Accessible Published by De Gruyter August 18, 2021

Radicalization potentials of young Muslims in prison: What role do religious factors play?

Sonja King, Johann Endres and Mark Stemmler


Wie steht es um das Ausmaß islamistischer Einstellungen unter jungen Muslimen im deutschen Justizvollzug? Ziel der vorliegenden Arbeit war es, Radikalisierungspotenziale und damit in Verbindung stehende mögliche religionsbezogene Einflussfaktoren sowie die Dauer der bisher verbüßten Haft anhand einer Stichprobe von 87 jungen muslimischen Inhaftierten explorativ zu untersuchen. Die Teilnehmer bearbeiteten hierfür einen Fragebogen, in dem wir Fragen zu religiösen Sozialisationserfahrungen und aktueller Religiosität stellten, sowie Zustimmung zu fundamentalistischen und religiös-militanten Aussagen erhoben. Wir untersuchten die Häufigkeitsverteilungen fundamentalistischer und militanter Einstellungen und verglichen die Skalenmittelwerte von Muslimen mit einer nicht-muslimischen Vergleichsgruppe (n = 255). Anschließend berechneten wir Regressionsmodelle und modellierten Kausalketten, die die von uns untersuchten Konzepte durch Mediation miteinander verknüpften. Unsere Ergebnisse geben Hinweise auf eine erhöhte Vulnerabilität von muslimischen Gefangenen für Radikalisierung. Weiterhin geht aus unseren Ergebnissen hervor, dass Sozialisationseinflüsse zwar offenbar keinen unmittelbaren Einfluss haben, sich aber indirekt auf das Ausmaß militanter Einstellungen auswirken. Die bisher im Gefängnis verbrachte Zeit spielt eine untergeordnete Rolle. Wir diskutieren die Verallgemeinerbarkeit unserer Ergebnisse und mögliche Implikationen für den Strafvollzug.


How widespread are Islamist attitudes in German prisons? The present research aims at exploring radicalization potentials as well as religion related factors among 87 young Muslim inmates in Bavarian prisons. Participants completed a survey in which we collected measures of religious socialization, religiousness, fundamentalism, and religious militancy. We investigated the distribution of fundamentalist and militant attitudes and compared the means of Muslims and non-Muslim prisoners (n = 255), calculated regressions, and modelled causal chains linking the relevant concepts through mediation. Our results indicate an increased vulnerability of Muslim inmates for religious radicalization. The results suggest that there is no immediate effect of religious socialization on religious militancy, but that it indirectly influences militant attitudes. The time spent in prison plays a subordinate role. Generalization and implications for the prison context are discussed.

1 Introduction

Almost 20 years after one of the most devastating terror attacks in the West, the topics radicalization and terrorism are still highly relevant. Against the background of an increase in right-wing populism in many countries and the recent protests against »Corona measures«, two particularly shocking attacks in late 2020 were a painful reminder of the vulnerability of Europe for Islamist motivated attacks. In October, a French teacher was beheaded for using Mohammed caricatures to discuss the freedom of expression. Not even three weeks later, Austria had to deal with the first major Islamist terror attack directed against the general population. Like in most cases of Islamist motivated terrorism attacks in Europe, the perpetrators had some things in common: They considered themselves Muslims, and they had a criminal past – many of them had even spent time in prison at some point in their lives. But are these characteristics interrelated, and, if they are, how? And what role does religion play?

With the current paper, we present a study among young Muslim inmates in Bavarian prisons. We were interested in how widespread aspects of Islamist ideologies are and explore the role of religion. The study was carried out as one part of a research project funded by the German Research Foundation (Deutsche Forschungsgemeinschaft (DFG); STE923/10).

We will use the terms as follows: Islamism is a movement with a political goal, which is to replace un-Islamic concepts in non-private spaces (Armborst 2014). As democracy is an »un-Islamic« concept Islamism rejects, Islamism is a form of political extremism (Backes 2007). Salafism is a fundamentalist movement related to Islamism, which is divided into several currents with different stances towards the use of violence (Wiktorowicz 2006). When we talk about militancy, we refer to the readiness to use violence, which is characteristic for Jihadism, the »most radical form of contemporary Islamism« (Armborst 2014, 235).

2 Background

2.1 The role of prison in radicalization and de-radicalization

Prisons are often discussed as breeding grounds, schools, or assembly lines for violent extremists (see, for example, Jones 2014; Mulcahy, Merrington & Bell 2013; Rappaport, Veldhuis & Guiora 2013; Sinai 2014). Since 9/11, in particular Muslim extremism [sic] has been a great concern (Sinai 2014). Yet to this day, there is no clarity regarding the severity of the problem. Whereas Ronco, Sbraccia and Torrente (2020) conclude that »the attitude of some countries toward prison radicalization appears like a form of moral panic (Cohen 1972)«, other authors do not completely dismiss the idea that there could be a link between imprisonment and radicalization. The concerns about »prison radicalization« were empirically supported by the results of research in Spanish prisons (Trujillo et al. 2009), a study among prisoners in a Philippine jail (Kruglanski et al. 2016), and the analysis of 79 terrorist profiles (Basra, Neumann & Brunner 2016), which indicated that a criminal past and serving time in prison was not uncommon for jihadists. There are some specific characteristics of prisons that seem to be related to radicalization, for example, prison gangs (Hamm 2009; Rushchenko 2019; Sinai 2014), the number of Muslim offenders or offenders convicted of terrorist offenses (Trujillo et al. 2009) and overcrowding (Hamm 2009; Jones 2014; Neumann 2010).

However, Neumann (2010) also points to positive potentials of imprisonment, and some empirical studies show that a period of imprisonment can have a protective function in terms of contributing to disengagement (Lösel et al. 2020). This is particularly interesting given the high number of foreign fighters who participated in the war in Syria and have since returned to Europe. It is an open question whether these individuals are better accommodated together with other prisoners to support their resocialization, or if it is better to separate them so that they cannot disseminate extremist attitudes and ideology, and if they can be deradicalized or not (Endres & King 2018a). To this point, there is only limited high-quality evidence for the effectiveness of interventions (Endres & King 2018 a; Jugl et al., 2021; Silke & Veldhuis, 2017). This research gap is one of the major problems with radicalization research (see, for example, Schuurman & Eijkmann 2013; Silke 2001), and is partially based on a substantial lack of data (Ronco et al. 2020).

2.2 Theoretical models of radicalization

There are quite a few plausible theoretical radicalization models that are widely accepted, to some extent empirically supported, and often referred to. On the one hand, there are models outlining radicalization as a process, a staircase, or stages (see, for example, McCauley & Moskalenko 2017; Moghaddam 2005; Silber & Bhatt 2007; Sinai 2014). All of them describe radicalization as an escalating development that may result in a terrorist act. On the other hand, root cause approaches try to identify the origins of radicalization, or the components that contribute to the making of an extremist (see, for example, Kruglanski et al. 2014; Kruglanski & Webber 2014; Veldhuis & Staun 2009; Webber & Kruglanski 2017). We are going to illustrate such models in the following section.

2.2.1 Radicalization processes

Partially drawing on work by Silber and Bhatt (2007), who proposed a model for radicalization in the West, Sinai (2014) suggested a model of prison radicalization that describes consecutive phases of the radicalization process in the prison environment. Both models begin with risk factors on the individual level. Whereas Silber and Bhatt describe the individuals on the first stage as »unremarkable« (Silber & Bhatt 2007, 1) Muslims under 35, who often live in segregated ethnic enclaves, Sinai identifies a history of violent behaviour, anti-social attitudes, low self-esteem, mental disorders and a sense of victimization as important characteristics that create a vulnerability and which are partially characteristic for the population of prison inmates. Sinai’s model then postulates situational and contextual risk factors such as the presence of an extremist ideology and/or a charismatic leader in the prison environment, before he draws level with Silber and Bhatt again. Both models describe the core process of radicalization as a phase of identification with a new worldview, which can lead to the adoption of extremist ideologies. The next phase is described as indoctrination, and finally, jihadization (Silber & Bhatt 2007) or militancy (Sinai 2014) concludes the development of a radical mindset. Over these phases, the individual adopts a religious-political ideology including the idea that violence can or even has to be a legitimate means to fight for the religiously defined goals. In the end, the individual considers him-/herself as a warrior for God who is willing to use violence to serve the higher cause. As Sinai (2014) focuses on prison radicalization, his model continues with post-release activities, which might lead to an attack and possible further developments.

2.2.2 Driving forces of radicalization

Conceptually, process models have a rather descriptive character. Other models attempt a more explanatory, dynamic approach, as for example the one proposed by Kruglanski and Webber (2014). We will relate to it as the »3N« model (Webber & Kruglanski 2017). The model comprises three levels of factors that critically contribute to radicalization and that are basically included in Sinai’s model as well: Individual factors (needs in the »3N« model), a culturally and religiously framed ideology (narrative in the »3N« model), and social factors (network in the »3N« model). As for the individual needs, Kruglanski emphasizes motivating factors, especially a »significance quest« (Kruglanski et al. 2009; Kruglanski & Orehek 2011; Kruglanski et al. 2013; Kruglanski et al. 2014) that can contribute to radicalization. The significance quest is often triggered by a significance loss – an experience of humiliation or deprivation (Kruglanski et al., 2014) – or by a possibility to gain more significance, for instance, by fighting or even dying for a higher cause (Dugas et al. 2016; Kruglanski et al. 2009; Kruglanski et al. 2014). Ideology, in turn, serves as a framework that is used to make sense of what is happening (Oliver & Johnston 2000), but also, to justify violence as a means to fight injustice. A composite narrative extracted from narratives of seven militant groups (Saucier et al. 2009, 65) summarizes ideology as follows:

«We (i.e., our group, however defined) have a glorious past, but modernity has been (...) a great catastrophe (...). (We) are (...) obstructed from reaching our rightful place, (...) by an enemy so evil that it does not even deserve to be called human. (...). Extreme measures are required; indeed, any means will be justified (...). It is a duty to kill the perpetrators of evil (...) Those who sacrifice themselves in our cause will attain glory (...).«

The most important functions of ideology illustrated here are legitimizing violence by pointing to the higher cause and the power of the enemy, and by dehumanizing the enemy, which is another technique of moral disengagement or neutralization (Ribeaud & Eisner, 2010). Besides, ideology also includes a clear definition of the in- and outgroup. This clear definition of the group relates to the third factor relevant to radicalization, the social group, which is a point of reference for identification as explained above. This model is a valuable contribution because it highlights the meaning of psychological factors and of mutual reinforcement of individual needs, social factors, and ideology.

2.3 What could make prisons a »breeding ground«?

Imprisonment comes along with psychological distress (Endres & King 2018 a; Hosser 2008). This can be perceived as threatening, which has been shown to increase support for extremist violence (Beller & Kröger 2017). Deprivations on many levels (»pains of imprisonment«; Sykes 1958) can be related to the driving forces described in the »3N« model. The specific hardships include being cut off from family and friends, or a lack of privacy and autonomy (Endres & King 2018 a; Rocheleau 2013; Sykes 1958); therefore, a loss of self-worth is a frequent first reaction to imprisonment (Hosser 2008). Sykes (1958) described how the individual begins to develop a self-image in which the morally acceptable idea of oneself fades. According to Webber and Kruglanski (2017), the transgression against a social norm can suffice to trigger the quest for significance that contributes to radicalization according to the »3N« model. Furthermore, prisoners are cut off from (protective) social relations, for example, family or romantic partners. Therefore, the meaning of interpersonal contacts and the influence of subcultures in prison (Endres & King 2018a) increases – for example, other inmates with the same cultural or religious background (Thomas & Zaitzow 2006). Mansour (2016, 143) refers to Salafists as the »better social workers«, who provide understanding, support, and the feeling of belonging, and can foster the entry into a radical network as described in the »3N« model. Religion as a coping strategy in the context of stressful life events is common (Ano & Vasconcelles 2005). Therefore, a religious search for meaning or even conversion in prisons is not rare (Maruna, Wilson & Curran 2006; Spalek & El-Hassan 2007). However, this does not necessarily have to lead to radicalization. An obvious starting point to engage prisoners with extremist ideology could be using feelings of unjust treatment, and setting up prison staff, other authorities, or the state as an enemy. Furthermore, religious ideology provides a new worldview and perspective for those who are facing social exclusion. Muslims make up a disproportionately high percentage of the prison population, especially among the young prisoners (Endres & King 2018a). However, the general minority status could be used to attribute for perceived injustice. Prisoners will always wear the stigma of a criminal, but a conversion to religion provides the chance for a new beginning (Endres & King 2018 a; Sinai 2014) in a new framework of reference, in which a superior power is the only one to judge them. The connectivity with extremist ideology represents the third »N« in the radicalization model.

2.4 The role of religion

This eschatological aspect of fundamentalist ideology like the one of Salafism (Wiktorowicz 2006) is the most significant difference to other extremist ideologies. The factor »religion« can be related to the three levels described in the model suggested by Kruglanski and Webber (2014). First, it can only put down roots if an individual is receptive to religious ideas, second, militant Islamist ideology draws on religion, and third, it connects the individual to a religious community that frames religious socialization processes.

The integration of Muslims in the West is facing complex challenges; partially, because in this population, religion makes up a substantial part of a person’s identity (Frindte 2011). As »Islam has a remarkable tendency to uphold high levels of religious identification as well as practice and belief from generation to generation« (Kühle 2012, 114), the tradition of religious values and practices is highly important. Religious socialization can be defined as »an interactive process through which social agents influence individuals’ religious beliefs and understandings« (Sherkat 2003, 151). Exposure to socialization experiences starts in early childhood (Cornwall 1988). This can include rituals and practices such as attending services, learning about religious scripture, prayer, and rules regarding what is right and what wrong (Cornwall 1988). We therefore assume that, like in other contexts (Ostroff & Kozlowski 1992), socialization can be seen as a learning process, in which individuals gather procedural or declarative knowledge (De Jong & Ferguson-Hessler 1996) by different types of learning, for example, model learning (Bandura 1969).

Regarding the ‘new generation’ of Salafists in Western countries, there are claims that they are religious analphabets lacking adequate religious socialization (Dantschke 2017). This implies the assumption that religiousness might protect against a deviant, radical understanding of religion. Empirical support for this argument can be seen in the fact that the protective effect of religion can not only benefit psychosocial health (e.g., Braam & Koenig 2019; Norko 2017), but that the protective effects of religion even extend to delinquency (Johnson et al. 2000). In contrast, survey data suggest that particularly very traditional religiousness is related to elevated levels of religious fundamentalism (Frindte 2011). Therefore, if we want to investigate the role of religion in the development of extremist attitudes, religious socialization is an important starting point.

2.5 Aims and research questions

In the project presented with this paper, we aimed at finding answers to the following questions:

  1. 1.

    How widespread are fundamentalist and religious-militant attitudes among young prisoners?

  2. 2.

    What religion related factors contribute to these attitudes?

  3. 3.

    What is the role of the duration of detention?

3 Methods

3.1 Data collection

We collected our data in four prisons for young offenders in Bavaria. The survey was always completed by individuals in a solitary cell or in a supervised group setting to avoid mutual influencing. The prisoners received a compensation with a value of approximately 3 Euros (chocolate, stamps, or a financial deposit). After excluding cases of response sets (Cronbach 1946), we could include N = 342 complete questionnaires into our analyses.

3.2 Sample

Our sample comprised N = 342 males, of which n = 87 (25.4 %) were Muslim, n = 198 (57.9 %) were Christian, n = 44 (12.9 %) had no religious denomination, and n = 9 (2.6 %) belonged to other religions. The youngest participants were 15 years old; the highest age was 27 (M = 19.82; SD = 2.46). Of the 258 cases (75.4 %) whose first citizenship was German, 58 (17.0 %) had a second citizenship. The individuals in our sample had served up to 60 months in the current facility so far (M = 10.54, SD = 10.43, n = 323).

3.3 Measures

To assess religious socialization, current religiousness, and fundamentalism, we used items we had developed and validated in a pilot study (Endres & King 2018b). Based on the resulting item parameters (difficulty, discriminatory power, reliability; see Lienert & Raatz 1998), we selected items to generate the final version of the scales. The items of the religious socialization and religiousness[1] scales were matched according to their content, which was based on the publications by Cornwall (1988) and Sherkat (2003). For example, whereas one item in the socialization scale regarded saying prayers in the family of origin, the religiousness scale asked for saying prayers these days. The measure for religious fundamentalism was inspired by Mansour (2016), who described the characteristics of the current neo-Salafist youth movement: Black-and-white thinking, feelings of superiority, negative attitudes towards sexuality, proselytizing, a literal interpretation of religious scripture, belief in heaven and hell, and pedagogics using fear. As we wanted to disseminate the questionnaires among all prisoners independently of their religious denomination to have a comparison group, we made sure that the items were not specific to Islam. For example, we did not use the term »Quran« or »mosque« but wrote about »religious scripture« or »place of worship«. To operationalize »religious militancy«, we used six items of the Assessment and Treatment of Radicalization Scale[2] (ATRS; Loza 2007), partially in a slightly simplified version. The scale was developed to measure Middle Eastern ideologies (Karimi, Cimbura & Loza 2019), but as we wanted to present the items to a non-Muslim comparison group, we made sure that all items we presented made sense independently of the cultural background. The items we chose addressed war, death, or fighting in combination with a justification related to religion. Since we had reason to assume that the sporadic missing values in the individual items did not remain systematically unanswered, we decided to use a listwise exclusion when 20 % or more of the items of a scale were missing. All scales correlated significantly; coefficients ranged from .44 to .70 (Table 1). Example items, item count, response format, and descriptive data of the instruments are provided in Table 2.

Table 1

Bivariate correlation coefficients

Religious socialization Religious-ness Fundamen-talism
Religiousness .64
Fundamentalism .54 .68
Extremism .44 .55 .70

Note. Sample sizes range from 310 to 338. All coefficients p < .001

Table 2:

Overview of measures, descriptive data, and reliability coefficients

Scale Items Range Descriptive Data Reliability
n M SD n α
Religious socialization

Example: In my family, we used to pray.
10 0 – 4 338 1.34 0.97 315 0.90

Example: I consider myself religious.
10 0 – 4 334 1.27 1.01 306 0.91

Example: Sinners go to hell.
25 0 – 4 311 1.43 0.79 262 0.91

Example: Dying while fighting for God is a great honour.
6 0 – 4 332 0.86 0.85 313 0.81

Our survey included many highly sensitive topics, which is why we decided to avoid further risk of perceived stigmatization by asking details about the participants’ offences. Regarding the duration of detention, we asked the inmates how long they had been in prison and in a separate question about prior periods of detention. For the present analysis, we used the current period of detention.

3.4 Analyses

We used SPSS 26 for the data analysis. For group comparisons, we used t-tests. Whenever necessary (violation of distribution assumption, uneven group sizes), we applied a bootstrapping procedure (Field 2013). The bootstrapping was based on the drawing of 1,000 samples with a 95 % confidence interval[3]. To test the influence of multiple factors, we used linear regression models. Since especially religious socialization, religiousness, fundamentalism, and militancy correlated highly[4], we always checked for variance inflation and tolerance values to rule out multicollinearity (Field 2013), which turned out to not be the case. We also checked the distribution of the residuals when we tested regression models. To run the mediation analysis, we followed Hayes (2018) and installed the PROCESS©[5] macros (Version 3.4) into SPSS.

4 Results

In a first step, we explored the distribution of reported fundamentalism and religious[6] militancy measures. Results are shown in Table 3.

Table 3

Distribution of fundamentalist and militant attitudes in the Muslim group

n Percentiles Min Max
25 50 75
Fundamentalism 68 1.77 2.16 2.63 0.28 3.52
Militancy 81 0.50 1.33 2.00 0.00 3.33

Each of the items of the scale » militancy« were at least partially endorsed by more than a quarter of the respondents. »Dying for God in battle is a great honour« and »People who die for God go to heaven« were both partially (scale value 3) or strongly (scale value 4) supported by 32 respondents (36.8 %); »I would not turn down the opportunity to die for God in battle« by 22 (25.3 %). The statement »I support the fight against the unbelievers« received the partial or strong agreement of seven persons (8.0 %), »The fight between believers and unbelievers is morally justified« as well as »Terrorists« are in fact fighters for the will of God« by four persons each (4.6 %).

We then contrasted the means of the group of Muslims against the comparison group of non-Muslim prisoners. Means, standard deviations and results of the t-tests are displayed in Table 4.

Table 4

Results of the group comparisons

n M SD df t
Religious socialization
Muslim 83 2.10 .99 122 8.39***
Not Muslim 255 1.09 .82
Muslim 81 1.77 .98 332 5.39***
Not Muslim 253 1.11 .96
Muslim 68 2.13 0.64 309 9.54***
Not Muslim 243 1.23 0.71
Religious militancy
Muslim 81 1.29 0.85 330 5.38***
Not Muslim 251 0.73 0.81

Note. *** = p < .001

To assess the extent to which the variables had predictive power, we included them in linear regression models. Our first model was supposed to predict religiousness. Based on theoretical assumptions, we included duration of detention and religious socialization as predictors, of which only religious socialization turned out to have a significant weight. The duration of detention, however, did not make a significant contribution to the prediction model (see Table 5a). We then calculated a regression model to predict religious fundamentalism, in which we included the previously used variables religious socialization, the duration of detention and, additionally, religiousness. The results in Table 5 b show that, whereas religiousness was a significant predictor, the other variables did not have a significant weight. The last model was to predict religious militancy. Besides all variables used in the previous analyses, fundamentalism was included as additional predictor. As can be seen in Table 5 c, fundamentalism made a strong and significant contribution to the prediction model, and, other than in the models predicting religiousness and fundamentalism, the time spent in prison did as well, with a negative coefficient.

Table 5a

Linear model of predictors of religiousness (n = 77)

Coefficients Model summary
b SE b β p R R2
Constant 1.42 .19 <.001 .59 .35
Time in prison .003 .01 .03 .75
Religious socialization .59 .09 .60 <.001
Table 5b

Linear model of predictors of fundamentalism (n = 65)

Coefficients Model summary
b SE b β p R R2
Constant 1.41 .19 <.001 .55 .31
Time in prison .005 .006 .08 .45
Religious socialization .03 .08 .04 .74
Current religiousness .35 .08 .53 <.001
Table 5c

Linear model of predictors of militancy (n = 64)

Coefficients Model summary
b SE b β p R R2
Constant -.35 .33 <.001 .63 .40
Time in prison -.02 .01 -.21 .04
Religious socialization .08 .10 .10 .86
Current religiousness -.02 .12 -.03 .85
Fundamentalism .77 .16 .58 <.001

Based on the theoretical process models, we assumed that we would be able to connect the variables in a mediation model. A mediation model links a causal antecedent variable with a consequent variable through an intermediary variable or moderator variable (see Hayes 2018, 77). It is also possible to include several moderator variables, which results in a model displaying the predictors’ effect transmitted through multiple mechanisms simultaneously as in a causal chain (Hayes 2018, 147). As the regression models suggested that militancy emerged from fundamentalism, that fundamentalism emerged from religiousness and that religiousness emerged from religious socialization, we arranged the variables as displayed in Figure 1. We had found that the time spent in the current facility was a significant predictor variable for militancy; hence, we included it as a covariate.

There was no direct effect of religious socialization on militancy; therefore, we detected a significant indirect effect that was mediated through current religiousness and fundamentalism. This represents a moderate effect, κ2 = .16, 95 % CI [.08, .31].

Figure 1 
Model of religious socialization as a mediated predictor of religious militancy

Figure 1

Model of religious socialization as a mediated predictor of religious militancy

5 Discussion

Our first research question regarded the presence of Islamist attitudes among prisoners, which we assessed in a survey among 87 Muslim inmates. To assess the fundamentalist (Salafist) mindset, we presented items that we had developed based on a thorough description of attitudes that describe worldviews of the contemporary neo-Salafist movement in Germany (Mansour 2016). Important characteristics are black-and-white thinking, a feeling of religious superiority, proselytizing, and a literal interpretation of religious scripture, to name just a few. Salafism comprises several currents with different stances towards political claims and the use of violence (Wiktorowicz 2006). Our measure of fundamentalism included anti-democratic attitudes, but no items measured a support for violent means. These were separated and combined in a measure for religious militancy, which assessed attitudes that supported the fight against non-believers and martyrdom[7].

Our research shows that, apparently, there are prisoners who display signs of Islamist attitudes. On a scale ranging from 0 to 4, Muslim inmates scored a mean of 2.13 on the fundamentalism scale, and 1.29 on the militancy scale. Although these means are not per se »extreme« in a statistical sense, our items represented relatively bold statements. Our results suggest that some Muslim prisoners clearly support Islamist ideas. Among one quarter of the inmates, the means of the scales measuring fundamentalist and militant attitudes were relatively high (scale centre or higher). The statements we presented included only one that directly asked for a readiness to participate in a fight for God. This was supported by a quarter of the Muslim inmates. Over a third of the Muslim inmates showed positive attitudes towards self-sacrifice, whereas attitudes in support of a fight against the »non-believers« were less frequent – but present in some inmates. These individuals represent prisoners, who have – theoretically – reached one of the final phases of the radicalization process, which has been described as a »(call) for violence against adversaries« (Sinai 2014, 39). According to this model, release might be followed by joining a terrorist organization or cell and plans to conduct an attack.

Based on this finding and theoretical assumptions, we then iteratively calculated regression models, beginning with a model predicting current religiousness, then fundamentalism, then militancy. Theoretically, the variables we used could be arranged in a plausible order. Religious socialization relates to characteristics of religious life in the original family, which are assumed to influence the degree to which religion is included in present life. This was our operationalization of religiousness. Fundamentalism, in turn, is a very literal, »back-to-the-roots« understanding of religion, and it is a framework of reference in which religion can be more than merely a private issue (Marty 1988), a guidance for social life and politics. Furthermore, religious fundamentalism does not necessarily include pro-violent attitudes. In the context of Salafism, Wiktorowicz (2006) distinguishes purists, who emphasize non-violent methods, from jihadists, who call for violence. Hence, positive attitudes towards religiously motivated violence represented our understanding of the »extreme« and the outcome we were most interested in.

Our first model was set up to predict religiousness from the time spent in prison and religious socialization; the second model additionally included religiousness to predict fundamentalism, and the last one fundamentalism to predict militancy. Interestingly, the religion-related variable that had made a significant contribution in a previous model did not turn out as a significant predictor in the next one. This suggested that the constructs »emerged« from one another, which we then tested in a mediation model. In this model, we arranged the variables according to our findings and plausible considerations. The strongest link in our model was the one between fundamentalism and militancy, which confirmed that the association between religion and violence is mostly rooted in fundamentalist movements of religions in general (Barkun 2003). In other words, accepting a fundamentalist understanding of Islam seems to be an important intermediate step for adopting a militant outlook and a proneness to use violence for political goals. A direct effect of religious socialization on militancy could not be identified. Instead, we identified a significant indirect effect, which was mediated through fundamentalism leading to attitudes that support militancy. Ten years ago, the influence of socialization factors was investigated in an extensive panel study of Muslims in Germany (Frindte 2011) and came to similar results: Traditional religiousness was associated with religious fundamentalism and anti-democratic attitudes. Also, their study showed that a relatively large number of German and non-German Muslims could be described as »strictly religious with strong dislike for the West, tendencies to accept violence and without any tendencies to integrate (into society)« (Frindte 2011, 431). These findings suggest that socialization experiences can be a vulnerability in Muslims – independently of the prison context.

Regarding imprisonment, our data suggest that the overall fundamentalist attitudes among Muslims do not vary with the time spent in prison. However, with more time in prison, militant attitudes seem to slightly decrease. Interviews with former extremists showed that a period of imprisonment often leads to disengagement from terrorism or extremism (Bubolz & Simi 2015; Ferguson 2016; Latif et al. 2020; Reinares 2011; Sikkens et al. 2017, Simi et al. 2019). However, disengagement – particularly involuntary disengagement after being physically cut off an extremist group because of imprisonment – does not necessarily mean deradicalization (Horgan 2008). Yet some studies reported details on how imprisonment can contribute to a change of mind (Ferguson 2016; Latif et al. 2020), which could also mean that militancy becomes less attractive. Nevertheless, our study cannot provide evidence for individual developments, but for the overall group of Muslims who completed our survey. We simply don’t know about the individuals’ mindsets before they went to prison, and, most importantly, the individuals we surveyed were not a sample of serious extremists who participated in violent activities.

To tell whether prisons are »breeding grounds« for radicalization, the data should be compared to those of a control group outside prison. To draw conclusions on the effect of imprisonment, a longitudinal study should be conducted. To enhance our knowledge regarding the exact factors and processes that contribute to radicalization or de-radicalization in prison, research needs to be able to explore the specific individual cases and pathways, which can be done best in interview studies. The factors that have been suggested to pose a certain risk – for example, the amount of Muslim or terrorist prisoners (Trujillo et al. 2009), could be hardly controlled.

In sum, there are limitations of our research with regard to the study design (cross-sectional, no comparison group outside prison) and the limitations of questionnaire surveys and self-report, for example, self-selection. Only male prisoners were included in our evaluation; however, female juveniles are also susceptible to Islamist radicalization (Goede 2019; Goede et al. 2020). Specific vulnerabilities and connecting points for ideological content should be researched in depth in order to elaborate differential radicalization processes. It is also important to note that we did not control for the general propensity to violence in the survey. This leaves the question unanswered whether the willingness to fight actually results from religious attitudes or whether existing violent propensities are legitimized by a religious interpretive framework. A similar direction of causality has been suggested by Basra and Neumann (2017), who show how terrorist organizations suggest that crime is a »form of worship« (p. 1). Also, there were no controls for perceived religious discrimination, which has been found to enhance understanding of the associations between religion and delinquency (Beller et al. 2019) and might also advance knowledge of radicalization processes in contexts with Muslim minorities. We only investigated prisoners in Bavarian prisons. The present analyses only include prisoners with sufficient German language skills who do not necessarily represent the overall Muslim population in prisons. In particular, this concerns prisoners who had only recently immigrated from Middle Eastern and Northern African counties and had not yet gained the language skills required to participate in the survey. Further research should be conducted in more prisons in Germany, with additional data from inmates who do not yet have sufficient German language skills. Most importantly, to assess changes in radical attitudes, a longitudinal design is required.

6 Conclusions

Overall, our study has been able to capture potentials of Islamist radicalization in the context of prisons. Prisons concentrate some risk factors for radicalization in general, starting with severe psychological distress and self-esteem issues in some inmates, as well as shifts in the social environment in favour of in-prison subcultures. However, the additional challenges of integration and the fact that negative experiences could be attributed to minority status makes the Muslim prison population particularly vulnerable. Moreover, the problem of Islamist radicalization is still far from settled – especially when it is realized that quite a few returned fighters for IS could still be allocated to the prison system and might there exert their influence as experienced »fighters« on impressionable Muslims.

The measures we used were reliable operationalizations of fundamentalist and militant attitudes. Although, overall, the political climate of opinion among Muslim prisoners can apparently not be characterized as very »extreme«, there were some cases that raise concerns. Whether these individuals would actually commit an attack, or if they intended to provoke, is of course uncertain, but past attacks and the fact that the perpetrators had been imprisoned beforehand give rise to concerns. Individuals who display negative attitudes against the »nonbelievers« should be observed and prevented from influencing other inmates, and – at best – be brought to participate in deradicalization programmes. Some Muslim prisoners seem to find the concept of martyrdom quite attractive. Even if they do not support the idea of a fight against non-believers, this gives reason for concern as martyrdom is an essential aspect of jihadist ideology. German prisons like the ones in Bavaria, for example, manage extremist offenders and suspected cases mainly by observation by specialized staff members, and cooperate with providers of deradicalization programmes.

Although we cannot draw conclusions regarding psychological processes that lead to radicalization, our mediation analyses can give us a hint on how militant attitudes emerge from early socialization influences in the Muslim prison population. Our results indicate that religious factors do play a significant role in the development of attitudes that support violence justified by Islamist ideology. There is no evidence yet regarding the effectiveness of religious or general spiritual guidance as a preventive means. Furthermore, the implementation of such programs requires considering many factors, as, for instance, security issues or resources. Yet there are many reasons to argue for religious services in prisons, such as their positive effects on well-being or beneficial social resources. For the religious inmates, a religious authority figure can contribute to keep religiousness on a moderate path. Additionally, preventive measures should focus on the »initial spark« of religious extremism, which is rooted in early socialization experiences.


This project was funded by the German Research Foundation(DFG; file number: STE923/10).

7 References

Ano, G.G. & Vasconcelles, E.B. (2005). Religious coping and psychological adjustment to stress: A meta-analysis. Journal of Clinical Psychology 61/4, 461–480. Search in Google Scholar

Armborst, A. (2014). Radicalisation and de-radicalisation of social movements: The comeback of political Islam? Crime, Law and Social Change 62/3, 235–255. Search in Google Scholar

Backes, U. (2007). Meaning and forms of political extremism in past and present. Středoevropské politické studie 9/4, 242–262. Search in Google Scholar

Bandura, A. (1969). Social-learning theory of identificatory processes, in: D.A. Goslin (ed.), Rand McNally sociology series. Handbook of Socialization Theory and Research. Chicago, 213–262 Search in Google Scholar

Barkun, M. (2003). Religious violence and the myth of fundamentalism. Totalitarian Movements and Political Religions 4/3, 55–70. Search in Google Scholar

Basra, R., Neumann, P.R. & Brunner, C. (2016). Criminal pasts, terrorist futures: european jihadists and the new crime-terror nexus. London. Search in Google Scholar

Basra, R., & Neumann, P. R. (2017). Crime as Jihad. CTC Sentinel, 10/9, 1–5. Search in Google Scholar

Beller, J., & Kröger, C. (2018). Religiosity, religious fundamentalism, and perceived threat as predictors of Muslim support for extremist violence. Psychology of Religion and Spirituality, 10/4, 345–355. Search in Google Scholar

Beller, J., Kröger, C. & Hosser, D. (2019). Prädiktoren der Delinquenz bei adoleszenten Muslimen. Forensische Psychiatrie, Psychologie, Kriminologie, 13/2, 188–198. Search in Google Scholar

Braam, A. W., & Koenig, H. G. (2019). Religion, spirituality and depression in prospective studies: A systematic review. Journal of Affective Disorders, 257, 428–438. Search in Google Scholar

Bubolz, B.F. & Simi, P. (2015). Leaving the world of hate: Life-course transitions and self-change. American Behavioral Scientist 59/12, 1588–1608. Search in Google Scholar

Cohen S. (1972). Folk devils and moral panic: The creation of the mod and rockers. Oxford. Search in Google Scholar

Cornwall, M. (1988). The influence of three agents of religious socialization: Family, church, and peers, in: D.L. Thomas (ed.), Religious studies center specialized monograph series, Vol. 3. The religion & family connection: Social science perspectives. Provo, 207–231. Search in Google Scholar

Cronbach, L.J. (1946). Response sets and test validity. Educational and Psychological Measurement 6/4, 475–494. Search in Google Scholar

Dantschke, C. (2017). Attraktivität, Anziehungskraft und Akteure des politischen und militanten Salafismus in Deutschland [Attraction, gravitational force and agents of political and militant Salafism in Germany], in: A. Toprak & G. Weitzel (eds.), Salafismus in Deutschland: Jugendkulturelle Aspekte, pädagogische Perspektiven [Salafism in Germany: Aspects of Youth Culture, Pedagogic Perspectives]. Wiesbaden, 61–76. Search in Google Scholar

De Jong, T. & Ferguson-Hessler, M.G. (1996). Types and qualities of knowledge. Educational Psychologist 31/2, 105–113. Search in Google Scholar

Dugas, M., Bélanger, J.J., Moyano, M., Schumpe, B.M., Kruglanski, A.W., Gelfand, M.J., . . . Nociti, N. (2016). The quest for significance motivates self-sacrifice. Motivation Science 2/1, 15–32. Search in Google Scholar

Endres, J. & King, S. (2018a). Radikalisierung und Deradikalisierung im Justizvollzug [Radicalization and de-radicalization in the prison context], in: M. Walsh, B. Pniewski, M. Kober & A. Armborst (eds.), Evidenzorientierte Kriminalprävention in Deutschland [Evidence-Oriented Crime Prevention in Germany]. Wiesbaden, 511–530. Search in Google Scholar

Endres, J. & King, S. (2018b). Welche Rolle spielt religiöses Wissen für islamistische Radikalisierung? [What is the role of religious knowledge in the context of islamist radicalization?], in: K. Boers & M. Schaerff (eds.). Kriminologische Welt in Bewegung [The world of criminology in motion]. Mönchengladbach, 218–227. Search in Google Scholar

Ferguson, N. (2016). Disengaging from terrorism: A Northern Irish experience. Journal for Deradicalization, 6, 1–23. Search in Google Scholar

Field, A. (2013). Discovering statistics using IBM SPSS statistics: And sex and drugs and rock ›n‹ roll. 4th ed. Los Angeles. Search in Google Scholar

Frindte, W., Boehnke, K., Kreikenbom, H. & Wagner, W. (2011). Lebenswelten junger Muslime in Deutschland: Ein sozial- und medienwissenschaftliches System zur Analyse, Bewertung und Prävention islamistischer Radikalisierungsprozesse junger Menschen in Deutschland. [Life realitites of young Muslims in Germany: An analytical framework based on social and media science to assess and prevent islamis radicalization processes in Germany]. Berlin. Search in Google Scholar

Goede, L.-R. (2019). Islamistische Einstellungen und Verhaltensweisen unter muslimischen Jugendlichen [Islamist attitudes and behaviors among Muslim youth.] Zeitschrift für Jugendkriminalrecht und Jugendhilfe [Journal for Juvenile Criminal Law and Juvenile Welfare] 31/3, 204-213 Search in Google Scholar

Goede, L.-R., Schröder, C. P. & Lehmann, L. (2020). Perspektiven von Jugendlichen. Ergebnisse einer Befragung zu den Themen Politik, Religion und Gemeinschaft im Rahmen des Projektes »Radikalisierung im digitalen Zeitalter (RadigZ)«. KFN-Forschungsberichte Nr. 151 [Perspectives of young people. Results of a survey on politics, religion and community within the project »Radicalization in the digital age (RadigZ)«. KFN Research Reports No. 151]. Hannover: Kriminologisches Forschungsinstitut Niedersachsen. Search in Google Scholar

Hamm, M.S. (2009). Prison islam in the age of sacred terror. British Journal of Criminology 49/5, 667–685. Search in Google Scholar

Hayes, A.F. (2018). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. 2nd ed. New York. Search in Google Scholar

Horgan, J. (2008). Deradicalization or disengagement? A process in need of clarity and a counterterrorism initiative in need of evaluation. Perspectives on Terrorism 2/4, 3–8. Search in Google Scholar

Hosser, D. (2008). Prisonisierungseffekte. [Effects of prisonization], in: R. Volbert & M. Steller (eds.), Handbuch der Rechtspsychologie [Handbook of Legal Psychology]. Göttingen, 172–179. Search in Google Scholar

Johnson, B. R., Li, S. de, Larson, D. B. & McCullough, M. (2000). A systematic review of the religiosity and delinquency literature. Journal of Contemporary Criminal Justice, 16/1, 32–52. Search in Google Scholar

Jones, C. R. (2014). Are prisons really schools for terrorism? Challenging the rhetoric on prison radicalization. Punishment and Society 16, 74–103. Search in Google Scholar

Jugl, I., Lösel, F., Bender, D. & King, S. (2020). Psychosocial prevention programs against radicalization and extremism: a meta-analysis of outcome evaluations. European Journal of Psychology applied to Legal Context 13/1, 37–46. Search in Google Scholar

Karimi, Y., Cimbura, A. & Loza, W. (2019). Assessing the prevalence of middle eastern extreme ideologies among some Iranians. Journal of Strategic Security 12/3, 157–169. Search in Google Scholar

Kruglanski, A.W., Bélanger, J.J., Gelfand, M., Gunaratna, R., Hettiarachchi, M., Reinares, F., . . . Sharvit, K. (2013). Terrorism – a (self) love story: Redirecting the significance quest can end violence. The American Psychologist 68/7, 559–575 Search in Google Scholar

Kruglanski, A.W., Chen, X., Dechesne, M., Fishman, S. & Orehek, E. (2009). Fully committed: Suicide bombers‹ motivation and the quest for personal significance. Political Psychology 30/3, 331–357. Search in Google Scholar

Kruglanski, A.W., Gelfand, M,J., Bélanger, J.J., Sheveland, A., Hetiarachchi, M. & Gunaratna, R. (2014). The psychology of radicalization and deradicalization: How significance quest impacts violent extremism. Political Psychology 35, 69–93. Search in Google Scholar

Kruglanski, A.W., Gelfand, M.J., Sheveland, A., Babush, M., Hetiarachchi, M., Ng Bonto, M. & Gunaratna, R. (2016). What a difference two years make: Patterns of radicalization in a Philippine jail. Dynamics of Asymmetric Conflict 9/1–3, 13–36. Search in Google Scholar

Kruglanski, A.W. & Orehek, E. (2011). The role of the quest for personal significance in motivating terrorism, in: J.P. Forgas, A.W. Kruglanski, and K.D. Williams (eds). The Psychology of Social Conflict and Aggression. New York, 153–164. Search in Google Scholar

Kruglanski, A.W. & Webber, D. (2014). The Psychology of radicalization. Zeitschrift für Internationale Strafrechtsdogmatik [Journal for International Criminal Law Doctrine] 9, 379–388. Search in Google Scholar

Kühle, L. (2012). In the faith of our fathers? Religious minority socialization in pluralistic societies. Nordic Journal of Religion and Society 25/2, 113–130. Search in Google Scholar

Latif, M., Blee, K., DeMichele, M., Simi, P. & Alexander, S. (2020). Why white supremacist women become disillusioned, and why they leave. The Sociological Quarterly 61/3, 367–388. Search in Google Scholar

Lienert, G.A. & Raatz, U. (1998). Testaufbau und Testanalyse [Test Construction and Test Analyssis]. 6. Aufl. [6th ed.] Weinheim. Search in Google Scholar

Lösel, F., Bender, D., Jugl, I. & King, S. (2020). Resilience against political and religious extremism, radicalization, and related violence: A systematic review of studies on protective factors, in: D. Weisburd, E.U. Savona, B. Hasisi & F. Calderoni (eds.), Understanding Recruitment to Organized Crime and Terrorism. Cham, 55–84. Search in Google Scholar

Loza, W. (2007). The Assessment and Treatment of Radicalization Scale (ATRS): A measure of Middle-Eastern extremism. Unpublished manuscript. Search in Google Scholar

Mansour, A. (2016). Generation Allah. Warum wir im Kampf gegen religiösen Extremismus umdenken müssen. [Generation Allah: Why the fight against religious fundamentalism requires us to re-think]. Frankfurt am Main. Search in Google Scholar

Marty, M. E. (1988). Fundamentalism as a social phenomenon. Bulletin of the American Academy of Arts and Sciences 42/2, 15–29. Search in Google Scholar

Maruna, S., Wilson, L. & Curran, K. (2006). Why God Is often found behind bars: Prison conversions and the crisis of self-narrative. Research in Human Development 3/2–3, 161–184. Search in Google Scholar

McCauley, C. & Moskalenko, S. (2017). Understanding political radicalization: The two-pyramids model. American Psychologist 72/3, 205–216. Search in Google Scholar

Moghaddam, F.M. (2005). The staircase to terrorism: A psychological exploration. American Psychologist 60/2, 161–169. Search in Google Scholar

Mulcahy, E., Merrington, S. & Bell, P.J. (2013). The radicalisation of prison inmates: A review of the literature on recruitment, religion and prisoner vulnerability. Journal of Human Security 9/1, 4–14­. Search in Google Scholar

Neumann, P. (2010). Prisons and terrorism. Radicalisation and de-radicalisation in 15 countries. London. Search in Google Scholar

Norko, M. A., Freeman, D., Phillips, J., Hunter, W., Lewis, R., & Viswanathan, R. (2017). Can religion protect against suicide?. The Journal of Nervous and Mental Disease, 205/1, 9–14. Search in Google Scholar

Oliver, P. & Johnston, H. (2000). What a good idea! Ideologies and frames in social movement research. Mobilization: An International Quarterly 5/1, 37–54. Search in Google Scholar

Ostroff, C. & Kozlowski, S.W. (1992). Organizational socialization as a learning process: The role of information acquisition. Personnel Psychology 45/4, 849–874. Search in Google Scholar

Rappaport, A.J., Veldhuis, T. & Guiora, A.N. (2013). Homeland security and the inmate population: The risk and reality of Islamic radicalization in prison, in: L. Gideon (ed.), Special Needs Ofenders in Correctional Institutions. Thousand Oaks, 431–458. Search in Google Scholar

Reinares, F. (2011). Exit from terrorism: A qualitative empirical study on disengagement and deradicalization among members of ETA. Terrorism and Political Violence 23/5, 780–803. Search in Google Scholar

Ribeaud, D. & Eisner, M. (2010). Are moral disengagement, neutralization techniques, and self-serving cognitive distortions the same? Developing a unified scale of moral neutralization of aggression. International Journal of Conflict and Violence 4/2, 298–315. Search in Google Scholar

Rocheleau, A. M. (2013). An empirical exploration of the »pains of imprisonment« and the level of prison misconduct and violence. Criminal Justice Review 38/3, 354–374. Search in Google Scholar

Ronco, D., Sbraccia, A. & Torrente, G. (2019). Prison de-radicalization strategies, programmes and risk assessment tools in Europe. Rome. Search in Google Scholar

Saucier, G., Akers, L.G., Shen-Miller, S., Kneževié, G. & Stankov, L. (2009). Patterns of thinking in militant extremism. Perspectives on Psychological Science: A Journal of the Association for Psychological Science 4/3, 256–271. Search in Google Scholar

Schuurman, B. & Eijkman, Q. (2013). Moving terrorism research forward: The crucial role of primary sources. Terrorism and Counter-Terrorism Studies. The Hague. Search in Google Scholar

Sherkat, D. E. (2003). Religious socialization, in: M. Dillon (ed.), Handbook of the Sociology of Religion. Cambridge, 151–163. Search in Google Scholar

Sikkens, E., van San, M., Sieckelinck, S., & de Winter, M. (2017). Parental influence on radicalization and de-radicalization according to the lived experiences of former extremists and their families. Journal for Deradicalization 12, 192–226. Search in Google Scholar

Silber, M.D. & Bhatt, A. (2007). Radicalization in the West: The homegrown threat. New York. Search in Google Scholar

Silke, A. (2001). The devil you know: Continuing problems with research on terrorism. Terrorism and Political Violence 13/4, 1–14. Search in Google Scholar

Silke, A. & Veldhuis, T. (2017). Countering violent extremism in prisons: A review of key recent research and critical research gaps. Perspectives on Terrorism 11, 2–11. Search in Google Scholar

Simi, P., Windisch, S., Harris, D. & Ligon, G. (2019). Anger from within: The role of emotions in disengagement from violent extremism. Journal of Qualitative Criminal Justice & Criminology 7/2. Search in Google Scholar

Sinai, J. (2014). Developing a model of prison radicalization, in: A. Silke (ed.), Prisons, terrorism and extremism: Critical issues in management, radicalisation and reform. London, 35–46. Search in Google Scholar

Spalek, B. & El-Hasan, S. (2007). Muslim converts in prison. The Howard Journal of Criminal Justice 46/2, 99–114. Search in Google Scholar

Sykes, G.M. (1958). The society of captives. Princeton. Search in Google Scholar

Thomas, J. & Zaitzow, B.H. (2006). Conning or conversion? The role of religion in prison coping. The Prison Journal 86/2, 242–259. Search in Google Scholar

Trujillo, H.M., Jordán, J., Antonio Gutiérrez, J. & González-Cabrera, J. (2009). Radicalization in prisons? Field research in 25 Spanish prisons. Terrorism and Political Violence 21/4, 558–579. Search in Google Scholar

Veldhuis, T. & Staun, J. (2009). Islamist radicalisation: A root cause model. The Hague. Search in Google Scholar

Webber, D. & Kruglanski, A.W. (2017). Psychological factors in radicalization: A »3N« approach, in: G. LaFree & J.D. Freilich (eds.), The Handbook of the Criminology of Terrorism. Vol. 14. Hoboken, 33–46. Search in Google Scholar

Wiktorowicz, Q. (2006). Anatomy of the Salafi movement. Studies in Conflict and Terrorism 29/3, 207–239. Search in Google Scholar

Published Online: 2021-08-18
Published in Print: 2021-09-14

© 2021 Walter de Gruyter GmbH, Berlin/Boston