previous sub-section
Kosovo's Changing Institutional Fate
next part


Despite its partial loss of empirical sovereignty over Kosovo, Serbia had not yet moved to full-scale ethnic cleansing by spring 1999, suggesting

that as long as Western powers recognized Serbian juridical sovereignty over the province, despotism was not an attractive option. Serbia, it seemed, still had too much to lose by pushing its own citizens out of territory that it lawfully ruled. Once Western powers questioned Serbia's juridical sovereignty over Kosovo by launching the air war, however, the region was no longer fully part of the Serbian core, and Belgrade lost all semblance of restraint. This is a controversial claim, as it suggests that NATO bears indirect responsibility for Serbian ethnic cleansing. It was hotly denied by NATO representatives, who say that Serbia was already in the process of ethnically cleansing the province when NATO intervened. Most available evidence, however, suggests otherwise.

The best data come from the OSCE, whose monitors were on the ground until four days before the NATO air war began. Its report says Serbian ethnic cleansing began in earnest when its monitors withdrew on March 20, and escalated dramatically when the air raids began.[102] A New York Times report makes a similar claim, saying the Serbian attack "kicked into high gear on March 24, the night NATO began bombing Yugoslavia."[103] My own interviews along the Albanian border lend credence to this view. According to dozens of refugees from Kosovo's urban centers, Serbian troops began emptying the region's towns for the first time on March 25. Drawing on my research, Human Rights Watch wrote on March 30, 1999, that "the Yugoslav government evidently made a decision over the weekend [of March 25–27] to ‘cleanse’ the region of ethnic Albanians."[104] Indeed, the rate of ethnic Albanian depopulation in April and May 1999 was ten times greater than during the most intense Serbian offensives of 1998.[105]

Western politicians and NATO officials were uncomfortable with these facts, as they suggested the air war endangered the very people they were trying to protect. On March 28, 1999, President Clinton denied the NATO bombings were accelerating Serbia's expulsions, and NATO officials said shortly thereafter that the air war had only pushed Serbian forces to speed up an existing expulsion plan.[106] As evidence, officials pointed to Serbian troop mobilizations in Kosovo in January 1999[107] and a secret Serbian plan, "Operation Horseshoe."[108] According to the German government, Belgrade devised Horseshoe in late 1998 and set it in motion during January 1999, months before the NATO air war began.[109] The plan allegedly ordered Serbian forces to begin attacking Kosovo from the north, east, and west, forcing the population to flee southward.[110] In June 1999, KLA soldiers and British reporters in Kosovo said they had discovered proof of Horseshoe amidst captured Serbian documents.[111]


The notion that NATO's intervention only slightly accelerated an ongoing Serbian expulsion campaign is radically at odds with Western intelligence assessments prior to the air war, however. In early 1999, U.S. intelligence officials believed that an upcoming Serbian offensive in Kosovo would be a limited attack. According to experts on U.S. military policy, the Central Intelligence Agency did not even "prominently raise the possibility" of systematic depopulation in the months leading up to the air assault. The commander of NATO's Serbia war operations agreed, saying, "We never expected that the Serbs would push ahead with the wholesale deportation of the ethnic Albanian population."[112] At least some U.S. officials continued to maintain this position after the NATO air war began, despite the embarrassment it caused their government. Five days after the air campaign began, a Pentagon spokesman said no one "could have foreseen the breadth of this [Serbian] brutality," contradicting his own president, who had stated one day earlier that the United States had intervened precisely because it knew a Serbian ethnic cleansing offensive was imminent.[113] In fact, the evidence suggests that prior to the war, most U.S. analysts believed Serbia would at the very worst expel some 350,000 persons from their homes, repeating their 1998 actions.[114] The dearth of humanitarian provision along Kosovo's borders lends credence to this view. According to UN relief workers, Western governments did not warn of a mass flow of refugees prior to the air war.[115] It is also true that no supplies had been pre-positioned along Kosovo's borders prior to the launching of NATO's air war.[116]

Was there, then, any Serbian expulsion plan at all? According to Serbian reporter Braca Grubačić, editor of a respected English-language newsletter in Belgrade, there was no preconceived plan. "There were vague ideas about expulsions" prior to the NATO attack, he said, but no premeditated ethnic cleansing scheme. Once the bombing began, however, Serbian troops and paramilitaries "just did it," since there was a broad Serbian attitude of "we'll fuck'em if they start."[117] This argument is indirectly supported by retired German general Heinz Loquai, who claims "Operation Horseshoe" was a German government invention aimed at legitimating its controversial participation in the Kosovo war.[118] According to Loquai, German intelligence obtained vague reports via Bulgarian security sources of Serbian plans for Kosovo and then repackaged the rumors as the full-blown "Operation Horseshoe."

The speed and efficiency with which Serbia carried out the expulsions, however, makes the notion of an entirely spontaneous effort seem unlikely. Serbian operations were too rapid, systematic, and coordinated to

have been thrown together, in the heat of war and at the last moment, by impetuous Serbian fighters. Some careful planning must have been in place before the air war, even as only one of several possible scenarios. A reasonable interpretation of events is that Serbian officers, like their counterparts worldwide, prepared different scenarios for different contingencies during late 1998 and early 1999. A substantial military offensive aimed at clearing certain pro-KLA areas of fighters and their civilian supporters is likely to have been prepared for spring 1999, as most Western intelligence officials seem to have anticipated. A broader and more comprehensive effort to empty Kosovo of all ethnic Albanians is also likely to have been on the drawing boards, however, just in case the opportunity arose. As I argued in the introduction to Part I, the Serbian national idea contained a bundle of multiple and conflicting interpretations, and there was no single, cohesive set of tactics for achieving Serbian national goals. Rather, there were multiple possibilities and interpretations, only some of which were translated into action at specific times by particular institutional settings.

Operation Horseshoe, or its functional equivalent, was set in motion in March 1999 because Kosovo's institutional setting had been dramatically transformed by NATO's determined military intervention. "NATO's bombing," a Brookings Institution study argues, "lifted a constraint on the Serb leader that may have been operative until that point. Before that point, [Milošević] had an incentive to keep NATO from attacking him. Once the attack was under way, however, he no longer had that same reason to hold back."[119] In institutional terms, the NATO air strike expelled Serbia from the Western-dominated "international community" and appeared deeply threatening to Serbia's juridical and empirical sovereignty over the province. Although Kosovo was still theoretically located within the Serbian core (a status NATO officially said it had no intention of changing), it appeared on the verge of escaping Serbia's orbit through a combined KLA/NATO effort. Kosovo's ghetto status was evaporating, transforming the region into a Bosnia-like "frontier." With the West finally bombing Belgrade, Serbia no longer could expect any benefits from holding back. It had been isolated from the Western-dominated international community and thus could not expect any gain from continuing to observe any vestige of norms against forced deportation.

Why, specifically, did Serbia pursue ethnic cleansing in Kosovo? First, the notion of changing the ethnic balance in disputed regions was a powerful strand of Serbian political discourse. Many in Serbia saw continued

Serbian rule over Kosovo as the only way to protect local ethnic Serbs and preserve the country's national heritage. Kosovo's ethnic Albanian majority, in this view, presented an acute political, military, and demographic threat. Beyond this fundamental point, however, Serbia seemed to have additional tactical considerations. First, it hoped to weaken NATO's resolve by destabilizing Macedonia and Albania.[120] With Italy, Germany, and other European countries fearful of being forced to accept more refugees, moreover, Serbian decision makers may have also gambled that the mass outflow might drive a wedge between the United States and its European allies. Third, Serbia hoped to defeat the KLA by reducing its pool of potential supporters.[121] Although in spring 1999 all areas of Kosovo experienced ethnic cleansing, pro-KLA regions were hardest hit.[122] And finally, Serbian military planners may have hoped to hinder a NATO ground invasion by crowding the roads with refugees. These tactical considerations all backfired, however, as televised images of refugees provided powerful justification for NATO's intervention, boosting popular support in Europe and the United States for the war.[123]

In previous chapters, we saw different manifestations of Serbian nationalism spread over different geographic locales. Chapter 3 discussed the most virulent manifestations of Serbian nationalism in Bosnia, arguing that the region's frontier-like setting led to ethnic cleansing. Chapter 4, by contrast, showed a less despotic manifestation of Serbian nationalism within the Serbian core, where semi-private paramilitaries and nationalist crisis committees were partially constrained by the Serbian state.

The events described in this chapter condense these experiences into one geographic locale, the disputed province of Kosovo, where institutional settings changed over time, not space. Between 1989 and 1997, Kosovo was firmly controlled by Serbia and was tightly integrated into the Serbian core. The West extended the norm of human rights to the province, but steadfastly refused to recognize ethnic Albanian pleas for Kosovo's independence. At the same time, there was a lack of a serious armed insurgency on the ground; the result of these combined circumstances was that Serbian empirical and juridical sovereignty was maintained. Although this left ethnic Albanians trapped under a harsh regime of ethnic policing, it also shielded them from Bosnia-style ethnic cleansing. As had been true in the Sandžak and Vojvodina, semi-private Serbian nationalists threatened extreme violence against Kosovo, but did not follow through with action—or not, at least, until conditions changed.


In 1998, Kosovo's institutional setting shifted dramatically following an effective ethnic Albanian insurgency threatening Serbia's empirical sovereignty over the province. Pockets of rural territory became no-go areas for Serbian forces, and KLA rebels managed to seize as much as 40 percent of the territory for a short time. Kosovo's ghetto status was disrupted, although not entirely dismantled, and Serbian ethnic cleansing was pernicious but piecemeal. Some rural areas were violently depopulated, but others were not, and Kosovo's urban concentrations were not touched. This all changed in 1999, when the withdrawal of OSCE monitors and the NATO air war cut Serbia off from its last ties to the world community and presented the regime with a serious challenge to its control over Kosovo. As had been true for Bosnia, the province was being dramatically and quickly "externalized" by Western action, and again, Serbia responded with ethnic cleansing. Having lost infrastructural control over Kosovo, ethnic cleansing, not ethnic policing, became Serbia's repertoire of violence.

Luckily for Kosovo's ethnic Albanians, the West had greater conviction in 1999 than in 1992. Whereas Bosnia was abandoned to its fate by vaguely well-meaning but entirely undercommitted Western powers in April 1992, Kosovo was not. NATO's intervention was indirectly responsible for the Serbian assault on civilians, but NATO troops eventually saw the task through and brought the refugees home. Some 10,000 ethnic Albanians paid with their lives, however, and hundreds of thousands had lost their homes and possessions.

Part I chronicled the rise of Serbian nationalism in the 1980s and early 1990s and the subsequent waves of state-supported or tolerated violence in Bosnia and Serbia proper. Those waves, however, were by no means uniform. Rather than seeing "Serbian nationalist violence" as a homogeneous phenomenon, I have sought to highlight the varieties in repertoires of Serbian state violence across different institutional settings. In Bosnia, Serbian nationalism led to ethnic cleansing during 1992–93, the first terrible year of the Bosnian war, but in the Serbian core, ethnic cleansing did not occur until 1998–99, and then only in Kosovo.

Contemporary Serbian nationalism contained both radical and more moderate strands. The most radical elements, which conventional Western wisdom has most come to associate with Serbia, defined membership in the Serbian community in purely ethnic terms and was committed to

establishing a Greater Serbian state. In that vision, non-Serbs had little hope of fair treatment. More moderate strands of Serbian political discourse were influenced by the longtime existence of a Serbian republic within the socialist and anti-nationalist Yugoslavia, and the insertion of that republic in a wider global context. For states in the post–World War II world, be they communist or liberal, membership in the global universe of moral obligation was at least formally defined by universalistic criteria, including notions of law and order and bureaucratic due process. This phenomenon was increasingly reinforced by global-level norms of human rights.

Serbian nationalism in the 1980s, like many similar ideologies, was an amalgam of both radical and moderate strands, or perhaps more accurately, of national-particularism and universalism. The two coexisted uneasily, and neither achieved lasting hegemony. Radical particularism was victorious in some areas at some times, while strands of universalism had the upper hand elsewhere. Within Serbia proper the state enjoyed unrivaled infrastructural powers, and was thus unwilling to let extreme national particularism reign supreme. Serbia's ethnic cleansing of areas under its juridical and empirical sovereignty would have demolished its identity as a modern, liberal, or socialist state, and would have led to its exile from the wider community of nations. Neither could the Serbian state abandon nationalism altogether, however, because the ruling elite maintained power through its legitimating discourse, and because a nationalist counter-elite was waiting impatiently in the wings. A Serbian government that openly repudiated the Serbian people and abandoned diaspora Serbs altogether risked sparking a substantial domestic political challenge.

In 1992 and 1993, a unique set of events created an institutional setting that led to a despotic regime of power in Bosnia, permitting the most radical strands of Serbian nationalism to predominate. The trigger was the ability of Slovenia and Croatia to win international recognition of their sovereignty, which paved the way for Bosnian independence. Once the West applied the norm of sovereignty, rather than human rights, to Bosnia, the stage was set for the creation of frontier-like conditions. In April 1992, Bosnia was internationally recognized as a sovereign entity distinct from rump Yugoslavia, composed of Serbia and Montenegro, creating two separate institutional entities. Western powers expected sovereignty would insulate Bosnia from Serbian nationalism, but the exact opposite took place.

By separating Bosnia from Serbia, the West provided the radical

strand of Serbian nationalism an opportunity to free itself from the constraints of international and domestic norms. Chaos reigned in Bosnia, but the new, Western-imposed border absolved Serbia from legal responsibility for events. As a result, Serbian nationalism could grow to its fullest and most awful dimensions. On the frontier, membership in the universe of moral obligation was defined by Serbian nationalists in purely particularistic terms, leading to the forced removal of non-Serbs from the newly sovereign Bosnia. The institutional setting facilitating this moment of ethnic cleansing was the frontier, while its mechanisms were the paramilitaries, crisis committees, and the covert, cross-border network linking Belgrade to Bosnia.

Inside the Serbian core, things were very different. Here, the Serbian state remained constrained by national and international norms, which, for a time, reinforced the universalistic strands of Serbian political discourse. With its infrastructural power still high and its juridical and empirical sovereignty ensured, Serbia allowed norms of responsibility, law, order, and universal citizenship to predominate.

Throughout most of the 1990s, communal membership in Serbia was a complex affair. Purely national criteria were still strongly supported, and many non-Serbs lived in a state of threat and discrimination. But as citizens of a Serbian state seeking to project an image of multiethnic harmony to itself and to the world, they could lay claim to protection of a sort from state authorities. Unlike non-Serb Bosnians, the non-Serb citizens of Serbia and Montenegro had a platform from which they could defend themselves. They had rights as citizens of the Serbian state, and these could not be entirely withdrawn without the collapse of the state's legitimacy on both the domestic and international fronts. In Kosovo, this system was undermined in 1998–99 due first to the KLA insurgency, and then to NATO's intervention. With both its empirical and juridical sovereignty over the province under threat, Serbia abandoned its remaining commitments to universalism and unleashed a cruelly intense wave of ethnic cleansing.

International diplomacy and norms played crucial roles in the entire process. Both the frontier and the core were internationalized arenas: On the Bosnian frontier, that internationalization took place through the rapid and surprising consolidation of Bosnia as a full-fledged sovereign state and the grafting of the sovereignty norm onto the Bosnia-Serbia border. Inside the Yugoslav/Serbian core, internationalization was already in place in the state's investment in its "citizenship" in the international community. Even non-Serb populations in Sandžak, Vojvodina,

and Kosovo were partially protected by international human rights norms. Internationalization held the Serbian state accountable for events in its territory, forcing it to restrain radical nationalism. On the Bosnian frontier, however, the opposite took place. In Kosovo, international human rights norms initially helped protect ethnic Albanians from forced depopulation, but as Serbian retaliations for KLA actions became increasingly brutal, human rights pressures forced the West toward military intervention.

Another possibility existed for dealing with the demographic and political challenges posed by non-Serbs in Bosnia, Kosovo, and elsewhere. If Serbs wanted to maintain a radically nationalist notion of communal membership without expelling the non-Serb population, they could have defined Muslims and Croats as subjects with few legal rights, creating a two-tier political hierarchy. Non-Serbs would have occupied the lower tier as subordinate subjects, being permitted to stay on their land, but with severely restricted political and legal rights. In order to create such a formalized two-tier system, however, Serbs would have needed at least the tacit permission of international forces, and more specifically, the large Western powers.

Part II deals with precisely one such system: Israel and the Palestinian territories it occupied beginning in 1967. Here, the Jewish-Israeli population, defined in purely ethno-national terms, controlled some 2 million non-Jews with few political and few legal rights. Palestinians were officially inscribed into the Israeli control system as subordinate subjects, leaving them trapped inside the bureaucratic fabric of the Jewish state, exposing them to harsh ethnic policing while simultaneously shielding them from ethnic cleansing. The institutional setting in which this took place was the Palestinian ghetto, whose construction and methods of operation are the topic of the following chapters.

previous sub-section
Kosovo's Changing Institutional Fate
next part