Ethnic Politics and Conflict/Violence

State of the Field and New Directions

Erika Forsberg, Jóhanna K. Birnir, Christian Davenport

• Publisher: Taylor & Francis
• Year: 2018
• eBook ISBN: 9781351725286
• Edition: 1
• Topic: Politics & International Relations
• Subtopic: Politics
• Index: Politics & International Relations

Ethnicity is one of the most salient and enduring topics of social science, not least with regard to its potential link to political conflict/violence. Despite, or perhaps because of, the concept's significant use, all too seldom has the field paused to consider the state of our knowledge. For example, how do we define and conceive of ethnicity within the context of political conflict? What do we really know about the causal determinants of ethnic conflict? What has been the most useful development within this literature, and why?

This volume comprises reflections from an international range of prominent political scientists all engaged in the study of ethnicity and conflict/violence. They attempt to synthesize what the field does and does not know with regard to ethnic conflict, as well as draw out the research directions for the immediate future in unique and interesting ways.

This book was originally published as a special issue of the journal Ethnopolitics.

The Field of Ethnic Conflict Studies: An Interplay of Theory with Reality

MONICA DUFFY TOFT

ABSTRACTAlthough there existed research on ethnic conflict prior to the 1990s, since then a combination of real-world events and advances in comparative case design and statistical modelling accelerated the field of inquiry. No longer were conceptual issues the core concern, but we witnessed more concerted efforts to build general explanations engaging a diverse set of questions; including the conditions under which ethnic identity is politicized, comes into conflict or conciliation, or is implicated in violent struggle and war. Statistical modelling and in-depth, sophisticated case comparisons have become the norm. Although definitive knowledge on the dynamics and mechanisms of ethnic conflict remains a work in progress, we are now in a position to make better generalizations.

Brief Introduction

As a scholar who has been working on ethnicity and violence for the past two decades, it is a welcome task to reflect on where the field has been and where it is going. When I started my dissertation in the mid-1990s, the field of ethnic politics and its relationship to large-scale violence was not well developed (important exceptions include Horowitz, 1985). In the international relations (IR) subfield of political science, scholars focusing on interstate and nuclear warfare largely overshadowed those who focused on identity-based violence, in large part because it was thought that sub-state violence was so much less likely to cause serious destruction by comparison. Moreover, in mainstream discussions, scholars tended to accept the view that most cases involving ethnicity were sui generis. Thus, the question ‘what causes ethnic violence?’ was rarely asked, much less answered. Little in the way of comprehensive theory on ethnic violence emerged, largely because the focus of inquiry centred on the causes of revolutions, riots, and societal discontent more generally, and tended to focus on particular cases with little comparative analysis (Eckstein, 1964). In terms of policy implications, most policy makers appeared to hope the thorny and complex issues raised by the observable rise in sub-state identity-based violence could be ignored.

What We Know: Reality Impacts Scholarship

With the end of the Cold War, which coincided with the start of my academic career, interest in sub-state violence captured the attention of policy makers, and later academics. Four overlapping factors made understanding of ethnic violence more important for policy makers, and more interesting to the security studies and IR communities.

First, the very nature of the demise of the USSR had been clearly nationalist and had come not from without, but from within (Beissinger, 2002; d’Encausse, 1981). The real capacity of ethnic identity to bring down a state, a major power, had been clearly demonstrated. Groups within a state had acted in a way that fundamentally—structurally—altered the reality of interstate politics (Roeder, 2007; Toft, 2003).

Second, ethnic war returned to continental Europe in a way not seen since WWII. Civil wars may not be as absolutely destructive as world or nuclear wars, but the horrors of Yugoslavia’s progressive disintegration were real, public, and proximate. Civil wars elsewhere (e.g. in Africa and Latin America) demonstrated an increased capacity to destroy life and eviscerate costly infrastructure. Ethnic wars were also seen as more difficult by nature to halt by means of intervention and negotiation, because the objects of such wars often seemed to be the physical destruction of an ethnic group either by killing or by forced migration, such that material inducements and security guarantees could no longer be relied upon to stop the violence.¹

Third, transportation and communications technology made it possible for the first time in history to cause the rapid movement of large numbers of people from the developing to the developed world. The costs of disease, poverty, and war in the developing world could no longer be so easily contained.²

Fourth, statistical testing of relationships around ethnicity, politics and violence was non-existent, and case-based, qualitative work had been largely anecdotal. Although sustained inquiry into the causes of ethnic conflict actually dates from the 1980s with the emergence of works by Donald Horowitz’s Ethnic groups in conflict (1985)—now considered the seminal text³—prior to the mid-1990s, questions about ethnicity were generally directed towards explaining ethnicity as such (and as it related to national identity, for example),⁴ rather than the relationship between ethnicity as a concept and the likelihood, frequency, or intensity of identity-based political conflict.

I would contend that two scholarly contributions—Roy Licklider’s ‘The Consequences of Negotiated Settlements in Civil Wars, 1945–1993’ (1994) and the Minorities at Risk (MAR) data set—profoundly shaped the next decade of scholarship both empirically and theoretically.

Licklider’s article accomplished two important tasks. First, by publishing in the flagship journal of the political science discipline, Licklider helped to bring the study of civil wars into the mainstream. Second, although he presented only a rudimentary data set of civil wars, Licklider divided them into two types; distinguishing between identity-based and non-identity-based civil wars.⁵ He then presented some preliminary findings about the differences between the two.

Licklider’s article proved a watershed in the study of ethnic wars for two reasons. First, sub-state violence became more widely accepted as a subject worthy of research in its own right. Second, his separation of civil wars into two types appealed to scholars who were witnessing the outbreak of ethnic violence in places that were supposed to have been immune; and more importantly, suggested there were useful—crucial even—generalizations that could emerge from a systematic comparison of civil wars across time and space.

At about the same time, the University of Maryland, under the guidance of conflict scholar Ted Gurr released the MAR data set (Gurr, 1993a, 1993b). This was the first such data set that isolated and coded ethnic minorities across the globe and since 1945. It allowed scholars to investigate the conditions under which ethnic groups rebelled, faced discrimination, and became engaged in political conflict, and sometimes violent confrontations with the state. MAR’s multitude of variables allowed scholars to investigate how ethnicity entered politics and its inclusion of different ethnic groups within one state or one group across different states allowed scholars to study the variation in the salience of ethnic identity in an array of political, social and economic environments.

Prior to this period of development, ethnic conflict itself as an outcome was sometimes considered, but not in a rigorous way. Again, consider David Horowitz’s seminal book, Ethnic groups in conflict, which broaches the issue of ethnic conflict. His analysis and insights are presented in broad terms. This is helpful, but only to a point. The book’s theorizing is often contradictory, and the multitude of historical cases serves as illustrations and not tests of key propositions.

Thus, the development of larger and better data sets allowed for the statistical testing of the conditions under which ethnicity as a variable might or might not contribute to political outcomes. These findings were then combined with the systematic application of the case study method, which is when the field really began to mature. I am a product of that time. I was trained as an IR scholar at the University of Chicago to develop general, parsimonious theories about the nature of politics in the style of Kenneth Waltz (Theory of international politics, 1979) and to deploy rigorous methods to test hypotheses in support of theory building. However, rather than explore the dynamics between states, I decided to focus on conflict and violence dynamics within states, and in particular, among ethnic groups and what that meant for global politics and the stability of states (Toft, 2003).

Findings have emerged from this initial period that held the test of time. These include geographic concentration of ethnic groups; a history of lost autonomy; political exclusion; and the resilience of ethnic identities over time (once a given ethnic identity becomes part of a political environment, it tends to remain a part of it) (cf. Gurr, 1993a, 1993b; Jenne, 2007; Roeder, 2007; Saideman, 2001; Toft, 2003).

Most Useful Developments

Perhaps the most useful development over time has been the creation of data sets exclusively for the testing of criteria related to ethnicity and political behaviour; with the forerunner being the aforementioned MAR (http://www.cidcm.umd.edu/mar/) data set. It was and remains one of only a few data sets that isolates ethnic ‘groups’ as a discrete political and social entity, allowing them to be interrogated as actors in the political arena.

Although some scholars have pointed to limitations of this data set (e.g. bias in terms of the cases in that only active or actively discriminated against groups are included), many of these critiques are misplaced, and result from the tendency of many critics towards interest in questions beyond the scope of the data set: for example, under what conditions do ethnic groups form? By design, MAR cannot address such questions, since the origins of a group are a starting condition for a group to be included in the data set.

A more recent and useful development is the Ethnic power relations data set (Cederman, Wimmer, & Min, 2010), and the geocoding of groups and events that allows for a more localized assessment of conflict dynamics in data sets (e.g. Armed Conflict Location and Event Data Project (ACLED): http://www.prio.no/CSCW/Datasets/Armed-Conflict/Armed-Conflict-Location-and-Event-Data/). What these data allow for is a critical assessment of the institutionalized power dynamics among ethnic groups within states.

As an explanation of ethnic violence, the institutionalist approach has a number of important advantages. First, institutions, unlike ‘identities’ and ‘charisma’ appear to be more easily quantifiable and comparable. To the extent this is true, this would facilitate testing and theory building. Second, the logic is sound: power-sharing institutions allow affected groups a say in their own government. It makes sense that groups within states that are more democratic should be able to resolve ethnic disputes short of violence (Hegre, Ellingsen, Gates, & Gleditsch, 2001), and that an ethnic group’s relationship to its political institutions influences that groups access to political and economic goods, levels of grievance, and capacity for mobilization (Cederman, Weidmann, & Gleditsch, 2011; Wicherpfennig, Metternich, Cederman, & Gleditsch, 2012).

Least Useful Developments

The least useful development over time also relates to data, and that is the use of more generalized civil war data sets that are incapable of differentiating ethnic and non-ethnic conflicts (e.g. initially the Correlates of War: http://www.correlatesofwar.org), or fail to include variables that accurately capture how ethnicity might play out in politics.⁶

This came about for data-driven and theoretical reasons. First, most data sets (even today) do not differentiate between different types of civil wars such as ethnic or non-ethnic. In part this was driven by data availability in that most of the indicators were state-level aggregate indicators, such as GDP per capita, but in other cases it was the result of theorizing that ethnicity provides nothing unique in understanding politics.

This bias against ethnicity as an autonomous factor in explaining political conflict was further compounded by the introduction and widespread adoption of an indicator designed to measure the fractionalization of a country: Ethnolinguistic Fractionalization or ELF (Reyna-Querol, 2002).⁷ Not only were the data used to derive ELF four decades old, but ELF was based on a source that codes only for linguistic groups. Language, however, is only one way in which ethnicity might define groups in a country, others might include race or religion (Toft & Zhukov, 2015).

Perhaps most importantly, ELF inaccurately captures the interactions that actually take place between ethnic groups and therefore how this might or might not lead to ethnic conflict. As a country-wide measure ELF is simply too lumpy to account for the multitude of geographies and settlement patterns of ethnic groups. To say that a member of one group is as likely to meet members of another group based on a measure of the relative proportions of the group populations countrywide fails to consider such factors as rural and urban environments, as well as intermingled or isolated, concentrated ones. Again, there is a lack of theoretical appreciation of what makes ethnic groups and ethnicity unique (e.g. attachment to localities an...

Table of contents

• Cover
• Half Title
• Series Page
• Title Page
• Copyright Page
• Contents
• Citation Information
• Notes on Contributors
• Introduction: State of the Field of Ethnic Politics and Conflict
• 1 The Field of Ethnic Conflict Studies: An Interplay of Theory with Reality
• 2 Disaggregating Identities to Study Ethnic Conflict
• 3 Inequalities Between Ethnic Groups, Conflict, and Political Organizations
• 4 Bringing Perpetrator and Victim Identities into Empirical Ethnic Conflict Research
• 5 Observations on the Study of Ethnic Conflict
• 6 Focus on the Fundamentals: Reflections on the State of Ethnic Conflict Studies
• 7 Above, Below and Beyond the State: New Directions in Ethnic Conflict Studies
• 8 What do We Know as a Field about the Causal Determinants of Ethnic Conflict?
• 9 New Directions in the Study of Ethnopolitics
• 10 ELF Must Die: Institutions, Concentration, the International Relations of Ethnic Conflict and the Quest for Better Data
• 11 Ethnic Conflict: Looking Inside Groups
• 12 Why Does Ethnicity Increase in Salience as Political Order Decays?
• 13 Ethnic Conflict: Questions and Answers
• 14 Old Wine, New Bottles: Synthesis and Integration in Ethnic Conflict Research
• Index