Ethnic Nationalism and International Conflict: The Case of Serbia*

by V.P. Gagnon, Jr.

A slightly revised version of this paper was published in International Security, vol.19, no.3 (winter 1994/95), pp.130-166

Does ethnicity affect the international system? What are the causes of violent conflict attributed to ethnic solidarity? Since the collapse of the Soviet Union and the outbreak of war in the Balkans, these questions have seized the attention of international relations scholars and policy makers.1 In the former Yugoslavia, war conducted in the name of ethnic solidarity has destroyed the Yugoslav state, leveled entire cities, and resulted in hundreds of thousands of casualties and millions of refugees.2 It has also brought NATO's first out-of-area actions, the largest United Nations peacekeeping operation in history, and the very real possibility of war spreading to other parts of the Balkans.

Is the Yugoslav case a look into the future of international relations? Are ethnically-mixed regions in the post-Cold War era inevitably the sites of violent conflict that will spill over into the international arena? If so, the only apparent solution would be the creation of ethnically pure states; yet the greatest threats to peace in this century have tended to come from those regions in which partitions along ethnic or religious lines have taken place.3 This apparent contradiction is in fact one of the major challenges to peace and stability in the international system, given the growing number of violent conflicts described and justified in terms of ethnicity, culture, and religion.

Despite the urgency of this issue, theories of international relations have until quite recently not addressed the question of ethnic nationalist conflict. The main challenge is conceptual: how to establish the causal link between ethnic nationalist sentiment and interstate violence.4 Existing approaches tend to assume either that ethnic sentiment itself is the main cause of violent conflict, or that external security concerns lead national decision makers to inflame such sentiment.5 In this paper I argue that such violent conflict is caused not by ethnic sentiments, nor by external security concerns, but rather by the dynamics of within-group conflict.6 The external conflict, although justified and described in terms of relations with other ethnic groups and taking place within that context, has its main goal within the state, among members of the same ethnicity.7

I argue that violent conflict along ethnic cleavages is provoked by elites in order to create a domestic political context where ethnicity is the only politically relevant identity. It thereby constructs the individual interest of the broader population in terms of the threat to the community defined in ethnic terms. Such a strategy is a response by ruling elites to shifts in the structure of domestic political and economic power: by constructing individual interest in terms of the threat to the group, endangered elites can fend off domestic challengers who seek to mobilize the population against the status quo, and can better position themselves to deal with future challenges.

The dominant Realist approach in international relations tells us very little about violent conflict along ethnic lines, and cannot explain the Yugoslav case. Focusing on external security concerns, this approach argues that conflictual behavior in the name of ethnic nationalism is a response to external threats to the state (or to the ethnic group).8 The general literature on ethnic conflict likewise uses the "ethnic group" as actor and looks to factors outside the group to explain intergroup conflict.9 But in fact, the Serbian leadership from 1987 onward actively created rather than responded to threats to Serbs, by purposefully provoking and fostering the outbreak of conflict along ethnic lines, especially in regions of Yugoslavia with histories of good interethnic relations.10

Although the Serbian leadership itself has justified its policies in terms of an external security threat to Serbia and Serbs, over the past thirty years a significant part of the Serbian political elite has advocated a very different strategy based on democratic pluralism, peaceful negotiations of political conflict and modernization of the Serbian economy.11 This strategy would probably have been much more successful and much less costly in ensuring the interests of Serbs and Serbia, even if the goal had been an independent, enlarged Serbia.12 It is difficult to argue that an objective security threat exists when even nationalistically-oriented elites in Serbia denounce the war and claim there was no need for it.

Another common explanation for violent conflicts along ethnic lines, particularly for the Yugoslav case, is that ancient ethnic hatreds have burst to the surface.13 But this too is unsupported by the evidence: in fact, Yugoslavia never saw the kind of religious wars seen in Western and Central Europe, and Serbs and Croats never fought before this century;14 intermarriage rates were quite high in those ethnically-mixed regions that saw the worst violence;15 and sociological polling as late as 1989-90 showed high levels of tolerance especially in these mixed regions.16 Although some tensions existed between nationalities and republics, and the forcible repression of overt national sentiment added to the perception on all sides that the existing economic and political system was unjust, the evidence indicates that despite claims to the contrary by nationalist politicians and historians in Serbia and Croatia, "ethnic hatreds" are not the essential, primary cause of the Yugoslav conflict.
In the following sections I lay out a theoretical framework and hypotheses about ethnic nationalist conflict that look to internal dynamics to explain external conflict. I then apply this to the specific case of Yugoslavia, concentrating on five episodes in which elites within the Serbian republic resorted to conflictual strategies described and justified in terms of the interest of the Serbian people.17 In the conclusion I look at how this framework can illuminate other cases, and what it says about strategies for conflict resolution.

Domestic Power and International Conflict: A Theoretical Framework
This section lays out a framework and proposes some hypotheses about the link between ethnicity (and other ideas such as religion, culture, class) and international conflict. It is based on the following premises: first, the domestic arena is of central concern for state decision makers and elites because it is the location of the bases of their power. Ruling elites will thus focus on preserving these domestic bases of power. Second, persuasion is the most effective and least costly means of influence in domestic politics. One particularly effective means of persuasion is to appeal to the interest of politically relevant actors as members of a group. Third, within the domestic arena, appeals for support must be directed to material and nonmaterial values of the relevant target audiences--those actors whose support is necessary to gain and maintain power. Ideas such as ethnicity, religion, culture, and class therefore play a key role as instruments of power and influence, in particular because of their centrality legitimacy and authority.

Finally, conflict over ideas and how they are framed is an essential characteristic of domestic politics, since the result determines the way political arguments can be made, how interests are defined, and the values by which political action must be justified. The challenge for elites is therefore to define the interest of the collective in a way that coincides with their own power interests. In other words, they must express their interests in the "language" of the collective interest.

Hypotheses about Nationalist Conflict and Domestic Politics
Given these premises, the following hypotheses are put forward to identify under what conditions national leaders resort to conflictual policies described and justified in terms of threats to the ethnic nation.

1. If ruling elites face domestic threats to their power, or to the political or economic structure on which that power is based, they will be willing to respond by undertaking policies that are costly to society as a whole, even if the costs are imposed from outside. Behavior vis-à-vis the outside may therefore have its main goal in the domestic arena. If the behavior of political actors is goal-seeking and if they weigh costs and benefits in making decisions, they will take into account costs and benefits not only internationally, but also within the domestic system. Since survival in power is a prerequisite to achieving any other goal, if that key goal is threatened those elites most affected will be willing to impose costs on other societal actors in order to gain benefits for themselves. If the most effective way to achieve domestic goals involves provoking costs from outside, as long as the net benefit to the threatened elites is positive, they will be willing to undertake such a strategy.

2. Responses to domestic threats will be made in a way that minimizes the danger to the bases of domestic power. If domestic legitimacy precludes the massive use of force against political opponents and depends on respecting certain political forms and "rules of the game," elites are very circumscribed in how they can respond to domestic threats. One effective response in this context is to shift the focus of political debate away from issues where they are most threatened--for example proposed changes in the structure of domestic economic or political power--towards other issues, defined in cultural or ethnic terms, appealing to interests other than those which clash with the elites' interests.18 By defining the collective interest in non-economic (cultural or ethnic) terms, such a definition can become the central political issue, replacing economic issues.
But ethnicity or culture in and of itself does not determine policies; the interest of the collective defined in ethnic terms can be defined in any number of ways. Indeed, if there are different groups within the elite competing for political support, they will compete over how to define the collective interest, and over how to achieve a specific definition, by drawing selectively on traditions and mythologies and in effect constructing a particular version of that interest. The elite faction that succeeds in identifying itself with the interest of the collective, and in defining the collective interest in a way that maximizes its own ability to achieve priority goals, wins an important victory. It has framed the terms of political discourse and debate, and thus the limits of legitimate policy, in a way that may delegitimize or make politically irrelevant the interests of challenger elites and prevent them from mobilizing the population on specific issues or along certain lines.

3. Images of and alleged threats from the outside world can play a key role in this domestic political strategy. A strategy relying on such threatening images can range from merely citing an alleged threat, to provoking conflict in order to create the image of threat; conflict in turn can range from political to military. Since political mobilization occurs most readily around grievances, in order to shift the political agenda elites must find issues of grievance unrelated to those issues on which they are most threatened, and construct a political context in which those issues become the center of political debate. It is at this point that focus on the interest of the group vis-à-vis the outside world proves to be quite useful. If the grievance or threat is to the collective rather than to individuals, it creates an image of potentially very high costs imposed on the group regardless of the direct impact on individuals. It therefore constructs the individual's interest in terms of a particular definition of the group's interest. If the threat or grievance is outside the direct experience of the majority of politically relevant actors, there is no way to verify whether the grievance is real, or indeed whether it is being addressed or not. And while such a strategy is costly in terms of reactions from the outside, the main costs (the effects of violent conflict) are taking place outside the domestic political arena, and are thus not imposed directly on politically relevant actors. It also becomes in effect a self-fulfilling prophecy, as the reactions provoked by the conflictual policies are themselves pointed to as proof of the original contention. Thus is created a grievance that, if violence is involved, is sure to continue for years.

The effect of creating an image of threat to the group is to place the interest of the group above the interest of individuals. This political strategy is crucial because, in the case of aggressive nationalism and images of threats to the ethnic nation, it creates a context where ethnicity is all that counts, and where other interests are no longer as relevant.

In addition, such an image of overwhelming threat to the group delegitimizes the dissent of those challengers who attempt to appeal to members of the relevant gsroup as individuals or who appeal to identities other than the "legitimate" identity in a "legitimate" way, especially if dissenters can be portrayed as selfish and uninterested in the well-being of the group, and can therefore be branded as traitors.19

Thus by using a strategy of agenda setting, shifting the focus of political attention toward the very pressing issue of threats to the group from outside, and by actively provoking and creating such threats, threatened elites can maximize the domestic benefits while minimizing the costs imposed on their own supporters and thus the danger to their own power bases.

4. In this domestic political context, information and control over information play a vital role. Control or ownership of mass media (especially television) therefore bestows an enormous political advantage where the wider population is involved in politics, and is a key element in the success of such a strategy. Information about the outside world in particular can play a central role, since it is beyond the direct experience of the vast majority of the population and cannot be directly verified. 5. Elites will tend to define the relevant collective in ethnic terms when past political participation has been so defined; when such a definition is encouraged by international circumstances; and when these elites are seen as credible defenders of ethnic interests and concerns. Clearly for grievances or threats to the group to be politically relevant, a majority of politically relevant actors must be able to be identified as members of that group. That does not mean, however, that their main or primary identity must be to the group; in fact, people have multiple identities and such identities are highly contextual. The key is to make a particular identity, and a specific definition of that identity, the only relevant or legitimate one in political contexts.

The group identity must have some link to past political participation. Ethnic nationalist ideology, relying on the idea of identity in terms of ethnicity and appealing to the collective interest of the nation defined in ethnic terms, has in particular proved a quite effective means of gaining and maintaining power in the domestic arena in the successor states of the great empires of Eastern and Central Europe. A major reason for this politicization of ethnic identity is external; from the nineteenth century onwards the great powers used the standard of national (usually ethnically-defined) self- determination to decide whether a territory merited recognition as a sovereign state--a practice that continues to today. Those elites in a territory who could make the best case for representing the interests of an ethnic group could increase their power vis-à- vis the domestic arena by being internationally recognized as the representative of their ethnic or national group.20 In Eastern and Central Europe this factor reinforced the Ottoman, Romanov and Habsburg empires' definition of political participation in terms of religion in the first two cases and language in the latter and the subsequent construction of politicized identities in the 19th century.21

This political reification of ethnic identity was even more firmly entrenched in the socialist world, where the regimes used the ethnic nation as a real unit of analysis. Thus, territorial administrative units were set up along ethnic lines, specific aspects of national culture were encouraged, and in Yugoslavia an ethnic "key" existed within republics which determined the distribution of certain positions by ethnic identity according to the proportion of each group in the republic's population.22 In addition, the repression of overt expression of ethnic sentiment and the oppression of entire groups of people solely on ethnic criteria created grievances which came to the fore when politics was opened up to the wider population. This combination of an ethnically- defined territorialization of power and ethnic grievances against the old regime encouraged elites in newly democratizing regions to look to ethnic national identity as a legitimating force.

It is these types of political circumstances, both internal and external, which determine which identity will be likely to predominate in domestic influence attempts which appeal to collective interests. Since conflictual policies tend to take place along these previously politicized lines of identity, they also tend to create the impression of continuity between past conflicts and current ones, and indeed are specifically portrayed in this way. But there is nothing "natural" about ethnic interest that requires it to be defined in a conflictual way.

6. The larger and more immediate the threat to the ruling elite, the more willing it is to take measures which, while preserving its position in the short term, may bring high costs in the longer term; in effect it discounts future costs. The intensity and thus costliness of a conflictual strategy depends on the degree of the threat to old elites. These factors include:

a) The time frame of the threat to power. While the conflictual policies may over the long run result in an untenable position and ultimately undermine the bases of political influence, current political behavior in a situation of immediate threat is motivated by that threat and the concern for keeping power in the short run, which at least leaves open the possibility of survival in the long run. This also gives them time to fashion alternative strategies for dealing with change, including shifting the bases of their power.

b) The strength of the opposing elites also affects the immediacy of the threat. If the opposing elites are successfully mobilizing the majority of the politically relevant actors against the status quo, ruling elites will feel quite threatened and be willing to incur high costs to preserve its position. Threatened elite will also attempt to recruit other elites, at the local and regional as well as national levels, to prevent such a mobilization, since the stronger the challenging coalition, the greater the threat.

c) The costs to the threatened elites of losing power is a further factor; that is, what resources and fallback positions will they have if change does take place. If they have everything to lose and nothing to gain, they will be much more likely to fight and to build a coalition with other elites than if they have resources that would allow them to remain involved in power to some degree. For example in the Soviet Union and in Yugoslavia many ideological workers and some managers of unprofitable factories felt very threatened by moves toward market economics (because they owned nothing, since their power, privilege, and economic security were tied up with their positions, and because they had no skills transferrable to market conditions), and have indeed been the strongest parts of coalitions using aggressive nationalism. In Hungary, on the other hand, the reform process of previous 25 years meant that there were many fewer purely ideological workers, and many party workers and managers had transferrable skills. The result is less resistance on their part to change, since it did not pose as much threat.23

d) For the conflictual strategy to include the use of military force, especially against other states, the status-quo coalition must include a dominant faction within the military.

7. Threatened elites may use marginal neo-fascist parties as part of their conflictual strategy in conditions where the wider population is included in the political system. Every country has small extremist groups whose mainstay is exactly this type of strategy of ethnic hatred and violence; their motivations may be political, personal, psychological. But the very existence of this option is clearly not enough for it to come to dominate state policy. For a sufficient part of the power elite to ally with or make use of this extremist fringe and give it access to weapons, media, etc., parts of the national and regional elite must feel threatened enough, as described above, to see the costs involved as worth the risk. An advantage is that by bringing extremists into the political realm, the "center" is moved to the right; so a statement that ten years earlier may have been too racist, after this kind of strategy may be relatively moderate.24 By making issues of ethnic nationalism the center of political discourse, this strategy also turns those who, if the axis were economic issues, would be labeled arch-conservatives, into moderate centrists.
8. Internal costs of a conflictual strategy are closely monitored, since they must be outweighed by benefits. Of particular importance is the need to prevent popular mobilization against costs of the conflictual external strategy. When conflict is in the realm of political rhetoric, it may have great support among the population, since it is basically costless. But if military conflict is involved, the costs to the general population rapidly start to mount. Despite assumption that ethnic political mobilization inevitably pushes politics towards extremistm, there is in fact little evidence of a natural progression from ethnic mobilization to violent ethnic conflict.25 Rather, conflict will be undertaken with an eye to minimizing the costs for those parts of the population which are key for support, and conflict will tend to be provoked outside the borders of the elite's power base, with great efforts taken to prevent war from spilling over into the domestic territory. Thus in the Soviet case, anti-reform conservatives provoked violent ethnic conflict outside of Russia, in Moldova, Georgia and the Baltics; in the Yugoslav case armed conflict has not taken place within Serbia itself, and the Croatian conservatives' conflictual strategy affected mainly central Bosnia.

Of course, if material conditions deteriorate enough and if the discrepancy between the interest of the collective group and the interest of the status quo power elite becomes great enough, parts of the elite may successfully lead the wider population to revolt violently against the power structure to such an extent that force has to be used against members of the group whose collective interest is supposedly being sought. In this case parts of the old elite may jump on the bandwagon of the new elites who lead such revolutionary revolts.

9. External costs are also key. Such a strategy is most likely when the potential international costs, in terms of how it would affect the status-quo elites' domestic power position, are minimal. But if the cost of external reaction threatens parts of the status- quo coalition, they may defect, since losses at hands of domestic elites may be less than at hands of external foes, especially if challenger elites are willing to offer a deal to the defectors. This strategy will thus be very sensitive to the kinds of costs it provokes from the outside.

This type of conflictual policy thus comes to dominate some states or regions and not others, depending on the degree of threat to the existing power structure and the size of the coalition (at both national and regional levels) of those within the power elite threatened by change. If a challenge to the existing power structure takes place in such a way that most of the old elite perceives a "way out," either by cooptation into the new system or by being allowed to fade away without being deprived of all privileges and benefits, a coalition will probably not be strong enough to impose a costly conflictual strategy as state policy. It may nevertheless incite conflict and violence in the hopes of gaining wider support.

The Case of Serbia
The violent conflict along ethnic lines in the former Yugoslavia was a purposeful and rational strategy planned by those most threatened by changes to the structure of economic and political power, changes being advocated in particular by reformists within the ruling Serbian communist party. In response, a wide coalition--conservatives in the Serbian party leadership, local and regional party elites who would be most threatened by such changes, orthodox marxist intellectuals, nationalist writers, and parts of the Yugoslav army--joined together to provoke conflict along ethnic lines. This conflict created a political context where individual interest was defined not in terms of economic well-being, but as the survival of the Serbian people. Their original goal was to recentralize Yugoslavia in order to crush reformist trends throughout the country, but especially in Serbia itself. By 1990, in a changed international context and with backlashes in other republics against their centralization strategy, the conservative coalition moved to destroy the Yugoslav state and create a new, Serbian-majority state from which a large percentage of non-Serbs have been expelled. By provoking conflict along ethnic lines this coalition deflected demands for radical change and allowed the elite to reposition itself and survive in a way that would have been unthinkable in the old Yugoslavia, where only 39 percent of the population was Serb.

Serbian conservatives relied on the particular idea of ethnicity in their conflictual strategy because political participation and legitimation in this region historically was constructed in such terms. The Serbian national myth, molded in the struggle against the Ottoman Turks and in the expansion of the Serbian state in the 19th and early 20th centuries, played a central role in Yugoslav politics between 1918 and 1941, and remained important for the communist partisans, who relied on popular support during the war. The ethnic national bases of the Yugoslav republics was the result of this wartime need for popular political support, and was maintained as more than a facade after the 1948 break with the Soviet Union again forced the communists to rely on some level of popular support. This political reification of ethnicity, along with the suppression of expressions of ethnic sentiment, combined to reinforce the historical construction of political identity in terms of ethnic identity.26

In addition, the rhetoric of threats to the ethnic nation was available to Serbian conservatives in a way that it was not in other republics, in part because the Serbian party was one of the few that was ethnically homogeneous enough that such a strategy would not automatically alienate a significant portion of the party membership. In addition, the Serbian republic (even without its provinces) had regional differences in economic development that were more extreme and significant than in any other republic. Thus while liberals there were stronger, conservatives were also more threatened, and had a grassroots base upon which to rely for support. Serbia's conservatives were also well-placed to oppose change, given Serbia's centrality to the Yugoslav federation and the often congruent interests between Serbian conservatives and conservative elements in the Yugoslav army.
Five episodes are described below in which conservative forces, especially those in Serbia, were threatened with the radical restructuring of political and economic structure of power. Using the concepts laid out above, each section looks at the threat to the conservatives and the status quo; looks at their responses; and looks at the effect of those responses.

1960s: Threats to the status quo
In the early 1960s, in response to an increasingly dysfunctional economic system, reformists in the Yugoslav party leadership, with Tito's support, began a radical restructuring of the Yugoslav political and economic system. At the micro level the 1965 reform was a direct attack on party bureaucrats in enterprises as well as those in local administrative positions,27 and also involved a loosening of party control of society, including tolerance of more open expression of national sentiment.28
At the macro level the reform radically decentralized the federation, and almost all decision making was given to the republics. This allowed the top leadership to bypass the conservatives who dominated the central bureaucracy and to rely instead on the republic-level leaders and central committees, which were dominated by young technocratically-oriented reformists. Indeed, this decentralization was enthusiastically supported by all the party leaderships, including the Serbian. By the summer of 1971 there was also discussion of decentralizing the party itself, a topic which was to be addressed at a party meeting in November 1971.29 If undertaken, the effect would have been to institutionalize reformism in each republic, remove all power from the conservatives who dominated the center, and to remove even the possibility of a conservative comeback.

The conservatives were clearly threatened by the popularity of the young republic-level reformist leaderships in their central committees, as well as among the wider population. Indeed, the goal of the reforms had been in part to broaden the legitimacy of the communist party by building a base in that wider population; this meant, however, that conservatives were faced with leaders who could mobilize the population in support of irreversible radical changes in the structure of power.30

Response
In response the conservatives at first tried to sabotage implementation of the reform. The result, however, was that in 1966 Tito purged conservatives from the leadership of the party, and the reform became even more radically threatening to conservatives. Some conservatives in the Serbian party then began publicly to argue that the reforms were harmful to the Serbian nation, and linked the reforms to the "historical enemies" of Serbia. Although they were expelled from the party in 1968, by 1971, as the party faced the possibility of radical decentralization, other conservatives in the Serbian party and army pointing in particular to the open expression of nationalist sentiment in Croatia, which included some extremist views. Conservatives blamed the Croatian leadership for revival of Croatian nationalism.31 These conservatives allied with some conservatives in the Croatian and Bosnian parties, party workers and war veterans who had been forced into retirement, members of the central bureaucracy, elements in the Yugoslav army, and Serbian nationalist intellectuals, to invoke the massacre of hundreds of thousands of Serbs by the Croatian Ustase leadership during World War II and to blame the reforms for undermining socialism and endangering Croatia's Serbs. Conservatives in the security forces and in the army, in particular, convinced Tito to act against the Croatian reformists.32 The Croatian reformists were purged and tanks were sent onto the streets of Zagreb.33 The following year the Serbian reformists were also purged, despite very strong resistance from the republic's central committee; the other republics and provinces followed. As a result, the local-level reforms were effectively reversed, and a renewed ideologization took place.34

Casting the threat posed by reform in terms of ethnic nationalism allowed the conservatives to shift the focus of political debate away from the cross-republic reformist project, and toward the alleged threats from Croatian nationalism; this allowed them to argue that radical reform had in fact brought the emergence of nationalism and thus of counterrevolution.35 By using the threat of external and internal enemies of socialism defined in ethnic national terms, they managed to divide the country's strong reformists and thereby to prevent the decentralization of the party and to reverse the essence of the reforms (although decentralization of the federation itself remained and was enshrined in the 1974 constitution).36 In addition, the Yugoslav Army now became a key political player, with the official role of ensuring the domestic order against external and internal enemies; this made the army the natural ally of conservatives in the party. By 1974, 12 percent of the federal central committee wer army officers, up from 2 percent in 1969.37

1980-1987: Threat to conservatives
When Tito died in May 1980 the debate over reform, which had been muffled, broke out into the open. The economic crisis triggered by the global recession of the late 1970s, the oil shock, and Yugoslavia's huge foreign debt burden ($20 billion by the early 1980s), as well as the negative results brought by ending reform in the early 1970s, all compelled radical systemic change. And indeed, the reformists' proposals were much more radical than in the 1960s and their audience--managerial elites, democratically-oriented intellectuals, and party rank-and-file--much more receptive. The proposals were therefore much more threatening to the conservatives than they had been in the 1960s, especially without Tito to moderate conflicts; the political conflict had become winner-take-all.

Serbian reformists were in the forefront of this struggle, and in the early 1980s the Serbian party was among the most liberal in the country. Members of the Serbian party leadership called not only for totally removing party influence at the local levels of the economy, but also for greater reliance on private enterprise and individual initiative; multiple candidates in state and party elections; free, secret elections in the party; recognition and adoption of "all the positive achievements of bourgeois civilization," i.e. liberal democracy.38 From within the party were also heard calls for private enterprise to become the "pillar of the economy," and even calls for a multi-party system. Reformists were also very critical of the Army's privileged political and budgetary position, and very early called for cutting that influence.39 Once again reformists were seeking to mobilize broader popular sentiment against conservative positions among party rank-and-file as well as the wider population at a time when the economic crisis had discredited the conservatives' ideological stance.40

Due to the consensus nature of federal decision making, the conservatives were at first able to hinder an outright reformist victory, but the terms of the debate nevertheless shifted in the favor of the reformists. Indeed, by the mid-1980s secret multi-candidate elections were being held for party officers, and even some state posts were chosen in multi-candidate popular votes.41

Conservative response to the threat
Conservatives in Serbia responded with a three-pronged strategy. The first was to reemphasize orthodox Marxist themes, in an attempt to delegitimize liberal trends at the lower levels of the party; the second was to attempt to defeat the reformists in the leadership by shifting the focus of attention toward ethnic questions, in particular the alleged "genocide" against Serbs in the province of Kosovo; the third was to portray Serbia as the victim of Tito's Yugoslavia, setting the stage for an attack on the autonomy of other republics.

Although the conservatives were not very successful in the political debates over reform at the leadership level, at the local level in Serbia they imposed an orthodox ideological line, while at the same time raising the issue of Serbian nationalism. Most notable was the Belgrade party organization, headed beginning in 1984 by Slobodan Milosevic. Soon after coming to power Milosevic began a campaign stressing ideological orthodoxy,42 and sent out warnings to all Belgrade party units urging vigilance against "the dangerous increase in anti-Yugoslav propaganda" from internal and external enemies, a warning that also dominated Yugoslav army leadership pronouncements.43

At the same time, a nationalist campaign began among Belgrade party members and "lefist" intellectuals, including Milosevic's sociologist wife Mirjana Markovic, who sought to defend "the national dignity of Serbia" and to protect its interests in Yugoslavia.44 Belgrade also saw growing numbers of protests by Serbs from the province of Kosovo, claiming to be the victims of ethnic Albanian "genocide."45 The fact that the demonstrations took place without police interference was a sign that they were at least tolerated by the Belgrade party.

In January 1986, despite very strong opposition from within the party leadership, Milosevic was elected head of the Serbian party's central committee.46 This period saw increased attention to the issue of Kosovo by a Belgrade-centered coalition of conservative party members, orthodox Marxist intellectuals, and nationalist-oriented intellectuals who repeated the charges of "genocide" against Serbs in Kosovo.47 Journalists who were allied with Milosevic, especially at the daily newspaper Politika, undertook a media campaign to demonize ethnic Albanians and to "confirm" the allegations of genocide.48 Indeed, the issue of Kosovo now became the conservatives' main weapon against reformist forces within Serbia and in the wider federation, as Serbian conservatives insisted that the issue be the priority concern not only of the Serbian party but at the federal level as well.49

However, it soon became clear that this coalition's goals were not limited to Kosovo and Serbia. An ideological manifesto written by some members of the Serbian Academy of Sciences and Arts in 1985, although claiming to call for democracy, actually advocated the restoration of the repressive, centralized socialist system that existed before the 1965 reforms; sharply attacked the 1965 reforms as the root of all evil in Yugoslavia and as being aimed against Serbs; declared Serbs in Kosovo and Croatia to be endangered; and denounced the "anti-Serbian coalition" within Yugoslavia.50 Indeed, given the nature of decision-making in Yugoslavia, to prevent radical reform in Serbia the conservatives would have to ensure that it did not take hold in the other republics and at the federal level.

Effect of the conservative response
The result of the Serbian strategy was that questions of radical reform were shunted aside in order to deal with the pressing issue of "genocide" in Kosovo, and through a combination of press manipulation, mass rallies and political manipulation, and a stress on Stalinist notions of democratic centralism, by September 1987 Milosevic managed to consolidate conservative control over the Serbian republic's party organization.51 Those parts of the Serbian media that had been relatively independent were taken over by conservative editors allied with Milosevic.

1988-1990: Threats to the status quo
The conservative coalition, although it had consolidated control over the Serbian party organization, the conservative coalition still faced threats from reformist forces in other Yugoslav republic and provincial organizations (Serbia was only one of eight) as well as in the federal government, especially as the economic situation continued to deteriorate. Slovenia, with strong liberal and democratic currents, was in the vanguard of increasingly vocal calls for an end to the one-party system and for Yugoslavia to move closer to the west, as well as very sharp criticisms of the Yugoslav army.52 Also threatening were the successes of Federal Prime Minister Ante Markovic, a strong reformer who, despite Serbian opposition, managed to get Federal Assembly approval for radical transformation of the Yugoslav economy.53

Conservative responses to the threat
Over the course of 1988 and 1989 Milosevic and his allies attempted to subvert the party leaderships in other Yugoslav republics and to weaken the federal government through a strategy of appealing to an aggressive version of Serbian nationalism. This strategy was viable despite the Serbs' minority status in Yugoslavia, because Serbs were overrepresented among politically relevant actors including communist party officials and members in other republics, and within the federal bureaucracies.54 As long as this remained the case, Serbian conservatives could "legitimately" gain power in all of Yugoslavia (and thereby legally recentralize the country) if they could dominate the federal party and state collective leaderships by controlling at least five of the eight votes.
To this end Serbian conservatives continued to focus on the image of threatened Serbs in Kosovo. They staged mob rallies of tens of thousands in every major town in Serbia as well as in other republics and in front of party headquarters and during party meetings; these rallies, decrying the "atrocities" in Kosovo, called for party leaders to step down.55 The result was that the party leaderships in Vojvodina and Montenegro were ousted in October 1988 and January 1989.56 The Kosovo party leadership (which had been hand-picked by the conservatives in Belgrade) was also pressured to resign and thereby allow the end of Kosovo's autonomy and the recentralization of Serbia. Although these moves provoked massive demonstrations and strikes among the province's Albanian population to protest the threat to its autonomy, in March 1989 the Kosovo assembly, subjected to fraud and manipulation by Belgrade, voted to end the province's autonomy.57

Similar pressure was also put on the Croatian government: massive rallies organized from Belgrade were held in the rural Serb majority region of Krajina, with the intention of eventually moving on to Zagreb to overthrow the Croatian party leadership.58 Likewise the ruling party in Bosnia-Hercegovina discovered that Serbia's secret police were active in the republic.59 In Slovenia the plan was cruder: hundreds of intellectuals and dissidents were to be arrested and the army was to be used to put down protests.60

The conservatives' strategy of consolidating control over the other republics through the use of aggressive Serbian nationalism was accompanied by increasingly vehement media demonizations not only of Albanians, but also of Croats,61 as well as an active campaign to portray Tito's Yugoslavia as specifically anti-Serbian.62 It claimed that an authoritarian, Serb-dominated and centralized Yugoslavia was the only way to ensure the security and interests of all Serbs; such a Yugoslavia also, not coincidentally, would ensure the power interests of the conservative Serbian elites.

One of the major concerns throughout the country, but especially in Serbia, was the deteriorating economy, which was an issue that could not be ignored. Milosevic blamed these problems on Markovic's reforms, and put forward his own program that rejected even the most modest of the reformists' proposals. Milosevic called for more efficient use of existing resources rather than any structural changes, emphasized "social ownership" rather than private property, stressed the priority of reforming (that is, strengthening) the federal organs, and rejected even the possibility of nonsocialist political parties.63

Meanwhile the army, under Defense Minister Branko Mamula, openly sided with orthodox conservative positions and harshly attacked the political opposition. Counter to trends in much of the country, orthodox indoctrination was stepped up within the military itself.64 The army also endorsed Milosevic's neo-socialist economic and political program, stressing in particular continued monopoly of the communist party and recentralization of the state.65 In cooperation with Serbian conservatives, the military openly attacked reformists' calls to democratize the country, reduce the military's political role and to reform the military-industrial complex. Moreover, statements by top army officers "made clear that they viewed the Army's internal mission in orthodox ideological terms."66

Effects of the conservative responses
Although this strategy gained Serbia control over four of the eight federal units, and placed the threats to Serbdom at the center of political discourse, it also provoked backlashes in the other republics. In Slovenia, publication of the army plans to crush dissent radicalized the party and the wider population in Slovenia, where by mid-1988 an unofficial referendum on independence was held and the party began advocating introduction of a multi-party system. In Croatia, a bastion of conservatism since 1971, the Serbian moves emboldened the reformist minority, so that by October 1988 the Croatian party proposed dismantling the communist party's leading role and encouraging private property.67 Even conservative Serbs within the Croatian leadership criticized Milosevic's strategy.68 Likewise in Bosnia, which had previously been supportive of Milosevic, the aggressive nationalist strategy and the threat to the Bosnian party leadership led it to distance itself from Serbia's positions.69

By the end of 1989, reformist forces had taken over the Croatian party, and both the Slovene and Croatian parties had scheduled multi-party elections for the spring of 1990 (in Croatia despite attempts by conservative Serb allies of Milosevic to prevent this).70 An attempt by Milosevic to recentralize the federal party at an extraordinary LCY Congress in January 1990 failed as the Slovene party walked out when its proposal for de jure party independence was rejected, and the Croatian, Bosnian and Macedonian parties refused to continue the meeting.

1990: Threats to the status quo
1990 saw the greatest threat yet to the conservative Serbian coalition and its allies: the emergence of a political system in which the wider population would choose political leaderships. The strategy of recentralizing Yugoslavia by use of mob rallies and aggressive Serbian nationalism to pressure communist party leaderships was clearly no longer feasible; likewise, there was little chance of winning an election in a country where only 39 percent of the population was Serb, especially since Milosevic's strategy had alienated most non-Serbs.

The specific threats were now from three directions. The first was the fact that in the spring 1990 elections in Slovenia and Croatia, openly anti-socialist parties committed to a loosening rather than tightening of political ties had taken power, due in large part to a backlash against Milosevic.71 Federal decision-making bodies thus now included representatives from these two republics, marking the introduction of an irreconcilable ideological difference in terms of economic and political viewpoints. Indeed, the Slovene and Croat governments soon put forward formal proposals for confederalizing the country, rejecting out of hand Serbia's calls for recentralization. Given the pressure for multi-party elections in the other Yugoslav republics, and the fall of communist parties throughout the rest of Eastern Europe, it seemed likely that other republics would join these calls.72

The second set of threats came from the policies of federal Prime Minister Markovic. By early 1990 these policies were quite successful in lowering inflation and improving the country's economic situation, and he was very popular, especially within Serbia.73 Taking advantage of these successes, and looking ahead to multi-party elections, he pushed bills through the Federal Assembly legalizing a multi-party system in the entire country, and in July 1990 formed a political party to support his reforms.74

The biggest challenge, however, came from within Serbia itself. Encouraged by the fall of communist regimes in the rest of Eastern Europe and the victory of noncommunists in Croatia and Slovenia, opposition forces in Serbia began organizing and pressuring the regime for multiparty elections, holding massive protest rallies in May. Although Milosevic argued that elections could not be held until the Kosovo issue was resolved, by June the Serbian regime recognized the unavoidability of elections.75

Response
Within Serbia, the regime again resorted to the issue of Kosovo, in particular working assiduously to provoke violent resistance from the Albanian population.76 Despite these actions and the fact that the new Serbian constitution, adopted in September, effectively stripped Kosovo of its autonomy, the Albanians pursued a policy of peaceful resistance.

While turning up the heat on Kosovo, the Serbian party also had to deal with opposition parties at home, both nationalist ones from the right (most notably the Serbian Renewal Movement, SPO, headed by writer Vuk Draskovic), as well as from civically-oriented democratic parties. In the face of anticommunist nationalist party opposition, and in order to win the necessary two-thirds of the Serbian vote (since the party had alienated the 33 percent non-Serb population of the republic), the Serbian conservatives first undertook a strategy of averting a split of the communist party into a large pro-reform social democratic party that would more credibly appeal to the population's economic interest, and a small hard-line party (as happened in the rest of Eastern Europe). The Serbian party was renamed the Socialist Party of Serbia (SPS). The regime continued its complete control over the mass media, and greatly limited access of opposition parties to television. In the election campaign the SPS successfully appealed to the material well-being of the voters, arguing for the continuation of a socialist system that provided social security and economic growth.77 Economic problems were blamed on the "anti-Serbian" policies of Yugoslav federal Prime Minister Markovic. The government also printed US$2 billion in dinars for overdue worker salaries just before the December elections, funds taken illegally from the federal treasury.

On issues of nationalism, the party had already very much distanced itself from the policies of Tito, especially those which forbade public expression of national sentiment. This fact, plus the fact that Yugoslav agriculture had remained in private hands, ensured the SPS of most of the vote of peasants and those one generation off the land, a majority of the voters, and thus dampened anticommunist sentiment against it.78 The SPS, linking the nationalist Serbian Renewal Movement (SRM) to Serbian extremists during World War II, portrayed the SRM as wanting to drag Serbia into war, and painted itself as a moderating and progressive force.79 The SPS managed to win an overwhelming majority of parliamentary seats with the support of 47 percent of the electorate (72 percent of Serbia's Serbs).80

But the challenge to the conservatives from outside of Serbia, in the context of the Yugoslav federation, continued. The Serbian conservatives' response was to continue to demonize other ethnic nationalities, and also to begin provoking confrontations and violent conflicts along ethnic lines and to discredit the very idea of a federal Yugoslavia, calling it the creation of a Vatican-Comintern conspiracy.81

Even before the 1990 elections the Belgrade media had stepped up its campaign against Croatia, and after the elections it accused the new Croatian ruling party, the Croatian Democratic Union (CDU) of planning to massacre its Serbian residents.82 Throughout the summer of 1990 the Serbian media also ran stories detailing the anti-Serb massacres of the World War II Ustasa regime, furthering the implicit link with the CDU.83 Following the Croatian elections, Belgrade and its allies also began to provoke violent conflict in the Serbian-populated areas of Croatia, although in the May 1990 elections only a small minority of Croatia's Serbs had supported the Serbian nationalist party, the Serbian Democratic Party (SDP).84 Between July 1990 and March 1991, Belgrade's allies took over the SDP, replacing moderate leaders with hard liners. It portrayed the CDU as genocidal Ustasa; rejected all compromises with Zagreb; held mass rallies and erected barricades; threatened moderate Serbs and non-SDP members who refused to go along with the confrontational strategy; provoked armed incidents with the Croatia police, and stormed villages adjacent to the regions already controlled by Serbian forces and annexed them to their territory.85 Throughout this period, conciliatory moves by the Croatian regime were rejected and moderate Serbs who disagreed with Belgrade's conflictual strategy were branded as traitors.86 Although the campaign rhetoric and the actions of hard-liners in CDU did give Serbs cause for concern, rather than fostering negotiation and compromise with Zagreb, Belgrade exacerbated the Croatian Serbs' concerns.

Following Milosevic's December 1990 victory in the Serbian elections, the situation in Croatia became even more confrontational as a hard line group within the SDP, working closely with Belgrade and armed by the Yugoslav Army, began to provoke armed conflicts with Croatian police in areas where Serbs were not in the majority.87 Croatian Serbs were increasingly pressured to toe the SDP line, and Croats in Krajina were besieged by Serbian armed forces and pressured to leave.88 These purposefully provoked conflicts were publicly characterized by Belgrade as "ethnic conflicts," the result of ancient hatreds, and the Yugoslav army was called in to separate the groups. At the end of February, Krajina proclaimed its autonomy from Croatia.

These Serbian moves provoked Croat hardliners to take repressive actions against Serbs in areas where the ruling party controlled the local government: these actions were pointed to by Belgrade's allies as proof of the threat to Serbs.89 Despite calls by Croatian hard-liners to use military force, Zagreb lacked significant stocks of weapons (although it was seeking sources), and Croatian president Franjo Tudjman clearly feared providing the Yugoslav army with an excuse to crush the Croatian government. He was thus forced to accept the army's gradually expanding occupation of the areas where the SDS's authoritarian rule prevailed. This period saw the groundwork for a similar strategy being laid in Bosnia by Belgrade's ally there, Radovan Karadzic, head of that republic's SDS.90

As conflict heated up in Croatia, in negotiations over the future of Yugoslavia, Milosevic and his allies refused to budge from his call for a more tightly centralized federation. He declared that if his demand was rejected, then the borders of Serbia would be redrawn so that all Serbs would live in one state.91

The result of this strategy of conflict was to further the destruction of Yugoslavia. The provocations and repression of even moderate Serbs in Croatia increased the territory under JNA control, and provoked reactions on the part of extremist Croats.

1991: Threat to the status quo
This apparently successful strategy was suddenly interrupted when the Serbian political opposition held massive protest rallies in Belgrade on March 9 and 10.92 Appealing to the wider population, the opposition, led by Serbian Renewal Movement chief Vuk Draskovic, threatened to oust the regime by force of street rallies. Initially called to denounce the regime's tight control and manipulation of the media, the rallies also condemned Milosevic's disastrous economic policies and his policy of provoking conflict with other republics.93 They called for the SPS to step down from power as other East European communists had done. Although Milosevic's immediate reaction was to call the army to put down the demonstrations (since the republic's police forces were all in Kosovo), the military refused to use massive force.94 This marked the start of the democratic opposition's rapid rise in popularity and the beginning of an open split within the ruling SPS by democratic, pro-reform forces.95 Shortly thereafter massive strikes (including one of 700,000 workers) aimed specifically against Milosevic's regime shook Serbia.96

Response
Given the refusal of the army to use force, Milosevic was forced to negotiate with his opponents. He accepted limited economic reform, printed more money to pay workers, and discussed the formation of a multi-party Serbian national council. At the end of March he secretly met with Croatian President Tudjman to agree on a division of Bosnia-Hercegovina, thus removing the possibility of Tudjman taking advantage of Milosevic's then-weak position. In April Milosevic finally accepted the principle of a confederation, and in early June talks over the future of Yugoslavia, he agreed to the principles on which such a confederation would be based.97 Belgrade also pressured its Serbian allies in Croatia to negotiate with Zagreb, although they refused to reach an agreement.98

Yet at the same time, the strategy of provoking conflict along ethnic lines was also stepped up. Milosevic himself labeled the protesters "enemies of Serbia" who were working for Albanians, Croats, and Slovenes to try to destroy Serbia, and he ominously stressed the "great foreign pressures and threats" being exerted on Serbia and which gave "support to the forces of disintegration of Yugoslavia."99 The media stepped up its portrayals of Croatia as a fascist Ustase state, and in April graphically reported on the opening of caves in Bosnia-Hercegovina filled with the bones of thousands of Serb victims of the Ustase; in August it broadcast the mass interment of the remains.100

This period was also one of close cooperation between the Yugoslav army, the Belgrade regime and the Bosnian SDP, as the three sides implemented "Project RAM," a plan to use military force to expand Serbia's borders westward and create a new Serbian Yugoslavia.101 Thus in Bosnia in the spring of 1991, the SDP set up "Serbian Autonomous Regions" which were declared no longer under the authority of the republic government, a repetition of the Krajina strategy.102

The SPS at this time also began an open alliance with the neo-fascist Serbian Radical Party led by Vojislav Seselj, ensuring Seselj's election to the Serbian parliament in a by-election.103 Seselj's guerrilla groups were active in the ensuing escalation of conflict in Croatia. In this period, Belgrade also exerted growing pressure on moderate Serb leaders in Croatia's ethnically-mixed Slavonia region (where Serbs were not in the majority) to accept its confrontational strategy; in May, Krajina held a referendum to join with Serbia, and Belgrade-supported guerrillas including Seselj's "Chetniks," flowed into Croatia, terrorizing both Serb and non-Serb populations in the more developed regions of Eastern and Western Slavonia (neither of which had Serb majorities).104 These forces attacked Croatian police, in at least one case massacring and mutilating them, and began a policy of forcible ethnic expulsions in areas coming under their control. Moderate SDP leaders denounced Belgrade for provoking and orchestrating this confrontational strategy.105

In the face of this pressure, and in preparation for the new confederal agreement, in late June the Croatian government declared the start of a process of disassociation from Yugoslavia, specifically stating that it was not an act of unilateral secession and that Zagreb continued to recognize the authority of federal organs, including the army.106 When the army attacked Slovenia following its own declaration of sovereignty, Croatia did not help the Slovenes, in order to avoid giving the army an excuse to attack Croatia.107

Nevertheless, Yugoslav army, despite its promises not to attack Croatia,108 escalated and intensified the conflict in Croatia, and Serbian forces continued their strategy of provoking conflicts in Slavonia and on the borders of Krajina, terrorizing civilian populations, destroying Croatian villages and Croat parts of towns, bombing cities to drive out the population, and forcing Serbs on threat of death to join them and point out Croat-owned houses.109 Serbs who openly disagreed with these policies were terrorized and silenced.110 Helsinki Watch noted that in the period through August 1991, when the Croats finally went on the offensive and Croat extremists themselves undertook atrocities against civilians, by far the most egregious human rights abuses were committed by the Serbian guerrillas and the Yugoslav army, including indiscriminate use of violence to achieve their goals of terrorizing the Serb population into submission and driving out the non-Serb population.111 This policy, by provoking extremists forces in Croatia into action, thus in effect became a self-fulfilling prophecy as the Serbian regime pointed to those atrocities as proof of their original charges.112

This war policy also destroyed the chances for Markovic's reforms to succeed. Although Slovenia and Croatia along with Serbia had been trying to block implementation of many aspects of his reform, the JNA and Serbian guerrilla attacks ended support for a continued Yugoslavia even among those who had advocated it, while Milosevic's moves to take over the Federal presidency and marginalize the federal government by September 1991 led Markovic to the conclusion that he had no choice but to resign.113 By the summer the army was also draining the federal hard currency reserves and taking up a vast proportion of the federal budget, which had been carefully managed by Markovic.

The war also helped Milosevic in his domestic crisis. In April 1991 the democratic opposition had been at a high point, the ruling party was facing a split, and commentators were predicting the imminent fall of the SPS. But the SPS used charges of genocide and the subsequent war in Croatia to suppress internal party dissent and to marginalize the democratic opposition by drowning out concerns about economic and political reform, and by charging those who questioned the war with treason. The regime also used the war to try to physically destroy the opposition: it sent first to the front reservists from counties that had voted for opposition parties. Opposition leaders and outspoken anti-war activists were sent to the front, and any criticism was met with physical threats and violence from neo-fascist gangs.114 The regime also targeted the Hungarian minority in Vojvodina (an absolute majority in seven counties), which, although only three percent of Serbia's population, represented seven to eight percent of reservists at the front and 20 percent of casualties.115

By September the army was attacking Dubrovnik, and thousands of reservists were wandering Bosnia-Hercegovina, terrorizing the Slavic Muslim population.116 But at this same time there was growing discontent in Serbia about the war.117 Thousands of young men hid or left the country to avoid being drafted, and whole units of reservists deserted from the front.118 It was clear that the SPS's hard-line allies in Moscow had failed in their attempts to seize power.119 By November 1991, when the European Community threatened economic sanctions against Serbia, and Croat forces began taking back territory, Serbia accepted the principle of UN peacekeeping forces in the areas it controlled in Croatia.

By this time, the opposition in Serbia was again gaining momentum, drawing on the anti-war sentiment and continued economic decline. Condemning the SPS's economic policy, the war in Croatia, and even the conflictual policy in Kosovo, the opposition by February was gathering hundreds of thousands of signatures calling for Milosevic's resignation and the convening of a constitutional assembly.120 Once again, the regime pulled back, finally allowed UN troops to move into Krajina, put pressure on hard-liners in Knin, allowed moderate Serbs to negotiate with Zagreb,121 set up meetings with the remaining four Yugoslav republics to negotiate a future Yugoslavia and called for talks with Croatia.122

But at this same time, Serbia also stepped up the pressure on Bosnia, instituting an economic blockade of the areas not controlled by its SDP allies.123 This time the ethnic enemy was the allegedly fundamentalist Muslim population of Bosnia, who were said to be seeking to impose an Islamic state on the Serbs of the republic.124 Indeed, the same scenario was beginning in Bosnia: in December the SDP declared it would form a republic,125 and in January 1992 the independent "Serbian Republic" was declared in the 66 percent of Bosnian territory that the SDS controlled there, the "Serbian Autonomous Regions" that had been formed the year before. SDP leader Radovan Karadzic at this time declared that Bosnia would never again be undivided.126 Objecting to a referendum to be held in those parts of the republic not under SDP control, Serbian guerrilla forces began armed attacks on Croat and Muslim civilians in early March.127 Despite this, the referendum, seeking approval for Bosnia-Hercegovina's independence, was approved by 63 percent of the republic's population, including a large proportion of those Serbs who lived outside of SDP control.128 Within the next two months Serbian guerrilla groups had carried out massive atrocities, expelling and murdering non-Serbs, mostly in areas already controlled by the SDS.129 In the Bosnian conflict, however, local Serbs were used in the fighting rather than Serbs from Serbia. By September 1992 Belgrade's Bosnian Serb allies had increased their territorial holdings by less than ten percent, to about 70 percent. As in Krajina, almost the entire non-Serbian population has been killed or driven out, and Serbian dissenters have been silenced and repressed.130

The Serbian conservatives' strategy had a short-term goal of survival in power and preserving the structure of economic and political power in Serbia. In the long term, their strategy initially had the goal of creating a centralized, authoritarian Yugoslavia where the conservatives would crush all attempts at radical change and enforce their own orthodoxy. But when in 1990 the bases of political power shifted to the wider population, the conservatives were forced to change this strategy. Having discredited themselves in the eyes of the 61 percent of Yugoslavia's population that is non-Serb, they resorted to destroying the old multiethnic Yugoslavia and creating on its ruins a new enlarged Serbian state with a large majority of Serbs in which they could use appeals to Serbian nationalism as a means of defining political interests, thereby preserving the existing power structure.131 The violence itself and the retaliatory violence against innocent Serb civilians in Bosnia and Croatia has created a situation in which grievances defined in ethnic terms are sure to continue to play an important role in Serbian politics. Meanwhile, in the conditions of international economic sanctions, the regime has taken firmer control of the economy.

Conclusion: Ethnic Conflict as a Political Strategy
Violent conflict described and justified in terms of ethnic solidarity is not an automatic outgrowth of ethnic identity, or even of ethnic mobilization. Violence on a scale large enough to affect international security is the result of purposeful and strategic policies rather than being driven by irrational masses. Indeed, in this case there is much evidence that the "masses," especially in ethnically-mixed regions, did not want war and that violence was imposed by forces from outside.
If such conflict is driven by domestic concerns, one way outside actors can try to prevent or moderate it is by making the external costs of such conflict so high that the conflict itself would endanger the domestic power structure. The most obvious way is the use of military force. But to prevent such conflicts, the threat of force must be made early, and it must be credible. In the Yugoslav case the international community has been incapable of fulfilling either condition.
Another way to try to modify or prevent this type of conflict is to influence the situation from within, to strike at the root cause of the conflictual behavior. Assuring minorities of their rights may of course be important; but that alone does not address the roots of the conflict in cases such as this one. Rather, the target must be the real causes of conflictual policy. This too must be done early, but it is much less costly than a military solution. The international community can undertake policies such as ensuring multiple sources of mass information and active and early support for democratic forces. But in cases where domestic structural changes are being fostered by international actors, those actors must also be very attentive to the domestic political context into which they are intervening, and in particular should take into account the concerns of those who are most negatively affected by domestic changes. An example is to ensure those elites most affected by change of fall-back positions.
As for negotiations, if violence along ethnic lines is caused by internal conflict, then talks over interests outside the domestic arena will be without effect, since the goal of the conflict is not in the international environment, vis-à-vis another state, but rather at home. To be truly effective these internal factors must also be brought into negotiations.

What are the implications of this approach for understanding the link between nationalism and violent conflict in other parts of the world? If domestic conflict drives external conflict, and if the potential costs in the outside world are a key part of the domestic calculus, then we would expect such types of external conflict to be less likely in a truly threatening international environment. If the risk of provoking high costs from outside is too high, threatened elites have more motivation to seek a compromise solution with challengers at home. On the other hand, in conditions where the external threat to security is minimal, threatened elites may be more tempted to use conflict in the external arena as one part of their domestic political strategy. In effect, the end of the Cold War may therefore have its primary effects not in the international arena, but rather in domestic spheres around the world.

What about absence of conflict? In the Russian case, Gorbachev's evolutionary style of incremental reform, where he brought conservatives step-by-step toward radical change, was one factor preventing a feeling of sudden threat among conservatives. Since then economic change has taken place gradually in Russia, and often the new owners of privatized enterprises are the former managers and party bureaucrats. Although this gives them a stake in the new system, if these firms are unprofitable or poorly run, a rash of bankruptcies may have a drastic effect. In addition, in the Soviet Union and then in Russia, because reformists are in control of the central government they also control the media, making it very difficult for hard-liners to create images via television. Extreme Russian nationalist Vladimir Zhirinovskii, who made a surprisingly strong showing in Russia's 1994 elections and whose expansionist rhetoric alarms many in Russia's 'near abroad,' is another example. His success is due to the fact that his rhetoric serves the domestic political purposes of threatened elites in the security forces as well as forcibly pensioned party workers and others. It is thus no coincidence that like Milosevic, Zhirinovskii and other extreme nationalist Russians speak in terms of threats to Russians outside of Russia; any conflict will thus most likely be outside of the Russian Federation's borders.

Methodologically, this case shows the importance of recognizing that political rhetoric is itself political behavior, and that a conflict described in ethnic terms and taking place along ethnic lines, while it may be about ethnic issues, may be caused by issues not related to ethnicity. The ability of violence to create specific political contexts means that those provoking violent conflict may have as their goal something quite outside the direct objects of conflict. It is thus important to realize that the rhetoric of ethnic nationalist purists is exactly that: rhetoric. Within every group the definition of group interest is contested, and in fact that definition is the key to power.