Abstract
While there is ample research on everyday cosmopolitanism, the relation with religion is less understood. This study examines the difference in everyday cosmopolitanism between Muslim, Christian and non-religious urban youth. Further, it studies the influence of religiosity, religious identification and perceived discrimination on cosmopolitanism. A one-way ANOVA analysis was conducted on data from 1039 students in 17 secondary schools in the super-diverse city of Antwerp. Multilevel regression analysis was conducted on a sample of Muslim (n = 496) and Christian (n = 225) youth. The results indicate no difference between religious and non-religious youth regarding their everyday cosmopolitanism. Moreover, for Muslim youth, intrinsic religiosity is positively associated with cosmopolitan orientations, while religious identification and discrimination negatively effect cosmopolitanism. For Christian youth, religious factors do not explain their cosmopolitan orientations. Overall, the article suggests that scholars and policy makers should discuss the potential of religion to foster cosmopolitan orientations.
This article examines the relation between religion and everyday cosmopolitanism among urban diverse youth. Although much of the debate on cosmopolitanism has been theoretical and normative, empirical research has investigated how cosmopolitanism is constructed ‘from below’ (Keating, 2016; Pichler, 2008; Werbner, 2015). Researchers are increasingly interested in questions concerning which individuals or groups are more likely to have cosmopolitan orientations, mostly understood as a certain openness to cultural difference or diversity (Skrbis and Woodward, 2007).
Much research has examined to what extent such a cosmopolitan openness is predicted by socio-demographic indicators like social class, gender, educational level, residential area, etc. (e.g. Keating, 2016; Olofsson and Öhman, 2007; Pichler, 2009; Skrbis and Woodward, 2007; Werbner, 1999). Some have briefly touched upon the potential effect of religion on cosmopolitanism (e.g. Woodward et al., 2008). In this article, we aim to deepen the understanding of the role of religion by exploring how religious factors can foster or weaken cosmopolitan orientations (see also Roudometof, 2005), in particular in the lives of youth in super-diverse cities.
We aim to achieve this by examining the following questions: (1) Do Muslim, Christian and non-religious urban youth differ in their everyday cosmopolitan orientations? (2) What are the effects of religiosity (religious practices and intrinsic religiosity), religious identification and perceived discrimination of ethnic/religious groups in school on everyday cosmopolitan orientations for Muslim and Christian youth? Our analysis is based on data collected from 1039 students in the 5th and 6th year from 17 secondary education schools in the super-diverse city of Antwerp.
Antwerp is the largest city in the Flemish region in Belgium and displays a high degree of cultural and religious diversity (Oosterlynck et al., 2017). It has citizens from 171 different nationalities, and in recent years the population with a migration background (50.1%) has become numerically larger than the share of ‘native’ Belgian residents (49.9 %); this makes it a good example of a ‘majority-minority city’ (Crul, 2016) in which there is no longer a numerical ethnic majority. The share of residents with a migration background is even expected to increase further, as in the group of children aged 10–19, only 29.6% children are considered ‘native’. Therefore, youth in Antwerp schools are confronted with religious and cultural differences on a daily basis. More generally, research suggests that young people in urban areas are more likely to report cosmopolitan identities and attitudes, but that there is a lack of in-depth empirical and comparative research (Keating, 2016; Norris and Inglehart, 2009). The context of Antwerp, then, seems well suited to explore cosmopolitan identities and orientations among urban youth.
First, we clarify our theoretical starting points and discuss what we understand by everyday cosmopolitan orientations. Second, we look into the relationship with religion, particularly Islam and Christianity, and discuss the specific role of religiosity, religious identification and discrimination. Third, we present the methodology and discuss the results. We conclude by discussing our findings and limitations.
Theory and hypotheses
Cosmopolitanism and everyday cosmopolitan orientations
Over the past two decades there has been a broad discussion about the notion of cosmopolitanism across a wide range of disciplines (for an overview, see Delanty, 2012). In the social sciences, cosmopolitanism is mostly understood as a conscious openness to cultural differences, loyalty to human kind and feelings of being a world citizen (Hannerz, 2004; Skrbis and Woodward, 2007; Vertovec and Cohen, 2002). It is associated with an increase in supra-national contacts and the emergence of post-national identities, and is often defined in opposition to ethnic or exclusive nationalism (Pichler, 2008). However, the field of cosmopolitanism studies has also received criticism for its too narrow focus on the archetypal, mobile elite that travels the globe easily, thereby reproducing the imagination of cosmopolitanism as an ‘elite’ identity feature (Calhoun, 2002; Werbner, 1999, 2015). Therefore, for a broader understanding of cosmopolitanism, research needs to widen its lens.
Accordingly, our study focuses on everyday cosmopolitan orientations among diverse youth. We build on research on so-called ‘ordinary’ or ‘everyday’ cosmopolitanism in which different scholars set out to examine how cosmopolitan ‘openness’ – as an everyday disposition – is constituted in quotidian contexts (Pichler, 2009; Skrbis and Woodward, 2007; Wang, 2018). It concerns a cosmopolitanism ‘from below’ or a ‘vernacular cosmopolitanism’, which is less elitist and less Eurocentric (Werbner, 2015). This is in line with Appiah’s (2010) argument on ‘rooted cosmopolitanism’, referring to cosmopolitans as members of morally and emotionally significant communities, and thus as ‘rooted’ in local allegiances, while embracing notions of tolerance and openness to the world and a shared humanity identity. The notion of ‘rooted cosmopolitanism’ enables an understanding of everyday cosmopolitanism as embedded in the interactions and negotiations between various belongings, loyalties and multiple identities of individuals in everyday life. It can grasp the ‘in-between’ situation of young people in super-diverse contexts, where cultural relations are actively reworked (Synnes, 2018; Turner et al., 2014).
Researchers have distinguished various features of such an everyday cosmopolitan disposition (ranging from the political over the aesthetic-cultural to the ethical) (Keating, 2016; Pichler, 2008; Vertovec and Cohen, 2002). Yet, it is argued that individuals who self-identify as cosmopolitan usually do not display all possible features. This article employs an empirical measure of everyday cosmopolitanism at the individual level, which includes two crucial dimensions. We focus on an ‘identity’ dimension (the extent to which individuals see and feel themselves as citizens of the world) and an ‘attitude’ dimension (the extent to which individuals hold attitudes and beliefs recognizing diversity) (Keating, 2016; Pichler, 2009). In line with Pichler (2009), who indicates a positive association between the identity and attitudinal components, we argue that both notions are closely related and needed for the measurement of everyday cosmopolitanism.
Religion and everyday cosmopolitanism
Literature emphasizes that being able to imagine and appreciate lives across social, cultural and ethno-religious boundaries requires a certain ‘cosmopolitan imagination’ (Delanty, 2009). In most literature, it has been largely presupposed that cosmopolitanism is intertwined with liberalism and secularism (cf. Appiah, 2017). Therefore, religion is usually not taken into consideration as a component of this cosmopolitan imagination.
While Western European countries have traditionally supported one or more versions of Christianity, in the last decades a more radical secularism has become influential, especially after 9/11, and most visibly in countries such as France, Denmark, the Netherlands and Belgium. This secularist discourse pits religious identities and group membership against secular-political authority and equal citizenship (Modood, 2019). Much like the popular discourse on the subject, researchers and the social sciences highlight how the different moral grammars and worldviews of religions can lead people to emphasize the importance of group differences, with religion sometimes informing political conflict and violence (e.g. Gorski and Türkmen-Dervişoğlu, 2013; Juergensmeyer et al., 2013). Koopmans (2015), for instance, studies how religiosity and fundamentalism among Muslims and Christians may incite hostile out-group views.
The negative social connotations of religion in Western Europe are particularly pronounced when it comes to Islam (Cesari, 2004; Foner and Alba, 2008). In the public and political debate, specific features of Islam are often problematized and seen as a threat to the liberal values of European countries. Subsequently, Muslims are frequently framed as the ethnic and religious other (Fadil et al., 2014; Zemni, 2011) and they are expected to not prioritize their religious identity, at least if they desire full membership to the national community (Modood, 2019). This is confirmed by scholarly work studying the salience of religious–secular symbolic boundaries in Europe (Foner and Alba, 2008; Trittler, 2019). Scholars mostly focus on Islam when discussing immigrant religion. While some focus on how Muslims actively manage stereotypes and prejudices (e.g. Lamont et al., 2002), others analyse Islam as a barrier to integration for minorities or discuss how it informs Muslims to be less accepting towards e.g. gender equality and gay rights (e.g. Norris and Inglehart, 2008). Generally, a lot of attention is dedicated to themes such as gender relations, freedom of speech, radicalism and the (in)compatibility between Islam and Western values.
Less research has focused on how religiosity among Muslims and Christians can incite positive out-group views and a cosmopolitan openness towards cultural differences. Religion is indeed rarely examined as positive contributor to the emergence of global citizenship and common humanity (Iqtidar, 2012; Levitt, 2008).
Nevertheless, there is also recognition of the role that religion has played in the cultivation of the cosmopolitan interest. At the macro-level, it is widely acknowledged that Islam and Christianity have contributed to the emergence of cosmopolitan ideals (Iqtidar, 2012; Turner, 2010). By forming transcultural sacred imagined communities (albeit through colonial conquest), Christianity has developed into a world religion, entailing the notion of a global community of believers under the concept of ‘Christendom’ (Beyer, 1994). In the case of Islam, many observers have highlighted its universalistic dimensions and cosmopolitan human morals. This ‘Muslim Cosmopolitanism’ is in part the legacy of the doctrine of the Ummah – expressing the belief that the Islamic community should transcend local, national or ethnic boundaries – and the modern development of a global Muslim diaspora (Bowen, 2004). Furthermore, through intense cultural contact, religions like Islam and Christianity have developed an ecumenical consciousness and a tolerance of difference, albeit in slow and fragile ways (Turner, 2010).
At the level of everyday interaction, religion can be used as a strategy of bridging group boundaries and engaging in equality claims-making based on a common identity. Although our study is based on survey data, it resonates with qualitative research demonstrating how different groups can employ universalistic religious repertoires to foreground a commitment to a common humanity. Researchers (e.g. Beaman, 2016; Jacobson, 1997; Lamont et al., 2002) have shown how Muslims tend to emphasize a commitment to a set of principles informed by Islamic moral universalism, asserting the intrinsic equality, morals and destiny of humans across all boundaries (nationality, ethnicity, religion, etc.). In an analogous way, Synnes (2018) has shown how Christian youth emphasize a universal understanding of their religion with values such as inclusiveness and transcendence of (ethnic) symbolic boundaries. Bayat (2009: 186) has demonstrated how Muslims and Christians in an Egyptian suburb have imaginary and prejudiced views of the ‘other’ but still develop an ‘everyday cosmopolitan coexistence’ among each other.
Furthermore, quantitative studies have shown the importance of redirecting collective identity projects to supra-national identities such as the cosmopolitan identity. Saroglou and Mathijsen (2007), for instance, have shown that Muslim and Christian youth with an immigration background identify less with the Belgian identity than with the world citizen identity. Likewise, Agirdag et al. (2016) and Clycq et al. (2020) have indicated that ethno-religious minorities in Belgium tend to redirect their identities towards cosmopolitan identities (here the European identity), and away from exclusive (sub)national identities. Moreover, as already indicated, research has shown that young people are more likely to identify as cosmopolitan, particularly when they live, or go to school, in urban areas (Keating, 2016; Norris and Inglehart, 2009).
In the case of Antwerp, we are interested in knowing whether there will be differences in cosmopolitan orientations between Muslims, Christians and non-religious youth. Based on the theoretical and empirical insights discussed above, we expect the following: (H1) Muslim, Christian and non-religious urban youth will have cosmopolitan orientations and there will be no significant difference between religious and non-religious youth.
The role of religiosity, religious identification and perceived discrimination
Further, we deepen our understanding of the relation between religion and everyday cosmopolitanism by examining the effects of religiosity, religious identification and perceived discrimination of ethnic and religious groups at school for Muslim and Christian youth.
Religiosity
As studies have indicated differing trends for various dimensions of religiosity, we will approach it as a multidimensional phenomenon (Molteni and Biolcati, 2018; Van Praag et al., 2016). Therefore, we will look at religious practices and intrinsic-personal religiosity.
With religious practices, we refer to the belonging of individuals to religious communities through (public) participation in religious rituals and communal activities (Huber and Huber, 2012). Research indicates a decline in the religious practices and beliefs of West European Christians (Molteni and Biolcati, 2018; Storm, 2011). However, for Christians with an immigration background religious practices can remain important (Maxwell and Bleich, 2014; Storm, 2011). Moreover, Christian practices and rituals still have a symbolic meaning in secularized European countries and are seen as part of the nation’s heritage (Fleischmann and Phalet, 2018; Storm, 2011). In contrast, religious practices of Muslims are often highly problematized in the public debate. National media pay quite a lot of (mostly negative) attention to Islamic rules and rites, such as the wearing of the headscarf (banned for public functions and in secondary schools), ritual slaughter, religious holidays, etc. (Ichau and d’Haenens, 2016). Research shows that (some) Muslims will simply continue these religious practices and use them as identity markers, partly in reaction to their problematization (Foner and Alba, 2008). It has indeed been demonstrated that religious traditions are highly valued for first and second generation Muslims (Fleischmann and Phalet, 2018). However, researchers also show a decline in religious practices of second generation Muslims (Maliepaard et al., 2010; Voas and Fleischmann, 2012).
This brings us to the second religiosity dimension, i.e. intrinsic-personal religiosity. This refers to the interest in religious values, the meaning of religion and the emotional dimension. It is the perception of oneself as religious, beyond practices, dogma or official membership, where leading religiously meaningful lives is a goal in itself (Allport and Ross, 1967; Ghorpade et al., 2006). This resonates with a widespread observation of the emergence of individualized forms of religiosity. Several sociologists of religion have described how an important segment of the Catholic world has distanced itself from the Church (especially in relation to religious practices) and increasingly define their religion in terms of so-called typical Christian values such as social justice, a humane approach towards people and solidarity (Cipriani, 2001; Dobbelaere and Voyé, 1990). For Muslims as well, it is increasingly argued that they negotiate their religiosity in the West European context by individualizing and privatizing their religious practices (Cesari, 2004). This trend would make religious public practices play a less important role, as Muslims tend to prioritize an ‘Islam of the heart’ (Beaman, 2016; Killian, 2007).
Various scholars studied the effects of different dimensions of religiosity on e.g. out-group perceptions, civic and social engagement, identification processes, etc. On the one hand, Verkuyten and Yildiz (2007) argue that Muslims who are involved in religious practices show a dis-identification with the national (Dutch) identity, which could indicate social closure. Likewise, Scheepers et al. (2002) argue that Christians who attend church more frequently are more prejudiced towards others. On the other hand, some researchers also indicate a positive association between churchgoing and open views to immigration and racial differences (Storm, 2011). In the case of intrinsic-personal religiosity, researchers argue that individual religiosity reduces negative out-group views and prejudices (Allport and Ross, 1967; Scheepers et al., 2002; Storm, 2011). Lastly, Grundel and Maliepaard (2012) show that personal religious values of Muslims are compatible with democratic values and tolerance towards difference. While these studies did not directly examine cosmopolitan orientations, we follow these empirical insights and expect that: (H2a) Religious practices and intrinsic-personal religiosity will be positively related to everyday cosmopolitan orientations of Muslim and Christian youth.
Religious identification. 1
We follow previous research in examining religion as a grounding for a social identification (Ysseldyk et al., 2010). Hence, it is constructed through social interactions and its embeddedness in the social context. Religious identification is often seen as a dimension of religiosity. However, researchers argue that it is primarily defined by internal or self-categorization and external or other-identification (i.e. how somebody is perceived by others in the social environment) (Jenkins, 2014) of an individual with a religious category rather than effective religiosity (Torrekens and Jacobs, 2016; Wolf, 2005).
Ribberink et al. (2017) argue that Muslim identity is constructed through the negotiation with non-Muslims within a broader and predominantly secular (or Christian), West European context, and is therefore ‘made in Europe’. As previously discussed, Islam is often presented as less compatible with so-called ‘western values’. Thus, it is to be expected that the problematization of their religion affects the self-understanding of Muslims. One recurrent finding is the emergence of a salient Muslim identity (Fleischmann and Phalet, 2018; Maxwell and Bleich, 2014; Torrekens and Jacobs, 2016). Muslims tend to emphasize their religious identity as a strategy of revalorizing their stigmatized identity (Lamont et al., 2002). As Muslims are often ‘held accountable’ – meaning they have to account not only for themselves, but also for what other Muslims do or say – this oppositional or defensive identity can emerge as a politicized rather than merely a religious identification (Brubaker, 2013). Researchers argue, then, that a Muslim – communal – identity is based on its external stigmatization and categorization, rather than specific intrinsic religious features (Beaman, 2016).
Christian identity is becoming less salient in Europe and it is a less contested identity for non-Western Christians (as their religion is seen as part of Europe’s cultural heritage). However, scholars also have indicated how, recently, Christianity has symbolically been culturalized in West European societies (Joppke, 2018; Storm, 2011). Hervieu-Léger (2000) has identified this trend as ‘ethnic religion’. It is used in secularizing countries to identify with national traditions and an ethnic heritage of Europe, rather than with faith and intrinsic religious beliefs. Researchers argue that the Christian heritage is often mobilized as an argument against the immigration of religious minorities, the latter being closely associated with Islam (Casanova, 2007; Storm, 2011). Elaborating on these findings, we expect that: (H2b) Religious identification will be negatively related to everyday cosmopolitan orientations of Muslim and Christian youth.
Perceived discrimination of ethnic/religious groups at school
Maxwell and Bleich (2014) show that both Muslims and Christians with an immigration background experience social exclusion. However, research indicates that anti-Muslim feelings are generally more widespread than anti-immigrant resentments (Torrekens and Jacobs, 2016). Likewise, in the case of Belgium, research shows that Muslim youths are more likely than other minorities to perceive discrimination at school (Clycq et al., 2014; Fleischmann and Phalet, 2018). The headscarf ban is set in almost all schools and teachers tend to have negative attitudes towards the Islamic religiosity of students (Agirdag et al., 2012; Van Praag et al., 2016). Not only is Islam seen as a barrier for general integration into the mainstream (Foner and Alba, 2008; Torrekens and Jacobs, 2016), it is also conceived to be a hindrance to educational success (Van Praag et al., 2016). As previously discussed, experiencing social exclusion and disadvantages at school may incite defensive and oppositional orientations. Wimmer and Soehl (2014) argue that the inability of immigrants for blurring symbolic boundaries towards others is associated with discrimination, rather than with their cultural or religious values. In this study, we will focus on perceived discrimination by teachers, as the teacher–student relation is one of the most important factors in students’ trajectories (Nouwen and Clycq, 2019). We can expect that: (H2c) Perceived discrimination will be negatively related to everyday cosmopolitan orientations. This will be mostly the case for Muslims, as they are more likely to perceive discrimination.
Method
Data
We used data collected in February and March 2017 from 1039 students in the 5th and 6th year from 17 secondary education schools in Antwerp, Belgium. A questionnaire was filled out in class with one or two researchers present; it was administered in Dutch and took approximately 40 minutes. The questionnaires were anonymous and analysed in confidentiality. The participating schools varied according to educational track (academic track [ASO], vocational track [BSO] and technical track [TSO]) and educational network (four belonged to the state network, 11 to the Catholic network and two to the city network). They varied according to ethnic composition, which is measured by the Flemish Ministry of Education by the home language of the students. Three schools had over 60% of students with another home language next to Dutch, 10 schools had 30–60% of students and four schools had fewer than 30% of students (AGODI, 2017).
The total sample consisted of Muslims (n = 496; 47.7%), Christians (n = 225; 21.7%), non-religious (n = 248; 23.9%) and other religious youth (n = 70; 6.7%). The latter will not be included in the comparative analysis due to the small numbers of respondents. In a second phase, we conducted a subset analysis on the sample of Muslim and Christian youth. The sample of Muslim youth consisted mostly of second generation migrants (n = 379; 82.2%). The sample of Christian youth 2 consisted of 34.5% (n = 76) Belgian ‘origin’ students, 43.6% (n = 96) second generation migrants and 21.8% (n = 48) first generation migrants.
Research design
First, we examined to what extent Muslim, Christian and non-religious youth differ in their everyday cosmopolitan orientations. We conducted a one-way ANOVA analysis with a post-hoc Bonferroni test and discussed the effect sizes (Cohen’s d) (using SPSS statistics 24).
Second, we examined the effects of religiosity, religious identification and perceived discrimination (and their interactions) on everyday cosmopolitanism for the subsets of Muslim (n = 496) and Christian (n = 225) youth. We did not include non-religious students, as non-religious practices and value systems were not measured in our study. We conducted a multilevel regression analysis as the data consist of a clustered sample of students nested within schools (using MLwiN, two level procedure). Unconditional models were estimated to determine the amount of variance that occurs on the school level for everyday cosmopolitanism. For the subset of Muslim students, there was no variance situated on the school level. For the Christian students, there was only 9.6% of the variance situated on the school level. This is in line with other studies, indicating that most of the variation occurs within schools and between pupils (Teddlie and Reynolds, 2000). We do belief it is important to report a multilevel analysis to account for the nested structure of the data. Further, we estimated random intercept models to explore the individual-level effects. We controlled for gender, educational track, migration status and the educational level of the parents. The metric predictors are grand mean centred and unstandardized effects are reported in the tables. Tests to see if the data met the assumption of collinearity indicated that multicollinearity was not a concern for the models.
Measures
The dependent variable, Everyday cosmopolitan orientations, was measured by a five-point scale (‘strongly disagree’ to ‘strongly agree’) based on five items (based on Pichler, 2009; Saran and Kalliny, 2012): ‘Above all, I see myself as world citizen’, ‘I believe that humans all over the world have a lot in common’, ‘I believe I respect the culture of others as much as my culture’, ‘I believe it is our duty to be solidary towards fellow citizens, whatever their ethnicity and religion’ and ‘I believe every human has the right to be respected’. Answers to the five categories were averaged. This component emerged from principal component analysis on 12 items. The scale acquired an adequate Cronbach alpha of 0.70. Previous research (Keating, 2016; Pichler, 2009) treated the identity and attitude dimensions as separate variables, both entailing various items. Our measurement consists both dimensions as (1) we were limited in the number of items measuring cosmopolitan identity, (2) the literature suggests a close and meaningful association between both dimensions and (3) PCA analysis indicated strong associations between the items.
We include several independents variables. First, Religious affiliation was measured by the question ‘to which religious tradition or belief system do you belong?’ Eight response categories were reduced to four: ‘Christianity’, ‘Islam’, ‘Non religious’ and ‘other’. Second, Religiosity of the respondents was measured by three dimensions: (1) Prayer that consisted of two categories: those who pray and those who do not pray. (2) Church/mosque attendance consisted of two categories: those who go to the mosque/church and those who do not go. (3) Intrinsic-personal religiosity was measured by a five-point scale based on four items (based on Ghorpade et al., 2006; Grundel and Maliepaard, 2012; Saroglou and Mathijsen, 2007): ‘I see myself as a religious person, even when I do not strongly practise my religion (by e.g. attending religious services, praying, etc.)’, ‘I feel proud when I think of my faith’, 3 ‘when I am worried about something or I feel nervous, my faith helps me to calm down’ and ‘my faith brings meaning to my life’. The scale acquired a Cronbach alpha of 0.79. Third, Religious identification was measured by the indication of respondents of how strongly they identified as Muslim or Christian on a five-point scale (Fleischmann and Phalet, 2018). Fourth, Perceived discrimination (of ethnic or religious groups) at school was measured by a five-point scale based on four items (Fleischmann and Phalet, 2018): ‘Most teachers favour students of their own ethnic or religious group’, ‘students are discriminated by some teachers, because of their ethnicity or religion’, ‘most teachers consider members of other ethnic groups, unfairly, as troublemakers’ and ‘not everyone is treated equally at school’. The scale acquired a Cronbach alpha of 0.78.
Lastly, we controlled for the migration status of the respondents. Respondents are defined as ‘1st generation’ when they came to Belgium after the age of 15 or between the age of six and 15. They are defined as ‘2nd generation’ when they came to Belgium before the age of six or when they are born in Belgium, but one or both of the parents are not. They are defined as having no migration status when they are born in Belgium, as well as their parents. The latter category is adapted as the reference category for the Christian sample, while for the Muslim sample the 1st generation is selected. In addition, we controlled for gender (female as reference category), educational track (academic ‘ASO’, vocational ‘BSO’ and technical ‘TSO’, the first being the reference category) and the educational level of the mother and father (low, middle, high and other, the first being the reference category). Table 1 presents the descriptive statistics for the respondents.
Descriptive statistics for the independent variables: frequencies (%), means and standard deviations.
Results
Everyday cosmopolitanism among Muslim, Christian and non-religious youth
The mean levels on everyday cosmopolitanism for all the groups suggest that Muslim students have a slightly higher level of cosmopolitan orientations (M = 4.1; SD = 0.61; n = 491) than Christians (M = 3.96; SD = 0.65; n = 224) and non-religious students (M = 4.0; SD = 0.59; n = 246). Christian students have slightly lower levels of cosmopolitanism than Muslim and non-religious students. The average scores of the three groups suggest that, on average, the respondents agree with orienting as cosmopolitan.
There is a statistically significant difference between the groups as determined by one-way ANOVA (F(3,882) = 4.923; p < 0.009). Post-hoc testing reveals that Muslim students are significantly more cosmopolitan than Christians (diff = −0.133; p < 0.05). The Cohen’s d (= 0.21) shows a rather small effect size. In line with our first expectations, there is no significant difference between Christian and non-religious students and Muslim and non-religious students regarding their level of everyday cosmopolitan orientations.
The effects of religiosity, religious identification and perceived discrimination on everyday cosmopolitanism for Muslim and Christian youth
In a second phase, we aim to find out the effects of religiosity, religious identification and perceived discrimination on everyday cosmopolitanism for the subsets of Muslim and Christian pupils. It must be noted that prior to these analysis, we tested the effects of the main variables and religious affiliation in a joint analysis. In this model, the effect of religious affiliation was significant and showed that Muslims are more cosmopolitan than Christians (b = 0.131; p < 0.01). Interaction effects between religious affiliation and the main variables were not significant, meaning that the difference on cosmopolitan orientations between Christians and Muslims could not be explained by religious factors. Therefore, we decided to further examine the effects of the main variables within the groups, and argue that a subset analysis will deepen our understanding of the questions at hand.
The results for Muslim youth (Table 2) show that an intrinsic-personal religiosity has a positive effect on everyday cosmopolitanism (b = 0.137; p < 0.05). Contrastingly, there is no significant effect of prayer and mosque attendance. Further, regarding the effect of religious identification for Muslims, there is a negative effect (b = −0.111; p < 0.01). Muslim students who identify more strongly as Muslim will have lower cosmopolitan orientations. In addition, the effect of perceiving discrimination of ethnic or religious groups at school has a negative effect (b = −0.092; p < 0.01). Interactions effects between the main variables are not significant.
Multilevel linear regression on everyday cosmopolitanism among Muslim youth: unstandardized coefficients (b) and standard errors (SEs).
p < 0.001; ** p < 0.01; * p < 0.05.
The control effects indicate that Muslim boys display a lower level of cosmopolitanism than Muslim girls (b = −0.203; p < 0.01). Students in BSO do significantly have lower cosmopolitan orientations than students in ASO (b = −0.320; p < 0.001), while students in TSO do not differ from ASO students. Regarding the effect of migration status, there is no difference between second and first generation of migration. Having a mother with a middle educational level will lead to higher cosmopolitan orientations than having a mother with a low educational level (b = 0.302; p < 0.001). Having a father with a high educational level will lead to lower cosmopolitan orientations than having a father with a low educational level (b = −0.300; p < 0.001).
The results for Christian students (Table 3, model 1) indicate that the main variables do not influence their everyday cosmopolitanism. In model 2, the negative effect of perceived discrimination (b = −0.223; p < 0.01) is moderated by going to church (b = 0.333; p < 0.01). Respondents who do not go to church experience a negative effect of discrimination on their cosmopolitan orientations.
Multilevel linear regression on everyday cosmopolitanism among Christian youth: unstandardized coefficients (b) and standard errors (SEs).
p < 0.001; ** p < 0.01; * p < 0.05.
The control effects (Table 3, model 1) show that there is no effect for gender on cosmopolitan orientations. Students in TSO (b = −0.230; p < 0.05) and BSO (b = −0.463; p < 0.05) do significantly display lower cosmopolitan orientations than students in ASO. Further, there is no significant difference in cosmopolitanism between Christians with no migration background and students from a first or second generation of migration. Students with a mother with a high educational level have higher cosmopolitan orientations than students with a mother with a low educational level (b = 0.345; p < 0.05). Students with a father with a middle educational level show lower cosmopolitanism, than students with a father with a low educational level (b = −0.434; p < 0.05).
Discussion and conclusion
This study contributes to the understanding of cosmopolitanism ‘from below’ by focusing on everyday cosmopolitan orientations, such as a world citizen identity and openness to cultural differences, among youth in the super-diverse city of Antwerp. In this context, young people regularly engage with people from diverse ethno-cultural and religious backgrounds. Therefore, these teens act upon various cultural inputs from diverse peer networks, families, and local urban and global popular culture (Turner et al., 2014). Our study contributes to the scarce knowledge on how this younger generation, socialized in this emerging super-diverse urban context, construct and constantly rework their multiple identities and cultural relations. In addition, it contributes to the scarce empirical research on the relation between religion and everyday cosmopolitanism. It is often argued that religious doctrines impose closure on social groups, compromising principles of equal citizenship and disallowing cross-cultural and cross-religious connectivities. Indeed, some scholars and public opinion leaders associate cosmopolitanism and liberalism with secularism, while religious values are perceived as incompatible with cosmopolitan beliefs. While secular boundaries are salient in Western Europe, and religious practices and beliefs in general are seen as illegitimate (Trittler, 2019), particularly Muslims are depicted as the cultural other in European countries (Cesari, 2004; Zemni, 2011).
However, our study shows that there is no difference between religious and non-religious youth regarding their everyday cosmopolitan orientations. While Muslims tend to be more cosmopolitan than Christians are, the difference (effect size) is rather small. Generally, our study suggests that Muslim, Christian and non-religious youth identify as world citizens and express openness to cultural differences to the same degree. In comparison to other Belgian studies that are concerned with (sub-)national and supra-national identities among youth (see e.g. Agirdag et al., 2016; Fleischmann and Phalet, 2018; Saroglou and Mathijsen, 2007), the mean levels on everyday cosmopolitanism in this study are high. Despite the differences in questionnaires between the studies, it might be concluded, then, that an everyday cosmopolitan identity (and associated attitudes) seems to be a viable collective identity for youth who negotiate multiple identity belongings within a super-diverse urban context. In addition, our study shows that young people can develop cosmopolitan attitudes, by studying orientations on how to engage with diversity on a daily basis, going beyond an elitist understanding of cosmopolitanism. Lastly, it must be argued that our study provides a perspective on the compatibility of religion, and more particularly Islam, with cosmopolitan orientations, something that has often been questioned.
Further, we aimed to deepen our understanding of the role of religion by examining the effects of religiosity (religious practices and intrinsic-personal religiosity), religious identification and perceived discrimination of ethnic/religious groups in school on everyday cosmopolitan orientations for Muslim and Christian youth. Our study shows that for Muslim youth an intrinsic-personal religiosity is positively associated with everyday cosmopolitanism, while religious practices did not have an effect. Scholars such as Roy (2014) have argued that the individualization and essentialist perception of religion as separate from culture fragments religious authority and facilitates religious extremism. However, such perspectives run the risk of overlooking ongoing processes of individualization by Muslims as part of everyday practices for negotiating and balancing multiple identity belongings and various cultural expectations in Western Europe (Jeldtoft, 2011). Indeed, research has shown that individualization is an active strategy for managing stigma related to Muslims’ religion and for equalizing and adapting themselves to mainstream society (Beaman, 2016; Killian, 2007; Synnes, 2018). In addition, our results show that religious identification and perceived discrimination are negatively associated with everyday cosmopolitanism for Muslims. As discussed, the problematization of Muslim identity in Western Europe can result in a defensive and politicized identity (Fleischmann and Phalet, 2018; Foner and Alba, 2008), and thus it is less associated with specific intrinsic religious features (Beaman, 2016). A stigmatized religious identity and discrimination can reinforce separation and intolerance (Cesari, 2004) and thus hinder a cosmopolitan openness. As argued by Werbner (2015), a cosmopolitan identity remains fragile, as institutional racism, xenophobia, discrimination and social exclusion hampers a cosmopolitan creativity. Our study demonstrates that religious beliefs can be used for Muslim youth to emphasize an openness to cultural differences, yet policy makers and teachers should be aware of the (negative) impact of stigmatization and discrimination related to Muslims’ religiosity. These results thus not only show that religion is a significant factor in predicting cosmopolitanism among Muslim youth, but also suggest that scholars and policy makers should discuss how religiosity can be used as an asset for emphasizing cosmopolitan orientations.
Contrary to the findings on our Muslim sample, our results demonstrate that cosmopolitan orientations among Christians cannot be explained by religious factors. This can be due to, firstly, the declining salience of Christian identity, practices and beliefs in Western Europe (Molteni and Biolcati, 2018; Storm, 2011). Secondly, Christian practices still have symbolic resonance in Western Europe and are seen as part of Europe’s cultural heritage (Fleischmann and Phalet, 2018; Storm, 2011). Therefore, Christian beliefs and identity are less contested and problematized, while for Muslims there is a pressing need of negotiating and using their religion for blurring symbolic boundaries with others. We did expect a negative effect of Christian identification, as it is increasingly used in Western Europe in a ‘culturalized’ or ‘ethnic’ way. We did not find this result. Interestingly, we did find that a negative effect of perceived discrimination was moderated by going to church. A negative association between perceiving discrimination and cosmopolitan orientations is therefore only true for those who do not go to church. Further research may explore what other variables can encourage or impede the cosmopolitan views of Christian youth.
Finally, we note some limitations of this study. First, it can be argued that the relation we have studied – the impact of religiosity on cosmopolitan attitudes – can be reversed. Yet, previous research has already examined extensively the influence of out-group perceptions and attitudes of non-Muslims on Muslims’ religiosity and identity (e.g. Maxwell and Bleich, 2014), showing that negative out-group views lead to a higher religiosity and salient religious identification for Muslims. Hence, we considered it interesting to research how Muslims’ religiosity and identity in their turn influence their out-group perceptions. Yet, more in-depth investigation of the different mechanisms driving the relationship between religiosity and cosmopolitan attitudes is needed. Second, this multilevel study did not find significant variance on the school level. However, based on previous research, we would expect that there are significant differences on the school level concerning the out-group views and relations of students (Vervaet et al., 2016). Thus, further research with a large sample of schools is needed (our sample size was limited to 17 schools). Third, our sample of Muslims consisted of predominantly North African (mostly Moroccans) youth of the second generation. As it did not strongly include variation on ethnic background and migration status, it is difficult to assess the impact of both variables on cosmopolitanism. Further research can explore if there is a significant impact, by adjusting the sample strategy. Lastly, our study focused on youth in the context of a super-diverse city. Further research can explore these new orientation processes and the generalization of these results to other institutional contexts and urban or rural contexts.
Footnotes
Funding
This work was supported by Universiteit Antwerpen (grant number DOCPRO 33174). No potential conflict of interest was reported by the author(s).
