Aggregations are wholes, yet not wholes—brought together,
yet carried asunder; in accord, yet not in accord;
[in] all, one; [in] one, all.

HERACLITUS

Yugoslavia from the devolutions of the early 1960s until 1989 could be seen as a nine-actor balance-of-power system that consisted of a federal actor (the federal government or, alternatively, the LCY), six socialist republics (Bosnia-Herzegovina, Croatia, Macedonia, Montenegro, Serbia, and Slovenia), and two socialist autonomous provinces (Kosovo and Vojvodina) that, while within the socialist republic of Serbia, were, in fact, autonomous actors. Within this universe, regional demands were aggregated along ethnic lines and articulated by republican and provincial authorities.

In a multiethnic state, diverse social problems also manifest themselves as interethnic problems. The involvement of ethnicity and the possibility for the mobilization of group loyalty and group resources transmogrify political processes. Social antagonisms are expressed differently in a multiethnic state from the way in which they are expressed in an ethnically homogeneous or binational state. Politics in a multiethnic state is essentially different from politics in a nation-state. Multiethnicity becomes the justification for introducing a federal system in which ethnic and republic boundaries coincide. Middle-level interest groups tend to follow the lead of their republican leaders during times of intercommunal crisis, since intercommunal stability is, to a large extent, hinged on the viability of the federal system as a mechanism of conflict regulation and crisis management.

Ethnically, Yugoslavia is one of the most heterogeneous polities in the world (see table 3). The largest group, the Serbs, makes up less than 40 percent of the population, and the second largest group, the Croats, represents roughly 20 percent of the total population. Although the three largest groups share a common language (known, alternatively, as Serbo-Croatian or Croato-Serbian), roughly one-fifth of all Yugoslavs are not native speakers of Serbo-Croatian, and many never learn the language. (This is especially true in the rural areas of the non-Serbo-Croatian-speaking federal units.) Yugoslavia also contains pockets of national minorities from all of its neighbors, although the numbers of Austrians and Greeks are minuscule.

Table 3. Population of Yugoslavia by Ethnic Group, 1981

Source: Statistički kalendar Jugoslavije 1982 (Belgrade: Savenzni zavod za statistiku, February 1982), 28:37.

Yugoslavia had some 22, 418, 331 inhabitants as of March 31, 1981, with Serbia, Croatia, and Bosnia-Herzegovina the most populous units. Although Bosnia and Vojvodina are both highly heterogeneous, their heterogeneity tends, with the exception of the largest towns, to be pluralist—that is, the two federal units are ethnically mixed, but most of their villages and smaller towns are ethnically homogeneous. In all of the other federal units, the titular nationality constitutes at least two-thirds of the population (in four of these, the titular nationality accounts for at least 75 percent of the population). For this reason, republican policy continues to be identified (in every federal unit except, arguably, Bosnia) with the interests of the principal ethnic group. This did not apply to Kosovo until the 1970s, but after the riots in 1968 and 1969, positions of authority in Kosovo were steadily relinquished by Serbs and Montenegrins and turned over to Albanians. As table 4 makes clear, all the federal units, except for Slovenia and Serbia, are relatively heterogeneous. In Bosnia, Vojvodina, and Macedonia no ethnic group composes more than two-thirds of the population. In Croatia, moreover, the Serbian minority (11.6 per cent in 1981) played a pivotal role in exacerbating interethnic tensions in 1970-71.

Table 4. Population of Yugoslavia by Republic, 1981

Sources: Statistički kalendar Jugoslavije 1982 (Belgrade: Savenzni zavod za statistiku, February 1982), 28:33; and author’s calculations from NIN, no. 1626, February 28, 1982, pp. 1920.

Ethnicity and Ethnocentrism

Today ethnicity is usually understood in two senses—as the objective characteristics that distinguish and define a group of people (common language, consciousness of being a group, sense of territorial homeland, and, sometimes, common religion) and also as the subjective response to the fact of collective identity, that is, as the pure phenomenon of group consciousness. Collective identity itself signifies the formation of an interest group in which the preservation of group characteristics and the safeguarding of territorial lines become invested with emotive value.

Ethnicity need not be accompanied by ethnocentrism. It is the presence of a perceived cultural threat—intensified by ethnic mixing—that gives rise to ethnocentrism. Ethnocentrism may be defined here as an attitudinal orientation that inclines one to condemn the culture, language, and customs of another ethnic group as inferior and “wrong.” Ethnocentrism is reinforced by stereotypes both of one’s own group and of other groups.

Every society has a degree of ethnocentrism, but the intensity may vary widely among groups within a single society. Religion plays an important role in nurturing ethnocentrism, and certain religious groups, such as Muslims, have shown a marked predisposition toward ethnocentrism. The dramatically lower incidence of intermarriage between members of Yugoslavia’s Muslim community and members of other ethnoreligious communities may indicate a greater ethnocentrism rating for that community.¹

Peter Klinar, in a 1971 study, found that Yugoslavia’s national groups display a strong proclivity to engage in ethnic stereotyping. The Croatian self-image is, of course, a flattering one and emphasizes the Croats’ love of justice, the peace-loving nature of the Croatian nation, its thousand-year-old culture, and Croatia’s “Westernness.’’ The Serbian self-image focuses on the heroic character of the Serbs and paints the Serbian nation as the guardian of Yugoslavia. The Croats and Serbs do not share such high regard for each other, however. The Croats tend to view the Serbs as expansionistic and arrogant, and the Serbs portray the Croats as passive, timid to the point of cowardice, and inclined to collaborate with (foreign) subversive elements. The Slovenes’ self-image contains some feelings of superiority (love of order, efficiency at work, and cleanliness) and of inferiority (principally in connection with their lack of a historical tradition of independent statehood). The Slovenes also have a tendency to look down on other Yugoslavs for their inefficiency and allegedly irrational use of resources. The Slovenes themselves are viewed by other Yugoslavs as unsociable, unfriendly, “Germans. “²

The Yugoslavs realized, even before Milosevic, that ethnocentrism remained a problem for them. Stipe Šuvar, criticizing ethnocentrism among Croats, noted that during 1976-77, Croatian theaters did not present a single work written by a dramatist from another republic, and only one book was translated into Croatian from the language of another Yugoslav nationality.³ Ethnocentrism is deeply ingrained and is reflected in a panoply of variables. For instance, there is a strong tendency for marriages to involve members of the same ethnic or language group (for example, Serbs often marry Croats). Those most likely to intermarry with members of other ethnic groups are members of the smaller minorities (particularly Italians, Ruthenes, and Bulgarians).⁴ Ethnocentrism may also lead a group to change its customs and behavior, accentuating or even creating differences to distinguish itself from a group with which it has confrontations. Šuvar wrote in 1978 that “a number of ignorant Serbs in Croatia have [recently] started to speak ekavian [spoken by the Serbs in Serbia], even though their mothers had taught them ijekavian,” the other principal variant of Serbo-Croatian.⁵

Yugoslavia’s brief flirtation with multicandidate elections confirmed the prominence of ethnocentrism in political choice. In local elections held in Derventa during July and August of 1968, for instance, voters overwhelmingly cast their ballots along ethnic lines—Serbs supporting Serbs, Croats voting for Croats, and Muslims for Muslims. As might be expected, however, very few (about 3 percent), when asked about their voting patterns, would admit that ethnicity had anything to do with their choices.⁶

Nationalism

Early theorists of modernization, especially Marxists, assumed that nationalism was obsolescent and moribund. The advancing integration of world culture was expected to result in an erosion of ethnic consciousness, and nationalism as a divisive force was often rated as secondary in importance to class conflicts and the urban/rural cleavage. But ethnicity, not ideology or class, appears to provide the ultimate source of political identity in intergroup conflict.

Yugoslavia’s Marxists have gradually accepted the permanence of the ethnic factor and now concede that, far from witnessing the unraveling of national consciousness, the contemporary period has seen the deepening of ethnic loyalties and a retrenchment among nationality groups.⁷ Some have gone even further, admitting that neither the content nor the role of a nation or nationality group is at all changed by socialist revolution.⁸

Nationalism can assume many forms within a multiethnic state. It may foster separatism, stimulate unitarism, provide a transregional focus of loyalty, or drive the component ethnic groups into civil strife. During the late eighteenth and nineteeth centuries, the nationalist awakening spread to all groups in the Habsburg empire except the Ruthenes, yet each group reacted in a different fashion. The Pan-Germans dreamt of union with the rest of the German nation. The Italians worked actively for secession. The Czechs strove for federalist concessions and attempted to move all of Austria in a federalist direction. The Poles enjoyed their autonomy, oppressed the Ruthenes, and cautiously waited, fearful that a miscalculation would undo the privileges they enjoyed and catapult them into the culturally and politically depressed condition of the Russian and Prussian Poles. The Croats worked calmly for autonomy. The Magyars, for their part, were bent on the Magyarization of their subjects and the revival of the Kingdom of St. Stephen. Nationalism clearly is not a constant that can be plugged into any equation with predictable effects. On the contrary, it is a variable whose influence may be critical at some times and under certain circumstances, and only marginally significant in other times and conditions.

Following E. K. Francis, nationalism will be understood here as collective affectivity “extolling the nation as a supreme value and representing it as a dominant principle of societal organization, “⁹ that is, as politicized ethnicity. Obviously, nationalism is potentially threatening to a multinational state. Hence, elites in multinational states have often tried to create transnational bases for collective consciousness. Soviet leaders were long fond of speaking of the “Soviet” people. The principle of Kaisertreue served a similar function in the Habsburg empire. In postwar Europe it was hoped that institutional cooperation would promote European unity, but it was also apprehended that nationalism was a divisive element that had to be overcome if support for a European state was to burgeon. Similarly, in Yugoslavia the Communist party originally hoped to erase all ethnic attachments, not only to the groups as they are currently defined—Serbs, Croats, Slovenes, and so forth—but also to the South Slav conglomerate.

But Serbia easily relapsed into the role of “big brother” in postwar Yugoslavia. Despite Stipe Šuvar’s assurance that “the Serbian people . . . has outgrown nationalism and is characterized by general internationalism,”¹⁰ non-Serbs felt threatened, not only because of memories of the Great Serbianism of the interwar period but also because assimilation of any kind, whether to a Serbian or to a Yugoslav model, was abhorrent both to culture-conscious elements and to regional politicians who had a stake in federalism. Yet the Serbs have felt comfortable in their self-appointed role. That Serbian chauvinism lives is not only readily verifiable, but was also admitted by politicians in the Serbian League of Communists (LC). Thus, Vladimir Jovičič, a member of the executive committee of the central committee of the Serbian LC, told the weekly news magazine NIN in 1978 that Serbian chauvinism, existing “beyond rationality,” remained a serious problem.¹¹ A high proportion of the lower- and middleranking civil servants in the federal government were, naturally, drawn from the Serbian community in the area surrounding Belgrade. Yet, whatever the reasons, the proliferation of Serbs in the federal government added to the wariness of a non-Serb population that has learned to identify the Belgrade regime with the interests of the Serbian nation. Indeed, it was in the hope of breaking this identification that the proposal was entertained, in the immediate postwar years, to move the federal capital to Sarajevo. The dismissal of Aleksandar Ranković from his post as chief of security in July 1966 and the subsequent purge of the secret police represented a decisive dissociation of the federal government from Serbian interests, and the crisis was generally viewed as a defeat for the Serbs. Later, in the wake of the Croatian crisis of 1971, Tito purged not only the Croatian party but also the Serbian party: such high-ranking Serbian liberals as Marko Nikezić and Latinka Perović were among the heavy casualties suffered by the Serbian party at that time.

Two questions may be raised, then, about the function of nationalism in Yugoslavia: (1) Was nationalism perceived as a problem by the LCY? and (2) Was nationalism the guiding impulse in republican policies, economic or otherwise? The most radical interpretation would answer both questions in the affirmative and conclude, as Milovan Djilas did in 1974, that Yugoslavia might not survive long beyond 2024.¹² In the most extreme view, nationalism is considered so intense and so rampant in Yugoslavia that the unity of the state is preserved only by authoritarian means and Yugoslavia is believed to be threatened by “the eventual anarchy of unstructured conflict as a result of primordial distrust.”¹³

Perhaps no one will dispute the importance of nationalism on the regime’s agenda. Public life regularly reverberates with its pulses. In the summer of 1978 a wave of Croatian nationalism forced a number of concessions in the selection of plays for the Split Summer Festival, for instance. Slobodna Dalmacija singled out Soljan’s play, “A Romance about Three Loves,” as a well-known but mediocre piece of nationalist drama. Evidently, attempts were also made to include plays by completely unknown authors in the program, simply because they were nationalistic.¹⁴ In the early 1980s, the regime made clear its fear of nationalistic manifestations and its determination not to let matters get out of hand. The LCY took regional nationalism most seriously. In February 1975, for instance, fifteen Croats were sentenced to terms of up to twelve years in jail for having formed a terrorist organization designed to bring about the secession of Croatia from Yugoslavia.¹⁵ Albanian nationalists were also rounded up and jailed in the course of 1975 and again in 1979, and in December 1975 Croatian separatist Miljenko Hrkac was sentenced to death for having planted a bomb in a Belgrade theater in July 1968.¹⁶ It could be argued that in these cases it was the violence employed or the advocacy of secession, rather than nationalism, that provoked sharp government reaction. Yet, in other instances, the LCY displayed a hostility to expressions of nationalism that stopped short of these extremes. In 1974, for instance, the semiofficial newspaper Oslobodjenje warned that Muslim nationalism was neither naive nor harmless and attacked the importance accorded Islam.¹⁷ Again, in February 1982, a Bosnian Muslim received a two-year prison sentence for having publicly claimed that there was ethnic discrimination in Bosnia and Kosovo.¹⁸ Nor have the Serbs been immune from antinationalist thrusts. On December 5, 1977, for instance, the LCY weekly, Komunist, upbraided the prominent Serbian author, Dobrica Ćosić, for insisting that the Serbs were being exploited and denigrated by the other Yugoslav nationalities and for making other inflammatory nationalistic statements.¹⁹

Nationalism is a very real phenomenon in Yugoslavia and animates much of social life. Tensions have reached the boiling point in nearly every republic in recent years. Typically, sporting events provide a ready spark for outbursts of ethnic hatred, and, according to NIN, Serb-Croat competitions such as bas ketball games and soccer matches are still frequently marred by outbursts of ethnic slurs and name calling. In Split, after one such sports match, young Croats pushed several cars bearing Belgrade license plates into the sea.²⁰ Even the old Serbian military and patriotic songs have resurfaced—beginning in 1965—although they have been banned since 1945. Surprisingly, as late as March 1966, Vladimir Bakarić, then president of the League of Communists of Croatia, seemed uncertain as to how important nationalism was and, at one point in the course of a long interview, blithely dismissed the importance of nationalism. “The “national question,’ “ he told Borba, “plays a very small role in the ordinary world. The youth are not very interested. And I think it can be said that nationalism, as a mass phenomenon, does not exist, and that it appears in the first place among the bureaucrats, [and] among the intelligentsia in our elites.”²¹ A few years later, however, a disabused Bakarić confessed that under socialism Yugoslavia had witnessed a reanimation of ethnic consciousness and ethnonationalism.²²

By the late 1960s, nationalism had once again become the central question in Yugoslav politics, and it remains of fundamental importance today. I had a chance to see this for myself while visiting Yugoslavia for the first time in August and September of 1978. I recall in particular one conversation that took place when I was visiting the island of Pag, off the Dalmatian coast. A middle-aged Croat heatedly insisted that Serbian and Croatian were “two completely different languages” and that “there is no such thing as Serbo-Croatian.” The idea that a self-respecting Croat should admit to having something as basic as language in common with the Serbs was a complete anathema to him.

However vital nationalism may be, and however important in molding people’s attitudes, whether it is the guiding impulse in the formulation of the republics’ policies or plays only a pivotal role in setting their context is an entirely different question. Certainly the feeling of relative deprivation had become strong among many Croats, including Croatian leaders and intellectuals, by 1971; moreover, ethnic prejudice impeded not only interrepublican migration but also interregional circulation of investment capital—hurting resource-rich, capital-poor regions in the south.²³ This has been a serious problem since, as William Zimmerman has noted, “some parts of Yugoslavia—Kosovo and Montenegro in particular—are barely integrated into [the] Yugoslav market.”²⁴

In multinational states, not only interregional rivalry but also “regional opposition against centralized political power tends to be expressed and mobilized in ethnic terms.”²⁵ Regional leaders in Yugoslavia have frequently adopted this tactic, even though economic and ethnic interests are not always equally well served by the same solutions. Regional elites deliberately play the role of national (i.e., ethnic) leaders and present issues in ethnic terms: Croatian leaders have done this vis-à-vis the Serbs, Kosovar-Albanian leaders vis-à-vis the Serbs, Macedonian party leaders vis-à-vis the Serbs, and Serbian leaders vis-à-vis almost everyone else. But Yugoslav politics is not reducible to Serb versus nonSerb, as we shall see in the course of this study. The Serbs, for instance, share many common interests with the Montenegrins as well as with the Macedonians, and they have repeatedly courted the Bosnian party as a natural ally. Nor would it be accurate to leave the impression that non-Serbs constitute a “camp.” Croats, for instance, became (at one time) increasingly worried about the steady Slovene infiltration of the northwest corner of Croatia’s Istrian peninsula, along the border with Slovenia. This was, at least in part, an economic issue, but the apprehension was ethnic in source and was voiced in ethnic terms.

Thus, Yugoslavia’s federal system, like the international state system, fixed boundaries, oriented the communication system, provided the pattern for economic life, restricted and channeled population movement, and set the parameters for political life and political conflict. The importance of the institutional framework must not be underestimated, since it can both reinforce and muffle inherent lines of conflict.

In this multipolar context there will be instances in which the nature of the issues will be unintelligible except in the light of nationalism. But, in this connection, it is important to remember that cultural and communications elites may have a vested interest in fueling cultural nationalism. Djilas himself denied that national feeling was the basis for interrepublican rivalry; rather, he felt that the elites manipulated national feelings to maximize their power. Yet five of the six republics (Serbia, Croatia, Slovenia, Macedonia, and Montenegro) were created around their regional ethnic majorities, and the federal system itself was a creature of the multinational configuration of Yugoslavia. Thus, whether primary or secondary, ethnicity is wedded to regionalism in Yugoslavia.

Cleavage in a Multiethnic State

The contours of politics are often defined or molded by lines of cleavage that lay the basis or set the framework for conflict behavior and accommodation. Cleavage may be economic, cultural, religious, linguistic, racial, or regional. It may be appropriate here to register an important caveat. Kaplan has warned that no pattern of international politics is permanent. Comparative power levels fluctuate, cultural affinities lose or gain importance, and ethnic conflict, which may be the crucial factor at one point, may be mere window dressing at another point. Even “the definition of groups is in constant flux. “²⁶ Thus, for instance, the Macedonian nationality, which never enjoyed recognition under the interwar government and did not enjoy Communist party recognition until November 1942, obtained republican status at the end of World War II and has been the object of serious courting by a Bulgarophobic Belgrade. Again, the Muslims, once regarded strictly as a religious group by the regime, later came into their own as a major Yugoslav nationality group.

In 1971, R. V. Burks argued that Yugoslavia was characterized by the polarization of northern provinces, which enjoyed economies comparable to Austria, and southern provinces, whose economies were similar to those of Albania or southern Italy. Burks associated the deep conflict of regional interests with differences in the levels and rates of development, among other things, and defined the north as Slovenia, Croatia-Slavonia, Vojvodina, and Belgrade, and the south as Dalmatia, Bosnia-Herzegovina, Serbia proper, Montenegro, Kosovo, Macedonia, and Croatia south of the Sava River.²⁷ After the war, however, industrial production in Montenegro grew at a rate that far eclipsed growth rates in other republics. Montenegro has modern airports at Titograd and Tivat; Titograd has now been connected by railway both with Nikšić and with Bar. The economy of Dalmatia, moreover, was so completely transformed in the 1970s, mostly as a result of the construction of the Adriatic highway and the development of the tourist industry there, that it came to be considered part of the developed north.²⁸ Yet, despite these changes, it can be argued that the basic problem in Yugoslavia remained the same, namely, how does a developed industrial region coexist in the same polity with an underdeveloped region that desires further industrialization? The dividing line between rich and poor still corresponds almost exactly to the old military frontier that once divided the Habsburg and Ottoman empires.

Politically, the so-called northerners have tended to be liberal, partial to decentralization, and opposed to programs that have been geared to the redistribution of resources, whereas the southerners have tended to be conservative, partial to centralized disbursement of investment funds, and hostile to anything smacking of laissez-faire economics. The Tito regime expressly recognized the cleavage and, for redistribution purposes, designated certain areas as “underdeveloped.”

The differences in social product are reflected in table 5 and regional differences in net personal income are shown in table 6. Given the tremendous disparity in wealth, it is perhaps not surprising that conflict over policy directions has flared up regularly. But the Serb-Croat rivalry that (for reasons a cursory examination of table 3 should make clear) remains the pivot of ethnic competition in Yugoslavia is widely viewed by both Serbs and Croats in religious terms. The identification of the churches with nationalism goes back to the time of the Ottoman occupation, when, under the millet system, the Orthodox churches were the guardians of national culture as well as the political viceroys. In independent Montenegro, too, the Orthodox Church arrogated to itself the role of protector of the Montenegrin nation. At the same time, the Catholic Church in Croatia, which had been absorbed into the Habsburg empire, maintained a political role. Thus, for example, when Orthodox Serbs in Dalmatia formed a political party in 1879, the Catholic Serbs replied by forming a movement of their own under the leadership of Baron Frano Gundulić of Dubrovnik. The chief ideologue of this movement was Ivo Stojanovié, a Catholic priest. More significant, when the exclusionist Croatian Bights party of Ante Starčević began to expand its activities into Bosnia and Herzegovina in the 1880s, its leading proponents in the region were none other than the Franciscan friars of Herzegovina, who were eager to transform Bosnian Muslims into Croatian Catholics. Nor should one fail to mention the church’s involvement in the creation of the Croatian Agricultural Bank in 1900, whose raison d’etre was resistance to the encroachment of Hungarian capital. Subsequently, in World War II, the Croatian-fascist government of Ante Pavelić found a ready ally in the Catholic Church. During the war, priests on both sides of the barricades played active roles in fomenting the mutual slaughter of Serbs and Croats. And in the 1980s, communist officials would complain that Croatianness tends to be identified with Catholicity and Serbianness with Orthodoxy. Contrary to what one would expect in a modernizing society, namely, secularization, Yugoslavia experienced a waxing xenophobia and a recrudescence of religious sentiment among the young in the late 1960s, especially in Slovenia and Croatia.²⁹ That this tumescent affectivity was associated with a reassertion of nationalist feelings and was, in both instances, centered in Slovenia and Croatia, underlines the closeness of the ethnic-religious relationship.

Table 5. Regional Economic Disparities in Yugoslavia, 1990 Five measures of relative advantage

Source: Danas, no. 459 (4 December 1990), p. 18.

Conflict of Interest

So deep were the divisions in Yugoslav society in 1945 that the Communist party had little with which to hold the country together except the partisan myth, promises of a future cornucopia, and coercive force. The break with the Cominform certainly helped in this respect, for even anticommunists rallied to Tito rather than risk Sovietization. But the promises remained important. The regime had pledged, for example, to intensify industrialization—a clearly perceived need—and to level interregional economic disparities. This was to be done by funneling a disproportionate amount of new investments into the poorer regions of the south.

This early period was, of course, also the period of the Stalinist model— with or without Stalin—and the decision to emphasize heavy industry naturally gave preference to the southern regions that were rich in such basic raw materials as minerals and energy sources. Kosovo is rich in minerals, with its mines at Trepča alone producing 25 percent of Europe’s lead and 13 percent of Europe’s zinc.³⁰ Bosnia-Herzegovina has important reserves of coal, iron ore, and lumber, and sizeable potential sources of hydroelectric power. Thus, it made good economic sense to locate heavy industry in the south. But, as Kardelj himself noted in 1960, the overall effect of this was “to contribute indirectly more to the rise in national income in other regions with developed manufacturing by supplying them with raw materials from less developed areas. “³¹ Slovenia, Croatia, and the Belgrade environs—and, to a lesser extent, Vojvodina³²—continued to be preferentially favored with the development of light and consumer industries. Understandably, politicians in the underdeveloped republics saw parallels to the classic model of imperialism. In addition, income differentials were widening. Macedonian per capita income, for instance, dipped from 31 percent below the national average in 1947 to 36 percent below the average in 1963, while, during the same period, the average per capita income in Slovenia soared from 62 percent above the national average to almost twice the national average.³³

Yet the more advanced republics were also dissatisfied. In addition to investments in raw material extraction and processing, the south was also the recipient of certain investments demonstratively made on political rather than economic grounds; Croatia and Slovenia were particularly irked by the duplication of industries already flourishing in their own republics. Furthermore, these new industries often proved glaringly inefficient. Thus, with the shift of the center of steel production from Slovenia to Bosnia, it became cheaper to import steel than to produce it in Yugoslavia. The market reform of 1965 “revealed that 600,000 industrial workers, nearly half the industrial labor force, were employed in enterprises that operated at a loss.”³⁴ Croatia’s share in the total Yugoslav industrial potential dropped from 33 percent in 1925 to 29 percent in 1946 and plunged to 19 percent in 1965.³⁵ Of course, since Croatia’s industry was expanding during this period and since percentages reflect a relationship to a greater whole, these figures do not necessarily mean that Croatia was being “oppressed”; the important point is that many people thought the figures did mean just that. Croatian scholars like Šime Djodan and Marko Veselica lashed out at the alleged exploitation of Croatia (and Slovenia). Djodan, who was later imprisoned, often hinted at secession. The Croats did not deny that Croatia had made progress under communism, but they adamantly maintained that progress would have been far greater had the Croatian economy not been milked to succor the less developed republics.

Table 6. Net Personal Income by Republic (In new dinars)

Source: Statistički Codišnjak 1979 (Belgrade, July 1979), p. 431.

As early as 1957, Bosnia’s share in industrial investment began to slip as investment in heavy industry (and, more particularly, in raw material extraction) reached the saturation point. Bosnia’s share of federal investment funds had declined from 19.4 percent under the first five-year plan to 12.6 percent under the second. Serbia’s share, meantime, jumped from 31.3 percent to 41.7 percent—the only increase among the six republics. Serbia was now garnering between 50 and 60 percent of all new investments in metalwork, machinebuilding, nonferrous metallurgy, and chemical industries. These facts convinced Joseph Bombelles that it was no longer possible to “speak of a conscious government policy of aiding underdeveloped republics after 1956. Rather, the western republics were now supposed to subsidize the eastern republics without regard to the level of development.”³⁶ The entire system of investments was overhauled in the 1965 reforms (to be treated in detail in chapter 6), and profitability became the crucial criterion. The central government ceased to channel funds to the south under a planned system, and, despite continued payments of subsidies to certain industries and the invocation of “special case” tax consideration for investments in the south, the reform generally favored the developed regions.

Table 7. Relative Importance of Heavy and Light Industry, 1977 (In percent)

Source: Momir Ćećez, “O efikasnosti društvenih sredstava u privredno nedovoljno razvijenim područjima,” Pregled 70 (1) (January 1980): 34-35.

It is certainly not insignificant that the four most prosperous republics are precisely those that have the highest proportions of light industry, whether this is measured, on the one hand, in terms of contribution to the social product or, on the other, in terms of employment (see table 7). The two federal units most reliant on heavy industry are the two poorest units, Kosovo and Montenegro. Clearly this complementarity might furnish the basis for identical policy aims. Yet a more detailed breakdown of the structure of industry (table 8), although it omits certain branches of light industry, fails to reveal any obvious axes of potential cooperation or competition.

Conflict Behavior

Despite the hypotheses originally set forth, we have seen that Yugoslavia’s regions can be divided into “developed” and “underdeveloped” categories. If this rift is the basis for the polarization of the political system, Yugoslavia ought to be analyzed as a loose bipolar, not a balance-of-power, system. The old Serbian-Croatian friction has not become suddenly irrelevant either. The SerbCroat civil war (1941-45) provides support for viewing pre-1965 Yugoslav politics in bipolar terms. War, the great polarizer, did in fact drive Serbs and Croats further apart. At the same time, however, there are reasons for thinking that the bipolar model is not relevant to the 1965-89 period. One might note, among other things, that the north-south divide does not explain enough to warrant the term bipolarity and that there are enough crossovers to justify construing Yugoslavia as a system characterized by flexible coalitions among its eight federal units.

Table 8. Structure of Industry as Percent of Social Product, 1977 (In percent)

Source: Momir Ćećez, “O efikasnosti društvenih sredstava u privredno nedovoljno razvijenim područjima,” Pregled 70 (1) (January 1980): 32-33.

William A. Gamson, in “A Theory of Coalition Formation,” outlined four conditions required for the flexible formation of coalitions: (1) there are issues to be resolved by three or more interested parties, each of whom stands to lose if the outcome assumes a given configuration; (2) no single possible outcome would maximize the payoff to all actors; (3) no single actor has the resources to impose a decision; and (4) no participant enjoys a veto power that would require that it be included in any winning coalition.³⁷ I would argue that these conditions were met in communist Yugoslavia and that, because of the degree of decentralism, republican leaders were in the position to strike bargains on national legislation.

On the basis of the north-south cleavage and the information provided in tables 5 and 6, we might graphically depict Yugoslavia as a four-box set:

Figure 1

The upper left-hand box contains those actors that, during most of their history, had formed a part of the Austrian empire. The lower left-hand box contains actors liberated from Ottoman rule in the nineteenth century. Bosnia-Herzegovina was placed under Austrian administration in 1878, the same year in which Montenegro achieved its independence from Istanbul. The lower right-hand box contains actors that remained under Ottoman rule until 1912: members of this set, as suggested by appendix 3, showed the most marked and consistent pattern of economic retrogression as measured by total investment in fixed assets. The upper right-hand box is less homogeneous: Serbia won its independence in 1878, having enjoyed internal autonomy since the early part of the nineteenth century, whereas Vojvodina constituted an integral part of the Kingdom of Hungary until the Peace of Trianon. These historical differences were reinforced by systematic geographic variations. Moreover, differences in political history also made for differences in economic development, cultural life, work ethic, and religious traditions.

The northerners in the upper half of the box—Slovenia, Croatia, Vojvodina, and Serbia—were the more developed regions and had the highest social products and per capita incomes. The southerners in the lower half of the box— Bosnia-Herzegovina, Montenegro, Kosovo, and Macedonia—were officially designated “underdeveloped areas” after 1965. Between 1956 and 1964, however, it appeared that the notion of underdevelopment was being manipulated by Serbian elements so that, in Bombelles’s words, “the policy of aid to underdeveloped republics was changed to a policy of aiding [the] eastern republics.”³⁸ Indeed, between 1961 and 1964 the republic of Bosnia-Herzegovina was considered “developed” for funding purposes, while portions of Serbia proper were redesignated as “underdeveloped” (albeit along with sections of Slavonia). During this period frequent complaints were heard about Serbian banking and industrial monopolies, especially vis-à-vis the concentration of corporations in Belgrade and its vicinity. Thus, an alternative, at least for the 1956-64 period, is to view the cleavage as an east-west split, with the republic of Serbia (including Vojvodina and Kosovo), Macedonia, and possibly Montenegro—the easterners—exploiting the westerners—Croatia, Slovenia, and Bosnia-Herzegovina.

The diagram in figure 1 predicts that policy alignments that cross diagonally should be rare. That is, if the system is bipolar and not balance of power, we will not expect to find Slovenia and Kosovo aligned, say, against Serbia and Montenegro. What we find, however, is exactly the opposite. We find flexible, ephemeral alignments; a tendency for republics to align in opposition to whichever federal unit is seen to threaten the stability of the system (whether Serbia, Croatia, or even Kosovo); and alignments that confute any simple predictions on the basis of cleavages.

This is not surprising in the least. After all, by May 1940 the CPY had denounced the Sporazum, under which Croatia had finally gained a degree of administrative and political autonomy in the old kingdom. The party depicted the agreement as a compact between Serbian and Croatian bourgeoisies made at the expense not only of the working classes but also of the other republics. Although the partisans probably never intended to go that far, the adoption of a federal constitution was an unmistakable concession (a) to nationalist sentiment and (b) to the balance-of-power principle. In this vein, the recognition of Montenegrins as a distinct nationality in the 1946 constitution makes sense from the perspective of balance-of-power politics. “It has been suggested that the creation of a Montenegrin republic, rather than the inclusion of Montenegro in Serbia, was intended to allay fears of Serbian hegemony in other republics.”³⁹ With Montenegro incorporated into Serbia, the Serbian republic would have been given a superordinate position in the new federation, but, established as a separate republic, Montenegro was available as an alignment partner to all other actors, even though the close cultural affinity and commonly shared economic interests continued to incline Montenegro to align, virtually automatically, with Serbia. Another factor, however, was at work—the postwar reemergence of the anti-Serbian faction of the Montenegrin elite. The CPY took advantage of this current to strengthen Montenegro as an independent republic.

Conflict Accommodation

As noted in chapter 1, the Tito regime (1945-80) was the ultimate arbiter in this system and, thus, was both the realm in which conflict accommodation most often took place and the chief agent for such accommodation. After Tito’s death, the federal center started to decay—at first slowly, but after 1987, rapidly. By 1989, the federal government was no longer capable of playing the role of rule-enforcer, and Serbia’s Miloševic stepped into the vacuum.

Tito had come to believe in decentralization, and he generally preferred to let the republics govern themselves, and, under Tito’s immediate successors, the federal government relinquished its position of preponderance and participated in interrepublican negotiations, if at all, as an equal (or, perhaps, as primus inter pares). The practical import is that conflict accommodation is accomplished by the parties in the conflict. This has certain drawbacks in terms of speed. It took the nine negotiating parties (the eight federal units and a federal deputy) a full year to agree on the level of aid to be provided to Montenegro for damage caused by the April 1979 earthquake. And the debate on the establishment of new criteria for the disbursement of resources from the Fund for the Accelerated Development of the Underdeveloped Republics and Kosovo (FADURK) missed two deadlines, forcing the parties to operate the program, temporarily at least, in the absence of formal criteria. Again, the debate on the balance-of-payments deficit, which was supposed to wind up in December 1979, dragged on to July 1980. Yet the process of interrepublican conflict accommodation expressed in the Serbo-Croatian words dogovaranje (literally “coming to an agreement” or, as I have translated it, “negotiation”) and usaglašavanje (literally “harmonization of views” or, as I have translated it, “mutual accommodation”) has been relatively successful. The “concert” style of accommodation has reduced the incidence and intensity of resentment at policy decisions, since those affected by the decisions have been consulted.

In a fragmented system with reinforced cleavages, whose federal units enjoy extensive powers, the danger of escalation of policy differences into crises is great. Yugoslavia has been characterized by reinforced cleavages, conflicts of interest that frequently follow cleavage lines, and conflict behavior that does not always follow lines of cleavage. Yet, as long as the communists were in charge, the system did not get out of control.

Conflict Regulation

Having set out to investigate the viability of democracy in plural societies, Rabushka and Shepsle concluded that differing ethnic groups tend to be self-segregating and intermingle only “in the marketplace”; force is necessary to maintain order in plural societies; crosscutting cleavages do not eliminate the political importance of ethnic distinctions; only balanced competition affords the possibility of working democracy; and, as a result, stable democracy is impossible in fragmented societies. In short, as sociologist Pierre van den Berghe concluded, in despair, “generally, the more pluralistic the society as a whole and the political institutions in particular, the more tyrannical the polity.”⁴⁰

At the same time, it is a well-known principle that repression, especially if unevenly or inefficiently applied, tends to provoke resistance, obstruction, and violence. Neither the repressive Serbian centralism of the interwar period nor the hegemonistic unitarism of the postwar period (1946-66) succeeded in producing the modicum of stability essential for the orderly functioning of Yugoslav society. Neither formula, in short, was adequate to the requirements of conflict regulation in multiethnic Yugoslavia. Yugoslavia’s post-1971 solution was to wed decentralization to democratization in a one-party context and to make that marriage the basis for conflict regulation. The Yugoslav experiment was based on the premise that conflict could be regulated if there was pluralism within the party, even if genuine pluralism was proscribed outside the party.

In a sense, conflict accommodation is subsumed under the broader category of conflict regulation. In this context, however, I have distinguished between them in order to highlight two discrete functions necessary for the survival of the system. By conflict accommodation, I mean the working out of solutions satisfactory to the major parties to a conflict. By conflict regulation, I mean the prevention, containment, and management of conflict in such a way that the basic parameters of the system are maintained. If the ground rules of the system have to be rewritten, a section of the society attempts violent secession, the society explodes in massive and sustained violence, or the very survival of the system becomes a matter of widespread doubt, then conflict regulation has failed. Conflict-regulating practices are those measures that make possible the normal functioning of the society. When conflict regulation fails, as happened in Yugoslavia in 1971, the result is system crisis. The rules of the system are called into question, and the orderly functioning of the system cannot be restored without a restructuring of system relationships and institutionalized patterns of behavior. The explosion of ethnic riots in Kosovo in 1981 proved to be an adumbration of impending system crisis.

Eric A. Nordlinger, whose 1972 study has given us a perspicacious analysis of conflict regulation, outlined six typical conflict-regulating practices: (1) a stable government coalition in which the political parties of all major ethnic groups are represented according to a fixed formula; (2) the principle of proportionality, according to which the representation accorded to each ethnic group in the ruling coalition is roughly proportional to its share in the overall population; (3) the mutual veto, which allows any coalition partner to veto legislation; (4) the purposive depoliticization of public policy areas of direct concern to the constituent ethnic groups qua separate groups, that is, the endeavor to treat them as administrative problems of the entire country rather than as political contests pitting one group against another; (5) reliance on compromise as a means to draw antagonists to a common solution; and (6) the encouragement of concessions as a method of compromise.⁴¹ Nordlinger claimed that at least one of these six practices had to be employed for conflict regulation to be successful.

The Yugoslav socialist polity, in fact, employed five of the six practices listed above (all but the first—unless the LCY itself is interpreted as a stable coalition). Proportional representation was more or less faithfully adhered to in the party, the government, and even, within limits, in the army officer corps and party organization (see table 9). The republics and provinces enjoyed equal representation in the collective presidency, the federal Executive Council, the Constitutional Court of Yugoslavia, and various other bodies. The mutual veto was also incorporated into the Communist system of decision making, which spurned majoritarianism as a species of tyranny and sought to ensure solutions satisfactory to all by requiring unanimity among delegations in the two houses of the Skupština. To make this practical, the Yugoslavs spent a great deal of time in committees hammering out compromises acceptable to all.

In the early postwar years, when all decisions of any consequence were made by the CPY politburo, and the republics had to be satisfied with consultation on questions affecting their interests, the Yugoslavs hoped to “depoliticize politics” by breaking down the sense of separateness that divided Croats from Serbs, Slovenes from Macedonians. But the endeavor failed, and by the late Tito era, the Yugoslavs were trying a new tack, namely, to purposively depoliticize public policy areas impinging on the basic values and interests of the chief nationality groups. Devolution was the technique whereby this was accomplished, since the greater the policy domain within the jurisdiction of the federal units, the fewer the occasions for dispute at the federal level. Where a common strategy was necessary, the decision-making style of usaglašavanje, incorporating the mutual veto, was at hand.

Table 9. National Composition of Office-Holders

Sources: Gary K. Bertsch, “Ethnicity and Politics in Socialist Yugoslavia, “ in Annals of the American Academy of Political and Social Science (September 1977), 433: 97; Slobodan Stanković, “On the Eve of the 12th Yugoslav Party Congress,” Radio Free Europe Research, June 25, 1982, p. 4; and Tanjug (November 24, 1989), trans. in Foreign Broadcast Information Service, Daily Report (Eastern Europe), December 14, 1989, p. 89.

For all the talk of class struggle and class interests, the Yugoslavs realized— no thanks to Marx and Engels—that the interests of the working class are not monolithic and that sections of the working class may have strongly vested interests that conflict with the interests of another section of the working class. The Yugoslavs therefore embraced the un-Marxist concept of compromise. “At this point in our socioeconomic development,” wrote Mahmut Mujačić, a candidate member of the central committee of the LCY, in 1978, “compromise (kompromis) is an extremely important means of reconciling viewpoints and a necessary element in the harmonization of interests among the republics and provinces.”⁴² It became, in fact, a routine element in interrepublican negotiation. Finally, the Yugoslav modus operandi also embraced the last item on Nordlinger’s list—the use of concessions offered to the less fortunate or weaker republics.

Nordlinger pointedly excluded federalism and semiautonomy from his list. Both, he claimed, may more properly be viewed as the goals of certain actors in the system rather than as processes assuring conflict regulation, and, far from regulating conflict, they might exacerbate it.⁴³ The Yugoslavs, however, disputed this assessment and placed the dual policy of federalism and far-reaching autonomy at the very heart of their solution to the problems of interethnic distrust and interregional rivalry—the key, in other words, to conflict regulation in Yugoslavia.⁴⁴

Conflict regulation in an international balance-of-power system involves the use of analogous and often identical techniques. Bearing in mind that the mechanism for conflict regulation in a balance-of-power system is the so-called concert system, it is apparent that such a mechanism presumes unanimity among the great power concert-participants (the mutual veto), assumes the use of compromises and concessions, and honors the principle of proportionality in the sense of one great power, one vote.

It should not be forgotten, however, that the Yugoslav communists were not elected to power but seized it, and they have never won free elections in any part of Yugoslavia. Nor should it be forgotten that, for all the talk of the sovereignty of the republics, the LCY was always determined to hold Yugoslavia together by whatever means necessary. Ultimately, Yugoslav stability has rested on force and the threat of force. The LCY never hesitated to use force when deemed necessary; it resolutely sent armed troops and tanks into Croatia in 1971 and into Kosovo in 1981. Similarly, in the weeks after Tito’s death, extraordinary security measures were taken in Bosnia, where a Croatian nationalist putsch was not ruled out. The events of 1991 are also of clear relevance here.

Behavioral Constants in a Balance-of-Power System

To present Yugoslavia as a balance-of-power system is neither to deny the reality of the deterioration of interethnic relations in the mid-1960s nor to assume that the antipathy that has sometimes characterized interethnic relations is a permanent, if periodically latent, feature of the system. I do not contend that it is merely because of the ethnic divisions that the federal system operates as a balance-of-power system. Yet, certainly, ethnicity is a potent factor that influences the choice of alliance partners and the endurance of coalitions, especially where the dictates of economic interest are ambiguous. This need not be a weakness; by a kind of Madisonian logic, it may even be a strength. This is evidently the interpretation that the president of the Chamber of Nationalities wished to place on things when he wrote, in 1967, that he considered “the existence of several peoples and nationalities in Yugoslavia . . . an advantage for our socialist community, rather than a necessary evil which has to be disposed of as soon as possible by some kind of supranational structures.”⁴⁵ No system of interactor behavior—whether international, interethnic, or interregional—is permanent, but an understanding of the prevailing patterns may help us predict likely outcomes of system breakdown, as Kaplan foresaw. When Djilas wrote that, by 2024, “Yugoslavia will become a confederation of four states: Slovenia, Croatia, Macedonia, and Serbia, with Serbia itself being a federative state,”⁴⁶ he was extrapolating current trends in an attempt to identify those that, of a number of contradictory tendencies, will prevail. Systems analysis does not start with what is possible and work backwards to link the future with the present; it starts with what is essential and derives models of change on the basis of institutionalized power. Thus, Djilas’s four emergent states were all expected to evolve from then-existing socialist republics. But there is not only an institutional dimension to Djilas’s prognostications; there is also an ethnic dimension. Djilas predicted that both Montenegro and Bosnia-Herzegovina would disappear as autonomous territories—the former, presumably, because of the close ethnic affinity between Serbs and Montenegrins and the latter, perhaps, because Muslim national self-consciousness is neither as developed nor as tenacious as the Serbian, the Croatian, or the Slovene. In the first edition of this book, I concluded this chapter by observing:

An awareness of the wide-ranging autonomy enjoyed by the federal units is sufficient to permit prediction of evolution in the direction of loose confederation. Establishing the precise patterns of interrepublican behavior would enable us— in the event we wished to undertake such an enterprise—to forecast, with some confidence, how many states might emerge from system breakdown and what their likely alignments would be.⁴⁷

Events have proven the accuracy of this forecast, just as they have highlighted the dangers associated with the resultant clash between centrifugal and centripetal forces.