Introduction

The United Kingdom public was shocked by the story of the “Bethnal Green Trio”, three teenage girls who went to Iraq in order to become wives of Islamic State fighters. Of the three, two are now presumed dead; the last one – Shamima Begum – currently resides in a Syrian refugee camp after being stripped of her citizenship by the UK government (Philips 2024). Some accuse Shamima Begum of actively assisting ISIS in a role of a combatant through activities like constructing suicide belts. Others argue that Shamima Begum had no such capabilities as she was not even permitted to leave her house by her husband (Baker et al. 2023). The story of Shamima Begum is just one among many. In fact, almost five thousand women from across the world travelled to Iraq and Syria to join ISIS, with an additional four and a half thousand minors comprising the ranks of foreigners among ISIS (Khomami 2018). Of the eight hundred and fifty British citizens among ISIS members, almost two hundred were women and children, yet of the four hundred and twenty-five repatriated to the UK, only six were women and children (ibid). The many young girls and women lured in through ISIS propaganda came from different backgrounds. In fact, some of them were not raised in Muslim households, only converting after radicalization (The Week 2018). Additionally, punishments faced upon returning back to the UK for these women varied. While Shamima Begum was stripped of her nationality, Linda Wenzel’s sentence was reconsidered from execution to prison-time (Ibid). Others were repatriated to their countries, such as in the case of Emilie Koenig who was brought back to France (Al-Monitor 2022) What’s more, few to none have considered as to why some female collaborators are judged as perpetrators, whilst others are ‘forgiven’ for their actions.

Extant research on terrorist or armed groups has largely focused on combatants, with experimental work largely focusing on reintegrating direct (often male) combatants (Agneman and Strömbom 2023; Godefroidt and Langer 2023). Conflict scholars have also explored the motivating factors that push female combatants to join armed groups (Thomas 2021; Wood and Thomas 2017; Thomas and Bond 2015; Giri and Haer 2024). In addition to different types of combatants, recent scholarly articles shed light on collaborators and those who fall into “gray” areas, acknowledging the fact that post-conflict reconstruction often focuses on ex-fighters whilst those in more ancillary roles remain understudied (Kao and Revkin 2023). Conflict scholars have also analyzed the role that child soldiers play in armed rebel and terrorist groups, given that their unique circumstances set them apart from adult combatants (Nyamutata 2020; Capone 2017; Mironova and Whitt 2024). However, there is are notable gaps in research specifically examining the experiences and reintegration challenges of ISIS brides - women and girls who travelled ISIS-controlled territories to marry fighters. These ‘brides’ occupy a complex position that defies binary categorization as combatant or victim, leading to their circumstances being overlooked. Their diverse backgrounds, limited roles within ISIS and legal challenges upon return, set them apart from other ISIS-affiliates. Looking at this issue broadly, I bring attention to public opinion on people which cannot be categorized within the victim-perpetrator binary. Secondly, I focus on ‘ISIS Brides’ in particular, a community that is relatively neglected in political conversations surrounding post-conflict transitional justice. I also explore the interaction between ethnicity, religion and collaborator status, bringing in theoretical explanations of why certain ‘brides’ are favoured for repatriation over others. Finally, v=by focusing on ‘ISIS brides’, I aim to fill critical gaps in the understanding of deradicalization and reintegration processes. The findings have important implications for developing effective rehabilitation programs for non-combatants, a group that remains discounted in discussions on post-conflict reconstruction and reconciliation.

Public opinion against Shamima Begum at least partially motivated the revocation of her citizenship. Home Secretary of the time, Sajid Javed, was advised that hostility against Shamima Begum and other ‘ISIS Brides’ was high enough that cancelling their citizenship would not impact community relations (Sabbagh and Bowcott 2020). The UK government’s factoring in of public opinion into decision-making on the topic makes it a worthwhile avenue of research to pursue. I argue that the public, based on limited public information they receive about these ISIS brides, make judgements that affect how sympathetic they are to these returnees. Firstly, I argue that returnees that are viewed as more vulnerable to coercion are more likely to be ‘forgiven’ than those who are not. I expect returnees that are younger and those who have said to have fled ISIS to be perceived as vulnerable, and more ‘deserving’ of public sympathy. Secondly, I propose that cultural discrimination and Islamophobia play a great role in determining whether the UK public wants to repatriate and rehabilitate a particular returnee. I expect that Muslim and non-white returnees will be treated more harshly as their place in UK society is already questioned by those on the political right-wing.

To test my theoretical expectations, I conduct a novel forced-choice and ratings-based conjoint experiment, in which I present respondents with hypothetical randomized profiles of ISIS returnees. The profiles combine visual and textual information, providing respondents with four attributes of returnees: age, ethnicity, , perceived religiosity and the manner in which they have departed from ISIS (voluntary or non-voluntary). I argue that the woman’s age upon joining ISIS, and her means of exit are a proxy for deciding whether she was agentic, or a victim of coercion. I employ the variables Ethnicity and perceived religiosity to test whether UK respondents are indeed less forgiving of non-white and visibly Muslim women due to their perceived incompatibilities with UK society. Such experimental designs improve external validity of the study as the vignettes reflect the way the stories of ‘ISIS brides’ are portrayed in the media.

Literature Review

– I first overview the literature on combatants and collaborators, I then examine current research on minor and female combatants, Finally, I discuss current findings on ‘ISIS Brides’ and then discuss gaps in the literature.

Combatants and Collaborators - What Do We Know?

While ‘ISIS Brides’ themselves are understudied, surrounding literature on combatants and collaborators establish an individual’s perceived volition as the main determining factor of their rehabilitative prospects. Previous experimental research on attitudes towards ex-combatants, finds that respondents do not discriminate between ex-fighters of different descents. However, any ethnic “penalty” may be neutralized by assumptions of minority fighters (e.g., Afro-Colombian in Colombia) having less agency (Agneman and Strömbom 2023, 19). In addition, civilian attitudes towards the reintegration of ex-combatants is driven by both risk assessment and fairness perceptions. Specifically, the public is more willing to forgive ex-combatants who were forced into fighting as opposed to those who joined of their own volition (Godefroidt and Langer 2023). Kao and Revkin (2023) also researched public perceptions towards ISIS collaborators in Mosul, Iraq finding public opinion to be shaped by assumptions of volition and responsibility, where people assign more responsibility and punishment to those showing greater autonomy. Secondly, respondent exposure to violence motivates them to prefer revenge, which affects the punishment they select for ex-collaborators. These studies establish that perceived combatant volition plays a key role in determining ‘forgiveness’, with the respondent’s ‘want’ for revenge, and ethnic discrimination being secondary factors.

Child and Female Combatants

In the literature on female combatants, scholars have studied both the “demand” and “supply” side of why women fight. On the “demand” side, scholars focus on organizational characteristics that incentivize female combatants to join. Among other things, some terrorist organizations view female combatants as presenting a tactical advantage (Thomas and Bond 2015). Particularly, perceptions of how women pose a ‘lesser threat’ are taken advantage of by terrorist groups, which is partly why Conservative terrorist groups recruit women despite their ideological reservations with female mobility (Thomas 2021). Additionally, female combatants also have lower desertion rates due to greater social repercussions, and they also serve as a bridge between civilian populations and combatants, which is essential for a group’s longevity (Giri and Haer 2024). Notably, female combatant presence is negatively correlated with Islamist ideology, and group religiosity (Wood and Thomas 2017). On the “supply” side, Weber (2021) applies the ‘Complex Political Perpetrators’ framework, arguing that female ex-combatants were both victims of repression, which often motivated them to join armed-groups in the first place, and perpetrators themselves. Thus, scholarly work on female fighters establishes their complexity, revealing group characteristics that facilitate their participation, strategic advantages they pose as combatants, and their dual nature as being both on the receiving and inflicting end of violence.

ISIS affiliated children, among whom many were female “ISIS brides”, fall under a legal gray area (Nyamutata 2020). While ostensibly fulfilling the child soldiering criteria, such as forcible recruitment, combative and non-combative roles, and designation as sex slaves, they also predominantly do not meet the age requirement to be considered child soldiers (which is fifteen and under). Thus, they can neither be considered fully victimized as child soldiers, nor given perpetrator status, as International Criminal Law bars the prosecution of minors under eighteen. Despite this, these children are viewed more sympathetically than adult ISIS affiliates. In Iraq, locals were more lenient towards minor ISIS affiliates than their adult counterparts, irrespective of whether they were combatants (Mironova and Whitt 2024). What sets apart foreign children in ISIS from traditional understandings of child soldiering are the circumstances under which they join. In most armed conflicts, child soldiers are inducted into groups through coercion, under the duress of severe poverty. These circumstances differ from the foreign girls that joined ISIS, who were raised in more stable environments (Capone 2017, 177). Thus, to these groups, coercion is a less convincing explanation than for other child soldiers. Shamima Begum, fifteen years old at the time of her departure from the UK, was a schoolgirl from a stable household (Cheng 2021). This is a far cry from the child soldiers that are studied in conflict literature, which lends to the idea that girls who became ‘ISIS brides’ are seen as more agentic, and by extension, more deserving of punishment.

The “ISIS Bride” Phenomenon

Western women who joined ISIS in large numbers played a variety of roles in the organization. While ISIS encouraged women to provide support by being good mothers and wives, many found other non-combative roles through which they could provide support (Pokalova 2019, 182). Many foreign women joined the Al Khansa Brigade, ISIS’s morality enforcement wing, where they enforced societal restrictions using punishments like beatings and lashings (Pokalova 2019, 184). Other women disseminated propaganda through social media. They documented their lives, provided advice on how to ‘immigrate’ to ISIS territory, and taught girls how to be wives for militant husbands (Saripi 2015, 27)., Though, there is little evidence to suggest that British women were taking part in combat despite the accusations (Whitehead 2016). Although at least something is known about the activities of ISIS brides, it is often difficult to profile these women, as variations in age, ethnicity, and education level made it difficult to ‘profile’ vulnerable individuals (Saltman and Smith 2015). According to one female guesthouse registry in ISIS territory, women staying there ranged aged eleven to seventy-six, with girls under twenty one constituting 20% of the sample (Milton and Dodwell 2018). Only seven percent of the women came from Western countries, yet this may not be an accurate representation of ‘ISIS brides’ given that the guesthouse seemed to cater specifically to Russian and Central-Asian women (Ibid).

There is a lack of consensus in the current literature on ‘ISIS Brides’ on how they are and should be perceived. Jackson (2022) explains how the term “grooming” has become popular with policymakers, academics, as well as the public to explain how online interactions between girls in Western and European countries result in radicalization. She finds that the language used in labeling these interactions is sexually charged, with implications of enticement, in order to solidify the explanation of grooming by likening it to CSE (Child Sexual Exploitation). On the other hand, the author disagrees with this general assessment, arguing that it strips women of their autonomy. She argues that instances of grooming were not well defined and how some of these women participated in propaganda dissemination and ISIS morality policing (Ibid). Additionally, Martini’s (2018) found the act of joining ISIS challenged Western perceptions of Muslim women as lacking agency, going against the perception of Muslim women as oppressed victims of the men in their communities. Jacoby (2015) places perceptions of “ISIS brides” into three main categories: as the victim, warrior, or the feminist - with the last focusing on the agency shown by these women. ISIS also deliberately disseminated propaganda on discrimination and hate crimes against Muslim women, arguing that they would never be truly accepted unless they underwent hijrah (or migration) to ISIS territory, providing incentives to women who felt alienated by Western expectations of dressing, socialization and ‘egalitarian feminism’ (Bjørgum 2016). This shows that the work surrounding ‘ISIS Brides’ has not definitively established the motivations for these women, nor how dangerous their presence in Western society is.

Gaps in the Literature

Despite the research done on ‘ISIS Brides’ in recent years, current literature could be expanded on in key ways. Firstly, there has been no experimental work on female collaborators. Secondly, many of the related experimental studies (Godefroidt and Langer 2023; Bauer, Fiala, and Levely 2018; Agneman and Strömbom 2023, Kao and Revkin 2023) were conducted in the terrorist group’s country of origin, giving the respondents a uniquely personal relationship (where they might have been directly exposed to terrorist activities) with the group that a Western respondent is unlikely to have. For example, Kao and Revkin’s (2023) findings may differ from my own, given that Iraqi and UK citizens have a different level of exposure to ISIS violence. By conducting this survey in the UK, I extend theories of post-conflict reintegration beyond war-torn societies, an area of study that remains causally under-explored. By virtue of being in the UK, it is likely that all the survey respondents are un-exposed to ISIS violence or administration, which may produce different findings than studies conducted in post-conflict environments like Iraq or Syria. Additionally, the survey location creates a focus on mechanisms of volition and ethnic discrimination against non-white communities. Finally, I offer a uniquely gendered lens of analysis that other experimental papers have not presented. The aforementioned papers focus on ex-combatants and collaborators irrespective of their gender, whereas I focus on a unique subset of female ISIS collaborators who have unique motivations and circumstances for joining ISIS.

The previous literature on ex-combatants and collaborators shows the role that ethnicity plays in reintegration prospects is mixed. Given Agneman and Strömbom’s (2023) discussion of ethnic “penalties” alongside assumptions on people’s agency, there is mixed evidence on how ethnic minorities would be treated by the public. My study seeks to incorporate research on biases towards young girls from ethnic minorities, factoring their ethnic origin into their repatriation prospects. Finally, scholarly work on ‘ISIS bride’ returnees is methodologically qualitative (Saripi 2015; Cook and Vale 2018; Martini 2018; Jackson 2022), and has not analyzed the role of ethnicity in reintegration. Thus, I propose a novel conjoint experiment on differing ISIS bride profiles in order to gauge public support for rehabilitative policy implementation.

Theory and Hypotheses

There are three key mechanisms at play that, I argue, affect a respondent’s sympathy towards returnees. The first mechanism is the reduction of sympathy that occurs as a result of viewing potential returnees as more responsible for their actions. The second mechanism is cultural discrimination, which posits that non-white communities are already seen as unable to integrate well into UK society — which gives the implication that they are more likely to ‘relapse’ into, or hide their radicalization. The final mechanism is the interplay between in-group vs. out-group attitudes. I argue that in-group members show favoritism and leniency towards each other — making them more sympathetic to returnees from their communities.

Volition

I argue that a respondent’s judgment of whether someone is responsible for their actions as a female ISIS affiliate is what impacts how positively or negatively they perceive the woman. It is reasonable to argue that, in the eyes of the public, if the girl joined ISIS of her own choice, her repatriation poses a threat to national security, and if she was groomed, she is deserving of reintegration into society. Thus, people’s perceptions on the degree of autonomy and responsibility exhibited by an ISIS bride is what should determine their opinions on her potential rehabilitation. I mainly aim to test this via a returnee’s age at which she joined ISIS, and how she exited the organization. The younger the girl is, the more likely that the “grooming” explanation is more convincing for the public. A fifteen year old girl is legally recognized as not just below the age of voting and drinking, but also under the age of consent for sexual activities. Thus, the public is likely to view a teenage girl joining ISIS as a victim, whereas the adult woman is more likely to be viewed as someone who made a conscious decision to join a terrorist organization. These ‘brides’, e.g., those interviewed and rehabilitated in Kyrgyzstan, spent approximately a decade under ISIS and subsequent detention camps (Kulenbekova and Wittenberg 2024). Therefore, upon returning, all would most likely be well into adulthood. Thus, their age upon return would be less salient a variable than their age upon leaving the country. The victim who was groomed would, logically, be viewed as less likely to ‘reoffend’ or continue acting upon radical beliefs. The person with more agency, would be viewed as more likely to ‘relapse’ or not fully deradicalize, making her seem a threat to society. For instance, when the time came to prosecute ISIS affiliates after the collapse of the caliphate, Iraqi locals clearly differentiated between child and adult combatants, recognizing that the former had limited understandings of consequence and reasoning-based judgements (Mironova and Whitt 2024). Although some find that age is not a consequential factor when it comes to attitudes towards ex-collaborators (Kao and Revkin 2021), I argue that age is still a characteristic that is worthwhile to test, especially given the differing context within which I conduct my experiment. Thus I propose that:

H1 (volition and age): UK respondents will be more supportive of rehabilitative measures for younger returnees.

Additionally, I argue that the means of exit from ISIS should give respondents cues on the returnee’s degree of volition and participation in the organization. Fleeing of one’s own agency implies a level of dissatisfaction with the organization that may be due to involuntary participation. While regret is possible amongst returnees who joined of their own will, I argue that respondents interpret the organizational exit attribute as a proxy for voluntary participation, given a lack of further information on the returnee’s motivations and actions. Measuring the degree of a person’s volition, being captured implies a level of support that compelled a returnee to stay within ISIS until they physically could not do so anymore. Thus, I argue that respondents would be more sympathetic of women that fled as opposed to those that were captured. The corresponding attribute for this hypothesis, “Exit from ISIS”, is drawn from El Masri et al’s (2024) conjoint experiment on ISIS collaborators.

H2 (volition and exit): UK respondents will be more supportive of punitive measures for people that were captured, as opposed to those who fled.

Another way volition plays a role is through the respondent’s judgment of what the “bride” did once she was in this territory. Due to the secrecy surrounding women in ISIS in general, it is quite difficult to ascertain the exact role these women played as ISIS “brides”, whether they remained solely as wives and mothers, or participated in ancillary activities like the morality enforcement through the Al-Khansa Brigade. Therefore, I argue that the public will, upon viewing a particular returnee, assume the bride’s extent of participation in ISIS activities. Thus, this would allow the concept of volition to play a strong role in determining whether a respondent is more or less sympathetic towards a potential female returnee. The secrecy surrounding these women’s extent of collaboration is why I deliberately chose not to add an additional attribute in my research design that described the role these women played once they had already found their way into ISIS territory. Such as in the case of Shamima Begum, respondents may receive contradicting information on the role ‘ISIS brides’ played in the organization, and may choose other more definitive attributes to make their judgements about them.

Discrimination Against Religious and Ethnic Minorities

There exists discrimination between British born citizens on the basis of race. The question of “where are you actually from?” exemplifies this discrimination, where non-white British-born citizens are not seen as truly belonging in the country. Citizenship and ethnicity are conceptualized as antithetical to one another, given that to be ethnically different means to immediately identify yourself as a non-citizen (Alexander et al 2007). Inter-group interaction stimulated by demographic change can influence people to have more exclusionary attitudes towards members of out-groups (Enos 2014). Given the rise in population of Muslims and ethnic minorities in the United Kingdom, coupled with negative stereotyping , other UK citizens discriminate against these communities as they feel they wouldn’t belong in the country alongside the rest of the population. I argue that respondents might factor in their out-group prejudices towards Muslims, as well as towards ethnic minorities whilst choosing “favourable” cases to reintegrate, rather than simply judging the extent of volition ‘shown’ by each returnee.

Additionally, it is difficult to disaggregate discrimination against minorities on the basis of religion and ethnicity, given the racialization of faith in Western countries. Joshi (2006) argues that in multicultural societies like the US, people associate certain religions with certain communities - making ethnicity and race become proxies for a person’s religion. Discrimination against a Pakistani may be because of their ethnic origin, but also tied to the assumption that their religion (Islam) is incompatible with the West. Thus, both religious and ethnic discrimination become intertwined. I address this issue by separating ethnicity and outward religion into two separate attributes in my research design. By having possible combinations of white Hijabis and non-white non-Hijabis, I obtain an AMCE (Average Marginal Component Effect) for each attribute distinct from each other. Thus:

H3 (discrimination on the basis of ethnicity): UK respondents will be more supportive of rehabilitative measures towards white ISIS bride returnees than returnees of other ethnicities, irrespective of the respondent’s own identification.

Both Social Identity Theory, as well as discriminatory preferences towards Muslims would factor into respondent decision-making in regards to religiosity. Not only would respondents favor non-Muslims over Muslims due to in-group preferences, they would also assume that deradicalization would entail forgoing religious paraphernalia. Here, the normative view of a ‘deradicalized’ Muslim would make respondents view hijabi returnees more negatively.

Additionally, normative perceptions of ‘deradicalized’ women which play into determining who is seen as less of a risk upon their return. For example, reintegration schemes for female returnees were deeply tied with an image of what a ‘reintegrated’ returnee should appear as. Stenger (2024), argues that in reintegration schemes, the removal of religious paraphernalia was seen as successful deradicalization, implying that Islamic faith is seen as contrasting to Western culture. Thus, to integrate back into society, one must abandon outward Muslim religiosity. In Kazakh deradicalization programs, religious expression was regulated, where girls were barred from wearing the hijab (Weine 2019). Shamima Begum and other ‘brides’ like Emily Koening were also portrayed as shedding their religious garb in exchange for a more “Westernized appearance”, an action described as a ploy to appeal to the public (Pleasance 2021). This means that religiosity and deradicalization are seen as antithetical to one another, which problematizes the right to expression for the Muslim community. This problematization of faith is also recognized by returnees in Balkan states, many of whom highlighted discrimination against Niqabis as a challenge in their reintegration - widening the alienation between them and their communities (Perešin et al. 2021, 38).

H4 (discrimination on the basis of religion): UK respondents will be more supportive of punitive measures for ISIS bride returnees that are pictured wearing Muslim religious paraphernalia (such as the hijab or the niqab).

Additionally, I expect the political alignment of the respondent to have an impact on their view on rehabilitation for returnees, across the board. The political right-wing of Britain has consistently argued for the repeal of the Human Rights Act, which they argue gives undue rights and liberties to undeserving groups through compliance with the European Court of Human Rights (Webber 2021, 113). the more right-wing a person has, I argue that they are more supportive of punishment as opposed to rehabilitation. This is opposed to those on the left-wing of the political spectrum, who are more in favor of prison reform as well as structuralist understandings of human behavior (as opposed to completely individualistic). Additionally, the British right-wing has also consistently shown isolationist tendencies when confronted with European Union regulations and immigration, which was a major driving factor for their support of Brexit. Here, I argue that political alignment should affect not only support for repatriation across all returnees, but also the repatriation of non-white, visibly Muslim returnees as well. I posit that:

H5a (Ideology and Ethnicity): The more right-wing a respondent identifies as, the more supportive they will be of white returnees over non-white returnees. H5b (Ideology and Islam): The more right-wing a respondent identifies as, the more supportive they will be of visibly non-Muslim returnees over visibly Muslim returnees.

In-Group vs. Out-Group Attitudes

Multicultural societies like the UK lead to environments where social groups may either practice in-group policing, or in-group favoritism. Due to social categorization into collectives, individuals seek to differentiate themselves from the out-group to justify their own social position (Ashforth and Mael 1989, 31). Consequently, I expect the sample to react less favorably to visibly Muslim returnee profiles. Moreover, rehabilitative schemes often promote a normative profile of what a “deradicalized” female returnee should ideally look like—one devoid of all visibly Muslim symbology or clothing (Weine 2019). This normative view likely reinforces societal biases against Muslim profiles, creating a second layer of discrimination. Additionally, the likelihood of engaging in inter-group favoritism versus in-group policing — defined as intra-group accountability mechanisms that promote intergroup harmony — varies with group size and power dynamics. For example, White people, constituting 82% of the population (“Population of England and Wales” 2022), represent a dominant group. Unlike smaller ethnic communities, which may rely on in-group policing to prevent internal dissent and avert external backlash (Fearon and Laitin 1996), larger groups like White or non-Muslim communities are less incentivized to form such mechanisms. Instead, they are more likely to display inter-group favoritism due to the absence of pressures associated with minority group dynamics.

However, I argue that due to circumstances surrounding the Muslim community in the country, inter-group favoritism will be the mechanism that prevails. In-group policing serves as a counter-mechanism to in-group favoritism. This involves members of a group enforcing cooperative behavior within their own in-group to prevent inter-ethnic conflicts (Fearon and Laitin 1996). For Muslims in the UK, this dynamic is especially salient, as anti-immigrant groups often portray Islam as antithetical to Western values (Pupcenoks and McCabe 2013). This stigmatization exposes the Muslim community to heightened risks of violence, as seen in the aftermath of the July 2024 Southport Stabbings (Mohamed 2024), which triggered anti-immigrant and anti-Muslim riots across the country. Consequently, Muslims may adopt stricter in-group policing practices, holding their peers to higher behavioral standards and imposing harsher punishments. By disincentivizing radicalization within the community, these measures aim to mitigate backlash from Islamophobic and white supremacist groups. Smaller ethnic or religious groups often adopt strategies to foster inter-ethnic harmony. These may include assimilating into broader societal norms or engaging in “boundary policing,” wherein interactions with other groups are restricted (Fearon and Laitin 1996). This is as, in the UK, Muslims practice varying degrees of segregation, driven partly by the desire to shield themselves from racial hatred and discrimination. Additionally, proximity to mosques, Halal eateries, and other religious accommodations often shapes residential choices (Tahar et al. 2023). This relative segregation facilitates a degree of in-group favoritism by reducing intergroup interactions, diminishing the necessity for in-group policing mechanisms., Therefore, I argue that:

H5a (Muslim Intra-Group Favoritism): Muslim respondents will be more forgiving of visibly Muslim returnees than visibly non-Muslim returnees.

H5b (White Intra-Group Favoritism): White respondents will be more forgiving of White returnees as opposed to non White respondents.

Research Design and Empirical Strategy

I employ a conjoint experiment, where I present respondents with two hypothetical profiles of “ISIS brides.” Conjoints designs are suitable for dealing with social desirability bias (Horiuchi et al. 2022) which is relevant for my study because respondents may feel the need to respond in ‘politically correct’ manners. For my experiment, I will randomize four attributes of hypothetical ‘ISIS Brides’, and analyze the effect that each attribute has on the womens’ public perception.

The hypothetical returnees’ profiles will be communicated via a hybrid of text-based and visual based information. The attributes to be randomized are listed in Table 1. From these attributes, “Exit from ISIS”, name and age at which they left the UK will be communicated through text description. Additionally, skin tone and religious paraphernalia will be telegraphed via visuals. The same ethnic features become more salient in determining political behavior when portrayed via a visual conjoint as opposed to a text-based conjoint experiment (López Ortega and Radojevic 2024). This means that ethnicity and wearing a hijab, two highly visible traits, would become more effective in eliciting respondent reactions. In addition, most UK respondents receive information about female ISIS returnees via the news, which would consist of a headline and an accompanying picture. Therefore, I mimic this format to closely match how people actually receive relevant information in order to increase the external validity of my findings. Whilst originally planning to use a more neutral term for these returnees, given that the more popular terms are “ISIS bride” or “Jihadi bride”, I argue that using these exact terms will, again, match the way the UK public receives information. My attributes and their levels are shown in the poster.

Whilst traditional conjoint experiments assign one attribute to a corresponding variable, I have assigned two — name and skin tone — in order to reduce racial ambiguity. People, especially when not exposed to a high number of inter-group interactions, have a lower capacity to distinguish between out-group members — a phenomenon known as out-group homogeneity bias (Ackerman et al. 2006). During the research design stage, testing only the pictures and names as racial indicators resulted in mis-identification by members outside of that group. Therefore, relying on both names and pictures as racial indicators reduces ambiguity. Table 1 shows each attribute to be randomized and their respective levels.
Profile 1 “Alice”, 21 at the time of joining ISIS, awaits repatriation to the UK after fleeing the organization.
Profile 2 “Abena”, 15 at the time of joining ISIS, awaits repatriation to the UK after being captured by the Syrian government.

Table 2 illustrates what profiles might look like to respondents. I also asked questions on potential rehabilitation, prison-time, and more severe punishments through the form of a ratings system.The structure of these questions are inspired by Kao and Revkin’s (2023) measure for punishing ISIS collaborators.

Examples include:

  • To what extent do you think this particular returnee can be rehabilitated into UK society?
    (This question would be fielded in reference to both shown profiles.)

  • To what extent do you find this person’s presence a threat to UK society, if repatriated?

  • Which “ISIS Bride” do you think is more likely to be rehabilitated?

In order to test my hypotheses, I use the following equation to compute the Ordinary Least Squares Regression (OLS) to calculate the Average Marginal Component Effect (AMCE) of each treatment. Here, i refers to the respondent, j refers to the round, and k refers to the returnee profile.

\[Favouribility_{ijk} = \alpha + \\ \beta_1[age_{ijk} = 21] + \beta_2[age_{ijk} = 15]\\ \beta_2[hijab_{ijk} = NoHijab] + \\ \beta_3[skincolor_{ijk} = Tan] + \beta_4[skincolor_{ijk} = Black] + \\ \beta_5[Exit_{ijk} = Fled] + \varepsilon_{ijk}\]

For my conditional hypotheses (Muslim and White Intra-group Favouritism), I will utilize the same equation while subsetting my sample into Muslim and White respondents respectively.

In order to have sufficient statistical power for my conditional hypothesis regarding Muslims (Muslim Intra-group favouritism), I attempted to oversample Muslim respondents at twenty percent of the sample, as opposed to the approximately seven percent they constitute of the UK population. However, due to time constraints, was only able to obtain an approximately 12% Muslim sample. In order to mitigate the complications of reduced observations, the final 247 of the total 2011 respondents were presented with 7 rounds of forced choice questions, as opposed to the initial decision to field 4 rounds.

Results

Table 3: Summary Statistics

Table 3: Summary Statistics

The sample has a mean age of 42, with about half of it being female, and almost ⅓ having a Bachelor’s degree. About 75% of the survey respondents are white, and about 12% are Muslim. Additionally, 28% of respondents have a Bachelor’s degree, 17% have a Masters degree, and about 5% have a PhD.

Figure 1: AMCEs for the overall sample, and then subsamples consisting of Muslims, Non-Muslims, and non-white people. Muslim respondents are more favourable towards ethnic minorities, whilst less favourable towards returnees who fled ISIS and are pictured without the hijab as opposed to non-Muslims. White respondents are more favourable towards younger returnees (both aged 15 and 21), and are less favourable towards ethnic minorities. Additionally, White respondents also positively view returnees who flee ISIS, as well as those not pictured with the hijab.

Figure 1: AMCEs for the overall sample, and then subsamples consisting of Muslims, Non-Muslims, and non-white people. Muslim respondents are more favourable towards ethnic minorities, whilst less favourable towards returnees who fled ISIS and are pictured without the hijab as opposed to non-Muslims. White respondents are more favourable towards younger returnees (both aged 15 and 21), and are less favourable towards ethnic minorities. Additionally, White respondents also positively view returnees who flee ISIS, as well as those not pictured with the hijab.

Figures 1 and 2 describe the findings for the outlined conjoint experiment. The AMCEs for the age attribute are revealing: The younger a hypothetical returnee was upon joining ISIS, the more favourable public opinion is towards her prospective rehabilitation. This is in line with H1, or the age and volition hypothesis. It seems that indeed respondents view younger girls as less agentic, and subscribe to the idea that these girls were groomed into joining ISIS. Hence, the public believes in their future reintegration and ‘redemption’. The difference in magnitude between a returnee being fifteen versus twenty one is likely due to the fact that a fifteen year old is legally under the age of consent. Having a legally defined boundary under which individuals are seen as unable to consent strengthens the argument that a returnee of such an age was unable to make an informed decision before becoming an ‘ISIS bride’.

For H2 (exit and volition), clearly the respondents are more lenient towards women who fled ISIS as opposed to those who were captured by the Syrian government. This results speaks to the importance of volition in the public’s assessment of returnees and threat they may pose. Fleeing the organization potentially signals to the public that the ‘bride’ may have been manipulated into joining ISIS under false pretenses. Alternatively, fleeing may also signal a tangible step towards deradicalization, which may make the returnee seem less of a threat were she to return to the UK.

Racial discrimination towards ethnic minorities clearly factors into repatriation prospects. Both South Asian and Black returnees are favoured less than White returnees to be brought back to the UK and rehabilitated. Given that ethnic discrimination is still rife in the UK, with minorities often not being seen as ‘truly’ belonging to the country, it follows that people would be less enthusiastic about an ‘ISIS Bride’s’ return if she belonged to a minority group.

Following my findings on racial discrimination, I analyze in Figure 2 whether ethnic minorities were punished more severely for the same transgressions than their white counterparts. Black returnees are seen in a more positive manner when they are younger, as opposed to White and South Asian profiles. If people assume girls from minority backgrounds have less autonomy due to ‘oppressive’ cultures, respondents should view both Black and South Asian profiles more positively when they are younger, though that does not seem to be the case. White profiles are also seen more positively when they’re pictured not wearing the hijab, as opposed to minority women. This may be because not wearing the hijab as a white woman may signal a complete commitment to deradicalization (especially if respondents already assume that minority women were brought up in a ‘radicalized’ culture), whereas the risk of ‘relapse’ still remains for South Asian and Black women.

Figure 2: Are Ethnic Minorities More Heavily Punished for the Same Transgressions?Black profiles are seen more favourably when they’re younger, and white profiles more favorably when they wear the hijab.

Figure 2: Are Ethnic Minorities More Heavily Punished for the Same Transgressions?Black profiles are seen more favourably when they’re younger, and white profiles more favorably when they wear the hijab.

I also find evidence of religious discrimination. Referring back to Figure 1, profiles without a hijab were viewed in a more positive light, with respondents favouring their rehabilitation much more than of hijabi profiles. Due to deradicalization becoming synonymous with abandoning outward religiosity, respondents may believe that a hijabi woman is not truly committed to reintegration. Given that existing deradicalization programs also operate on the need for removing religious paraphernalia, it is likely that the conflation of religiosity and radical ideology makes hijabi women significantly less attractive candidates for repatriation.

To test the interplay between respondents’ ideological stance and attributes of the returness , I categorized respondents based on their ideological leanings. If a respondent scored between 1 and 5, I classified them as left-wing, and if they responded between 6 and 10, I classified them as right-wing. After doing so, I created an AMCE plot with the aforementioned sub-groupings, shown in figure 3. For H5a (Ideology and Ethnicity), there seems to be little to no difference between left and right-wing individuals, showing that political leanings have no bearing on whether respondents discriminate against returnees of ethnic minorities. Similarly, for H5b (Ideology and Islam), there does not seem to be a significant difference between respondents for hijabi or non-hijabi returnees. Although there is no difference in how left-wing and right-wing respondents react to returnees’ (ostensive) religiosity and ethnicity, Interestingly, left-wing respondents are significantly more sympathetic towards returnees who joined ISIS as minors and those who have fled the organization. This may be due to the general left-wing belief in rehabilitative justice over punitive justice, as both fleeing the organization and being underage are indicators of remorse and/or a lack of full decision-making capacity at the time

Figure 3: AMCEs for Right-Wing and Left-Wing Respondents. Left-Wing respondents highly favour the reintegration of 15 year old returnees, as well as those who fled ISIS. They also slightly favor non hijabi profiles more than right-wing respondents. ight-wing respondents are more favourable towards ethnic minorities, though the magnitude of this difference is relatively small.

Figure 3: AMCEs for Right-Wing and Left-Wing Respondents. Left-Wing respondents highly favour the reintegration of 15 year old returnees, as well as those who fled ISIS. They also slightly favor non hijabi profiles more than right-wing respondents. ight-wing respondents are more favourable towards ethnic minorities, though the magnitude of this difference is relatively small.

For H6a and H6b, which posit that both Muslims and White people exhibit in-group favoritism, I conduct subgroup analyses shown in Figures 1 and 2 respectively. Muslims do not differentiate between visibly Muslim and non Muslim returnees. Additionally, respondents identifying as white see South Asian and Black returnees in a less favourable light. Thus, my subgroup analyses show strong support for the fact that both Muslim and white communities exhibit favouritism towards in-group members. Though as a word of caution, due to a lower response rate from Muslims than hoped, my results may not be sufficiently powered.

Discussion

Respondents are more likely to support the ‘rehabilitation’ of a potential returnee if they view her as a victim, or as redeemable. A returnee’s age and means of organizational exit clearly signal remorse and victimhood, which is why respondents seem more favourable towards rehabilitating younger returnees who fled ISIS. However, it also seems that a returnee’s ‘redemption’ in the eyes of the public is also reliant on their ethnicity and religion. By virtue of being a minority, people may need to ‘perform’ in front of their peers to demonstrate their loyalty to the UK as a country, and British as an identity. If someone does the opposite, and commits an act that goes against the perceived values of their national identity, they may be more heavily punished for their transgressions as the threshold for belonging is higher for them than it is for White Christian citizens. When the 2024 Southport Stabbings occurred, disinformation spread about how the attacker was a Muslim immigrant, when in reality he was a Christian and citizen by birth (Mohamed 2024). When crime and transgression is associated with being a minority and not belonging, it creates an environment where racist and white supremacist ideas can proliferate, putting entire communities at risk of discrimination, profiling, and hate-crimes.

My findings on religious discrimination also have important implications for designing deradicalization programs for potential returnees. Previous reintegration schemes, such as in Kazakhstan, focused on deradicalization through abandoning outward religiosity (Weine 2019). Niqabi women in the Balkans felt more alienated upon their return due to their outward appearance, making reintegration more difficult for them than it would be for other women ) (Perešin et al. 2021, 38). Additionally, this is likely why some media outlets accused the lawyers of Shamima Begum and Emily Koening of appearing in the media in Westernized clothing, as that would help convince the public that they ‘deserved’ to return (Pleasance 2021). Given that ISIS preyed upon these women’s fears of not being truly accepted in their home country, discrimination against Hijabi and Niqabi women might justify the same fears that pushed these women to leave in the first place. Therefore, it is imperative that such rehabilitation facilities focus on reconciling Islamic religiosity with ‘Western’ living, otherwise fundamentalist groups will continue to use the apparent irreconcilable differences with the two as a recruitment tactic.

My experiment also comes with two limitations to keep in mind. Firstly, I was not able to get as many Muslim respondents as I aimed for (12% of the sample vs. the desired 20%), This means that my sub-group analysis on Muslim intra-group favoritism may not be sufficiently powered. Secondly, the photos used in my experiment were AI generated, and then subsequently manipulated in photoshop. As a word of caution, there may be a difference in responses based on whether people are able to recognize the manipulation, which may bias the results. Therefore, in order to see whether using AI faces in future experiments is a worthwhile endeavor, replicating the experiment with both AI and real model headshots could reveal whether the former can be a low cost alternative to the latter.

Conclusion

My study sheds light on the public perception of ‘ISIS Brides’, a subset of women who occupy a gray area between perpetrator and victim, some who possibly occupy both positions simultaneously. Given the absence of concrete information on most ‘ISIS Brides’, as well as the complicated circumstances behind their departure from Britain, the public uses the limited knowledge it has to ascertain whether these women are deserving of forgiveness and rehabilitation. My findings show that the public uses age, organizational exit, ethnicity and religiosity as heuristics to make judgements about the level of rehabilitation to be afforded to these women. Not only do respondents take into account whether a returnee was reasonably autonomous in decision-making, they also ‘punish’ brides for being ethnic and religious minorities. Decisions on whether a returnee is deserving of forgiveness are influenced by ethnic and religious discrimination, with those seemingly conforming to ‘Western’ ideals being more deserving of repatriation. Post conflict literature, and restorative justice programs should delve into how best to reintegrate those who already suffer some degree of alienation due to their ethnicity or religion, and how to balance accountability with adequate amounts of social support. Otherwise, governments risk letting such women slip between the cracks once more, creating another generation lost to fundamentalist and terrorist groups.

Bibliography

Ackerman, Joshua M., Jenessa R. Shapiro, Steven L. Neuberg, Douglas T. Kenrick, D. Vaughn Becker, Vladas Griskevicius, Jon K. Maner, and Mark Schaller. 2006. “They All Look the Same to Me (Unless They’re Angry): From Out-Group Homogeneity to Out-Group Heterogeneity.” Psychological Science 17 (10): 836–40. https://doi.org/10.1111/j.1467-9280.2006.01790.x.

Agneman, Gustav, and Lisa Strömbom. 2023. “How Ethnic Discrimination Shapes Political Reintegration After War: Insights From a Conjoint Experiment in Colombia.” Journal of Conflict Resolution, December, 00220027231222252. https://doi.org/10.1177/00220027231222252.

Alexander, Claire, Rosalind Edwards, and Bogusia Temple. 2007. “Contesting Cultural Communities: Language, Ethnicity and Citizenship in Britain.” Journal of Ethnic and Migration Studies 33 (5): 783–800. https://doi.org/10.1080/13691830701359223.

Ashforth, Blake E., and Fred Mael. 1989. “Social Identity Theory and the Organization.” The Academy of Management Review 14 (1): 20. https://doi.org/10.2307/258189.

Baker, Josh, Sara Obeidat, and Joe Kent. 2023. “Friend Who Inspired Shamima Begum to Join IS Mocks Her as Non-Believer.” March 22, 2023. https://www.bbc.com/news/uk-65031943.

Bottoms, Bette L., Suzanne L. Davis, and Michelle A. Epstein. 2004. “Effects of Victim and Defendant Race on Jurors’ Decisions in Child Sexual Abuse Cases.” Journal of Applied Social Psychology 34 (1): 1–33. https://doi.org/10.1111/j.1559-1816.2004.tb02535.x.

Capone, Francesca. 2017. “‘Worse’ than Child Soldiers? A Critical Analysis of Foreign Children in the Ranks of Isil.” International Criminal Law Review 17 (1): 161–85. https://doi.org/10.1163/15718123-01701003.

Cheng, Amy. 2021. “After Long Legal Battle with ‘ISIS Bride,’ U.K. Pushes for Power to Cancel Citizenship without Notice.” The Washington Post (Online). Washington, D.C., United States Washington, D.C.: WP Company LLC d/b/a The Washington Post. 2599293002. ProQuest Central. November 19, 2021. http://proxy.library.nyu.edu/login?qurl=https%3A%2F%2Fwww.proquest.com%2Fblogs-podcasts-websites%2Fafter-long-legal-battle-with-isis-bride-u-k%2Fdocview%2F2599293002%2Fse-2%3Faccountid%3D12768.

Clayton, Katherine, Jeremy Ferwerda, and Yusaku Horiuchi. 2021. “Exposure to Immigration and Admission Preferences: Evidence from France.” Political Behavior 43 (1): 175–200. https://doi.org/10.1007/s11109-019-09550-z. Cole, Samantha. 2019. “This Company Thinks It Can Solve Diversity With 100,000 Fake AI Faces.” VICE (blog). September 20, 2019. https://www.vice.com/en/article/generated-photos-thinks-it-can-solve-diversity-with-100000-fake-ai-faces/.

Cook, Joana, and Gina Amy Vale. 2018. From Daesh to “Diaspora”: Tracing the Women and Minors of Islamic State. International Centre for the Study of Radicalisation.

De Coninck, David, Lise-Lore Steeman, and Leen d’Haenens. 2024. “Uncovering Covert Discrimination: A Conjoint Experiment.” Ethics & Behavior, September, 1–16. https://doi.org/10.1080/10508422.2024.2400973.

Donnaloja, Victoria. 2022. “British Nationals’ Preferences Over Who Gets to Be a Citizen According to a Choice-Based Conjoint Experiment.” European Sociological Review 38 (2): 202–18. https://doi.org/10.1093/esr/jcab034.

El Masri, Mirna, Jeyabraba, Maiyoraa, and Welsh, Blair. 2024. “Former ISIS Collaborators in the Middle East and North Africa: Repatriation and Reconciliation in Comparison.” Pre-Analysis Plan. Open Science Framework.

Enos, Ryan D. 2014. “Causal Effect of Intergroup Contact on Exclusionary Attitudes.” Proceedings of the National Academy of Sciences 111 (10): 3699–3704. https://doi.org/10.1073/pnas.1317670111.

Fearon, James D., and David D. Laitin. 1996. “Explaining Interethnic Cooperation.” American Political Science Review 90 (4): 715–35. https://doi.org/10.2307/2945838.

Findor, Andrej, Matej Hruška, Petra Jankovská, and Michaela Pobudová. 2022. “Who Should Be Given an Opportunity to Live in Slovakia? A Conjoint Experiment on Immigration Preferences.” Journal of Immigrant & Refugee Studies 20 (1): 79–93. https://doi.org/10.1080/15562948.2021.1902597.

Fisher, Lucy. 2023. “UK under Fresh Pressure to Bring Home Nationals Held in Syria.” July 27, 2023.https://www.ft.com/content/39adea5e-5fe9-4042-8b27-fc0789c22d8d.

“France Repatriates 51 from Syria Camps in Policy Change.” 2022. July 5, 2022. https://www.al-monitor.com/originals/2022/07/france-repatriates-51-syria-camps-policy-change

Freitag, Markus and Julian Schuessler (2020). “cjpowR – A Priori Power Analyses for Conjoint Experiments,” R Package.

Giri, Keshab, and Roos Haer. 2024. “Female Combatants and Durability of Civil War.” Studies in Conflict & Terrorism 47 (5): 526–47. https://doi.org/10.1080/1057610X.2021.1980982.

Godefroidt, Amélie, and Arnim Langer. 2023. “What Drives Attitudes towards the Reintegration of Former Fighters? Insights from a Conjoint Experiment in Nigeria.” Journal of Peace Research 60 (3): 410–27. https://doi.org/10.1177/00223433221087070.

Hainmueller, Jens, Daniel J. Hopkins, and Teppei Yamamoto. 2014. “Causal Inference in Conjoint Analysis: Understanding Multidimensional Choices via Stated Preference Experiments.” Political Analysis 22 (1): 1–30. https://doi.org/10.1093/pan/mpt024.

Hainmueller, Jens, and Daniel J. Hopkins. 2015. “The Hidden American Immigration Consensus: A Conjoint Analysis of Attitudes toward Immigrants.” American Journal of Political Science 59 (3): 529–48. https://doi.org/10.1111/ajps.12138. Jackson, Leonie B. 2022. The Monstrous & the Vulnerable: Framing British Jihadi Brides. 1st ed. Oxford University Press. https://doi.org/10.1093/oso/9780197647332.001.0001.

Jacoby, Tami Amanda. 2015. “Jihadi Brides at the Intersections of Contemporary Feminism.” New Political Science 37 (4): 525–42. https://doi.org/10.1080/07393148.2015.1089028.

Joshi, Khyati Y. 2006. “The Racialization of Hinduism, Islam, and Sikhism in the United States.” Equity & Excellence in Education 39 (3): 211–26. https://doi.org/10.1080/10665680600790327.

Kao, Kristen, and Mara R. Revkin. 2023. “Retribution or Reconciliation? Post‐Conflict Attitudes toward Enemy Collaborators.” American Journal of Political Science 67 (2): 358–73. https://doi.org/10.1111/ajps.12673.

Khomami, Nadia. 2018. “Number of Women and Children Who Joined Isis ‘Significantly Underestimated.’” The Guardian, July 23, 2018, sec. World news. https://www.theguardian.com/world/2018/jul/23/number-of-women-and-children-joining-isis-significantly-underestimated.

Krzyzowski, Lukasz, and Magdalena Nowicka. 2021. “European Solidarity as Boundary-Making: A Conjoint Analysis of Attitudes towards Islam in the Context of the ‘Refugee Crisis.’” Journal of Sociology 57 (2): 305–24. https://doi.org/10.1177/1440783320902945.

Kulenbekova, Eleanora, and Daniel Wittenberg. 2024. “Women and Children Who Went to Live with IS in Syria Are Being Brought Home,” January 16, 2024. https://www.bbc.com/news/world-asia-67926617.

López Ortega, Alberto, and Marco Radojevic. 2024. “Visual Conjoint vs. Text Conjoint and the Differential Discriminatory Effect of (Visible) Social Categories.” Political Behavior, June. https://doi.org/10.1007/s11109-024-09953-7.

Martini, Alice. 2018. “Making Women Terrorists into ‘Jihadi Brides’: An Analysis of Media Narratives on Women Joining ISIS.” Critical Studies on Terrorism 11 (3): 458–77. https://doi.org/10.1080/17539153.2018.1448204.

Milton, Daniel, and Brian Dodwell. 2018. “Jihadi Brides? Examining a Female Guesthouse Registry from the Islamic State’s Caliphate.” Combating Terrorism Center at Westpoint. https://ctc.westpoint.edu/jihadi-brides-examining-female-guesthouse-registry-islamic-states-caliphate/.

Mironova, Vera, and Sam Whitt. 2024. “Retribution versus Rehabilitation for Children within Insurgency: Public Attitudes Toward ISIS-Affiliated Youth in Mosul, Iraq.” Terrorism and Political Violence, March, 1–19. https://doi.org/10.1080/09546553.2024.2319254.

Mohamed, Edna. 2024. “Southport Stabbing: What Led to the Spread of Disinformation?” Al Jazeera. August 2, 2024. https://www.aljazeera.com/news/2024/8/2/southport-stabbing-what-led-to-the-spread-of-disinformation.

Nyamutata, Conrad. 2020. “Young Terrorists or Child Soldiers? ISIS Children, International Law and Victimhood.” Journal of Conflict and Security Law 25 (2): 237–61. https://doi.org/10.1093/jcsl/krz034.

Perešin, Anita, Melisa Hasanović, and Kujtim Bytyqi. 2021. “Female Returnees from Syria to the Western Balkans.” Perspectives on Terrorism 15 (5): 29–45. Philips, Aleks. 2024. “Shamima Begum Loses Final UK Court Bid over Citizenship.” August 7, 2024. https://www.bbc.com/news/articles/cm2n8xv61x3o.

Pokalova, Elena. 2019. Returning Islamist Foreign Fighters : Threats and Challenges to the West. Cham, SWITZERLAND: Springer International Publishing AG. http://ebookcentral.proquest.com/lib/nyulibrary-ebooks/detail.action?docID=5967899.

“Population of England and Wales.” 2022. December 22, 2022. https://www.ethnicity-facts-figures.service.gov.uk/uk-population-by-ethnicity/national-and-regional-populations/population-of-england-and-wales/latest/.

Porter, Jon. 2019. “100,000 Free AI-Generated Headshots Put Stock Photo Companies on Notice.” The Verge. September 20, 2019. https://www.theverge.com/2019/9/20/20875362/100000-fake-ai-photos-stock-photography-royalty-free.

Powell, Amber Joy, Heather R. Hlavka, and Sameena Mulla. 2017. “Intersectionality and Credibility in Child Sexual Assault Trials.” Gender & Society 31 (4): 457–80. https://doi.org/10.1177/0891243217716116.

Pupcenoks, Juris, and Ryan McCabe. 2013. “The Rise of the Fringe: Right Wing Populists, Islamists and Politics in the UK.” Journal of Muslim Minority Affairs 33 (2): 171–84. https://doi.org/10.1080/13602004.2013.826451.

“Runaway German Schoolgirl Linda Wenzel Jailed in Iraq.” 2018. Theweek. February 19, 2018. https://theweek.com/91749/runaway-german-schoolgirl-linda-wenzel-jailed-in-iraq. Sabbagh, Dan, and Owen Bowcott. 2020. “Javid Advised of ‘hostile Public Sentiment’ to Shamima Begum, Court Told.” The Guardian, November 24, 2020, sec. UK news. https://www.theguardian.com/uk-news/2020/nov/24/shamima-begum-cannot-communicate-with-legal-team-court-told.

Saltman, Erin Marie, and Melanie Smith. 2015. “‘Till Martyrdom Do Us Part’ Gender and the ISIS Phenomenon.” Institute for Strategic Dialogue. https://www.isdglobal.org/wp-content/uploads/2016/02/Till_Martyrdom_Do_Us_Part_Gender_and_the_ISIS_Phenomenon.pdf.

Saripi, Nur Irfani Binte. 2015. “Female Members of ISIS: A Greater Need for Rehabilitation.” Counter Terrorist Trends and Analyses 7 (3): 26–31.

“Shamima Begum Refused UK Supreme Court Appeal of Citizenship Removal.” Al Arabiya English. August 7, 2024. https://english.alarabiya.net/News/world/2024/08/07/shamima-begum-refused-uk-supreme-court-appeal-of-citizenship-removal.

Sobolewska, Maria, Silvia Galandini, and Laurence Lessard-Phillips. 2017. “The Public View of Immigrant Integration: Multidimensional and Consensual. Evidence from Survey Experiments in the UK and the Netherlands.” Journal of Ethnic and Migration Studies 43 (1): 58–79. https://doi.org/10.1080/1369183X.2016.1248377.

Steele, Liza G., Lamis Abdelaaty, and Nga Than. 2023. “Attitudes about Refugees and Immigrants Arriving in the United States: A Conjoint Experiment.” Ethnic and Racial Studies 46 (10): 2163–91. https://doi.org/10.1080/01419870.2022.2159475.

Tahar, Farouq, Asma Mehan, and Krzysztof Nawratek. 2023. “Spatial Reflections on Muslims’ Segregation in Britain.” Religions 14 (3): 349. https://doi.org/10.3390/rel14030349.

Tanno, Sophie, Jessie Gretener, and Benjamin Brown. 2024. “Woman Who Joined ISIS as a Teen Loses Challenge against the Removal of Her UK Citizenship.” CNN. February 23, 2024. https://www.cnn.com/2024/02/23/europe/shamima-begum-appeal-loses-intl/index.html.

Thomas, Jakana L. 2021. “Wolves in Sheep’s Clothing: Assessing the Effect of Gender Norms on the Lethality of Female Suicide Terrorism.” International Organization 75 (3): 769–802. https://doi.org/10.1017/S0020818321000035.

Thomas, Jakana L., and Kanisha D. Bond. 2015. “Women’s Participation in Violent Political Organizations.” American Political Science Review 109 (3): 488–506. https://doi.org/10.1017/S0003055415000313.

Whitehead, Tom. 2016. “One in Three Jihadists Now Women, Warns Police Chief.” The Telegraph, April 27, 2016. https://www.telegraph.co.uk/news/2016/04/27/one-in-three-jihadists-now-women-warns-police-chief/.

Webber, Frances. 2021. “Britain’s Authoritarian Turn.” Race & Class 62 (4): 106–20. https://doi.org/10.1177/0306396821989181.

Weber, Sanne. 2021. “Defying the Victim-Perpetrator Binary: Female Ex-Combatants in Colombia and Guatemala as Complex Political Perpetrators∞.” International Journal of Transitional Justice 15 (2): 264–83. https://doi.org/10.1093/ijtj/ijab006.

Wood, Reed M, and Jakana L Thomas. 2017. “Women on the Frontline: Rebel Group Ideology and Women’s Participation in Violent Rebellion.” Journal of Peace Research 54 (1): 31–46. https://doi.org/10.1177/0022343316675025.