Status in the World System and Ethnic Mobilization.

Daily reports of Arab-Israeli violence in the West Bank, IRA bombings in Northern Ireland, attacks on gypsies in Hungary, rebellions in Sri Lanka, and the ethnic civil war in Rwanda document the fact that ethnic violence can erupt spontaneously nearly anywhere in the world.(1) Comparative research also suggests that ethnically based movements have constituted the modal form of group conflict since World War II (Gurr 1993). For policy makers, rising waves of ethnic violence are seen as an uncontrollable threat to world peace (Esman 1995). Taken together, the evidence suggests that ethnic movements appear extremely resilient in the modern world, despite predictions to the contrary (Jalali and Lipset 1992-1993).

Although international news headlines reflect heightened concern about ethnic movements, few systematic analyses exist across multiple countries over a long period of time.(2) Given the consensus that ethnic mobilization has diffused broadly across the world's states, it is even more curious that social scientists have not systematically considered the role of world-level processes in generating ethnic movements. To remedy this shortcoming, we examine the argument that the increasing integration of a world economic and political system has encouraged ethnic fragmentation within states. We explore two key arguments about the consequences of global integration. In particular, we argue that integration of the world's states creates a context for comparing distribution processes among groups within states. The result is that existing ethnic inequalities have generated demands for the redistribution of resources and political rights. Both of these ideological and material processes generate movements based in ethnic competition for civil rights and for economic resources within states (see also Olzak 1998).

We seek to explain why ethnic mobilization varies in intensity from small-scale, local protest to mob violence and civil war. Ethnic mobilization is collective action that is based predominantly on some set of ethnic markers, such as skin color, language, migration history, and residential segregation. Different forms of ethnic collective action can potentially emerge at any stage. For example, one can easily imagine a relatively peaceful nationalist movement turning violent over time and then gradually becoming subdued or repressed. Ethnic movements also fail. This could alternatively indicate that their membership declined, interest disappeared or was co-opted, or the movement has been repressed and moved underground or into exile. We begin first with an analysis of some key historical trends associated with the dynamics of ethnic mobilization.

INTEGRATION OF THE WORLD SYSTEM AND ETHNIC MOBILIZATION

We build our integration argument on research that suggests that the diffusion of nationalism can be systematically linked to the rising salience of ethnicity in the modern period. For example, consider Anderson's (1991) observation that both nationalism and ethnic movements involve moral claims over a territory or population (see also Hechter 1987). As Smith (1986), Anderson (1991), Tilly (1975, 1993), and others have demonstrated, the ideological roots of nationalism shaped the nature and legitimacy of modern demands for territorial, administrative, economic parity, and political rights of ethnic groups (Smith 1981, 18).(3)

These trends can be recast sociologically as integrative processes that reinvigorate ethnic boundaries in a variety of ways. Prior to World War I, nationalist themes of ethnic sovereignty undercut the legitimacy of empires in much the same way that separatist movements threaten contemporary states (Gellner 1983). We define nationalism broadly in terms of a process of granting citizenship rights, which diffused first to individuals, then to ethnic, language, gender, and sexual orientation groups that were deemed more and less deserving in a variety of countries. We argue that nationalism and ethnic movements are both consequences of the integration of states and territories into an interdependent world system. In other words, an increasingly integrated system based on nationalist ideology reinforced the integrity of states while encouraging its antithesis in the form of ethnic splinter movements.

Although it has its drawbacks, we think that world system theory provides a useful way to organize our thinking about some of the economic and political constraints on states as they face challenges from ethnic groups. It also provides a coherent explanation for patterns of diffusion of nationalism. It seems useful to consider the link between nationalism and ethnic mobilization using this perspective as well. For this reason, we apply some of the most useful concepts from world system theory to link nationalism with ethnic mobilization. Following this logic, we share Snyder and Kick's (1979) view that arguments from this tradition gain leverage only when they are evaluated empirically.

According to a world system perspective, the density of economic links among the world's states has created a hierarchy of more and less powerful countries. This perspective partitions countries into three sets of dependent relations: core, semiperipheral, and peripheral states (Wallerstein 1976). Core states can be defined as having (1) centrality in trade and military interventions, (2) dominance through the use (or threat) of a superior armed force, and (3) centrality in a network of diplomatic information and exchange, specifically in their role of sending diplomats and authoring treaties (see Snyder and Kick 1979). Periphery states are those that score lowest on these same dimensions of centrality in military, economic, trade, and diplomatic domination. Semiperipheral states occupy the middle ground between more central and less dominant states.

According to a world system theory of stratification, the economic integration of the world system linked various regions, polities, and markets together into a dense and interdependent system. Economic recessions, bank failures, or labor shortages now have repercussions in vastly different and formerly unconnected regions and states. Political turbulence, including actions of ethnic social movements, can produce serious reactions across national borders within minutes or even seconds. It seems reasonable to carry this argument one step further. We consider whether integrative processes have specific, centrifugal consequences for ethnic politics. Put differently, we ask if the magnitude of ethnic violence and nonviolence varies systematically among core, semiperipheral, and peripheral countries.

Some scholars assert that the gradual diffusion of human rights now underlies all modern social movements, including ethnic ones (Smith 1995; Ramirez, Soysal, and Shanahan 1997). Most advocates of this world culture argument do not view it as the result of a simple convergence process (Meyer et al. 1997). Emerging trends in this human rights ideology can be expected to penetrate states that are most integrated into the world system long before they affect trends in isolated countries. In core countries, this would have the consequence of encouraging multiple movements that use conventional political avenues for protesting, lobbying, and voicing grievances.

In contrast, in many peripheral countries, especially those with authoritarian regimes, authorities tend to suppress even mild forms of collective protest (Olivier 1990; Francisco 1995; Rasler 1996). This suggests that, on average, peripheral countries would experience less ethnic protest. At the same time, a threshold effect may build slowly in these settings, so that when protest does erupt in less democratic states, it is more likely to be violent, secessionist, and confrontational (Koopmans 1995; Kriesi et al. 1995).

ETHNIC PROTEST VERSUS ETHNIC VIOLENCE

Ethnic mobilization takes many forms. This section provides a catalogue of the relevant types. Researchers commonly distinguish ethnic protest from intergroup conflict. In an ethnic protest, a group expresses a racial or ethnic grievance, usually to government officials or to the public at large. Examples include a march for voting rights, a sit-in targeted at a discriminating firm, or a boycott against a company that has discriminated against some ethnic group. Conflicts generally involve two or more ethnic groups in contention, as in public hostilities, attacks, lynchings, and skirmishes.

A second distinction can be made between violent and nonviolent ethnic events. Ethnic events can be considered violent if some participants wield weapons or use them to harm persons or property (e.g., in the takeover of hostages or of public buildings).

Ethnic violence can also be arrayed along a continuum of categories, with sustained, armed rebellions at one end and small-scale symbolic attacks at the other end of the continuum (Gurr 1993). Ethnic rebellions are generally distinguished from other kinds of ethnic/racial protest as involving more sustained actions that continue over weeks, months, and years (Olzak and Olivier 1998).

Protest movements that claim to be deserving of special attention by the state but stop short of demands for outright secession are referred to as autonomy movements. These movements often seek to negotiate special regional or group (i.e., reservation) status, but they do not aim for complete separation from the administrative state (at least initially). Unlike rebellion and ethnic violence, they rarely use or advocate armed conflict and terrorist tactics (Hechter 1992).

At the other extreme, violence is directed against a specific ethnic target. Such action is often called exaggerated nationalism and "ethnic cleansing." In the past, this violence has involved goals of ethnic or racial purity that required exclusion of some other group of undesirables. Most recently, this form of violence has arisen as former states (or modern empires) fragment, dissolve entirely, or attempt to forge new and ethnically homogeneous identities (Jalali and Lipset 1992-1993; Akbar 1995). The consequence is often a combination of pogroms, terrorist movements, disenfranchisement, and various types of physical attack, such as rape, genocide, and hostage taking.

Despite these conceptual differences, one ideological undercurrent links ethnic protest, violence, and rebellion. Leaders and supporters of all these types of ethnic protest and conflict use the language of nationalism and self-determination to justify ethnic movements.(4) Examples include recent antiforeigner violence in Germany, France, and Denmark; nationalist movements within Chechnya and Azerbaijan; ongoing separatist movements in Eritrea and Sri Lanka; Serbian nationalism; and tribal civil war in Rwanda. Do these different instances have any features in common? The question prompts us to specify the causal links among ethnic movements, forms of conflict, and levels of violence.

THEORETICAL ARGUMENTS AND HYPOTHESES

This section argues that characteristics of states in the world economic system generate distinctive patterns of ethnic violence, nonviolent protest, and rebellion. We concentrate on (1) the peripheral status of countries and ethnic mobilization, (2) the states with more and less open policies granting civil rights, and (3) the centrality of a state in the world system network of international organizations. We briefly review the relevant literature and then formulate hypotheses for each of these three dimensions. First, patterns of violence and nonviolent ethnic protest ought to differ in peripheral and nonperipheral countries. A second argument suggests that policies of civil liberties shape the form of ethnic protest and violence within countries. Our third argument is that integration into the network of international associations moderates ethnic violence.

PERIPHERAL STATUS AND ETHNIC MOBILIZATION

World system theory rests on the historical argument that the world's states were gradually transformed into economically and politically dominant core nations, a less-developed semiperiphery, and increasingly dependent peripheral nations. This change emerged as the diffusion of a world capitalist system reinforced the dependence of the peripheral nations on core nations. The consequence is that the system of inequality among nations retards a variety of indicators of development in those poor countries, including the diffusion of minority rights. According to this view, the triumph of an integrated world economic and political system widened even small gaps that once existed between richer and poorer regions within and between countries. Global competition among nations intensifies as a result.

Ethnic mobilization can occur in two ways: as dominant groups attempt to reassert their dominance over newly competing groups or as formerly disadvantaged ethnic groups challenge the existing power structure.(5)

In this view, declining inequality provides the resources and organizational infrastructure for new leaders and followers to mobilize ethnic protest. As groups come to compete for newly acquired rights, resources, and political avenues of access, ethnic competition, conflict, and fragmentation within nations occur (Olzak 1992). One organizing principle of this study is that increasing global integration has intensified ethnic competition for a variety of resources.

Various theories on social movements also suggest that a change in levels of inequality intensifies all types of protest and conflict. For example, the literature on Western democracies in Europe leads one to expect that, in most core countries, decreasing ethnic inequality will lead to ethnic protest that employs rather conventional repertoires of protest. Put differently, civil rights tactics of sit-ins, boycotts, and marches are widely legitimated and institutionalized (Smith 1995). Thus, decreasing gaps in ethnic inequality lead to a high frequency of unthreatening protest in countries where the system is open to collective protest. The converse argument is that systems that are relatively closed to protest and public expressions of grievances will be relatively unaffected by changes in ethnic inequality.

There is evidence …