What Killed Yugoslavia?

Social Determinants of Political Collapse

Djordje Stefanovic

A thesis submitted in conformity with the requirements

for the degree of Doctor of Philosophy

Graduate Department of Sociology

University of Toronto

395 Wellington Street 395, rue Wellington Ottawa ON K1A0N4 Ottawa ON K1A0N4 Canada Canada

Your file Votre reference ISBN: 978-0-494-44820-5 Our file Notre reference ISBN: 978-0-494-44820-5

NOTICE: AVIS: The author has granted a non L'auteur a accorde une licence non exclusive exclusive license allowing Library permettant a la Bibliotheque et Archives and Archives Canada to reproduce, Canada de reproduire, publier, archiver, publish, archive, preserve, conserve, sauvegarder, conserver, transmettre au public communicate to the public by par telecommunication ou par Plntemet, prefer, telecommunication or on the Internet, distribuer et vendre des theses partout dans loan, distribute and sell theses le monde, a des fins commerciales ou autres, worldwide, for commercial or non sur support microforme, papier, electronique commercial purposes, in microform, et/ou autres formats. paper, electronic and/or any other formats.

The author retains copyright L'auteur conserve la propriete du droit d'auteur ownership and moral rights in et des droits moraux qui protege cette these. this thesis. Neither the thesis Ni la these ni des extraits substantiels de nor substantial extracts from it celle-ci ne doivent etre imprimes ou autrement may be printed or otherwise reproduits sans son autorisation. reproduced without the author's permission.

In compliance with the Canadian Conformement a la loi canadienne Privacy Act some supporting sur la protection de la vie privee, forms may have been removed quelques formulaires secondaires from this thesis. ont ete enleves de cette these.

While these forms may be included Bien que ces formulaires in the document page count, aient inclus dans la pagination, their removal does not represent il n'y aura aucun contenu manquant. any loss of content from the thesis. Canada What Killed Yugoslavia? Social Determinants of Political Collapse

Djordje Stefanovic Department of Sociology University of Toronto Doctor of Philosophy 2008 ABSTRACT

This dissertation develops an empirically based analysis of the role of culture and key political institutions in the popular support for ultra-nationalists in the process of the

Yugoslav disintegration and its aftermath. The dissertation responds to three key substantive questions. First, is there any validity in the often-repeated claims about the importance of "ancient hatreds" or "clash of civilizations" in the Yugoslav collapse?

Second, did the erstwhile Yugoslav federalism contain or deepen ethnic tensions and conflicts? Third, does the enduring strong popular support for ultra-nationalists in some ex-Yugoslav republics mean that ethnicity has "trumped" class?

To answer these questions, I have used a variety of recent quantitative data

(election results data sets, census results, and survey data sets) as well as historical evidence (internal policy documents, secret diplomatic correspondence, and diaries of officials). On the basis of the statistical and comparative historical analysis, my dissertation arrives at several important findings. First, Yugoslavia was neither undermined by "ancient hatreds" nor torn apart by "the clash of civilizations." Instead, the political elites exploited a tradition of intolerance (especially negative visions of the

Other) formed in the pre-communist period. Second, the poorly designed federalist institutions in Yugoslavia (and other Communist federations) unintentionally undermined

ii political unity and strengthened the nationalism they were supposed to contain. Finally, the rise of the far right did not happen because "ethnicity trumps class." Rather, the appeal of class-specific welfare chauvinism - the demand that only the ethnic majority receives social protection - enabled ultra-nationalists to mobilize economically vulnerable sections of the majority population.

iii TABLE OF CONTENTS

ABSTRACT II LIST OF TABLES VI LIST OF FIGURES VII

CHAPTER ONE: What Killed Yugoslavia? Existing Answers and Their Limitations 1 Introduction 1 Ancient Hatreds or Elite Manipulation? 2 Subversive Institutions? 9 Ethnicity Trumped Class? 11 Overview of the Chapters 13

CHAPTER TWO: Seeing the Albanians through Serbian Eyes: The Inventors of the Tradition of Intolerance and Their Critics, 1804-1939 17 1804 -1839: The Liberation and the Cleansing of Serbia of Muslims 18 Different Visions of National Identity and State Formation 21 1878: A 'Model' Ethnic Cleansing? 25 1878 — 1912: The Persecution of Kosovo Serbs and Hardening of Anti-Albanian Resentment in Serbia 28 The Liberations/Conquests of Kosovo: 1912-1924 34 The Failure of the Colonial Policy, 1924-1929 42 Planning the Cleansing of Kosovo from the Albanians, 1929-1940 45 Assimilate, Deport, or Kill: The Formation of the Cleansed National States at the Balkans 49

CHAPTER THREE: The Unintended Consequences of Ethno- Federalism: How Yugoslav Communists Dug Their Own Graves 56 The Ethno-Federal Dilemma. 56 The Containment Effect of Ethno-Federalism?. 57 The Fragmentation Effect of Ethno-Federalism?. 59 Where the Medicine turned into Poison: The Yugoslav Case. 62 Yugoslavia's Three Waves of Decentralization 63 Yugoslavia as a "Facade Federation" 63 The Confederal Turn in the mid-1960s 64 The 1974 Decentralization: Killing Ethnic Nationalism with Kindness. 66 Effects of Decentralization 68 Incompleteness of Ethnic Republics and Slovenian Exceptionalism 68 Yugoslavia against Yugoslavs 69 Ethnicized Economic Equalization Policy. 72 The Rise of Titular Nationalisms 75 Ethnic Nationalism within the "Vanguard" 76 The Centre's Response to Nationalism: Purge the Sinners,

IV Sweeten the Sin 79 The Withdrawal to Home Republics 80 From Unimaginable to Inevitable: The Disintegration of Yugoslavia 83 The Economic Crisis of the 1980s 83 Failure of Attempts to Escape Ethnic Fragmentation 85 Three Paths Out of Yugoslavia 86 Discussion 89 The Unintended Consequences of Yugoslav Ethnic Federalism 89 Generalizability of Yugoslav Experience? 92

CHAPTER FOUR: The Path to Weimar Serbia? Explaining the Resurgence of the Serbian Far Right After the Fall of Milosevic 95 Introduction 95 The European Far Right: Defenders of 'White Christian' Europe 98 The Crisis of Serbian Society and the Rise of Serbian Radicals 99 Theoretical Explanations of Intolerance and Support for Ultra-Nationalists 103 Ethnic Threat. 103 Economic Vulnerability. 107 Data and Measures 110 Results 113 Discussion 118 Appendix 4A Ethnicity and Support for the Radicals 123 Appendix 4B. Variable Descriptions, Data Sources, and Hypothesized Effects 124

CHAPTER FIVE: Conclusion 125 Understanding the Weight of History 125 The Impact of Ethno-Federalism: How the Yugoslav Communists Dug Their Own Graves 128 Why Does the Public Support Ultra-Nationalists? The Welfare Chauvinism Thesis 130 Conclusion: Better Answers, New Limitations, Comparative Implications 132

REFERENCES 136

v List of Tables

Table 3.1: Ethnic Completeness of the Republics in 1948 and 1981 69 Table 3.2: Per Capita Social Product of the Regions as a Percent of Yugoslav Average 73 Table 3.3: Major Ethno-nationalist Mobilizations in the Yugoslav Communist Party....77 Table 3.4: Economic Factors and Internal Migration, 1971-1981 82 Table 3.5: Concentration in Home Republics by Internal Migration, 1961-1981 83 Table 3.6: Personal Attachment of Yugoslav Citizens to Territorial Organization, May-June 1990 88 Table 3.7: Predictors and Timing of Secessions of Yugoslav Republic and Provinces....89 Table 4.1: Regression of the Percentage Majority Vote for the SRS on Predictors, December 2000 Elections 114 Table 4.2: Regression of the Percentage Majority Vote for the SRS on Predictors, September 2002 Elections 116 Table 4.3: Regression of the Percentage Majority Vote for the SRS on Predictors, 13 June 2004, Presidential Elections (Round One) 117 Table 4.4. Regression of the Percentage Majority Vote for the SRS on Predictors, 27 June 2004, Presidential Elections (Round Two) 118 Table 5.1. Peak Political Performance of Major European Far Right Parties, 1989-2002 131

vi List of Figures

Figure 3.1: Withdrawal to Home Republics 81

vii 1

Chapter One: What Killed Yugoslavia? Existing Answers and Their Limitations

Introduction

The fall of Yugoslavia and the horrific violence accompanying its disintegration have spawned an abundance of publications attempting to explain these tragic developments.

Although laudable in its efforts to respond in a timely fashion to the rising public and political demand for an instant understanding of the region's complex history and politics, much of the literature does not meet scholarly standards. Seeking to address this lacuna in scholarship, my work aims to offer a nuanced and empirically based analysis of the role of culture, key political institutions, and popular support for ultra-nationalists in the process of the Yugoslav disintegration and its aftermath. To this end, I will respond to three key substantive questions. First, is there any validity in the often repeated claims about the importance of "ancient hatreds" in the Yugoslav collapse? Second, did

Yugoslav federalism contain or deepen the ethnic tensions and conflicts? Third, does the enduring strong popular support for ultra-nationalists in some ex-Yugoslav republics mean that the ethnicity has "trumped" class?

While these questions are certainly fascinating for area specialists, the attempt to answer them carries a certain risk. The danger lies in providing responses that might be analytically sophisticated and empirically grounded, but lacking any clear implications beyond this region.1 This danger is well illustrated by the exasperated tone of a recent

1 Indeed, oft-lamented problems of works in the area studies include lack of solid theoretical foundations, poor knowledge of statistical methodologies, and a tendency to "wallow in the facts" of a favorite case study without abstracting any useful generalizations (Shea 1997: 4-5). 2 observation made by two comparativists in their review of the research on the Bosnian war:

There are many rival explanations of the onset of civil war in Bosnia. Most explanations cannot fit neatly in a theoretical framework that tries to explain more than just Bosnia. Reading case studies or reports on the war, it is hard to know what we might learn from Bosnia that we can generalize to other wars. (Kalyvas and Sambanis 2005: 191)

To deal with the issue of generalizability, my dissertation approaches the substantive question in such a way as to address the major international scholarly debates. What makes the Yugoslav case(s) worth studying sociologically is that it meets Ragin's criteria for a good sociological case - it has significance for theory-development (Ragin 1992:

15). Rather than trying to represent a comprehensive alternative account of the Yugoslav collapse, the dissertation focuses on specific puzzles. Yet it selects those with an empirical payoff well beyond this particular case study. My approach to the question of generalizability thus follows Bunce's argument, made in response to the criticism of

"area studies" in general, namely, that we do not have to choose between "knowing places and using theory" (Bunce 1999: 163). In other words, the study not only answers the key questions in a way that the area specialist will find useful, but it helps to advance comparative and theoretical understanding of the three main sets of causes of ethnic conflict: culture, political institutions, and class politics.

Ancient Hatreds or Elite Manipulation?

Early interpretations of the Yugoslav conflicts are mostly culturally determinist arguments about the "explosion" of "ancient hatreds" that were "unleashed" by the fall of communism. Primordialists claim the wars in the former Yugoslavia resulted from the 3 explosion of simmering "ancient hatreds," the "fierce social character" of some ethnic groups, or the inevitable "clash of civilizations" of different ethno-religious groups

(Huntington 1996; Kaplan 1993; Mestrovic 1993). According to this perspective, numerous historical instances of massive inter-ethnic violence show that the Balkans is a violence-prone region (Cohen 2001: 380). The only extended periods of inter-ethnic peace and stability were produced by repressive, authoritarian regimes: the Ottoman

Empire, the Austrian Empire, the Yugoslav royal dictatorship, and the Yugoslav communist dictatorship. According to some primordialists, the communist "super-ego's" harsh repression of the primordial nationalist "id" only increased the frustration, and, inevitably, the intensity of the "explosion" once the repression was removed (Brubaker

1998: 285-286). For example, Henry Kissinger argues that "for the Bosnians, the overwhelming reality is their historical memory, which has sustained their ineradicable and unquenchable aspirations for centuries... The deep-seated hatred of each party for all the others exists" (Kissinger 1997:19).

Samuel Huntington has developed probably the most influential primordialist explanation of ethnic violence. In his controversial book, Clash of Civilizations, he argues that in a post-Cold War world, the obsolete paradigm of the three worlds needs to be replaced with a paradigm of civilizations (Huntington 1996: 13). A "civilization" is the most encompassing cultural group, short of the human race: in his definition, civilizations roughly correspond to the major world religions (43). Due to fundamental value differences, "the relations between groups from different civilizations ... will almost never be close, usually cold, and often hostile" (207). Thus, countries that contain

2 While the "Clash of Civilizations" argument was generally rejected as cultural determinism at its worst by many academics who study ethnic conflicts in the 1990s, the same argument and Huntington's terminology are now widely embraced by a number of politicians and journalists. 4 large groups associated with different civilizations (such as the former USSR, the former

Yugoslavia, Nigeria, Sudan, India, or Sri Lanka) will experience huge "internal strain" and become "good candidates for dismemberment" (28).

In Huntington's terminology, the violent dismemberment of such countries and other clashes between groups or states from different civilizations are "fault line conflicts" (Huntington 1996: 253). Fault line conflicts are rooted in memories of sustained and ongoing cultural distance, antagonistic relations, intense hatreds, and mutual violence (Huntington 1996: 291-2). This primordialist perspective tends to produce a sense of hopelessness and inevitability about the violence. For example, the former US Secretary of State, Warren Christopher, argues that the war in Bosnia was "an intractable problem 'from hell' that no one can expect to solve ... less a moral tragedy... and more a tribal feud that no outsider could hope to settle" (Friedman 1993:1).

A number of scholars have criticized the culturally deterministic account of ethnic violence. As they note, primordialists have a serious problem explaining "realistic" conflicts between groups, such as struggles over scarce resources (for example, land and oil), as these conflicts are not motivated by prejudice or cultural animosity. The primordialists' focus on cultural factors makes these types of conflicts practically unintelligible. Second, there is empirical evidence suggesting that identities and animosities are actually modifiable. Huntington is unable to account for long periods of peaceful coexistence and extensive intermarriage between the Serbs and Muslims in

Bosnia (Huntington 1996: 254), as it flies in the face of his theoretical argument.

3 Huntington's assumption that ethno-cultural differences will necessarily lead to different policy preferences seems overly deterministic. Even if they do, in some cases, there is still no reason to assume that these differences are necessarily irreconcilable. 5

Furthermore, even in cases where there is considerable cultural animosity, the primordialist argument fails to explain the timing of the ethnic violence. For example, from 1988 to 1998, the ethnic Albanian minority in the Serbian province of Kosovo was subject to systematic discriminatory policies and police persecution. A variety of indicators, such as inter-ethnic marriage rates, residential segregation, and migration patterns indicate high levels of ethnic distance between Kosovo Albanians and Kosovo

Serbs since at least the 1960s. Still, because of Serbian military superiority and the lack of international support, the Kosovo Albanian leadership opted for the non-violent and gradual pursuit of independence. Thus, even a combination of cultural animosities with direct discrimination and repression will not necessarily "trigger" an escalation. Cultural animosities are neither necessary nor sufficient for ethnic violence.

Responding to the weaknesses of the primordialist perspective, the instrumentalist school has gained ground since the mid 1990s. Instrumentalists argue that nationalistic leaders use hate propaganda to manufacture ethnic insecurities and conflicts. In the

Balkans, for example, they argue that Balkan peoples have a long tradition of multiethnic coexistence (Donia and Fine 1994). Instrumentalists assume that the traumatic memories of previous short episodes of inter-ethnic conflicts had healed by the 1980s (Malcolm

1998: 6). In their view, the recent violence is an aberration caused by Machiavellian political leaders, in particular, Serbia's President Milosevic and Croatia's President

Tudjman, who promoted diversionary ethnic wars as a way of deflecting demands for genuine democratization and social justice (Gagnon 2004). They imply that the "normal" pattern of inter-ethnic relations in the region is not "barbaric" violence, but multiculturalism and inter-religious cooperation. The instrumentalist perspective tends to 6 be optimistic about the future prospects of the region. It assumes that once these politicians and their mass media "factories of hate" are removed, the region will return to its "normal" pattern of inter-ethnic cooperation.

The instrumentalist approach has certain flaws. With its focus on scheming

Machiavellian politicians, it fails to explain why "ordinary people" voted for these politicians for so long. Nationalist mobilization projects can succeed only if their message resonates with enough people. The approach also fails to explain the variability across time and space in the success of the nationalist appeals (Brubaker 1998: 291, 304).

Instrumentalists do not explore whether the radical nationalists projects actually reflect deeply held and genuinely felt popular intolerance of the ethnic Others. While the instrumentalist perspective improves upon the simplistic historicism of primordialists, then, it introduces new problems.

Like primordialists, instrumentalists tend to read history selectively. While primordialists focus on the episodes of inter-ethnic violence to support the "ancient hatreds" thesis, instrumentalists tend to ignore such episodes, focusing instead on periods of inter-ethnic cooperation. Since the history of inter-ethnic relations in the Balkans is long and complex, both schools can find supportive examples for their respective theses.

However, in both cases, such selective use of historical evidence is methodologically indefensible.

I will use an alternative theoretical approach that represents the Balkans as neither more violence-prone nor more genuinely multicultural than the rest of Europe.

Drawing on the major comparative historical works of Charles Tilly and Miroslav Hroch 7 on the formation of national states in Western and Eastern Europe, I will analyze the

Balkan cases in the context of wider European processes.

In his analysis of the formation of national states in Western Europe, Charles Tilly

(1975: 43-4,49, 78) notes that almost all successful states (i.e. those that survive) have implemented some form of deliberate cultural homogenization policy, such as the adoption of a state religion, the expulsion of minorities, the institution of a national language, or educational standardization. While Northwestern Europe was gradually culturally homogenized, the Ottoman system of non-territorial autonomies and tolerance for religious minorities effectively ensured that Southeastern Europe remained without clear ethno-cultural boundaries.

As the early Balkan national states expanded into the political vacuum left by the collapsing empires, they encompassed ethnically diverse - and potentially irredentist - populations. The continued presence of these minorities was incompatible with the coercive Utopia of a centralized and homogeneous nation-state. Alternative roads to modernity were beyond imagination. Thus, minorities and ethnic pluralism had to become victims of the Utopia. As Maria Todorova observes:

... the Balkans were becoming European by shedding the last residue of an imperial legacy, widely considered an anomaly at the time, and by assuming and emulating the homogeneous European nation-state as the normative form of social organization. It may well be that what we witness today [in the 1990s], wrongly attributed to some Balkan essence, is the ultimate Europeanization of the Balkans. (Todorova 1997: 13)

While this line of reasoning captures some of the reasons for exclusionary ideas and policies in the Balkans, analysis of historical evidence shows major flaws in the ideology-centered explanation. My first empirical chapter therefore develops a comparatively informed and historically grounded analysis of the traditions of 8 intolerance, while avoiding the selective use of history that has plagued primordialism and instrumentalism.

In order to analyze the role of history in the formation of ideology and the practice of exclusion, I have focused on the formation of the Serbian tradition of intolerance towards Albanians. In the regional context, the history of Albanian-Serbian relations is particularly troubled. One would be hard-pressed to find any instances of major inter-ethnic conflicts between Serbs and Croats in Croatia before the 20th century.

Similarly, the relations between different groups of Bosnians were remarkably good in the major urban areas in the late communist period, with high levels of intermarriages and the gradual emergence of a shared Yugoslav identity. But in the case of Albanian-

Serbian relations in Kosovo province, even in the best days of federal Yugoslavia, intermarriage was practically non-existent, ethnic distance was considerable, and almost nobody self-declared as "Yugoslav only."4 Furthermore, immediately after the death of

Tito, Yugoslavia's communist authoritarian leader, there were massive ethnic Albanian demonstrations in Kosovo against the Yugoslav state, followed by a crackdown by security forces.

Recent ethnographic research shows that in the early 21st century, the nationalist

"historical" narratives of Kosovo Serbs and Albanians tended to emphasize a long history of victimization by the other side (Zdravkovic 2005). In his recent comparative historical analysis of ethnic violence in 20th century Eastern Europe, Petersen concludes that while

4 Geographic distribution of ethnic groups as reported in the 1981 Yugoslav Census indicates that pre-war Yugoslavia, Bosnia in particular, was a very complex society, with major urban/rural and regional differences in inter-ethnic composition (Goldstein 1986: 315). Yugoslav identity was clearly emerging in Vojvodina, parts of Croatia, urban Bosnia, and coastal Montenegro. At the same time, in the areas of contact between Serbs and Albanians (Kosovo), Albanians and Macedonians (western Macedonia), and Serbs and Muslims (eastern Bosnia), it appears that Yugoslav identity was hardly present. 9

"ancient hatreds" do not seem to drive many conflicts, the Serbian expulsion of

Albanians from Kosovo in 1999 is a case with "an excellent fit" (2002: 255). Indeed, it would seem that if there is a place in the Balkans where the ancient hatreds hypothesis has some validity, it is Kosovo.

Subversive Institutions?

Why did the Yugoslav federal institutions fail to stop the escalation of ethnic tensions into violence? The failure of Yugoslav federalism, followed by the collapse of the

Czechoslovak and Soviet federations, triggered a major international debate on the long- term sustainability of the federalist solution of the national question. While some scholars argue that federal institutions can contain ethnic violence, others are increasingly skeptical.

The federalist camp includes Michael Hechter, the celebrated author of Internal

Colonialism, and a senior proponent of the rational choice theory. Hechter's recent book,

Containing Nationalism, argues that political decentralization (primarily by means of introducing territorial ethno-federalism), combined with a reduction in the ethno-cultural division of labor, might be the best way to contain nationalist demands for sovereignty and, hence, to reduce the potential for nationalist conflict. Hechter is addressing a major international debate whose implications reach far beyond the Balkans. The separatist movements found in a number of countries today could conceivably lead to violent ethnic partitions. Hechter's argument might offer the hope of a peaceful resolution of ethnic tensions and conflicts in such places as Quebec, Northern Ireland, Corsica, the Basque 10 region, Kosovo, Western Macedonia, Republika Srpska, Transylvania, Crimea, Iraq,

Chechnya, and Tibet, to name only a few.

Hechter supports his argument for ethno-political decentralization by means of a comparative historical analysis of the spread of nationalism in Europe and a quantitative analysis of some recent cases of nationalist rebellions worldwide. He claims that the spread of direct rule (political, legal, and fiscal centralization) and state nationalism

(loyalty to the "nation-state" and forced cultural homogenization) in Europe triggered the

"peripheral nationalism" of minorities who resisted incorporation or attempted secession

(Hechter 2000a: 71). Hechter's quantitative analysis indicates that Yugoslavia is a highly decentralized country mostly without national conflicts. Clearly aware that this claim is likely to raise eyebrows in light of the bloody wars of the 1990s, Hechter argues that the relationship between decentralization and nationalist violence is likely to be U-shaped rather than linear (Hechter 2000a: 148). In the Yugoslav case, "extreme decentralization" made it difficult for the federal government to adapt to exogenous economic shifts and courted fragmentation (Hechter 2000b: 325-5). More generally, "[i]f too little decentralization causes rebellion, then too much is likely to engender fragmentation," and this can be "another source of intergroup violence" (Hechter 2000a: 151-2)

Contrary to Hechter's thesis, several leading regional specialists have argued that the introduction of ethnic federalism has unintentionally undermined communist states in

Eastern Europe. According to Bunce, "The irony of the collapse of socialism, then, was that the very institutions that had defined these systems and were there, presumably, to defend them as well, ended up functioning over time to subvert both the regime and the state" (Bunce 1999: 2). Similarly, Brubaker comments on the fate of the USSR's attempt 11 to create a unified Soviet people through ethno-federalism: "The annals of unintended consequences are rich indeed, but seldom have intention and consequence diverged as spectacularly as they did in this case" (Brubaker 1994: 49).

Instead of simply endorsing or rejecting ethno-federalism, in Chapter Three, I seek to identify specific conditions under which ethno-federalism leads to fragmentation instead of containment. Departing from Hechter's insight that limited doses of decentralization can contain nationalism, but that excessive decentralization leads to fragmentation, I identify specific institutions and processes that had disintegrative effects in the Yugoslav case (1945-1991). Unlike Hechter who focuses on the fiscal decentralization, I assume that the decentralization has a plurality of dimensions that cannot be easily expressed as a single continuous variable. Going beyond Hechter, therefore, I seek to identify specific ethno-federal institutions and processes that had disintegrative effects. An in-depth analysis of the Yugoslav case is justified by its critical position in Hechter's analysis. As Hechter himself argues, the excessive Yugoslav decentralization unintentionally turned ethno-federalism from helpful medicine into the poison of fragmentation.

Ethnicity Trumped Class?

When the introduction of the multiparty system finally enabled the Yugoslavs to choose their own leaders, many citizens voted for belligerent ultra-nationalists. The ability of ultra-nationalists to capture power and control the coercive apparatus of the state led to the escalation of inter-ethnic conflicts into military confrontations and ethnic cleansing.

Why did the Yugoslavs contribute to their own misfortune by voting for such parties? 12

The troubling vulnerability of democratizing polities to an upsurge of nationalism has met with increased attention from comparative social scientists. Snyder, for one, argues that recently democratised regimes are exceptionally conflict-prone, especially if an authoritarian elite facing strong democratic challengers tries to de-mobilize the democratic constituency by pursuing nationalist mobilization for diversionary wars

(Snyder 2000:. 28-9, 198; Mansfield and Snyder 2005: 29). Inter-ethnic warfare can produce strong domestic nationalist and militarist institutions and ideologies, as well as lasting external rivalries. In the end, repeated failed democratisations, strong ultra- nationalist institutions, and recurring inter-ethnic warfare can become a vicious circle

(Mansfield and Snyder 2005).

Some scholars are inclined to explain the strength of the vote for the ultra- nationalists as a sign of a general decline in the significance of class.5 Ivarsflaten argues that blue-collar workers who vote for the far right are "discounting" their own economic interests (Ivarsflaten 2005: 468). This should particularly be the case in post-communist

Europe, where the weakness of socialist ideas has supposedly de-legitimised class cleavages in favour of "identity-based" cleavages (Ost and Crowley 2001: 229).7 There is some empirical support for the "insignificance of class" thesis. For example, there was no statistical association between party choice and social class of voters in the first post- communist elections in Hungary (Szelenyi et al. 1997: 209).

5 These interpretations echo the New Political Culture thesis of Clark and Lipset (1991, 1993) on the declining political significance, or even "dying" of classes. 6 Ivarsflaten's (2005) analysis focuses on the New Radical Right Parties that combine racism with economic neo-liberalism. She des not address the welfare chauvinist ideological formula of other far right parties. 7 Since class can also be identity (in the sense of class for itself), these authors are probably referring to non-class identities, such as ethnicity, religion, or gender. 13

However, others point out that voting for far right ultra-nationalists might be a new form of class politics. Kitschelt argues that individuals with resources that could be economically beneficial in a market economy will tend to support market transformation, while those who are more dependent on extensive welfare state policies will vote against pro-market parties. That is, potential winners (professionals, small business people, the well-educated, the urban, and the young) are more likely to vote for pro-market parties than are potential losers (pensioners, the poorly educated, and unqualified workers)

(Kitschelt 1995a). Kitschelt goes so far as to predict that in economically depressed, post- communist Europe, "welfare chauvinism" (targeting social services to the ethnic majority) might be an even more successful strategy than in Western Europe (Kitschelt

1995b: 23).

But can the welfare chauvinism thesis account for plural support for ultra- nationalists in the post-Yugoslav region? Or should we conclude that class politics has been eclipsed by ethnic conflict? Chapter Four provides an empirical response to these questions.

Overview of the Chapters

This Chapter is intended to serve as a brief introduction to the dissertation. Chapter Two, the first empirical chapter, focuses on the role of distant events and historical processes on the Yugoslav inter-ethnic relations. To investigate whether a path dependency exists in the formation of exclusionary ideology and policy and how it might operate if it does, I conduct a theoretically informed historical case study of a particularly difficult inter- ethnic relation, characterized by recurring violent confrontations. In order to analyze the 14 role of history in the formation of ideology and the practice of exclusion, I focus on the formation of the Serbian tradition of intolerance towards the Albanians.

Methodologically, my focus is on identifying formative periods, turning points, and causal mechanisms which maintain the process over time. In terms of evidence, my work combines the use of existing historical studies with my own primary research into such sources as diaries and biographies of leading participants in the historical events, as well as diaries of Western travelers in the region. My main finding is that the nationalist elite's vision of the "essential" characteristics of the Other determined which policies towards them would appear realistic and feasible. Those minorities whose "essential" character was perceived as culturally distant (in terms of religion, language, and socio-economic development) and physically threatening (in terms of the previous history of conflict and the current presence of major foreign supporters) were more likely to be perceived as logical candidates for deportation or murder.

In Chapter Three, I focus on the failure of the key political institutions to prevent the escalation of inter-ethnic conflicts. Both social movement theory and Hechter's argument on U-shaped relationship between decentralization and ethnic conflict imply that excessive forms of decentralization can have a disintegrative effect. This chapter identifies a number of specific institutions and processes that had disintegrative effects in the Yugoslav case. In specific, carving a country into a set of supposedly ethnic republics with multi-ethnic populations can lead to institutionalized majority racism against minorities. Ethnic equalization policies can buy the loyalty of less developed regions but trigger the secessionism of the wealthy, while ethnic equalization policies based on ethno-territorial autonomies can lead to ethnification of the redistribution policy. A 15 central government which repeatedly responds to secessionist pressures with decentralization can create a never-ending cycle of decentralization and nationalist mobilization. And attempts to accommodate secessionists by increased ethnification of federal institutions and reduced representation of non-ethnic cleavages can unintentionally eliminate institutions with the potential to transcend ethnic cleavage, thus creating systemic paralysis. As the Yugoslav case shows, certain forms of ethno- federalism can empower secessionist nationalists to break up a country.

Chapter Four seeks to explain the strength of the popular support for the extreme nationalists. The chapter evaluates the validity of the thesis of the "insignificance" of class politics in post-communist Europe by analyzing the quantitative evidence on the social support for the Serbian far right. This study analyses the results of five post-

Milosevic elections and finds support for the argument that the majority's economic vulnerability and a sense of "ethnic threat" are the major contextual predictors of far right support. In short, class-specific welfare chauvinist appeal has enabled ultra-nationalists to mobilize economically vulnerable sections of the majority population.

In Chapter Five, the concluding chapter, I draw together the three strands of the argument and suggest directions for future research. I argue that Yugoslavia was neither undermined by "ancient hatreds" nor torn apart by "the clash of civilizations." Instead, the political elites exploited a tradition of intolerance (especially negative visions of the

Other) formed in the pre-communist period. The poorly designed federalist institutions in

Yugoslavia (and other communist federations) unintentionally undermined political unity and strengthened the nationalism they were supposed to contain. Finally, I indicate that the rise of the far right did not happen because "ethnicity trumps class." Rather, the 16 appeal of class-specific welfare chauvinism - the demand that only the ethnic majority receives social protection - enabled ultra-nationalists to mobilize economically vulnerable sections of the majority population.

I conclude the dissertation by pointing to possibilities for future research. In order to establish the exact effect of the tradition of intolerance on the Yugoslav disintegration, it is necessary to develop an historical analysis that will connect the formative period with the escalation of the Albanian-Serbian ethnic conflict in the early 1980s. And to determine whether the negative effects of specific ethno-federalist institutions occur outside the post-communist region, it is necessary to test these hypotheses on cases from other regions. Finally, the generalizability of the welfare chauvinist appeal can be established only by testing this hypothesis as an explanation of the variability in support for the FR in post-communist Europe.

The empirical chapters of this dissertation are written in article format, with the intention of being publication-ready upon defense. Chapter Two is published and Chapter Four is accepted for publication. 17

Chapter Two: Seeing the Albanians through Serbian Eyes: The Inventors of the Tradition of Intolerance and Their Critics, 1804-1939

The emergence of the ideologies of'moral murder' is one of the conditions for the escalation of ethnic conflicts into ethnic cleansing or a genocide (Mann 2001: 237). By studying the social construction of Albanians in Serbian political discourse and the corresponding state policies, this article fills some of the gap in the existing knowledge on the treatment of Albanians by the Serbian and Yugoslav state in the period before the

Second World War (Gallagher 2001a: 151). The troubled incorporation of the Albanian minority into the Serbian state, when compared with similar nation-state building processes in the Balkans, enables us to reassess and extend Hroch's theory of the rise of minority nationalisms in Eastern Europe.

An elite's visions of minority populations and state policies cannot alone explain the opening or closing of the opportunities for ethnic cleansing.9 However, they do enable us to understand why and how the political elite might try to pursue such opportunities.

The evolution of the Serbian vision and policies towards the Albanians was influenced by wider European nation-building models, as well as by the clientelistic relationship with some of the Great Powers. The Serbian elite faced a fundamental tension between an uncritically accepted model of the homogeneous national-state and the complex multi ethnic reality of the Balkans. Since the entrenched Serbian visions of the Albanian

Muslims were particularly negative, the Serbian elite applied the most exclusionary policies towards the Albanians.

9 Ethnic cleansing refers to the removal of members of an ethnic group from a territory against their will. For a detailed discussion of the concept, see Mann 2005: 10-18. 18

1804 -1839: The Liberation and the Cleansing of Serbia of Muslims

The First Serbian Uprising (1804-1813) created the first Serbian political authority that had to deal with a Muslim question on its territory. 10 The Uprising was initially a local rebellion that only gradually escalated into a struggle for national independence from the Ottoman Empire. The immediate goal of the rebels was to end the arbitrary and repressive regime of Dahije, n but they also felt a preexisting animosity towards Muslims. Matija Nenadovic, one of the leaders of the uprising, wrote down the last message he received from his father — who was taken as a prisoner and soon murdered during the execution of the notables:12 'Say to him [Matija] that neither he nor any of my people should ever again believe the Turks' (Nenadovic 1969: 50). Extant class hatred of the Serbian peasants towards urban Muslim merchants and nobility was clearly a major motivator for mass violence. 13 Nenadovic describes the take-over of the

Valjevo township by the Serbian rebels:

The initial treatment of the local Muslims during the formative phase of the Serbian state has crucial significance for the later visions and treatment of the Muslim Albanians. Since the Ottoman millet system entrenched the central importance of the religious identity for social position, the Serbian elite members of this period would frequently refer to all Balkan Muslims as the 'Turks' regardless of their actual ethnic background. Petar Petrovic Njegos, the ruler of Montenegro, writes in a 1843 letter 'We [the Christian Orthodox Montenegrins] are now at war with all the Turks [sa svijem Turcima] around us, and that is almost always the case. They are always united against us [slozni na nase zlo], both the Bosniacs and the [Muslim] Albanians. (Letter to Jeremije M. Gagic, August 12, 1843, (Njegos 1967: 84). It appears that some Balkan Muslims also saw themselves as the 'Turks' in this sense. In a work published in 1849 Vuk Karadzic argued that many 'Serbs of the Turkish faith' [Srbi turskoga zakond] need to receive proper education so that they will 'learn and recognize' that 'they not the Turks, but the Serbs' (Karadzic 1972 [1837]: 119). 11 Dahije were a group of Ottoman janissaries who managed to usurp control of the Belgrade region, while ignoring the orders of the central government in Istanbul. 12 In the hope of pre-empting the likely Serbian rebellion, the Dahije ordered the execution of the Serbian notables and village chiefs in 1804. 13 Pavlowitch (2002: 28) estimates the pre-1804 population of the Belgrade Pashalik at 400,000. About 10% were mostly urban Muslims, including about 900 feudal lords. 19

At that time in Valjevo there were twenty-four mosques and it was said that there were nearly three thousand Turkish and some two hundred Christian houses.... Any house that had not been burnt the Serbs tore to bits and took their windows and doors and everything else that could be removed (Nenadovic 1969: 7).

Nenadovic indicated that the Serb rebels had a clear awareness of the Kosovo myth, 14 but at this point the awareness did not seem to transfer into any clear territorial agenda per se. Nenadovic's text has occasional references to the epics of the Kosovo battle. Still, he seems to interpret them as a symbolic and inspirational story of a 'golden age', its heroes, and the fall of a civilization, rather than as a call for the 'liberation' of

Kosovo. A major Serbian national intellectual and participant in the Uprising Vuk

Karadzic, even wondered whether the whole of Kosovo 15 should be seen as a part of

Serbia or Albania.16 Furthermore, independence from the Ottoman Empire was initially not even proposed, not only because it was clearly beyond reach without major support from a Great Power, but also because a number of leaders could still remember a fairly good life before the Dahije and under the rule of the Hadzi Mustafa Pasha. Kardorde

Petrovic and other rebel leaders escalated their goals and the Serbian forces began to attack all the 'Turkish' settlements, not just the Dahije and their allies .

In 1389, the Ottoman Imperial forces defeated a Christian alliance led by the Serbian Prince Lazar Hrebljanovic on Kosovo polje. The defeat led to the fall of the Serbian Empire and advance of the Ottoman Empire into Europe. Serbian folk epic and national historiography portray the Kosovo battle as the central event in the Serbian history - the end of the Golden Age and beginning of the slavery under the Ottoman yoke. An example of this genre is Petrov 1991: 43- 64. 15 In this text, 'Kosovo' refers to the territory of the present day Serbian Province of Kosovo whose 'final' status at this time remains hotly disputed. 16 Specifically, Karadzic (1972c [I860]), 10) wonders whether the south-western part of Kosovo, that Serbs call Metohija, should be seen as part of Serbia. 17 Pavlowitch explains that the escalation happened following the Ottoman-Russian peace settlement in Bucharest and the failure of the Serbian-Ottoman negotiations. Presumably, the Serbian rebels expected an imminent Ottoman attack and wanted to consolidate their hold on the territory (Pavlowitch 2002: 30). Similarly, during the Second Uprising in 1815, Serbian leader Milos Obrenovic ordered the rebels to attack anyone wearing green, i.e. all the Muslims (Mazower 2000: 89). 20

Dositej Obradovic, an educated and influential Serb who established the first secular school in free Serbia in 1808 and became Serbia's first Minister of Education in

1811, pleaded in vain to the leaders of the Uprising not to attack the 'peaceful Turks.' In a letter written to the leaders in 1806, Obradovic acknowledged that the Turkish soldiers and feudal lords could not be allowed to stay in the Serbian state. However, he also urged that:

...all those Turks who supports themselves and do not want to rule over the raya [Christians] you should allow to live peacefully in their homes, like they did before, just put them under your rule....[a]nd spread among them your language, knowledge, and literature (Stojancevic 1980: 242).

Despite such humanistic pleas, the First Uprising and several other failed

Christian uprisings in the period of Ottoman decline seem to follow a pattern of escalating violence. Oppression and lawlessness triggered an uprising that escalated into killing, plundering, and burning of Muslim settlements. The Ottoman reinforcements

soon arrived, crushed the rebels, committed various crimes, burned villages, and re

established Ottoman rule. 18 It is significant that the Serb rebels were trying to permanently cleanse the Muslims, while the Ottoman forces engaged in exemplary

repression, aimed at teaching the Serbs a lesson but without driving them out through

permanent mass expulsion into the Austrian Empire. Why did their actions differ?

For example, following the failure of the 1814 rebellion, about 5,000 Serb rebels believed an Ottoman promise of amnesty and surrendered. Instead, the Ottoman forces enslaved them and sent them away to Istanbul (Karadzic 1972 [I860]: 230). They were lucky, relatively speaking, as about 40 other captured rebels were impaled (Karadzic 1972 [I860]: 268). Karadzic did note, and further Serbian intellectuals will not forget that the Kosovo Albanian Muslims came to help the Ottoman forces (Karadzic 1972 [I860]: 280). On later interpretations, see Markovic 1991: 211. 19 Following the Austrian occupation of Belgrade Pashalik 1788-1791, the Ottoman authorities tried to prevent further Serbian mass emigration by proclaiming a general amnesty and delegating tax collection to the local self-government (Pavlowitch 2002: 28). Prior to the First Uprising, the local Serb leaders occasionally threatened the Ottoman authorities that they would go to Germany (i.e. Habsburg Empire) unless they were treated better. 21

Arguably there was a class hatred, as the Ottomans perceived the Christian peasants as a source of tax revenue and forced labour, while the Serb peasants saw the mostly urban Muslims as a generally useless class of'gluttons' [izjelice] as some of the

Muslims were referred to in the Serbian folk epic song that describes the First Uprising

(Djuric 1977: 651). Another factor leading to the Serbs' eliminationist attempts might be the inspiring influence of Austrian policies towards the Muslims in the newly acquired territories. As Ivo Goldstein (1999: 46) puts it, traces of signs of Ottoman rule

'disappeared as mosques and other structures were pulled down...' In addition, the

Muslim presence in the cities was justifiably perceived as a source of internal threat in

01 the event of an Ottoman counter-attack. The perception of urban Muslims as the economic exploiters and as a major security threat was not auspicious for the other

Muslims that would soon be included in the expanding Serbian state.

Different Visions of National Identity and State Formation.

In the emerging Serbian national state there were also proponents of religious tolerance. One of them was Ilija Garasanin (1812-1874), who served as Minister of

Internal Affairs and then Minister of Foreign Affairs (Jovanovic 1932: 215-34). While

Garasanin's 1844 Nacertanije explicitly defines the re-establishment of the Serbian

Serbian historian Rados Ljusic (1994: 17) also argues that the Serbian rebels' actions against the local Muslims were motivated by a combination of class tension and religious intolerance. 21 Thus, the Serbian leaders managed in 1834 to negotiate the withdrawal of the Muslim urban population from Serbia. Following the 1862 clashes, all Turkish civilians were evacuated and Serbia had to pay annual compensation to the Porte for their holdings. Serbs coming fromth e remaining Ottoman provinces moved to Serbia to settle the lands throughout the period (Pavlowitch 2002: 33, 54). 22

Empire as the primary national goal (Garasanin 1998: 75), he simultaneously advocated religious freedoms, in order to make a union with Serbia more attractive for

Bosnian Catholics and Muslims (Garasanin 1998: 91):

... Serbia has to be the natural protector of all the Turkish Slavs, and if she genuinely takes this role upon herself, the other Slavs will allow her to speak and act in their name. If Serbia showed to her neighbors that she only thinks of herself, and ignores the ill fortune and needs of others, they would not listen to her and thus instead of unity there would be mistrust, envy and misfortune (Garasanin 1998: 93).

Garasanin advocated that Serbia should gain access to ~ not occupy/liberate — some of the Albanian sea ports (Garasanin 1998: 96). This would imply that he does not see Albania as a future part of the Serbian state. Interestingly, Garasanin does not mention Kosovo once in his Nacertanije. This is particularly important in light of the more recent nationalist claims of the primeval sanctity of Kosovo as the 'sacred Serbian land'. Looking at the writings of Nenadovic, Karadzic or Garasanin, it does not appear that Serbian intellectuals and politicians embraced a religious claim to Kosovo until very late in the 19th century or even early in the 20th century.

In 1847 The Mountain Wreath by Petar Petrovic Njegos was published. It is a work of indisputable literary value, while still containing passages which reflect his negative attitude towards the Tslamicized' Slavs. As Njegos put it, they are the 'disgusting traitors' \pogani izrodi] who 'denied the faith of the forefathers' (Njegos 1978: 6, 28).

Their existence was thus incompatible with a Christian Montenegrin state: the Muslims a

'snake within our breast', a 'devil in our Christian land' (22). For their own good, they

22 Serbian nationalists frequently viewed the medieval Serbian Empire, which existed for a brief period of time before the Ottoman invasion, as the golden age of Serbian culture and history. 23 Petar Petrovic Njegos (1812-1851) was the Prince-Bishop of Montenegro from 1830 until 1851. During his rule Montenegro was basically a tribal alliance of Orthodox Christian groups that remained outside the Ottoman rule. 23 must return to their authentic, original, 'European' path (62). Thus, the solution of the

Muslim question was supposed to come through coercive conversion.

However, those who refused to convert wasted their chance to join the nation. For their ~ repeated — treacherous behavior they would be punished with expulsion or murder. Montenegrin land must be 'cleansed' of the non-Christians [ociscena od nekrsti]

(6). As the Montenegrin Bishop in the poem proclaimed, 'those who blasphemed the

Holy Name of Christ we will baptize, with water or with blood!' (50) Building on the tradition of intolerance, Njegos dedicated The Mountain Wreath to the Serbian leader

Karadorde, who 'cleansed' Serbia of Muslims during the First Uprising. Thus, he attached his own work to the pre-existing eliminationist ideas and practices.

The inspiration for Njegos's work was in actual historical developments. After the

Ottoman defeats in Central Europe and general decline of the Ottoman power in the second half of the 18th century, Montenegrin peasants began organized attacks on the local Slavic Muslim settlers, who gradually withdrew into the lowlands and the urban areas (Erdeljanovic 1978: 63, 65, 82). Njegos's poem probably describes a major

Montenegrin attack on the local Muslims that seems to have occurred on Christmas Eve of 1709 (Erdeljanovic 1978: 79). While some Muslims managed to escape, many Muslim men were killed, and their women and children were coercively converted (Karadzic

1972a [1837]: 23-4).

Still, it is important to avoid simplistic or singular interpretations of Njegos's work and life. Calling Gorski Vijenac — one of the most complex and finest literary works in the Montenegrin and Serbian literature ~ a 'hymn to genocide' — is about as fair as labeling Shakespeare's Merchant of Venice — an 'ode to anti-Semitism.' As the 24 political and religious leader of Montenegro, Njegos worked to reduce the tensions between the Christian Montenegrin tribes and the neighboring Muslim populations. 24

Moreover, in his actual diplomatic correspondence with Osman-Pasha Skopljak, one cannot fail to see Njegos's profound individual struggle to come to some meaningful reconciliation with the Slavic Muslims. Njegos writes how much he misses all the Serb nobility that converted to Islam, how hard it is for him to see his Montenegrins at constant war with their own 'islamicized brothers' [bracorn isturcenom], that he would love 'more than anything else in the world' to see the unity [slogu] among the brothers.

He writes to Skopljak, who was a Bosniac:

When you speak to me as my brother Bosniac, I am your brother and your friend, but when you speak as an alien, as an Asian, as an enemy of our people and our name, I cannot accept that [menije to protivno] and no reasonable man could accept that.

Despite Njegos's genuine desire to think of a way to incorporate the Slavic

Muslims into the Serbian/Montenegrin state, the central element of his thinking remains that the Muslims — seen as a residue of the Ottoman and 'Asiatic' imperial subjugation ~ could not be allowed to exist in the liberated, European, and Christian Balkan states.

Tragically, his work fuses the commitment to the national liberation with extreme intolerance towards the Balkan Muslims.

See, for example, his Letter to the Turks of Podgorica and Spuze, sent on April 27, 1832 (Njegos 1967: 46). 25 From his Letter to Osman-Pasha Skopljak, October 5, 1847 (Njegos 1967: 153). 25

1878: A 'Model' Ethnic Cleansing?

In 1878, following a series of Christian uprisings against the Ottoman Empire, the

Russo— Turkish war, and the Berlin Congress, Serbia gained complete independence, as well as new territories in the Toplica and Kosanica regions adjacent right next to

Oft

Kosovo. These two regions had a sizable Albanian population which the Serbian government decided to deport (Horvat 1988: 71). The Serbian Army Commander insisted that Serbia 'should not have its Caucasus' and the Prime Minister argued that the

Albanian minority might represent a security concern (Bogdanovic 1985: 139).

In 1909, Serbian intellectual Jovan Hadzi-Vasiljevic explained that the major motivation for the 1878 deportation was also to 'create a pure Serbian nation state' by

'cleansing' the land of the non-Christians, as 'the great Serbian poet Njegos argued'

(Milosavljevic 2002: 81). Hadzi-Vasiljevic was here interpreting Njegos rather loosely, as Njegos's work focused on the Slavic Muslims and not on Albanian Muslims. The ominous implication was that Albanians, as non-Slavs, were not even capable of assimilation. While the Serbian state authorities repeatedly attempted to assimilate the

Slavic Muslims, they refrained from attempting to 'Serbianize' he Albanians (Lampe

2000: 93, 97). While both security concerns and the exclusive nationalist ideology influenced the government's policies, there was also some Serbian resistance to the

'cleansing' of the Albanians. General Jovan Belimarkovic opposed the deportation and offered his resignation to the government over this issue (Milosavljevic 2002: 80) and

During the Great Eastern Crisis, major fighting took place in Bosnia in which Serbian and Montenegrin regular troops and irregulars clashed with Ottoman forces. About 150,000 lives were lost in Bosnia and about 200,000 non-combatants sought refugee in the Habsburg Empire (Pavlowitch 2002 63-64). 26 journalist Manojlo Bordevic also condemned these policies and argued that Serbia should have pursued a policy of peaceful reconciliation towards the Albanians:

In Toplica the Albanians were encountered, and we had nothing more important to do but to expel these warlike, but hard working people from their homes. Instead of making a peace with them as the defeated side - they were without any good reason pushed across the border ~ so that they'll settle on the other side as the enemies of everything Serbian, to become the avengers towards those who pushed them from their homes (Milosavljevic 2002: 79).

Despite some voices of dissent, the Serbian regime 'encouraged' about 71,000

Muslims, including 49,000 Albanians 'to leave' (Pavlowitch 2002: 67). The regime then gradually settled Serbs and Montenegrins in these territories (Rudic 1978: 68). Prior to

1878, the Serbs comprised not more than one half of the population of Nis, the largest city in the region; by 1884 the Serbian share rose to 80 % (Pavlowitch 2002: 68).

According to Ottoman sources, Serbian forces also destroyed mosques in Leskovac,

Prokuplje, and Vranje. 27

The 'cleansing' of Toplica and Kosanica would have long-term negative effects on

Serbian-Albanian relations. The Albanians expelled from these regions moved over the new border to Kosovo, where the Ottoman authorities forced the Serb population out of the border region and settled the refugees there (Popovic 1987: 80). Janjicije Popovic, a

Kosovo Serb community leader in the period prior to the Balkan Wars, noted that after the 1876-78 wars, the hatred of the Turks and Albanians towards the Serbs 'tripled'

(Popovic 1987: 227).. A number of Albanian refugees from Toplica region, radicalized by their experience, engaged in retaliatory violence against the Serbian minority in

Kosovo (Milosavljevic 2002: 79). In 1900 Zivojin Peric, a Belgrade Professor of Law, noted that in retrospect, 'this unbearable situation probably would not have occurred had

27 Letter of the Turkish Minister of Foreign Affairs to the Serbian Minister of Foreign Affairs, October 17, 1898 (Ministarstvo Inostranih Dela Kraljevine Srbije [1899] 1998: 87). 27 the Serbian government allowed Albanians to stay in Serbia'. He also argued that conciliatory treatment towards Albanians in Serbia could have helped the Serbian government to gain the sympathies of Albanians of the Ottoman Empire (Zivojin Peric, quoted in Horvat 1988: 71).

Thus, while both humanitarian concerns and Serbian political interests would have dictated conciliation and moderation, the Serbian government, motivated by exclusive nationalist and anti-Muslim sentiments, chose expulsion. The 1878 cleansing was a turning point because it was the first gross and large scale injustice committed by

Serbian forces against the Albanians. From that point onward, both ethnic groups had recent experiences of massive victimization that could be used to justify 'revenge' attacks.

Furthermore, Muslim Albanians had every reason to resist the incorporation into the

Serbian state.

Particularly ominous was that some Serbian intellectuals later concluded that the

1878 model 'worked' and that it should be reproduced elsewhere. In 1917, during the

Corfu negotiations with Ante Trumbic and other Croat politicians on the future of Bosnia,

Serbian politician Stojan Protic reportedly stated that the Serbs have a violent solution for

Bosnia, which was conversion or extermination of Slavic Muslims (Mestrovic 1993: 66).

While the actual Serbian policies after the liberation/take-over of Bosnia in 1918 were not nearly that extreme, Protic' ideas are a sign that an important section of the Serbian elite saw the 1878 cleansing as 'normal' way to solve the 'Muslim question' in the

Balkans. By 1937 a leading Serbian intellectual Vaso Cubrilovic argued that the 'model cleansing' of Toplica and Kosanica regions in 1878 should be reproduced against the

Kosovo Albanians. 28

1878 ~ 1912: The Persecution of Kosovo Serbs and Hardening of Anti-Albanian

Resentment in Serbia

Following the Ottoman defeat in the 1878 war, Muslim Albanians faced an increasingly dangerous political environment. Kosovo townships were full of thousands of Muslim refugees expelled from the territories captured by Serbia (Bogdanovic 1985:

143). The Treaty of San Stefano between Ottoman and Russian Empire envisioned the creation of a Greater Bulgarian state that would encompass territories with a large

Albanian presence (Austin 2002: 249-50). The Albanian leadership responded to these challenges by forming the Albanian League in Prizren in June 1878, which demanded the creation of an Albanian territorial autonomous unit within the Ottoman Empire, which was to include ethnically mixed regions, such as Kosovo and Macedonia (Bogdanovic

1985: 144).

While the League was soon suppressed by the Ottoman authorities, these demands reflected a complex position of the Muslim Albanians in the collapsing Ottoman Empire.

The Muslim-dominated Ottoman Empire provided considerable opportunities for the

Muslim Albanians (Stavrianos 2000: 501). A number of Albanian chieftains rose to high positions in the Ottoman service, such as a Muhammad Ali, pasha of Egypt, or Ali Pasha of Janina (Jelavich 1983: 219). In light of the religious cleansings by Orthodox nationalists in Bulgaria, Greece, Serbia, and Montenegro, the defense and preservation of the Ottoman Empire seemed as the best way to protect Muslim security and privileges

(Skendi 1967: 193, 202). At the same time, the Ottoman regime was increasingly trying to implement some elements of the West European 'nation building' models. A component of the nationalization policies was the attempt to 'Turkicize' the Muslim 29 ethnic groups within the Empire (Roshwald 2001: 57-63). Thus, while various potentially irredentist Christian minorities were allowed to establish their own schools, the Ottoman authorities banned not only Albanian-language schools but — after 1902 — even private correspondence in Albanian (Skendi 1967: 144).

Not surprisingly, Kosovo Muslim Albanians strongly resisted the Tanzimat reforms and then the Young Turks' push for religious equality which further eroded the remaining Muslim privileges (Durham 2001 [1913]: 247, 274). To add an insult to injury, the Young Turk regime also intensified attempts to assimilate Muslim Albanians into the

'Ottoman nation' (Skendi 1967: 391, 400). In 1908 and 1912, Kosovo Albanian Muslims revolted against the new Constitution that guaranteed religious equality. Instead, they demanded the return to the Islamic Law and the exemption from taxation (Skendi 1967:

343, 348-9; Rakic 1985: 182).

The most vulnerable of all, caught in the middle of the struggle between the declining Empire and the rising nationalist states and movements, were the Kosovo Serbs and the Catholic Albanians. Their old feudal duties towards the Muslim lords (Popovic

1987: 156) seemed even heavier after the peasants in neighboring Serbia were freed from these obligations (Durham 1985 [1909]: 263). The Muslim Albanian population was using its privileged position — including the right to bear arms, denied to Christians — to put pressure on the Catholic Albanians to convert (248). Edith Durham reported that 'the

Albanians are almost solely Albanophone, whereas the scattered Serbs usually speak both languages, and when addressed in Serb often replied first in Albanian' (294). The anti-

Serbian anger of the Albanian Muslims cleansed from Serbia in 1878 combined with the lack of will of the local Ottoman authority to protect the Christians from the loyal 30

Muslims, created increasingly hostile and insecure environment for the local Serb population. Even the clearly Albanophile Durham noted that Albanian militants were using 'medieval methods, for this is the Land of the Living Past,' to push the Serbs out

(296). From 1880 to 1889, more than 60,000 Serbs moved from European Turkey to

Serbia (Ministarstvo 1998 [1899]: 136). Jovan Cvijic estimated that from 1876 to 1912,

150,000 Serbs from Sandzak and Kosovo moved to Serbia (Bogdanovic 1985: 150).

Just as the 1878 cleansing fed the anti-Serbian sentiments and Albanian nationalism, the 1878-1912 discrimination, abuse, and exodus of the Serb minority fed the anti-Albanian feelings and Serbian nationalism. An 1899 Memorandum of Serbian leaders from Kosovo and Macedonia claimed that 'The Serbs of the Kosovo vilayet live like beings without any rights, as a flock of sheep without a guardian, under the paws of wild Albanian animals' (Batakovic 1991: 104). Kosovo Serb leader Janjicije Popovic described the movement of Albanian settlers into previously homogenously Serbian villages as a 'cancer' (Popovic 1987: 242). Some of the officials of the Serbian Ministry of Foreign affairs used similar language in official correspondence, describing Albanians as a 'wild tribe' with 'cruel instincts' (Ministarstvo 1998 [1899]: 1, 16).

Enraged by the persecutions of the Kosovo Serbs, a number of Serbian intellectuals and journalists joined the angry hate propaganda that seemed to culminate during the preparations for the Balkan Wars. In 1912 Serbian demographer Jovan Cvijic argued that 'there is a general consensus that the Albanians are the most barbarous tribes of Europe' (Milosavljevic 2002: 225). One Serbian intellectual of this period described the Albanians as 'European Indians' and 'lazy savages,' and even claimed that in 'the

Albanian language there is no word for love' (226, 229). 31

Still, in the middle of the rising anger towards the Albanians, some Serbian intellectuals and political activists spoke against the hegemonic discourse. Dimitrije

Tucovic, a leader of the Serbian social-democrats, developed a strong critique of anti-

Albanian stereotypes and rising Serbian imperial ambitions. Tucovic found inspiration in the writings of the Montenegrin-Serbian tribal leader, Marko Miljanov (1833-1901) who described Montenegrin- Albanian cooperation and friendly relations and called on the

Serbs and Montenegrins to realize their similarities with Albanians (Miljanov 1963;

Tucovic 1975-1981: 38). 28 Tucovic and Serbian Social Democrats 29 were also inspired by the German Social Democracy's critique of German and Austrian imperialism.

However, unlike the German SDP, Serbian Social Democrats voted against the military budget in 1914. Like some of their Russian counterparts, the Serbian Social-Democrats were among the most truly international of all Marxist parties (Connor 1984: 130).

Tucovic and the Social Democrats were already able to see something that

President Wilson and many others were to realize much later: in the regions with complex ethnic composition and multi-layered cultural identities there was simply no way to apply the principle of self-determination without creating new injustices. Tucovic pointedly asked: 'Who will then establish the boundary line between the Serbians and

Due to the small population and relative isolation of some settlements in mountainous Montenegro, ethnographers were able to trace the Montenegrin/Serb/Slavic relations with the Albanians/Illyrians rather deep into history. While some of the Illyrian clans of Montenegro have been Slavicized, some of the Slav tribal groups in northern Albania have been Albanianized. Some of the clans, such as the Kuci clan, ended up as partially Catholic Albanian, and partially Orthodox Montenegrin (Dvornikovic 1990 [1939]: 300). Since the 18 century, Montenegrin Orthodox Christians frequently cooperated with the Catholic Albanians in their common struggle against the Muslim Slavs and the Ottomans (Karadzic 1972 [1837]: 13). It seems that in this period and region, religious identities were more politically relevant than the linguistic ones. 29 Tucovic and his fellow social democrats were following the ideas of Svetozar Markovic, one of the first Serbian socialists. Markovic rejected the project of the 'Greater Serbia' ~ a state that he said would oppress other Balkan peoples ~ and instead advocated Serbian-led liberation of the oppressed Serbs and all other other Balkans from the Ottoman Empire (Markovic 1960: 221). 32

Bulgarians? How could all Macedonian Slavs be brought into one state without at the same time enslaving the Greeks and the other peoples?' (Tucovic [1914] 104) Tucovic criticized the propaganda of those Serbian newspapers and intellectuals 'who have proclaimed that the Albanians are wild and superfluous inhabitants of the Balkan peninsula and that it is the duty of the Serbian people to remove them from this world'

(Tucovic [1913a]: 213).

Instead of fatal project of the quasi-independent, ethnically cleansed, and mutually hostile national states, Serbian Social Democrats advocated the establishment of a federation of Balkan peoples. They hoped that a Balkan Federation would prevent the imperial intrusions of the European powers and contain interethnic conflicts (Tucovic

[1910a]: 257). They tried to forge better cooperation among socialists in the region and organized the First Conference of Balkan social democratic parties, held in Belgrade

January 7-9, 1910(151,155).

Tucovic was also explicit in his rejection of the dominant anti-Albanian sentiments: 'Other Balkan peoples have no interest in living in hostility with the

Albanians. The union is a salvation for all of them. And in that union there is a place for

Albanians, too' ([1910c] 238-9). Abdul Frasheri, one of the leaders of Albanian cultural and political national movement, also argued that the union of an Albanian state in a

Balkan Confederation would serve Albanian interests (Skendi 1967: 165).

Unfortunately, this was not to be. Some elements of the idea of Balkan solidarity also existed in official Serbian state policy, but it was combined with nationalist ambitions and the tradition of intolerance towards the Muslims of the Balkans. During the preparation for the Balkan Wars, Serbian Prime Minister Nikola Pasic proclaimed his 33 commitment to the anti-imperial solidarity of the Balkan national states (encapsulated in the 'Balkan for the Balkan peoples' slogan) and support for ethnographic boundaries among the Balkan states (Stankovic 1988: 227).

While preparing for the Balkan wars, the Montenegrin state established military cooperation with the Catholic Albanians, concentrated in Northern Albania (Skendi 1967:

409, 413). Greece pursued a fairly successful assimilationist policy towards the Orthodox

Albanians in the South (108, 126-7, 154-6). However, the Serbian government was remarkably unsuccessful in achieving practically any kind cooperation with the predominately Muslim Albanians in Kosovo.

Milan Rakic, Serbian Consul in Kosovo, offered a meaningful explanation for the failure of Serbian attempts to collaborate with Muslim Albanians. According to Rakic,

Albanian Muslims were intolerant towards Serbian Christians, and, as Muslims, they were concerned about losing the religious privileges they enjoyed in the Ottoman Empire.

They also remembered well the expulsion of the Muslim Albanians from Serbia in 1878

(Rakic 1985: 95-7). Despite Rakic' conclusion that the Serbian government did not have any chance of achieving a political agreement with the Muslim Albanians, in 1912 Prime

Minister Pasic offered to the Albanian leaders guarantees of religious freedoms and respect for cultural rights in exchange for support in the coming Serbian military offensive. As Rakic correctly predicted earlier, the Albanian leaders rejected Pasic's offer

(Batakovic 1991: 173; Lampe 2000: 92, 97).

Serbian pre-war propaganda represented the Albanians as 'incapable' of forming a state on their own, and thus good candidates for the 'civilizing' influence of the Serbian state (Milosavljevic 2002: 218-9). Conveniently, the 'civilizing mission' of the Serbian 34 state was combined with a plan to gain access to the Albanian Adriatic ports by annexing large parts of Albania itself. Thus, in a manner characteristic of European imperialism at the time, a paternalistic justification for the expansion was combined with the pursuit of economic interests. In preparation for the First Balkan War, the Serbian government concluded a secret agreement with Bulgaria on the division of the territory that promised northern Albania to Serbia.

The Liberations/Conquests of Kosovo: 1912-1924

The First Balkan War brought surprisingly rapid victories for the Serbian and

Montenegrin forces and their Bulgarian and Greek allies. The atrocities committed during this war by far overshadowed the 1878 cleansing. According to Durham's reports, the Montenegrin soldiers were telling her that 'when the land is once ours there will be no

Mohammedan question,' because the Muslims would die or go to Asia (Durham 2001

[1913]: 31). Whereas the Montenegrin forces also engaged in forced conversions of

Catholics and Muslims, the Serbian forces did not even attempt to convert the Albanian

Muslim population (Bogdanovic 1985: 174; Lampe 2000: 97). A Serbian journalist later interviewed by Leon Trotsky recalled his visit to the region by train following the Serbian offensive:

'By five pm. we were approaching Kumanovo. The sun had set, it was starting to get dark. But the brighter the sky became, the more brightly the fearful illumination of the fires stood against it. Burning was going on all around us. Entire Albanian villages have been turned into pillars of fire — far and near, right up to the railway line... In all the fiery monotony this picture was repeated all the way to Skopje' (Trotsky 1993: 267). 30

Trotsky worked as a war reporter in the region during the First Balkan War. 35

The Western observers estimated that approximately 7,000 - 10,000 Albanians were killed in the Kumanovo and Pristina areas (Horvat 1988: 34; McCarthy 1995: 141).

The agents of the British-Macedonian Relief Fund estimated that in Province of Monastir

(occupied by the Serb and Greek forces) about 80% of Muslim villages were burned

(International Commission 1993: 72). Over 25,000 refugees from Kosovo fled to northern Albania (Durham 1985 [1909]: 29). According to the observers, while the local

Christian population also engaged in inter-communal violence, the worst atrocities against the civilians were usually committed by the paramilitary units (International

Commission 1993: 148, 169). According to a Serbian officer, the paramilitaries were mostly recruited 'from the dregs of society' and have 'made murder, robbery, and violence a savage sport' (Trotsky 1993: 271). Another Serbian officer noted that while some paramilitaries were intellectuals and 'nationalist zealots,' the rest 'had joined the army for the sake of loot' (120). These units systematically killed and plundered Albanian and

Turkish shop owners after the fall of Skopje (268). Moreover, regular forces also frequently engaged in the executions of the POWs. The Serbian officer whom Trotsky interviewed explained: 'I strictly forbade my men to kill them [the POWs] but I must say frankly that this order was not obeyed' (119-20). Even the Serbian civilian population took part in the plunder:

From the area around Vranje the population has crossed en masse into the Albanian villages, to pick up whatever may catch their eye. Peasant women carry away on their shoulders even the doors and windows of Albanian homes (269).

There was a clear logic in this madness, as the Serbian forces engaged in permanent demographic engineering (Mazower 2000: 124) of the population in the region. The Orthodox Slavs frequently experienced the arrival of the Serbian units as a liberation, which is understandable in light of the repression of the late Ottoman era. On 36 the other hand, the armies of the Balkan orthodox states subjected the Muslim Slavs to coercive conversions, expulsions, or even murder. After the fall of Sandzak under

Montenegrin control in 1912, about 13,000 Muslims were forced to convert to

Orthodoxy, comparable numbers were forced to leave for Bosnia, and the Montenegrins took their revenge for the losses suffered during the 1876-1878 uprising (Lampe 2000:

97).

While the motives of enrichment and 'revenge' were probably omnipresent at the level of direct perpetrators, it is important to distinguish them from more ideological and long term motivations of the military and the political elite. The paramilitaries were probably less motivated by the writings of Hegel or Fichte on the nation and history, and more by the hopes of stealing an ox or a cow. Trotsky's informers clearly state that the

Serbian (and Bulgarian) army officers were also encouraging their soldiers with the promises of free land in the newly liberated territories. Destruction, plundering, and mass expulsions destroyed the economy and the tax base in the newly acquired territories and thus hardly served the economic interest of the state institutions. Besides, the very existence of the paramilitaries was out of question without the consent, encouragement, supply, and toleration of the authorities. When some paramilitary units forgot who set the rules of the game and started plundering the Serbian villages as well, they were quickly crushed by the regular forces (Trotsky 1993: 121). Thus, Cathie Carmichael seems right to argue that the causes of ethnic cleansing policies ~ especially at the level of the political elite — were almost entirely ideological (Carmichael 2002:1). 37

Tucovic, who was drafted into the Serbian Army prior to the Balkan Wars, explained the crimes against Muslims as a part of the eliminationist projects of the

Balkan states:

These atrocities were neither initiated by individuals nor were due to personal distress, but were a constitutive part of the 'national' programs of the Balkan states. The Serbian Army was exterminating Albanians in Old Serbia (Kosovo) and Albania, the Bulgarian Army ~ the Turks in Thrace, and the Greek Army ~ the Turks and the Albanians in Devol ~ all in the criminal belief that they were attaining a 'national' aim — that by removing these innocent people off the face of the Earth [sa lica zemlje] they were removing an enemy with whom they would otherwise have to deal in the future (Tucovic [1913b] 161).

In line with the long-term plan of taking Albanian ports, the Serbian and

Montenegrin units occupied, and on November 29, 1912, tried to annex large parts of northern and central Albania (Bogdanovic 1985: 169). They were soon forced out by

Italy and the Habsburg Empire. The Serbian government, having recently committed itself to the principle of national self-determination for the Balkan peoples, was now desperately searching for an ideological rationale for the annexation of territories where ethnic Albanians were an overwhelming majority. In the negotiations with the Austrian representatives and later at the London peace conference in January 1913, Serbian Prime

Minister Pasic argued that many Albanians were Serbs 'by blood', who had changed their religion and nationality due to the Ottoman era persecutions (Stankovic 1988: 231-4).

Pasic further argued that due to the concentration of Serbian religious monuments in Metohija (southwestern and overwhelmingly demographically Albanian part of the present day Kosovo) this region had been a 'Holy Land of the Serbian People.. .since time immemorial' [oduvek] (Bogdanovic 1985: 172). Karadzic, Njegos and Garasanin all failed to make a 'religious' claim on any part of the present day Kosovo. Thus, it does appear that the 'Holy Land of the Serbian People' was a relatively recent invented 38 tradition, probably formulated in the hope of mobilizing Christian solidarity from the

Western Powers against the predominantly Muslim Albanians. The use of the 'religious' argument enables Pasic to claim for Serbia a part of the present day Kosovo that was clearly demographically Albanian.

Pasic even decided to use the Serbian military to 'arrange' that the local Albanian leaders ~ in the disputed areas with large Albanian majority — 'ask' the Great Powers to be left in the Serbian state. In December 1913, the London Conference of Ambassadors decided to recognize an independent Albania, with borders that were yet to be determined

(Stavrianos 2000: 541). In response, Pasic promptly sent a letter to the Serbian High

Command with the following request:

As soon as you can, if possible tomorrow, try to arrange that the Albanian leaders from Pec, Dakovica, Prizren, and Debar send telegrams to London and to ask — in their own name and in front of their communities — the French, the Italian, the Russian, the German, and the Italian ambassadors and Grey [the British Foreign Secretary] to leave them in the Serbian state and not to join them to the coastal Albania (Pasic 1986: 631).

This sudden desire to quickly ('if possible tomorrow') manufacture the consent of the Albanians for their incorporation into the Serbian state is very interesting since in the same time the Serbian Ambassador in Berlin was trying to persuade other diplomats that the borders between Serbia and Albania cannot be established by local plebiscites

'because the Albanians are not a people for itself [narodza sebe] who could vote freely and consciously' (Jovanovic 1986 [1912]: 636).

Finally, during the diplomatic negotiations, Pasic argued that economic and military prerogatives of Serbia demand an access to the sea, as a way to break the encirclement by the Habsburg Empire and its allies (Stankovic 1988: 231-233, 236). This argument seems sincere. Since the fall of the Obrenovic dynasty in 1903 and the re- 39 orientation of the new Serbian government towards France and Russia as the new patrons, the Habsburg Empire and the landlocked Serbia were clearly on the collision course. Pasic was desperately searching for an access to sea ports that would reduce

Serbia's economic dependence on the increasingly hostile Habsburg Empire and provide secure supply lines for military materials. While sincere, this argument shows how far the Serbian government had departed from its earlier commitment to the principle of national self-determination for the peoples of the Balkans. Certainly, trampling upon the basic interests of putatively inferior peoples was hardly a local specialty. According to

Tucovic, the true inspiration and the model for the Serbian state were the successful

European imperial powers:

Serbian capitalists have opened their account of colonial murders and horrors and now they can proudly join the capitalist company with the English, the Dutch, the French, the Germans, the Italians, and the Russians (Tucovic [1914]: 102).

While a number of Western journalists did report on the crimes committed by the

Christian armies, it does not seem that there were major Western diplomatic protests against these abuses. Again, it seems that the sense of Christian/European solidarity made it very hard to feel sorry for 'the Turks.' Barbara Jelavich observes that the major Western

European newspapers had a tendency to report the Ottoman atrocities against the Balkan

Christians and to simply ignore the Christian atrocities against the Balkan Muslims

(Jelavich 1983: 217). In 1915, the British pro-consul in Egypt wrote: 'We are fighting in order that the Turkish hordes, who for 500 years have camped in Europe, should be driven back into Asia' (Gallagher 2001b: 84). 40

From 1914 to 1924, Serbian/Yugoslav troops repeatedly invaded parts of Albania.

In the winter of 1916-1917, during the Serbian withdrawal through Albania, Albanian guerrilla units did not miss an opportunity for revenge attacks against Serbian troops

(Pavlowitch 2002: 97). In 1916 and 1917, during the Serbian uprising against the

Austrian occupation in the Toplica region, the Albanian paramilitaries attacked the

Serbian rebels (Bogdanovic 1985: 179). When the Serbian Army returned to Kosovo in

1918, the roles of the persecutors and the victims were reversed again.

At the Paris Peace conference, the Serbian delegation refused to extend the minority rights protections on the Albanian minority in Kosovo. In line with the

Albanian inferiority' argument, the Chief of the Serbian Military Mission at the peace conference suggested that the 'uncivilized' people of Albania were incapable of an

'independent existence'. Thus, he claimed the Yugoslav state should be given northern

Albania to play a 'civilizing' role there (Krizman and Hrabak 1960: 321; Lampe 2000:

166). A 1921 report by a Serbian Radical Party member from Kosovo states that the local

Serbs developed the 'crazy idea' that 'Muslims will not be allowed to live in a Serbian state. On the basis of that idea, the local Serbs are committing various crimes against the

Muslims' (Horvat 1988: 38).

From 1918 to 1924, the new Kingdom fought against Kosovo Albanian guerillas that were supported by the Kosovo Committee in Albania. The goal of the Kosovo

Committee was Albanian national unification, by military means if necessary (Austin

2002: 255). In 1921, in his speech to the League of Nations, the Prime Minister of

Albania claimed that 'the Albanian nation has suffered cruelly from the dismemberment of which the country was victim in 1913' (Austin 2002: 265). The Yugoslav Army used 41 artillery in the attacks on the rebel-held villages, the property of the families of the

Albanian rebels was often confiscated by the state, and the Army routinely sent families of the rebels into concentration camps (Bogdanovic 1985: 190). In 1924, the Yugoslav

Army supported the establishment of a client regime in Tirana, which then quickly suppressed the Kosovo Committee (Austin 2002: 256). Faced with the ruthless tactics of the Yugoslav military and without the support from the 'motherland', the Albanian guerilla in Kosovo soon lost military significance.

Throughout this long period of almost non-stop warfare in Kosovo, Macedonia, and parts of Albania, several Serbian politicians and intellectuals continued to criticize

Serbian atrocities and its colonial policy. In 1911 a Serbian social-democrat Dragisa

Lapcevic argued: 'Our efforts must not be led astray by the anger of the chauvinists .... who in a chorus numbering one hundred thousand madmen demand a 'great empire' of this or that ancient ruler' (Horvat 1988: 17). During the 1913 Serbian invasion of Albania, social-democrats Kosta Novakovic and Dimitrije Tucovic criticized the Serbian Army's repression of the Albanian civilian population (37). Dragisa Vasic, a moderate conservative and a Yugoslav military officer on duty in Albania and later secretary of the elite Serbian Cultural Club, saw the crimes committed in 1921 and described 'poor

Albania' as 'our shame' in which Serbian soldiers 'die defending plunderers and criminals'

(39). In 1921, Stojan Protic moderated his reported previous views and suggested that

Kosovo and Sandzak should become one Muslim-majority province (91). Still, as

Lapcevic's earlier statement aptly illustrates, it seems that the majority of the politically active Serbs were supportive of the colonial policy towards the Albanians. 42

The Failure of the Colonial Policy, 1924-1929

With the consolidation of the new Kingdom during the 1920s, there were some attempts and successes in including Albanians into the political party system. Just to be on the safe side, careful gerrymandering before the 1920 elections for the Constitutional

Assembly ensured that Serbs outnumbered Albanians in all electoral units (Lampe 2000:

121, 124). Nevertheless, by 1924 the Albanian and Turkish party Cemiyet was in coalition with the (Serbian) Radical Party and had 14 MPs in parliament (Horvat 1988:

43). However, once the leaders of the Radical Party concluded that Cemiyet was secretly pursuing a secessionist policy, they broke off the coalition and had the leader of Cemiyet arrested. Cemiyet was defeated in 1925 parliamentary elections and stopped functioning the same year (Banac 1984: 378). A major chance to incorporate genuine Albanian representatives in the parliamentary life of the Yugoslav state was thus lost.

The long lasting Albanian guerrilla warfare probably confirmed a belief that the

Albanians were hopelessly disloyal. A 1924 Police Report provides a memorable description of the Albanians: 'The Arnauts [Albanians] live in many municipalities of southern Serbia — brutal, primitive, distrustful, smart, and disloyal people' (Hodza 1984:

6). While it is probably true that the Albanians were not loyal to the Serb-dominated state, it is hard to imagine that it could have been otherwise, given the way that the state had treated them in 1878, in 1912 and again in the 1920s. In this case the assumptions of disloyalty had a tendency to work as a self-fulfilling prophecy. The Serbian elite seemed unable to understand that the loyalty of Albanian minority could be unconditionally demanded, but has to be earned and maintained with negotiations and fair treatment. 43

In 1924, Albanian emigre organizations appealed to the League of Nations to enforce the Minority Treaties and to protect the Albanian minority in Kosovo. The official response of the Yugoslav Royal representatives deserves to be quoted at length:

If one could trust to the authors of the petition, it would seem that the Royal government threatens the survival, the material welfare, and the cultural life of the peaceful and civilized Albanian population. However, that population is neither peaceful nor civilized. It lacks national consciousness, and it is only capable and willing to work as hired murderers (Horvat 1988: 47).

Even after such a scandalous response there were no sanctions imposed on the

Yugoslav Royal government. The central problem was probably a lack of commitment of the Great Powers to the enforcement of minority rights, especially in the cases where the offenders were the local client states. In 1925, the British Foreign Secretary observed that the minority treaties serve only to 'keep alive the differences which otherwise might be healed in time.' Thus, in these circumstances the treaties were hardly 'anything but an evil for all concerned' (Gallagher 2001b: 85).

Probably to ensure that Albanian national consciousness did not develop, the

Yugoslav government prevented the opening of Albanian-language schools in Kosovo

(88). During the inter war period, the illiteracy rate for the Yugoslav Albanians remained at 90 percent (Banac 1984: 299). From 1919 to 1941 Macedonia and Kosovo remained under the direct ministerial administration of Belgrade (Lampe 2000: 10).

The government tried to 'improve' the ethnic composition of Kosovo by promoting colonization policies, such as the confiscation of Albanian land and introduction of the Serbian and Montenegrin settlers (Horvat 1988: 40). The combination of discrimination, repression, and colonization did produce some of the desired results.

By 1921 about 40,000 Kosovo Albanians had left for Albania, and by 1930 another 44

45,000 left for Turkey (Bogdanovic 1985: 193). By the late 1930s about 60,000 colonists, mostly Serbs, were introduced into Kosovo (189). However, according to the official

Yugoslav census data, the ethnic-cultural composition of Kosovo showed only minor changes. In 1921, about 27% of Kosovo inhabitants reported Serbian or Croatian mother tongue and 65 % reported Albanian mother tongue. By 1931, 33% reported one Slavic

o 1 mother tongue and about 60% reported Albanian mother tongue.

Why did the Serbian policy in the 1920s shift from policed deportations (as in the

1830s or in 1878) to a more gradual colonization? A possible reason might be the radically different demographic make up of the newly acquired region. The estimates of the 'national' identity of the populations have to be taken cautiously, but it does seem that

Serbia's more recent territorial acquisitions could hardly be seen as 'Serbian-majority' areas. While the Belgrade Pashalik before 1804 was reported to be about 90% Serbian, and the Toplica and Kosanica regions were about 80% Serbian before 1878, in 1921

Serbs and Montenegrins comprised at most 40 % of the population of Kosovo (Lampe

2002: 10; Pavlowitch 2002: 28, 84). In addition, with the establishment of a client regime in Tirana in 1924, the Belgrade authorities were probably less likely to see Kosovo

Albanians as a major secessionist or security threat. However, the growing influence of the fascist Italy in Albania would soon force the Belgrade elite to re-think its policy on the Kosovo Albanian question.

Author's computations from the 1921 and 1931 Yugoslav Census Data. Yugoslav Federal Statistics Bureau (1998). The 1921 and 1931 Censuses reported mother tongue and religion, not self-declared ethnic identity. The 1921 Census definition did not allow for reporting of Macedonian language and most Slavic Muslims probably reported being Serbo-Croat speakers. Thus, it is highly unlikely that all 27% of the reported Serbo-Croat speakers thought of themselves as Serbs. The 1931 Census lumped together various Slavic languages, and it again makes no sense to assume that all of these people thought of themselves as Serbs. 45

Planning the Cleansing of Kosovo from the Albanians, 1929-1940

As the transition from democracy to right wing authoritarianism spread across

Eastern Europe, leading Serbian intellectuals and politicians of the Kingdom of

Yugoslavia developed increasingly exclusionary plans for the Albanian minority. In

1929, King Alexander proclaimed a Royal dictatorship and implemented a set of measures aimed at a speedy creation of a unified Yugoslav nation, basically modeled after the French one. The country was divided into new administrative areas, banovine, which ignored almost all existing ethnic boundaries (Lampe 2000: 167). Conveniently, the boundaries of the new 'banovine' ensured that none had a Slavic Muslim or Albanian majority (167). Kosovo region was divided among three different banovine (Hodza 1984:

28).

In the 1930s Milan Stojadinovic' conservative right government made major attempts at reconciliation and power sharing with several non-Serb groups. Stojadinovic carefully distanced himself from the coercive 'ethnic amalgamation' ideas and policies of

King Alexander (Bulatovic 2002: 293), and brought leading Slovenian and Bosnian

Muslim politicians into his government. He also removed Serbian hardliner General

Zivkovic, who objected to the partnership with the Muslims (Lampe 2000: 177-8).

However, Stojadinovic's government also planned to bring the anti-Albanian policies to a new extreme.

Probably the most infamous articulation of the anti-Albanian sentiments of the

Serbian elite in this period was the 1937 proposal from Vaso Cubrilovic, the Secretary of 46 the elite Serbian Cultural Club, regarding the cleansing of Kosovo Albanians.32

Cubrilovic argued that, 'in the 20th century, only a country inhabited by its own people can be confident of its security' (Cubrilovic 1997b [1937]: 405). Thus, the Albanian

'wedge' in Kosovo represent a major threat to the security of the Yugoslav state.

Assimilation is not a realistic option 'because they base themselves in Albania their national awareness is awakened' (402). Thus, 'if we do not settle the score with them at the proper time, within 20-30 years we shall have to cope with terrible irredentism' (407).

The 'slow and sluggish gradual colonization policy' had failed to fundamentally shift the demographic composition of the region. As Protic reportedly argued in 1917,

Cubrilovic claimed that in the Balkans, 'European' approaches to the minority rights were entirely ineffectual: 'The fundamental mistake made by the [Serbian] authorities in charge at that time was that, forgetting where they were, they wanted to solve all the major ethnic problems of the troubled and bloody Balkans by Western methods' (401). More radical methods were needed. 'We hold the view that the only effective method for solving this problem is mass expulsion of the Albanians' (424). There were already numerous examples of the effectiveness of this method. The 1878 cleansing of Toplica and Kosanica demonstrated that the method worked. Moreover, 'all the Balkan states, since 1912, have solved or are on the point of solving their problems with national minorities through mass population transfers' (402). The Great Powers of Europe are doing this too:

If Germany can re-settle tens of thousands of Jews, if Russia can move millions from one end of the continent to another, there will be no world war for the sake of a few hundred thousands re-settled Arnauts (408).

32 The Serbian Cultural Club was not an agency of the Stojadinovic government. The Club was the major Serbian elite institution, bringing together leading Serbian academics, senior bureaucrats and major industrialists. (Dimic 1996: 508-9, 510-1). 47

Cubrilovic goes into the detail in the elaboration of the policy, describing specific measures used against the Albanians in southern Serbia after 1878 in order to 'create a new Toplica out of Kosovo' (415). The best lands should be taken away from Albanian peasants, taxes should be 'ruthlessly' collected, licenses of Albanian professionals should be revoked, etc. In addition, Serbian paramilitary units should be organized and secretly assisted by the state, and local riots should be incited and then suppressed by the local

Serbian militias. Finally, the Albanian villages and city quarters should be secretly burned (410). He continued, 'nationalizing the regions around Sar Mountain [Southern

Kosovo and Western Macedonia] would mean that we can stifle irredentism once and for all, and ensure our control over these territories forever' (405).

Cubrilovic's idea is clearly intolerant to the extreme. His basic aim was to induce the Albanians to move, preferably to Turkey, by making their lives unbearable and by selective massacres of the Albanian civilians. In 1938, the Yugoslav Military HQ joined the ultra-nationalist intellectuals by advocating the resettlement of the ethnic Albanians into Turkey, as soon as possible (Bogdanovic 1985: 194). The HQ was probably concerned with the growing Italian presence and influence in Albania (Stavrianos 2000:

725-6), which had the potential to turn 'disloyalty' of the Yugoslav Albanian minority into a 'security concern.' Possibly inspired by Cubrilovic's proposal and reflecting the spirit of the times elsewhere in Europe, Stojadinovic signed an agreement with Turkey to accept 40,000 Albanian families in 1938 (Elsie 1997: 427).33 The Communist Party of

Yugoslavia organized a protest by Belgrade University students against this

'Convention Regulating the Emigration of the Turkish Population from the Region of Southern Serbia in Yugoslavia', [1938] in Elsie, 1997: 427. 48 eliminationist plan (Cohen 2001: 14). The agreement was not implemented at that point, due to a lack of funds (Lampe 2000: 192).

Early in 1939, following discussions between Stojadinovic and Count Ciano (the

Italian Foreign Minister) the Yugoslav Ministry of Foreign Affairs prepared an internal analysis on the possible division of Albania between Italy and Yugoslavia (Krizman

1977: 77-90). The analysis offers the following justification for the invasion and annexation of parts of Albania: 'The Albanians never showed solidarity with the other

Balkan nations. As well, we never felt any sympathies for them' (80). A section of the memorandum was written by Ivo Andric, then Yugoslav diplomat and a later Nobel Prize winner for literature. Andric advised against the annexation and also argued that it might be possible for the Yugoslav state to assimilate the Kosovo Albanian minority. In the context of the times and the intellectual climate, these were relatively enlightened ideas.

However, he also does imply that if the Albanian state vanished, the out migration of the

Muslim Albanians into Turkey ~ presumably a goal of the Stojadinovic government ~ would be facilitated (89).

Surprisingly, some of these eliminationist ideas would actually be put into practice by the multinational Yugoslav Communist regime. Vaso Cubrilovic again wrote a memorandum in 1944, this time advocating massive deportations of all major non-

Slavic minorities that 'proved' their disloyalty by massive collaboration with the fascist powers — Germans, Hungarians, Italians, and of course, Albanians. Since this time he was addressing the Yugoslav Communists, Cubrilovic was now using the 'model' of

Soviet cleansing of ethnic Germans and the 'population exchange' between Ukraine and

Poland in 1944 (Cubrilovic 1997a [1944]). Indeed, the Yugoslav Communists did deport 49 the German and the Italian minorities at the end of the war (Carmichael 2002: 54-5). In the 1950s the Yugoslav Communists did implement the agreement with Turkey on the re settlement of Turks' (Hodza 1984: 68-70). Vaso Cubrilovic, who died in 1990, joined the

Serbian Academy of Arts and Sciences, an elite cultural institution that played a crucial role in the 1980s Serbian national mobilization.34

Assimilate, Deport, or Kill:

The Formation of the Cleansed National States at the Balkans

Why were the Serbian goals and means of action towards the Albanians so exclusionary and why did they gradually worsen throughout the 19th and the early 20th century? A simple answer would be to argue that a large part of the Serbian political elite was simply following the trends in continental European and regional thought and the policy of the times. In his analysis of the formation of national states in Western Europe,

Charles Tilly notes that almost all successful states (i.e. those that survived) adopted some forms of deliberate cultural homogenization policies, such as the adoption of a state religion, expulsions of minorities, institution of a national language, and educational standardization (Tilly 1975: 43-44, 49, 78.). The Ottoman system of non-territorial autonomies and tolerance for religious minorities effectively ensured that while

Northwestern Europe was gradually culturally homogenized, Southeastern Europe remained without clear ethno-cultural boundaries.

As a number of European nationalist movements radicalized towards fascism

(Alter 1994: 16-38), the Serbian national programme gradually shifted from Garasanin's

34 Still, Cubrilovic did publicly distance himself from the controversial 1986 Memorandum of the group of the members of the Serbian Academy of Arts and Science. (Dragovic-Soso 2002: 186). 50 paternalistic 'imperial restoration' towards Cubrilovic' ethnic cleansing. As the early

Balkan national states expanded into the political vacuum left by the collapsing empires, they encompassed ethnically diverse — and potentially irredentist ~ populations. The continued presence of these minorities was simply incompatible with the coercive Utopia of a centralized and homogeneous nation-state. The alternative roads to modernity were beyond imagination. The minorities and ethnic pluralism had to become the victims of the Utopia.

It is tempting to argue that Miroslav Hroch's excellent account of the rise of minority nationalism in Eastern Europe fails to follow the process to its logical conclusion. Hroch distinguishes just the first three phases. Phase A is characterized by a scholarly interest and codification of history, language, and unique customs. In the Phase

B, 'patriotic groups' use the definition of'the people' from Phase A to 'awaken the nation', create national identity, and gain state recognition for the 'nation'. Finally, in Phase C, a mass national movement arises, especially if the ethnic activists manage to convincingly frame an existing conflict of interests as a 'national conflict' (Hroch 1985). While this account seems to convincingly describe the formation of collective 'national' political identity of a number of minority nations, it does not provide systematic analysis of the developments of the minority national movements after they becoming the majority nationalisms ~ that is, after they capture the state power.

The formation of the national states in the Balkans appears to advance from the point where Hroch's account stops. A 'Phase D' might be the achievement of an independent state, usually at the 'core' and homogeneous ethnic territory (Serbia, Bulgaria and Montenegro before 1912, Romania before 1918). Early autonomous states were used 51 as a stepping stone to a full independence by Romania, Serbia, Bulgaria, and Greece

(Mazower 2000: 101). 'Phase E' would then be the acquisition of new territories with mixed populations, usually following the collapse of the Empires. Here the focus was on

'national unification' by territorial expansion (101). For example, the pre-1918 Romania was 90% ethnically homogeneous, but the territorial gains increased the percentage of the minorities to 27% (Rieber 2000: 12). Finally, 'Phase F' would be the coercive ethnic homogenization of the acquired territories, by assimilation, expulsion or mass murder.

Following the collapse of empires, the new national regimes aimed to achieve unitary, ethnically cleansed polities (Livezeanu 1995:4). Between 1911 and the first post-WWI census, Greece, Yugoslavia and Bulgaria managed to dramatically reduce the Muslim population from the territories captured during the wars. From the Muslim population living there in 1911, by the early 1920s, 55% remained in Bulgaria, 46% in Yugoslavia, and only 17% in Greece (McCarthy 1995: 155). From 1912 to 1922, about half a million

Muslims, 1.5 million Greeks, and 250,000 Bulgarians [Orthodox Slavs] were 'relocated'

(Todorova 1997: 175).

Maria Todorova observes:

...the Balkans were becoming European by shedding the last residue of an imperial legacy, widely considered anomaly at the time, and by assuming and emulating the homogeneous European nation-state as the normative form of social organization. It may well be that what we witness today [in the 1990s], wrongly attributed to some Balkan essence, is the ultimate Europeanization of the Balkans (13).

While this line of reasoning does capture some of the reasons for the Serbian anti-

Albanian ideas and policies, it still has major flaws. It downplays the crucial importance of Great Power politics on the ability of the Balkan nationalist leaders to pursue their coercive homogenization programmes. The major European Powers exercised crucial 52 influence, initially by treating some of the atrocities against the Christians as

'humanitarian disasters' and simply ignoring the atrocities towards the Muslims, and later by frequently ignoring the violations of the minority rights by the local client states, including Yugoslavia.

Furthermore, the explanation that centers just on the nationalist ideology represents the Serbian political and cultural elite as unimaginative imitators, whose major fault was that they were not ahead of the rest of Europe. The situation was more complex.

Even the most cursory look at the process of the formation of the Yugoslav state shows that the Serbian political elite was capable of designing and negotiating institutions and identities different from the standard nation-state model.

Serb leaders were somewhat constrained in their thinking by the dominant ideas of their times, but when faced with complex realities they were able ~ usually via the process of trial and error — to come up with better solutions. The perception of the Croats gradually evolved from assimilationist plans (Karadzic 1972b [1837]: 114-125) to the recognition of and the ethno-territorial autonomy for the Croat people in 1939. While the Serbian elite denied Slavic Macedonians the right to define their own collective identity, they still focused on coercive assimilation instead of more exclusionary policies.

Serbian policies were even less tolerant towards Slavic Muslims, and in several periods this group was subjected to coercive conversions, expulsions, and massacres. Yet, the

Stojadinovic regime was prepared to include the leading Bosnian Muslim politician

In the late 1930s, the Serbian Cultural Club rejected the 'unitarism' (implicitly referring to the ideas of integral Yugoslavism) and recognized the legitimacy of the Croat claim for territorial unit. However, the leaders of the Club strongly objected to the involuntary inclusion of the Serbian majority regions into the Croat unit and also insisted on the need for a Serbian territorial unit to be explicitly defined (Dimic 1996: 529, 534). Also see Bakic 2004: 476-7. 53

Mehmed Spaho in the government, even if that meant alienating Serbian hard-liners, such as General Zivkovic.

The Yugoslav Albanians, however, were generally treated worse than the Slavic

Muslims. They were neither recognized as one of the constitutive 'tribes' ~ as the

Slovenians and the Croats eventually were — nor were they seen as potentially

'assimilable' into the Serbian nation — like the Macedonians and the Slavic Muslims.

This is not to idealize the profoundly discriminatory policies of coercive assimilation and conversions towards the Macedonians and the Slavic Muslims. While the pressure to assimilate usually does indicate a strong sense of cultural superiority, it still allows for the possibility that the Others can become equal to Us, if they only make the effort.

Thus, the important limitation of the argument which centers just on the coercive

Utopia of nation state is that it fails to explain why were some minorities singled out for assimilation and others for more exclusionary treatment? Seemingly no amount of assimilation could turn Albanians into Serbs. In the Serbian case, those minorities whose 'essential' character was perceived as culturally distant (in terms of religion, language, and socio-economic development) and physically threatening (in terms of the previous history of conflict and the current presence of major foreign supporters) were more likely to be perceived as logical candidates for deportation or murder. Thus, the nationalist elite's visions of the 'essential' characteristics of the Other determined which policies towards Them would appear as realistic and feasible.

Similarly, the inability or unwillingness to assimilate the Muslim populations seems to have been frequent trouble in the process of the creation of nation-states in the Balkans. While the non- territorial Orthodox Christian minorities (such as the Vlachs or the Orthodox Roma) were often easily assimilated into a rising nation, attempts at assimilation of Muslims had rarely been successful (Poulton 2000: 52, 61). 54

The perception of Albanians as unassimilable and barbarous was formed as a result of the persecution of the Kosovo Serbs by the Albanian extremists in the 1878-

1912 period. The recurring episodes of Kosovo Albanian massive support for the imperial invaders of Serbia ~ first the Ottoman and later the Austrian Empire — further strengthened the belief that they were an disloyal minority and a security threat.

The Serbian elite's animosity towards the Albanians was built on a tradition of intolerance towards Albanians and Muslims. Tragically, in this tradition the commitment to the national liberation was fused with extreme intolerance towards the Balkan

Muslims. The new contributors to the tradition of intolerance were clearly building on the genealogy of ideas and actions created by their predecessors. Njegos' verse looked back to the 1709 cleansing and he dedicated The Mountain Wreath to Karadorde, who liberated but also cleansed Serbia from the 'Turks'. Hadzi-Vasiljevic justified the 1878 cleansing of Albanian Muslims with reference to Njegos ideas. Cubrilovic used the 1878 cleansing as a blueprint for his own proposal on Kosovo. Cubrilovic's conceptualizations of Albanians as a 'wedge' in the Yugoslav lands influenced in turn the Serbian historian

Dimitrije Bogdanovic, who had a decisive impact in turn on the intellectual framing of the Kosovo crisis in the 1980s (Dragovic-Soso 2002:127-139).

Still, against the current, the counter-hegemonic discourse and practices also exist in this history. In the thoughts and lives of Marko Miljanov and Dimitrije Tucovic, in the resistance of General Belimarkovic to the 1878 cleansing, in the Balkan federalist projects of Serbian social-democrats, in the demonstrations of Belgrade youth against the deportation of Albanians in 1938 ~ the Montenegrins and the Serbs have a valuable 55 heritage of a different vision of the Albanians: not as 'inferior' or 'savages,' but as respected neighbors and fellow Balkan peoples. 56

Chapter Three: The Unintended Consequences of Ethno-Federalism: How Yugoslav Communists Dug Their Own Graves

The Ethno-Federal Dilemma

Ethno-federalism is a form of political power sharing within a single country between the central government and the government of one or more ethno-territorial autonomous regions. Each such region is explicitly designed as an ethnic homeland for an ethnic group ("titulars"), with special responsibilities for that group's cultural survival and advancement. Ethnic federalism implies that neither level of government can legally abolish the other; each has a constitutionally defined and separate jurisdiction.

Should multi-ethnic countries try to contain ethnic conflicts by introducing ethnic federalism? This is a question with important policy implications. As separatist movements found in a number of multi-ethnic countries could conceivably lead to violent ethnic partitions, many such countries are either implementing or considering the implementation of ethno-federal institutions. Ethno-federalism might offer the hope of a peaceful resolution of ethnic tensions and conflicts in such places as Quebec, Northern

Ireland, Corsica, the Basque region, Kosovo, Western Macedonia, Republika Srpska,

Transylvania, Northern Cyprus, Crimea, Iraq, Chechnya, and Tibet, to name only a few.

My alternative approach builds on insights from studies that draw attention to resources and opportunities, not grievances in the formation of nationalist movements.

By noting the importance of state-recognized identities and existing organizations for the emergence of social movements, these studies provide a key link between political process models and historical institutionalism. In specific, instead of blanket endorsement

37 Following Soviet terminology, "titulars" refer to the members of the ethnic group for whom ethno-territorial autonomy is established. 57 or rejection of the ethno-federal solution, my goal is to identify specific conditions under which it does not work. Michael Hechter qualifies his general endorsement of ethno- federalism with a warning, based on the failure of Yugoslav federalism, that excessive decentralization has disintegrative effects. Using Hechter's insight as a point of departure, I identify specific institutions and processes that had disintegrative effects in the Yugoslav case (1945-1991). Due to the spatial distribution of Yugoslav ethnic groups, the assumption that inter-ethnic equality demands that each nation have titular status in its own ethnic republic unintentionally led to institutionalized racism of titulars against minorities. Further, while ethnic equalization policies bought the loyalty of less- developed regions, they triggered the secessionism of the wealthy. Finally, the dogmatic pursuit of Leninist teaching led Yugoslav communists to respond to nationalist mobilizations with more decentralization, effectively creating a vicious circle of decentralizations and nationalist mobilizations.

The Containment Effect of Ethno-Federalism?

Michael Hechter has recently (2000) claimed that a central government's introduction of ethnic federalism can weaken nationalist demands for independence. The reasoning, which generally follows the logic of the deprivation theories of social protest (Smelser

1962; Davies 1962, 1969; Gurr and Moore 1997), is that minorities protest and try to secede when they are deprived of autonomy and equal treatment. Transfer of resources to ethnic autonomous units by political and economic decentralization should reduce the strength of the secessionist threat. In this fashion, "federalism undercuts nationalist conflict by providing minorities with a more optimal mix of government-provided goods" 58

(Hechter and Okamoto 2001: 204).

(loyalty to the "nation-state" and forced cultural homogenization) in Europe triggered the

"peripheral nationalism" of minorities who resisted incorporation or attempted secession

(Hechter 2000a: 71). Hechter further claims that a U-shaped relationship between level of decentralization and national conflict can be observed by looking at the association between fiscal centralization and national rebellion in the 1980s (Hechter 2000a: 148).

Hechter's quantitative analysis indicates that Yugoslavia was a highly decentralized country mostly without national conflicts. Clearly aware that this claim is likely to raise eyebrows in light of the bloody wars of the 1990s, Hechter accounts for

Yugoslavia by developing his argument on the U-shaped relationship. In the Yugoslav case, "extreme decentralization" made it difficult for the federal government to adapt to exogenous economic shifts and courted fragmentation (Hechter 2000b: 325-5). More generally, "[i]f too little decentralization causes rebellion, then too much is likely to engender fragmentation," and this can be "another source of intergroup violence"

(Hechter 2000a: 151-2). The problem with this modification of the theory is that we are given no clear instructions on the threshold point. At what point decentralization stops being a medicine (preventing rebellion) and becomes a poison (engendering fragmentation)? For policy makers and citizens considering ethno-federal experiments, this is not an abstract question. 59

The Fragmentation Effect of Ethno-Federalism?

Several social scientists have claimed that ethno-federalism unintentionally facilitates nationalist mobilization and state disintegration (Connor 1984; Vujacic and Zaslavsky

1991: 137; Beissinger 1998: 182; Bunce 1999: 49; Gorenburg 2003: 25). Often building on social movement theory, they draw attention to resources and opportunities, not grievances in the formation of nationalist movements. By noting the importance of state- recognized identities and existing organizations for the emergence of social movements, they provide a key link between political process models and historical institutionalism

(Brubaker 1996; Bunce 1999; Gorenburg 2003: 9). For example, in the Soviet case, once

Gorbachev's reforms reduced the threat of direct repression of protest activity, the deep institutionalization of ethnicity determined the form and expression of nationalist demands (Zaslavsky 1997: 89; Gorenburg 2003: 9, 24).

These scholars argue that ethnofederalism fragments national identity, freezes ethnic identities, and legitimizes titular racism. Ethno-territorial autonomy encourages titulars to perceive ethno-territorial autonomy as ethnic "homeland" or their collective ethnic property (Connor 1984: 501), in which they can "naturally" expect - and demand

- preferential treatment vis-a-vis non-titulars in terms of access to jobs, cultural resources, and so on (Zaslavsky and Brym 1983: 98). As Khrushchev explained in 1956,

"[s]hould the Jews want to occupy the foremost positions in our republics now, it would naturally be taken amiss by the indigenous inhabitants" (Zaslavsky and Brym 1983: 105; emphasis mine). An ethnicity-based national government can thus fail to either integrate or tolerate ethnic Others (Kalin 2004: 306) and instead degenerate into a "government of ethnic majority, by ethnic majority, for ethnic majority" (Hayden 1999: 15). 60

Consequently, titulars typically see a strong presence of non-titulars in good jobs as inherently unfair. Yet the demand to remove minorities from good jobs appears blatantly unfair to non-titulars who typically frame their opposition as a demand for equal

("merit-based") treatment, regardless of ethnic background. When the distribution of jobs is openly along ethnic lines, even those who previously did not care about their or others' ethnicity are more likely to notice it (Gorenburg 2003: 12). Finally, with the specification of geographic borders and official political symbols (flag, coat of arms, capital city) for an ethno-territorial autonomy, it becomes easier for titulars to perceive "their" autonomy as a state-in-waiting (Connor 1984: 300-1, 497; Bunce 1999: 49).

In addition to fears of disintegrative ideological effects, critics argue that ethno- federalism provides key institutional resources for secessionists. Because ethno-territorial autonomy creates a bureaucratic apparatus that has a vested interest in the maintenance and strengthening of titular ethnic identities, the weakening or disappearance of such identities might threaten the bureaucracies' reason to exist. Thus, if secessionist politicians manage to capture a provincial government, they have at their disposal a number of key institutional resources to further their own cause: tax revenue, patronage

TO positions in the provincial officialdom, and even coercive apparatuses.

Studies of post-communist Europe have produced a growing body of evidence on the institutionally disintegrative effects of ethno-federalism. As Bunce notes: "The irony of the collapse of socialism, then, was that the very institutions that had defined these systems and were there, presumably, to defend them as well, ended up functioning over time to subvert both the regime and the state" (Bunce 1999: 2). Studies of the collapse of 38 In addition, as Donald Horowitz (1998) explains, in the post-communist period, international law has evolved in the direction of the recognition of the right of secession of territorial autonomies from multi-national federal states. 61 the USSR find that the ethnic groups with the highest level of autonomy (union republics), most wealth, highest average level of education, and the least linguistically assimilated population were the first to secede (Hale 2000). As these union republics used their well-developed ethnic institutions to cultivate ethnic identities and facilitate nationalist mobilization, Hale concludes that "appeasing restive regions by decentralization is unlikely to succeed" (Hale 2000: 32, 49). Similarly, Beissinger (2002) finds that Soviet ethnic groups with union republics, the lowest level of linguistic assimilation, and highest level of urbanization were "early risers" on the path to secession. He concludes that the Soviet federal system, instead of preventing secessionism, actually promoted it by creating a sense of territoriality (even for groups that previously lacked bounded territory) by providing ethnic cultural and educational institutions, and by creating national middle-class party cadres and intellectuals

(Beissinger 2002: 119).

While the communist ethno-federations failed to contain nationalist conflicts

(Vujacic and Zaslavsky 1991: 137; Emizet and Hesli 1995: 504; Gorenburg 2003: 25), it cannot be argued that ethno-federalism has always produced state disintegration and fragmentation. A number of democratic ethnic federations - such as Switzerland, Canada, and India - have relatively successfully weathered secessionist challenges over long periods. The danger is that overgeneralizing from the dramatically negative experience of

Communist federations might make us downplay the more successful records of their democratic counterparts. Even so, Hechter is cautious in his endorsement of ethno- federalism, noting that "there is little consensus about the kinds of political institutions that are most likely to contain nationalism." (2001: 203-4). 62

Where the Medicine turned into Poison: The Yugoslav Case

Departing from Hechter's insight that limited doses of decentralization can contain nationalism, but that excessive decentralization leads to fragmentation, I will identify specific institutions and processes that had disintegrative effects in the Yugoslav case

(1945-1991). Unlike Hechter who focuses on fiscal decentralization, I assume that decentralization has a plurality of dimensions that cannot be easily expressed as a single continuous variable. Going beyond Hechter, therefore, I seek to identify specific ethno- federal institutions and processes that had disintegrative effects. An in-depth analysis of the Yugoslav case is justified by its critical position in Hechter's analysis; as Hechter himself argues, the excessive Yugoslav decentralization unintentionally turned ethno- federalism from helpful medicine into the poison of fragmentation.

This chapter uses an extended historical case study of the evolution of ethnic federalism in Communist Yugoslavia (1945-1991) analyzing the effects of three waves of

Yugoslav decentralization on collective political identities, inter-ethic relations, and institutional settings. An analysis looking only at the last phase of the crisis cannot possibly capture slow-moving, cumulative changes in the political environment (Pierson

2004: 141, 166). As Pierson notes, the identification of causal mechanisms requires the analysis of big and slow-moving processes that can only be studied successfully over the long term (2003); he discusses changes in the conception of nationhood (2003: 181) and in the level of a federation's centralization (2004: 120) as typical examples. Tracing such processes enables us to understand the underlying causal mechanisms.39

As Pierson explains, much of contemporary research in political sciences is geared towards short time horizons (2003: 180). This approach is of limited use if we are trying to understand the "slow-moving" causal processes, where the change is continuous but extremely gradual. For example, many key 63

Yugoslavia's Three Waves of Decentralization

This historical analysis outlines evolution of Yugoslav federalism and discusses the effects of ethno-federal institutions on national unity. It shows that the institutional structure of the Yugoslav federal system was not static: from 1945 to the 1980s,

Yugoslavia evolved from "facade federalism" into a very decentralized federation.

Yugoslavia as a "Facade Federation "

The first step in Yugoslav post-war decentralization was replacement of pre-war unitary

Royal Yugoslavia (1918-1941) with a Communist federation composed of ethnic republics. A Leninist vision heavily influenced Yugoslav communists and their

"solution" to the national question (Guzina 2000:22). As Tito observed in 1935, "[t]he

Soviet Union became the model for the multinational state, the model for what

Yugoslavia with its varied races must become" (Connor 1984: 146). Believing that a nationalist demand for independence arises from a minority's bitter experience with discrimination and oppression by the majority (Connor 1984: 201), Lenin developed a deprivation-based explanation of minority nationalism. In his view, by creating ethno- national autonomies, the Soviet government would protect minorities from a Greater

Russian chauvinism and enable them to experience full cultural development. Over time, minorities would overcome their distrust of the central government. Similar levels of socio-economic development and economic growth would bring various ethnic groups closer until they "merged" into a new national collectivity - a "Soviet people" (Connor

1984:50,201,245,277).

demographic conditions (such as levels of fertility and urbanization) typically change very slowly, but their gradual transformation over time can eventually completely transform social structure (Pierson 2003: 181). 64

In 1935, Yugoslav communists decided that the post-revolutionary Yugoslav state should be a multi-national federation. The pre-war Kingdom of Yugoslavia was clearly dominated by the Serbian monarchy and political elite. The communists promised that a new Yugoslavia would replace "Greater Serbian hegemony" and the oppression of non-

Serbs with the national equality of all Yugoslav peoples. This promise was a crucial element of their mobilizing appeal to non-Serbs in World War II (Connor 1984: 147,

158). To this end, Yugoslav communists implemented a system that closely followed the

Soviet federalism (Crawford 1998: 254). Following the Soviet model, each republic received a legislative assembly, government, judiciary, anthem, coat of arms, and flag

(Stanovcic 1988: 33). Despite its formal federal character, the centralized organization of the Communist Party, which wielded political and economic power, insured that

Yugoslavia was a unitary state with some cultural autonomy for recognized ethnic groups

(Shoup 1968: 119,122; Vucinich 1969: 253). In this period, Communist Yugoslavia was not a genuine federation.

The Confederal Turn in the mid-1960s

While economic decentralization started in the 1950s, a more dramatic move towards major institutional decentralization started after the 1966 purge of Aleksandar Rankovic and his supporters, mostly ethnic Serbs. Rankovic was a federal leadership member in charge of the secret police and internal party cadre policy (Burg 1983: 32, 36). Found guilty of systematic persecution of the Albanian and Hungarian cadre, Rankovic and his group were portrayed as coercive centralizers. The Party was now on course to rectify its previous mistakes. 65

To prevent future centralist ("unitarist") abuses, control over the republics' secret police and cadre policy was shifted to the republican communist parties (Burg 1983: 36-

7, 82). In November 1967, the central Party leadership agreed that Republican Party congresses should take place before the federal party congress, thereby enabling republican delegations to formulate and defend a common position (Burg 1983: 62).

Moreover, it was decided that most members of the federal Party organs should be representatives of the republics and that there should be "appropriate national and republican representation" in the central party organs (Rusinow 1977: 255; Burg 1983:

34, 64).

In the aftermath of Rankovic's downfall, leading Slovenian communist Edvard

Kardelj claimed that each Yugoslav republic was a nation with its own culture and that each republic should build its own national economy to sustain its culture (Vucinich

1969: 266, 274). 40 In the economic sphere, the hollowing out of the centre's economic powers (and responsibilities) that began in the late 1950s was rapidly completed. Federal level central economic planning was eliminated, and almost all remaining federal economic powers were shifted to the republics (Rusinow 1985: 136). The republics gained control over cadre policy in the banks, companies, and social services (Ramet

1984: 77). Confederation even made inroads into the military sphere: by the late 1960s, the republics and provinces gained control of Territorial Defence units (Rusinow 1977:

285; Bilandzic 1986: 24).

The effects of these changes in institutional framework were dramatic. A later

Party report lamented that the "CP Y [Communist Party of Yugoslavia] was reduced to a

40 Kardelj's argument seems a major revision of the Leninist doctrine on the national question. A Leninist would argue that in some (more or less distant) future, different Yugoslav nations should fuse into one, but it appears that in Kardelj's thinking such fusion was not just postponed, but cancelled. 66 coalition of republican and provincial organizations" (Burg 1983: 81). Since the major source of social power in a Communist society is the centralized power to appoint managers in state agencies and companies (Bilandzic 1986: 41), the shift of that power to the level of the republican Central Committees greatly strengthened the republics. The

Yugoslav economy disintegrated into eight "mercantilist and protectionist regional fiefdoms" (Rusinow 1985: 141). By March 1969, there was no independent "party centre" left other than Tito (Burg 1983: 64). Once state security was "republicanized," the Army was the only all-Yugoslav institution directly controlled by the centre (Burg

1983: 67). While the usual criticism of the communist federations as "federal in form, unitary in content" (Connor 1980: 52, 225, 468; McGarry 1998: 226) rings true for the early post-war period, by the late 1960s, Yugoslavia was clearly a decentralized state.

The 1974 Decentralization: Killing Ethnic Nationalism with Kindness

In the aftermath of the 1971-72 Croatian nationalist mobilization, the party centre introduced a new wave of decentralization, and the 1974 constitution confirmed the "de facto confederal structure of the state" (Rusinow 1985: 141). Further economic decentralization allowed Croatia to retain a much greater share of tourism revenue and created republican control over the banking system (Rusinow 1977: 313, 323). By the end of 1974, the only significant economic powers left in the hands of the federal government were the coordination of foreign economic relations (including loans from the World Bank), the Federal equalization fund, and monetary policy (Ramet 1984: 79).

Federal laws were supposed to be implemented by republican bureaucracies, and this enabled considerable variability in implementation (Sekelj 1990: 187). Moreover, veto 67 powers over constitutional changes were granted to every republic and province (Burg

1983: 206, 210). The parity principle for the representation of republics was extended to the collective state Presidency, Federal government, and Constitutional Court (Rusinow

1977: 285). The Territorial Defence forces (republican militia), introduced in 1969, were constitutionally enshrined (Rusinow 1977: 285), and the Army was now supposed to follow ethnic proportionality in the higher ranks. Yugoslavia had no common educational system, student exchange program, or system of inter-republic labour mobility (Jovic

2003: 65). While the republican communist parties controlled appointments to the federal intuitions via a quota system, federal institutions had no corresponding influence over such policies within the republics (Burg 1988: 10, 18-19). Furthermore, the central party and state organs were increasingly supposed to make decisions on the basis of consensus

(Bilandzic 1986: 21).

The death of Tito in 1980 removed the final arbiter who had enough country-wide charisma to impose binding decisions on the republican parties and, if needed, to purge republican parties that "degenerated" into nationalist positions. After the previous waves of decentralization, other "federal" leaders were basically representatives of their home republics and hence lacked the capacity to make unpopular decisions for the country as a whole. The extent of decentralization and isolation was such that if communist leaders from one republic criticized nationalism in another, it was seen as meddling in the other's internal affairs. The cumulative result of the waves of decentralization, including introduction of multiple veto points, was that Yugoslavs became "prisoners of a system we cannot change" (Rusinow 1988: 5).

Thus, Yugoslavia evolved from heavily centralized "facade federation" into a 68 highly decentralized state, with extremely weak central government. In effect, the belief that decentralization would eliminate the structural causes of nationalism led the communist leadership to respond to each national mobilization with more decentralization. However, decentralization and nationalism became mutually reinforcing.

Effects of Decentralization

Incompleteness of Ethnic Republics and Slovenian Exceptionalism

If national equality required recognition of Yugoslav ethnic homeland republics, then it was necessary to decide where to draw the borders of these republics. Even with the best of intentions, this was remarkably difficult, since most Yugoslav ethnic groups were geographically dispersed. Only in the case of Slovenia was it fairly easy to create an almost "ideal" Slovenian national republic: all Slovenians were in one republic, and only

Slovenians lived in that republic.

The issue of "completeness" of the national republics had decisive importance. If the existing republican borders were ever to become inter-state borders, Slovenia would emerge as a homogenous nation-state, but all other republics would not. As Slovenia was, in demographic terms, a nation-state in the waiting, when Slovenian nationalists started pushing for secession in the late 1980s, they did not have to worry that a large number of

Slovenians would be left behind in other parts of Yugoslavia, or that Slovenia's internal minorities would use the opportunity to secede from Slovenia. All other national republics had to face at least one if not both issues. Not surprisingly, the attempt to draw borders of national republics from an enormously complex ethnic patchwork resulted in 69 tensions between the titular principle and multi-ethnic reality (Rusinow 1985: 158), except in the case of Slovenia (see Table 3.1).

Table 3.1 Ethnic Completeness of the Republics in 1948 and 1981 (in percent) 1948 1981 Republic Titulars Internal External Internal External Bosnia Bosniacs 30.73 97.46 39.52 81.50 Croatia Croats 79.20 78.62 75.08 78.02 Macedonia Macedonians 68.49 97.47 67.01 95.49 Montenegro Montenegrins 90.67 80.34 68.54 69.17 Serbia Serbs 73.89 73.68 66.38 75.94 Slovenia Slovenians 97.00 95.39 90.52 97.66 Notes: Internal Completeness: percent of titulars in the population of "their" republic. External Completeness: percent of ethnic groups members who live within "their" republic. Source: Yugoslav Federal Statistics Bureau, Census Data CD

Yugoslavia against Yugoslavs

In many ethnic federations, there is on-going tension between the affiliation of minority nations with "their" territorial autonomy and with the common state. While many citizens frequently feel some affinity to both national identities (Keating 1997), federal government typically strives to strengthen the national identification with the federal state. A peculiar feature of the Yugoslav federation was that the federal government worked to discourage the rise of the Yugoslav national identity.

The early 1960s saw the culmination, followed by a sudden reversal, of the

Yugoslav communists' attempts to form a common nation from Yugoslav ethnic groups.

In line with Leninist theory, they had a vague hope that a common socialist culture and economy would merge the ethnic groups without fully replacing the separate identities

(Rusinow 1977: 134-5). To this end, the 1958 Party program endorsed the creation of a common Yugoslav identity (Burg and Berbaum 1989: 538), and the 1961 census allowed respondents to declare themselves "Yugoslavs" in an ethnic sense (Ramet 1984: 55). But in 1963, Tito complained, "There are even some individuals who don't dare call 70 themselves Yugoslavs" (Shoup 1968: 224).

In general, non-Serbian Yugoslavs had reservations about the spread of Yugoslav national identity. Minority group members often see attempts at the creation of a common, amalgamated identity as nothing a more sophisticated attempt at assimilation by the majority. In the Yugoslav case, non-Serbian suspicions were strengthened by negative experiences of coercive assimilation into the Yugoslav nation in Royal

Yugoslavia, perceived as Serbianization under Yugoslav disguise (Bakic 2004b: 491,

506,516).

Predictably, republican communist leaderships were hostile to the spread of

Yugoslavism, as they worried it would undermine their ethnic power base (Burg and

Berbaum 1989: 538). Surprisingly, the federal communist leadership also decisively turned against the formation of the Yugoslav national identity. In a 1962 speech, leading

Slovenian communist and federal party ideologist Kardelj proclaimed, "The federation of the Yugoslav peoples is not a framework for the creation of a new Yugoslav nation"

(Jovie 2003: 135). Speaking at a Communist Party of Yugoslavia [CPY] Congress in

December 1964, Tito condemned "the confused idea that unity of our peoples means the elimination of nationalities and the creation of something new and artificial" (Ramet

1984: 55). He described attempts to create a unified Yugoslav nation as an expression of

"bureaucratic centralism" and "unitarism" (Ramet 1984: 55-6). From this point on, trying to create conditions for the formation of a common nation was a major ideological sin.

The later rise of Yugoslav identity happened in spite of'th e party-state's policies.

In the 1980s, more than 3 million people (more than 13% of the total population) were the product of ethnically mixed marriages or were married to someone of different 71 ethnicity (Woodward 1995: 36). In some regions of the country (especially urban Bosnia,

Croatia, and Vojvodina), a high number of individuals entered inter-ethnic marriages or expressed Yugoslav national identity. Analysis of 1980s survey data indicates that individuals from intermarriages were 11.58 times more likely to self-declare as ethnic

Yugoslavs than other respondents, net of other predictors (Sekulic, Massey, and Hodson

1994a: 92). Moreover, individuals in interethnic marriages and self-declared ethnic

Yugoslavs expressed the highest levels of interethnic tolerance (Massey, Hodson, and

Sekulic 1999: 682-3, 685). Fourteen percent of children from ethnically homogenous marriages also self-declared as ethnic Yugoslavs (Burg and Berbaum 1989: 540).

In accordance with these trends, some Yugoslav intellectuals argued that "if the forced creation of a Yugoslav nation was bad, it does not follow that a spontaneous emergence of the Yugoslav nation should be banned" (Horvat 1985: 319). However, the republican leaderships strongly resisted formation of a Yugoslav national identity, which they saw as threatening to assimilate the nations they claimed to represent (Burg and

Berbaum 1989: 538). When the 1981 census showed that more than 1.2 million citizens

(more than 5.5% of the total population) self-declared as Yugoslavs, a member of the leadership of the Communist Party of Croatia wrote in the regional newspaper that "there is absolutely no possibility of a Yugoslav nation being formed" (Ramet 1984: 57), thereby directly denying the choice that had just been made by many Yugoslav citizens.

Nor did Yugoslav state institutions, following a Stalinist understanding of national identity, allow recognition of multiple ethnic identities (for example, Croat-Yugoslav), insisting that everyone must choose a single identity. Yugoslav federalism undermined common national identity by encouraging open hostility to Yugoslav national formation 72 and prohibiting the recognition of multiple identities of individuals. As we will see later, the limited appeal of Yugoslav national identity had a major negative impact on the strength of pro-Yugoslav political forces during their 1980s showdown with exclusionary nationalists.

Ethnicized Economic Equalization Policy

As explained earlier, the Leninist theory of nationalism argues that in a socialist country nationalism arises out of memories of oppression and current experience of economic inequality among ethnic groups. Thus, a major Yugoslav economic effort was dedicated to the economic development of less developed republics, with the aim of eliminating uneven development as the "objective" cause of nationalism (Shoup 1968: 227; Vucinich

1969: 257; Rusinow 1977: 247). The policy essentially consisted of increased taxation of more developed republics and provinces, and the transfer of these funds to the governments of less developed republics and provinces. While the policy had some positive results, it failed to meet equalization targets and produced a host of unanticipated problems. The more developed republics resisted the policy and tried to keep redistributed funds at a minimum, while the beneficiaries insisted on higher contributions.

From the point of view of the centre, the issue of equalization was often framed as an efficiency vs. equity dilemma: whether to invest scarce funds in the most developed republics to achieve better returns, or in the least developed republics to help them to catch up (Shoup 1968: 233, 237-8). An alternative solution to the problem of underdevelopment - allowing profitable companies from more developed republics to 73 open branches in less developed republics and thus take advantage of cheap and abundant labour - was seen as a form of exploitation and, thus, was ideologically unacceptable

(Shoup 1968: 239, 259-260; Zarkovic-Bookman 1992:130).

Despite redistributive efforts, instead of catching up, the less developed regions fell farther behind. By the late 1970s, almost 3 percent of Yugoslav's gross material product (GMP) was transferred to the less developed republics and provinces; this represented 11 percent of the GMP of these regions (Tome 1988: 7). As Table 3.2 shows, economic inequality between the more and less developed regions was worse in 1978 than 1947. While strong population growth (due to higher fertility) in less developed regions was partly to blame, the equalization policy also lacked an effective mechanism for keeping the leaders of the LDRs accountable for the funds transferred to them.

Table 3.2 Per Capita Social Product of the Regions as a Percent of Yugoslav Average Region 1947 1965 1978 Economic Populati Growth Growth Slovenia 162 177 205 5.8 35.92 Croatia 105 120 127 5.7 22.48 Vojvodina 100 122 115 5.5 22.34 Central Serbia 101 95 98 4.9 37.65 Macedonia 70 70 68 4.9 65.58 Montenegro 94 71 71 4.1 54.91 Bosnia 86 69 64 4.1 60.77 Kosovo 49 39 29 3.2 117.70

MDRs 110 118 124 5.5 29.90 LDRs 77 64 59 4.1 70.05 Notes: Economic Growth is average annual growth rate of per capita social product, 1947-1978; Population growth refers to average annual per cent change, 1948-1981; MDRs are more developed regions; LDRs are less developed regions. Sources: Ramet (1984: 183); Yugoslav Census Data CD, 1947-1981

The initially uncontroversial decision to designate whole republics (or provinces) as "less developed" - and eligible for financial support - had unforeseen negative effects.

To keep receiving equalization funds, the leaderships of the republics designated as "less 74 developed" struggled to maintain this status (Ramet 1984: 197-8). Since most republics and provinces had a clearly specified titular ethnic group, the Yugoslav-level debates about equalization funds soon gained unsavoury "why should we be paying for them?" undertones, with both "we" and "they" defined in ethnic terms. The question of development and general economic policy gradually became ethnicized (Kuljic 1997).

The leaderships of the more developed republics argued that their profitable industries were heavily taxed to build ventures of dubious value elsewhere, such as prestige projects or unprofitable plants (Vucinich 1969: 262-3). In 1958, Slovenia had 8.6 percent of the Yugoslav population, but contributed 37.2 percent to the national budget

(Vucinich 1969: 263). A 1965 Yugoslav survey found that the anger of the better-off towards redistribution was not just a leadership phenomenon. Citizens of more developed regions were not satisfied with the redistribution scheme. While 46.9 percent of all

Yugoslavs agreed that the "developed should help the less developed," only 29.7 percent of respondents in Slovenia agreed (Burg 1983: 49). About 10 percent of Slovenian respondents argued that the developed republics had been "victimized enough" by the equalization program. While these sentiments were appearing even in the decades of high

Yugoslav economic growth, they emerged more fully in the 1980s when an economic crisis shook the Yugoslav system to the core. Instead of bringing Yugoslavs closer, as

Leninist theory prescribed, the equalization policy failed to close the regional development gaps. It also ethnicized economic policy debates, and strengthened conditions for the secession of the rich. 75

The Rise of Titular Nationalisms

Constitutional designations of republics, with the exception of Bosnia, defined the titular ethnic group for a given territory. However, as no republic, other than Slovenia, was ethnically homogenous, the ethnic supremacy of the titulars had dramatic consequences for non-titular ethnic groups. Once republics are represented as the national states of the titulars, non-titulars begin to feel like second-class citizens (Tadic 1986: 67). This is borne out by Yugoslav survey results from the late 1980s which indicate that in each republic titulars were always more ethnically intolerant than non-titulars. Especially intolerant were titulars who lived in a municipality of "their own" republic in which non- titulars were a local demographic majority (Massey, Hodson, and Sekulic 1999).

The demand for "national economies," enshrined by leading Party ideologist

Kardelj, not only had disintegrative effects on the common Yugoslav economy, but led to the demand that the majority nation should enjoy economic domination in "its" republic

(Horvat 1985: 294). As Horvat explains, the concept of national economies implies "the demand for ethnic discrimination" - in favour of the majority (Horvat 1985: 294).

Members of other ethnic groups increasingly feel like foreigners, and are more inclined to relocate to "their own" republics. The 1974 Yugoslav constitution confirmed the definition of titulars for each republic; by the 1980s, most ethno-territorial entities

(provinces and republics) were regarded as their national states by the titular ethnic groups (Rusinow 1988: 1).

The significance of the loss or gain of titular status is well demonstrated by the effects of the reversal of the titular status of Kosovo. After 1966, when Albanians replaced Serbs as titulars, Party leaders instituted dramatic changes in cadre and 76 employment policy. Kosovo Albanians represented half of the Communist party members in the 1950s, and by 1978, they were about two-thirds (Cohen 1989: 364, 366).

In 1974, 58.2 percent of those employed in the public sector were ethnic Albanians, and

31 percent were ethnic Serbs; by 1980, 92 percent were ethnic Albanians and only 5 percent ethnic Serbs (Ramet 1984: 161). In 1966, ethnic Albanians were only 13 percent of the provincial state security personnel, but by 1981, local security forces were 75 percent Albanian (Cohen 1989: 366; Lukic 1990: 228). As Kosovo Albanian intellectual

Shkelzen Maliqi observed in 1966, "the Kosovo administration changed its owner, but not its character" (Maliqi 1989: 72). The use of the titular position to discriminate against non-titulars confirms that introduction of ethnic federalism can enable new forms of discrimination and poison inter-ethnic relations.

Ethnic Nationalism within the "Vanguard"

Another unintended disintegrative effect of Yugoslav ethno-federalism was the creation of institutional resources and incentives for nationalist mobilizations - by the communists. Despite Lenin's dictum that the Party be kept free of nationalism (Connor

1984), after the 1960s several major nationalist mobilizations were ledby the leaderships of the republican communist parties (Table 3.3).

The communist elite wanting to use nationalism typically started by creating a sphere safe from police persecution for non-communist nationalist activists, usually based in supposedly apolitical ethno-cultural institutions (Rusinow 1977: 290; Sekelj

1990: 213). A key role in Croatian nationalist mobilization (1971) was played by Matica

Hrvatska, that republic's ethno-cultural institution (Ramet 1984: 128, 130); in Serbian 77 nationalist mobilization (1986), a similar role was played by the Serbian Academy of

Arts and Sciences (Dragovic-Soso 2002). Thus, supposedly apolitical institutions became jumping-off points for non-communist nationalists.

Table 3.3 Major Ethno-nationalist Mobilizations in the Yugoslav Communist Party Period Region Major Demands Outcomes 1970-72 Croatia Fiscal Decentralization Police Repression Affirmative Action for Titulars Party Purge Threat of use of Federal Army 1981-82 Kosovo Republican Status for Kosovo Police Repression Party Purge Threat of use of Federal Army 1983-91 Slovenia Formal Confederalization War against the Federal Army Republican Military Forces Successful secession Secession 1986-91 Serbia Abolition of Kosovo and Vojvodina Autonomies Abolished autonomies Police Repression of Albanians Re-Centralization of Yugoslavia Purge of Albanians from Public Greater Serbian Unification Employment War in Croatia

The next step was to validate at least part of the nationalist framing of contentious issues by arguing that the only way to contain nationalism in an ethnic group is by addressing the issues that inflame nationalists. As a moderate Serbian communist noted in 1981, the communist leadership was "expressing concern for the interest of their own nation to such an extent and in such form that it became impossible to distinguish them from nationalist rhetoric" (Spiro Galovic, cited in Jovic 2003: 289-90).

The third step was to openly articulate these issues and demand a better deal for the nation, thereby inflaming nationalist mobilization and representing themselves as the defenders of the nation's interests to titulars (Horvat 1985: 288). As Mika Tripalo,

President of the CP of Croatia, explained in 1970, "[t]he League of Communists in every republic expresses and is obliged to express the interest of the working class of its nation and of its republic" (Ramet 1984: 233). As communists turned into nationalists, they modified their ideological discourse, replacing class with nation as the revolutionary 78 historical agent and replacing "class enemies" with "enemies of the nation" (Vujacic

2003: 360, 368). Communist party leaders represented themselves to the communist leaders of other republics as "moderates" who should be heeded, lest the "extremists" - genuine non-communist nationalists - take over (Rusinow 1977: 252; Woodward 1995:

70, 89).

Instrumental use of ethnic nationalism was a high risk gamble. While the communist leaders of Croatian (1971) and Albanian (1981) ethnic mobilizations were purged by the centre, the communist leaders of Slovenian (1983) and Serbian (1986) ethnic mobilization managed to become first presidents of independent countries. The potential gain for the communist leadership was strengthened popularity in their own republics (Vujacic 2003: 359) and a stronger bargaining position vis-a-vis the leadership of other republics (Woodward 1995: 89). The risk was twofold, however. On the one hand, they could be outflanked by genuine non-communist nationalists, who could make more radical demands and push the Communists into real ethnic conflict (Rusinow 1977:

291). On the other hand, they risked confrontation with the centre which would quickly be fed up with disruptive nationalist mobilizations and purge them for being too soft on nationalists.

It is important to recognize that the institutional incentives for the nationalist gamble were created by earlier ethno-federal decentralization. The high level of institutional autonomy (particularly in terms of cadre policy) of the republican communist parties after the mid-1960s decentralization became a resource for nationalist mobilizations within the CPY. The effect of the decentralizations, at least in the case of the four ethnic mobilizations, was to strengthen the hand of ethnic nationalists. 79

The Centre's Response to Nationalism: Purge the Sinners, Sweeten the Sin

The response of the senior leadership to the early nationalist mobilizations was in line with the Leninist theory, but had unexpected effects. The leaders of ethnic mobilizations were purged, and some more radical ethnic activists were jailed. As Tito said at a 1966

Central Committee meeting, "There can be no place in the League of Communists for those who have degenerated to a nationalist position" (Vucinich 1966: 267). However, while the communists making nationalist demands were (eventually) purged, their demands were typically met later (Rusinow 1985: 141). Ultimately, the new wave of decentralization and weakening of the central powers increased the ability of republican leaders to play the nationalist card. Overall, the centre treated the nationalist "deviation" as individual pathology, not a predictable and rational response to institutional incentives, an approach in line with Lenin's theory of nationalism. The problem was that Lenin's theory was wrong.

In December 1971, Tito met the Croat nationalist mobilization with police repression and a massive Party purge. Over 500 student demonstrators were arrested, 741 people were purged from the Party, 131 were removed from positions of power, and 280 were pressured to resign (Burg 1983: 157). This wave of repression was backed up by the open threat of the use of the Federal Army, should Croat communists refuse to comply.

Tito warned in subsequent meetings that there had been a "misunderstanding" of

"democratic centralism" in the Party, with too much democracy and not enough centralism (Burg 1972). While nationalist mobilization was crushed by extra- legal police repression and mass purges, nationalist causes were not eliminated (Horvat 1985: 323), and the 1974 decentralization met almost all major demands of the Croat nationalists, 80 further weakening the centre and strengthening the republics.

In essence, the Yugoslav political system kept producing ethnic nationalism

(Vucinich, 1969: 260; Vujacic and Zaslavsky 1991: 131; Stambolic 1995: 61-2, 235).

Ongoing decentralization - without democratization - established a plurality of communist parties, each controlling its own republic (Kostunica 1988: 89). The leading proponents of revived nationalism were not members of the residual bourgeoisie, but the communists themselves (Vucinich 1969: 282). In the late 1960s, Vucinich correctly predicted that since communists continued to treat national problems by satisfying ethnic demands, Yugoslavia would evolve into a confederation (1969: 283). Attempts to appease nationalists by devolution and decentralization appeared to be short-term remedies, as they fed ever-increasing appetites for autonomy and could be exploited as tools of rebellion (Ramet 1984:137, 169, 235). Thus, the dynamics of decentralization and nationalist mobilization were mutually reinforcing.

The Withdrawal to Home Republics

Instead of bringing Yugoslav peoples closer together, as Lenin's theory predicted, the logic of Yugoslav ethno-federalism led to a quiet "unmixing of the peoples" that started long before the wars of the 1990s. Yugoslav federalism gave a privileged position to titulars as the first among equals in their republics. The dilemma for non-titulars was the following: why be a vulnerable minority in someone else's republic, when you can be a titular in your own republic? As Figure 3.1 shows, many non-titulars resolved this dilemma by moving to their own republics: Serbs moved to central Serbia and Vojvodina,

Croats to Croatia, Albanians to Kosovo, Muslims to Bosnia (Bilandzic 1986: 136-7; 81

Stanovcic 1988: 28).

But ethnic factors affected migration even before the 1960s reforms. A multivariate analysis of the 1961 census shows the importance of ethnic affinity in the direction of migration: the higher the percentage of an ethnic group in the destination, the greater the migration of that ethnic group to that destination. For example, Serbs of north western Bosnia moved to Serbia rather than the economically more attractive and geographically closer Croatia (Hawrylyshyn 1977: 112).

Figure 3.1: Withdrawal to Home Republics

1961 1971 1981

Source: Petrovic (1987: 121-3) based on 1981 Yugoslav census data Note: Y-axis shows the number of migrants of the given titular ethnic group moving into the "home" republic as a percent of all inter-republic migrants from that titular group in the given period. As Table 3.4 shows, a combination of economic and ethnic factors was responsible for internal population movements. As one would expect on the basis of economic logic, the overall migration balance was negative for less developed republics

(Montenegro, Macedonia, Bosnia) and positive for more developed ones. If their republic had high unemployment, even titulars (such as Montenegrins or Macedonians) were likely to move out. At the same time, as Figure 1 shows, the rates of immigration to one's 82 own republic slowly increased over time for most titular ethnic groups.

As seen in Table 3.4, and as economic logic would imply, migrants moved from less developed regions with high unemployment into more developed regions with lower unemployment. However, to which more developed region they moved depended on ethnic factors. Apparently, the direction of migration was more influenced by ethnic factors than employment opportunities, wages, or climate (Bilandzic 1986: 136). Croats from impoverished parts of Bosnia migrated towards Croatia, Serbs from rural Croatia and Bosnia moved to Serbia, and so on (Petrovic 1987). The impact of non-economic

Table 3.4 Economic Factors and Internal Migration, 1971-1981 Social Product per Unemployment Rate, Migration Balance, Republic capita, 1975 1971-1981 Average 1971-1981 Slovenia 201 1.73 2.30 Croatia 124 5.77 1.09 Serbia n/a 15.67 0.69 Macedonia 69 25.76 -0.50 Montenegro 70 15.12 -2.67 Bosnia 69 13J3 -3.31 Note: Social product per capita is given as a percent of Yugoslav average. Migration Balance is the difference between number of immigrants and out-migrants, divided by the total population in 1971. Source: Petrovic (1987); Woodward (1995: 383-4); Ramet (1984: 183). factors is especially visible for the Kosovo Albanian population, who experienced very high unemployment - as high as 50 percent in 1984 - and rapidly increasing density of settlement, but very low out-migration rates. While there are no studies of the issue, it seems possible that the main barriers to internal migration were various forms of discrimination in housing and employment against non-titulars. Titular nationalism thus may have helped maintain and increase the "ethnic purity" of the republics. Instead of

"merging" into a new national collectivity, Yugoslav peoples were growing farther apart. 83

Table 3.5 Concentration in Home Republics by Internal Migration, 1961-1981 1951-1961 1962-1971 1972-1981 Ethnicity Bosniac 1.00 1.04 0.92 Croat 1.58 2.10 1.05 Macedonian 0.50 0.59 0.56 Montenegrin 0.29 0.38 0.41 Serbian 3.19 3.84 4.09 Slovenian 0.65 1.02 2.27

Yugoslav Average 1.20 1.49 1.55 Note: Concentration is the ratio of number of titulars who moved into the home republic divided by the number of titulars who moved out in the 10 years before the census in question. (This only refers to internal migration within Yugoslavia.) Numbers above 1 indicate increased concentration in the home republic, and between 0 and 1 indicate increased dispersion. Source: Petrovic (1987)

From Unimaginable to Inevitable: The Disintegration of Yugoslavia

The cumulative disintegrative effects of Yugoslav ethno-federalism came to the forefront during the severe economic crisis of the 1980s. With the death of Tito in 1980, the country lacked an arbiter who could transcend ethnic divides. Faced with difficult choices and unpopular decisions, the Yugoslav political elite was increasingly paralysed. Pro-

Yugoslav civil society organizations tried to overcome ethnic cleavages, but ethno- federalism's deep institutionalization of ethnicity gave an insurmountable advantage to their ethnically exclusive competitors. Some smaller Yugoslav nations, who benefited materially and symbolically from the federal system, remained overwhelmingly pro-

Yugoslav to the bitter end, but wealthier and larger ethnic groups used their institutional resources to pursue separate agendas, regardless of the effect on Yugoslav unity.

The Economic Crisis of the 1980s

The use of foreign loans to finance projects of dubious economic and social importance

during the 1970s, combined with the effects of the second oil shock at the end of the 84

1970s, produced a major debt crisis in the 1980s (Rusinow 1988: 2; Woodward 1995:

47). The crisis intensified debates between more and less developed regions over economic equalization. At the same time, the end of the Cold War led to the end of US support for Yugoslav unity, as the Soviet threat that Yugoslavia was supposed to counterbalance in Southern Europe disappeared (Woodward 1995; Jovic 2003).

By the late 1970s, the Yugoslav government was facing seriously depleted foreign reserves, falling exports, and foreign debt of about $20 billion (Woodward 1995: 48).

Under IMF pressure, the federal and republic governments had to implement painful austerity measures aimed at reducing domestic consumption and increasing exports

(Woodward 1995: 51). Investments in social services and infrastructure were frozen.

Food subsidies were abandoned in 1982. By the end of 1984, the population living below the poverty line was 25 percent (Woodward 1995: 51-2). By 1985, 39 percent of

Yugoslav citizens under 25 were unemployed (Woodward 1995: 57). By Spring 1988,

Yugoslavia's official unemployment stood at 1.2 million, annual inflation was 150 percent, and living standard improvements of the prosperous 1970s had evaporated. In

October 1988, only 6 million were employed out of 23 million Yugoslav citizens

(Rusinow 1988: 2, 76). Falling real incomes (average real salaries dropped by 24 percent in 1988) and consumer goods shortages, combined with the weakening of the social safety net, radicalized social tensions (Cohen 1989: 439; Woodward 1995: 52, 383).

As the crisis worsened throughout the 1980s, the inability to make and implement painful decisions became glaringly obvious. Economic decentralization and the requirement that key decisions be made by consensus at the federal level made the decision-making process slow to the point of paralysis (Burg 1988: 13). Any change in 85 the rules of the game was practically impossible, as each republic and both provinces had veto power over constitutional change (Woodward 1995: 60). Without Tito, no procedure or person existed who could resolve differences between incompatible projects of different republics (Woodward 1995: 84).

Failure of Attempts to Escape Ethnic Fragmentation

In the late 1980s, some civil society organizations, such as the Society for Yugoslav

Democratic Initiative, tried to overcome ethnic and regional divides and organize at the country level. The last Federal Minister, Ante Markovic, successfully organized the only all-Yugoslav political party to contest elections in the late 1980s: the Alliance of the

Reform Forces of Yugoslavia.

Still, given the institutional framework of the Yugoslav federation, the mobilizing potential of non-ethnic organizations and parties was constrained. For one thing, the 1974 constitution recognized as legitimate political subjects "the working class and all working people" and "nations and nationalities" (Hayden 1999: 75). Communist ethnic policy legitimated and tolerated a plurality of ethno-regional identities and interests, but neither institutionalized nor tolerated pluralism of collective identities and interests, including classes (Burg 1983: 307; Tome 1988: 75; Vujacic and Zaslavsky 1991: 133, 139). There was no collective bargaining at the federal level between independent associations of unions and employers, but there was constant bargaining between ethnicized Communist parties over economic and social policies. Members of federal organs frequently claimed to represent the interests of their respective nations or republics (Burg 1983: 192), but it was ideologically unacceptable for them to represent middle class interests. At the same 86 time, alternative, non-ethnic political identities were delegitimized (Sekelj 1990: 150).

Deep institutional entrenchment of ethnic identities and lack of institutional recognition of non-ethnic collective identities channelled political protest along the lines of ethnicity.

The fatal weakness of the Yugoslav arrangement was its inability to encourage and represent country-wide interests, such as economic interests, in federal decision making (Burg 1988: 20, 21; Vujacic and Zaslavsky 1991: 133, 138). Lacking a strong institutional foundation, genuinely pro-Yugoslav forces were no match for their ethnically exclusive competitors.

Three Paths Out of Yugoslavia

The combination of Yugoslav institutional dynamics and the external shocks of the 1980s created three distinct paths of ethnic conflict escalation: secession of the rich (Slovenia and Croatia), central secession (Serbia and Montenegro) (Conversi 2000), and staying till the bitter end (Bosnia and Macedonia). The developed republics had long complained about the waste of equalization funds by the poor and about the transfer of resources; these debates became increasingly intense with the economic crisis (Burg 1983: 187,

294). In 1977 a long and bitter debate took place at the federal level over equalization

support for Kosovo, with Slovenia and Croatia insisting it be a loan, and other republics

arguing it be a grant (Ramet 1984: 74).

In the late 1980s, faced with rising Serbian nationalism, the Slovenian Communist

Party decided to push for ethnic nationalism and independence (Jovic 2003: 448-9),

enabling this republic to escape future transfer payments and exit Yugoslavia without

leaving a large number of fellow Slovenians behind. Instead of struggling to construct a 87 democratic federal Yugoslavia, the Slovenian leadership correctly calculated they would be better off on their own. The timing of Slovenia's secession supports the resource mobilization theory, not the containment argument. In light of the evidence, it is impossible to sustain the claim that Slovenians seceded first because they had the lowest standard of living or the least self-government. In fact, Slovenia was the wealthiest republic, very ethnically homogenous, with no linguistic assimilation. Slovenian nationalists had the resources necessary for secession - a wealthy and educated population, combined with a high level of ethnic completeness. While all republics eventually seceded, those with the most resources were the first to go.

The titular nationalism of Kosovo Albanians triggered out-migration of the

Kosovo Serbian minority. This, in turn, created a major opportunity for nationalism within the Serbian Communist Party. Having gained power in Serbia thanks to Serbian ethnic mobilization on the Kosovo issue, Milosevic used his position to push for strengthening the Yugoslav central government. Initially, he had the support of poorer republics, who hoped that an effective central government might reinvigorate the economic equalization program. Inside Serbia, his base of support came from a hard-line pro-Soviet group in the Communist party and non-communist Serbian nationalists

(Cohen 1995: 54). Milosevic's tactic was to use Serbian overrepresentation in the CPY to drive a Serb-led recentralization of the Party via the Party Congress (Jovic 2003: 458,

470). However, non-Serbs rapidly became alarmed at the repressive and discriminatory policies imposed by the Milosevic regime on Kosovo Albanians (Gligorov 1994: 70).

The regime's high level of tolerance for Serbian nationalism also triggered fears that a

Yugoslavia constructed by Milosevic would be Serb-dominated, similar to the inter-war 88

Kingdom (Bilandzic 1986: 207-8).

Table 3.6 Personal Attachment of Yugoslav Citizens to Territorial Organization, May- June 1990 (in percent) Level of Territorial Affiliation Ethnicity Republican/Provincial Yugoslavia Bosniacs 50 84 Montenegrins 47 80 Yugoslavs 32 71 Serbs 51 71 Hungarians 62 69 Macedonians 52 68 Others 43 58 Albanians 47 49 Croats 51 48 Slovenes 66 26

Total Sample 52 62 Source: Cohen (1995: 173) Note: Percent of respondents in each ethnic group who felt a particular level of affiliation was "quite important" or "very important" for them personally, based on interviews with 4,230 randomly selected adults throughout the country.

When the attempt to gain control over all of Yugoslavia failed, Milosevic used his influence on the Yugoslav military to carve up Croatia and Bosnia and to create, if necessary, by war and ethnic cleansing, a rump Yugoslavia encompassing Serb-majority territories. Consequently, the disintegration of Yugoslavia, unlike the break up of

Czechoslovakia, was a bloody and tragic process. While Milosevic and Serbian ultranationalists bear the lion's share of responsibility for the brutality of the Yugoslav disintegration, the disintegration itself was made possible by excessive decentralization, which enabled the ethnic republics to bring down the federal government and tear the country apart.

Some smaller Yugoslav nations (Bosniac, Montenegrin, and Macedonian), showed a strong level of support for a Yugoslav union (see Table 3.6). Their republics were the chief beneficiaries of the federal equalization fund, so their loyalty reflected 89

their economic interests (see Table 3.7). In addition, their high level of internal ethnic

incompleteness made them vulnerable to second order secessionist movements if the

disintegration of Yugoslav triggered multiple secessionist movements. The continued

support of Bosnia and Macedonia for Yugoslav unity supports Hechter's expectations

that recognition and decentralization can contain nationalist excesses or the demand for

an independent state. Intermarriage rates and percent of self-declared Yugoslavs in the

given republic/province do not have a clear effect on the timing of the secessions.

Table 3.7 Predictors and Timing of Secessions of Yugoslav Republic and Provinces

Republic Equalization Social Inter Percent Completeness, 1981 Date of Fund Use Product marriage Yugoslav, (percent) Secession (percent) rate 1981 Internal External Slovenia 0 205 10.9 1.4 90.52 97.66 June 25,'91 Croatia 0 127 17.0 8.2 78.62 75.08 June 25,'91 Bosnia 27.9 64 12.0 7.9 97.46 39.52 March 1,' 92 Macedonia 9.9 68 8.2 0.7 67.01 95.49 Sep 8,' 91 Montenegro 19.6 71 13.8 5.3 68.54 69.17 June 3, '06 Kosovo 42.6 29 6.1 0.2 77.42 70.89 Feb 17, '08 Notes: Equalization fund use: disbursement of equalization funds, 1981-85 Schedule. Social Product: region's per capita social product as a percent of Yugoslav average, in 1978. Percent Yugoslav: percent of self-declared ethnic Yugoslavs in the population of the region, 1981 census. Intermarriage rate: intermarriage as a percent of all new marriages in the region in 1981. Sources: Ramet(1984: 199); Petrovic (1985: 30, 60).

Discussion

The Unintended Consequences of Yugoslav Ethnic Federalism

Drawing on Hechter's theory, this study identifies the institutional arrangements and

policy responses that turned an ethno-federal "cure" into "poison" in Yugoslavia. My

analysis points to the peculiar institutional logic of Yugoslav ethno-federalism as the root

cause of the country's disintegration. The assumption that inter-ethnic equality demands

that each nation have titular status in its own ethnic republic did not make much sense in 90 light of the multi-ethnic character and geographic dispersion of Yugoslav ethnic groups, a situation made worse by the dogmatic pursuit of Lenin's version of the deprivation theory of minority nationalism. In response to each wave of nationalism, the centres imposed new doses of decentralization, a "cure" according to the logic of deprivation theories, which assume that increased decentralization will reduce minority deprivation and, hence, the reason for secessionism.

Unfortunately for Yugoslavia, the effect was the opposite, and decentralization became a self-reinforcing process. As decentralization created incentives and resources for nationalist mobilization, the centre reacted with purges and further decentralizations.

Waves of decentralization and nationalist mobilizations became a vicious circle.

Ethnification of representation in central institutions led to the disappearance of any institution that could transcend ethnic cleavages. Alternative paths to ethnification were closed off by weak institutionalization of non-ethnic identities (class, region, gender), a lack of recognition of nested national identities (such as Albanian-Yugoslav), and a lack of state support for Yugoslav national formation. The institution of multiple veto points at the federal level made any escape from the disintegrative downward spiral extremely difficult. Communist ethnic policies, instead of containing nationalism, continued to produce it (Vujacic and Zaslavsky 1991: 131)

Ethnic equalization policies such as those advocated by Hechter can strengthen support by recipients of funds, but they are also likely to strengthen secessionist tendencies among the federal units that fund the programs. In other words, these policies can buy the loyalty of less developed regions but trigger the secessionism of the wealthy.

Since regions are associated with specific ethnic groups, economic conflict easily 91 becomes ethnicized in the refusal of wealthy ethnic regions to pay for "lazy" ethnic others. An alterative approach to the problem of unequal regional development would be to increase capital and labour mobility, thereby enabling investment in regions with a cheaper labour force and a concomitant movement of the unemployed to regions with low unemployment. However, this approach would be incompatible with the institutional logic of titular nationalism and national economies. Another alternative would be to develop social welfare policies at the federal level and engage in non-ethnically specific policies aimed at supporting the economically vulnerable, regardless of their ethnicity or region of residence (Wilson 1978; 1987). In the case of Yugoslavia, this would have been incompatible with the decentralizing drive of the Yugoslav system after the mid-1960s, which shifted responsibility for social policy away from the centre.

In Yugoslavia, titular policies and the designation of ethnic homelands poisoned ethnic relations by effectively institutionalizing majority racism against minorities.

Arguably, the focus should have been on "de-titularization," or the creation of a more inclusive civic sense of collective ownership of territory and resources, instead of policies institutionalizing ethnically-specific first and second-class citizenships. The Yugoslav federal experience also suggests that titular nationalism in ethnic republics can produce far-reaching counter-mobilization by the most numerous ethnic group at the country level. Anger over discrimination against Serbs in Kosovo and their flight from non-titular status in other parts of Yugoslavia was a major resource for Milosevic and Serbian nationalists. 92

Generalizability of Yugoslav Experience?

While the present analysis indicates major disintegrative effects of certain ethno- federalist institutions and policies in the critically important Yugoslav case, these findings cannot justify a total rejection of ethno-federal forms. We need to distinguish between Yugoslavia's idiosyncratic features and the general problems of other ethno- federal states. To assess the comparative and theoretical significance of my findings, it is necessary to establish which causal patterns might explain other cases and if so, how their significance can be tested elsewhere.

To begin, nationalism in the vanguard might also be found in other Communist ethno-federal systems. Two other communist federations (USSR and Czechoslovakia) also massively decentralized in the 1960s (Bunce 1999: 51). Bunce (1999: 51, 85) argues that declining rates of growth and a corresponding decline in economic resources led

Breznjev, Husak, and Tito to attempt to gain the loyalty of regional elites by allowing them a wider sphere of "action" in local matters. That is, the centre tried to compensate for declining resources by granting increasing autonomy. It would be interesting to see whether this finding applies to the management of the sensitive issue of decentralization and cadre policy in Tibet, Xinjiang, and Inner Mongolia by the Communist Party of the

People's Republic of China. Still, the vanguard nationalism argument cannot be logically extended to multiparty democracies, as they (by definition) feature a multiplicity of parties that have to deal with the national question in (at least somewhat) transparent

inter-party competition, as opposed to internal one-party struggle.

The scope of other causal mechanisms appears more promising. In principle,

ethnic equalization policies could trigger secessionism of wealthy regions in democratic 93 ethno-federations. Similarly, titular nationalism and discrimination against regional minorities that trigger withdrawal to safer territory (for example, non-Francophone migration from Quebec) might be a general structural problem of ethnic federations.

Ethnicity-based sub-national governments may fail to integrate or tolerate ethnic Others

(Kalin 2004: 306) and degenerate into a "government of ethnic majority, by ethnic majority, for ethnic majority" (Hayden 1999: 15). Finally, democratic federal governments may unintentionally destroy unity by excessively institutionalizing ethnic representation and weakening institutions otherwise able to transcend ethnic cleavage.

Without further empirical testing, these are only plausible hypotheses. The next

step would be a comparative historical analysis of other federal systems with similarly high levels of decentralization who have managed, at least for the time being, to avoid the path to disintegration. A logical candidate would be Switzerland, as this ethnic federation had levels of fiscal decentralization similar to Yugoslavia in the 1980s (Hechter 2000a:

148), but managed to contain ethnic conflicts in the 1990s. Systematic comparisons of

this type would certainly help to refine the hypotheses.

While the historical sociology method is useful to test competing arguments about

long-term causal processes in a few cases, it is of limited use in generalizing findings.

Even if historical analysis indicates the existence of hypothesized causal mechanisms, the

generalizability of these mechanisms must be tested by large N quantitative studies of the

effects of ethno-federalism on multi-ethnic societies. The single-country study tests

causal mechanisms by explaining change over time; the cross-country study explains

cross-country variations, thereby establishing the scope of the applicability of the

mechanisms. As mentioned above, even if the historical analysis of critical cases 94 indicates the existence of hypothesized causal mechanisms, their generalizability must be tested by large N quantitative studies of ethno-federalism in multi-ethnic societies.

Nevertheless, both social movement theory and Hechter's U-shaped relationship between decentralization and ethnic conflict imply that excessive forms of decentralization can have a disintegrative effect. This study identifies a number of specific institutions and processes that had disintegrative effects in the Yugoslav case. As the case shows, carving a country into a set of supposedly ethnic republics with multi ethnic populations can lead to institutionalized majority racism against minorities. Ethnic equalization policies can buy the loyalty of less developed regions but trigger the

secessionism of the wealthy, while ethnic equalization policies based on ethno-territorial autonomies can lead to ethnification of the redistribution policy. A central government which repeatedly responds to secessionist pressures with decentralization can create a never-ending cycle of decentralization and nationalist mobilization. And attempts to

accommodate secessionists by increased ethnification of federal institutions and reduced representation of non-ethnic cleavages can unintentionally eliminate institutions with the

potential to transcend ethnic cleavage, thus creating systemic paralysis. As the Yugoslav

case shows, certain forms of ethno-federalism can empower secessionist nationalists to

break up a country. 95

Chapter Four: The Path to Weimar Serbia? Explaining the Resurgence of the Serbian Far Right After the Fall of Milosevic

Introduction

During the 1990s, the Serbian Radical Party [SRS] was the most successful far right party in post-communist Eastern Europe. It was the most radical partner in Milosevic's governments in the 1992-1995 and 1998-2000 periods, pushing the Serbian state towards inter-ethnic confrontations in both Bosnia and Kosovo. However, in October 2000, the

Serbian democratic opposition overwhelmed the Milosevic regime, Milosevic and a number of close collaborators were sent to the International Court of Justice in The

Hague, and support for his party collapsed. One would expect the far right Radicals to also fade away.

Instead, Serbian ultra-nationalists are enjoying a surprising resurgence in political

support. In the December 2003 elections, the SRS emerged as the largest single party in the Serbian Parliament. Then, in the June 2004 Serbian presidential elections, the SRS

candidate won 45 per cent of the popular vote, coming close to victory. More recently, in

the September 2004 local elections, the SRS won control of Novi Sad, Serbia's second

largest city. Finally, in the January 2007 parliamentary elections, the Radicals gained

even more votes than in 2003. Instead of collapsing, the Radicals are close to regaining

power. If this trend continues, it will have profound implications on the process of

democratic transition in Serbia, on inter-ethnic relations within Serbia, and on regional

political stability in Southeastern Europe.

It is puzzling that so many Serbs vote for the ultra-nationalists, given the miseries

ultra-nationalism brought them during the 1990s. While media hegemony and electoral 96 fraud were indisputably among the reasons for the electoral successes of ultra-nationalists during the 1990s, with the fall of Milosevic, these two factors have been essentially irrelevant. There have been few reports of irregularities in recent elections. In fact, the respected Belgrade Centre for Free and Democratic Elections [CeSID] declared the

December 2003 elections to be the most properly implemented elections in Serbia's history. Furthermore, after 2000, the Radicals (and Milosevic's Socialists) lost control of the state media; nor were they able to afford much time on commercial stations. Indeed, during the December 2003 elections, the major television stations showed a slight bias towards the governing Democratic Party.41 Yet in spite of mass media messages (or the lack of them), the Radicals received twice as many votes as the Democrats. To understand the re-emergence of the Radicals after 2000, then, new explanations are needed.

To deal with the SRS resurgence, I draw upon two well-established research traditions that have rarely been applied here: studies of the Western European far right and North American inter-racial relations. Specifically, I test the economic vulnerability hypothesis, based on Kitschelt and Wimmer's analysis of 'welfare chauvinism,' and the ethnic threat hypothesis, based on Blalock and Olzak's analysis of ethnic competition.

These hypotheses argue that if local contextual factors increase a majority's sense of

economic vulnerability and ethnic threat, then the majority group members are more

likely to respond positively to the far right message.

The Medium Group agency's monitoring of Serbian mass media from December 8 to December 24, 2003, shows that the (highest ranking) Democratic Party was the subject of television reporting for 7,151 seconds, while the Radical Party received only 4,070 seconds (Media Monitoring (Novi Sad, Serbia: Medium Group, 2003), p. 4). 97

Empirically, I use a multivariate analysis of municipal-level data on the post-2000 election results. Previous research on ultra-nationalism in Serbia includes political economy, historical sociology, institutionalist analysis, and survey-based social psychological studies (Woodward 1995; Cohen 2001; Goati 2001; Antonic 2002; Pantic

2002). However, multivariate statistical analyses have rarely been used, with the notable exception of several studies by Sekulic and associates (1994a, 1994b, 1999). Even so,

Sekulic's studies use attitudinal data, not election results.

Arguably, by studying the successes of the Serbian ultra-nationalists, we may better understand the vulnerability of democratising polities to an upsurge of nationalism.

On the basis of a comparative historical study of fascist takeovers, Juan Linz (1980) points to the apparent paradox that fascist movements can only grow in liberal democracies. Since fascist parties do not function as Leninist-type insurrections, but rather, as nationalist-populist movements, they require political liberty to mobilise support. In other words, fascists need democratic institutions to capture power and then suppress democracy (Payne 1995: 490). Similarly, Snyder argues that the recently democratised regimes are exceptionally conflict-prone, especially if the authoritarian elite facing strong democratic challengers try to de-mobilize the democratic constituency by pursuing nationalist mobilization for diversionary wars (Snyder 2000: 28-9, 198;

Mansfield and Snyder 2005: 29). Inter-ethnic warfare can produce strong domestic nationalist and militarist institutions and ideologies, as well as lasting external rivalries.

In the end, repeated failed democratisations, strong ultra-nationalist institutions, and recurring inter-ethnic warfare can become a vicious circle (Mansfield and Snyder 2005). 98

In the Serbian case, the ability of ultra-nationalists to capture power and control the coercive apparatus of the state has led to the escalation of inter-ethnic conflicts into military confrontations and ethnic cleansing. In short, whenever Serbian Radicals enter government, Serbia is soon at war.42 The larger implication is that as democratic institutions spread into previously authoritarian multi-ethnic countries, the sustainability of the transition depends on the ability of democrats to prevent ultra-nationalist takeovers.

The European Far Right: Defenders of 'White Christian' Europe

While existing theoretical and empirical works on the Western European far right are useful for similar research on Eastern Europe, researchers need to take into account certain important differences in regional conditions. For one thing, while Western

European far right parties have focused on the 'defence' of their own 'homogenous' nations as well as 'Fortress Europe' from the arrival of the allegedly inassimilable immigrant 'ethnic other,' the Eastern European far right parties have pursued territorial disputes with neighbouring 'European' nations (Eatwell 2004: 9).43 In addition, the

Eastern European far right has often focused on supposedly disloyal and secessionist national minorities. For example, the Serbian SRS has openly called for the repression of

42 The Radicals supported a Socialist minority government from 1992 to 1993. In this period, Serbia participated in wars in Croatia and Bosnia. From 1998 to 2000, the Radicals and Socialists formed a coalition government. This time, Serbia was at war with Kosovo Albanians and NATO. 43 There are exceptions to these tendencies. The Vlaams Blok in Belgium and the Northern League in Italy have focused on territorial disputes with the central governments, while German Republicans disputed recognition of the border with Poland (Betz 1994, p. 133; Gibson et al. 2002, p. 839). Yet the intolerance towards Roma minorities by Czech Republicans, the anti- Semitism of the Party of Greater Romania and Russian far right militants, or the hostility of Zhirinovsky's LDPR towards immigrants from the Caucuses are similar to the anti-immigrant hostilities of the Western European far right (Brym 1994; Pehe 1996; Shafir 1999, p. 214). 99 the ethnic Albanian minority (Cohen 2001: 227). Just as the increased presence of immigrant minorities has been a political resource in the far right mobilisation in Western

Europe, so too territorially concentrated national minorities have been indispensable in the creation of far right propaganda in Eastern Europe. Unlike their Western European

counterparts, however, the Balkan far right parties have put their ideas of ethnic purity

into practice - forming paramilitary units and 'cleansing' territories of ethnic others

during the post-Yugoslav wars.

Despite these differences, the far right parties in both parts of Europe have a

considerable ideological affinity, mostly in terms of their shared animosity towards 'non-

Europeans,' primarily European Muslims and Americans. Muslims are often portrayed as

anti-Western, intolerant, inassimilable, and highly fertile, thereby threatening a

demographic take-over of European countries. Unlike the supposedly 'barbaric' and

'backward' Muslims, the US is portrayed as the promoter of 'corrupted' forms of

modernity, such as the despised move towards globalisation or the 'horrors' of

multiculturalism (Weinberg 2003).

The Crisis of Serbian Society and the Rise of Serbian Radicals

The rise of the Serbian Radical Party needs to be placed in the context of the protracted

and traumatic process of the disintegration of multi-national Yugoslavia. During the

escalation of the Kosovo conflict and the Milosevic rule (1981-2000), Serbia's once

promising economy and stable ethnic relations experienced a catastrophic deterioration.

In the late 1980s, the Milosevic government unilaterally abolished the autonomy of

Kosovo province and used police repression and state discrimination against ethnic 100

Albanians (Cohen 2001; Poulton 2003:132-135). Then, during the 1990s, Serbian regular troops and paramilitaries participated in the wars and ethic cleansing campaigns (Mann

2005).

Wars and mutual ethnic cleansing in Croatia, Bosnia, and Kosovo have brought waves of mostly Serbian refugees into Serbia. In January 2004, there were 514,513 refugees out of a total population of 7,498,001 in Serbia and Montenegro (without

Kosovo) (UNHRC 2004; Serbian Statistics Bureau 2005). The UN economic sanctions, the Milosevic regime's mismanagement of the economy, and NATO air raids in 1999 have devastated Serbia's economy. In 1993, Serbia experienced hyperinflation, almost 60 per cent unemployment, GDP per capita of $1,250 US (in 1989 it was $3,300 US) (Cohen

1995: 353-4; Cohen 2001:162). In 2001, only about 39.7 per cent of the working-age population was registered as employed (Serbian Statistics Bureau, 2005).

Despite the deterioration of economic conditions, the Milosevic regime tried to maintain the Communist era welfare state. During most of the 1990s, the state prohibited firing of workers, which provided job security, but combined with declining output, also led to a fall in productivity (Mijatovic 2005: 295, 325). The Serbian welfare state continued to provide a wide variety of universal social services, such as health care and child care, but the quality has declined over time, because of lack of new investment in the infrastructure and arrears in transfer payments.

In the cultural sphere, the failure of a more inclusive Yugoslavism coincided with the 'return' to a more ethnically-exclusive Serbian identity. The ultra-nationalistic discourse of leading Serbian intellectuals (Dragovic-Soso 2002), the hate propaganda in government-controlled mass media (Thompson 1999), and escalating ethnic conflicts led 101 to a decline in census self-identification as Yugoslav,44 the 'brightening' of once-blurred boundaries with other ethnic groups (especially Croats and Bosniacs), and a dramatic increase in ethnic distance (Golubovic 1995).

As some former 'Yugoslavs' were turning into 'Serbs,' the formation or 're discovery' of the more exclusive national identity was often accompanied by an increasingly hostile perception of their former fellow Yugoslavs. Serbian ultra- nationalists vehemently denied that the Serbs had a privileged position in Communist

Yugoslavia,45 denied or downplayed the major responsibility of the Serbian political elite in the crisis of Yugoslav federalism, and used biased reinterpretations of the common history to justify intolerant treatment of other ex-Yugoslav ethnic groups (Dragovic-Soso

2002). The crimes of Croat fascists during WWII against Serbian and Jewish populations, and the Ottoman era persecutions of Orthodox Christians by Bosnian and Albanian

Muslims were cited as examples of the immutably evil collective essence of these other nations, which, in turn, justified Serbian distrust, intolerance, or 'revenge' (Brubaker

1998: 304; Skopljanac et al. 2000; Thompson 1999). Finally, the recognition and national formation of Macedonians, Montenegrins, and Bosniacs in the Communist period was represented as a 'false consciousness' of'artificial nations' who should now 'return' to their Serbian essence (Milosavljevic 2002: 191-231).

While sections of Serbian civil society and democratic opposition parties rejected this kind of Serbian national formation, Serbian Radicals distinguished themselves by

44 In the 1981 Yugoslav Census, 5.76 percent of Yugoslav citizens self-declared as 'Yugoslavs in the ethnic sense.' In the 1991 Census, this declined to 2.98 percent (author's computations from Census tables) (Yugoslav Federal Statistics Bureau 1998). 45 While it is debatable whether the Serbs were privileged vis-a-vis the Slovenes or the Croats, the ethnic composition of the key party and state agencies leaves little room for doubt that traditionally Christian ethnicities (Slovenes, Croats, Serbs, and Montenegrins) were in a privileged position vis-a-vis Muslim ethnicities (Bosniacs or Albanians) (Szayna 2000). 102 their ultra-nationalism, even in this xenophobic atmosphere. The Radicals repeatedly criticised the Milosevic regime for betraying the cause of Serbian unification (Markotich

1995a; Skrozza 2005), presenting themselves as firm, uncompromising, and totally committed 'Serbian patriots' (Markotich 1995b; Smajlovic 1997). The Radicals consistently supported the 'liberation' and unification of all 'Serbian lands' into a strong

Greater Serbia (Vucic 1995: 25; Zemunske Novine 1997: 19).

From 1991 to 1993, the SRS sent party paramilitaries to the front lines in Croatia and Bosnia. Reportedly, Seselj incited the mass murder of Croats in November 1991

(ICTY 2003). Just before the outbreak of war in Bosnia, during an SRS rally close to the

Serbian-Bosnian border, Seselj promised, 'We are going to clean Bosnia of pagans

[Bosnian Muslims] and show them a road which will take them to the East, where they belong' (ICTY 2003). Within Serbia, the SRS was calling for confrontation with 'disloyal minorities' and insufficiently 'patriotic' Serbs. In 1991, the SRS activists were reportedly

involved in the expulsion of ethnic Croats from Hrtkovci, a mostly Croat village in northern Serbia. And the Radicals and Milosevic were in the coalition government during the 1999 Kosovo war, when Serbian forces cleansed at least 862,979 ethnic Albanians

from Kosovo (HRW 2001: 134).

The electoral defeat and the downfall of the Milosevic regime in October 2000

seemed to mark the end of the horrors of ethnic cleansing and a new beginning for

Serbia. Milosevic and a number of his senior associates were sent to the International

War Crimes Tribunal in The Hague, and the Serbian criminal justice system began

prosecuting a number of lower-level perpetrators of war crimes. The new democratic

government achieved macro-economic stabilisation, signed an agreement on association 103 with the EU, placed a number of ethnic minority politicians in key cabinet positions, negotiated power sharing between Albanians and Serbs in the Presevo valley, and started negotiations on the future status of Kosova.46

However, the dark legacy of the Milosevic era lives on, making Serbia's road to a successful multi-ethnic democracy and EU membership extremely difficult. For example,

Zoran Djindjic, Serbia's charismatic democratic Prime Minister, was assassinated by ultra-nationalist state security troops in March 2003. The ultra-nationalist groups associated with the Army of Serbia and Montenegro continued to shelter major Serbian fugitives from The Hague's War Crimes Tribunal (Whitmore 2005). Further, by 2005, economic transition was entering a critical phase, with the IMF demanding dramatic cuts in social spending, regardless of the effects on economically-vulnerable sections of the population (IMF 2006: 29). In 2006, Serbian democrats failed to persuade the majority of citizens of Montenegro to stay unified with Serbia, and influential Western governments rejected their demand to keep Kosovo in Serbia. In January 2007, the Radicals achieved their best results in Parliamentary elections in the post-Milosevic period.

Theoretical Explanations of Intolerance and Support for Ultra-Nationalists

Ethnic Threat

What predicts majority intolerance and support for ultra-nationalists? Blalock, Olzak, and

Nagel have argued that the increased presence of minorities leads to a more intensive inter-group competition for scarce resources, such as jobs (Blalock 1967; Olzak and

Nagel 1986). Stemming from this argument, the 'realistic conflict' theory implies that

461 use interchangeably the Albanian (Kosova) and Serbian (Kosovo) names for Serbia's southern province whose future status remains bitterly disputed. 104 higher levels of minority presence will be positively related to majority intolerance and discrimination.

In support of the realistic conflict argument, multi-level US studies have found that whites' anti-Black attitudes are positively related to the per cent of Blacks in the region (Fossett and Kiecolt 1989, p. 828; Kalin 1996, p. 171; Oliver and Wong 2003, p.

568). Bergesen and Herman (1998) also report that the rapid migration of an Asian-

American minority into predominately African-American neighbourhoods was associated with local backlash violence during the 1992 Los Angeles riots. In Western Europe,

studies have found that the per cent of ethnic minorities in a region is one of the strongest contextual predictors of the percentage vote for far right parties (Betz 1994; Eatwell

2000a, p. 178; Lubbers and Scheepers 2002). In repeated surveys, French National Front

sympathisers report that immigration policy is a primary cause of their decision to

support the National Front (Golder 2003). David Ost even argues that a major reason for the failure of the Polish far right to make an electoral breakthrough is the lack of locally

available ethnic minorities. In his view, without a few credibly threatening ethnic others -

'politically indispensable demons' - the ability of the far right to mobilise the ethnic

majority can be severely constrained (Ost 1999).

While it is generally recognised that the sense of ethnic solidarity with Serbs

outside Serbia was a major resource for the Milosevic regime (Goati 2001), the sense of

ethnic threat experienced by Serbs in Serbia has not received much attention. A major

exception is a study by Massey, Sekulic, and Hodson. Using 1989-1990 Yugoslav survey

data, collected just before the outbreak of war in Croatia, these researchers found the

highest levels of ethnic intolerance were expressed by an 'enclaved majority,' a group 105 that is a majority at the level of the republic, but a local (municipal-level) minority. They argue that the ethnic majority might experience its local minority status as 'victimization'

(Massey, Sekulic, and Hodson 1999). Following this argument, I expect that the Serbs of

Serbia who live in municipalities in which they have reason to fear other ethnic minorities or neighbouring states are more likely to support the ultra-nationalists than are those in 'safe' or 'undisputed' Serbian core areas.

Serbs represent a majority in Serbia as a whole,47 but they are a local minority in

Raska/Sandzak (a large Bosniac/Muslim minority), Kosovo and Presevo (a large

Albanian minority), and Backa (a large Hungarian minority). This, plus a further rapid reduction in the percentage of Serbs in a region, caused either by ethnic differences in fertility rates, or by ethnic differences in out-migration rates, can lead to Serbian insecurity. In fact, after their loss of control of Kosovo, many ethnic Serbs see any kind of territorial autonomy for minorities as a threat to Serbia's territorial unity (Bakic 2004a: p. 13). In a 2001 Serbian survey, 81.8 per cent of respondents rejected ethno-territorial autonomy for ethnic Albanians in the Presevo region, and only 3.6 per cent supported an introduction of ethno-territorial autonomy for Hungarians in Northern Serbia (CPIJM

Data Set, Studies JMS-160 and JMS-162).48 Finally, the sense of'ethnic threat' is likely to be heightened if members of the locally-present minority are engaged in secessionist conflict with the Serbian state, as was the case with some members of the Kosovo

Albanian minority. Arguably, Serbs who feel threatened by the ethnic 'other' might be

47 In the 1991 Census, the total population of Serbia was 9,778,991: Ethnic Serbs about 65 percent, Albanians 17 percent, Hungarians 4 percent, Yugoslavs 3 percent, and Bosniacs/Muslims 3 percent (Yugoslav Federal Statistics Bureau 1998). 48 The CPIJM Data Set has been collected by the Centre for Political Studies and Public Opinion Research, Belgrade Institute for Social Sciences. The Centre collaborates with the Institute for Social Research at Michigan University, collecting country level data for the 'World Values Survey.' 106 more inclined to support ultra-nationalist parties, who promise to defend the unity and territorial integrity of Serbia by confronting the perceived secessionist minorities and aggressive neighbouring states. In sum, the percentage of minorities in the municipality is likely to be positively associated with the percentage of majority voting for the SRS.

Scholars examining American evidence have argued that the effect of minority percentage on the majority's intolerance is likely non-linear. Blalock (1967) posits that the relationship between minority percentage and majority discrimination should be positive with decreasing slope. Blalock's proposition has been supported by subsequent empirical research. Schuman et al., for example, found that white Americans are much more positive about school and residential integration involving a small number of

African Americans, than they are about integration when African Americans are a majority or almost a majority (Schuman et al. 1985). Hostility seems to peak when

Blacks represent 40-60 per cent of the local population (Kalin 1996, p. 172). Similarly, an analysis of French election results in 1984 and 1997 discovered that the departments with the largest concentration of people of North African descent consistently attract about twice the level of support for the National Front than those with the smallest

concentrations (Minkenberg and Schain 2003). The researchers therefore argue that the

majority is more likely to accept a minority if its own demographic hegemony is not

threatened (Blalock 1967; Schuman et al. 1985; Kalin 1996, p. 172.). In the case of

Serbia, I expect that the relationship between the percentage of minorities and the

percentage of the majority voting for the SRS will be positive with a decreasing slope.

Finally, municipalities with high percentages of refugees might be more likely to

support ultra-nationalists. Research on other inter-ethnic conflicts shows that refugees 107

(often expelled from their homes by 'ethnic enemies') are likely to be radicalised by their traumatic experiences and, thus, are more likely to support ultra-nationalists. For

example, refugees from 'unredeemed' lands were a major constituency of the extreme nationalist parties in inter-war Hungary, and of far right parties in the West German

elections of 1953 and 1955 (Kolsto 1995, p. 298; Brubaker 1996, p. 178). Further, the

associations of the pieds-noirs, the European settlers forced to flee to France after Algeria

achieved independence, have provided strong support of the French National Front

(Veugelers 2005).

In the Serbian case, the vast majority of refugees are ethnic Serbs who have fled

Croatia, Bosnia, or Kosovo. In 1995, about 650,000 refugees were registered in Serbia: of

these, 78 per cent were ethnic Serbs, 10 per cent ethnic Yugoslavs, and 10 per cent

Muslims (Raduski 2001, p. 118). In many cases, refugees joined extended family

members in Serbia. Ninety-five per cent of the Serb refugees from Bosnia were

accommodated outside state refugee camps (Raduski 2001). Arguably, the presence of

newly-arrived victimised family members might have a radicalising effect on the native

Serbian population. Thus, I expect to find a positive relationship between the percentage

of refugees and the percentage of the majority population voting for the SRS.

Economic Vulnerability

Kitschelt argues that individuals with resources that could be economically beneficial in a

market economy will tend to support market transformation, while those who are more

dependent on extensive welfare state policies will vote against pro-market parties. That

is, potential winners (professionals, small business people, the well-educated, the urban, 108

and the young) are more likely to vote for pro-market parties than are potential losers

(pensioners, the poorly-educated, and unqualified workers) (Kitschelt 1995a). Kitschelt

goes so far as to predict that in economically depressed, post-communist Europe, 'welfare

chauvinism' (targeting social services to the ethnic majority) might be an even more

successful strategy than in Western Europe (Kitschelt 1995b, p. 23).

However, one might wonder whether the welfare chauvinism thesis can be

credibly extended to Eastern Europe. Several post-communist states have experienced a prolonged and extremely severe economic recession, coupled with a breakdown of the

state institutional capacities (Manning 2004, p. 219). It makes little sense to debate who

should be entitled to receive state services, if the state in question is unable to provide

any meaningful services in the first place. However, while the Serbian welfare state is

certainly not as well developed as some Western and Northern European states, economic

and social indicators show that it has continuously engaged in a major welfare effort.49

As the Serbian state provides valuable social services, the political debates about

entitlement are meaningful.

Welfare chauvinism in Serbia is the ideology of an economically-vulnerable

majority population that tries to monopolize access to scarce resources by portraying

some national minorities as 'disloyal' to the Serbian nation-state, and therefore

illegitimate competitors for these resources. Western European welfare chauvinists

49 In 2004, Serbian public expenditures were 42.3 percent of the GDP, within the average 40-45 percent of advanced reformers of Central Europe and the Baltic, and comparable to advanced EU countries (World Bank 2002, p. 81; OECD 2003, p. 64; Manning 2004, p. 220; Begovic 2005, p. 448). This is a serious welfare effort, significantly above the levels of countries experiencing state disintegration. Since the Serbian GDP per capita (in 2004, US$ 5,200 in purchasing power parity) is much lower than in the EU countries, the amount of spending per citizen that the state can afford is also lower (Popovic 2005, p. 36). Serbia experienced short but traumatic periods of state failure to provide basic services during the 1993 hyperinflation and the 1998 war against NATO. 109 typically argue that ethnic others are 'illegitimate competitors' for the scarce resources

such as jobs and public housing, which should be reserved for the 'real' members of the nation (Wimmer 1997a, p. 21; Wimmer 1997b, p. 655). Serbian welfare chauvinists agree. In a 1996 survey, respondents were asked whether they agree that 'when there is great unemployment, Serbs and Montenegrins should have a priority in getting jobs.'

Thirty-four per cent of all respondents agreed, compared to 73 per cent of those who were

strong supporters of the SRS (CPIJM Data Set, Study JJM- 129). After their electoral victory in Zemun, a suburb of Belgrade, the Radical municipal government took public housing apartments away from ethnic Croat families.

In the Serbian case, empirical support for the Kitschelt economic vulnerability thesis from individual-level data is remarkably strong. Studies of the electoral base of the

Radicals and Milosevic's Socialists indicate that the most economically-vulnerable

groups of ethnic Serbs have recognised these parties as the major champions of their

economic interests (Goati 2001). Notably, the inability of the democratic opposition to

formulate an economic transition program that clearly addresses the social security of the

economically-vulnerable section of the Serbian population has enabled the Socialists and

the Radicals to include these groups among their supporters.

In the post-Milosevic period, the Radicals have tried, and largely succeeded, to

capture traditional Socialist supporters by criticising privatisation and calling for

economic security. At a pre-election rally in Belgrade in December 2002, the Radical

leader received enthusiastic applause for stating that the new government 'is selling out

cheaply the property that generations have created with hard work.' Not surprisingly, the 110 majority of supporters attending the rally appeared to be elderly manual workers.50 By

August 2002, manual workers were more likely to support the Radicals than the

Socialists, while educated middle-class voters remained the base of the Democratic Party

(Slavujevic 2002).

While the studies based on individual-level data have found support for the welfare chauvinism argument, in this study I will test its validity with contextual factors.

That is, the underlying reasoning here is not that the unemployed are necessarily more likely to vote for the far right, but that in municipalities experiencing economic deprivation (such as high levels of unemployment), everyone - including the employed - is more likely to be fearful of the economic situation and thus more likely to vote for the far right. This argument follows the Western European theoretical arguments on the link between the subjective sense of economic vulnerability and far right voting (Eatwell

2000b, p. 418; Wimmer 2006, p. 66), as well as the repeated empirical finding that the unemployment rates and the far right vote are positively related at the local level, net of other variables in the models (Lubbers and Scheppers 2002, p. 123; Dulmer and Klein

2005, p. 250). Thus, I expect municipalities with higher levels of unemployment to be more likely to vote for the SRS.

Data and Measures

The dependent variable - electoral support for the SRS - is based on the election results reported by the Centre for Free and Democratic Elections [CeSID]. Only majority voters are taken into account, because survey findings indicate that minorities are extremely unlikely to vote for Serbian Radicals. Fewer than 4 per cent of ethnic Albanians,

50 Field Observations, December 2002. Ill

Muslims, or Hungarians have a positive opinion of the SRS leader, whereas a considerable percentage of ethnic Serbs, Montenegrins, and Yugoslavs (hereafter, the majority) express a positive opinion (see Appendix l).51 Estimates of majority registered voters are based on the assumption that the per cent of the majority in a municipality

(based on Census results) is the same as the per cent of registered voters who belong to the ethnic majority. The registered voter figures were collected from the CeSID data set.

Using municipality-level majority vote as the dependent variable enables a study based on a rich election data set, but it also poses unusual restrictions in terms of the meaningful interpretations of results. On the positive side, from December 2000 to June

2004, Serbia had seven national-level parliamentary or presidential elections. All

municipal-level results of these elections are in CeSID's data set, thereby enabling a time-sensitive analysis of a large body of data. On the negative side, correlations among

contextual-level variables cannot be used to make assertions about individual-level

relations. Thus, to avoid the ecological fallacy, analysis has to focus on macro-level

explanations (Snijders and Bosker 1999).52

I was unable to locate survey findings on the party preferences of ethnic Bulgarians in Serbia. Two mostly Bulgarian municipalities are outliers in terms of the values of the dependent variable. While it seems unlikely that ethnic Bulgarians would massively vote for the Serbian Radicals, this possibility cannot be dismissed. Opinion polls report that during the 1999 NATO air attacks on Serbia, about 50 percent of Bulgarians expressed support for fellow Orthodox and Slavic Serbs, and only up to 24 percent supported NATO (RFE, Newsline, March 30, 1999). 52 While it would be advantageous to incorporate individual level data and to conduct multilevel analysis of the SRS election results, inclusion of individual-level data would prevent the analysis of the impact of important municipal level predictors. The samples used in Serbian national level surveys are typically representative of the level of major regions of Serbia, not of all 160 municipalities without Kosovo. Thus, multilevel analysis of contextual effects would stay at the level of regions, and municipal-level effects would be lost. Moreover, Serbian survey samples routinely underestimate the SRS vote. This methodological problem is usually explained as a result of the social desirability effect: SRS voters are less likely to reveal their electoral preferences because of negative discourse about Radicals. 112

Ethnic threat is measured by the municipal per cent minority, derived from 2002

Census data, and the municipal presence of refugees.53 While a number of refugees escaped from the wars in Croatia and Bosnia in the early 1990s, two main waves of refugees arrived in Serbia in 1995 and 1999. According to CeSID data, early in 1995, there were 448,477 registered refugees in Serbia; they were joined by another 239,841

Serbs from Bosnia and Croatia after the Croatian army's 1995 offensive. According to the United Nations High Commissioner for Refugees (UNHCR) data, the 1999 wave consisted of 187,129 persons, predominantly Kosova Serbs (also 19,551 Kosova Roma) who were escaping from revenge attacks by Albanian extremists during and after the

NATO troops' arrival in Kosovo (UNHCR, p. 25).

Economic vulnerability is measured by several indicators. The unemployment variable captures the per cent of the working-age population registered as unemployed, as reported by the Monthly Bulletin of the Serbian Labour Market Bureau.54 This study uses data on municipal-level unemployment collected one to two months before the

While using the per cent minority as an indicator of ethnic threat is standard procedure, several researchers have expressed concerns about this approach. Since the realistic conflict theory assumes that what matters is what is perceived, using attitudinal measures of perceived threat is methodologically superior to the simple equation of threat to group size (Oliver and Wong 2003). While one has to concur with these arguments, the appropriate municipal-level attitudinal data are not available in this case. Still, the findings of previous studies lend credibility to the argument that many Radical party supporters have a high level of ethnic distance from minorities. Further, in an already-cited 2001 Serbian survey, 81.8 per cent of respondents rejected ethno-territorial autonomy for ethnic Albanians in Presevo, and only 3.6 per cent supported an introduction of ethno-territorial autonomy for Hungarians in Northern Serbia. Thus, it seems plausible to assume that the perception of ethnic threat will increase with the increased local presence of ethnic minorities. 54 It would be methodologically preferable to use municipal-level survey results on individual's fear of unemployment rather than contextual data on per cent unemployed. However, such data are rarely available at a local level, as national samples representative at a municipal level would be prohibitively expensive. Several recent studies of the Western European far right use the percent unemployed as a measure of local level feelings of economic vulnerability (for example, Lubbers and Scheppers 2002; Dulmer and Klein 2005), reasoning that the increased local presence of the unemployed will make others - including the employed - more fearful about their economic prospects, thus raising the general sense of economic vulnerability. 113 elections in question. Details on other indicators of economic vulnerability, namely, the availability of physicians, average wages, phone subscriptions, and municipal revenue, are provided in Appendix 4B. The 'physicians' variable refers to number of inhabitants

[in hundreds] per physician in the municipality. High levels of the variable imply low availability of physicians and greater economic vulnerability. 'Wage' refers to municipality-specific annual average monthly (after-tax) wages per employee, and captures dimensions of economic vulnerability distinctly different from the per cent unemployed. Since indicators were not available for all years, the exact mix of indicators in the model varies from election to election. Limited data availability is an impediment to the comparison of the models for different years. In line with the economic vulnerability argument, however, the expectation is that less-developed municipalities will be more likely to vote strongly for the SRS. The data analysis was conducted in

SPSS 14, using the ordinary least squares regression.

Results

The first elections after the fall of the Milosevic regime were held in December 2000. At this point, Milosevic was out of power but still in the country, his party was still united, and there was continued low-intensity violence between ethnic Albanian militants and

Serbian security forces in the three municipalities with a large ethnic Albanian population that border Kosovo. Popular approval of the democratic parties was very high, however, and the Radicals received only 8.59 per cent of all votes.

As expected, Model I in Table 4.1 shows that ethnic threat explains much of the support for the Radicals. The presence of minorities was a strong predictor of the 114 majority vote for the Radicals. Similarly, the presence of refugees from 1995 (refugees from Croatia and Bosnia) was positively associated with SRS electoral performance.

Unexpectedly, the presence of refugees from 1999 (refugees from Kosovo) was not associated with the level of support for the SRS.

Table 4.1. Regression of the Percentage Majority Vote for the SRS on Predictors, December 2000 Elections I II III Predictors B t Sig. B t Sig. B t Sig.

Ethnic Threat Minorities 0.140 15.010 *** 0.134 14.473 *** 0.221 7.409 *** Refugees 1995 0.053 3.092 ** 0.050 2.793 ** 0.039 2.216 * Refugees 2000 -0.033 -0.769 NS -0.020 -0.460 NS 0.024 0.550 NS

Economic Vulnerability Unemployed 0.129 3.448 ** 0.122 3.320 ** Physicians 0.021 0.467 NS 0.039 0.858 NS Wages -0.111 -0.363 NS -0.155 -0.519 NS

Non-Linear Effects Minorities Quadratic -0.001 -3.045 **

Constant 4.260 14.758 *** 2.294 2.077 * 1.782 NS Adjusted R2 0.651 0.673 0.690 N of municipalities 159 157 157

Significance: NS= not significant, fp<.10, *p<.05, **p<.01, *** p<.001.

As Models II and III show, the effect of economic vulnerability is more complex than anticipated. The effect of unemployment was straightforward: as expected, per cent unemployed had a strong positive association with the SRS vote. The number of physicians per 100 inhabitants was statistically insignificant, and it had an unexpected sign. Model IV shows that there is a significant quadratic term in the relationship between the per cent minority and the SRS vote. This finding supports Blalock's expectation that the relationship between the presence of minorities and the majority's intolerance will be positive with decreasing slope. Overall, the models for the 2000 115 elections explain between 65 per cent and 69 per cent of the variation among municipalities.

The September 2002 elections were held in a different political and economic context. By now, inter-ethnic violence in the Presevo region was replaced by carefully- negotiated power-sharing arrangements, Milosevic and a number of senior members of his regime had been sent to the International War Crimes Tribunal, and Milosevic's party had split into two factions. The reform of Serbia's economy had brought macroeconomic stability and some major foreign investments, but also the first waves of bankruptcies and massive downsizing, especially in the banking sector.

Table 4.2 shows that the results of the September 2002 Presidential elections reflect some of these changes. It seems that reduction in the objective ethnic threat (the end of ethnic violence in the South) reduced the significance of ethnic threat, while the increased pain of market reform, combined with the skilful SRS increase in anti- privatisation messages, helped the SRS capture more support from the now economically-vulnerable ethnic Serbs. As Model I shows, the ethnic threat indicators now explain only 45.3 per cent of the variability of the dependent variable, a drastic decline when compared to the Table 4.1 results. Furthermore, this time, neither of the refugee variables reaches statistical significance. While the 2000 models predict that a 10 per cent increase in unemployment will produce only a 1.3 per cent increase in the SRS vote, the 2002 model predicts that the same increase of unemployment will result in a 3.8 per cent increase, net of other variables in the model. As the economic vulnerability hypothesis would predict, greater scarcity of physicians is associated with higher levels of support for the SRS. Surprisingly, average wages, phones per capita, and municipal 116 revenue per capita do not have a statistically significant effect. Model III provides further support for Blalock's hypothesis.

Table 4.2. Regression of the Percentage Majority Vote for the SRS on Predictors, September 2002 Elections I II III Predictors B t Sig. B t Sig. B t Sig. Ethnic Threat Minorities 0.232 10.241 *** 0.174 8.616 *** 0.328 5.486 *** Refugees 1995 0.059 1.407 NS 0.052 1.324 NS Refugees 2000 0.042 0.402 NS 0.105 1.135 NS

Economic Vulnerability Unemployed 0.383 5.122 *** 0.387 5.271 *** Physicians 0.366 3.143 ** 0.404 3.845 *** Wages -0.312 -1.530 NS Phones -0.006 -0.926 NS Municipal Revenue 0.000 1.099 NS

Non-Linear Effects Minorities Quadratic -0.002 -2.515 *

Constant 13.596 19.466 *** 6.295 2.336 * 3.316 1.950 f Adjusted R2 0.453 0.529 0.539 N of municipalities 159 159 159 Significance: NS= not significant, fp<.10, *p<05, **p<01, *** p<.001.

The November 2003 Presidential elections were held in a generally similar political and economic context, and the results are similar to those for the 2002 elections.55 Per cent minority and per cent unemployed are again statistically significant predictors, and the quadratic term supports Blalock's hypothesis. There are, however, several new findings. The economic vulnerability variables have a suppressor effect on the 1995 refugee variable. An increase in the municipality's average wage of 10,000

Dinars (about 150 Euros in November 2003) reduced the predicted majority vote for the

SRS by 8.4 per cent. These findings support Kitschelt's argument.

The table showing November 2003 election results (available from the author upon request) basically repeats the findings shown in the Table 2. 117

Table 4.3. Regression of the Percentage Majority Vote for the SRS on Predictors, 13 June 2004, Presidential Elections (Round One) I II III Predictors B t Sig. B t Sig. B t Sig.

Ethnic Threat Minorities 0.245 9.828 *** 0.219 8.906 *** 0.276 3.356 ** Refugees 1995 0.078 1.697 t 0.132 2.726 ** 0.125 2.517 * Refugees 2000 0.297 2.562 * 0.279 2.514 * 0.309 2.605 *

Economic Vulnerability Unemployed 0.229 2.708 ** 0.221 2.585 * Wages -0.414 -2.513 * -0.423 -2.559 *

Non-Linear Effects Minorities Quadratic -0.001 -0.722 NS

Constant 13.266 17.239 *** 13.583 5.002 *** 13.468 4.944 *** Adjusted R2 0.470 0.519 0.517 N of municipalities 158 158 158 Significance: NS= not significant, fp<.10, *p<.05, **p<.01, *** p<.001.

The Presidential elections held on June 13, 2004, reflect a changed political

environment: anti-Serbian riots in Kosovo and anti-Muslim riots in Serbia in March 2004

(Human Rights Watch 2004) had shifted the political discourse. The defence of the

remaining Kosovo Serbs, the return of those who had been 'ethnically cleansed,' the

protection of Serbian religious monuments, and the future status of Kosova, were all back

at the centre of Serbia's political discourse.

Table 4.3 shows the decline in the significance of economic vulnerability and the

increased significance of ethnic threat predictors of SRS support. As Model I shows, for

the first time, the percentage of 1999 (Kosovo) refugees has a statistically significant

impact. Contrary to Blalock's argument, Model III shows that the quadratic effect of

percentage minority is no longer significant. 118

Table 4.4. Regression of the Percentage Majority Vote for the SRS on Predictors, 27 June 2004, Presidential Elections (Round Two) I II III Predictors B t Sig. B t Sig. B t Sig.

Ethnic Threat Minorities 0.335 11.725 *** 0.311 11.039 *** 0.424 4.525 *** Refugees 1995 0.011 0.211 NS 0.101 1.819 f 0.086 1.526 NS Refugees 2000 -0.052 -0.390 NS -0.049 -0.386 NS 0.010 0.076 NS

Economic Vulnerability Unemployed 0.108 1.112 NS 0.092 0.938 NS Wages -0.706 -3.741 *** -0.725 -3.838 ***

Non-Linear Effects Minorities Quadratic -0.001 -1.265 NS

Constant 23.604 26.746 *** 29.302 9.421 *** 29.073 9.350 *** Adjusted R2 0.496 0.543 0.545 N of municipalities 158 158 158 Significance: NS= not significant, tp<10, *p<.05, **p<.01, *** p<.001.

Since none of the Presidential candidates captured an absolute majority, the two top candidates went into a second round of voting on June 27, 2005. As the opponent of the Radicals was from the moderate left Democratic Party, one would assume that the

Radical candidate might capture some of the moderate right vote. Indeed, the SRS improved its overall vote by 15 per cent. As Model I shows, while the overall model fit for ethnic threat variables remains the same, the importance of different predictors shifts dramatically. While per cent minority is still highly significant, refugee variables no longer have any effect. Model II shows that economic vulnerability variables suppress the

1995 refugee variable. As in the first round, the quadratic effect in Model III is no longer statistically significant.

Discussion

This study finds that certain theories used to explain Western European far right and

North American inter-racial relations are of considerable use in explaining the rise of the 119 far right in Serbia. Both ethnic threat and economic vulnerability hypotheses are

supported, albeit with some qualifications.

While Blalock's expectation with respect to threshold effects is mostly supported, the effect of the presence of the refugees is not as consistent as expected. Surprisingly, in most of the elections, municipalities with a concentration of earlier refugees from Bosnia and Croatia appear more likely to vote for the SRS than municipalities with a strong presence of more recent refugees from Kosova. More targeted research on the electoral preference and contextual effects of the various groups of refugees might help to explain this.

While the findings for four out of the six elections analysed support Blalock's

expectation that the relationship between the percentage of a minority group and the level

of majority intolerance should have a threshold effect, one wonders why the thresholds

are roughly 15 and 40 per cent. The first threshold is very similar to the 10-15 per cent

non-white threshold for white economic discrimination, identified by Blalock (1957, p.

678) for US metropolitan areas in the 1950s. Blalock explains the first threshold as the

effect of a minority's visibility - the point at which the prejudiced members of the

majority initially 'realize' there might be 'a problem.' One explanation for the second

threshold is that the majority's anxiety about the possibility of minority ethnic secession

peaks as a minority approaches the 50 per cent mark locally. If the minority becomes the

local majority in a municipality that lies on an inter-state border, a possible demand for

the 'correction' of inter-state borders (the incorporation of the municipality into the

potentially hostile neighbouring state) would become an easily imaginable threat to local

Serbs. 120

Kitschelt's and Wimmer's hypothesis of 'welfare chauvinism' is supported by the findings on the success of the SRS after 2000 in municipalities with high unemployment and/or low average employment incomes. My findings and the results of previous studies

(Slavujevic 2002) suggest that the SRS has expanded into the political space left by the collapse of Milosevic's party and the re-alignment of its working-class base. In line with

Kitschelt's expectations for post-communist Europe, the promise of social security for the majority seems to be a winning formula, even in relatively ethnically-homogenous, but economically-vulnerable municipalities. The prevalence of welfare chauvinism in Eastern

Europe is further supported by earlier studies of East Germany, Poland, Hungary, and the

Czech Republic.56

The phenomenon of working-class support for the far right is not limited to post- communist Eastern Europe. However, Serbian welfare chauvinism differs from the

Western European variety. While the targets of exclusionary policies in Western Europe are typically recent (especially non-White and non-Christian) immigrants, in Serbia the primary targets are ethnic groups who have lived in Serbia and in the Balkans for centuries. Whereas Western European welfare chauvinists attempt to legitimise exclusion by arguing that only the 'long established' population is entitled to social services since they created these services themselves through heavy taxes (Wimmer 2004, p. 12),

Serbian welfare chauvinists argue that presumably disloyal minorities do not deserve to benefit from the Serbian welfare state.

According to a 1998 survey, 39 per cent of East Germans (and only 23 per cent of West Germans) agree that 'foreigners' access to welfare services should be restricted (Minkenberg 2002). A comparative study of electoral politics in Poland, Hungary, and Czech Republic finds that anti-market parties did best in regions with high unemployment, low educational attainment, and poor living conditions (Tworzecki 2003). 121

The important implication for Serbia is that a neo-liberal economic transition, characterised by dramatic increases in unemployment and the further undermining of the social safety net, might create an opportunity for the electoral breakthrough of the

Radicals. Indeed, in the fall of 2005, IMF negotiators insisted that the Serbian government accelerate structural reforms, in particular by keeping average wage increases below inflation, privatising all remaining socially-owned enterprises, initiating further bankruptcies, and implementing large and permanent cuts in health insurance benefits (IMF 2006). According to the IMF negotiators, 'the implementation of these reforms requires more political determination than seen so far. Agreed measures should be executed vigorously, despite resistance from vested interests' (IMF 2006, p. 29).

Apparently, the IMF has little interest in the predictable political consequences of such harsh restructuring policies. Unfortunately, this is a recurring problem. Susan

Woodward has persuasively argued that earlier IMF-induced economic reforms led, albeit unintentionally, to economic decline and the disintegration of the Yugoslav social security systems, which paved the way for the ultra-nationalist movements of the late

1980s and early 1990s (Woodward 1996).

There is no need to repeat these mistakes, as there are better options for reform.

Swank and Betz (2003) have found that Western European countries with strong welfare states absorb high levels of trade openness, capital mobility, and immigration rates without experiencing a major rise in the vote for the far right. They hold that a strong welfare state reduces economic risks and insecurities for the ethnic majority, thus decreasing the likelihood of support for the far right. Similarly, Mansfield and Snyder

(2005, p. 276) conclude their comparative analysis of the causes of warlike behaviour 122 among new democracies by arguing that the democratising regimes need to develop

social welfare policies to buffer the costs of transition. Just as the development of the welfare state in some Western European countries arguably helped to undermine popular

support for Leninist parties, so too the maintenance and development of the welfare state

in Eastern European countries might play a decisive role in the reduction of the electoral

appeal of authoritarian ultra-nationalists. 123

Appendix 4A Ethnicity and Support for the Radicals

Respondent's Self-Declared Ethnicity Opinion on the Serbs Montenegrin Yugoslav Hung;aria n Muslim Albanian SRS Leader (Bosniac) Very Positive 9.4 4.11 3.95 0 0 0 Positive 26.69 14.24 19.74 3.45 1.39 3.13 Negative 22.47 20.89 15.79 19.54 8.33 21.88 Very Negative 26.5 39.65 47.37 40.23 83.33 50 DK/Other 13.78 21.2 13.6 36.78 6.94 25 Number of Respondents 1,255 316 76 87 144 32 Source: CPIJM Database, Survey of Yugoslavian (Serbia & Montenegro) Public Opinion, Fall 1996 124

Appendix 4B. Variable Descriptions, Data Sources, and Hypothesized Effects

Variable Description Year of Data Impact on the Collection Majority Vote for the SRS DEPENDENT VARIABLE Majority vote Percentage of the Registered Dec/2000 for the Radicals Majority Voters in the given Sep/2002 municipality who voted for the SRS Nov/2003 on the specified elections. a Dec/2003 Jun13/2004 Jun27/2004 INDEPENDENT VARIABLES Ethnic Threat Minorities Per cent of Minority Population in Positive, with the municipality b 2002 decreasing slope

Refugees Refugees as a per cent of majority 1995 Positive population.c 2000 Economic Vulnerability Unemployment Per cent of Working Age Population Dec/2000 that is registered as unemployed. d Dec/2002 Positive July/2003 Jun/2004

Physicians Number of Inhabitants [in 100s] per 2000 Positive one physiciane 2001

Wages Average monthly wage (after tax) 2000 per employee, annual average, in 2001 1000 Dinars6 2002 Negative 2003 2004 Phones N of phone subscribers per 100 Negative inhabitants, as a per cent of Serbian 2001 average (Serbia =100) e

Municipal Local Municipal Revenue per 2001 Negative Revenue Inhabitant, in Dinars e

Note: Due to lack of 1991 and 2002 census data, Kosova results are excluded from the analysis. Sources: a Estimates calculated from the CESID Elections data set and 2002 Serbian Census results; b Linear extrapolation from the 1991 and 2002 Censuses; 1991 and 2002 Censuses ;c CESID Refugees Data and UNHCR Registration Papers; Serbian Monthly Labor Market Reports;e Municipalities in Serbia data set, Serbian Statistics Bureau web page 125

Chapter Five: Conclusion

My dissertation offers new, empirically tested explanations of the role of culture, institutions, and class politics in the recent Yugoslav collapse. Yugoslavia was neither undermined by "ancient hatreds" nor torn apart through the manipulations of ignorant masses by Machiavellian elites. Instead, the elites exploited a tradition of intolerance

(especially negative visions of the Other) formed in the pre-communist period. The communist ethno-federalist institutions in Yugoslavia (and other communist federations) unintentionally undermined political unity and strengthened the nationalism they were supposed to contain. And the rise of the far right did not happen because "ethnicity trumps class." Rather, class-specific welfare chauvinist appeal enabled ultra-nationalists to mobilize economically vulnerable sections of the majority population.

Understanding the Weight of History

The tradition of all dead generations weighs like a nightmare on the brains of the living. (Marx 1852)

To investigate whether a path dependency exists in the formation of the exclusionary ideology and policy and how it might operate if it does, I conducted a theoretically informed historical case study.57 Methodologically, my focus was on

The concepts of "tradition of intolerance" that I use has some similarity with the "repertoire of contention" concept used in the social movement theory (SMT). "Repertoires of contention" are ways that people act together in pursuit of shared interests, such as strikes, peaceful demonstrations, roadside bomb attacks, etc. The repertoires are durable cultural innovations that emerge from successful struggle (Tilly 1995: 15, 26-8). However, while the SMT looks at the repertoire of means seen as effective and justifiable by non-state actors, I analyze the repertoire of state exclusionary policies (land expropriations, colonization by settlers, ethnic cleansing, etc.) seen as effective and justifiable by the state and the actors closely linked to the state. In addition, 126

identifying formative periods, turning points, and causal mechanisms which maintain the process over time. In terms of evidence, my work combines the use of existing historical

studies with my own primary research into such sources as diaries and biographies of

leading participants in the historical events, as well as diaries of Western travelers in the region.

Admittedly, historical sociology has some methodological limitations compared to the sociological study of more recent events and contemporary actors. For example, there is no possibility of using survey methods or conducting interviews. Key documents

are decayed or intentionally destroyed, inter-generational oral history is of questionable

reliability, and so on. However, historical sociology provides some less-appreciated

methodological advantages, such as access to previously secret and now declassified

official documents (Bryant 1994: 7). I have made extended use of this advantage by

analyzing the previously secret diplomatic correspondence of the Serbian state.58

While the existing ideology-centered explanation, based on Tilly (1975) and

Hroch (1985), that focuses on the importance of the coercive Utopia of nation state,

captures much of the historical dynamic, it has major flaws. It downplays the crucial

impact of Great Power politics on the ability of the Balkan nationalist leaders to pursue

coercive homogenization programs. In fact, the major European Powers exercised crucial

influence, initially by treating some of the atrocities against Christians as "humanitarian

the tradition of intolerance not only encompasses the means used, but includes the ideological representations of the Other and justifications for the exclusionary polices. 58 The extensive use of primary historical sources helps to deal with a frequently mentioned methodological limitation of historical sociology. As Burawoy and Goldthorpe argue, a problem with historical sociology using secondary sources is that the sociologist ends up interpreting the interpretations of a historian. With such distance from primary sources, the possibilities of partial or complete misinterpretation are increased. The method is even more problematic when the sociologist uses few secondary sources, thereby taking for granted interpretations of one historian that other specialist historians might dispute (Bryant 1994: 12-13; Goldthorpe 1994: 67-70). 127 disasters" and simply ignoring the atrocities towards Muslims, and later by frequently overlooking violations of minority rights by the local client states, including the Kingdom of Yugoslavia.

The coercive utopia thesis also fails to explain why some minorities were singled out for assimilation while others received more exclusionary treatment. Seemingly, no amount of assimilation could turn Albanians into Serbs. In the Serbian case, those minorities whose "essential" character was perceived as culturally distant (in terms of religion, language, and socio-economic development) and physically threatening (in terms of the previous history of conflict and the current presence of major foreign supporters) were more likely to be perceived as logical candidates for deportation or murder. Thus, the nationalist elite's vision of the "essential" characteristics of the Other determined which policies towards them would appear realistic and feasible.

The findings support the argument that hegemonic discourse and practices and exclusion have self-reinforcing characteristics. For example, the perception of Albanians as unassimilable and barbarous resulted from the persecution of Kosovo Serbs by

Albanian extremists in the 1878-1912 period. The recurring episodes of Kosovo Albanian massive support for the imperial invaders of Serbia - first the Ottoman and later the

Austrian Empire - further strengthened the belief that they were a disloyal minority and a security threat. The new contributors to the tradition of intolerance were clearly building on the genealogy of ideas and actions created by their predecessors. The outcome of this process has been a solidified vision of Albanians as a "proven" disloyal minority and a security threat to the Serbian state. 128

While the legacy of dead generations does indeed "weigh like a nightmare on the brains of the living," this legacy also contains subversive thoughts, with the potential to undermine exclusionary practices and ideas. Notably, my historical analysis demonstrates the existence of an alternative vision (to which Serbian social-democrats made a major contribution) of Albanians not as "inferior" or "savages," but as respected neighbors and fellow Balkan peoples.

The Impact of Ethno-Federalism: How the Yugoslav Communists

Dug Their Own Graves

Based on earlier ideas of social-democrats and on the Soviet multinational model, the

Yugoslav communist solution to the national question was multi-national federalism.

However, while some claim that ethnic federalism contains nationalist excesses, a number of historical institutionalists argue that communist ethno-federalism was a

"subversive" institution which fatally undermined the communist states (Brubaker 1994;

Crawford 1998: 198-9, 217; Bunce 1999; Guzina 2000: 21-22; Hechter 2000a). It is true that the communist federations (USSR, Yugoslavia, Czechoslovakia) have disintegrated, while none of the unitary communist states have broken apart, but this observation tells us neither whether the association is spurious nor the exact causal mechanism by which federalism leads to state disintegration. The question, then, is whether the specific institutional design of communist institutions had slow but cumulative unintended effects which ultimately led to the disintegration of some of these states in the early 1990s. As shown earlier, Hechter has major difficulties portraying Yugoslavia as a highly 129 decentralized country lacking national conflicts. Thus, Yugoslav federation is particularly interesting as an "anomalous case."

The historical-sociological method is perfectly suited for this puzzle, as its lengthy time horizon enables the study of big, cumulative, and slow-moving processes, with considerable temporal distance between the cause (or formative period) and effect

(Pierson 2003: 179,181). Therefore, Chapter Three follows the method of analytic historical sociology, explaining recurring causal patterns by comparing the testable implications of hypotheses with the historical evidence reported by area specialists

(Skocpol 1984: 375, 378-82).

Instead of blanket endorsement or rejection of the ethno-federal solution,

I identify specific conditions under which it does not work. Because several countries are either implementing or considering the implementation of ethno-federal institutions, understanding the effects of these institutions on Eastern Europe should be helpful to these policy discussions. Certainly, a number of federal countries (such as Canada,

Switzerland, and India) have been relatively successful in containing excessive nationalism, so the East European experience helps to identify pathological forms of federalism, rather than announcing the "inevitable" failure of the federalist project. For example, Michael Hechter qualifies his general endorsement of ethno-federalism with a warning, based on the failure of Yugoslav federalism, that excessive decentralization has disintegrative effects. Using Hechter's insight as a point of departure, I identify specific institutions and processes that had disintegrative effects in the Yugoslav case (1945-

1991). Due to the geographic dispersion of Yugoslav ethnic groups, the assumption that inter-ethnic equality demands that each nation have titular status in its own ethnic 130 republic unintentionally led to institutionalized racism of titulars against minorities.

Further, while ethnic equalization policies bought the loyalty of less-developed regions, they triggered the secessionism of the wealthy. Finally, the dogmatic pursuit of Leninist teaching led Yugoslav communists to respond to nationalist mobilizations with more decentralization, effectively creating a vicious circle of decentralizations and nationalist mobilizations.

Why Does the Public Support Ultra-Nationalists? The Welfare Chauvinism Thesis

During the 1990s, the Serbian Radical Party (SRS) was the most successful far right party in Europe (see Table 5.1). Kitschelt's and Wimmer's "welfare chauvinism" hypothesis is supported by the findings on the success of the SRS after 2000 in municipalities with high unemployment and/or low average employment incomes. My findings and the results of previous studies (e.g., Slavujevic 2002) suggest that the SRS has expanded into the political space left by the collapse of Milosevic's party and the re-alignment of its working-class base. In line with Kitschelt's expectations for post-communist Europe, the promise of social security for the majority seems to be a winning formula, even in relatively ethnically homogenous but economically vulnerable municipalities. The prevalence of welfare chauvinism in Eastern Europe is further supported by earlier studies of East Germany, Poland, Hungary, and the Czech Republic.59

According to a 1998 survey, 39 percent of East Germans (and only 23 percent of West Germans) agree that "foreigners'" access to welfare services should be restricted (Minkenberg 2002). A comparative study of electoral politics in Poland, Hungary, and Czech Republic finds that anti-market parties do best in regions with high unemployment, low educational attainment, and poor living conditions (Tworzecki 2003). 131

Table 5.1. Peak Political Performance of Major European Far Right Parties, 1989-2002 Country Main Far Right Party Peak Electoral Support Participation in (Year)a National Government? (Year) Austria Austrian Freedom Party 27% (1999) Yes (2000) _ Belgium Vlaams Blok ^ 20 (2001)b No" ZI Croatia Croatian Party of Rights 7(1992) No Czech Rep. Republican Party 9 (1996) No Netherlands List Pirn Fortuyn J_7 (2002) Yes (2002) ~ France National Front ^ 17(2002) No Italyc Northern League 10(1996) Yes (1994, 2001) Norway Progress Party 13 (1989) JSFo Romania Party of Greater Romania 33 (2000) No Russia Liberal Democratic Party 23 (1993) Jv[o Serbia Serbian Radical Party 49(1997) Yes (1993 and 1998) Slovakia Slovak National Party 14 (1992) Yes (1992and 1994)

Sources: Eatwell (2004: 1; 2003: 47); Giordano (2003: 218); Goati (2001); Grdesic (1999); Ignazi (1997: 54); Merkl (2003: 24-5); Minkenberg (2002: 351); Mudde (2000: 13); Penc and Urban (1998: 40); Schutze and Bigelow (1996: 3, 29); Szayna (1997); Cibulka (1999: 117); Veugelers and Magnan (2005: 839). a Refers to country-level elections only; b The Vlaams Blok results are for Flanders only; c The results for Italy do not include the National Alliance, as they have moved towards the centre since 1995 (Ignazi 2005: 336-7).

Working-class support for the far right is not limited to post-communist Eastern

Europe, but Serbian welfare chauvinism differs from the Western European variety.

While the targets of exclusionary policies in Western Europe are typically recent

(especially non-White and non-Christian) immigrants, in Serbia the primary targets are

ethnic groups who have lived in Serbia and in the Balkans for centuries. Whereas

Western European welfare chauvinists attempt to legitimise exclusion by arguing that

only the "long established" population is entitled to social services, since they created

these services themselves through heavy taxes (Wimmer 2004: 12), Serbian welfare

chauvinists argue that presumably disloyal minorities do not deserve to benefit from the

Serbian welfare state. 132

Conclusion: Better Answers, New Limitations, Comparative Implications

Regardless of its contributions, the dissertation has some important limitations. Most notably, while Chapter Two provides a detailed analysis of the formative (pre- communist) period of exclusionary ideas and practices, it covers only the period up to

WW II. Thus it contains a very limited analysis of the "mechanisms of reproduction" of that tradition in the communist period, when blatant forms of ethnic intolerance were

severely prohibited. We cannot simply see the processes and institutions as "frozen" after the formative periods (Thelen 2003: 211). Since the factors responsible for the genesis of

an institution may not be the same as those that sustain it over time (Stinchcombe 1968), we need to understand mechanisms of institutional reproduction over time (Thelen 2003).

Therefore, it is necessary is to develop an historical analysis that will connect the

formative period of the tradition of intolerance with the new escalation of the Albanian-

Serbian ethnic conflict in the early 1980s.

Furthermore, theoretical implications drawn on the basis of a single case study

(even combined with some regional contrasts) have to be taken with a grain of salt. The

next phase is to test the causal factors and the mechanisms identified (especially the

importance of nation-state ideology, clientelistic relations with great powers, and the

crystallization of negative images of the Other) in other comparable cases (such as the

formation of other homogenized nation-states in the Balkans). A brief review of

secondary evidence from the region (as discussed in Chapter Two) supports the

plausibility of my argument. However, systematic testing with a greater range of

historical evidence is required to verify the scope of its explanatory power. 133

As the chapter on federalism (Chapter Three) uses more comparative evidence, the findings can claim more generalizability, but not beyond the post-communist region.

Comparative historical sociology is a fruitful way to establish causal mechanisms by testing hypotheses on historically sensitive evidence (Rueschemeyer and Stephens 1997:

63, 66), but it has limited use in establishing the generalizability of the mechanisms discovered (Ragin 2000: 90). Large-n quantitative studies perform better in this aspect.

Ideally, an historical investigation of causal mechanisms in a small number of theoretically justified cases should be followed by quantitative testing and verification of the mechanisms discovered on a large number of cases.60 The first step would be a comparative historical analysis of other federal systems with similarly high levels of decentralization who have managed, at least for the time being, to avoid the path to disintegration. A logical candidate would be Switzerland, as this ethnic federation had levels of fiscal decentralization similar to Yugoslavia in the 1980s (Hechter 2000a: 148), but managed to contain ethnic conflicts in the 1990s. Overall, the only way to determine whether the negative effects of ethno-federalism occur outside the post-communist region is to test these hypotheses on cases from other regions.

Finally, while the Serbian case discussed in Chapter Four gives strong support to the welfare chauvinism hypothesis, it presents some contradictory findings. For example, the effect of the presence of refugees is not as consistent as expected. Surprisingly, in most elections, Serbia's municipalities with a concentration of earlier refugees from

Bosnia and Croatia appear more likely to vote for the far right than municipalities with a strong presence of more recent refugees from Kosova. More targeted research on the

60 Walker and Cohen argue that once the explanatory value of a theory has been demonstrated on a small set of cases, it is justifiable to try to extent the scope of the theory by generalizing and testing it on a larger number of cases (Walker and Cohen 1985: 300). 134 electoral preference and contextual effects of various groups of refugees might help to explain this finding. Nor does the chapter directly test the spatial effects (i.e. the position of the municipality on the "threatened border") on the far right vote, although the underlying theory of ethnic threat implies that such effects should be observable. Multi level analysis incorporating data on the geo-political position of the municipality should be used to test for these effects.

Could other European Far Right parties electorally profit from the use of the welfare chauvinism appeal? The experience of multi-ethnic settler societies implies that a shift from exclusionary ethnic identity to a more inclusive civic national identity, combined with a full political incorporation for minorities, might block the Far Right's path to power (Myles and St-Arnaud 2007: 354). I also expect that a strong social- democratic political party and civil society - committed to ethnic equality, economic egalitarianism, and social rights - can immunize vulnerable sections of the ethnic majority population from the welfare chauvinist appeal. My future post-doctoral research aims to identify valid and generalizable causal explanations of the variability in support for the Far Right in Eastern Europey

Despite its limitations, the dissertation has implications beyond the case study. By refining and extending key existing theories, it points to a way of analyzing the impact of culture on ethnic conflicts without the selective reading of historical evidence that plagues ancient hatreds and elite manipulation perspectives. My analysis of the unintended and cumulative effects of ethno-federalism identifies specific ethno-federal institutions and policies that should be avoided if possible. As the findings support the 135 welfare chauvinism hypothesis, pronouncements on the death of class politics, particularly in the post-communist region, are simply wrong. 61

61 These findings also support Geoffrey Evans's conclusions on the similarity of key political cleavages in Western and Eastern Europe. Evans argues that while ethnic cleavages are important in a number of (ethnically diverse) Easter European countries, the class structure also matters a lot for voting (Evans 2006, p. 250). Similarly, Szelezy et all (1997, p. 216) found that a dramatic growth of the support for social-democratic parties among the blue collar workers after the first post-communist elections in Poland and Hungary. 136

References

ALTER, PETER 1994 Nationalism, 2nd edition, London: Edward Arnold ANTONIC, SLOBODAN 2002 Zarobljena Zemlja, Belgrade, Serbia: Otkrovenje AUSTIN, ROBERT 2002 'Greater Albania: The Albanian State and the Question of Kosovo,' in John Lampe and Mark Mazower, (eds) Ideology and National Identity in The 20th Century Southeastern Europe (Manuscript), pp. 235-253 BERGESEN, ALBERT and HERMAN, MAX 1998 'Immigration, race, and riot,' American Sociological Review 63: 39-54 BAKIC, JOVO 2004a 'Zwischen Rechtsliberalismus und -extremismus.' Ost-West Gegeninformationen 16 (3): 9-15 — 2004b Ideologije jugoslovenstva izmedu srpskog i hrvatskog nacionalizma, Zrenjanin: Gradska Narodna Biblioteka BANAC, IVO 1984 The National Question in Yugoslavia, Ithaca: Cornell University Press BATAKOVIC, DUSAN 1991 Kosovo i Metohija u srpsko-arbanaskim odnosima, Pristina, Serbia: SANU BEGOVIC, BORIS 2005 'Cetiri godine tranzicije u Srbiji,' in Boris Begovic and Bosko Mijatovic (eds.), Cetiri Godine Tranzicije u Srbiji, Belgrade: CLDS, pp. 439-468 BEISSINGER, MARK 2002 Nationalist Mobilization and the Collapse of the Soviet State, Cambridge: Cambridge University Press — 1998 "Nationalism that Bark and Nationalisms that Bite: Ernest Gellner and the Substantiation of Nations" in John A. Hall (ed.) The State of the Nation, Cambridge: Cambridge University Press, pp. 169-190 BETZ, HANS-GEORG 1994 Radical Right-Wing Populism in Western Europe, New York: St. Martin's BILANDZIC, DUSAN 1986 Jugoslavia Poslije Tita 1980-1985 Zagreb: Globus BLALOCK, HUBERT 1967 Towards a Theory of Minority-Group Relations, New York: Capricorn — 1957 'Per cent non-White and discrimination in the South,' American Sociological Review 22 (6): 677-682. BOGDANOVIC, DIMITRIJE 1985 Knjiga o Kosovu, Belgrade: SANU BRUBAKER, ROGERS 1998 'Myths and misconceptions in the study of nationalism,' in John A. Hall (ed.) The State of the Nation, Cambridge, UK: Cambridge University Press, pp. 272-305 — 1996 Nationalism Reframed, Cambridge: Cambridge University Press — 1994 "Nationhood and the National Question in the Soviet Union and Post-Soviet Eurasia: An Institutionalist Account." Theory and Society 23 (l):47-78. BRYANT, JOSEPH 1994 "Evidence and explanation in history and sociology" British Journal of Sociology, vol 45, 1, March: 3-19 BRYM, ROBERT 1994 The Jews of Moscow, Kiev and Minsk, London: MacMillan BULATOVIC, MARKO 2002 'Struggling with Yugoslavism: Dilemmas of Interwar Political Thought', in John Lampe and Mark Mazower, (eds) Ideology and National Identity in The 20th Century Southeastern Europe (Manuscript), pp. 254-276 137

BUNCE, VALERIE 1999 Subversive Institutions: the Design and the Destruction of Socialism and the State, Cambridge: Cambridge University Press BURG STEVEN L. and BERBAUM, MICHAEL, L. 1989 "Community, Integration, and Stability in Multinational Yugoslavia" American Political Science Review, June, 83,2: 535-554 BURG, STEVEN L. 1988 Political Structures in Dennison Rusinow (ed) Yugoslavia: A Fractured Federalism Washington: Wilson Center Press — 1983 Conflict and Cohesion in a Socialist Yugoslavia Princeton: Princeton University Press CARMICHAEL, CATHIE 2002 Ethnic Cleansing in the Balkans: Nationalism and the Destruction of Tradition, London: Routledge CENTAR ZA POLITIKOLOSKA ISTRAZIVANJA IJAVNO MNENJE 1991-2002 CPIJM Dataset CENTAR ZA SLOBODNEIZBOREIDEMOKRATIJU(CeSID) 2005 Online 'Rezultati Izbora', URL: www.cesid.org/rezultati/index.jsp CIBULKA, FRANK 1999 'The radical right in Slovakia', in Sabrina P. Ramet (ed.) The Radical Right in Central and Eastern Europe Since 1989, University Park, PA : Pennsylvania State University Press, pp. 109-132 CLARK, TERRY and LIPSET, SEYMOUR 2001 [1991] "Are social classes dying?" in Terry Clark and Seymour Lipset (eds) The Breakdown of Class Politics Washington, D.C.: Woodrow Wilson Center Press, pp. 39-54 CLARK, TERRY et. al. 2001 [1993] "The declining political significance of social class" in Terry Clark and Seymour Lipset (eds) The Breakdown of Class Politics Washington, D.C.: Woodrow Wilson Center Press, pp. 77-104 COHEN, LENARD J. 2001 Serpent in the Bosom: The Rise and Fall of Slobodan Milosevic, Boulder, Colorado: Westview — 1995 Broken Bonds, Boulder, Colorado: Westview — 1989 The Socialist Pyramid: Elites and Power in Yugoslavia, Oakville, Ontario: Mosaic Press CONNOR, WALKER 1984 The National Question in Marxist-Leninist Theory and Strategy Princeton: Princeton University Press CONVERSI, DANIELE 2000 "Central Secession: Toward a New Analytical Concept? The Case of Former Yugoslavia" Journal of Ethnic and Migration Studies, Vol. 26, No.2: 333-355 CRAWFORD, BEVERLY 1998 "The Causes of Cultural Conflict: An Institutionalist Approach," in Beverly Crawford and Ronnie D. Lipschutz (eds.) The Myth of "Ethnic Conflict": Politics, Economics, and "Cultural" Violence Berkeley: International and Area Studies, University of California at Berkeley, pp. 3-43 CUBRILOVIC, VASO 1997a [1944] 'The Minority Problem in the New Yugoslavia', in Kosovo: In the Heart of the Powder Keg, R. Elsie, New York: Columbia University Press, pp. 449-464 — 1997b [1937] 'Iseljavanje Arnauta' [The Expulsion of the Albanians] in Kosovo: In the Heart of the Powder Keg, R. Elsie, ed, New York: Columbia University Press, pp. 400-424 DAVIES, JAMES C. 1969 "The J-Curve of Rising and Declining Satisfactions as a Cause of Revolution and Rebellion" in Violence in America: Historical and 138

Comparative Perspectives, edited by Hugh Graham and Ted Gurr, Beverly Hills, CA: Praeger, pp. 547-576 — 1962 "Toward a Theory of Revolution" American Sociological Review 27: 5-19 DIMIC, LJUBODRAG 1996 Kulturna politika Kraljevine Jugoslavije, 1918-1941, Vol 1, Belgrade: Stubovi Kulture DJURIC, VOJISLAV 1977 Antologija narodnih junackih pesama, Belgrade: SKZ DONIA, ROBERT and FINE, JOHN 1994 Bosnia and Hercegovina : a Tradition Betrayed, New York : Columbia University Press DRAGOVIC-SOSO, JASNA 2002 'Saviours of the Nation', Montreal: McGill-Queen's University Press DULMER, HERMANN and KLEIN, MARKUS 2005 'Extreme right-wing voting in Germany in a multilevel perspective,' European Journal of Political Research 44: 243-263 DURHAM, EDITH 2001 [1913] Letter to the Editor of The Near East,' in Albania and the Albanians: Selected Articles and Letters of Edith Durham, Bejtullah Destani, ed., London: Centre for Albanian studies — 1985 [1909] High Albania London: Virago DVORNIKOVIC, VLADIMIR 1990 [1939] Karakterologija Jugoslavena Belgrade: Prosveta EATWELL, ROGER 2004 'The new extreme right challenge,' in Roger Eatwell and Cass Mudde (eds.), Western Democracies and the New Extreme Right Challenge, London, New York: Routledge, pp. 1-16 —2000a 'The extreme right and British exceptionalism,' in Paul Hainsworth (ed.), The Politics of Extreme Right, London: Pinter, pp. 172-192 —2000b 'The Rebirth of the 'extreme right' in Western Europe,' Parliamentary Affairs 53: 407-425 ELSIE, ROBERT 1997 Kosovo: In the Heart of the Powder Keg, (ed) New York: Columbia University Press EMIZET, KISANGANI and HESLI, VICKI1995 "The Disposition to Secede: An Analysis of the Soviet Case," Comparative Political Studies, vol 27, no 4, Jan: 493-536 ERDELJANOVIC, JO VAN 1978 Star a Crna Gora, Belgrade: Slovo ljubve EVANS, GEOFFREY 2006 "The social bases of political divisions in post-communist Eastern Europe," Annual Review of Sociology, vol.32: 245-270 FOSSETT, MARK and KIECOLT, JILL 1989 'The relative size of minority populations and white racial attitudes,' Social Science Quarterly 70: 820-35 FRIEDMAN, THOMAS L. 1993 'Bosnia Reconsidered' New York Times, April 8, p. Al. GALLAGHER, TOM 2001a 'A Balkan History Learning Curve', European History Quarterly,Vo\.3\:\4\-\55 — 2001b Outcast Europe: the Balkans, 1789-1989, from the Ottomans to Milosevic, London: Routledge GAGNON, V. 2004. The Myth of Ethnic War: Serbia and Croatia in the 1990s. London: Cornell University Press. GARASANIN, ILIJA 1998 [1844] Nacertanije, Belgrade GIBSON, RACHEL, McALLISTER, IAN, and SVENSON, TAMI 2002 'The politics of race and immigration in Australia,' Ethnic and Racial Studies 25 (5): 823-844 139

GIORDANO, BENITO 2003 "The Continuing Transformation of Italian Politics and the Contradictory Fortunes of the Lega Nord" Journal of Modern Italian Studies, vol. 8, no. 2, pp. 216-230 GLIGOROV, VLADIMIR 1994 Why do Countries Break-Up? Uppsala, Sweden: Uppsala University GO ATI, VLADIMIR 2001 Izbori u SRJ, Belgrade, Serbia: Centar za Slobodne Izbore i Demokratiju GOLDER, MATT 2003 'Explaining variation in the success of extreme right parties in Western Europe,' Comparative Political Studies 36 (4): 432-466 GOLDSTEIN, IVO 1999 Croatia: A History, London: Hurst GOLDSTEIN, SLAVKO (ed) 1986 The Times Atlas Svjetske Povijesti Ljubljana: Cakarijeva Zalozba GOLDTHORPE, JOHN 1994 "The uses of history in sociology: a reply" British Journal of Sociology, vol 45,1, March: 211-230 GOLUBOVIC, ZAGORKA 1995 'Stavovi prema nacionalnim problemima,' in Zagorka Golubovic et al. (eds.), Drustveni Karakter i Drustvene Promene u Svetlu Nacionalnih Sukoba, Belgrade: Institute for Philosophy and Social Theory, pp. 225-246 GORENBURG, DMITRY 2003 Minority Ethnic Mobilization in Russia, Cambridge: Cambridge University Press GRDESIC, IVAN 1999 'The radical right in Croatia and its constituency', in Sabrina P. Ramet The Radical Right in Central and Eastern Europe Since 1989, University Park, PA : Pennsylvania State University Press,, pp. 171-192 GURR, TED and MOORE, WILL. 1997 "Ethnopolitical Rebellion: A Cross-Sectional Analysis of the 1980s with Risk Assessments for the 1990s" American Journal of Political Science 41:1079-1103 GUZINA, DEJAN 2000 "The Self-Destruction of Yugoslavia" Canadian Review of Studies in Nationalism, vol. 27: 21-32 HALE, HENRY 2000 " The Parade of Sovereignties: Testing Theories of Secession in the Soviet Setting" British Journal of Political Science, 30: 31-56 HAWRYLYSHYN, OLI 1977 "Ethnic Affinity and Migration Flows in Postwar Yugoslavia" Economic Development and Cultural Change, Vol. 26, No. 1, Oct.: 93-116 HAYDEN, ROBERT M. 1999 Blueprints for a House Divided: The Constitutional Logic of the Yugoslav Conflicts Ann Arbor, Mich.:University of Michigan Press HECHTER, MICHAEL and OKAMOTO, DIN A 2001 "Political Consequences of Minority Group Formation" Annual Review of Political Science, 2001, 4: 189-215 HECHTER, MICHAEL 2000a Containing Nationalism. New York: Oxford University Press. 2000b "Nationalism and Rationality" Journal of World Systems Research, VI, 2, Summer/Fall, 308-329 HODZA, HAJREDIN 1984 Afirmacija albanske nacionalnosti u Jugoslaviji, Pristina, Yugoslavia: Rilindija HOROWITZ, DONALD 1998 " Self-Determination: Politics, Philosophy, and Law" in Margaret Moore (ed) National Self-Determination and Secession New York: Oxford University Press, pp. 181-214 140

HORVAT, BRANKO 1988 Kosovsko Pitanje, Zagreb: Globus — 1985 Jugoslavensko Drustvo u Krizi, Zagreb: Globus HROCH, MIROSLAV 1985 Social Preconditions of National Revival in Europe Cambridge: Cambridge University Press HUMAN RIGHTS WATCH [HRW] 2004 Failure to Protect: Anti-Minority Violence in Kosovo, March 2004, URL: http://hrw.org/reports/2004/kosovo0704 — 2001 Under Orders: War Crimes in Kosovo, New York: Human Rights Watch HUNTIGNTON, SAMUEL P. 1996 Clash of Civilizations and the Remaking of the World Order New York: Simon & Schuster IGNAZI, PIERO 2005 ' Legitimation and evolution on the Italian right wing: social and ideological repositioning of Alleanza Nazionale and the Lega Nord', South European Society and Politics, vol. 10, no. 2, July, pp. 333-349 —1997 'The extreme right in Europe: a survey', in Peter Merkl and Leonard Weinberg (eds) The Revival of Right-Wing Extremism in the Nineties, London: Frank Cass, pp. 47-64 INTERNATIONAL COMMISSION TO INQUIRY INTO THE CAUSES AND CONDUCT OF THE BALKAN WARS 1993 [1913] 'Report' in George F. Kennan (ed), The Other Balkan Wars: 1913 Carnegie Report, Washington DC: Carnegie Endowment for International Peace INTERNATIONAL MONETARY FUND 2006 'The IMF country report no. 06/58,' February, IMF, URL: www.imf.org/external/country/SCG/index.htm THE INTERNATIONAL CRIMINAL TRIBUNAL FOR THE FORMER YUGOSLAVIA [ICTYJ 2003 The Indictment against Vojislav Seselj, January 15, The Hague, URL: www.un.org/icty/indictment/english/ses-ii030115e.htm IVARSFLATEN, ELIZABETH 2005 "The vulnerable populist Right parties: no economic realignment fueling their electoral success," European Journal of Political Research vol 44: 465-492 JELAVICH, BARBARA 1983 History of the Balkans: Eighteenth and Nineteenth Centuries Cambridge: Cambridge University Press JOVANOVIC, JO VAN 1986 [1912] "Letter of the Serbian Ambassador in Berlin to Serbian PM Nikola Pasic," 16/29 Dec. In Mihailo Vojvodic (ed), Dokumenti o spoljnojpolitici Kraljevine Srbije, 1903-1914 Vol 5, Book 3, Belgrade: SANU JOVANOVIC, SLOBODAN 1932 Politicke ipravne rasprave, Belgrade: Geca Kon JOVIC, DEJAN 2003 Jugoslavija — Drzava kojaje Odumrla: Uspon, kriza ipad Cetvrte Jugoslavije Belgrade: Samizdat B92 KALIN, RUDOLF 1996 'Ethnic attitudes as a function of ethnic presence,' Canadian Journal of Behavioral Science 28 (3): 253-261 KALIN, WALTER 2004 "Decentralized Governance in Fragmented Societies: Solution or Cause of New Evils?" In Andreas Wimmer (ed.) Facing Ethnic Conflicts Lanham, Md: Rowman & Littlefield Publishers, pp. 301-314 KALYVAS, STATHIS and SAMBANIS, NICHOLAS 2005 "Bosnia's civil war: origins and violence dynamics" Understanding Civil War, Vol 2, Washington, DC: The World Bank, pp. 191-230 KAPLAN, ROBERT 1993 Balkan Ghosts: a Journey through History New York, St. Martin's Press KARADZIC, VUK 1972a [1837] Crna Gora i Crnogorci Belgrade: NOLIT 141

— 1972b [1837] Crna Gora i Boka Kotorska Belgrade: NOLIT — 1972c [1860] Srpska Istorija Naseg Vremena Belgrade: NOLIT KEATING, MICHAEL 1997 Stateless Nation-Building: Quebec, Catalonia and Scotland in the Changing State System, Nations and Nationalism 3 (4): 689-717. KISSINGER, HENRY 1997 'Limits to What the US can do in Bosnia*, Washington Post, 22 September: A19 KITSCHELT, HERBERT 1995a 'Formation of party cleavages in post-communist democracies,' Party Politics 1 (4): 447-472 — 1995b The Radical Right in Western Europe, Ann Arbor: University of Michigan KNIGGE, PIA 1998 'The ecological correlates of right-wing extremism in Western Europe,' European Journal of Political Research 34: 249-279 KOLSTO, PAL 1995 Russians in the Former Soviet Republics, Bloomington : Indiana University Press KOSTUNICA, VOJISLAV 1988 "The Constitution and the Federal States," Dennison Rusinow (ed) Yugoslavia: A Fractured Federalism, Washington: Wilson Center Press, pp. 78-92 KRIZMAN BOGDAN and HRABAK, BOGUMIL. 1960 Zapisinici sa sednica delegacye Kraljevine SHS na Mirovnoj konferenciji u Parizu, Belgrade: Kultura KRIZMAN, BOGDAN 1977 'Elaborat dra Ive Andrica o Albaniji iz 1939. godine,' Casopis za suvremenu povijest, Vol 2: 77-90 KULJIC, TODOR 1997 "Titova Nacionalna Politika," Istorija 20. Veka, Br 2. LAMPE, JOHN R. 2002 'Reconnecting the 20th Century Histories of Southeastern Europe' in Ideology and National Identity in The 20th Century Southeastern Europe, J.R. Lampe and M. Mazower, eds., (Manuscript), pp. 1-14 — 2000 Yugoslavia as History, Twice There was a Country, 2nd edition, Cambridge: Cambridge University Press LINZ, JUAN 1980 'Political space and fascism as a late-comer,' in Stein Larsen et al. (eds.), Who Were the Fascists, Bergen: Universitetforlaget, pp. 153-189 LIVEZEANU, IRINA 1995 Cultural Politics in Greater Romania, 1918-1930, Ithaca: Cornell University Press LJUSIC, RADOS 1994 Srbija 19. Veka, Belgrade: Vojnaknjiga LUBBERS, MARCEL and SCHEEPERS, PEER 2002 'French Front National voting: a micro and macro Perspective,' Ethnic and Racial Studies 25 (1): 120-49 LUKIC, VOJIN 1990 Brionski Plenum, Beograd: Strucna Knjiga LUTHER, KURT 2003 'The FPO,' in Peter Merkl and Leonard Weinberg (eds), Right- Wing Extremism in the Twenty-First Century, London: Frank Cass, pp. 191-219 MANN, MICHAEL 2005 The Dark Side of Democracy, Cambridge: Cambridge University 2001 'Explaining the Murderous Ethnic Cleansing: The Macro Level,' in Monserrat Guibernau and John Hutchinson (eds), Understanding Nationalism Cambridge: Polity Press MANNING, NICK 2004 'Diversity and change in pre-accession Central and Eastern Europe since 1989,' Journal of European Social Policy 14 (3): 211-232 MARKOTICH, STAN 1995a 'Serbian Radical denounces Milosevic' RFE/RL Newsline, November 14 —1995b 'Is Serbia still manufacturing poisonous gas?' RFE/RL Newsline, November 27 142

MARKOVIC, MIHAJLO 1991 "Problem Kosova" in Aleksa Djilas ed. Srpsko Pitanje, Belgrade: Politika, pp. 207-224 MARKOVIC, SVETOZAR 1960 'Srbija na Istoku' in Sabrani Spisi, Vol. Ill, Belgrade: NOLIT MARX, KARL 1991 [1852] The 18th Brumaire of Lois Bonaparte New York: International Publishers MAZOWER, MARK 2000 The Balkans: From the End of Byzantium to the Present Day London: Weidenfeld & Nicolson MCCARTHY, JUSTIN 1995 Death and Exile: The Ethnic Cleansing of Ottoman Muslims, 1821-1922, Princeton: Princeton University Press McGARRY, JOHN 1998 "Orphans of Secession: National Pluralism in Secessionist Regions and Post-Secession States" in Margaret Moore National Self- Determination and Secession Oxford: Oxford University Press, pp. 215-232 MEDIUM GROUP 2003 Media Monitoring: Parlamentarni Izbori 2003 Novi Sad, Serbia: Medium Group MERKL, PETER 2003 'Stronger than ever', in Peter Merkl and Leonard Weinberg (eds) Right-Wing Extremism in the Twenty-First Century, London: Frank Cass, pp. 23- 46 MESTROVIC, IVAN 1993 Uspomene napoliticke ljude i dogadaje, Zagreb: Nakladni Zavod Matice Hrvatske MESTROVIC, STJEPAN 1993 Habits of the Balkan Heart: Social Character and the Fall of Communism, College Station : Texas A&M University Press MIJATOVIC, BOSKO 2005 'Reforma radnih odnosa i trzista rada,' in Boris Begovic and Bosko Mijatovic (eds.), Cetiri Godine Tranzicije u Srbiji, Belgrade: CLDS, pp. 293-336 MILJANOV, MARKO 1963 Sjaj Legende Titograd, Yugoslavia: Graficki Zavod. MILOSAVLJEVIC, OLIVERA 2002 U Tradiciji Nacionalizma, Belgrade: Helsinski Odbor za Ljudska Prava u Srbiji MINKENBERG, MICHAEL and SCHAIN, MARTIN 2003 'The Front National in context,' in Peter Merkl and Leonard Weinberg (eds), Right-Wing Extremism in the Twenty-First Century, London: Frank Cass, pp. 161-190 MINKENBERG, MICHAEL 2002 'The radical right in post-socialist Central and Eastern Europe: comparative observations and interpretations,' East European Politics and Societies 16 (2): 335-362 MINISTARSTVOINOSTRANIHDELA KRAUEVINE SRBIJE 1998 [1899] Prepiska o Arbanaskim Nasiljima u Staroj Srbiji, 1898-1899. Belgrade: Nikola Pasic MUDDE, CAS 2000 ' Extreme-right parties in Eastern Europe', Patterns of Prejudice, 34, 1, pp. 5-27 MYLES, JOHN and St-ARNAUD, SEBASTIEN 2007 "Population Diversity, Multiculturalism, and the Welfare State" in Keith Banting and Will Kymlicka (eds.), Multiculturalism and the Welfare State, Oxford: Oxford University Press, pp. 339-354 NACIONALNA SLUZBA ZA ZAPOSLJAVANJE 2002-2004 'Mesecni statisticki bilten,' URL: www.rztr.co.vu/statistika.htm NENADOVIC, MATIJA 1969 Memoirs (Translated by Lovett F. Edwards) Oxford: Claredon Press 143

NJEGOS, P. PETROVIC 1978 The Mountain Wreath [Gorski Vijenac], Toronto: Jugoslavika — 1967 Izabrana Pisma Titograd, Yugoslavia: Graficki zavod OLIVER, ERIC and WONG, JANELLE 2003 'Intergroup prejudice in multiethnic settings,' American Journal of Political Science 47 (4): 567-82 OLZAK, SUSAN and NAGEL, JOANE 1986 'Competitive ethnic relations: an overview,' in Susan Olzak and Joane Nagel (eds), Competitive Ethnic Relations, Orlando: Academic Press, pp. 1-12 OST, DAVID and CROWLEY, STEPHEN 2001 "Making sense of labor weakness in post communism," in David Ost and Stephen Crowley (eds.) Workers After Workers' States : Labor and Politics in Postcommunist Eastern Europe Lanham, MD : Rowman & Littlefield Publishers OST, DAVID 1999 'The Radical Right in Poland,' in Sabrina P. Ramet (ed.), The Radical Right in Central and Eastern Europe Since 1989, University Park, PA: Pennsylvania State University, pp. 85-108 PANTIC, DRAGOMIR 2002 'Vrednosti biraca pre i posle demokratskog preokreta 2000,' in V. Goati (ed.), Partijska Scena Srbije posle 5. Oktobra 2000, Belgrade, Serbia: IDN, pp. 79-130 PASIC, NIKOLA 1986 [1912] "Letter to the High Command of the Army of the Kingdom of Serbia," 16/29 Dec In Mihailo Vojvodic (ed), Dokumenti o spoljnoj politici Kraljevine Srbije, 1903-1914 Vol 5, Book 3, Belgrade: SANU PAVLOWITCH, STEVAN K. 2002 Serbia: The History Behind the Name, London: Hurst & Co PAYNE, STANLEY 1995 A History of Fascism, 1914-1945, Madison: University of Wisconsin PEHE, JIRI1996 'Czech extreme-right leader causes another scandal,' OMRIDaily Digest, Radio Free Europe/Radio Freedom, July 30 : 146 PENC, STANISLAV and URBAN, JAN 1998 'Czech Republic's extremist acts galvanize Roma population', Transitions, vol. 5, no. 7, July: 39-40 PETERSEN, ROGER 2002 Understanding Ethnic Violence: Fear, Hatred, and Resentment in Twentieth-Century Eastern Europe, Cambridge: Cambridge University Press. PETROV, ALEKSANDAR 1991 'Kosovo ~ Sveta prica srpskog naroda' in Aleksa Djilas, ed. Srpsko Pitanje, Belgrade: Politika PETROVIC, RUZA 1987 Migracije u Jugoslaviji i Etnicki Aspekt, Beograd: Radisa Timotic — 1985 Etnicki Mesoviti Brakovi u Jugoslaviji Beograd: Institut za Socioloska Istrazivanja PIERSON, PAUL 2004 Politics in Time Princeton: Princeton University Press 2003 "Big, Slow-Moving, and ... Invisible: Macrosocial Processes in the Study of Comparative Politics," in James Mahoney and Dietrich Rueschemeyer (eds) Comparative Historical Analysis in the Social Sciences, Cambridge: Cambridge University Press, pp. Ill-201 POPOVIC, DANICA 2005 'Privredna aktivnost i makroekonomska politika u tranziciji,' in Boris Begovic and Bosko Mijatovic (eds.), Cetiri Godine Tranzicije u Srbiji, Belgrade: CLDS, pp. 35-117 144

POPOVIC, JANJICIJE 1987 Zivot Srba na Kosovu, 1812-1912, Belgrade: Knjizevne Novine POULTON, HUGH 2003 'Macedonians and Albanians as Yugoslavs,' in Dejan Djokic (ed.), Yugoslavism, London: Hurst, pp. 115-135 -— 2000 'The Muslim Experience in the Balkan States, 1919-1991', Nationalities Papers, Vol. 28, No.l: 45-66 RADUSKI, NADA 2001 'Nacionalne manjine i medjuetnicki odnosi na Balkanu,' Socioloski Pregled 35 (1-2): 75-91 Radio Free Europe 1999 'RFE Newsline,' March 30 RAGIN, CHARLES 2000 Fuzzy-Set Social Science Chicago: The University of Chicago Press — 1992 "Introduction" in Charles Ragin and Howard Becker (eds.) What is a case? : Exploring the Foundations of Social Inquiry Cambridge: Cambridge University Press, pp. 1-18 RAKIC, MILAN 1985 Konzulska Pisma, 1905-1911, Belgrade: Prosveta RAMET, PEDRO 1984 Nationalism and Federalism in Yugoslavia, 1963-1983 Bloomington:Indiana University Press RIEBER, ALFRED J. 2000 'Repressive Population Transfers in Central, Eastern and Southeastern Europe: A Historical Review,' in A. J. Rieber, ed., Forced Migration in Central and Eastern Europe, 1939-1950, London: Frank Cass, pp. 1-27 ROSHWALD, AVIEL 2001 Ethnic Nationalism and The Fall of Empires: Central Europe, Russia, and the Middle East, 1914-1923, London: Routledge RUDIC, VUJADIN 1978 Stanovnistvo Toplice, Belgrade: SANU RUESCHEMEYER, DIETRICH AND STEPHENS, JOHN 1997 "Comparing Historical Sequences — A Powerful Tool for Causal Analysis," Comparative Social Research, vol. 16, pp. 55-72. RUSINOW, DENNISON 1988 "History and Status of Ethnic Conflicts," Dennison Rusinow (ed) Yugoslavia: A Fractured Federalism, Washington: Wilson Center Press, pp. 23-40 — 1985 Nationalities Policy and the "National Question" in Pedro Ramet (ed), Yugoslavia in the 1980s Boulder: Westview Press — 1977 The Yugoslav Experiment, 1948-1974 Berkeley and Los Angeles: UCLA Press SCHUMAN, HOWARD, STEEH, CHARLOTTE, and BOBO, LAWRENCE 1985 Racial attitudes in America, Cambridge, MA: Harvard University SCHUTZE, F. and BIGELOW, A. 1993 Russia's Parliamentary Elections 1993 and 1995, Columbus, OH: The Current Digest of the Soviet Press, 1996 SEKELJ, LASLO 1990 Jugoslavia: Struktura Raspadanja, Belgrade: Rad SEKULIC, DUSKO, MASSEY, GARTH and HODSON, RANDY 1999 'Ethnic enclaves and intolerance,' Social Forces 78 (2): 669-93 — 1994a 'Who were the Yugoslavs?' American Sociological Review 59: 83-97 — 1994b 'National tolerance in the former Yugoslavia,' American Journal of Sociology 99(6): 1534-1558 SERBIAN STATISTICS BUREAU [SSB] 2000-2004 'Opstinskipokazatelji' [Cumulative dataset], URL: http://web.rzsstatserb.sr.gov.yu/axd/index.php — 2002 Opstine u Srbiji, 2002, Belgrade, Serbia: Republicki Zavod za Statistiku SHAFIR, MICHAEL 1999 'The mind of Romania's radical right,' in Sabrina P. Ramet 145

(ed.), The Radical Right in Central and Eastern Europe Since 1989, University Park, PA: Pennsylvania State University Press, pp. 213-232 SHEA, CHRISTOPHER 1997 "Political scientists clash over value of area studies" The Chronicle of Higher Education, vol 43, 18:4-5 SHOUP, PAUL 1968 Communism and the Yugoslav National Question New York: Columbia University Press SKENDI, STAVRO 1967 The Albanian National Awakening, 1878-1912 Princeton: Princeton University Press SKOCPOL, THEDA 1984 "Emerging agendas and recurrent strategies in historical sociology," in Theda Skocpol (Ed.) Vision and Method in Historical Sociology, Cambridge: Cambridge University Press, pp. 356-391 SKOPLJANAC, NED A, GREDELJ, STJEPAN, and HODZIC, ALIJA (eds) 2000 Media and War, Zagreb: Centre for Transition and Civil Society Research SKORZZA, TAMARA 2005 'Seselj i Milosevic,' Vreme 764, August 25 URL: http://www.vreme.com/ SLAVUJEVIC, ZORAN 2002 'Socijalna utemeljnost politickih stranaka pre i posle izbora 2000,' in Vladimir Goati (ed.), Partijska Scena Srbije posle 5. oktobra 2000, Belgrade, Serbia: Institute for Social Sciences, pp. 159-194 SMAJLOVIC LJILJANA 1997 Tntervju: Vojislav Seselj,' Vreme, April 12 URL: http://www.vreme.com/ SMELSER, NEIL 1962 Theory of Collective Behavior, New York: Free Press SNIJDERS, TOM and BOSKER, ROEL 1999 Multilevel Analysis, London: Sage SNYDER, JACK and MANSFIELD, EDWARD 2005 Electing to Fight Cambridge: MIT Press SNYDER, JACK 2000 From Voting to Violence, New York: WW Norton STAMBOLIC, IVAN 1995 Put u Bespuce Beograd: Radio B-92 STANKOVIC, DJORDJE 1988 'Nikola Pasic i stvaranje albanske drzave,' in Dj. Stankovic, ed., Iskusenja Jugoslovenske Istoriografije, Belgrade: Rad STAVRIANOS, LEFTEN S. 2000 The Balkans Since 1453, New York: New York University Press STINCHCOMBE, ARTHUR L. 1968 Constructing Social Theories, New York: Harcourt STOJANCEVIC, VLADIMIR 1980 Srbija u vreme Prvog Srpskog Ustanka 1804-1813, Leskovac: Biblioteka narodnog muzeja SWANK, DUANE and BETZ, HANS-GEORG 2003 'Globalization, the welfare state and right-wing populism in Western Europe,' Socio-Economic Review 1: 215-45 SZAYNA, THOMAS 2000 Identifying Potential Ethnic Conflict, Santa Monica: RAND SZELENYI, IVAN et al. 1997 "Left turn in post-communist politics: bringing class back in?"East European Politics and Society vol 11, (1) Winter: 190-224 TADIC, LJUBOMIR 1986 Da lije Nacionalizam nasa Sudbina? Belgrade: Knjizevne Novine THELEN, KATHLEEN 2003 "How institutions evolve: insights from comparative historical analysis," in James Mahoney and Dietrich Rueschmeyer (eds) Comparative Historical Analysis in the Social Sciences, Cambridge: Cambridge University Press, pp. 208-240 THOMPSON, MARK 1999 Forging War (Revised edition), London: University of Luton 146

TILLY, CHARLES 1995 "Contentious Repertoires in Great Britain, 1758-1834," in Mark Traugott (Ed.) Repertoires and Cycles of Collective Action Durham: Duke University Press, pp. 15-42 — 1975 'Reflections on the History of European State-Making', in Charles Tilly, ed. The Formation of National States in Western Europe, Princeton: Princeton University Press, pp. 17-31 TODOROVA, MARIA 1997 Imaging the Balkans New York: Oxford University Press TOMC, GREGOR 1988 "Classes, Party Elites, and Ethnic Groups" Dennison Rusinow (ed) Yugoslavia: A Fractured Federalism, Washington: Wilson Center Press, pp. 58-77 TROTSKY, LEON 1993 The Balkan Wars, 1912-13: The War Correspondence, Edited by G. Weissman and D. Williams, New York: Pathfinder TUCOVIC, DIMITRIJE [1914] 'Srbija i Arbanija' in Sabrana Dela, (1975-1981) Belgrade: Rad — [1913a] 'Rezim u novim krajevima,' Borba, Nov 16 in Sabrana Dela, (1975-1981) Belgrade: Rad — [1913b]' Krvna osveta soldateske,' Radnicke Novine, Oct. 22 in Sabrana Dela, (1975-1981) Belgrade: Rad — [1910a] Interview for La Jeune Turquie, Sep 4 in Sabrana Dela, (1975-1981) Belgrade: Rad — [1910b] "Prva Socijaldemokratska Balkanska Konferencija, Die Neue Zeit, March 11, in Sabrana Dela, (1975-1981) Belgrade: Rad — [1910c] 'Albansko Pitanje,' Borba, May 1, 1910, in Sabrana Dela, (1975-1981) Belgrade: Rad TWORZECKI, HUBERT 2003 Learning to Choose, Stanford, CA: Stanford University UNHCR 2001 'Registration of internally displaced persons from Kosovo and Metohija (Report),' URL: www.internal-displacement.org/8025708F004CE90B /(httpCountries)6A36317DA8D29A1802570A7004CAB90?opendocument&coun t=10000 VEUGELERS, JOHN 2005 'Ex-colonials, voluntary associations, and electoral support for the contemporary far right,' Comparative European Politics 3: 408-431 VEUGELERS, JOHN and MAGNAN, ANDRE 2005 'Conditions of far-right strength in contemporary Western Europe: an application of Kitschelt's theory', European Journal of Political Research, vol. 44, pp. 837-860 VOJVODIC, MIHAILO 1986 Dokumenti o spoljnojpolitici Kraljevine Srbije, 1903-1914 Vol 5, Book 3, Belgrade: SANU VUCIC, ALEKSANDAR 1995 'Poslanickapolitika,' Vreme, August 21: 15 VUCINICH, WAYNE, S. 1969 "Nationalism and Communism" in Wayne Vucinich, Contemporary Yugoslavia Berkeley: University of California Press VUJACIC, VELJKO 2003 "From Class to Nation" East European Politics and Societies Vol. 17, No. 359-392 VUJACIC, VELJKO and ZASLAVSKY, VICTOR 1991 "The Causes of Disintegration in the USSR and Yugoslavia," Telos, Number 88, Summer: 120-140 WALKER, HENRY and COHEN, BERNARD 1985 " Scope statements: imperatives for evaluating theory," American Sociological Review, June, vol. 50, pp. 288-301 WEINBERG, LEONARD 2003 'Conclusion,' in Peter Merkl and Leonard Weinberg 147

(eds.), Right-Wing Extremism in the Twenty-First Century, London: Frank Cass, pp. 293-302 WHITMORE, BRIAN 2005 'Serbian Army reportedly protected Mladic until 2002 as list of those who helped Mladic made state secret,' RFE/RL Newsline 10 (20) URL: http://www.rferl.org/ WILSON, WILLIAM J. 1987 The Truly Disadvantaged, Chicago: The University of Chicago Press — 1980 [1978] The Declining Significance of Race, 2nd edition, Chicago: The University of Chicago Press WIMMER, ANDREAS 2006 'Ethnic exclusion in nationalizing states', in Gerard Delanty and Krishan Kumar (eds), The Sage Handbook of Nations and Nationalism, London: Sage, pp. 334-344 — 2004 'Does ethnicity matter? Everyday group formation in three Swiss immigrant neighborhoods,' Ethnic and Racial Studies 27 (1): 1-36 — 1997a 'Explaining xenophobia and racism: a critical review of current research approaches,' Ethnic and Racial Studies 20 (1): 17-41 — 1997b 'Who owns the state? Understanding ethnic conflict in post-colonial societies,' Nations and Nationalism 3 (4): 631-665 WOODWARD, SUSAN 1995 Balkan Tragedy, Washington, DC: Brookings Institution WORLD BANK 2002 Transition: The First Ten Years, Washington, DC: World Bank, URL: http://lnwebl8.worldbank.org/ECA/eca.nsf/Attachments/Transitionl/ $File/complete.pdf YUGOSLAV FEDERAL STATISTICS BUREAU \99% Yugoslav Censuses, 1921-1991 [CD], Belgrade: RTS ZARKOVIC-BOOKMAN, MILICA 1992 "Economic Issues Underlying Secession: the Case of Slovenia and Slovakia" Communist Economies and Economic Transformation, Vol 4, No. 1:111-134 ZASLAVSKY, VICTOR and BRYM, ROBERT 1983 Soviet-Jewish Emigration and Soviet Nationality Policy London : Macmillan ZASLAVSKY, VICTOR 1997 "The Soviet System and the Soviet Union: Causes of Collapse" in Karen Barkey and Mark von Hagen (eds) After Empire: Multiethnic Societies & Nation-Building, the Soviet Union & Russian, Ottoman & Habsburg Empires Boulder, Colo.: Westview Press, pp. 73-98 Zemunske Novine 1997 'Interview with Vojislav Seselj,' December, pp. 15-19 ZDRAVKOVIC, HELENA 2005 Politika Zrtve na Kosovu, Belgrade: Cigoja Stampa.