Muslim-immigrant & Refugee Integration in Europe Cultural and - - PDF document

Muslim-immigrant & Refugee Integration in Europe

Cultural and Biological Racism, Muslim immigrants, and refugee exclusion in the context

• f the European refugee crisis.

Abstract This study sets the focus on Muslim immigrant exclusion and provides deeper insight to what are the determinants of these exclusionary sentiments in the context of Europe’s refugee crisis. Many scholars theorize that Muslim exclusion in Europe is based on Cultural Racism –i.e. immigrants are excluded based on culture, language or religion differences. As opposed to the US in which Biological Racism – i.e. the individual is reduced to the phenotypical characteristic of skin color. This study takes on a “bright and blurry” boundary for immigrant integration approach in order to assess the Cultural/Biological Racism theory. Contrary to what previous scholars have theorized, the findings suggest that race is a brighter boundary than language or religion for

• Muslims. However, when compared to other immigrants, Muslims show an added exclusionary

effect based on religion. Hence, we find evidence for both Cultural and Biological Racism. In light

• f these results, Muslim exclusion is better understood adopting an intersectionality framework in

which both, religion and race interact – among other factors. These findings prove to be relevant to Muslim immigrant integration in light of the current influx of refugees into Europe. Future steps consist of finding detailed variation across different countries in Europe. Introduction Muslim discrimination is a highly complex and understudied field within the social sciences. There is little agreement throughout the literature that touches upon this issue. Most studies that tackle the Muslim question assume discrimination is based on cultural traits such as language, religion,

• r cultural values. Others view Muslim discrimination as a consequence of resource competition.

Very few studies argue race is a source of discrimination for Muslims. This is the first study to address Muslim discrimination taking on an eclectic view of all three sources of discrimination: race, culture, and resources competition. The focus is set on Europe in light of the massive influx

• f refugees of Muslim origin.

The recent trends in European immigration during the last decade are unprecedented in migration

• history. In 2011 a number of contentious events of revolutionary nature emerged in several

countries south and east of the Mediterranean. This phenomenon, known by many as the Arab

Spring, spread through countries like Tunisia, Lybia, Egypt, Syria, Argelia, and Morocco, creating very unstable political and economic situations. Economic crisis, civil unrest, toppling governments, and in many cases civil war, compelled many to migrate to Europe (De Bel-Air 2016). Since 2011, and especially after 2014 on account of the Syrian civil war, the number of migrants traversing the Mediterranean routes towards Europe has increased many-fold. As can be seen in figure 1, the numbers have been increasing steadily since 2011, reaching in 2015 a total of 1,012,957. Parallel to this high number of immigrants flowing to Europe, the concerns of EU citizens for immigrants has become the most important one, even surpassing the economic situation –this is especially noteworthy, as Europe still finds itself in the midst of an economic recession. As can be seen in figure 2, in 2014 33% of EU citizens responded the highest concern was the item Economic Situation, followed by Unemployement (29%), The state of Member States’ public finances (25%), and in fourth place Immigration (23%). In 2015, these items were still in the top four highest concerns, however; Immigration escalated to the first position with 38% of the EU citizens indicating it is the most pressing concern for Europe. Additionally, in fifth place we find Terrorism is an important concern for Europeans. Source: Eurobarometer 2014 & 2015.

5 10 15 20 25 30 35 40 None Other Pensions The environment Taxation Energy supply Crime Climate change Don't know EU's influence in the world Rising prices/inflation/cost of living Terrorism Immigration The state of Member States' public finances Unemployment Economic situation

Percent

Figure 2. Most important concerns for EU citizens , 2014-2015.

2015 2014

Over the last 20 years exclusionary feelings towards Muslim immigrants has increased drastically (Street et al. 2008).This anti-Muslim sentiment coexists with the fact that a large number of these refugees come from countries of Arabic culture, and are Muslims. However, very little is known about the nature of this anti-Muslim sentiment. This study sets the focus on Muslim immigrant and refugee exclusion and provides deeper insight to what are the determinants of these exclusionary sentiments. To do so, the study adopts Richard Alba’s “Bright and Blurry” theoretical framework to European Social Survey Data. Theoretical background The literature on the sources of Muslim discrimination is very divided. Part of the reason is the absence of acknowledgment of the complexity of Muslim discrimination. The term “Muslim” is highly conflated with racial, cultural, and religious differences with Europeans. Some scholars address the issue by focusing on the religious aspect of Muslim discrimination and dub it Islamophobia (Hopkins and Kahani-Hopkins 2006). Some argue that the term Muslimophobia is more appropriate due discrimination towards Muslims as people, rather than Islam the religion (Cheng 2015). Others make the case for a racialization of Muslims on account of recent political contexts such as the rise in terrorist attacks in the name of Islam (Sander 2006). This study adopts a more eclectic vision of Muslim discrimination by conceptualizing the category “Muslim” as one that includes racial, cultural, and religious differences at once. There are three main theoretical bodies that explain Muslim exclusion: Ethnic competition theory, Intergroup contact theory, and Cultural Racism. Perceived threats: Ethnic competition theory. Ethnic competition theory posits that the larger a migrant group is, the higher the level of perceived threat on behalf of the host population. In turn, this leads to higher levels of anti-Muslim sentiment (Connor 2010). Perceived threat is operationalized in the literature in the form of perceived cultural threat or perceived economic threat. Hence, in the case of a cultural threat, the habitual interpretation is that if the “outgroup” is large enough, it will undermine cultural practices and norms of the host society (Blake 2014). In the case of perceived economic threat, the prevailing idea is that the larger the number of “outgroup” members, the higher the competition over the scarce resources. There is a number of studies that have found perceived economic and cultural threats –at the individual and country level- were significant predictors of both immigrant and Muslim exclusion (Mclaren 2003; Quillian 1995). The Contact Hypothesis: Intergroup contact theory.

Intergroup contact theory states that larger groups of immigrants induce higher probabilities that the host population will have contact with the immigrants and minorities. As a consequence, any prejudice the host population may have is refuted or reduced due to this exposure to “reality” (Allport 1979). In line with this theory, anti-Muslim sentiment will be reduced in the presence of higher numbers of Muslim immigrants. Results from previous research suggest high numbers of Muslim immigrants induce both intergroup friendships as well as perceptions of ethnic threat (Savelkoul et al. 2011). Bright and Blurry Boundaries In order to gain deeper insight on Muslim exclusionary sentiment the study adopts Richard Alba’s approach to exclusionary feelings based on Frederik Barth’s concept of social boundaries. In the seminal article Bright vs. blurred boundaries: Second-generation assimilation and exclusion in France, Germany, and the United States (2005), Alba argues social boundaries are specific cultural elements that determine the boundary between majority and minority. However, these boundaries are not independent from each other, and coexist in the realm of exclusionary feelings. Alba emphasizes that the most useful approach in studying social boundaries is by conceiving social boundaries in terms of degree. Bright boundaries are those for which distinction is clear and unambiguous, so that individuals easily know which side of the boundary they are on. Other boundaries are blurry, meaning the location of whether an individual is majority or minority is

• ambiguous. Whether a boundary is bright or blurry is determined on both sides of the boundaries,

which is to say; both minority and majority have an active role in defining these boundaries (Allport 1979). Alba argues there are four crucial boundaries in determining whether an individual is majority or minority –citizenship, religion, language, and race. Given in this study I will focus

• n the construction of boundaries by the majority –i.e. European citizen- I will adopt the same

boundaries Alba argues about, with the exception of citizenship. Cultural Racism vs Biological Racism In a theoretical study on the role of religion in the assimilation process, Alba suggests social boundaries may be brighter or blurrier in different manners for the US and Europe (Foner and Alba 2008). Alba argues there is somewhat of a consensus on the positive role of religion for the incorporation of immigrants into the US. Religious membership offers a community that may be used as a network hub of mutual support with co-ethnics and native-born. In Europe, however; this is not the case. Religion, especially Islam, is analyzed as a barrier to integration and a source of conflict with mainstream institutions and cultural practices. Alba and Foner (2008) argue the reasons behind why religion might be such a bright boundary for immigrants –and more specifically Muslims- are several. First, majority populations in Europe tend to be mainly secular, and are suspicious of potential claims based on religion and its requirements (Meer and Modood 2009). Hence, immigrants who are thought to be intense believers will find a bright boundary during their assimilation process. Second, although Europe tends to be secular, its institutions and national identities have been built on Christianity and remain

anchored to it. Alba and Foner argue this is an important contrast with the US, for in the US, immigrant groups and native population share the same religious core –i.e. Christianity. A big proportion of immigrants to Europe are Muslim and, hence; do not share commonalities with Christian-based religions. Third, popular belief and a small group of scholars support the idea that Muslim cultural values are a potential threat to European cultural values. These particular points

• f view flag gender relations and cultural practices particular to Muslims – e.g. the headscarf or

the jilbab- as the main conflicting points with European culture. According to previous research, perceived cultural threats operate at both the individual level –in the form of everyday interaction with minorities- and at the country level –i.e. the individual’s concerns revolve around the “greater good” of its country (Mclaren 2003). Overall, Alba and Foner argue Europe shows what they call Cultural Racism; culture or religion are essentialized to the point that individuals are reduced to a certain religion or cultural trait and are seen inferior because of it (Algan 2012). As opposed to the US in which Biological Racism – i.e. the individual is reduced to the phenotypical characteristic

• f skin color- seems to be prevalent, in Europe Cultural Racism seems to be the main driver of

immigrant and Muslim exclusionary sentiment (Foner and Alba 2008). Refugees A recent study in Science was the first to quantitatively relate the anti-Muslim sentiment in Europe to the recent refugee inflow (Bansak, Hainmueller, and Hangartner 2016). Bansak et al. found that asylum seekers who have higher employability, have more consistent asylum testimonies and severe vulnerabilities, and are Christian rather than Muslim received the greatest public support. Hence, we arrive at the set of research questions: 1) Is Muslim discrimination a function of ethnic competition, intergroup contact, or racism? 2) Is European anti-Muslim sentiment driven by Cultural Racism or Biological Racism? What social boundaries are bright or blurry for European citizens regarding Muslim immigrants? 3) Is the context-specific factor of terrorism a predictor of high anti-Muslim sentiment? Is the context specific factor of refugee crisis a predictor low anti-Muslim sentitment? Data, sample, and analytical approach

For this study I use round 7 of the European Social Survey, for which the field work was carried

• ut during 2015. Round 7 is particularly interesting as it is a specialized module developed to

capture public opinion on immigration, discrimination, and prejudice. In this particular round 20 countries participated in the survey, and provides a total of 31,561 respondents. Given the interest resides in describing bright and blurry boundaries set by the majority population, it is important to exclude from the sample those members of a minority, those who were born outside of Europe, and second generations. Missing data is substantial in this dataset, reaching an overall level of approximately 65% of complete cases. Following the literature on best practices, the prior steps to analysis estimated 20 implicates of missing values following Multivariate Imputation by Chained Equations techniques (Dong and Peng 2013). After multiple imputation and exclusion of cases that are not of interest to the study, the total number of respondents is 24669. Across country this number is considerably balanced, as there is a minimum of 834 respondents per group and a maximum of 2320 respondents per group –i.e. country. Dependent variables Muslim exclusion. The main dependent variable is dichotomous and it measures weather European citizens are willing to allow Muslim migrants into their country or not. The question of which it stems reads as follows: To what extent you think [country] should allow Muslims from other countries to come and live in [country]? It is measured on a 1-4 scale in which 1 is “Allow many Muslims” , 2 is “Allow some Muslims”, 3 is “Allow few Muslims” and 4 is “Allow none”. This variable is recoded into a dichotomous variable which separates whether the respondent is willing to allow Muslims into the country or not. Exclusion of immigrants of different ethnicity/race. An additional variable is introduced in the study regarding immigrant exclusion without

• specification. To what extent you think [country] should allow immigrants of different

ethnicity/race from other countries to come and live in [country]? It is measured on a 1-4 scale in which 1 is “Allow many” , 2 is “Allow some ”, 3 is “Allow few ” and 4 is “Allow none”. Individual Level Independent variables Bright and Blurry Boundaries As an indicator of specific social boundaries the study uses a battery question the ESS has which asks for certain qualifications towards accepting immigrants into their country. The question reads as follows: “Please tell me how important you think each of these things should be in deciding whether someone born, brought up and living outside [country] should be able to come and live here”. As a follow up they ask about specific aspects. As an indicator of whether religion is a bright or blurry boundary I use the question that asks about Christianity as a qualification: How

important should it be for them to come from a Christian background? As an indicator of whether race is a bright or blurry boundary I use the question that asks about whether being white is important or not for acceptance: How important should it be for them to be white? As an indicator

• f whether language is a bright or blurry boundary I use the question in the ESS that asks

specifically about the importance of language for acceptance: How important should it be for them to be able to speak [country’s official language(s)]? These three variables are measured on a scale

• f 0-10 where 0 is Extremely unimportant and 10 is Extremely important.

In Favor of refugees The ESS asks a specific question that serves as an indicator to refugee acceptance. It asks How much do you agree with the following statement? Government of [country] should be generous judging applications for refugee status. It is an ordinal variable in which 1 is “Disagree strongly” and 5 is “Agree strongly”. Individual level Controls Economic threat: Unemployment In order to capture whether the respondents feel economically threatened at an individual level, I establish a comparison using a dummy variable between the employed and unemployed in the last 12 months. Given the current situation in Europe, this is a good indicator towards individual economic threat. Furthermore, this variable has been used in this exact manner in the previous research (Mclaren 2003). Intergroup contact: Contact with people of other race or ethnicity In order to address intergroup contact theory at the individual level, the models contain a variable that gauges the frequency a respondent has with individuals of different race or ethnicity. Demographic controls As controls I will use basic demographic characteristics such as Age, Gender, Household income, and Years of education. Following the contact hypothesis, which states contact lessens the degree

• f prejudice one may have on a minority, I introduce the amount of contact with minority an

individual has (Binder et al. 2009). It is important to notice this variable has a limitation as to the causal ordering. It is possible people who are less likely to exclude Muslim immigrants tend to have more contact with with minorities (Enos 2014). Country level controls Economic threat: Unemployment rate. Previous research focusing on anti-immigrant sentiment has consistently found evidence of a strong association between perceived realistic threat and anti-immigrant sentiment (Mclaren 2003). Realistic threat theory stems off of the idea that members of the majority group perceive they are entitled to specific resources. When these resources are scarce and the competition

between minority and majority increases, the dominant group is likely to react with exclusionary feelings (Strabac and Listhaug 2008). Europe is pulling out of a long lasting recession in which levels of unemployment have reached unprecedented levels. This is especially notorious for countries in southern Europe. Spain, Greece, and Italy show unemployment rates of approximately 7- 8% in 2007. In 2014, Spain had an unemployment rate of 24.9%, Greece reached 26% and Italy increased its unemployment to 13%. Intergroup contact: Percent of Muslims in the country As a control for minority of interest at the country level, the multilevel models estimate the effects

• f percent of Muslims in the country in 2014 as a level 2 variable. This percentage is provided by

the PEW- Templeton Global Religious Futures Project. Number of Terrorist Attacks It is possible the number of terrorist attacks in the country plays a significant role in public opinion. The Global Terrorism Database provides data on the number of terrorist attacks that have taken place during the last 20 years for each country. It is worth mentioning most of these attacks are of unknown claim. This means the data does not register what terrorist groups these attacks are attributable to. Analytical approach TALK ABOUT THE LACK OF DATA ON RACE IN EUROPE AND HOW THIS APPROACH INDIRECTLY GETS AT IT As for the analytical approach, I begin by establishing a paired T-test to capture the differences in means of anti-immigrant sentiment. This will provide information on whether the anti-immigrant sentiment is larger for Muslims than for other immigrants of different race and ethnicity. The second step consists on using a random intercept multilevel logistic regression to estimate the level of association between the cited independent variables and anti-immigrant sentiment. Level 1 is set at the individual level, whereas level two is set to countries. By allowing intercepts to vary randomly (across countries), it is possible to account for variation at the country level and obtain a slightly more accurate representation of the anti-Muslim sentiment. The third part of the analysis consists of looking at individual countries and establish bright and blurry boundary differentials for each one. In order to asses these differentials, the study assesses standardized coefficients. This renders the relative size of the coefficients allowing for and indications to which qualifications for Muslim acceptance are brighter or blurrier.

Results Table 1 shows the results of the difference of means test is statistically significant for most countries, with the exception of France, Denmark, and the UK. The results reveal that anti- immigrant sentiment for Muslims is generally higher than for immigrants of different race/ethnicity, as all differences are positive. Given both variables are measured on the same scale, it is possible to directly compare the difference between the means as an indicator of social

• distance. East Europe seems to have the highest disparity when it comes to immigrant
• discrimination. This is particularly true for Poland, which shows a difference in means of 0.57,

indicating that Muslim anti-immigrant sentiment is higher. Conversely, West Europe seems to show the lowest discrimination disparities for Muslims and Immigrants of different race/ethnicity. Austria, Germany, Belgium and Denmark seem to show a thinner gap between anti-immigrant sentiment for Muslims and all immigrants. These results indicate that discrimination against Muslims is overall higher than it is for immigrants of different race and ethnicity. At this stage it seems the evidence we encounter in table could support Muslim discrimination in Europe is based on cultural racism instead of biological racism. If biological racism would be the base of exclusion, we would see no differences between these two groups. However, this is not evidence enough to conclude this.

• Table1. Proportion of people who agrees with allowing no Muslim immigrants and no immigrants
• f different race/ethnicity by country.

Allow no Muslim immigrants in country. Allow no immigrants

• f different

race/ethnicity in country Difference West Europe Austria 20.9% 13.7% 7.2% *** Belgium 20.1% 13.4% 6.7% *** Germany 8.0% 4.2% 3.7% *** France 13.2% 10.2% 2.9% *** Netherlands 13.2% 5.7% 7.5% *** East Europe Czech Rep 59.3% 27.5% 31.8% *** Hungary 56.7% 31.8% 24.9% *** Poland 32.0% 9.5% 22.5% *** North Europe Denmark 10.3% 5.7% 4.6% *** Finland 17.5% 7.9% 9.6% *** UK 17.9% 13.4% 4.5% *** Ireland 27.1% 14.7% 12.4% *** Norway 7.7% 1.2% 6.4% *** Sweden 3.6% 0.5% 3.1% *** Switzerland 13.4% 4.2% 9.1% *** Lithuania 40.6% 14.8% 25.8% *** South Europe Spain 21.7% 11.9% 9.8% *** Portugal 30.7% 15.9% 14.8% *** Slovenia 20.3% 9.7% 10.7% *** Note: *p < .05; **p < .01; ***p < .001 Paired T-test

Table 2 shows the estimates for the Multilevel Logistic regression for Muslim exclusion. In this model, the The intra-class correlation coefficient is 0.269, indicating that approximately 27% of the total variance is accounted for by countries in level 2. This is a considerable ICC, making the use of multilevel logistic regression appropriate for the analysis. Likelihood ratio tests comparing the null model with models 1 through 3 show models 1, 2 and 3 are a better fit than the actual null model.

Table 2. Multilevel Logistic Regression Estimates for the Willingness to Not Allow Muslim Immigrants in the Country (1 = Not Allow, 0 = Allow , Odds Ratio). Individual level Null Model Model 1 Model 2 Model 3 Acceptance Qualifications Language Qualification 1.077*** 1.077*** 1.078*** (0.01) (0.01) (0.01) Religious Qualification 1.085*** 1.085*** 1.085*** (0.01) (0.01) (0.01) Race Qualification 1.150*** 1.150*** 1.150*** (0.01) (0.01) (0.01) Refugee 0.567*** 0.568*** 0.567*** (0.01) (0.01) (0.01) Contact Often 0.894*** 0.894*** 0.894*** (0.01) (0.01) (0.01) Hoiusehold income 0.955*** 0.955*** 0.955*** (0.01) (0.01) (0.01) Education 0.948*** 0.947*** 0.947*** (0.00) (0.00) (0.00) Female 1.118*** 1.118*** 1.119*** (0.02) (0.02) (0.02) Age (decades) 1.083*** 1.083*** 1.083*** (0.01) (0.01) (0.01) Religiosity 0.945*** 0.945*** 0.945*** (0.01) (0.01) (0.01) Unemployed 1.056 1.057 1.057 (0.05) (0.05) (0.05) Country Level Unemployment Rate 1.056† (0.03) Number of Terrorist Attacks 0.999 (0.00) Percent Muslim 0.809*** (0.03) Constant

• 1.407***
• 0.482
• 0.004

0.195*** (0.20) (0.33) (0.21) (3.20) Random-effects Parameters Country: Identity sd(_cons) 0.887 0.599 0.638 0.382 (0.15) (0.10) (0.11) (0.07) N (Individuals) 24669 24669 24669 24669 N Groups 20 20 20 20 Exponentiated coefficients; Standard errors in parentheses † p<0.1, * p<0.05, ** p<0.01, *** p<0.001

The first interesting results are located in the acceptance qualifications. By looking at the size of the coefficients in terms of odds ratios, it is possible to see that Race qualification has the strongest

association with Muslim exclusion sentiment. In other words, the exclusionary feelings towards Muslims is most associated with the requirements that immigrants are white. A one-unit increase in Race qualification increases the odds of not allowing Muslims by 15.3 %. Language qualification increases the odds of excluding Muslims, suggesting the European citizens require Muslim immigrants to know the language of their country in order to accept their integration into society. Whether the immigrants are Christian or not is also highly associated with Muslim exclusion. More importantly, the increase in odds of Religion qualification is lower than Race qualification, suggesting race is a brighter boundary than religion and language. A very interesting finding is the incredibly high association between Refugee and Muslim

• exclusion. Those who believe the government should be generous to refugee application, are much

more unlikely to not allow Muslim immigrants. A one-unit increase in refugee acceptance decreases the odds of not allowing Muslim immigrants by 43.6%, all else equal. As for controls at the individual level, the contact with minority friends shows a negative association with Muslim exclusionary feelings. However, due to the issues with potential reverse causality, I limit my analysis to introducing it as a control. An interesting result that comes from the comparison between males and females. When asked about Muslims, women show increased

• dds of 8.8% of choosing to allow no Muslims. Possibly, women are particularly less accepting of

Muslim gender relations. The country level controls are introduced sequentially in models 1 through 3. Model 1 shows unemployment rate at the country level increases the odds of Muslim exclusion by 5.2%, all else

• equal. This contrasts with the unemployment variable at the individual level, which was

statistically insignificant. This suggests economic threat at the country level is of concern for European citizens. Terrorist attacks is not statistically significant, however; as mentioned in the variable description section, the terrorist attacks that this variable captures contains mostly unknown attributions. Furthermore, it is possible media coverage of terrorism and not terrorism itself is the driver of public opinion in this matter. Conversely, percentage of Muslims in the country is statistically significant and reduces the odds of Muslim exclusion by 19%. However, caution interpreting this variable as there may be reversed causality issues with this variable as well. The Multilevel models show the overall estimates for Europe. Even though it accounts for cross- country variation, it is important to distinguish between regional and country variation in terms of proneness to Muslim exclusion. Table 3 shows the logistic regression estimates for Muslim exclusion by European regions. The division in regions follows the United Nations official division

• f Europe. Overall, there is consistency in results when looking at these regions separately. The

race qualification is statistically significant in all regions. The religious qualification is significant for all regions, except for the South in which it is marginally statistically significant. The language

qualification is statistically significant in the West and the North, yet; it is marginally statistically significant in the East and the South.

Table 3. Multilevel Logistic Regression Estimates for the Willingness to Not Allow Immigrants of Different Race/Ethnicity in the Country (1 = Not Allow, 0 = Allow , Odds Ratio). Individual level Null Model Model 1 Model 2 Model 3 Acceptance Qualifications Language Qualification 1.091*** 1.091*** 1.092*** (0.01) (0.01) (0.01) Religious Qualification 1.019 1.019 1.018 (0.01) (0.01) (0.01) Race Qualification 1.173*** 1.174*** 1.173*** (0.01) (0.01) (0.01) In favor of refugees 0.939*** 0.939*** 0.939*** (0.01) (0.01) (0.01) Contact Minorities 0.544*** 0.545*** 0.544*** (0.01) (0.01) (0.01) Household income 0.920*** 0.920*** 0.920*** (0.01) (0.01) (0.01) Education 0.937*** 0.937*** 0.937*** (0.00) (0.00) (0.00) Female 1.065** 1.065** 1.065** (0.03) (0.03) (0.03) Age (decades) 1.091*** 1.091*** 1.092*** (0.01) (0.01) (0.01) Religiosity 0.946*** 0.946*** 0.946*** (0.01) (0.01) (0.01) Unemployed 1.190*** 1.192*** 1.192*** (0.06) (0.06) (0.06) Country Level Unemployment Rate 1.072* (0.04) Number of Terrorist Attacks 1.001 (0.00) Percent Muslim 0.888 (0.06) Constant

• 2.36***
• 1.02
• 0.48†
• 0.04

(0.23) (0.37) (0.25) (0.3) Random-effects Parameters Country: Identity 1.011 0.640 0.705 0.650 sd(_cons) (0.17) (0.12) (0.13) (0.12) N (Individuals) 24631 24631 24631 24631 N Groups 20 20 20 20 Exponentiated coefficients; Standard errors in parentheses

† p<0.1, * p<0.05, ** p<0.01, *** p<0.001

• 60%
• 40%
• 20%

0% 20% 40% 60% 80% Language qualification Religious qualification Race qualification In favor of refugees Frequency of contact with minorities Household Income Years of education Female Age (decades) Religiosity Unemployed Unemployment rate Number of terrorist attacks Percent Muslim

Figure 3. Percent increase in the likelihood of Muslim immigrants and immigrants of different race/ethnicity exclusion by individual/country characteristics (standardized)

Decreases likelihood Increases likelihood

Individual level

Country level

Acceptance qualifications

Table 3. Logistic Regression Estimates for the Willingness to Not Allow Muslim Immigrants in the Country by European Regions (1 = Not Allow, 0 = Allow, Odds Ratio). East West North South Acceptance Qualifications Language Qualification 1.02† 1.158*** 1.114*** 1.046† (0.02) (0.03) (0.02) (0.03) Religious Qualification 1.075*** 1.075*** 1.112*** 1.035† (0.02) (0.02) (0.02) (0.02) Race Qualification 1.091*** 1.208*** 1.171*** 1.140*** (0.02) (0.02) (0.02) (0.02) In Favor of Refugees 0.627*** 0.516*** 0.588*** 0.582*** (0.03) (0.02) (0.02) (0.03) Household income 1 0.951** 0.931*** 0.904*** (0.02) (0.01) (0.02) (0.00) Education 0.957*** 0.934*** 0.957*** 0.947*** (0.01) (0.01) (0.01) (0.01) Female 1.143*** 1.098* 1.089* 1.108 (0.05) (0.05) (0.04) (0.06) Age (decades) 1.085*** 1.076** 1.105*** 1.102** (0.03) (0.03) (0.02) (0.04) Unemployed 0.966 1.139 0.98 1.148 (0.13) (0.15) (0.10) (0.15) Religiosity 0.967* 0.959** 0.923*** 1 (0.02) (0.02) (0.01) (0.02) Contact Often 0.551*** 0.542*** 0.822*** 0.589*** (0.02) (0.02) (0.02) (0.02) N 3521 7631 10521 2958 Exponentiated coefficients; Standard errors in parentheses † p<0.1, * p<0.05, ** p<0.01, *** p<0.001

In order to see what are the brighter and blurrier boundaries, I have graphed the standardized coefficients in figure 3. In all regions of Europe, Race is the brightest of the boundaries. Conversely, Religion tends to be either insignificant or the blurriest of the boundaries for Muslim

• acceptance. This suggests Alba’s portrayal of Cultural Racism over Biological Racism for Europe

is partially true. While religion seems to be a bright boundary for Muslim immigrant acceptance, race tends to be the largest contributor to Muslim exclusion. In order to fully assess Alba’s argument, the same data should be available for the US, allowing a comparison with Europe.

Table 4. Logistic Regression Estimates for the Willingness to Not Allow Immigrants of Different Race/Ethnicity in the Country by European Regions (1 = Not Allow, 0 = Allow, Odds Ratio). East West North South Acceptance Qualifications Language Qualification 1.04 1.200*** 1.112*** 1.036 (0.02) (0.04) (0.03) (0.03) Race Qualification 1.101*** 1.189*** 1.268*** 1.155*** (0.02) (0.03) (0.03) (0.03) Religious Qualification 1.024 1 1.023 0.984 (0.02) (0.02) (0.02) (0.03) Refugee 0.613*** 0.488*** 0.545*** 0.674*** (0.03) (0.03) (0.02) (0.05) Household income 0.973 0.945** 0.922*** 0.854*** (0.02) (0.02) (0.02) (0.03) Education 0.931*** 0.928*** 0.950*** 0.934*** (0.01) (0.01) (0.01) (0.01) Female 0.99 1.154** 1.096* 0.971 (0.05) (0.06) (0.05) (0.07) Age (decades) 1.081** 0.976 0.985 1.047 (0.03) (0.03) (0.03) (0.04) Unemployed 0.966 1.139 0.98 1.148 (0.19) (0.20) (0.15) (0.24) Religiosity 0.991 0.930*** 0.920*** 0.982 (0.02) (0.02) (0.02) (0.03) Contact Often 1.04 1.04 1.04 1.04 (0.07) (0.06) (0.05) (0.08) N 3521 7631 10521 2958 Exponentiated coefficients; Standard errors in parentheses; country dummies omitted.. † p<0.1, * p<0.05, ** p<0.01, *** p<0.001

Figure 4. In order to see individual country variation, map 1 shows the standardized coefficients for the race qualification when the logit model is estimated for individual countries. As can be seen, the coefficient is the highest in France, Portugal, Ireland, and Finland. Map 1. Standardized Race Qualification Coefficients for each country.

1 1.2 1.4 1.6 1.8 2 Language Qualifications Race Qualification Religious Qualification Odds ratios East N=3521 1 1.2 1.4 1.6 1.8 2 Language Qualifications Race Qualification Religious Qualification Odds ratios West N = 7631 1 1.2 1.4 1.6 1.8 2 Language Qualifications Race Qualification Religious Qualification Odds Ratio North N = 10251 1 1.2 1.4 1.6 1.8 2 Language Qualifications Race Qualification Religious Qualification Odds Ratio South N=2958

Conclusions This study sets the focus on Muslim immigrant and refugee exclusion and provides deeper insight to what are the determinants of these exclusionary sentiments. It is the belief of many scholars that Muslim exclusion in Europe is based on Cultural Racism –i.e. culture or religion are essentialized to the point that individuals are reduced to a certain religion or cultural trait and are seen inferior because of it. As opposed to the US in which Biological Racism – i.e. the individual is reduced to the phenotypical characteristic of skin color. To do so, the study adopts Richard Alba’s “Bright and Blurry” boundary theoretical framework. Bright boundaries are those for which distinction is clear and unambiguous, so that individuals easily know which side of the boundary they are on. Other boundaries are blurry, meaning the location of whether an individual is majority or minority is ambiguous. Alba argues there are four crucial boundaries in determining whether an individual is majority or minority –citizenship, religion, language, and race. This study focused on European citizens, hence; religion, language, and race are examined. The findings suggest race is a consistent bright boundary throughout Europe. In other words, for Europeans it is important that Muslims are white in order to allow them into their countries. Conversely, religion and language are less consistent and of lower importance than race for Muslim exclusion. Although there is substantial variation across countries, it seems both cultural and biological racism take place in Europe. Yet, biological racism seems to have a higher impact than cultural racism. That being said, refugee condition of

immigrants seems to decrease the possibilities of Muslim exclusion. In general, Europeans favor refugee acceptance. According to the results of the study, there is a strong association between refugee and Muslim acceptance. These results coincide with recent findings on refugee acceptance in Europe. There are limitations to the study. The first and most important one is the cross-sectional nature of the data. In absence of longitudinal data, it is not possible to make a causal claim regarding Muslim

• exclusion. This is especially true in the absence of data that captures life courses of actual Muslim
• immigrants. Additionally, data on race and ethnicity is severely limited, as the only available

variable refers to the condition of whether the immigrant is white or not. The same happens for religion, as the only available variable refers to the condition that an immigrant is Christian for acceptance. That being said, there are important policy implications in the results for this study. First of all, at the very least this study shows that race is an important factor in Muslim exclusion. Europe tends to not include race data in their surveys, and when they do it is severely limited. It is the intention

• f this study to push for the inclusion of detailed race data in future surveys. Furthermore, the

implications of the results show integration in Europe for this new influx of Muslim refugees may play out to be very difficult. For both of these reasons, further research on the nature of anti- Muslim sentiment in Europe is needed. As many demographers argue, this new influx of Muslim refugees can be of great benefit to Europe.Europe presents a rapidly aging population structure due to stable below-replacement fertility rates, and the aging of the Baby-boomers (Rowland 2003). Some European countries have begun to face population decline, and their age structure brings the sustainability of the welfare state into serious questioning (Jimenez-Ridruejo Ayuso et al. 2009). Most of the immigrants who make the journey to Europe bear kids –i.e. 38% of the total number are under-aged- and are far from the retirement age of 65-67 (De Bel-Air 2016). The migration profile of this inflow can be the potential solution to Europe’s demographic projection. For this reason it is essential to understand what the assimilation process may be like for this new inflow of migrants.

References Algan, Yann. 2012. Cultural Integration of Immigrants in Europe. Oxford: Oxford University Press. Allport, Gordon Willard. 1979. The Nature of Prejudice. Basic Books. Bansak, Kirk, Jens Hainmueller, and Dominik Hangartner. 2016. “How Economic, Humanitarian, and Religious Concerns Shape European Attitudes toward Asylum Seekers.” Science aag2147. Binder, Jens et al. 2009. “Does Contact Reduce Prejudice or Does Prejudice Reduce Contact? A Longitudinal Test of the Contact Hypothesis among Majority and Minority Groups in Three European Countries.” Journal of Personality and Social Psychology 96(4):843–56. Blake, Michael. 2014. “The Right to Exclude.” Critical Review of International Social and Political Philosophy 17(5):521–37. Cheng, Jennifer E. 2015. “Islamophobia, Muslimophobia or Racism? Parliamentary Discourses on Islam and Muslims in Debates on the Minaret Ban in Switzerland.” Discourse & Society 26(5):562–86. Connor, Phillip. 2010. “Contexts of Immigrant Receptivity and Immigrant Religious Outcomes: The Case of Muslims in Western Europe.” Ethnic and Racial Studies 33(3):376–403. De Bel-Air, Françoise. 2016. “Migration Profile: Syria.” Retrieved May 2, 2016 (http://cadmus.eui.eu/handle/1814/39225). Dong, Yiran and Chao-Ying Joanne Peng. 2013. “Principled Missing Data Methods for Researchers.” SpringerPlus 2. Retrieved June 5, 2016 (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3701793/). Enos, Ryan D. 2014. “Causal Effect of Intergroup Contact on Exclusionary Attitudes.” Proceedings of the National Academy of Sciences 111(10):3699–3704. Foner, Nancy and Richard Alba. 2008. “Immigrant Religion in the U.S. and Western Europe: Bridge or Barrier to Inclusion?” The International Migration Review 42(2):360–92. Hopkins, Nick and Vered Kahani-Hopkins. 2006. “Minority Group Members’ Theories of Intergroup Contact: A Case Study of British Muslims’ Conceptualizations of Islamophobia and Social Change.” British Journal of Social Psychology 45(2):245–64. Jimenez-Ridruejo Ayuso, Zenon, Carlos Borondo Arribas, Julio Lopez Diaz, Carmen Lorenzo Lago, and Carmen Rodriguez Sumaza. 2009. “The effect of immigration on the long run sustainability of the public pension system in Spain.” Hacienda Publica Espanola (188):73–121. Mclaren, Lauren M. 2003. “Anti-Immigrant Prejudice in Europe: Contact, Threat Perception, and Preferences for the Exclusion of Migrants.” Social Forces 81(3):909–36.

Meer, Nasar and Tariq Modood. 2009. “Refutations of Racism in the ‘Muslim Question.’” Patterns

• f Prejudice 43(3-4):335–54.

Quillian, Lincoln. 1995. “Prejudice as a Response to Perceived Group Threat: Population Composition and Anti-Immigrant and Racial Prejudice in Europe.” American Sociological Review 60(4):586–611. Rowland, Donald T. 2003. Demographic Methods and Concepts. Edición: Pap/Cdr. Oxford ; New York: OUP Oxford. Sander, Åke. 2006. “Experiences of Swedish Muslims after the Terror Attacks in the USA on 11 September 2001.” Journal of Ethnic and Migration Studies 32(5):809–30. Savelkoul, Michael, Peer Scheepers, Jochem Tolsma, and Louk Hagendoorn. 2011. “Anti-Muslim Attitudes in The Netherlands: Tests of Contradictory Hypotheses Derived from Ethnic Competition Theory and Intergroup Contact Theory.” European Sociological Review 27(6):741–58. Strabac, Zan and Ola Listhaug. 2008. “Anti-Muslim Prejudice in Europe: A Multilevel Analysis

• f Survey Data from 30 Countries.” Social Science Research 37(1):268–86.

Street, 1615 L., NW, Suite 800 Washington, and DC 20036 202 419 4300 |. Main 202 419 4349 |. Fax 202 419 4372 |. Media Inquiries. 2008. “Unfavorable Views of Jews and Muslims on the Increase in Europe.” Pew Research Center’s Global Attitudes Project. Retrieved October 23, 2016 (http://www.pewglobal.org/2008/09/17/unfavorable-views-of-jews-and- muslims-on-the-increase-in-europe/).

Appendix