Princes do keep due sentinel, that none of their neighbours do overgrow.

FRANCIS LORD BACON, “Essay on Empire”

Among the many paradoxes of Yugoslav politics that long fascinated Western observers were the apparent disjunction between Titoism’s seemingly efficient mechanism for conflict regulation and its occasional lapses into acerbic crisis behavior and the contradiction between the quasi-confederal relations long prevailing among the republics and the system’s long failure to fall apart. This book is, in part, an effort to explain how these ostensibly incompatible elements not only coexisted but in fact characterized a unified system of behavior, a system better known as the “balance of power.” I have argued that interrepublican relations in Yugoslavia were characterized by flexible coalitions and that the political behavior exhibited by these federal units closely resembled the political behavior of states in the European balance-of-power system.

In the domestic, just as in the international, balance-of-power system, one cannot expect component actors to be moved by exhortations to the good of the community. Each group will attempt to pursue its interests by subordinating the common good to its own communal interests and also by subordinating the good of other communal groups to the good of the whole, (if not, in fact, to its own private good). Of course, an element of restraint is built into a balance-of-power system, as long as it lasts, viz., a sense that in most cases the general equilibrium (recall Kaplan’s rules 5 and 6) serves the interest of each player.¹ But it is the shape of conflict that gives the system its distinctive contours. A requirement for the stability of the system is that no single actor is indispensable to the formation of a winning coalition (and therefore capable of imposing its will on decision making). Such a condition prevailed in Yugoslavia until 1990. Thus, policy outcomes in matters of interrepublican importance came to depend on the free combination of republican actors in shifting alliances and flexible coalitions. The existence of group interests may be presumed; the politicization of group interests may not. The argument has therefore consisted of three parts: establishment of politicized group interests that coincided with nationality and/or republic boundaries, demonstration that the republics and autonomous provinces enjoyed sufficient autonomy to be able to pursue these interests in meaningful ways, and confutation of simple bipolar views of Yugoslav politics.

In an ideal realm of pure abstractions, actors have no vested interests and combine in a completely free and unfettered way with no desire but to “win.” And, in this spirit, the uncommitted and disinterested are expected to ally with the weaker bloc to prevent change: an international system, abstractly conceived, is said to have a built-in tendency to stagnation.² But, in the realm of real political behavior, the flexibility of a system is almost inevitably compromised to a greater or lesser degree by the existence of stable patterns of interests. The flexibility of a system stems in part from the overlap of interests, in part from internal contradictions (that is, competing and conflictual interests within a single republic), and in part from the willingness of partners to engage in trade-offs in the particular policy areas that most concern them (e.g., Croatian and Bosnian cooperation on the latter’s eligibility for FADURK assistance). Stable communal interests among Bosnia, Macedonia, Montenegro, and Kosovo in maintaining the FADURK apparatus is an example of the overlapping of interests, while the common anti-Serbian current found in Croatian, Albanian, and Montenegrin nationalisms illustrates the role of internal contradictions in the system.

Obviously, flexibility in interrepublican coalitions may be a factor of changes in the orientation of the republican elites in power (whether because of changing issues, changing context, changing interests, or changing perceptions). It may also, however, be a factor of changes in the composition of the republican elites as certain factions edge aside their rivals. Since different factions (the chief ones being the so-called liberals and conservatives) have different perceptions of republican interests, their coalition choices will inevitably be affected by the composition of elites in the other republics. The Croatian-Slovenian-Macedonian axis of 1967-69 was possible because liberals occupied the top leadership positions in these republics. (This was not enough to vouchsafe the coalition— as was shown by the consecutive defections from the coalition of Slovenia and Macedonia, which left Croatia to seek new alliance partners.) Similarly, Croatian conservatives united with conservative forces in Serbia and Slovenia to change the balance of forces in the system; in the process, the alignments of coalitions also changed.

In recounting the several controversies and crises that have sparked interrepublican relations during the 1962-91 period, I have tried to highlight political behavior relevant to the hypotheses derived from my adaptation of Kaplan’s balance-of-power model. It is now possible to take stock and see which hypotheses are confirmed and which discounted. I have discussed some thirty-two controversies (in some cases overlapping) in greater or lesser detail. These are:

Cl. The reform crisis, 1961-66

C2. The status of the autonomous provinces, early 1960s-1974

C3. The Belgrade-Bar railroad, 1963-76

C4. The disbursement of FADURK’s resources, 1965-72

C5. The debate on republican tax limits, 1967

C6. The status of Albanians and civil disorders, 1967 to the present

C7. The struggle over devolution, 1967-71

C8. The Slovenian highway crisis, 1969

C9. The debate on banking reform, 1970

C10. The controversy over Macedonia’s Muslims, 1970-71

C11. The status of Serbs in Croatia, 1971

C12. The Croatian crisis, 1971-72

C13. The controversy over fishing compensation from Italy, 1972

C14. The debate over scope and conditions of aid to Kosovo, 1977

C15. The debate on the draft law on customs tariffs, 1978

C16. The debate over FADURK criteria, 1978-82

C17. The air transport controversy, 1979-82

C18. The controversy over inflation control—wages, 1979

C19. The controversy over inflation control—prices, 1979

C20. The controversy over compensation to Montenegro, 1980

C21. The “Bratstvo-Jedinstvo” controversy, 1980

C22. Liberalization versus retrenchment, 1980-87

C23. Decentralism versus recentralization, 1980-87

C24. Dispute over the retention of foreign currency earnings, 1985

C25. Dispute over the constitutional proposals, 1987

C26. Dispute over reform of the Assembly, 1988-89

C27. Dispute over reform of the SFRJ presidency, 1988-89

C28. The anti-Mikulić coalition, 1988

C29. Clash over the 1989 federal budget, 1989

C30. Support for Marković’s economic program, 1989

C31. The “Slovenia versus Serbia” debate, 1989

C32. The confederalization debate, 1989-90

Throughout these crises, the republics and autonomous provinces displayed autonomy of decision making and willingness to realign in shifting coalitions according to perceived changes of interest (compare the patterns in appendix 1).

I have also examined ten hypotheses, set forth in chapter 1. In brief, these are:

H1. Federal units promote their own interests even when they conflict with the interests of the federal community.

H2. Federal units negotiate but provoke crisis if necessary to protect their interests.

H3. Federal units resist encroachment or ethnic assimilation and form coalitions to oppose any threatening actor.

H4. All federal units are acceptable coalition partners (though in some cases this may apply only in a faction-to-faction sense).

H5. Alliance between a federal unit and an external power is viewed as illegitimate.

H6. The federal government tries to reduce the disparities in power and wealth among the federal units.

H7. The federal government takes whatever action is necessary to resolve interrepublican disputes that threaten to upset the system.

H8. The federal government strives to be ethnically neutral.

H9. Social mobilization increases the pressure for decentralization.

H10. Conflict is effectively regulated under a concert system.

A cross-tabulation of these hypotheses with the thirty-two controversies is shown in figure 2. The results show a strong corroboration for hypotheses 1, 2, 3, 4, 7, 8, and 9—derived, in part, from Kaplan’s rules 1, 2, 4, and 6—and reliable confirmation of hypothesis 9. Only twelve of the twenty-one cases were relevant tests of hypothesis 6, and, of this number, ten cases confirmed the hypothesis. But the apparent disconfirmation of the supposition of an “equalizing federal actor” in two cases is not as serious as might be supposed, since, in at least one of the two cases, the federal government “broke the rule” by limiting itself to making recommendations—a policy in harmony with the quasiconfederal character of the system. Hypothesis 5, the disallowance of alliance with foreign powers, was only explicitly confirmed in three cases, but there can be little doubt of its validity. As for the remaining hypothesis (10), it is actually neither a part of Kaplan’s model nor essential to the stable operation of a balance-of-power system. Ironically, hypothesis 10, the evaluational hypothesis that purports to corroborate Nordlinger’s favorable estimation of Yugoslavia’s capabilities of conflict regulation, remains, at least insofar as the confirmatory evidence of these controversies is concerned, the most problematic.

More than seventy years ago, Arthur Bentley observed that “on any political question which we could study . . . we should never be justified in treating the interests of the whole nation as decisive. There are always some parts of the nation to be found arrayed against other parts.”³ And the foregoing case studies have manifestly demonstrated that Yugoslavia’s republics do have frequent conflicts of interest. A casual observer of the Yugoslav scene might expect to find profoundly polarized alliance behavior, with the underdeveloped republics invariably coalescing around a common position, Slovenia and Croatia consistently mutually supportive, and Croatia and Serbia generally at loggerheads. Such inflexible alignments would be incompatible with a true balance-of-power system. Moreover, when the federal government plays a role, we should expect, under the conditions of the model, that more often than not it will intervene to preserve the balance. This might mean that the federal authority will come down on the weaker side, if one construes the federal government as a surrogate “balancer”; it could also come down on the stronger side, if one emphasizes instead the “concert” aspect of the federal government, that is, its self-presentation as an interrepublican arena.

Table 26 reveals the trends that emerge from a review of the twenty-one controversies. Note that virtually all of the economic controversies documented are relatively specific issues, in contrast to the more all-encompassing issues found under the “political” category (such as the Croatian crisis or the crisis over devolution). As often as not, Croatia and Serbia are allied in economic issues (usually as parts of a larger bloc)—reflecting system flexibility. They have been consistently at odds, however, on political issues. Second, it appears that the underdeveloped republics oppose each other more often than they stick together (in conformity with the model), and, in five of the eight cases in which they did act as a bloc, the issue was strictly economic. Indeed, little else would predispose these four units to act in unison. Third, Croatia and Slovenia appear to be mutually supportive in political issues, but, in economics, they oppose each other almost as often as they are allied. Fourth, as far as federal interventions are concerned, the federal government appears inclined to go with the preponderant bloc, though this perception may be a distortion of the issue, since what is at stake is not defeat and annexation but the general interest of the majority versus the exclusive interests of one or two republics. Seen in this light, the fact that the federal government favored the majority in 75 percent of the cases is a reflection of the workings of consensual politics within a concert mechanism. Finally, there were at least three instances among the twenty-one cases in which an actor switched sides in midstream—a recourse that lies at the heart of a balance-of-power system.

Figure 2

Table 26. Trends in Alliance Behavior among Yugoslav Republics

Because, as I have noted, concrete interests are at stake rather than the mere preservation of balance, it is not uncommon to find blocs of six or seven actors uniting against one or two recalcitrant federal units. The number of republics in each alliance is not so important as maintenance of the approximate equality (and formal equality) of the actors (no republic may rise to a position of preeminence analogous, even crudely, to that of the Soviet RSFSR). The system must remain at (or be restored to) equilibrium, and that equilibrium must be valued by the actors in the system. Since these conditions prevail in socialist Yugoslavia, I do not hesitate to identify Yugoslavia as a concert variant of the balance-of-power system.

I have traced the development of interrepublican relations in Yugoslavia from 1962 to 1991. During these three decades, the Yugoslav federal system passed through seven behavioral phases: (1) 1962-66, system change; (2) 1966-69, balance of power, semiconfederal; (3) 1969-71, balance of power, quasi-confederal; (4) 1972-80, balance of power, quasi-confederal, with a concert mechanism; (5) 1980-87, balance of power, quasi-confederal; (6) 1987-89, balance of power, fissiparous, with pressure toward recentralization; (7) 1989-91, confed-eralization. The years 1962-66 were catalytic, years when the behavioral patterns associated with the “unitarist” period broke down and dissolved the neat polarity of an age when most of the republics were politically dormant. Through the mobilization of the other six federal units into the system, the Serb-Croat rivalry ceased to be the lodestone of intra-Yugoslav politics, and the republics began to engage in a pattern of shifting coalitions. Moreover, beginning with the Janko Smole affair and the Slovenian road crisis, the republics began to consider opposition to the federal government in areas where their interests were threatened.

Mobilization of ethnic-group actors spawned a revival of nationalisms, however. Dawning awareness of the harshness of Ranković’s policies, moreover, heightened the sense of collective cognitive dissonance that had already been aroused, in some cases, by feelings of relative economic deprivation. Spokespersons for the republics no longer spoke in terms of Yugoslav interests but, increasingly, in terms of republican interests. The crucial turning point came at the Fourth Plenum of the central committee of the LCY in July 1966. This plenum marked the fall of Ranković and Stefanović and the beginning of a period of increased autonomy for the republics. It also signaled a change in interrepublican dynamics. For the first time, the underdeveloped republics allied with Croatia and Slovenia and against Serbia. The Brioni plenum thus completed the process of system change begun by the economic reforms.

Between 1966 and 1969, the Yugoslav parliament was reinvigorated, the Chamber of Nationalities came back into its own, and the republics discovered that they had more political clout than previously realized. Yet this confederal tendency was limited by institutional factors and by the continued resistance of unitarist forces. The passage of far-reaching constitutional amendments between 1967 and 1968 overcame the first hurdle, and the latter retreated in the face of the growing strength of the national-liberal-technocratic coalition. After 1969, the central apparatus appears to have lost control of the situation—it was internally divided and lacked clear direction—and, consequently, the republics aggrandized their power. Republican self-interest was unabashedly touted as the highest good. In a typical comment, Miko Tripalo told Vjesnik in 1970 that “the League of Communists in every republic expresses and is obliged to express the interests of the working class of its nation and of its republic.”⁴ Ultimately, this attitude, when wedded to the nationalist oestrus, led to the formulation of nonnegotiable and totally impracticable demands and precipitated a severe confrontational crisis in the system.

Tito’s “compromise,” which ended the Croatian crisis and swept the liberals out of office in Croatia, Slovenia, Serbia, and Macedonia, affected interrepublican relations in an important way. Tito refused to scuttle the liberal reforms, yet he did not trust the liberals to administer them. He let the conservative factions carry out the liberals’ vision, thus confirming the quasi-confederal nature of the system without allowing autonomous centers of power to become focuses of politicized collective affectivity. Ultimately, Tito hoped to hold together a “liberal” system not by force but by a common ideology, the ideology of “conservatism.” In his appreciation of the importance of system homogeneity Tito may be likened to Metternich, who also stabilized a system by suppressing one of two rival ideologies. Finally, in the wake of Tito’s death, the central party, shorn of the arbiter who had created unity among rival factions, proved unable to agree on a common policy and watched its control slacken as the republican parties jumped to take advantage of the new confederalism.

One of the points that I hope has emerged from this study is that multiethnic configuration and level of modernization present a regime with a narrowed field of policy options. The regime’s policy “choice” may sometimes be dictated from below, by the systemic environment, including the actors in it. This does not mean the regime is helplessly determined by its environment; it does mean that the regime’s ability to affect political development is limited. The LCY did choose to terminate discussion of federalization of the party on the grounds that such a debate might resurrect the multiparty system; in a multiethnic state, there is a great danger that a multiparty system (which is likely to be ideologically heterogeneous) will provoke instability.

It is incontestable that the presence of a central government distinguished the Yugoslav federal system from the classical European balance-of-power system of the eighteenth and nineteenth centuries, at least until the end of 1989. Yet, just as it makes sense to speak of polarized politics in a binational state, so is it meaningful to speak of multipolar politics in a multinational state. Where multipolar politics are characterized by a shifting pattern of flexible coalitions, the system may accurately be characterized as a balance-of-power system. Furthermore, the federal government in Yugoslavia often functioned as primus inter pares in a nine-actor universe (as in the interrepublican committees) and, in any case, is to a considerable degree an interrepublican body itself—thus warranting its characterization as a concert mechanism.

The argument of this study has proceeded at three levels to these conclusions:

(1) Yugoslavia was able to attain a measure of temporary stability by virtue of the regime’s federal formula, which, one could argue, gives the system its peculiar “quasi-legitimate” character;⁵

(2) “the communists substituted regional pluralization (administrative decentralization) for political pluralization (multiparty democracy),”⁶ with the result that interelite conflict, during the era of the LCY monopoly, was always intra-LCY conflict, sometimes pitting factions existing within the same regional party organization against each other;

(3) the pattern of interrepublican relations that has resulted from (a) the presence of eight autonomous federal units among which no single factor predominates, (b) the combination of a viable federal system with various interrepublican bodies, and (c) the fear of the alternatives to the status quo closely approximates the pattern best known as the “balance of power”; and

(4) there are close parallels, in this instance, between the constants of internal politics and those of international politics.

The more specific findings of this work may be summarized as follows:

(1) policy is often the result of coalition politics in which republics are the actors;

(2) republican party organizations consist of rival factions with rival perceptions of interest, and coalition patterns are a function of the vicissitudes of interfaction rivalry;

(3) there is strong corroboration for the applicability of four of Kaplan’s central rules to Yugoslavia, namely, (a) all actors are acceptable coalition partners, (b) actors act to restrain actors subscribing to supranational organizing principles, (c) actors unite to oppose any unit that tends to threaten the autonomy of the others or the stability of the system, and (d) actors provoke crisis rather than forego the opportunity to increase capabilities;

(4) certain trends in interrepublican behavior are identifiable, specifically: (a) Croatia and Serbia are often allied in economic issues but are consistently at odds in political issues, (b) the underdeveloped republics oppose each other as often as they stick together, and, in four out of the five cases in which they acted as a bloc, the issue was strictly economic; (c) Croatia and Slovenia are mutually supportive in political issues but often oppose each other in economic issues, and (d) the federal government, interested in stability, is inclined to side with the preponderant bloc (a factor of the concert mechanism); and

(5) the Yugoslav claim to have affected interethnic relations must be taken seriously.

Yugoslavia’s communists tried for more than forty years to “solve” the national question. They placed their trust in strict proportionality in federal organs (though not in the military), in ethnic quota systems (which were not applied in Kosovo or always in Croatia, or even in Bosnia), in massive decentralization to the point of verging on de facto confederalization, and in a mythology that drew inspiration from the partisan war, the nonaligned movement, and, for as long as he was alive, the charismatic personality of Tito. Ultimately, they failed. But their failure was not so much a failure of confederalism, but rather a failure of the concept of limited democracy, of the idea that democracy can emerge out of one-party rule.

Models of behavior, of course, cannot substitute for either research or analysis, nor are they meant to. The best that can be expected from a model is that it will highlight patterns in real behavior, sketch relationships and state linkages, serve as a guide to meaningful and useful comparative analysis, and establish a focus for research. There is no intrinsic reason why a model of behavior developed in one context might not be adaptable to another. This study has been founded on that supposition.