Research Note: Pathways of Foreign Fighters: An in-Depth and Comparative Study Based on Dutch Probation Files

Abstract

Foreign fighter (FF) backgrounds and their pathways have been studied before, yet most work is based on open sources. In this article, we examine FFs by examining uniquely acquired data from Dutch probation files. Focusing on profound life events and using insights from social control theorists and the concept of redemption, we provide a general overview of the group through different themes. Moreover, unlike most previous studies, we make comparisons between the FFs based on age, sex, and timing of (attempted) travel. Besides empirical findings regarding profound events, mental health and criminal histories, results demonstrate that a radical social (online) environment was critical in becoming a FF. Also, the explorations of the age and sex comparisons demonstrate the necessity of looking beyond the FF group as a whole.

View correction statement:

Correction

With the perceived threat posed by foreign fighters (FFs), researching pathways and characteristics of FFs has received increased attention, particularly in relation to the most recent Syrian war. In the European Union, for example, researchers estimated in 2016 that between 3,922 and 4,294 FFs had left for Syria/Iraq, of which around 30% had returned to their home countries, potentially posing a safety threat.¹ In a comparative analysis of 34 articles using empirical data on Western FFs in Syria and Iraq, Dawson summarizes findings on FF backgrounds: the majority were young men with an average age of 26 years (about 18% pertains women); most were single, a third were married and had children; and about 15% were converts. According to Dawson, data on socioeconomic variables, mental health, and the nexus between crime and terrorism are largely inadequate, limited, and sometimes conflicting, leading to the conclusion that data on FF’s backgrounds are still inconclusive in important ways and more systematic, methodically rigorous and comparative research is needed.²

Moreover, although there are some studies based on primary data (e.g. interviews with FFs), research on pathways and characteristics of FFs is predominantly based on open sources. Over the last decade, researchers have examined the FF phenomenon related to the Syrian case by addressing, for instance, motives and/or backgrounds of jihadist FFs³, non-jihadist FFs⁴, and (perspectives on) dealing with returnees.⁵ There are also studies looking specifically at FFs from the Netherlands.⁶ However, besides studies by Weenink⁷ and Leyenhorst and Andreas⁸, official file data from Dutch government agencies regarding FFs have not been examined yet. Additionally, the FF group is oftentimes examined as a whole rather than exploring the heterogeneity within the group.

To fill this research gap, we will (I) examine uniquely acquired Dutch probation file data, which allows us to incorporate a range of factors measured at the time of becoming an (attempted) FF, and (II) include an exploration of subcategories within the researched FF group. Unlike most previous studies, we include differentiations based on age (youths versus adults), sex (male versus female), and timing—between those travelling either prior to or after the declaration of a caliphate in 2014 by jihadist group ISIS (also known as ISIL or Daesh). Consequently, we demonstrate the necessity of looking beyond the foreign fighter group as a whole. Our research question is: What do FFs from the Netherlands look like regarding signs/triggers leading up to their terrorist crime, life domains, social network, health and personality characteristics, criminal background, and online activities, and can we detect differences comparing FFs based on age, sex, and timing of travelling?

Foreign Fighters

As frequently mentioned by researchers, foreign fighting is not a new phenomenon; there are numerous examples of conflicts, such as the War in Afghanistan and the recent War in Ukraine, where FFs or foreign “volunteers” were part of an armed struggle.⁹ According to Malet, most participants mobilize in response to a perceived threat and recruiters—across highly varied conflicts in space and time—persistently utilize the same frame of defensive mobilization.¹⁰ In terms of motives to join, scholars mention: adventure, vindication of a group, lack of better alternative opportunities,¹¹ an environment free from stigma and discrimination regarding their faith, a wish to further the Islamic caliphate (ISIS),¹² outrage at what is allegedly happening in the country they wish to travel to, devotion to the ideology of the group they aspire to join, and searching for their identity and meaning.¹³ Although motivations regarding practicing their faith and furthering the Islamic caliphate¹⁴ seem more context specific, Frenett and Silverman have demonstrated that motives are rather universal across conflicts and time.¹⁵

What is different about the foreign fighting phenomenon related to the most recent Syrian war is the attraction it had on a diverse demographic group. Even before the declaration of an Islamic caliphate by ISIS in the summer of 2014, like in other militant populations, the bulk of recruits were young men in their early twenties. However, there were relatively more women, and very young and very old people joining.¹⁶ By comparing jihadist FFs from before and during the Syrian war and before and after ISIS, Tezcür and Besaw also demonstrate that the Syrian war has led to the mobilization of new and broader demographic groups lacking earlier experiences of political activism.¹⁷

Importantly, regarding the road to violent extremism and terrorism, FFs might differ to a certain degree from other terrorist categories, like recruiters, financers, or attackers. Participation in violence abroad as a FF can be viewed as one specific type of political violence, which nature can vary depending on context and group; violence could take the form of insurgency or joining a traditional civil war, or it could lead to participation in acts of ethnic cleansing and terrorism.¹⁸ Additionally, the question remains whether those travelling abroad actually engage in violence, and if so, to what extent. Therefore, FFs could be considered a specific group who may differ from other violent (e.g. attackers, hostage takers) and nonviolent (e.g. recruiters, financers) terrorists who remain in their home country.

Like in other countries, FFs—those apprehended prior to travel, those still abroad, and returnees—are prosecuted in the Netherlands. As Paulussen and Pitcher describe about the Dutch context, a wide range of crimes (e.g. training for terrorism, participation in a terrorist organization) is used to prosecute FFs.¹⁹ As gathering evidence in war-torn countries like Syria is difficult, prosecutors have mainly focused on charges such as membership of a terrorist organization—since the evidentiary threshold is considerably lower here. As such, prosecutors do not need to prove the other crimes, like war crimes, but “solely” that the suspect joined a terrorist organization.²⁰

Theoretical Focus

In order to examine the pathways of FFs, we use insights from social control theorists who claim that the inclination to be deviant is all-present but people can be “protected” from deviance when bonds or ties connect them to conventional society.²¹ Deviating from conforming behavior is more probable when ties to mainstream society—in the form of bonds to prosocial values (e.g. in favor of society’s values and norms), prosocial people (e.g. family members), and prosocial institutions (e.g. school)—are absent or weakened.²² In their age-graded theory of informal social control, Sampson and Laub emphasize the importance of life-course “turning points” or milestones that could alter or redirect behavioral trajectories in life.²³ They state that adult social bonds to family and employment are significantly related to changes in adult crime, regardless of early differences in childhood experiences and delinquency.²⁴

Relatedly, Feddes and colleagues have identified several potential trigger events on the personal, group, and societal level with regard to radicalization processes.²⁵ Although not all FFs may have undergone such a process, we believe similar triggers might have existed prior to their (attempted) travel. On the personal level, Feddes and colleagues describe the following triggers: problems at home, confrontation with death, loss of (perspective on) employment or problems in school, confrontations with authorities, direct experiences of discrimination, racism or marginalization, and detention. On the group level, they distinguish: meeting a radical person, joining a radical group, joining a training camp, breaking of social ties, getting married, and confrontation with propaganda. Lastly, on the societal level, they mention: perceived lack of influence and calls to violence, perceived attacks on “own” group, and government policy. These factors are regularly also referred to as personal, pull, and push factors.²⁶ Feddes and colleagues stress that the manifestation of trigger factors does not necessarily imply that an individual radicalizes; that depends on (I) the property of the triggers—the more enduring and powerful the triggers are, the higher the chance of radicalization—and (II) individual characteristics.²⁷

Another related concept that might apply to the FF group is the idea of “redeeming” oneself.²⁸ Experiencing some form of collective injustice (e.g. injustices in Syria due to Western interference) and/or moral or religious guilt (i.e. struggling with lacking a pure life) could lead to a need for leaving a problematic past (e.g. criminal involvement) and their sins (e.g. using substances) behind, and doing something “good” by joining an armed struggle of their people.²⁹ This idea of redemption is somewhat similar to Maruna’s description of “making good,” where, different from persisting offenders, desisting ex-offenders not only “changed,” but they also “reconstituted.”³⁰ Similarly, FFs can experience such an idea of reconstitution through committing themselves to an armed struggle and longing for radical redemption, in the form of gratification or rewards.³¹

Exploring Categories

Besides examining the FF group as a whole, terrorism studies have shown the benefits of analyzing subgroups (e.g. based on roles or Internet usage) within terrorist and extremist samples to distinguish pathways and background characteristics among these populations.³² It allows for a more detailed understanding of factors and processes contributing to different forms of deviant radical behavior. Furthermore, it does justice to the heterogeneity of individuals within terrorist and extremist populations.³³ Therefore, we discuss three comparisons below that we aim to explore in more depth.

Age Differences

One of the most consistent findings within developmental criminology is that there is a relationship between crime and age.³⁴ The age-crime curve has been regularly observed, where criminal engagement increases in adolescence, peaks in late adolescence, and decreases rapidly thereafter.³⁵ Farrington argues that the age crime curve “reflects decreasing parental controls, a peaking of peer influence in the teenage years, and then increasing family and community controls with age.”³⁶ In a systematic review, he and his colleagues give an overview of explanatory risk factors of crime, violence, and delinquency, such as adverse family environment (e.g. parental incarceration, child maltreatment) and socioeconomic status.³⁷ Hence, we are interested in whether there are differences between those who (attempted) to become FFs during their youth (<25 years) and those who did during adulthood (25 years or older).

Sex Differences

Although males dominate the FF group, a significant part (of European FFs) pertains women.³⁸ Albeit the outcome (i.e. (attempted) travel to Syria and Iraq) is the same, developments prior to that outcome may differ between men and women, which could be crucial knowledge in terms of prevention. From studies into delinquency we know that there can be different factors and mechanisms at play for women and men regarding their involvement in crime.³⁹ Thus, we are interested in comparing male and female FFs, to see in what ways they are similar and different from each other.

Prior and After the Declaration of a Caliphate by ISIS

From the start of the Syrian war in 2011, the conflict attracted FFs from across the world.⁴⁰ Among the opposition against the Assad-regime were also jihadist inspired groups, like Jabhat al-Nusra, al-Qaeda, and ISIS.⁴¹ The latter became more prominent over the course of the war and sieged territory in both Iraq and Syria. During the summer of 2014, the then leader of ISIS declared his territory a “caliphate,” claiming ISIS was now the only Islamically legitimate state in the world and all Muslims were obliged to support its fight to hold and expand its territory.⁴² Consequently, ISIS succeeded in attracting far more recruits than al-Qaeda. Although Malet nuances this statement, claiming that only half of the Sunni jihadi FFs were loyal to ISIS and that there were many other FFs fighting along Christian, Kurdish and other minority armed groups in the region,⁴³ Bergema and Van San demonstrated that there was a wave of mostly women and whole families who travelled to ISIS territory after the declaration.⁴⁴ Hence, it may be worthwhile to see whether the establishment of a caliphate led to a more diverse influx of FFs.

Method

Sample

After approval from the Ethics Committee for Legal and Criminological Research of the Vrije Universiteit Amsterdam,⁴⁵ we gathered data at the national office of the Dutch Probation Service (DPS). The DPS writes recommendations for judges, prison authorities, and public prosecutors and gives advice. Additionally, probation officers supervise and monitor clients (i.e. suspects and offenders) during their probation time. Since 2012, a special team, referred to as TER-team,⁴⁶ deals with probation clients who are convicted or suspected of preparing or committing terrorist crimes.⁴⁷ Our dataset contains all individuals who were supervised by the TER-team between 2012 and 2019 after being charged with a crime with terrorist intent (n = 184).⁴⁸ For the current study, we focused on individuals categorized as (attempted) FFs.⁴⁹ Additionally, we only included cases in which there was one or more advisory report(s) to collect data from (excluded n = 3). The total sample is n = 76. Data were collected in February, March, September 2020 and from July until the beginning of September 2021.⁵⁰

Notably, our sample comprises both suspects and convicts. We included both for four reasons: (I) the indictment states concrete suspicions about actions with terrorist intent, compared to more broad suspicions for other types of suspects, (II) building a court case (i.e. collecting evidence) is often more challenging in terrorist suspect cases (e.g. establishing evidence committed in foreign countries rather than The Netherlands), (III) the regularly long duration of the court cases connected to terrorist suspects, leads to cases still being in progress during our research, (IV) we focused on advisory reports (see also below) which were produced prior to the court case and court decision, giving us no information on whether or not someone was convicted. Hence, our sample includes both suspects and convicts.

Files

DPS delivered an anonymized overview of all TER-clients up until the end of 2019. Based on unique numbers (instead of names) listed in this overview, we could identify the necessary files in their secured systems. During data collection, we focused on advisory reports produced by probation officers. An advisory report includes information on the legal charges, a description of the criminal events that (allegedly) transpired, a brief outline of a client’s criminal background, and information on various life domains (e.g. social network, employment situation). Regularly, an advisory report is based on information that a client provides during conversations with TER-team probation officers. When available, information from informants (e.g. family members, health care professionals), police files, and (mental) health reports are used to complement the information they receive from the client.

Codebook and Coding

Based on theoretical insights and earlier studies, the development of our codebook underwent various stages. First, we created a codebook that: (I) incorporated all different life course domains, including risk and protective factors and possible triggers, (II) gave insight into ideological beliefs, and (III) addressed the type of terrorist crime and enabled us to differentiate between terrorist categories (e.g. recruiters, financers, FFs). The second step was to familiarize with the files, after which all coders separately coded similar cases and immediately discussed the codes, for both the open and closed questions. By discussing our interpretation of the data and our codes, we were able to improve the codebook.⁵¹ Once the codebook was finalized, 16 files were randomly selected for five coders to independently code. Subsequently, we tested all quantitative variables in our dataset that lacked complete agreement by calculating the Krippendorff Alpha. Variables with an alpha of lower than 0.67 were either modified or deleted from the codebook. Finally, after distributing all cases among the coding team and coding separately, we collectively assessed and discussed all our quantitative codings per theme for each case and made final modifications when necessary. During coding, we solely used data pertaining to an individual’s situation before the first (alleged) terrorist crime handled by the probation’s office.

Importantly, the files were produced for different reasons than research, and some files contained more data than others. When coding the files, we frequently ran into missing information. Nonetheless, we retrieved extensive data on a variety of variables, allowing us to provide a general description of the FF group and shed light on potential differences between the researched subgroups.

Variables

Dependent Variables

Our dependent variables differ based on the exploration of the comparisons we included in our study. For the first comparison, our dependent variable is age. We divided the group into youths (n = 41), who were 24 years or younger at the time of the (alleged) terrorist crime, and adults (n = 35), who were 25 years or older. The second comparison was based on the biological sex of the suspects, with females (n = 26) and males (n = 50). For our last comparison, we divided the group into (attempted) travelling prior (n = 25) and after (n = 50) the declaration of a caliphate by ISIS.

Independent Variables

Independent variables were grouped per subtheme and the variables measured the situation of an individual prior to committing their (alleged) terrorist crime. Primarily dummy (0 = not in file or no, 1 = yes⁵²) and categorical variables were included. Variables covered the following themes: indications of extreme ideological beliefs (e.g. changes in appearance), feelings of injustice (e.g. direct experience with discrimination), profound events/triggers (e.g. problems at the parental home), childhood experiences, housing situation, religious affiliation, educational background, employment, financial situation, social network (family, friends, partner/children, other social contacts), (mental) health, personality characteristics, IQ, substance use, support (i.e. social care), criminal history, interventions, probation contact, and online activities related to the terrorist crime. More information about the included variables can be found in the tables. Besides the dummy and categorical variables, we also included open variables. These variables provide more context to each of the themes and the related (dummy) variables, and allowed us to give more qualitative meaning to our findings.

Analysis

Our analysis is divided into two parts. First, based on descriptive data, we outlined the entire group on all included themes to get a general view of their backgrounds and personal characteristics. Second, we assessed through the use of Fisher’s exact tests (two tailed) whether there are significant differences between comparison groups based on age, sex, and timing of travelling. Additionally, where possible, we qualitatively examined the differences more closely.

For our analysis, we used SPSS, version 8, and Atlas.ti, version 16.1. Due to the sensitivity of the data and to prevent privacy issues, we report our findings on a group level. Consequently, we decided to refrain from reporting exact numbers below five and percentages lower than 10% or 15%, depending on the sizes of the comparison groups. In those cases, we use a “<” or “>” to indicate that we cannot report the exact percentages and where we found significant differences, we made the numbers bold in the tables to indicate which group is showing a higher percentage.

Findings

General Findings

Demographics

The average age of FFs was 24.1 years (SD = 5.9), with a range from 15 to 46 years. Regarding ideological background, 73.7% was suspected of a jihadi-inspired ideological offense. For 15.7%, it was difficult to categorize their ideological beliefs due to lacking a clear ideology, inconclusive information, or a different type of ideological beliefs (e.g. separatist) than extreme right, extreme left, or jihadi-inspired. Data on ideological beliefs was missing for 10.5% of the sample.

Signs and Triggers

As Table 1 demonstrates, over half of the examined individuals (57.9%) showed one or more indications of extreme ideological beliefs prior to their terrorist crime. Indications present were change(s) in appearance and/or behavior, expressed concerns from family members, and possessing propaganda. The latter was most prominent; almost a third possessed propaganda material (e.g. jihadist magazines).

Table 1. Percentage signs and triggers leading up to terrorist crime (n = 76).

Subthemes and related variables	Total (%)	Age (%)		Sig	Sex (%)		Sig	ISIS (%)		Sig
		Youth (n = 41)	Adult (n = 35)		Male (n = 50)	Female (n = 26)		Prior (n = 25)	After (n = 51)
Indications extreme ideological beliefs^1
General (one or more indications)	57.9	65.9	48.6		50.0	73.1	#	48.0	62.7
Change(s) in appearance	18.4	24.4	11.4		<15.0	>15.0	***	16.0	19.6
Change(s) in behavior	22.4	24.4	20.0		16.0	34.6	#	24.0	21.6
Family members express concern	14.5	14.6	14.3		>15.0	<15.0	#	<15.0	>15.0
Possessing propaganda	31.6	36.6	25.7		30.0	34.6		24.0	35.3
Indications of feelings of injustice
General feelings of injustice	43.4	46.3	40.0		40.0	50.0		56.0	37.3
Direct experience with exclusion (e.g. discrimination)	11.8	>10.0	<10.0		<15.0	>15.0		>15.0	<15.0
Worries about “own” group (us vs. them)	19.7	22.0	17.1		<15.0	>15.0		20.0	19.6
Criticizing government policies	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Distrust in the “system” (e.g. government)	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Responding to call for action	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Profound events and/or triggers
General (sum of events, one or more)	77.6	73.2	82.9		76.0	80.8		76.0	78.4
Divorce parents seen as problematic	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Problems at parental/foster home	21.1	22.0	20.0		20.0	23.1		20.0	21.6
Confrontation with own mortality	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Confrontation with death of loved one	13.2	12.2	14.3		>15.0	<15.0		>15.0	<15.0
Negative experience with law enforcers	<10.0	<10.0	>10.0		<15.0	<15.0		<15.0	<15.0
Negative experience with other governmental/social institutions	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Victim of crime	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Health problems loved one	<10.0	>10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Experiences with violence	17.1	17.1	17.1		<15.0	>15.0	#	<15.0	>15.0
Sexual violence	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Domestic violence	11.8	12.2	11.4		<15.0	>15.0	#	<15.0	<15.0
Other profound event(s)	64.5	65.9	62.9		66.0	61.5		64.0	64.7

^# p < .1,

*** p<.001

¹We also checked whether suspects ever visited a training camp, but this was not the case.

Indications of feelings of injustice were present for less than half of the FF group (43.4%). We looked into direct experiences with exclusion (e.g. discrimination), worries about their “own” group (e.g. us versus them thinking), criticizing government policies, distrust in the “system,” and responding to a call for action. Our analysis shows that worries about their “own” group (19.7%) and direct experiences with exclusion (11.8%) were most common in the reports. Frequently, FFs mentioned examples of experiences with discrimination (e.g. being bullied for being Muslim, negative experiences with police), proclaimed feelings of injustice regarding the most recent Syrian war (e.g. towards the victims, due to Western involvement), and described the injustice towards Muslims in the Netherlands compared to other (religious) groups.

Also, we examined events considered to be profound events that potentially functioned as triggers leading up to the terrorist crime: family youth experiences (like problems at home or divorce), confrontation with death or illness, negative encounters with governmental institutions (e.g. law enforcement, care facilities), being a victim of crime, and experiencing (domestic and/or sexual) violence. A significant part, about three quarters of the group, experienced one or more profound events. In the cases where we identified profound events, these events happened during the life course prior to the terrorist crime. Hence, they could have played a vital role in becoming a FF. Importantly, however, only in less than five cases, a profound event was mentioned as being directly related to (the attempt of) becoming a FF. Consequently, we cannot draw firm conclusions about the effects of profound events in their pathways to becoming a FF.

As we zoom in on the specific profound events, none of the events seems particularly common. Although modest in number, problems at the parental or foster home were most prominent (e.g. confrontations with parent(s)) (21.1%), followed by experiences with violence (17.1%), and confrontation with the death of a loved one (13.2%). Next to the events we included in our study based on theory, we also added a variable to identify other potential trigger events (i.e. where the event was experienced as profound), which we found in 64.5% of the cases. We identified, for example, the following events: escaping a war zone and becoming a refugee, radical family members (who became FFs), unstable living situation after leaving the parental home (mostly before the age of 18 years), problems with (ex-)partner and/or family in law, being unable to see own child(ren), and/or (unplanned) teen pregnancies.

Life Domains

Regarding life domains, as shown in Table 2, we first examined childhood experiences (which partly overlaps with profound events examined in the previous section). Although childhood information was regularly lacking, we notice that 22.4% experienced some form of trauma during childhood (age 0-11) and 21.1% during puberty (age 12-17). Regularly, it was reported that parents were divorced and the father left the family, which was experienced as difficult and led to problematic home situations, for instance, in cases where mothers were unable to educate their child or children on their own.

Table 2. Percentage life domains before terrorist crime (n = 76).

Subthemes and related variables	Total(%)	Age(%)		sig	Sex(%)		sig	ISIS(%)		sig
		Youth (n = 41)	Adult (n = 35)		Male (n = 50)	Female (n = 26)		Prior (n = 25)	After (n = 51)
Youth^1
Trauma during childhood (incl. bullying)	22.4	22.0	22.9		18.0	30.8		24.0	21.6
Trauma during adolescence (incl. bullying)	21.1	29.3	11.4	#	<15.0	>15.0	*	24.0	19.6
Living situation^2
Household
Single	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
With a partner	>10.0	<10.0	>10.0		<15.0	>15.0		<15.0	>15.0
At parental house	28.9	43.9	11.4	**	20.0	46.2	*	28.0	29.4
Other	30.3	24.4	37.1		34.0	23.1		36.0	27.5
Unknown	28.9	24.4	34.3		36.0	15.4	#	32.0	27.5
Ever moved	64.5	56.1	74.3		66.0	61.5		68.0	62.7
Emigrated	31.6	29.3	34.3		34.0	26.9		28.0	33.3
Ever had a refugee status	14.5	14.6	14.3		>15.0	<15.0		<15.0	>15.0
Ever moved out of the parental home	53.9	41.5	68.6	*	58.0	46.2		60.0	51.0
Religious affiliation
Identifies with a religion (in general)	72.4	63.4	82.9	#	76.0	65.4		72.0	72.5
Convert	14.5	<10.0	>10.0		<15.0	>15.0	*	<15.0	>15.0	#
Identified with a religion before terrorist crime	77.6	75.6	80.0		74.0	84.6		76.0	78.4
Knowledge of religion (according to TER-team)	11.8	<10.0	>10.0		<15.0	>15.0		<15.0	<15.0
Religious upbringing	31.6	36.6	25.7		34.0	26.9		28.0	33.3
Education
Finished…
Lower	13.2	12.2	14.3		>15.0	<15.0		<15.0	>15.0
Middle	38.2	43.9	31.4		36.0	42.3		52.0	31.4
Higher	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Other	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Unknown	39.5	36.6	42.9		40.0	38.5		32.0	43.1
School problems	32.9	43.9	20.0	*	40.0	19.2	#	48.0	25.5	#
Dropout of school	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Employment
Employed in year before terrorist crime	39.5	31.7	48.6		48.0	23.1	*	36.0	41.2
Ever problems at work	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Ever became unemployed	26.3	17.1	37.1	#	32.4	15.4		<15.0	>15.0	#
Professional military training	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Financial situation
Main type of income in year before terrorist crime
Employment	27.6	14.6	42.9	**	34.0	15.4		24.0	29.4
Social benefits	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Other	>10.0	29.3	11.4	#	<15.0	>15.0	#	>15.0	>15.0
Unknown	42.1	46.3	37.1		40.0	46.2		48.0	39.2
Ever received benefits	15.8	17.1	14.3		>15.0	<15.0		<15.0	>15.0
In debt before terrorist crime	22.4	17.1	28.6		22.0	23.1		<15.0	>15.0
Received debt assistance	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0

^# p<.1,

* p<.05,

** p<.01.

¹Information on childhood and adolescence was limited. We also checked for other variables (instability, friendship), but these variables were poorly filled.

²Initially, we also included variables on sheltered housing, but these variables were poorly filled.

Living situations were diverse in the year prior to their terrorist crime. FFs lived at the parental home or in what we categorized as “other” living situations, like living together with roommates or with other family members than parents. However, in some cases it was unclear whether someone lived alone or with others. Furthermore, some moved around in a short time span, e.g. living with a partner, then moving back to the parental home, and subsequently staying with a friend. In terms of moving around: 31.6% ever emigrated, 14.5% ever had a refugee status, 53.9% ever left their parental home, and 64.5% has ever moved.⁵³ For 28.9% of the sample, we had no information about their living situation.

Over three quarters of the group identified with a religion prior to their terrorist crime. For the rest of the cases, information on this theme was missing or inconclusive. The majority held beliefs derived from Islamic religious movements. Almost one third mentioned a religious upbringing, while a smaller part of the group pertained converts (14.5%). Converts were mostly women. The TER-team occasionally also reported about whether a suspect had, in their opinion, sufficient or any knowledge of their religion, and in 11.8% of the cases there was an indication of having (more than superficial) knowledge about their religion.

In terms of socioeconomic status, reports showed that 38.2% had finished middle level education. One third of the group experienced school problems, of which motivational problems (e.g. skipping class) were most dominant. Importantly, however, for 39.5% we lacked information on their educational background. In the year prior to the terrorist crime, 39.5% was employed. We did not find any striking commonalities regarding stability or type of employment. For slightly more than a quarter of the group, it was reported that they ever experienced becoming unemployed. Reasons differed. In seven cases, unemployment was due to the economic climate (e.g. reorganizations) or their personal performance (e.g. conflict about work performance). In fewer cases, it had to do with physical or psychological problems. Lastly, in five cases, they decided to quit their jobs themselves. For a relatively large part of the group, 42.1%, it was unknown what their main source of income was in the year prior to the terrorist crime. When reported, we notice that income through employment was most common (27.6%). Regarding financial dependency, 15.8% ever received social benefits and 22.4% ever had debts.

Social Network

During childhood, most FFs grew up in relatively large families, with multiple siblings (i.e. two or more). As shown in Table 3, in a third of the cases, it was reported that the parents got divorced or separated at some point. For about two thirds of the group, both parents were still alive and 59.2% of the group had contact with one or both (foster) parents—of which 17 cases described the relationship with their parents as difficult, 13 cases as predominantly good, and for 23 cases the nature of the relationship was missing or only described for one parent as good and the other as difficult. In 38.2%, FFs had experienced a sudden negative event (i.e. both physical and mental change) with their family; in 16 cases, the divorce of parents was mentioned, followed by 11 cases in which someone had a difficult relationship with (a) parent(s). In fewer cases, we found it had to do with: the death of a family member, a family member being part of a terrorist organization, having no permanent residence, and descriptions of violence. For 11.8%, there was a death of a family member; in 5 cases it pertained a sister or brother, and in less than 5 cases a father died. Interestingly, a fifth (21.1%), had a radical family member. In those cases, it mostly pertained a male sibling.

Table 3. Percentage social network before terrorist crime (n = 76).

Subthemes and related variables	Total(%)	Age(%)		sig	Sex(%)		sig	ISIS(%)		sig
		Youth (n = 41)	Adult (n = 35)		Male (n = 50)	Female (n = 26)		Prior (n = 25)	After (n = 51)
Family^1
Parents divorced	32.9	41.5	22.9	#	26.0	46.2		48.0	25.5	#
Parents alive?
Both alive	65.8	78.0	51.4	*	66.0	65.4		60.0	68.6
One or both died	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Unknown	>10.0	>10.0	>10.0	#	>15.0	>15.0		>15.0	>15.0
Contact with (foster) parent(s)	59.2	73.2	42.9	*	54.0	69.2		60.0	58.8
Death of a family member	11.8	<10.0	>10.0		<15.0	>15.0		>15.0	<15.0
Sudden negative event with family (e.g. father leaves the house)	38.2	43.9	31.4		34.0	46.2		44.0	35.3
Domestic violence in parental home	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Radical family member	21.1	19.5	22.9		24.0	15.4		24.0	19.6
Friends^2
Radical friend(s)	32.9	39.0	25.7		34.0	30.8		44.0	27.5
Criminal friend(s)	11.8	<10.0	>10.0		>15.0	<15.0	*	>15.0	<15.0	#
Loss of a friend(s)	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Part of group	10.5	<10.0	>10.0		<15.0	<15.0		<15.0	<15.0
Partner and children^3
Ever a partner	55.3	41.5	71.4	*	48.0	69.2	#	52.0	56.9
Ever had (presumably) a radical partner?
Yes	21.1	24.4	17.1		<15.0	>15.0	**	28.0	17.6
Unclear	38.2	22.0	57.1	**	42.0	30.8		28.0	43.1
Ever married (incl. Islamic)	34.2	22.0	48.6	*	30.0	42.3		40.0	31.4
Ever divorced	<10.0	<10.0	>10.0		<15.0	<15.0		<15.0	<15.0
Child(ren)	28.9	<10.0	>10.0	***	26.0	34.6		16.0	35.3
Custody of child(ren)	17.1	<10.0	>10.0	**	<15.0	>15.0		<15.0	>15.0
Other type of contact related to terrorist crime	35.5	61.0	68.6		30.0	46.2		76.0	58.8
Total radical/terrorist crime related social contacts	72.4	78.0	65.7		70.0	76.9		76.0	70.6
One contact	43.4	43.9	42.9		48.0	34.6		40.0	45.1
Two or more contacts	28.9	34.1	22.9		22.0	42.3		36.0	25.5

^# p<.1,

* p<.05,

** p<.01,

*** p<.001

¹We checked if someone was an only child, had siblings, parents or other family members with had problems (e.g. mental, criminal background), but due to small numbers we excluded them from the table.

²The numbers were too small regarding variables on the quality of friendships (e.g. leader/follower). Furthermore, variables related to group activity were poorly filled.

³We checked for partner characteristics (criminal background, substance abuse, mental problems, partner violence) and whether they suffered the loss of a child or experienced domestic violence, but the numbers were small.

In terms of friends, for 32.9%, it was reported that he or she had (a) radical friend(s) and 11.8% had criminal friend(s). Being part of a group (10.5%) and the loss of a friend (<10%) were relatively uncommon.

In 55.3% of the cases, a suspect had or has had a partner. In 21.1% of the cases, they had been involved with a radical partner and for 38.2% it was unclear, based on the information in the reports, whether they had a radical partner. Little over a third got married at some point and less than 10% got divorced. In 28.9% of the cases, a suspect had (a) child(ren) prior to their suspicion and in 17.1% the suspect had custody over the child(ren). In the cases with children, we found a few descriptions where it was stated that the partner was radical, also wanted to become a FF or was already staying in Syria/Iraq.

Little over a third of the group, 35.5%, had other types of social contacts that were specifically related to the terrorist crime. Generally, these contacts were (co-)FFs and (a network of) so-called “sisters.” In fewer cases, mention was made of: ISIS-fighters, potential partners in the Caliphate, “brothers,” radicalized contacts and sympathizers. Essentially, most terms overlap and describe contacts that (almost) became FFs or who were pushing the terrorist suspect to become one (i.e. recruiters).

When taking the variables of radical family member(s), radical friend(s), radical partner, and social contacts related to the terrorist crime together, we find that 72.4% of the FF group had one (43.4%) or more (28.9%) radical social contacts. Hence, a significant part was involved with individuals who likely influenced their path towards becoming a FF.

(Mental) Health and Personality Characteristics

Healthwise, less than 10% had indications of problems with their physical health, as shown in table 4. For 38.2%, the reports showed indications of mental health problems, and 17.1% of the group was officially diagnosed with a mental health disorder prior to becoming a FF. None of the different mental health disorders, like ADHD, depression or PTSS, were strikingly common among the FFs. Most dominant were indications of narcissism and depression, followed by psychoses, anti-social personality disorder/traits, and PTSS. For less than 10% of the group, the mental health problems could be linked directly to the terrorist crime, as forensic research assessments showed. In those cases, a suspect was reduced or not accountable for his or her actions at the time of the terrorist crime.

Table 4. Percentage (mental) health and personality characteristics before terrorist crime (n = 76).

Subthemes and related variables	Total (%)	Age (%)		Sig	Sex (%)		Sig	ISIS (%)		Sig
		Youth (n = 41)	Adult (n = 35)		Male (n = 50)	Female (n = 26)		Prior (n = 25)	After (n = 51)
Health
Problems with physical health	<10.0	<10.0	<10.0		<15.0	>15.0		<15.0	<15.0
Problems with mental health	38.2	39.0	37.1		40.0	34.6		36.0	39.2
Official diagnose of mental health problems (besides forensic report)	17.1	17.1	17.1		16.0	19.2		<15.0	>15.0
Link mental health problems and terrorist crime	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Forensic report on terrorist suspect	34.2	41.5	25.7		34.0	34.6		40.0	31.4
Sort of mental health problems…
ADHD	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Attachment disorder	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Autism	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Borderline	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Depression	<10.0	>10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Narcissism	10.5	>10.0	<10.0		<15.0	<15.0		>15.0	<15.0
Paranoia	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Psychoses	<10.0	>10.0	<10.0		<15.0	<15.0	#	<15.0	<15.0
Psychopathy	<10.0	<10.0	>10.0		<15.0	<15.0	#	<15.0	<15.0
PTSS	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Schizophrenia	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Suicidal	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Personality characteristics
Aggressive	<10.0	<10.0	<10.0		<15.0	>15.0		<15.0	<15.0
Impulsive	15.8	19.5	11.4		<15.0	>15.0		>15.0	<15.0
Loneliness/social isolation	13.2	14.6	11.4		<15.0	>15.0		<15.0	>15.0	*
Superiority	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
IQ
Indications intellectually challenged	10.5	>10.0	<10.0		<15.0	>15.0		>15.0	<15.0
Officially diagnosed	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Substances^1
Drugs
Ever used	44.7	31.7	60.0	*	58.0	19.2	**	32.0	51.0
Drug problem	15.8	12.2	20.0		>15.0	<15.0		<15.0	>15.0
Link drug use and terrorist suspicion	<10.0	<10.0	<10.0	#	<15.0	<15.0		<15.0	<15.0
Alcohol^2
Ever used	44.7	41.5	48.6		52.0	30.8	#	36.0	49.0
Alcohol problem	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Support (in the form of social work)^3
Overall	38.2	39.0	37.1		34.0	46.2		36.0	39.2
Health support (e.g. mental health care)	21.1	19.5	22.9		22.0	19.2		<15.0	>15.0
Other type of support (e.g. for work, living, education, etc.)	23.7	29.3	17.1		18.0	34.6		24.0	23.5

^# p<.1,

* p<.05,

** p<.01

¹We initially included gambling and other addictions, but these variables were poorly filled.

²We also checked whether there was a link between alcohol use and terrorist crime. This was not the case.

³We checked for out-of-home placement, but these variables were poorly filled.

Data on personality characteristics and behaviors were scarce in the reports. However, although the numbers are small, we did find that probation officers reported that for 15.8% of the group there were indications of impulsive behavior and that 13.2% had dealt with loneliness or social isolation. 10.5% of the (attempted) FFs had indications of being intellectually challenged, but for less than 10% there was an official diagnose.

Regarding substance use, data demonstrate that 44.7% ever used drugs and/or alcohol. 15.8% ever had a drug problem and less than 10% ever an alcohol problem. In relation to the terrorist crime, for less than 10% there was a link between their crime and the use of drugs, and we found no indications that alcohol problems were linked to the terrorist crimes. In some cases, suspects stopped (suddenly) with substance use due to the (increased) interest in their religion. However, some also stopped due to non-religious reasons, such as the birth of their child, the death of a loved one, or becoming more serious about school or work.

Receiving support in the form of social care or social work was the case for 38.2% of the group. 21.1% received a type of health support, which pertained mostly mental health care, and 23.7% received other types of support, for instance help with getting a house or successfully obtaining a degree.

Crime

Table 5 shows that almost half of the group (48.7%) has a criminal background and 39.5% was ever convicted of a crime prior to becoming a FF. Most dominant was property crimes (26.3%), followed by violent crimes (21.1%), openly violent crimes (14.5%), and miscellaneous crimes (10.5%). Types of interventions most common were community service and related interventions (22.4%), followed by being detained (13.2%) and/or getting fined (11.8%). In 18.4% of the cases, suspects were ever in contact with the probation services prior to their terrorist crime, and 13.2% with the youth probation services.

Table 5. Percentage criminal background before terrorist crime (n = 76).

Subthemes and related variables	Total (%)	Age (%)		Sig	Sex (%)		Sig	ISIS (%)		Sig
		Youth (n = 41)	Adult (n = 35)		Male (n = 50)	Female (n = 26)		Prior (n = 25)	After (n = 51)
Criminal history
Criminal past	48.7	39.0	60.0		64.0	19.2	***	48.0	49.0
Ever convicted	39.5	31.7	48.6		50.0	19.2	*	36.0	41.2
Another terrorist crime	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Crime categories suspected of…
Violent crimes	21.1	12.2	31.4	#	>15.0	<15.0	**	24.0	19.6
Property crimes	26.3	17.1	37.1	#	>15.0	<15.0	*	36.0	21.6
Against life crimes (e.g. murder)	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Threatening crimes	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Public disorder crimes	<10.0	<10.0	<10.0		<15.0	<15.0	#	<15.0	<15.0
Openly violent crimes	14.5	12.2	17.1		>15.0	<15.0	*	>15.0	<15.0
Weapon crimes	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Drug crimes	<10.0	<10.0	>10.0	*	<15.0	<15.0		<15.0	<15.0
Sex crimes	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Traffic crimes	<10.0	<10.0	>10.0		<15.0	<15.0	#	<15.0	<15.0
Other crimes	10.5	>10.0	<10.0		>15.0	<15.0	*	<15.0	<15.0
Interventions
Fine	11.8	<10.0	>10.0	#	<15.0	<15.0		<15.0	<15.0
Prison	13.2	<10.0	>10.0	*	>15.0	<15.0	*	>15.0	<15.0
Mental institution (TBS)	<10.0	<10.0	<10.0		<15.0	<15.0		<15.0	<15.0
Community service, etc.	22.4	22.0	22.9		>15.0	<15.0	**	32.0	17.6
Other	10.5	<10.0	>10.0		<15.0	<15.0		<15.0	<15.0
Probation contact
Ever had contact with probation	18.4	12.2	25.7		>15.0	<15.0		20.0	17.6
Youth probation	13.2	14.6	11.4		>15.0	<15.0		>15.0	<15.0

^# p<.1,

* p<.05,

** p<.01,

*** p<.001

Online Activities

Table 6 demonstrates that online activities related to the terrorist crime were identified for 44.7% of the group. In 27.6% of the cases, the activities could be categorized as passive (e.g. having or watching jihadi-inspired content) and for 21.1% it was typified as active (e.g. distributing content themselves or posting ideologically motivated content). During our analysis, we identified five categories of online activities. First, searching for information was common. Mostly, their search was focused on being a Muslim and about the Syrian war, which in some cases led to feelings of wanting to help and support their “brothers and sisters.” In their search, they then ended up on websites and Facebook pages where extremist messages were being shared.

Table 6. Percentage online activities related to the terrorist crime (n = 76).

Subthemes and related variables	Total (%)	Age (%)		Sig	Sex (%)		Sig	ISIS (%)		Sig
		Youth (n = 41)	Adult (n = 35)		Male (n = 50)	Female (n = 26)		Prior (n = 25)	After (n = 51)
Online activity	44.7	70.7	42.9	*	52.0	69.2		56.0	58.8
Passively (e.g. having or watching content)	27.6	31.7	22.9		33.0	23.1		20.0	31.4
Actively (e.g. distributing or posting/messaging)	21.1	>10.0	<10.0	*	<15.0	>15.0	#	<15.0	>15.0

^# p<.1,

* p<.05

A second category we found is having interactions with social contacts. They were connecting with like-minded people, such as “brothers”/“sisters”/strangers (mostly those who already travelled to Syria/Iraq or at least claimed to be there) and in some cases discussed with them their wishes to become a FF and practical matters related to this. Also, they were members of (secret) Facebook groups where propaganda material was distributed (sometimes having an active role, such as being co-administrator of a Facebook page); they helped others with their travels; had contact with a family member who was already a FF; made plans with others to travel abroad or actually travelled together. The third activity identified was being in contact with a (future) partner. Some went online to search for a suitable partner, concluded an Islamic marriage through Skype, and travelled to Syria/Iraq after an online request from their (potential) partner.

A fourth category we observed is having and/or distributing propaganda material. Mostly, it pertained (jihadi inspired) images, magazines of ISIS or Dabiq, videos, or music. Some FFs were themselves quite active, for instance by posting messages or giving lectures or prayers. There were also videos in possession that would show how good life was in the caliphate. Additionally, although less prominent than the other activities, a final activity was searching for practical information about, for instance, routes to travel or getting a visa.

In sum, a reoccurring pathway was identified among the files we examined. First, the potential FF will search the internet, for instance on how to become a good Muslim, to read about the Syrian war or to find a partner. Due to this internet search, he or she gets into contact with like-minded individuals, which often leads to being invited to join (secret) chat groups or pages on Facebook for instance. Through conversations with one or more radicals, other FFs, and/or potential partners, they subsequently become more and more convinced they should travel abroad and become FFs. Several times it was mentioned that the reason for finally making the step of becoming a FF was through requests or calls of FFs or potential partners to join them.

Comparisons

The next part of our analysis focuses on comparisons between earlier described categorizations. We also examined whether these categories might be correlated between each other. Only the cross tabulation of youths and adults versus prior and after declaration of a caliphate shows a significant correlation of p<.1, demonstrating that adults were more likely to join the ISIS caliphate after its declaration rather than prior to its declaration, compared to youths.

Exploring Life Phases: Youths versus Adults

Based on a division between youths and adults, we notice that the first subgroup, compared to the adult group, had significantly more often: experienced a difficult time or trauma during puberty, had school problems, had divorced parents, and were more engaged in online activities (and were significantly more active online) related to becoming a FF. The adult group, on the other hand, had comparatively more often: moved outside of the parental home, identified with a religion, became ever unemployed, and ever had a partner, got married, and had children. Furthermore, although it pertained small numbers, we also found that for the adult group there was more often a link between drug use and their terrorist crime. Finally, in terms of criminal history, they were significantly more often suspected of a violent-, property-, and drugs crime. Consequently, they received more interventions in the form of fines and detention.

Based on the above, it appears that the youth group experienced more problems (e.g. difficult time or trauma during puberty, school problems, divorced parents) that more generally explain deviant behavior, whereas the adult group seemed to experience life events (e.g. losing employment, drug use, criminal history) that put them in a negative spiral and led them to becoming a FF.

Some of the differences we find between both groups can, to a large extent, be attributed to these different age categories. For instance, since the adult group is older, they will more likely have a longer employment history and therefore a higher chance of experiencing a period of unemployment at some point. The same holds for having a criminal history, moving away from the parental home, and being married. However, we also find differences that are not related to age (e.g. having divorced parents, identifying with a religion), showing the importance of such a comparison.

Exploring Differences between Women and Men

Compared to male FFs, female FFs significantly more often: showed changes in appearance and behavior related to their terrorist crime, ever experienced (domestic) violence, had experienced a difficult time or trauma during puberty, still lived at the parental house, were converts, had a different type of income than through employment or social benefits, ever had a (radical) partner, and were active online (e.g. distributing content or posting messages) related to their terrorist crime.

Male FFs differed significantly from the female FFs in the following way: more often family members expressed their concern as a sign leading up to their terrorist crime, their living situation was more regularly unknown, they had school problems, were employed in the year prior to their suspicion, had criminal friends, had more often indications of psychoses and psychopathy (although the numbers are small), ever used drugs and/or alcohol, had a criminal past, were ever convicted, and were more often suspected of particular crimes (e.g. violent, property, et cetera), and, consequently, received more interventions in the form of detention and community service.

In sum, females seem to have experienced adversity during puberty and/or experienced violence, were potentially more extreme in their ideological conviction (e.g. by showing changes in appearance and behavior, being more often converts), and were more immersed in a potentially problematic social environment, through having a (radical) partner and being active online related to their terrorist crime (e.g. spreading extremist content). This might have increased their chances to travel abroad. Males, on the other hand, seemed to have had more disadvantaged positions in society, by having more regularly school problems, an unknown housing situation, by being slightly more often mentally unstable, and by having a criminal past (including more often having criminal friends and experiencing interventions).

Exploring Periods before and after ISIS’ Caliphate Disclosure statement

In this comparison, before and after the declaration of the ISIS caliphate, we find the least amount of significant differences. Nonetheless, we notice that, compared to the after group, the before group had significantly more often school problems, divorced parents, and friends with a criminal background. The after group, compared to the before group, showed significantly more often indications of loneliness/social isolation, were more regularly converts, and became more often ever unemployed. As with the youths versus adults comparison (and the earlier mentioned correlation between age and timing), it seems that the before group can be more categorized as “risk takers”—with factors that explain more generally deviant behavior—whereas the after group seemed to be more exposed to negative experiences over the life course.

Discussion

In this article, we examined suspected and convicted (attempted) FFs from the Netherlands by analyzing data from probation files. We provided a detailed general overview of the group and made comparisons based on age (youths versus adults), sex (male versus female), and timing (before and after the ISIS’ caliphate declaration in 2014). The research question we aimed to answer is: What do FFs from the Netherlands look like regarding signs/triggers leading up to their terrorist crime, life domains, social network, health and personality characteristics, criminal background, and online activities, and can we detect differences comparing FFs based on age, sex, and timing of travelling?

Within the themes that we examined, we observe that a substantial part of the variables in the tables demonstrate small numbers, indicating that some of the researched factors and processes mentioned in earlier research and theoretical contributions are relatively rare among the Dutch FF group. For instance, we did not find any striking results regarding certain life domains (e.g. living situation, religion, socioeconomic situation), personality characteristics, and substance use.

Our study matches the findings of Wilner and Yar, who identified differences when comparing Canadian, European, and United States’ FFs, showing that the European FF group contained relatively more women and a substantial part had already a criminal background.⁵⁴ Moreover, our findings largely align with the overview study on Western FFs by Dawson. Our studied FF group contained a roughly similar percentage of women and converts, and showed that indeed about a third was married and had children. Hence, our findings confirm earlier (open source) research concerning these characteristics, while it also provides some new insights that we will discuss below.

One of the key findings is that indications of extreme ideological beliefs (57.9%), feelings of injustice (43.4%), and having experienced profound events (77.6%) were relatively common within the FF group. As De Graaf described, profound events could have led to feelings of collective injustice and/or moral or religious guilt triggering a need for redeeming oneself.⁵⁵ As most people will experience negative profound events at some point in their life, future research should try to examine how experiencing profound events might influence these particular feelings and, subsequently, how this leads to extreme ideological beliefs and sometimes related actions. For instance, a qualitative study comparing terrorist- and regular detainees, showed that terrorist detainees seem to perceive adverse experiences as more severe and as signs of social injustice, while the regular detainees seemed to accept their problems as part of life and something they should cope with personally.⁵⁶ Future research should examine if a similar tendency is present among FFs specifically.

A second key finding is that a substantial part of the FF group, 72.4%, had one or more radical social contacts. As both Farrington and Feddes and colleagues pinpoint, delinquent or radical peers can be key risk factors for the onset of deviant (radical) behavior. Based on our finding and in line with a study by Bergsma and Van San, it seems that the connection with other radicals could be considered one of the more important features in the pathways of (attempted) FFs.⁵⁷ It is therefore not the absence or weakening of social ties. Rather, it is the bonds with radicals that seem vital in these pathways.

Third, for 38.2% of the group, there were indications of mental health problems, of which 17.1% was officially diagnosed with a mental health disorder already prior to becoming a FF. In line with research by Weenink, mental health problems seem to be relatively common among the FF group.⁵⁸ Our research therefore contributes to the debate on the relationship between mental health problems and terrorism.⁵⁹

A fourth key finding is that a relatively large part of the group, 39.5%, was ever convicted of a crime prior to their alleged terrorist crime. According to Basra and Neumann, becoming a jihadist after criminal involvement could legitimize crime, offer redemption from past sins or, in the case of white collar or petty crime, be a way to finance extremism.⁶⁰ Once again, this could be connected to the idea of redemption: these individuals were seeking to redeem themselves through becoming FFs, leaving their “bad” pasts behind, and putting their skills to “good” use.⁶¹

Lastly, the online environment seemed to have played an important role for a substantial part of the group (44.7%): it facilitated FFs in gaining (practical and ideological) information, in interacting with like-minded people, such as other FFs and potential partners, and introduce them to propaganda material. Importantly, consciously searching for, and interacting with, like-minded people and potential partners deflects from the idea of social control theory that deviance—in this case becoming a FF—would be prevented through (strong) ties with others. Our research shows that, in some cases, online contacts with radicals seemed to have been the final push in becoming a FF.

Regarding the exploration of comparisons based on age, sex, and timing of (attempted) travelling, we notice that particularly the age and sex comparisons indicate important differences within the FF group. The youth group experienced more problems (e.g. school problems, divorced parents) that Farrington describes as “key indicators” that generally explain youths’ deviant behavior.⁶² The adult group, on the other hand, seemed to experience negative life events (e.g. losing employment, drug use, criminal history) prior to becoming a FF. Hence, the idea of redemption or reconstitution, as stated by De Graaf and Maruna⁶³, might be more applicable to the adult group. These age categories indicate that different approaches are necessary to prevent people from becoming FFs.

Similarly, we find that women differ from men. Whereas men tended to have a more disadvantaged position in society (e.g. unknown housing situation, criminal past), women seemed to have experienced more internal hardship (e.g. adversity during puberty, experiencing violence) and were immersed in a problematic social environment (e.g. network of “sisters,” radical partner). The importance of belonging and sisterhood (i.e. having peers) as a pull factor for women, as described by Saltman and Smith, and the significance of virtual communities, as Veilleux-Lepage and colleagues emphasize, seems in line with our findings regarding online participation and activities (through online networks of “sisters” for instance).⁶⁴ Prevention efforts should incorporate the identified differences, leading to more tailor-made approaches to prevent both men and women from becoming FFs.⁶⁵

Limitations

While our file analysis is based on unique empirical data about (attempted) FFs, there are limitations that should be kept in mind. First, there is a potential bias in the data due to the use of information from a governmental organization. The last decades, law enforcement has focused on specific types of ideological beliefs (i.e. primarily jihadism). Subsequently, there might be an overrepresentation of jihadi-inspired FFs in our sample. A second limitation pertains to using file data. The reports produced by the probation office are created for different purposes (e.g. advising a judge) than research, leading to certain themes being less prominent (e.g. childhood experiences) in the reports than others (e.g. criminal history). Although probation officers use a standard form and base their advice on multiple information sources, the reports reflect what they deem important to report, rather than offering full (and unbiased) life stories. Furthermore, we found reports with very little information, primarily due to suspects refraining from talking to probation officers and police, leading to missing data. Nonetheless, based on our assessment, most reports were sufficiently filled with valuable information, by making use of different sources (e.g. informants, police files) to gather information on a suspect’s situation. Finally, we only used data from the Netherlands, limiting generalizability of our findings. Future studies should apply a broader scope and include a multitude of countries in their research.

Conclusion

Thus far, FF’s backgrounds and pathways were predominantly studied through open sources and comparisons within FF samples were rarely conducted. Through the probation file data, we could confirm earlier research on demographics of FFs and extend current knowledge. We detected five factors that stand out and need further examination in future research: experiencing profound events (combined with feelings of injustice and signs of extreme ideology), radical social contacts, mental health, criminal history, and the online environment. Especially, social (online) contacts and the facilitating function of the online environment seem to have been highly important in taking the actual step of becoming a FF. Also, the explorations of the age and sex comparisons demonstrate how important it is to look beyond the FF group as a whole. Although the outcome might be similar—becoming (attempted) FFs—the factors and mechanisms that lead to that outcome may differ substantially. Therefore, more in-depth comparisons within terrorist and extremist samples, and more specifically, FF samples, would create a better understanding of what leads to this particular outcome and what might be ways to prevent people from ending up there.

Author Statement

Fabienne Thijs performed the probation file research and wrote the manuscript. Elanie Rodermond and Edward Kleemans provided feedback on the drafts and final version of the manuscript.

Availability of Data and Material

The dataset is not available publicly but is stored within the protected environment of the NSCR. The codebook can be provided upon request.

Acknowledgements

We thank the Dutch Probation Service, particularly Ada Andreas and her TER-team and SIBA department colleagues, for their cooperation and dedication to help and provide us with the necessary facilities to conduct our research. Also, a special thanks to the coders: Melissa Eijking, Eva Hortensius, Bodine Koopen, Romy van Kuik, Ike Lahaut, and Annabel de Rooij. Their critical remarks and dedication improved the coding process, both in terms of efficiency as well as quality. Finally, we thank the reviewers for their valuable comments.

Disclosure Statement

The authors report there are no competing interests to declare.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This work was supported by the Dutch Research Council (NWO) Research Talent grant [406.18.527].

Notes

1 B. Van Ginkel and E. Entenmann, eds., The Foreign Fighters Phenomenon in the European Union. Profiles, Threats & Policies (The Hague: ICCT, 2016), 3.

2 L.L. Dawson, A Comparative Analysis of the Data on Western Foreign Fighters in Syria and Iraq: Who Went and Why? (The Hague: ICCT, 2021), 39–41, https://icct.nl/publication/a-comparative-analysis-of-the-data-on-western-foreign-fighters-in-syria-and-iraq-who-went-and-why/.

3 e.g., E. Bakker and R. de Bont, “Belgian and Dutch Jihadist Foreign Fighters (2012–2015): Characteristics, Motivations, and Roles in the War in Syria and Iraq,” Small Wars and Insurgencies 27, no. 5 (2016); S. Harris-Hogan, A. Amarasingam, and L.L. Dawson, “A Comparative Analysis of Australian and Canadian Foreign Fighters Traveling to Syria and Iraq,” Studies in Conflict & Terrorism (2022); Dawson, A Comparative Analysis; A. Speckhard and M. Ellenberg, “The effects of Assad’s Atrocities and the Call to Foreign Fighters to Come to Syria on the Rise and Fall of the ISIS Caliphate,”" Behavioral Sciences of Terrorism and Political Aggression 14, no. 2 (2020); D. Weggemans, E. Bakker, and P. Grol, “Who Are They and Why Do They Go? The Radicalisation and Preparatory Processes of Dutch Jihadist Foreign Fighters,” Perspectives on Terrorism 8, no. 4 (2014); A. Peliger and D. Milton, From Cradle to Grave: The Lifecycle of Foreign Fighters in Iraq and Syria (West Point: Combating Terrorism Center, 2016), https://apps.dtic.mil/sti/pdfs/AD1021545.pdf; G.M. Tezcür and C. Besaw, “Jihadist waves: Syria, the Islamic State, and the Changing Nature of Foreign Fighters,” Conflict Management and Peace Science 37, no. 2 (2017); S.C. Reynolds and M.M. Hafez, “Social Network Analysis of German Foreign Fighters in Syria and Iraq,” Terrorism and Political Violence 31, no. 4 (2019); L.L. Dawson and A. Amarasingam, “Talking to Foreign Fighters: Insights into the Motivations for Hijrah to Syria and Iraq,” Studies in Conflict & Terrorism 40, no. 3 (2017); L. Lindekilde, P. Bertelsen, and M. Stohl, “Who Goes, Why, and With What Effects: The Problem of Foreign Fighters from Europe,” Small Wars and Insurgencies 27, no. 5 (2016); A.W. Weenink, “Adversity Criminality and Mental Health Problems in Jihadis in Dutch Police Files,” Perspectives on Terrorism 13, no. 5 (2019); A.W. Weenink, “Behavioral Problems and Disorders among Radicals in Police Files,” Perspectives on Terrorism 9, no. 2 (2015); K. Rękawek, Not Only Syria? The Phenomenon of Foreign Fighters in a Comparative Perspective (Amsterdam: IOS Press, 2017).

4 A. Koch, “The Non-Jihadi Foreign Fighters: Western Right-Wing and Left-Wing Extremists in Syria,” Terrorism and Political Violence 33, no. 4 (2019); J. Fritz and J.K. Young, “Transnational Volunteers: American Foreign Fighters Combating the Islamic State,” ibid 32, no. 3 (2020).

5 e.g., C. Paulussen and K. Pitcher, Prosecuting (Potential) Foreign Fighters: Legislative and Practical Challenges (The Hague: ICCT, 2018); R. Coolsaet and T. Renard, Returnees: Who Are They, Why Are They (Not) Coming Back and How Should We Deal With Them? Assessing Policies on Returning Foreign Terrorist Fighters in Belgium, Germany and the Netherlands (Brussels: Egmont Institute, 2018), http://hdl.handle.net/1854/LU-8547880; E. Bakker, C. Paulussen, and E. Entenmann, “Returning Jihadist Foreign Fighters Challenges Pertaining to Threat Assessment and Governance of this Pan-European Problem,” Security and Human Rights 25, no. 1 (2014); H.B. Lazreg, “Western Foreign Fighters: The Threat to Homeland and International Security,” Terrorism and Political Violence 32, no. 4 (2020); A. Zammit, Australian Foreign Fighters: Risks and Responses (Lowy Institute, 2015), https://www.lowyinstitute.org/publications/australian-foreign-fighters-risks-responses#_edn1; D. Malet, et al., The European Experience with Foreign Fighters and Returnees (Egmont Institute, 2018), http://www.jstor.org/stable/resrep17406.4; D. Malet and R. Hayes, “Foreign Fighter Returnees: An Indefinite Threat?,” Terrorism and Political Violence 32, no. 8 (2018); E. Strømmen, Jihadi Brides or Female Foreign Fighters? Women in Da'esh – from Recruitment to Sentencing (Oslo: PRIO, 2017), https://www.prio.org/publications/10546; S. Wittendorp, et al., Measures against Jihadist Foreign Fighters: A Policy Comparison between the Netherlands, Belgium, Denmark, Germany, France, the UK and the US (2010 to 2017) (ISGA, 2017), https://hdl.handle.net/1887/64739; E. Azani and L. Koblentz-Stenzler, “Muslim Converts Who Turn to Global Jihad: Radicalization Characteristics and Countermeasures,” Studies in Conflict & Terrorism 45, no. 2 (2022)

6 e.g., Weenink, “Adversity Criminality”; Weggemans, Bakker, and Grol, “Who Are.”; Bakker and de Bont, “Belgian and Dutch Jihadist”; R. Bergema and M. van San, “Waves of the Black Banner: An Exploratory Study on the Dutch Jihadist Foreign Fighter Contingent in Syria and Iraq,” Studies in Conflict & Terrorism 42, no. 7 (2019); M. Van San, “Lost Souls Searching for Answers? Belgian and Dutch Converts Joining the Islamic State,” Perspectives on Terrorism 9, no. 5 (2015); E. Bakker and P. Grol, Motives and Considerations of Potential Foreign Fighters from the Netherlands (Den Haag: ICCT, 2015); M. Van Leyenhorst and A. Andreas, “Dutch Suspects of Terrorist Activity: A Study of Their Biographical Backgrounds Based on Primary Sources,”" Journal for Deradicalization Fall, 12 (2017).

7 Weenink, “Adversity Criminality.”; Weenink, “Behavioral Problems.”

8 Van Leyenhorst and Andreas, “Dutch Suspects.”

9 e.g., D. Malet, “Why Foreign Fighters?,” Orbis 54, no. 1 (2010); M.G. Donnelly, T.M. Sanderson, and Z. Fellman, Foreign Fighters in History (Center for Strategic and International Studies, 2017), https://www.jstor.org/stable/resrep23304; D. Malet, “A History of Foreign Fighters,” in Foreign Fighters (Oxford University Press, 2013); K. Rękawek, Foreign Fighters in Ukraine: The Brown-Red Cocktail (Routledge, 2022).

10 Malet, “Why Foreign Fighters?,” 114.

11 Ibid.

12 Strømmen, Jihadi Brides, 2.

13 R. Frenett and T. Silverman, “Foreign Fighters: Motivations for Travel to Foreign Conflicts,” in Foreign Fighters under International Law and Beyond, ed. A.d. Guttry, et al. (2016).

14 Strømmen, Jihadi Brides, 2.

15 Frenett and Silverman, “Foreign Fighters.”

16 T. Hegghammer, “Syria’s Foreign Fighters,” Foreign Policy, 2013, https://foreignpolicy.com/2013/12/09/syrias-foreign-fighters/.

17 Tezcür and Besaw, “Jihadist waves,” 216.

18 E.D. Løvlien, “Domestic Support for Foreign Fighting: Attitudes Towards the Use of Violence Abroad,” European Journal for Security Research 6, no. 1 (2021): 74.

19 Paulussen and Pitcher, “Prosecuting (Potential) Foreign Fighters,” 18–19.

20 H. Cuyckens and C. Paulussen, “The Prosecution of Foreign Fighters in Western Europe: The Difficult Relationship Between Counter-Terrorism and International Humanitarian Law,” Journal of Conflict and Security Law 24, no. 3 (2019): 538.

21 T. Hirschi, Causes of delinquency (Berkeley: University of California Press, 1969).

22 Ibid.

23 R.J. Sampson and J.H. Laub, Crime in the Making: Pathways and Turning Points through Life (Cambridge, MA: Harvard University Press, 1993).

24 J.H. Laub, Z.R. Rowan, and R.J. Sampson, “The Age-Graded Theory of Informal Control,”" in The Oxford Handbook of Developmental and Life-Course Criminology, ed. D.P. Farrington, L. Kazemian, and A.R. Piquero (New York: Oxford University Press, 2018), 296.

25 A.R. Feddes, L. Nickolson, and B. Doosje, “Triggerfactoren in Het Radicaliseringsproces,” Justitiële verkenningen 2 (2016).

26 M. Vergani, et al., “The Three Ps of Radicalization: Push, Pull and Personal. A Systematic Scoping Review of the Scientific Evidence about Radicalization Into Violent Extremism,” Studies in Conflict & Terrorism 43, no. 10 (2018): 3–4.

27 Feddes, Nickolson, and Doosje, “Triggerfactoren,” 38.

28 B. de Graaf, Radicale verlossing. Wat terroristen geloven. (Amsterdam: Prometheus, 2021); R. Basra and P.R. Neumann, “Criminal Pasts, Terrorist Futures: European Jihadists and the New Crime-Terror Nexus,” Perspectives on Terrorism 10, no. 6 (2016).

29 de Graaf, Radicale verlossing, 255–72.

30 S. Maruna, Making Good: How Ex-Convicts Reform and Rebuild their Lives (Washington, DC: American Psychological Association, 2001), 10.

31 de Graaf, Radicale Verlossing, 262–63, 70–71.

32 e.g., A. Perliger, G. Koehler-Derrick, and A. Pedahzur, “The Gap Between Participation and Violence: Why We Need to Disaggregate Terrorist ‘Profiles',” International Studies Quarterly 60, no. 2 (2016); R. Scrivens, et al., “Differentiating Online Posting Behaviors of Violent and Nonviolent Right-Wing Extremists,” Criminal Justice Policy Review 33, no. 9 (2022); K. Jasko and G. LaFree, “Who is More Violent in Extremist Groups? A Comparison of Leaders and Followers,” Aggressive Behavior 46, no. 2 (2020); C. Herath and J. Whittaker, “Online Radicalisation: Moving Beyond a Simple Dichotomy,” Terrorism and Political Violence (2021); J. Kenyon, J.F. Binder, and C. Baker-Beall, “Online Radicalization: Profile and Risk Analysis of Individuals Convicted of Extremist Offences,” Legal and Criminological Psychology (2022); V.M. Hudson and K.B. Hodgson, “Sex and Terror: Is the Subordination of Women Associated with the Use of Terror?,” Terrorism and Political Violence 34, no. 3 (2020); M. Jensen, P. James, and E. Yates, “Contextualizing Disengagement: How Exit Barriers Shape the Pathways Out of Far-Right Extremism in the United States,” Studies in Conflict & Terrorism (2020).

33 Jasko and LaFree, “Who is More Violent,” 1.

34 e.g., E.P. Shulman, L.D. Steinberg, and A.R. Piquero, “The Age-Crime Curve in Adolescence and Early Adulthood is Not Due to Age Differences in Economic Status,” Journal of Youth and Adolescence 42, no. 6 (2013); T. Hirschi and M. Gottfredson, “Age and the Explanation of Crime,” American Journal of Sociology 89, no. 3 (1983); T.E. Moffitt, “Adolescence-Limited and Life-Course-Persistent Antisocial Behavior: A Developmental Taxonomy,” Psychological Review 100, no. 4 (1993).

35 Shulman, Steinberg, and Piquero, “The Age-Crime Curve,” 848.

36 D.P. Farrington, “Age and Crime,” Crime and Justice 7 (1986): 236.

37 D.P. Farrington, H. Gaffney, and M.M. Ttofi, “Systematic Reviews of Explanatory Risk Factors for Violence, Offending, and Delinquency,” Aggression and Violent Behavior 33 (2017).

38 For studies on female FFs, see for instance: A.N. Spencer, “The Hidden Face of Terrorism: An Analysis of the Women in Islamic State,”" Journal of Strategic Security 9, no. 3 (2016); Strømmen, Jihadi Brides; H. Mondani, et al., Women in Violent Extremism in Sweden (Copenhagen: N.C.o. Ministries, 2021).

39 See for example: T.E. Moffitt, et al., Sex Differences in Antisocial Behaviour: Conduct Disorder, Delinquency, and Violence in the Dunedin Longitudinal Study (New York, NY: Cambridge University Press, 2001).

40 Zammit, Australian Foreign Fighters, 2.

41 A. Ibrahim, “Ending the Syrian Civil War: Is There a Federal Solution?,” Geopolitics, History, and International Relations 9, no. 2 (2017): 149–51.

42 Z. Beauchamp, “ISIS, A History: How the World's Worst Terror Group Came To Be,” VOX, Date 19 Nov 2015, https://www.vox.com/2015/11/19/9760284/isis-history.

43 Malet, et al., The European Experience, 6; ibid.

44 Bergema and van San, “Waves,” 653–54.

45 Based on article 6 of the GDRP, the Ethics Committee approved our data collection without gathering informed consent.

46 TER-team refers to Terrorism, Extremism, and Radicalization team.

47 See also: https://www.reclassering.nl/over-de-reclassering/wat-wij-doen/aanpak-radicalisering (in Dutch only).

48 The TER-team has also a category of clients that were initially monitored and supervised because of a nonterrorist crime but who were, during their probation time, identified as becoming more radicalized. These cases were excluded.

49 As we focus on information prior to becoming a foreign fighter, we do not make a distinction between those who attempted to leave, those who reside abroad, and those who returned.

50 Data collection was interrupted a few times due to Covid-19 measures.

51 Our codebook includes definitions for each variable, with instructions and examples of how to apply the different codings consistently.

52 “1” was coded when it was stated (rather than implied) that a feature was present. For some variables, however, we also coded “1” (or “3” meaning “highly likely,” eventually added to the “yes” category) when it was highly likely that a certain feature was present based on the information. For instance, when a suspect was a parent and lived with his/her children, we assumed he/she had custody over the children.

53 There is some overlap between the categories.

54 A. Wilner and I. Yar, Canadian Terrorists by the Numbers: An Assessment of Canadians Joining and Supporting Terrorist Groups (MacDonald-Laurier Institute, 2019), 26, https://macdonaldlaurier.ca/files/pdf/20190205_MLI_Canadian_Terrorists_Wilner_PAPER_WebFinal.pdf.

55 de Graaf, Radicale Verlossing, 255–72.

56 I. Versteegt, et al., Terrorism Adversity and Identity. A Qualitative Study of Detained Terrorism Suspects in Comparison to Other Detainees (Amsterdam: NSCR, 2018), 120–21, https://nscr.nl/app/uploads/2019/01/NSCR-Rapport_Terrorism-adversity-and-identity_181218-2.pdf.

57 Bergema and van San, “Waves,” 656.

58 Weenink, “Adversity Criminality.”; Weenink, “Behavioral Problems.”

59 See for example: P. Gill, et al., “Systematic Review of Mental Health Problems and Violent Extremism,” Journal of Forensic Psychiatry & Psychology 32, no. 1 (2021).

60 Basra and Neumann, “Criminal Pasts.”

61 Ibid.; de Graaf, Radicale verlossing; Maruna, Making good.

62 D.P. Farrington, “Developmental and Life-Course Criminology: Key Theoretical and Empirical Issues – The 2002 Sutherland Award Address,” Criminology 41, no. 2 (2003): 235–36.

63 de Graaf, Radicale verlossing; Maruna, Making good.

64 E.M. Saltman and M. Smith, Till Martyrdom Do Us Part. Gender and the ISIS Phenomenon (London: Institute for Strategic Dialogue, 2015), 15–16, www.strategicdialogue.org/wp-content/uploads/2016/02/Till_Martyrdom_Do_Us_Part_Gender_and_the_ISIS_Phenomenon.pdf; Y. Veilleux-Lepage, A. Phelan, and A.D. Lokmanoglu, “Gendered Radicalisation and ‘Everyday Practices’: An Analysis of Extreme Right and Islamic State Women-Only Forums,” European Journal of International Security (2022).

65 For policy recommendations, see for example: Veilleux-Lepage, Phelan, and Lokmanoglu, “Gendered radicalisation.”