likely occurs in proximate states by (1) the observation of others using self-determination as a basis of organisation (nonrelational diffusion or emulation through watching and learning from the experience of others), (2) inspiring greater attention to own related grievances and (3) affecting or increasing perceptions of potential gains based on the observed accommodations by the central governments and success of the groups making the self-determination claim. These globally developed notions of self-determination have been able to diffuse regionally based on the proximity of states (Cunningham Gallagher and Sawyer 2017). This corresponds to another fact, namely that ethnic conflicts have been contagious as they spread from one state to another based on ethnic kinship. This was also seen in the global data for the period from 1946 to 2009. The data suggest that dispersed or transnational ethnic groups act as transnational transmitters of ethnic conflict and that the probability of ethnic conflict in one state increases the probability of ethnic conflicts in another state if they share the same ethnic group. Members of a dispersed ethnic group can share a joint sense of homeland, monitor the behaviour and status of each other across borders and may become inspired to replicate the behaviour of kin members involved in a conflict. This research also suggests that a good example is the link between the ethnic conflict in Kosovo with the one in North Macedonia (Forsberg 2014). Civil wars have also been contagious based on the principle of proximity among states and the fact that they create disruptions and demonstration effects and attract other recruits or rebels. The higher the infection risk by a neighbouring civil war, the higher the likelihood of third-party (neighbour state) intervention in this war (Kathman 2010).

Further evidence of interconnections among secessions is the fact that historically speaking, secessions come in waves. The number of states in the international system had dramatically risen from 25 in 1816 to around 200 in 2008 predominantly due to secessions (see Coggins 2014, 2016). However, secession, as quite a rare outcome of political struggle, requires a unique window of opportunity. When this window is opened in one state, it can produce a significant effect on other secessionist conflicts in the region. A wave of secessions swept across Europe after the end of the First World War and the number of states in the world has risen considerably since the end of the Second World War. The UN Charter incorporated the principle of equal rights and self-determination of peoples, which was followed by a big wave of decolonisation in Africa and elsewhere in the 1950s and 1960s. The last wave came after the end of the Cold War with the dissolution of the Soviet Union, Yugoslavia and Czechoslovakia on top of the secessions of Eritrea from Ethiopia, Namibia from South Africa and Timor Leste from Indonesia. In this context, Beissinger (see Pavkovic and Radan 2016a, p. 191) stressed that secessionist mobilisation may emerge as a transnational tidal force and not as an isolated collection of movements.

A group of authors doubt that secession is a solution because it paradoxically leads to the threat of further secession. Freeman (1998, p. 19) stressed that ethnic states created by secession are likely to discriminate unjustly on ethnic grounds and Glaser (2003, p. 371) identified the problem of trapped minorities, i.e. a minority in a secessionist state which opposes the break-up, finds itself deprived of membership of the association it prefers and is forced to become a member of a new association. In addition, Bishai (1998, pp. 93–94) found that the solution promised by secession is illusory and morally empty because it merely draws new boundaries around the same problems of shifting populations and identities. It is futile to match territorial boundaries to the many possible types of human group affiliation. Secession is thereby at most a temporary and incomplete solution to the problems of political consent. Secession creates new types of minority groups trapped within either the seceding territory or the territory of the parent state. Secession simply recreates the original problems inherent within the state structure and creates the potential for new secessions. This is also confirmed statistically when we see that the anti-colonial wave of independence did not create an enduring equilibrium in society and, instead, led to an even stronger desire for independence and sovereignty (see Coggins 2016, p. 40).

In contrast, some authors regard the interconnected secessions within a state as potentially useful. Beran (1984, p. 30, see also Siroky 2016, p. 65) developed the domino theory of secession based on the recursive use of the majority principle. He argued that an initial successful secession is likely to lead to a series of secessions resulting in unviable political entities. To prevent this, he proposed a very specific liberal solution in the form of recursive use of the majority principle where a minority within a seceding entity would also be given a chance to secede. Costa (2003, p. 78) similarly discussed “secessions within secession” or “secession within the seceding states”. This domino spread of secessions is only possible if the internal minorities of the seceding states are allowed to secede. Namely, a minority should be permitted to secede and establish its own state, but only with the use of the majority principle which is an essential part of democracy. Beran (1998, p. 38) also indicated that this recursiveness of referendums might continue until they are applied to a single territorial community (i.e. a community no longer composed of a number of communities) to determine its political status. Consequently, some authors have understood the purpose of self-determination as the eventual accomplishment of the optimal structure of the international system. There would be fewer conflicts and wars in the international system if each nationality were a separate nation and thus satisfied with its lot (Waltz, 1959 in Bartkus 1999, p. 108).

Empirical evidence on secessionist interconnections in the Western Balkans

The above-mentioned models of contagion of ethnic conflicts, civil wars, self-determination and policy convergence or emulation models suggest that some patterns of interdependence exist among secession conflicts. In summary, they suggest that dispersed ethnic groups in proximate states in any regional environment might observe the behaviour, status and success of their kin, learn from each other, affect their perceptions, inspire and emulate or replicate successful behaviour. These theories of contagion and emulation can be usefully explored in the Western Balkans.

The former Yugoslav space underwent several “sequential secessions” (Slovenia, Croatia, Bosnia, Macedonia and Kosovo) and consequently also attempts at “recursive secessions” (Serb Krajina from Croatia, Republika Srpska from Bosnia, North Macedonia from FYROM). Some secessionists influenced each other: the timing of Croatia’s declaration of independence was, for example, inspired by that of Slovenia (see Coggins 2014, p. 78). Pavkovic and Radan (2016a, p. 129) explain that some of the mentioned secessions “triggered” other secession attempts. Kosovo's secession resembles that of Bosnia and Herzegovina as it appears to have been stimulated by declarations of independence by Croatia and Slovenia (Pavkovic 2000, p. 492). There is also evidence that Great Power leaders were very concerned about the possible demonstration effects of secessions among the Yugoslav republics and aware that Yugoslav separatism could potentially influence separatism in the USSR (Coggins 2014, p. 96, 112). Indeed, a similar recursive use of secessions took place in the USSR when Georgia's secession was followed by secessionist violence in Abkhazia and South Ossetia and when Moldova's independence was followed by a secessionist move in Transdniestria (see Siroky 2016, p. 55). In addition, Sakwa and Pavkovic (2016, p. 147) observed that the emergence of Yugoslav secessionist movements “provoked” entrenched conflicts in the former Soviet space that had assumed a frozen character until Kosovo's independence was recognised in 2008. It was also noted that there was considerable emulation and assistance among secessionist movements in the former Soviet Union (travel by representatives, legal assistance, media support, emulation of the content of declarations of independence, mutually agreed secessions, etc.). Similarly, in the space of the former SFRY, there were coordinated proclamations/declarations of independence (Slovenia and Croatia), with the Slovenian and Croatian declarations of independence being emulated by Bosnia and Kosovo (Pavkovic and Radan 2016a, pp. 136–157, Sakwa and Pavkovic 2016, p. 164, Coggins 2014, p. 94).

The case of Western Balkans was chosen because it is a small and closely interconnected region with a common history and plenty of shared problems, including a relatively high number of successful and unsuccessful attempts over the past 30 years. This makes it even more likely that secessionist tendencies there are interconnected. Several secessionist conflicts are considered in our analysis. Bosnia and Herzegovina was pacified by the Dayton Agreement in 1995 although serious secessionist and autonomous tensions in Republika Srpska continue even today. The long-term political vision of some actors in Republika Srpska has been to allow Bosnia to function in exchange for growing regional autonomy and, if the opportunity arises, to organise a referendum on secession and declare independence. Like Bosnia, Kosovo has been a kind of mild “protectorate” of the international community. NATO’s military intervention in 1999 brought an end to Belgrade’s physical control and Kosovo declared independence in 2008. Negotiations on the status of Kosovo between Belgrade and Pristina have largely focused on the status of the Serbs remaining within the borders of Kosovo. There was a serious possibility that Serbs from North Kosovo would rebel, secede and join Serbia. Albanians in southern Serbia (the municipalities of Presevo, Bujanovac and Medvedja between 1999 and 2011) and in North Macedonia (in 2001) have already had armed rebellions in efforts to secede to join Kosovo. Sandjak is mainly a Muslim-populated part of Serbia on the border with Bosnia. Potential secession was discussed at the time of the war in Bosnia but since then religious and economic factors have become stronger drivers of autonomy in the region. Vojvodina is an autonomous province in the north of Serbia that is populated by foreign minorities. This province has been moving in the direction of autonomy in the context of a not particularly well-substantiated fear of secession.² Our analysis below aims to show that these secessionist conflicts are interconnected in several ways.

Qualitative evidence on the potential for domino secessions

Several domino scenarios for the region were assembled from existing media sources and our interviews in the region. In the first scenario, Kosovo’s declaration of independence could have been followed by the declaration of independence by the Serbs living in the north of Kosovo and their accession to Serbia (Jane’s 2007b). Judy Batt (2007, p. 1) stressed that no one can really stop these Serbs from declaring independence from Kosovo if they really decide to do so. Accordingly, Kosovo’s proclamation of independence would probably trigger these Serbs’ declaration of secession from Kosovo, then likely leading to Albanian violence and revenge in the town of Mitrovica and other Serb enclaves. Another question was what could KFOR actually do to prevent this violence. The international community might also very quickly become the target of radical Albanians (e.g. from Vetvendosje or the Self-determination Movement). The second scenario concerns the spread of instability from Kosovo to Bosnia and Herzegovina. The Socialist Party of Serbia, for example, had continuously been calling on the Bosnian Serbs to declare independence should Kosovo secede from Serbia (SE Times 2007a). Such statements could also be heard in Republika Srpska itself. The third scenario considers the expansion of the conflict from Kosovo to Macedonia, where early signs of potential unrest were emerging (SE Times 2007c).

Our interviews with experts in the region confirm these interconnections among the cases. In one scenario described by one interviewee, the North Kosovo Serb territorial defence force organises attacks on the international community and Serbs formally declare independence from Kosovo, the Serbian government officially announces its support for the secession of North Kosovo and its incorporation into Serbia. In reaction to this, the Albanians living in the south of Kosovo would attack Serbian municipalities and the Albanians living in Presevo would rebel, followed by a rebellion in Western Macedonia. Albanians would announce that they also have the right to secede from Serbia and Macedonia. Tirana would claim the rights of the Albanians living in southern Serbia. Bulgaria would then become active in relation to Macedonia (Interview 11). The violent division of Kosovo would definitely bring about the division of Macedonia, the separation of Presevo and part of Montenegro – they would all initially join Kosovo and then Albania at a later stage (Interview 9). As a result, Albanians and Serbs would be forced to flee (Interview 12). In the most radical scenario, the independence of Kosovo would be just one step in the joining of all Albanians within a single state. In this context, the International Commission on the Balkans gauged the mood of the population of the Balkans on the issue of Kosovo and Albania unifying to create one country (Great Albania). It turned out that any such potential unification was positively perceived by the Kosovo Albanians and residents of Albania, while the populations of other countries (Bosnia and Herzegovina, Serbia, Macedonia and Montenegro) viewed this negatively. Moreover, the Kosovo Albanians and residents of Albania saw the disintegration of both Bosnia–Herzegovina and Macedonia as positive. This led the Commission to conclude that Europe must still carefully consider the Albanian factor in the region (Centre for Liberal Strategies 2005, p. 17, 43).

The qualitative analytical framework in this paper is based on the indicators of interconnections among secession conflicts in the Western Balkans. We created indicators from the existing theory and empirical data (as suggested by the grounded theory; see Strauss and Corbin 1998, p. 12). This means that several indicators were created from (1) the keywords from the above studies of secession conflicts; (2) the collected media material from the region and (3) interviews with experts. We reduced the existing indications of interconnections among the secessionist conflicts in the Western Balkans to nine general indicators (see Table 1). The collected empirical material offered a chance to see some deeper patterns and converging perceptions or practices, and create informative and useful indicators. The purpose of identifying the indicators in our study was to obtain a broad, multidimensional range of indications regarding the links among secessionist processes and to better see their patterns.

Table 1. Indicators of interconnections among secessionist conflicts.

Indicators of interconnection	Explanation
Exchange of observers to determine the best practices	Secession actors monitor other similar cases by occasionally sending representatives to observe and learn
The transfer of violence from one conflict to another	Incidents and various violent activities in one case of secession can contribute to incidents and violence in other cases
Compensatory behaviour in terms of the transfer of motivations and strategies	Secession actors create their motives and action strategies based on the outcomes achieved in other secessionist conflicts
Problematic recognition of independence due to potential interconnections and domino effects	Other states have problems recognising secessions due to fearing a chain reaction within their territory
Spillover of diplomatic recognition battles into the sports arena: which secession entity may participate in which international competition	The battle for recognition can transfer into the sport arena where secessionists compete to participate in official competitions
Artificial pacification of a rebellion in one case to give other cases better chances of success	Secessionist leaders in one case could influence secessionists in other cases to reduce their activity at a time when it might harm their chances of success
Conservative policy of the international community to stop the escalation of secessions to prevent domino effects	The international community is afraid of domino effects that might threaten the existing order
Ideas of exchanging territory in the event of interconnected secessionist conflicts	If secessionist cases are proximate, ideas of controlled simultaneous secessions that include the exchange of territory can emerge
Pan-ethnic movements as a motivational glue for connecting secessions	Ideas of a common future in a single state can simultaneously fuel several secessionist conflicts, making them more interconnected

Exchange of observers to determine best practices. In the region of former Yugoslavia, such exchange visits have been underway since Slovenia’s independence. It is known that both Catalans and Kosovars sent several informal delegations to Slovenia to meet with various stakeholders who presented the Slovenian success story. In our fieldwork, we also learned that the President of Republika Srpska, Dodik, had been closely monitoring Montenegro’s separation from Serbia and paid great attention to the fact that Montenegro had a right to a secession referendum in its state contract. Dodik sent an adviser to Montenegro to study how this was being done. Since then, the “secession referendum” appeared in Dodik’s vocabulary (Interview 5) and he followed a similar strategy of making the state dysfunctional by blocking state institutions and using the legislative process (Interview 3).

The transfer of violence from one conflict to another. The possibility of the conflict being transferred from Kosovo to BiH was shown in 2004 during the unrest in Kosovo. A Serbian teenager was killed in a drive-by shooting and three Albanian kids drowned in a river. Mass opinion wrongly attributed the drownings to Serbs, with mass protests and violence by Albanians against Serbs and their cultural heritage starting across entire Kosovo, leading to many people being killed and wounded. These incidents were politically condemned in Bosnia, yet the Serbian Orthodox Church in Bugojno was set on fire and Islamic community offices in Banja Luka were stoned in retaliation for the violence against the Serbs in Kosovo.

The transfer of violence became even more obvious after 1999 when the conflict in Kosovo ended and NATO intervened. This was when elements of the Kosovo Liberation Army (KLA)³ withdrew from Kosovo and started a rebellion in southern Serbia, which saw the Liberation Army of Presevo, Medvedja and Bujanovac (UCPMB) being established. They benefited from the vacuum created in the 5-kilometre border safety zone between Serbia and Kosovo that was controlled neither by NATO nor the Serbs. When the security control in this zone was returned to Yugoslav forces, some fighters escaped and came to North-West Macedonia in 2001 to launch a rebellion there. Their motivation was built on a combination of independence and organised crime aims. The fighters who arrived in Macedonia also came from Kosovo and its KLA. According to Jane’s security assessment, evidence shows that the local National Liberation Army (NLA) of Albanians in Macedonia was led by experienced UCPMB and KLA fighters while drawing its rank-and-file from disaffected local ethnic Albanians in the country’s north (Jane’s 2007c). The fighting lasted six months. The lesson of this conflict is that situations in Kosovo and southern Serbia determine the secessionist conflict in Macedonia. If these two situations deteriorate, the experience in 2001 demonstrates that the threat of a spillover jeopardising the territorial integrity of Macedonia is a reality. However, such a transfer of violence did not occur in 2004 during or after the violence against the Serbs in Kosovo. This means that we are talking about the possibility of transfer if the political leaders in Kosovo are not satisfied with how the secessionist conflict related to Kosovo is being resolved.

Compensatory behaviour in terms of the transfer of motivations and strategies. Republika Srpska’s possible secession from BiH is strongly motivated by the compensatory logic related to Kosovo achieving independence from Serbia. President of Republika Srpska, Dodik, has frequently stated that the way of arranging Kosovo’s status will set a precedent for the Republic of Srpska. If the current policy trajectory in Kosovo is maintained, the Serbs will also move in the direction of becoming independent from BiH. He stated that only the Republic of Srpska is a durable category, not BiH, and that he intended to organise a referendum on independence if Kosovo were to become independent. In 2008, the Parliament of Republika Srpska passed a resolution declaring the right to organise a referendum on the legal status of the entity Republika Srpska in the event of Kosovo unilaterally declaring its independence. The dissolution of Serbia and Montenegro has revived aspirations to maintain the entity of Republika Srpska, and even create new states within BiH. Some of those in Serbia’s team negotiating on Kosovo suggested in 2006 that if Serbia had to change its borderline (due to the loss of Kosovo), then new borders in BiH should also be considered. This scenario would have directly led to the disintegration of BiH (Jane’s 2007a). All the negotiations surrounding Kosovo coupled with an emboldened Republika Srpska have seen the Croats in BiH starting to demand greater autonomy through the idea of establishing a third Croat entity, a step that would require altering the Constitution. Not surprisingly, Dodik was in support of the Croat endeavours, stating that the efforts to establish a third entity in Bosnia are natural (Delo 2015a). Experts believe that if Serbs declare independence in BiH then the people in Sanđak will, by way of compensation, then declare independence from Serbia (Interview 13).

Interestingly, the start of the rebellion in Kosovo against the Serbs is also connected to the way the Bosnian crisis was resolved in the 1990s. Disappointed because the Dayton Peace Agreement of 1995 did not resolve Kosovo’s status, the Kosovo Liberation Army started with guerrilla and terrorist activities in 1998 with the aim for Kosovo to secede from Serbia and then create a Greater Albania. From that point on, the Serbs in Kosovo have sought to compensate for their actual/potential loss of Kosovo by making increased efforts to secede and join Serbia. As part of the complex ethnic distribution of Serbs in Kosovo, the Serbs in northern Kosovo were mostly in favour of annexing their part of Kosovo to Serbia, whereas Serbs in enclaves located elsewhere were more cautious and did not support that “solution” to such a degree. Accordingly, the Serbs in Kosovo did not recognise the authority of UNMIK after the war in 1999 and thus created separate and parallel administrative structures, courts, schools, a health system and even security structures. Local experts believe this logic was supported in Serbia where the majority wanted to compensate for the loss of Kosovo by having the North secede from Kosovo (Interview 11). Today, the Serbian goal in the North is not so much to bring about secession but to achieve some kind of autonomy within Kosovo. This might one day then be used as an instrument for independence (Interview 7).

Already in 1992, after a cascade of secessionist impulses in the former Yugoslavia, another smaller-scale secession referendum was held in the Presevo valley, organise by the local Albanian authorities but not recognised by the central government in Belgrade. The majority of ethnic Albanians voted for independence from Serbia and expressed their desire to join Kosovo. They started to refer to the Presevo valley as Eastern Kosovo. By doing so, they immediately drew an implicit (even explicit) link between southern Serbia and the Serb-inhabited territory in the north of Kosovo (north of the Ibar River) (Preshevajone 2016, GlobalSecurity, n. d.).

Success in the case of Kosovo and the almost unlimited support of the USA and NATO led the Albanians to believe they would be able to continue with secessions in southern Serbia and Macedonia (i.e. that similar international support would then be provided). The Liberation Army of Presevo, Medveda and Bujanovac was, therefore, established and immediately became active once the Kosovo conflict came to an end. This paramilitary and separatist organisation was fighting for independence from the Federal Republic of Yugoslavia and was seen as a successor to the KLA, which was officially disbanded following the war in Kosovo (BBC News 2001). After signing the Konculj Agreement and disarming, these Albanians repeated several times that they would unite with Kosovo if the Serbian villages in the north of Kosovo were to join Serbia. A CEPS policy report revealed that Albanians living in the Presevo valley would be both pleased and disappointed if Kosovo were to gain full independence: pleased that Kosovo has gained its freedom from Serbia and disappointed because they could not participate in the new state (CEPS Policy Brief 2007). Their biggest political goal since then is to achieve greater autonomy within Serbia and to enjoy the same rights as held by ethnic Serbs in Kosovo.

A similar transfer occurred in North Macedonia. First, in 1992 the autonomous Republic of Ilirida was declared in the western part of Macedonia. Certain Albanian politicians considered this idea a solution to the Albanian minority problem and the problem with rights in Macedonia. In 2014, Ilirida was declared again. The Macedonian government declared the idea unconstitutional. In 2015, armed terrorists from Kosovo clashed in the Macedonian village of Kumanovo. The Army of the Republic of Ilirida was said to be involved (Delo 2015b). Albanians in Macedonia are also pursuing the following compensatory logic: “if Serbs are to get an entity in the North, then why can’t they get one as well?” (Interview 1). Former president of the Albanian political party PDP Nevzat Halili openly argued that Macedonia needs to be federalised or disintegrated in the event Kosovo disintegrates. The idea of the federalisation of Macedonia is perceived by many Albanians as the partition of Macedonia (Interview 10).

Problematic recognition of independence due to potential interconnections and domino effects.⁴ The most telling example from the region concerns the recognition of Kosovo after it declared independence in 2008. Despite Kosovar authorities engaging in a long and painful process of seeking support and recognition, Kosovo was still only recognised as a state by 101 of the 193 UN member states and 23 of the 28 EU member states by the end of 2020. A typical reason given by other states for recognition was respect for the popular will of the Kosovars in the context of past Serbian repression, while a typical reason for not recognising Kosovo was that it would violate international law and the fear of triggering a domino effect on similar situations in their own state. Among EU members, Cyprus often stated it would never recognise the unilaterally declared independence of Kosovo, even if all other EU members did, due to the similarity with its own circumstances and because of the principles that were violated in Kosovo (Al Jazeera 2008). Greece supports this view due to its links with Cyprus. Slovakia also did not recognise Kosovo since it was always in support of a final settlement of its status based on an agreement between Serbs and Kosovars and within the UN Security Council in line with UN-approved principles (Ministry of Foreign Affairs of the Slovak Republic 2008). Yet, it is well known that Slovakia does not recognise Kosovo due to the potential precedent it would set for the Hungarian minority living in southern Slovakia. Spain has not recognised Kosovo officially since it has objections regarding the legality of a unilateral declaration of independence from the international law perspective. A second reason given concerns possible implications for Spain’s own situation at home related to the autonomous communities of Catalonia, Galicia and the Basque country (The Kosovo Times 2009). Bosnia and Herzegovina is clearly divided on the topic of the recognition of Kosovo: the Parliament of Republika Srpska adopted a resolution demanding that Bosnia not recognise Kosovo and not cooperate officially with Kosovo; otherwise, the Bosnian Serbs will begin making concrete steps towards affirmation of their state status (B92 2008). Consequently, BiH has not recognised Kosovo. The government of Macedonia supported Kosovo’s independence from Serbia, but the SE Times (2007b) reported that the risks of transferring separatism to Macedonia and damage to its own economic interests meant that it did not support the idea of Kosovo being ethnically divided.

Russian authorities frequently declared that Kosovo’s unilateral declaration of independence runs against the UN Charter, resolution 1244, the principles of the Helsinki Final Act and the sovereignty of Serbia. Russia has since then acted in support of Serbia’s territorial sovereignty and integrity and blocked all related themes in the Security Council. Russian President Vladimir Putin labelled Kosovo’s declaration of independence a “terrible precedent” that will come back to hit the West “in the face” and “de facto blow apart the whole system of international relations, developed not over decades, but over centuries”. Putin also stated that the move was not well considered. Other Russian officials added that this move could boost the claims made for independence in separatist regions in Western Europe (Ministry of Foreign Affairs of Russia 2008). The secession of Kosovo also affected the process of the dissolution of Ukraine. Russia often drew historical and international interconnections between the conflicts in Crimea and Kosovo. The Russian Foreign Ministry cited the case of Kosovo in 2008 as a precedent in support of the legitimacy of the declaration of Crimean independence (Center for Strategic and International Studies 2018). Putin stated that Crimea did exactly the same as Kosovo had done with the assistance of Western colleagues; unlike Kosovo, Crimea had even organised a referendum and the Albanians have been permitted to do more than the Russians (Presidential Information Office 2014). Alongside these interconnections, Russia recognised the breakaway Georgian regions of Abkhazia and South Ossetia as independent in 2008, explicitly referring to Kosovo in its decision (see Summers 2014, p. 235).

Spillover of diplomatic recognition battles into the sports arena.⁵ Sometimes the battle for recognition has crossed over into the sport arena, especially when it comes to the seceding entity seeking to obtain permission to participate in official competitions. In 2012, the international football association FIFA allowed a football team from Kosovo to play official friendly international games. Yet, Michel Platini, the president of UEFA (The Union of European Football Associations) was seriously critical and protested against this decision by the president of FIFA. Platini announced that this was a political decision in breach of the UEFA statute, made despite all five of UEFA’s members in FIFA have been against it, despite the facts that UEFA did not allow Kosovo to play such official games and that the UN did not recognise Kosovo as an independent state. FIFA’s president replied that it is not the task of international football authorities to prohibit football games but to open the door for football playing. A consequence of this was the response of the Football Association of Republika Srpska (part of the football association of BiH), whose president asked FIFA to allow the football team from the Serbian entity to also play international friendly games. He expected help with lobbying, including Serbian assistance (24ur 2012). This illustrates how interconnections among secessionist conflicts can play out in the sports arena.

Artificial pacification of a rebellion in one case to give other cases better chances of success. It turned out that politicians from Kosovo had not only fuelled other secessionist efforts but also acted to calm them down to give Kosovo a better chance of maintaining and implementing its own secession. There is evidence that such calls were made to rebels in southern Serbia because, at the time, any conflicts there might have disturbed Kosovo’s drive towards independence (CEPS Policy Brief 2007). In North Macedonia, when a group of some 100 men occupied the village of Kondovo in 2004 and 2005 the conflict was stopped after the intervention of Albanian politicians from Macedonia (Ali Ahmeti) and politicians from Kosovo (Hardinaj). Hardinaj convinced the armed men to surrender, and the violence was successfully halted. While some thought that this rebellion was an outcome of the poor economic situation in the village, evidence shows that it was directed by the Albanian National Army, a pan-Albanian radical group (Jane’s 2007c).

Conservative policy of the international community to stop the escalation of secessions to prevent domino effects. Considerable data confirm the conservatism and protectionism shown by the international community. After the call by the Socialist Party of Serbia to Bosnian Serbs in October 2007 to declare independence should Kosovo secede from Serbia, diplomats of some Western countries sent official protests to the Serbian Ministry of the Interior, pointing out the consequences of linking BiH with the status of Kosovo. High representatives of the international community in BiH several times reminded the president of Republika Srpska to stop with and be very careful with his secession-associated statements and provocations that could hold broader ramifications. Should such statements continue, the international community promised to take harsh measures like replacing the ministers and other high-ranking officials. Dodik then publicly declared he was not afraid of such threats by the High Representative. Furthermore, the domino-effect hypothesis is also seen in the Serb reactions to the international community’s position. Serbian Prime Minister Vojislav Kostunica expressed Serbia’s full support for Republika Srpska in its fight against the threat to the fundamental interests of the Serbian people. Moreover, he added that the High Representative had himself created a crisis to abolish Republika Srpska, which is only part of a broader conspiracy that wants to grant independence to Kosovo. He declared that Serbia would defend the rights of the Serbs in BiH with the same determination as he was defending the right of Serbia to keep Kosovo (SE Times 2007a).

In relation to Kosovo, the international community had rejected the idea of any further partition due to potential domino or spill-over effects elsewhere (Interview 7; Interview 8). Everybody is aware that this is Pandora’s box: Macedonia, southern Serbia … (Interview 11). When negotiations on the final status of Kosovo began in February 2006, the work of the Troika and the Contact group followed these principles: Kosovo cannot be divided, a return to the pre-1999 situation is unacceptable and the territory of Kosovo cannot be annexed by any neighbouring country. KFOR with its mandate was also ready to prevent both a violent secession by the Serbs and any violent intervention by the central government from Pristina against the Serbs. One of our sources (Interview 14) stated that the military escalation of violence in Kosovo did not worry KFOR so greatly due to its existing military capabilities on the ground. The bigger problem was any comprehensive popular uprising, which KFOR would have been unable to stop.

Aiming to prevent domino effects, a 5-kilometre safety zone was created in southern Serbia after the war in Kosovo ended. The Yugoslav military was not allowed in this zone, only lightly armed Serbian local police forces could operate there. This zone prevented the Yugoslav military and Kosovars from coming into direct contact, yet it also unexpectedly gave an opportunity for Albanian rebels to become organised.

In the area of Macedonia, UNPROFOR (United Nations Protection Force) prevented the spread of violence until 1995 when UNPREDEP (United Nations Preventive Deployment Force) was established. Its mandate was similar: monitoring and reporting on developments in the border areas that could threaten the territorial indivisibility, confidence and stability in Macedonia. This mission was considered to have been successful while it lasted. Yet, the Macedonian government recognised Taiwan in exchange for some financial investments, prompting the Chinese government’s strong reaction by vetoing the prolongation of this mission in the Security Council (see Matovski 2007). Had it existed, UNPREDEP would have been in a position to stop the spread of violence from Kosovo to Macedonia in 2001.

Ideas of exchanging territory in the event of interconnected secessionist conflicts. The International Crisis group mentioned after the war in Kosovo that the idea of swapping certain parts of Serbia and Kosovo should not be excluded (After Milosevic 2001, p. 41). The idea of partitioning Kosovo and the related swapping of territories between Serbia (it would get north Kosovo) and Kosovo (it would get southern Serbia) has been alive since then and more or less formally expressed by stakeholders in both Belgrade and Pristina. This idea might appear to some as a creative way of honourably freeing up Serbia from the burden of Kosovo (EUobserver 2010). Ideas of partitioning Kosovo were again brought to life when the former CIA Deputy Chief for the Balkans, Steven Meyer, publicly advised Serbian Prime Minister Vucic to withdraw from the dialogue with Pristina and demand a more effective negotiation framework. In these negotiations, the Serbs had informally proposed recognising Kosovo, but the North would have to secede (Dacic) and Vuk Jeremic informally proposed an exchange of territory (the Presevo valley for North Kosovo) (Interview 11). The existence of this logic of swapping territory in the case of a Serbian secession from Kosovo was also confirmed by the Albanian experts (Interview 9; Interview 6; Interview 2). Swapping areas of land would entail very organised and controlled simultaneous secessions and territorial exchanges.

Pan-ethnic movements as a motivational glue for connecting secessions. The region has moved from the idea of a Great Serbia in the 1990s to the idea of a Great Albania. The former was recently again expressed by the former president of Republika Srpska. He stated that, together with the majority of ethnic-Serb parts of Kosovo, Republika Srpska should create a greater Serbia. He said: “There is a wonderful idea … and that is to separate RS [from Bosnia] and to form a union with Serbia and that is together with four municipalities of Kosovo” (EUobserver 2017). A more radically growing idea of Great Albania has built on the myth of pan-Albanianism and connects several secessionist conflicts in where Albanians play a role. Many observers have seen pan-Albanianism as a serious threat to Balkan stability, while the Albanians downplay its relevance and talk about the importance of Euro-Atlantic integrations for their countries. Following long negotiations with Serbia and frustration because the Kosovo people still need a visa to enter the EU, Albanian Prime Minister Edi Rama stated that Albania and Kosovo will unite in a classic way if the EU does not establish the conditions for the European integration of Kosovo and the liberalisation of visas. He contends that unification is unavoidable and will take place either in the EU or outside of it. Kosovo’s Prime Minister commented on this statement by saying this was not a threat to the EU, but a reality arising from Kosovo’s isolation from the EU (Delo 2015c). In 2001, the Albanians in Macedonia were actually taken advantage of by the Albanians from Kosovo in the direction of the romantic notion of a Great Albania (Interview 4). To support this dream, the Albanian National Army (ANA) was established in 1999 as a pan-Albanian radical group calling for the unification of all Albanians living in Albania, Macedonia, Serbia, Montenegro and Greece to form the state of Greater Albania. However, in reality, this group has grown to become a web of criminal, terrorist and extremist organisations (see Jane’s 2007b, 2007c, International Crisis Group 2004, p. 9). In addition, Albanian parliamentary member Mesila Doda stated that Macedonians as a nation never really existed and that Macedonia is an artificial state structure. On its territory lies the state of Ilirida, which will once again be created by the Albanians (Delo 2017). Furthermore, several years ago a prominent high representative of the Albanian government wrote a book on Great Albania.

Quantitative evidence on the potential for domino secessions

Applying network theory to the relationships among secessionist conflicts inevitably leads to a network of connections among nodes (secession conflicts) with relatively strong connectivity and density (cases are supposed to be relatively linked) and with some cases more central than others (more strongly linked to others). Our fieldwork in the Western Balkans yielded sufficient quantitative evidence of the interconnections among different secessionist conflicts. A sample of 33 interviewed experts on Bosnia, Serbia, Kosovo and North Macedonia answered a question about the strength of the links among seven secessionist conflicts (Republika Srpska – Bosnia, North Kosovo – Kosovo, North-West Macedonia – North Macedonia, Southern Serbia – Serbia, Sandjak – Serbia, Vojvodina – Serbia, Kosovo – Serbia) on the scale shown in Table 2.⁶

Table 2. The strength of interconnections among secessionist conflicts.

No connection	Some/weak connection	Medium connection	Strong connection
0	1	2	3
0	0%–33%	33%–66%	66%–100%
	One-third of the situation depends on another situation	Half the situation depends on another situation	Most of the situation depends on another situation

A database of 187 entries was created based on the answers given. Assessments such as 2 +  and 0–1 were recoded as 2.25 and 0.5, respectively.

The average strength calculated for interconnection among all secessionist conflicts is 1.44 (somewhere between weak and medium). This means that our respondents on average regarded them as not being strongly interconnected. However, this is only the average view and certain pairs are much more connected than others. Our calculations of the strength of the links among all potential secession pairs show that the most interconnected secessions are North Kosovo from Kosovo with southern Serbia from Serbia (2.88), Republika Srpska from Bosnia with Kosovo from Serbia (2.5) and Republika Srpska from Bosnia with Sandjak from Serbia (2.36), whereas the least interconnected cases are southern Serbia from Serbia with Vojvodina from Serbia (0.25). This asymmetric distribution of the estimated strength of the interconnections suggests that an escalation of secessionist efforts in some cases would have a stronger effect on other cases.

Accordingly, if all secessionist conflicts are perceived to be more or less interconnected, they may be seen as a cognitive network. To visualise such a network among secession impulses in the Western Balkans, we created an additional database with the average interconnections among all pairs of secession conflicts. We used software for social network analysis (Spider – Pajek, see http://mrvar.fdv.uni-lj.si/pajek; De Nooy et al. 2005). A node in this graph (see Figure 1) represents a specific secessionist conflict, while the boldness of a line shows the strength of their average mutual interconnection on a 0–3 scale based on the opinion of the interviewed experts.

Figure 1. Interconnected secessionist conflicts in the Western Balkans on a scale from 0 to 3 (each node represents a secessionist conflict between a secessionist and related central government).

The ties among the various secessionist conflicts actually create a very dense network (network density of 0.76). The figure suggests that all secessionist cases are interconnected (this is a connected and almost complete network) except for Kosovo’s secession from Serbia, which is still an unfinished process in the minds of the Serbs from Republika Srpska. The most dangerous escalations with the biggest average impact on all other situations are the secession of the Serbs in North Kosovo from Kosovo (average link strength of 1.7 with all other cases) and the secession of Republika Srpska from Bosnia (average link strength of 1.56). The above visualisation also suggests that the escalation of one secession in the Western Balkans is likely to affect other secessionist conflicts. In addition, this network graph shows that the escalation of one secession could affect other secession efforts not only directly (1st-order effects) but also indirectly as 2nd- or n-th order effects. For example, the escalation of the secession crisis in BiH could directly affect the situation in North Macedonia and indirectly by escalating the crises in North Kosovo, southern Serbia, Sandjak and Vojvodina. The final effect on North Macedonia therefore might be much bigger than initially expected.

Conclusion

This paper corroborates previous general studies on the contagion of self-determination and ethnic conflict in the case of the Western Balkans. We can confirm our argument that secessionist conflicts in different Western Balkan countries are interconnected in complex ways. Our empirical model showed that the outcomes in one secession conflict influence the expectations, motivations, strategies and outcomes in others. Empirically, we list several indicators that reflect the multiplexity of interconnections among secessionist conflicts. It is of interest that interconnections exist in different dimensions: they span from direct effects in terms of the transfer of violence to the legal, political or diplomatic dimension of interconnected battles for recognition, bargaining on the exchange of areas of territory, battles in the sports arena through to dreams of bringing all representatives of one ethnic group together in a single state, which could be accomplished by several interconnected simultaneous or consecutive secessions. Identified qualitative indicators can be used for studying relations among secession conflicts also in other European regions, Asia, America and Africa.

We also found that the perceptions held by the elites in the Western Balkans on the interdependence of secessionist conflicts may entail a complex cognitive network. Our quantitative findings suggest that the cases are sufficiently connected for the escalation of tensions to be transferred from one case to another. Some cases are particularly strongly linked such that an escalation in one directly affects the escalation of others (North Kosovo from Kosovo – southern Serbia from Serbia, Republika Srpska from Bosnia – Kosovo from Serbia, and Republika Srpska from Bosnia – Sandjak from Serbia).

This means that allowing one secessionist conflict to freely develop the international community risks a domino effect on other secessionist conflicts. These effects will be directly transferred among the most strongly connected cases, but the biggest risk arises from the combined impact of direct and indirect nth-order effects. This means that, besides the relatively predictable direct effects, several unpredictable and unimaginable indirect effects could put the wider region on “secession fire”. The recursive use of secession referendums might seem to be the most democratic way of changing the current state order, but it is unlikely that everyone would conduct such a referendum. Still, their application would be revolutionary in practice.

Our findings usefully supplement the explanatory theories of secession (which examine what causes secessions and how actors act) and deepen the existing knowledge on conflict contagion by showing how different secessions may “feed each other”. An inter-secession perspective can be usefully added to the social, political and psychological preconditions of secessions and also be helpful in the process of predicting secessions. A useful predictor in the secession process is not merely what secession movements and central governments want and do, but also how other secession situations affect their desires and actions. There is probably no better place to learn lessons from interconnected secession conflicts than the Western Balkans. It was very valuable to use a bottom-up approach by incorporating local experts and their views on interconnections among the various secessionist conflicts. The innovative application of social network analysis to visualise interconnections among secessions in the Western Balkans has also significantly improved the present understanding of the complexity related to secessionism and the unfolding risks of armed conflicts in the future. Finally, we also found that secessionist movements can be interconnected by the way they justify their secession, implying that the connections can additionally be further explored in the normative theory of secession (investigating the justifications of secession).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are openly available in Social Science Data Archives (ADP) at https://doi.org/10.17898/ADP_SOZB18_V1, reference number [ADP – IDNo: SOZB18].

Additional information

Funding

This work was supported by Javna Agencija za Raziskovalno Dejavnost RS: [Grant Number BI-US/18-19-047].

Notes

1 In the empirical part, we use articles from recognised foreign media in the English language (EU Observer, SE Times, The Kosovo Times, BBC) and also recognised regional media in local languages (Delo, B92, 24ur).

2 Vojvodina is used more as a control case in our study. There, secessionism was never particularly strong and was more financially driven. However, as an autonomous republic in the former Yugoslavia, it was legally in a similar position to Kosovo. Under the 1974 constitution, like other republics it held many rights, except the right of secession (see Ker-Lindsay 2012, p. 33).

3 The Democratic League of Kosovo, led by Ibrahim Rugova, was a mainstream secessionist party modelled on the Polish Solidarity, which organised a shadow government and attempted to secede by peaceful means. After the international community’s refusal to include the Kosovo issue in the Dayton Peace Agreement because Kosovo was not a former administrative unit of Yugoslavia, and after other disappointments, the KLA, as a radical and minor secession movement, started a more direct and violent rebellion, attracting many new members and supporters (see Pavkovic and Radan 2016a, p. 153; Sakwa and Pavkovic 2016, p. 161; Kubo 2016, p. 177, Coggins 2014, p. 110, 116).

4 The theory of international relations of secession shows that states respond to ethnic conflicts in other states based on their own vulnerability to secession, domestic politics and logic of power. Empirical studies relying on the Minorities and Potential Supporters Dataset also show that countries facing separatism at home are more likely to assist separatists, especially in proximate weaker states and according to their ethnic similarity (see Saideman 2016). Recognition is vital for secessionist entities and also includes the legitimacy conferred by membership or participation in a number of international organisations (Ker-Lindsay 2012, p. 6). Recognition can be bi-lateral or collective. In the latter case, states coordinate and consult with each other, and observe, wait for each other in the process of recognising (Ker-Lindsay 2012, p. 10, 13, Coggins 2014, p. 38). The trend is toward collective recognition, as clearly demonstrated in the case of the former Yugoslavia (see Oeter 2014).

5 Participation in sporting events and associations can often be a source of competition between the parent state and the breakaway territories. Kosovo has been blocked by Serbia in several important sports associations (Ker-Lindsay 2012, pp. 152–155).

6 The full list of interviewees is provided in the supplementary file.