Prejudice and Negative Behavioral Intentions towards Refugees: The Role of Threat and Different Forms of National Identification
Levin Zühlke – van Hulzen
Universiteit van Amsterdam Programmagroep Sociale Psychologie
Studentnummer: 10004362
begeleid door: Allard R. Feddes
datum: 01-07-2016
aantal woorden abstract: 148
Abstract
The research tested a newly developed model that integrated strength of national identification and different national identity types (ethnic, cultural and civic) to explain the relationship between inter-group threat and prejudice and behavioral intentions towards refugees. The distinction between ethnic and cultural variables was unique and new measure for identity types was employed for the first time. It was shown that in a highly representative example of Dutch nationals strength of national identification had in indirect effect on the outcomes through ethnic identification but a direct effect by itself. Type of national identification but not strength of national identification mediated the relationship of realistic threat and inter-group anxiety with prejudice and behavioral intentions. Symbolic threat had an indirect effect on prejudice and behavioral intentions through its relationship with strength of national identification which in turn related to cultural and civic identification. Not all expected relationships could be demonstrated.
Introduction
Inter-ethnic tensions are rising all over Europe (Mieriņa & Koroļeva, 2015). Many nationals seem to develop negative attitudes and even violent behavior towards immigrants, especially refugees, whereas others turn to acceptance and helping behavior. In the present study we want to investigate causes for prejudice and negative behavioral intentions towards refugees. We focus on whether the previously established relationship of perceived inter-group threat (Stephan, Stephan, Demitrakis, Yamada, & Clason, 2000) and strength of national identification (Negy, Shreve, Jensen, & Uddin, 2003; Voci, 2006) with prejudice and behavioral intentions is mediated by the type of national identification. National identification describes the extend to which a person identifies with her own national group. National identity types are different concepts that people use to define their national in-group and therefore different ways in which people identify with their national group (Billiet, Maddens, & Beerten, 2003). Billiet et al. have suggested that national identification types explain the relationship between national identification and prejudice towards immigrants. We will differentiate between ethnic national identification, cultural national identification and civic national identification (see Table 1). Most research has focused on the distinction between ethnic and civic identification, but the, to our knowledge, only study that distinguished ethnic, cultural and civic identification (Reijerse, Van Acker, Vanbeselaere, Phalet, & Duriez, 2013) found that
controlling for cultural identification the relationship between ethnic identification and prejudice lost significance. We aimed to continue research on this topic. We also tested a new scale for national identity types.
Table 1. Definitions and Characteristics of National Identity Types.
Definition Inclusiveness Attitudes towards
immigrants Ethnic ethnicity, ancestry, “blood”;
shared culture based on ethnicity
low negative correlation Cultural shared culture and values, culture
preservation;
ethnicity not relevant
low negative correlation
Civic participation in society, obeying the law; ethnicity and culture not relevant
high no correlation
Research on national identity types is still limited. It is not established with certainty how different types of identification relate to the more general concept of strength of national
identification. Because of this, there also seems to be a debate on whether national identification types are a mediator of the relationship between national identification and negative attitudes and behavioral intentions towards immigrants (Pehrson, Brown, & Zagefka, 2009; Pehrson, Vignoles, & Brown, 2009), or a moderator (Meeus, Duriez, Vanbeselaere, & Boen, 2010; Tsukamoto, Enright, & Karasawa, 2013; Verkuyten & Martinovic, 2015). We tested the model as stated in Figure 1.
Before providing details about our hypotheses we will introduce the theoretical explanations for the relevance of inter-group threat and national identification in explaining attitudes and
behavioral intentions towards other ethnic groups. We will then discuss previous research and provide some details about relevant studies before concluding with our hypotheses.
Figure 1. Predicted Model. Mediation of the Influence of Inter-group Threat on Attitudes and Behavioral Intentions towards Refugees by National Identification and National Identity Types.
Perceived Threat. Inter-group threat theory (Stephan et al., 2000) offers an account for negative outcomes of group relations. It states that realistic threat, symbolic threat and inter-group anxiety cause prejudice (see Stephan, Ybarra, & Bachman, 1999). Realistic threat relates to a
perception that material interests of one self or one's in-group are in danger, symbolic threat relates to a perception that beliefs or values of the in-group are at risk. Immigrants are often perceived as a threat to welfare resources, employment opportunities (realistic threat) and the national culture (symbolic threat). The link between inter-group threat and prejudice and behavioral intentions is well established (see Stephan et al., 2000).
National Identification. An alternative approach to inter-group relations is social identity theory (see Social Identity Theory, Tajfel & Turner, 1986; Self Categorization Theory, Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). Social identities (group memberships) are considered an important part of peoples personal identities. Once groups are formed, people go through a process of self-categorization that includes self-stereotyping to create an in-group prototype. Research shows that the existence of an in-group, that one identifies with, in relation to an out-group, that one does not identify with, can lead to in-group favoritism (see Gramzow & Gaertner, 2005). Still, for out-group derogation rather than in-group favoritism, the link with identification is not that consistent and social identity theory might not be a good explanation (Brown, 2000).
National identities are important social identities but Pehrson et al. argue that while there is a lot of research demonstrating a relationship between national identification and prejudice and negative behavioral intentions towards immigrants (Negy, Shreve, Jensen, & Uddin, 2003; Voci, 2006). (Bourhis and Dayan 2004; Negy, Shreve, Jensen, & Uddin, 2003; Pettigrew 2006; Pettigrew, Wagner, & Christ 2007; Verkuyten 2004; Voci, 2006), on a conceptual level this may not be
meaningful and also other research should be considered that contradicts these findings. It seems valid to assume that who identifies highly with the nation has a greater concern for the nation and thus is more prejudiced against immigrants. This implies that immigration is perceived as a threat, though, which may not be the case for every high identifier. Pehrson et al. point out that a study by Jackson et al. (2001) did not find a relationship between national pride and rejection of immigrants in a study of 15 European nations with more than 12000 participants.
Belgium is another example that illustrates the complexity of the relationship between national identification and prejudice and behavioral intentions towards immigrants or refugees. Being a bi-national country with especially in the Flemish part high levels of prejudice towards the other nation, Wallonia, national identification has a different meaning for the two sub-nations (see Maddens et al., 2000). Maddens et al. showed that for Flemish a high level of national identification was related to a higher level of hostility towards immigrants whereas for the Walloons the opposite was true. They argue that this is caused by a predominantly ethnic representation of the nation in Flanders in contrast to a predominantly republican (i.e. civic) representation of the nation in Wallonia (Van Dam, 1996). In the next section we will first describe different types of national
identificatoin and their relationship with prejudice and behavioral intentions. We will then further integrate these findings with research on national identification and inter-group threat.
Types of national identification
Most researchers have distinguished only between an inclusive identity, which usually is called civic or republican and an exclusive identity focusing on ethnicity and culture (Billiet, Maddens, & Beerten, 2003). The exclusive national identity can be split into a cultural and an ethnic component, though (Reijerse, Van Acker, Vanbeselaere, Phalet, & Duriez, 2013). People with an ethnic national identity view their nation as a group that shares biological features and has a common ancestry (Smith, 2001). People with a cultural national identity (Kymlicka, 2001) define their nation as a community of people who share a common culture rather than ethnicity. They consider preserving this culture an important goal for the nation. People with a civic national identity 'view their nation as a group that is committed towards each other and the institutions of their nation' (Smith, 2001), citizenship is key rather than ancestry or culture. See Table 1 for a summary.
Although conceptually different, some correlations, especially between ethnic and cultural identity, are assumed. Research has demonstrated inter-correlations between all three identity types (see Pehrson, Vignoles et al., 2009; Reijerse, Van Acker, Vanbeselaere, Phalet, & Duriez, 2013). The same literature shows that ethnic or cultural identification are linked to prejudice but not civic identification is not. In the following section research on the relationship between national identity type, national identification and attitudes towards immigrants will be presented.
National identity types and strength of national identification
Researchers offer different reasons for the proposed explanation of the relationship between national identification and negative attitudes towards refugees (Billiet et al., 2003). The previously cited study in Belgium (Maddens et al., 2000) points to a moderation. The strength of the national identification would only be related to prejudice or negative behavioral intentions for people with an exclusive (ethnic, cultural) national identity but not for people with an inclusive national identity (civic). Likewise, Pehrson, Brown and Zagefka (2009) showed that national identification was linked to negative attitudes and behavioral intentions towards refugees only if indigenous people conceived their nation in terms of ethnicity but not if they conceived their nation in terms of citizenship. They also demonstrated that a higher level of ethnic identification was related to a higher level of prejudice at a later time and that ethnic identification was related to negative behavioral intentions and a lower level of positive behavioral intentions towards refugees.
Others have proposed that people who identify highly with their nation adopt an ethnic or cultural national identity which in turn relates to negative attitudes and behavioral intentions towards immigrants (Meeus, Duriez, Vanbeselaere, & Boen, 2010; Tsukamoto, Enright, & Karasawa, 2013; Verkuyten & Martinovic, 2015). In this case national identity type would be a mediator. For example, Meeus et al. (2010) showed that in a correlational study ethnic or civic identification mediated the relationship between national identification and prejudice. In a longitudinal study they furthermore demonstrated that national identity caused higher levels of ethnic identity which caused higher levels of prejudice. For moderation we could not identify a longitudinal study, thus we assume mediation in our hypotheses.
All research previously discussed focused on ethnic (sometimes inflated with cultural) identification in comparison to civic identification. As mentioned in the introduction, cultural rather than ethnic identification may be the relevant predictor of prejudice (Reijerse et al., 2013). Reijerse et al. argue that since World War Two culture has replaced ethnicity in the discourse on nationality. It seems that in different European countries culture serves as a more important symbolic boundary between ethnic groups than race (Bail, 2008). In relation to inter-group threat theory this is relevant because symbolic threat seems to relate to cultural aspects of identification whereas realistic threat or inter-group anxiety do not appear to relate to specific aspects of national identity, although for an ethnic identity the mere presence of another ethnic group may be threatening (Smith, 2001). In the next section we will discuss research that links inter-group threat theory to national identification types.
National Identity Types and Inter-group Threat
There is not much research on the relationship between national identity types and inter-group threat. It seems that higher perceived threat is related to lowered levels of civic national identity (Ha & Jang, 2015) and heightened levels of ethnic/cultural national identity (Wright, 2011). Strength of national identification seems to be related to higher levels of perceived threat (Falomir-Pichastor & Frederic, 2013). Also it has been demonstrated that people tend to increase group homogeneity in response to threats (Castano & Yzerbyt, 1998; Doosje et al., 1995, Doosje, Spears, Ellemers, & Koomen, 1999; Simon & Klandermans, 2001). Because of this, we hypothesize that strength of national identification and national identity type are mediators of the relationship between inter-group threat and prejudice and behavioral intentions.
Present Research
between national identification, national identity type (ethnic, cultural, civic) and prejudice and behavioral intentions towards immigrants. We proposed that inter-group threat (realistic, symbolic and inter-group anxiety) would be positively related to strength of national identification which would be positively related to ethnic and cultural national identity but negatively related to civic national identity. These national identification types would in turn predict prejudice and behavioral intentions.
We expected that perceived inter-group threat, strength of national identification and ethnic and cultural national identity would bepositively related to prejudice and behavioral
intentions towards immigrants (hypothesis 1). Civic national identity was predicted to be unrelated to prejudice or behavioral intentions towards immigrants (hypothesis 2). We expected that perceived inter-group threat was positively related to strength of national identification (hypothesis 3).
Strength of national identification was predicted to be positively related to ethnic and cultural national identity (hypothesis 4), but not related to civic national identity (hypothesis 5). We expected that the relationship between strength of national identification and prejudice and
behavioral intentions towards immigrants would be mediated by national identity type (hypothesis 6): the stronger one's national identification the higher one's score on ethnic and cultural national identity and the lower one's score on civic national identity.
METHOD Participants
Based on power-analysis we aimed to recruit 200 participants, at least 161. After selection the final sample consisted of 175 Dutch nationals with an average age of 49 (range 19 to 87, SD = 16) of which 56.9% were male. Five participants had a second nationality, thirteen were born outside the Netherlands and 22 had parents that were born outside the Netherlands.
Procedure
Before data-collection started we asked for approval of the ethical commission of University of Amsterdam which was granted (2016SP6623). Participants were drawn from a random sample of the adult Dutch population of the city of Haarlem. We chose Haarlem because unlike other big cities its ethnic diversity is not much higher than the Dutch average (“Bevolking; generatie, geslacht, leeftijd en herkomstgroepering”, 2016; “Haarlem in cijfers”, 2016). Also, like most
municipalities in the Netherlands, Haarlem currently houses refugees. This sparked some protest by inhabitants who tried to prevent the arrival of refugees by starting a lawsuit. After this lawsuit was
rejected, no collective protests have been reported (see Informatierapport Politieregistraties op opvanglocaties van het COA, 2016).
A randomized list of 100 streets was made and people living in the houses numbers one to twenty were approached at their home for participation. People were asked to participate online, via e-mail or on paper. Participation was voluntary. Some streets were in the proximity of refugee housing. Every house was approached once between 16.30 and 21.00 o'clock on weekdays. Ten non-residential streets were excluded (industry areas and malls). In total 1338 houses were
approached, 706 people could be invited to participate and 367 agreed to participate of which 242 actually participated. Only participants that had Dutch nationality were included. Participants who had Dutch nationality but also an immigrant background were included. Thirty-nine participants were excluded because the survey was not completed, others because no information on age or nationality was provided.
The research was cross-sectional. All predictor and outcome variables were measured using an online or on-paper questionnaire that employed scales for the different variables. The online version was a survey programmed in Qualtrics, which was identical in content to the paper version. It contained ten sub-scales to measure perceived threat (symbolic, realistic, inter-group anxiety), strength of national identification, national identity type (ethnic, cultural, and civic), prejudice, and behavioral intentions (positive and negative). At the end some demographic variables were
measured, namely sex, age, highest level of education, nationality, country of birth, parents country of birth, employment status, job security, religion and political affiliation.
Participants first answered the scale for national identification, then the scale for national identity type, the scale for perceived symbolic and realistic threat, the scale for inter-group anxiety, the scale for prejudice towards refugees and finally the scale for behavioral intentions towards refugees. After completing the questionnaire, participants were thanked for their participation and provided with the researchers' contact information to keep them updated about the research outcome and for possible complaints.
Material
All measures were five-point Likert-scales ranging form 'strongly disagree' to 'strongly agree'.
National identification. To measure strength of national identification a translated version of the scale developed by Pehrson, Brown and Zagefka (2009) was used. The scale included three items (‘I see myself as a Dutch person’, ‘I feel good about the Dutch’, ‘I’m glad to be Dutch’). The scale had a reliability of ɑ = .88.
National identity type. The measure of ethnic, cultural and civic national identity was based on the citizenship representations scale from Reijerse et al. (2013). A question for ethnic national identity was “To consider someone a Dutch person it is important that this person has Dutch ancestors.” A question for cultural national identity was “To consider someone a Dutch person it is important that this person preserves the Dutch culture”. A question for civic national identity was “To consider someone a Dutch person it is important that this person actively participates in Dutch society.” See appendix for the full scale.
In comparison to the original scale some questions were rewritten or split into two questions and questions dropped in the original scale were included. Then a pretest was conducted to select items relevant in a Dutch population. We decided not to use the original scale because it included some questions that contained two concepts like “The person dedicates himself/herself to preserving the Dutch culture and to further developing it.”. Questions like this were split up into two separate questions because it is possible that participants have conflicting ideas about each part of the question which could result in inconsistent answers. Questions that were excluded in the previous research because they did not correlate well with one specific identification type were included again to check their validity in a Dutch sample. Regarding civic identification some questions were dropped because the original questionnaire measured citizenship representations rather than
identification. For ethnic and cultural identification this did not make a difference but the questions about civic citizenship representations included questions about political attitudes rather than civic identification. An example is 'The person accepts that members of all cultural groups may
participate in the political process.'. The original scale, the scale that we developed, the results of the pretest and the scale that was derived from the pretest are reported in appendix A.
Perceived threat. Realistic threat, symbolic threat and inter-group anxiety were measured using a translated version of the scales from Stephan, Ybarra and Bachman (1999). A sample item for realistic threat was 'Refugee immigration has increased the tax burden on Dutch people.' The scale had a reliability of ɑ = .84. A sample item for symbolic threat was 'Refugees should learn to conform to the rules and norms of Dutch society as soon as possible.' The scale had a reliability of ɑ = .75. Inter-group anxiety was assessed by emotion words like 'worried' and 'confident'. The scale had a reliability of ɑ = .88.
Prejudice. A translated and slightly adjusted version of the scale of Pehrson, Vignoles and Brown (2009) was used. A sample item was "Immigrants increase crime rates". The scale had six items. In addition, a translated version of two sub-scales from the Blatant and Subtle Prejudice Scale (Pettigrew & Meertens, 1995) was employed, namely the intimacy scale that measures willingness to be associated with immigrants, and the positive emotions scale. A sample item for
intimacy was “I would be willing to have an intimate relationship with a refugee.”; for positive emotions, participants were asked how often they had felt sympathy and admiration for refugees living in the Netherlands. Because the inter-scale correlations for all three measures of prejudice were high (r > .50, see Table 3) we decided to aggregate the scale for prejudice, intimacy and positive emotions into one measure for prejudice against refugees. The aggregated scale had a reliability of ɑ = .89.
Behavioral intentions. Behavioral intentions were measured using a translated version of the scales developed by Pehrson, Brown and Zagefka (2009). “(...) participants were asked to imagine a scenario in which a group of people in their neighborhood had formed to prevent asylum seekers from living in the area. Participants indicated on a [five-point] scale the likelihood (1 = definitely not; 3 = maybe; 5 = definitely) that they would support such a group by signing a petition, writing to [a] member of Parliament, donating money to the group and attending a demonstration (…).” (Pehrson, Brown, & Zagefka, 2009). This negative behavioral intentions scale had a reliability of ɑ = .85. The measure of positive behavioral intentions asked about support for a group that sought to 'protect the rights of asylum seekers', but was otherwise identical to the measure of negative behavioral intentions. The reliability was ɑ = .84.
Results Preliminary Analyses
Means. First we analyzed means and standard deviations of all predictor and outcome variables (see Table 2). The average for national identification was remarkably high, indicating an overall high identification with the Dutch. The average level of inter-group threat and prejudice that was reported was quite low with only symbolic threat being higher than the mid-point of the scale. It should be noted that most questions were explicitly asking about negative attitudes or emotions towards refugees. Thus, any response that does not indicate a strong disapproval could be
interpreted as a moderate expression of fear or prejudice. Regarding behavioral intentions negative intentions were very low whereas the average for positive intentions was at the mid-point of the scale indicating that almost half the participants would engage in some form of positive behavior. Table 2. Means and Standard Deviations of Outcome Variables and Predictor Variables.
M SD Range
National Identification 4.27 .90 1 - 5
Symbolic Threat 3.2 .64 1 - 5
Inter-group Anxiety 2.21 .65 1 - 5
Prejudice 2.56 .88 1 - 5
Intimacy 2.1 .86 1 - 5
Positive Emotions 2.41 .82 1 - 5
Prejudice Aggregated 2.38 .74 1 - 5
Positive Behavioral Intentions 2.99 1.02 1 - 5 Negative Behavioral Intentions 1.57 .74 1 - 5
Correlations. In Table 3 we show the inter-correlations between all predictor and outcome variables. Symbolic threat, realistic threat and inter-group anxiety were highly inter-correlated. They showed a high positive correlation with prejudice and negative behavioral intentions and a negative correlation with positive behavioral intentions. Prejudice was positively related to negative behavioral intentions and negatively related to positive behavioral intentions. Positive and negative behavioral intentions were moderately negatively related. Thus, the correlations are in line with our expectations. Only national identification did not correlate with inter-group threat or prejudice and behavioral intentions which is contrary to our expectations. The only exception was a positive association between national identification and symbolic threat. We expected positive correlations with all other measures except from positive behavioral intentions which national identification was expected to be negatively related to.
Table 3. Correlations between all predictor and outcome variables National Identification Symbolic Threat Realistic Threat Inter-group Anxiety Prejudice Negative Behavioral Intentions Positive Behavioral Intentions National Identification 1 Symbolic Threat .16* 1 Realistic Threat .05 .67** 1 Inter-group Anxiety -.08 .37** .50** 1 Prejudice .10 .74** .81** .57** 1 Negative Behavioral Intentions .07 .34** .43** .37** .48** 1 Positive -.07 -.50** -.53** -.39** -.66** -.24** 1
Behavioral Intentions
* p < .05; ** p < .01
Identity Types. We conducted a confirmatory factor analysis (three factors, Varimax) to investigate whether our predicted model of national identity types could be demonstrated. Factor correlations for ethnic identification were as predicted. All items for ethnic identification loaded on one factor. The second factor, though, included all other items except from “language proficiency”. This single item emerged as a third factor, together with “accepting Dutch Norms and Values” which also loaded high on the second factor (see Table 4).
Table 4. Items and Factor Loadings in Confirmatory Factor Analysis before Selection. Item
Dutch ancestors (Ethnic) .176 .764 -.090
Lives by laws and rules (Civic) .654 .028 -.090
Maintaining Dutch culture (Cultural) .758 .232 .203
Born in the Netherlands (Ethnic) -.046 .733 .136
Participation in society (Civic) .605 -.118 .173
Defend culture against fast change (Cultural) .742 .226 -.028 Good enough Dutch to communicate (Civic) .118 .017 .887
Grown up in Dutch family (Ethnic) .218 .740 .023
Hands culture down to next generation (Cultural) .717 .147 .307 Further developing Dutch culture (Cultural) .584 .259 .307 Accepts Dutch Norms and Values (Civic) .552 .036 .554 Cultural background and origin (Ethnic) .071 .808 .027
We decided to drop 'language proficiency' and conducted another factor analysis (three factors, Varimax). Now, the expected pattern of factor correlations occurred with exception of 'norms and values'. This item loaded on the cultural factor, not on the civic factor. We decided to accept this because this was also the case in the study which our list of items was based on (Reijerse et al., 2013), which also excluded 'language proficiency' because of low factor correlations (see appendix A). The scale now included four items for ethnic identification, five items for cultural identification and two items for civic identification. See Table 5 for the final list of items and factor loadings.
Item
Dutch ancestors (Ethnic) .033 .787 .260
Lives by laws and rules (Civic) .217 .087 .863
Maintaining Dutch culture (Cultural) .746 .215 .262
Born in the Netherlands (Ethnic) .109 .710 -.177
Participation in society (Civic) .480 -.155 .455
Defend culture against fast change (Cultural) .647 .221 .255
Grown up in Dutch family (Ethnic) .226 .736 .074
Hands culture down to next generation (Cultural) .721 .130 .302 Further developing Dutch culture (Cultural) .784 .208 -.079 Accepts Dutch Norms and Values (Cultural) .764 -.013 .059 Cultural background and origin (Ethnic) .087 .808 .021
Table 6 shows means and standard deviations for identity types. The average of cultural identification was higher than the other identity types, almost as high as the average of national identification. The average for civic identification was at the mid-point of the scale indicating that only half of the participants showed civic identification and the average for ethnic identification was a full scale-point lower than the mid-point, indicating that most people did not report ethnic national identification. This fits with the notion of Bail (2008) that culture is a more important boundary marker for ethnic groups than race.
Table 6. Means and Standard Deviations of Identification and Identity Types.
M SD
National Identification 4.27 .90
Ethnic Identification 2.09 .89
Civic Identification 3.06 .82
Cultural Identification 4.16 .81
Table 7 shows the correlations between the different types of identification and all other predictor and outcome variables. Ethnic and cultural identification were positively related but unexpectedly there also was a high positive correlation between cultural and civic identification.
National identification was not related to ethnic identification but positively related to both cultural and civic identification. For civic identification this was unexpected. Like we expected, ethnic and cultural identification were highly positively related to inter-group threat and prejudice and
modestly to negative behavioral intentions but negatively related to positive behavioral intentions. Notably, ethnic identification was particularly highly related to realistic threat and inter-group anxiety, whereas cultural identification was most highly related to symbolic threat and only
modestly related to inter-group anxiety. This seems to fit with the shared focus on culture and social norms in symbolic threat and cultural identification. Unexpectedly, civic identification was
positively related to symbolic and realistic threat and prejudice. The latter was only modest when compared to ethnic or cultural identification, but the correlation with cultural identification was high and previous studies indicated only a modest correlation between civic identification and othe identification types (Reijerse et al., 2013). Like we expected civic identification was not related to behavioral intentions.
Table 7. Correlations of National Identification and National Identity Types with Threat, Prejudice and Behavioral Intentions
Ethnic Identification Cultural Identification Civic Identification National Identification Ethnic Identification 1 Cultural Identification .335** 1 Civic Identification .124 .502** 1 National Identification .082 .244** .268** 1 Symbolic Threat .383** .461** .324** .158* Realistic Threat .477** .337** .203** .049 Inter-group Anxiety .405** .192* .126 -.083 Prejudice aggregated .523** .398** .247** .097 Negative Behavioral Intentions .314** .272** -.025 .067 Positive Behavioral Intentions -.297** -.173** -.144 -.070
* p < .05; ** p < .01
Control variables. To check for the influence of control variables on our results we conducted regression analyses that included all control variables as predictors for every predictor and outcome measure. National identification (adjusted R² = –.01; p = .48), ethnic identification (adjusted R² = .02; p = .26) and inter-group threat (adjusted R² = –.05; p = .96) appeared to be unaffected by control measures just as symbolic threat (adjusted R² < –.01; p = .47), although there, sex demonstrated a non-significant trend (ß = –.17; p = .054). Thus, men probably perceived a higher level of symbolic threat than women. Realistic threat appeared to be affected by employment status (ß = –.304; p < .05), indicating that unemployed participants experienced a higher level of realistic threat (M = 2.70) than employed people (M = 2.21) or participants that did not look for work (M = 2.10). Prejudice (ß = –.24; p < .01) appeared to be affected by highest level of education. The higher the level of education the lower the score on prejudice. Positive behavioral intentions were related to sex (ß = .24; p < .01) indicating that women showed higher levels of positive
behavioral intentions. Negative behavioral intentions were related to highest level of education (ß = –.21; p = .03) and at an almost significant level by sex (ß = –.17; p = .054), indicating that a higher level of education and being female was related to lower levels of negative behavioral intentions. Cultural identification was higher for males than for females (ß = –.18; r = .03) and lower for participants whose father was born outside of the Netherlands (ß = –.24; r = .02). Civic
identification was shown to be higher for participants that were born abroad (ß = .22; r = .02) or with a father born abroad (ß = .23; r = .03).
Because political preference and religion had many dimensions we decided to reduce
political preference into a dichotomous variable indicating left-wing or right-wing affiliations and to reduce religion into the dimensions 'christian', 'other' and 'non-religious'. We conducted another series of regression analyses and found that religion did not predict any other measure whereas political affiliation was a predictor for cultural identification (ß = -.27; r < .01), ethnic identification (ß = -.19; r = .04), realistic threat (ß = -.34; r < .001), symbolic threat (ß = -.37; r < .001), inter-group anxiety (ß = -.22; r = .02), prejudice (ß = -.43; r < .001), positive behavioral intentions (ß = .21; r = .02) and with a non-significant trend for negative behavioral intentions (ß = -.17; r = .07). This means that right-wing political identification did not relate to civic identification and national identification but was associated with a higher level of realistic and symbolic threat, inter-group anxiety, prejudice and ethnic and cultural identification and probably with negative behavioral intentions. It was associated with lower levels of positive behavioral intentions. In the following analyses we always included only the control variables that were shown to be related to the
predictor and outcome variables that were analyzed.
Hypotheses
Hypotheses 1&2. To investigate which predictors were related to the outcome measures, for every outcome measure a regression analysis was conducted that included all predictor variables.
As expected, realistic threat (ß = .39; p < .001), symbolic threat (ß = .32; p < .001), inter-group anxiety (ß = .19; p < .01) and ethnic identification (ß = .10; p = .04) emerged as predictors of prejudice. This means that a higher level of threat and ethnic identification was related to a higher level of prejudice. Civic identification (ß = .01; p = .81) did not predict prejudice. Unexpectedly, national identification (ß = .04; p = .42) and cultural identification (ß = -.02; p = .67) did not predict prejudice either.
For negative behavioral intentions inter-group anxiety (ß = .21; p = .02) and civic
identification (ß = -.24; p < .01) were predictors while realistic threat (ß = .20; p = .07) and cultural identification (ß = .17; p = .07) showed a non-significant trend. Contrary to our expectations, national identification (ß = .11; p = .18), symbolic threat (ß = .06; p = .58) and ethnic identification (ß = .03; p = .71) were no predictors of negative behavioral intentions.
Inter-group threat was a predictor of lower positive behavioral intentions with realistic threat (ß = –.28; p < .01), symbolic threat (ß = –.26; p = .01) and inter-group anxiety (ß = –.19; p = .03). Contrary to our expectations, national identification (ß = –.05; p = .47), ethnic identification (ß < -.01; p = .99) and cultural identification (ß = .14; p = .11) did not predict positive behavioral intentions. For civic identification (ß = –.01; p = .90) this was expected.
Hypothesis 3. To investigate whether inter-group threat could predict national identification a series of regressions was conducted.
For symbolic threat the model was not significant (R² = .03; adjusted R² < .01; p = .89) when including the control variable sex (ß = .01; p = .89), exclusion of this highly insignificant factor showed that collinearity between symbolic threat and political preference was a problem. Exclusion of political preference showed that symbolic threat was a predictor (ß = .16; p = .04) of national identification although the explained variance was rather small (R² = .02). This means that symbolic threat is related to national identification but this might be due to its correlation with political preference. Realistic threat was no predictor of national identification (ß = .02; p = .86). The same was found for inter-group anxiety (ß = –.10; p = .23).
Hypotheses 4&5. To measure the predictive power of strength of national identification for identity type regression analyses were conducted. Unlike we expected, ethnic identification was not predicted by national identification (ß = .06; p = .44). For cultural identification analysis showed
that national identification (ß = .20; p = .01) was a predictor. National identification was shown to be a positive, not negative like expected, predictor of civic identification (ß = .25; p < .01).
Therefore our hypothesis could not be confirmed for ethnic identification and for civic
identification the opposite relationship than expected was found. Higher levels of cultural and civic identification were related to national identification.
Hypothesis 6 was that the direct effect of national identification on prejudice and behavioral intentions would be due to mediation by national identity type. According to the Baron and Kenny (1986) two conditions need to be met for mediation analysis to be executed. First, there should be a direct effect of the expected predictor variable and the expected mediator variable on the outcome variable. Second, the expected predictor variable should be a predictor of the expected mediator variable. In the analysis of hypothesis 1 it was already established that national identification was no predictor for prejudice or behavioral intentions. To be extra sure we repeated that analysis excluding all other predictor variables and control variables. Even so, national identification did not have a direct effect on prejudice (ß = .25; p < .01), negative behavioral intentions (ß = .25; p < .01) or positive behavioral intentions (ß = .25; p < .01). Therefore the conditions of Baron and Kenny's were not met and mediation analysis was not possible.
Exploratory Analyses
Preacher and Hayes (2004) demonstrated a possibility for mediation to occur even in the absence of direct effects. Indeed, it is possible that a variable has an indirect effect on another variable through its effect on a third variable. Even if no direct effects can be observed it is still possible that the indirect effect is significant. This can be tested by bootstrapping analysis (see Preacher & Hayes, 2004). We decided to do bootstrapping analyses for the possible influence of strength of national identification on prejudice and behavioral intentions via mediation by national identity type but to also to do bootstrapping analysis for the influence of inter-group threat on prejudice and behavioral intentions checking for mediation of national identity type. A table with all results can be found in appendix C. Table 8 only shows significant results.
Table 8. Significant Results of Bootstrapping Indirect Effects Analyses for National Identification and Inter-group Threat.
Independend Variable
Dependend Variable
Mediator Lower Boundary of Confidence Interval
Upper Boundary of Confidence Interval Symbolic Threat Cultural Id. National Id. .0052 .1090
Symbolic Threat Civic Id. National Id. .0044 .1049
Symbolic Threat Prejudice Ethnic Id. .0625 .2013
Symbolic Threat Negative Intentions Ethnic Id. .0250 .1930
Realistic Threat Prejudice Ethnic Id. .0322 .1555
Realistic Threat Prejudice Cultural Id. .0171 .0864
Realistic Threat Prejudice Civic Id. .0008 .0448
Anxiety Prejudice Ethnic Id. .0849 .2612
Anxiety Negative Intentions Ethnic Id. .0137 .1771
Anxiety Prejudice Cultural Id. .0149 .1376
Anxiety Negative Intentions Cultural Id. .0049 .1202
Anxiety Prejudice Civic Id. .0016 .0853
National Id. Prejudice Cultural Id. .0242 .1642
National Id. Negative Intentions Cultural Id. .0135 .1234 National Id. Positive Intentions Cultural Id. –.1183 –.0068
National Id. Prejudice Civic Id. .0186 .1081
National Id. Positive Intentions Civic Id. –.1132 –.0032
Bootstrapping Mediation Analyses. It was shown that there was an indirect effect of national identification on prejudice and positive behavioral intentions through cultural and civic
identification but not through ethnic identification. The stronger participants identified with the Dutch, the higher they scored on prejudice and the lower they scored on positive behavioral intentions due to positive correlations of both predictor and outcome variables with cultural and civic identification. National identification had a positive indirect effect on negative behavioral intentions only through cultural identification. These results are support for hypothesis 6 although not all predicted relationships could be demonstrated, namely ethnic identification seems to be no moderator of the relationship between national identification and prejudice and behavioral
intentions whereas cultural and civic identification seem to be. For civic identification this was unexpected since it was not expected to have a direct or mediating effect.
Realistic threat and inter-group anxiety had a positive indirect effect on prejudice through all three types of national identification. Inter-group anxiety also had a positive indirect effect on negative behavioral intentions through ethnic and cultural but not civic identification. We did not formulate hypotheses regarding this relationship but these results seem to be in line with our general expectation that ethnic and cultural identification are mediators of the relationship between threat and prejudice and behavioral intentions while civic identification is not.
only through ethnic identification. A higher level of symbolic threat was related to a higher level of prejudice and negative behavioral intentions due to positive correlations of both predictor and outcome variables with ethnic identification. Also, symbolic threat had a positive indirect effect on ethnic and cultural but not civic identification through strength of national identification, indicating that a high level of symbolic threat was related to a higher level of ethnic and cultural identification because of positive correlations of all three variables with strength of national identification. These findings are in line with our model. As mentioned earlier, there was an indirect effect of national identification on prejudice and negative behavioral intentions through cultural and civic
identification, here it was shown that there was an indirect effect of symbolic threat on ethnic and cultural identification through strength of national identification. This implies that for symbolic threat and cultural identification the full series of effects as stated in our model were found. Figure 2. Results of a Series of Independent Regression Analyses and Bootstrapping Analyses for Indirect Effects
*Direct lines between to variables indicate a direct effect. Lines passing another variable indicate in indirect effect through that variable. All effects are positive, negative effects are marked by a minus in a circle.
relationship between strength of national identification and prejudice and behavioral intentions. We tested for this Results are reported in Table 9. The results of Pehrson et al. could not be replicated. None of the indirect effects of strength of national identification on prejudice or behavioral
intentions were caused by moderation since none of the interaction terms were significant.
Table 9. Results of Moderation Analyses for significant Indirect Effects of National Identification through National Identity Types on Prejudice and Behavioral Intentions
Independent Variable Dependent Variable Possible Mediator Beta Significance National Identification Negative Intentions Cultural Identification .04 .66
National Identification Positive Intentions Cultural Identification .08 .36 National Identification Prejudice Civic Identification -.01 .86 National Identification Positive Intentions Civic Identification .08 .39
Conclusion and Discussion
The results of this study show that national identity types might be more important in explaining prejudice and behavioral intentions towards refugees than strength of national
identification in general. They also might be more relevant in explaining the relationship between inter-group threat and prejudice and behavioral intentions. There was no direct effect of national identification on any outcome measure. Regarding our model most of the direct effects that we expected were not found, but further analysis showed that strength of national identification was related to prejudice and behavioral intentions towards immigrants through cultural and civic identification. Furthermore, the relationship of symbolic threat with prejudice and behavioral intentions was possibly mediated by strength of national identification and national identification type as predicted in our model. There seems to be a special relationship between symbolic threat and cultural identification. For realistic threat and inter-group anxiety the relationship with prejudice and behavioral intentions was mediated by national identity type without a link through strength of national identification.
It should be noted, though, that these are overall conclusions and quite some details of our model were not significant. The results seem promising nevertheless, especially considering that this was the first study testing an overall model that integrates strength of national identification, national identity types and inter-group threat to predict prejudice and behavioral intentions towards refugees.
Culture vs ethnicity. It seems that cultural identification was more central in explaining negativity towards refugees than ethnic identification. Ethnic identification was more related to
prejudice whereas cultural identification was more related to negative behavioral intentions. Negative behavioral intentions comprise a stronger measure of negativity towards refugees than prejudice because it measures active anti-refugee behaviors. Also, ethnic identification did not mediate the relationship between national identification and prejudice or behavioral intentions. It seems that indeed, like Bail (2008) proposed, culture has replaced race as a marker of national groups. Still ethnic identification appeared to be a factor in explaining the relationship between inter-group threat and prejudice.
Civic Identification. Civic identity played a rather unexpected role. Is was related to less negative behavioral intentions which is in line with the notion of civic identity being not a negative factor in inter-group attitudes. But indirect effect analyses showed that strength of national
identification, realistic threat and symbolic threat were indirectly and positively related to prejudice through civic identification. This does not fit with theory or previous research (Billiet et al., 2003). It might be due to the fact that our measure did not include political attitudes about citizenship as the measure of Reijerse et al. (2013) did, like 'accepting political participation of all groups'. Our items 'active participation in society' and 'abiding to laws and rules' might be interpreted in an exclusive way by some participants.
Moderation or mediation. The results fit with previous research that found a mediation of the effect of strength of national identification on attitudes and behavioral intentions (see Meeus et al., 2010). In contrast to previous research the mediation was due to an indirect effect rather than partial mediation since strength of national identification had no direct effect on the outcomes. Indirect effects can be considered a weaker form of mediation than partial mediation.
Limitations and Suggestions for Future Research
The research was a correlational study. This implies that no causality can be inferred. The study therefore provides an indication that there is a mediation of the effect of inter-group threat and strength of national identification on prejudice and behavioral intentions towards refugees by
national identity type but this should be tested in experimental or longitudinal research. Although our scale for identity types provided a measure for ethnic, cultural and civic identification, we think that the measure for civic identification needs to be improved. It included only two items and some results were found that are not compatible with any previous research. A first step might be to re-introduce the item 'origin and culture are not important in being Dutch'. We reversed this item to 'origin and culture are important' because generally negatively worded
questionnaire items should be avoided, but considering the high factor loading that was reported by Reijerse at al. this might have been to strict of an application of a general rule. It should be noted,
though, that scales like these are highly context specific and researchers should always consider the specific characteristics of the country that the measure will be used in. From personal experience we know that the presumed inability of living up to the law and integrating/participating in society can be an aspect of anti-immigrant discourse in the Netherlands.
Another limitation might be the sensibility of the highly political topic. It is likely that people with negative attitudes or behavioral intentions towards refugees did not fully report their true answers due to conscious or unconscious social desirability effects. This is especially likely regarding the measure of behavioral intentions which measured explicit political behaviors. A more subtle measure of behavioral intentions could include helping behaviors and avoidance behaviors. An alternative approach to measuring attitudes could be implicit tests.
Conclusion
Summing up, this study showed that national identity types can be important in predicting prejudice and behavioral intentions and serve as a possible explanation for the relationship between inter-group threat and prejudice and behavioral intentions towards refugees. General
group-identification measures like national group-identification may only be interpreted meaningfully when considering the way in which people conceive of their group. In dealing with the refugee influx in Europe targeting perceived threats still seems to be the most important, but promoting a more inclusive national identity could be helpful too.
Literature
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of personality and social psychology, 51(6), 1173.
Bevolking; generatie, geslacht, leeftijd en herkomstgroepering, 1 januari (2016, May 30). Retrieved from
http://statline.cbs.nl/Statweb/publication/?DM=SLNL&PA=37325&D1=1-4&D2=a&D3=0&D4=a&D5=1,3-4&D6=17-20&VW=T
Billiet, J., Maddens, B., & Beerten, R. (2003). National identity and attitude toward foreigners in a multinational state: A replication. Political Psychology, 24(2), 241-257.
Cohen, J. (1992). A power primer. Psychological bulletin, 112(1), 155.
Falomir-Pichastor, J. M., & Frederic, N. S. (2013). The dark side of heterogeneous ingroup
identities: National identification, perceived threat, and prejudice against immigrants. Journal of Experimental Social Psychology, 49(1), 72-79.
as an evaluative base. Journal of personality and social psychology, 88(5), 801.
Ha, S. E., & Jang, S. J. (2015). Immigration, threat perception, and national identity: Evidence from South Korea. International Journal of Intercultural Relations, 44, 53-62.
Haarlem in cijfers (2016, June 13). Retrieved from
http://www.cbsinuwbuurt.nl/#gemeenten2015_aantal_inwoners
Informatierapport Politieregistraties op opvanglocaties van het COA (2016). Retrieved from https://static.politie.nl/binaries/content/assets/politie/wob/00-korpsstaf/incidenten-asielbeleid/20160201---informatierapport-politieregistraties-coa-locaties.pdf
Kymlicka, W. (2001). Politics in the vernacular: Nationalism, multiculturalism and citizenship. Oxford: Oxford University Press.
Meeus, J., Duriez, B., Vanbeselaere, N., & Boen, F. (2010). The role of national identity
representation in the relation between in ‐group ide
Ethnic versus civic representation. British Journal of Social Psychology, 49(2), 305-320.
Mieriņa, I., & Koroļeva, I. (2015). Support for far right ideology and anti ‐migrant at youth in Europe: A comparative analysis. The Sociological Review, 63(S2), 183-205.
Negy, C., Shreve, T. L., Jensen, B. J., & Uddin, N. (2003). Ethnic identity, self-esteem, and ethnocentrism: a study of social identity versus multicultural theory of development. Cultural Diversity and Ethnic Minority Psychology, 9(4), 333.
Pehrson, S., Brown, R., & Zagefka, H. (2009). When does national identification lead to the
rejection of immigrants? Cross ‐sectional a
in ‐group definitions. British Journal of Social Psychology, 48(1), 61-76.
Pehrson, S., Vignoles, V. L., & Brown, R. (2009). National identification and anti-immigrant prejudice: Individual and contextual effects of national definitions. Social Psychology Quarterly, 72(1), 24-38.
Pettigrew, T. F., & Meertens, R. W. (1995). Subtle and blatant prejudice in Western Europe. European journal of social psychology, 25(1), 57-75.
Reijerse, A., Van Acker, K., Vanbeselaere, N., Phalet, K., & Duriez, B. (2013). Beyond the Ethnic ‐Civic Dichotomy: Cultural Citizenship as a New Way of Excluding Immigrants. Political Psychology, 34(4), 611-630.
Riek, B. M., Mania, E. W., & Gaertner, S. L. (2006). Intergroup threat and outgroup attitudes: A meta-analytic review. Personality and Social Psychology Review, 10(4), 336-353.
Smith, A. (2001). Nationalism: Theory, ideology, history. Cambridge, MA: Polity Press.
Snauwaert, B., Soenens, B., Vanbeselaere, N., & Boen, F. (2003). When integration does not necessarily imply integration different conceptualizations of acculturation orientations lead to
different classifications. Journal of cross-cultural psychology, 34(2), 231-239.
Stephan, C. W., Stephan, W. G., Demitrakis, K. M., Yamada, A. M., & Clason, D. L. (2000). Women's attitudes toward men: An integrated threat theory approach. Psychology of Women Quarterly, 24(1), 63-73.
Stephan, W. G., Ybarra, O., & Bachman, G. (1999). Prejudice toward immigrants1. Journal of Applied Social Psychology, 29(11), 2221-2237.
Tsukamoto, S., Enright, J., & Karasawa, M. (2013). Psychological essentialism and nationalism as determinants of interethnic bias. The Journal of social psychology, 153(5), 515-519.
Verkuyten, M., & Martinovic, B. (2015). Dual identity, in ‐group pro
feelings among ethnic minority groups. European Journal of Social Psychology.
Verkuyten, M., & Thijs, J. (2002). Multiculturalism among minority and majority adolescents in the Netherlands. International Journal of Intercultural Relations, 26(1), 91-108.
Voci, A. (2006). The link between identification and in ‐group fav
social identity and trust ‐related emotions. British Journal of Social Psychology, 45(2), 265-284.
Wright, M. (2011). Diversity and the imagined community: Immigrant diversity and conceptions of national identity. Political Psychology, 32(5), 837-862.
Zagefka, H., Nigbur, D., Gonzalez, R., & Tip, L. (2013). Why does ingroup essentialism increase prejudice against minority members?. International Journal of Psychology, 48(1), 60-68.
Appendix A. Scales for National Identity Types
Table 10. Items with Factor Loadings for National Identity Type from original Scale (Reijerse et al., 2013)
Items as tested in the pre-test: To consider somebody a Dutch person
1. … it is important that the person has Dutch ancestors. (ethnic)
2. … it is important that the person was born in the Netherlands. (ethnic)
3. … it is important that the person was raised in a Dutch family from a young age. (ethnic) 4. … it is important that the person respects the christian origin of Dutch culture. (cultural) 5. … it is important that the person respects the Dutch symbols (for example the flag). (cultural) 6. … it is important that the person respects the traditional Dutch lifestyle. (cultural)
7. … it is important that the person behaves according to Dutch norms and values. (cultural) 8. … it is important that the person protects Dutch the culture against to rapid changes.
(cultural)
9. … it is important that the person passes the Dutch culture on. (cultural) 10. … it is important that the person preserves the Dutch culture. (cultural)
11. … it is important that the person dedicates him/herself to developing the Dutch culture. (cultural)
12. … it is important that the person has been living in the Netherlands for a longer period of time. (civic)
13. … it is important that the person actively participates in the Dutch society. (civic)
14. … it is important that the person respects the fundamental laws and rules of Dutch society. (civic)
15. … it is NOT important what the persons origin or cultural background is. (civic)
(civic)
Participants in the pretest were 34 Dutch nationals with parents born in the Netherlands. Participants were not asked to provide data on sex and age. Most participants confirmed their participation afterwards, they were then asked to provide these data. Those 25 participants were between 19 and 72 years old (M = 25), thirteen were male and twelve were female. A confirmatory factor analysis for three factors (PCA with varimax rotation) was conducted to establish that the items relate to three different factors, namely ethnic, cultural and civic national identity. The Bartlett's test showed sphericity in the data (p < .000) but re-analysis with z-scores did not change this, so the analysis was continued with the original data set. The three factor solution explained 63,35% of the total variance. There were twelve items that loaded almost exclusively on either the cultural, the civic or the ethnic factor (factor loadings between .694 and .896). Four items were dropped (item 4; item 5; item 6; item 12) because they loaded on two factors. Then, a second
confirmatory factor analysis was conducted. Results are presented in Table 11. The final selection of items was:
To consider somebody a Dutch person
1. … it is important that the person has Dutch ancestors. (ethnic)
2. … it is important that the person was born in the Netherlands. (ethnic)
3. … it is important that the person was raised in a Dutch family from a young age. (ethnic) 4. … it is important what the persons origin or cultural background is. (ethnic)
5. … it is important that the person behaves according to Dutch norms and values. (cultural) 6. … it is important that the person protects Dutch the culture against to rapid changes.
(cultural)
7. … it is important that the person passes the Dutch culture on to the next generation. (cultural) 8. … it is important that the person preserves the Dutch culture. (cultural)
9. … it is important that the person dedicates him/herself to developing the Dutch culture. (cultural)
10. … it is important that the person actively participates in the Dutch society. (civic)
11. … it is important that the person respects the fundamental laws and rules of Dutch society. (civic)
12. … it is important that the person speaks Dutch well enough to communicate with others. (civic)
Item Ethnic Identification Cultural Identification Civic Identification Dutch Ancestors .218 -.090 .770
Accepts christian origin of Dutch culture .376 .430 .251
Lives according to Dutch laws and rules .199 .797 -.052
Born in the Netherlands -.121 .093 .853
Respect Dutch symbols .637 .464 -.066
Participate actively in society .335 .792 .083
Speak Dutch well enough to communicate .190 .764 .172
Grown up in Dutch family .127 .181 .727
Respects traditional Dutch lifestyle .454 .334 .039
Has been living in Netherlands for longer time .055 .427 .464 Origin and cultural background NOT important -.188 .016 -.694 Behaves according to Dutch norms and values .123 .809 .017 Protects Dutch culture against to fast changes .844 -.012 .251 Passes Dutch culture on to next generation .821 .326 .209
Further develops Dutch culture .815 .237 .078
Preserves Dutch culture .896 .170 .073
Appendix B. Measures
In hoeverre bent u het eens met de onderstaande stellingen? Geef uw antwoord op een schaal 1 tot 5 (1 = helemaal mee oneens; 3 = neutraal; 5 = helemaal mee eens).
Symbolische bedreiging
Vluchtelingen moeten leren om zich aan te passen aan de regels en normen van de Nederlandse samenleving zo snel mogelijk na hun aankomst
Immigratie van vluchtelingen ondermijnt de Nederlandse cultuur.
De normen en waarden van vluchtelingen met betrekking tot werk zijn in principe best vergelijkbaar met die van de meeste Nederlanders.
De normen en waarden van vluchtelingen met betrekking tot kwesties van moraal en religie zijn niet compatibel met de normen en waarden van de meeste Nederlanders.
De normen en waarden van vluchtelingen met betrekking tot familiezaken en de opvoeding van kinderen zijn in principe best vergelijkbaar met die van de meeste Nederlanders.
De normen en waarden van vluchtelingen met betrekking tot sociale relaties zijn niet compatibel met de normen en waarden van de meeste Nederlanders.
Vluchtelingen zouden de Nederlandse manier van omgang niet hoeven te accepteren.
Realistische bedreiging
Vluchtelingen krijgen meer van de samenleving dan dat ze bijdragen.
De kinderen van vluchtelingen zouden de zelfde rechten moeten hebben om in Nederland naar school te gaan als Nederlanders hebben.
De komst van vluchtelingen heeft de belastingdruk voor Nederlanders verhoogd. Vluchtelingen verdringen geen Nederlanders van hun banen.
Vluchtelingen zouden de zelfde rechten op gezondheidszorg moeten hebben als Nederlanders. De sociale voorzieningen voor Nederlanders zijn minder toegankelijk geworden vanwege de
immigratie van vluchtelingen.
De kwaliteit van de sociale voorzieningen voor Nederlanders is het zelfde gebleven ondanks de immigratie van vluchtelingen.
Vluchtelingen hebben het zelfde recht op sociale huurwoningen en toeslagen als arme Nederlanders.
Inter-group anxiety
Als ik in contact kom met een vluchteling voel ik me 1) bevreesd. (apprehensive)
2) onzeker. (uncertain) 3) bezorgd. (worried)
5) angstig. (anxious) 6) bedreigd. (threatened)
7) op mijn gemak. (comfortable) 8) vertrouwend. (trusting) 9) vriendelijk. (friendly) 10) zelfverzekerd. (confident) 11) veilig. (safe)
12) ontspannen. (at ease)
Vooroordeel
1) Vluchtelingen verhogen de criminaliteitscijfers.
2) Vluchtelingen zijn doorgaans goed voor de Nederlandse economie. 3) Vluchtelingen nemen banen af van mensen die in Nederland zijn geboren.
4) Vluchtelingen verbeteren de Nederlandse samenleving door nieuwe ideeën en culturen te introduceren.
5) De overheid besteedt te veel geld aan het ondersteunen van vluchtelingen.
6) De hoeveelheid vluchtelingen in Nederland vandaag de dag zou gereduceerd moeten worden.
Intimiteit
1) Als een kind van mij kinderen zou hebben met een vluchteling dan zou ik dat vervelend vinden.
2) Ik zou bereid zijn om een intieme relatie te hebben met een vluchteling.
3) Ik zou het niet erg vinden als een voldoende gekwalificeerde vluchteling werd aangenomen als mijn leidinggevende.
4) Ik zou het niet erg vinden als een vluchteling met een vergelijkbare economische achtergrond als ik lid zou worden van mijn familie door een huwelijk.
Positieve emoties
Hoe vaak heeft u zich op de volgende manier gevoeld tegenover vluchtelingen. Geef uw antwoord op een schaal van 1tot 5 (1 = zeer zelden; 3 = soms; 5 = zeer vaak).
1) Hoe vaak heeft u sympathie gevoeld voor vluchtelingen die hier wonen? 2) Hoe vaak heeft u bewondering gevoeld voor vluchtelingen die hier wonen?
Gedragsintenties
Verplaats u in het volgende scenario: In uw buurt heeft zich een groep bewoners aaneengesloten om te verhinderen dat asielzoekers worden gehuisvest in de regio. Beantwoordt de onderstaande
vragen. Geef uw antwoord op een schaal van 1 tot 5 (1 = zeer kleine kans; 3 = misschien; 5 = zeer grote kans).
Hoe groot acht u de kans dat u de groep zou ondersteunen... 1) door een petitie te tekenen?
2) door een brief naar een gemeenteraadslid te sturen? 3) door geld te doneren?
4) door deel te nemen aan een demonstratie?
Verplaats u in een ander scenario: In uw buurt heeft zich een groep bewoners aaneengesloten om asielzoekers te helpen. Beantwoordt de onderstaande vragen. Geef uw antwoord op een schaal van 1 tot 5 (1 = zeer kleine kans; 3 = misschien; 5 = zeer grote kans).
Hoe groot acht u de kans dat u de groep zou ondersteunen... 1) door een petitie te tekenen?
2) door een brief naar een gemeenteraadslid te sturen? 3) door geld te doneren?
4) door deel te nemen aan een demonstratie?
Appendix C. Full Results of Bootstrapping Indirect Effects Analyses
Table 12. Full Results of Bootstrapping Analyses for Indirect Effects of Inter-group Threat and National Identification
Independend Variable
Dependend Variable Mediator Lower Boundary of Confidence Interval Upper Boundary of Confidence Interval
Symbolic Threat Ethnic Id. National Id. –.0198 .0512
Symbolic Threat* Cultural Id. National Id. .0052* .1090*
Symbolic Threat* Civic Id. National Id. .0044* .1049*
Realistic Threat Ethnic Id. National Id. –.0055 .0376 Realistic Threat Cultural Id. National Id. –.0208 .0593
Realistic Threat Civic Id. National Id. –.0213 .0614
Anxiety Ethnic Id. National Id. –.0693 .0057
Anxiety Cultural Id. National Id. –.0914 .0138
Anxiety Civic Id. National Id. –.0848 .0157
Symbolic Threat* Prejudice Ethnic Id. .0625* .2013*
Symbolic Threat* Negative Intentions Ethnic Id. .0250* .1930* Symbolic Threat Positive Intentions Ethnic Id. –.1971 .0198
Symbolic Threat Prejudice Cultural Id. –.0226 .1033
Symbolic Threat Negative Intentions Cultural Id. –.0236 .1719 Symbolic Threat Positive Intentions Cultural Id. –.0454 .1577
Symbolic Threat Prejudice Civic Id. –.0331 .0433
Symbolic Threat Negative Intentions Civic Id. –.1285 .0116 Symbolic Threat Positive Intentions Civic Id. –.0500 .0799
Realistic Threat* Prejudice Ethnic Id. .0322* .1555*
Realistic Threat Negative Intentions Ethnic Id. –.0139 .1562 Realistic Threat Positive Intentions Ethnic Id. –.1379 .0728 Realistic Threat * Prejudice Cultural Id. .0171* .0864* Realistic Threat Negative Intentions Cultural Id. –.0028 .1205 Realistic Threat Positive Intentions Cultural Id. –.0539 .0668
Realistic Threat* Prejudice Civic Id. .0008* .0448*
Realistic Threat Negative Intentions Civic Id. –.0765 .0061 Realistic Threat Positive Intentions Civic Id. –.0530 .0254
Anxiety* Prejudice Ethnic Id. .0849* .2612*
Anxiety* Negative Intentions Ethnic Id. .0137* .1771*
Anxiety Positive Intentions Ethnic Id. –.2314 .0074
Anxiety* Prejudice Cultural Id. .0149* .1376*
Anxiety* Negative Intentions Cultural Id. .0049* .1202*
Anxiety* Prejudice Civic Id. .0016* .0853*
Anxiety Negative Intentions Civic Id. –.0648 .0105
Anxiety Positive Intentions Civic Id. –.0824 .0035
National Id. Prejudice Ethnic Id. –.0192 .1095
National Id. Negative Intentions Ethnic Id. –.0121 .0718 National Id. Positive Intentions Ethnic Id. –.0947 .0118
National Id.* Prejudice Cultural Id. .0242* .1642*
National Id.* Negative Intentions Cultural Id. .0135* .1234* National Id.* Positive Intentions Cultural Id. –.1183* –.0068*
National Id.* Prejudice Civic Id. .0186* .1081*
National Id. Negative Intentions Civic Id. –.0421 .0356 National Id.* Positive Intentions Civic Id. –.1132* –.0032* * significant results. 95% confidence interval does not include zero (equals p < .05).
