-
This article was downloaded by: [King's College London]On: 23
September 2013, At: 11:01Publisher: RoutledgeInforma Ltd Registered
in England and Wales Registered Number: 1072954Registered office:
Mortimer House, 37-41 Mortimer Street, London W1T 3JH,UK
Security StudiesPublication details, including instructions for
authorsand subscription
information:http://www.tandfonline.com/loi/fsst20
Why they fight: Hypotheses onthe causes of contemporarydeadly
conflictDaniel Byman a & Stephen Van Evera ba Policy analyst at
the RAND Corporation,b Associate professor of political
science,Massachusetts Institute of Technology,Published online: 24
Dec 2007.
To cite this article: Daniel Byman & Stephen Van Evera
(1998) Why they fight:Hypotheses on the causes of contemporary
deadly conflict, Security Studies, 7:3, 1-50,DOI:
10.1080/09636419808429350
To link to this article:
http://dx.doi.org/10.1080/09636419808429350
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy
of all theinformation (the Content) contained in the publications
on our platform.However, Taylor & Francis, our agents, and our
licensors make norepresentations or warranties whatsoever as to the
accuracy, completeness,or suitability for any purpose of the
Content. Any opinions and viewsexpressed in this publication are
the opinions and views of the authors, andare not the views of or
endorsed by Taylor & Francis. The accuracy of theContent should
not be relied upon and should be independently verified withprimary
sources of information. Taylor and Francis shall not be liable for
anylosses, actions, claims, proceedings, demands, costs, expenses,
damages,and other liabilities whatsoever or howsoever caused
arising directly orindirectly in connection with, in relation to or
arising out of the use of theContent.
This article may be used for research, teaching, and private
study purposes.Any substantial or systematic reproduction,
redistribution, reselling, loan,
-
sub-licensing, systematic supply, or distribution in any form to
anyone isexpressly forbidden. Terms & Conditions of access and
use can be found
athttp://www.tandfonline.com/page/terms-and-conditions
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
WHY THEY FIGHT:HYPOTHESES ON THE CAUSES
OF CONTEMPORARY DEADLY CONFLICT
DANIEL BYMAN AND STEPHEN VAN EVERA
THE COLD WAR'S end in 1989 evoked both euphoria and gloom
aboutthe prospects for a peaceful world. Many greeted the fall of
theBerlin Wall as a harbinger of a tranquil new millennium. Then
anopposite view emerged as violence erupted in the Persian Gulf,
the Balkans,the Caucasus, Central Asia, and elsewhere: the world
was falling into a "newworld disorder." One scholar claimed in 1992
that "while the end of diecold war has greatly reduced the chance
of global nuclear catastrophe, ithas, inadvertendy, increased the
chances for lesser disasters such as regionalwars."1 Anodier argued
in 1993 that "the key narrative of the new worldorder is the
disintegration of nation-states into ethnic civil war."2
What pattern has in fact emerged? Specifically, has violence
increased ordiminished since die end of die cold war? Where is
postcold war violencelocated, and what form does it assumecivil or
international? What are itscauses? What future for war can we
extrapolate from current conditions?These are the questions diis
paper addresses.
We argue that the optimists of 1989 were closer to the truth
than theCassandras: there is no "new world disorder." While the
post-cold war
Daniel Byman is policy analyst at the RAND Corporation; Stephen
Van Evera is associateprofessor of political science at the
Massachusetts Institute of Technology.
The authors would like to thank Michael Brown, Taylor Seybolt,
Jeremy Shapiro, BenjaminValentino, and the reviewers of Security
Studies for comments on earlier versions of this work.
1. Kim R. Holmes, "The New World Disorder," The Heritage
Lectures no. 42, 22October 1992.
2. Michael Ignatieff, Blood and Belonging (New York: Noonday
Press, 1993), 5. DanielSchorr likewise noted in 1994 that the cold
war's end had spawned "conflict and miserymore horrible than the
theoretical visions of superpower collision. The danger now is
notbombs but people, people in rage against each other and people
fleeing from the rage"(Daniel Schorr, "End of Cold War Leads to
Ethnic Strife," 6 September 1992, broadcast onNational Public
Radio).
SECURITY STUDIES 7, no. 3 (spring 1998): 1-50Published by Frank
Cass, London.
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
2 SECURITY STUDIES 7, NO. 3
world is hardly tranquil, the number of active wars worldwide
has fallenmarkedly since 1989. The number of civil wars has fallen
significantlyeleven states experienced civil wars in 1996 compared
with seventeen in1989and international conflict has nearly
vanished, at least for now.
We hypothesi2e that seven causes of civil violence stand out
inimportance (that is, in their potence and prevalence) in the cold
war andpostcold war periods. They are: 1. The collapse of
postSecond WorldWar empires; 2. A lack of regime legitimacy; 3.
State weakness; 4.Communal hegemonism; 5. Revolutionary ideology;
6. Aristocraticintransigence; 7. Superpower proxy wars. Together,
we argue, these sevencauses account for most civil violence in
recent times. These hypotheseswere inferred by studying all the
civil wars that have occurred since 1989.Some also were borrowed
from existing works on civil violence.
Our first hypothesis is that the collapse of empire, embodied in
the 1991collapse of the Soviet Union and Yugoslavia, generated a
number of recentwars. Imperial collapse spawns successor states
that are primed for civilwar. These states often have illegitimate
regimes, commingled hostilepopulations, and contested borders.
Moreover, no "rules of the road"define the rights and
responsibilities of the former metropole or otherexternal powers in
the former imperial zone; hence outsiders oftenintervene to claim a
sphere of influence or to disrupt another power'ssphere, sparking
civil conflict. These dangers caused considerable carnageafter many
past imperial collapses. Massive killing occurred after the
3. We use the terms "civil war," "civil violence" and "civil
conflict" interchangeably inthis paper. By these terms we include
political conflicts with the following attributes: (1) atleast
1,000 deaths during the total span of the conflict; (2) the people
involved in the violenceare geographically contiguous (to exclude
European colonial wars); and (3) the peopleinvolved are concerned
about living together in the same political unit. This
definitionincludes organized civil wars and also communal riots,
pogroms, ethnic cleansing, and otherinstances where bloodshed was
high but only one side was responsible for the killings whilethe
other side (or sides) suffered disproportionately.
This definition resembles the definitions other scholars use for
civil war but is broaderthan two commonly accepted definitionsthose
of Roy Licklider and of J. David Singer andMelvin Small. See Roy
Licklider, "How Civil Wars End," in Roy Licklider, ed., Stopping
theKilling (New York: New York University Press, 1993), 9, and
Melvin Small and J. DavidSinger, Resort to Arms: International and
Civil Wars, 1816-1980 (Beverly Hills: Sage, 1982), 210.Licklider
defines civil war to include "large-scale violence among
geographically contiguouspeople concerned about possibly having to
live with one another in the same political unitafter the
conflict." For Licklider, however, civil war must also involve
multiple sovereigntya fact that distinguishes civil wars from other
types of domestic violence. We do not,however, include multiple
sovereignty in our definition. In their definition of civil war,
Singerand Small are careful to exclude both "regional internal
war"a situation where subnationalgovernments clashand communal
violence, where there is no government. Our definitionof civil
conflict would include both of these phenomena. Small and Singer
also limit the casesthey examine to wars where over 1,000 people
died in one year, while we also examine caseswhere 1,000 people
died during the total span of the conflict.
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Hypotheses on the Causes of Contemporary Deadly Conflict 3
British, French, Ottomans, and Portuguese withdrew from their
empiresearlier in this century. Chaos likewise followed the
collapse of ancientempires from China to Rome. The collapses of the
Soviet and Yugoslavempires resulted in massive bloodshed for
similar reasons: they spawnedfrail successor governments that ruled
hostile intermingled populations andoften faced destructive
interference by the former metropole.
A lack of regime legitimacy caused many civil wars both during
and afterthe cold war. Civil-war-causing crises of government
legitimacy have sprungfrom two main causes: the growth of restive
middle classes in authoritarianstates, and a lack of regime
accountability, which leads in turn to corrupt orincompetent state
policies. Such legitimacy crises cause violence both whena regime
tries to regain legitimacy and when a regime tries to
suppressdissenta dilemma we label "the reform trap."
To regain legitimacy, besieged authoritarian regimes may move
todemocratize. Democratization, however, can spawn conflict as old
elitesinflame and manipulate hatreds in an effort to gain or remain
in power.Democratization can also spawn conflict if majoritarian
democratic rulesare adopted that cast all power to tyrannical
majorities, driving oppressedminorities to rebel. Finally,
democratization can spawn war by givingpolitical space to hardened
secessionist groups that cannot be appeased bypower-sharing and
instead exploit democratic freedoms to organize forwar. Democracy
causes peace between mature democracies, butdemocratization is a
dangerous cause of war in multiethnic authoritarianstates; hence
crises of regime legitimacy that trigger democratization arealso
causes of war.
Alternately, regimes that eschew democratization and instead
attempt tosuppress dissent often "hunker down," relying on an
increasingly narrowcore to defend them. Hunkering down may enable a
regime to survive ashort-term challenge to its legitimacy, but it
can cause violent resistance byexcluded social groups and does
little to ease the original legitimacy crisis.
4. For works on the politics of empire, see Michael W. Doyle,
Empires (Ithaca: CornellUniversity Press, 1986); D. K. Fieldhouse,
The Colonial Empires: A Comparative Survey from theEighteenth
Century (New York: Dell, 1966); Richard L. Rudolph and David F.
Good, eds.,Nationalism and Empire: The Habsburg Monarchy and the
Soviet Union (New York: St. Martin's,1992); and Jack Snyder, Myths
of Empire: Domestic Politics and International Ambition
(Ithaca:Cornell University Press, 1991). For an examination of the
relationship between the collapseof empires and war, see Jonathan
Ladinsky, "After the Fall: The Collapse of Empires and theCauses of
War" (Ph.D. diss., MIT, forthcoming).
5. The "reform trap" is similar to the "King's dilemma" analyzed
definitively by SamuelHuntington. Samuel P. Huntington, Political
Order in Changing Societies (New Haven: YaleUniversity Press,
1968), 177-91. Huntington argues that a traditional leader seeking
tomodernize may inadvertently create instability that, in the end,
causes the collapse of thetraditional order.
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
4 SECURITY STUDIES 7, NO. 3
This second war-cause grew more prevalent with the cold war's
end, asthe collapse of communism and the triumph of free-market
liberalismdiscredited repressive and statist ideologies throughout
the world. Regimesthat embraced these ideologies lost legitimacy
and often undertookdestabilizing reforms.
A third leading cause of civil conflict is state weakness. The
spread ofmodern small arms among Third World populations has
weakenedgovernments' relative position vis-a-vis potential civil
opponents over thepast few decades. The end of superpower aid to
beleaguered governmentshas further weakened many authoritarian
regimes since 1989. Thisweakening, in turn, has made it easier for
groups to challenge governmentsby force.
A fourth cause is communal hegemonism the aspiration of
ethnic,religious, clan or class groups for hegemony over other
groups. Violenceoften results when hegemonistic groups seek to
impose their way of life onothers, particularly on peoples that
have a well-developed group identity oftheir own. Peace is
strongest when groups adopt a live-and-let-live stancetoward
others.
The war-causing effects of communal hegemonism are catalyzed by
thefirst three causes. Submerged hegemonistic groups are freer to
go on arampage, and are more likely to provoke defensive violence
by others, whenan empire collapses, regime legitimacy declines, or
central power weakens.
Fifth, revolutionary political ideas have caused civil war by
leadinggroups to adopt extreme goals and tactics that precluded
peacefulcompromise with others. Marxist-Leninist ideas have fueled
civil warsworldwide since 1917, by leading movements of the
disenfranchised toembrace extreme communist political programs.
Muslim revolutionarymovements have likewise triggered civil
violence in the Arab world andSouth Asia in recent years. By
delegitimizing Marxist-Leninist politicalideas, the Soviet collapse
vastly reduced the prevalence of this importantcause of war.
Sixth, aristocratic intransigencethe refusal of elites in some
steeplystratified states to share power and wealthhas triggered
several recentcivil wars (for example, in Nicaragua in the 1970s
and in El Salvador and
6. We use the term "communal" in this essay to encompass ethnic,
religious, tribal, andlinguistic groups. A communal group is a
group of people bound together by a belief ofcommon heritage and
group distinctiveness, often reinforced by religion, perceived
kinshipties, language, and history. Examples of communal groups are
Turks (a common language,perceptions of a shared history) and Jews
(belief in common ancestry reinforced by acommon religion and
history). Large tribal groups and clans that perceive themselves
ashaving a common identity fall under this category as well.
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Hypotheses on the Causes of Contemporary Deadly Conflict 5
Guatemala in the 1980s). This danger has diminished with the
cold war'send, as democratic norms have spread to conservative
elites and as theseelites have lost unqualified U.S. backing. This
loss has forced them tomoderate their behavior.
The fourth, fifth and sixth causes are all examples of a more
generalphenomenonextremism in political ends and means. Peace
amonggroups, classes and movements is most threatened when they
insist ondominance and adopt take-no-prisoners tactics.
Seventh, the superpower competition for influence in Third World
stateswas a major cause of civil conflict during the cold war. A
number of ThirdWorld states (Afghanistan, Angola, Cambodia,
Nicaragua, El Salvador,Guatemala, and Ethiopia, among others)
became cold-war battlegroundswhen one or both superpowers backed
proxies or directly intervened tosupport client groups. The cold
war's end stopped this competition andslowed or ended the civil
violence it spawned.
All seven causes operated both during and after the cold war,
but somewere more prevalent during the cold war while some have
been moreprevalent afterward. Specifically, the first three causes
listed above (collapseof empire, regime illegitimacy, and state
weakness) grew more abundantafter 1989, the fourth (communal
hegemonism) stood roughly constantacross both periods, and the last
three (revolutionary ideology, aristocraticintransigence, and
superpower competition) abated sharply after 1989. Theabatement of
these last three causes largely explains the net decline in
civilviolence since 1989.
The most important of these causes of war since 1989 were the
loss ofgovernment legitimacy and communal hegemonism. Of
thirty-sevencountries that have suffered conflicts since the fall
of the Berlin Wall,conflict between a hegemonistic ethnic group and
other groups helped fueltwenty-five conflicts. The loss of regime
legitimacy was a major cause ofnineteen wars. In fourteen conflicts
the governments were too weak tosuppress or appease even minor
rebellions, a weakness that led to a largerconflagration. The
collapse of empire also helped precipitate thirteen
recentconflicts, including five of the eleven "new" conflicts which
were not activebefore 1989. Revolutionary ideologies were a major
cause of elevenconflicts (with communism being the culprit ideology
in five of theseeleven). Superpower competition was a major causes
of eight conflicts;aristocratic intransigence a major cause of
four.
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Table 1
LOCATIONS OF RECENT CIVIL VIOLENCE
Conflictlocation
AfghanistanAlgeria
Angola
Azerbaijan/ArmeniaBurma
Burundi
CambodiaChad
Colombia
El Salvador
Ethiopia
Georgia
GuatemalaHaiti
1989
Yes
Yes
No*
Yes
?*
Yes
Yes
Yes
No*
1990
Yes
Yes
?*
No*
Yes
Yes
Yes
Yes
No*
1991
Yes
Yes
?*
No*
No*
No*
Yes
No*
?*
Conflict active in year
1992 1993 1994
Yes
Yes
Yes
Yes
No*
Yes
Yes
?*
Yes
Yes
Yes
Yes
No*
Yes
No*
Yes
Yes
No*
?*
Yes
Yes
Yes
Yes
Yes
Yes
No*
No*
?*
No*
?*
1995
Yes
Yes
No*
P*
Yes
No*
No*
No*
1996
Yes
Yes
No*
Yes
No*
Yes**
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
India***
IndonesiaIraq
Lebanon
Liberia
Mozambique
Pakistan
Peru
Philippines
Romania
Russia (Chechnya)Rwanda
Sierra Leone
Somalia
SouthAfrica****Sri Lanka
Yes(Pun-jab)No*
No*
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No*
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No*
Yes
?*
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No*Yes**
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes(Kash-mir)
Yes**
?*
Yes
No*
Yes
Yes
Yes
Yes
No*
Yes**
?*
?*
No*
No*
Yes
?*
?*
Yes
Yes
No*
Yes**
?*
Yes
No*
Yes
Yes
?*
?*
Yes
?*
No*
Yes'
?*
No*
No*
Yes
Yes
?*
?*
Yes
(see key on pp. 8-9)Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Conflictlocation
Sudan
TajikistanTurkey
Uganda
Yemen
Yugoslavia(and successorstates)Zaire
Total Wars
Total conflictsactive*****
1989
Yes
No*
Yes
17
23
Table 1 (continued)
1990
Yes
?*
?*
17
23
1991
Yes
?*
Yes
Yes
14
23
Conflict active in year
1992
Yes
Yes
Yes
Yes
20
23
1993
?*
Yes
Yes
Yes
17
24
1994
?*
No*
Yes
Yes
Yes
12
25
1995
Yes
No*
Yes
Yes
11
23
1996
Yes**
No*
Yes
?*
Yes
11
22
Countries in italics are home to recurring or constant conflicts
(that is countries that suffered internal wars on and offor
continually in the 1970s and 1980s.) Of the thirty-sevn conflicts
listed above, twenty-six were recurring or constantconflicts.
* Conflict was active in the year in question but probably did
not reach 1,000 deaths a year. A "no" in the boxindicates that data
are available, while a "?" indicates that precise figures are not
available. Conflicts with a "?"and a "no" are not included in the
"Total Wars" figure, but are included in the "Total Conflicts
Active" box.
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
** Conflict deaths probably reached 1,000 deaths a year, but
precise figures are unavailable. In some cases, suchas the Sudan,
it is highly likely that deaths from civil violence, exceeded the
1,000 death figure considerably.
*** India is home to both recurring conflicts (Punjab) and
nonrecurring ones (Assam).**** The South African conflict changed
from an ANC-government struggle to one between Zulu groups and
ANC partisans.***** An "active" conflict includes conflicts that
reached the 1,000 deaths a year criteria and those that did not
reach this level but were not completely resolved.
We list each conflict by location even though several locations
(such as India) are home to multiple conflicts that oftenhave
highly different causes. When a conflict occurred in an area under
different sovereignties (for example the Croat-Serb conflict
occurred in both "Yugoslavia" and "Croatia") we list it according
to where the conflict began. Thus,conflicts in the former Soviet
Union are generally listed under their successor states, as the
fighting did not break outuntil after the Soviet Union collapse. On
the other hand, we count the former Yugoslavia as one location
because thefighting began there when the union was intact. Thus,
other descriptions of civil violence might list more conflicts
orfewer, depending on how they code various conflicts. A change in
coding, however, would not significantly change theconclusions of
this paper.
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
10 SECURITY STUDIES 7, NO. 3
The next section defines civil violence and describes post-cold
wartrends in its incidence and location. The subsequent section
describes theseven prime causes of recent (that is, since 1989)
civil violence. Ourconclusion argues that civil war seems likely to
diminish further in thedecades ahead, but one major cause of civil
wardemocratization inmultiethnic stateswill raise serious risks
down the road.
There is no widely used source that codes the causes of the
civil wars wediscuss. Lacking such a source, our judgments on these
wars' causes areauthors' estimates based on our survey of
mainstream press accounts andother secondary sources. Others would
code many of these casesdifferently, but we think our coding fairly
reflects press and othersecondary accounts.
RECENT TRENDS IN THE INCIDENCE OF CIVIL CONFLICT
J-ylHREE SEPARATE measures of civil conflict indicate that it
brieflyJ . increased after the cold war ended, but it then quickly
faded back to
levels at or below those of the late cold war. Although each
measure usesdifferent criteria, all three show the same overall
trend. The "new worlddisorder" was short-lived, and the world today
is becoming more peaceful.
The number of states experiencing civil conflict. It offers our
count of thenumber of states with major civil conflicts under way.8
Table 1 reveals thatthe number of states with ongoing conflicts
increased right after the end ofthe cold war but then declined
sharply, falling to levels well below late cold-war levels by 1995.
In 1989 seventeen countries suffered civil conflictsinvolving more
than 1,000 deaths. The number of states with active civilconflicts
peaked in 1992, when twenty countries had major civil
conflictsunder way. By 1996, however, the total number of countries
experiencing
7. We examine all the instances of widespread civil violence
active after the end of thecold war. For the purposes of this
article, the fall of the Berlin Wall in November 1989 marksthe end
of the cold war. Any widespread civil violence that broke out after
that point isexamined in this study.
8. Our count was compiled primarily from descriptions in The
Europa World Book series,the Economist, the New York Times, Jane's
Intelligence Review, articles in academic journals such asProblems
of Communism, Conflict Studies, and Current History, and selected
works notedspecifically in the text. We also drew on Peter
Wallensteen and Margareta Sollenberg, "TheEnd of International War?
Armed Conflict 1989-95," Journal of Peace Research 33, no. 3(August
1996): 353-70; Patrick Brogan, The Fighting Never Stopped: A
Comprehensive Guide toWorld Conflicts Since 1945 (New York:
Vintage, 1990); Ruth Leger Sivard, World Miltary andSocial
Expenditures 1996 (Leesburg, VA: WMSE Publicaitons, 1996); SIPRI
Yearbook 1996 (NewYork: Oxford University Press, 1996); and R. J.
Rummel, Death by Government (NewBrunswick: Transaction, 1996).
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Hypotheses on the Causes of Contemporary Deadly Conflict 11
such major conflicts had fallen to eleven. Civil warfare had
hardlydisappeared, but it was down from late cold war period.
The number of separate civil conflicts. A count of the number of
dyadic civilconflicts (that is, a count of each separate feud) by
Peter Wallensteen andMargareta Sollenberg reveals the same pattern.
They report that the numberof civil conflicts under way rose from
forty-four in 1989 to forty-six in1990 and fifty in 1991, and then
peaked at fifty-four in 1992. (See Table 4below.) The number of
conflicts then fell to forty-six in 1993, to forty-twoin 1994, and
to thirty-fourwell below the 1989 countin 1995.9
Table 2WORLDWIDE REFUGEES, 1980-9510
Year
1980198119821983198419851986198719881989199019911992199319941995
Number ofrefugees
(in millions)5.78.29.8
10.410.910.511.612.413.314.814.917.217.018.216.414.4
9. Wallensteen and Sollenberg, "The End of International War?"
table 2 on p. 354. Theirwar-count is higher than ours because they
count war-dyads instead of states with wars andbecause they use a
more inclusive definition of civil conflict than we do, including
someminor wars involving fewer than 1,000 total deaths.
10. United Nations High Commissioner for Refugees, The State of
the World's 'Refugees: InSearch of Solutions (New York: Oxford
University Press, 1995), 248.
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Table 3
CAUSES OF RECENT CIVIL CONFLICTS
ConflictLocation Major Parties Involved Causes of the
Conflict
Afghanistan
Algeria
Angola
Azerbaijan/Armenia
1. Mujahedin v. Soviet-backed government
2. Mujahedin v. otherMujahedin factionsIslamist (FIS, GIA)
v.government (FLN)UNITA v. government(MPLA)Armenians v. Azeris
Collapseof empire(post-1945)
Yes
Yes
Lack ofregimelegitimacy
Yes
Yes
Stateweakness
Yes
Yes
Super-Communal Aristocratic Revolutionary powerhegemonism
intransigence ideology proxy*
Yes Yes Yes(v. Sovietsonly)
Yes
Yes Yes
Yes
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Burma 1. Government v. Yesdemocratic opposition
2. Government v.Karen ethnic group
3. Government v.Mong Tai Army
4. Government v. otherethnic groups
Burundi Hutu v. Tutsi Yes
Cambodia Khmer Rouge v. rivalorganizations (KPNLF,FUNCINEC)
Chad 1. Government versusmilitary faction andMovement for
theNational Salvation ofChad
2. Clan infighting
Yes(crushingdemocracyforces)
Yes(Karen, MongTai, otherethnicstruggles)
Yes
Yes Yes
Yes
Yes Yes
Yes(clan fighting)
(see key on p. 21)
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Table 3 (continued)
ConflictLocation Major Parties Involved Causes of the
Conflict
Colombia 1. Government v. M-19,FARC, EPLO, ELN, andsplinter
groups
2. Guerrilla groupsfighting one another
El Salvador FMLN v. government
Ethiopia 1. Government versus YesEritrean People'sLiberation
Movement
2. Government versusTigray People'sLiberation Front
3. Government versusother communalfactions
Collapseof empire Lack of(post- regime State1945) legitimacy
weakness
Yes Yes
Communalhegemonism
Super-Aristocratic Revolutionary powerintransigence ideology
proxy*
Yes
Yes
Yes Yes
Yes
Yes
Yes
Yes
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Georgia
Guatemala
Haiti
India
Indonesia
Iraq
1. Abkhaz v. Georgians Yes Yes2. Ossetians v.Georgians
Government v. leftistguerrillas
Military government vs.Aristide supporters
1. Government v. Yes YesKashmiri separatists (Kashmir)
(Kashmir)2. Hindu v. Muslim3. Government v.Assamese separatists(and
Bengalis v.Assamese)
1. Aceh separatists v.government forces
1. Kurds (PUK, KDP) v. Yes**Sunni Arab government2. Shi'a v.
Sunni Arabgovernment
Yes
Yes(Assam,Hindu-Muslimfighting)
Yes
Yes
Yes
Yes
Yes Yes
Yes
(see key on p. 21)Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Table 3 (continued)
ConflictLocation Major Parties Involved Causes of the
Conflict
Collapseof empire Lack of Super-(post- regime State Communal
Aristocratic Revolutionary power1945) legitimacy weakness
hegemonism intransigence ideology proxy*
Lebanon Sunnis, Shi'a, Druze, Yes Yes YesMaronite
Christians,others against oneanother and themselves
Liberia Krahn, Gio, Mano, and Yes Yes Yesother tribes and
theirassociated militias
Mozambique RENAMO vs Yes Yesgovernment (FRELIMO)
Pakistan 1. Violence among Yes Yespolitical parties, often
(politicalethnically linked violence2. Ann-mobajir violence
onty)
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Peru
Philippines
Romania
Russia
Rwanda
Sierra Leone
Government versusShining Path, TupacAmaru
RevolutionaryMovement1. Government v. NPA2. Government v.Muslim
groups (MNLF,MILF)
Armed forces/securityservices v. NationalSalvation Front
andpopular backers
Russia v. Chechen Yesseparatists
1. Hutu v. Tutsi Yes2. interha/mMuxn v.Hutu moderates
Government versusRevolutionary UnitedFront forces
Yes
Yes
Yes
Yes(Hutuv.Hutu)Yes Yes
Yes(v. Musigroups)
Yes
Yes(Hutuv.Tutsi)
Yes
Yes
Yes Yes
(see key on p. 21)Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Table 3 (continued)
ConflictLocation Major Parties Involved
Collapseof empire Lack of(post- regime1945) legitimacy
Causes of the Conflict
Super-State Communal Aristocratic Revolutionary powerweakness
hegemonism intransigence ideology proxy*
Somalia Clan-based politicalfighting (USC factions,SSDF,
SPM)
South Africa 1. Apartheidgovernment versusANC, other
anti-apartheid forces
2. ANC-Inkatha fighting
Sri Lanka 1. Government v. Tamilinsurgents (LTTE, etc.)2.
Government v.Sinhalese radicals (JVP)
Yes Yes Yes
Yes
Yes
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Sudan Islamist, Arab Yes Yes Yesgovernment v. Christianand
animist AfricanSPLA
Tajikistan Tajik "old guard" and Yes Yes YesUzbeks
v.democratic/religiouscoalition (United TajikOpposition)
Turkey Govt. v. Kurds (PKK) YesUganda 1. Government v. Yes
Yes
Uganda People's (Holy SpiritDemocratic Army Movement)2.
Government v. HolySpirit Movement
Yemen Former North Yemen Yesv. forces of formerSouth Yemen
(see key on p. 21)
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Table 3 (continued)
ConflictLocation Major Parties Involved Causes of the
Conflict
Collapseof empire Lack of Super-(post- regime State Communal
Aristocratic Revolutionary power1945) legitimacy weakness
hegemonism intransigence ideology proxy*
Yugoslavia(andsuccessor
states)
Zaire
Total number
1. Croat governmentand militias v. Serbforces2. Serb
paramilitaryforces v. Muslimmilitias and govt. forces3. Muslim
governmentand militias v. Croatgovernment and militias
Government v. Kabila'smovement.
37
Yes
13
Yes
19
Yes
14
Yes
Yes
25 11
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
* We consider a conflict to be caused by a superpower proxy
struggle if the superpower intervention caused, widened, or
sustained the civilwar in question.
** The Iraqi state in general was strong, but it hovered on the
brink of collapse in 1991 after the Gulf War. This weakness
encouragedseveral repressed minorities to rise up.
The causes identified above were major and active factors at the
time of the outbreak of the latest round of fighting. As such, they
played aimportant role in the concerns or motivations of the
combatants. Determining the cause of a conflict is difficult and
involves subjective judgments.Moreover, several of the causes we
identify function concurrently, making it hard to decide which one
was the most important. To avoid "doublecounting," we do not list
"lack of regime legitimacy" as a cause when the lack occurs because
an empire has collapsed, leaving illegitimate successorstates, or
because hegemonistic communal groups or intransigent aristocrats
hold power, alienating other groups. If these excluded cases were
doublecounted as cases of "lack of regime legitimacy," that
category would be much larger.
The list of major parties involved includes only the primary
movements or groups involved in the fighting. Many, indeed perhaps
most, conflictsinvolved a staggering array of small militias and
factionsan array we often agglomerate into broad descriptive
communal or political labels. Separategroups are noted if the
country experienced multiple, largely unrelated conflicts or if the
groups in question had highly different motivations or areeasily
distinguishable. Thus Burma, where the government v. democratic
opposition conflict is quite different in nature from the
government'sstruggle against the Karen people, has multiple
listings. Colombia's many guerrilla groups, while quite different
in their particular agendas, all felt theregime was illegitimate,
considered themselves revolutionary, and took advantage of state
weakness to carry out their struggle. Thus we list them asone entry
while noting the major groups. When a conflict cause applies to
only one of the parties involved in the fighting, it is so noted in
the table.The purpose of this list is to describe the current state
of violence and to help the reader understand our coding of certain
conflicts, not to provide acomprehensive account of the identity of
the parties in each conflict.-)-
. Although the patterns in this table might be examined further
by more sophisticated quantitative techniques, we have chosen not
to do so given theambiguous nature of the data and the high degree
of uncertainty regarding many of the conditions necessary for the
various causes to function. SeeGary King, Robert O. Keohane, and
Sidney Verba, Designing Social Inquiry: Scientific Inference in
Qualitative 'Research (Princeton: Princeton University Press,1994),
44, for an argument that quantitative indexes that do not relate
closely to the events in question can actually increase measurement
and causalinference problems.
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
22 SECURITY STUDIES 7, NO. 3
Total casualties, measured by counting total refugees. Counts of
civil warsmeasure global civil violence imperfectly because great
and small warsweigh the same in the measure. Estimates of total
worldwide civil warcasualties would be a better measure, but
casualty figures are unavailable orunreliable for most civil wars.
Hence any measure of total worldwidecasualties is also
unreliable.
Global casualties can be measured indirectly, however, by
counting theglobal war refugee population. The refugee population
is a useful surrogatemeasure of war casualtiespeople flee in rough
proportion to the violencethey sufferand like our "number of states
with wars" measure it indicatesthat the "new world disorder" is a
myth. As Table 2 reveals, the globalrefugee population rose
slightly after the end of the cold war, from 14.8million in 1989 to
a peak of 18.2 million in 1993.11 Then the refugeepopulation fell
back to 14.4 million in 1995, that is, roughly to late cold
warlevels. Thus this refugee measure, like our "number of states
with wars"measure and Wallensteen and Sollenberg's "number of wars"
measure,indicates that the "new world disorder" was a spike
phenomenon of theearly 1990s that quickly faded. Specifically, it
suggests that violence in themid-1990s was above mid-1980s levels
but slightly below the level of 1989.
Are the conflicts of the mid-1990s old or new? Of the
thirty-seven warsduring the period 198996 listed in Table 1,
twenty-six are "recurring orconstant," meaning that they easily
span the cold war and postcold warperiods. The remaining eleven are
"new" conflicts, meaning that they brokeout after the cold war
ended and their causes are not rooted deeply in pre-1989 events in
their countries. Five of these eleven new conflicts erupted inthe
former Soviet and Yugoslav empires and reflect the war-causing
effectsof imperial collapse.
11. Refugee flows are a crude measure of civil violence. One
important measure ofrefugeesinternally displaced refugeesis not
listed here though a more complete accountof refugee totals would
include these individuals. Civil wars often generate massive
refugeeflows within a country's borders, as individuals flee areas
of fighting for relatively saferregions. Historic data on such
flows, however, are incomplete and probably would bemisleading for
comparison purposes, as flows in wealthier states that receive more
mediaattention are more likely to be recorded. Moreover, refugees
often remain in the country ofrefuge even after a civil war in
their country of origin ends. Furthermore, many refugees fleefor
economic reasons, not because of civil violence. Refugee flow data
also may be biaseddue to changes in the policies of receiving
states, which may take fewer refugees even thoughthe number of
people wanting to flee remains unchanged. In general, however,
there is ahigh correlation between internal wars, particularly
ethnic conflicts, and refugee flows. SeeMyron Weiner, "Bad
Neighbors, Bad Neighborhoods: An Inquiry into the Causes ofRefugee
Flows," International Security 21, no. 1 (summer 1996): 5-42.
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Hypotheses on the Causes of Contemporary Deadly Conflict 2 3
Table 4
INTERSTATE AND INTRASTATE ARMED CONFLICT, 1989-95
Type of Conflict
IntrastateIntrastate withforeign intervention
InterstateAll armed conflict
1989
431
3
47
1990
44
2
3
49
1991
491
1
51
1992
522
1
55
1993
42
4
0
46
1994
42
0
0
42
199
34
0
1
35
Source. Wallensteen and Sollenberg, "The End of International
War?" 354.
Where are recent civil wars occurring? Africa and Asia have had
the most(thirteen states in Africa and eleven in Asia have
experienced civil warssince 1989). Trailing are the Middle East
(five states), the Westernhemisphere (five states), and Europe
(three states).
What proportion of total warfare do these civil conflicts
represent? Inrecent years, civil war has replaced international war
as the dominant formof war and has nearly replaced it as the only
form of war. Wallensteen andSollenberg report that purely
intrastate wars outnumbered purely interstatewars worldwide by 43-3
in 1989, 44-3 in 1990, 49-1 in 1991, 52-1 in 1992,42-0 in 1993 and
again in 1994, and 34-1 in 1995.12 Moreover, most ofthese few
interstate wars were small affairs: the Persian Gulf war of 199091
has been the only major old-fashioned interstate war since
1989.
12 .Wallensteen and Sollenberg, "The End of International War?"
354. Wallensteen andSollenberg also classify several wars as
"intrastate with foreign intervention" and count themas follows:
one in 1989, two in 1990, one in 1991, two in 1992, four in 1993,
and none in1994 and 1995 (ibid.). For other works noting the
importance of internal conflict since theend of the cold war, see
Ted Robert Gurr, "People Against States: Ethnopolitical Conflictand
the Changing World System," International Studies Quarterly 38, no.
3 (September 1994):347-77; Stephen R. David, "Internal War: Causes
and Cures," World Politics 49 0uly 1997):552-76; Ted Robert Gurr,
Minorities at Risk: A. Global View of Ethnopolitical
Conflicts(Washington: U.S. Institute of Peace Press, 1993); and
S'laughter Among Neighbors: The PoliticalOrigins of Communal
Violence, Human Rights Watch (New Haven: Yale University Press,
1995).Good discussions of the interplay between international
conflict and internal conflict can befound in Michael E. Brown,
ed., Ethnic Conflict and International Security (Princeton:
PrincetonUniversity Press, 1993); Myron Weiner, The Global
Migration Crisis: Challenge to States and toHuman Rights (New York:
HarperCollins, 1995); Ted Robert Gurr and Barbara Harff,
EthnicConflict in World Politics (Boulder: Westview, 1994); and
Michael E. Brown, ed., TheInternational Dimensions of Internal
Conflict (Cambridge: MIT Press, 1996).
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
24 SECURITY STUDIES 7, NO. 3
This pattern continues a striking change in the nature of
warfare thatbegan in the 1960s, a change revealed in data collected
by Frank WhelonWayman, J. David Singer, and Meredith Sarkees and
summarized in Table5. It reveals a sharp drop in the number of
international wars beginning inthe 1980s and a sharp rise in civil
warfare beginning in the 1960s, causing amarked rise in the
proportion of all warfare worldwide that is civil innature. Some 60
percent of the 171 wars of the nineteenth century
wereinternational. Some 51 percent of the 115 wars of 1900-60
wereinternational. The percent of wars that were international then
plummets to36 percent in the 1960s, 26 percent in the 1970s, and 17
percent in the1980s. Only 10 percent of the conflicts in the 1990s
were international. Thewar problem is now largely synonymous with
the civil war problem.
CAUSES OF CONTEMPORARY CIVIL CONFLICT
CIVIL VIOLENCE has many causes, but several stand out in
importance.This section examines seven common and potent causes
that togetherexplain the bulk of civil conflict since 1989. These
seven causes are notwholly exclusive or unrelated. Several overlap
or cause each other in waysnoted below.
13 .The count of recent civil wars by Wayman, Singer, and
Sarkees is lower than oursbecause they define civil war more
restrictively than we do. See Frank Whelon Wayman, J.David Singer,
and Meredith Sarkees, "Intra-State, and Extra-Systemic Wars,
1816-1995"(paper presented at the annual meeting of the
International Studies Association, San Diego,Calif., April 1996),
Table 1, 10.
14. Scholarly work on the causes of civil conflict is vast.
Recent works include RoyLicklider, "The Consequences of Negotiated
Settlements in Civil Wars, 1945-1993,"American Political Science
Review 89, no. 3 (September 1995): 681-90; Chaim Kaufmann,"Possible
and Impossible Solutions to Ethnic Civil Wars," International
Security 20, no. 4(spring 1996): 136-75; Stuart J. Kaufman,
"Spiraling to Ethnic War: Elites, Masses, andMoscow in Moldova's
Civil War," International Security 21, no. 2 (fall 1996): 108-38;
David A.Lake and Donald Rothchild, "Containing Fear: The Origins
and Management of EthnicConflict," International Security 21, no. 2
(fall 1996): 41-75; Barry R. Posen, "The SecurityDilemma and Ethnic
Conflict," in Brown, Ethnic Conflict and International Security,
103-24; andStephen Van Evera, "Hypotheses on Nationalism and War,"
International Security 18, no. 4(spring 1994): 5-39. Classic works
of value on internal war include Donald Horowitz, EthnicGroups in
Conflict (Berkeley: University of California Press, 1985); Myron
Weiner, Sons of theSoil (Princeton: Princeton University Press,
1978); and Robert H. Bates, "Modernization,Ethnic Competition, and
the Rationality of Politics in Contemporary Africa" in State
VersusEthnic Claims: African Policy Dilemmas, ed. D. Rothchild and
V. Olorunsola (BoulderWestview, 1985).
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Hypotheses on the Causes of Contemporary Deadly Conflict 25
COLLAPSE OF EMPIRE
Much post-cold war violence has occurred in the successor states
offormer empires, especially the Soviet and Yugoslav empires.15 The
reasonlies in the powerful war-causing effects of imperial
collapse. Before the endof the cold war, the collapse of the
British, French, Portuguese, and othercolonial empires caused many
conflicts in Asia and Africa, including someconflicts that
continued into the 1980s and 1990s. Imperial collapses alsoaccount
for a large share of recently erupting deadly conflicts.
Specifically,wars caused by the 1991 collapse of the Soviet and
Yugoslav empiresexplain much of the "spike" in conflict observed in
1992 and 1993.16
The collapse of an empire can cause conflict in five ways. (1)
Thegovernments that emerge in the successor states often lack
legitimacy,hence suffering the neuroses of illegitimate regimes.
(2) The successorstates' governments, even if legitimate, are often
too weak to deter citizensfrom violence or to reassure them that
they need not use violence in self-defense. (3) Successor states
often have artificial borders that are unrelatedto local demography
and are unsetded by formal agreement; this generatesborder
quarrels. (4) The populations of successor states are
oftencomprised of hostile groups who intermingled during imperial
times. Theirproximity breeds mutual fear, hostility, and violence.
(5) The rights andduties of major powers in the zone of imperial
collapse are often undefined.As a result the former metropole and
other outside powers often collide asthey contend for power in the
zone of imperial retraction. The metropoleinterferes to recover
lost influence; outside powers interfere to preventdisorder or to
expand their influence. These causes are detailed below.
Illegitimate governments. Many of the successor governments that
emergedafter the fall of the Soviet Union and the collapse of
Yugoslavia had littlelegitimacy. This lack of legitimacy encouraged
a violent scramble for poweramong leaders and interest groups and
encouraged minorities to resistincorporation into die successor
state.
During the days of empire, local leaders depended on ties to
themetropole, not local communities, for their power and authority.
When theempires collapsed, these leaders suddenly found themselves
governingwithout institutions or a popular mandate, under challenge
from rival elites.
15. For a complete treatment of the relationship between the
collapse of empire and theoutbreak of war, see Ladinsky, "After the
Fall."
16. Thus Kim Holmes notes that "we have witnessed the collapse
of the world's greatestland empire. As with the demise of other
great empires in historywhether they be Spanish,French, Turkish,
Austro-Hungarian, Germany, or Britishwar is the fruit of
disorder"(Holmes, "The New World Disorder").
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
26 SECURITY STUDIES 7, NO. 3
In Tajikistan, for example, Rakhmon Nabiyev and associated
apparatchiksessentially stumbled into power after his predecessor
left due to popularprotests following the failed coup in Moscow. To
stay in power, Nabiyevplayed up nationalism, armed selected
followers, suppressed opposition,and otherwise strove to find
substitutes for his regime's lack of legitimacy.These substitutes
failed to satisfy many residents. Democrats, Islamists, andrival
communal groups all rejected Nabiyev's bona fides and took up
armsto oust his government.
This lack of legitimacy invites rebellion by minorities in the
new states.Minorities that accepted a subordinate status in a
large, multiethnic empireoften reject a minority status when an
empire's collapse empowers anethnic rival. In Georgia, for example,
Georgian nationalists led by ZviadGamsakhurdia took power as the
Soviet Union collapsed, with widespreadsupport among ethnic
Georgians. Two large minority communities inGeorgia, however, that
had apparently accepted their minority status in theSoviet Unionthe
Abkhaz and the Ossetianstook up arms to preventtheir incorporation
into the Georgian-dominated state. Similarly,Moldova's nationalist
movement alarmed residents in the Transdniesteriaregion, which is
60 percent ethnic Russian and Ukrainian. TheseTransdniestrian
Russians and Ukrainians proclaimed the TransdniesterianMoldovan
Soviet Socialist Republic and tried to remain attached to theSoviet
Union. Only the dispatch of Soviet troops prevented
massiveviolence.20 The Chechens in Russia and the Armenians in
Azerbaijan alsoresisted incorporation into a new state dominated by
what they fearedwould be their community's persecutors. In all
these cases, minorities in amultiethnic empire sought their own
state after the empire collapsed in partbecause they rejected the
legitimacy of the successor government. (For a
17. Due to this illegitimacy problem, some parts of the Soviet
empire resisted the empire'scollapse.
18. See Barnett Rubin, "The Fragmentation of Tajikistan,"
Survival 35, no. 4 (winter 1993-1994): 71-91; and "Tajikistan:
Islam wins," Economist, 21 October 1992, 32.
19. For an analysis of minority tension in the Georgian quest
for independence, seeDarrell Slider, "The Politics of Georgia's
Independence," Problems of Communism 40, no. 6(November 1991):
63-79.
20. V. Solnar, "Hatred and fear on both banks of the Dniester,"
New Times International,no. 14 (April 1992): 8-9; and William
Crowther, "Moldova after Independence," CurrentHistory 93, no. 585
(October 1994): 342-47. The Gagauzi, a Turkic-speaking,
OrthodoxChristian people from Bulgaria, also resisted incorporation
into the new state and proclaimedtheir independence.
21. See Mark Saroyan, "The 'Karabakh Syndrome' and Azerbaijani
Politics," Problems ofCommunism 39, no. 5 (September 1990): 14-29,
for information on the origins of the conflictin Azerbaijan. For
background on the conflict in Chechnya, see Christopher Panico,
Conflictsin the Caucasus: Russia's War in Chechnya, Conflict
Studies no. 281 (Research Institute for theStudy of Conflict and
Terrorism, 1995).
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Hypotheses on the Causes of Contemporary Deadly Conflict 27
more detailed discussion of the problems that illegitimate
governmentsface, see "Loss of Legitimacy" below.)
Weak governments. Even legitimate successor governments often
emergefrom collapsed empires in a frail condition. State security
services and themilitary must often be purged, reformed, or created
out of whole cloth.Frequently, the successor state relied on
subsidies from the metropole andnow must do without. Thus,
successor regimes often lack the resources todeter, suppress, or
buy off dissent. Tajikistan, for example, becameindependent without
a clear national identity, a strong economic base, ornational
security forces.22 It quickly collapsed into civil war.
Successor governments also may have little control over
governmentinstitutions. In Georgia, the nationalist successor
governments often lackedcontrol of the military and other
institutions. Tengis Kitovani, whocontrolled the Georgian National
Guard, deployed troops to the capital ofAbkhazia and shelled the
capital of South Ossetia, despite GeorgianPresident Shevardnadze's
desire for peace talks. Similarly, paramilitaryleaders in Georgia
often controlled the supply of food, fuel, and othernecessities.
(We examine government weakness as a source of conflict ingreater
detail in the section "Weak States" below.)
Artificial borders. Successor states often inherit artificial
borders thatcorrespond poorly to natural boundaries or to local
demography and havenot been settled by agreements with neighbors.
These borders often followadministrative boundaries that were
imposed by the metropole withoutregard for local feelings. As a
result these borders bisect national groupsand create
ethnic-minority enclaves. Thus, the European powerspartitioned
Africa at the 1878 Congress of Berlin with little regard for
theunity of African peoples, drawing lines that seldom followed
geographic orcommunal boundaries.2 Later, Stalin drew borders that
split the Turkic andMuslim peoples of the Soviet Empire into
different administrative units inorder to weaken their political
strength.
Cursed with such borders, several successor states to the Soviet
andYugoslav empires have fought bloody wars to resolve questions
raised by
22. See Rubin, "The Fragmentation of Tajikistan," 75-78.23. For
an assessment of these divisions on conflict in Georgia, see
"Georgia: Unholy
Trinity," New Statesman and Society 5, no. 219, 11 September
1992, 19-20; and "Georgia:Tearing Apart," Economist, 3 October
1992, 55.
24. At the conference, German chancellor Otto von Bismarck
"continually warned therepresentatives of the Great Powers that
their principal business was to reach a settlementamong themselves
and not to worry unduly about the happiness of lesser breeds
without thelaw" (Gordon A. Craig, Germany: 1866-1945 (New York:
Oxford University Press, 1978],112). For more on the European role
in the creation of borders in Africa, see ThomasPakenham, The
Scramble for Africa (New York: Avon, 1991).
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
28 SECURITY STUDIES 7, NO. 3
maldesigned boundaries. The Armenia-Azerbaijan conflict stemmed
fromArmenian demands that Nagorno-Karabakhan Armenian enclave
withinAzerbaijanbe transferred to Armenia. The Serb-Croat-Muslim
war of199195 likewise stemmed from arbitrary borders that bisected
the Serbianand Croat nations, leaving each with large diasporas
living as minoritiesoutside the national state. Each then fought to
recover its diaspora.
Intermingled populations. Empires foster national intermingling
that canplague the politics of their successor states. During the
imperial era,individuals can more easily move about within the
empire, causingintermingling. Moreover, some empires deliberately
intermingle theempire's national groups by inducing or compelling
cross-migration amonggroups. As a result, the empire's successor
states may have populationscomposed of mutually antagonistic
peoples. Stalin's forced marches ofmillions of subjects are the
most famous example of such enforcedintermingling, and they sowed
the seeds of current conflict. The recentconflict in Chechnya, for
example, stems from Stalin's 1944 deportation ofthe Chechens.
Moscow allowed these Chechens to return to Chechnya in1956, but on
returning they met a hostile welcome from new, largelyRussian
setders, who had been encouraged by Moscow to migrate there.This
settler-native tension fueled Chechen nationalism and
secessionism.
Intermingling causes conflict by shoving antagonistic groups
togetherand by producing an ethnic security dilemma between diem
(mat is, asituation where the security of two groups is mutually
incompatible, andeach group's efforts to secure itself reduce the
other's security). Hostileintermingled groups each must fear that
the other may turn on them at anopportune moment, leading each to
think in turn that it should strike at atime of its own advantage.
Such thinking played a major role in fuelingSerbia's attacks on the
Croatians and Bosnians in 199192 and inmotivating Armenia's war
against Azerbaijan.
Metropole interference. Former metropoles often intervene in
dieir formerempire, sometimes triggering new colonial wars. The
metropole may beanimated by perceptions of a security threat, by
claims to a sphere ofinfluence, or by the need to protect or
recover diaspora populations in theperiphery. Moscow today, for
example, claims the right to intervene in
25. "Azerbaijan," Europa World Book 1994 (London: Europa
Publications Limited, 1993),438-40.
26. For works noting the importance of the security dilemma, see
Posen, "The SecurityDilemma and Ethnic Conflict"; James Fearon,
"Rationalist Explanations for War,"International Organization 49,
no. 3 (summer 1995); Kaufmann, "Possible and ImpossibleSolutions to
Ethnic Civil Wars"; and Barbara F. Walter, "The Resolution of Civil
Wars: WhyNegotiations Fail" (Ph.D. diss., University of Chicago,
1994).
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Hypotheses on the Causes of Contemporary Deadly Conflict 29
parts of its former empire to protect Russian citizens. Many
Russiannationalists also assert that this area remains their sphere
of influence, andsome Russians fear the possibility of former
republics allying with outsidepowers. Such thinking led Russia to
intervene in conflicts in Moldova,Georgia, and Tajikistan. In
Moldova, Russian intervention probablyreduced civil violence, but
in Tajikistan it probably fed a bloody civil war.In Georgia its
effects were mixed, sometimes fueling violence byencouraging Abkhaz
separatism, yet intervening in the end to help enforcea cease-fire
after Schevardnadze agreed to join the Commonwealth ofIndependent
States.
External intervention. Outside powers intervene in collapsed
empires forthree reasons. First, these powers often see the
successor states of acollapsed empire as easy prey for their
imperial ambitions. After thecollapse of the Ottoman Empire, for
example, Britain quickly abandonedwartime promises to Middle
Eastern leaders and divided much of theOttoman lands with France.29
Similarly, the European powers often clashedover the Balkansmost
notably in 1914after the Balkan states escapedOttoman rule. In 1975
Indonesia seized East Timor after the Portuguesewithdrawal,
triggering a bloody neocolonial war.
Second, outside powers may intervene to avert threats that the
imperialcollapse creates. Thus, after Portugal abandoned its
empire, South Africafought bloody interventionary wars in Namibia,
Angola, and Mozambique,largely because it feared that the new
black-ruled governments wouldsupport South Africa's antiapartheid
resistance.
Finally, outside powers may intervene to protect embattled
co-ethnics inthe former empire. Thus, ethnic Tajiks in Afghanistan
have lately providedarms and a haven for Islamic rebels in
Tajikistan, while Uzbekistan hasbolstered the Tajik government for
fear that any alternative regime wouldoppress ethnic Uzbeks living
in Tajikistan. Mozambique supported blackliberation forces waging
war against the white minority regime in Rhodesiaduring the 1970s.
This spurred Rhodesia to create the Renamo insurgencythat spread
death and destruction in Mozambique into the 1990s.
27. Bruce D. Porter, "A Country Instead of a Cause: Russian
Foreign Policy in the Post-Soviet Era," in Order and Disorder after
the Cold War, ed. Brad Roberts (Cambridge: MIT Press,1995),
7-8.
28. Although accurate information is scarce, it appears that
local Russian forcesperhapswith encouragement from Moscowaided
Ossetian and Abkhaz separatists in their struggleagainst Georgian
forces.
29. A good account of this process is David Fromkin, A Peace to
End All Peace: The Fall ofthe Ottoman Empire and the Creation of
the Modern Middle East (New York: Avon, 1990).
30. See Rubin, "The Fragmentation of Tajikistan."
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
30 SECURITY STUDIES 7, NO. 3
Table 5
CIVIL AND INTERNATIONAL WARS, 1816-1995: WARS INITIATED PER
DECADE
Decade
1816-19
1820-29
1830-39
1840-49
1850-591860-69
1870-79
1880-89
1890-99
19th Ctotal
1900-1909
1910-19
1920-29
1930-39
1940-49
1950-59
1960-69
1970-79
1980-89
Inter-statewar
0
2
0
4
5
8
4
3
4
30
2
8
3
3
6
7
4
Extra-systemic(largelycolonial) Subtotalinter- for
inter-national nationalwar
2
6
57
9
510
12
16
72
4
6
6
2
5
6
3
2
0
war
(2)(8)(5)
(11)(14)(13)(14)(15)(20)
(102)
Intra-statewar
1
7
10
9
7
15
8
3
9
69
Grandtotal
3
15
15
20
21
28
22
18
29
171
Percentof allwars
that areinter-national
67
53
33
55
67
46
64
83
69
60
(10)(14)
(8)(10)
(8)(9)(9)(9)(4)
7
10
12
8
9t
10
16
25
19
17
24
20
18
17
19
25
34
23
59
58
40
56
47
47
36
26
17
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Hypotheses on the Causes of Contemporary Deadly Conflict 31
1990-95 2 0 (2) 19 21 10
20th Ctotal 49 34 (83) 135 218 38
Grandtotal 79 106 (185) 204 389 48
Source: Frank Whelon Wayman, J. David Singer, and Meredith
Sarkees, "Inter-State, Intra-State, and Extra-Systemic Wars
1816-1995" (paper presented at the annual meeting of
theInternational Studies Association, San Diego, Calif., April
1996), Table 1, p. 10.
LOSS OF LEGITIMACY
Much of today's violence occurs in states whose governments
havesomehow lost their legitimacy. ' This section discusses how
legitimacy islost and why such losses cause civil conflict.
Causes of Legitimacy LossFour factors are frequent sources of
lost legitimacy in states recently at war:the discrediting of the
Soviet model; poor economic performance; a lack ofregime
accountability (which gives rise in turn to incompetent and
corruptgovernance); and the rise of a restive new middle class that
seeks a greaterpolitical power. Each of these underlying factors,
alone or in conjunctionwith others, can discredit a regime and lead
to civil violence.
The discrediting of the Soviet model. When the Berlin Wall fell,
regimes thatrelied on the Soviet Union as a model for their
economies and politicssuffered a blow to their legitimacy. Indeed,
throughout the Third World thecollapse of Communism discredited
authoritarian regimes of all stripes,33for example, many regimes in
sub-Saharan Africa, and also those in Algeria
31. An illegitimate regime is one broadly believed by the public
to have lost its right torule because of its perceived failure to
provide for the common good.
32. Defeat in a war can also cause a government to lose its
legitimacy. The dearth ofinternational war in the post-cold war
period, however, has reduced the importance of thiscause of regime
legitimacy loss.
33. Peter W. Rodman, More Precious Than Peace (New York:
Scribner's, 1994), 532-33.Rodman notes that Syria's dictator Hafez
al-Asad was equated with Romanian dictatorNicolae Ceaucsescu and
East Germany's Erich Honecker. Similarly, the collapse of
EasternEuropean regimes strengthened African democratic forces and
disheartened Africa'sautocrats. See Copson, Africa's Wars in the
1990s, 167. For information on the impact of thespread of the
liberal democratic model, see Samuel P. Huntington, The Third
Wave:Democratization in the Late Twentieth Century (Norman:
University of Oklahoma Press, 1991).
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
32 SECURITY STUDIES 7, NO. 3
and the former South Yemen, which relied heavily on the
state-led34
economy and society.Poor economic performance. Regimes that
preside over stagnating or declining
economies may lose legitimacy. Poor economic performance
directlygenerates popular unhappiness. Remedies to poor performance
can also stirpublic opposition if they require bitter medicine,
such as trimminggovernment subsidies and payrolls. Such measures
further decrease aregime's legitimacy.
A lack of regime accountability. Regimes that are not
accountable to theirpopulations have less incentive to serve their
publics well, so they servethem poorly. Incompetence and corruption
are their hallmarks.3 Theyoften cannot be removed peacefully,
moreover, compelling their opponentsto resort to force.
The rise of a middle class and resulting demands to democratize.
Industrializationand the spread of literacy are potent causes of
democracy. The literatemiddle classes that these processes create
nearly always demand politicalpluralizationa reality reflected in
the close correlation worldwide betweenlevels of democratization
and the size of literate middle classes.37 Theemergence of educated
middle classes in authoritarian states is
thereforeregime-delegitimating: it brings on the scene middle-class
voices that willreject the authoritarian old political order.
Response to Legitimacy Loss
The loss of legitimacy is the underlying cause of conflict, but
the conflict'sproximate causes are the regime's responses to this
loss of legitimacy.Regimes losing their legitimacy often choose
between two responses:
34. George Joffe, "YemenThe Reasons for Conflict," Jane's
Intelligence Review (August1994): 369; John P. Entelis, "The Crisis
of Authoritarianism in North Africa: The Case ofAlgeria," Problems
of Communism 41, no. 3 (May 1992): 71-82.
35. Corruption often comes with a lack of accountability. In
Pakistan and the Philippines,for example, widespread corruption has
discredited governments and led to the growth ofopposition
movements. Philippine dictator Ferdinand Marcos stole with such
abandon thathe became one of Asia's most wealthy men before he was
ousted in 1986. In 1996Transparency International ranked Pakistan
the second most corrupt country in the worldafter Nigeria. New York
Times, 28 November 1996, C1.
36. In short, regime accountability often determines whether
"voice" is expressed inballots or bullets. The definitive
description of this tradeoff remains Albert O. Hirschman'sExit,
Voice, and Loyalty: Responses to Decline in Firms, Organisations,
and States (1970; Cambridge:Harvard University Press, 1981).
37. Robert A. Dahl, Polyarchy: Participation and Opposition (New
Haven: Yale UniversityPress, 1971), 62-80; Seymour Martin Lipset,
Political Man: The Social Bases of Politics, exp. ed.(Baltimore:
Johns Hopkins University Press, 1981), 27-63.
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Hypotheses on the Causes of Contemporary Deadly Conflict 33
democratizing in an attempt to gain broader support and
hunkering downin order to weather any disgruntlement. 8
By democratizing, regimes can increase government accountability
andpopular participation and thus dampen dissent. Democratization,
however,can also raise the risk of civil war. A large political
science literature hasshown that democracy causes peace among
mature democracies.Democracy, however, is a Janus-faced phenomenon:
democratization isalso a potent cause of civil conflict in
multiethnic authoritarian states.Democratic institutions are often
poor vehicles for organizing the equaldivision of power and
privilege among hostile ethnic groups. Hence loser-groups are often
even more dissatisfied under democracy than they wereunder the
previous authoritarian regime. Democratic freedoms (of
speech,press, assembly, etc.) also give political space to
determined secessionistgroups, allowing them more room to organize
for war. Hencedemocratization can unleash communal conflicts that
lay dormant underprevious authoritarian regimes.
The alternative to democratization is hunkering down:
relyingincreasingly on one edinic, tribal, or religious group or
one sector ofsociety, such as the military or members of wealthier
social classes.Hunkering down, in the short term, can allow a
regime to weather a crisisas it can count on the loyalty of key
elites. In the long term, however,hunkering down can provoke
greater popular discontent with the regime.
Democratization
Democratization offers four paths to civil war. Incumbent
authoritarianelites may crush emerging democratic forces because
they fear that the firstdemocratic victors will exploit state power
to impose a new dictatorship)what has been called "one person one
vote once." Minorities may fightbecause they fear that majority
rule would install in power a permanentelected majority that allows
the minority no voice in decision making. Afterdemocratic
transitions, victorious groups may fight over the division
ofspoils. Finally, democratization may empower hardened
secessionists whoexploit democratic freedoms to organize
secessionist rebellion. Severalrecent conflicts, including those in
.Algeria, Azerbaijan-Armenia, Chechnya,Georgia, India (Kashmir),
Pakistan, South Africa, Tajikistan, and the united
38. Regime responses to legitimacy loss are not limited to
democratizing and hunkeringdown. Regimes at time promote economic
reform to recover their legitimacy. Other,bloodier, alternatives
include provoking an international conflict and blaming problems
onminorities or other scapegoats at home.
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
34 SECURITY STUDIES 7, NO. 3
Yemen, stemmed in part from attempts at democratization and
followedone of these paths from democratization to war.
The Algerian case suggests a general lesson noted by Adam
Przeworski,who argues that a necessary condition of successful
democratization is therealistic expectation that relinquishing
power now will not require doing sofor ever. In Algeria the ruling
National Liberation Front (FLN) elites refusedto concede a lost
election in 1993 in part from fear that the Islamistwinners would
impose a dictatorship once in power. This led the FLN tohunker down
instead of accepting the results of democratization.
In Georgia, democratization produced war by causing minority
fears ofmajority tyranny.40 As noted above, the minority Abkhaz
feared that theirdistinct cultures would be overrun by a
power-monopolizing Georgianmajority. Hence they opted for violent
resistance when Georgiannationalists appeared poised to win
elections. The experiences of Sri Lankaand Northern Ireland teach
the same lesson. In Sri Lanka, the majoritySinhalese long
monopolized power at the expense of the minority Tamils,provoking
the bloody Tiger rebellion.41 In Northern Ireland, the
Protestantmajority monopolized power at the expense of the Catholic
minority during192269, provoking violent Catholic nationalism.
The victors of democratization can also quarrel over the spoils.
SouthAfrica's transition to democracy led to increased tension
between Inkathaand the ANC over the division of power within the
victorious African
39. Adam Przeworski, Democracy and the Market (Cambridge:
Cambridge University Press,1991), 26-34. The fear of the FLN was
not without some merit. For discussions of whyAlgeria's leaders
were reluctant to surrender power and doubted the good faith of
theIslamists, see "Shooting or voting for Islam," Economist, 28
August 1993, 39; and ClaireSpencer, "Algeria in Crisis," Survival
36, no. 2 (summer 1994): 149-63. For an overview onthe general
question of the tension between democratic ideals and Islamic
movements, seeJohn L. Esposito and John O. Voll, Islam and
Democracy (New York: Oxford University Press,1996). The authors
discuss Algeria on pages 151-72.
40. Discussion of majority tyranny traces back to James Madison,
"The Same SubjectContinued..." (Federalist no. 10), The Federalist
Papers, intro. Clinton Rossiter (New York:New American Library,
1961), 77-84. Madison discusses the risks that arise when "a
majorityis included in a faction" (80) and the dangers of tyranny
by "the superior force of aninterested and overbearing majority"
(77). Discussing remedies are Arend Lijphart, "ThePower-Sharing
Approach," in Conflict and Peacemaking in Multiethnic Societies ed.
Joseph V.Montville (Lexington, Mass.: Lexington Books, 1990),
491-509; Kenneth D. McRae,"Theories of Power-Sharing and Conflict
Management," in Montville, Conflict andPeacemaking, 93-106; Jurg
Steiner, "Power-Sharing: Another Swiss 'Export Product?"
inMontville, Conflict and Peacemaking, 107-14; and Timothy D. Sisk,
Power Sharing and InternationalMediation in Ethnic Conflicts
(Washington, D.C.: United States Institute of Peace, 1996),
34-45,58-63.
41. An account is Brogan, Fighting Never Stopped, 221-34.
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Hypotheses on the Causes of Contemporary Deadly Conflict 35
community. In Yemen, the former southern leaders felt cheated
out of theirshare of power after elections there.
The Chechen experience illustrates the risk that hardened groups
willexploit democratic freedoms to promote separatism. After
sufferingrepeated cruelties by Russian rulers, many Chechens want
no part of theRussian state, regardless of its government type or
its respect for minorityrights. When given the right to assemble
and speak freely, they found aconsensus on rejecting any ties to
Moscowa position that triggered abrutal Russian crackdown in which
tens of thousands of Chechens andRussians died.
Hunkering Down
Instead of democratizing, some regimes respond to pressure to
pluralize by"hunkering down"relying on a narrower base of support
to stay inpower. Such a move, however, can provoke further dissent
and violence.The very problems that provoke a legitimacy crisis in
the first placecorruption, a lack of accountability, demands for a
say in decision making,etc.are exacerbated by the hunkering-down
process. Thus, even thoughin the short term a regime may survive a
challenge, in die long term diescope and scale of dissent is likely
to grow.
Hunkering-down behavior fueled many current disputes, including
bodithose that began before and after the end of die cold war. In
Rwanda andBurundi, for example, regimes have relied increasingly on
edinic kinsmenwith no pretense of including others. Similarly,
regimes in Burma, Chad, ElSalvador, Ediiopia, Guatemala, Iraq,
Kenya, Lebanon, Liberia, Pakistan, thePhilippines, Somalia, Uganda,
and Zaire have relied more and more on oneparticular ethnic,
religious, or tribal group or one sector of society, usually
42. Joffee, "YemenThe Reasons for Conflict," 370-71.43. The road
to democracy contains other potential perils. In their work
discussing the
relationship between democratization and interstate war,
Mansfield and Snyder note that theinitial stage of democratization
is extremely dangerous for several reasons. First, threatenedelites
from the autocratic regime often use chauvinistic rhetoric as they
compete for alliesamong the populace. Second, social groups that
might be losers in a mature democracy oftenmanipulate information
and otherwise distort the democratic process. Third, a lack of
stronginstitutionsthe checks and balances that places power in the
hands of a responsible, well-informed votercan further increase the
chances of war. All these reasons whydemocratization can cause
interstate war apply to internal conflict as well. Elites'
chauvinisticrhetoric can be targeted at ethnic minorities,
particularly if they are traditional enemies, aswell as other
countries. Beleaguered social groups often manipulate information
anddemonize their opponents, making power-sharing extremely
difficult. The lack of stronginstitutions can allow a small number
of individuals to take the steps necessary to bring aboutinternal
war. Edward D. Mansfield and Jack Snyder, "Democratization and the
Danger ofWar," International Security 20, no. 1 (summer 1995):
5-38.
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
36 SECURITY STUDIES 7, NO. 3
the military, to govern after challenges to their legitimacy
arose. Thisreliance has created a self-sustaining cycle: as dissent
increases, regimes fallback more and more on "trusted" individuals
from the same communalgroup or sector; this reliance in turn
increase resentment among theexcluded groups and engendered charges
of corruption and favoritism.
The "Reform Trap"Hunkering down often follows democratization
when regimes face alegitimacy challenge. Efforts to democratize
lead to a growing governabilitycrisis, which in turn leads the
regime to hunker down and abort itsdemocratization experiment. This
"reform trap" generates further dissentand leads to civil war.
The democratization process and subsequent hunkering down
oftenshare the same cause: a regime attempting to stay in power
while gainingpopular support for painful reforms. Algeria, for
example, began itsdemocratization process after food riots in 1988,
and Sierra Leone allowedelections after it adopted unpopular
economic reforms under IMFpressure. The pain engendered by the
economic reform process, however,often leads the regime to lose the
elections, causing it to hunker down inorder to stay in
power.45
The "reform trap" is a common factor in civil wars throughout
the worldtoday. Algeria, Burma, Pakistan, Somalia, and Tajikistan
all initiatedhesitant democratization and then, when the results
were not to the likingof the regime or a powerful group, chose
instead to hunker down andignore the elections. Algeria's
experience illustrates the reform trap neatly.In 1989 the ruling
FLN leaders authorized elections in order to reach out toa hostile
society disgruntled by regime corruption, a lack of
accountability,
44. On Sierra Leone, see Christopher Clapham "Recent History,"
Europa World BookAfrica 1995 (London: Europa Publications Limited,
1994), 803-7.
45. Angola's recent return to violence in 1992 illustrates the
other side of this coin. ThereUNITA, the leading opposition group,
expected to defeat the government in elections andreturned to
violence when it lost. For articles that note Savimbi's reluctance
to accept theobvious verdict of the polls, see Alex Vines, Angola
and Mozambique: Aftermath of Conflict,Conflict Studies 280
(Research Institute for the Study of Conflict and Terrorism, 1995);
andAndrew Meldrum, "Lessons from Angola," Africa Report 38, no. 1
Qanuary-February 1993):22-24.
46. Such a tension is common in collapsed empires. As noted
above, the governments ofthe successor states often lack legitimacy
and thus face the choice of democratizing to try togain popular
support or hunkering down if they fear popular rejection.
Furthermore, manyformer empires, including both the Yugoslav and
the Soviet Empire, often contain hostileminorities. Thus, the
polarization that makes democratization extremely difficult often
ispresent.
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
Hypotheses on the Causes of Contemporary Deadly Conflict 37
and economic stagnation. Reformers within the FLN hoped to use
electionsto regain popular support and to acquire a mandate to
carry out difficulteconomic changes. Not surprisingly, the Islamic
Salvation Front (FIS) wonthe elections, much to the horror of the
FLN. The FLN then decided to"hunker down." It nullified the
elections and tried to disband the FIS.Bloody civil war soon
followed as the FIS resisted these measures.
WEAK STATES
Weakness in the capacities of states has been an important cause
of currentcivil violence. Defeat in war, loosened central control
over economicactivity, and other factors that lessen a state's
strength can spark civilviolence. Recently, states have been
weakened by the end of the cold war,which ended superpower aid to
client Third World regimes, and by thespread of powerful small
armaments. The weakening of the state increasedthe incidence of
civil war in two ways: it decreased the state's coercivepower, and
reduced its ability to co-opt opponents and rival groups.
A decrease in a state's coercive ability fosters civil conflict
in two ways.First, if the state is weak, restive ethnic groups or
other threats to peace areno longer reassured or deterred from
organizing. Predatory groups plot warbecause they are less deterred
by fear of state repression. This alarms othergroups who then
mobilrze for war in self-defense, taking security into theirown
hands because they no longer trust the central state to provide it.
Thusas the weakness of the state in Lebanon became apparent in the
early1970s, various communal groups began forming militias for
self-defense.The second, related, impact of decreased state
coercive ability is an inabilityto defeat groups committed to
violence. Even unpopular regimes canstamp out potentially violent
opposition when they have enough resourcesto overwhelm the
insurgents directly, arrest their leaders, or otherwise
47. A weak state is one that lacks financial, military, and
institutional resources toimplement its policies. For works that
note the importance of the strength of the state, seePeter B.
Evans, Dietrich Rueschemeyer, and Theda Skocpol, eds., Bringing the
State Back In(Cambridge: Cambridge University Press, 1985); Charles
Tilly, ed., The Formation of NationStates in Western Europe
(Princeton: Princeton University Press, 1975); and Joel S.
Migdal,Strong Societies and Weak States: State-Society Relations
and State Capabilities in the Third World(Princeton: Princeton
University Press, 1988).
48. One traditionally important cause of government weakness is
defeat in aninternational war. In the post-cold war period,
however, only the Iraqi case fits thispatternan unsurprising
development given the overall dearth of international conflict
inthis period. The Second World War era conflicts in Vietnam,
Yugoslavia, and Greece,however, are examples of how an
international conflict can weaken (or remove) stategovernments,
thus catalyzing groups for civil conflict.
49. Dilip Hiro, Lebanon: Fire and Embers (New York: St.
Martin's, 1992), 12.
Dow
nloa
ded
by [K
ing's
Colle
ge L
ondo
n] at
11:01
23 Se
ptemb
er 20
13
-
38 SECURITY STUDIES 7, NO. 3
interfere with group organization before the violence becomes
widespread.When the state weakens, however, insurgencies become
more difficult todefeat. In Iraq, for example, Shi'a Muslim
organizations had long opposedthe Ba'athist regime, but it was only
after the near collapse of SaddamHussein's government following
Operation Desert Storm that they gainedwidespread support and
almost toppled the government.50
Weakened states also have fewer resources with which to buy
offopposition.51 Somalia's economy collapsed in the 1980s as income
fromremittances fell and the regime's appalling human rights record
led to adecrease in international aid.52 Hence, Siad Barre's
government lost itsability to play off various clans by dangling
aid in front of them. As a result,he was forced to consider
elections, and the country soon unraveled.
Cold war factors. During the cold war both superpowers bolstered
frailThird World client regimes with arms, military training,
money, and attimes troops. This aid lost its rationale as the cold
war faded, hence thesuperpowers sharply reduced their largesse.
This caused a marked decline inthe strength of the superpowers'
client regimes.53 Between 1981 and 1984,the United States gave or
sold $800 million (in current dollars) in arms toAfrica, while the
Soviet Union delivered $11.1 billion. This patterncontinued from
1985 to 1988, widi Moscow sending over $13.5 billion inarms and
Washington sending $900 million. Deliveries plummeted after thecold
war ended. Between 1992 and 1994, the United States delivered
$395million in arms to Africa, while Russia sent $610 million worth
of arms.54
If Western aid was given, it was now often conditioned on
democratic ormarket reforms. This reform pressure, in turn, often
precipitated
50. For a discussion of the impact of the Gulf War on Iraqi
Shi'a, see Yitzhak Nakash, TheShi'is of Iraq (Princeton: Princeton
University Press, 1994), 273-81. For a detailed analysis
ofpolitical Shi'ism in Iraq, see Joyce N. Wiley, The Islamic
Movement of Iraqi Shi'as (BoulderLynne Rienner, 1992).
51. If governments can win elites to their side and prevent them
from encouragingconflict, fighting may be mitigated despite
widespread hostility on the part of the populationat large. Robert
Dahl notes the importance of political activists in the
stability