An intense political and intellectual debate is taking place in Europe regarding the pace at which recent immigrants are (or are not) integrating to European cultural values. Huntington's notion of clash of civilizations as well as Sen’s analysis of multi-cultural societies, for instance, prominently feature in many discussions in the media.

In particular, the perceived slow integration of Muslim immigrants to European values –  as exemplified by (for example) the riots in Paris, the violent reactions to the publications of the vignettes on the prophet Mohammed, and of course the terrorist attacks in Madrid and London – is at the forefront of the public debate.

It should be noted, however, that historical examples of religious groups mistakenly singled out for their ``inability to assimilate'' abound . Between 1815 and 1860, for instance, the large inflow of Catholic immigrants to the US, and their ''clannishness and separatism'' aroused intense anti-Catholic and anti-foreign resentment among large fractions of the Protestant majority.

Is this the case for Muslims in Europe? Is the resentment against the Muslim immigrants in Europe nowadays also a transient reaction to the generally very resilient character of cultural and especially religious values? Or are the Muslims really resisting integration, and more successfully so than other immigrant groups?

In our research, we aim to shed some light on this issue using a unique UK data set, the Fourth National Survey of Ethnic Minorities (FNSEM), which asks a variety of questions on individuals' ethnic and religious preferences (such as importance of religion, attitudes towards inter-marriage, and relevance of ethnicity in choosing the children's school). We encode the answers to these questions to produce various measures of intensity of each individual's religious identity, distinguishing Muslims and non-Muslims. We then examine relevant differences in the determinants of religious identity across groups, to investigate whether it's true to assert that the process of cultural integration is different for the Muslims1 than for the other UK minorities in the sample2 .

First of all, there is no doubt that Muslims can be characterised by the intensity of their religious identity. Indeed, regardless of the dimension of identity we construct, the percentage of Muslims having an intense religious identity is roughly twice as much as that of non-Muslims. For instance, 79% of Muslims in the sample answer that religion is 'very important' to how they live their life, compared to only 42% of non-Muslims. Similarly, 70% of Muslims claim that they would 'mind very much' if a close relative married a white person, compared to 37% of non-Muslims. A greater resistance of Muslims to cultural integration is also signaled by the percentage of Muslims speaking English at home or with friends, which is always significantly lower than that of non-Muslims.

Muslims and non-Muslim immigrants, however, differ in terms of several demographic and socio-economic characteristics which could in principle be correlated with their different observed attitudes towards religious identity and integration, and hence explain the stronger resistance to integration for Muslim immigrants in the UK. In particular, Muslims are on average less educated than non-Muslims, with a lower household income, and with more than a double probability to be unemployed. Muslims also live in more ethnically-segregated areas, which have higher unemployment rates.

To what extent do these and other demographic and socio-economic contextual characteristics of Muslim immigrants in the UK explain their differential attachment to their religion and associated cultural traits?

Our analysis reveals that the integration pattern adopted by Muslim immigrants in the UK contains in fact several important specific aspects. For non-Muslims, a high level of education (being highly educated in Britain) and a high level of  job qualification (being a manager) are among the most important factors that decrease their sense of identity. For Muslims, instead, education does not seem to have any effect on the attenuation of their identity and, on the contrary, being a manager as well as having a high income seems to strengthen their religious faith. Consistently, Muslims living in areas with lower unemployment rates seem to show a higher identity. The picture that emerges from the data is that although Muslims are less likely to become managers than non-Muslims and are poorer, those who succeed show a stronger religious identity.

Most importantly, even after conditioning on our rich set of individual and contextual demographic and socio-economic characteristics, the speed of cultural integration is slower for the Muslims than for the non-Muslims. For non-Muslims, the longer they have been in the UK, the more attenuate is the attachment to their culture of origin, whereas for Muslims, the number of years since their arrival in the UK does not show any significant effect on their inclination to assimilate.

Figure 1 : Integration patterns over time

Figure 1 reports the marginal effects, i.e. the changes in the probability of having a strong religious identity following a one-year increase in the time spent in the UK. They decline for both Muslim and non-Muslim with time spent in the UK, but the average effect over time is less than 1% for Muslim and more than 3% for non-Muslims. Figure 1 also shows that being born in the UK has an effect of decreasing the intensity of religious identity for all immigrants, but this effect is more than two times greater for non-Muslims than for Muslims.

According to our estimates, when the effects of our large set of individual and contextual characteristics have been accounted for, a Muslim born in the UK who has spent more than 50 years there has on average the same probability of having a strong religious identity as a first generation non-Muslim who has been in the UK for fewer than 20 years. Second generation Muslims never achieve the (lower) level of probability of having a strong religious identity of second generation non-Muslims at any point in time.
Interesting (and perhaps surprising) results are also obtained with regards to the dependence of identity on the ethnic/religious composition of the neighbourhood where immigrants live. We find that living in a more integrated neighbourhood (with a lower percentage of own ethnic/religious minority group) and speaking English at work, which signals a mixed working environment, are both associated with a higher sense of identity. This integration pattern is common to both Muslims and non-Muslims, but it appears more marked for Muslims. It suggests that intense forms of identities appear to be formed in social contexts in which the minority ethnic/religious trait is more exposed to the interaction with the majority norm of behaviour, perhaps as a reaction to the norms of the majority or its demand for integration of immigrant populations.

It should be noted in this respect that episodes of harassment and discrimination tend to have relatively higher frequency in less segregated neighborhoods. In our data, the frequency of serious episodes of racial harassment (e.g. attacks) is more than double in mixed than in segregated neighbourhoods (19% and 9% respectively). Consistently, instances of discrimination suffered by individuals in the sample (on average by ethnic group) have a positive effect on identity, stronger for Muslims than for non-Muslims.

We conclude that the evidence about Muslim integration in the UK speaks in favour of its specificity. But the specific pattern of integration we document in our sample for Muslim immigrants in the UK stands somewhat in contrast to the intellectual foundation of most immigration policy in Europe, with its focus on economic achievement and geographic integration: a stronger resistance to cultural integration on the part of Muslim immigrants (in the form of intense ethnic/religious identity) is relatively more prevalent between the more successful and better educated immigrants, and in the less segregated and better-off neighbourhoods.

We should be very careful before drawing policy implications from our analysis. First of all, we only have data for the mid-90s. Most importantly, we are simply documenting differences in behaviour across groups rather than causal relationships. Our analysis of integration should in any case obviously not be interpreted to favour shying away from economic support to the immigrants and/or to favour the formation of ghettoes. It does, however, cast some doubts on simple-minded policies favouring geographic and social integration (as well as school integration, for instance) under the optimistic belief that geographic and social integration breed cultural integration. It is possibly true that "when we leave work, most of us leave multi-ethnic Britain behind" (Trevor Phillips – head of the Commission on Racial Equality in the UK). But our analysis suggests that this may not be the real problem. The presence of unresolved forms of ethnic and religious discrimination, which appear predominant in culturally diverse and mixed environments, and perhaps more acute towards successful members of the minorities, may actually be the initial locus to look at for the design of a sensible integration policy .


1 In our sample, Muslims are mostly Pakistani, Bangladeshi, and Indians.

2 In our sample, non-Muslim minorities include, for example, Carribean, Chinese, and non-Muslim Indians.