Leonardo R. Arriola

PS 311

Prof. D. Laitin

29 November 2001

Bringing the Second Image Back In:An Application of IR Theory to the Study of Ethnic Conflict

Although the field of international relations (IR) has been traditionally concerned with the uses of violence and power among states, its scholars have long ignored these processes when they occur within the boundaries of the very same states.The field’s lack of attention to internal conflict– whether ethnic rebellion, religious war, or ideological revolution – should be surprising, for internal conflict fundamentally affects the integrity of the units that make up the international system as it is studied and understood. The dean of the realist school, Kenneth Waltz, has noted: “We easily lose sight of the fact that struggles to achieve and maintain power, to establish order, and to contrive a kind of justice within states, may be bloodier than wars among them.”[1]

The work of theorizing and accumulating the empirical evidence on internal conflict has generally been left to comparativists, but since the end of the Cold War, IR scholars have also begun to participate in this area of inquiry. Can IR theory, which primarily offers systemic-level explanations for state behavior, provide any insights on what is essentially, from the field’s perspective, a unit-level phenomenon? Does it fill any theoretical gaps that comparativists have overlooked? These questionsare addressed here in the context of ethnic conflict. This paper: (1) briefly discusseswhy ethnic conflict has been ignored by the IR field; (2)examineshow IR scholars conceptualizeethnic conflict; and (3) assesses the limitations of IR theory inaccounting for the origins of ethnic conflict.The field of IR, as is made clear in this paper,has yet to develop hypotheses or approaches that can satisfactorily explain one of the more challenging puzzles in contemporary world politics.

Ignoring Ethnic Conflict

The study of ethnic conflict has been ignored in IR, in part, because the field tends to be driven by foreign policies and current eventsthat primarily concern the interests of great powers.Although the incidence of ethnic conflict was rising throughout much of the Cold War, it rarely affected the strategic interests of either the United States or the Soviet Union – superpowers primarily concerned with maintaining stability in a bipolar world.[2] If the absence of border changes can be considered a measure of the international system’s stability, then ethnic conflictwithin independent states produced little volatility, for it resulted in few, if any, significant border changes throughout the period. American policymakers, and therefore scholars,easily ignored the problem.

The IR field’s disregard for ethnic conflict has been further reinforced by the orientation of its dominant theoretical paradigm – realism.In determining the sources of state behavior, realism privileges systemic-level (third-image) over unit-level (second-image) analysis; the distribution of power in the international system, for instance, is as a systemic variable that receives pride of place among all potential explanatory variables.In his study of the relationship between revolutions and war,for instance, Walt acknowledges that“[s]ystemic theories such as neorealist balance-of-power theory focus on the constraining impact of international anarchy and the effects of different distributions of power, while downplaying the impact of domestic policies, ideology, or other unit-level factors.”[3]Realism’s core assumptions provide the rationale for eschewing unit-level analysis: states are the key actors in world politics; states are unitary-rational actors; states act in their self-interest; international anarchy – the absence of central government to offer security, enforce agreements, or prevent the use of violence – shapes state preferences and actions.

But since the end of the Cold War, ethnic conflicts have demanded greater attention from IR scholars. This type of internal struggle has been associated with the violent emergence of newly-independent states such as Croatia (1992), Bosnia-Herzegovina (1992),and Eritrea (1993), andthe now-transitioning East Timor (1999);[4]moreover, ethnic conflict within weak states has been linked to greater regional instability and conflict in Central Africa, Central Asia, Southeast Europe, and South Asia.

Conceptualizing Ethnic Conflict

IR scholars, in turning their attention to ethnic conflict, have sought to analyze the phenomenon by employing the “security dilemma,” a realist concept typically used to explaininterstate war.[5]Posen, for instance, argues that the security dilemma accurately captures “the special conditions that arise when proximate groups of people suddenly find themselves newly responsible for their own security.”[6]Anarchyis a condition under which ethnic groups, like states in the international system, fear for their basic survival; their overriding interest is to enhance their capabilities for self-defense.

In providing for their own security, however, ethnic groups also present a threat to their rivals for power and resources. Herein lays the security dilemma encountered in anarchy: “What one does to enhance one’s own security causes reactions that, in the end, can make one less secure.”[7]This security dilemma is produced by two conditions: the indistinguishability of offensive and defensive measures, and perceptions of an offensive advantage. First, ethnic groups find it difficult to determine whether the actions of their potential adversaries are intended for offensive or defensive purposes. Posen and Kaufmann concur that this factor is particularly acute in ethnic conflicts, since the military capability of each side is largely dependent on group solidarity, which is often built on a distorted history of rivalry and conflict with their neighbors. “[T]he nationalist rhetoric that accompanies mobilization often seems to and often does indicate offensive intent.”[8]Second, ethnic groups perceive that they have an offensive advantagein the uncertainty produced by anarchy,thereby increasing the likelihood of a preventive war. “If one side has an advantage that will not be present later and if security can best be achieved by offensive military action in any case, then leaders will be inclined to attack.”[9]

IR scholars contend that it is the security dilemma that creates the escalating cycle of violence often observed in ethnic conflicts. The security dilemma thus serves as the independent variable that accounts for variation in the dependent variable of inter-group relations. For example, Posen claims that the conditions of the security dilemma led to ethnic warbetween Croatia and Serbiaas Yugoslavia disintegrated. The Serbs had several incentivesto launch a preventive war: the historical enmity between the two groups led Serbs to attribute offensive intentions to Croat actions; Serbs had an excuse to intervene on behalf of their ethnic brethren in Croatia; and Serbiafearedthat their offensive advantage would erode as Croatia gained international recognition and allies.In contrast, Russia and Ukraine experienced a relatively peaceful parting because the conditions for the security dilemma were absent during the Soviet Union’s dissolution: there was no history of deep animosity between the two groups; Russiaenjoyed considerable advantages in the military balance of power, which wasfurther stabilized by nuclear weapons in both countries; Russia did not fear the erosion of its power; and the Russian diaspora could be easily defended by Russianmilitary forces in the event of a conflict.

The IR treatment of ethnic conflict is supplemented by prescriptive recommendations. According to the realist analysis,cooperation between ethnic groups is difficult to sustain under anarchy. All fear that cheating by the other side will leave them weaker. Theorists like Mearsheimer, Van Evera, and Kaufmannconclude that a durable solution to ethnic conflict can only be attained by going beyond the partitioning of splintered states; ethnic groups must be physically separated to minimize perceptions of threat and maximize their capacities for self-defense.[10]Kaufmann argues that war cannot end until ethnic groups are separated into defensible, homogenous regions: “Ethnic separation does not guarantee peace, but it allows it. Once populations are separated, both cleansing and rescue imperatives disappear; war is no longer mandatory.”[11]Taking the concentration of ethnic groups as the independent variable to explain how ethnic wars are resolved, Kaufmann finds that those ethnic civil wars which ended in a negotiated agreement to preserve the state generally involved regionally-concentrated minorities, resolutions that provided for regional autonomy, and a significantly lower fatality rate than conflicts ending in suppression.[12]

The Limits of IR Theory

IR theorists contend that the security dilemma shows considerable explanatory power when applied to recent cases of ethnic conflict; however, these cases also underscore the limitations of their theory. Realist assumptions become problematic when transferred to unit-level analysis. By blindly assuming ethnic groups are unitary actors and by taking anarchy as an exogenous condition, IR scholars fail to identify the causal mechanisms involved in ethnic conflict or to address the potentially more fruitful research question of why states fall apart in the first place.

Just as states are assumed to be unitary-rational actors in international affairs, ethnic groups are understood to be cohesive and organized. This assumption is of central importance to the security dilemma explanation, for it is the ability of an ethnic group to act as a unit that presents a threat to its adversary: “the ‘groupness’ of the ethnic, religious, cultural, and linguistic collectivities that emerge from collapsed empires gives each of them an inherent offensive military power.”[13]This assumption of groupness, however, requires more explication than IR scholars currently provide. A more comprehensive theory of ethnic conflict should address the question that Walt puts forth in the specific context of revolutions: “how are individuals persuaded to bear the costs and risks of revolutionary activity, and how do they sustain their commitment in the face of prolonged uncertainty, danger, and other difficulties?”[14]Kaufmann and other scholarsrespond by accepting that not every individual “may be mobilized as an active fighter his or her own group,” yet they continue to merely assert that “ethnic identities are fixed by birth.”[15]As long as such claims remain unsubstantiated, the explanatory power of the security dilemma is limited.

According to realism, the security dilemma only exists under anarchy, or in the case of state dissolution, anarchy-like conditions. The issue to determine, therefore, is how anarchy comes about in the first place.Why do states break down to produce anarchic conditions? Does ethnic conflict lead to state dissolution, or does the breakdown of the state lead to ethnic conflict?The security dilemmacannot answer these questions. IR theorists, once again, give unsatisfactory answers.Kaufmann ignores the question by claiming that “solutions to ethnic wars do not depend on their causes.”[16]Posen gives a partial answer in arguing that ethnic conflict is produced by the process of “imperial collapse” rather than the “nationalist card.”[17]Posen’s arguments may be limited to ethnic conflict as it occurs between states rather than groups, for his units of analysis are state-dyads, i.e., Croatia-Serbia or Russia-Ukraine, rather than inter-group relations within a single state, i.e., Croats versus Serbian Croats or Ukrainians versus Russian Ukrainians.

Conclusion

Can IR theory complement the work comparativists have already done in the area of ethnic conflict? The security dilemma, at best,is a partial model. It provides some insight on how inter-group conflicts are perpetuated through mutual distrust and insecurity, and offers some provocative prescriptions for ending these conflicts. Realism’s conceptualization of ethnic conflict, however, does not satisfactorily explain why such wars are initiated, and its assumptions – the unitary actor and anarchy – may, in fact, be part of the research question to be explored.The lesson here is not that realism’s explanatory variables are invalid, but what should be clear is that IR scholars have not adequately specified the dependent variable of this puzzle.Should inter-group relations or the onset of ethnic conflict be on the left-hand side of the equation, or would it be more profitable to investigate the causes of state breakdown?

1

[1] Kenneth N. Waltz, Theory of International Politics (New York: McGraw-Hill, Inc., 1979), 103.

[2] Fearon and Laitin show that the occurrence of civil wars has risen steadily since the 1950s and 1960s; it is not a post-Cold War phenomenon, but a difficult one to terminate. See James D. Fearon and David D. Laitin, “Ethnicity, Insurgency, and Civil War,” paper presented at the annual meeting of the APSA, San Francisco, September 2001. Available at

[3]Stephen M. Walt, “Revolution and War,” World Politics, 44, 3 (April 1992), 321.

[4]Correlates of War Project, “Interstate System Membership List, 1816-1993.” Available at

[5] For a detailed discussion of the security dilemma, see Robert Jervis, “Cooperation Under the Security Dilemma,” World Politics, 30, 2 (January 1978), 167-213.

[6] Barry R. Posen, “The Security Dilemma and Ethnic Conflict,” in Michael Brown, ed., Ethnic Conflict and International Security (Cambridge: MIT Press, 1997), 103.

[7] Ibid., 104.

[8] Chaim Kaufmann, “Possible and Impossible Solutions to Ethnic Civil Wars,” International Security, 20, 4 (Spring 1996, 147.

[9] Posen, 109.

[10]Mearsheimer and Van Evera apply the logic of the security dilemma to the Bosnian case to argue for the separation of Muslims, Croats, and Serbs into separate states. See John J. Mearsheimer and Stephen Van Evera, “When Peace Means War: The Partition That Dare Not Speak its Name,” The New Republic (18 December 1995). Available through Lexis-Nexis.

[11] Kaufmann, 150.

[12] Ibid., 161.

[13] Posen, 106.

[14] Walt, 335.

[15] Kaufmann, 140-141.

[16] Ibid., 137.

[17] Posen, 105.