Ethnic Conflict and International Politics: Explaining Diffusion and Escalation
Edited by Steven E. Lobell and Philip Mauceri
palqrave
macmiUan
ACKNOWLEDGMENTS
CHAPTER 1
Diffusion and Escalation of Ethnic Conflict
Steven E. LobeU and Philip Mauceri
Many people and institutions contributed to the development of this volume. We would particularly like to thank our colleague Dhirendra Vajpeyi who encouraged us to think about ethnic conflict and international relations, and whose continued support was vital. The project took shape at a workshop on "Ethnic Conflict and Human Rights," sponsored by the United States Institute of Peace and held at the University of Northern Iowa in 2001. We would like to thank Jeffrey Helsing from United States Institute of Peace, Susan Koch, Catherine Palczewski, and Kent Sandstrom, for sharing their ideas with us during and after the workshop. In addition, various parts of this project were presented over the years at professional meetings, including the International Studies Association's Annual Convention in Los Angeles, CA. Neal Jesse, Kristen Williams, and Henry Teune offered us important insights at this meeting. For assistance in bringing this volume to completion, we wish to thank the Department of Political Science and the College of Social and Behavioral Sciences at the University of Northern Iowa, our reviewers, including Sandra joireman and Donald Rothchild, as well as David Pervin at Palgrave MacMillan. Finally, we both wish to thank our families for the support and patience they offered.
Introduction
In this book we offer an approach to understanding the internationalization of ethnic conflict in different regional contexts that integrates international relations and comparative analysis. We examine four core explanatory frameworks that contribute to the diffusion and the escalation of ethnic conflicts in divided states and societies. These explanations are at the nexus of comparative and international understandings of conflict. Much of the literature on ethnic conflict focuses on the origins of ethnic identity (instrumentalists versus constructivists; Smith 1986, 1993; Kaplan 1993; Connor 1994; Arfi 1998) or on the sources of ethnic conflict (Lake and Rothchild 1998).1 We focus on the link between ethnic and interstate conflict, and specifically, on the internationalization of ethnic conflicts. We recognize the internationalization of ethnic conflict as requiring an understanding of "intermestic" structures and processes.2
By treating the state as a unitary actor that pursue national interests and international relations as the interaction of sovereign states, many international relations theorists ignore the domestic political environment that characterizes most divided states. For instance, realists assume that the central government commands the obedience of different groups under its authority and thereby controls what goes on within its borders (Waltz 1979; Grieco 1990; Mearsheimer 2001). International conflicts are viewed largely as interstate conflicts involving unitary state actors. Yet, internal armed conflict is much more prevalent than interstate conflict. Since the end of World War II, most wars have been internal conflicts, a trend that accelerated in the aftermath of the end of the Cold War. Between 1989 and 1994, there were no fewer than 99 civil wars, with 80,000 deaths in the 1993-1994 period alone (Gurr 1994; Wallensteen and Sollenberg 2000). The majority of these internal conflicts had a strong ethnic dimension. Although few states have ceased to exist, even in the third world (David 1991), states experiencing internal ethnic conflict face serious threats to their structures and regimes, which are often of a much higher order in comparison with external threats (Ayoob 1991; Gause 1992).What has increasingly been referred to as "state failure,"
2
Steven Lobell and Philip Mauceri
Diffusion and Escalation of Ethnic Conflict 3
that is the collapse of the central state apparatus and its ability to control national territory, has been largely the result of spiraling internal ethnic conflicts and not interstate wars (Carnegie Commission 1997).
Few comparative political analyses focus on the role of international structures and politics to explain domestic conflicts, as Theda Skocpol (1979: 18-19) long ago pointed out. Whether analysts focus on culture or structure, what often results is an approach that comes close to suggesting the enclosed and sui generis nature of the conflict being examined (McAdam, Tarrow and Tilly 1997: 143). Conflict is viewed as the result of purely internal disputes and histories, only indirectly influenced by external dynamics.Yet international conditions can provide favorable or unfavorable opportunities for access to resources, legitimacy and coalition partners. Most theoretical approaches that focus on the impact of international politics in domestic conflict remain limited largely to the economic policy making process. Dependency school theorists (Amin 1976; Cardoso and Faletto 1979; Evans 1979) were the first to draw attention in systematic ways to the important influence exercised by multinational corporations, foreign governments, and international financial organizations in the internal policy making processes and political coalitions of underdeveloped countries. More recently, growing market integration and technological changes have renewed attention to the links between policies and their outcomes in different states (Keohane and Milner 1996). Still, even in the area of political economy, many, such institution-oriented analyses, continue to focus on domestic economic structures in determining policy outcomes and coalitions.
Many realist accounts accentuate systemic pressures as determining but ignore the influence of domestic politics. Comparative analyses that emphasize internal factors grant domestic coalitions primacy but neglect the importance of international politics. In this book we integrate this literature. We argue that "inter-mestic" forces, and especially the entangling of domestic and international pressures, and their internal and external reverberations, contribute to the diffusion and the escalation of ethnic conflict (Gourevitch 1978; Putnam 1988; Muller and Risse-Kappen 1993; Risse-Kappen 1995; Keohane and Milner 1996; Lobell 2003).
We believe that there are significant differences within states, with multiple actors that influence a state's international policies and its position in the global system. Moreover, international actors play a major role in the possibility of constructing domestic political coalitions, not only through direct intervention but by supporting änd/or opposing them. In an analysis of the relation between internal ethnic conflict and international relations, it is useful to view the state as consisting of a ranked or unranked divided ethnic systems (in most system there are more than two ethnic groups; Horowitz 1985:21-36). In the ranked system, one ethnic group is superordinate and the other is subordinate. In this system subordinates can try to displace süperordinate groups or raise their own position without denying the legitimacy of the hierarchy. In the latter system, neither is subordinate to the other: Instead, parallel ethnic groups coexist, perhaps autonomously,; with each "peak" group internally stratified. In this system there is strong pressure for educational and occupational proportionality by both ethnic groups.
We examine two types of interstate ethnic conflict. Diffusion (or contagion) entails igniting conflict in other states or the spillover processes by which conflicts in one country directly affect neighboring countries (Migdal 1988; Gurr 1993; Lake and Rothchild 1998). Escalation involves the drawing or pulling in of other states, non-state actors, or outside ethnic groups into the internal conflict. We examine four explanatory frameworks to understand the escalation and diffusion of interstate ethnic conflict.
The Weakening of State Institutional Structures, Particularly Where There has been a Preexisting "Ethnic Contract"^
Ethnic contracts in divided societies entail formal guarantees of power sharing, minority rights, state patronage, and jobs in the public and private sector.4 Ethnic contracts specify the rights and responsibilities, political privileges, and access to resources of each ethnic group. Some of the safeguards include power sharing arrangements, proportionality, electoral rules, and mutual vetoes (Lake and Rothchild 1998: 13—14). When such safeguards for nnnority ethnic groups are effective, they feel secure in relation with the state and the ethnic majority about their future (Lake and Rothchild 1998). As Van Evera notes (1994), such a minority-respecting nationalism will grant rights to other nationalities (while a minority-oppressing nationalism denies such right to these other nationalities, subjugating them instead).
One of the basic functions of the modern state is to provide internal order and security. Posen (1993), Snyder (1993), and Saideman (1998) emphasize the emerging anarchy that results from the weakening of the state. Severe economic crises, demographic pressures, deterioration of infrastructure and public services, external conflicts, or internal conflicts between actors regarding the legitimacy of existing institutional arrangements can undermine the ability of the state to guarantee security within its territory (Migdal 1988; Baker and Ausink 1996). At the extreme, the result is the complete collapse of the state.
The weakening of state institutional structures, particularly where there has been a preexisting ethnic contract, will create insecurities on the part of vulnerable ethnic groups. When the central authority declines, groups become fearful for their survival. Under such conditions, each ethnic group will look to their own devices for protection against others; safeguards are absent to guarantee their political status or economic and even physical security.
Because the state has a monopoly of force and authority, its resources can be an ethnic group's greatest ally or its adversary. One outcome of the weakening of state structures is competition among ethnic groups for control of the state, its apparatus, and its resources. The position of each group is that if my group does not capture the state, another group will, putting my group at the mercy of the state. As Saideman notes, "If the state cannot protect the interests of all ethnic groups, then each group will seek to control the state, decreasing the security of other groups and decreasing the ability of the state to provide security for any group" (Saideman 1996: 23). Thus, the collapse of the state and the loss of its
4
Steven Lobeil and Philip Mauceri
Diffusion and Escalation of Ethnic Conflict 5
control over society will create a commitment problem that arises when two ethnic groups find themselves without a third party that can credibly guarantee agreements between them (Fearon 1998).
Escalation
Outside ethnic groups will join ongoing internal ethnic conflicts when they possess opportunities. Where the central authority has recendy collapsed (i.e., competition waged among different ethnic groups for control of the state), predatory outside ethnic groups will take advantage of windows of vulnerability and opportunity in order to capture the spoils (Posen 1993; Van Evera 1999). In this instance, predatory ethnic groups consider states with significant internal conflicts to be an easy target. External support can also be used in order to prolong the conflict and to drain its enemies economy, thereby weakening it from within.
Diffusion
Ethnic security dilemmas can lead to the escalation and the diffusion of ethnic conflict (Posen 1993). With the collapse of the state structure, ethnic groups will take measures to protect themselves, practicing self-help, to ensure their security. These domestic measures may be seen as threatening by neighboring ethnic groups, especially in instances of enduring external ethnic rivalries. The outcome can be a hostility spiral of action and reaction, resulting in the diffusion of the conflict. Ethnic security dilemmas are particularly intense when offense and defense balances are indistinguishable and when offensive military operations are more effective than defensive military operations.5 Under such conditions, ethnic groups will engage in offensive strategies in order to survive.
The Role of Non-State Regional and International Actors in Reducing State Sovereignty and Empowering Domestic Actors
Intergovernmental organizations (IGOs) are composed of states and the individuals who are sent as delegates to such organizations representing the interests and policies of their home governments. Nongovernmental organizations (NGOs) consist of individuals or national groups who are not official representatives of national governments. Important NGOs include religious bodies, ethnic and religious political parties, and advocacy and militant groups.
Divided states, weak states, and states undergoing formation are likely to have relatively porous borders (Gause 1992).6 In such states, trans-border communities from kin states can reach out to their ethnic diasporas through transnational ideologies, IGOs, and NGOs to promote dissent against the ruling ethnic group and to challenge the Westphalian sovereign state system (King and Melvin 1999/2000). Expatriates and foreign activists can intentionally heighten a nation's awareness of its divisions. Such states will have a lowered ability to fend off military interventions and domestic subversions from the outside.
Escalation
Transnational penetration and foreign meddling in domestic politics will be more effective against divided societies, becoming the batdeground for outside powers, and thereby contributing to the escalation of ethnic conflict. Members of external ethnic groups are concerned about the welfare and condition of ethnic kin in neighboring states. The status of the ethnic group can have an important impact on the intensity of conflictual relations between the states. Transnational ethnic linkages will have a greater influence on the interactions of states if the members of the ethnic group in the target state are in the minority and the members of that same minority group are in power in the neighboring state (Moore and Davis 1998). Escalation is most likely to occur when a disadvantaged ethnic minority is politically active in an effort to modify their status. In such instances, neighboring kin will intervene through IGOs, NGOs, or state action to support their brethren's cause. Support can take many forms, ranging from popular encouragement to covert training, arms supply, and financing of armed groups.
Diffusion
Ethnic kin can use NGOs to transcend weak state boundaries and threaten ruling ethnic groups in neighboring states (Gause 1992). Few states have the ability to project their power over long distances; divided states lack the weapons and the logistical capability for direct protracted conflict beyond their borders. As Walt notes, most threats are from approximate neighbors (1987).
Heightened ethnic awareness allows outsiders to appeal to their ethnic members in neighboring states through IGOs and NGOs. Ethnic groups are likely to appeal to neighboring ethnic kin to destabilize the majority ethnic group. This will be accomplished through supporting dissident groups by groundswells of popular support, questioning the legitimacy of the ruling regime, redrawing state boundaries, and/or overthrowing the ruling regime that opposes them. The dominant ethnic group will view such manipulation as a direct threat to the independence of the state, reacting negatively toward the state directing the campaign. The more intrusive the penetration, the greater the backlash from the target state.
Changes in the Ethnic Balance of Power that Have Real or Potential Repercussions in the Competition for the Distribution of Social, Economic, and
Political Resources
When the balance of ethnic power is stable, conflict is unlikely. Problems arise when the balance of ethnic power shifts, resulting in demands by the new majority for a redistribution of the economic, political, social, and military resources and a revision of the existing ethnic contract (Lake and Rothchild 1998). Shifts in ethnic power differentials can result from demographic changes due to migration and differential birth rates, and uneven rates of economic development such as ethnically differentiated modernization (Horowitz 1985: 102).
6
Steven Lobeil and Philip Mauceri
Diffusion and Escalation of Ethnic Conflict
7
Underrepresented in politics and the public and private sector, the new majority is likely to demand a reallocation of political and economic entitlements commensurate with its new position or in proportion to the changing numbers. These changes will embolden the new majority and threaten the newly but entrenched minority.
Escalation
As the influence of one ethnic group declines, there is the incentive for aggression by both the new ethnic majority and the new ethnic minority, and outside ethnic kin to engage in preventive aggression.7 There is ethnic apprehension that decline will be accompanied by a weakening of one's bargaining position and a corresponding decline in the political, economic, cultural, and other benefits that one receives from the status quo. The weakened side is tempted to engage in preventive aggression before its relative power is expected to decline. By attacking immediately, the group can enhance its chances for long-run security by blocking or retarding the rise of the ethnic challenger while that opportunity is still available (Levy 1987).The concern is that changes in the distribution of ethnic power will embolden die new ethnic majority to impose its will on the new ethnic minority. The distrust of the concentration of power in the hands of the new ethnic majority and the belief that power corrupts will contribute to the escalation of the conflict. As well, the rising challenger might initiate aggression rather than wait to be pushed down its power curve.
Through a process of omni-balancing, the threatened groupfs) will appeal to secondary outside groups in order to balance against the primary internal threat (David 1991). In such instances, the threatened group will protect itself, at the expense of the interests of the state. Alliances, especially ethnic alliances, will draw neighboring kin into the internal ethnic conflict.
Power shifts are most likely to result in escalation under several conditions: when the ethnic challenger is of such a size that at its peak it will roughly equal the dominant ethnic group; if the rise of the ethnic challenger is rapid; if the dominant ethnic group is inflexible in its policies; if there is not tradition of friendship between the ethnic groups; and if the ethnic challenger sets out to revise the status quo (Organski 1968: 376).
Diffusion
Upsetting the balance of power in one state can have a chain reaction or fusion reaction on neighboring power distributions. The new ethnic majority are likely to assist their brethren in neighboring states, especially if they are an abused minority, upsetting the domestic balance of power.
The Degree of Economic, Social, and Cultural Integration within the Regional or Global System (Especially Reinforcing Versus Crosscutting Interstate Cleavages Between/Among Ethnic Groups)
Political and economic differentials exist among ethnic groups. These include access to positions of political power, civil service, inequalities in income, access
to higher education, and the presence in professions (Gurr 1993). Horowitz distinguishes between those groups that enjoy privileged access to resources and an interest in economic modernization, and those that do not (1.985: 166).8 A group may be "modern" because it is disproportionately educated and represented in the civil service and the independent professions, disproportionately wealthy and well represented in business. A "modern" group is one that has benefited from opportunities in education and nonagricultural employment. They are represented above the mean in number of secondary and university graduates, in bureaucratic, commercial, and professional employment, and per capita income.
A group may be more "traditional" because it is disproportionately rural, in subsistence rather than the cash economy, or disproportionately poor or uneducated. Such groups are less favorably situated in terms of educational attainment, high-salaried employment, and per capita income (Horowitz 1985: 233). While the modernizing ethnic group is more interested in education and new opportunities to be tied into the modern economy and its international dimensions, especially in the West, traditional ethnic- groups have some inhibitions on taking up new opportunities and remain somewhat apart from full participation in the modern sector of the economy (Horowitz 1985:148). It should be noted that these differences cannot usually be solely ascribed to culture or region. Traditional groups have most often been the historical victims of political, ethnic, or religious oppression, either domestically an/or internationally (e.g., colonialism) and their disadvantageous position in a particular state clearly reflects that prior history. Moreover ongoing discrimination and oppression may also be a factor limiting access to resources.
An additional factor in this dynamic is the process of globalization, which will create domestic winners and losers. Globalization, understood in its most basic sense of a growing integration of political, social, economic and cultural structures and relations across borders, involves a reduction in identity with territoriality and jurisdictional boundaries. Nonetheless it can also be boundary-heightening (Rosenau 1994; Rothchild and Groth 1995) by producing a backlash of nationalism and religious fundamentalism, or what is termed localization. Where globalization allows people, goods, information, norms, practices and institutions to move about despite boundaries, localization involves narrowness and a pressure to withdraw, highlighting borders and attachment to land. This latter process inhibits the movement of people, goods, information, norms, practices, and institutions. Winners from globalization will benefit from the expansion of production, trade, and investment beyond national borders. Losers from globalization will be harmed due to loss of jobs, the displacement of native industries by foreign ones, the loss of cultural icons, and the erosion of belief systems. Not surprisingly, according to Horowitz, "there is much evidence that so-called backward groups are more frequently initiators of ethnic violence and advanced groups more frequent victims" (1985:166).
■ Diffusion
; It is clear that traditional ethnic groups are losers from globalization. ::;Apprehensive about their survival (economic,political, and cultural), such groups :-are likely to respond to globalization by engaging in ethnonationalism to
8
Steven Lobell and Philip Mauceri
Diffusion and Escalation- of Ethnic Conflict
9
undermine advanced groups (Rothchild and Groth 1995). Ethnonationalism strengthens the sense of identity and mutual belonging to an ethnic group by its emphasis upon past sufferings and present or future threats at the hands of outsiders to the group. Traditional groups will use diversionary conflicts with opposing internal and/or external ethnic groups to facilitate the process of localization. Such a policy will create cohesion among these groups, through conflict with the modernizing and globalizing out-group and their foreign kin. Localization includes mobilizing communal groups by contributing to their sense of common identity, cohesion, and organization. Through diversionary conflicts, traditional groups will formulate an image of "others which is characterized by hostility, malevolence, suspicion, and mistrust" (Rothchild and Groth 1995: 71).9 Ethnic groups will castigate those members of ones own national core group who cooperate with the enemy.
Escalation
Traditional ethnic groups that are losers from the globalization process will most likely press for closer ties to kin in and outside their borders in order to undermine the modernizing groups (especially contiguous). Such traditional groups will retrieve group members across a territorial border in order to shore, up their position for localization.
Summary of the Book
This book examines ethnic conflicts in several regional contexts.10 We review four explanatory frameworks as they relate to escalation and diffusion of interstate ethnic conflict: (1) the weakening of state institutional structures, particularly where there has been a preexisting ethnic contract; (2) the increased role of non-state regional and international actors (NGOs and IGOs) and the reduction of state sovereignty and the empowerment of domestic actors and interest groups; (3) changes in the ethnic balance of power have real or potential repercussions in the competition for the distribution of social, economic, and political resources; and (4) the degree of economic, social, and cultural integration within the regional or global system (especially reinforcing versus crosscutting interstate cleavages between/among ethnic groups).
This volume examines several cross-regional cases. The richness of this book is that while individual authors adopt different approaches and methods, not all of which address each explanatory framework, the total project provides a comprehensive discussion of the internationalization of ethnic conflict. In this way, the volume provides a series of case studies that highlight both the differences and similarities in the internationalization process. We make no claims about offering a new theoretical explanation on the origins and development of ethnic conflict. Nor are we testing specific hypotheses. Our purpose is to examine the internationalization of ethnic conflict in light of a series of explanatory frameworks that helps us integrate and entangle domestic and international factors. We believe that the case studies explored here highlight the links between
the domestic and international arenas in trying to understand conflict and intend for this enterprise to be a first step toward theory building in the area of ethnic conflict.
The remainder of the book includes chapters that focus on specific cases. Chapter 2, by David Carment and Patrick James, compares the cases of Post-Communist Yugoslavia and Post-Independence Somalia, focusing on intervention by third-party states. Chapter 3, by Neal Jesse and Marc Simon, use a game-theoretic approach to compare ethnic violence in Northern Ireland and the Republic of South Africa. Based on the framework developed in this chapter, they address why diffusion and escalation occurred in the latter case but not in the former. In chapter 4, Shale Horowitz examines ethnic conflicts in the Soviet successor states of Azerbaijan, Georgia, Moldova, and Tajikistan. While all escalated to involve other states, none diffused into neighboring states. In chapter 5, Kristen Williams, examines the internationalization of ethnic conflict in the Balkans. Chapter 6, by Philip Mauceri, examines Latin America. He contends that a mix of demonstration effects, the increased role and interconnectedness of the international media, a shift in the international agenda of major powers toward human rights and the protection of minorities, and the growing presence and sophistication of NGOs explain the diffusion of ethnic-based conflict throughout Latin America. In chapter 7, John Quinn discusses what many have called Africa's First World War, and the links between ethnic conflict, regime change, and internationalization. In chapter 8, Jeffrey Helsing examines the internationalization of ethnic conflict in Lebanon, Iraq, and the Israeli-Palestinian-Arab conflict. Finally, in chapter 9, Bob Oberst examines the expansion of the Sri Lankan conflict from a low-level and sporadic warfare to a war with strong international connections and ties.
Notes
1. On constructivists (socially constructed identities), see Young (1993); and Brubaker (1995).
2. This term is coined by Abraham Lowenthal.
3. Lake and Rothchild (1998:13) credit Leonard Binder for coining this term.
4. Ethnic contracts can be formal or informal. One such solution is consociational democracy (Lijphart 1977). Consociationalism involves executive power sharing and grand coalitions, formal and informal separation of powers, balanced bicameralism and special minority representation in the upper chamber, a multiparty system, proportional representation, territorial as well as nonterritorial federalism and decentralization, and a written constitution protecting minority rights through difficult procedures for amendment.
5. On the offense-defense balance, see Posen 1984; Lynn-Jones 1995; Van Evera 1998.
6. External intervention of NGOs can have a positive dimension as well. The establishment of such transnational networks allow "ethnic groups with greater opportunities for transnational interactions: the exchange of ideas, information, wealth, and political strategies" (Stack 1997: 21).
7. Escalation can occur even if: (1) the chances of success for the preventive attack are long provided that the chances of success later would still be longer; and (2) the stronger group promises not to exploit the weaker group in the future.
Steven Lobell and Philip Mauceri
Horowitz refers to those that have privileged access and interest in obtaining greater resources as "advanced" and to those who do not as "backward." Given the changes in comparative terminology since Horowitz developed his framework, we prefer to use the terms "modernizers" and "traditionalists" to capture die differences outlined by Horowitz.
See Snyder and Ballentine (1996) on myths.
On die concept of a region, see Lake and Morgan (1997); Stein and Lobell (1997).
CHAPTER 2
Third-Party States in Ethnic Conflict: Identifying the Domestic Determinants of Intervention
David Carment and Patrick James
The Problem of Ethnic Intervention
Why .are some states predisposed to intervene in other states' ethnic conflicts?1 Internal wars continue to outpace interstate wars both in intensity and duration, so questions about ethnic intervention are becoming central to the study of violent conflict (Saideman 1997, 1998, 2001a,b; Wallensteen and Sollenberg 2000; Carment and Rowlands 2001; Kaufman 2001; Sislin and Pearson 2001). Far from resolving internal disputes, third-party involvement frequently translates into interference-prolonging and even intensifying such conflicts. Indeed, many of the most intractable civil conflicts in recent decades are marked by high-profile and disastrous interventions by outside states. Examples include India's 1987 efforts to mediate Sri Lanka's protracted conflict, Russia's intervention in the civil wars in Georgia,Tajikistan, Moldova and Azerbaijan,Turkey and Greece's involvement in the Cypriot conflicts, and confrontations between India and Pakistan over successive Kashmiri insurgencies (Carment 1994). Major issues in both theory and policy related to ethnic intervention and its aftermath continue to challenge the field (Hayden 1996a,b;Wallace 1996;Woodward 1996).
Instances of ethnic intervention continue to accumulate. Unfortunately, students of conflict management are only beginning to construct theories about third-party interventionism with a general range of application, most notably with respect to ethnic conflict (Gurr 1992; Saideman 1998, 2001b; Tenorio 2001).2This oversight is particularly glaring given the fact that such involvement continues to spoil or delay peace settlements in so many civil conflicts (Weiner 1971; Gurr 1992; Carment 1992, 1993; Brown 1996b; Lobell and Mauceri, chapter 1, this volume). Moreover, with the acceleration of globalization and the growth of supranational organizations, so-called internal conflicts have become more "internationalized" than ever before (Midlarsky 1992; Brown 1993, 1996a,b,c; Carment and James 1997, 1998).3 The ongoing gap in knowledge about intervention in ethnic strife undoubtedly reflects the position of this problem at the intersection of comparative and international politics: a
12
David Carment and Patrick James
Third-Party States in Ethnic Conflict
13
comprehensive explanation will require a network of factors that reaches across that traditional divide.
Ethnic conflicts are "quintessential political conflicts involving incompatible \ preferences over distributional outcomes" (Mozaffar 1999: 47).Thus ethnic inter- < vention is defined here as economic or political interference by one sovereign j state in the internal affairs of another for the stated purpose of assisting an ethnic | group residing in that state. In other words, intervention as studied here is redis-tributive in its goals and favors one ethnic antagonist over others. This chapter | seeks to identify domestic determinants of ethnic interventionism by third-party 1 states, which reflects a growing awareness that interstate and even system-level factors can go only so far in accounting for the development of individual cases. I
Recent literature on ethnic conflict, perhaps in reaction to the upheavals at i the outset of the 1990s, outlines a host of causes for third-party interventions. j Most of the factors noted point to characteristics within the conflict's host-state, j an especially interesting tendency because the sheer number of conflicts in j progress at about the same time would seem to encourage interest in more encompassing explanations. Hypotheses about host-states include the depth of ; ethnic cleavage or animosities, institutional breakdown and resultant security .: j
dilemmas that provoke third-party involvement, opportunistic "ethnic elites" ) who successfully lobby for outside assistance, and so on (Lake and Rothchild 1998; see also Kaufman 2001).4 While such propositions have the potential to shed important light on the ways in which ethnic conflict becomes internationalized, the respective arguments cannot account for why some states are more prone than others to intervene in ethnic wars outside their borders. Prominent examples would include Pakistan's interference in Kashmir and Rwanda's interference in the Congo. To answer questions about why the propensity toward intervention varies among members of the international polity, it will be useful and even essential to examine characteristics within intervening states.
Two broad schools of thought exist on ethnopolitics in general and the domestic determinants of ethnic interventionism in particular: affective versus instrumental (Gurr 1992:16—17;Wieland 2001: 209).Each is introduced in turn, with more detailed summaries appearing at a later point.
According to the affective school of thought, ethnic allegiances between one state and an embattled group in another will encourage intervention. While such tendencies exist, the empirical record demonstrates that cross-border ethnic ties do not always "bind"—these ties inspire third-party intervention in some cases but not others. Although states like Serb-dominated Yugoslavia have intervened repeatedly on behalf of Serb minorities elsewhere, others, such as China, generally have refrained from intervening on behalf of their ethnic counterparts abroad (Yu 1999). Furthermore, a state may show significant variation both over time and in relation to different diaspora; Somalia serves as a prominent example (Saideman 1998, 2001b).This difference becomes even more interesting when cast in terms of opportunity versus willingness (Most and Starr 1989; Ciom-Revilla and Starr 1995). As a Great Power, China has ongoing opportunities to intervene with military force but generally does not, while "Serboslavia," a weak state, interfered destructively with its neighbors on multiple occasions over the last decade.
Another school of thought, termed instrumentalist, asserts that leaders intervene in ethnic wars to obtain access to valued resources or to create diversion or "rally around the flag" effects when they face defeat at home. However, this argument cannot explain why leaders of many other states refrain from opportunistic interventions or utilize alternative methods of regaining popular support when faced with internal challenges.5
Both schools of thought seem to beg the question: why do affective ties or elite incentives lead to a foreign policy of intervention in some states but not in others? This puzzle may be solved, at least in part, by adding an institutional component to the study of ethnic interventionism. In other words, it is useful to consider the possibility that affective ethnic ties and elite incentives are filtered through domestic institutions to yield differential foreign policies with respect to ethnic conflicts in other states.This study develops a model of domestic determinants of ethnic interventionism and presents preliminary evidence in the form of focused case studies to assess its predictions.
Determinants of ethnic interventionism include: (1) ethnic domination versus diversity—the extent to which institutions of the state are in the hands of a single ethnic group; and (2) high versus low institutional constraints—the extent to which leaders enjoy discretionary power over state policies (Carment and James 1995, 1996, 2000).6 The model yields a prediction for each type of state with respect to interventionism: (1) ethnically dominant, low-constraint states are predisposed to bellicose policies of ethnic intervention; (2) ethnically dominant, high-constraint states are prone to policies of sporadic ethnic intervention; (3) ethnically diverse, low-constraint states are likely to pursue pacific intervention through diplomacy; and (4) ethnically diverse, high-constraint states are likely to engage in realpolitik policies toward interventionism.
This chapter proceeds in six further parts. The first part provides an overview of the existing literature on ethnic interventionism. The second part presents a model used to identify states that are more or less likely to intervene in conflicts outside their borders. In the third part, predictions for foreign policy as related to ethnic intervention are generated for states on the basis of the preceding model. In the fourth part, focused case studies are used to illustrate the model's predictions. The fifth part relates the findings just noted to the hypotheses about escalation that are the focus of the volume as a whole. In the sixth part the results of the analysis are summarized and directions are offered for further research.
Perspectives on Ethnic Interventionism
Recent waves of theorizing on ethnic interventionism have focused primarily on the state in which the conflict takes place. The working assumption is that ethnic conflict is generated internally and then externalized or "internationalized" as the conflict draws in external actors (Suhrke and Noble 1977; Brown 1996a,b,c; Lobell and Mauceri, this volume). Several hypotheses are put forward about why this occurs. One idea is that ethnic conflict may weaken state structures, which in turn encourages intervention by outside actors who hope to obtain resources or influence the outcome of the strife to their advantage
14
David Carment and Patrick James
(Cooper and Berdal 1993; Hislope 2001; Lobell and Mauceri, this volume). Another hypothesis is that ethnic conflicts create harmful side effects (or, in the language of economics, "negative externalities"; Sandler 1997) in the form of regional security dilemmas that neighboring states attempt to resolve through intervention (such as Rwanda's intervention in Zaire; Posen 1993). A more liberal view, by contrast, would infer that third parties intervene in ethnic conflicts in reaction to gross violations of human rights and produce "good offices and mediation, peace making and peace keeping, protection of human rights, humanitarian assistance, and stigmatization of regimes deemed guilty of flagrant violation of international norms" (Esman 1995: 21)? NATO's intervention to prevent the Yugoslav government from ethnically cleansing Kosovar Albanians in 1999 may serve as evidence for the liberal point of view (Carment and Harvey 2000; see also White 1995 on British humanitarian relief, as a specific example, in the former Yugoslavia).8
One very useful summary of the external dimensions of internal (and frequently ethnic) conflicts is provided by Brown (1996c: 592, 596-99; see also Kaufman 2001: 40-41). In a more general treatment of the actions of neighboring states in internal conflicts, Brown identifies five kinds of intervention that are linked to respective motivations. At one extreme is humanitarian intervention, where refugee problems, for example, provide the rationale for getting involved. At the other extreme are opportunistic interventions and invasions, which come about as a by-product of the intervening state's experience with instability and warfare spilling over from its neighbor (Brown 1996c: 592), Examples of opportunistic interventions and even invasions, which generally appear in line with a more realist point of view, "are plentiful": South African support for insurgents in Mozambique and Angola in the 1980s; Angola versus Zaire, and Sudan and Uganda, respectively, against each other; Pakistan's assistance to insurgents in India's Punjab and India's support for insurgents in Pakistan's Sindh; and the Syrian invasion of Jordan during the latter's civil war in 1970 (Brown 1996c: 598-99). In all of these instances the intervening or invading state pursued some combination of political, economic and military interests.
While the preceding discussion sheds important light on the proximate causes, timing, and pathways of ethnic interventions, they cannot explain why some states are more likely than others to intervene in conflicts outside their borders. This is because the target state is at most a part of the story; the decision to intervene, in the end, must come from within a potential third-party state. To answer the question of who is disposed toward third-party action, we must explore characteristics of the intervening states themselves. As noted earlier, the two basic approaches toward the study of interventionist foreign policy are affective and instrumental.
Instrumentalist explanations hold that ethnic conflict spreads with purpose rather than inevitability (Suhrke and Noble 1977; Heraclides 1991; Cooper and Berdal 1993; Snyder 1993; Gagnon 1994/1995;Tishkov 1997;Weingast with de Figueiredo 1999; Kaufman 2001; Ross 2001). Instrumentalists view external intervention as neither a direct outcome nor a by-product of centuries-old hatred. Instead, the emphasis is on benefits that.accrue to leaders who engage in ethnic interventionism. Political leaders are expected to respond like rational
Third-Party States in Ethnic Conflict
15
actors to external opportunities to maximize their security, wealth, and power (Meadwell 1991, 1993). Priorities among these objectives, of course, will vary from one leader to the next and perhaps even for the same decision-maker over time. In this formulation, political elites faced with a leadership challenge may engage in adventurous foreign policies to create a diversionary effect as a means of shoring up domestic support (James and Oneal 1991; Morgan and Bickers 1992; Miller 1995; James and Rioux 1998;Weingast with de Figueiredo 1999).
Further understanding may He in systematic assessment of the politics of cross-border ethnic affinities. Although instrumentalism may account for why elites themselves engage in interventionist policies, affective ties—based on emotion and ascriptive criteria—are crucial for explaining why "ethnic rescue" can be such an effective political strategy for elites (Smith 1986: 75; see also Chazan 1991; Connor 1994; Brown 1996c). Ethnic ties exist as a result of significant cross-border perceptions of a shared past and common destiny. Today it is almost a truism that ethnicity is one of the primary social identities around the world (Stack 1997; Dombrowski 1998; Ross 2001).True or not, this universal perception almost certainly is self-reinforcing. Indeed, ethnic identities increasingly trump state identities where the two are incongruent; as primordialists point out, there is a limit to the "plasticity" of ethnic identity (Kaufman 2001: 24; see also Stack 1997, 2000). Ethnic or confessional identities become more salient and even inclusive as national systems of order collapse (Zartman 1997: 4), Even more compelling is the imposition of identity by extremist members of an opposing group; under conditions of intense conflict, opting out of a particular ethnic designation may simply prove to be impossible (Kaufinann 1996:144).The political importance of such identities—particularly when they span state borders— should not be underestimated. Cross-border ethnic ties involve a range of emotional linkages, including shared grievances, language, customs, religion, principles, racial-cultural affinities and common humanitarian concerns. Due to the power of these linkages, conflicts involving a group's co-ethnics may be imagined to involve the group itself. Intervening on behalf of the group's co-ethnics in crisis may therefore be an effective means of obtaining the political support of this group.
While certainly important, even these insights concerning the rise of transnational group identification still do not explain why some states are more likely than others to intervene on behalf of their co-ethnics abroad. Some interesting ideas and empirical evaluations from Saideman (1997,1998) on secessionism and irredentism, respectively, can be used to provide structure for further discussion of this issue. As will become apparent, ethnic ties are significant in accounting for third-party conduct, but other factors enter into the equation as well.
Saideman (1997) demonstrates that ethnic allegiances are an important factor in motivating interventions in secessionist conflicts abroad. In particular, vulnerability to secession does not seem to be a determining factor with respect to support for the government of a state versus a secessionist movement in the following cases: the Congo crisis (1960), the Nigerian civil war (1967-1970) and the disintegration of Yugoslavia (1991-1995). A review of potential interveners (N = 46) across these observations produces the following key findings, among others: states vulnerable to secession supported secessionists versus host-states in
16
David Garment and Patrick James
20 and 11 instances, respectively, while the numbers for states not vulnerable to secession are, in turn, 3 an 7 (Saideman 1997: 747).9 In contrast with this very weak connection between vulnerability and self-restraint with respect to support for secessionists (which is at variance from conventional treatments such as Jackson and Rösberg 1982), consider the following result: among states with ethnic ties to secessionists, 21 supported the secessionists and one supported the host-state, while among states with ethnic ties to host-states, 15 supported the host-state and only one supported the secessionists (Saideman 1997: 748).10The underlying message of these findings is clear: vulnerability to secession does not seem to deter a potential third party from intervention, while ethnic ties seem to matter a great deal in figuring out which side will receive support if it is provided.
Irredentism, the natural complement to secessionism in conceptual terms, is found to result from "competition for support of groups who have ethnic ties in neighboring territories" (Saideman 1998: 52; see also Saideman and Ayres 2000). This basic insight accounts for why irredentism is not a constant factor in cases where ethnic ties always are present. The key to the timing of irredentism is the need for support from essential constituents. The cases of Somalia (1960-1980) and Serbia (1991-1997) are used by Saideman (1998: 56) to demonstrate this point. In Somalia, a noteworthy effort toward a Pan Somalia comes in 1963-1964, right before the March elections. Competition focused on the clans with connections to ethnic Somalis in Kenya and Ethiopia. By contrast, in the late 1960s, electoral pressure subsided. While it governed a still-democratic regime, the ruling Somali Youth League (SYL) faced decreasing competition in the elections of 1967 and 1969 because the opposition became weaker and ultimately disintegrated. After the election of 1969, for example, the SYL and its coalition partner held the vast majority of seats, which greatly reduced pressure for irredentist policies (Saideman 1998: 72).The same variation over time is seen for Serbia: the Serb government at one time offered considerable support for Serbs in Croatia, for example, when it faced serious political competition from the extremist Serbian Radical Party in 1991-1992, but actually imposed his own sanctions on the Bosnian Serbs in summer 1994. The major opposition parties approved of a peace plan and Milosevic, by that time under heavy pressure from economic sanctions imposed by the major Western states, faced no significant lobby in favor of irredentism (Saideman 1998: 83-84). In sum, the Somali and Serbian cases are instructive in showing that variation in political competition can explain the sometimes inconsistent nature of irredentism.
Why, then, do ethnic ties lead to interventionist foreign policy in some cases but not others? When do cross-border ethnic ties influence state foreign policy? In general terms, the answers offered by this chapter are as follows: when ethnic ties exist, the institutional constraints and ethnic composition of a state play the primary roles in determining whether that state will attempt to intervene on behalf of its co-ethnics abroad. Our model draws on the insights from Saideman—in particular, that ethnic interventions are driven significantly by domestic factors—and hypothesizes the ways in which two such factors (ethnic composition and level of institutional constraint) interact to produce a certain type of ethnic foreign policy: Importandy, whereas Saideman's analysis aims to identify the conditions under which states follow active irredentism, this model uses the same case studies to test
Third-Party States in Ethnic Conflict
17
predictions concerning the type of foreign policy each state is likely to pursue. While affect and instrumentalism are critical components in this process, these factors are filtered through domestic institutions that largely determine whether and in what way ethnic ties influence foreign policy formation.
The Model
This section presents a model that can be used to predict a state's disposition toward its co-ethnics abroad.The section begins by exploring how ethnic composition and domestic institutional constraints are expected to influence a state's propensity to intervene on behalf of its ethnic brethren. Interactions between these two variables are analyzed in the context of a two-by-two matrix. The model depicts four ideal types of states, each of which is associated with a distinct ethnic foreign policy.These predictions subsequently are tested.
Ethnic Composition: Domination versus Diversity
Under conditions of cross-border ethnic ties, leaders interested in gaining and retaining political power, and a conflict between a state's co-ethnics and their host-state, what policy is the state expected to adopt toward its co-ethnics?11 To answer this question, it is essential to consider the ethnic composition of the intervening state—specifically, that of the ruling elite. A states ethnic composition may take many different forms. In the interest of parsimony, two ideal types are considered here: (a) ethnic domination and (b) ethnic diversity.
Ethnic domination occurs where a single ethnic group enjoys political control over the state such that elites can improve their position—and that of their ethnic group—without the support of other ethnic groups. Ethnic foreign policy can assume a special significance in the domestic arena. This is because both agenda-setting powers and jurisdiction over foreign policy are in the hands of elites who have incentives to emphasize and reinforce ties with their ethnic group—generally the dominant ethnic group in the state and thus the main domestic constituency. In most ethnically dominated states, political competitors may only defeat the ruling elite by positing stronger linkages between themselves and the principal ethnic constituency. This can initiate rounds of "ethnic outbidding," as the ruling elite responds to these challenges by introducing issues and policies that suggest an even greater loyalty to their ethnic constituency (Horowitz 1985). This process usually results in redistributive policies in favor of the dominant ethnic group, even on issues in which the group has only a marginal interest. Somalia, for example, is the most ethnically homogeneous state in sub-Saharan Africa, yet clan-based competition for power is the norm, with highly destructive effects over the long term (Brown 1996b: 15),
Leaders of ethnically diverse states, by contrast, generally cannot rule without the support of other ethnic groups. Consequently, the leaders of such states are likely to prefer policies that appeal to more than one ethnic group as a means of attracting interethnic political support (Saideman 2001b: 25). In this way, elites
18 David Carment and Patrick James
can be expected to seek support on the basis of both ethnic and crosscutting identities. All other things being equal, ethnic outbidding is less likely to occur in such states because elites are aware that competing excessively for the favor of one ethnic group may alienate other ethnic groups, who could form a coalition to oust the leadership. Policies and political stances therefore are likely to involve significant interethnic compromise; the ongoing brokerage politics of the Democratic and Republican Parties in the United States would be a prime example. Ethnic intervention is not ruled out for diverse states but, regardless of regime type, this behavior is expected to be a product of special circumstances rather than an ongoing feature of foreign policy.
Institutional Constraints: High versus Low
Domestic institutions are a second major determinant of elite policy-making. Institutional constraints are defined as the broad, underlying patterns of political authority and constitutional structure of the state (Gurr 1974,1990; Morgan and Campbell 1991). For the purposes of this analysis, institutional constraints will denote the extent to which these structures limit the actions of elites in accordance with the preferences of citizens, broadly defined. These constraints then may be equated with democratic institutions, which include (a) regular, free, and fair elections; (b) universal voting franchise; (c) referenda on matters of high constitutional importance; (d) constitutionally guaranteed and enforced civil liberties; (e) an independent judiciary; and (f) various checks and balances on executive decision-making, such as federal structures, shared authorities, congressional or parliamentary override and procedures for executive impeachment.
Analysis of foreign policy in the context of domestic institutions might seem wrong-headed from the outset. After all, unlike domestic policy, foreign policy is presumably more resistant to the vagaries of public opinion, especially during periods of interstate conflict and war. However, if elites must compete for power in the electoral arena, the masses can be expected to influence elite decisionmaking, regardless of issue-area (Tsebelis 1990: 167). (One example is the previously noted diversionary model of conflict.) Given that possibility, two types of political institutionalization are considered in this analysis: (i) low institutional constraints—commonly associated with authoritarianism or totalitarianism; and (ii) high institutional constraints—usually associated with democracy.
Leaders in many states have no popular mandate, but instead rule on the basis of force or coercion, elite pacts, heredity, or even divine authority. These elites enjoy low institutional constraints by virtue of the fact that their power does not depend upon the support of the population at large. Elite decision-making in such states is relatively unconstrained by popular opinion or constituent interests. States with low institutional constraints include military dictatorships such as Chile under Pinochet, monarchies and revolutionary republics of the contemporary Middle East, and one-party totalitarian regimes such as the former Soviet Union and China today. Where such regimes exist, policies result from elite bargaining over the distribution of resources—lower levels of government operate primarily through coercion or patron-client relations. Leaders of such states may pursue
\ Third-Party States in Ethnic Conflict 19
I bellicose or rent-seeking policies since they have greater incentives to serve their
\ own interests and those of their patrons than to serve the interests of the state as a
E whole—even when doing so entails weakening the state through war or inter-
' national sanctions (Lake 1992). While institutional constraints may be low, either
\ intra-ethnic or inter-ethnic factional pressures could produce intervention.
t Where leaders, by contrast, face high institutional constraints, they are account-
| able directiy to the people (or at least an identifiable and presumed majority
j coalition), who can either punish or reward elite behavior through regular
[ statewide elections. In a federal system, for instance, the relationship between
[ elites and masses consists of a two-tier game involving legislative and electoral
f politics.These institutions largely determine the extent to which the states eth-
t nic composition—and the interests of its various ethnic constituencies—plays a
I role in elite decision-making. Whatever the states ethnic makeup, its elites will have powerful incentives to curry favor with the masses. The leaders of such states
} therefore are less likely to pursue unnecessarily reckless policies—either foreign
| or domestic—that put the interests of the state and its citizens as a whole at risk,
i The same cannot be said of leaders of low-constraint states, where the interests
I of the masses are far less influential in elite pohcy-making. *-•■. ■•
Interaction Effects: Predictions for Foreign Policy
A Typology of States
Which states are most likely to adopt belligerent or adventurous foreign policies? A state's foreign policy is generally a function of multiple factors, which can be either internal or external to the state in question {Morgenthau 1957; Huntington 1968; Skocpol 1979; Gurr 1980; Snyder 1991b). For example, a state may adopt a bellicose foreign policy when social or economic pressures threaten to overwhelm its political institutions. Interventionist policies also may be motivated by a leader's "gamble for resurrection" in the face of defeat or usurpation (Weingast with de Figueiredo 1999). Conversely, cross-border ethnic affinities may draw one state into the internal conflict of another in order to "rescue" its ethnic compatriots in crisis (Van Evera 1994). Finally, ethnic conflicts produce "cleavages" that external actors can exploit to increase their security and gain access to valued resources (Carment and James 1995,1996).
Although propositions such as these can address when interventions take place and why they unfold as they do, an exclusive focus on the triggers of interventionist policies excludes consideration of deeper, perhaps even structural, enabling conditions. Moreover, these explanations do not address the question as to why some states are more prone to interventionism than others. The model presented here is founded on the idea that such questions can be answered only through structural analysis of the intervening state itself (Brass 1991;Kohli 1990).
As noted earlier, the structural factors most relevant to ethnic foreign policy include: (1) ethnic composition of the state; and (2) the extent to which leaders are constrained by their ethnic constituencies. These variables interact to produce different policy "win-sets" at the state level (Putnam 1988). A win-set is
20
David Carment and Patrick James
Ethnic composition Dominant (a) Diverse (b)
Low (1)	Dominant-low (la): belligerence	Diverse-low (ib): passive lobbying
institutional		
constraints		
High (II)	Dominant-high (lia): sporadic interventionism	Diverse-high (llb): realpoiitik policies
Figure 2.1   Domestic determinants of ethnic intervention.
defined here as all of the successful foreign policy strategies that would "win," or satisfy the minimum number of people or institutions the leader needs to gain or retain power. Interactions between the variables are depicted in figure 2.1, which conveys four ideal state types, each of which is associated with a distinct foreign policy orientation (see figure IX)}2 Briefly, incentives to adopt ethnic foreign policies are minimal for high-constraint, ethnically diverse states (which have the smallest win-sets), and maximal for low-constraint, ethnically dominant states (which have the largest win-sets). The expected foreign policy profile for each state type is elaborated here.
Foreign Policy Profiles
Leaders may be confronted with the need to prove their status as the "best nationalist" from either (i) in a dictatorship, a mass- or elite-led social movement with revolutionary potential; or (ii) in a democracy, electoral considerations. In both instances, of course, ethnic interventionism could occur either in response to observed pressure or a belief that such forces must be headed off. The profiles that follow assess the interaction effects that make each of the preceding processes more or less likely to occur.
Dominant-Low (IJ-.Belligerence With low institutional constraints, elites have considerable discretion in the formation of foreign policy. Many of these elites will have come to power through force (e.g., Somalia at the end of the 1960s) and frequently owe their position to the support of the relatively small proportion of the general public who actively prop
Third-Party States in Ethnic Conflict
21
up the regime (e.g., the military, bureaucracy, an economic class, ethnic group, or foreign actors such as governments and multinational corporations}. With a limited constituency to satisfy, all other things being equal, these leaders have larger policy win-sets and are thus relatively unconstrained in their decisionmaking. Checks on executive authority—when present—tend to be "on paper only." The de facto versus de jure operation of the Soviet regime could be cited as the archetypal example. Power tends to be concentrated in the hands of a numerically small elite with relative immunity to domestic pressures. Foreign and domestic policies can be expected to appeal to the dominant ethnic group (and especially those with military standing), but not in a way that would threaten the power base of elite members. The interests of the masses thus are subordinated to those of elites.
Elite foreign policy is influenced primarily by the identity of the constituency. When leaders depend upon economic elites, multinational corporations or international organizations, their foreign policies are more likely to resemble that of the security- and wealth-maximizing states envisioned by structural realists (James 2002).These rulers have longer time horizons and are therefore less likely to engage in rent-seeking or belligerent foreign policies that yield only short-term gains. Indeed, they may include a number of standard realist factors into their decision-making calculus, such as absolute and relative capabilities, alliance structures, and so forth.
When elites depend upon the support of an ethnic group, however, foreign policies are much more likely to take on an ethnic component. Leaders retain or gain power by demonstrably giving precedence to the interests of this group. Political challengers then may attempt to usurp power by demonstrating an even greater willingness to champion the interests of the dominant group. This process of ethnic outbidding may lead to an ethnically oriented foreign policy as elites attempt to protect themselves from charges that they have "sold out" their ethnic constituents. Intervention and even invasion may take place as a means toward undermining efforts by rivals to seize power through either elite-based conspiracy or mass mobilization.Thus a ruling elite could choose to intervene on behalf of its co-ethnics across a border even when broader economic or security interests are ill-served by doing so. In such cases domestic considerations therefore trump the interests of the state as a whole. Consistent with this prediction, the empirical record shows that the propensity to intervene in ethnic conflicts increases with the ethnic homogeneity of the intervening state (Horowitz 1991; Carment and James 1995).13
Diverse-Lotv (Ib): Passive Lobbying Leaders of ethnically diverse, low-constraint states, by contrast, are unlikely to pander to the interests of any one ethnic group.This is because the expected payoffs from ethnic politics are quite small at best and perhaps even entail probable losses from the outset. For example, if the military and bureaucracy are controlled by different ethnic groups, or if the ruling elite is multiethnic, then policies that serve any single group could threaten the regime status quo—a kind of internal "balance of power." Moreover, although groups may negotiate informally over state resources, there is an understanding that open politicization of ethnic
22
David Garment and Patrick James
Third-Party States in Ethnic Conflict
23
issues—like formal access to group entitlements or support for ethnic insurgencies in neighboring states—could well provoke ethnic violence within the state itself, thus weakening state institutions and possibly upsetting the regime.
Ethnic interventionism is expected to take the form of passive lobbying. Leaders are likely to forego risky foreign policies in favor of nationalist rhetoric or token gestures aimed at placating the ethnically dominant group without doing damage to the state's international reputation or its broader economic and security-related interests. For example, elites may be rewarded for strong public statements that affirm the precedence of the dominant ethnic group in the state, while acting in a way that seeks to maximize overall wealth and security for the state. In such states, the leadership's "bark" is generally worse than its "bite."
Dominant-High (IIJ: Sporadic Intervention Leaders in ethnically dominant, high-constraint states are expected to ignore the interests of the primary ethnic group at their own peril. Since replacement is a real possibility, elites are constrained highly by mass sentiment.When the majority of the electorate belongs to a single ethnic group, leaders are vulnerable to challengers' claims that they have betrayed the interests of this crucial constituency. Leaders therefore must protect themselves from political attacks that play on issues of nationalism, historical injustice, ethnic grievances and internal or external threats to the "purity" of the nation. Whether real or mythical in nature, the past is more than just a prologue to the politics of today.
Ethnic outbidding occurs in ethnically dominant, high-constraint states as political elites compete for the support of the key constituency. The process goes in the same direction as in ethnically dominant, low-constraint states, but the masses "lead" and elites "follow." The electoral support of the primary ethnic bloc is crucial to gaining and retaining power. Awareness of that point means that elites respond to surges of nationalism at the grassroots level out of fear that they otherwise will be replaced. Thus a kind of intermittent hypernationahsm can emerge as elites attempt to outbid one another for the support of this group.
Ethnically dominant, high-constraint states are likely to engage in sporadic rather than persistent ethnic interventionism. The leaders of such states depend upon support from a large proportion of the public through regular, free and fair statewide elections. People vote not only with their hearts but also with their pocketbooks, so foreign policies that weaken the state by producing international sanctions, war or other costly outcomes may result in elite turnover (Bueno de Mesquita and Siverson 1995). Ethnic intervention is expected to be sporadic as the tendencies toward outbidding normally will be subordinate to the need for good politico-economic management. When specific circumstances combine to make other options unattractive, however, intervention comes into play. A sense of crisis and limited alternatives both at home and abroad will be needed to evoke a forceful ethnic foreign policy.
Diverse-High (IIh): Realpolitik Policies . Elites in ethnically diverse, high-constraint states have extremely narrow policy win-sets. The leaders of such states must satisfy a broad constituency that also is ethnically diverse and therefore entails many different—and often
'Li :
:■■■}!!.:■
m
m. ■
":
it;:;
countervailing—policy preferences. At the same time, high institutional constraints oblige these leaders to retain at least a plurality of popular support in order to gain and (especially) maintain political power. These two restrictions greatly limit the range of policies elites can adopt that will satisfy the requisite number of constituents for retaining office. Cognizant of this, leaders of such states are unlikely to pursue policies—formal or informal, foreign or domestic— that serve the interest of any single ethnic group at the perceived or real expense ■ of others. Leaders also must appeal across ethnic groups to obtain sufficient support for office, so they instead will adopt policies that enjoy broad, interethnic support.
For similar reasons, leaders of ethnically diverse, high-constraint states are unlikely to pursue ethnic interventions. Electoral considerations and the attendant need to maintain interethnic support at home will cause elites to downplay ethnicity as a source of foreign policy. Ethnic interventions are likely to occur only when there are no strong preferences among any of the state's ethnic groups or where there is a general consensus that the state has strong real-politik-related reasons for involvement abroad. In sum, leaders of ethnically diverse, high-constraint states are unlikely to intervene on behalf of co-ethnics in other states, particularly if doing so fails to serve the state's overall economic or security concerns. The rulers of such states have incentives to respond to the interests of a broader cross-ethnic constituency (which usually entails the citizenry as a whole) over those of a single ethnic group that may have a special interest in the fate of their co-ethnics abroad.
Figure 2.1 summarizes the foreign policy predictions for the four state types. Since leaders of Type Ia states must satisfy relatively few constituents, belligerent ethnic interventions may easily be preferred. Types Ib and II, fit between the extremes.Type Ib states are inhibited because of ethnic diversity, while Type IIa states are constrained by institutions, although rhetorical outbidding may increase IIa s propensity toward an ethnic foreign policy. Elite policy win-sets are narrowest for Type IIb state, the leaders of which are most likely to eschew ethnic interventions.
From this analysis, a general hypothesis emerges about relative likelihood of intervention:
General Intervention Hypothesis (GIH), The potential for intervention in ethnic conflicts varies with the structural characteristics within the intervening state itself: ethnically dominant, low-constraint states (Type Ia) have the highest disposition toward intervention and are inclined toward belligerence; ethnically dominant, high-constraint states (Type ILJ are likely to engage in sporadic interventionism; ethnically diverse, low-constraint states (Type Ib) are inclined toward passive lobbying for their co-ethnics in neighboring states; and ethnically diverse, high-constraint states (Type IIb) are least likely to adopt ethnic foreign policies and tend toward realpolitik policies.
Figure 2.1, which focuses on domestic determinants, is not intended to explain every aspect of ethnic interventionism. Clearly, there are factors beyond those enumerated in the GIH that influence the propensity to intervene and tactics employed, along with the timing, intensity, and course of such interventions (Kaufman 2001: 6).14 Despite such limitations, this model is expected to shed
24
David Carment and Patrick James
important light on the question of why some states are more prone to ethnic interventions than others.
Case Illustrations
Overview
Two case studies will be used to offer a preHminary sense of how the model works in application.The cases,Yugoslavia/Serbia and Somalia, provide a total of three "observations" in the sense described by King, Keohane, andVerba (1994: 117).While Serbia remains aType IL, state throughout the period of study (i.e., from 1990 to 1995), Somalia shifts from Type IIa to Type Ia (i.e., from 1960 to 1978, with the transition in 1969). Thus, a total of three observations exist and, as will become apparent, they are sufficient to produce significant variation in the dependent variable, that is, ethnic interventionism.
One limitation to the cases selected is with respect to ethnic composition, which remains a constant across all three observations. Thus the purpose of the present chapter is to focus on variation in institutional constraints and use the evidence that follows to facilitate further theory building (Fox 1999: 442); ethnic composition will play a role in future research that takes the form of aggregate data analysis (Carment et al. 2002). In sum, analysis of the three observations is intended less as a formal testing of the model than as an illustration of how it can be applied in a preliminary way to account for ethnic interventionism.
In post-communist Yugoslavia, the Serb-dominated Yugoslav government at first actively supported Serb rebellions in Bosnia and Croatia but later abandoned its co-ethnics under the pressure of economic sanctions. (This case is complicated somewhat by the fact that Yugoslavia's substate borders became interstate borders during the course of this conflict.) Yugoslavia may be considered an ethnically dominant, high-constraint state throughout the 1990s. Thus its leaders would be predicted to intervene sporadically on behalf of Serb minorities abroad. The period of the case study of Serbia is from the transition out of communism in 1990 to the Dayton Accords of 1995.
The second case is that of post-independence Somalia, which embarked on successive irredentist projects in Ethiopia. Somalia is categorized here as both an ethnically dominant, high-constraint state (1960-1969) and an ethnically dominant, low-constraint state (1969-1978).These designations suggest that Somali leaders would be likely to intervene sporadically in the first period and with belligerence in the second period on behalf of co-ethnics over the border. Although it may be objected that Somalia is ethnically diverse due to its clan divisions, it may be considered homogeneous in the sense that its inhabitants are all ethnically Somali, speak the same language, practice the same religion, and, most importantly, identify with ethnic Somahs in Ethiopia, Kenya, and Djibouti. The period of the case study of Somalia is from independence in 1960 to defeat in the war with Ethiopia in 1978.
Third-Party States in Ethnic Conflict Post-CommunistYugoslavia
25
Prerevolutionary Yugoslavia is classified as an ethnically diverse, low-constraint state, or Type Ib. Yugoslavia's six ethnic republics—Serbia, Croatia, Bosnia-Herzegovina, Slovenia, Macedonia, and Montenegro—shared power. Its leaders responded in 1974 to complaints of Serb chauvinism by transforming the republic from a centralized state to an explicitly federal structure that included the six republics and two autonomous provinces inside Serbia—Kosovo andVojvodina. It may have been due partly to this delicate interethnic balance that the Yugoslav government refrained from intervening in other states' internal affairs, although the imposing power of the USSR within the near abroad also must be recognized for its obvious importance.
These confederal structures gradually devolved to Yugoslavia's constituent republics in the decade following Josip Titos death in 1980.This phase proved crucial to development of particularist identities and competing visions of Yugoslavia's future. The growing divisiveness of the Yugoslav state also may be traced to the rise of Serb nationalism and gradual Serb domination of the Yugoslav National Army (JNA) in the 1980s.Yugoslavia's economic failures later discredited liberal reformers who believed that its constitution and economy could be restructured within existing political arrangements. Against a backdrop of economic collapse, nationalist leaders gained tremendous popular appeal and easily overwhelmed their antinationalist counterparts. Croatian and Slovenian national leaders touted the virtues of national independence as a means of integrating into the Western economic and political structures; meanwhile, Serb nationalist elites railed against Serb underrepresentation in Yugoslavia's political institutions. For example, in 1986, a "Memorandum" from the Serbian Academy of Arts and Sciences called for a more centralized system in order to address economic problems and hurled accusations about an anti-Serbian conspiracy (Kaufman 2001:179).
Politics in post-communist Yugoslavia thus took on an ethnic character that had been established in the period leading up to the collapse of the Soviet bloc. At the same time, ethnocentric leaders, each of whom had powerful political incentives to cater to the "national interests" of their respective constituencies, gradually replaced Yugoslavia's pro-federal elites. Croatian and Slovenian leaders therefore favored national independence as the surest route to prosperity for Yugoslavia's ; richest republics. At the same time, Serb national leaders strove to contain these : secessionist impulses in the interests of preserving a federation in which the Serbs had a controlling stake, with the potential for income redistribution and similar kinds of ethnically oriented policies. Serb leaders portrayed Slovenian and Croatian secessionism as a conspiracy against the Serbs and resistance to these actions as essential to Yugoslavia's (read "Serbia's") national interests. In particular, Slobodan Milosevic used state-controlled media in the late 1980s to fuel Serb resentment against other ethnic groups; Muslim Albanians, for example, became identified with a renewed Islamic menace to Serbs (Kaufman 2001:180).
January 1990, witnessed the effect end of the Yugoslav Communist Party, as Milosevic's efforts to grab power ultimately caused the Party Congress and later