The composition of the minority population as a threat: Can real economic and cultural threats explain xenophobia

http://iss.sagepub.com/International Sociology

http://iss.sagepub.com/content/early/2011/05/26/0268580910394004The online version of this article can be found at:

DOI: 10.1177/0268580910394004

published online 26 May 2011International SociologyMikael Hjerm and Kikuko Nagayoshi

and cultural threats explain xenophobia?The composition of the minority population as a threat: Can real economic

Published by:

http://www.sagepublications.com

On behalf of:

International Sociological Association

can be found at:International SociologyAdditional services and information for

http://iss.sagepub.com/cgi/alertsEmail Alerts:

http://iss.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

What is This?

- May 26, 2011Version of Record >>

at Umea University Library on October 20, 2011iss.sagepub.comDownloaded from

http://iss.sagepub.com/

http://iss.sagepub.com/content/early/2011/05/26/0268580910394004

http://www.sagepublications.com

http://www.isa-sociology.org/

http://iss.sagepub.com/cgi/alerts

http://iss.sagepub.com/subscriptions

http://www.sagepub.com/journalsReprints.nav

http://www.sagepub.com/journalsPermissions.nav

http://iss.sagepub.com/content/early/2011/05/26/0268580910394004.full.pdf

http://online.sagepub.com/site/sphelp/vorhelp.xhtml


Corresponding author:Mikael Hjerm, Department of Sociology, Umeå University, 901 87 Umeå, Sweden.Email: [email protected]

Article

International Sociology1–29

© The Author(s) 2011Reprints and permission: sagepub.

co.uk/journalsPermissions.navDOI: 10.1177/0268580910394004

iss.sagepub.com

The composition of the minority population as a threat: Can real economic and cultural threats explain xenophobia?

Mikael HjermUmeå University, Sweden

Kikuko NagayoshiTohoku University, Japan

AbstractThis article sets out to develop a classical theme of empirical research within group threat theory, namely the argument that the size of the minority population threatens the majority population. To be able to clarify the mixed empirical results within this version of group threat theory, the article focuses on the composition of the immigrant population. The article tests both objective sources of cultural threats (linguistic composition and the Muslim population) and economic threats (the proportion of working-class individuals and the unemployed among the immigrant population). The study concludes that, first, the composition of the immigrant population is of utter importance for the size argument to be valid for cultural threats (proportion of Muslim population), whereas for economic threats it does not matter. Second, compositional economic threats matter strongly to the group that genuinely competes for scarce resources – the working class is more xenophobic when the immigrant working class is large. Third, the study brings some clarity with regard to the cultural composition of the immigrant population: it is shown that the relationship between Muslims and European majority populations mirrors the relationship between whites and African-Americans in the US.

Keywordsattitude, culture, economic, ESS, Europe, group threat, majority, minority, Muslim, prejudice, xenophobia



2 International Sociology

Introduction

A special case of group threat theory stipulates that the majority is threatened by the size of one or more minority group(s) that share the same territory (Blalock, 1957; Dixon, 2006; Fosset and Kiecolt, 1989; Taylor, 1998). In other words, the larger the minority, the more threatened the majority feels and, therefore, the more averse the majority becomes. The majority becomes averse since it perceives that the minority is a threat to its dominant position as the groups compete over scarce resources.

The size argument can explain the relationship between whites and African-Americans in the US (Fosset and Kiecolt, 1989; Glaser, 1994; Quillian, 1996; Smith, 1981; Taylor, 1998, 2000), but becomes questionable when we consider other group relations. Hood and Morris (1997) show that attitudes towards immigration and immi-grants relate negatively to the concentration of immigrants in the US. The latter result concurs with Taylor (1998), who demonstrates that size matters when dealing with African-American minorities, but does not influence whites’ prejudice towards immi-grant groups in the US. Dixon and Rosenbaum (2004) present similar evidence. Researchers in Europe have shown that the relationship between the size of various classifications of immigrant groups and anti-immigrant attitudes is far from straight-forward (Hjerm, 2007b; McLaren, 2003; Quillian, 1995). Pettigrew (2000) claims that individual-level predictors of prejudice (or anti-immigrant attitudes), such as age, education and political conservatism, are consistent over geographical areas, across groups and over time. On the group level, this consistency across geographical areas has not been obvious since the relationship between whites and African-Americans in the US is not fully mirrored elsewhere.

The lack of clarity, or homogeneity, in the effect of the size of the minority popula-tion is related to the lack of a clear understanding of what kind of threat different minority populations really pose in different circumstances. Previous research has often taken the size of minority groups per se as a valid indicator of group threat, which is somewhat problematic. Dixon (2006: 2181) claims that ‘the presence of a racial/ethnic group is only the first step in the causal chain to prejudice’. The question is whether the threat related to size is due to the competition over available jobs, concerns about the welfare state being undermined, or whether immigrants are perceived to threaten the national way of life of the majority population. Any of these possible explanations is plausible in contemporary empirical research. The diverse effects of the size of the minority population on anti-immigrant attitudes in previous research reflect this diversity; we therefore need to look into the specifics of the composition of the minority population that stimulate anti-immigrant attitudes. Another point that has not been clarified is how different compositions of minority populations affect different groups of people, and whether this depends on the direct threat to their position. To shed some light on these questions, we set out to examine whether anti-immigrant attitudes are triggered by objective sources of economic and cultural threats related to the composition of the minority population, as well as whether the effects of these threats vary depending on individual exposure to different forms of the threats. Data come from the European Social Survey round 1 (ESS1).



Hjerm and Nagayoshi 3

Theoretical discussion

The theoretical departure point is found in so-called group threat theory (Blumer, 1958), which assumes that individuals identify with one or more group(s) and that the diverse interests of different groups generate conflicts that, in turn, generate negative attitudes. Blumer identified group identity, out-group stereotyping, preferred group status and per-ceived threat as being intrinsic to prejudice. The fourth precondition, the group threat condition, states that prejudice is a reaction to explicit or implicit challenges to the domi-nant group’s position – it is the catalyst that makes the other three preconditions lead to prejudice. Put more precisely, people make group classifications that will have political and economic saliency in that they are more likely to want to benefit their own group over other groups in the struggle for scarce resources due to the desired status hierarchy of different groups. This means that one or more minority groups threaten the majority group, which elicits anti-immigrant attitudes among the members of the majority group.

The classic empirical example of group threat theory is Bobo’s (1983) study of the bussing of white schoolchildren to black neighbourhoods in the US. Bobo showed that white parents strongly opposed bussing when it concerned their own children despite their advocacy of general principles of equality between ethnic groups. The interpretation is that they experienced a threat towards their group position and reacted accordingly. A vast body of empirical evidence supports group threat theory by showing that various forms of group threat affect negative attitudes towards minority groups (McLaren, 2003; Quillian, 1995, 1996).1 We can conclude that threat brought by a group that is subordi-nate to the majority has come to be seen as the single best indicator of negative attitudes towards out-groups (Paxton and Mughan, 2006; Stephan and Stephan, 2001).

There is a theoretical difference between the ‘realistic group threat theory’ school (see Bobo, 1983; Sears and Jessor, 1996), where anti-immigrant attitudes are a result of real experiences and interests, and the school of group threat theory that claims that what matters is not whether threats are real or not, but that they are perceived as such. This theoretical difference is manifested in different empirical studies, where some place an emphasis on perceived threat by using subjective indicators (Fetzer, 2000; McLaren, 2003; Sniderman and Hagendoorn, 2007), whereas others try to examine the effects of real threat by using objective indicators (Quillian, 1995; Semyonov et al., 2006). The ‘perceived threat’ school is correct in the point that it is quite difficult, if not impossible, to identify ‘real’ impacts that immigrants bring to a host society. Nevertheless, ‘realistic group threat theory’ leads to an important insight: anti-immigrant attitudes might have objective sources. Even when members of the majority group do not have strong dis-criminative views against immigrants, they may become xenophobic because some specific objective conditions incline them to feel threatened about losing the resources they have. A focus only on perceived threat might obscure our understanding of what triggers anti-immigrant attitudes. Therefore, we focus explicitly on real or objective sources of threat. One of the frequently examined objective sources of threat has been, as indicated in the introduction, the proportion of the minority population.2 Two articles, one by Blalock (1957) and the other by Pettigrew (1957), have laid the foundation for this. Blalock (1967) points out two reasons why the majority group is expected to feel threatened by the relative size of a minority group. First, the larger the group, the more fierce the competition for scarce resources. Second, the larger the group, the greater




the potential is for collective (political) action against the majority. In other words, the majority group can perceive large numbers of immigrants as competitors in the economic domain and/or in the domain of identity politics.

Despite the theoretical assumption, the effect of size is not clear-cut except for in the relationship between whites and African-Americans in the US. Thus, it is not the size of the foreign population, but rather the size of the African-American population in the US that has an important meaning. The ambiguity of results may stem from the indicator used to measure the immigrant population. In group threat theory, people become averse towards out-groups when they feel threatened by members of those groups. In other words, it is not the existence of immigrants that matters but the existence of immigrants who threaten the host residents’ positions. If immigrants are culturally similar to the host society, they may not be culturally threatening. Moreover, wealthy immigrants who cre-ate job opportunities in the host society may not be regarded as an economic threat. From this point of view, we can assume that it is not the size of the immigrant population as a whole that is meaningful, but rather the composition of the immigrant group itself. Immigrants are perceived as a threat in two different ways, culturally and economically, and the size of the immigrant population matters only when they bring these threats to the host society.

Sources of economic threat

Ever since the days of Sherif and Sherif (1953), the existence of prejudice has been explained by the competition for scarce resources: that immigrants pose a threat to the material well-being of the majority population. As Giles and Hertz (1994) put it, the relationships between different ethnic groups are viewed ‘as a function of their competi-tive positions’. Many studies similarly conclude that economic threats are of importance. Quillian (1995) shows that low GDP and a large proportion of immigrants produce anti-minority prejudice in Europe. The less the majority feels that their jobs are in jeopardy, the more likely they are to be in favour of, or at least not against, increased levels of immigration (Espenshade and Hempstead, 1996). Moreover, anti-immigrant sentiments rise during economic downturns (Olzak, 1992) and are also dependent on the respondents’ perception of national economics (Burns and Gimpel, 2000). A problem with previous studies of economic circumstances and anti-immigrant attitudes is that they have focused on either very general economic contexts (i.e. GDP, unemployment rates) or perceptions of economic contexts; as a consequence, they have neglected the direct economic threat possibly posed by the immigrant population. In other words, the examined factors are better viewed as contexts that make threats from the minority population salient, not as threats per se.

The way in which immigrants become sources of economic threats is explained in the split labour market theory (Bonacich, 1972; Boswell, 1986), which assumes that xenophobia is strengthened when immigrant workers work for decidedly lower wages than the majority population. Many immigrants are more willing to work for low wages with poor working conditions in a receiving country due to low incomes and living standards in their home countries. As a result, immigrants diminish job opportunities for the majority populations as well as undermine the wage standard in the host society – thereby becoming a threat to the majority population. The split labour markets are




reproduced because of the reduced labour market mobility of minorities relative to that of the dominant workers (Boswell, 1986). From this viewpoint, two groups of immi-grants are especially likely to pose an economic threat: foreign-born manual workers and foreign-born unemployed. The former group can be a threat because the influx of low-skilled and low-waged immigrants may distort the national labour market. The unemployed can be seen to undermine the welfare state since immigrants are perceived to ‘free-ride’ on the welfare state (Crepaz, 2008). Thus, our first hypothesis is:

H1: The larger the proportion of low-skilled immigrants (i.e. foreign-born manual workers and foreign-born unemployed) in a country, the more xenophobic the majority population is.

Competition for scarce resources with immigrants can be perceived not only as a collective competition (‘they stole our jobs’), but also as individual competitions (‘they stole my job’). Group threat theory stipulates and empirical research shows that those who have low socioeconomic status have stronger antipathies towards immigrants (Coenders and Scheepers, 2003; Esses et al., 2001; Scheepers et al., 2002; Svallfors, 2006).3 It is assumed that the reason for this antipathy is because people with low socio-economic status are more exposed to competition over scarce resources. However, the fact that low socioeconomic status predicts anti-immigrant attitudes does not automati-cally imply that this is due to competition over economic resources or even that it relates to economic threat at all. To be sure that the economic threat that is posed by the immi-grant population really matters, we will need to find that the working class has stronger anti-immigrant attitudes than other classes and that these attitudes are dependent on the proportion of immigrants from the working class (i.e. the number of those who threaten to take their jobs). Or, we will need to find that the working class or the unemployed become more xenophobic when unemployment within the immigrant population is large (i.e. a situation where competition over welfare state resources is salient). The second hypothesis is phrased as:

H2: The larger the proportion of low-skilled immigrants, the more xenophobic are those who directly compete with them.

Sources of cultural threat

Cultural aspects are stressed as important in accounts of the existence of prejudice (e.g. McLaren and Johnson, 2007). The origin of anti-immigrant attitudes relates to cultural tensions between ethnic groups where foreigners are viewed as potential threats to national identity, social order and, most importantly, to the values cherished by the majority population. The focus on cultural threat has either been somewhat neglected or, alternatively, has been overtly on perceptions of cultural threat instead of their objective sources. However, the existing empirical studies do suggest that cultural threat is of importance. For example, Van der Brug et al. (2000) find that social and economic variables were mostly insignificant in explaining support for anti-immigrant parties in seven European countries. Moreover, Scheepers et al. (2002) and Dixon (2006) show that cultural threat is a better analytical tool for explaining prejudice than economic threat. Still, we do not know what cultural differences trigger perceptions of




cultural threats. The reason for this is that cultural threats are often viewed in terms of perceived threats where it is not obvious if people feel threatened due to their being prejudiced or if it is the other way around. The size of the population of non-European/non-western immigrants is often used as an indicator of objective sources of cultural threat (e.g. Schneider, 2007), but ‘non-European’ implies various differences: linguistic, religious, or perhaps physical appearance. As a result, we need to examine what kind of differences matters. A problem is that, in comparison to economic threat, it is less clear how to measure cultural threat in a more objective way than what has been common in previous research. We therefore approach cultural threat from two different angles.

Paxton and Mughan (2006) point out that the key concept of cultural threat is assimila-tion, and show that people feel culturally threatened by the lack of assimilation of the minority population. In the classical assimilation theory, all immigrants were predicted to irreversibly assimilate into the host society. Even though there were different standpoints as to whether the acculturation was one-way or not, it was viewed as an important and inevitable stage in assimilation (Gordon, 1964; Park, 1950). For example, Gordon (1964) thought that over years and generations, immigrants in the US would be Americanized and would become indistinguishable to the middle-class cultural patterns of largely white Protestant, Anglo-Saxon origins. However, the assumption that acculturation is inevitable is now strongly questioned. Assimilation as a linear process occurred in specific eco-nomic and political conditions during the first mass migration to the US in the early 20th century (Rumbaut, 1997). Compared to most immigrants in that period, today’s immi-grants vary with regard to their socioeconomic status and cultural backgrounds. The latter is also the case in Europe today, where difficulties in acculturation become salient as the number of non-Europeans increases (Ålund and Schierup, 1991). Moreover, the transi-tion from a manufacturing-based economy to a service-based post-industrial economy makes the form of assimilation of immigrants different (Rumbaut, 1997; Zhou, 1997). As Glaser (1994) states, the prediction of assimilation has been discredited and the new real-ity of cultural pluralism has appeared. Portes and Zhou (1993) elaborate the assimilation theory and establish the segmented assimilation theory. The latter assumes that there are three possible patterns of assimilation:

. . . one of them replicates the time-honored portrayal of growing acculturation and parallel integration into the white middle-class; a second leads straight into the opposite direction to permanent poverty and assimilation into the underclass; still a third associates rapid economic advancement with deliberate preservation of immigrant community’s value and tight solidarity. (Portes and Zhou, 1993: 82)

A low level of cultural assimilation might bring anxiety to the majority population. For the majority population, immigrants’ preservation of their native cultures is per-ceived to undermine the majority’s way of life, which can be seen in the case of linguis-tic assimilation. A common language is one of the strongest markers of cultural difference, as well as one of the best predictors of the probability of a successful inte-gration process (Dustmann, 1994; Esser, 2006; Jasinskaja-Lahti, 2008). In other words, linguistic differences are likely to be a source of cultural threat (cf. Zárate et al., 2002). Sniderman and Hagendoorn (2007) analyse the remarks of the Dutch participant in the group discussion, and find that the majority feels uneasy when immigrants use their




native languages in their presence (Sniderman and Hagendoorn, 2007). Therefore, if specific compositions of immigrant groups trigger xenophobia by being sources of a cultural threat, we hypothesize that:

H3: A large proportion of immigrants with different cultures (i.e. those who do not linguistically assimilate) strengthens xenophobia.

In addition to linguistic diversity, we focus on religious differences. Blumer (1958) emphasized the importance of historical contexts in prejudice. In his words, ‘the sense of group position is the very heart of the relation of the dominant to the subordinate group’ (Blumer, 1958: 4). This partly explains why the size of African-American groups affects anti-African-American prejudice while the size of Asian-Americans or Hispanic groups does not affect anti-Asian-American or Hispanic prejudice (Dixon, 2006; Taylor, 1998). The long history of racial discrimination and slavery in the US shaped whites’ perception of the relationship between African-Americans and whites. On this basis, prejudice is triggered if there is a large proportion of African-Americans. In a European context, ethno-religious aspects play a similar role to that of race in the US. Goldberg (2006) explains this in his arguments of racial Europeanization; anti-Semitism was one example of racism in Europe in the past, and anti-Muslim sentiment is a more recent example. In Europeans’ colonial occupation of the Middle East, orientalism towards Muslims became racially historicized, and Muslims were perceived as inferior to Europeans. They were regarded as irrational, aggressive, emotional and strange in their ways and habits (Goldberg, 2006). The perception that there is a serious cleavage between liberalistic – or perhaps Christian – values and Islamic values is strong in Europe (Sniderman and Hagendoorn, 2007). We can thus assume that a large proportion of Muslims becomes a source of cultural threat and thereby triggers xenophobia:4

H4: A large proportion of Muslim immigrants strengthens xenophobia.

As in the case of economic threat, cultural threat is also assumed to have different effects, depending on individual dispositions. There are two factors that are of particular importance. The first is derived from contact theory (e.g. Jackman and Crane, 1986; Park, 1950; Sigelman et al., 1996). Contact theory assumes that contact with members of the out-group reduces averse attitudes towards them because it corrects stereotypes. The empirical evidence supporting contact theory is vast, and it has been proven that high-quality contacts are more likely to have a positive effect than low-quality contacts. For example, previous research has found that friendship reduces xenophobia, espe-cially when the size of the immigrant population is large (McLaren, 2003; Pettigrew, 1998). The modes of contact that we investigate are those related to having friends as well as living close to immigrants. However, there is a selective bias: the more tolerant people are, the more willing they are to live among immigrants or to be friends with them. The selective bias is important, but previous research has found that selective bias is modest, while there are stronger effects of contact on xenophobia (Dixon, 2006; Pettigrew, 1998). In other words, those who have prejudice are likely to avoid contact with immigrants, but do become more tolerant once they have contact with them. The fifth hypothesis is:




H5: People who do not have contact with immigrants are more strongly affected by cultural threat than others.

The second individual disposition is a preference for cultural homogeneity.5 A prefer-ence for homogeneity or sense of strong in-group solidarity combined with out-group stereotyping is fundamental to anti-immigrant attitudes. The threat against the group’s position is of the greatest importance, since xenophobia is closely related to nationality and identity (Yakushko, 2009). When people attribute great value to national solidarity, large populations of immigrants who have different cultures or who are perceived to be in a lower group position may be more likely to be seen as a threat. From this viewpoint, we can assume that those who place importance on cultural homogeneity within a state are more likely to feel threatened. The sixth hypothesis is:

H6: People who value cultural homogeneity within a state are more strongly affected by cultural threat than others.

In summary, H1 and H2 deal with economic threat whereas H3 to H6 deal with cultural threat. The reason to include more hypotheses for cultural threat relates to the concept being theoretically less clear as well as to problems related to the operationali-zation of cultural threat. Moreover, H1, H3 and H4 deal with general threat, whereas H2, H5 and H6 deal with different effects of the threat depending on individual dispo-sitions. By examining these six hypotheses, we set out to examine both the effects of economic and cultural threats in general as well as if these threats affect people who are more likely to be exposed to them.

Data and method

Data come from the European Social Survey (ESS) 2002 (www.europeansocialsurvey.org). ESS 2002 includes 22 countries, but we use only 21 countries in this case. Israel has been excluded from our analysis for two reasons. First, Israel has adopted a unique immigration policy that focuses on the repatriation of Jewish people; as a result, its resi-dents may have different tendencies concerning anti-immigrant attitudes. Second, there are limits to the accessibility of macro data for Israel. The countries included in the analysis are: Austria, Belgium, the Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Luxembourg, the Netherlands, Norway, Poland, Portugal, Slovenia, Spain, Sweden, Switzerland and the UK. The samples were selected through a random sampling method of eligible residents over 15 years of age. The respondents were asked to answer an in-person questionnaire. The response rate varied between countries, with the lowest in Switzerland (33.5 percent) and the highest in Greece (80.0 percent).6 Despite this variance, there seem to be few deviations of samples, and response rates are generally better in ESS than in other comparable national surveys. Since this research analyses attitudes towards immigrants, we exclude the data of those who were born outside the country of analysis. This leaves a new file containing 27,399 cases, after missing cases were sorted out.

We used hierarchical linear modelling as the analytical method. In analysing social science data within a nested structure, this model has three advantages (Garner and




Raudenbush, 1991: 253). First, it avoids violating the assumption of independence of observations that traditional ordinary least square analysis suffers from. Second, it is effective not only in estimating group-level effects but also in estimating cross-level effects. In other words, it is useful to examine the effects of group-level factors on both the average level of outcome within the group and on the certain structural relationships within groups. This point is quite important for our research because we focus on the effects of the group-level threat and the differences of those effects according to indi-vidual conditions. Third, this model can separate the variance between levels and provides true variance of the micro-level parameter that is separated from the sampling variance. Although the large N is small (i.e. 21 countries), it is nevertheless possible to perform this kind of modelling with this number of cases. Snijders and Bosker (1999) provide a rule of thumb, stating that if N is equal or greater to 10, a random intercept model (i.e. a linear hierarchical model) is preferable over a fixed model as a means of regression analysis. We used the software program HLM version 5.05 (Raudenbush et al., 2001) to model the effects.

Indicators

The dependent variable: Xenophobia

Group threat theory implies that the stronger the threat against group interests, the more averse the attitudes towards other groups. These antipathies or feelings can take the form of negative attitudes such as prejudice, xenophobia and racism, as well as an unwilling-ness to support actions to remedy segregation and increase equality between ethnic groups. As symbolic theories have shown, there

is a qualitative difference between such negative

attitudes and the support for legal frameworks to remedy discrimination. Still, group threat theory implies that group threat should influence attitudes towards immigrants as well as policies. Wilson (2001: 486) is correct in his interpretation when he says that:

Group threat theory is a general perspective, so the link it specifies between dominants’ threat perceptions and their attitudes toward subordinates should apply to native-born Americans’ attitudes towards immigrants and immigration policies.

Therefore, in applying this reasoning in a broader context, group threat theory should be able to predict a variety of antagonistic attitudes towards immigrants. This is also obvi-ous in the way the outcome has been measured in relation to group threat theory. Blumer used the notion of prejudice as the dependent variable in his group threat theory, but the notion of prejudice would, in his context, be better understood as antipathies or a general set of negative attitudes towards others. Blalock (1957) used racial discrimination, Fosset and Kiecolt (1989) employed views on integration, Scheepers et al. (2002) applied ethnic exclusion and Halperin et al. (2007) drew on social distance. The general idea is that group threat relates to a general set of negative attitudes about minority groups, not that it relates to a specific set of beliefs. In this article we use xenophobia.

The etymological meaning of xenophobia comes from the Greek words xeno, which means ‘stranger’, and phobia, which means ‘fear’. Thus, xenophobia literally means ‘fear of strangers’. The lexical definition of xenophobia has been developed in scientific literature in two ways. First, the stranger has come to be equated with foreigners or




immigrants, and second, xenophobia has come to always include a derogatory under-standing of immigrants.

Thus, a contemporary understanding of xenophobia is to view it as an antipathy towards immigrants or foreigners based upon fear (Hjerm, 2007a). Xenophobia is partly an attitude, but also includes an affective part: fear, which is intrinsic to xenophobia. The fear is a result of being in a situation where individuals perceive that their individual or their group’s position is being threatened. The threat against the group’s position is of the greatest importance because xenophobia is closely related to nationality and identity (Yakushko, 2009) – whereby people may perceive foreigners or immigrants as a threat to the national solidarity.

Six statements have been used to operationalize xenophobia:

1. Would you say that people who come to live here generally take jobs away from workers in [country], or generally help to create new jobs?

2. Most people who come to live here work and pay taxes. They also use health and welfare services. On balance, do you think people who come here take out more than they put in or put in more than they take out?

3. Would you say it is generally bad or good for [country]’s economy that people come to live here from other countries?

4. Would you say that [country]’s cultural life is generally undermined or enriched by people coming to live here from other countries?

5. Is [country] made a worse or a better place to live by people coming to live here from other countries?

6. Are [country]’s crime problems made worse or better by people coming to live here from other countries?

These items concern fear about immigrants and of immigrants bringing troubles to a society. In this sense, this indicator shows the perception of threat, and we analyse how xenophobia, or threat perception, is affected by the composition of minority popula-tions. All of the responses to the statements range from 0 to 10 on an 11-point scale. The six statements produce a one-factor solution (principal component analysis) in all coun-tries (see the Appendix). The indicators also load very similarly on the factor in all countries. As expected, Cronbach’s alpha scores are also high in all of the included countries (see Appendix). Thus, it is clear that the six items measure one and the same dimension in all countries. Of course, the same result occurs when testing the indicators in a latent class analysis. Similar items in the International Social Survey Programme also produce one dimension/cluster/class for all of the included countries. People simply do not make a distinction between the items included here: they are simply positive or negative towards immigrants in general. The index also correlates strongly with atti-tudes towards immigration. The more xenophobic one is, the fewer immigrants one is willing to accept.7 Correlations between single items and the index are very similar and indicate that they have a similar impact. Moreover, the omission of one or more items from the index does not produce any difference in the results. It is perhaps possible to object to single items, but together they clearly make up a valid indicator of xenophobia. The items were aggregated to create one indicator, whereby the higher the value, the stronger the xenophobia.




Independent variables: Macro level8

We used two indicators of economic threat, overrepresentation rate of foreign-born in manual work and in unemployment. The former measure is the quotient of dividing the portion of manual9 workers within the foreign-born workers group by the portion of manual workers within the native-born workers group. The latter measure is the quotient of dividing the portion of the unemployed within the active foreign-born population by the proportion of the unemployed within the active native-born population. These variables indicate the extent to which the foreign-born population is likely to work in low-skilled jobs compared to the native-born population. In other words, they indicate the degree to which foreign-born populations are crowded into low-skilled jobs. These indicators are calculated from OECD (2008) except for Slovenia, where these indicators are derived from the average of three rounds of the ESS.

We also used two indicators for cultural threat. The first is the rate of linguistic non-assimilation, which is operationalized as the proportion of those who do not speak the official language at home, and the second is the proportion of Muslims within a country. The language indicator is made from raw data from the fractionalization index of Alesina et al. (2003);10 the ethno-religious indicator comes from the data of the Association of Religious Data Archive.11 These indicators are arguably good indicators of cultural threat, but are not unproblematic as measurements. For example, we cannot identify whether immigrants can actually speak the official language or not – only if they do not speak it at home (and the second-generation immigrants who speak their parents’ native language at home are included in the proportion of those who do not speak the official language). Moreover, the Muslim population is not limited to immi-grants, as there are others who also believe in Islam. Therefore, we need to take into account that these indicators do have their limitations.

These threats relate to several other contextual variables, which need to be controlled for. We control for the proportion of foreign-born population. The figures come from the OECD (2008) except in the case of Slovenia, which is derived from Eurostat (2008). The economic condition of a country is assumed to relate both to the social condition of immigrants and to anti-immigrant attitudes. If the economic condition of a country is poor, anti-immigrant attitudes increase (Quillian, 1995), and wealthy countries are more likely to attract immigrants (O’Connell, 2005). The economic conditions are measured via gross domestic product. GDP is measured in US$1000 per capita (International Monetary Fund, 2002).

Independent variables: Micro level

Our hypotheses assume that people who are directly threatened by the immigrant popula-tion will be more xenophobic than those who are not directly threatened. The position related to economic threat is occupational status, since we are measuring economic threat coming from the labour market. Occupational status is divided into four categories: non-manual workers, manual workers, the unemployed and those outside the labour market.

To be able to test whether cultural threat has different effects depending on the risk of being threatened, we use preferences for cultural homogeneity and two types of contacts with immigrants. As an indicator of preferences for cultural homogeneity, we made a




subtraction index out of six items related to who people want to let into the country.12 The higher the value is (the scale runs from –9 to +9), the stronger the preference for cultural homogeneity.

Contacts with immigrants are measured on two levels of propinquity: contacts as a friend, which represents whether one has any foreign friends who have come to live in [country] (‘none’ = 0, ‘a few/several’ = 1), and neighbourhood contacts, which represents whether the area they currently live in is described as containing ‘many/some’ (scored as 1) or ‘almost none’ (scored as 0) people of a different race or ethnic group.13

In addition to these variables, social demographic variables are also considered because these variables are not always equally dispersed across countries. If these indi-vidual variables have an effect on anti-immigrant attitudes, and if we do not control for such effects, we may overlook or exaggerate the effects of contextual variables. Age is assumed to have an effect on people’s attitudes towards ethnic minorities or immigrants (Mulder and Krahn, 2005; Semyonov et al., 2006); the older the respondents are, the more xenophobic they become. In the present study, the age of the respondents ranges from 16 to 89 years old.14

Education is a strong predictor of anti-immigrant attitudes. The more educated the host residents are, the less likely they are to hold a negative attitude towards immigrants (Coenders and Scheepers, 2003; Hello et al., 2004). Education can be interpreted as an indicator of social vulnerability because it determines the opportunity to acquire jobs with high income in industrial societies (Hello et al., 2004). However, there are also other aspects of education that affect anti-immigrant attitudes: the conveyance of democratic and multicultural values to children (Hjerm, 2005), or increases in the degree of contact with foreigners (Nukaga, 2006). In this research, we only examine the main effects of education on xenophobia. Education is measured by the number of years that the respond-ents attended full-time schooling.

In some studies, gender has been assumed to have an effect on attitudes towards ethnic minorities, as females are considered to be more tolerant than males (Hello et al., 2004). Gender is also included in the analysis (male = 1, female = 0).

We also control for immigrant origin. In other words, we control for whether at least one of the respondent’s parents was born outside the resident country or not (immigrant origin = 1, native origin only = 0).

Results

We first examine the national differences of average levels of xenophobia, which differ between countries. As Figure 1 shows, the level of xenophobia is relatively low in Scandinavian countries (Sweden, 27.2; Finland, 29.5; Denmark, 31.2; Norway, 31.4) and relatively higher in Southern European or Eastern European countries (Greece, 42.7; Czech Republic, 37.5; Hungary, 36.8), although there are exceptions such as Italy (31.6) and Spain (31.2). It might be surprising that immigrant-sending countries show higher xenophobia than immigrant-receiving countries, although these results concur with previous research (Green, 2007; Schneider, 2007; Semyonov et al., 2006).

The question is how these differences in xenophobia are affected by threat. Before looking into the effects of economic or cultural threats, correlations between country-level variables are shown in Table 1. There are strong correlations between the rate of linguistic




non-assimilation and the proportion of foreign-born populations (0.78). We can say that immigrants are likely to maintain their native language when they are many. On the other hand, correlations between the rate of Muslims and the proportion of foreign-born popula-tions are quite weak (0.29). Moreover, GDP has strong correlations with the proportion of foreign-born populations as well as the overrepresentation rate in unemployment. Immigrants are attracted to the wealthy countries despite being more likely to become unemployed there than in economically poorer countries.

How do these country-level factors affect xenophobia? Table 2 shows the results of the analyses. Before examining our hypotheses, we need examine if there are significant differences in xenophobia that can be referred to the country level of analysis. This is examined in Model 0 (the null model), which includes only error terms. Model 0 indi-cates that there are significant random effects on the country-level intercept, which means that the average levels of xenophobia differ between countries. Ten percent of the variance can be referred to the country level.

We add individual variables to Model 0 in Model 0a. As expected, the effect of the individual indicators on xenophobia mirrors previous research. Education has a negative effect on xenophobia and people of immigrant origin are, obviously, less xenophobic than people of native origin. Contrary to most previous research (Hello et al., 2004), female respondents show higher levels of xenophobia than male respondents, but the effect is limited (–0.39).15 Occupational status has a significant effect on xenophobia because manual workers and the unemployed are more xenophobic than non-manual workers. According to the result, we can assume that those who are in more competitive conditions with immigrants become more xenophobic than those who are not. Preferences for cultural homogeneity have significant effects on xenophobia and indicate that the more people prefer cultural homogeneity, the more xenophobic they are, even though causality is not obvious. In addition, both contact with immigrants as a friend and neigh-bourhood contacts have significant negative effects on xenophobia. In contact theory, superficial contact is assumed to have no, or rather positive, effects on xenophobia. The result presented here indicates that even superficial contacts may weaken xenophobia.

Figure 1. Difference in the level of xenophobia between European countries




However, our main aim is to examine whether objective sources of economic and cul-tural threats related to the composition of the immigrant population affect xenophobia in general and/or have different effects depending on individual exposure to economic and cultural threat. Model 1 and Model 2 examine Hypotheses 1 and 2 by including our two macro-level indicators of economic threats and the interaction terms of these macro-level indicators and individual occupational status. We also control for a number of general macro-level indicators. We see that GDP is negatively related to xenophobia (i.e. people are less xenophobic in countries with better economic conditions). The proportion of foreign-born population has no impact on the level of xenophobia, indicating once again that the size of the immigrant population is by itself not automatically threatening to people.

Hypothesis 1 is faulty in that neither the level of manual workers (M1) nor the level of unemployed foreign-born population (M2) has any effect on xenophobia. The latter implies that people do not in general become more xenophobic because low-skilled immigrants threaten to undermine the labour market by lowering average wages or undermining the welfare state. In general, people are simply not threatened by this.

Hypothesis 2 stated that people who are likely to be personally threatened economi-cally were more likely to be xenophobic – we expected manual workers and the unem-ployed to be more xenophobic when immigrants are likely to be their competitors. Model 1 shows that the manual workers are more xenophobic when the proportion of foreign-born manual workers is large, whereas no such relationship exists for the unemployed or for those who are not in the labour market. The overrepresentation of immigrants in manual jobs only affects the level of xenophobia among manual workers. In other words, people who are in direct competition for economic resources with the immigrant popula-tion are more likely to be xenophobic. The interaction effects in Model 2 show milder effects than in Model 1. The working class, as well as those not in the labour market, tend to be more xenophobic when immigrants are more likely to be unemployed. However,

Table 1. Correlations between country-level variables

Overrepresentation rate in manual work

Overrepresentation rate in unemployment

Rate of linguistic non- assimilation

Proportions of Muslims

GDP

Proportions of foreign-born population

0.545 * 0.347 n.s. 0.781 ** 0.285 n.s. 0.684 **

Overrepresentation rate in manual work

- 0.277 n.s. 0.387 + 0.521 * 0.322 n.s.

Overrepresentation rate in unemployment

- - 0.067 n.s. 0.390 + 0.697 **

Rate of linguistic non-assimilation

- - - –0.082 n.s. 0.428 +

Proportions of Muslims

- - - - 0.252 n.s.

Notes: N = 21, ** p < .01, * p < .05, + p < .1, n.s. = not significant.




those who are unemployed are not affected by high levels of unemployment within the immigrant population. The results from Model 2 indicate that the direct threat of unem-ployed immigrants is of minor importance. It was reasonable to assume that the number of unemployed immigrants was in fact threatening, since unemployment wears on the resources of the welfare state. However, such potential use of the welfare state by the immigrant population is not seen as being a severe threat compared to when immigrants compete over the same jobs as the working-class population.

Hypothesis 3 assumed that xenophobia is strengthened when the proportion of immi-grants who do not speak the country’s official language becomes large. This hypothesis is examined in Model 3. Model 3 shows no significant effect concerning the level of linguistic assimilation on xenophobia, implying that Hypothesis 3 is invalid. The sce-nario in which immigrants use their native languages instead of the official languages of the host society is sometimes described as a symbol of cultural threat to the host society (Kymlicka, 2001; Sniderman and Hagendoorn, 2007). The latter cannot be empirically corroborated in our results, as people are not more xenophobic in linguistically homoge-neous countries.16 However, we need to pay attention to the limitations of our indicator, which shows the rate of linguistic non-assimilation in the private sphere. Arends-Tóth and Van de Vijver (2003) show that the Dutch population prefers assimilation almost at the same level in the private sphere as in the public sphere. However, there is a possibil-ity that the majority population feel threatened only when immigrant minorities speak their native language in the public sphere. Therefore, we need further research to exam-ine linguistic assimilation and xenophobia before dismissing it.

Hypothesis 4 stated that ethno-religious differences trigger xenophobia. This hypoth-esis is supported in Model 4, which shows that the proportion of Muslims does have a positive effect on xenophobia: the larger the proportion of Muslims in the society, the higher the risk for a xenophobic majority population. An exception to the results in Model 4 is that it is not the proportion of Muslims that matters, but rather the proportion of visible immigrants (immigrants from the non-western world). The latter is not the case as the proportion of non-European immigrants only has a marginal effect on interaction with cultural threat perception and such a model has a worse fit than Model 4.17 The limited effect of ethnic/racial divisions can also be found in other research (Hjerm, 2007a; Quillian, 1995). Therefore, it seems that it is ethno-religious differences that mat-ter. A likely interpretation is that people feel that their national way of life is threatened when the culturally most distant group in Europe (Muslims) is large enough.

Hypothesis 5 stated that those who favour cultural homogeneity are more likely to be xenophobic than others when the cultural threat is more imminent. According to Model 3 and Model 4, this hypothesis appears to be faulty. The interaction terms between the evaluation of cultural homogeneity and cultural threat do not have significant effects in these models. Those who value cultural homogeneity within a state have xenophobic attitudes regardless of the level of cultural threat. In other words, those who are willing to promote cultural homogeneity still negate immigrants and do so even when immi-grants assimilate linguistically or religiously into the host society.

Hypothesis 6 stated that those who do not have contacts with immigrants are more likely to be xenophobic than others. This hypothesis is valid with regard to cultural threat regarding ethno-religious differences but not linguistic differences. Model 3 shows that there is no significant interaction effect between the proportion of immigrants who do




not speak the official language and the evaluation of cultural homogeneity or contacts with immigrants. Model 4 shows that the effect of the proportion of Muslims increases if people do not have immigrant friends. The latter may be due to a self-selection bias in which those who are willing to make immigrant friends are originally less xenophobic. This tendency is stronger when the Muslim population is large. However, the result is still important because the interaction terms between the size of the Muslim population and the neighbourhood contact does not have any effect, even though the indicator of neighbourhood does not refer to contact with any specific group. This result suggests that casual acquaintance with Muslims does not weaken negative attitudes towards them. The obvious next step is to interpret these results using an American research tradition, where Dixon (2006) points out that prejudice towards African-Americans is not weakened through casual acquaintance with African-Americans. In order to reduce the prejudice, friendly feelings towards the out-group are required, whereas prejudice towards Asian-Americans and Hispanics is weakened through casual relations. Muslims might therefore occupy the same position in Europe as African-Americans do in the US.

Concluding discussion

In this study, we set out to develop a classical theme of empirical research with group threat theory, namely the argument that the size of the minority population threatens the majority population. By analysing cross-national data, we develop it in two aspects. First, we examine what kinds of threat matter. Previous research has taken us a long way in showing how various threats affect anti-immigrant attitudes, but we still lack a solid understanding of the sources of those threats. For example, we know that anti-immigrant attitudes tend to be more prevalent in poor economic contexts, but why is this? Is the explanation the obvious one that it is about the competition for scarce resources between groups, or can it be because the articulation related to cultural differences between groups becomes more salient during a recession? Although there have been some studies that elaborate the categorization of various threats, most of them examine the effects of these differences in terms of perception (McLaren, 2003; Sniderman and Hagendoorn, 2007), but not the effects of different objective sources of threats. In the end, previous research has had difficulty identifying what stimulates threat perception and thereby strengthens xenophobia. Therefore, it has not been proven thus far whether these different sources of threats actually affect anti-immigrant attitudes or are in fact mediated by other factors. Second, we contribute to explaining how the size of the immigrant population affects xenophobia. In previous research, the effect of the size of the minority population has not provided any clarity because the results point in different directions depending on both the type of population and the unit of analysis. We tried to develop group threat theory in these aspects by examining the effects of the specific threats (economic threat and cul-tural threat) that are directly related to the composition of the immigrant group: manual workers and the unemployed among immigrant populations, immigrants who do not linguistically assimilate into the majority society and Muslim immigrants.

The main findings can be summarized in three points. First, we have shown that the composition of the immigrant population is of utter importance for the population size argument to be valid. Objective cultural threat (the size of the Muslim population) does matter in terms of strengthening xenophobia among the majority population. The latter



Hjerm and Nagayoshi 17T

able

2.

Mul

tilev

el a

naly

sis

of x

enop

hobi

a M0

M0a

M1

M2

M3

M4

BSE

BSE

BSE

BSE

BSE

BSE

Inte

rcep

t49

.82

**0.

71

34.4

3 **

0.68

34

.51

**0.

54

34.5

4 **

0.57

34

.43

**0.

61

34.4

3 **

0.52

In

divid

ual l

evel

Age

.00

0.00

0.

00

0.00

0.

00

0.00

0.

000.

00

0.00

0.

00

Mal

e-0

.39

**0.

11

-0.3

8 **

0.11

-0

.39

**0.

11

-0.3

8 **

0.11

-0

.39

**0.

11

Educ

atio

n-0

.51

**0.

16

-0.5

2 **

0.02

-0

.52

**0.

02

-0.5

1 **

0.02

-0

.51

**0.

02

Man

ual w

orke

rs1.

62 *

*0.

18

1.52

**

0.23

1.

47 *

*0.

23

1.61

**

0.18

1.

61 *

*0.

18

Une

mpl

oyed

2.38

**

0.27

1.

92 *

*0.

45

1.90

**

0.46

2.

36 *

*0.

27

2.36

**

0.27

N

ot in

labo

ur m

arke

t0.

25 +

0.13

0.

15

0.23

0.

12

0.21

0.

23 +

0.13

0.

23 +

0.13

C

onta

ct a

s a

frie

nd-3

.09

**0.

12

-3.0

7 **

0.12

-3

.07

**0.

12

-3.0

6 **

0.25

-3

.03

**0.

23

Nei

ghbo

urho

od c

onta

cts

-0.5

4 **

0.11

-0

.54

**0.

11 -0

.54

**0.

11

-0.5

2 *

0.19

-0

.55

**0.

19

Imm

igra

nt’s

ori

gin

-1.1

0 **

0.21

-1

.09

**0.

21 -1

.09

**0.

21

-1.0

6 **

0.21

-1

.07

**0.

21

Eval

uatio

n of

cul

tura

l hom

ogen

eity

0.86

**

0.05

0.

86 *

*0.

05

0.86

**

0.05

0.

86 *

*0.

12

0.86

**

0.11

G

roup

leve

lSi

ze o

f for

eign

-bor

n po

pula

tion

0.01

0.

10

0.07

0.

09

GD

P (U

S$10

00/p

er c

apita

)-0

.19

**0.

06 -0

.15

**0.

09

–0.1

5 **

0.04

–0

.18

**0.

04

Ove

rrep

rese

ntat

ion

rate

in m

anua

l wor

k2.

88

1.68

O

verr

epre

sent

atio

n ra

te in

une

mpl

oyed

-0.8

3 1.

12

0.00

0.06

Rat

e of

ling

uist

ic n

on-a

ssim

ilatio

n0.

72

0.25

R

ate

of M

uslim

sIn

tera

ctio

n te

rmO

verr

epre

sent

atio

n ra

te in

man

ual w

ork

M

anua

l wor

kers

1.25

*0.

60

U

nem

ploy

ed1.

82

1.16

Not

in la

bour

mar

ket

0.57

0.

59

Ove

rrep

rese

ntat

ion

rate

in u

nem

ploy

ed

Man

ual w

orke

rs0.

64 +

0.32

Une

mpl

oyed

0.84

0.

66

N

ot in

labo

ur m

arke

t0.

66 *

0.29

(Con

tinue

d)




Tab

le 2

. (C

ontin

ued)

M0

M0a

M1

M2

M3

M4

BSE

BSE

BSE

BSE

BSE

BSE

Rat

e of

ling

uist

ic n

on-a

ssim

ilatio

n

Belie

f in

cultu

ral h

omog

enei

ty-0

.02

0.01

Con

tact

as

a fr

iend

-0.0

1 0.

02

ne

ighb

ourh

ood

cont

acts

0.03

0.

02

Prop

ortio

ns o

f Mus

lims

Be

lief i

n cu

ltura

l hom

ogen

eity

0.08

0.

05

C

onta

ct a

s a

frie

nd–0

.24

*0.

11

N

eigh

bour

hood

con

tact

s–0

.01

0.09

Ra

ndom

effe

cts

Cou

ntry

leve

l 10

.88

**3.

21

9.32

**

2.72

5.

78 *

*2.

40

6.35

**

2.52

7.

51 *

*2.

74

5.33

**

2.31

(% e

xpla

ined

com

pare

d to

the

in

terc

ept

mod

el)

14.3

4 46

.85

41.6

7 30

.98

51.0

6

Indi

vidu

al le

vel

86.2

69.

29

77.2

2 8.

84

76.9

8 8.

77

76.9

9 8.

77

76.6

9 8.

76

76.7

0 8.

76

(%

exp

lain

ed c

ompa

red

to t

he

inte

rcep

t m

odel

)10

.48

10.7

6 10

.75

11.1

0 11

.09

Mod

el fi

tD

evia

nce

1999

87.1

719

6976

.93

1969

21.3

1 19

6925

.52

1968

66.5

919

6848

.36

Para

met

er2

211

1111

11

Not

es: N

= 2

7,39

9, *

* p <

.01,

* p

< .0

5, +

p <

.1. A

ge, e

duca

tion,

and

per

cept

ion

of c

ultu

ral t

hrea

t ar

e ce

ntre

d ar

ound

gra

nd-m

ean.




is an important finding in itself, because the sources of cultural threat have not been the focus of much research. Economic threats that are related to the composition of the immigrant population showed no effect on xenophobia. The latter does not refute the idea that economic threats are of no significance (since there is a negative correlation between GDP and xenophobia), but it implies that the economic threat that comes from the specific composition of the immigrant population does not tend to increase xenopho-bia. In other words, when it comes to the general effects of compositional threats on the whole population, it is obvious that it is the cultural threat that matters. In previous research, the perception of economic threat as a collective-level threat has been shown to have an effect on xenophobia (McLaren and Johnson, 2007; Sniderman and Hagendoorn, 2007). However, this collective economic threat does not stem from the actual existence of split labour markets. Moreover, Semyonov et al. (2006) and Hjerm (2007b) find that the political environment affects xenophobia. From these results, we can assume that worsening economic conditions stimulate the anxiety of host residents about the future of a country, and that radical right-wing parties associate this anxiety with immigrants. In other words, immigrants are perceived as an economic threat not just because their numbers are increasing, but also because they are articulated as an economic threat by, for example, political parties (e.g. see Berezin, 2009). Therefore, the size of the immi-grant population itself does not need to have any general effect on xenophobia. For example, Skenderovic (2007) argues that Swiss radical right-wing parties employed a set of arguments that were basically aimed at presenting immigration as a threat to Switzerland and inducing xenophobia. The parties of the ‘movement against over-for-eignization’ in Switzerland that emerged in the 1960s and early 1970s demanded the reduction of the immigration rate by applying economic and labour market arguments; these parties claimed that competition on the job market would reduce wages and that Swiss workers would be laid off more easily (Skenderovic, 2007: 168–9). They had some success with this platform in national elections and referendums. Whether this explanation, the perception of an economic threat, and xenophobia itself is influenced by radical right-wing parties, is valid or not remains a task for future research.

Second, we have shown that compositional economic threats matter to the group that really competes for the scarce resources. We have seen that general economic threats, such as the economic prosperity of a country, function to diminish anti-immigrant atti-tudes. We have also seen that specific threats matter. The idea that the working class is more threatened by a larger share of manual workers among the immigrant population is predicted by group threat theory, but has so far only received limited empirical sup-port. Xenophobic attitudes increase when people face increased competition over scarce resources. The conclusion is that the economic composition of the immigrant population is of little general importance, but is of quite substantial importance for the working class. This conclusion further corroborates group threat theory, particularly realistic group conflict theory (see Blalock, 1957; Sherif and Sherif, 1953).

Third, we have brought some clarity to the effect of the cultural composition of the immigrant population. Previous research from the US has generally shown that whites are more prejudiced when the African-American population is large, whereas similar results are not as obvious in a European context. In this context, the scenario more closely reflects the relationship between whites and immigrants in the US. We have shown that it is not the language differences that matter, because the proportion of immigrants who




do not speak the majority language had no effect on xenophobia; nor did the language indicator interact with an individual level of exposure to those threats. On the contrary, it was the proportion of the Muslim population that mattered because the larger the Muslim population is, the more xenophobic the people are, which indicates that the immigrant population can be perceived as a cultural threat under specific conditions. Moreover, people who have immigrant friends become far less xenophobic if the proportion of the Muslim population is large, whereas no such effect exists for people who live in immi-grant-dense neighbourhoods in our analyses. Our interpretation of the latter is that casual contacts do not matter in relation to Muslims but deeper contact does, which means that the relationship between populations of Muslims and the majority populations in Europe mirrors the relationship between whites and African-Americans in the US. Blumer (1958) emphasized that the sense of group position, which is the basis of prejudice, is a historical product. It is formed through communication between members of dominant groups and stimulated by a ‘big event’ and interest groups. Arguments that encompass Muslims in Europe and conflicts between them may work to put Muslims in a special position in the group perceptions of Europeans. However, we need further research to examine the extent to which we can state that the positions of African-Americans in the US and those of Muslims in Europe are similar.

Pettigrew (2000) claims that the individual-level predictors of prejudice (or anti-immigrant attitudes) are consistent over geographical areas, across groups and over time. On the group level, this consistency across geographical areas has not been obvious, since the relationship between whites and African-Americans in the US could not be fully mirrored elsewhere until now. Further research is required to examine the extent to which these two relationships are similar, but our results provide us with one piece of the puzzle of how to think about consistency of the group-level predictors.

We admit that there are some limitations to this research. Our indicators of threat may require further elaboration. There is a problem with our indicators of sources of cultural threat. For example, it would be better to recreate the rate of linguistic non-assimilation to indicate the frequency of use of official language in public space, but we currently have no such variable at our disposal. So, even though we do not find any effects on xenophobia it is too early to dismiss all effects of language diversity. Still, we think we are making an important contribution in trying to examine objective sources of cultural threats that are tangible, and not only examining the subjective threat that people per-ceive just because they may be prejudiced from the start. Moreover, we would like to have a larger second-level sample size to be able to generalize the results. For example, our finding about the role of Muslims in Europe is interesting, but it would be fruitful to examine this in other countries outside Europe and/or, alternatively, at another political level such as regions.

Bearing in mind the limitations we must also acknowledge that our three main find-ings are worth repeating. We refer to the results that the cultural threat has objective sources relating to the composition of minority populations, whereas economic threat on a collective level does not. Moreover, people who are directly threatened both economi-cally and culturally are more likely to be xenophobic. Furthermore, Muslims in Europe may be perceived by the majority as African-Americans are in the US. The latter contrib-utes to the support of the universal applicability of group threat theory.




App

endi

x: S

umm

ary

of c

ont

extu

al le

vel v

aria

bles

and

xen

oph

obi

a

Cou

ntry

NEc

onom

ic t

hrea

tC

ultu

ral t

hrea

tSi

ze o

f fo

reig

n-bo

rn

popu

latio

n

GD

P (U

S$10

00

per

capi

ta)

Ove

rrep

rese

ntat

ion

ra

te in

man

ual w

ork

Ove

rrep

rese

ntat

ion

rate

in u

nem

ploy

edR

ate

of li

ngui

stic

no

n-as

sim

ilatio

nPr

opor

tion

of M

uslim

s

Aus

tria

1309

1.75

2.

00

8.01

2.

95

13.8

3 24

.82

Belg

ium

1284

1.16

2.

17

7.12

3.

59

12.0

0 24

.47

Switz

erla

nd13

351.

50

2.94

8.

91

4.30

25

.10

36.8

4 C

zech

Rep

ublic

722

1.09

1.

54

18.8

2 0.

01

5.20

6.

95

Ger

man

y22

321.

71

2.22

8.

69

4.47

12

.46

24.1

3 D

enm

ark

1095

1.29

2.

32

5.43

2.

12

7.36

32

.96

Spai

n10

251.

23

1.20

2.

32

1.04

5.

50

16.5

9 Fi

nlan

d17

271.

12

2.20

1.

82

0.30

2.

65

24.5

8 Fr

ance

1166

1.35

1.

64

6.36

8.

21

11.6

5 23

.20

UK

1636

0.77

1.

51

1.62

2.

24

9.44

25

.57

Gre

ece

1912

2.20

0.

97

1.52

4.

65

10.7

8 12

.02

Hun

gary

1036

0.78

0.

67

1.50

0.

24

3.24

5.

92

Irel

and

1384

0.85

1.

33

0.00

0.

52

10.9

7 30

.21

Ital

y84

61.

30

1.14

2.

23

1.20

4.

13

20.4

0 Lu

xem

bour

g51

81.

80

1.58

56

.49

1.01

36

.58

46.7

5 N

ethe

rlan

ds18

111.

40

2.53

4.

23

4.93

11

.19

25.8

9 N

orw

ay16

581.

24

2.76

3.

45

1.93

8.

34

39.8

1 Po

land

1408

0.70

0.

66

2.38

0.

01

2.40

4.

67

Port

ugal

951

0.83

1.

13

1.00

0.

24

6.74

11

.56

Swed

en14

101.

41

2.68

10

.46

2.01

14

.45

25.4

6 Sl

oven

ia93

41.

42

0.85

12

.12

1.29

10

.68

11.0

8 27

,399

Ave

rage

1.28

1.

72

7.83

2.

25

10.7

0 22

.57 (C

ontin

ued)




App

endi

x: (

Con

tinue

d)

Cou

ntry

Xen

opho

bia

Cro

nbac

h’s

αFa

ctor

load

ing

12

34

56

Aus

tria

0.86

8 0.

740

0.80

8 0.

797

0.77

9 0.

801

0.73

7 Be

lgiu

m0.

828

0.71

8 0.

727

0.79

1 0.

672

0.80

8 0.

687

Switz

erla

nd0.

804

0.64

9 0.

663

0.79

6 0.

761

0.76

8 0.

631

Cze

ch R

epub

lic0.

849

0.72

0 0.

728

0.84

4 0.

770

0.79

0 0.

668

Ger

man

y0.

834

0.73

7 0.

721

0.74

2 0.

723

0.83

5 0.

683

Den

mar

k0.

844

0.61

5 0.

710

0.84

0 0.

803

0.82

5 0.

684

Spai

n0.

821

0.73

1 0.

701

0.76

8 0.

756

0.79

1 0.

605

Finl

and

0.80

9 0.

693

0.62

1 0.

814

0.68

0 0.

813

0.67

4 Fr

ance

0.86

9 0.

724

0.73

7 0.

830

0.83

3 0.

852

0.68

8 U

K0.

883

0.71

0 0.

796

0.84

2 0.

811

0.85

6 0.

744

Gre

ece

0.86

7 0.

793

0.73

2 0.

813

0.81

4 0.

769

0.74

9 H

unga

ry0.

841

0.75

0 0.

726

0.82

3 0.

726

0.81

0 0.

648

Irel

and

0.83

5 0.

642

0.68

7 0.

832

0.79

9 0.

846

0.61

6 It

aly

0.75

0 0.

549

0.65

0 0.

757

0.73

5 0.

736

0.57

9 Lu

xem

bour

g0.

706

0.57

9 0.

691

0.72

6 0.

722

0.69

4 0.

361

Net

herl

ands

0.80

0 0.

600

0.72

1 0.

816

0.69

0 0.

752

0.65

5 N

orw

ay0.

801

0.61

5 0.

701

0.79

7 0.

762

0.76

3 0.

595

Pola

nd0.

813

0.72

1 0.

742

0.80

3 0.

694

0.74

6 0.

602

Port

ugal

0.83

4 0.

753

0.61

1 0.

802

0.78

5 0.

830

0.67

1 Sw

eden

0.81

9 0.

665

0.74

1 0.

845

0.73

1 0.

829

0.52

9 Sl

oven

ia0.

805

0.72

4 0.

658

0.78

5 0.

685

0.78

0 0.

636

Ave

rage




Funding

This research is sponsored by the Swedish Council for Working Life and Social Research (DR 2007-0004) and by Vetenskapsrådet (DR: 2008-1314).

Notes

1. See Esposito and Murphy (1999) for a critique of contemporary ways of empirically testing Blumer’s group threat theory.

2. The size argument was developed much earlier by Moreno (1953 [1934]) in his 1934 theory of saturation point, where a certain proportion of the minority group is needed for the majority to treat them negatively. This idea has now disappeared and the size argument is simply put forward in a supplementary way.

3. There are, of course, studies that hint that the socioeconomic status indicators do not always affect anti-immigrant attitudes (e.g. Fetzer, 2000; Halperin et al., 2007).

4. It is important to note that the proportion of Muslims is an operationalization of a possible cultural threat that can trigger xenophobic attitudes not only towards Muslims but towards all immigrants. The latter can be compared to economic threat where the presence of unemployed immigrants may or may not trigger xenophobia against all immigrants, not only against those who are unemployed.

5. Some may feel that the preference for cultural homogeneity and xenophobia overlap widely or that causality between these two attitudes is opposite – that people prefer cultural homoge-neity, since they have xenophobic attitudes. However, in this article we regard preference for cultural homogeneity as in-group (i.e. Europeans) preference, while xenophobia is regarded as negative attitudes towards immigrants. Since we operationalized the preference as described in note 12, those who hate immigrants regardless of country of origin become as low in the preference as those who like immigrants regardless of country of origin. Social identity theory assumes that people favour their own group over out-groups because people seek to raise their self-esteem (Tajfel and Turner, 1979). According to Blumer (1958), this in-group favouritism does not always provoke xenophobia even though it works as a basis for it – which implies that the preference exists independently of actual threat. McLaren and Johnson (2007) have found that racist attitudes do not affect xenophobia. In fact, in our research, the correlation between xenophobia and the preference for cultural homogeneity is quite small: only 0.100 (N = 27,399). Moreover, our results indicate that individual economic conditions as well as level of economic threat affect xenophobia even after controlling for this preference. In addi-tion, the social demographic variables and the group-level variables in this research show only weak, if any, effect on the preference for cultural homogeneity (the results are available from the authors). Furthermore, the main results in this research do not change even after omitting this indicator. Therefore, we use the preference as an independent variable.

6. Although the low response rate of Switzerland seems to be problematic, the main results in this research do not change even after eliminating Switzerland. Detailed results are available from the authors.

7. Correlations with six different statements about immigration and the index vary between 0.5 and 0.7.

8. A summary of macro-level variables is shown in the Appendix. GDP and unemployment rate are from 2002. Proportions of foreign-born population are derived from 2003 data for Slov-enia and from a census data of 2000 an average of 1998–2002 for the other countries.

9. ‘Craft and related trade workers’, ‘plant and machine operators and assemblers’ and ‘elemen-tary occupations’ are categorized as ‘manual workers’.

10. Specifically, we define official language in each country in the following way: German for Austria; Dutch, French and German for Belgium; Czech for the Czech Republic; Danish for




Denmark; Finnish, Sami and Swedish for Finland; French for France; German for Germany; Greek for Greece; Hungarian for Hungary; Irish and English for Ireland; Italian, German, French and Sardinian for Italy; Luxembourgian for Luxembourg; Dutch for the Netherlands; Norwegian for Norway; Polish for Poland; Portuguese for Portugal; Spanish, Catalan and Galician for Spain; Swedish for Sweden; Slovenian for Slovenia; English, Welsh and Scottish Gaelic for the UK. For all countries, ‘others’ is included for those who do not have official languages. In the Netherlands, the answers are given as multiple answers. Thus, we need to notice that the meaning of the proportion is different from those in other countries.

11. The data were downloaded from the Association of Religion Data Archives (www.thearda.com) and were collected by the Association of Religion Data Archives, by Roger Finke and Brian J Grim.

12. The six items are: (1) To what extent do you think [country] should allow people of the same race or ethnic group as most [country] people to come and live here? (2) How about people of a different race or ethnic group from most [country] people? (3) To what extent do you think [country] should allow people from the richer countries in Europe to come and live here? (4) How about people from the poorer countries in Europe? (5) To what extent do you think [country] should allow people from the richer countries outside Europe to come and live here? (6) How about people from the poorer countries outside Europe?

Statements range from ‘allow many to come and live here’ (= 1) to ‘allow none’ (= 4). The three questions referring to non-European immigrants (2, 5 and 6) and the other three ques-tions referring to European immigrants (1, 3 and 4) were added separately before the latter was subtracted from the former. This creates a theoretical index that varies between –9 and 9. A suspicion is that this measure of cultural threat in fact measures the same thing as the xeno-phobia index. The latter is not the case as the two measures clearly make up two different factors if run in a factor analysis. Results are available from the authors.

13. The original question on friendship contacts is ‘Do you have any friends who have come to live in [country] from another country?’ The three possible answers are ‘yes, several’, ‘yes, a few’, or ‘no, none at all’. The original question on neighbourhood contacts is ‘How would you describe the area where you currently live?’ The three possible answers are ‘an area where almost nobody is of a different race or ethnic group from most [country] people’, ‘some people are of a different race or ethnic group from most [country] people’, or ‘many people are of a different race or ethnic group’.

14. Original data include samples who are under 16 years old or over 90 years old. However, to keep cross-country equivalence and reliability of answers, these samples are excluded from the analysis.

15. This result stems from the difference between men and women in relation to the possibility of facing an economic threat. If we take a closer look at this effect, we find that women are more averse to immigrants than men are if the overrepresentation rate of immigrants in manual work is low (results not in the table). In contrast, native men are more negative towards immi-grants when there is an increased possibility that immigrants are manual workers. Females are more likely to work in non-manual work such as office workers, receptionists, or service workers. They therefore feel less threatened by a large proportion of foreign manual workers.

16. Some may suspect that the degree of linguistic assimilation affects xenophobia intermediated by preference for cultural homogeneity. However, when we analyse the effects of the threat of linguistic assimilation on preference for cultural homogeneity, there is no significant effect. Furthermore, a similar result was measured in relation to the percentage of Muslims. This means that preference for cultural homogeneity is independent of these real cultural threats.

17. Results available from the authors.




References

Alesina A, Devleeschauwer A, Easterly W, Kurlat S and Wacziarg R (2003) Fractionalization. Journal of Economic Growth 8(2): 155–194.

Ålund A and Schierup C-U (1991) Paradoxes of Multiculturalism: Essays on Swedish Society. Aldershot and Brookfield: Avebury.

Arends-Tóth J and Van de Vijver FJR (2003) Multiculturalism and acculturation: Views of Dutch and Turkish-Dutch. European Journal of Social Psychology 33(2): 249–266.

Berezin M (2009) Illiberal Politics in Neoliberal Times: Culture, Security and Populism in the New Europe. Cambridge: Cambridge University Press.

Blalock HM (1957) Percent non-white and discrimination in the South. American Sociological Review 22(6): 667–682.

Blalock HM (1967) Towards a Theory of Minority Group Relations. New York: Wiley.Blumer H (1958) Race prejudice as a sense of group position. Pacific Sociological Review

1(1): 3–7.Bobo L (1983) Whites’ opposition to busing: Symbolic racism or realistic group conflict. Journal

of Personality and Social Psychology 45(6): 1196–1210.Bonacich E (1972) A theory of ethnic antagonism: The split labor market. American Sociological

Review 37(5): 547–559.Boswell TE (1986) A split labor market analysis of discrimination against Chinese immigrants,

1850–1882. American Sociological Review 51(3): 352–371. Burns P and Gimpel JG (2000) Economic insecurity, prejudicial stereotypes, and public opinion on

immigration policy. Political Science Quarterly 115(2): 201–225.Coenders M and Scheepers P (2003) The effect of education on nationalism and ethnic exclusionism:

An international comparison. Political Psychology 24(2): 313–343.Crepaz M (2008) Trust beyond Borders: Immigration, the Welfare State, and Identity in Modern

Societies. Ann Arbor, MI: University of Michigan Press.Dixon JC (2006) The ties that bind and those that don’t: Towards reconciling group threat and

contact theories of prejudice. Social Forces 84(4): 2179–2204.Dixon JC and Rosenbaum M (2004) Nice to know you? Testing contact cultural, and group threat

theories of anti-black and anti-Hispanic stereotypes. Social Science Quarterly 85(2): 257–280.Dustmann C (1994) Speaking fluency, writing fluency and earnings of migrants. Journal of

Population Economics 7(2): 133–156. Espenshade TJ and Hempstead K (1996) Contemporary American attitudes toward US immigra-

tion. International Migration Review 30(2): 559–576.Esposito L and Murphy WJ (1999) Desensitizing Herbert Blumer’s work on race relations:

Recent applications of his group threat theory to the study of contemporary race prejudice. Sociological Quarterly 40(3): 397–410.

Esser H (2006) Migration, language and integration. AKI Research Review 4.Esses V, Dovido L, Jackson LM and Armstrong TL (2001) The immigration dilemma: The role of

perceived group competition, ethnic prejudice, and national identity. Journal of Social Issues 57(3): 389–412.

Eurostat (2008) Population by Citizenship and Country of Birth. Available at: epp.eurostat.ec.europa.eu/portal/page/portal/population/data/database

Fetzer JS (2000) Public Attitudes towards Immigration in the United States, France and Germany. Cambridge: Cambridge University Press.

Fosset M and Kiecolt JK (1989) The relative size of minority populations and white racial attitudes. Social Science Quarterly 70(4): 820–835.

Garner CL and Raudenbush SW (1991) Neighborhood effects of educational attainment: A multilevel analysis. Sociology of Education 64(4): 251–262.




Giles M and Hertz K (1994) Racial threat and partisan identification. American Political Science Review 88(2): 317–326.

Glaser JM (1994) Back to the black belt: Racial environment and white racial attitudes in the South. Journal of Politics 56(1): 21–41.

Goldberg DT (2006) Racial Europeanization. Ethnic and Racial Studies 29(2): 331–364.Gordon MM (1964) Assimilation in American Life: The Role of Race, Religion, and National

Origin. New York: Oxford University Press. Green EGT (2007) Guarding the gates of Europe: A typological analysis of immigration attitudes

across 21 countries. International Journal of Psychology 42(6): 365–379.Halperin E, Canetti-Nisism D and Pedahzur A (2007) Threatened by the uncontrollable:

Psychological and socio-economic antecedents of social distance towards labor migrants in Israel. International Journal of Intercultural Relations 31(4): 459–478.

Hello E, Scheepers P, Vermulst A and Gerris JRM (2004) Association between educational attainments and ethnic distance in young adults. Acta Sociologica 47(3): 253–275.

Hjerm M (2005) What the future may bring: Xenophobia among Swedish adolescents. Acta Sociologica 48(4): 292–307.

Hjerm M (2007a) Are we the people? National sentiments in a changing political landscape. In: Svallfors S (ed.) The Political Sociology of the Welfare State: Institutions, Cleavages, Orientations. Stanford, CA: Stanford University Press, 223–257.

Hjerm M (2007b) Do numbers really count? Group threat theory revisited. Journal of Ethnic and Migration Studies 33(7): 1253–1276.

Hood MV and Morris IL (1997) Amigo o enemigo? Context, attitudes, and Anglo public opinion toward immigration. Social Science Quarterly 78(2): 309–323.

International Monetary Fund. (2002) World Economic Outlook Database 2002. September. www.imf.org/external/pubs/ft/weo/2002/02/data/

Jackman MR and Crane M (1986) Some of my best friends are black . . . interracial friendship and whites’ racial attitudes. Public Opinion Quarterly 50(4): 459–486.

Jasinskaja-Lahti I (2008) Long-term immigrant adaptation: Eight-year follow-up study among immigrants from Russia and Estonia living in Finland. International Journal of Psychology 43(1): 6–18.

Kymlicka W (2001) Politics in the Vernacular: Nationalism, Multiculturalism, and Citizenship. Oxford: Oxford University Press.

McLaren LM (2003) Anti-immigrant prejudice in Europe: Contact, threat perception, and preferences for the exclusion of immigrants. Social Forces 81(3): 909–936.

McLaren LM and Johnson M (2007) Resources, group conflict and symbols: Explaining anti-immigrant hostility in Britain. Political Studies 55(4): 709–732.

Moreno JL (1953 [1934]) Who Shall Survive? Foundations of Sociometry, Group Psychotherapy and Sociodrama. New York: Beacon House.

Mulder M and Krahn H (2005) Individual- and community-level determinants of support for immigration and cultural diversity in Canada. Canadian Review of Sociology and Anthropology 42(4): 421–444.

Nukaga M (2006) Xenophobia and the effect of education: Determinants of Japanese attitudes toward acceptance of foreigners. Japan General Social Surveys/Kenkyu Ronbun Shu 5: 191–202.

O’Connell M (2005) Economic forces and anti-immigrant attitudes in Western Europe: A paradox in search of an explanation. Patterns of Prejudice 39(1): 60–74.

OECD. (2008) Database on Immigrants in OECD Countries. Available at: stats.oecd.org/Index.aspx.Olzak S (1992) The Dynamics of Ethnic Competition and Conflict. Stanford, CA: Stanford

University Press.




Park RE (1950) Race and Culture. Glencoe, IL: Free Press.Paxton P and Mughan A (2006) What’s to fear from immigrants? Creating an assimilationist threat

scale. Political Psychology 27(4): 549–568.Pettigrew TF (1957) Demographic correlates of border-state desegregation. American Sociological

Review 22(6): 683–689.Pettigrew TF (1998) Intergroup contact theory. Annual Review of Psychology 49: 65–85.Pettigrew TF (2000) Systematizing the predictors of prejudice. In: Sears DO, Sidanius J and

Lawrence B (eds) Systematizing the Predictors of Prejudice. Chicago: The University of Chicago Press, 280–301.

Portes A and Zhou M (1993) The new second generation: Segmented assimilation and its variants among post-1965 immigrant youth. Annals of the American Academy of Political and Social Science 530: 74–98.

Quillian L (1995) Prejudice as a response to perceived group threat: Population composition and anti immigrant racial prejudice in Europe. American Sociological Review 60(4): 586–611.

Quillian L (1996) Group threat and regional change in attitudes toward African-Americans. American Journal of Sociology 102(3): 816–860.

Raudenbush S, Bryk A and Congdon R (2001) HLM Version 5.05. Lincolnwood, IL: Scientific Software International Inc.

Rumbaut RG (1997) Assimilation and its discontents: Between rhetoric and reality. The International Migration Review 31(4): 923–960.

Scheepers P, Gijsbert M and Coenders M (2002) Ethnic exclusionism in European countries: Public opposition to civil rights for legal migrants as a response to perceived ethnic threat. European Sociological Review 18(1): 17–34.

Schneider LS (2007) Anti-immigrant attitudes in Europe: Outgroup size and perceived ethnic threat. European Sociological Review 24(1): 53–67.

Sears DO and Jessor T (1996) White’s racial policy attitudes: The role of white racism. Social Science Quarterly 61: 16–53.

Semyonov M, Raijman R and Gorodzeisky A (2006) The rise of anti-foreigner sentiment in European societies, 1998–2000. American Sociological Review 71 (June): 426–449.

Sherif M and Sherif C (1953) Groups in Harmony and Tension: An Integration of Studies on Intergroup Relations. New York: Harper.

Sigelman L, Bledsoe T, Welch S and Combs MW (1996) Making contact? Black–white social interaction in an urban setting. American Journal of Sociology 101(5): 1306–1322.

Skenderovic D (2007) Immigration and the radical right in Switzerland: Ideology, discourse and opportunities. Patterns of Prejudice 41(2): 155–176.

Smith AW (1981) Racial tolerance as a function of group position. American Sociological Review 46(5): 558–573.

Sniderman PM and Hagendoorn L (2007) When Ways of Life Collide: Multiculturalism and its Discontents in the Netherlands. Princeton, NJ: Princeton University Press.

Snijders TAB and Bosker RJ (1999) Multilevel Analysis. London: Sage.Stephan WG and Stephan CW (2001) Improving Intergroup Relations. Thousand Oaks, CA: Sage.Svallfors S (2006) The Moral Economy of Class. Stanford, CA: Stanford University Press.Tajfel H and Turner JC (1979) Integrative theory of intergroup conflict. In: Austin WG and Worchel S

(eds) The Social Psychology of Intergroup Relations. Monterey, CA: Brooks/Cole, 33–47.Taylor MC (1998) How white attitudes vary with the racial composition of local populations:

Numbers count. American Sociological Review 63(4): 512–535.Taylor MC (2000) The significance of racial context. In: Sears DO, Sidanius J and Bobo L (eds)

The Significance of Racial Context. Chicago: The University of Chicago Press, 118–136.




Van der Brug W, Fnennema M and Tillie J (2000) Anti-immigrant parties in Europe: Ideological or protest vote? European Journal of Political Research 37(1): 77–102.

Wilson TC (2001) Americans’ views on immigration policy: Testing the role of threatened group interest. Sociological Perspectives 44(4): 485–501.

Yakushko O (2009) Xenophobia: Understanding the roots and consequences of negative attitudes toward immigrants. The Counseling Psychologist 37(1): 36–66.

Zárate MA, Garcia B, Garza AA and Hitlam RT (2002) Cultural threat and perceived realistic group conflict as dual predictors of prejudice. Journal of Experimental Social Psychology 40(1): 99–105.

Zhou M (1997) Segmented assimilation: Issues, controversies, and recent research on the new second generation. International Migration Review 31(4): 975–1008.

Biographical notes

Mikael Hjerm is a Professor at the Department of Sociology at Umeå University, Sweden. He cur-rently focuses on xenophobia as well as on diversity and the welfare state.

Kikuko Nagayoshi defended her doctoral thesis in at Osaka University in March 2010 were she focused on the risks and possibilities for the multicultural welfare state. She is currently working at Tohoku University.

Résumé

Le présent article vise à explorer un thème classique d’investigation empirique au sein de la théorie des groupes, selon lequel la taille de la population minoritaire représente une menace pour le groupe majoritaire. Afin de clarifier les résultats ambivalents issus de cette perspective analytique, nous nous intéressons plus précisément à la composition du groupe minoritaire. Nous examinons des sources objectives de menaces culturelles (composition linguistique et part de musulmans) et des menaces économiques (la part d’ouvriers et de chômeurs parmi la population migrante). L’analyse aboutit à trois conclusions. Tout d’abord, dans le cas des menaces culturelles, l’effet de la taille du groupe minoritaire varie significativement selon la composition de la population migrante (part de musulmans), alors que cela ne se vérifie pas en matière de menaces économiques. Deuxièmement, en matière de menaces économiques, la composition du groupe minoritaire joue un rôle déterminant auprès du sous-groupe majoritaire qui est en situation de concurrence réelle autour de ressources rares ; les ouvriers du groupe majoritaire manifeste davantage d’attitudes xénophobes quand le groupe minoritaire comporte une part importante d’ouvriers. Enfin, nous avons clarifié quelques débats à propos de la composition culturelle des groupes minoritaires ; les relations entre les musulmans et la population européenne majoritaire reflète clairement la relation entre les blancs et les Afro-Américains aux Etats-Unis

Mots clés: Race, relations interraciales, sociologie

Resúmen

Este artículo trata de desarrollar un tema clásico de la investigación empírica dentro de la teoría de la amenaza grupal: el argumento de que el tamaño del grupo minoritario amenaza al grupo mayori-tario. Con el fin de explicar los resultados empíricos contradictorios dentro de esta versión de la teoría de la amenaza grupal, el artículo se centra en la composición de la población inmigrante. Para ello contrastamos el efecto de fuentes objetivas tanto de amenaza cultural (la composición lingüística de la población musulmana) como de amenaza económica (la proporción de individuos




de la clase trabajadora y el desempleo entre la población inmigrante). Hemos extraído tres conclu-siones principales. Primero, la composición de la población inmigrante es de gran importancia para que el argumento del tamaño grupal sea válido en el caso de las amenazas culturales (proporción de población musulmana), mientras que no tiene importancia para las amenazas económicas. Segundo, las amenazas económicas derivadas de la composición sí que importan para el grupo que compite genuinamente por los recursos escasos (la clase trabajadora es más xenófoba cuando la clase trabajadora inmigrantes es más grande). Tercero, hemos arrojado alguna luz sobre la compos-ición cultural de la población inmigrante: así, hemos mostrado que la relación entre los musul-manes y las poblaciones mayoritarias europeas reproduce la relación entre blancos y afroamerica-nos en los Estados Unidos.

Palabras clave: Raza, relaciones raciales, sociología



The composition of the minority population as a threat: Can real economic and cultural threats explain xenophobia

Documents