0% found this document useful (0 votes)
245 views11 pagesSeymour - Identity Issues and Civil War
Seymour - Identity Issues and Civil War
Uploaded by
dnmq86Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
0% found this document useful (0 votes)
245 views11 pagesSeymour - Identity Issues and Civil War
Seymour - Identity Issues and Civil War
Uploaded by
dnmq86Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
/ 11
WARNING CONCERNING COPYRIGHT RESTRICTIONS
The copyright law of the United States (Title 17, United States Code) governs
the making of photocopies or other reproduction of copyrighted materials.
Under certain conditions specified in the law, libraries and archives are
_—__authorized to furnish a photocopy or other reproduction.
One of these specified conditions is that the photocopy or reproduction is
not tobe used for any purpose other than private study, scholarship, or
research. If electronic transmissions of reserve materials are used for
purposes in excess of what constitutes “fair use,” that user may be liable for
infringement.
Copyright permissions apply only to the current semester. Downloading
and/or distributing documents is not permitted.
Individuals with administrative rights in blackboard are not allowed to
access documents that are marked as “item is no longer available.” What Do We Know
about Civil Wars?
Edited by T. David Mason and
Sara McLaughlin Mitchell
Jc
328.5
| ROWMAN & LITTLEFIELD .W53
Lanham + Boulder * New York * London 2016 _ TET TE SBOE. 2 + — RR
j
Published by Rowman & Lilefeld
AA‘wholly owned subsidiary of The Rowman & Litlefield Publishing Group, Inc
44501 Forbes Boulevard, Suite 200, Lanham, Maryland 20706
Unit A, Whitacre Mews, 26-34 Stannaty Steet, London SEI1 4AB, Uniced Kingdom
Copyright © 2016 by Rowman 8 Littlefield
All rights esered, No pare ofthis book may be reproduced in any form of by any electronic
‘Gr mechanical means including information seorage and rcrieval systems, without weten|
petmission from the publisher, except by a reviewer who may quote passages ina review.
British Library Cataloguing i Publication Information Available
Library of Congress Cataloging in-Publication Data Available
ISBN 978-1-4422-4224-1 (cloth: alk paper)
ISBN 978-1-4422-4225-8 (pbk. alk. pape)
ISBN 978-1-4422-4226-5 (eleconic)
a
23 The paper used in this publication meets che minimum requirements of American
"National Sandard for Information Sciences—Permanence of Paper for Printed Library
Materials, ANSI/NISO 239.48-1992,
Printed in the United States of America
Contents
List of Figures
Acknowledgments
List of Acronyms
‘What Do We Know about Civil Wars? Introduction and Overview
T. David Mason, Sara McLaughlin Mitchell, and Alysa K. Prorok
PART I: FACTORS THAT BRING ABOUT CIVIL WAR
1 Introduction—Parterns of Armed Conflict since 1945
Nile Peter Gleditic, Erik Melander, and Henrik Urdal
2. Antecedents of Civil War Onset: Greed, Grievance, and
ate Repression
Joseph K. Young
3. Identity Isues and Civil War: Ethnic and Religious Divisions
Lee J. M. Seymour and Kathleen Gallagher Cunningham
4 State Capacity, Regime Type, and Civil War
Karl DeRouen Jr. and David Sobek
5 Transnational Dimensions of Civil Wars: Clustering, Contagion,
and Connectedness
Erika Forsberg
33
4%
0 a Joseph K Young
A challenge i, however, how do we scale these results up? A benefit ofthe quan-
‘itative studies in the eatly 2000s was their generalizability and the possibility that
they represent big trends in post-World War Il conflic. An obvious trade-off was
the internal validity of the research as data were collected at higher levels of spatial
and temporal aggregation. Micro-level studies tend to privilege the exact opposie
‘We can be more confident chat sweep operations in Chechnya from 2000 to 2005
hhad a precise ffect on violence in the region during this time period. We cannot
bbe confident thar this resule can be applied to other kinds of counterinsurgeney ot
applied to Afghanistan or Traq or even to Chechnya from 1993 10 1999. We need
studies that examine both micro- and macro-results, but macro-results work because
lower barriers to entry have been more common,
Relatedly there has been a recent turn coward studying strategic nonviolence. Che-
snoweth and Stephan (2011) and Stephan and Chenoweth (2008) along with work
by Schock (2005) argue that strategic nonviolence is a more effective strategy when
dissidents us it as compared to violence, when challenging a state. Even when faced
with a repressive tate, possibly the critical driver ofa civil war, the tactic of nonvio-
lence can lead to strategic suecess. Since nonviolence attracts greater international
support and sympathy as well as domestic involvement (Chenoweth and Srephan
2011), an increase in the use ofthis strategy could lead to larger reductions in civil
vwar onseis while sil allowing citizens to mobilze in the face of an unresponsive or
repressive regime. In Tunisia, a pro-democracy, mainly nonviolent movement in 2011
broughe down the regime. While the dictatorship in Tunisia did not repress is citizens
as much as other states in the region, other nonviolent movements from South Aca
(1983-1990) to the Philippines (1983-1986) brought down more repressive dictators,
without civil war (Schock 2005). This interest n nonviolence is inherently linked to a
more process-oriented approach. If dissidents can choose these various tactics and be
-met by different responses, chen the state, then only thinking about the big bang of
civil wa, will miss many ofthe antecedents and drivers of the conflict.
Finally, asthe need for predictive models grows, incorporating dissident and stare
interactions along with structural models is likely che mosc comprehensive approach,
‘to explaining the onset of civil war. The quantitative civil war licracure rightly pointed
analysts, journalists, and observers to weak or poor states. Beyond this sorting mech-
nism, we stil need to understand when these conflicts will develop to strengthen
institutions and protec individuals. Early-warning models and prediction based on
real-time data thac deal withthe interactions beeween the state and dissidents secmed
highly unlikely a decade ago, bue the reality is tha this approach is both feasible and
pethaps the best way to evaluate the factors chat lead to civil war onset. For example,
Jay Ullelder, a political scientist and statistical modeler, developed an early-warning
system paid for by the U.S. Holocaust Memorial Museum to provide policy makers
with a predictive tool co head of these atrocities. These daca and models can help
resolve academic debates and also improve the human condition. While debates over
hhow we best think about and model the problem of civil war sometimes sem abstract,
the consequences, unfortunately, for many people in affected areas are all too real
kL
3
Identity Issues and Civil War
Ethnic and Religious Divisions
Lee J. M. Seymour and Kathleen Gallagher Cunningham
Ethnic and religious identities are entangled in most civil wars. By some accounts,
as many a8 64 percent of civil wats have been fought along ethnic lines in recent
decades (Deny and Walter 2014: Themner and Wallensteen 2012). As we write in
Jate 2014, a bloody sectarian war in Syria fought mainly between Alawis and Sun-
nis has spilled over the borders into Iraq and threatens Lebanon's delicate sectarian
politics. The explosive rise of che Islamic State in parts of Syria and Iraq demonstrates
the potency of religious mobilization. Russia has annexed Crimea and intervened
in Ukraine’ civil war in defense of ethnic Russians. In South Sudan, feuding elites
have mobilized tribal miliias and disaffected army elements, with much fighting
‘on the ground pitting rival Nuer and Dinka ethnic groups. These cases underscore
the prevalence of ethnic and religious conflict, reinforce popular beliefs about the
prevalence of identity conflict, and suggest che limits of ethnic and religious accom-
‘modation in many societies. Indeed, approximately 14 percent of the ethnic minor
ties in the world have been involved in significant violence against the state (Gurr
1996; Fearon 2008).
‘Yet most ethnic groups in most places live without experiencing large-scale, orga
nized ethnic violence (Fearon and Laitin 1996). Indeed, much research suggests that
the entanglement of ethnic and religious identity in political violence and civil war
is highly complex. Even cil wars viewed as “ethnic” or “religious” in character are
fought in ways chat complicate popular understandings of the role being played by
such identities (Kalyvas 2003). For instance, in the wars engulfing Syria and Iraq,
notwithstanding che zeal of Sunni Salafists and Shi'a and Alawite radicals on either
side, alliances between armed groups and the regional governments backing them
appear largely opportunistic in nacure. In eastern Ukraine, the roots of the current
conflict between Ukrainians and Russians are arguably more economic than ethnic
(Zhukov 2014). And in South Sudan, the logie of violence is essentially material
a “ Lee JM, Seymour and Kathleen Gallagher Cunninghame
rather than ethnic: “On the surface these appear to be ethnic conflicts, but that isa
product of ethnic patronage char constitutes military units, not deep-rooted tribal
Snimosities’ (de Waal 2014: 362). Whereas most students would quickly point to
thine idenvcies asthe main cause of civl was, che empirical findings aze actually
suite mixed.
Under what condivions are we likely co sce violence organized around ethnic and
religious cleavages, and shat implications does this have for the dynamics of civil
war? We answer these questions by reviewing the voluminous literature on the
Identity dimensions of civil war! We fist define our terms, noting controversies over
fundamental concepts such as ethnicity, identity, and nationalism. The second sec
tion then reviews what we think we know about the conditions under which identity
matters for civil war, examining the role of identity in how wars begin, how they
te fought, and how they end, We conclude by noting exciting diretions for farure
research, While the findings are mixed and che mechanisms connecting identity 0
violence remain debated, the past three decades have seen considerable progress in
out understanding of identity in civil was.
ETHNIC AND RELIGIOUS
IDENTITIES IN CIVIL WAR RESEARCH
The politics of identity came relatively late ro the academic study of civil war. In
1970, for example, a decade aftr the tuen to identity polis in the social sciences,
“Ted Gurts path-breaking contribution Why Men Rebel made lite direct reference to
ethnic or religious identities. Literatures on political violence and social revolution,
fon one hand, and studies on ethnicity and nationalism, on the other hand, devel-
‘oped largely in parallel with cach other. Only with Horowites (1985) Eidmic Groups
in Conflict did the two lage literatures come together. In the following rwo decades
(1985-2005), che number of articles dealing with ethnicity and nationalism in
politcal science roughly quadrupled (Cederman 2012). Civil war has been a major
focus ofthis literature, alongside other violent episodes such as rots, terrorism, mass
killing and genocide, and state violence and repression.
Several factors converged to drive his interest inthe confluence of identity polities
and civil war. These included che realization chat violence is a key impediment to
development, which raised the profile of identity politics in economics and political
‘economy. Another reason for the focus on ethnic conflict was the marked decline
in interstate war that saw a number of international relations specialists warn theie
tention t0 “internal” conflicts. Events themselves were a major drives, in particular
the breakup of the Soviet and Yugoslav states in a burst of ethno-nacionalism, wich
shocking violence in the intertlated separatist and irredentist wars in Yugoslavia
“This coincided with emergence of a number of failed states racked by identity con-
flices and incidents of echnically motivated violence and mass killings, especially che
1994 Rwandan genocide.
entity howe and Civil War 6
Similarly, interest in religious identity, a relatively neglected area of study aside
from the debate sparked by Huntington's Clash of Civilizations (1996), surged after
the terrorist atacks of September 11, 2001. The emergence and persistence of relic
giously inspired and organized insurgent groups in places such as Somalia, Kenya,
Yemen, Nigetia, Chechnya, Syria, and Iraq has driven a renewed focus on religious
identity. Despite echoes of a “clash of civilizations.” these conflicts arguably have
their roots in parochial tribal rivalries and historical relations beeween center and pe-
riphery rather than transnational conflict between religiously organized civilizations.
‘The murkiness and complexity ofthese conflicts underscores the need to shed light
con the particular identities and grievances entangled in these wars.
‘Much research has focused on problems of defining and operationalizing key
terms such as identity, ehnicty and religion. An ident is a social category denoting
some fundamental and consequential sameness” in which an individual is eligible
to be a member (Brubaker and Cooper 2000). Hale (2004) conceives of identity as
a kind of “Social radar” that provides individuals with a point of reference, allowing
them co situate themselves within a wider group and understand hows their member-
ship affees them in the social world, including relations with other groups.
Ethnic identity is correspondingly conceived as a subsee of identities in which
cligibility for membership is determined by accibures based on common descent
(Horowitz. 1985). A crucial property of these descent-based attibutes is that they
are both difficult wo change in the shore term and visible o other members of society
(Chandra 2006). More specifically, “atribuces associated with or believed to be as-
sociated with” ethnicity inchude
those acquited genetically (eg, skin color, ender hair eype, eye color, height, and
physical feacures),chrough cultural and historical inheritance (eg, name, language,
place of birth, and origi of one's parents and ancestos), or inthe course of one life-
time as markers of such an inberiance (eg, lst name or tribal markings). Atributes
“Telived to be associated with descent” are aeibutcs around which credible myth of
association with descent has been woven, whether or not such an asociation exists in
fact. (Chandra 2006: 400)
Following Horowitz (1985: 46) and Rogowski (1985), we can distinguish the
“surength of ethnic markers’ as a function of how recognizable they are and how
costly they are ro change. This leads to two expectations: first, the stronger the ethnic
_markers that distinguish members ofa group from its rival, the more easily members
of one group can be singled out for discriminatory treatment by the members of the
rival group; and second, the stronger those markers, the les likely is assimilation and
‘the more likely conflict beeween the groups
Religious identity is sometimes synonymous with ethnicity but, conceived in
terms of the sctength of ethnic markers, is generally harder to identify (or easier to
conceal) and less costly to change. For some groups, religious and ethnic identities
may be more of less the same, as with Armenians who are overwhelmingly Arme-
nian Orthodox, or Ashkenazi Jews, whose ethnic identity is wrapped up in Judaism. 6 Lee J. M. Seymour and Kathleen Gallagher Cunninghame
However, as Ashutosh Varshney notes, the distinction between ethnic and religious
identicy “becomes critical... when ethnicity and religion clash (East and West Paki-
sean before 1971, Kashmiri Hindus and Muslims, Irish Protestants and Catholics,
black and white American Christians)” (Varshney 2009),
Thus, while religion is often an important feature of ethnicity, the exo concepts
are not synonymous. One way of thinking of religious identities as sepa
tional or echnic identities is to emphasize the element of practices such as “surifice,
prayer, meditation, pilgrimage, war, proselytization and charitable acts’ (Toft 2011:
674), Religious identities are thus rooted in
from na
a system of practice and belie in the attainment of beneficial personal or collective
shift i existence (heaven, nirvana, paradise, salvation, ecstasy, transcendence, oneness,
peace), by means of ating ot aot acting in specific ways which are constitutive of an
‘established community practice, and for which empirical referent are either unnecessary
0% indeed, anathema. (Toft 2011: 674)
Religion shares ethnicity’s descent-based character insofar as one is often born into
a religious identity, chough the intensity of religious identities are potentially ted co
the extent one ceases to practice and believe in a way that ethnicity is not. We un-
derstand what it might mean to be a lapsed Catholic who no longer attends church,
cor nonobservant Muslim who does not pray and enjoys the occasional whiskey, oF
an atheist Jew who eats bacon and shellfish. The same individuals can have tightly
‘overlapping cthnic identities as, say, Poles, Somalis, or Jewish Americans, but ethnic:
ity is noc necessaily diminished by che same actions. Conversely, because ethnicity
involves descent-based attributes, converting to Catholicism, Sunni Islam, or Juda~
ism, does nor necessarily make one ethnically Polish, Somali, or American Jewish.
These definitions are not withoue their erties. As an analytic category, identity
is “riddled with ambiguity, riven with contradictory meanings, and encumbered by
reifying connotations” so much so that some scholars propose dropping the erm
altogether (Brubaker and Cooper 2000: 34). Chandra has argued that many of the
causal claims made for ethnicity in political science research concern properties that
are not intrinsic co ethnic identity (2006). Such claims concern characteristics such
a fxis territorial concentration, dense social neeworks, cultural cohesiveness, or the
‘emotional reetions that identitis elicit, all of which vary across both ethnicities and
other identity groups. Similarly, the relationship of clan case, tribal, sectarian, and
linguistic identities o ethnicity, and of ethnicity to nationalism, remains contested
(Calhoun 1993; Schaez 2013).
The ambiguity of identity complicates measurement and comparison, particularly
in large-N datasets. Many arguments about ethnic identiy stress the importance of
diversity, with various measures aiming to capcure polarization and factonalization
in a given population (Easterly and Levine 1997). Early measures of “ethno-linguis-
tic frationaliation,” or ELF, calculated a concentration index on the basis of Soviet,
‘ethnographic data collected in the 1960s and published in the Arlas Narodov Mina
(1964) number of studies using ELF found chat both very homogenous societ-
dentty Ics and Coil War ”
ies (with litde to no diversity) and very diverse societies (where no single group can
dominate the state easily) have a much lower rsk of civil war. However, cis measure
has been much criticized as static and failing to accoune for che political salience
of specific identities. Alternative measures emphasize different ways of calculating
fractionaliztion or restricting which populations are included. For example, Fearon
(2003) relies on linguistic difference co measure the overall diversity of a country.
Posner (2004) offers an index of politically relevane ethnic groups (PREG) for forty-
‘wo countries in Aftica. Montalvo and Reynal-Quetol (2005) create an index of
ethnic polarization thac measures how far the distribution of groups is from a bipolar
discibution, which captures the maximum degree of polarization.
‘While ELF and its successors all caprure ethnic (and sometimes religious) diver-
sity asa property of a country, another set of data projects centers on the ethnic or
religious group as the central actor, rather than the state. Some of the eatiest work
to collect data about identity groups took place in the Minorities at Risk (MAR)
project. While representing an important step in collecting data on groups at risk of
rebellion, protest, or repression, che dataser effectively examined only a set of cases
where identity already corresponded to mistreatment by the state, complicating
testing against a wider ser of ethnic groups not at tsk. More recent data collection
efforts, such as those behind the Ethnic Power Relations dataset, employ country
expertise to capture the political relevance of diferent ethnic groups and patterns of
horizontal inequality across them (Wimmer, Caderman, and Min 2009). This has
also been paired with spatial modeling techniques to geographically reference iden-
ticy groups. Thus, we now know much mote about the geographic dstibution of
etic identity, even down to sub-cthnic segments defined by language and religion
‘This makes hypothesis resting possible on the bass of much more fine-grained data,
enabling researchers to connect causal mechanisms to particular identity cleavages.
Disaggregating among the elements of ethnic identity, for instance, one study finds
that intrastate conflict is more likely within linguistic dyads than religious ones (Bor-
‘mann, Cederman, and Vogt 2015)
‘Academic debates over how to define and measure these concepts have real-
world relevance. One particular danger is that highly aggregated notions of identity
shift from being categories of analysis employed by scholars ro investigate social
phenomena into categories of practice inherent in everyday social experience used
by ordinary actors (Brubaker and Cooper 2000). By uncritcally adopting certain
identities as categories of analysis, scholars can reinforce and reproduce the efforts of
politcal entrepreneurs who use ethnicity or religion as if they existed primordially
and essentially, rather than as socially constructed categories—either as “invented
traditions" (Hobsbawm and Ranger 1983) or “imagined communities” (Anderson
1991). For instance, in the Soviet Union, designation of certain ethnic groups as
“nations” entiled co varying degrees of institutional starus, a practice reflecting
prevailing nationalist thinking, arguably contributed to the USSR’s collapse chrough
a series of national revolutions that led to still simmering civil wars (Bunce 1999;
Beissinger 2002; Roeder 2007). Campbell (1998) argues that academic portrayals 8 Lee J. M. Seymour and Kashlen Gallagher Cunningham
of ethnic violence in Bosnia ding the bloody cv wat of 1992-1995 played into
the hands of ehno-nationals who sought portay the coextence of muliple
rational groups on Bosi teitory os unnatural and undesiabe. Mahmood Mar
dni (2010) makes similar pont about the way violence in Darfur was uncitcaly
porcayed bythe media and academics, Rather than viewing the warn Darras @
roy nny hing cnr wi comple enn i
violence wat portayed ata one-sided genocidal campaign pitting ethnle“Aicans™
vers “Arabs in ways that sought to legimate Westra inervencon (Seymour
2014). In sum, the at of dexgnating certain ethic or religious communis 35
{r0ups and making clams abou the consequences of roupnes cls for crf at
tention, especially where violence is concerned (Brubaker 2004).
UNDER WHAT CONDITIONS AND HOW,
DOES IDENTITY MATTER IN CIVIL WARS?
‘Turning to explanations that situate identity in important civil war processes, we
‘examine arguments examining ethnicity and teligion in civil war onses, dynamics
within wars, and their duration and termination, Taken together, this work makes
« convincing case that identity is central to evil wars and political violence, bu the
causal connections are complex and often contingent,
Identity and the Causes of Civil Wars
Ldeny-based arguments have played an important ein explanations for how
er
sei summary of thie ey mechanism inking etic deny toc war “Rebel
er ————C
[i] more ape be aggre, (2] bee able to mobile, and 3] moc ikl ace
dial bargaining challenges computed o other groupe (2014: 200), Following
this insight, we focus on diferent theresa popostons explaining how deni
saps ptf rns bon, an ging en thao
eee ————C
covlement that coked pasion, amet, and apprehensions as aval group
contested thei relative supesixity wishin a state, “Ethnic conflict arises from the
commen erliatvesgnfeance corded by the poups o acknowledged ferences,
tien played out in pbc tub of aman se conescton 1985227)
Other authors make similar argument linking ehnie pievanes to onteseion
over the sate or state-based dicrimiasion (Gur 1993; Gut and Moore 1997
The scalpychologal elements of Horowis argument rest on view of echnie
iy wooed in iship chat, while fective loaned by co-ethnic This bel hes
parila levance fs how group member pecive ets As Hlrowie we,
entity Isues amd Ctl War o
group members are potential kinsmen, a threat ro any member of the group may be
seen in somewhat the same light asa chreat to the family” (1985: 65). The eonfla
tion of ethnicisy with kinship accounts for the increased probability and intensity
of ethnic conflict in societies where political contestation becomes organized slong
ethnic lines.
Primordilist or essentialist arguments take this insight further, emphasizing the
itrational psychological and emotional mechanisms behind the sense of Kinship that
pervades ethnically and religiously organized violence. As Connor argues, “people do
rot voluntarily die for chings that are rational” (1994: 206), The symbols, poetry
songs, metaphors, and recurrent images behind ethnic appeals—"blood, family,
brothers, sisters, mother, forefathers, ancestors, home”—tap into the sense of Kinship
that underlies ethnic and religious slidarities (Connor 1994; also Kaufinan 2001)
Scructurl changes can trigger emotional mechanisms of fear, hatred, and resentment
thar promote ethnic violence and conflier. Roger Petersen (2002) has taken this
insight Fureher in arguing thatthe same “ancient hatreds” many academics are quick
to dismiss do in fact serve as “schemal that shape ethnic violence, with emotions
catalyzing grievances and sustaining collective action (eg, Petersen 201 1; MeDoom
2012; Pearlman 2013)
‘An important alternative to arguments based on seemingly innate identity
emphasizes instrumentalization, particularly by entrepreneusialclites who exploit
cthnicicy’s mobilizational advantages (Bates 1983). Against an essentialist view that
identities endure and give rise to deep-seated grievances between groups compet-
ing for satus, the instrumentalist argument porerays identity and che grievances
to which ic give rise as malleable, or, at the extreme, even epiphenomenal. "Pure
instrumentalists emphasize individual calulus in an identity markerplace in which
ethnic entrepreneurs can create and sell new identity categories to willing buyers”
(Gambanis and Shayo 2013: 299). The focus here shifts from the emotions and
passions that ethno-nationalism and religion evoke to the strategic rationale behind
promoting mobilization around particular identities according to self-incerest. For
‘example, V. P Gagnon argues that “the violence inthe former Yugoslavia was a strate
fic policy chosen by lites who were confronted with political pluralism and popular
mobilization” (2006: 7). Valentino (2000) makes a comparable argument about
the cite calculus behind mass killing, arguing that events like the 1994 Rwandan
genocide reflec brutal strategies designed to counter threats zo laden’ power and
advance their interests. Another rationalist argument emphasizes that social mecha-
nisms of sanctioning within groups promote cooperation and mobilization without
dlite manipulation. An experiment in a poor neighborhood in Kampala, Uganda, for
instance, found that “co-ethnics play cooperative equilibria, whereas non-co-ethnics
do nos,” and that ‘co-ethnics are more closely linked on social networks and thus
plausibly better able to support cooperation through the threat of social sanction”
(Habyarimana et al. 2007). Ethnicity is thus a potent cleavage around which co oF-
ganize collective action because its prone co manipulation from above and because
‘ethnic homogeneity induces more cooperative behavir from below. 50 Lee J. M, Seymour and Kathleen Gallagher Cunningham
[A different set of perspectives downplay the role of identity aleogether. A number
of scholars have emphasized the degree to which group mobilization and individual
pamicipation in violence can instead be linked to opporcunty. While ethnicity gives
would-be rebels a stronger base of suppore ftom which ro recruit, a social base in
particular identity is insufficient without the opportunity to rebel afforded by
financing or sate weakness, Many ethnic (and some religious) groups tend co be
geographically concentrated, making the physical act of mobilizing easier (Gates
$3002; Toft 2003), Fearon and Laitin (2003) argue chat state capacity isa key variable
‘explaining rebellion. Insurgencis have the opportunity to mobilize in peripheral at-
‘av of weak state eapacity with minim police or military presence (sce also Buhaugs
Gates, and Lujala 2009), Others emphasize “greed” as a motive for rebellion (Collet
and Hoeffler 2004), downplaying the role of grievances and emphasizing che gains
to be had from lootable resources and diaspora funding that provide the key to f=
hrancing wats (Le Billon 2001; Ross 2004a, 2004b), Mueller (2000) shows chat even
the disputes commonly considered “echnic”(Yagoslavia and Rwanda) include high
fevels of apparendly opportunistic participation in violence by antisocial elements
of saciery. The role of opportuni factors, broadly construed, rather than identity
factors, has been borne out in empitial work on civil wa. “Cross national saitical
studies find surprisingly few differences berween che determinants of civil war onset
in general, versus ‘ethnic’ civil was in particular” (Fearon 2008, 857-58), However,
a series of studies have presenced counterevidence to this. For example, Hegre and
Sambans (2006) find chat identity is associated with conflict at lower levels of armed
conflict than ate oen used to test his empirically.
"Another set of explanations for conflict outbreak and perseverance focus on the
bargaining process from a rationalise perspective, observing the escalation ¢o civil
svar as an outcome ofa fulure to achieve an ex ante setlement (Powell 2006; Walter
53002, 20096). War, in chis view, is “inefficienc” in the sense cha the costs incurred
