Nationalist and Power-Seeking Leadership Preferences in Ethno-Territorial Conflicts: Theory, a Measurement Framework, and Applications to the Breakup of Yugoslavia

Abstract

Theoretically, variation in leadership preferences is often taken to be an important predictor of ethno-territorial conflict outcomes. Yet there is a significant gap when it comes to applying this theory. Case studies do not take a consistent approach to measuring leadership preferences, while statistical research tends to omit the variable altogether. This paper suggests a standardized approach to measuring leadership preferences along two dimensions – a dimension that captures the weight given to achieving ideal nationalist goals as against minimizing conflict costs and downside conflict risks, and a dimension that indicates how much intrinsic nationalist goals are valued relative to the goal of taking and maintaining political power. The resulting measurement template is then applied to seven potential ethno-territorial conflicts in Yugoslavia in the early 1990s. These cases indicate that leadership preferences may help to explain ethno-territorial conflict outcomes and, at the same time, are sometimes not well predicted by other important variables, such as the balance of power and the pre-conflict status quo.

Notes

^1. There are many variations on this basic pattern of ethnic conflict over rival homeland claims. Sometimes there is more than one politically dominant ethnic group, sometimes the largest ethnic group does not control the state, sometimes conflicts occur between sub-national ethnic groups, and so on.

^2. Joseph Rothschild, Ethnopolitics: A Conceptual Framework (New York: Columbia UP 1981) p.150.

^3. Michael Brown, ‘The Causes of Internal Conflict: An Overview’ in Michael Brown, Owen Coté, Jr, Sean Lynne-Jones and Steven Miller (eds) Nationalism and Ethnic Conflict (Cambridge, MA: MIT Press 2001) p.15.

^4. Decisions in conflicts are usually made by leaders. Both structural and institutional constraints must work through leaders with various preferences. Studies on individual decision-making have demonstrated that the same material conditions may have varying ramifications for different leaders. See Daniel Kahneman and Amos Tversky, ‘Prospect Theory: An Analysis of Decision under Risk’, Econometrica 47/2 (1979) pp.263–91; Jack Levy, ‘Applications of Prospect Theory to Political Science’, Synthese 135/2 (2003) pp.215–41. A large literature on political institutions posits that, because institutional constraints must be internalized by leaders, the nexus of domestic political institutions and decisions in war and conflict can only be understood through the lens of leaders' characteristics and preferences. See Bruce Bueno de Mesquita, Alastair Smith, Randolph Siverson, and James Morrow, The Logic of Political Survival (Cambridge, MA: MIT Press 2003); Giacomo Chiozza and H. E. Goemans, ‘International Conflict and the Tenure of Leaders: Is War Still “Ex Post” Inefficient?’ American Journal of Political Science 48/3 (2004) pp.604–19; Margaret Hermann and Charles Kegley, Jr, ‘Rethinking Democracy and International Peace: Perspectives from Political Psychology’, International Studies Quarterly 39/4 (1995) pp.511–33; Jameson Lee Ungerer, ‘Assessing the Progress of the Democratic Peace Research Program’, International Studies Review 14 (2012) pp.1–31.

^5. For example, see William Greene, Econometric Analysis (New York: Macmillan 1990) pp.259–61.

^6. In describing leadership preferences in ethno-territorial conflicts, it is common to refer broadly to these two dimensions. For general discussions, see Rothschild (note 2) pp.137–47 and pp.160–7, and Stephen John Stedman, ‘Spoiler Problems in Peace Processes’, International Security 22/2 (1997) pp.5–53. For examples of applied discussions, see Paul Anastasiou, ‘Nationalism as a Deterrent to Peace and Interethnic Democracy: The Failure of Nationalist Leadership from the Hague Talks to the Cyprus Referendum’, International Studies Perspectives 8/2 (2007) pp.190–205; Paul Brass, Ethnicity and Nationalism: Theory and Comparison (New Delhi: Sage 1991) pp.167–219; Erik Melander, ‘The Nagorno-Karabakh Conflict Revisited: Was the War Inevitable?’ Journal of Cold War Studies 3/2 (2001) pp.48–75; Emrullah Uslu, ‘Turkey's Kurdish Problem: Steps toward a Solution’, Studies in Conflict and Terrorism 30/2 (2007) pp.157–72. The references are listed in the online Appendix.

^7. Bargaining models of war onset also account for war duration – in the sense that a war that does not break out has zero duration. Thus, if a war has already broken out, the same model can be used to explain whether or not a war ends or continues under the prevailing conditions. See Robert Powell, ‘Bargaining and Learning While Fighting’, American Journal of Political Science 48/2 (2004) pp.344–61. Similarly, avoiding war or ending war involves explaining the outcome chosen instead of war. Thus, if war has already broken out, explaining why it ends at a certain time also explains the outcome when it ends. See Harrison Wagner, ‘Bargaining and War’, American Journal of Political Science 44/3 (2000) pp.469–84. The basic version of the model discussed here assumes that there is only one way of fighting the war – one war strategy. If this assumption is dropped and two or more war strategies are allowed for each side, then the framework can also predict choice of war strategies.

^8. The Addendum shows results for a one-period model. The same logic applies in two- or infinite-period models. Proofs are available from the authors upon request. See Ariel Rubinstein, ‘Perfect Equilibrium in a Bargaining Model’, Econometrica 50/1 (1982) pp.97–109 for the basic model. Well-known applications include James Fearon, ‘Rationalist Explanations of War’, International Organization 49/3 (1995) pp.379–414 and Robert Powell, In the Shadow of Power: States and Strategies in International Politics (Princeton, NJ: Princeton University Press 1999), who use a model like that of the Addendum's Proposition 1. Bruce Bueno de Mesquita and David Lalman, War and Reason: Domestic and International Imperatives (New Haven, CT: Yale University Press 1992) distinguish more hawkish leaders, which in our context are analogous to leaders with more extreme nationalist preferences.

^9. For example, concessions necessary to prevent war might themselves cause changes in the balance of power, which in the absence of credible commitments to the status quo would then lead to new crises; or there might be uncertainty about the true balance of power, leading one or both sides to believe that war will deliver greater gains than the other side is willing to accept in a bargain short of war.

^10. Note that such a significant change in the status quo is typically due to significant changes in relative power or in leadership preferences, and thus would be considered an effect rather than a cause of crisis.

^11. Although continuous measures would be theoretically ideal, we were not able to distinguish more than five intervals along the nationalist goals dimension and three intervals along the principled–unprincipled dimension.

^12. Laura Silber and Allan Little, The Death of Yugoslavia (revised ed.) (London: Penguin 1996) p.86.

^13. Montenegro is excluded from the analysis, because Montenegrins are not ethnically distinct from Serbs. There was a potential conflict here, but it involved regional rather than ethnic separatism. To analyze this case, it would be necessary to adapt the analysis and the template. For space reasons, we also do not consider the much smaller ethnic minorities, such as the Sandzhak Muslims and the Vojvodina Hungarians and Croats.

^14. See the Appendix for sources.

^15. As noted in the Appendix, Milošević sometimes appeared to moderate his nationalism, speaking and acting more like an ordinary nationalist.

^16. Milošević defined his nationalist goals in terms of restoring Serbia's full internal control over the autonomous provinces of Kosovo and Vojvodina and pursuing self-determination for the concentrated Serb communities in Croatia and Bosnia. Milošević's nationalist goals did not involve Slovenia or Macedonia. For Tudjman, independence within Croatia's Yugoslav boundaries was the primary objective, while self-determination of the Bosnian Croats was a secondary objective. For Izetbegović, the preference is the same vis-à-vis Serbia and Croatia, given his strong commitment to a unified Bosnia as the future religious-nationalist homeland of the Bosnian Muslims. Gligorov's cautious, moderate nationalism applied to both Serbia and the Macedonian Albanians. For more details, see the Appendix.

^17. This might not be so in all institutional and practical settings. In Slovenia, Peterle as the prime minister had more power, and could not have been easily constrained by Kučan in the event of a serious disagreement. For the Macedonian Albanians in 1994–1995, Halili and Xhaferi led separate, large and well-organized parties. So either could have acted independently of the other. We classify the leaders of the Croatian Serbs and Bosnian Serbs as dependent on Milošević, and those of the Bosnian Croats as dependent on Tudjman. As discussed in the Appendix, Milošević wanted the appearance of Croatian and Bosnian Serb independence. But when he wanted to, he did not hesitate to impose new leadership, or to exclude the Croatian and Bosnian Serb leaders from negotiations and decisions about when and on what terms to end the conflict (most notably, in the Dayton Accords). Tudjman openly appointed and replaced Bosnian Croat leaders at his pleasure.

^18. The institutions were adopted in the 1974 constitution, but were not fully operative while Tito remained the de facto absolute ruler.

^19. Here and elsewhere, we use the term ‘substantially all’ to indicate that no large regions claimed as parts of homelands were outside federal territorial limits. In all cases, there are small communities near borders that might ideally be added to federal territories.

^20. Dijana Plestina, Regional Development in Communist Yugoslavia: Success, Failure, and Consequences (Boulder, CO: Westview 1992).

^21. The 1–5 range of variation used here is sample-specific. Larger or different samples would typically require greater variation on one or both ends of the range. Developing such a scale is beyond the scope of this paper.

^22. For example, for the dyads Croatia–Bosnian Muslims and Serbia–Bosnian Muslims, the economic strength of the Bosnian Croats and Bosnian Serbs are deducted from the figure for Bosnia as a whole and added to the respective figures for Croatia and Serbia. At the same time, the Croatia figure excludes Croatian Serbs and the Serbia figure excludes Kosovo Albanians. Where separate statistics for ethnic minorities are not available – as with Croatian Serbs, Bosnian Croats, Bosnian Serbs, and Macedonian Albanians – we estimate that their per capita incomes were the same as the republic average. We do not use the well-known Correlates of War measure of material capabilities, which has missing data and also seems to have significant measurement errors for these cases. For example, Slovenia is estimated to have about one-fifth the capabilities of Serbia – a gross underestimate largely due to excluding sizeable paramilitary (reserve and police) forces from estimates of military manpower and spending.

^23. For overviews of the various conflict onsets and non-onsets, see Steven Burg and Paul Shoup, The War in Bosnia-Herzegovina: Ethnic Conflict and International Intervention (Armonk, NY: M.E. Sharpe 1999); Lenard Cohen, Broken Bonds: The Disintegration of Yugoslavia (Boulder, CO: Westview 1993); James Gow and Cathie Carmichael, Slovenia and the Slovenians: A Small State and the New Europe (Bloomington: Indiana UP 2000); Tim Judah, Kosovo: What Everyone Needs to Know (Oxford: Oxford UP 2008); Hugh Poulton, Who Are the Macedonians? (Bloomington: Indiana UP 2000), and the other sources are given in the Appendix.

^24. War broke out in Kosovo in 1998, as did crisis (but not war) in Macedonia in 1999. This later time period is beyond the scope of this paper.

^25. Information about preferences and possibly other factors is necessary to make such predictions.

^26. If the political benefits are not significant, a strong power-seeker can be expected to avoid the significant potential downside political risks of war. Significant diversionary political benefits may be captured by using harsh public rhetoric, and by initiating a crisis while stopping short of war.

^27. In 1991, after Slovenia's unilaterally declared independence, Milošević stopped a nascent conflict between Slovenia and the Yugoslav People's Army (YPA), preferring to let Slovenia leave the Federation in order to gain control over the rump Yugoslav state and the YPA.

^28. James Fearon, ‘Signaling Foreign Policy Interests: Tying Hands versus Sinking Costs’, Journal of Conflict Resolution 41/1 (1997) pp.68–90; Jessica Weeks, ‘Autocratic Audience Costs: Regime Type and Signaling Resolve’, International Organization 62/1 (2008) pp.35–64.

^29. The implication is that, other things equal, a power-seeker always gains more politically from a better deal, or .