1 Gurr, Ted Robert, “The Internationalization of Protracted Communal Conflicts since 1945: Which Groups, Where and How,” in Midlarsky, Manus, ed., The Internationalization of Communal Strife (London: Routledge, 1992), 4–24Google Scholar; Ryan, Stephen, Ethnic Conflict and International Relations (Brookfield, Vt.: Gower, 1988)Google Scholar; and Ryan, Stephen, “Emerging Ethnic Conflict: The Neglected International Dimension,” Review of International Studies 14 (1988), 161–77.CrossRefGoogle Scholar

2 Putnam, Robert D., “Diplomacy and Domestic Politics: The Logic of Two-Level Games,” International Organization 42 (1988), 459.CrossRefGoogle Scholar

3 Davis, David, Jaggers, Keith and Moore, Will, “Ethnicity, Minorities and International Conflict Patterns,” in Carment, David and James, Patrick, eds., The International Politics of Ethnic Conflict: Theory and Evidence (Pittsburgh: University of Pittsburgh Press, 1996).Google Scholar

4 Irredentism means the claim to the territory of an entity, usually an independent state, wherein an ethnic ingroup is a numerical minority but forms a regional plurality (or even a majority). Theoretically, ethnic linkages are not a condition for irredenta, for claims can be based solely on territory ( Vasquez, John A., The War Puzzle [Cambridge: Cambridge University Press, 1993], 310–11)CrossRefGoogle Scholar. In reality, however, many irredenta are mixed and disputes about the nature of the claim usually involve mobilization of groups on the basis of the principle of reuniting ethnic kin. For that reason, and because the focus of this investigation is on ethnic factors that encourage interstate conflict, irredenta are defined as territorial and ethnic in nature. In other words, there must be an attempt to detach land and people from one state in order to incorporate them in another, as in Somalia and the Ethiopian Ogaden, Serbia and parts of Croatia and Bosnia, or Germany and the Sudetenland. Secessionist ethnic conflict refers to formal and informal aspects of political alienation in which one or more ethnic groups pursue, through political means, reduced control by a central authority. Such strife may involve politically mobilized, organized, ethnic insurgency movements and the use of force.

5 This follows Weiner's classic exegesis of irredentism, which assumed the existence of a “shared” ethnic group crossing the international boundary between two states. Subvariants occur when (1) an ethnic group transcends multiple borders but does not itself constitute a state (with the Kurds in Turkey, Iran, Iraq and Russia, for example) or (2) irredentist claims are made only about territory that has been “cleansed” of the ethnic claimant (for example, Armenian claims to eastern Anatolia) ( Weiner, Myron, “The Macedonian Syndrome: An Historical Model of International Relations and Political Development,” World Politics 23 [1971], 668)CrossRefGoogle Scholar. This study focuses on irredenta in which both an ethnic group and a territory are to be redeemed; as Horowitz and Chazan have argued, these cases are most likely to result in group mobilization that leads to interstate conflict and possibly war ( Horowitz, Donald, Ethnic Croups in Conflict [Berkeley: University of California Press, 1985]Google Scholar, and Chazan, Naomi, ed., Irredentism and International Politics [Boulder: Lynne Rienner, 1991])Google Scholar.

6 For a perspective on peaceful secession see Young, Robert, “How Do Peaceful Secessions Happen?” this Journal 27 (1994), 774–92Google Scholar, and Carment and James, eds., International Politics of Ethnic Conflict.

7 Although secession also has been defined more broadly to include ethnic and non-ethnic patterns of political separation, this inquiry uses secession as a synonym for separatism only in the sense of ethnic claims ( Gurr, Ted Robert, “Resolving Ethnopolitical Conflicts: Exit, Autonomy or Access,” in Gurr, Ted Robert, ed., Minorities at Risk: Origins and Outcomes of Ethnopolitical Conflicts [Washington: United States Institute of Peace Press, 1993])Google Scholar. Wood regards “secession” as a more precise term than “separatism” because it refers to a demand for formal withdrawal by a member unit (or units) from a central authority on the basis of claims to independent, sovereign status ( Wood, John, “Secession: A Comparative Analytical Perspective,” this Journal 14 [1981], 107–34)Google Scholar. Separatism, by contrast, covers all aspects of political alienation that include a desire for reduction of control by a central authority. This definition includes the anticolonial struggles (violent and nonviolent) of groups within a newly independent state (like Indonesia or Algeria) and more well-known insurgencies (such as Tamil separatism in Sri Lanka and Bangladesh) ( Carment, David, “The International Dimensions of Ethnic Conflict: Concepts, Indicators and Theory,” Journal of Peace Research 30 [1993], 137–50).CrossRefGoogle Scholar

8 Some argue that ethnic conflicts are self-limiting by virtue of their particularist traits. Others suggest that affinities can pose security dilemmas for entire regions. In either instance, ethnic affinity is viewed as an opportunity to be exploited by external states. Groups on which elites rely for support also perceive these international ethnic links as potentially useful (see Fearon, James D., “Ethnic War as a Commitment Problem” [unpublished ms., University of Chicago, June 1993]).Google Scholar

9 One initially thinks of intervention as an armed assault on one state by another. However, any effort to interfere with or disrupt the internal affairs of the states constitutes intervention, including both covert and overt activities, would be included in this definition. Intervention may also be the calculated use of political, economic and military instruments by one country to influence the domestic or the foreign policies of another country. Intervention will not be, and indeed has not been, confined solely to intervention by states through military means. It encompasses a broader range of activities, although it is hard to find agreement on what these might be.

10 Heraclides, Alexis, The Self-Determination of Minorities in International Politics (Portland: Frank Cass, 1991)Google Scholar, and Heraclides, Alexis “Secessionist Minorities and External Involvement,” International Organization 44 (1990), 341–78.CrossRefGoogle Scholar

11 The choice of aggressive tactics is usually influenced by an ethnic ideology (Panslavism), a sense of historic injustice (Danzig) or even a perceived menace to values that justify some kind of future society (perhaps one in which all of the relevant territory and people are reclaimed, Greece and Macedonia).

12 This summary of the research enterprise on two-level games is based on Evans, Peter B., “Building an Integrative Approach to International and Domestic Politics,” in Evans, Peter B., Jacobson, Harold K. and Putnam, Robert D., eds., Double-Edged Diplomacy: International Bargaining and Domestic Politics (Berkeley: University of California Press, 1993).Google Scholar

13 Mesquita, Bruce Bueno de, Siverson, Randy and Woller, Gary, “War and the Fate of Regimes: A Comparative Analysis,” American Political Science Review 86 (1992), 638–64.CrossRefGoogle Scholar

14 Preferences arise from a decision maker's role as leader within a specific institutional framework, so the two-level approach shares some basic similarities with foreign policy research that links structure and orientation. However, this study goes further by suggesting that affect represents an additional enabling condition in shaping foreign policy decisions in ethnic conflicts; it provides a crucial link between elite and mass behaviour and domestic politics and international ambition.

15 Tsebelis, George, Nested Games: Rational Choice in Comparative Politics (Berkeley: University of California Press, 1990).Google Scholar

16 For statistical testing of the model see David Carment and Patrick James, “Secession and Irredenta in World Politics: The Neglected Interstate Dimension,” in Carment and James, eds., The International Politics of Ethnic Conflict, and Carment, David, “The International Politics of Ethnic Conflict: The Interstate Dimensions of Secession and Irredenta in the Twentieth Century” (unpublished doctoral dissertation, McGill University, 1994).Google Scholar

17 Rothschild, Joseph, Ethnopolitics: A Conceptual Framework (New York: Columbia University Press, 1981).Google Scholar

18 Institutional constraint is not synonymous with democracy. Instead, three dimensions relate to the executive, legislature and general public: (1) executive constraint ranging from seizure of power to competitive elections; (2) executive regulation ranging from unlimited authority to legislative parity; and (3) regulation of participation ranging from no formal arrangements to formally institutionalized ( Morgan, Clifton T. and Campbell, Sally H., “Domestic Structure, Decisional Constraints, and War,” Journal of Conflict Resolution 35 [1991], 187–211).CrossRefGoogle Scholar

19 Lijphart, Arend, “Consociation and Federation: Conceptual and Empirical Links,” this Journal 22 (1979), 499–522.Google Scholar

20 Horowitz, Donald, “Patterns of Ethnic Separatism,” Comparative Studies in Society and History 23 (1981), 165–95.CrossRefGoogle Scholar

21 Lustick, Ian, “Stability in Deeply Divided Societies: Consociationalism versus Control,” World Politics 31 (1986), 352–74.Google Scholar

22 Putnam, “Diplomacy and Domestic Politics,” 426–60.

23 In theoretical terms, this relationship can be best understood through the use of two overarching concepts: opportunity and willingness. Ethnic affinities and cleavage provide the opportunities for states to intervene. The willingness to take an extreme action is counterbalanced by a state's internal ethnic composition and institutional arrangements. Thus it would be expected that low-constraint states are more likely to utilize force in exploiting the opportunities afforded by cleavage and affinity. Varying levels of willingness distinguish different responses to the opportunity, and danger, represented by ethnic cleavages and affinities.

24 This initial exploration of the typology within the context of research on two-level games is consistent with George's advice concerning “systematic progression of well selected controlled comparisons.” A subsequent priority would be to probe for differences between the intermediate categories, that is, Types I_b and II_a ( George, Alexander, “Case Studies and Theory Development: The Method of Structured, Focused Comparison,” in Lauren, Paul Gordon, ed., Diplomacy: New Approaches in History, Theory and Policy [New York: Free Press, 1979]).Google Scholar

25 Gagnon, V. P., “Ethnic Nationalism and International Conflict: The Case of Serbia,” International Security 19 (1994), 130–66.CrossRefGoogle Scholar

26 Banac, Ivo, The National Question in Yugoslavia (Ithaca: Cornell University Press, 1984)Google Scholar, and Flere, Sergej, “Explaining Ethnic Antagonisms in Yugoslavia,” European Sociological Review 7 (1991), 183–93.CrossRefGoogle Scholar

27 Ramet, Sabrina P., Nationalism and Federalism in Yugoslavia, 1962–1991 (2nd ed.; Bloomington: University of Indiana Press, 1992).Google Scholar

28 Ethnicity as the primary means of mobilization, and the subsequent polarization of political issues based on Serb-Croat rivalries played key roles in the initial escalation of the conflict. Originally, the idea of “panslavism” maintained Yugoslavian cohesion. The idea at one time had both domestic and international imperatives. Under Tito, the unity of the South Slavs provided a politically viable internal structure. It insulated the region from outside interference and, in that way, promised a coherent vision of the future that was better and more peaceful than the past. Among Yugoslavia's various ethnic groups, particularist identities and competing visions of the future developed in the decade after Tito's death.

29 Equally important has been Yugoslavia's demography, which has ramifications not only for Yugoslavia as an entity but also for the various republics and provinces. All its ethnic groups, including Croatian, Slovenian, Muslim and Serb, tend towards geographic distribution and, in the other states, including Croatia and Bosnia, Serbs constitute significant minorities. Although not a majority of the former Yugoslavia's population, Serbs always constituted its single largest ethnic group (see Ramet, Nationalism and Federalism in Yugoslavia).

30 The Bosnian and Croatian conflicts also featured short-lived attempts at second-order secession by Serbs in Croatia and Serbs and Croatians in Bosnia (minorities within minorities). As a result, three interstate ethnic conflicts unfolded simultaneously. The first consisted of the secessions of Yugoslavia's republics. The second was the irredentist struggle, meaning territorial retrieval by the Serbdominated JNA and irregular forces. The third process included simultaneous declarations of independence by the self-styled minority Serb and Croatian governments. A fourth process may well include the formal absorption of parts of Bosnia into Croatia and Serbia.

31 Ultranationalist views seemed to have a more direct influence on the formation of Serbian foreign policy. For example, Serb intellectuals played a prominent role in the development of a pan-Slavic doctrine, which was used as the basis for intervention in Bosnia and repression in Kosovo. A key contributor to this manifesto, written in 1986, was Dobrica Cosic, a Serbian novelist and fervent nationalist who had been expelled from the Serbian Communist party for opposing recognition of the Bosnian Muslims. On June 15, 1992, Cosic was elected president of the new Yugoslavia (Serb-Montenegro) ( Ramet, Sabrina P., “War in the Balkans,” Foreign Affairs 71 [1992], 79–98).CrossRefGoogle Scholar

32 Ramet, Nationalism and Federalism in Yugoslavia, 264.

33 Of Yugoslavia's three crisis areas, Bosnia-Hercegovina is the most complex. During the years of Turkish rule, the majority of the area's Christians were converted to Islam. In 1870, Bosnia fell under Austro-Hungarian occupation, in 1908 it became a full part of the empire and in 1918 it incorporated into the new Kingdom of Serbs, Croats and Slovenes. In 1946 it became one of Yugoslavia's six socialist republics. Bosnia is the one state in which Slav Muslims make up most of the population. Although the Muslims are a numerical majority (44% of the population), they never possessed the equivalent political clout and military power of their numerically smaller Serbian and Croatian counterparts. At crisis onset, Bosnia was led by a coalition government comprising representatives of all three ethnic groups. Operating outside this coalition, the Bosnian SDS, a Serb breakaway group led by Radovan Karadzic, armed with JNA equipment, had already proved successful in stalling any political solution to the future of Bosnia. This stalling tactic was aided by the fact that the majority of the JNA was stationed in Bosnia-Hercegovina before the conflict, and the republic was the site of most of the Federal Army's weapons factories.

34 Three stages of interaction can be identified. First, an aggressive Serbian foreign policy emerged with respect to the newly created states of Bosnia, Croatia and Slovenia. This policy depended on the ability of Serb leader Milosevic to build a coalition of forces willing to restructure the neighbouring states so as to create a “Greater Serbia.” At stage two, Serb aggression, which arose largely from domestic-policy concerns, is re-interpreted as a security issue for Serbia. This is because of the perceived threat to Serbs living in Croatia and Bosnia. At stage three, Serb leaders took advantage of the cleavages created by Serb-held enclaves in Croatia and Bosnia and the crisis escalated into war. The conflict in Croatia had decisive spillover effects for the conflict in Bosnia-Hercegovina, not the least of which was the forcing of Bosnian leadership (as well as Macedonia) to choose either independence or incorporation into a truncated Yugoslavia.

35 None of these linkages proved strong enough for regional warfare, but they have been instrumental in shaping the course of events. Hungary, for example, placed its air force on alert during the Croatian war. Arms shipments quickly flowed from Germany to Slovenia and Croatia. Albania responded by placing its armies on a state of alert and declared full support for the creation of an independent Kosovo. Bulgaria issued a statement to the effect that its army would not threaten Yugoslav security but would recognize an independent Macedonian state. When Macedonia declared independence, Greece and Albania mobilized their troops along the borders. The fear was that Belgrade would next turn to mass expulsions of Albanians from the province of Kosovo and Macedonia. Bitter dispute with Athens over the independence sought by the Yugoslav republic of Macedonia also created heightened tension between the two countries. As a result, the UN deployed 1,700 troops in Macedonia as a trip wire against possible expansion of the conflict (UNPREDEP).

36 Ramet, “War in the Balkans,” 79–98.

37 Glenny, Misha, “What Is to Be Done?” New York Review of Books, May 1993, 14–17.Google Scholar

38 Gagnon, “Ethnic Nationalism and International Conflict.”

39 Cohen, Lenerd J., “The Disintegration of Yugoslavia,” Current History 91 (1992), 369–75.Google Scholar

40 Space does not permit case study of the two middle types. Within Type I_b (meaning diversity and low constraint), issues important to a specific ethnic group (such as retrieval of ethnic brethren from another state) are unlikely to generate interest because the payoffs are small. If the military is from a group different than the majority, or the elite represents more than one ethnic group, then it normally is impolitic to pursue a confrontational, interventionist strategy. The size of the stage one win-set will be smaller, with suboptimal outcomes. Granting ethnicity a central place in foreign policy is dangerous because it may incite potential enemies within the state to seek support from neighbours. Type I_b states are not immune to involuntary defection; ethnic diversity increases the chance that leaders of a group cannot or will not cooperate on certain policies. These states can be expected to support means to monitor and control defection. This would include international rules and institutions that derive from mutual vulnerability.

41 Type II_a cases include high institutional constraints and a dominant, “like-minded” ethnic group. The elite is expected to go along with the ethnically oriented sentiments of the masses, because replacement is quite possible. The main problem is to manage the discrepancy between the expectations of political friends and enemies on one hand, and the outcome on the other. A leader must be ready for opposition tactics that are influenced heavily by ideology, a sense of historical injustice, perceived grievance or a threat to values that justify a future society. Concerns about involuntary defection from an agreement will be para-mount. As opposed to either Type I_a or I_b, the elite does not lead. “Like-minded” refers to those cases in which even political opponents offer all-purpose backing for an overt interventionist strategy. This support is crucial to expanding the stage one win-set. Indeed, if it promises domestic payoffs to all elites, leaders may pursue an aggressive foreign policy even if it appears to be costly (below a certain threshold) in the international arena.

42 For an analysis of the domestic dimension of the commitment problem see Fearon, James D., “Commitment Problems and the Spread of Ethnic Conflict,” IGCC-San Diego Project on the International Spread and Management of Ethnic Conflict, (unpublished ms., 1995)Google Scholar, and Weingast, Barry R., “Constructing Trust: The Political and Economic Roots of Ethnic and Regional Conflict” (unpublished ms., Stanford University, 1995).Google Scholar

43 Downs, Anthony, An Economic Theory of Democracy (New York: Harper-Collins, 1957), 141.Google Scholar

44 Both the AIADMK and the DMK called for some form of Indian intervention, United Nations mediation and self-determination for Sri Lankan Tamils. The more moderate AIADMK had to play the ethnic game in order to prevent the DMK from capitalizing on its inaction. For example, Ramachandran pressured the Indian government to issue Indian passports to some of the rebel leaders in order to facilitate their movement within India and abroad ( Khory, Kavita R., “Separatism in South Asia: The Politics of Ethnic Conflict and Regional Security,” paper presented at the annual meeting of the American Political Science Association, Chicago, 1992).Google Scholar

45 Silva, K. M. De, “The Making of the Indo-Sri Lanka Accord: The Final Phase, June-July 1987,” in Silva, K. M. de and Samarasinghe, S. W. R. de A., eds., Peace Accords and Ethnic Conflict (New York: Pinter, 1993), 112–55Google Scholar, and Sivarajah, A., “Indo-Sri Lanka Relations and Sri Lanka's Ethnic Crisis: The Tamil Nadu Factor,” in Kodikara, Shelton, ed., South Asian Strategic Issues: Sri Lankan Perspectives (New Delhi: Sage, 1990), 135–46.Google Scholar

46 Gandhi had an election coming up in 1985 and Congress preferred an electoral pact with the ruling AIADMK, which hoped to make political mileage out of the Tamil issue, to the point of demanding that India take direct military action against Sri Lanka. The rebels were allowed to build up arsenals of arms in Tamil Nadu, run training camps and ship military hardware across the Palk straits. Retired Indian officers trained the militants in guerrilla warfare ( Venkateshwar, P. Rao, “Ethnic Conflict in Sri Lanka: India's Role and Perception,” Asian Survey 28 [1988], 419–36).Google Scholar

47 India Today (New Delhi), December 15, 1985.

48 Asiaweek (Hong Kong), November 23, 1986.

49 Ibid.

50 In this arrangement the ethnic percentage of the Tamils would go up to 48 per cent, Muslims to 37 per cent and the Sinhalese would drop to 14 per cent, ensuring a Tamil majority in the Eastern Province but not an absolute majority. Institutional links would also be established between the Northern and Eastern Provinces.

51 The Hindu (New Delhi), February 3, 10 and 11, 1987.

52 Ibid., February 12, 1987.

53 Pfaffenberger, Bryan, “Sri Lanka in 1987: Indian Intervention and Resurgence of the JVP,” Asian Survey 28 (1988), 137–47.CrossRefGoogle Scholar

54 Kohli, Atul, Democracy and Discontent: India's Growing Crisis of Governability (Cambridge: Cambridge University Press, 1990)Google Scholar, and Brass, Paul, Ethnicity and Nationalism: Theory and Comparison (Newbury Park: Sage, 1991).Google Scholar

55 On June 3, 1987, the Indian government sent relief supplies to the people of Jaffna in a flotilla of 19 fishing boats flying the Red Cross flag. Colombo rejected the supplies and blocked the Indian flotilla's entry into Sri Lankan waters. The government of India once again condemned the government of Sri Lanka's actions, warning that it would not remain an indifferent spectator to the plight of the Tamils in Sri Lanka. Tamil militants inflamed the situation by slaughtering 29 Buddhist monks on their way to an ordination ceremony. Indian newspaper editorials called for an armed invasion of Sri Lanka (Pfaffenberger, “Sri Lanka in 1987”). On June 4, 1987, five Indian Air Force supply planes, escorted by Mirage 2000 fighter jets, entered Sri Lanka's airspace and dropped relief supplies in and around Jaffna. The government of Sri Lanka condemned the Indian air drop, known as “Operation Eagle,” as a “naked violation of our independence” and “an unwarranted assault on our sovereignty and territorial integrity” ( Carment, David, “Les dimensions internes des comportements en temps de crise: étude de cas entre l'Inde et le Sri Lanka, 1983–1990,” Études Internationales 23 [1992], 253–76).CrossRefGoogle Scholar

56 After the departure of the IPKF in 1990, and throughout 1991 and 1992, a multiparty parliamentary committee began to consider ways to offer Tamils alternatives to supporting the Tigers. A plan was devised with the proviso that the Northern and Eastern Provinces would not merge. More power would be devolved to the north and east, to protect minority Muslims living in the region. The plan was contingent on Tamil agreement to abandon merging the north and east, which generated significant hope among the participants for a peaceful solution to the conflict. The assassination of President Premadasa on May 1, 1993, possibly by Tamil Tigers, dealt a significant blow to the plan. In less than two weeks, assassins had struck down the country's president and the only other man to challenge him, opposition leader Lalith Athulathmudali. In 1994, Sri Lanka's domestic protracted conflict underwent two significant changes: (1) the election of SLFP leader Kumaratunga, who was dedicated to resolving the Tamil issue, and (2) efforts by a United Nations mission to oversee a reduction in violence in the north. After the collapse of negotiations in the spring of 1995, the SLFP stepped up its military operations against the Tigers with unprecedented success.

57 Hardin, Russell, One for All: The Logic of Group Conflict (Princeton: Princeton University Press, 1995).Google Scholar

58 Leng, Russell J., Interstate Crisis Behavior, 1816–1980: Realism versus Reciprocity (Cambridge: Cambridge University Press, 1993).CrossRefGoogle Scholar

59 For example, in 1993, Karadzic said that “there will be no more negotiations” on further Serb concessions at Geneva peace negotiations (The Globe and Mail [Toronto], August 24, 1993). Ethnic leaders must choose either a settlement imposed on them by third parties or a negotiated solution. For the belligerents these choices must be preferable to continued fighting. Therefore, leaders must be able to convince their supporters to accept the alternative and they must enforce their followers’ observance of it. The problem is that, in Bosnia, disunity and lack of cohesion within the ranks of the adversaries made it difficult for the various factions and the UN to engage in any meaningful form of negotiation. Ethnic leaders lacked both the power and the authority to enforce their decisions and concessions. On December 15, 1995, the United Nations Security Council—acting under Chapter VII of the Charter of the UN—adopted Resolution 1031, which authorized the member states of NATO to establish the NATO implementation force (IFOR), under unified command and control and composed of ground, air and maritime units from NATO and non-NATO nations, to ensure compliance with the relevant provisions of the Dayton Peace Agreement.

60 The United Nations and NATO were working on the premise that if the Bosnian Serbs could be persuaded to reduce their conflictual behaviour, possibly the other sides would reciprocate. International sanctions in combination with discriminant and proportional force were used to induce a “hurting stalemate” and to influence the behaviour of all the participants in the conflict.

61 In effect, the international community sent separate signals to Serb leader, Milosevic, who held a tight reign on SDS leader Karadzic: resolve the conflict through negotiations or face tighter sanctions. The pressure on Milosevic did bear fruit but at a high cost. As before, the Croatian leader Mate Boban and Muslim leader Itzebegovic were receptive to the idea of the plan comprising a ceasefire plan, a political agreement as well as a map reorganizing the former Yugoslav republic into 10 separate regions under a central government. Karadzic's willingness to sign the agreement came after only immense pressure from Milosevic. The concession coincided with the arrival of Milosevic at the Geneva Peace Talks. The proposed map of the new Bosnian state showed that both the European Community and the US, which was then party to the talks, were willing to allow Bosnia to be redrawn along ethnic lines, with the Serbs controlling almost 70 per cent of the former Yugoslav territory. On May 6, 1993, the Bosnian-Serb parliament met and rejected the Vance-Owen peace initiative (The Globe and Mail [Toronto], January 10, 1993, and The International Herald Tribune [Paris], May 7, 1993). The subsequent 1995 Dayton Peace Accord merely reaffirmed what had long been recognized: Bosnia would be two effectively independent nations. Territorial boundaries gained through military means were considered legitimate, but the Serb's stranglehold would be reduced to 49 per cent of Bosnian territory.

62 Specific ethnic groups within the military and the bureaucracy can dominate the state through several means, notably skewed recruitment, and when the ethnic composition of military and civilian leadership is congruent. The key problem is that soldiers who remain on the sidelines will have difficulty putting ethnic affiliations aside, and leaders may use intervention as a means of shoring up domestic support. Consider in this context the military within the states of eastern and central Europe. The military is suffering from extremely poor social conditions, low morale, high levels of absenteeism and low conscription, corruption, inadequate funding and a general loss of purpose. The inability of current governments to address these problems may become a prime reason for the armed forces to give their support to ethnic leaders who promise that their concerns would be addressed. For ethnic leaders there are obvious political benefits.

63 “Ethnic restructuring” of armed forces can reduce the salience of ethnicity as a basis for mobilization, recruitment and advancement, and should be a component of defence conversion. So long as the militaries of these states remain in social disarray they are susceptible to appeals from zealous politicians.

64 In 1992, to ensure safe havens for civilians ensnared in the conflict, NATO began to implement plans for the military enforcement of a no-fly zone. However, the actual enforcement could not take place until the United Nations Security Council had approved it (The Globe and Mail [Toronto], December 6, 1992). Achieving consensus among the NATO member states was painfully slow. For example, in November 1994 United Nations peacekeepers were detained by Bosnia-Serb forces in retaliation for the NATO bombing of Serb-controlled depots near the town of Bihac. Less than a year later, on August 30, 1995, NATO initiated a week-long series of selective air strikes against Serb positions. The operation, called “Deliberate Force,” consisted of attacks by over 60 aircraft and the artillery of the newly created Rapid Reaction Force. The strikes were implemented more than two full years after Serb forces laid siege to Sarajevo and were intended to force the Serbs to remove their artillery aimed at United Nations-designated safe areas. NATO affirmed its determination to act in the same way for the other safe areas. The six “safe areas” (Gorazde, Sarajevo, Bihac, Zepa, Tuzla and Srebrenica) were decreed as such by Resolutions 824 and 836 of the Security Council (NATO Press Release, Internet version, Office of Information and Press, Brussels, August 30, 1995).