JUST ANOTHER CIVIL WAR?

The United States and its Western partners have long uttered the words “never again” when confronted with the mass suffering of distant strangers at the hands of a foreign government. ¹ Yet, given the inconsistent application of humanitarian norms, such as the Responsibility to Protect (R2P), the phrase has often rung hollow (International Commission on Intervention and State Sovereignty 2001; Doyle 2011). When do Western powers acknowledge humanitarian suffering and intervene militarily on behalf of strangers? In part, it may be a matter of perception. Even uttering “never again” demands that foreign policy actors agree that mass atrocities are occurring. Policy elites’ perceptions of distant violent conflicts may prove imperative to understanding third-party policy choices, contributing to our collective experience of the selectivity gap of humanitarian military interventions (HMI) and third-party conflict management. This article explores the influence of conflict perceptions in changing the likelihood of Western humanitarian military intervention during the Kosovo Crisis of 1999, as also interacting with geopolitical interests and humanitarian norms.

The Kosovo Crisis stands as an international precedent for humanitarian military intervention, partially because it symbolizes both Western power and benevolence, as Hehir (2019) argued. NATO's military intervention in Kosovo in March 1999 was the first time since the founding of the United Nations (UN) that a set of states outside of international and domestic institutions breached another state's sovereignty based on humanitarian justifications. As Bellamy (2009) highlighted, the 1999 NATO intervention prompted growing political consensus on state sovereignty as a responsibility. Yet, it also enflamed the debate on armed third-party interventions for humanitarian purposes. As the years followed, however, a similar humanitarian impetus did not envelop the world consistently. For example, why did NATO wage a humanitarian military intervention in Kosovo while ignoring the graver atrocities in South Sudan and other global conflicts?

NATO's decision to intervene between ethnic Albanian separatists and the Serbian military and paramilitaries came in the backdrop of years of intrastate violence within Kosovo and after almost a decade of such violence within the Balkans (Hehir 2019). Thus, the violence engulfing Kosovo was not always at the top of Western policy agendas. This article asks how the Kosovo Crisis transformed from a relatively ignored domestic conflict in 1997 into a systematic atrocity that threatened to destroy the fabric of Western identity by late 1998.

Before the late 1990s, the international community perceived the Kosovo Crisis as either a typical ethnic civil war or a national government fighting against a terrorist paramilitary group, the Kosovo Liberal Army (KLA). As this article will argue, a robust global response was improbable as long as this “equal culpability” civil war perception dominated among international actors. Conflict perceptions of the Kosovo Crisis required a dramatic transformation before the Western community considered a strong humanitarian response.

To untangle the mechanisms behind the selectivity of humanitarian military interventions, the article uses process tracing of the 1999 precedent-setting Kosovo Crisis through archival evidence and quantitative data on Western media coverage of the progressing conflict. Relying on NATO archives and elite dialogue—mainly via the Foreign Affairs Oral History (FAOH) Collection—and data on media conflict perceptions, I analyze how Kosovo moved from the periphery of Western attention to becoming the litmus test of Western security and values.

The scope of this article's contribution is to explore how the NATO 1999 intervention in Kosovo required shifting conflict perceptions, not geopolitical interests or humanitarian outrage alone. I argue that the perceptions activated humanitarian norms and interests alike in Kosovo. Contrary to the standard literature that focuses on humanitarian norms or national interests as the primary drivers of the intervention, this article highlights the transformation in the Western elite perception of the Kosovo Crisis from “just another civil war” to a systematic killing/genocide. The Kosovo Crisis “earned” a Western military intervention because of shifting, favorable conflict perceptions which encouraged Western institutions’ involvement.

The West knew of the brutal crimes occurring in Kosovo long before the 1990s. Thomas P. H. Dunlop, U.S. Political Counselor in Belgrade (1978–1982), recalls the atmosphere in the 1970s and 1980s when the Yugoslav government responded only “to give minimal lip service to negotiations and then hit them [the Albanians] just as hard as they could” (Foreign Affairs Oral History 2011). Dunlop adds, “[a]s I see it, the only way that it can change now is for the lid to blow off and a bloodcurdling, Ruanda-like pogrom to occur” (Foreign Affairs Oral History 2011). Ambassador Rudolf V. Perina, U.S. Chief of Mission in Belgrade (1993–1996), echoed a similar sentiment by the mid-1990s: “the fact that it had not blown up, however, invariably led to it being relegated to the back burner” (Foreign Affairs Oral History 2014). Fast forward to the beginning of 1999, the Kosovo Crisis prompted a U.S.-led military intervention. ² The international community has labeled the Kosovo Crisis as a robust case of humanitarian intervention, primarily driven by humanitarian need and, perhaps secondarily, by underlying national interests. But these standard intervention explanations present many puzzles, as discussed below.

In this article, I introduce two critical conditions that made Kosovo a prime candidate for Western intervention by 1999 and not sooner. First, Kosovo had proven to Western audiences that it was experiencing ethnic cleansing, not another endless ethnic civil war between equally culpable parties. Once this “favorable” conflict perception had been set in place, it interlinked with Western values of liberalism, the norm of protecting innocents, and institutions of collective security. Now the lack of response to the Kosovo Crisis paralleled the failure of Western ideals and principles. Second, since Kosovo unraveled at the edge of the Western neighborhood, the one-sided violence (often labeled as impending genocide) activated the value-based missions of Western institutions. Ultimately, such interactions may be generalizable to past and future patterns of humanitarian military intervention and non-intervention.

The evidence to test this argument originates from declassified NATO archives, press releases, conferences, and formal speeches, supplemented by U.S. Department of State and Congressional archives. Moreover, this article is one of the first to quantitatively analyze the transformation in conflict perceptions before and during the Kosovo Crisis. It does so by searching the New York Times historical archives for key thematic terms related to the Kosovo Crisis, such as “ethnic cleansing,” “massacres,” “mass killings,” and “genocide,” versus “civil war,” “terrorists/separatists,” and “ethnic conflict.”

This article proceeds in five parts. The next section introduces the study's theoretical framework, followed by the methods section. I then trace the limitations of standard explanations of the NATO intervention in Kosovo. At the same time, the discussion offers an archival and media analysis of conflict perceptions as a compelling factor for NATO's intervention in Kosovo. The article concludes with theoretical and policy implications for the selectivity gap of humanitarian military interventions beyond Kosovo.

Theoretical Approach

While underexplored in the political science literature, the narratives that political elites perpetuate during times of crisis may impact the nature of third-party responses. Moreover, international responses to a distant, foreign conflict often depend on the strategic manipulation of identity (Saideman 2002). As Kaufman (2015) argues, the complexities of identity-based civil wars, conflicts driven by ethnic, nationalist, or other identity-based grievances, may prompt unique dimensions for outside actors. Yet few academic studies directly integrate these perceptions-based effects within the debates on humanitarian intervention.

In this article, I consider that international audiences can perceive foreign, violent crises as outcomes of civil wars perpetuated by primordial, ancient hatreds or as outcomes of systemic international aggression against one innocent party, altering responses to the crisis. Campbell (1998) touches upon these types of conflict perceptions when describing the international response to the Bosnian conflict. He mentions that Serbia's campaign of mass violence in Srebrenica in 1995 was “ethnic cleansing” or “genocide” in the eyes of the Bosnian Muslims. But for the Serbs, the violence was perceived as an expected component of regular battle during a civil war. Depending on their conflict perceptions of the Bosnian violence, outside parties diverged on what to do in response.

Most global political elites and analysts at the time understood the Bosnian conflict as an identity-based civil war. Thus, the international community assigned equal culpability to all sides of the civil war (Garton 1995; Wheen 1997; Campbell 1998, 52–53). The understanding of the conflict as one of a complicated, inevitable, centuries-old ethnic civil war also implied low chances of third-party success in attaining security and liberalization in Bosnia. Instead, this conflict perception allowed the international community to look the other way since a “civil war” is deemed a state's internal business. Had the international community perceived Bosnia as a case of international aggression, it may have mounted a faster, more robust response due to a sense of responsibility to stop the violence and a perceived higher likelihood of success.

As the case of Bosnia shows, international perceptions about an internal conflict can alter foreign policy options. The perspective that Campbell (1998) applied to Bosnia can also apply to other internal crises, making conflict perceptions an important variable when discussing Western actors’ selectivity of humanitarian military interventions. In fact, in a recent statistical analysis of all cases of intrastate conflict (1989–2014) matched to humanitarian intervention and non-intervention, I have shown that conflict perception, from perceptions of civil war to systematic killings, significantly alter the probability of intervention, and often more so than traditional factors of national interest or human rights (Kushi 2022). I now employ a mixed-methods analysis to process-trace one precedent-setting case, the Kosovo Crisis, and examine the following hypothesis:

The figure below illustrates the hypothesized relationship between conflict perceptions and humanitarian military interventions that apply to cases beyond the Kosovo Crisis. (Figure 1).

OPEN IN VIEWER

Figure 1.

Pathway to Humanitarian Military Intervention.

If this hypothesis is supported, I would expect to see the following patterns within the process of tracing the Kosovo case. After Kosovo meets the threshold condition of a humanitarian crisis, most international political actors will initially perceive the Kosovo Crisis as another Balkan civil war between equally guilty parties, with little Western intervention under consideration. Months before the 1999 NATO military intervention, however, conflict perceptions shifted to that of systematic killings. This will prompt immediate, new discussions of military intervention as a NATO, U.S., and E.U. policy choice, ultimately prompting a humanitarian military intervention in Kosovo.

I search for such a trajectory by exploring archival dialogues and measuring media narratives of conflict perceptions in Kosovo. I also examine several alternative hypotheses as part of this process-tracing, including the impact of national interests on NATO's intervention in Kosovo and the result of human rights abuses on the likelihood of intervention. For the national interest argument, I would expect the United States and European countries to intervene in Kosovo if their national interest calculations changed from 1998 to1999 to favor intervention. The humanitarian argument would expect the United States and the European policy communities to respond directly to increasing human rights abuses in Kosovo with progressively heavier measures, including humanitarian military intervention.

Methods

First, I process-trace how conflict perceptions altered the trajectory of the Kosovo Crisis and third-party responses via archival evidence. Then, I quantify New York Times coverage of Kosovo in the years before the intervention, searching for keywords that represent distinct Western conflict perceptions, as revealed through the qualitative analysis. Last, I present graphical trends of the relative number of media narratives throughout the months progressing to the NATO military intervention.

Standard Explanations of The Kosovo Intervention

Some scholars interpret trends of humanitarian military intervention as an example of growing normative, multilateral standards in international society in which human rights protection becomes a part of state security and interests (Wheeler 2001; Finnemore 2003). They find that democratic values, economic ties, human rights norms, similar ideologies, or previous involvements drive humanitarian military intervention patterns, especially in the case of U.S. interventions (Choi 2013; Lyon and Dolan 2007; Mullenbach and Matthews 2008; Finnemore 2003). For instance, Choi and James’ (2016) cross-national, time-series analysis shows that the United States is likely to undertake military campaigns for humanitarian reasons rather than for its security interests, such as democratization or terrorism reduction. Despite such a humanitarian intent, Choi and James (2017) also show in a later study that all forms of U.S. foreign policy intervention, including humanitarian military interventions, do not relate to improved levels of human rights in the target state.

Therefore, many remain skeptical of the humanitarian drivers, arguing that powerful states merely coopt the humanitarian label to pursue national interests within their spheres of influence (Mearsheimer 2001 1973; De Jonge Oudraat 1996; Fordham 2008; Dunne and Schmidt 2011). While both perspectives explain some patterns of humanitarian military intervention, they do not fully resolve the puzzle of intervention selectivity: why did the United States intervene to protect human rights in Kosovo but not in Darfur, South Sudan, Rwanda, or even Bosnia (until post-genocide) where human rights violations were arguably more widespread and severe? Why intervene to end famine in Somalia in 1993 in the absence of traditional national interests?

Is it Human Rights or National Interests?

Kosovo had suffered state-sponsored human rights abuses for years before the 1999 NATO military intervention. Both human rights and interests-based narratives in Kosovo are dimmed by this evidence that U.S. and European elites were initially unsupportive of intervening, even after constant updates on mass human rights violations. As the Political Terror Scale in Figure 1A in the Appendix reveals, no sudden increased intensity of the regional humanitarian crisis might have prompted intervention in Kosovo in 1999 rather than years prior.

As Victor C. Comras, part of the Kosovo War Sanctions Team in 1998, recalls in a 2002 interview, “We appeared to be less and less interested in events in Kosovo and more and more eager to get out of the Balkans simply.” But that was not due to a lack of knowledge of what the Serbs were doing. He continues, “The Serbian forces shot first and asked questions later. More and more Albanian villages went up in smoke… The Contact group was still concentrating on Bosnia and was slow to pick up on the growing Kosovo crisis” (Foreign Affairs Oral History 2011).

In March 1998, when General Wesley Clark warned the Pentagon of trouble brewing in Kosovo, Vice Chairman of the Joint Chiefs of Staff Joseph Ralston responded with, “Look, Wes, we’ve got a lot on our plates back here… We can’t deal with any more problems” (Clark 2001, 109). Even by the summer of 1998, when Alexander Vershbow, U.S. Ambassador to NATO, pushed for a Dayton-style solution to Kosovo (with U.S. troops as part of an international peacekeeping force), he had no support in the White House (Harris 2005, 362). Even by December 1998, U.S. Major General Dennis Reimer, the Army Chief of Staff, responded to another warning from Clark that there may be an impending war in Kosovo with, “But we don’t want to fight there” (Clark 2001, 165). Most U.S. policy makers were reluctant to send the smallest of peacekeeping forces to Kosovo in 1998, even against the backdrop of the Srebrenica genocide in Bosnia and Herzegovina a few years prior (Frontline 2000e).

While most standard explanations of the Kosovo intervention rely on the existence of vital U.S. national interests, such a narrative was crafted on the eve of intervention and post-crisis. It did not exist during the ongoing intrastate conflict in the region. On the first night of Operation Allied Force, President Clinton argued that “ending this tragedy is a moral imperative,” he also emphasized that it was “also important to America's national interest” (Clinton 1999a). A “prosperous, secure, undivided, and free” Europe was essential to the U.S.'s prosperity (Clinton 1999a, 453). NATO's credibility was also at stake: “Imagine what would happen if we and our allies instead decided just to look the other way, as these people were massacred on NATO's doorstep. That would discredit NATO, the cornerstone on which our security has rested for 50 years now” (Clinton 1999a, 451–452). This rhetoric, uttered on the day of the intervention, colors much of the perceived causes of the intervention while intentionally ignoring the dynamics of years prior.

But what changed between 1998 and 1999 that made Kosovo a strategic target for the United States and its Western partners? Some say that having initiated diplomacy and military threats, NATO members could have lost confidence in NATO's capability to guarantee regional security if Milošević's actions in Kosovo were ignored (Roberts 1999). ³ However, this discussion lacks an important temporal context. Merely a few months before the intervention, the Kosovo conflict was understood as a civil war between terrorists and government agents. Then, NATO did not want to aid terrorists, despite previous Western diplomatic engagements and threats in both Bosnia and Kosovo. As detailed in the following sections, these standard accounts skip over an important step that suddenly put NATO's credibility on the line in Kosovo and led to post-intervention discussions of vital interests where none previously existed. In other words, NATO's credibility was a key driver of Kosovo's intervention, but only in the aftermath of changing conflict perceptions.

Before late 1998, the U.S. government did not appear to hold geopolitically solid policy incentives to act in Kosovo. Instead, it had placed all its Balkans resources on the Bosnia settlement. As neorealist John Mearsheimer (quoted in Miller 1999) stated at the time, “Most realists are offshore balancers, and most of us do not see any such threat emerging as a result of Kosovo.” In the most robust realist circles, the eventual NATO intervention in Kosovo was understood as a fundamental rejection of state principles and an example of imprudent foreign policy, given the lack of security interests and exacerbated hostilities with Russia and China (Mandelbaum 1999). Recalling the United States’ recent inactions in Rwanda and Krajina, Croatia, Charles Krauthammer (1999a) denounced the intervention as “righteous self-delusion” and “impossibly moralistic and universal.” To many, instead of preventing further conflict, the U.S. intervention in Kosovo internationalized a small civil war without stopping the violence or refugee flows (Krauthammer 1999b).

In addition to preserving NATO's credibility, the other standard, interests-based argument is that a predicted refugee spill-over motivated Western powers to intervene in Kosovo. According to the U.N. High Commissioner for Refugees (UNHCR), after the rejection of Kosovo's autonomous status in 1989, about 350,000 ethnic Albanians left Kosovo for Western Europe. With the beginning of the conflict in 1998, another 100,000 people fled to nearby countries such as Albania, Montenegro, and Macedonia (Del Mundo and Wilkinson 1999, 10–11). Since recipient countries were already experiencing crises, the United States and its allies worried that further refugee flows could destabilize the whole of Europe. But then again, the refugee problem stemming from the Balkans existed years before discussions of military intervention. What changed?

On the other hand, had the primary motivation behind the Kosovo intervention been humanitarian, not geopolitical, the United States and its allies would have been expected to protect as many lives as possible through their actions. However, in a one-on-one briefing with Secretary of State Albright, General Clark told her that the Serbs would undoubtedly attack the civilian population if NATO launched airstrikes. When asked whether he thought the administration should go through with the air strikes, he replied, “Yes, we have to. We put NATO's credibility on the line… There's no real alternative now” (Clark 2001, 171). As U.S. political elites had predicted, during the 11 weeks of bombardment until June 10, an estimated 10,000 people died violently in Kosovo (Mandelbaum 1999, 3; Human Rights Watch 2002, 16). Most of them were ethnic Albanian civilians murdered by Serbs, but an estimated 488 to 527 were casualties of the NATO aerial attacks (Mandelbaum 1999, 5).

The standard arguments for either national interests or humanitarian intent are not without merit, but each presents inconsistent timelines. Neither of these traditional views can explain why the Kosovo Crisis rose to the top of the Western agenda by 1999, not years before. Why did the humanitarian context suddenly become politically important to Western elites, media, and leaders, when it was muted years prior? For the realists, why did geopolitical interests manifest in Kosovo after repeated discussions of Kosovo as an unimportant periphery? Why are most of the realist arguments founded upon post-conflict justifications?

Moreover, isolating interests from human rights dimensions becomes problematic in Kosovo's timeline of events. Clinton's (1999a) speech in 1999, interpreted through the lens of humanitarianism by some observers and geopolitical interests by others, hints at a deeper story of intervention. The president repeats phrases like “defenseless people”; he discusses historical lessons and parallels to Bosnia. Every promise relates solely to the European continent and the preservation of Western values. What connects the humanitarian arguments to the realist ones in Kosovo are the interactions between understanding the conflict and the Western ideals and missions that such a conflict perception activated. Kosovo received third-party support only once it tied its suffering to the Western narrative. Then, inaction would have again marked the West's failure to protect its neighbors from potential genocide. After the West's humanitarian failures in Bosnia and Rwanda, Kosovo also benefitted from faster “lessons learned” shifts in narratives. As the following sections will discuss, NATO's credibility only became relevant in Kosovo once the internal violence was recognized as ethnic cleansing or impending genocide, in the backdrop of growing elite and media analogies to past genocides.

Transformed Conflict Perceptions during the Kosovo Crisis

On the surface of the dialogue introduced above, the military intervention in Kosovo held security and humanitarian dimensions, which made it more likely to occur. Still, neither one of these dimensions explains the Kosovo intervention or the others that followed. Kosovo was not a critically strategic region for the United States at the beginning of the intrastate conflict, nor was Kosovo's humanitarian plight greater than or equal in intensity to other past and concurrent internal wars. The Kosovo Crisis became worthy of consideration through the lenses of security and humanitarianism only once conflict perceptions shifted. This opened Kosovo to Western sympathies and finally to humanitarian intervention. After all, the intrastate conflict within Kosovo had always occurred and intensified next to the Western neighborhood. Still, it was not until late 1998 that the conflict perception underwent a dramatic transformation that prompted NATO to assign blame and envelop Kosovo in Western protection.

Even by December 1998, U.S. political elites opposed any one-sided involvement. “Apparently, our Yugoslav team had already decided on a somewhat different course. They did not believe that we should appear to choose sides in the Kosovo crisis,” said Comras (Foreign Affairs Oral History (FAOH) 2002). Western actors first needed to re-label the conflict as systematic killings, not the typical ethnic civil war.

From Terrorists to Innocent Civilians

Long before Kosovo had made headlines, when violence raged within Bosnia, most Western actors perceived the Bosnian conflict as another Balkan civil war, defined by three equally culpable ethnic sides battling for ancient territorial claims. As Campbell (1998) has examined in nuance, Western actors hesitated to intervene in Bosnia until the narrative changed from the ethnoreligious civil war to systematic killings of civilians and ongoing genocide. Until 1998, even while the violence in Kosovo escalated, the Western approach to Milošević wavered between denouncing his atrocities in the region to perceiving him as the main guarantor of peace in the Balkans (given his negotiating role in Bosnia's Dayton Conference in 1995). In fact, in 1998, Robert Gelbard, then the United States’ special envoy to the Balkans, publicly declared the Kosovo Liberation Army (KLA) as “without any questions, a terrorist group” to distance Milošević from blame (Judah 2000, 138). For conflict perceptions in Kosovo to change in favor of military intervention, Milošević had to be seen as the primary aggressor.

This shift required a re-interpretation of the KLA, the separatist paramilitary of the ethnic Albanians in Kosovo. Until 1998, the West had labeled the KLA as a terrorist organization per Serbian leadership standards, perceiving the Kosovo Crisis as another case of civil war. Such narratives of equally guilty parties engulfed in another “ethnic” or “religious” civil war were the norm in Western elite dialogue and press briefings before 1999.

Even after the Serbian government had repressed peaceful demonstrations in Kosovo, in March of 1998, the North Atlantic Council's statement affirmed the equally culpable narrative, calling “on all sides to take immediate steps to reduce the tensions” (North Atlantic Treaty Organization 1998a). Similar statements followed, condemning “the unacceptable use of force by Serbian police against peaceful demonstrators as well as terrorist actions by the Kosovo Liberation Army or any other group” (North Atlantic Treaty Organization 1998b, c). The statements frequently used the term “Kosovar extremists.”

As this narrative persisted in the Western media and elite dialogues, the Kosovo humanitarian crisis remained untouched. By June 1998, the only third-party responses amounted to symbolic air exercises in Albania and Macedonia to demonstrate NATO power, observer missions in neighboring countries, and more talks of funding UN and OSCE monitoring missions in Serbia to promote reciprocal political dialogue. There were no talks of moving beyond lukewarm diplomatic stances that promoted third-party neutrality.

This diplomatic route produced a short-lived ceasefire in 1998, promoted by the several UN Security Council Resolutions discussed above. During the truce, the civil war narrative continued to dominate Western discussions of the crisis, with NATO and Western state officials holding to the same neutral chains of blame, as evidenced in The Statement on Kosovo at the Meeting of the North Atlantic Council in the Foreign Ministers Session of the National Atlantic Treaty Organization (North Atlantic Treaty Organization 1998d). Due to Serbian military advances, the ceasefire's failure prompted a NATO air verification mission but no option of direct intervention. Daalder explained in an interview for Frontline (2000b) that the West was at first wary of serving as the “airforce of the KLA … a bunch of thugs …” against an established, sovereign state. “we’ve got to find a political solution—not because Milošević might otherwise do bad things—but otherwise, the KLA will win.”

By the end of 1998, the understanding of Kosovo as a civil war between two equally culpable factions began to fade, paralleling changes in media attention to Kosovar refugees and a direct elite witness to the civilian massacres, as will be detailed below. In the aftermath of the Western failure to prevent the Srebrenica genocide in Bosnia, this shift in perceptions of the Kosovo crisis evolved more quickly. Within the year, the dominant narrative of civil war transformed into one of ethnic cleansing, followed by potential genocide in the heart of Europe. Secretary Albright's spokesperson James Rubin put it simply in 1999, “In order to move towards military action, it has to be clear that the Serbs were responsible” (Rubin 1999; Frontline 2000c). Even President Clinton recognized the benefits of such a shift in perception in the backdrop of the Bosnian tragedy worsened by Western passivity in the face of “ancient hatreds.” President Clinton would eventually come to criticize the primary narrative on Kosovo and those who excused the violence because “somehow they were intrinsically made to murder one another” (Clinton 1999b).

The Raçak Massacre as the Turning Point

Though Kosovo benefitted from the “lessons learned” following the failure to prevent the Bosnian and Rwandan “holocausts,” Kosovo still needed the smoking gun (Bates 2009; Steinweis 2005). After the Raçak massacre in January 1999, when Serbian forces murdered 45 ethnic Albanians, the media reports and elites alike latched on to a one-sided genocide narrative for good (Organization for Security and Cooperation in Europe 1999; Frontline 2000d). Ambassador William Walker, serving as the head of the Kosovo Verification Mission, personally witnessed the aftermath of the Raçak massacre and held an anger-fueled press conference. In this public briefing, he declared that Serbian security forces had committed crimes against humanity against ethnic Albanian civilians (Frontline 2000f).

Before this massacre, U.S. and NATO elites had relied on empty threats against Milošević. Until the day of the Raçak massacre, Western actors negotiated on Kosovo with the unified assumption that the use of force would not be required, nor would it be warranted, given the nature of the conflict. The West did not want to act as the de-facto military of the KLA, a group of thugs perceived as equally as violent and guilty as their Serbian counterparts. Daalder explains in his interview in Frontline (2000b):

Serb forces are engaged in yet another massacre of 45 people in Raçak. That's on the front page of the newspaper. The next day, the OSCE monitors are on the scene and report what happened. The head of the OSCE monitors says this is a crime against humanity… Within four days, there was an immediate agreement in the White House that the same option Madeleine Albright put on the table four days earlier, which was not acceptable, now becomes acceptable.

“My Mindset Is Munich”; The Role of Holocaust Analogies

In Bosnia, the West had remained passive until many foreign journalists reported from “concentration camps” within Bosnia itself, activating a wide range of Holocaust analogies (Akrivoulis 2015). In the Western collective identity at the time, Bosnia was the first instance of non-prevented genocide on European soil since the Holocaust (Cushman and Mestrovic 1996). But by 1999, following Ambassador Walker's declarations against the Serbian state, media reports from Kosovar refugee camps in nearby countries predicted another Bosnia.

The outcome of the Rambouillet Conference in March of 1999 solidified this dynamic—with the ethnic Albanian factions signing on to a peace deal with several concessions and Serbian factions refusing to discuss key terms of the agreements. This portrayed ethnic Albanians as the peace-seekers while Serbia continued to mobilize its military around Kosovo. So, the Holocaust imagery only grew. Madeleine Albright remarked of Kosovo in 1999, “My mindset is Munich”—hinting at the legacy of European appeasement of ruthless, genocidal leaders over 60 years ago in the late 1930s (Dobbs 1999, 34). This Holocaust imagery would also frequently find its way into Clinton's (1999b) speeches on the Kosovo conflict, founded upon the main lessons learned in Bosnia, “the genocide in the heart of Europe”:

they’ve now started rolling from village to village, predominantly in north central Kosovo, shelling civilians and torching their homes so they can’t return. In several villages, Serbian police have dragged the male members of Kosovar families from their homes, lined up fathers with sons, and shot them in cold blood. (Clinton 1999b)

U.S. Ambassador David Scheffer's statement showcases this dynamic and the selection of Kosovo for humanitarian intervention while muting other international calls for help. “There are hideous crime scenes elsewhere recently—Sierra Leone, the Democratic Republic of the Congo, and Sudan come to mind—but Kosovo represents a government-planned campaign to eliminate, either through forced deportation or killing, most of an ethnic population from its home” (NATO HQ 1999a). The analogies to past genocides helped place the blame of the Kosovo conflict solely on Milošević's shoulders, revoking any gradations of victimhood from the Serbian state. As Air Commodore David Wilby stated for NATO in April 1999, “All too clearly, we have seen who the victim is and who the victimizer is, and he [Milošević] remains the person who bears the full responsibility for the human tragedy that we have witnessed in Kosovo” (NATO HQ 1999d).

By 1999, analogies to genocide were circulating within U.S. Congress as well. Members discussed “a level of atrocity not seen since World War II… in the heart of Europe” (145 Cong. Rec. H1185, H1190, H1199, H1201 1999). In Senate discussions, many equated Milošević to Hitler (145 Cong. Rec. S2668 1999; 145 Cong. Rec. S4141 1999). Multiple times, it was claimed that Hitler and Stalin served as Milošević's “role models” (145 Cong. Rec. H2424–2425 1999). Then-Senator Joe Biden (D-DE) often referred to the Raçak massacre as an echo of Hitler's campaigns (145 Cong. Rec. S3050 1999; 145 Cong. Rec. S3048 1999). These new framings of the Kosovo Crisis through the lens of genocide increased U.S. domestic support for military intervention when none existed a year prior (see 145 Cong. Rec. H1667, 1668 1999; 145 Cong. Rec. E682 1999; 145 Cong. Rec. H2231–2238 1999; 145 Cong. Rec. S4518, 4545, 4531, 1999; 145 Cong. Rec. S4536, 4567 1999; 145 Cong. Rec. H2411, 2436, 2448–2450 1999).

A shift in conflict perception enveloped all talks of Kosovo within NATO, the United States, the United Kingdom, and the NATO alliance. Official press briefings would now repeat how “the grave humanitarian crisis was brought on by the actions of President Milošević's forces in Kosovo,” without mention of Kosovo separatists or terrorist non-state actors (North Atlantic Treaty Organization 1999d). Such a narrative allowed NATO to join in humanitarian relief efforts in Kosovo, coordinating with the OSCE, UNHCR, and the Council of Europe. By the end of January 1999, this shift in perceptions would prompt threats of airstrikes in FRY if political negotiations did not yield a lasting peace between the Serbian state and its ethnic Albanian population. By the end of March 1999, as the months passed without the cessation of violence, this new narrative permitted a consensus between Western countries and institutions on the urgency of airstrikes against FRY (North Atlantic Treaty Organization 1999e).

As General Wesley Clark summarized in an interview for Frontline (2000a), “People had finally come to terms with the fact that it was not just an ordinary reaction to some domestic terrorism that was taking place—they were seeing the first stages of unfolding drama that was likely to culminate in ethnic cleansing.” As the year progressed, NATO repeatedly claimed the justification of ethnic cleansing for its military involvement (Solana 1999).

As President Clinton also affirmed in 1999, the Western community no longer saw the intrastate conflict in Kosovo as a petty ethnic or religious civil war fought by equally guilty parties. By then, they perceived the violence in Kosovo as an act of aggression by the Serbian authorities against a “largely defenseless people” (Clinton 1999a). Repeated meetings with President Milošević further aided this narrative, as Milošević's talk of ethnic cleansing the “bandit” Albanians, blocking international investigators from sites of violence, and breaking diplomatic promises to NATO became the norm of such encounters (Frontline 2000a). Non-intervention in Kosovo had now become synonymous with Western complicity in campaigns of systematic killing. Before the military mission had begun, the rhetoric had swung fully in the direction of stopping campaigns of systematic violence against ethnic Albanian civilians. Western counterparts had forgotten their previous talks of “terrorist KLA” members. The official NATO tagline, paralleling all elite discussions and press briefings for the next several months, was to stop “a deliberate policy of oppression, ethnic cleansing, and violence pursued by the Belgrade regime” (North Atlantic Treaty Organization1999c).

In sum, the historical perceptions of the Kosovo conflict changed entirely in the months before the military intervention, turning “terrorists” and “extremists” into innocent Kosovar civilians fighting against a terrorist state. This perception came from more significant learning from the media and political elites on the origins of the Kosovo Crisis and Serbian military action, especially after Ambassador Walker's conference in January 1999. The Kosovar Albanians were now justified in the eyes of the Western elites in their rebellion against the Serbian government, hence deserving intervention. Key to such shifts was the changing linguistic arsenals that political elites used in context with one another, the media, and international organizations. They moved away from discussions of civil war, ethnic hatreds, separatists, and terrorists to narratives centered on mass graves, executions, civilian exodus, concentration camps, helpless refugees, and scorched villages. NATO briefings consistently mentioned the displacement of ethnic Albanian families, the revocation of passports and identities by Serbian authorities, the selling of property in exchange for “train tickets to nowhere,” and the use of human shields by Serbians in battles. The narrative shifted away from terrorist ethnic Albanians to hotels full of injured ethnic Albanian women, disappearing military-age men, and children massacred in front of their parents.

By March of 1999, NATO Spokesman Jamie Shea was quoting statistics on refugees and civilian killings on a tri-daily basis (NATO HQ 1999d). For example,

800,000 Kosovar Albanians have fled Kosovo since March 1998, 650,000 are internally displaced … at least 4,000 victims of summary executions are reported since the beginning of the year, nearly 1.5 million Kosovar Albanians—or 90 percent of the population—have been expelled from their homes. We have reported mass executions in 65 towns and villages and mass graves in at least seven locations.

This perspective only grew internationally post-intervention as UN Secretary-General Kofi Annan, in his keynote speech to the 54^th session of the UN General Assembly in September 1999, tempered his argument on Security Council-approved use of force in humanitarian endeavors to legitimize non-UN sanctioned intervention in Kosovo. He posed the following question: “If, in those dark days and hours leading up to the genocide [in Rwanda], a coalition of states had been prepared to act in defense of the Tutsi population, but did not receive prompt Council authorization, should such a coalition have stood aside and allowed the horror to unfold?” (Annan 1999). This question only arose from the new perception that the Kosovo conflict was like the Rwandan conflict—a crisis characterized by systematic killing from one side and massive civilian deaths on the other. Until such a conflict perception of “systematic violations of human rights” had monopolized the Kosovo Crisis narrative in both media and elite political circles, intervention remained muted across most relevant arenas.

Civil War vs. Systematic Killings in the Media

In this section, I present the media trajectory of the Kosovo Crisis across Western audiences. I measure the total and relative article counts in the New York Times archives as related to different narrations of the crisis, spanning the spectrum of civil war to systematic killings. The Appendix provides public opinion data to parallel the timeline of this media exploration. It is important to note that this section only explores the temporal associations between conflict perceptions and the NATO humanitarian military intervention in Kosovo. It does not seek to explore the origins of conflict perceptions but instead evaluates whether changing perceptions can significantly impact foreign policy choices.

Figure 2 introduces the central conflict perception framings relative to one another and the total number of monthly mentions of “Kosovo” between January 1998 and July 1999. Figure 3 offers relative trends of the specific narrative framings across the months. As revealed below, the Kosovo Crisis did not enter the Western consciousness and media coverage in a meaningful way until March 1998. Before 1998, the New York Times practically ignored Kosovo despite the ongoing crisis since the early 1990s. For example, when I searched the New York Times archives for September 1997, I found no mention of Kosovo in the coverage. As the months progressed that year, the New York Times mentioned Kosovo an average of one to three times monthly until 1998. Therefore, I began the trend analysis in January 1998.

OPEN IN VIEWER

Figure 2.

All Kosovo Framings by Month, Jan. 1998–July 99, NYT.

OPEN IN VIEWER

Figure 3.

All Substantive Framings by Month, Jan. 1998–July 99, NYT.

Throughout 1998, Figure 3 shows small spikes in the coverage of the Kosovo Crisis—in March 1998, June 1998, and October 1998. All these increases in range parallel significant developments regarding human rights abuses, detailed in the Appendix. The graph also shows that Kosovo became a priority for Western media and audiences in January and February of 1999 when the Raçak massacre hit Western airwaves through the anger of Ambassador William Walker. A week after the Raçak massacre, articles entitled “NATO Threatens Military Action to Stem the Violence in Kosovo” and “Kosovo Massacre Is Called Revenge” raged throughout the New York Times and grew in numbers as the weeks progressed (Whitney 1999; Perlez 1999). Kosovo then spiked in interest before and during the NATO military intervention from March to June 1999. It was not until January and February of 1999 that the Ethnic Cleansing narrative began outpacing the Civil War narrative for the Kosovo violence, as will become more apparent in the following graphics.

In Figure 4, the Civil War narratives appear as line graphs, while the systematic killing narratives appear as bar graphs. The Civil War narrative dominated New York Times article counts on Kosovo from January 1998 until January 1999. This was due to the Massacre/Mass Killings narrative spiking in the aftermath of the Raçak massacre. Steadily, the other systematic killing narratives grew in the following months. By April 1999, the framings of Massacre, Ethnic Cleansing, and Genocide had risen above the Civil War narrative framing of Kosovo.

OPEN IN VIEWER

Figure 4.

Civil War versus Systematic Killings Framings in Kosovo, Article Count, Jan. 1998–July 99, NYT.

The graph in Figure 5 compares the percentage of Civil War narratives on Kosovo to the rate of Systematic Killings per month in New York Times coverage. Once again, the systematic killing narrative spikes during the time of the Raçak massacre, where it overtakes the civil war narrative for good. Before 1999, however, the civil war narrative dominated the media coverage of the Kosovo Crisis, except for the slight blip in March of 1998.

OPEN IN VIEWER

Figure 5.

Percent Civil War Narratives Compared to Systematic Killings Narratives in Kosovo, Jan. 98–July 99, NYT.

Figure 6 presents some additional dimensions of the Civil War narrative that are not captured by other measures—that of the separate Ethnic War/Conflict framing in which ethnic Albanian rebels and separatists are pitted against the Serbian state equally culpably. While the Ethnic War narrative declined during and after the Raçak massacre in January, the Ethnic Conflict narrative dramatically increased. This represents a curious finding. Do the words “war” and “conflict” denote different levels of the legitimacy of violence or culpability across actors in this case? Indeed, war may imply an expected, traditional type of symmetrical or institutionalization of violence, whereas conflict is vague in its judgment of blame or structures of power. Figure 6 also shows a gradual decline in narratives of ethnic Albanian rebels since January 1999, further confirming the dynamics implicated in the Ethnic Conflict versus Ethnic War patterns.

OPEN IN VIEWER

Figure 6.

Percentage Ethnic War-related Framings in Kosovo relative to all Framings, Jan. 98–July 99, NYT.

Figure 7 presents the indexed article count measures per narrative for a more straightforward comparative interpretation. The index value is 100, pegged to March 1998 numbers. The measurement below then reflects the percent change over time of a particular narrative while keeping the starting points constant. I have selected March 1998 as the baseline measure because it is at the beginning of the series, following a zero value (which cannot be adequately indexed), and because March 1998 marked the first moment Kosovo appeared on the New York Times agenda in a meaningful way. Thus, the changes over time in the following graph measure the relative narratives since the West learned of the first big humanitarian disaster in Kosovo. The graph shows that all measures denoting systematic killings in Kosovo jumped dramatically in article count after January 1999. The most interesting takeaway is that the Genocide article counts increase by 100 percent between January and February and then by 2,000 percent by March, while all other narratives in this group follow suit shortly afterward.

OPEN IN VIEWER

Figure 7.

Index of Narrative Changes (March 1998  =  100), Jan. 98–July 99, NYT.

Conclusion

As confirmed by the archival evidence and media content analysis, the shift in the narrative after January 1999 transformed Kosovo into a case of ethnic cleansing or even genocide. This conflict perception opened the way for a robust regional reaction to the crisis. Now the West had to fend against acts of potential genocide on its doorstep, not simply tolerate another Balkan ethnic war. Inaction, in this case, amounted to the non-enforcement of liberal Western values and the futility of Western institutions.

The exact detail of how Kosovo's narrative changed from civil war to ethnic cleansing in 1999 is an important question that naturally follows. Still, it is not within the scope of this article. Based on the evidence following the Raçak massacre, some plausible hypotheses are the impact of Western regionality and Western political elite attention. Western political elites already on the ground, with previous experiences in Bosnia, may have facilitated the narrative shift. Previous NATO-led operations in the region made it easier for Kosovo to move away from the typical civil war narration and receive international support. The complex interactions between Western states, organizations, elites, and various local actors on the ground solidified a favorable conflict perception of “systematic killings” for the Kosovo Crisis. The perception shift was sudden, mainly because media and leadership interest in Kosovo was also sudden, relying on the reports of acute massacres to spark humanitarian and security interests. But once the massacres in Kosovo entered the Western consciousness, it was not long before analogies to Bosnia, the Holocaust, and past Western mistakes came to the rescue—increasing the probability of third-party military intervention.

Kosovo's location within the “heart of Europe” and its role as a test of NATO's post-Cold War functionalities may have also spurred greater international support. For instance, had Kosovo been located outside the sphere of Western influence, without the institutional benefits provided by NATO, the EU, and other regional actors, its favorable conflict perceptions may not have altered the intervention pathway. But then again, had the Kosovo Crisis remained within the narrative of ethnic civil war, it may have taken the West many more years to intervene with military force, or intervention would have never come.

This article encourages additional research on the interactions between conflict perceptions, elite mobilizations, and media attention in other cases of humanitarian military intervention or non-intervention. The research presented here does not aim to define causal relationships between conflict perceptions and humanitarian intervention as a policy choice. Instead, it seeks to show that changing conflict perceptions do impact the foreign policy-making process and intervention trajectories in a meaningful way. Future research may wish to examine the various mechanisms that can impact conflict perceptions, including media attention, the actions of Western political elites responsible for placing military intervention on the table, and a range of strategic considerations. An essential set of questions to pose are: do conflict perceptions change due to deliberate, strategic actions by Western or local elites, or do they primarily respond to new events and media attention on the ground? Why do conflict perceptions change to favor intervention in certain regions of the world, such as Kosovo, and not others, such as Darfur?

The NATO intervention in Libya and the international non-intervention in Darfur would prove extreme cases to explore in the future, especially as they would draw attention to a significant regional dimension beyond the West. Changing conflict perceptions may bring about a range of secondary interests to third-party actors responsible for critical foreign policy decisions. Beyond delving into standard accounts of national interests, economic benefits, and humanitarian rhetoric, we must also understand how the power of perception may explain decision making across global crises and how perceptions are formed and transformed in the international and domestic spheres.