Content uploaded by Milos Popovic
Author content
All content in this area was uploaded by Milos Popovic on Oct 12, 2021
Content may be subject to copyright.
Purifying the Religion: An Analysis of Haram Targeting among Salafi
Jihadi Groups
David S. Siroky, Emil Aslan Souleimanov, Jean-François Ratelle, and Milos Popovic1
Keywords: jihad, Salafi ideology, religious violence, haram targeting, Russia, Caucasus,
global analysis.
The Islamic State (ISIS/ISIL) publicly celebrated the carnage at a gay nightclub in Orlando,
Florida on June 12, 2016—the deadliest terrorist attack on U.S. soil since 9/11, and the
deadliest mass shooting by a single gunman in U.S. history with 49 dead and over 52
injured—calling it the “best Ramadan gift”. Simultaneously, the group issued a new video
inviting more lone-wolf attacks across the United States, especially in places where haram
(i.e., forbidden or proscribed by Islamic law) activities are taking place. Other haram
2
locations, such as the twin “Sin Cities”—Las Vegas and Macau—are also on the target “wish
list” of many Salafi-jihadi groups. In early 2017, ISIS threatened to attack the European
3
Women's Championship football match in Galgenwaard Stadium in Utrecht on July 19, 2017,
causing widespread fear in the oldest religious epicenter of the Netherlands, where over 10%
of the current inhabitants are Muslims.
While lone-wolf ISIS-inspired attacks have drawn a lot of attention, local jihadi
groups often cause even more destruction. For example, Boko Haram-affiliated jihadis
3The terms Salafi-jihadi and jihadi are used interchangeably in this article.
2Ben Hubbard, “ISIS Uses Ramadan as Calling for New Terrorist Attacks,” New York Times, July 3,
2016. Lone-wolf homegrown ISIS-inspired attacks are different than the group-based attacks examined here.
1Milos Popovic acknowledges that this project has received funding from the European Union’s Horizon
2020 research and innovation programme under grant agreement No 798392.
1
attacked a brothel on June 28, 2014, killing ten people in Nigeria's city of Bauchi. Two weeks
4
later, Islamic State-affiliated jihadis in Iraq attacked a brothel in Baghdad, murdering a total
of 32 people. In addition to clubs and brothels, jihadi groups have attacked other haram
5
targets, such as liquor stores, pre-Islamic statues, mixed-sex schools, tourist resorts and
non-halal restaurants. On December 7, 2013, for instance, unidentified Iraqi jihadis targeted
liquor stores across the country, killing 15 people. Boko Haram and Taliban-affiliated jihadi
6
groups in Afghanistan have frequently used violence against schools offering mixed-sex
education as a way of broadcasting their disapproval, while increasing the risk for students to
attend.7
Taliban-affiliated jihadi groups have targeted popular resort areas in Afghanistan,
citing “illicit fun”, and have made threats to do the same in the West on the grounds that they
are haram. Jihadi groups across the Middle East have also targeted non-halal stores and
8
restaurants that violate Islamic dietary rules with increasing frequency. Asbat al-Ansar, a
9
Salafi-inspired and Al Qaeda-linked terrorist group operating in Lebanon in the mid-1990s,
routinely bombed non-halal restaurants, liquor stores, along with nightclubs, theaters, and
hostile religious leaders. As early as in the 1980s, Mujama al Islamiya, a vigilante group
10
established by the controversial Sheikh Akhmed Yassin in Gaza, routinely attacked liquor
stores, casinos, cinemas, and restaurants selling alcohol.11
11 Beverley Milton-Edwards. Islamic politics in Palestine. London: IB Tauris, 1999: 115.
10 Country reports on terrorism 2004, U.S. Department of State: Office of the Coordinator for
Counterterrorism, Washington, D.C. 2005, April 2005: 95.
9Katharina Von Knop. "The female jihad: Al Qaeda's women." Studies in Conflict & Terrorism 30.5
(2007): 397-414.
8Tom A. Peter, ‘Taliban attack Kabul resort, citing ‘illicit fun’ and alcohol,’ Christian Science Monitor,
June 22, 2012.
7Global Coalition to Protect Education from Attack (GCPEA), Country Profiles: Afghanistan, (New
York City, NY: GCPEA, 2015).
6Robert Spencer, ‘Sharia in action in Iraq: Islamic supremacists attack liquor stores, murdering 15
people,’ Jihadi Watch, December 7, 2013.
5‘Iraqi Jihadists slaughter 32 at Baghdad brothel,’ No More Cocktails, July 14, 2014.
4Robert Spencer, ‘Nigeria: Islamic Jihadists murder 10 and injure 10 more with bomb blast at brothel’,
Jihadi Watch, June 28,2014.
2
However, haram targeting is not limited to the Middle East and Africa. In Russia, for
example, Kavkazskii uzel reported that two bombs exploded in different “illicit” cafés in
Khasavyurt, killing seven people in January 2011. On August 4th, 2013, another blast
12
occurred in a store selling alcohol in Dagestan’s capital city of Makhachkala. One of the
13
most infamous Dagestani jihadi groups, Jamaat Shariat, warned business owners that, if they
do not stop their haram activities, “we’ll set fire to your brothels, blow up places where you
do haram, destroy your properties and shoot at your stores and casinos, …[and] at your
saunas, where adultery is practiced.”14
Despite the attention that such attacks receive in the media, most haram targeting is
much less violent, and not all Salafi-jihadi groups engage in haram targeting–indeed most do
not–and that raises a crucial question. This article suggests that there is a clear logic to haram
targeting that can be derived from Salafi-jihadi groups’ ethnic structure. Haram-centered
violence, we posit, has the important function of uniting ethnically mixed jihadi groups by
fostering a superordinate Islamic identity that is necessary for fighters to overcome collective
action problems and ethnic divisions. As a result, leaders of ethnically mixed Salafi jihadi
groups are more likely to encourage haram targeting as a means of socializing recruits from
distinct ethnic backgrounds into a unified fighting force. The observable empirical implication
14 “V Dagestane rasprostranayutsa listovki s ugrozami v adres torgovtsev alkogolem i
narkotikami,”KavkazskiyUzel, 18 May 2010, http://www.kavkaz-uzel.ru/articles/168958/, accessed July 11,
2013.
13 Dvachelovekago spitalizirovany v Makhachkaleposlevzryva v gastronome "24 chasa", Kavkazskiyuzel,
5 August 2013 http://www.kavkaz-uzel.eu/articles/228133/ accessed March 24, 2017; “Twin bomb blasts rip
through two shops in Dagestan,” Euronews, 31 October 2013,
http://www.euronews.com/2013/10/31/twin-bomb-blasts-rip-through-two-shops-in-dagestanaccessed March 24,
2017
12 “Explosion in Khasavyurt is qualified as an act of terror” Kavkazskiy uzel, 27 January 2011, available at
http://www.eng.kavkaz-uzel.ru/articles/15938/, accessed March 24, 2017; “Three persons lost in explosion in
Khasavyurt” Kavkazskiy uzel, 15 January 2011, available at http://www.eng.kavkaz-uzel.ru/articles/15871/,
accessed March 24, 2017. The majority of the attacks against haram targets in the North Caucasus happened
between 2010 and 2014 in Dagestan.
3
of this logic is that haram targeting is much more common among ethnically mixed groups
than among homogenous ones.
The article develops and tests this core proposition against alternatives using two
original samples. First, drawing on ethnographic fieldwork along with roughly 200 original
interviews, we developed a unique dataset of all jihadi groups and haram attacks (against
liquor stores, casinos and bathhouses) in Dagestan between 2010 and 2014. We draw on
15
original interviews with ex-jihadi combatants in Dagestan, eyewitnesses of haram targeting,
and members of the Ministry of Internal Affairs (MVD) and the Federal Security Service
(FSB) to provide additional insight and context on the logic of haram targeting. Second, we
created a cross-national sample covering all Salafi jihadi groups across the globe operating
between 1998 and 2015, which enables us to offer the first global assessment of haram
targeting.
We selected Russia’s peripheral region of Dagestan for an in-depth study of haram
targeting since it hosts numerous jihadi groups that have engaged in a large number of haram
attacks as well as some that have done so only rarely, or not at all. Compared to many other
places in the world, Dagestan witnessed a significant amount of haram targeting, especially in
the years 2010-2014, and also exhibits immense internal variation among the various jamaats
that can be used to assess the theory. Dagestan is also a multi-ethnic polity, composed of 14
major ethnic groups and more than 30 local languages, including the Russian language as its
lingua franca. Since the early 2000s, the republic has faced an upsurge of Islamic insurgent
violence, driven largely by local factors. Following Moscow’s successful counter-insurgency
16
16 These include corruption and clan competition for local resources, human rights abuses, religious
repression, economic decline and the spillover effects of the Chechen conflicts. Campana Aurélie and
Jean-François Ratelle. "A Political Sociology Approach to the Diffusion of Conflict from Chechnya to Dagestan
and Ingushetia." Studies in Conflict & Terrorism 37.2 (2014): 115-34; Jean-François Ratelle, and Emil Aslan
Souleimanov. "Retaliation in Rebellion: The Missing Link to Explaining Insurgent Violence in
Dagestan." Terrorism and Political Violence 29.4 (2017): 573-92; Emil Souleimanov and David Siroky, Random
15 Bathhouses (saunas) are commonly considered brothels due to the proliferation of sexual services there.
4
campaign and the emigration of hundreds of young Salafists to live under the Islamic State in
Syria and Iraq, the local insurgency has significantly receded. This wave of departures
17
underlined long-lasting religious tensions between the Russian state-sponsored form of Islam
and the Salafi communities within Dagestan, often depicted by the government as extremist
groups.
As a clan-based society, political and religious groups in Dagestan have to overcome
cleavages and engage with multiple identities in order to pursue collective action. After the
fall of the Soviet Union, one solution to these challenges was the creation of a consociational
model of power-sharing that sought to “manage” ethnic and clan tensions between Avars,
Dargins, Laks, Kumyks, and other less numerous ethnicities. Overcoming conflicts and
18
grievances among the republic’s main ethnic groups has been one of the main challenges of
Dagestan’s political system. Similar difficulties can be observed inside insurgent and religious
groups, where ethnic and clan-based identities have often conflicted with the need for a
common identity among the fighters.
For these reasons, studying haram targeting in Dagestan, especially during this time
period, affords an attractive opportunity to leverage original micro-level data for
theory-testing. We complement this in-depth analysis with a new disaggregated global dataset
on haram attacks among jihadi groups across the entire universe of locations where jihadi
groups operated (e.g., Iraq, Pakistan, Afghanistan, Somalia, Nigeria) from 1998 to 2015.19
19 See Appendix.
18 Robert Bruce Ware and Enver Kisriev, “Ethnic Parity and Democratic Pluralism in Dagestan: A
Consociational Approach,” Europe-Asia Studies 53.1 (2001), pp. 105-131.
17 Jean-François Ratelle. "North Caucasian Foreign Fighters in Syria and Iraq: Assessing the Threat of
Returnees to the Russian Federation." Caucasus Survey 4.3 (2016): 218-38.
or Retributive? World Politics (2016), 68.4: 677-712; David Siroky and Valery Dzutsati, The Empire Strikes
Back: Ethnicity, Terrain, and Indiscriminate Violence in Counterinsurgencies, Social Science Quarterly,
96.3(2016), 807-829.
. Domitilla Sagramoso. "The Radicalisation of Islamic Salafi Jamaats in the North Caucasus" Europe-Asia
Studies 64.3 (2012): 561-595; Marya Rozanova and Akhmet Yarlykapov. "The Islamic Religion and Cultural
Diversity in Contemporary Russia: Case Study of North Caucasus Region, Dagestan." OMNES: The Journal of
Multicultural Society 5.1 (2014): 22-47.
5
This serves to provide a sense of the extent to which the in-depth Dagestan results have
external validity.
In the next section, we dissect the concept of haram and its role in Salafi-jihadi
ideology, and then present our theory of haram targeting and its testable implications. Using
new datasets, first in Dagestan and then globally, we assess all the empirical evidence in view
of the theory and contextualize it using ethnographic data. The last section highlights the
findings, the limitations and several future directions for research on religious violence.
What We Know about Salafi-jihadi Violence and Haram
In response to the increasing destruction caused by Salafi-jihadi organizations, scholarship on
jihadi violence has grown substantially over the last decade and a half. Whereas there was
20
once a dearth of research, there is now an abundance of studies on different dimensions of
Salafi jihadism: the pathways of individual radicalization, motivations to join or exit jihadi
21
21 Clark McCauley and Sophia Moskalenko, “Mechanisms of Political Radicalization: Pathways Toward
Terrorism,” Terrorism and Political Violence,20:3 (2008): 415-433.
20 On jihadi organizations and radical Islamic movements, see Faisal Devji, Landscapes of Jihad:
Militancy, Morality and Modernity (Ithaca, NY: Cornell University Press, 2005); Mary Habeck, Knowing the
Enemy (New Haven: Yale University Press, 2006); Rudolph Peters, Jihad in Classical and Modern Islam
(Princeton, NJ: Markus Wiener Publishers, 2005); Bassam Tibi, The Challenge of Fundamentalism: Political
Islam and the New World Disorder (Berkeley: University of California Press, 2002); Emmanuel Sivan, Radical
Islam: Medieval Theology and Modern Politics (New Haven, CT: Yale University Press, 1990);
6
groups, the logic of jihadi-based suicide terrorism, as well as the jihadi ideology,
22 23 24
networks, financing and– most relevant for this study–targeting.
25 26 27
Previous research on terrorist targeting suggests that religious ideology shapes the
decision to attack “soft targets” (generally understood as undefended objects and unaware
civilians), but there is no study addressing the phenomenon of jihadi violence against haram
28
targets, a specific form of soft target that embodies behaviors and goods considered sinful,
unlawful, and forbidden according to Salafi ideology and doctrine. While there are many
29
anecdotes and newspaper articles about haram attacks, which have periodically captured
29 James J.F. Forest, Homeland Security: Protecting America's Targets (Greenwood, 2006), pp. 39-41;
Glenn P. McGovern, "Securitization After Terror" in Encyclopedia of Transnational Crime and Justice (ed.
Margaret Beare), Sage, 2012, p. 371.
28 Ranya Ahmed. "Terrorist ideologies and target selection," Journal of applied security research 13.3
(2018): 376-390; Drake, 1998b. Gus Martin considers soft targets as places where there is “a large number of
civilians gather as well as military targets that generally are not on alert (passive), and not likely to offer
confrontation”. Gus Martin (2013) Understanding terrorism (4th ed.), Sage Publishing, 1. Roughly 70 percent of
all terrorist attacks worldwide since 1968 have been against soft targets, and the trend has apparently increased
over time. Both practitioners and scholars of terrorism have suggested that terrorists tend to attack the most
vulnerable “soft” targets. Quoted in Mia Bloom, Dying to Kill: The Allure of Suicide Terror (New York:
Columbia University Press, 2005). When authorities harden previously attacked targets, terrorists pursue less
protected ones (“the substitution effect”).Consistent with this logic, non-combatants are often selected as targets
because of their vulnerability and the psychological impact –“the image of civilians dying can be much more
powerful than the image of an attack on soldiers or police officers, since the risk of death for them is considered
to be part of their job as soldiers and police.”
27 We build on studies of terrorist target selection, see Charles J.M. Drake, Terrorists' target selection,
Basingstoke: Macmillan 1998a; Charles J.M. Drake, "The role of ideology in terrorists’ target selection,"
Terrorism and Political Violence 10.2 (1998b): 53-85; Martin Libicki, Peter Chalk and Melanie Sisson,
Exploring Terrorist Targeting Preferences, RAND, 2007; Patrick Brandt and Todd Sandler, “What Do
Transnational Terrorists Target? Has It Changed? Are We Safer?,” The Journal of Conflict Resolution 54.2
(2010): 214–36; Peter Toft, Arash Duero, and Arunas Bieliauskas, "Terrorist targeting and energy security,"
Energy Policy 38.8 (2010): 4411-4421.
26 Jacob N. Shapiro and David A. Siegel. “Underfunding in Terrorist Organizations. ” International
Studies Quarterly 51.2 (2007): 405–29.
25 Mette Eilstrup-Sangiovanni and Culvert Jones. “Assessing the Dangers of Illicit Networks: Why
Al-Qaida May Be Less Threatening Than Many Think.” International Security 33.2 (2008): 7–44; Scott
Helfstein and Dominick Wright. “Covert or Convenient? Evolution of Terror Attack Networks.” The Journal of
Conflict Resolution 55: 5 (2011): 785–813; Robert F. Trager and Dessislava P. Zagorcheva. “Deterring
Terrorism: It Can Be Done.” International Security 30.3 (2005): 87–123.
24 Victor H. Asal, Karl Rethemeyer, Ian Anderson, Allyson Stein, Jeffrey Rizo, and Matthew Rozea, “The
Softest of Targets: A Study on Terrorist Target Selection,” Journal of Applied Security Research 4.3 (2009):
258-278.
23 Carlos Pestana Barros and Isabel Proenca, “Mixed Logit Estimation of Radical Islamic Terrorism in
Europe and North America: A Comparative Study.” The Journal of Conflict Resolution 49: 2 (2005): 298-314;
Mario Ferrero. “Martyrdom Contracts.” The Journal of Conflict Resolution 50.6 (2006): 855–877.
22 Max Abrahms. “Why Terrorism Does Not Work.” International Security 31: 2 (2006): 42–78; cf. Erica
Chenoweth, Nicholas Miller, Elizabeth McClellan, Hillel Frisch, Paul Staniland, and Max Abrahms. “What
Makes Terrorists Tick.” International Security 33: 4 (2009): 180–202.
7
headlines, to the best of our knowledge there has not yet been a systematic analysis of haram
attacks among Salafi jihadi groups, either in one region or globally.
The Salafi movement has haram targeting–including by violent means–engrained in
its origins, and indeed emerged as such to cleanse the religion of everything inconsistent with
its understanding of the Islamic dogma. It calls for a strict and literal reading of Muslim holy
30
texts. Theologically, Salafis advocate the closest observance of monotheism (tawhid).
31
Proponents of the Salafi doctrine refer to themselves as believers in the unity of god
(muwahhidun), and refuse to acknowledge any other source of identity–e.g., ethnicity, class,
tribe, race–than their religious identity as part of the global umma (the community of fellow
Muslim believers). Proponents of Salafi-jihadism call for the establishment (or, revival of the
state from the late Mohammad era and its successors) of an Islamic theocracy based on
Islamic law (shariah), and emphasize the fight against infidels (kuffar),hypocrites
(munafiqun),and apostates (murtaddun). In order to purify Islam and return it to the path of
their righteous forefathers (as-salaf as-salih, hence the term “Salafism”), Muslims have to put
an end to pre-Islamic practices (jahiliyya) as well as later innovations in Islam (bid'a).Most
famously, Salafi-jihadis have emphasized the need for a “holy war” or struggle (jihad), which
specifically targets anyone and anything that is considered haram.32
32 Often jihad is understood as an offensive military struggle for the sake of expanding the realm of Islam
at the expense of the non-Islamic world (dar al-harb), and less frequently as a struggle for individual
self-improvement (the greater jihad or jihad al-akbar) or as a defensive war through military struggle (jihad
al-asghar) for the sake of the liberation of Islamic lands(dar al-islam). See Shiraz Maher, Salafi-Jihadism: The
History of an Idea, Oxford: Oxford University Press, 2016. Pp. 31-70; Farhad Khosrokhavar, Inside Jihadism:
Understanding Jihadi Movements Worldwide, Abingdon: Routledge, 2015.
31 Alexander Knysh, “Contextualizing the Salafi-Sufi Conflict (from the Northern Caucasus to
Hadramawt),” Middle Eastern Studies 43.4 (2007): 503-530; Gabriel R. Warburg, “From Sufism to
Fundamentalism: The Mahdiyya and the Wahhabiyya,” Middle Eastern Studies 45.4 (2009): 661-672.
30 While Salafis and jihadis usually share the common theological tenets of Salafism, they may differ in
their approach to the nature of armed jihad, or jihad by sword (jihad bissaif) as a duty for Muslims. Jihadis
usually consider it a duty for Muslims, while Salafis may differ in their attitude. Therefore, some observers
differentiate between moderate and militant Salafism, and place jihadis in the latter category. In practice, the
term jihadi is commonly used to designate those involved in armed struggle; Salafi to denote a puritanical
theological interpretation of Islam.
8
As a puritanical branch of Sunni Islam, Salafism is notorious for its uncompromising
stance towards haram, especially compared to the more relaxed attitudes of other
interpretations of Sunni Islam. Islamic tradition invokes the term haram to describe the
unlawful, immoral, and forbidden practices stipulated in the Quran and in the Sunnah. These
33
practices include, but are not limited to, adultery, premarital sex, mixed-sex communication
(unless between close relatives), gambling, homosexuality and prostitution. They also prohibit
the consumption of pork and alcohol, or any non-halal food. In addition to these practices
uniformly deemed haram according to most Islamic scholars, Salafi doctrine also outlaws
tobacco, music, dancing, singing and other forms of “illicit fun.”34
In other words, the Islamic juridical framework recognizes degrees of haram, and this
shapes both how jihadis and broader society perceive these acts. Some behaviors and acts are
considered makrooh (disliked by God), whereas others are considered explicitly haram
(forbidden by God). Most Muslims categorize smoking, for example, as makrooh rather than
haram. Within the haram category, these acts can be divided into major sins (kabira) and
minor sins (saghira), and these can be further divided between those acts that hurt only the
individual who is doing them (zatihi), versus haram acts that hurt all of society (ghairihi).
These distinctions are not merely theoretical, but can have a tangible impact on how the
haram act is perceived by its audience, and on its utility for the perpetrators. While it is
important to understand the full scope of the concept of haram, it is equally crucial that it is
understood differently by various Salafi-jihadi groups across cultural contexts. In Dagestani
jamaats, haram is primarily centered around alcohol consumption, which represents a critical
34 Some other practices are deemed haram for Salafis, including abiding by the principles of liberal
democracy, or implementing secular –and thus non-Islamic –laws.
33 The collection of teachings and practices ascribed to the life of Prophet Muhammad.
9
social problem, whereas in other regions jihadis are focused on purifying the religion from
other pressing issues.
Given the centrality of haram to Salafi ideology, and yet the massive variation in the
extent to which it is actually used - it seems clear that a careful examination of the extent to
which the various Salafi jihadi groups around the world actually engage in it is warranted.
This study investigates why some jihadi groups use violence against places, people and
practices deemed haram, whereas other jihadi groups refrain from it. The next section
proposes a preliminary theory of haram targeting, which leads to the empirical expectation
that ethnically mixed jihadi groups are much more likely to engage in haram attacks. We then
assess this conjecture in Dagestan and globally, while considering alternative explanations,
before drawing general conclusions and discussing promising directions for future research.
Related Literature and Theory
“Terrorists select their targets because of their symbolic and propaganda value,” writes Martin
(2014: 122). “High-profile, sentimental, or otherwise significant targets are chosen with the
expectation that the target’s constituency will be moved, and that the victims’ audience will in
some way suffer.” Examples include embassies and diplomatic personnel, international
35
symbols, symbolic buildings and sites, emblematic people, and personnel carriers. In
36
36 Ibid.; In their typology of terrorist targets, Newman and Clarke as well as Drake designate such targets
as iconic and symbolic in that they are representative of a target group’s identity and strength. The 9/11 attacks
on the World Trade Center, the Pentagon, and the failed attack on the White House are illustrative of the desire to
shake the symbolic economic, military, and political foundations of American power. See U.S. Department of
Justice, Office of Community Oriented Policing Services (COPS), Policing Terrorism: An Executive’s Guide, by
Ronald Clarke and Graeme Newman, (National Criminal Justice Reference Service, 2008); Drake, Terrorist’s
Target Selection, 10-11; Brigitte L. Nacos, “The Terrorist Calculus behind 9-11: A Model for Future Terrorism,”
Studies in Conflict and Terrorism 26.1 (203): 1-16.
35 Gus Martin, Essentials of Terrorism: Concepts and Controversies (Thousand Oaks: Sage, 2014): 122.
10
general, experts on terrorism suggest that terrorist targets are selected to “inform, educate, and
ultimately rally the people behind the revolution.”37
Building on this work, other scholars have focused on the ideologies of terrorist
groups and how they shape their perception of enemies, while also determining who or what
is considered a legitimate target. Asal et al. (2009) examine the role of ideology in target
38
selection—specifically, religious ideology—and focuses on civilian targeting. Religious
39
terrorism is distinct from other kinds of terrorism because it emphasizes divine punishment
and redemption, so targets are selected because they represent “the forces of evil”.
40
Juergensmeyer (2000) links religious violence to the targeting of “evil” civilians, objects and
practices. Salafi jihadis, like other religious fundamentalists, appeal to a godly rather than to a
worldly judge to legitimize their target selection. While religious ideology fuels
41
41 Daniel P. Hepworth, “Analysis of Al-Qaeda Terrorist Attacks to Investigate Rational Action,”
Perspectives on Terrorism 7:2 (2013). On the role of political theology in conflict, see: John D. Carlson and
Jonathan H. Ebel, eds. 2012. From Jeremiad to Jihad: Religion, Violence, and America. University of California
Press; Bruce Lincoln. 2003. Holy Terrors. Thinking about religion after 9/11. Chicago: University of Chicago
40 Drake, 1998b; Simon Perry and Badi Hasisi, “Rational Choice Rewards and the Jihadist Suicide
Bomber,” Criminological Theory and Terrorism 27.1 (2015): 53-80; Mark Juergensmeyer, Terror in the Mind of
God: The Global Rise of Religious Violence, 3rd ed, (Oakland: University of California Press, 2003); Daniel S.
Gressang, "Audience and Message: Assessing Terrorist WMD Potential," Terrorism & Political Violence 13.3
(2001): 83.
39 Asal et al. 2009. Ideologies serve as guides for target selection by rationalizing terrorist violence. For
rational actor models of terrorism, see Dwight R. Lee, “Free Riding and Paid Riding in the Fight Against
Terrorism,” The American Economic Review 78:2 (1998): 22-26; Todd Sandler and Walter Enders, An Economic
Perspective on Transnational Terrorism, (2002) https://www.diw.de/sixcms/detail.php/39116. Accessed 20
February 2015; David Lake, “Rational extremism: Understanding terrorism in the twenty-first century,”
Dialog-IO 1:1 (2002): 15-29; Eli Berman and David Laitin. Hard Targets: Theory and Evidence on Suicide
Attacks. NBER Working Paper No. 11740, November 2005; Todd Sandler, John Tschirhart, and Jon Cauley, “A
Theoretical Analysis of Transnational Terrorism,” American Political Science Review 77.1 (1983): 36-54. Some
scholars have questioned the worldly rationality of religious terrorists. Jacob N. Shapiro, David A. Siegel,
“Underfunding in Terrorist Organizations.” In: N. Memon et al (eds.) Mathematical Methods in
Counterterrorism (Springer: New York, 2009).
38 Drake, 1998b; Teun van Dongen, “The lengths terrorists go to: perpetrator characteristics and the
complexity of jihadist terror attacks in Europe, 2004-2011,” Behavioural Sciences of Terrorism and Political
Aggression 6: 1 (2014): 58-80; Drake, 1998a; Lisa McCartan, Andrea Masselli, Michael Rey, and Danielle
Rusnak, “The logic of terrorist target choice: An examination of Chechen rebel bombings from 1997-2003,”
Studies in Conflict and Terrorism 31.1: 60-79. Focusing on ETA, de la Calle and Sánchez-Cuence (2004)
demonstrate that ethno-separatist terrorists are materially constrained by the resources they have, and
ideologically inhibited by the preferences of their supporters. Luis De la Calle and Ignacio Sánchez-Cuenca, “La
selección de víctimas en ETA,” Revista Española de Ciencia Política 10 (2004): 53–79
Clarke and Newman suggest terrorists select targets that elicit positive reactions from sympathizers/followers.
37 Bruce Hoffman, Inside Terrorism (New York: Columbia University Press, 2008): 17.
11
haram targeting, and haram is an explicitly religious concept central to Salafi jihadism, it is
also not the whole story, since we clearly observe variation in haram targeting among Salafi
jihadi groups that share the same ideology. If more religiously oriented jihadis were more
likely to engage in more haram targeting, and also more attracted to ethnically mixed jamaats,
it might be difficult to disentangle these reasons. While we cannot think of a reason why more
religiously oriented jihadists would be disproportionately drawn to ethnically mixed jamaats,
we do know that, in Dagestan, some of the most religiously fundamentalist (purist) groups
were mono-ethnic rather than mixed-ethnic jamaats, and further that these groups largely
refrained from haram attacks. Purist rural jamaats led by Salafi-jihadi icons were if anything
42
more likely to attract religious fundamentalists. Although religious ideology is clearly
important, if it were the whole story behind target selection, haram targeting would be far
more widespread than it actually is among Salafi jihadi groups.43
In practice, haram violence is rare (and typically not as brutal as the events that have
captured headlines), and are usually too limited in scope and intensity to defeat the opponent
outright and generate substantial political change. Instead, haram targets operate as largely
44
non-verbal messages; like other forms of terrorism, the power of haram targeting is primarily
psychological and sociological. It both serves to send a signal to the local population and, at
45
the same, to cohere a group of fighters from different ethnicities around a central religious
objective, fostering group solidarity that will help them overcome collective action problems.
45 Alex Schmid, “Terrorism as Psychological Warfare,” Democracy and Security 1: 2 (2005): 137-146.
44 According to Abrahms 2006 (p. 56), terrorist attacks are a “coercive instrument intended to
communicate to target[s]… the costs of noncompliance with their policy demands, while simultaneously seeking
to undermine the resolve of the target.
43 While the total number of haram targets is not known, we recognize that there is a chance of there being
an unequal distribution of haram targets, which would influence the extent of haram targeting. This is an
important topic for future research.
42 See later discussion of Gubden jamaat.
Press; Daniel Philpott. 2007. “Explaining the Political Ambivalence of Religion.” American Political Science
Review 101/3 (August): 505-526.
12
One might wonder whether the reason some groups engage in more haram targeting is
that they cannot control, monitor or sanction their rank-and-file members well. This
“organizational control” approach has been used successfully to understand the use of
violence and civilian targeting, and could be applied to understanding and explaining
46
variation in haram targeting. Abrahms and Potter, for example, argue that decentralized
47
groups face more principal-agent problems in which the rank-and-file fighters that are
subordinated to weak leaders are more likely to target civilians. In this logic, targeting
decisions often come from the bottom-up, sometimes in defiance of the leaders' preferences.
There is, however, an important distinction between the logic of civilian targeting, which is
usually a mix of strategy and pillaging, and haram targeting, which is almost always an
intentional strategy with a very specific religious message, and rarely the result of
recklessness and “bad apple” soldiers. Moreover, haram attacks are rare, especially compared
to civilian victimization, which alas occurs in nearly all civil wars. Haram targeting is, of
course, limited to civil wars with Muslim combatants. Further, even among such conflicts, it
is further restricted to those with a Salafi-Jihadi ideology, which is a minority, and only a
fraction of Salafi jihadi groups actually engage in haram targeting.
Ethnographic research with former Dagestani jihadi also indicates that the amirs
(commanders) of jamaats were the principal decision-makers. According to an ex-jihadi,
ordinary jihadis were usually treated as “infantry [soldaty], with the leadership [nachal’stvo]
making decisions about what to do, how to do, who to kill and ally yourself with […] how to
make or spend money. Our role was auxiliary [vspomogatel’nyi].” Akhmet Yarlykapov, a
48
48 Interview with “Daud.” Carrying out these less-challenging attacks at the onset of their careers provides
fresh jihadis a sense of purpose as well as confidence in their own skills, while binding them to the larger group.
Just as an attack on a haram target cements the sense of godly purpose and group solidarity among group
47 Max Abrahms and Philip Potter. "Explaining terrorism: Leadership deficits and militant group
tactics." International Organization (2015) 69.2, 311-342.
46 Devorah Manekin, Regular Soldiers, Irregular War: Violence and Restraint in the Second Intifada,
Cornell University Press, 2020.
13
reputed Russian-Dagestani anthropologist and an expert on the regional insurgency, observed
this pattern across the North Caucasus jihadi groups and elsewhere. According to Yarlykapov,
ordinary jihadists in Dagestan were subject to all sorts of manipulation and rarely in charge of
anything else than fulfilling tasks ascribed to them by the jamaat leadership; “matters of
importance were concentrated in the hands of amirs.” The leadership of jihadi groups
49
decided to socialize recruits using haram targeting, mostly in ethnically diverse jamaats. In
50
short, low-level members in jihadi groups did determine whether or which haram targets to
attack.
Rather than religious infusion per se or principal-agent problems, this article
51
suggests that the ethnic composition of jihadi groups influenced their likelihood of using of
haram attacks. We hypothesize that ethnically mixed jihadi groups are more likely to engage
in haram attacks because a critical function of haram targeting is to help overcome
conflicting ethnic and clan-based loyalties in mixed Salafi-jihadi groups, which can impede
collective action in the future (for instance, if members are unwilling to hit targets where their
ethnic kin reside). This is particularly true in a multi-ethnic setting. By confronting a
52
common enemy, we argue that leaders in jihadi groups sought to overcome the more parochial
identities of their members and create a greater group solidarity. As a cohesion-building
52 On ethnicity and collective action, see Henry Hale. The Foundation of Ethnic Politics, Cambridge
University Press, 2008.
51 In the discussion of Dagestan, we also consider urban-rural cleavages as an explanation, and, in the
global statistical analysis, we consider the rebel group’s age and strength.
50 Many of these attacks were part of initial rituals for new recruits to lock in complicity. Available
literature corroborates these findings. See, for example, Emil A. Souleimanov, "Making Jihad or Making
Money? Understanding the Transformation of Dagestan’s Jamaats into Organised Crime Groups," Journal of
Strategic Studies (2018), 41.4, 604-628.It is important to notice that jamaat leaders routinely deployed the
custom of initiation violence: fresh recruits were tasked with attacks on law enforcement, which were recorded,
putting jihadist “foot soldiers” in blackmailed positions and thus critically increasing their dependence on jamaat
leaders. Authors’ interview with Irina Gordiyenko, a reputable Russian journalist covering Dagestan, June 2014.
49 Authors’ interview with Akhmet Yarlykapov, 26 July 2020.
members, it also boosts the perpetrator's reputation outwardly vis-a-vis other jihadi groups. As a result, Jihadis
are usually quick to broadcast successful targeting of haram entities on their webpages, since conducting attacks
against symbols of non-Islamic values boosts a jihadi group’s reputation within the Salafi-jihadi community by
demonstrating their commitment to the common cause.
14
exercise, haram targeting, therefore, focused on what united the different ethnicities–Islam
and the umma–rather than on what divided them.
Attacking haram sites can help cement the group by providing a sense of godly
purpose, fostering solidarity among members and unifying recruits from different ethnic
groups. Among the various opportunities for violence, haram targets foster a superordinate
religious identity that transcends ethnicity while serving as a bonding experience for young
fighters. Through these mechanisms, haram targeting can help to overcome collective action
problems that are frequently greater in more diverse groups. For these reasons, we surmise,
haram targeting is more often used as a strategy by the leadership of multi-ethnic jamaats.53
Consistent with this logic, we hypothesize that:
H1: Multi-ethnic Salafi jihadi groups will be more likely than mono-ethnic ones to use
haram targeting.
In the next section, we examine this proposition using new data and interviews from
ethnographic research in Dagestan; after that, we use a new disaggregated global data set to
assess its external validity.
Data and methods
Using new data on all known incidents of haram-centered violence over a 50-month period
from 17 May 2010 to 27 July 2014 in Dagestan, we systematically examine the role of ethnic
53 Jamaat is a term for jihadi groups widely used throughout Russia’s North Caucasus and in the
post-Soviet republics, but is used here to refer to Salafi jihadi groups generally.
15
diversity in jihadi groups in haram targeting. We select 2010 as the start date because it
coincides with the date when Jamaat Shariat, Dagestan's largest jihadi network, began
54
disseminating leaflets across the republic, declaring war on those involved in “seeding spoiled
morals and multiplying sins.” During this 50-month period, we identified 41 distinct haram
55
attacks (out of a total of 386 terrorist attacks in total), and approximately 20 unique
56
Salafi-jihadi groups and sub-groups. Each incident was verified in two Russian-speaking
trackers of jihadi activities in Dagestan: The Caucasian Knot (KavkazskiiUzel),an Internet
news portal specializing in the Caucasus, and Chernovik, a Dagestani daily. After identifying
all mentioned events, we further validated them through other media outlets (Dagestanskaya
Pravda,RIA Dagestan, Makhachkalinskie izvestiya), together with Dagestan-based jihadi
groups’ websites, such as vdagestan.com,ummanews.com, and kavkazcenter.com, to confirm
the haram attack and to fill in any available details.
In addition, we also draw on numerous interviews conducted over the last 10-year
period with ordinary Dagestanis living in both rural and urban areas where insurgent groups
have operated. Our interviews cover a broad set of demographic groups inside Dagestan,
including Soviet and post-Soviet educated Dagestanis, males and females, and various social
classes, e.g., intelligentsia, religious figures, businessmen, human rights workers, journalists,
manual workers, farmers, and retirees. Each interview lasted between 30 minutes and two
hours and focused on the causes and dynamics of insurgent violence in Dagestan as well as
56 During the same period, the GTD (Global Terrorism Database) indicates that there were 386 terrorist
attacks in Dagestan, which suggests that attacks against haram targets represent about 10% of all attacks.
55 This appeal marked the start of a campaign of haram attacks across Dagestan. Quoted in Emil
Souleimanov, “Dagestan's Jihadists and Haram Targeting,” Central Asian and Caucasus Analyst 17.3 (2015):
10-13. Leaflets available upon request from authors.
54 In contemporary Dagestan as elsewhere in the North Caucasus, jamaat is a term commonly used to
designate local jihadi groups.
16
ethnicity and religion. Regarding the identification of the perpetrators, we asked at least two
informants from different groups to independently code each attack and its perpetrator.57
In the period from February to June 2017, we conducted an additional series of 20
follow-up interviews with our informants using various internet-based communication
channels (WhatsApp, Skype, Signal, Viber, etc.). The bulk of these interviews were conducted
with former jihadis from Dagestan; a minority was conducted with current jihadis from
Dagestan. Experts on jihadi violence in Dagestan, including local journalists, scholars,
politicians, and officers of the Dagestani Ministry of Interior and the Federal Security Service
were also consulted. Interviews with former and current jihadis were conducted both inside
and outside Dagestan in Istanbul, Tbilisi, and Prague, where multiple ex-combatants now live.
Interviews were conducted in most major Dagestani cities (including Makhachkala,
58
Kaspiysk, Derbent, Khasavyurt, Izberbash, and Buynaksk) as well as in rural areas where
insurgent were active (such as Karabudakhkentsky, Gergebilsky, Gunibsky, Levashinsky,
Shamilsky, Tabasaransky, and Untsukulsky districts).
To the best of our knowledge, the final dataset of haram attacks in Dagestan is the
most complete and accurate set of haram terrorist incidents in a single region, validated by
multiple independent sources, and with relatively detailed information regarding attacks, the
perpetrators of violence, along with characteristics of each jamaat.
Ethnicity and the logic of haram targeting in Dagestan
58 Former and current jihadis, as well as the officer of the Dagestani Ministry of Interior, consented to be
interviewed on the grounds of strict confidentiality. Their identities are therefore concealed in this study, whereas
the identities of scholars and journalists are disclosed.
57 In cases where our informants contradicted each other regarding the jamaat responsible for the attack,
we contacted additional informants for more information to arrive at a final coding.
17
Dagestan’s rural jihadi groups largely operate on a territorial-clan (tukhum) basis. Given the
59
salience of tukhum-based kinship in the republic's rural areas, authorities and law enforcement
are often manned by the members of the same family (or clan) as the members of locally
operating jihadi groups. Endogamy, pervasive in Dagestan’s rural areas, reinforces these clan
60
and ethnic loyalties; with Dagestan’s ethnic groups numbering from a few thousand to several
hundred thousand members, many co-ethnics are close or distant relatives. Moreover, against
the backdrop of Dagestan’s highly ethnically fractionalized society, ethnicity serves as the
principal source of self-identification, with its relevance stretching well beyond established
kinship ties. In their quotidian lives, individuals place enormous importance on ethnicity in
general and ethnic solidarity in particular: marriage, communal life, employment, and politics
all revolve around the notion of ethnicity-centered in-group solidarity. As many Dagestanis
61
have confessed, ethnicity is seen as an extended family mainly because in the not-too-distant
past, members of the republic’s ethnic groups were all relatives.62
Ethnicity has retained its significance even among regional jihadi groups, both in
Dagestan and across the multiethnic North Caucasus. Some jamaats are dominated by Avars
63
and others by Kumyks, Laks, Dargins, and Lezgins, whereas other jamaats are ethnically
63 Tekushev, “Triumph of the Caucasus Emirate,” CSIS:
https://csis-website-prod.s3.amazonaws.com/s3fs-public/legacy_files/files/publication/120217_Hahn_IIPER_52.
pdf A similar observation holds for Chechen Salafis, culturally proximate to their Dagestani neighbors.
62 Although clan and nationalism constitute essentially non-Islamic cleavages, known by the derogatory
term of asabiyya (tribalism or nationalism), which are contrasted to the highly revered concept of umma, the one
Islamic nation, religious, ethnic and tukhum loyalties are intertwined in Dagestani society. Virginie Collombier
and Olivier Roy, eds. Tribes and Global Jihadism. New York: Oxford University Press, 2018.
61 Olga Tsapieva and Tagir Muslimov, “Etnopoliticheskaya I etnosocialnayasituatsiya v Dagestane I
noveyshie konflikty,” Rossiya I musulmanskiy mir, 3.51 (2007):
https://cyberleninka.ru/article/n/etnopoliticheskaya-i-etnosotsialnaya-situatsiya-v-respublike-dagestan-i-noveyshi
e-konflikty/viewer.
60 Interview with prof. Akhmet Yarlykapov of the Moscow State Institute of International Relations, 30
January 2015. Akhmet Yarlykapov observed that virtually everybody is someone's relative in the rural areas, so
those who went to the woods [joined jamaats] are not eager to liquidate their relatives in the local administration
and police (or haram attacks). The instances of killing relatives […] became notorious just because they were so
rare.
59 This holds specifically for (ethnically homogenous) jihadi groups operating in rural areas, while jihadi
groups operating in urban centers are usually ethnically mixed. Souleimanov, 2018.
18
mixed. While targeting relatives is considered taboo in Dagestani society, killing a co-ethnic,
although less extreme, is also frowned upon, due to the widespread view in Dagestan of ethnic
groups as a form of extended kinship. This attitude, while pervasive throughout Dagestan, is
particularly strong in the more rural areas where ethnicity is most closely interwoven with
kinship and most, though certainly not all, jamaats tend to be monoethnic.
Rural and mono-ethnic jamaats sometimes do target military, civilians, and other soft
targets outside of their areas of operations, including in major cities and other rural areas, yet,
compared to urban multi-ethnic jamaats, rural mono-ethnic jamaats tended to strike against a
haram target less frequently, even when they had the capacity and opportunity to do so.
Although haram businesses are more widespread in urban areas, it is important to appreciate
that they also exist in rural areas, and are typically even easier to attack - since there is less
security - compared to urban areas. Yet, as an empirical matter, mono-ethnic jamaats almost
never attacked haram targets in their own rural areas. As an insurgent supporter from a remote
village in mountainous Dagestan explained how local haram is mostly addressed: “in villages
like this one, everyone knows everyone. Forest brothers don’t need to target “bad” Muslims;
we ensure that fellow villagers respect Islam, their families, and themselves. We know who
smokes, drinks or gambles. If they want to engage in sinful behaviors, they can leave and live
in Makhachkala. If not, we can take care of them. They would not dare do that in public
here”!64
By contrast, multi-ethnic jamaats do engage in a significant amount of haram
targeting in their own backyard (mostly in cities) and (to a lesser extent, for some of the
reasons just mentioned) in rural areas. This further underscores that the dynamics we describe
in Dagestan are not as much attributable to the greater density of haram targets in urban areas,
64 Interview with “Sultan.”
19
or to other urban-rural differences, so much as they are to the ethnic composition of the
jamaat itself, and to the leadership’s decision in ethnically-mixed jamaats to select haram
targets.
In our dataset, we identified 36 insurgent groups active in Dagestan during the period
of our study, including 12 mono-ethnic and 24 multi-ethnic groups. From that population,
65 66
we identified 24 jamaats that engaged in haram attacks (5 monoethnic and 19 multi-ethnic).
Even accounting for the fact that there are twice as many multi-ethnic jamaats as mono-ethnic
ones in Dagestan (24 vs. 12), multi-ethnic jamaats have carried out nearly four times as many
attacks as mono-ethnic groups (33 versus 9 attacks, or 6.6 versus 1.8 attacks per year).
One of the key reasons for this disparity seems to be that the leaders of multiethnic
urban jamaats were concerned with the need to overcome the strong ethnic identity among
group members by fostering a uniform Salafi-jihadi identity. As one former jihadi observed,
“disputes have often been conditioned by [ethnic] nationalism,” so ethnically-mixed jamaats
67
appear to have been more exposed to top-down Salafi-jihadi indoctrination for the sake of
transcending dangerous ethnic divisions. Attacking targets commonly identified as
68
un-Islamic (e.g., haram) serves as one of the main ways jihadi groups overcome parochial
identities for the sake of fighting a common “evil” enemy.
Some interviewees underlined how acting against what they perceived as “spoiled
morals” and “anti-Islamic” behaviors helped to unite fighters of different ethnic backgrounds.
“Ethnicity [natsional’nost’] doesn't matter,” explained one insurgent supporter, “Islam does.
68 Souleimanov, “A Perfect Umma.”
67 Interview with “Idris.”
66 Some of the most well-known multi-ethnic jamaats include the Sharia, the Khasavyurtovskiy, the
Shamilkalinskiy (Makhachkala), the Derbentskiy, the Buynakskiy, the Kizlyarskiy, and Southern (Yuzhniy)
jamaats.
65 Some of the most well-known mono-ethnic jamaats include the Gubdenskiy, the Karabudakhkentskiy,
the Tsumadinskiy, and the Sergokalinskiy jamaats.
20
Targeting ‘them’ unites us together.” As summarized by a former jihadi, “What is against
69
Islam is the enemy of us, Muslims. No matter whether you’re an Avar, Dargin, or Lak […] we
have to fight against the evil [kufr and haram]. It all brings us together. This is the beauty of
it.” At the same time, as emphasized by a former member of an ethnically-mixed jamaat,
70
“the good thing” about attacking haram businesses (without casualties) was that “you would
avoid [ordinary] people hating you because you specifically targeted a group of, say, Avars.”71
These small-scale and usually non-lethal attacks helped ethnically fragmented jamaats
become more cohesive, while avoiding lethal violence, necessarily aimed against members of
some of the republic’s ethnic community. Overcoming ethnic bias–and thus cementing group
cohesion using the common goal of fighting evil–is a central theme in haram targeting. A
leading figure of a formerly urban jamaat acknowledged that, “if you’re a Rutul [member of a
small Dagestani ethnic group], you’re somewhat inclined to avoid targeting business run by
another Rutul: the likelihood of them being members of your tukhum(clan) are higher. You
would rather prefer targeting someone else. An Avar-run business, for example, but even if
you target one of your own…at least the business is the target, and no one is killed.72
Another former jihadist admitted that as an uncontroversial and non-lethal form of
violence, haram targeting fostered group cohesion in that it inserted a sense of divine purpose
into individual jihadis who “felt we did what was needed to be done on the path of Allah.”73
After each incident of haram targeting, as one former jihadist put it, foot soldiers “felt
friendlier [druzhnee] toward each other,” underscoring the role of haram targeting as an
74
instrument of cementing group cohesion and religiously-imbued group solidarity. Multi-ethnic
74 Interview with “Akhmed.”
73 Interview with “Akhmed.”
72 Interview with “Nurmagomed.”
71 Interview with “Said.”
70 Interview with “Daud.”
69 Interview with "Akhmed".
21
jamaats leaders often specifically selected haram targets from one of the key group’s to break
clan ties and reinforce the jihadi loyalties.
As a former jihadi recollected, “when you carry out your first ever task, and you're
successful, it boosts your self-confidence incredibly. You get convinced that you're on the
right (religious) path and that you're the right man.” The “religious high” here is just as
75
important (especially for new recruits) as group cohesion is to leaders. In many interviews,
individuals underlined how acting against what they perceived as “spoiled morals” and
“anti-Islamic” behaviors helped to unite fighters from different ethnic groups inside Dagestan.
“Carrying my first attack showed me the righteous path and confirmed my bond with fellow
brothers and what I knew about Pure Islam.(…) Islam transcends everything in life.”76
While former insurgents and their supporters agree in principle that Islam is superior
to, and should take precedence over, clan and ethnic loyalties, the reality on the ground is
often different from the established rhetoric. Ethnic favoritism (asabiyya) is a widely
recognized problem in multiethnic Dagestan, and it has also pervaded the republic’s jihadi
77
groups, with leaders of multiethnic jamaats preferring the recruitment of their ethnic kin or
relegating important positions within their jamaats to their ethnic kin a means of ensuring
loyalty. This, too, has caused tension within jamaats, with members of disadvantaged ethnic
groups expressing grievances over discriminatory, un-Islamic practices. In 2010, an Islamic
judge of Dagestan’s largest jamaat Shariat even issued an internal fatwa calling the leader of a
minor jamaat, operating in the Khasavyurt area, to abandon ethnic favoritism or leave the
jihad entirely. In short, ethnic favoritism and bias in ethnically mixed jamaats is a serious
78
78 Souleimanov, 2018.
77 Ware and Kisriev 2001; Souleimanov, “A perfect umma?” Ethnicities, 18.3 (2018), 434-453.
76 Interview with “Abdul.”
75 Interview with “Abdulla.”
22
issue that the leaders need to overcome if they are to be successful, and haram targeting is one
clear and often effective means to achieve trans-ethnic group cohesion.79
The Makhachkala and Derbent jamaats–both ethnically-mixed jamaats–sought to
overcome identity-based cleavages through haram targeting in order to create a sense of
cohesion. It appears that the leaders saw haram targeting as a rather non-controversial means
of fostering supra-ethnic religious solidarity and group cohesion, and often deployed it as a
form of initiation violence to integrate new and inexperienced recruits. By contrast, the
80
Gubden jamaat was a mono-ethnic group well-known for its terrorist and insurgent activities.
Its members were known for their religious fervor and their dedication for what they called
“Pure Islam.” Although they regularly targeted security forces linked to “anti-Wahhabist”
81
activists and “moderate” clerics in Dagestan and in Russia (e.g. the 2010 Moscow Metro
bombings), the mono-ethnic Gubden jamaat never once engaged in any haram targeting.82
We observe these same patterns at large in the quantitative data. First, in absolute
terms, multi-ethnic groups launched more haram attacks than mono-ethnic outfits: 5 of the 12
mono-ethnic jamaats (41%) engaged in a total of 9 haram attacks, whereas 19 out of 24
multi-ethnic jamaats (79%) carried out a total of 33 attacks. Mono-ethnic jamaats in Dagestan
82 Magomedali Vagabov, leader of the Gudben jamaat in 2010, launched a series of suicide attacks in
Dagestan and across Russia including in the Moscow Metro. See, Gordon Hahn. The Caucasus Emirate
Mujahideen: Global Jihadism in Russia’s North Caucasus and Beyond. (Jefferson, NC: McFarland & Company,
2014).
81 Such statements were often made throughout our fieldwork in rural Dagestani villages. Our
interviewees discussed about the religious fervor of rural jamaats’ fighters including the Gubden jamaat and
their dedication to what is described as a purified form of Islam.
80 We thank a reviewer for noting an interesting parallel with the use of sexual violence as an initiation
strategy for new recruits, which we mention in the conclusion. See Dara Cohen. Sexual Violence in Civil War,
Cornell University Press, 2016.
79 Haram targeting in Dagestan was most often used as an “educative” and non-lethal tactic rather than a
punishing one. The reason behind this decision is the potential cost associated with haram targeting such as the
risk to antagonize local supporters as well as the cost of targeting hard targets for identity-based purpose
resulting in severe retaliation at the hands of law enforcement. Although the cost-benefit analysis is not part of
our argument, it is central to take into consideration to understand why jamaats opted for non-lethal haram
targeting to resolve their collective action problem.
23
were responsible for only 21% of all haram attacks (9/42), while multi-ethnic jamaats
conducted the remaining 79% (33/42) of the haram attacks.
Figure 1 also shows that multi-ethnic jamaats also engaged in significantly more
frequent attacks (1.4) than homogeneous ethnic jamaats (0.8). Consistent with our
ethnographic evidence, we find that ethnically mixed jamaats carried out the bulk of their
attacks on businesses selling liquor (convenience stores, restaurants, supermarkets, grocery
stores, pubs, and pharmacies). Multi-ethnic jamaats carried out most of haram targeting, and
did so mostly in urban areas, since those targets were closest, although they did occasionally
83
attack haram targets in rural areas (e.g., Shamkhal, Novy Sulak, Ashaga Stal, Aknada, and
Stalskoe). As a member of multi-ethnic urban jamaat told us “we will hit the sinners wherever
they are (…) even if they are hiding outside of the city. We have resources for that. Sinners
can only understand bombs and fear.”84
84 Interview with “Sultan.”
83 Although the majority of mono-ethnic jamaats in our dataset waged insurgent warfare across urban
areas in the republic, only one of those mono-ethnic jamaats, the Endireyskaya jamaat, engaged in any
significant amount of haram targeting. The Endireyskaya jamaat is composed of Kumyk fighters who are active
in Khasavyurt and Kizilyur, as well as their surrounding rural areas, and committed five minor attacks against
haram targets in urban areas (with no casualties). In spite of the fact that the jamaat engaged in five haram
attacks, the vast majority of their other attacks targeted police forces, state representatives, and religious figures.
Moreover, based on a discussion with an MVD officer in Dagestan, we learned that these attacks were done in
partnership with the (a multi-ethnic jamaat).
24
Figure 1 : Haram Attacks in Dagestan by Ethnic Homogeneity of Jamaat
Since multi-ethnic jamaats are more likely to operate in city environments, they are
much closer to haram targets (because there are more haram targets in cities than in villages),
and this could partly explain why multi-ethnic jamaats are more likely to engage in haram
targeting. While it is certainly true that multi-ethnic groups are more likely to operate in cities,
and that haram is more prevalent in urban areas, a few caveats are in order. First, mono-ethnic
village-based jamaats sometimes did hit targets outside of their direct vicinities, including in
major cities, so the assumption that jamaats only hit targets in their vicinity in dubious (e.g.,
the Gudben jamaat as well as the Endireyskaya Jamaat). Conversely, urban jamaats
sometimes targeted rural haram objects. In short, the jihadi group’s proximity to more haram
25
objects did not directly correlate with the location of their targeting, since jamaats often
engaged in haram targeting outside the areas in which they resided. To determine whether
these results were driven by urban versus rural jamaat differences, we conducted a t-test for
difference in means. While urban jamaats engaged in more haram attacks than their rural
counterparts, this difference was not statistically significant (t = 1.4, p-value = 0.17). Another
way to look into this issue is to examine only urban jamaats. Limiting the sample to urban
jamaats, there is still significant variation in haram targeting, which is due largely to their
ethnic structure. While the urban-rural cleavage is important to take into account, her and
elsewhere, it does not change our main results or conclusions.
We also considered the possibility that a higher frequency of haram targeting by
ethnically mixed jamaats was due to attracting more religious fundamentalists to their ranks.
Religious fundamentalists were, if anything, more attracted to groups led by jihadi icons,
which were most often headed mono-ethnic jamaats in rural areas. For example, the head of
85
the mono-ethnic rural Gubden Jamaat, Ibrahimhalil Daudov, was considered one of the
leaders of Dagestani jihadism. A “founding co-father” of the jihadist “Caucasus Emirate”
across the North Caucasus in 2007, he became the head of the all-Dagestani “Jamaat Shariat”
in 2010. During its existence, the Gubden Jamaat carried out only a single attack against
haram objects (and none while being under Daudov’s command). It was dozens of non-haram
but brazen-faced attacks across the republic that made the group infamous and attracted many
religious fundamentalists. Similarly, the leader of the mono-ethnic Levashinskiy jamaat,
Rappani Khalilov, was also considered an important jihadi ideologue, and attracted religiously
motivated recruits, but this group avoided violence against haram targets altogether. In spite
85 Our interviewees, however, emphasized that strangers were not welcome in tight-knit rural jamaats.
26
of harboring religious fundamentalist members and even leaders, these largely mono-ethnic
jamaats generally avoided haram targeting and pursued other targets and objectives.
In short, these findings show that ethnically heterogeneous jihadi groups engage more
than homogeneous ones in haram attacks. The qualitative evidence indicates that ethnically
mixed jamaats did so primarily for the purpose of creating a common purpose, overcoming
ethnic and tribal cleavages, especially among new recruits, and fostering a more unified
fighting force capable of engaging in risky collective action.
Global Data on Haram Attacks
To assess the argument’s potential external validity and broader applicability, we created a
new global cross-national data set of all haram attacks from 1998 to 2015. Drawing on
background information in Jones (2018), Big, Allied and Dangerous (BAAD), and Crenshaw
(2013), we identify 113 out of 167 (67%) Islamist groups that can be classified as Salafi (or
Deobandi). Of these 113, we find that 36 of them (or 32%) carried out one or more haram
86
attacks during the observed period, for a total of 196 distinct haram attacks.87
87 Most of the attacks in our data targeted girls’ schools in Afghanistan, Dagestan and Pakistan (114),
followed by convenience/grocery stores (23), music shops (10), liquor stores (9), Sufi shrines (9), cafes (7),
mausoleums and tombs (4), restaurants (4), brothels (3), cinemas (3), barber shops (2), resorts (2), Shia mosques
(2), a health center (1), van with textbooks (1), a livestock fodder manufacturer (1), nightclub (1), and a film
studio (1). Three groups - Tehrik-i-Taliban (TTP) in Pakistan, the Taliban in Afghanistan, Daesh in Iraq and
Syria, and Boko Haram in Nigeria - were responsible for the bulk of these attacks.
86 We consulted the BAAD database as well as the External Actor Conflict Dataset (EACD) to identify a
total of 119 Islamic militant groups that have operated in the period 1998-2015. We then drew extensively on
Jones (2014, 2018), Crenshaw (2013) and a number of online books and reports to code Salafi/Deobandi groups.
We also added 10 Dagestani outfits to arrive at a total of 142 Islamist militant groups. Next, we collected the data
on terrorist attacks for all these groups using the GTD. We filtered these attacks by selecting “soft targets”,
including businesses (restaurants, bars, cafes, retail stores, bakeries, hotels and resorts), religious institutions,
educational institutions, international NGOs, beach, museums, cultural centers, cultural houses, tourism travel
agencies, and tourist sites. We then examined the news related to every attack using the sources reported in the
GTD. Those attacks that had a clearly stated haram motive were coded 1 while those with no motive or
instrumental motive were coded 0. Importantly, many haram attacks were committed by unknown perpetrators
according to GTD. Given the lack of the identity of the group, we did not include these instances in our dataset.
27
To maximize comparability, we coded the ethnic composition of groups for the global
data in the same way as we did for Dagestan. Jihadi groups with two or more main ethnic
88
groups were coded as mixed, while those with more than half of its members belonging to a
single ethnic group were coded as homogenous. In total, we identified 61 ethnically
89
homogeneous groups (53%), 49 ethnically mixed ones (43%) and 4 that we could not discern
clearly (4%).
Figure 2 shows displays these new global data, and shows thta haram attacks are
much more common among ethnically mixed than among ethnically homogeneous jihadi
groups. Ethnically mixed groups were a minority of the Salafi jihadi groups, but carried out
two-thirds of all haram attacks (66%). Of the 61 ethnically homogenous groups, 11 engaged
in haram attacks (18%), compared to 25 of the 49 ethnically mixed groups (51%). Finally, the
average mixed group carried out more than twice as many attacks (2.6) than the average
homogenous group (1.1).
The two most active groups in the global data set, the Taliban and Tehrik-i-Taliban
(TTP), appear to illustrate this pattern. While both the Taliban and TTP share many
similarities, including their cooperation and disdain for girls’ school, they have different
ethnic structures, thereby offering a useful way to examine the core hypothesis. The Taliban is
largely composed of ethnic Pashtun tribes, while the TTP is comprised of dozens of tribal
90
factions as well as Arabs, Uzbeks, Afghans, Chechens and Punjabis. Assuming an equal
91
number of targets, our theory suggests that the TTP should conduct more attacks against
haram targets than the Taliban, as an intentional means of building solidarity and cohesion
91 Marta Crenshaw, 2013. Mapping Militant Organizations, source:
http://web.stanford.edu/group/mappingmilitants/cgi-bin/groups/view/105 (accessed on 3/12/2019)
90 Brenda Shaffer. The Limits of Culture: Islam and Foreign Policy. MIT Press. 2006, p. 277.
89 Given that there is no available dataset on ethnic composition of terrorist groups, we often had to rely
on anecdotal evidence, but our categories are also very coarse.
88 To code the ethnic composition of jihadi groups, we relied heavily on Crenshaw (2013), militant
profiles portals (Global Security and Tracking Terrorism) as well as online news.
28
across ethnically distinct and sometime divided members. The data indeed show that TTP
launched 79 haram attacks, whereas the Taliban launched “only”47.92
Figure 2: Ethnic Homogeneity and Haram Attacks, Global Sample (including the Dagestan sample)
Needless, taking into account rebel group characteristics aside from its ethnic
composition is critical. There is some evidence that older groups become deadlier through
learning and adaptation to counterterrorism, acquiring more specialized skills as they mature.
To account for this possibility, the model estimation controls for the age of jihadi groups.
93 94
Strength is another factor: some studies have found that weaker militant groups are more
94 We use the age of militant groups from the BAAD database: the number of years in existence as of
2015.
93 Bruce Hoffman. “Terrorism trends and prospects”, In: Ian Lesser, et al., Countering the New Terrorism.
(Rand: Santa Monica, CA 1999): 7-13.
92 The Taliban and TTP haram attacks are largely focused on girl schools, which suggests a strategy of
deterrence from modernization through fear. Other groups engaged in haram as a form of blackmail or taxation.
While this falls outside of the scope of our article, it is open for further research.
29
prone to target civilians due to a lack of incentives or resources to provide services to the local
population, while others find that stronger groups are more violent. Although haram
95 96
attacks are not aimed against civilians, weaker groups might utilize them to show that they are
more radical than their competitors in anticipation that such spectacular attacks would attract
new members. Recent research suggests that Salafi-jihadist groups are also likely to modify
97
their target selection in response to competition from other groups with similar ideology. If
98
this view holds, then weaker militants should be more likely to engage in haram attacks.99
However, a different alternative mechanism suggests that more experienced militant groups
are more versed in haram attacks.
We thus model the occurrence of haram attacks as a function of three covariates with a
logistic regression estimator: group ethnic composition, group strength, and group age. We
also account for the fact that some groups originate from the same country by clustering the
robust standard errors at the state level, and include country fixed effects to account for
unobserved cross-national heterogeneity.
Consistent with the theoretical expectations, and reinforcing the findings from our
analysis of data from Dagestan, Figure 3 (top) shows that ethnically mixed groups are
positively associated with a significantly higher probability of engaging in haram attacks. It
also shows that rebel capabilities and group age have no statistically significant effect on the
99 We measure this using the size interval of the organization’s membership from the BAAD database.
This covariate ranges from 0 (up to 100 members) to 1 (100-1000 members) to 2 (1000-10,000 members) and 3
(over 10,000 members). This variable is labeled “ordsize” in the BAAD database.We transform this categorical
measure into binary variable with 1, 2 and 3 treated as strong militants or 1, and weak militants or 0.
98 Megan Farrell. "The logic of transnational outbidding." Journal of Peace Research 57.3 (2020):
437-451.
97 Bloom, 2005; Andrew Kydd and Barbara Walter. "The strategies of terrorism." International Security
31.1 (2006): 49-80.
96 Victor Asal and R. Karl Rethemeyer. "The nature of the beast: Organizational structures and the
lethality of terrorist attacks." The Journal of Politics 70.2 (2008): 437-449.
95 Lisa Hultman. "Battle losses and rebel violence: Raising the costs for fighting." Terrorism and Political
Violence 19.2 (2007): 205-222; Reed Wood. "Rebel capability and strategic violence against civilians." Journal
of Peace Research 47.5 (2010): 601-614; Idean Salehyan, David Siroky, and Reed Wood. "External rebel
sponsorship and civilian abuse." International Organization 68.3 (2014): 633-661.
30
Figure 3: (Top) Full Model of Haram Attacks on global sample with mean point estimates and 90%
confidence intervals (thick horizontal line) and 95% confidence intervals (thin horizontal line). Robust
standard errors are clustered on the state level. The model includes country fixed-effects, which are not
shown here. The vertical dashed line is the line of “no effect”. The number of observations (Salafi
groups) is 114. (Bottom) Marginal effects of ethnically mixed Salafi groups on haram attacks of
model under (a).
likelihood of haram attacks, casting doubt on two possible alternative explanations while
reinforcing the plausibility of the main result. Figure 3 (bottom) displays the predicted
31
probability of a jihadi group conducting any haram attacks, separately for mixed and
homogeneous groups, and shows that ethnically mixed Salafi outfits were more than twice as
likely to conduct haram attacks compared to mono-ethnic ones.
Conclusion
Jihadi attacks against haram targets have made headlines around the world. As Salafi-jihadi
groups arise in more and more theatres of conflict, a better understanding of the strategic and
tactical choices behind this form of violence is warranted. With repeated threats of further
attacks across the U.S. and Europe, along with the massive influx of young male migrants
from Muslim countries, and European citizens returning from war and training, haram
targeting could easily become a major security issue in Western countries in the near future.
This article examines patterns of haram targeting using original empirical material
gathered from Dagestan and a new global data set that we compiled on haram attacks. In the
Dagestan analysis, we relied on dozens of original interviews with ex-combatant jihadis,
government officials and eyewitnesses to assess the core hypothesis both quantitatively and
qualitatively. Despite adhering to a uniform Salafi-jihadi dogma and rhetoric, not all Salafi
jihadi groups engage in haram targeting. It is mostly a phenomenon that ethnically mixed
jihadi groups practice, and we conjecture that leaders of mixed groups utilize it primarily for
overcoming internal parochial divisions based on ethnicity, tribe and clan by fostering a
superordinate identity based on the common purpose of purifying the religion from everything
un-Islamic. In a clan-based society like Dagestan, and in some parts of the Middle East,
haram targeting serves as a ritual of initiation carried out to cement loyalty among new
recruits, to overcome parochial and ethnic affiliation, and to increase the jamaat’s cohesion.
32
Usually non-lethal attacks against explicitly non-Islamic targets help to boost the sense of
divine purpose among newly recruited jihadis without creating a backlash among the local
population.
Scholars have shown that sexual violence can create cohesion in rebel groups by
separating individuals who violate such taboos from their former lives, making it harder for
them to leave and reintegrate. Among Salafi-jihadi groups, and particularly among more
100
ethnically mixed jihadi groups, haram targeting serves a similar but in some ways more
central social function, since it is tied to the group’s ideology in a way that sexual violence
usually is not. It also cuts off members from their old lives, like other violence, but it does not
do so by engaging in taboo behavior. Its behavior is explicitly sanctioned by the religious
ideology. Moreover, it unites different ethnic groups within the umma—thereby reinforcing
their decision to focus on their religious identity (and subordinate, to some extent, their ethnic
identity) with a sense of divine sanctioning for punishing the sinful and creating a more
Islamic society. While it is certainly not the only reason for haram targeting, we find a
consistent pattern in Dagestan and at the global level, which we hope provides a solid starting
point for advancing our understanding of haram targeting in the future.
We have suggested that haram targeting is particularly enticing for ethnically mixed
jamaats in Dagestan, for it represents a method to increase internal cohesion and boost
external reputation without incurring high costs from the local population or potential
retaliation from ethnically or kinship-related security forces. Yet it is important to note that
the tendency of multi-ethnic jamaats to engage in more haram attacks is only the tip of this
research frontier. Much progress has been made in our understanding of religious terrorism in
general, and we hope to have contributed to it further through this original study of haram
100 Cohen 2016.
33
targeting, though much remains to be done to better understand its logic across different
strategic settings. Further research is required, for instance, to better understand the role of the
control-collaboration model in civil war as well as the importance of membership in
international jihadist networks may influence the use of haram-based violence.
34
