Helping or hurting? The impact of foreign fighters on militant group behavior

ABSTRACT How do foreign fighters affect militant group behavior? Recent studies have examined the impact of foreign fighters on broader conflict outcomes, but we explore their specific impact on group behavior. Using foreign fighter data, we find that the presence of foreign fighters is positively related to group longevity, use of suicide operations, and the geographic spread of its operations. We elaborate on these findings in a case study of foreign fighters in al-Shabaab. This article provides important contributions to both scholarship and policy, elucidating the ideological and logistical impact of foreign fighters on tactics, target choices, and group lifespan.


Introduction
As more than 40,000 foreign fighters flowed into the civil conflict in Syria, 1 that began in 2011, questions arose regarding the threat they might eventually pose upon leaving the battlefield, as well as why they joined conflicts in which they had no direct stake in the first place. Comparatively less attention was placed on two related questions: How do foreign fighters influence the militant organizations they join and Do they expand the group's operational capabilities, change its attack behavior, and prolong its existence? These questions are important, as the number of foreign fighters who have mobilized internationally has grown in recent years. 2 On the other hand, foreign fighters are usually a small minority of combatants in conflicts; therefore, so there may also be constraints on their influence in the groups they join.
Despite the renewed interest on foreign fighters in academia and government, significant gaps remain in our understanding of the impact of foreign fighters specifically on organizations' trajectories throughout conflicts.
This article address this gap in the following manner. First, we examine the literature, highlighting the impact of foreign fighters on conflict outcomes and militant groups. This examination shows how previous research has laid a foundation for work that explores how foreign fightersaffect group behavior and outcomes. Second, we articulate our theoretical expectations regarding how foreign fighters affect grouplevel behavior and outcomes. Third, we discuss the research design, including data and methods, of our quantitative approach. Fourth, we conduct the quantitative analysis, which reveals that groups with foreign fighters are more likely to use suicide attacks, tend to carry out terrorist attacks in a larger number of countries, and last longer than groups that do not have foreign fighters. Fifth, we present a case study of the involvement of foreign fighters in al-Shabaab, a jihadist-oriented group operating in Somalia with a long-standing relationship with foreign fighters. This case study provides important context to the quantitative findings, highlighting the nuanced relationship between foreign fighters and the groups they join. We then offer some concluding comments based on our findings and directions for future research.

Foreign fighters and their impacts
In scholarship to date, there is ongoing debate about how to define foreign fighters. As Mendelsohn pointed out, 'the concept is still clouded by ambiguities'. 3 Drawing from Malet, Hegghammer, and Arielli's foundational works, we define foreign fighters as individuals who operate within the confines of an insurgency based on a personal decision; are not affiliated with an official military organization and have not been sent by their government; are not primarily motivated by personal gain; and lack citizenship in the country of the conflict. 4 Overall, the academic literature on foreign fighters has grown significantly over the past decade. In general, this research is divided into three stages of the foreign fighter 'lifecycle': 1) why individuals become foreign fighters, the pre-war phase; 2) what they do in conflicts, the war phase; 3) what those individuals do after they participate in a conflict, the postwar phase. 5 To date, most attention has focused on the pre-and postwar phases. 6 Work on the pre-war question has grown more rapidly due to events in Syria, including the However, others see foreign fighters as offering strategic-level advantages, even shaping conflict outcomes to the benefit of insurgents. Malet found that foreign fighter 'recruitment occurs when local insurgents attempt to broaden the scope of a conflict so as to increase their resources and maximize their chances of victory'. 14 Malet observed that opposition movements with foreign fighters are disproportionately likely to win, or at least not lose. 15 Chu and Braithwaite found evidence in their analysis of civil conflicts from 1946 to 2013 that conflicts involving foreign fighters saw a decrease in the likelihood of government victory by more than 430%. Thus, they conclude that foreign fighters represent 'more of a resource [for domestic insurgents] than a liability'. 16 Still, Arielli dissents, claiming that foreign fighters are not more likely to be found on the winning side, let alone be responsible for the overall outcome. He points out that foreign fighters often join multiple sides in a conflict, concluding that 'the mere incorporation of foreigners into a fighting force does not tell us much about that side's effectiveness or chances of success '. 17 Some qualitative work on specific cases supports this argument as well. In an examination of the role of Arab fighters in the anti-Soviet jihad -one of the largest mobilizations of foreign fighters to date -Williams assessed that foreign fighters 'did not make much of an impact' on the eventual defeat and withdrawal of Soviet forces. 18 Although these works and debates advance our understanding of foreign fighters' impacts, their focus has been on the overall conflict. The current study emphasizes a different level of analysis: the individual organizations within opposition movements. Conclusions about foreign fighters' effect on insurgent movements writ large may be premature without examining the impact these fighters have on individual organizations. Foreign fighters sometimes flock to minor organizations with little sway over opposition movements or conflict outcomes. In addition, they do not necessarily join all the groups in an opposition movement or at least join them all equally, which can be particularly consequential because the power distribution within opposition movements can shape conflict outcomes. 19 In other words, in addition to examining foreign fighters' effects on opposition movements and conflict outcomes overall, we propose examining whether and how they affect the organizations they join compared to organizations with no foreign fighters.

The effect of foreign fighters on militant organizations
What impact might we expect foreign fighters to have on organizational behavior? One point of consensus within the literature is that foreign fighters can lead to more extreme behavior, in terms of tactics and targets, in insurgencies. As Byman put it, 'foreign fighters often make war harsher '. 20 This makes sense because foreign fighters, who have already made the decision to leave home to volunteer in a far-away conflict, tend to be more ideologically committed and willing to go to extremes to achieve their objectives. 21 Thus, ideological gaps may arise between locals, who may be driven more by circumstances and proximity, and foreign insurgents. This gap was acutely evident in Afghanistan, where foreign fighters' willingness and even eagerness to die in combat -both during the 1980s anti-Soviet jihad and the Taliban's war against the Northern Alliance during the 1990sconfused and disturbed locals. 22 In addition to the ideological commitment foreign fighters bring, they also may provide resources and manpower. Desire for manpower in particular encourages competition between groups in fragmented insurgencies 23those with two or more competing groups -and groups face pressures to accommodate foreign fighters' preferences, even if more violent and less discriminate, in order to retain them. Indeed, foreign fighters can be less constrained and disciplined in their violence. 24 For example, foreign fighters in Bosnia operated a unit during the 1990s that became notorious for being ruthless and engaging in violent excesses. 25 Further, foreign fighters may also encourage more extreme tactics. Drawing from the social movement literature, Bakke traced how foreign fighters can influence local insurgents' 'repertoires of actions' by offering more extreme tactical innovations. 26  of understanding of local dynamics, culture, and language, foreign fighters' ruthlessness may alienate the local population and reduce local support, an accusation leveled against foreign fighters in Iraq, Somalia, and elsewhere.
Within the growing the literature surrounding foreign fighters' propensity for extreme violence, no scholarship has yet examined whether groups with foreign fighters specifically are more likely to adopt different tactics and targets than their counterparts without foreign fighters. This is the question that drives our hypotheses: do groups with foreign fighters engage in more extreme tactics and forms of violence? We argue that this tendency may manifest itself in three ways: attacks on soft targets, suicide attacks, and external attacks.
One form of more extreme violence is a greater willingness to target civilians, especially by striking soft targets, meaning nonmilitary, police, or government locations. Asal et al. point out that '[a]ttacks on "soft targets" inspire a different level of fear and convey a different ideological and political message'. 27 They find that groups with a religious ideology and more capable organizations have a higher probability of attacking soft targets. 28 Foreign fighters' ideological commitment and connections can contribute to both of these dynamics. Indeed, both Moore and Doctor found that the presence of foreign fighters in a conflict is associated with higher levels of anti-civilian violence. 29 Does this finding translate to variation in group-level behavior? In other words, are groups with foreign fighters more likely to deliberately target civilians?
It is possible that foreign fighters tend to join conflicts in which insurgent movements already have a propensity to engage in violence against civilians. Yet, foreign fighters may also influence groups' willingness to deliberately target civilians by reducing organizations' connection to the public. Doctor found that the presence of foreign fighters reduces rebel organizations' need to rely on the civilian population for support. 30 By extension, this may increase the propensity of groups with foreign fighters to strike soft targets. Overall, if groups with foreign fighters are less restrained in their violence toward civilians, then we should see them strike soft targets more than groups without foreign fighters. This leads to the first iteration of the hypothesis that fighters impact the operational activities of militant organizations: Striking soft targets is only one way in which foreign fighters may escalate the violence of the militant organizations they join. Another way through which groups can escalate their tactics is the use of suicide attacks. Suicide operations are attacks in which the death of the operative is the means by which the attack is accomplished. 31 Not only do such attacks signal organizations' high levels of commitment, 32 suicide terrorism also causes a significantly larger number of casualties per attack than other types of attacks. 33 In the literature examining suicide operations, studies have focused on everything from societal conditions to individual-level motivations. For example, Harmon et al. found that the organizational-level factors contributing to the use of suicide attacks received less attention compared to individual-level motivations or societal factors. 34 However, some exceptions indicate that group-level factors may influence the adoption of this tactic. Bloom found that suicide attacks can be the product of outbidding as militant groups compete for resources, recruits, and public support. 35 Ayers argued that organizations use suicide terrorism strategically to gain tactical advantage over adversaries and aid in recruitment efforts. 36 Of course, militant groups without foreign fighters have used suicide attacks, most notably, the Liberation Tigers of Tamil Eelam between 1982 and2000. 37 Still, Horowitz argues that the use of suicide operations depends on a group having the organizational capacity to use suicide bombing, as well as the desire to adopt it. 38 Foreign fighters may bolster an organization's ability and motivation to use the tactic through their willingness to conduct such attacks and their ideological zeal. Thus, if foreign fighters contribute to the extremity of organizations' violence, then groups with foreign fighters may also be more likely to use suicide attacks. While we are not arguing that foreign fighters are the sole or even primary motivating cause of suicide attacks, we hypothesize that the incorporation of foreign fighters may increase a group's willingness and ability to use this tactic. In some cases, foreign fighters have clearly had this impact. Bakke finds that the introduction of suicide attacks in Chechnya followed the Beyond the tactical impact, foreign fighters also bring a transnational dimension to groups. By definition, foreign fighters have interests outside the conflict zone, and, thus, some harbor political motivations beyond the immediate conflict. Byman found that foreign fighters tend to 'regionalize' conflicts. Not only does this expand the theater of operations, it also risks provoking outside interventions in conflicts. 42 Foreign fighters can expand a group's area of operations in two ways. First, foreign fighters, especially when they are part of a group's senior ranks, can influence an organization's decision to carry out attacks in other countries, leading it to deploy operatives and sponsor attacks elsewhere. For instance, in 2005, al-Qaida in Iraq conducted coordinated attacks on three hotels in Jordan, the country of origin of the group's leader Abu Musab al-Zarqawi. In addition, Rich and Conduit found that foreign fighters expanded insurgents' political goals in Chechnya and Syria. 43 In these cases, foreign fighters contributed to a shift that broadened those conflicts to become part of a larger project, a narrative that provided ideological justification for groups to conduct operations beyond the immediate conflict zone.
Second, foreign fighters sometimes return home from conflicts and carry out attacks in the name of the group. 44 Indeed, in some cases, foreign fighters have an equal or greater commitment to causes outside of the conflict they joined, as was the case for many Egyptian foreign fighters involved in the 39  anti-Soviet jihad in Afghanistan during the 1980s. 45 In both scenarios, groups with foreign fighters are more likely to conduct operations in multiple countries, leading to Hypothesis 1c:

Hypothesis 1c: Groups that incorporate foreign fighters conduct attacks in a larger number of countries than groups that do not incorporate them
Finally, we consider the impact that foreign fighters have on the survival of terrorist organizations. As discussed above, foreign fighters reflect a different type of variable in that they are a part of the group, intended to add strength, resources, and credibility to the organization. If it is the case that foreign fighters bring these things with them, then it follows that groups with foreign fighters should last longer than groups without them.
There is some evidence that supports this conclusion. As Malet has noted, it is difficult to find cases in which foreign fighters (and, by extension, the groups they join) have not been funded on some level by states, let alone the financial interest of private parties. 46 This funding inevitably finds its way into the hands of groups willing to accept these fighters, thereby strengthening the organization. When al-Qaida in Iraq/the Islamic State of Iraq brought in foreign fighters from 2006-2007, the group's internal documents showed that about a quarter of incoming fighters brought some monetary donation. Other declassified documents from this time show that nearly 70% of one entity's funding came from foreign donations. 47 While these numbers may not hold across all groups, certainly attracting foreign fighters at least has the possibility of a financial upside, which should ultimately provide support for a group's continued existence.

Research design
To test these hypotheses, we first created our dependent variables to measure the different group-level indicators potentially impacted by the presence of foreign fighters, as well as group-level data on the presence of foreign fighters. This information was then combined with previously collected data on a number of control variables for each terrorist organizations. For our initial list of groups, we used the static version of Acosta's Revolutionary and Militant Organizations Dataset (REVMOD) on the use of suicide attacks by militant organizations to create our initial list of organizations. 48 REVMOD contains 310 groups from a variety of sources from 1980-2014, including the Global Terrorism Database (GTD). However, because we relied exclusively on the GTD for the coding of some of our main variables of interest (as described below), we were only able to find data for 250 of the REVMOD organizations. We took these groups and added the various outcome variables related to the above enumerated hypotheses.
In the case of Hypothesis 1a, we created the variable Soft Targets. For each group we took the total number of attacks against soft targets and divided over the group's total number of attacks. The attack data came from the GTD, and we utilized the target type variable within GTD to identify attacks against soft targets. We considered any attack that was not against government facilities or personnel, military targets, or police facilities or personnel to be a soft target attack. This variable is a proportion that ranges from 0 to 1, with lower values indicating that a group carries out fewer attacks against soft targets and a higher value suggesting more attacks against soft targets.
For Hypothesis 1b, we used information on suicide attacks already present within the Acosta dataset. The variable Suicide Attacks was created by taking the total number of suicide attacks and dividing it over the group's total number of attacks. To test Hypothesis 1 c, we created a count variable, Countries Attacked, which is simply a tally of the number of countries in which a group carried out a terrorist attack in a given year according to the GTD. Finally, because Hypothesis 2 examined the duration of terrorist groups, we utilized a binary variable, Group Ended, created by Acosta to account for groups that had terminated by the end of his dataset. 49 Approximately 48% (119 organizations) ended during the time period covered by the dataset and are no longer active.
Next, we collected information on the use of foreign fighters by each of the REVMOD organizations. A research assistant reviewed previous scholarly work and conducted group-by-group searches in Lexis Nexis to identify if groups had foreign fighters within their ranks. 50 We then created a binary variable to account for the groups that incorporated foreign fighters. Of 250 organizations in our study, we found evidence that 44 of them utilized foreign fighters, or about 18% of the total. 51 48 We initially coded our foreign fighter data on top of the original release of REMVOD. Benjamin Acosta, 'Dying for survival: Why militant organizations continue to conduct suicide attacks', Journal of Peace Research 53/2 (2016), 180-196. Acosta has since released a more updated version that includes more organizations. 49 In the codebook for the REVMOD dataset, Acosta notes that the end of a group was coded as 'the date of elimination by force or an officially declared date of dissolution'. 50 Malet,Foreign Fighters. 51 In our initial data collection effort, we had also hoped to find information on two more key variables: the timing in which foreign fighters participated in the organization and the size of the foreign fighter contingent within the organization. Unfortunately, we were unable to find enough quality data on these two variables to have sufficient confidence in the data to include it in our analysis.
We also include several control variables, most of which, unless specifically noted, come from Acosta's dataset. 52 Group Age is the number of years during which the group existed as of 2014, assuming that it did not end. Size is measured by two binary variables, One Thousand and Ten Thousand, for groups with more than 1,000 and more than 10,000 members, respectively. The ideological orientation of groups is captured by two binary variables, Islamist and Leftist, with all other types of organizations being relegated to the reference category. Another binary variable, Self-Determination, accounts for organizations whose ultimate aims were to achieve a separate geographic space over which they could exercise political control. Attacks is a count of the total number of attacks attributed to each specific organization and comes from the GTD. State Sponsor is a binary variable, which indicates whether the militant group in question received state support, including through active assistance such as arms or less tangible forms of diplomatic support. Safe Haven is a count variable and contains the number of places in which the terrorist organization, either because of state support or open territory due to lack of policing, can find respite from counterterrorism pressure. The final two measures, Polity and GDP, account for two important characteristics, political institutions and economic wealth, of the organization's primary state adversary. Polity measures the state's political institutions on a 20-point scale, with countries that score low being more autocratic and countries that score high more democratic. GDP is a measure of the state's gross domestic product, which some scholars have argued may serve as a rough proxy for counterterrorism capability.
Because these outcomes require different variables to test the underlying logic, the quantitative method utilized varies from one analysis to the next. In the analysis of Hypothesis 1a, which uses a proportion as the dependent variable, we employed a generalized linear model with a logit link function. The dependent variable in Hypothesis 1b is a binary variable, so we ran that analysis with a probit model. To test Hypothesis 1c, which is based on a count variable of the number of countries in which the terrorist group has carried out operations, we opted for a negative binomial model. Finally, because Hypothesis 2 examines the mortality of terrorist groups, we use a Cox proportion hazards model. Each of these approaches is consistent with previous research utilizing similar types of data.

Analysis
We present the results of the analyses for Hypotheses 1a-1 c using coefficient plots as shown in Figure 1. 53 The first column, which presents coefficient estimates and standard errors for the generalized linear model that tests Hypothesis 1a, do not provide support for the hypothesis. The foreign fighter variable has a negative sign, which is contrary to what we anticipated, but is statistically insignificant. In other words, there does not appear to be a statistically meaningful relationship between the presence of foreign fighters and the tendency of groups to target a higher proportion of soft targets. Beyond the main variable, there is only one variable that was statistically significant in that model. The more operational activity a group engages in overall, as represented by the number of attacks that it carries out, the less it targets soft targets.
Next, the results in the middle column of Figure 1 provide support for Hypothesis 1b. Groups that incorporate foreign fighters utilize suicide terrorism at a higher rate than those that do not, and the difference is statistically significant. To offer a more substantive interpretation, we computed the predicted probability of a group employing suicide attacks when they included foreign fighters as opposed to when they did not. 54 This result is that groups with foreign fighters are about 54% more likely to use suicide attacks than groups without foreign fighters. This is a significant impact and suggests that the relationship between foreign fighters and terrorist groups' tactics is worth exploring further.
The control variables in this model also yielded several statistically significant results, which were consistent with previous literature. Groups that have been around longer, are modestly sized, adhere to an Islamist or Leftist ideology, or have access to a safe haven are all more likely to employ suicide bombing.
The final analysis in Figure 1 is of Hypothesis 1c, which posited that the number of countries in which a group conducted operations would be positively correlated with the presence of foreign fighters. As shown in the results indicated by the hollow circle in Figure 1, this result is also statistically significant, showing that groups with foreign fighters are also groups that carry out attacks in more countries. Calculations of the substantive impact of this effect, show that the expected number of countries attacked for a group without foreign fighters is 2.15, while the expected number for a group with foreign fighters is 3.22. 55 This may seem like a minor difference, but it is important to keep in mind two pieces of contextual information. First, 70% of 53 A table of results for these analyses appears as Table A1 in the online appendix. 54 This calculation was made using Gary King's 'clarify' software. In each instance in which we simulated predicted values to calculate substantive effects, the rest of the variables had to be held at specific levels. All continuous variables were held at their means, while binary variables were set at fixed levels as follows: One thousand = 0, Ten thousand = 0, Islamist = 1, Leftist = 0, Self-Determination = 0, State sponsor = 0, Safe Haven = 1). 55 This calculation was made using the 'margins' command in Stata. the groups in our dataset carry out attacks in one or two countries. Second, combating a militant organization that operates across borders is challenging, so even an expected expansion of one country is not trivial.
The results for the control variables in this analysis suggest some important relationships. Groups that execute more attacks, have a state sponsor, or have access to safe havens all conduct attacks in a larger number of countries. Interestingly, large groups with more than 10,000 members tend to carry out their operations in a smaller number of countries.
The final analysis explores how foreign fighters impact an organization's longevity. As discussed above for Hypothesis 2, our argument is that although groups with foreign fighters will face challenges, overall the additional resources and publicity should be a net positive for the organization's survival. Before discussing the results, it is important to remember that the coefficients in hazard models represent the impact of the variables on the likelihood that a group ends. Thus, coefficients under 1 show factors that increase the likelihood that a group survives, while coefficients over 1 are for factors that increase the likelihood that a group terminates. 56 The results in Figure 2 show that foreign fighters do seem to have a negative impact on the likelihood that a group ends, which means that groups with foreign fighters seem to last longer. It is worth noting that the result is just outside conventionally accepted levels of statistical significance (p-value = 0.053). Nevertheless, this finding indicates that the incorporation of foreign fighters does not seem to have a negative impact on a group's survivability, notwithstanding the other challenges such fighters may pose to a group.
To get a better picture of the substantive impact of foreign fighters on the demise of terrorist organizations, Figure 3 plots two survival rates, one for groups with foreign fighters (dashed and dotted line) and one for groups without foreign fighters (solid line). The x-axis of the Figure shows the age of the terrorist organization, such that the line's progression from left to right represents the general decreasing trend in group's chances of ending over its lifetime. As this graphical representation makes clear, foreign fighters have a substantively important effect on the survival rate of terrorist groups.
In summary, the preceding analyses have shown that there is a relationship between the presence of foreign fighters and both group behavior and duration. When foreign fighters are present in an organization, it lasts longer, is more likely to conduct suicide attacks, and carries out operations in a larger number of countries than an organization without foreign fighters. To examine these quantitative results at a more nuanced level, we now transition to a discussion of the role of foreign fighters in al-Shabaab. Table A1 in the online appendix.

Foreign fighters and al-Shabaab
In this section, we examine the role of foreign fighters in al-Shabaab, specifically examining the hypotheses tested above. Al-Shabaab offers a useful case to examine because it has had significant within-case variation over time. Foreign fighters have been present since its founding and remain present today. Al-Shabaab was the first organization in Somalia to adopt suicide operations and is consistently among the most lethal groups in Africa. In addition, al-Shabaab expanded its operations beyond Somalia in 2010 to Uganda and has since targeted Kenya, Ethiopia, and Djibouti. But what, if any, was the role of foreign fighters in these developments? Finally, al-Shabaab is sixteen years old as of this writing with no signs of its demise on the horizon. How have foreign fighters shaped its longevity? We explore these questions below to offer more insight into the relationship between foreign fighters and the behavior of terrorist organizations.

Introduction to al-Shabaab
Since its inception in 2005, al-Shabaab has adopted a range of organizational forms ranging from a faction within a broader umbrella organization to a quasi-state. Beginning as the radical fringe of the Islamic Court Union (ICU), al-Shabaab emerged as an independent organization after the Ethiopian invasion, which overthrew the ICU in early 2007. It became the leading insurgent group opposing Ethiopian forces and the Somali Transitional Federal Government and gained control of an estimated 80% of Somalia after the Ethiopian withdrawal in 2009. The African Union Mission in Somalia (AMISOM) rolled back the group's territorial gains between 2011 and 2015, and al-Shabaab evolved into a hybrid organization. It now controls rural areas, exerts influence through its courts system and 'taxation' collection, and terrorizes Mogadishu and AMISOM-contributing countries, particularly Kenya. 57 Al-Shabaab has several aims, although within-group variation exists around the importance of local grievances versus global jihadist goals. Overall, it seeks to install an Islamic state in Somalia and the surrounding Somali-dominated regions of East Africa in accordance with its interpretation of Islam and as part of al-Qaida's envisioned caliphate. The first step to achieving this is the expulsion of foreign forces, specifically AMISOM, and the overthrow of the 'apostate' Somali Federal Government.

Foreign fighters in Somalia
Al-Shabaab has welcomed foreign fighters since its inception. As a faction of the ICU, al-Shabaab protected a cell of East African al-Qaida operatives, whom the group proudly touted once the ICU took control of Mogadishu and welcomed more foreign fighters into its ranks. 58 As an independent group, al-Shabaab has consistently called for foreign fighters, especially after the Ethiopians invaded in 2007, and again as the Ethiopians withdrew and AMISOM's forces expanded throughout 2009. The group featured -and continues to feature -foreign fighters in its propaganda, urging others to join them.
However, foreign fighters have struggled in Somalia. Their presence provoked heightened counterterrorism pressure. They were caught in al-Shabaab's internecine conflict between 2011 and 2013. Those who opposed the group's now deceased founding leader Ahmed Abdi Godane were killed, detained, or fled. Moreover, Somalia became a less enticing location during the Arab Spring and once Syria became the focal point for foreign fighters. 59 Since 2014, the foreign fighters that remain in al-Shabaab are primarily from East Africa, mostly Kenya, and to a lesser extent Yemen. Defining foreign fighters in Somalia is particularly complex. 60 As discussed earlier, this article defines foreign fighters as individuals who (1) operate within the confines of an insurgency based on a personal decision, (2) are not affiliated with an official military organization and have not been sent by their government, (3) are not primarily motivated by personal gain, (4) and lack citizenship in the country of the conflict. But all the complexities Mendelsohn noted about the concept of foreign fighters are present in Somalia case. 61 We exclude residents of Somaliland from our definition of foreign fighters as the international community has not recognized the region's 1991 declaration of independence. However, we include other ethnic Somalis in the region and diaspora members because colonial borders, the state collapse, and the mass exodus that followed left many ethnic Somalis without Somali citizenship. Thus, while they retain clan and familial ties in Somalia, technically they fit the definition of foreign fighters used in this article: a very expansive definition in this case. Including such diaspora in the same category as fully foreign fighters, i.e., non-Somalis, risks conflating the different roles they may play in the group; however, to ensure consistency across the qualitative and quantitative elements of this article, we include them in this analysis while noting the caution required in interpretation.
Still, even including ethnic Somalis, it becomes clear when examining al-Shabaab's attacks on soft targets, its suicide operations, and its external operations that Somalis from Somalia drive strategy and decision-making. Overall, foreign fighters do not appear to determine al-Shabaab's targets, tactics, or theater of operations; but they do help execute the extreme violence the Somali leadership establishes.

Soft targets
Did foreign fighters contribute to al-Shabaab's propensity to attack soft targets? In the quantitative analysis, we did not find a meaningful relationship in this regard, which held in the qualitative process tracing. Because al-Shabaab has hosted foreign fighters and attacked soft targets since its inception, it is difficult to establish a causal relationship. However, al-Shabaab did increase its propensity to strike soft targets after 2012, at a time when foreign fighter presence and interest in Somalia was actually declining and the group was experiencing significant military defeats. Moreover, by this point, the most influential foreign fighters in al-Shabaab, the East African al-Qaida operatives, were all dead. Moreover, of note, one such operative even criticized al-Shabaab for its attacks that killed Muslim civilians, suggesting that foreign fighters were not driving this soft targeting. 62 In examining al-Shabaab's attacks on soft targets inside of Somalia, there are three main types: strikes against hotels, operations against restaurants, and targeted assassinations of civilians. We examine each in turn.
Al-Shabaab began systematically and strategically targeting Mogadishu hotels and restaurants in January 2015 as part of its bid to reanimate its insurgency and intimidate officials, diplomats, and businessmen. 63 And although it had conducted multiple hotel attacks before 2015, between October that year and October 2016, al-Shabaab increased attacks on '"soft" civilians targets, such as hotels and restaurants' refer to (Figure 4). 64 The group views hotels in particular as legitimate targets since 'they provide lodgings for members of the apostate government, certain members of the diaspora, foreigners and other infidels'. 65 Hotels in Mogadishu provide al-Shabaab access to targets that would otherwise be difficult to strike, though even hotels in Mogadishu tend to be semi-hardened in part because of al-Shabaab's attacks.
Compiling the data from both ACLED and the GTD datasets, we identified 48 al-Shabaab attacks on hotels since 2007. Of those 48 attacks on hotels, nine of them had participants with more than Somali citizenship or residency; however, all participants were ethnic Somalis. 66 The group's Amaliya Inquimas unit of the Amniyat, which is responsible for complex suicide operations, conducts most of the group's hotel attacks. 67 In addition to an already selective and independent recruitment process for the Amniyat, members of the Amaliya Inquimas were carefully chosen and given rigorous training with recruits competing for the right to carry out suicide operations. 68 Overall, the Amniyat is primarily Somali, 69 with 'family and clan connections to high-level Al-Shabaab commanders'. 70 Another major soft target al-Shabaab frequently strikes are restaurants. Unlike hotel attacks, al-Shabaab does not identify restaurant attackers in its propaganda, and there is little news reporting available to identify attackers, so foreign fighters' role as operatives in these attacks is more difficult to determine. However, since many of these attacks are simple suicide operations or IED attacks, they likely come under the purview of the Amniyat's Amaliya Istishhad unit, which is responsible for al-Shabaab's 'traditional' suicide bombings. 71 The available information indicates that this division of the Amniyat is also mostly Somali. 72 Further, as noted earlier, by 2013 when the group's hotel and restaurant attacks began to grow, al-Shabaab's foreign fighter ranks were in significant decline.
Since al-Shabaab's inception, targeted assassinations of civil servants, businessmen, civil society activists, government officials, and security figures have been a central tactic. 73   also target parliamentarians and individuals participating in the 2016 elections. 76 Between October 2015 and November 2018, al-Shabaab sustained its campaign of targeted assassinations; there was no significant escalation in number of attacks but lethality of the attacks spiked in June 2016. 77 Because assassins are rarely caught, identifying those who are foreign fighters is difficult. But foreigners tend to be conspicuous in Somalia, thus it is unlikely they are used as assassins regularly. However, foreign fighters did provide training in assassination techniques while the group was under the ICU. 78 The successful assassinations of high-level Transitional Federal Government officials provide evidence of the ICU/al-Shabaab's growing expertise in this tactic. 79 Al-Shabaab reopened training camps in April 2007, and foreign volunteers trained fighters on assassination techniques along with other military tactics. 80 Despite foreigners' early involvement in training al-Shabaab on assassination techniques, the direction to increase assassinations and other insurgency style tactics came from al-Shabaab's Somali leadership. Specifically, in February 2011, Godane directed al-Shabaab leaders to conduct 'supplementary activities such as vehicle-borne improvised explosive device attacks, assassinations, and roadside bombs'. 81 Additionally, two Somalis were identified as the Amniyat's head of assassinations in 2013: Mohamud Nuh Aden 82 and Abdilaahi Haji Daud. 83 Al-Shabaab has also recognized when its attacks on soft targets are mistakes or miscalculations, and sometimes refuses to claim such attacks or scapegoats foreigners. One unclaimed attack was the October 2017 twin truck bombing in Mogadishu that killed at least 587. In, a December 2009 attack on a graduation ceremony, the group blamed the attack on a foreigner, probably referring to the diaspora member who conducted the suicide mission. 84 This examination of al-Shabaab's attacks on soft targets suggests that foreign fighters have played a limited, auxiliary role, not a causal one. They did not drive the group's targeting decisions and only participated in a minority of operations. While they assisted al-Shabaab with developing its assassination capabilities, the group now possesses an indigenous capability independent of these fighters.

Suicide operations
Second, did foreign fighters influence al-Shabaab's decision to use suicide operations? The quantitative analysis found that groups that incorporate foreign fighters are more likely to use suicide bombing than groups without them. For many years, al-Shabaab was the first and only organization in Somalia to adopt suicide operations, until the Islamic State in Somaliaanother organization with foreign fighters -conducted its first suicide attack in 2017.
When al-Shabaab began using suicide operations -a tactic once considered unthinkable in Somalia -the East African al-Qaida operatives were cooperating closely with the group. As veteran al-Qaida members involved in prior suicide operations, many government officials and analysts point to them as facilitating the introduction to the tactic. 85 The subsequent role of foreign fighters appears to be some involvement in training and indoctrination. 86 While there is evidence of some foreign fighter involvement in training for suicide operations, as with soft targeting, the available information indicates that Somalis ultimately decide when to use the tactic and are the primarily operatives. Somalis have been named as the commanders of al-Shabaab's suicide operations, including Ibrahim Ali Abdi who reportedly 'planned the group's suicide missions and directed the preparation of attackers' vests, road-side explosives and car bombs' at least after 2009 until his death in 2013. 87 In 2013, the U.N. found al-Shabaab's Amniyat was responsible for al-Shabaab's suicide operations. Specifically the same Somali responsible for the Amniyat's assassinations unit, Abdilaahi Haji Daud, was also the leader of its suicide squad. 88 Later, when Abdishakur Tahlil was killed in December 2014, a Somali was identified as the leader of al-Shabaab's suicide attack unit. 89 As noted earlier, the Amniyat has two suicide operations units, one dedicated to 85 Abdinasir Mohamed and Sarah Childress, 'Suicide Bombing Kills Somali Ministers, Students', Wall Street Journal, 4 Dec. 2009. 86 Hansen, Al-Shabaab in Somalia, 93;David Shinn, 'Al Shabaab's Foreign Threat to Somalia', Orbis 55/2 (2011), 203-15, 209. 87 Mike Pflanz, 'US Drone Strike in Somalia Killed Al-Shabaab "chief Bomb-Maker"', The Telegraph, 29 Oct. 2013;Heidi Vogt andAbdalle Ahmed Mumin, 'U.S. Kills Somali Militant', Wall Street Journal, 29 Oct. 2013. 88 S/2013/413, 93. 89 Eric Schmitt, 'Drone Strike in Somalia Is Said to Kill Shabab Leader', The New York Times, 30 Dec. 2014, sec. Africa. traditional suicide operations and another dedicated to suicide assaults with gunmen and a suicide bomber, 90 and is dominated by Somalis with familial or clan connections to high-level al-Shabaab commanders. 91 In terms of operatives, foreign fighters rarely serve as suicide attackers. AMISOM General Mugisha remarked in 2010 that foreigners were not 'too interested in dying themselves, instead they would prefer to train others for martyrdom'. 92 Among the few foreign fighters who have been suicide operatives, the majority are ethnic Somalis. 93 In 2013, the U.N. noted that there have been no foreign fighters used in suicide operations; however, they exclude individuals of Somali descent from being classified as foreign fighters. 94 While finding that foreign suicide operatives were unusual, Warner and Chapin note that East African foreign fighters have traveled to Somalia to become suicide bombers, but the years-long waiting list may mean these regional recruits have yet to be trained or deployed. 95 In sum, foreign fighters had a role in importing the tactic and subsequent training. However, al-Shabaab institutionalized the use of it, and foreign fighters have had a limited role in the execution of such attacks. Also of note, the inclusion of ethnic Somalis increases the role of foreign fighters in these operations, as they are the primary foreign fighters who become suicide operatives.

External operations
Overall, the quantitative analysis found that groups with foreign fighters carry out attacks in more countries, and al-Shabaab's track record is consistent with that finding. However, given the fact that foreign fighters were a part of al-Shabaab since its inception, the fact that it did not begin to conduct operations outside of Somalia until 2010 suggests a nuanced relationship. Since then, it has targeted AMISOM-contributing countries to compel their withdrawal from Somalia, including in Uganda, Kenya, Djibouti, and attempted several attacks in Ethiopia. Of those, non-Somali citizens have been involved in attacks in Uganda and Kenya. We have been unable to determine the nationality of some individuals, and the available reporting may be incomplete in terms of all the planners and facilitators. However, based on the information available, we can conclude that the planning of external operations has generally come from the Somali leadership, and the decision to expand operations outside of Somalia rested with the group's now deceased leader, Godane.
In 2010, al-Shabaab conducted suicide bombings of a restaurant and bar in Kampala during the World Cup, killing over 70 people. The operatives and facilitators of the attack in Kampala were mostly non-Somali, including Kenyans, Ugandans, a Tanzanian, and a Rwandan. Additionally, there is information to suggest that al-Shabaab relied on the operational expertise of the sole surviving al-Qaida East African operative, Harun Fazul, in the attack planning. While one of the rare exceptions to the tendency for the Somali leadership to control attack planning, Fazul's relationship with Godane was strained by this point, suggesting Godane was already inclined to execute external operations and did not do so at Fazul's behest. 96 In September 2013, four al-Shabaab gunmen attacked the Westgate shopping mall in Nairobi, killing nearly 70 people. All the attackers were Somalis, though one had Norwegian citizenship and thus counts as a foreign fighter. There is conflicting reporting on the nationalities of the facilitators, though at least one appears to be a Kenyan citizen. While most of planners were Somali, reports conflict about whether Abdirahman Sandhere, aka 'Ukash', was Kenyan or Somali. 97 As al-Shabaab's senior operations commander at the time, he reportedly reviewed the attack plan and submitted it to Godane for final approval. 98 In April 2015, four al-Shabaab gunmen conducted a siege of Garissa University, killing nearly 150 people. 99 Details on the gunmen are limited, but one attacker was identified as an ethnic Somali Kenyan; in fact, he was the son of a chief in Mandera County. 100 In addition, a Tanzanian carrying explosives was arrested immediately after the attack, suggesting he was an attacker that survived the police response. 101 Two Kenyans were arrested and convicted on terrorism charges for delivering weapons to the attackers. 102 The Kenyan government identified an ethnic Somali Kenyan as the mastermind of the Garissa attack. 103 In January 2019, al-Shabaab struck a hotel and office complex in Nairobi with gunmen and a suicide operative. During the overnight siege, 21 people were killed. Bryden and Bahra noted that the DusitD2 attack was unique because it was the first 'martyrdom operation' that non-ethnic Somali Kenyans planned, led, and executed. 104 Other than the 2010 attack in Kampala, this is a rare instance in which non-Somalis were involved in directing an operation.
Al-Shabaab has also sought to establish external units, with varying success. Godane established two military wings in late 2013 to grow al-Shabaab's external operations. The first, focusing on Ethiopia, was led by Ethiopian-Somali Ali 'Diyaar', but it struggled to execute operations there. The second, Jaysh Ayman, was to focus on Kenya, Uganda, and Tanzania. 105 It has had considerably more success. Jaysh Ayman's foot soldiers draw from the areas it was tasked to attack, 106 with a core contingent of Kenyan fighters and a smaller number of Tanzanian, European, and Somali members. 107 It has conducted a series of attacks in Kenya, particularly in Lamu county.
Overall, foreign fighters helped al-Shabaab conduct its operations outside of Somalia, often serving as operatives for the group and the backbone of the external operations units. However, while al-Shabaab's access to such operatives expanded its external operations, the impetus for external operations came from al-Shabaab's Somali leadership, rather than foreign fighters refer to ( Figure 5).

Resilience
Al-Shabaab has proven an unusually resilient organization in the Somali context, and it shows no signs of demise in the foreseeable future. In tracing the group's evolution, foreign fighters played a role in helping the group to launch an insurgency against invading Ethiopian forces in 2007. In a matter of months, al-Shabaab underwent a rapid tactical evolution that took years elsewhere. 108 Foreign fighters assisted this adaptation by offering insurgency expertise, particularly relating to fabricating IEDs, suicide vests, and vehicleborne IEDs, which contributed to the rapid evolution of IED use in Somalia -a tactic that continued over the next few years. 109 More broadly, Shinn explained how in the early years foreign fighters served an important role as al-Shabaab field commanders because of their military experience. They brought specialized skills . . . They teach the techniques of suicide attacks, remote-controlled roadside bombings, kidnappings and assassinations of TFG officials, journalists, and humanitarian and civil society workers that constitute a threat to al Shabaab's goal of establishing a strict Islamic state completely under its control. 110 However, over time, al-Shabaab incorporated these tactics and had less need or use for foreign fighters.
The contributions were not always positive, and foreign fighters had drawbacks to al-Shabaab's longevity as well. In particular, foreign fighter involvement significantly increased counterterrorism pressure against the group. In the absence of foreign fighters, the U.S. would have had a limited interest in the conflict in Somalia -a country the U.S. studiously ignored after its disastrous intervention in 1993. But the involvement of East African al-Qaida operatives in Somalia spurred the U. S.' early counterterrorism campaign, actions that backfired and led to the rise of the ICU. Later, as some Americans joined al-Shabaab as foreign fighters and the group allied with al-Qaida, the U.S. had significant concerns about the potential threat to its homeland. The result is that the U.S. pushed for the creation of AMISOM, conducted airstrikes against al-Shabaab, and supported the Somali Federal Government, particularly the special operations unit Danab, which is the most effect anti-Shabaab force in Somalia. Today, al-Shabaab's continued resilience is most clearly derived from its divided and weak adversaries, as well as the group's ability to capitalize on local grievances in Somalia. 111 Foreign fighters are almost exclusively from East Africa. They provide supplementary manpower for the group, but the crux of its personnel still come from Somalia, and Somalis still lead the group.

Case study conclusion
The examination of al-Shabaab provides a nuanced understanding of foreign fighters, refining the takeaways of the quantitative analysis (Table 1). While it suggests that foreign fighters do play some role in soft targeting, suicide operations, external attacks, and resilience, it also provides a caution to the idea that foreign fighters drive such actions. Their influence is clearest at two points: early in the conflict and after years of contributions. Foreign fighters helped impart skills and provide manpower for early attacks, especially assassinations, suicide operations, and the first external attack in Kampala. Once domestic insurgents integrated the skills foreign fighters taught, they had less use for foreigners. Conversely, as the attack in Kenya on Dusit illustrates, foreign fighters can gain more responsibility for external operations as they gain experience and the trust of their domestic counterparts. In sum, although the al-Shabaab case highlights the centrality of domestic insurgents, it also shows that they used foreign fighters to bolster and implement the strategies determined by domestic insurgents.

Conclusion
Although the foreign fighter problem in Iraq and Syria has not disappeared, the inward flow has receded. Despite this fact, there is no reason to think that foreign fighters joining militant organizations is a phenomenon of the past and that individuals will stop being willing to travel in support of groups that invite such individuals into the fold. As a result, research which aims to explore all aspects of the foreign fighter problem set is much needed. Our aim here was to examine one part of this phenomenon: how foreign fighters impacted the behavior of the groups they joined across a number of different outcomes. In this way, we sought to build upon previous scholarly work that demonstrated how foreign fighters can have both positive and negative influences on the groups they join. While our quantitative and qualitative analysis did not show that foreign fighters affect every aspect of an organization, there were a few key results it highlighted: the use of suicide bombing, the number of countries in which a group carries out operations in, and the longevity of the organizations themselves. Far from concluding that foreign fighters have a causal role in leading groups to act in these ways, however, the case study of al-Shabaab offered a nuanced perspective on how the impact of foreign fighters may change over the lifespan of a group. This question of the dynamic impact of foreign fighters, something our quantitative analysis was not structured to answer, is something worth additional future research. Future research may also benefit from examining other organizational-level impacts using other datasets. 112 It is important to recognize, however, that aside from the finding related to increased longevity, it is not clear from our results whether the other relationships between foreign fighters and groups are ultimately a positive or negative factor for the organizations in which they participate. Even if foreign fighters help train operatives to engage in suicide bombings or encourage greater ideological extremism, it remains to be seen how this affects the group's sustainability and appeal. In short, we believe that our findings here provide ample evidence of the need to continue this line of examination to better understanding the impact of foreign fighters.

Disclosure statement
No potential conflict of interest was reported by the author(s).
112 For example, the Big, Allied and Dangerous (BAAD) 2 dataset would be a good candidate for such a comparative analysis in the future. Unfortunately, because the groups contained in BAAD 2 do not completely overlap from those included in the Acosta dataset, collecting that data would have required more time and resources than were available.

Data availability statement
The data supporting the results presented in the paper can be found at https:// danieljmilton.wordpress.com/.