The Global Jihadist Movement: The Most Lethal Ideology?

Abstract

Prior research has established a link between ideology and lethality, both within the homicide and terrorism literatures. We examine this relationship as it pertains to the Global Jihadist Movement (GJM). Using a series of logit and negative binomial models with a sample from the Global Terrorism Database, we find that the GJM is indeed more deadly. However, this relationship does not seem to differentially affect Americans, despite their role as the GJM’s defined “other.”

Keywords

terrorism Global Jihadist Movement ideologically motivated homicide quantitative lethality

Introduction

Executive heads of state and government officials alike have established a unified discourse that terrorism, defined as “the threatened or actual use of illegal force and violence to attain a political, economic, religious, or social goal through fear, coercion, or intimidation” (LaFree & Dugan, 2007, p. 184), is a significant threat. Although many forms of terrorism exist, policymakers have been primarily focused on that which is perpetrated by the Global Jihadist Movement (GJM)¹ (Cameron, 2015; Hollande, 2015; Merkel, 2015; Obama, 2015, 2016). For example, President Obama (2016) recently stated that his “top priority is to defeat ISIL and to eliminate the scourge of this barbaric terrorism . . . around the world” (para. 52). Following the Paris attacks by al Qa’ida, Prime Minister David Cameron declared, “The terrorist aim is clear. It is to divide us and to destroy our way of life” (2015, para. 15). James Clapper (2016) recently briefed to Congress that since 1970, “Sunni violent extremism has been on an upward trajectory,” and that Islamic State of Iraq and the Levant (ISIL) is “the preeminent terrorist threat” (p. 4).

Perhaps as an artifact of the debate behind what to call this phenomenon, the extant literature lacks clear and systematic quantification of its manifestation (Freilich & LaFree, 2016; Lum, Kennedy, & Sherley, 2006; Schmid, 2011; Young & Findley, 2011). Instead, previous work has taken a more focused approach to understanding the GJM through the evaluation of specific policies such as targeted killings (Hafez & Hatfield, 2006; Hepworth, 2014; Wilner, 2010; Zussman & Zussman, 2006), investigations into complex insurgency–counterinsurgency relationships (Fielding & Shortland, 2010; Linke, Witmer, & Loughlin, 2012), and assessments of both conciliatory and punitive interventions (Benmelech, Klor, & Berrebi, 2010; Condra & Shapiro, 2012; Dugan & Chenoweth, 2012; Lyall, 2009). While incredibly valuable contributions, these studies are often centered on one country (Linke et al., 2012), one group within the movement (Forest, 2012; Stenersen, 2010), one time period (Haddad, 2004), and/or they fail to examine terrorism as the primary outcome (Mousseau, 2011; Torres, Jordán, & Horsburgh, 2006). Exceptions in the literature (Asal & Rethemeyer, 2008; Barros & Proença, 2005; LaFree & Dugan, 2016; Piazza, 2009) assess the GJM’s broader impact and find collectively that this ideology is, at the very least, unique. However, and given this work’s similar limitations in scope, we seek to examine the broader question, “Have GJM-related incidents been more lethal than those perpetrated by other groups?”

Informed by a rational choice framework (Clarke & Neuman, 2006; Cornish & Clarke, 1986) and with insight from both the ideology–lethality and homicide literatures (Asal, Gill, Rethemeyer, & Horgan, 2013; Asal & Rethemeyer, 2008; Enders & Sandler, 2000; Gruenewald, 2011; Gruenewald & Pridemore, 2012; Parkin & Freilich, 2015; Parkin, Freilich, & Chermak, 2015; Piazza, 2009), the current study hypothesizes that the GJM has indeed been more lethal than that of other ideologically-based movements. We theorize that the GJM’s ideology, which involves both a supernatural audience and a defined “other” (Asal & Rethemeyer, 2008), has increased the perceived benefits associated with lethality. This investigation, first and foremost, starts by systematically operationalizing what constitutes the GJM. We then present basic descriptive trends and patterns of the movement over the last 20 years and then examine whether ideology predicts a variety of undesirable outcomes. Overall, we find that incidents follow similar patterns over time regardless of ideology. However, attacks associated with the GJM are more lethal in a number of ways, although these trends do not seem to be as pronounced with American casualties.

Literature Review

The Quantification of the GJM

Although much has been written on the GJM’s origins and key actors from a qualitative or philosophical standpoint (Agbiboa, 2013; Bergen, 2001, 2011; Borum & Gelles, 2005; Braniff & Moghadam, 2011; Byman, 2015; Forest, 2012; Hegghammer, 2006; McCauley & Moskalenko, 2014; Moghadam, 2008 Rabasa et al., 2006; Stenersen, 2010; Zelin, 2014), as noted, considerably less work has examined the quantitative impact of these groups. Two important exceptions are the contributions of LaFree and Dugan (2015) and Barros and Proença (2005), which examined the broader sphere. The former found that 13 of the 20 deadliest terror groups are what they refer to as “Islamic,” with these same groups much more likely to target people than their non-Islamic counterparts (LaFree & Dugan, 2015). They also discovered that both subsets (Islamic and non-Islamic) preferred bombings and armed assaults tactically (LaFree & Dugan, 2015). The latter, that of Barros and Proença, determined that “radical Islamic attacks” used explosives more than all other attack types and unduly employed assassinations when compared with non-Islamic groups.

However, even these important contributions are limited. Specifically, LaFree and Dugan (2015) employed a small sample of groups (n = 20), with only 13 of them tied to the GJM. Furthermore, Barros and Proença’s (2005) research focused exclusively on the Western operational environment. A related trajectory of research, while not explicitly focused on the GJM, has examined a larger universe (Asal et al., 2013; Asal & Rethemeyer, 2008; Enders & Sandler, 2000; Gruenewald, 2011; Parkin & Freilich, 2015; Parkin et al., 2015; Piazza, 2009). The next section reviews what is known about this literature—namely, the ideology–lethality connection—and how these findings can be viewed from within a rational choice perspective.

The Ideology–Lethality Connection and Rational Choice Theory

Prior work examining ideologically-motivated homicide has established its uniqueness in a variety of ways (Gruenewald, 2011; Gruenewald & Pridemore, 2012; Parkin & Freilich, 2015; Parkin et al., 2015). For example, Gruenewald (2011; Gruenewald & Pridemore, 2012) found that although largely similar to “average” U.S. homicides, those perpetrated by right-wing terrorists were more likely to involve both multiple offenders and victims, who were also strangers. Follow-up work centered on victim characteristics (Parkin & Freilich, 2015; Parkin et al., 2015) has discovered that the far-right is more likely to perpetrate interracial homicide, targeting minorities outside a private residence. Collectively, this line of research suggests that an ideological motivation can lead to distinctive homicide characteristics.

Another trajectory within the literature suggests that it is not just ideology, but the type of ideology, that can produce more detrimental outcomes. As Drake (1998) has stated, the manner in which a group targets is instructive of their worldviews, which can both legitimize targets and justify violence. Asal and Rethemeyer (2008), in one of the most rigorous tests of this premise, assessed the role of group characteristics including ideology, specifically ethnonationalism, religion, and a combination of both motivations in determining fatalities. This investigation posited that the ideology–lethality connection could be based on mechanisms similar to Drake’s premise: the organization’s audience (supernatural vs. earthly) and their ability to dehumanize their targets (othering). They found mixed support for these ideas but concluded that religious groups do kill more people than their nonreligious counterparts. Piazza (2009), theorizing similar constructs in explaining the ideology–lethality association, discovered that Islamist groups were not any more likely to perpetrate high casuality terrorist attacks. The exception was al Qa’ida, whose “universal/abstract” typology lends itself to a more communicative objective. Such an objective, according to Piazza, justifies the use of soft targets including members of the community, along with the use of suicide tactics.

Enders and Sandler’s (2000) study speaks to similar processes, but in the framework of what they term “choice-theoretic models.” These models essentially elucidate modern-day rational choice principles, namely, that terrorists are rational actors who operate based on their individual, risk-reward calculation (Clarke & Neuman, 2006; Cornish & Clarke, 1986; Dugan & Chenoweth, 2012). Religious groups, according to Enders and Sandler (2000), are less risk adverse because they often “place a high (infinite) value on reaching heaven through martyrdom” (p. 312). The process of othering further amplifies this value and consequently rationalizes acts of violence. Related work on target selection has found that those terrorists who adhere to a religious ideology, particularly manifestations of Islam, are more likely to direct violence toward the most vulnerable (Asal et al., 2009; Brandt & Sandler, 2010).

Although several studies have employed a rational choice framework to the study of terrorism (e.g., Carson, 2017; Clarke & Neuman, 2006; Dugan & Chenoweth, 2012; Dugan, LaFree, & Piquero, 2005; LaFree, Dugan, & Korte, 2009), Perry and Hasisi’s (2015) contribution is of particular relevance. These scholars contend that even suicide bombing can be considered a rational act, perpetrated for “self-gratifying” rewards. Such rewards include those associated with religion, like an afterlife for both the perpetrator and their family, or meeting the Prophet Muhammad and Allah. Perry and Hasisi (2015) concluded that even terrorists who kill themselves consider the costs and benefits of their behavior, based on the belief that the best possible path to jihad is through martyrdom.

This work on the ideology–lethality connection suggests that there is a relationship between the two, with caveats like that of Piazza’s (2009). Furthermore, this connection fits nicely within a rational choice framework as certain ideologies, particularly those based in religion, increase benefit perception. Based on this theoretical framework, the next section lays out a series of hypotheses regarding the GJM and lethality.

Current Study

This research examines the GJM from the broader, incident-based approach as in Gruenewald (2011) and Dugan and LaFree (2015), while focusing on the ideology–lethality connection as addressed in works like that of Asal and Rethemeyer (2008) and Asal and colleagues (2013). Specifically, we address, “Have GJM-related incidents been more lethal than those perpetrated by other groups?” Informed by a rational choice framework and the ideology–lethality literature, elements inherent to the GJM’s ideology should increase benefit perception associated with acts of terrorism. Specifically, we posit the following hypothesis:

Hypothesis 1: Incidents associated with the GJM will be more lethal and injurious than those perpetrated by other groups.

Related, and specific to the premise of Americans embodying the GJM’s “other,” we also hypothesize the following:

Hypothesis 2: Incidents associated with the GJM are more likely to hurt Americans than other ideologies.

Given the aforementioned research and similar work that speak to the unique harm yet distinct rationality (Brym & Araj, 2006; Piazza, 2009), we also examine suicide attacks as a separate outcome. We expect that GJM-related incidents will disproportionately involve this type of tactic, particularly given the focus on martyrdom.

Hypothesis 3: Incidents associated with the GJM are more likely to take the form of suicide attacks than other ideologies.

Thus, we examine the relationship between ideology and the outcomes of lethality, injury, American casualties, and suicide attacks. The next section reviews how we plan to operationalize these important constructs.

Data

The Global Terrorism Database (GTD), an open-source collection of terrorist incidents managed by National Consortium for the Study of Terrorism and Responses to Terrorism (START) and funded by the Department of Homeland Security (DHS), is the most comprehensive and inclusive database currently available. The GTD has a larger scope than any other terrorism dataset, including attacks that qualify as both domestic and international. This study utilized a January 2016 download of the GTD, which encompasses data through 2014 and identifies 141,966 incidents collected through open-source methods ranging from newspapers, wire services, and government reports. The GTD operates from a terrorism definition of “the threatened or actual use of illegal force and violence to attain a political, economic, religious or social goal through fear, coercion or intimidation” (LaFree & Dugan, 2007, p. 186). This definition translates to a mandatory set of criteria; namely, a rational act of the threat of/violence committed by a subnational actor (START, 2013). In addition, there are three screening criteria a researcher can use (goal, larger audience, and against international humanitarian law), although two of the three must be present for the incident to be represented in these data (and in our sample).²

Sample

From the GTD, this study first generated a global sample of incidents from 1994 to 2014 (N = 87,317; see Appendices A and B for a complete list of identified groups). This date range was chosen to avoid the documented issues with the 1993 data,³ along with the inherent importance of this time frame to the GJM (LaFree & Dugan, 2015; National Commission on Terrorist Attacks, 2004; START, 2013). Next, based on McCauley and Moskalenko’s (2014) conceptualization,⁴ we identified 287 GJM groups⁵ from the 3,188 total organizations represented in this sample.⁶ These coding procedures resulted in 17,026 GJM-related incidents from 1994 to 2014.

Carson (2017) utilized four main sources (Terrorist Organization Profiles, William Brantiff’s Congressional Testimony, Stanford’s Mapping Militant Organizations, and the Taliban/al Qa’ida Sanctions United Nations Committee), which identified 74 groups for inclusion. All groups in Carson’s study were included with the exception of four. Ansar Allah, Ansar Sarallah, Lashkar-e-Balochistan, and Pattani United Liberation Organization. Ansar Allah, Lashkar-e-Balochistan, and Pattani United Liberation Organization were determined to be more nationalistic or separatist than GJM; the lack of source documentation about Ansar Sarallah led to its exclusion. LaFree and Dugan (2015) also named 13 violent Islamic groups in their study, which are all included in this study.

Table 1 lists the 15 most prolific GJM terror groups between 1994 and 2014 by frequency of attacks. Strikingly, the top 15 groups in the GJM are responsible for 82.5% of attributed attacks. The Taliban, who are active mainly in the Afghanistan–Pakistan region, are uncontested as perpetrators with regard to frequency of incidents. ISIL, who have also operated under the names Tawhid and Jihad, al Qa’ida in Iraq (AQI), and the Islamic State of Iraq (ISI), is ranked with the second highest amount of attacks in this 20-year period. This is particularly troublesome because ISIL have only come into operational maturity in about the last 3 years. al-Shabaab and Boko Haram are ranked third and fourth, as two terror groups operating in Africa. al-Shabaab is one of al Qa’ida’s main affiliates in Africa, with their operational area as Somalia/Horn of Africa. Boko Haram was an al Qa’ida affiliate (until they pledged allegiance to ISIL) and operates primarily in Nigeria, accounting for nearly all of the country’s terror acts. Tehrik-i-Taliban Pakistan, a group that also conducts operations in Afghanistan–Pakistan region against the Pakistani government, rounds out the top five groups and those with over 1,000 attacks.

Table 1.

Top 15 Groups Tied to the GJM by Frequency.

Group name	Number of attacks	%
Taliban	4,258	30.3
ISIL^a	2,442	17.4
Al-Shabaab	1,737	12.4
Boko Haram	1,301	9.3
TTP	1,042	7.4
AQAP	755	5.4
Algerian Islamic Extremists	373	2.7
Hamas (Islamic Resistance Movement)	352	2.5
MILF	330	2.4
ASG	311	2.2
IG	250	1.8
GIA	236	1.7
AQLIM	221	1.6
Hizballah	218	1.6
GSPC	214	1.5
Total	14,040	100

Note. GJM = Global Jihadist Movement; ISIL = Islamic State of Iraq and the Levant; TTP = Tehrik-i-Taliban Pakistan; AQAP = Al-Qa’ida in the Arabian Peninsula; MILF = Moro Islamic Liberation Front; ASG = Abu Sayyaf Group; IG = al-Gama’at al-Islamiyya; GIA = Armed Islamic Group; AQLIM = Al-Qa’ida in the Lands of the Islamic Maghreb; GSPC = Salafist Group for Preaching and Fighting; AQI = Al-Qa’ida in Iraq.

Tawhid and Jihad, AQI, and Islamic State of Iraq (AQI) were earlier group names of ISIL. Their attack numbers are combined.

Measures

Affiliation

Our primary independent variable is GJM ideology for incidents tied to the aforementioned group list.⁷ This variable is operationalized through a dummy variable, where events with a GJM affiliation are coded as 1. As shown in Table 2, 44% of incidents are associated with the GJM when unknown perpetrator cases are eliminated.

Table 2.

Summary Statistics Without Unknowns (n = 38,330).

	M or %	SD	Minimum	Maximum
Dependent variables
Killed	0.544	0.498	0	1
Number killed	3.600	19.033	0	1,500
Wounded	0.408	0.491	0	1
Number wounded	4.711	41.786	0	5,500
U.S. killed	0.008	0.089	0	1
Number U.S. killed	0.116	10.677	0	1,358
U.S. wounded	0.005	0.073	0	1
Number U.S. wounded	0.054	3.832	0	650
Suicide	0.052	0.221	0	1
Independent variables
GJM	0.444	0.497	0	1
Success	0.924	0.265	0	1
Private citizens	0.258	0.437	0	1
Bomb use	0.458	0.499	0	1
Number of perpetrators	16.764	350.298	0	25,000
International	0.215	0.418	0	1
Claimed	0.259	0.438	0	1

Note. GJM = Global Jihadist Movement.

Dependent variables

We examine several dependent variables⁸ in our analyses: fatalities, injuries, American fatalities, American injuries, and whether an incident could be classified as a suicide attack.⁹ The GTD inputs the data for the first four as continuous variables, which we also recoded as dummy variables. Suicide is already coded as a dichotomous outcome, where 1 indicates evidence that the perpetrator did not expect to escape alive from the attack. Again, Table 2 demonstrates descriptive statistics for these outcomes, separated by affiliation measurement.

Attack characteristics

Also shown in Table 2, we control for a series of variables that could potentially be responsible for the relationship between ideology and our dependent outcomes. First, we include a measure for success, defined in the GTD, based on whether the incident’s attack type was achieved¹⁰ (for instance, a bomb that does not go off would be coded as unsuccessful and coded as 0). Next, we include measures for the target types and weapons¹¹ that are most prevalent in our outcomes of interest. These were the same across dependent variables (private citizens and use of bomb/explosive) with one exception; suicide attacks were most likely to involve a military target. We also incorporate a measure for the number of perpetrators with the assumption that a higher number would lead to differential dependent outcomes. Informed by prior research (Alakoc, 2017; Enders & Sandler, 2000; LaFree, Yang, & Crenshaw, 2009), we include a variable that designates whether an attack was international and whether it was claimed.¹² Finally, we employ a measure to indicate whether an attack was suicide based on the aforementioned research that finds these incidents to be unique.

Analytic Plan

We employed a two-part analysis to evaluate a myriad of outcomes. Given the first set of dependent variables was dichotomous, we first utilized logistic regression. We then analyzed the remaining dependent variables through count models. Such models are often employed in terrorism research given the overdispersion of zeros due to the rareness of the phenomenon. This was certainly the case here, where all four count dependent variables were highly skewed. Specifically, we utilized zero-inflated negative binomial models given that the means and variances were not equivalent, in which case a Poisson distribution would have been a better fit. This decision was reinforced by the α, a reflection of overdispersion, which yielded significance in all of our models.

Zero-inflated analyses allow for two independent decisions to be modeled—in this case, whether someone was killed or wounded in an attack, and second, the number of people killed or wounded in an attack. As Asal and colleagues (2013) described, this provides the opportunity for researchers to consider the intricacies involved in terrorism decision-making. In their example, they describe the possibility that terrorists might choose not to kill or even be incapable of killing. The Vuong (1989) tests all revealed that the zero-inflated models were superior to their baseline counterparts.

Results

Descriptive Trends

Figure 1 demonstrates the frequency of incidents by affiliation across time. Until 2012, unaffiliated events were been relatively steady, but at that point witness a 199% increase through the end of the series. The trends of GJM incident frequency follow a similar arc. Unsurprisingly, 2001 seems to be the year where an upward trend begins, rising exponentially in 2012 through 2014. From 2012, GJM-affiliated incidents have a 470% increase through 2014.

Figure 1.

Terror incidents by affiliation, 1994-2014.

Figures 2 and 3 show the average killed and wounded in attacks over time as disaggregated by affiliation. There are three major outliers in these data that represent the majority of fatalities and injuries. In 1998, the embassy bombings in Nairobi and Dar es Salaam are representative of the spike in average wounded per attack. The September 11 attacks explain the high average killed per attack for 2001. Last, the increase in average wounded for 2004 is likely representative of the War in Iraq turning from a decisive action conflict to a counterinsurgency, with the Battles of Fallujah and Mosul, among others, occurring in 2004.

Figure 2.

Average of individuals killed per terror event by affiliation from 1994 to 2014.

Figure 3.

Average of individuals wounded per terror event by affiliation from 1994 to 2014.

Table 3 illustrates the top 15 countries most affected by terrorism. Of particular note is the presence of two countries that have experienced frequent terrorism, but not any associated with the GJM. In the cases of Colombia (narco-terrorism, for example, Revolutionary Armed Forces of Colombia [FARC]) and Sri Lanka (separatists, for example, Liberation Tigers of Tamil Eelam [LTTE]), all of their terror incidents are unaffiliated. For India, the country with the highest incidence of terror acts, the overwhelming majority (90.6%) are unaffiliated with the movement. Turkey and the United Kingdom follow India’s pattern. Countries with the majority of terror acts associated with the GJM are as follows: Afghanistan (98.5%), Iraq (92.1%), Pakistan (67.3%), Somalia (96.7%), Nigeria (79.2%), Algeria (88.4%), and Yemen (68.3%).

Table 3.

Top 15 Countries of Terror Incidents by Frequency (n = 38,330).

Country	GJM		Unaffiliated
India	431	2.9%	4,140	31.9%
Afghanistan	4,355	29.1%	67	0.5%
Iraq	2,416	16.1%	207	1.6%
Pakistan	1,666	11.1%	811	6.3%
Colombia	0	0.0%	2,453	18.9%
Philippines	685	4.6%	1,224	9.4%
Somalia	1,574	10.5%	53	0.4%
Nigeria	1,251	8.3%	328	2.5%
Algeria	1,187	7.9%	156	1.2%
Yemen	782	5.2%	363	2.8%
Sri Lanka	0	0.0%	1,044	8.1%
Turkey	23	0.2%	787	6.1%
The United Kingdom	10	0.1%	686	5.3%
Israel	367	2.4%	272	2.1%
West Bank and Gaza Strip	240	1.6%	370	2.9%
Total	14,987	100.0%	12,961	100.0%

Note. GJM = Global Jihadist Movement.

Tables 4 to 6 represent the target, attack, and weapon types, respectively, separated by affiliation. As previously noted, the movement favors bomb use (and the use of bombing as a tactic), but so too do unaffiliated groups. Also noted, private citizens are a frequent target for both GJM and non-GJM classifications, but the military and police are the most common for attacks affiliated with the GJM. Contrarily, it is perhaps surprising that the targeting of businesses and transportation systems is actually skewed toward unaffiliated acts, especially given the intention of 9/11.

Table 4.

Target Types of Terror Incidents (n = 38,330).

Target type	GJM		Unaffiliated
Military and police	6,634	39.0%	5,495	25.8%
Private citizens and property	4,703	27.6%	5,782	27.1%
Government	1,969	11.6%	2,972	13.9%
Business	933	5.5%	2,582	12.1%
Transportation and systems	507	3.0%	1,357	6.4%
Educational institution	500	2.9%	476	2.2%
Religious figures/institutions	513	3.0%	422	2.0%
Terrorists/nonstate militia	423	2.5%	259	1.2%
Other	844	5.0%	1,961	9.2%
Total	17,026	100.0%	21,306	100.0%

Note. GJM = Global Jihadist Movement.

Table 5.

Attack Types of Terror Incidents (n = 38,330).

Attack type	GJM		Unaffiliated
Armed assault	4,499	26.4%	5,959	28.0%
Assassination	1,206	7.1%	1,126	5.3%
Bombing/explosion	8,173	48.0%	8,197	38.5%
Hostage taking (kidnapping)	1,496	8.8%	2,190	10.3%
Facility/infrastructure attack	480	2.8%	2,085	9.8%
Other	1,172	6.9%	1,747	8.2%
Total	17,026	100.0%	21,304	100.0%

Note. GJM = Global Jihadist Movement.

Table 6.

Weapon Types of Terror Incidents (n = 38,330).

Weapon type	GJM		Unaffiliated
Explosives	8,968	52.7%	8,576	40.3%
Firearms	5,384	31.6%	7,054	33.1%
Incendiary	453	2.7%	2,164	10.2%
Melee	431	2.5%	718	3.4%
Unknown	1,737	10.2%	2,641	12.4%
Other	53	0.3%	151	0.7%
Total	17,026	100.0%	21,304	100.0%

Note. GJM = Global Jihadist Movement.

Logistic Regression and Zero-Inflated Negative Binomial Results

Table 7 demonstrates the results from the logistic regression. As shown, ideology is a significant predictor of whether an incident involved a fatality, injury, or could be labeled as a suicide attack.¹³ In fact, the odds of an incident causing death is over 3 times more probable when tied to the GJM. Similarly, GJM incidents are over 1.7 and 6 times more likely to involve an injury and be classified as a suicide attack, respectively. These findings do seem to marginally translate to American casualties. GJM incidents were significantly more likely to kill Americans but less likely to injure them.

Table 7.

Logistic Regression Odds Ratios and Standard Errors Without Unknowns (n = 38,330).

Variable	Killed	Wounded	American killed	American wounded	Suicide
GJM	3.108*** (0.078)	1.677*** (0.039)	1.349* (0.199)	0.539*** (0.093)	5.837*** (0.397)
Success	8.068*** (0.495)	3.688*** (0.191)	7.113*** (4.152)	2.329* (0.909)	0.943 (0.086)
Private citizens	1.726*** (0.046)	1.130*** (0.030)	0.414*** (0.086)	0.502** (0.114)	—
Military	—	—	—	—	1.421*** (0.084)
Bomb use	0.379*** (0.009)	1.992*** (0.047)	0.439*** (0.071)	0.747 (0.139)	50.552*** (8.189)
Number of perpetrators	1.000* (0.051)	0.999 (0.000)	0.994 (0.006)	0.987 (0.011)	0.998 (0.002)
International	0.624*** (0.018)	0.816*** (0.024)	11.684*** (1.785)	5.597*** (0.897)	1.251*** (0.071)
Claimed	1.005 (0.028)	1.040 (0.028)	2.230*** (0.304)	1.276 (0.211)	2.697*** (0.137)
Suicide	38.8007*** (0.125)	5.189*** (0.353)	4.963*** (0.932)	6.872*** (1.526)	—

Note. GJM = Global Jihadist Movement.

p > .05. **p > .01***p > .001.

Table 8 exhibits the results from the count models. Interestingly, the GJM ideology is significant in only one of the four models, that of number killed. This indicates that as an incident becomes associated with the movement, its total fatality count increases. However, ideology does not seem to significantly influence the number of injuries nor number of American casualties.

Table 8.

Zero-Inflated Estimates and Standard Errors (n = 38,330).

Variable	Number killed	Number wounded	Americans killed	Americans wounded
GJM	0.105*** (0.019)	0.005 (0.022)	0.199 (0.263)	0.212 (0.311)
Success	0.573*** (0.066)	0.847*** (0.061)	0.201 (0.823)	1.950** (0.678)
Private citizens	0.234*** (0.019)	0.210*** (0.024)	2.097*** (0.276)	−1.371*** (0.385)
Bomb use	0.076*** (0.019)	0.579*** (0.022)	−0.785*** (0.184)	0.261 (0.232)
Number of perpetrators	0.005*** (0.000)	0.001*** (0.000)	0.048*** (0.009)	0.002 (0.012)
International	0.170*** (0.023)	0.165*** (0.027)	0.577* (0.280)	0.026 (0.341)
Claimed	−0.003 (0.006)	0.002** (0.007)	0.085 (0.056)	0.010 (0.069)
Suicide	0.810*** (0.031)	1.193*** (0.033)	1.545*** (0.193)	0.151 (0.238)

Note. GJM = Global Jihadist Movement.

p > .05. **p > .01***p > .001.

Discussion

As a whole, our hypotheses are partially supported indicating that, at the very least, the GJM is distinct. Despite the fact that the majority of terror acts in the last two decades is unaffiliated with the movement, these incidents are more likely to involve a casualty and have a larger fatality count. Interestingly, this only marginally translates to Americans as the GJM is less likely to injure this group, likely a mechanism of the rareness of GJM attacks on U.S. soil. In fact, only 0.0004% of incidents occurred within the U.S. Homeland. The GJM also favors suicide attacks, which present their own set of issues (Brym & Araj, 2006; Piazza, 2009).

The GJM has, in the last two decades, become a dominant manifestation of terrorism. This is not necessarily through frequency nor through a distinctive targeting strategy. The movement also does not appear to utilize unique weapons or attack types than that of other ideologies save suicide attacks similar to findings in prior research (Barros & Proença, 2005; Dugan & LaFree, 2015; Spaaj & Hamm, 2015). Rather, the GJM has a distinct ability to kill and maim civilians. al Qa’ida alone has averaged 63.2 deaths per incident, which is a 761% higher average than the next group.

These findings, viewed from within a rational choice framework, may very well be evidence that the GJM’s ideology increases the perceived benefits associated with acts of terrorism. As Asal and Rethemeyer (2008) described, the GJM’s ideology is “intrinsically linked” to deadliness through its constructs of a supernatural audience and a distinct “other.” According to Piazza (2009), the nexus of religion and ideology in terrorism set the conditions for such lethality, with suicide attacks viewed as a redemptive or “purifying” opportunity. Other work on the religious ideology–lethality connection reinforces the notion that the GJM’s ideology is unique in its ability to justify violence or, in rational choice terms, lower the costs of terrorism (Drake, 1998; Enders & Sandler, 2000; Laqueur, 1998). In contrast, movements that adhere to more pacifist philosophies, like that of the radical eco-movement, rarely harm civilians (Carson, LaFree, & Dugan, 2012).

While ideology is one important factor, there may be other mechanisms driving this GJM–lethality connection. One such factor is the GJM’s shift from a restrictive Salafi-Jihad model employed by al Qa’ida to the inclusive and redemptive “Call to Arms” model promoted by ISIL. This shift has created an extremely challenging environment for security services to operate in given the difficulty of identifying whether or not an individual presents a credible terrorist threat (Braniff & Moghadam, 2011; Rabasa et al., 2006). The current reality of isolated individuals conducting terror attacks in accordance with a terror group’s ideology with no preexisting contact yields a distinct capacity to affect mass casualty incidents without early warning.

Regardless of whether ideology or increasing decentralization is responsible, these findings provide important policy implications for crafting and executing a counterterrorism strategy. Our research supports the sentiment of policymakers, like that of Former Director Clapper, that GJM-related terrorism is a threat to national security. The GJM has demonstrated an ability to be more lethal and as such policymakers should continue to focus their counterterrorism efforts. When considering a target hardening strategy, countering small arms and explosives against soft targets should be a policy priority given the GJM’s operational capacity to inflict casualties in this method of attack. Undoubtedly, plans and training for mass casualty incidents that max response resources, such as a counter-WMD (weapons of mass destruction) event, are important. However, our study indicates that the likelihood of an attack from small arms and explosives is far greater.

Another counterterrorism suggestion is to shift focus from these traditional target hardening strategies to those centered on ideology. This has long been the focus of various radicalization programs, which have yielded success. A related line of research finds conciliatory actions as the most effective way to counter terrorism in other contexts (Dugan & Chenoweth, 2012). Although controversial, these strategies may hold promise for combating the GJM above and beyond punitive strikes. It would also appear that such punitive measures have not been effective thus far and may even produce backlash effects (Dugan & Chenoweth, 2012; Hafez & Hatfield, 2006; LaFree et al., 2009).

As with all empirical research, this study is not without its limitations. First, it is difficult to arrive at a consensus when defining the GJM. While this investigation relied on McCauley and Moskalenko’s (2014) inclusive definition to guide our systematic operationalization, there very well could be omissions. Alternatively, and while several quality control measures were employed, groups included in our sample may fit this overly broad definition while only tangentially connected to the GJM’s ideology. Future research should examine strength and variation of associations, as done by one recent ISIL report (START, 2016). Second, and relatedly, there are a large number of cases that cannot be attributed to a terror group and therefore be assigned affiliation. To address this issue, we ran our analyses with and without such groups’ inclusion, which yielded similar findings. Last, this research employs the GTD for its data. Inherently, there is the possibility of measurement error in using such a database in two manners. First, the data could be skewed based on modern technological advances; better communication, the Internet, and television media coverage could all be responsible for the higher number of incidents. Certainly, as Behlendorf, Belur, and Kumar (2016) maintained, open-source databases can overrepresent those incidents that cause causalities. Second, as Ackerman and Pinson (2016) stated, the possibility of credibility and validity issues exists. With regard to terrorism research, there are no existing data sets that evaluate such factors (Ackerman & Pinson, 2016). Given these limitations, the GTD is still arguably the most comprehensive and well-defined open-source database of terrorist incidents in existence. Parkin and Green (2016) concluded as much by stating the GTD “collect[s] valid and reliable data from open-sources” (p. 683). This study also incorporates Freilich and LaFree’s (2016) strategy of disaggregating the incident data by creating two separate terror groups to provide comparative study. Future research should persist in doing so, along with continuously looking for ways to measure and define the dynamic phenomenon that is the GJM.

Conclusion

As the homicide literature has previously established, an ideological motivation leads to distinct outcomes. The terrorism research has reinforced this notion by finding a connection between not just ideology, but type of ideology, and the outcome of lethality. Our research has discovered this to be the case with the GJM, an ideology with a supernatural audience and a defined other.

The study of ideology, with particular attention to the modern phenomenon of the GJM, is likely to become increasingly vital to the formulation of effective counterterrorism policy. Despite finding a majority of terror incidents unaffiliated with the GJM in the last two decades, the last 2 years of the timeline indicate otherwise. This pattern, a majority of attacks from 2012 to 2014 as GJM-related, is compounded by such incidents being more lethal and injurious than those perpetrated by other movements.

The leadership position of al Qa’ida in the movement has been recently challenged by ISIL, when they declared a new Caliphate. Not only is there a core difference in ideology, but there are operational differences between the two groups as well. al Qa’ida focuses on attacks requiring a high amount of resources and a large degree of planning, which yield high payoffs. On the other side, groups like ISIL, Boko Haram, and al-Shabaab are more interested in frequency of attacks, which allow them to appear more prolific and present. How these important intergroup dynamics play out will likely impact the threat of the GJM in future. All in all, empirical research from a criminologist perspective is vital to a deeper understanding of the complexities of the global security environment.

Footnotes

Appendix A

Appendix B

Appendix C

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Jennifer Varriale Carson is an associate professor of Criminal Justice and the Coordinator of Undergraduate Research at the University of Central Missouri. She holds a Ph.D. from the University of Maryland in Criminology and Criminal Justice. Her work focuses on policy evaluation, particularly the use of quasi-experimental methods in assessing counterterrorism efforts, and can be found in a number of outlets including Criminology and Public Policy, Journal of Quantitative Criminology, and Deviant Behavior.

Matthew Suppenbach received his B.S. (2011) and M.S. (2016) in Criminal Justice from the University of Central Missouri.

References

Ackerman

G. A.

Pinson

L. E.

(2016). Speaking truth to sources: Introducing a method for the quantitative evaluation of open sources in event data. Studies in Conflict & Terrorism, 39, 617-640.

Agbiboa

D. E.

(2013). Why Boko Haram exists: The relative deprivation perspective. African Conflict & Peacebuilding Review, 3, 144-157.

Alakoc

B. P.

(2017). Competing to kill: Terrorist organizations versus lone wolf terrorists. Terrorism and Political Violence, 29, 509-532.

Asal

Gill

Rethemeyer

R. K.

Horgan

(2013). Killing range explaining lethality variance within a terrorist organization. Journal of Conflict Resolution, 59, 401-427.

Asal

Rethemeyer

R. K.

(2008). The nature of the beast: Organizational structures and the lethality of terrorist attacks. The Journal of Politics, 70, 437-449.

Barros

C. P.

Proença

(2005). Mixed logit estimation of radical Islamic terrorism in Europe and North America: A comparative study. Journal of Conflict Resolution, 49, 298-314.

Behlendorf

Belur

Kumar

(2016). Peering through the Kaleidoscope: Variation and Validity in Data Collection on Terrorist Attacks. Studies in Conflict & Terrorism, 39(7-8), 641-667.

Benmelech

Berrebi

Klor

(2010, October). Counter-suicide-terrorism: Evidence from house demolitions. Cambridge, MA: National Bureau of Economic Research.

Bergen

(2001). Holy War, Inc.: Inside the secret world of Osama Bin Laden. New York, NY: Free Press.

10.

Bergen

(2011). Al-Qaeda, the Taliban, and other extremist groups in Afghanistan and Pakistan: Testimony before the Senate Foreign Relations Committee. Retrieved from http://www.operationspaix.net/DATA/DOCUMENT/4119~v~Al_Qaeda_the_Taliban___Other_Extremist_Groups_in_Afghanistan_and_Pakistan__Testimony_before_the_US_Senate_Committee_on_Foreign_Relations_by_Peter_Ber.pdf

11.

Borum

Gelles

(2005). Al-Qaeda’s operational evolution: Behavioral and organizational perspectives. Behavioral Sciences and the Law, 23, 467-483.

12.

Braniff

Moghadam

(2011). Towards global jihadism: Al-Qaeda’s strategic, ideological and structural adaptations since 9/11. Perspectives on Terrorism, 5(2), Article 52. Retrieved from http://www.terrorismanalysts.com/pt/index.php/pot/article/view/braniff-towards-global-jihadism/html

13.

Brandt

P. T.

Sandler

(2010). What do transnational terrorists target? Has it changed? Are we safer? Journal of Conflict Resolution, 54(2), 214-236.

14.

Brym

R. J.

Araj

(2006). Suicide bombing as strategy and interaction: The case of the second intifada. Social Forces, 84, 1969-1986.

15.

Byman

(2015). Al Qaeda, the Islamic state, and the Global Jihadist Movement: What everyone needs to know. New York, NY: Oxford University Press.

16.

Cameron

(2015, November). Prime Minister statement on Paris terrorist attack. Retrieved from https://www.gov.uk/government/news/prime-minister-statement-on-paris-terror-attack

17.

Carson

J. V.

(2017). The role of targeting killings in the “War on Terror.” Criminology & Public Policy, 16(1) 191-220.

18.

Carson

J. V.

LaFree

Dugan

(2012). Terrorist and non-terrorist criminal attacks by radical environmental and animal rights groups in the United States, 1970–2007. Terrorism and Political Violence, 24, 295-319.

19.

Clapper

J. R.

(2016). Statement for the record: Worldwide threat assessment of the US Intelligence Community. Retrieved from http://www.dni.gov/files/documents/Newsroom/Testimonies/HPSCI_Unclassified_2016_ATA_SFR-25Feb16.pdf

20.

Clarke

R. V. G.

Newman

G. R.

(2006). Outsmarting the terrorists. Portsmouth, NH: Greenwood Publishing Group.

21.

Condra

L. N.

Shapiro

J. N.

(2012). Who takes the blame? The strategic effects of collateral damage. American Journal of Political Science, 56, 167-187.

22.

Cornish

Clarke

(1986). Situational prevention, displacement of crime and rational choice theory. Situational Crime Prevention: From Theory into Practice, 1-16.

23.

Drake

C. J. M.

(1998). The role of ideology in terrorists’ target selection. Terrorism and Political Violence, 10(2), 53-85.

24.

Dugan

Chenoweth

(2012). Moving beyond deterrence: The effectiveness of raising the expected utility of abstaining from terrorism in Israel. American Sociological Review, 77, 597-624.

25.

Dugan

LaFree

Piquero

A. R.

(2005). Testing a rational choice model of airline hijackings. Criminology, 43(4), 1031-1065.

26.

Enders

Sandler

(2000). Is transnational terrorism becoming more threatening? A time-series investigation. The Journal of Conflict Resolution, 44, 307-332.

27.

Fielding

Shortland

(2010). “An eye for an eye, a tooth for a tooth”: Political violence and counterinsurgency in Egypt. Journal of Peace Research, 47, 433-447.

28.

Forest

J. J.

(2012). Confronting the terrorism of Boko Haram in Nigeria (Joint Special Operations University Report No. 12-5). Retrieved http://jsou.libguides.com/ld.php?content_id=2876918

29.

Freilich

J. D.

LaFree

(2016). Measurement issues in the study of terrorism: Introducing the special issue. Studies in Conflict & Terrorism, 39, 569-579.

30.

Gruenewald

(2011). A comparative examination of homicides perpetrated by far-right extremists. Homicide Studies, 15, 177-203.

31.

Gruenewald

Pridemore

W. A.

(2012). A comparison of ideologically-motivated homicides from the new extremist crime database and homicides from the supplementary homicide reports using multiple imputation by chained equations to handle missing values. Journal of Quantitative Criminology, 28, 141-162.

32.

Haddad

(2004). A comparative study of Lebanese and Palestinian perceptions of suicide bombings: The role of militant Islam and socio-economic status. International Journal of Comparative Sociology, 45, 337-363.

33.

Hafez

M. M.

Hatfield

J. M.

(2006). Do targeted assassinations work? A multivariate analysis of Israel’s controversial tactic during Al-Aqsa Uprising 1. Studies in Conflict & Terrorism, 29, 359-382.

34.

Hamm

Spaaij

(2015). Lone wolf terrorism in America: Using knowledge of radicalization pathways to forge prevention strategies. Retrieved from https://www.ncjrs.gov/pdffiles1/nij/grants/248691.pdf

35.

Hegghammer

(2006). Global jihadism after the Iraq War. Middle East Journal, 60, 11-32.

36.

Hepworth

D. P.

(2014). Terrorist retaliation? An analysis of terrorist attacks following the targeted killing of top-tier al Qaeda leadership. The Journal of Policing, Intelligence and Counter Terrorism, 9, 1-18.

37.

Hollande

(2015). Speech by the President of the Republic before a joint session of Parliament. Retrieved from http://www.diplomatie.gouv.fr/en/french-foreign-policy/defence-security/parisattacks-paris-terror-attacks-november-2015/article/speech-by-the-president-of-the-republic-before-a-joint-session-of-parliament

38.

LaFree

Dugan

(2007). Introducing the Global Terrorism Database. Terrorism and Political Violence, 19, 181-204.

39.

LaFree

Dugan

(2016). Global terrorism and the deadliest groups since 2001. In Backer

Ravinder

Huth

(Eds.), War and peace 2016 (pp. 67-78). New York, NY: Routledge.

40.

LaFree

Dugan

Korte

(2009). The impact of British counterterrorist strategies on political violence in Northern Ireland: Comparing deterrence and backlash models. Criminology, 47, 17-45.

41.

LaFree

Yang

S. M.

Crenshaw

(2009). Trajectories of terrorism. Criminology & Public Policy, 8(3), 445-473.

42.

Laqueur

(1998). The new face of terrorism. The Washington Quarterly, 21(4), 167-178.

43.

Lewis

(1996). Islam and liberal democracy: A historical overview. Journal of Democracy, 7(2), 52-63.

44.

Linke

A. M.

’Witmer

F. D. O.

Loughlin

(2012). Space-time granger analysis of the War in Iraq: A study of coalition and insurgent action-reaction. International Interactions, 38, 402-425.

45.

Lum

Kennedy

L. W.

Sherley

A. J.

(2006). The effectiveness of counter-terrorism strategies. Campbell Systematic Reviews. Retrieved from http://campbellcollaboration.org/lib/project/11/

46.

Lyall

(2009). Does indiscriminate violence incite insurgent attacks? Evidence from Chechnya. Journal of Conflict Resolution, 53, 331-362.

47.

McCauley

Moskalenko

(2014). Toward a profile of lone wolf terrorists: What moves an individual from radical opinion to radical action. Terrorism and Political Violence, 26, 65-91.

48.

Merkel

(2015). Speech by federal chancellor Angela Merkel at the reception for the diplomatic corps. Retrieved from https://www.bundesregierung.de/Content/EN/Reden/2015/2015-07-13-diplomatisches-corps_en.html

49.

Moghadam

(2008). The Salafi-Jihad as a religious ideology. Combating Terrorism Center Sentinel, 1(3), 14-16.

50.

Mousseau

(2011). Urban poverty and support for Islamist terror: Survey results of Muslims in fourteen countries. Journal of Peace Research, 48, 35-47.

51.

National Commission on Terrorist Attacks. (2004). The 9/11 commission report: Final report of the National Commission on Terrorist Attacks Upon the United States. New York, NY: W.W. Norton.

52.

National Consortium for the Study of Terrorism and Responses to Terrorism. (2013). Codebook: Inclusion criteria and variables. Retrieved from http://www.start.umd.edu/gtd/downloads/Codebook.pdf

53.

Obama

(2015). Remarks by the President in closing of the summit on countering violent extremism. Retrieved from https://www.whitehouse.gov/the-press-office/2015/02/18/remarks-president-closing-summit-countering-violent-extremism

54.

Obama

(2016). Remarks by President Obama and President Macri of Argentina in joint press conference. Retrieved from https://www.whitehouse.gov/the-press-office/2016/03/23/remarks-president-obama-and-president-macri-argentina-joint-press

55.

Pantucci

(2011). A typology of lone wolves: Preliminary analysis of lone Islamist terrorists. The International Centre for the Study of Radicalisation and Political Violence. Retrieved from http://icsr.info///a-typology-of-lone-wolves-preliminary-analysis-of-201104lone-islamist

56.

Parkin

W. S.

Freilich

J. D.

(2015). Routine activities and right-wing extremists: An empirical comparison of the victims of ideologically- and non-ideologically-motivated homicides committed by American far-rightists. Terrorism and Political Violence, 27, 182-203.

57.

Parkin

W. S.

Freilich

J. D.

Chermak

S. M.

(2015). Ideological victimization homicides perpetrated by far-right extremists. Homicide Studies, 19, 211-236.

58.

Parkin

W. S.

Green

D. A.

(2016). Terrorism in the news: The efficiency and impact of sampling methods on data collection and content analysis. Studies in Conflict & Terrorism, 39, 668-686.

59.

Perry

Hasisi

(2015). Rational choice rewards and the jihadist suicide bomber. Terrorism and Political Violence, 27(1), 53-80.

60.

Piazza

J. A.

(2009). Is Islamist terrorism more dangerous? An empirical study of group ideology, organization, and goal structure. Terrorism and Political Violence, 21, 62-88.

61.

Rabasa

Chalk

Cragin

Daly

S. A.

Gregg

H. S.

Karasik

T. W.

. . . Rosenau

(2006). Beyond al-Qaeda: Part 1. The Global Jihadist Movement. Santa Monica, CA: RAND.

62.

Schmid

A. P.

(2011). Introduction. In Schmid

A. P.

(Ed.), The Routledge handbook of terrorism research (pp. 1-38). New York, NY: Routledge.

63.

Spaaij

Hamm

M. S.

(2015). Key issues and research agendas in lone wolf terrorism. Studies in Conflict & Terrorism, 38(3), 167-178.

64.

Stanford Mapping Militant Organizations. (2012). Ahlu Sunna Wal Jama. Retrieved from http://web.stanford.edu/group/mappingmilitants/cgi-bin/groups/view/109

65.

Stenersen

(2010, February 5). The Taliban insurgency in Afghanistan—Organization, leadership and worldview. Kjeller: Norwegian Defence Research Establishment.

66.

Terrorism Research and Analysis Consortium. (2016b). Ahlu Sunna Waljama’a. Retrieved from http://www.trackingterrorism.org/group/ahlu-sunna-waljamaa-aswj

67.

Torres

M. R.

Jordán

Horsburgh

(2006). Analysis and evolution of the Global Jihadist Movement propaganda. Terrorism and Political Violence, 18, 399-421.

68.

Vuong

Q. H.

(1989). Likelihood ratio tests for model selection and non-nested hypotheses. Econometrica: Journal of the Econometric Society, 57, 307-333.

69.

Warrick

J. A.

(2014). Radical Islamic terrorism. Small Wars Journal, 10(3). Retrieved from http://smallwarsjournal.com/jrnl/art/radical-islamic-terrorism

70.

Wilner

A. S.

(2010). Targeted killings in Afghanistan: Measuring coercion and deterrence in counterterrorism and counterinsurgency. Studies in Conflict & Terrorism, 33, 307-329.

71.

Young

J. K.

Findley

M. G.

(2011). Promise and pitfalls of terrorism research. International Studies Review, 13, 411-431.

72.

Zelin

A. Y.

(2014). The war between ISIS and al-Qaeda for supremacy of the Global Jihadist Movement. Retrieved from http://www.washingtoninstitute.org/uploads/Documents/pubs/ResearchNote_20_Zelin.pdf

73.

Zussman

(2006). Assassinations: Evaluating the effectiveness of an Israeli counterterrorism policy using stock market data. The Journal of Economic Perspectives, 20, 193-206.