The Extremist’s Advantage in Civil Wars

The Extremist’s
Advantage in
Civil Wars

Barbara F. Walter

One of

the puzzles
of the current wave of civil wars is that rebel groups espousing extremist
ideologies—especially Salaª jihadism—have thrived in ways that moderate
rebels have not.1 Groups such as Jabhat al-Nusra and the Islamic State (also
known by the acronym ISIS) have attracted more recruits, foreign soldiers, and
ªnancing than corresponding moderate groups such as the Free Syrian Army,
Ahlu Sunna Waljamaa, or Jaysh Rijaal al-Tariqa al-Naqshbandia (JRTN).2 The
proliferation and success of extremist groups is particularly surprising given
that their goals are far more radical than those of the populations they seek
to represent.3 Salaª jihadists aim to establish a transnational caliphate using
military force, an objective the vast majority of Muslims do not support.4
Why have so many extremist groups emerged in countries experiencing civil
wars since 2003, and why have they thrived in ways that moderate groups
have not?

Barbara F. Walter is Professor of Political Science at the School of Global Policy and Strategy at the Univer-
sity of California, San Diego.

The author thanks Jesse Driscoll, Isaac Gendel, Dotan Haim, Ron Hassner, Allison Hodgkins,
Joshua Kertzer, Aila Matanock, William McCants, Assaf Moghadam, Richard Nielsen, Emily
Ritter, Michael Stohl, and Keren Yarhi-Milo for their willingness to read the manuscript and offer
helpful feedback. She is especially grateful to Gregoire Phillips for answering an endless series of
questions with enormous good cheer. Finally, she thanks the participants of the International Rela-
tions Faculty Colloquium at Princeton University for inviting her to present this work and follow-
ing up with thoughtful suggestions.

1. See Barbara F. Walter, “The New New Civil Wars,” Annual Review of Political Science, Vol. 20
(2017), pp. 469–486.
2. In 2016, Jabhat al-Nusra changed its name to Jabhat Fateh al-Sham. For simplicity, I use Jabhat
al-Nusra throughout when referring to this organization.
3. For data on the growth of Salaª jihadist movements since 1946, see Nils Petter Gleditsch and
Ida Rudolfsen, “Are Muslim Countries More Prone to Violence?” Research and Politics, April/June
2016, pp. 1–9; and Seth G. Jones, “A Persistent Threat: The Evolution of al Qa’ida and Other Salaª
Jihadists” (Santa Monica, Calif.: Rand Corporation, 2014), http://www.rand.org/pubs/research
_reports/RR637.html.
4. There were many varieties of radical Islamist movements, but the most extreme—the Salaª
jihadists—were the most numerous and successful. Note that the conceptualization of “caliphate”
is hotly debated by Muslim scholars and intellectuals. The “caliphate” that groups such as al-
Qaida and the Islamic State seek to build is ahistorical, not determined by the different caliphal
states that have existed in the past. For an in-depth discussion of these differences, see Hugh Ken-
nedy, The Prophet and the Age of the Caliphates: The Islamic Near East from the 6th to the 11th Century
(London: Longman, 1986); Madawi Al-Rasheed, Carool Kersten, and Marat Shterin, eds., Demys-
tifying the Caliphate: Historical Memory and Contemporary Contexts (London: Hurst, 2012); and
Hamid ‘Inayat, Modern Islamic Political Thought (New York: American Council of Learned Societies,
2014).

International Security, Vol. 42, No. 2 (Fall 2017), pp. 7–39, doi:10.1162/ISEC_a_00292
© 2017 by the President and Fellows of Harvard College and the Massachusetts Institute of Technology.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

International Security 42:2 8

This article has two goals. The ªrst is to provide a theoretical framework for
understanding the role of ideology, especially an extreme ideology, in civil
wars. It argues that an extreme ideology can offer rebel entrepreneurs sig-
niªcant organizational advantages over more moderate groups, especially in
environments with multiple competing rebel groups, weak rule of law, and
bad governance. The second goal is to apply this theory to current civil wars
in Muslim countries in order to explain the rise of radical Islamist groups
since 2003.

The article argues that under certain conditions, rebel leaders (such as Abu
Bakr al-Baghdadi of the Islamic State) have strong incentives to embrace an ex-
treme ideology even if they do not believe the ideas that underlie it.5 The same
is true of more moderate citizens who choose to support or associate with
these groups.6 Average citizens may have incentives to join or collaborate
with an extreme rebel group if they feel that such a group is more likely to win
a war and resist corrupting inºuences once in power. The initial leaders of
ISIS, for example, were a mix of religious zealots (Abu Ayyub al-Masri and
Abu Bakr al-Baghdadi), disgruntled tribesmen, and members of Saddam
Hussein’s secular Baathist regime.7 These individuals then built their organ-
ization by recruiting foot soldiers largely ignorant of Salaª tenets and prac-
tices.8 The recent rise of radical Islamist groups, therefore, does not necessarily
indicate an increase in support for radical ideas in Muslim countries experi-
encing civil war as much as it may reveal average Sunnis behaving strategi-
cally during uncertain and difªcult times.

The core of the argument is that an extreme ideology can help rebel elites
overcome three organizational challenges that may arise as they seek to build
movements in highly competitive, low-information environments.9 The ªrst is
a potentially debilitating collective-action problem.10 To mobilize and maintain

5. In addition, there were other key ªgures in the early organization, such as Abu Anas-al-Shami,
who claimed to be leaders of religious schools. Other midlevel ªgures in the organization’s Sharia
councils, even if lacking convincing credentials, also seem to have become more outwardly reli-
gious in the post-2003 Salaª jihadist movement.
6. For an excellent and nuanced discussion of the reasons individual citizens choose to support
and collaborate with insurgents, see Stathis N. Kalyvas, The Logic of Violence in Civil War (New
York: Cambridge University Press, 2006).
7. One possible exception was Abu Musab al-Zarqawi.
8. William McCants, The ISIS Apocalypse: The History, Strategy, and Doomsday Vision of the Islamic
State (New York: St. Martin’s, 2015), pp. 121, 150.
9. These organizational advantages may be so great that individuals are willing to absorb the
sacriªces associated with membership in such groups. For an in-depth discussion of these
sacriªces, see Lawrence R. Iannacone and Eli Berman, “Religious Extremism: The Good, the Bad,
and the Deadly,” Public Choice, Vol. 128, No. 1 (2006), pp. 109–129.
10. Jeremy M. Weinstein, Inside Rebellion: The Politics of Insurgent Violence (Cambridge: Cambridge
University Press, 2006); Stathis N. Kalyvas and Michael Adam Kocher, “Ethnic Cleavages and Ir-

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

The Extremist’s Advantage in Civil Wars 9

an army, rebel leaders must convince at least some individuals to pay the high
costs of ªghting, a requirement that most people would prefer to avoid.
Normally, rebel groups attempt to overcome collective-action problems by of-
fering recruits private rewards in the form of money, security, or access to
plunder. Extremist groups, especially faith-based groups, have a particular
advantage in the provision of private goods, because they can offer cheap, de-
ferred compensation in the form of an “eternal afterlife” or “rewards in para-
dise,” which nonreligious ideologies cannot do.11 They can also promise a
potentially devastating type of personal punishment (eternal damnation in the
case of Christianity and excommunication in the case of Islam) that is both
costless to enforce and impossible for targets to escape.

The second organizational challenge comprises two potentially harmful
principal-agent problems.12 To evade capture and avoid defeat, rebel leaders
must recruit soldiers who will remain committed and loyal to the organization
even after they are deployed to the ªeld. An extreme ideology can help over-
come this agency problem by screening out less committed soldiers, which
then reduces the problem of poor performance, side-switching, and betrayal.
Rebel leaders can also use an extreme ideology to signal their own dedica-
tion to a cause, allowing them to attract more devoted and hard-ªghting sol-
diers to their ranks.

The third organizational challenge is a potentially destabilizing commitment
problem. To successfully compete against other factions ªghting a war, rebel
leaders must reassure their soldiers and supporters that they will resist corrup-
tion once in power. This is especially important in countries with few insti-
tutional constraints on government elites and a history of exploitation. An
extreme ideology, especially one that demands personal sacriªces from its
leaders, could serve as a way for rebel elites to credibly commit to more princi-
pled behavior once in ofªce in a way that competing groups cannot. Taken to-
gether, these three challenges could help explain why ideologically extreme
rebel groups might outperform moderate groups even in environments where
the preferences of most citizens are not extreme.

regular War: Iraq and Vietnam,” Politics & Society, Vol. 35, No. 2 (June 2007), pp. 183–223; Scott
Gates, “Recruitment and Allegiance: The Micro-foundations of Rebellion,” Journal of Conºict Reso-
lution, Vol. 46, No. 1 (February 2002), pp. 111–130; Pablo Policzer, The Rise and Fall of Repression in
Chile (Notre Dame, Ind.: University of Notre Dame Press, 2009); and Francisco Gutiérrez Sanín and
Elisabeth Jean Wood, “Ideology in Civil War: Instrumental Adoption and Beyond,” Journal of Peace
Research, Vol. 51, No. 2 (2014), pp. 213–226.
11. Interestingly, nationalist groups may attempt to create a similar inducement by promising
members long-term hero status. Statues, monuments, and national celebrity could be the secular
equivalent of eternal salvation. I am indebted to Ron Hassner for this insight.
12. See Gates, “Recruitment and Allegiance”; Weinstein, Inside Rebellion; and Sanín and Wood,
“Ideology in Civil War.”

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

International Security 42:2 10

This argument represents a signiªcant departure from the conventional lit-
erature on civil wars in three important ways. First, it starts from the assump-
tion that adopting a speciªc ideology is a strategic decision made by rebel
leaders and not simply an extension of the underlying beliefs or grievances in
a society. Until now, most studies of civil war have treated the choice of ideol-
ogy as exogenous to the war; the goals and beliefs of rebel groups were as-
sumed to reºect preexisting societal cleavages and the political disagreements
that emerged from them.13 In contrast, this article argues that rebel leaders
choose an ideology based in part on the cleavage structure of society but also
on a real desire to outmaneuver and outcompete other factions ªghting a
war.14 The choice of ideology, therefore, is likely to be at least partly inºuenced
by the endogenous dynamics of the war itself: the number of rebel groups
vying for power, the quality of local information networks, and the strength
of institutional checks on government behavior.15

Second, this article views the stated ideology of a rebel group as a form of
cheap talk. Until now, most studies of civil war have assumed that the goals
and ideas endorsed by warring factions were genuine. This article argues in-

13. The terrorism literature considers ideology to be more malleable than does the civil war litera-
ture. On the role of ideology in how wars end, see, for example, Audrey Kurth Cronin, How Terror-
ism Ends: Understanding the Decline and Demise of Terrorist Campaigns (Princeton, N.J.: Princeton
University Press, 2009); and Assaf Moghadam, “Motives for Martyrdom: Al-Qaida, Salaª Jihad,
and the Spread of Suicide Attacks,” International Security, Vol. 33, No. 3 (Winter 2008/09), pp. 46–
78. The social movement literature discusses the role of entrepreneurs in the strategic framing of
ideology. See especially Doug McAdam, John D. McCarthy, and Mayer N. Zald, eds., Comparative
Perspectives on Social Movements: Political Opportunities, Mobilizing Structures, and Cultural Framings
(Cambridge: Cambridge University Press, 1996).
14. For an excellent review of theories related to ideological belief systems, see John T. Jost, Chris-
topher M. Federico, and Jaime L. Napier, “Political Ideology: Its Structure, Functions, and Elective
Afªnities,” Annual Review of Psychology, Vol. 60 (2009), pp. 307–337.
15. The literature on the construction of ethnic politics makes a related claim that the choice of a
leader to mobilize an “ethnic” agenda (and the incentive for individuals to follow this agenda) is
based on rational organizational challenges. See David A. Lake and Donald Rothchild, “Con-
taining Fear: The Origins and Management of Ethnic Conºict,” International Security, Vol. 21, No. 2
(Fall 1996), pp. 41–75; Barry R. Posen, “The Security Dilemma and Ethnic Conºict,” Survival,
Vol. 35, No. 1 (Spring 1993), pp. 27–47; Paul R. Brass, Theft of an Idol: Text and Context in the Repre-
sentation of Violence (Princeton, N.J.: Princeton University Press, 1997); James D. Fearon and David
D. Laitin, “Violence and the Social Construction of Identity,” International Organization, Vol. 54,
No. 4 (Autumn 2000), pp. 845–877; David D. Laitin, Nations, States, and Violence (Oxford: Oxford
University Press, 2007); Joan Esteban and Debraj Ray, “On the Salience of Ethnic Conºict,” Ameri-
can Economic Review, Vol. 98, No. 5 (2008), pp. 2185–2202; Nicholas Sambanis and Moses Shayo,
“Social Identiªcation and Ethnic Conºict,” American Political Science Review, Vol. 107, No. 2 (May
2013), pp. 294–325; Stanley J. Tambiah, Sri Lanka: Ethnic Fratricide and the Dismantling of Democracy
(London: I.B. Tauris, 1986); Stanley J. Tambiah, “The Nation-State in Crisis and the Rise of
Ethnonationalism,” in Edwin N. Wilmsen and Patrick McAllister eds., The Politics of Difference: Eth-
nic Premises in a World of Power (Chicago: University of Chicago Press, 1996), pp. 124–143; and
Bruce Kapferer, Legends of People, Myths of State: Violence, Intolerance, and Political Culture in Sri
Lanka and Australia (Washington, D.C.: Smithsonian Institution Press, 1988).

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

The Extremist’s Advantage in Civil Wars 11

stead that rebel elites, their soldiers, and rank-and-ªle sympathizers likely un-
derstand that public declarations about ideology are potentially empty words
that may or may not be backed by real action. The leaders of the Kurdish-Iraqi
group Ansar al-Islam, for example, had invested for years in public propa-
ganda designed to distinguish their group from the Islamic State, only to
merge with the group in 2014.16 The public’s understandable skepticism about
elite motives may create an opening for political entrepreneurs to adopt an ex-
treme ideological position as a way to signal their true dedication to a cause.
Salaª jihadists may be perceived as more committed to real reform than more
moderate groups, because adhering to a set of extreme principles inºicts at
least some costs on leaders.17 As I argue below, it is the cheap-talk nature of
most ideological platforms that makes an extreme ideology, with its excessive
dictates and demands, potentially informative.

Third, this article treats rank-and-ªle supporters of extremist groups as stra-
tegic actors in their own right and not simply blind followers to a cause. Until
now, most people have viewed the supporters of extremist movements as
true believers, ignoring the motivational processes of these individuals. Even if
rebel elites were deemed to be strategic or self-interested in their choice of ide-
ology, the “masses” were assumed to be more ideologically driven. This article
argues, however, that not all participants in these movements need to be truly
committed to an ideology to be willing to support a particular group. They
need only believe that an extreme group is better positioned to deliver an
outcome closer to their desired preferences than other groups competing for
their support.

Still, this does not mean that all rebel leaders use ideology instrumentally or
that true believers do not exist. In fact, rebel entrepreneurs require at least
some true believers for an extreme ideology to be an effective strategy in war.
To be successful, elites require passionate devotees to a cause who are willing
to join a movement in its infancy and ªght especially hard. These ideologues
allow the group to outcompete rival organizations, winning early battles and
making the group potentially more attractive to moderate individuals seeking
to assess whom to support. An extreme ideology, therefore, can make a group
attractive to both ideological and non-ideological individuals. True believers

16. Aymenn Jawad Al-Tamimi, “Key Updates on Iraq’s Sunni Insurgent Groups,” Brown Moses
blog, May 11, 2014, http://brown-moses.blogspot.co.uk/2014/05/key-updates-on-iraqs-sunni-
insurgent.html; and Aymenn Jawad Al-Tamimi, “Jamaat Ansar al-Islam: Fighting in North Alep-
po Countryside—Translation and Analysis” (Herzliya, Israel: Rubin Center, IDC Herzliya,
June 3, 2015), http://www.rubincenter.org/2015/06/jamaat-ansar-al-islam-ªghting-in-north-
aleppo-countryside-translation-and-analysis/.
17. Eli Berman and David D. Laitin, “Religion, Terrorism, and Public Goods: Testing the Club
Model,” Journal of Public Economics, Vol. 92, Nos. 10–11 (October 2008), pp. 1942–1967.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

International Security 42:2 12

join because they are genuinely committed to the goals and principles of a
movement. More practical individuals join because they believe the group is
better organized, more likely to win, and less corruptible than rival groups.

Note that the theory does not claim that ideas do not matter. Even the most
strategic rebel entrepreneur understands the beneªt of choosing a set of ideas
that already resonates with local citizens.18 It would make no sense to embrace
an Islamist ideology in a majority-Christian country or white nationalism in
the Democratic Republic of Congo. To have any chance of success, rebel entre-
preneurs must choose a set of ideas that coincides with the interests of at least
a subset of society. It is no surprise, for example, that most rebel groups in Iraq,
Libya, Syria, and Yemen have embraced political Islam as their base ideology,
given that it is favored by large majorities of Muslim citizens in those coun-
tries. Instead, this article argues that rebel entrepreneurs will strategically
emphasize those elements of an ideology that provide them the greatest com-
petitive advantage given the strategic conditions on the ground.

The ªnding that an extreme ideology can be an effective tool of war under
certain conditions could have potentially far-reaching implications for policy-
makers struggling with the rise of violent extremist groups around the world.
Since the terrorist attacks of September 11, 2001, the United States has been
grappling with how to reduce the growth and inºuence of Salaª jihadist
groups whose goals directly threaten U.S. interests. If such groups are growing
in number because Muslim citizens are becoming more radical in their beliefs,
then successful counterstrategies must attempt to change these beliefs—a
monumental task. If, however, citizens choose to support these groups for
more practical reasons—because Salaª jihadist groups are better able to build
an effective ªghting force or offer greater reassurance against corruption—
then effective counterstrategies need only address the conditions that make
support for these groups advantageous. Eliminate the underlying conditions
that make an extreme ideology expedient to embrace, and you eliminate the
incentives elites and moderates have to back it.

What follows is divided into ªve sections. In the ªrst section, I explain what
this article does and does not do. In the second, I consider the existing litera-
ture on the role of ideology in civil war and deªne the concepts of “ideology”
and “extremism.” In the third section, I lay out a theory of the strategic use of
extreme ideology in civil war. This includes the three strategic problems that
elites face when attempting to build a rebel organization, the conditions under
which these problems are likely to be especially severe, and the ways in which
an extreme ideology may help to reduce or eliminate them. In the fourth sec-

18. Sanín and Wood, “Ideology in Civil War,” p. 222.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

The Extremist’s Advantage in Civil Wars 13

tion, I use this theoretical framework to explain the rise of radical Islamist
groups in civil wars since 2003, especially Salaª jihadist groups. In the ªnal
section, I outline the implications that this research could have both for schol-
ars studying civil war and for policymakers struggling to address the rise of al-
Qaida, ISIS, and other Salaª jihadist groups around the world.

Scope Conditions

This is an article about the potential beneªts that rebel elites may gain by
adopting an extreme ideology, especially a religious ideology, and the practical
beneªts moderate citizens may gain for backing these elites. The theory, there-
fore, applies to extremism that emerges in the context of a civil war, by a group
pursuing violence, and that is religiously rather than nonreligiously based.
The article does not directly address extremist ideology outside the civil war
context, such as the use of extremism in the give-and-take of conventional pol-
itics. It also does not consider nonviolent extremism practiced by groups such
as the Amish or nonreligious extremism such as Adolf Hitler’s use of Nazism
in the 1930s and 1940s. Still, I suspect that the theory is likely to apply to situa-
tions with similar characteristics to civil wars: circumstances where groups are
competing for power under conditions of rapid political change and weak en-
forcement mechanisms.19 These are the environments in which collective-
action problems, principal-agent problems, and commitment problems are
likely to be particularly severe and where advantages are likely to accrue to or-
ganizations that best solve these problems. The degree to which the theory
generalizes beyond religious extremist groups in internal conºicts, though ex-
tremely important, is left to future research.

This is also not an article about the origins of ideology in a given society. It
does not address where ideas come from, why they resonate with certain pop-
ulations but not others, or how they change over time.20 Moreover, it is not an

19. The Shining Path in Peru, the Tamil Tigers in Sri Lanka, Maoists in China, the Lord’s Resis-
tance Army in Uganda, and the Revolutionary Armed Forces of Colombia all espoused more radi-
cal ideas and goals than the average citizens in their countries, and all competed successfully in
their respective wars.
20. The following is a sampling of the literature on the origins of ideology. In political science, see
Morris P. Fiorina, Samuel J. Abrams, and J.C. Pope, Culture War? The Myth of a Polarized America
(New York: Pearson Longman, 2005); Keith T. Poole and Howard Rosenthal, Congress: A Political-
Economic History of Roll Call Voting (Oxford: Oxford University Press, 1997); and John Zaller, The
Nature and Origin of Mass Opinion (Cambridge: Cambridge University Press, 1992). In psychology,
see T.W. Adorno et al., The Authoritarian Personality (New York: Harper and Row, 1950); John T.
Jost, “The End of the End of Ideology,” American Psychologist, Vol. 61, No. 7 (October 2006),
pp. 651–670; and Charles M. Judd and Jon A. Krosnick, “The Structural Bases of Consistency
among Political Attitudes: Effects of Political Expertise and Attitude Importance,” in Anthony R.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

International Security 42:2 14

article about the means by which citizens may be indoctrinated with one set of
ideas over another; that is a topic best left to experts in psychology or educa-
tion. All of these topics are worthy of serious study but are beyond the scope
of this article.

Existing Literature and Deªnitions

An extensive review of the literature on civil wars ªnds no book and only
three articles on the role of ideology in civil war.21 One of the articles uses
quantitative analysis to examine whether ideology inºuenced different players
in Colombia’s civil war.22 The authors found that it did: ideology allowed
paramilitaries and guerrillas to distinguish themselves from each other, and it
inºuenced the beliefs of soldiers over time. The second article summarizes the
ways in which ideology could affect the behavior of rebel groups in civil war,
including how it might inºuence their organization, activities, and mobiliza-
tion.23 A third article argues that ideology, especially radical ideologies, is nec-
essary to mobilize individuals.24 A number of other articles mention ideology,
but in ways not central to their main arguments.25 This article offers a theoreti-

Pratkanis, Steven J. Breckler, and Anthony G. Greenwald, eds., Attitude, Structure, and Function
(New York: Psychology, 1989), pp. 99–128.
21. The literature on terrorism has been more proactive than the literature on civil war in address-
ing the role of ideology in terrorist violence and targeting. Most of this literature focuses on the
role of religion in the behavior of individuals and terrorist organizations. Mark Juergensmeyer, for
example, explores how religion helps legitimize violence. See Juergensmeyer, Terror in the Mind of
God: The Global Rise of Religious Violence (Berkeley: University of California Press, 2000). Mia Bloom
argues and that suicide bombings are motivated by social and political concerns, not religious
ones. See Bloom, Dying to Kill: The Allure of Suicide Terror (New York: Columbia University Press,
2007). C.J.M. Drake examines the role of ideology in terrorists’ target selection. See Drake, “The
Role of Ideology in Terrorists’ Target Selection,” Terrorism and Political Violence, Vol. 10, No. 2 (Sum-
mer 1998), pp. 53–85, doi:10.1080/09546559808427457. Andrew H. Kydd and Barbara F. Walter re-
veal the ways in which an extreme ideology can create incentives for groups to intentionally
scuttle peace agreements. See Kydd and Walter, “Sabotaging the Peace: The Politics of Extremist
Violence,” International Organization, Vol. 56, No. 2 (Spring 2002), pp. 263–296, doi:10.1162/
002081802320005487.
22. Juan E. Ugarriza and Matthew J. Craig, “The Relevance of Ideology to Contemporary Armed
Conºicts: A Quantitative Analysis of Former Combatants in Colombia,” Journal of Conºict Resolu-
tion, Vol. 57, No. 3 (2013), pp. 445–477, https://doi.org/10.1177/0022002712446131.
23. Sanín and Wood, “Ideology in Civil War.”
24. Stefano Costalli and Andrea Ruggeri, “Indignation, Ideologies, and Armed Mobilization: Civil
War in Italy, 1943–45,” International Security, Vol. 40, No. 2 (Fall 2015), pp. 119–157, doi:10.1162ISEC
_a_00218.
25. Kalyvas, The Logic of Violence in Civil War; Weinstein, Inside Rebellion; Andreas Wimmer, Waves
of War: Nationalism, State Formation, and Ethnic Exclusion in the Modern World (Cambridge: Cam-
bridge University Press, 2013); Laia Balcells, Rivalry and Revenge: The Politics of Violence during Civil
War (Cambridge: Cambridge University Press, 2017); and Lars-Erik Cederman, Nils B. Weidmann,
and Kristian Skrede Gleditsch, “Horizontal Inequalities and Ethnonationalist Civil War: A Global

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

The Extremist’s Advantage in Civil Wars 15

cal framework for the role of ideology in civil war that serves as a blueprint for
a more ambitious research program on the subject.

The article deªnes ideology, in its simplest form, as a set of beliefs about
the proper order of society and how it can be achieved.26 It is, according to
Anthony Downs, “a verbal image of the good society and of the chief means of
constructing such a society.”27 All rebel groups pursuing political violence em-
brace some form of ideology, whether implicit or explicit; political contestation
at its heart is the ªght over how a country should be governed. The core ideol-
ogy of communism, for example, is economic equality through the elimination
of private property. The core ideology of Salaª jihadism is a return to a “pure”
state of Islam via the institutionalization of the caliphate.28

Extremism is trickier to deªne.29 If one deªnes a group as “extremist” based
on whether it uses violence to achieve its goals, as many people do,30 then this
deªnition encompasses every rebel group in every civil war that has ever been
fought. Such a deªnition is unhelpful in explaining variation in the level of ex-
tremism across groups ªghting the same civil war. Both ISIS and the Free

Comparison,” American Political Science Review, Vol. 105, No. 3 (August 2011), pp. 478–495, https://
doi.org/10.1017/S0003055411000207.
26. See especially Robert S. Erikson and Kent L. Tedin, American Public Opinion: Its Origins, Con-
tent, and Impact (New York: Pearson Longman, 2007). This parallels the deªnition of ideology fre-
quently found in both psychology and political science. For the former, see William Dember,
“Motivation and the Cognitive Revolution,” American Psychologist, Vol. 29, No. 3 (December 1974),
pp. 161–168; and Phillip E. Tetlock, “Cognitive Style and Political Ideology,” Journal of Personality
and Social Psychology, Vol. 45, No. 1 (1983), pp. 118–126. For the latter, see Daniel Bell, The End of Ide-
ology: On the Exhaustion of Political Ideas in the Fifties (Glencoe, Ill.: Free Press, 1960); Pamela
Johnston Conover and Stanley Feldman, “The Origins and Meaning of Liberal/Conservative Self-
Identiªcations,” American Journal of Political Science, Vol. 25, No. 4 (November 1981), pp. 617–645;
Robert E. Lane, Political Ideology: Why the American Common Man Believes What He Does (New York:
Free Press, 1962); and Herbert McClosky and John Zaller, The American Ethos: Public Attitudes to-
ward Capitalism and Democracy (Cambridge, Mass.: Harvard University Press, 1984).
27. Anthony Downs, An Economic Theory of Democracy (New York: Harper and Row, 1957), p. 86.
28. For in-depth reading on the core ideology of Salaª jihadism, including the role that militant ji-
had plays in its underpinnings, see Moghadam, “Motives for Martyrdom”; Nelly Lahoud, The
Jihadis’ Path to Self-Destruction (London: Hurst, 2010); and Shiraz Maher, Salaª-Jihadism: The History
of an Idea (London: Hurst, 2016).
29. In fact, few datasets on terrorism have any coding of groups’ ideologies. See Ivan Sascha
Sheehan, “Assessing and Comparing Data Sources for Terrorism Research,” in Cynthia Lum and
Leslie W. Kennedy, eds., Evidence-Based Counterterrorism Policy (New York: Springer Science and
Business Media, 2012), pp. 13–40, at p. 30. The Terrorist Organization Proªles database provides
the most descriptive information about the goals of terrorist organizations, but only classiªes
them by category: anarchist, anti-globalization, communist/socialist, environmental, left-wing,
nationalist/separatist, racist, religious, right-wing conservative, right-wing reactionary, and other.
See Victor Asal and R. Karl Rethemeyer “The Nature of the Beast: Organizational Structures and
the Lethality of Terrorist Attacks,” Journal of Politics, Vol. 70, No. 2 (April 2008), pp. 437–449,
doi:10.1017/S0022381608080419. The Terrorism in Western Europe Events Data Set, http://folk
.uib.no/sspje/tweed.htm, has a similar set of categories.
30. See Eric Hoffer, The True Believer: Thoughts on the Nature of Mass Movements (New York: Harper
Collins, 1951), p. xi.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

International Security 42:2 16

Syrian Army, for example, are rebel groups involved in the Syrian civil war,
yet they are not comparable in their level of ideological extremism. The chal-
lenge for scholars is to formulate a measure of “extremism” that does not rely
on speciªc modes of conduct or on value judgments made by the researcher
about the radical nature of a group’s goals and ideas.

This article deªnes ideological extremism in relation to the majority opinion
of the affected population on a key ideological dimension.31 An important di-
mension in the Muslim world is the proper interpretation of the role of Islam
in society, with the policy space ranging from more conservative and strict to
more liberal and tolerant. The majority position on the role of Islam in Muslim
society is consistent around the world. According to a survey of 38,000
Muslims in thirty-nine countries by the Pew Research Center, most Muslims
(including those in the Middle East and North Africa) favor both democracy
and a prominent role for Islam in the political life of their country.32 Clear ma-
jorities of Muslims around the world also strongly reject violence in the name
of Islam.33 This ªnding extends even to Iraq, where 91 percent of Muslims in
2013 said that suicide bombing was not justiªed, even during war.34 These ma-
jority opinions offer a baseline for what “moderate” and “extreme” mean in
the Muslim world. Groups such as the Muslim Brotherhood that believe
that modern democratic institutions are compatible with Islam and sup-
therefore, would be classiªed as “moderate
port democratic processes,
Islamists.”35 Groups such as the Salaªsts, which reject democracy and adhere
to a very narrow and conservative version of Islamic law, would be classiªed
as “extreme Islamists.”

In a competitive environment, where groups are ªghting over the deªnition
of an ideal government, choosing an ideology on the extreme end of societal
preferences seems destined to fail. By deªnition, the number of citizens hold-
ing these extreme views is likely insufªcient to ensure that these movements
prevail over more moderate ones. The rise of Salaª jihadists deªes this logic.

31. I am indebted to Andrew Kydd for this deªnition.
32. Pew Global Attitudes Project, “Views of a Changing World, June 2003” (Washington, D.C.:
Pew Research Center, June 3, 2003), p. 1; and Pew Global Attitudes Project, “Most Muslims Want
Democracy, Personal Freedoms, and Islam in Political Life” (Washington, D.C.: Pew Research Cen-
ter, July 10, 2012, pp. 15–16.
33. Pew Forum on Religion and Public Life, “The World’s Muslims: Religion, Politics, and Soci-
ety” (Washington, D.C.: Pew Research Center, April 30, 2013), p. 13.
34. Ibid., p. 14. Note that no polling data exist for Syria.
35. This characterization of the Muslim Brotherhood may be too generous and optimistic. Note
that not all branches of the Muslim Brotherhood support the democratic process to the same de-
gree. Some, such as the Muslim Brotherhood in Egypt, are certainly “moderate” by this deªnition.
Others, such as Hamas (which is the Muslim Brotherhood in Gaza), are not. For a more detailed
examination of this issue, see Marc Lynch, “Islam Divided between Salaª-Jihad and the Ikhwan,”
Studies in Conºict and Terrorism, Vol. 31, No. 6 (2010), pp. 467–487, doi:10.1080/10576101003752622.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

The Extremist’s Advantage in Civil Wars 17

Groups such as ISIS, Jabhat Fateh al-Sham, al-Qaida, al-Shabaab, and Boko
Haram have all succeeded in outcompeting other groups to conquer and con-
trol territory, an achievement that requires at least the implicit support of the
local population. It is the extreme nature of their beliefs relative to the majority
opinion that makes the growth and survival of these groups so perplexing.

The Strategic Use of Extreme Ideology in Civil War

This article offers a theory for the rise of violent extremist groups that re-
veals the conditions under which individuals will have strategic incentives to
behave as if they believe in an extreme ideology even if they do not. Assume
that there are three main types of actors in a civil war: political entrepreneurs
who are ªghting to gain political power, moderate citizens who must deter-
mine which of the competing political elites to support, and ideologues who
are devoted to a particular cause.36 The goal of political entrepreneurs is to win
the war and gain power. The goal of moderate citizens (beyond basic survival)
is to align with those political elites most likely to win the war and institute
policies most closely associated with their own political preferences. The goal
of ideologues is to establish their perfect vision of society. To win the war, po-
litical entrepreneurs must gain the support of at least some of the moderate cit-
izens who represent the majority of the population; the more soldiers and
sympathizers a group is able to attract, the more likely it is to win the war.

Political entrepreneurs who seek to build and sustain a rebel organization
confront three organizational challenges that, under certain conditions, may
create strong incentives for them to embrace an extreme ideology. First, rebel
entrepreneurs often face a substantial collective-action problem—what Mark
Lichbach called the “rebel’s dilemma.”37 To be successful, rebel entrepreneurs
must ªnd a way to recruit individuals willing to ªght and die for a cause even
when the beneªts of victory will be shared with those who do neither. Sec-
ond, rebel entrepreneurs face a classic principal-agent problem. To be success-
ful, rebel elites must be able to recognize and attract soldiers who will remain
trustworthy and loyal once deployed to the ªeld.38 Third, rebel entrepreneurs
face a commitment problem that creates incentives for soldiers, donors, and
rank-and-ªle sympathizers to defect from the organization unless leaders can
reassure them that future spoils will be shared.

Moderate citizens also face at least two challenges when seeking to deter-

36. Note that these political entrepreneurs can be rebel or government elites.
37. Mark Irving Lichbach, The Rebel’s Dilemma (Ann Arbor: University of Michigan Press, 1995).
38. See Gates, “Recruitment and Allegiance”; Weinstein, Inside Rebellion; Jacob N. Shapiro, The Ter-
rorist’s Dilemma: Managing Violent Covert Organizations (Princeton, N.J.: Princeton University Press,
2013); and Sanín and Wood, “Ideology in Civil War.”

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

International Security 42:2 18

mine which party to support in the midst of war. These challenges could create
incentives for them to embrace a more extreme group even if their preferences
are more moderate. The ªrst is related to the outcome of war. All else equal,
citizens would prefer to support the group that is most likely to win the war,
because doing so would shield them from punishment related to backing the
opposition. The second is related to the quality of governance. Citizens would
also prefer to back the rebel group that is most likely to institute reforms, even
if its ideology is not exactly aligned with theirs. In weakly institutionalized set-
tings, citizens understand that rebel leaders have incentives to abandon costly
political reforms once they are safely in power. The challenge is to determine
as early as possible which political elites are more likely to resist corruption
once in ofªce.

These two challenges encourage moderate citizens to try to ascertain which
of the competing rebel groups could potentially win the war and which is
more likely to be trustworthy over time. Information about the relative
strength of a group is fairly easy to gather; citizens can observe how groups
perform on the battleªeld. The more battles a group wins, the more attractive
it becomes. Information about the trustworthiness of elites is more difªcult to
obtain; all rebel elites have an incentive to claim that they will institute reforms
once in power, even if they know they will not. The desire to back the winning
faction and the one least likely to sell out once in power could lead moderate
citizens to shift their support in favor of more extreme groups. An extremist
group that institutes at least some reforms is preferred to a moderate group
that institutes none.

In what follows, I explore three organizational hurdles that elites face in the
context of civil war, explain how an extreme ideology could be an effective tool
for solving them, and reveal why moderates might gravitate toward these
groups even if their policy preferences are less radical.

extreme ideology and collective-action problems

One of the ªrst organizational hurdles that rebel entrepreneurs face when at-
tempting to build a ªghting force has to do with collective-action problems.
Most citizens would prefer to reap the rewards of living under their preferred
political system without having to pay the costs of ªghting to install it. To be
successful, however, rebel elites must convince some individuals to ªght for
them, a task that is especially hard in the nascent phase of a movement.39

Theoretically, rebel elites can reduce collective-action problems by pursu-

39. See Lichbach, The Rebel’s Dilemma.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

The Extremist’s Advantage in Civil Wars 19

ing two possible strategies. First, they can change the payoffs that recruits
receive for volunteering for battle, offering them inducements to join.40 Sol-
diers can be paid a high salary or provided education and health care beneªts.
They can be granted permission to supplement their salaries through loot-
ing, taxing, or exploiting conquered territories.41 Soldiers can also be given
things as simple as food, housing, or security. One of the attractions of the
Taliban in Afghanistan, for example, was that it offered citizens protection
from the arbitrary violence of warlords.42 Each of these inducements serves to
increase the private beneªts that a person enjoys by supporting an organiza-
tion, enhancing incentives to ªght and reducing incentives to free-ride.

A second strategy that elites could pursue is one that would impose punish-
ments for nonparticipation in a movement.43 Rebel leaders can inºict heavy
costs on individuals who choose to support a competitor or remain unin-
volved in a war.44 Such action could include the forced expropriation of an
individual’s property or the destruction of a family’s home. It could also
include the targeted assassination of individuals who join the enemy or
refuse to ªght against it. According to one member of Boko Haram, “I ofªcially
joined them when they started killing indiscriminately. I needed protection
and immunity from persecution by them so I could continue with my
business.”45 Collective-action problems disappear when the costs of nonpartic-
ipation are high.46

Both moderate and extreme rebel groups have the ability to offer in-
ducements and inºict punishments to solve the collective-action problem
associated with recruitment. An advantage, however, will accrue to groups
able to craft rewards and punishments that do not require a signiªcant outlay
of resources. The lower the economic and material costs of attracting a steady
stream of recruits, the more resources that can be devoted to the war effort,
and the more battleªeld success a group is likely to have.

40. Samuel L. Popkin, The Rational Peasant: The Political Economy of Rural Society in Vietnam (Berke-
ley: University of California Press, 1979); and Lichbach, The Rebel’s Dilemma.
41. Elisabeth Jean Wood, Insurgent Collective Action and Civil War in El Salvador (Cambridge: Cam-
bridge University Press, 2003); and Gates, “Recruitment and Allegiance.”
42. Iannacone and Berman, “Religious Extremism,” p. 120.
43. See Kalyvas and Kocher, “Ethnic Cleavages and Irregular War.”
44. For an excellent discussion about the organizational costs involved in attempting to coerce
rebel recruitment, see Kristine Eck, “Coercion in Rebel Recruitment,” Security Studies, Vol. 23,
No. 2 (2014), pp. 364–398, doi:10.1080/09636412.2014.905368.
45. Charlotte Alfred, “Why People Join Nigeria’s Boko Haram: Former Members of the Militant
Group Offer Some Answers,” Hufªngton Post, January 7, 2017, http://www.hufªngtonpost.com/
entry/boko-haram-recruitment-tactics_us_571265afe4b06f35cb6fc595. This article was based on a
report by Mercy Corps, a U.S.-based aid agency that interviewed forty-seven former members of
Boko Haram regarding the reasons why they joined the group.
46. Kalyvas and Kocher, “Ethnic Cleavages and Irregular War.”

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

International Security 42:2 20

An extreme ideology based on religion is uniquely positioned to compete in
this regard in at least four ways. First, religious organizations have the ability
to offer costless deferred compensation in the form of promises of rewards
in the afterlife. Second, many religions already include dictates that rebel lead-
ers can use to enforce behavior in ways that beneªt the organization. Salaª
jihadists, for example, have invoked Islam’s principle of takªr as license to kill
dissenters and defectors of their movement.47 In this way, religion could make
it easier for a rebel group to engage in coercion in order to attract recruits.
Third, religious leaders are often viewed as having greater moral authority
and legitimacy than ordinary citizens, making it easier for them to dispense
punishments without incurring civilian backlash.48 The ruling royal family of
Saudi Arabia aligned with Wahhabism for exactly this reason; the alliance
gave the monarchy the status of spiritual and political heads of the land, gain-
ing the deference of Saudi citizens. Fourth, religious organizations have the
unique ability to inºict a potentially devastating type of personal punishment:
eternal damnation in the case of Christianity and excommunication in the case
of Islam. Not only is this punishment costless to enforce, but it is also im-
possible for targets to escape, as punishment follows an individual even if
he or she chooses to switch sides or defect from the battleªeld. Religion, thus,
could create the opportunity to offer cheap, deferred inducements and punish-
ments, enabling rebel leaders to potentially outcompete less extreme or more
secular organizations. It also allows them to enter a war late, because they can
undercut competitors by producing rewards and punishments more cheaply.
An important outcome of these features is that they are likely to attract a dis-
proportionate share of ideologues to a group, especially in the early phases of
mobilization. Rebel groups with a cadre of highly committed, dogged ªghters
not only win battles, but gain a reputation for disciplined, high-quality
ªghting units. This reputation then serves to attract more moderate citizens
who are drawn to the group because of its dominance on the battleªeld.49 Ac-
cording to William McCants, the success of ISIS convinced many former
ªghters of the Free Syrian Army to defect for what many soldiers felt was the
better-funded and better-organized jihadist group.50 Rebel entrepreneurs who

47. It is important to note that takªr is debated even among Salaª jihadists and radical Islamists.
See Jessica Stern and J. M. Berger, ISIS: The State of Terror (New York: Ecco, 2016).
48. Max Weber, “The Three Types of Legitimate Rule,” Hans Gerth, trans., Berkeley Publications in
Society and Institutions, Vol. 4, No.1 (1958), pp. 1–11; Juergensmeyer, Terror in the Mind of God; and
Ron E. Hassner, War on Sacred Grounds (Ithaca, N.Y.: Cornell University Press, 2009).
49. Fotini Christia argues that rebel leaders choose alliances in part based on which factions they
believe are likely to win a war. See Christia, Alliance Formation in Civil Wars (Cambridge: Cam-
bridge University Press, 2012).
50. McCants, The ISIS Apocalypse, p. 115.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

The Extremist’s Advantage in Civil Wars 21

embrace an extreme ideology, therefore, could gain an early competitive ad-
vantage over more moderate groups in recruiting the very best ªghters, which
then translates into an advantage in recruiting less ideologically driven sol-
diers over time. Abu Bakr al-Baghdadi and other ISIS leaders appeared to have
understood this two-pronged effect. In an attempt to attract both zealots and
moderates, ISIS implemented its dictates and punishments inconsistently.
When popular support waned, the leadership often became more lenient to-
ward the local population. According to McCants, “[B]eing so strict was good
for impressing puritans, but it wasn’t terribly crowd pleasing.”51

extreme ideology and principal-agent problems

Rebel entrepreneurs also face two potential principal-agent problems that an
extreme ideology could help solve. The ªrst stems from the uncertainty rebel
leaders have regarding the dedication and reliability of soldiers volunteering
for service.52 The second stems from the uncertainty that recruits have regard-
ing the sincerity and devotion of rebel elites.

dedication and reliability of recruits. One of the challenges that rebel
entrepreneurs face is that they cannot identify ex ante which volunteers will
remain loyal to the organization over time. They must accept or reject soldiers
based on incomplete information on how individuals are likely to perform in
the ªeld once free of elite monitoring. This lack of information is potentially
problematic because soldiers (1) do not always have the same preferences as
their leaders, (2) could behave in ways that hurt the organization and its elites,
and (3) are difªcult to monitor once deployed.53

Rebel leaders understand that not all soldiers join rebel movements for
equally honorable reasons. Some join because they are truly committed to the
organization and its goals; others join for more self-interested reasons such as
adventure, revenge, or camaraderie.54 To simplify, assume that two types of
rebel recruits exist: those who are dedicated to the organization and those who
are not. Dedicated recruits are likely to remain loyal to the organization, be re-
liable, and ªght hard. Less dedicated recruits are more likely to perform
poorly, be lazy, exhibit predatory behavior, and betray the movement and its
leaders to the enemy. Potential recruits know whether they are dependable.
Rebel leaders do not.

One of the keys to a successful organization is the ability of rebel leaders to

51. Ibid., p. 139
52. This problem is central in Weinstein’s examination of civilian abuse during civil war. See
Weinstein, Inside Rebellion.
53. See ibid.
54. Kalyvas and Kocher, “Ethnic Cleavages and Irregular War.”

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

International Security 42:2 22

distinguish between dedicated and undedicated recruits before they are incor-
porated into the organization where they can inºict harm. One way to do this
is for rebel leaders to demand a costly signal from potential recruits.55 Costly
signals can come in many forms. Leaders can demand that potential recruits
go through a rigorous training period or “boot camp.” They can require great
sacriªce—for example, renouncing one’s former country or making a public
pledge to the opposition, which then marks them as traitors to the govern-
ment. Organizations could require new recruits to commit an egregious act
such as rape or a gratuitous killing.56

Once again, an extreme ideology could help determine which volunteers are
likely to be committed to the organization and which are not. The costly sig-
nals described above all require that the organization expend resources. Boot
camps demand money and personnel and are an inefªcient way to determine
commitment because only a certain percentage of recruits are able to complete
the program. Public denouncements put recruits at risk of government and
community reprisals, reducing the numbers of dedicated recruits ultimately
available to ªght. Rape and gratuitous killing may be a costly signal of com-
mitment, but these actions alienate the local population, making local opera-
tions and occupation more difªcult.

The most advantageous signals are those in which the costs are borne
entirely by recruits and not rebel leaders. An extreme ideology could be such a
signal. By forcing volunteers to pay high upfront costs in the form of socially
alienating, group-entrapping practices, an extremist group ensures that it has
chosen individuals who are more likely to remain true over time.57 Salaª
jihadists, for example, must renounce friends and family who do not embrace
Salaª jihadism. They must then follow a strict interpretation of Islamic doc-
trine that often levies harsh, nonnegotiable punishments for even slight indis-

55. For a detailed discussion on the many ways that terrorist organizations attempt to screen out
less committed recruits, see Shapiro, The Terrorist’s Dilemma, especially pp. 54–56.
56. Mohammad Hafez and Quintan Wiktorowicz argue the perpetration of atrocities reduces the
chances that recruits have to defect because, as a result of their actions, they are less likely to be ac-
cepted back into mainstream society. See Hafez and Wiktorowicz, “Violence as Contention in the
Egyptian Islamic Movement,” in Wiktorowicz, ed., Islamic Activism: A Social Movements Theory Ap-
proach (Bloomington: Indiana University Press, 2004), pp. 61–88. See also Thomas Hegghammer,
“Should I Stay or Should I Go? Explaining Variation in Western Jihadists’ Choices between Do-
mestic and Foreign Fighting,” American Political Science Review, Vol. 107, No. 1 (February 2013),
pp. 1–15, doi:10.1017/S0003055412000615; Dara K. Cohen, “Explaining Rape during Civil War:
Cross-national Evidence (1980–2009),” American Political Science Review, Vol. 107, No. 3 (August
2013), pp. 461–477, doi:10.1017/S0003055413000221; and Dara K. Cohen, Rape during Civil War
(Ithaca, N.Y.: Cornell University Press, 2016).
57. See Eli Berman, “Hamas, Taliban, and the Jewish Underground: An Economist’s View of Radi-
cal Religious Militias,” National Bureau of Economic Research Working Paper No. 10004 (Cam-
bridge, Mass.: National Bureau of Economic Research, September 2003), doi:10.3386/w10004.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

The Extremist’s Advantage in Civil Wars 23

cretions. Jabhat al-Nusra went one step further, requiring all new recruits
to immediately join the front line, where they were “tested for bravery,
dedication, and loyalty to the organization.”58 Individuals who intend to
shirk their duties or subvert the organization are less likely to join under
these conditions.59

In addition, an extreme ideology naturally appeals to ideologues in the
population—individuals at the extreme end of the preference distribution who
have more to gain by ªghting.60 Because the difference between winning and
losing is larger for individuals with extreme preferences than for moderates,
these individuals are willing to absorb heavier costs to attain victory.61 More
committed recruits, therefore, naturally gravitate toward more extreme ideo-
logical groups, creating a helpful separating equilibrium.

Finally, Jacob Shapiro argues that religious groups may have an additional
advantage screening potential recruits given their large social networks.
Hamas, for example, was better able to identify, monitor, and control its mem-
bers as a result of its rich array of social services than was its secular competi-
tor, Fatah.62

trustworthiness of rebel leaders. Rebel

leaders, however, may also
face a reverse principal-agent problem. High-quality, dependable soldiers will
attempt to identify which of the competing rebel factions have equally honor-
able leaders. Highly dedicated soldiers understand that rebel leaders have in-
centives to claim they are dedicated to political change even if they are really
motivated by ambition or greed. How do these citizens determine which lead-
ers to back and which to avoid?

Dedicated and determined citizens can identify more dedicated leaders via a
basic signaling game. Assume that there are two types of rebel leaders: those
who believe in the goals and principles of the group and those who do not.
Rebel leaders know their type; citizens do not. Elites can claim that they are
devoted to a particular set of goals, but because all leaders have incentives to
present themselves this way, these claims are the equivalent of cheap talk. One
opposition leader in Syria, for example, bluntly admitted that many Free
Syrian Army commanders lacked discipline and that the question every Syrian

58. Norman Benotman and Roisin Blake, “Jabhat al-Nusra: A Strategic Brieªng” (Lodon: Quilliam
Foundation, n.d.).
59. For an excellent discussion of agency problems that can arise during war, see Peter Schram,
“Managing Insurgency,” Ph.D. dissertation, Stanford University, 2017.
60. For a deeper discussion of the beneªts of selecting individuals who are autonomously moti-
vated, see Robert D. Petersen, Resistance and Rebellion: Lessons from Eastern Europe (Cambridge:
Cambridge University Press, 2001).
61. Kydd and Walter, “Sabotaging the Peace.”
62. Shapiro, The Terrorist’s Dilemma, p. 248.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

International Security 42:2 24

had for the opposition was, “Are you going to bring chaos or order?”63 Deter-
mining which leaders are truly devoted to reform becomes a critical task not
only for the most fervent supporters but for moderate supporters as well.

Rebel leaders who are able to devise costly signals of their dedication are
likely to attract a greater number of high-quality soldiers needed to win a war.
Rebel leaders have at least two ways to do this. The ªrst is to choose a set of
ideals on the extreme end of the ideological spectrum.64 Leaders who pick an
extreme ideology suggest that they are in it not just for the power or self-
enrichment but because they truly care about political change. The second is to
choose an ideology that demands personal sacriªce from the leaders, some-
thing that religious extremism often requires.65 Strict guidelines on elite behav-
ior signal to potential followers that the leaders may be equally committed to a
cause. Osama bin Laden’s decision to embrace Salaªsm imposed signiªcant
behavioral and ªnancial constraints that allowed him to distinguish himself
from many of his competitors.

extreme ideology and commitment problems

Rebel leaders face one ªnal organizational hurdle. Even if they are able to
solve collective-action problems (by offering inducements or punishments
to individuals to ªght), and even when they are able to solve principal-agent
problems (by identifying dedicated supporters and creating costly signals of
their own devotion), they may still have difªculty maintaining the long-term
support of citizens because of commitment problems.

Potential supporters of rebel groups know that rebel leaders, once in power,
have incentives to use their ofªce for personal enrichment even if they had no
intention of doing so during the war.66 That is because the payoffs for central-
izing power and stealing from the state radically change once a group wins the
war, especially in weakly institutionalized settings.67 Rebel leaders who are
best able to commit to good governance in the future, therefore, are likely to
gain additional moderate support during the war.68

63. David Ignatius, “How ISIS Spread in Syria and Iraq—And How to Stop It,” Atlantic, October,
2015, https://www.theatlantic.com/international/archive/2015/10/how-isis-started-syria-iraq/
412042/.
64. See Andrew Kydd, “Polarization and the Cost of Participation,” University of Wisconsin,
Madison, 2016.
65. Patrick Cockburn, Muqtada Al-Sadr and the Fall of Iraq (London: Faber and Faber, 2008).
66. Barbara F. Walter, “Bargaining Failures and Civil War,” Annual Review of Political Science,
Vol. 12 (2009), pp. 243–261, doi:10.1146/annurev.polisci.10.101405.135301.
67. Barbara F. Walter, “Why Bad Governance Leads to Repeat Civil War,” Journal of Conºict Resolu-
tion, Vol. 59, No. 7 (2014), pp. 1242–1272, doi:10.1177/0022002714528006.
68. See the literature on noncooperative theory of endogenous coalitions that explores the distri-
butional basis of group formation: Santiago Sánchez-Pagés, “Endogenous Coalition Formation in
Contests,” Review of Economic Design, Vol. 11 (2007), p. 139, doi:10.1007/s10058-007-0033-4;

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

The Extremist’s Advantage in Civil Wars 25

Again, an extreme ideology may have three advantages here. First, groups
on the ideological extreme are likely to have an easier time committing to more
far-reaching reform than moderate groups. In bargaining contexts, it is often
useful to be represented by an agent who is more hard-line.69 Hard-line agents
will reject deals that more moderate agents would accept, forcing an adversary
to make a better offer.70 Palestinians, for example, might prefer Hamas as a ne-
gotiating agent with Israel to Fatah, because Hamas is more likely to reject
mediocre deals.71

Second, extreme ideologies frequently attempt to control all aspects of peo-
ple’s lives, making their behavior more predictable over time.72 In her analysis
of totalitarian movements, for example, Hannah Arendt emphasizes the ways
in which extremist groups seek to eliminate all forms of spontaneous human
action and freedom, reducing the unknowns associated with future behavior.73
The same concept helps violent extremist movements credibly commit to
future reform.

Third, moderate citizens understand that ideologically extreme rebel groups
include true believers who could punish leaders who renege on promises. The
presence of ideologues within the party, therefore, helps leaders credibly com-
mit to at least some accountability after the war has been won.

Religious extremist groups, however, may have an added advantage over
secular groups in terms of committing to good behavior over time. Centralized
religions often have their own infrastructure for justice. Radical Islam, for ex-
ample, comes with its own jurisprudence—a system of laws interpreting
nearly every facet of governance with a wide variety of interpretative forums
from which to draw. The leadership of al-Qaida appeared to understand this
advantage by purposively targeting “regions beset by anarchy, where local
populations would welcome their ability to institute basic governance and
Islamic sharia law.”74 Religion, especially Islam, also has trustworthy third

Michelle R. Garªnkel, “Stable Alliance Formation in Distributional Conºict,” European Journal of
Political Economy, Vol. 20 (2004), pp. 829–852, doi:10.1016/j.ejpoleco.2003.06.002; and Christopher
Blattman and Edward Miguel, “Civil War,” Journal of Economic Literature, Vol. 48, No. 1 (March
2010), pp. 3–57, doi:10.1257/jel.48.1.3.
69. See Andrew H. Kydd and Barbara F. Walter, “The Strategies of Terrorism,” International Secu-
rity, Vol. 31, No. 1 (Summer 2006), pp. 49–80. See also Tanisha M. Fazal, “Rebellion, War Aims, and
the Laws of War,” Daedalus, Vol. 146, No. 1 (Winter 2017), pp. 71–82, doi:10.1162/DAED_a_00423.
70. Abhinay Muthoo, Bargaining Theory with Applications (Cambridge: Cambridge University
Press, 1999), p. 230.
71. See Mia M. Bloom, “Palestinian Suicide Bombing: Public Support, Market Share, and Outbid-
ding,” Political Science Quarterly, Vol. 119, No. 1 (Spring 2004), pp. 61–88, doi:10.2307/20202305.
72. Hannah Arendt, The Origins of Totalitarianism (New York: Harcourt, Brace, and Jovanovich,
1973).
73. Ibid., p. 466.
74. From Abu Bakr Najj, “The Management of Savagery: The Most Critical Stage through Which
the Islamic Nation Will Pass,” trans. William McCants (Cambridge, Mass.: John M. Olin Institute

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

International Security 42:2 26

parties willing to step in to arbitrate disputes and punish leaders—imams,
muftis, and ayatollahs are just some examples.75 Their presence could play an
important role in allowing Islamist rebel leaders to outcompete more moderate
groups on this dimension.

conditions conducive to extremism

The theory outlined above explains how an extreme ideology may help rebel
entrepreneurs overcome the three organizational hurdles discussed above and,
in the process, recruit and retain more soldiers and sympathizers. It also ex-
plains the beneªts that an extreme ideology may offer to more moderate citi-
zens if they believe it will help a group win a war and resist corruption. The
theory does not, however, explain the conditions under which embracing an
extreme ideology is likely to be more or less advantageous to both groups. In
what follows, I argue that three conditions are likely to produce particularly
severe collective-action, principal-agent, and commitment problems and that it
is these conditions that favor the emergence and success of extremist groups.
These conditions—periods of rapid political change, intense competition, and
limited constraints on executive power—frequently occur in countries with
authoritarian regimes in the throes of civil war. These conditions also exist in
Muslim countries experiencing civil war today.

what explains the choice of ideology?

Elites seeking to change the political status quo cannot pick just any ideology.
Those who hope to have some chance to attract recruits are limited to political
ideas that resonate with some portion of the population they hope to govern.76
Populations will have preexisting ideas about how things “ought to be” and
what an “ideal” government might look like.77 These ideas are likely to be
inºuenced, in part, by historical lines of political contestation, which often par-
allel ethnic, religious, sectarian, or class divisions.78 The choice of ideology,

for Strategic Studies at Harvard University, 2004), https://azelin.ªles.wordpress.com/2010/08/
abu-bakr-naji-the-management-of-savagery-the-most-critical-stage-through-which-the-umma-
will-pass.pdf.
75. Note, however, that these can be the rebel leaders themselves (or someone close to them).
76. Sanín and Wood emphasize the importance of what they call a “normative commitment to an
ideology.” See Sanín and Wood, “Ideology in Civil War,” p. 214.
77. See Paul Staniland, Networks of Rebellion: Explaining Insurgent Cohesion and Collapse (Ithaca,
N.Y.: Cornell University Press, 2014); Robert D. Benford and David A. Snow, “Framing Processes
and Social Movements: An Overview and Assessment,” Annual Review of Sociology, Vol. 26 (2000),
pp. 611–639, doi:10.1146/annurev.soc.26.1.611; and Maurice Duverger, Political Parties: Their Orga-
nization and Activity in the Modern State, Barbara and Robert North, trans. (New York: John Wiley
and Sons, 1954).
78. See John Levi Martin, “What Is Ideology?” Sociologia, Problemas E Práticas, No. 77 (2015), pp. 9–
31, doi:10.7458/SPP2015776220.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

The Extremist’s Advantage in Civil Wars 27

therefore, will depend on these preexisting lines of contestation; this is the ele-
ment of ideology that is exogenous to the war itself and does not change rap-
idly over time.

Every ideology, however, falls along a continuum from moderate to extreme.
Rebel entrepreneurs, therefore, must also choose where along this continuum
to place their movement. This is where the endogenous civil war environment
comes into play. Three contextual factors are likely to create incentives to move
to the ideological extreme. The ªrst involves the level of competition in a given
war. Collective-action problems are likely to be particularly severe when mul-
tiple rebel groups exist, all of which seek to appeal to the same subset of the
population. After Saddam Hussein was deposed in 2003, multiple insurgent
forces emerged within both the Sunni and Shia populations to ªght for control
of Iraq. The greater the number of rebel factions seeking to represent the same
population, the more intense the collective-action problems, and the greater
the inducements and punishments leaders will need to offer to compete.79

The second factor relates to the level of uncertainty that exists in a civil war.
Principal-agent problems are likely to be particularly problematic (and an ex-
tremist ideology particularly helpful) when rebel leaders and potential recruits
have limited information about each other. This situation is likely to occur
when a rebel group and its leaders are unknown to the local population, either
because they originated outside the country or because they are operating far
away from the individuals they seek to recruit. In both situations, rebel entre-
preneurs must draw from a population with whom they have had limited con-
tact and where local information networks are likely to be weak. Not only do
leaders have less knowledge about the trustworthiness of individuals who
volunteer for their organization, but local citizens have equally limited knowl-
edge about which of the many emerging faction leaders is particularly well
qualiªed to ªght and win a war. Costly signals of dedication and trustworthi-
ness from both leaders and recruits are likely to be particularly valuable under
these circumstances.

The third factor is related to the level of institutional constraints on exec-
utive power. Commitment problems are likely to be acute for rebel entrepre-
neurs in countries where citizens know there is limited rule of law, few checks
on government power, and endemic corruption. These are the civil wars in
which citizens are likely to demand stronger commitments to good behavior
over time in exchange for their support. An ideology that provides its own

79. This is the basic logic behind the strategy of outbidding discussed in Kydd and Walter, “The
Strategies of Terrorism”; and Bloom, “Palestinian Suicide Bombing.”

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

International Security 42:2 28

strict rules and laws, and its own mechanisms for justice, becomes quite useful
under these circumstances.

These three conditions suggest that extremism will not emerge and thrive in
all civil wars. Rebel groups that face limited or no competition for local sup-
port, that have deep local networks from which they can draw ªghters, and
that operate in a country with a strong rule-of-law tradition have far fewer in-
centives to move to the ideological extreme. These are the cases where the local
population has little need for especially costly signals or credible commit-
ments, and where they are more likely to gravitate toward moderate groups
whose ideologies better match their own.

Why Salaª Jihadism?

Seen in this light, the theory offers insights into the rise of Salaª jihadist
groups in the post-2003 civil wars. According to the theory, when faced with
the opportunity to compete for power, rebel entrepreneurs ªrst choose ideolo-
gies likely to appeal to the needs and desires of a particular population. They
then determine the group’s level of extremism based in part on how severe
collective-action, principal-agent, and commitment problems are likely to be.
Although not deªnitive, my theory suggests that the emergence of so many
Islamist groups in recent years appears to follow this pattern. Political
contestation in the Middle East began in earnest in 2003 with the fall of
Saddam Hussein and expanded with the Arab Spring protests of 2010–11.
These two events undermined the post–World War II imperial order and cre-
ated an opening for political entrepreneurs to assert new visions for govern-
ment and society. Not surprisingly, every rebel group that subsequently
formed in the region embraced political Islam (“Islamism”) as its main ideol-
ogy.80 As discussed above, large majorities of citizens in the Muslim world fa-
vor a prominent role for Islam in the political life of their countries, and
political identity in the Muslim world has long been vested in religion rather
than ethnicity.81 In addition, an Islamic form of government served as an
attractive counterpoint to the corrupt, mainly secular regimes that had trig-
gered the protests.

The question, however, is why so many groups chose to embrace a fringe
form of Islamism rather than a more moderate one. According to the theory,

80. Michael Cook, Ancient Religions, Modern Politics: The Islamic Case in Comparative Perspective
(Princeton, N.J.: Princeton University Press, 2014).
81. Pew Global Attitudes Project, “Views of a Changing World, June 2003,” p. 1; and Cook, Ancient
Religions, Modern Politics.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

The Extremist’s Advantage in Civil Wars 29

rebel entrepreneurs have incentives to move to the ideological extreme when
competition for new recruits is intense, when uncertainty about the dedication
of rebel leaders and recruits is high, and in countries with weak political insti-
tutions and a history of bad governance. All three of these conditions existed
in the Arab world when civil wars broke out.

competition and the rise of salaª jihadism

The ªrst distinguishing feature of the current wave of civil wars is the large
number of Sunni factions ªghting in each one. Four major Sunni factions are
currently ªghting in Syria, three in Iraq, three in Libya, two in Yemen, and
three in Somalia (this does not include the many smaller Sunni factions that
also exist).82 Sunni factions proliferated in these wars for at least three reasons.
First, Sunnis represented the largest sectarian group in all but one of these
countries (Iraq) and the largest group excluded from power in Syria and Iraq.
Sunni citizens, therefore, had the greatest incentive to ªght for political
change. Second, Sunnis represent between 85 and 90 percent of Muslims
worldwide, creating an exceptionally large pool of potential recruits from
which rebel groups could draw. Third, Sunni-based rebel groups had the
added opportunity to attract ªnancing from wealthy Persian Gulf countries
keen to install Sunni-led governments throughout the Middle East. Thus, the
combination of large local Sunni populations, even larger global Sunni popula-
tions, and wealthy Sunni patrons made it possible for multiple political entre-
preneurs to attempt to form their own separate factions.

The proliferation of Sunni rebel factions created heavy competition and,
thus, incentives for some of them to embrace an extreme version of Sunni
Islamism—Salaª jihadism. The usual collective-action problem associated
with recruiting soldiers was exacerbated by the need to outcompete rival
groups, making relatively costless inducements and punishments particularly
important. Salaª jihadism’s veneration of the “martyr” and its aggressive use
of takªr were tools that less extreme groups did not have when vying for the
same set of Sunni citizens.83

82. These groups include prominent actors such as ISIS branches in three countries; JRTN and the
Mujahideen Army in Iraq; Hay’at Tahrir al-Sham and Ahrar al-Sham in Syria; and al-Shabaab,
Ahlu Sunna Wal Jama, and al-Qaida in Somalia. Mapping Militant Organizations Project, Stanford
University, http://web.stanford.edu/group/mappingmilitants/cgi-bin/. Note that this project
only included prominent rebel groups in each conºict. To be classiªed as a rebel group, an organi-
zation had to meet three criteria. It had to be (1) a militant organization (with militant organiza-
tions deªned as “the most relevant and signiªcant groups for a given conºict theatre”), (2) actively
opposed to the state for at least one month within the time period covered by this dataset, and
(3) militarily or politically active within the time frame covered by this dataset.
83. See Moghadam, “Motives for Martyrdom.” The Free Syria Army and other non–Salaª jihadist

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

International Security 42:2 30

So, why did citizens gravitate toward the most extreme of the extreme rebel
groups (ISIS or Jabhat al-Nusra) when they could have opted for slightly
less extreme groups such as Ahrar al-Sham, the Jaysh al-Fath coalition, or
Jaysh al-Islam? Citizens could have gained similar beneªts and reassurances
from these competing groups, yet opted for the most extreme ideological
choice. The reason, I argue, has to do with timing. ISIS and al-Qaida solved the
collective-action problem earlier than these other groups, drew from a ready-
made and more devoted external network of followers willing to grant ªnan-
cial and material support, and more rapidly coordinated effective strikes
against the state. Thus, these groups appeared more capable at an earlier point
in the war, consolidated power, and distinguished themselves from a ªeld of
similar-looking groups.84

uncertainty and the rise of salaª jihadism

The second distinguishing feature of the post-2003 civil wars was the emer-
gence of rebel groups with weak local connections. ISIS, for example, did not
have deep roots in Iraqi or Syrian society, or in Libya, Mali, Nigeria, Somalia,
or Yemen. In addition, all of the most extreme Salaª jihadist groups—al-Qaida
in the Islamic Maghreb, al-Shabaab, al-Nusra, and ISIS—split from preexisting
movements. These groups were second-movers in their respective civil wars,
forced to play competitive catch-up with existing Sunni factions. Jabhat al-
Nusra, for example, formed ten months after the uprising in Syria began.85 In
highly competitive environments, where little information is known about a
new group, Salaª jihadism can provide a relatively quick and easy way to ap-
peal to the most devoted citizens and potentially hardest ªghters. In addition,
recall that moderate citizens have incentives to support the group that is likely
to win the war. Salaª jihadist groups start out weak but steadily grow stronger
as they are able to recruit these dedicated and strong ªghters. In so doing, they
increase their appeal to more moderate citizens, who gain more information
about the relative strength of competing groups as the war progresses.

weak institutions, bad governance, and the rise of salaª jihadism

A third distinguishing feature of these civil wars was the authoritarian nature
of the incumbent regimes and the absence of the rule of law in each of the af-

groups also celebrated their martyrs, but al-Qaida and ISIS leaders went a step further and pro-
moted martyrdom as an end in itself.
84. Ahrar al-Sham is actually considered to be stronger than ISIS or Jabhat al-Nusra in terms of
operational sophistication, military size, and political inºuence, but it seems to mix strategies,
sometimes leaning toward Salaªsm and at others adopting more moderate stances.
85. David Ignatius, “Al-Qaeda Afªliate Playing Larger Role in Syria Rebellion,” Washington Post,
November 30, 2012.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

The Extremist’s Advantage in Civil Wars 31

fected countries. The weak political restraints on Arab leaders and the extraor-
dinary corruption with which they ruled created severe commitment problems
for rebel leaders claiming to seek to serve the Sunni population. It is no coinci-
dence that an ideology that emphasized morality and justice emerged in a re-
gion that had been dominated by repressive and shockingly bad governments.
Salaªsm promised to represent Islam in all its purity. One of the founda-
tional beliefs of this ideology is that adherents should live like their “pious
predecessors”—those who lived prior to any Western interference or corrup-
tion in the region. The call for a return to early Islamic heritage was a call for a
return to values that have disappeared in the modern era of despotism,
patrimonialism, and bad government.86 ISIS won followers by promising to es-
tablish a theologically authentic state rather than a modern political one. Ac-
cording to one expert, “It is the group’s doctrinally consistent bypassing of
Western political culture that has allowed it to pick up so many recruits.”87

In addition, war destroyed any semblance of law and order, making the
Salaª jihadist groups even more attractive to moderate citizens as violence and
disorder progressed. This helps explain why Salaª jihadists in Iraq and Syria
gained adherents in the midst of war and why they remained militarily rele-
vant much longer than their moderate rivals. War dramatically reduced what
few institutions existed and offered an advantage to those groups who could
best promise and deliver order and justice.

Did the same occur on the Shia side? Although Shia factions proliferated,
especially in the Iraq civil war, these groups did not move to the ideological
extreme to the same degree as did Sunni factions. The same is true of the
Kurds in both Iraq and Syria. There are two reasons for this. First, the competi-
tion among Shia factions was not nearly as intense as it was among Sunni
factions; dramatically fewer Shia factions formed in both the Iraq and Syrian
civil wars. Shia rebel leaders, therefore, did not have the same incentives
to shift to a more extreme version of Islamism because collective-action prob-
lems were not as severe. Second, Shia organizations in Syria and Iraq were
homegrown, unlike their Sunni counterparts, which transcended interna-
tional borders.88 Shia factions, therefore, were not under the same burden to

86. For an excellent discussion of the rise of the military state and morally decadent dictatorial
leaders in Sunni Islam–majority countries, and of the effect this had on the corresponding rise in
Islamic fundamentalism, see Emmanuel Sivan, Radical Islam: Medieval Theology and Modern Politics
(New Haven, Conn.: Yale University Press, 1985).
87. James Olidort, “What Is Salaªsm? How a Nonpolitical Ideology Became a Political Force,” For-
eign Affairs, November 24, 2015, https://www.foreignaffairs.com/articles/syria/2015-11-24/what-
salaªsm.
88. See Mapping Militant Organizations Project, Stanford University, http://web.stanford.edu/
group/mappingmilitants/cgi-bin/.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

International Security 42:2 32

provide quick information about their dedication to the cause for which they
were ªghting.

In short, Salaª jihadism succeeded in winning over local populations be-
cause its particular ideology represented the sweet spot for the strategic use of
extremism discussed above. First, Islamism tapped into the desire of the popu-
lation for a more just political system based on religion.89 Justice, honor, and a
lack of corruption were the key ideals the ideology promoted and this mir-
rored the ideals sought by a majority of the population in Iraq, Libya, Syria,
and Yemen. Second, Salaª jihadists helped citizens identify which of the many
competing Sunni leaders was most likely to be devoted to this cause, notwith-
standing their claims. In a highly competitive world of rapid change, citizens
were looking to identify those leaders who had some chance to remain true to
a just Islamic state. Finally, this extreme ideology addressed the concerns of
more moderate citizens that leaders would become corrupt once in ofªce.
In weakly institutionalized environments with a history of corruption—
especially those in which institutions were further degraded by war—citizens
understood that leaders could not be held accountable once in power. The lan-
guage of Islam had resonance with the population, while the extreme version
of Salaª jihadism reassured the population that a more principled Islamic state
might actually be installed.

why do all groups not become extreme?

If embracing an extreme ideology at a particular point in time is so beneªcial,
why have all groups ªghting these civil wars not adopted Salaª jihadism?
Many groups have increasingly adopted a more extreme version of Islamism
over time. Figure 1 reveals that starting in the early 1990s, the number of Salaª
jihadist groups active worldwide began to increase, with a particular surge in
growth after the Arab Spring. This growth suggests that political entrepre-
neurs increasingly recognized the beneªts of embracing an extreme version of
Islam given current conditions, and progressively adopted it.

Still, not all groups have moved to the ideological extreme. As of 2016,
30 percent of all major militant groups in Syria, 50 percent of all major militant
groups in Somalia, and 65 percent of all major militant groups in Iraq are not
Salaª jihadist. Four reasons likely account for this.

The ªrst reason is that ideology is not entirely changeable; not all groups can

89. This may explain why radical Islamist groups failed to gain prominence in the civil war in
Bosnia from 1992 to 1995. Unlike citizens in the Middle East, North Africa, and Central Asia,
Bosnian Muslims had a blueprint for a Muslim democratic society that went beyond the Islamist
model. See Gilles Kepel, Jihad: The Trail of Political Islam (London: I.B. Taurus, 2002), p. 240.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

The Extremist’s Advantage in Civil Wars 33

Figure 1. Number of Active Salaª Jihadist Groups Worldwide, 1988–2013

1990

1995

2000

2005

2010

SOURCE: Data taken from Seth G. Jones, “A Persistent Threat: The Evolution of al Qa’ida and
Other Salaª Jihadists” (Santa Monica, Calif.: RAND Corporation, 2014), http://www.rand
.org/pubs/research_reports/RR637.html.

easily switch ideology to become more extreme. A group that desires to be-
come Salaª jihadist must gain the support of clerics with an established radical
interpretation of Islam compatible with Salaª jihadism. To do so, rebel leaders
must ªrst prove their internal devotion to the creed, something not all elites
are willing to do. Osama bin Laden, for example, was credible in part because
he was willing to adhere to a strict interpretation of everyday living prescribed
by Salaªst doctrine.90 Thus, while rebel leaders may come to instrumentally
prefer a different ideology as a result of wartime processes, not all groups can
successfully adopt any ideology they wish.91

Second, rebel entrepreneurs understand that there are decreasing marginal
returns for embracing the same ideology as the rest of the ªeld. Once four
Salaª jihadist groups exist in a given civil war, becoming the ªfth Salaª
jihadist group is less likely to lead to success. Rebel entrepreneurs, therefore,
may have incentives to differentiate their group from a crowded ªeld of Salaª
jihadist groups by staking out a different ideological position. Ahrar al-Sham,
for example, has made signiªcant efforts to distinguish itself from more ex-
treme groups since 2015. The more a rebel group can differentiate itself from

90. Lawrence Wright, The Looming Tower: Al-Qaeda and the Road to 9/11 (New York: Alfred A.
Knopf, 2006).
91. See Paul Staniland, “Militias, Ideology, and the State,” Journal of Conºict Resolution, Vol. 59,
No. 5 (2015), pp. 770–793, on local social structures and armed group emergence.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

International Security 42:2 34

other rebel groups, the less competition it will face, and the easier it may be to
recruit its own unique set of soldiers and supporters.

Third, there are beneªts for capturing and maintaining the middle ground in
a crowded ªeld of competing rebel factions. These rewards can come in many
forms, including support from moderate foreign sponsors, such as the United
States, or aid from moderate sympathizers unwilling to help more extreme fac-
tions.92 The Syrian rebel group Ahrar al-Islam pivoted from a more extreme
jihadist brand toward ideological moderation in early 2015 in an effort to con-
vince Turkey and the United States that the organization could serve as a pal-
atable alternative to the regime of Bashar al-Assad.93 Moderate groups may
also be allowed to organize and operate within countries in ways that radical
groups are not; governments may be more willing to tolerate them; and out-
side states may be more willing to endorse them. The Muslim Brotherhood in
Egypt, for example, has succeeded and survived in large part because it has
avoided the extremist label.

Fourth, there are at least three types of costs groups must pay to embrace an
extreme ideology that some leaders might seek to avoid. For one, groups that
adopt Salaª jihadism are more likely to be targeted in the United States’ war
on terror, creating a powerful enemy with which they must simulta-
neously battle. The leaders of Hayat Tahrir al-Sham—formerly known as
Jabhat al-Nusra—for example, sought to rebrand themselves and distance
their organization from al-Qaida in part to avoid U.S. air strikes.94 Second,
there is a short-term/long-term trade-off for leaders who embrace an extremist
ideology. An extreme ideology could help elites achieve short-term organiza-
tional goals during a war, only to make a negotiated settlement more difªcult
over the long term. Extremist rebel groups are more likely to engage in acts of
terror, which then make them less likely to be invited to the negotiating table.95
They are more likely to experience splintering,96 and their leaders could face

92. This might also be an incentive for rebel leaders motivated solely by the desire for ªnancial
gain: as long as they remain moderate, the United States will send money their way.
93. Labib Al-Nahhas “The Deadly Consequences of Mislabeling Syria’s Revolutionaries,” Wash-
ington Post, July 10, 2015; and Bassem Mroue, “One of Syria’s Most Powerful Rebel Groups Is
Rebranding Itself with Turkey’s Backing,” Business Insider, October 8, 2015.
94. “Hay’at Tahrir al-Sham (Formerly Jabhat al-Nusra)” (Palo Alto, Calif.: Mapping Militant Orga-
nizations Project, Stanford University, n.d.).
95. Virginia Page Fortna, “Do Terrorists Win? Rebels’ Use of Terrorism and Civil War Out-
comes,” International Organization, Vol. 69, No. 3 (Summer 2015), pp. 519–556, doi:10.1017/
S0020818315000089.
96. J. Bowyer Bell, The Dynamics of the Armed Struggle (London: Frank Cass, 1998); and Cyrus
Ernesto Zirakzadeh, “From Revolutionary Dreams to Organizational Fragmentation: Disputes
over Violence within ETA and Sendero Luminoso,” Terrorism and Political Violence, Vol. 14, No. 4
(2002), pp. 66–92, doi:10.1080/714005641.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

The Extremist’s Advantage in Civil Wars 35

retribution from true believers in their party for any concessions they make.
Al-Qaida’s leadership, for example, at times sought to limit its extremism but
was unsuccessful because of more extreme subsidiaries.97 The beneªts of ex-
tremism in the midst of civil war, therefore, must be weighed against the chal-
lenges of extremism when trying to end a conºict and transition to peaceful
political rule.

alternative explanations for pursuing an extremist ideology

There are a number of different explanations for why rebel leaders might em-
brace an extreme ideology and why these groups might then attract more sup-
porters during war than moderate groups. The ªrst set of explanations
emphasizes the intrinsic appeal of ideology. It is possible that Salaª jihadist
groups have grown in number and inºuence because Muslim citizens have be-
come more radical since 2003. The empirical evidence, however, does not ap-
pear to support this. Surveys by the Pew Research Center found that Muslim
views of ISIS are overwhelmingly negative. Ninety-nine percent of Muslims in
Lebanon, for example, said that they had “an unfavorable opinion of the
Islamic militant group in Iraq and Syria known as ISIS.”98 It is also possible
that the act of ªghting a war serves to polarize a population, making citizens
more likely to shift to the ideological extreme as the war progresses. If this
were the case, then citizens would gravitate toward more extreme groups, not
for any practical reason, but because their beliefs have become more extreme
over time.

Also, it is possible that elites embrace Salaª jihadism not because of any or-
ganizational advantages, but because they are able to tap into the preestab-
lished network created by al-Qaida and funded by wealthy Gulf states.
Zarqawi, for example, clearly understood that aligning his Sunni rebel group,
Jamaat al-Tawhid wal-Jihad, with al-Qaida in Iraq in 2001 would gain him ac-
cess to al-Qaida’s network of private Gulf funders, operational expertise, and
recruitment mechanisms. Finally, it is possible that extremist groups have
thrived because they are willing to engage in more brutal tactics than moder-
ate groups, allowing them to outcompete rival groups in terms of coercion.

All of these explanations are plausible, and some almost certainly played a
role in the rise of radical Islamist groups in Muslim countries around the
world. The value of this article, however, is to point out that an extreme ideol-
ogy can still have great value even if none of these other factors is at play.

97. See McCants, The ISIS Apocalypse.
98. Michael Lipka, “Muslims and Islam: Key Findings in the U.S. and around the World” (Wash-
ington, D.C.: Pew Research Center, 2017).

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

International Security 42:2 36

Conclusion

This article helps to explain one of the major international phenomena of our
time: the rise of Salaª jihadist groups in civil wars in the Muslim world. It be-
gan with the observation that the number of Salaª jihadist groups ªghting in
civil wars has grown and, in many cases, these groups have outperformed and
outlasted more moderate groups. By 2016, Salaª jihadist groups accounted for
most of the militant groups in Syria, half of all such groups in Somalia, and a
third of all militant groups in Iraq.

Currently, the scholarly literature has no theory for why rebel groups would
embrace an extreme ideology or why more extreme groups would perform
better than more moderate groups in war. The academic literature has tended
to ignore the role of ideology in civil war.99 Instead, scholars have assumed
that the core features of rebel groups—their identity, goals, and beliefs—were
exogenous to the war in which they were ªghting and not part of an ongoing
strategy of the war itself.

This article has attempted to ªll this gap. It has argued that the choice of ide-
ology, especially the level of extremism, is likely to be endogenous to speciªc
features of a civil war: the level of competition, a group’s local support net-
work, and the institutional constraints on state power. When competition is
high, information poor, and institutional constraints weak, an extremist ideol-
ogy could help rebel groups mitigate difªcult collective-action, principal-
agent, and commitment problems. An extremist ideology, therefore, can give
rebel entrepreneurs an organizational advantage under certain conditions.
All three of these conditions were present in the post-2003 civil wars in
Muslim countries, and all help explain the emergence and growth of Salaª
jihadist groups.

This theory challenges a number of assumptions about extremist ideology
and its adherents. The popular press is ªlled with stories about the increasing
radicalization of Muslims around the world. The assumption is that Salaª
jihadism has grown in popular support and strength because the number of
true believers has grown. This article offers a different interpretation of the rise
of these groups. Muslims need not believe in an extreme version of Islam to
sympathize and support groups that embrace such a vision. They need only
operate in a world where political power is heavily contested and the quality
and commitment of competing rebel groups is uncertain. An extremist ideol-
ogy will always attract citizens who believe in the ideology, but during times
of uncertainty and insecurity it will attract more moderate citizens as well.

99. One notable exception is Sanín and Wood, “Ideology in Civil War.”

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

The Extremist’s Advantage in Civil Wars 37

Finally, the popular press often portrays individuals who join or collaborate
with extremist groups as either true believers or hapless victims. The
formation of so many Salaª jihadist groups and the ability of groups such as
Jabhat al-Nusra, ISIS, and al-Qaida to capture and administer territory sug-
gests a deeper level of local support than many Westerners have been willing
to acknowledge. This article reveals the very practical incentives that moderate
citizens might have to support such groups in environments of uncertainty,
competition, and state abuse of power. The seeming popularity of Salaª
jihadist groups, therefore, may have less to do with the ideas Salaª jihadism
promotes than with the security and assurances it offers.

The framework reveals the role that an extreme ideology can play in orga-
nizing a rebel group and ªghting a war. It suggests that many rebel leaders
may have embraced Salaª jihadism because this particular ideology allowed
them to use cheap inducements and punishments to attract supporters. The
extreme nature of the ideology helped them recruit a core group of particularly
tough ªghters—zealots—who elevated the quality of
the ªghting force
and helped enlist additional soldiers attracted to the success of the group.
Salaª jihadism then allowed rebel elites to reassure Muslims that they would
remain true to the rule of Islam and avoid aligning with the West, making
them additionally attractive to a more mainstream audience. Al-Baghdadi
and al-Zarqawi effectively used these advantages to outcompete other
Sunni groups.

This article is the ªrst step in understanding the role that ideology, espe-
cially extremist ideology, plays in civil war. It reveals the incentives that indi-
viduals may have to promote an extreme ideology even if they do not believe
in its ideals. Much more work needs to be done. On the theoretical side, sig-
niªcantly more analysis is needed on the societal bases of ideology and on the
initial choice of ideology by rebel entrepreneurs. Why do certain ideas reso-
nate more with some societies than others? Additional theorizing is also
needed on the strategic choices made by individual citizens and the conditions
under which their support for one ideology over another may change and
evolve over time.

On the empirical side, this framework lends itself to rigorous testing. The
theory proposes a number of conditions purported to encourage the emer-
gence of extremist rebel groups. These include periods of heavy political com-
petition, low information, and weak political institutions. It also identiªes
additional factors that could cause elites and citizens to embrace an extreme
ideology, including rapid social change, lengthy civil wars, and powerful
Salaª jihadist networks. At the macro level, data could be collected on each of
these variables to determine whether extremist groups are more likely to

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

International Security 42:2 38

emerge under any of these conditions. At the meso level, data could be col-
lected to determine whether the ideology of rebel groups ªghting in a civil war
changes as conditions on the ground change. Do rebel groups shift to more ex-
treme ideologies if the number of rebel groups increases over time, or if
ªnancing from radical organizations suddenly becomes available? Finally, at
the micro level, data could be collected to determine whether moderate citi-
zens are strategic, shifting their support to more radical groups based on cer-
tain predicted conditions on the ground. Do moderate individuals behave as
the theory predicts or are they more heavily inºuenced by other factors? It is
my hope that this article triggers a larger research program on ideology and
civil war that addresses many of these theoretical and empirical issues.

The theory, if correct, could have three potentially important implications
for policymakers seeking to counter the rise of radical Islamist groups around
the world. The most important concerns the sources of support for extremist
groups, especially from local populations. The type of counterstrategy the
United States pursues will depend heavily on whether U.S. leaders assume
that Muslim citizens support Salaª jihadist groups because they truly be-
lieve in a seemingly “medieval” and fringe interpretation of Islam, or because
they believe these groups are more likely to deliver justice and reform in
countries where both characteristics are missing. This article reveals the incen-
tives that moderate citizens may have to support extreme groups and the con-
ditions under which these incentives are likely to be strong and weak. It
also reveals the parts of an extremist strategy that are designed to appeal to
more moderate citizens. Dissuading true believers from joining these move-
ments will be a huge challenge, but convincing more moderates to defect
might not.

This article suggests three ways in which moderate support for these move-
ments might be undercut. The ªrst involves the competitive nature of these
wars and the collective-action problems this creates for rebels. Wars with mul-
tiple competing rebel factions create incentives for rebel entrepreneurs to go to
the ideological extreme, but this competition also creates opportunities for out-
siders to bolster less extreme groups. This can be done by helping moderate
groups build high-quality ªghting units that would then attract the support of
a larger percentage of moderate citizens. The second has to do with agency
problems. Outsiders can also help to reveal the fraudulent nature of extremist
elites and their claims. Religious authorities of a more moderate persuasion
(often tribal leaders), for example, can use the politics of religious legitimacy as
a check against radical leaders who attempt to use their religiosity as a costly
signal of principle and devotion. Finally, outsiders can address the commit-

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

The Extremist’s Advantage in Civil Wars 39

ment problems that encourage extremism by helping to reform state institu-
tions and refusing to support and sustain corrupt regimes.100

Scholars can no longer close their eyes to the role of ideology in civil war.
The rise of Salaª jihadist groups, with their global aims and their condem-
nation of the United States and its allies, has become one of the biggest secu-
rity threats in the world today. To date, the United States has not been
particularly effective at reducing their number. Indeed, these groups are grow-
ing over time. This growth suggests an underlying level of support from
Muslim communities around the world that we do not yet understand. Only
by analyzing the ways in which an extreme ideology may beneªt both elites
and more moderate citizens in civil war can we begin to understand how it
might be strategically used, and the conditions under which it is likely to at-
tract support. This article is a ªrst step in that direction.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u

/
i
s
e
c
/
a
r
t
i
c
e
–
p
d

f
/

4
2
2
7
1
8
4
3
8
5
9

/
i
s
e
c
_
a
_
0
0
2
9
2
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

100. Kenneth M. Pollack and Barbara F. Walter, “Escaping the Civil War Trap in the Middle East,”
Washington Quarterly, Vol. 38, No. 2 (Summer 2015), pp. 29–46, doi.org/10.1080/0163660X.2015
.1064708.
Download pdf