1 Variation in the Relative Size of Immigrants and Anti-Immigrant Attitudes
Explaining the Role of the Ethnic Threat Theory, the Intergroup Contact Theory and the Halo Effect in Dutch Local districts
Master Thesis
International Relations
Graduate School of Social Sciences
Name: Kasper Baarda (10823522) Supervisor: S. Tanaka
Second reader: A. van Heelsum
2
Abstract
This study analysed the contradicting ethnic threat theory and intergroup contact theory, by taking the halo effect into account. When focusing on small communities, the intergroup contact theory and halo effect could play an important role in explaining variance of anti-immigrant attitudes (Himmelroos & Leino, 2015; Schlueter & Scheepers, 2010). The halo effect confirms with the intergroup contact theory by stating that anti-immigrant attitudes are most likely to be highest in a residential context nearby a residential context with a high relative size of immigrants (Rydgren & Ruth, 2013; Bowyer, 2009). In order to examine this argument, this study tested three local districts that differ in the relative size of immigrants within the municipality of Utrecht. All selected local districts were located next to the same surrounding local district with the highest relative size of immigrants for taking into account the halo effect. The study used data that were gathered by the author on anti-immigrant attitudes of the Dutch natives in these local districts using structured surveys. This makes it possible to analyse whether the relative size of immigrants in a local district affects anti-immigrant attitudes of Dutch natives. Using a multiple regression analysis, the study did not find a relationship between the relative size of immigrants and anti-immigrant attitudes. In line with the halo effect, this study did not find support for the ethnic threat theory in this situation. On the other hand, this study did also not find significant support for the intergroup contact theory in explaining the relationship between the relative size of immigrants in a local district and anti-immigrant attitudes of Dutch natives. The study found that education was the most important predictor for explaining anti-immigrant attitudes; respondents with a lower level of education had a higher level of anti-immigrant attitudes.
Keywords: anti-immigrant attitudes, intergroup contact, ethnic threat, relative size of immigrants
3
Table of Content
1. Introduction ... 5 2. Literature Review ... 8 3. Theoretical Framework ... 13 3.1 Anti-immigrant attitudes ... 133.2 Ethnic threat theory ... 14
3.2.1 Economic threat ... 15
3.2.2 Cultural threat ... 16
3.3 Intergroup contact theory ... 18
3.4 Conceptual Model ... 21
4. Methods ... 22
4.1 Data ... 22
4.2 Research design ... 23
4.3 Procedure ... 24
4.4 Population and sample ... 25
4.5 Measurement instruments... 26
4.5.1 The dependent variable ... 27
4.5.2 The independent variables ... 27
4.5.3 Control variables ... 29
5. Results ... 31
5.1 Descriptive characteristics ... 32
5.2 Multiple regression analysis ... 34
4
5.4 Limitations and recommendations ... 48
6. Conclusion ... 51
6.1 Theoretical implications and future research ... 52
6.2 Practical implications ... 53
7. References ... 54
5
1. Introduction
Migration flows have resulted in multi-cultural societies in Western European societies (Castles & Miller, 2009). The arrival of ethnic minority groups has led to heterogeneous cities (Semyonov & Glikman, 2012). At the same time, these ethnic minority groups are not evenly spread within the cities and within the country. In the Netherlands, almost half of the immigrants live in the Western part of the country, concentrated in the four biggest cities (Amsterdam, Rotterdam, Utrecht and The Hague). Within these cities the ethnic minority groups are mostly living in specific neighbourhoods and local districts (Raets, 2010). As a consequence, the extent to which the Dutch native population comes into contact with immigrants varies in their daily lives, because the residential context (e.g. neighbourhood or local district) functions as “a pool”, where people can meet and have contact with others (Gesthuizen, 2009).
The presence of immigrants in the Netherlands and in other Western European countries has been accompanied by rising interest in anti-immigrant attitudes the last decades (e.g. Scheve and Slaughter, 2001; Mayda, 2006; Davidov et al., 2008; Gorodzeisky & Semyonov, 2015). Ethnic tensions have increased within the country due to concerns over material and economic interests and conflicting norms and values (Coenders et al., 2008). Therefore, this has become more explicitly a subject of debate and political controversies (Tolsma et al., 2009; Scholten & Van Nispen, 2008; Gijsberts & Dagevos, 2005). In other words, anti-immigrant attitudes have become politicized (Bohman, 2014).
An example of the politicized anti-immigrant attitudes is that the extreme right party in the Netherlands (“Partij van de Vrijheid”), whose selling point is their anti-immigrant position, is on the rise in the Netherlands and gained electoral successes (Boomgaarden, 2006). Struggles of extreme-right parties cover sociocultural issues as national identity, social cohesion, citizenship and the sovereign state (Rydrgen & Ruth, 2013; Castles & Miller, 2009; Putnam, 2007). A
6 consequence of these rising politicized anti-immigrant attitudes is that it is leading to preferences of more restrictive immigration policies (Cornelius & Rosenblum, 2005; Meuleman et al., 2009).
The massive influx of mainly Syrian refugees who have come to Europe and the Netherlands is another development that has made anti-immigrant attitudes a subject of interest in present-day politics. This has put even more pressure on the ethnic relations in the Netherlands, because it brought problems to light, like the costs of refugees and deviating norms and values of refugees from the majority population in the Netherlands (Gordon, 2016). Therefore, it seems it is more important than ever to focus on anti-immigrant attitudes (Callens et al., 2015).
This study would like to focus on how anti-immigrant attitudes are shaped by the residential context, as the presence of immigrants differs in people’s daily life. This has led to the following research question of this study: “How and to what extent does the relative size of immigrants in a local district influence anti-immigrant attitudes?”
Literature connects the presence of immigrants in a residential context to anti-immigrant attitudes with two conflicting theories. Firstly, the ethnic threat theory states that more presence of immigrants (a higher relative size of immigrants) in a residential context will lead to more competition over scarce resources, especially for people with less socio-economic resources as they are more vulnerable. This, in turn, induces people’s anti-immigrant attitudes. Secondly, the intergroup contact theory states that a higher relative size of immigrants in a residential context will increase the possibility to have contact with immigrants, which in turn reduces perceived ethnic threat and anti-immigrant attitudes as it increases people’s affability with immigrants (Forbes, 1997). Moreover, the halo effect confirms with the intergroup contact theory by stating that anti-immigrant attitudes are most likely to be highest in a residential context nearby a residential
7 context with a high relative size of immigrants (Rydgren & Ruth, 2013; Bowyer, 2009).
This study analysed the contradicting ethnic threat theory and intergroup contact theory, by taking the halo effect into account. When focusing on small communities, the intergroup contact theory and halo effect could play an important role in explaining variance of anti-immigrant attitudes (Himmelroos & Leino, 2015; Schlueter & Scheepers, 2010). In order to examine this argument, this study tested three local districts that differ in the relative size of immigrants within the municipality of Utrecht. All selected local districts were located next to the same surrounding local district with the highest relative size of immigrants, for taking the halo effect into account. The study used data that were gathered by the author on anti-immigrant attitudes of the Dutch natives in these local districts using structured surveys. This makes it possible to analyse whether the relative size of immigrants in a local district affects anti-immigrant attitudes of Dutch natives. Using a multiple regression analysis, the study did not find a relationship between the relative size of immigrants and anti-immigrant attitudes. In line with the halo effect, this study did not find support for the ethnic threat theory in this situation. On the other hand, this study did also not find significant support for the intergroup contact theory in explaining the relationship between the relative size of immigrants in a local district and anti-immigrant attitudes of Dutch natives. The study found that education was the most important predictor for explaining anti-immigrant attitudes; respondents with a lower level of education had a higher level of anti-immigrant attitudes.
This study contributes to existing knowledge by focusing on the level of the local district. This has not widely been analysed yet as most studies focused on larger geographical units (Himmelroos & Leino, 2015; Tolsma et al., 2009; Shvets, 2004). Therefore, this study provides an insight in the micro-social processes that play a role when focusing on the level of the local district.
8 The second contribution of this study is the empirical application to the Dutch case. The residential context could influence intergroup relations and community life, and that is why the government invests in developing policy measures that aim to improve characteristics of the residential context (Tolsma, 2009). Therefore, it is important to empirically test how the residential context influences anti-immigrant attitudes in a Dutch empirical context.
The following section presents the literature review to provide an insight into the research field and to identify the gap in the literature. Secondly, this research provides a theoretical framework, where definitions of the concepts, relationships, underlying theories and hypotheses of this study will be presented. The method section presents the research design and operationalization of the variables. The analysis section of this study presents the results of the multiple regression analysis, but as well the discussion and limitations of the analysis. Lastly, the conclusion section presents the concluding remarks of this study, and pays special attention to the theoretical implications for future research and the practical implication of this study.
2. Literature Review
Literature within the field of social science attempted to understand the indicators of anti-immigrant attitudes of natives in a host society towards ethnic minority groups. It provided a broad array of perspectives focusing on causes, effects and prevention of anti-immigrant attitudes (Blalock, 1967; Allport, 1954). The focus in previous studies has been on explaining individual and contextual factors that are necessary for understanding anti-immigrant attitudes (Pettigrew, 2006; Semyonov et al., 2006). Contextual indicators refer to the size of the immigrant population, the economic conditions and the political climate (Semyonov et al., 2008). These structural-contextual indicators are the so-called “context of reception” of immigrants in the host society (Semyonov et al., 2008). In the studies about anti-immigrant attitudes, the relative size of immigrants has been used the
9 most as determinant of anti-immigrant attitudes (Schlueter & Scheepers, 2010; Semyonov et al., 2008; Strabac & Listhaug, 2008).
Most of these studies about anti-immigrant attitudes originally derive from the United States and Canada, as these countries have a rich history for being a host society for immigrants. The focus was mainly on anti-black prejudice (Strabac & Listhaug, 2008). Since the 1970s also European literature about anti-immigrant attitudes increased significantly with most attention being paid on prejudice against immigrants (Strabac & Listhaug, 2008). Interesting is that most of the literature deriving from the United States found that the relative size of the out-group, blacks in this case, is positively related to discrimination or prejudice (see for example Pettigrew et al, 2011; Quilian, 1996). However, European literature showed more contradicting results between the relative size of immigrants and anti-immigrant attitudes of people (Semyonov et al., 2009; Rocha & Espino, 2009). Therefore, it has been a subject of debate among social scientists (Semyonov et al., 2009; Rocha & Espino, 2009).
European studies could basically be divided into three different groups, offering two plausible, contradicting explanations of how the residential context is influencing anti-immigrant attitudes. On the one hand, previous studies found that a higher relative size of immigrants in a geographical unit leads to prejudices towards immigrants and anti-immigrant attitudes in comparison with people (see for example Semyonov et al., 2008; Tolsma et al., 2009; Scheepers et al., 2002; Gijsberts, 2005; Quillian, 1995).
This effect is explained by the ethnic threat theory. The ethnic threat theory (also called the ethnic conflict theory) derives from Blalock’s (1967) writing on discrimination in the American South. The theory is grounded in the social psychology and has been used as a central explanation for anti-immigrant attitudes in previous studies (see for example Levine & Campbell, 1972; Jackson, 1993).
10 The theory of Blalock (1967) argues that when different groups compete for the same scarce resources, this could result in the experience of threat of the in-group towards the out-in-group as immigrants are considered as outsiders. More presence of immigrants would induce this competition over scarce resource, which could lead to resistance or reticence against this out-group. Therefore, immigrants are often blamed for the declining socio-economic conditions in a country (e.g. increasing unemployment (Sniderman & Hagendoorn, 2007). They are seen as the cause of unfair competition over scarce resources, which could result in out-groups being blamed and negatively valued and may affect exclusionary reactions of natives (Lubbers et al., 2002). It could stimulate people’s willingness to remove competition of the out-group, which could be done by increasing the competitiveness of the in-group, decreasing competitiveness of the out-group or lastly by avoidance of the out-group (Esses et al., 1998). In other words, immigrants could become target of discrimination, prejudice, hostility and violence (Semyonov et al., 2006; Semyonov et al., 2008).
Previous studies revealed that individual factors also affect anti-immigrant attitudes. People with less socio-economic resources or religious affection might be more prone to competition over scarce resources and people with stronger religious attachments (Semyonov et al., 2008; Gijsberts et al., 2004; Scheepers et al., 2002). For example, people with a lower level of education or income are found to be more related to anti-immigrant attitudes in comparison with people who have a higher level of education or income (Manevska & Achterberg, 2011).
On the other hand, a group of studies found a negative relationship between the relative size of immigrants and the level of negative feelings towards immigrants and prejudices (see for example Wagner et al., 2006; McLaren, 2003). This group found that, contrary to the ethnic threat theory, people who live in a residential context with a higher relative size of immigrants show less anti-immigrant attitudes.
11 This effect is clarified by the intergroup contact theory, which explains the social psychological mechanisms that relate the relative size of immigrants in a residential context to anti-immigrant attitudes (Schlueter & Scheepers, 2010. Intergroup contact has been a central topic for researchers in the field of intercultural relations and is an important determinant for explaining anti-immigrant attitudes on a subnational level (Pettigrew et al., 2011; Himmelroos & Leino, 2015). The theory derives from the 1940s and focused on the racial conflicts between black and white people in the United States (Pettigrew et al., 2011). This theory states that when the relative size of immigrants in a residential context increases the possibility to have intergroup contact (McLaren, 2003). This will improve people’s knowledge about immigrants and creates a condition to understand the immigrant better, which in turn reduces the perceived ethnic threat of in-group members (Paas & Halapuu, 2012). Direct intergroup contact stimulates affability, because intergroup contact makes it more difficult for people to accept negative stereotypes (Forbes, 1997). Study of McLaren (2003) showed that intergroup contact reduces feelings of natives to expel legal immigrants from the country. Intergroup contact could take the perceived ethnic threat away by an experience of similar belief systems (McLaren, 2003).
Lastly, another group of studies did not find any significant relationship between the relative size of immigrants in a residential context and anti-immigrant attitudes (see for example Hjerm, 2007; Himmelroos & Leino, 2015). As explained, the interethnic conflict theory and the intergroup contact theory are contradicting. This could be an explanation of the contradicting results and the often not particularly strong effect of the relative size of immigrants (see Wagner et al., 2003; Gijsberts, 2005). In addition, Rothbart and John (1993) mentioned that there are equally numbers of studies showing favourable, unfavourable and no effects of intergroup contact.
12 Thus, literature has found contradicting results for explaining the relationship between the relative size of immigrants and anti-immigrant attitudes. The ethnic threat theory states that a higher relative size of immigrants is leading to more competition over scarce resources, which induces the perceived ethnic threat to the majority population. This, in turn, induces the level of anti-immigrant attitudes. On the other hand, the intergroup contact theory assumes that a higher relative size of immigrants is leading to more possibilities to have intergroup contact, which is reducing the perceived ethnic threat. This, in turn, reduces the level of anti-immigrant attitudes. However, previous studies applied these theories mainly focused on a large geographical unit for the explanation of the relationship between the relative size of immigrants and anti-immigrant attitudes (Himmelroos & Leino, 2015; Schlueter & Scheepers, 2010). Smaller geographical units, like the residential context, are not widely analysed yet (Himmelroos & Leino, 2015; Tolsma et al., 2009). It is not clear whether the results of the studies focusing on large geographical units will also hold when focusing on a smaller geographical unit. This is, because the ethnic threat theory and intergroup contact theory lacks in clarifying the influence of the halo effect on the level of the local district. This study will extend the line of quantitative studies by testing how and to what extent micro-social processes play a role in explaining anti-immigrant attitudes when focusing on the level of the local district.
A focus on smaller geographical units could also lead to more homogeneous results, because when focusing on smaller geographical units it is likely that people live close to immigrants, which a study cannot assume when it is focusing a larger geographical unit. The local district could be crucial in explaining anti-immigrant attitudes as at this level provides the possibility to meet immigrants on a daily basis. The residential context creates a filter, which may be an important predictor of how people look towards immigrants (Schlueter & Scheepers, 2010).
13 This study will also fill the gap in the literature through an empirical application to the Dutch case. It provides a more up-to-date empirical analysis of anti-immigrant attitudes in the Netherlands by providing unique data gathered using structured surveys including individual characteristic indicators and people’s attitudes towards immigrants. The residential context could influence intergroup relations and the quality of the community life (Tolsma et al., 2009). Therefore, by gaining insight in the relationship between the relative size of immigrants in a local district and anti-immigrant attitudes in a Dutch empirical context, policy measures to develop characteristics of the residential context, could be matched (Tolsma et al., 2009).
The aim of this research is to provide a better understanding of the relationship between the relative size of immigrants, perceived ethnic threat, intergroup contact and anti-immigrant attitudes. This study will focus on the local district in explaining anti-immigrant attitudes of the Dutch majority, as not all explanations for anti-immigrant attitudes are found at the individual-level. Individuals are also influenced by the residential context they live in. It is possible that some environments are more prone to anti-immigrant attitudes.
3. Theoretical Framework
In this section, this study will specify the variables used to relate the relative size of immigrants in a local district to anti-immigrant attitudes and elaborate the variables into useful theoretical concepts. Finally, this section also includes the hypotheses of this study, which are theoretically justified.
3.1 Anti-immigrant attitudes
The concept “anti-immigrant attitudes” is broad; it is referring to xenophobic and racist attitudes of the majority population in a host country towards different ethnic minority groups (Hjerm, 2009). It is not about a specific set of beliefs, but it is a predisposition to act in a particular way towards an ethnic minority group, or
14 members of that group (Himmelroos & Leino, 2015; Allport, 1954). As elaborated in the literature review, anti-immigrant attitudes are affected by the residential context and this is explained by the contradicting ethnic threat theory and intergroup contact theory.
3.2 Ethnic threat theory
Firstly, the ethnic threat theory is applied in this study to understand the relationship between the residential context and anti-immigrant attitudes. The concept “perceived ethnic threat” concerns about different aspects of life and different assumptions of group status (Bohman, 2014). Ethnic threat, explained as an output of competition over scarce resources, could be split into two components: an economic component and a cultural component. This is because the competition over these scarce resources can be about tangible and intangible scarce resources (Manevska & Achterberg, 2011). The realistic group threat theory explains the tangible competition over scarce resources (perceived economic threat) and the intangible competition over scarce resource (perceived cultural threat) is explained by a synthesis of two theories: the realistic group threat theory and the social identity theory. The realistic group threat theory explains how people develop anti-immigrant attitudes, while the social identity theory explains the situational differences between individuals (Schneider, 2008).
Both components pose a perceived threat to the position of the in-group population. These components of ethnic threat influence the mechanisms of identification with the in-group and contra-identification with the out-group, which means group members evaluate their in-group members more positively and out-group members more negatively (Sniderman et al., 2004). This could in turn induce negative feelings towards the out-group population as the in-group feels threatened by the out-group (Hjerm, 2009). It is important to mention that it is about the perceived ethnic threat and not about the actual ethnic threat people are facing,
15 while their perceptions may differ from reality (Rustenbach, 2010; Esses et al., 1998).
3.2.1 Economic threat
Most of the studies in the literature focused on the economic component of the ethnic threat theory in explaining anti-immigrant attitudes (see for example Burns & Gimpel, 2000). Economic threat is an interest-based component which states that hostile attitudes towards ethnic minority groups, such as racism and xenophobia, are the result of competition of different ethnic groups over tangible scarce resources (Jackson, 1993; Esses et al., 1998; LeVine & Campbell, 1972). Groups can compete about different scarce economic resources that are related to power, money, jobs, housing, social services and education for people’s children (McLaren & Johnson, 2007).
The ethnic threat theory states that the competition over scarce resources could be related to the individual and the in-group as a whole. Firstly, individual economic threat is referring to competition over scarce resources, which could lead to personal loss, such as the loss of income and the loss of a job (Strabac & Listhaug, 2008). This could develop anti-immigrant attitudes to retain people’s own position and to have grip on the good life as Gibson (2002) argues. The driving force behind this is a cost-benefit analysis people make based on their perceived individual threats, which shapes their immigrant attitudes (Espenshade & Hempstead, 1996).
Secondly, perceived economic threat could also refer to scarce economic resources that are taken away from their in-group, but do not necessarily directly affect the person itself (Funk, 2000). Previous studies state negative attitudes towards immigrants are best explained by economic threats influencing the in-group as a whole (see for example McLaren & Johnson, 2007; Funk, 2000). In their research, McLaren & Johnson (2007) state that anti-immigrant attitudes is not
16 strongly related to self-interest, but that perceived economic threats to the in-group as a whole are the main source of anti-immigrant attitudes. Thus, there does not need to be an immediate risk for the individual, but it is enough if an individual perceives that the in-group is competed (Bohman, 2014). Anti- immigrant attitudes are thus the result of immigrants taking away economic resources that do not belong to their group following members of the in-group, because immigrants are seen as newcomers (Quillian, 1995).
In addition, socio-economic conditions of people are important explaining individual differences in perceived economic threat. The strength of the competition over scarce resources is often based on economic uncertainty and the extent one has to do with the ethnic out-groups (Jaspers et al., 2009). Some people are more prone in the competition over scarce resources, as the ethnic threat theory does not influence all people in the same way (Scheve & Slaughter, 2001). Low-skilled natives are more likely to have negative attitudes towards immigrants as they are in a competition with low-skilled immigrants. In the Netherlands, immigrants are often overrepresented in low-skilled jobs and therefore low-skilled natives have similar socio-economic conditions as most of the immigrants (Manevska & Achterberg, 2011). As a consequence, immigrants are competing over the same scarce resources, like jobs, housing, schools and social services as low-skilled natives (McLaren & Johnson, 2007). Therefore, it is expected that natives with a less socio-economic resources will perceive more socio-economic threat than natives with a more socio-economic resources.
3.2.2 Cultural threat
On the other hand, the presence of immigrants could also affect the perceived cultural threat, which in turn could induce negative attitudes towards immigrants (Campbell, 1965; Manevska & Achterberg, 2011). Non-economic determinants were often ignored in literature explaining anti-immigrant attitudes, but the cultural component of ethnic threat gained more attention in recent literature
17 changing the idea that anti-immigrant attitudes are not only economically related, but culturally related as well (Jaspers et al., 2009; Taylor, 1998). Cultural perceived threat of people is referring to concerns over norms and values, but as well identity (Iversflaten, 2005). Ethnic minority groups in European countries often have a diverse ethnicity, linguistics, culture, religion and race, which could diminish the homogeneous identity and solidarity of natives in the host society (Himmelroos & Leino, 2015; Card et al., 2005). Thus, it could affect the social cohesion in a country and that is why non-Western immigrants are seen as more problematic than immigrants from a more similar culture in Western European countries (Lubbers et al., 2002; Rydgren & Ruth, 2013; Hagendoorn, 1995).
To fully understand the mechanism of the cultural component of the ethnic threat theory it is necessary to explain the social identity theory of Tajfel (1979). The social identity theory of Tajfel (1979) helps to understand the psychological mechanisms of developing, and maintaining a group identity. Social identity could be defined as “a person’s sense of who they are based on their group membership(s)” (McLeod, 2008). The theory states that people are creating a perceived distinction between their own group and other groups (Coenders, 2001). In this case, Dutch native see members of their own group as more heterogeneous as immigrants (members of the out-group). This process is called social categorization (Tajfel & Turner, 1979). Members of the in-group apply favourable characteristics to their group to create group distinctiveness and to enhance their self-image (Tajfel & Turner, 1979). In addition, the in-group tries to limit the opportunities of other groups to achieve and maintain positive distinctiveness (Esses et al., 2001; Chandler & Tsai, 2001). Thus, following the social identity theory of Tajfel (1979), stereotyping is based on exaggerated differences between groups and similarities within the out-group (McLeod, 2008). The social identity theory shows that perceived ethnic threat is not only a result of competition about tangible, economic resources, but also a consequence of competition about group identities (McLeod, 2008).
18 On the other hand, literature also states that the social identity theory is incomplete, because it lacks to clarify social categories in societies (Coenders, 2001). The cultural capital explanation of Manevska and Achterberg (2011) does recognize social categories in the society. Socio-economic conditions of people play a dominant role in explaining anti-immigrant attitudes (see for example Semyonov et al., 2004). The cultural capital explanation also states that the level of education possesses the cultural capital of people, because people with a higher level of education are more able to observe cultural expressions of people and to understand their meanings (Houtman, 2001). In addition, people with a higher level of education value more cultural diversity and are more willing to accept differences (Houtman, 2001).
Thus, in line with the ethnic threat theory, this study argues that a higher relative size of immigrants will induce economic and cultural threat to the majority population due to increasing competition over tangible and intangible scarce resources. This could in turn lead to higher anti-immigrant attitudes. This has led to the following hypotheses of this study:
Hypothesis 1: “In a local district, a higher relative size of immigrants will induce perceived economic threat to Dutch natives, which in turn will lead to higher anti-immigrant attitudes”
Hypothesis 2: “In a local district, a higher relative size of immigrants will induce perceived cultural threat to Dutch natives, which in turn will lead to higher anti-immigrant attitudes”
3.3 Intergroup contact theory
The intergroup contact theory is applied in this study to understand the relationship between the relative size of immigrants in a local district and
anti-19 immigrant attitudes. Intergroup contact could be defined as “the face-to-face interaction between members of clearly defined groups” (Pettigrew & Tropp, 2006, p. 754). Hamberger and Hewstone (1997) distinguished three dimensions of intergroup contact. Firstly, there are intimate contacts that are the consequence of friendship and kindship. Secondly, occasional contacts are the result of superficial and sporadic relationships. These first two dimensions refer to casual-intimate or quality of intergroup contact (Allport, 1954). This includes closeness and duration of intergroup contact. Study of McLaren argued that friendship is the most important determinant of the quality of intergroup contact (McLaren, 2003). Lastly, Hamberger and Hewstone (1997) argue that there are neighbourhood contacts, which is providing the opportunity to have shallow contacts with immigrants. It is thus not referring to causal-intimate contacts, but to the frequency to have intergroup contact. In other words, all these dimensions intergroup contact determine the strength of intergroup contact
.
However, the residential context people live in not shaped randomly, because it depends on socio-economic factors, like housing, labour market, income and living standards of people (Semyonov & Glikman, 2009). Therefore, intergroup contact is also not completely random, because ethnic minority groups mostly do not share the same residential area with natives, because they have less economic resources (Semyonov & Glikman, 2009). This shapes the living environment of people and is influencing the possibility to have intergroup contact. The segregation of groups in residential areas could lead to a blockage of the possibility to have intergroup contact, which in turn stimulates mutual stereotypes, hostility and anti-immigrant attitudes (Allport, 1954). This has been shown by the study of Wagner et al. (2006), which revealed that an increasing relative size of immigrants is increasing the likeliness that people from different ethnic groups would establish positive and constructive contacts, but spatial segregation across ethnic neighbourhoods and local districts decreases the chances of intergroup contact (Wagner et al., 2006; Semyonov & Glickman, 2009).
20 Moreover, a variant of the intergroup contact theory is the halo effect, which is an important effect to take into account when focusing on the level of the local district. The halo effect states xenophobic and racist people are not located in local districts with a high relative size of immigrants, but in local districts close to local districts with a high relative size of immigrants. The surrounding districts create a halo around the district with a relatively high size of immigrants. The research of Miles (1989) argues that we see this effect when focussing on smaller geographical units due to a difference between ‘meeting the imagined other’, and ‘experiencing the imagined other’. The experienced other is referring to local districts with higher sizes of immigrants having more possibilities to have immediate contact with immigrants, which reduces anti-immigrant attitudes according to the intergroup contact theory. On the other hand “meeting the imagined other” refers to surrounding local districts, who are likely to meet the imagined other, the immigrant, in the surrounding local district on a daily basis, but have less possibilities to have immediate social contact with immigrants. Thus, the two small communities are close enough to perceive the differences, but are too far away to recognise similarities (Hannah, 2012). In other words, it is likely surrounding local districts, close to a local district with a high relative size of immigrants, perceive the same ethnic threats due to the presence of the immigrants, but have fewer possibilities to have immediate contacts to reduce these ethnic threats. The intergroup contact theory is bounded to the residential context, because it is based on personal connections that are established through immediate contact and interaction (Rojon, 2013). The people who live in a local district with a higher relative size of immigrants could therefore be relatively more accepting towards immigrants, seeing them as friends and not as a threat or enemy (Martin, 2000).
Thus, the intergroup contact theory is another mechanism that is linking the relative out-group size to anti-immigrant attitudes. In local districts with a higher relative size of immigrants people will have more possibility to have intergroup contacts, which in turn would decrease ethnic threats people perceive. In this study
21 both quality and frequency will be taken into account as indicators of perceived ethnic threat according to Hamberger and Hewstone (1997). This has led to the following hypotheses:
Hypothesis 3: “In a local district, a higher relative size of immigrants will induce possibilities to have immigrant friends, which in turn will decrease the perceived economic and cultural threat to Dutch natives”
Hypothesis 4: “In a local district, a higher relative size of immigrants will induce the possibilities to have shallow contact with immigrants, which in turn will decrease economic and cultural threat to Dutch natives”
3.4 Conceptual Model
This study is testing the relationship between the relative size of immigrants in a local district and anti-immigrant attitudes of Dutch natives. The relationship is explained in the literature by two contradicting theories: the ethnic threat theory and the intergroup contact theory. This study expects that when local districts are surrounded by local district with a high relative size of immigrants, intergroup contact will play a decisive in explaining variance in anti-immigrant attitudes of Dutch natives. This has led to the following expected relationships of this study that are presented the conceptual model (Figure 1).
Figure 1: Conceptual model
+
++
+_
+ -The relative size of immigrants
Perceived ethnic threat
Economic threat Cultural threat
Anti-immigrant attitudes
Intergroup contact
Contact with immigrant friends Shallow contact with immigrants
22
4. Methods
4.1 Data
In order to fulfil the objectives of this research, the study needed a dataset which provides information on individuals’ attitudes towards immigrants, socio-economic conditions and socio-demographic characteristics. The goal of this study was to test how the relative size of immigrants in a local district is influencing the level of anti-immigrant attitudes of Dutch natives within a local district and therefore the unit of observation is the individual.
For testing the conceptual model this study used structured surveys. These surveys were personally distributed in three different local districts within the municipality of Utrecht. The study has a cross-sectional design, which means that the data of the analysis are gathered at one specific point of time. The surveys were handed out in May 2016. The questions and scales used in the survey are provided in the Appendix of this study.
This study also used existing observational data for some of the variables in the analysis. Data for the variable “the relative size of immigrants” (total non-Western immigrants divided by total population) were gathered using the statistics of the Dutch Centre of Statistics (CBS). This study focused on local districts instead of neighbourhoods, because the neighbourhood may be the most direct geographical environment where people spend most of their time, friends may very well live in other neighbourhoods (Tolsma et al., 2009). Meeting places as sport clubs, schools and churches are likely to be regulated at a bigger geographical unit, the level of the local district. Therefore, the local district is the most likely area in which people have their social networks. This study used the latest data of CBS that was available, which derives from 2014. For taking into account the halo effect, the surrounding local district with the highest relative size of immigrants was used as a
23 measure for every local district studied. Also this data was gathered by using data of the CBS (which also derives from 2014).
The statistical calculations were gathered with help of software program SPSS Statistics. The hypotheses were tested using a multiple regression analysis, which makes it possible to measure the influence of independent variables on a dependent variable (Babbie, 2004).
4.2 Research design
This study selected three local districts within the municipality of Utrecht that were categorized in three groups, according to the example of Rydgren and Ruth (2012), ranging from a local district with a low relative size of immigrants (0-10%), a local district with a medium relative size of immigrants (10-20%) to a local district with a high relative size of immigrants (20% or more).
As elaborated in the theoretical framework, there are good arguments to assume that the halo effect plays an important role when studying anti-immigrant attitudes on the level of the local district. This study controls for the halo effect by selecting local districts that have the same surrounding local district with the highest relative size of immigrants. In this way, the study assumes the halo effect will have the same influence on all local districts.
Based on these criteria, the study focused on three local districts within the municipality of Utrecht who met the criteria: “Utrecht Oost”, “Utrecht Zuid”, and “Utrecht Leidsche Rijn”. CBS (2016) showed that “Utrecht Zuidwest” was the surrounding district with the highest relative size of immigrants for each local district. Therefore, the surrounding local district with highest relative size of immigrants was the same for all local districts (see Table 1).
24 Table 1: Measured local districts
Variables Local districts
1* 2** 3*** Relative size of immigrants 8% 20% 27%
Surrounding local district with
highest relative size of immigrants 39% 39% 39% (Utrecht Zuidwest)
4.3 Procedure
The data were collected using random sampling in the three local districts within the municipality of Utrecht. The structured survey included questions measuring the variables of the conceptual model and individual characteristics of people that were added as control variables. Most of the items derived from existing questions, used in the European Social Survey (E.S.S., 2015). This study used questions from the survey of E.S.S. (2015), because this survey measured the same variables and control variables as used in this study. On the other hand, some additional questions were used, derived from other studies to create more accurate measurements that correspond with the operationalisations of the variables in this study. This will be further explained in the section “measurement instruments”.
The surveys were personally distributed by the author with informed consent of the participants of this study. This was done in public space at randomly picked times between 07:00 and 22:00 o’clock at random locations by picking random locations on the city map within the local districts. An instruction letter was provided about the purpose and confidentiality of the study. Also anonymity was guaranteed; respondents had to leave their completed surveys in a closed envelope and they were also not allowed to write their names on the surveys.
*1 = Utrecht Oost **2 = Utrecht Zuid
25
4.4 Population and sample
The population of interest was the majority population in the Netherlands, Dutch native people, living in the local districts studied within the municipality of Utrecht. Therefore only Dutch native people were included in this research and other respondents were excluded from the analysis. The sample consisted of three local districts, which had together a population including more than 30.000 individuals.
Important for the selection of the sample is the definition of “in-group”. Therefore, people are part of the in-group when they have the Dutch nationality and when they and their parents are born in The Netherlands. This was checked in the survey using the following items: “Do you have the Dutch nationality?”, “Were you born in The Netherlands?” and “Were both of your parents born in The Netherlands?” Answer possibilities were “yes” and “no”.
This study also excluded people who are born in the Netherlands and have the Dutch nationality, but whose parents are not born in the Netherlands. The reason for this is that these people are likely to have a stronger connection to immigrants as other natives do and this could influence their attitudes towards immigrants. Also people that were not part of the labour force, meaning everyone outside the age of 18 and 67 were not taken into account in the analysis. This is, because these people are not likely to experience economic threats of immigrants (competition over scarce resources) and could therefore have a lower level of anti-immigrant attitudes following the ethnic conflict theory. Thus, all the respondents that missed the selection criteria mentioned above were excluded from the analysis.
In total, 205 respondents participated in this research who met the selection criteria of this study. Of these respondents, 64 were living in Utrecht Oost, 71 in Utrecht Zuid, and 70 in Utrecht Leidsche Rijn. The average age of the respondents was 48 years old in the local district Utrecht Oost, 46 years old in Utrecht Zuid and
26 44 years old in Utrecht Leidsche Rijn. The youngest respondent of this study was 18 years old and the oldest participant 67 years old. Table 2 shows the full descriptive statistics of the main control variables gender, age, education and income by local district.
Table 2: Descriptive statistics
Variables N Min. Max. Mean Std.
dev. Utrecht Oost Gender* 64 1 2 1.58 .498 Age 64 19 67 48.47 14.582 Education ** 63 1 7 4.33 1.656 Income*** 62 1 5 2.98 1.373 Utrecht Zuid Gender 71 1 2 1.52 .503 Age 71 18 67 46.18 16.558 Education 71 1 7 4.76 1.617 Income 71 1 5 2.44 1.458
Utrecht Leidsche Rijn
Gender 70 1 2 1.47 .503
Age 70 18 67 43.73 15.435
Education 69 1 7 4.83 1.740
Income 68 1 5 2.65 1.423
4.5 Measurement instruments
The scales and items that were included are mentioned in the section below, where this study elaborates the operationalization of the independent and dependent variables. The factor loadings of the variables are also added in the appendices of this research. A factor analysis and reliability analysis were executed to test the respectively validity and reliability of the scales and questions used in the sample of this study. Components that were included had the following selection criteria:
*age: 1 is man, 2 is woman
**education was measured on a 7-point scale. Higher scores indicate higher levels of education
*** income was measured on a 5-point scale. Higher scores indicate higher levels of income
27 Eigenvalue > 1, and KMO value > .5 (Kaiser, 1974). The scale reliability was evaluated by using Cronbach’s α.
4.5.1 The dependent variable
Anti-immigrant attitudes: The dependent variable of this study is ‘anti-immigrant attitudes’ of Dutch natives. As mentioned before, anti-‘anti-immigrant attitudes refer to xenophobic and racist attitudes. This study used two items, which should capture the variable (E.S.S., 2015). The respondents were asked to answer the following question: “To what extent should the Netherlands allow people of a different race or ethnic group as most people in the Netherlands belong to, to come and live here?” Secondly, the respondents were asked: “To what extent should the Netherlands allow people from poorer countries outside Europe to come and live here?” The variable was measured on a four-point scale (ranging from “allow all”,” allow some”, “allow few” to “allow none”). The scores of the respondents on these questions were aggregated. As a consequence, the four-point scale changed into a seven-point scale. A higher the score on this variable means a higher level of anti-immigrant attitudes of a respondent. Also the answer possibility “do not know” was added. A factor analysis was performed to control for construct validity of the scale, but no separate factors within the scale were found. The scale reliability was good (α = .730).
4.5.2 The independent variables
The relative size of immigrants: The independent variable “the relative size of immigrants” was measured using the total non-Western immigrants divided by the total population in a local district. CBS (2014) refers to non-Western immigrants as immigrants with an ethnic background in Africa, Latin America and Asia (excluding Indonesia and Japan) or Turkey.
Intergroup contact: As mentioned before, quality and frequency of contact are two important indicators of intergroup contact. Therefore, the study used two
28 items to capture this concept, distinguishing close contact from daily contact with immigrants (E.S.S., 2015). Contact with colleagues, often used as a determinant of intergroup contact was not taken into account, because the study expected that it was not strongly related to the residential context of Dutch natives, because many people work outside their local district. CBS (2009) shows that 29 percent of the people who have a paid job work outside their region. In addition, Dixon and Rosenbaum (2004) argued that contact at the workplace is not a good predictor of anti-immigrant attitudes.
The first question asked to the respondents was referring to contact with immigrant friends: “Do you have any close friends who are of another race or ethnic group than most Dutch people?” (Three-points answer scale ranging from “none”, “a few” to “several”). The second question was referring to ever-day contact with immigrants or so-called shallow contact with immigrants: “How often do you outdoors contact with people of a different skin colour or ethnic group than most Dutch? For example: in public transport, in the street, in the store or in the local district?” (Seven-point answer scale ranging from “never”, “less than once a month”, “once a month”, “several times a month”, “once a week”, “several times a week” to “every day”. A higher the score on a question means more intergroup contact a respondent has.
Perceived economic threat: Several questions were used in this study to cover the variable “perceived economic threat”. Questions referred to issues related to the influence of immigrants on the economy in general and jobs, housing, welfare benefits and education. All these questions refer to competition over economic scarce resources that affect the society as a whole, but not necessarily the individual. This is, because perceived economic threat is not based on self-interest and therefore not a good determinant for explaining anti-immigrant attitudes (McLaren & Johnson, 2007). It also shows that perceived economic threat is a broad concept, but by using more items this study should be sufficient in measuring
29 this variable. Respondents had to respond to five statements using a five-point Likert scale ranging from “totally disagree”, “disagree”, “do not agree/do not disagree”, “agree” to “totally agree”. One question was mirrored (indicated contra-indicative) to avoid influence on the answer tendency of the respondent. This statement was in the end re-coded in the dataset, which means that for all questions a higher score means a higher level of perceived economic threat. No separate factors were found in the factor analysis, and the reliability of the items was sufficient (α = .711). The questions that were used to measure perceived economic threat, derived from E.S.S. (2015) and the research of McLaren & Johnson (2007), were: "Non-Western immigrants are generally good for the Dutch economy?", "Non-Western immigrants who come to live here take jobs away from Dutch workers", "Non-Western immigrants take precedence over the Dutch housing market", "Non-Western immigrants abuse the social welfare system" and "In schools with many non-Western immigrants the quality of education suffers".
Perceived cultural threat: This study measured perceived cultural threat by By asking the respondent to what extent respondents think that immigrants undermine the cultural life in general. Although only one statement was used, it should be sufficient enough to cover the concept of perceived cultural threat. The variable was measured on a five-point Likert scale ranging from “totally disagree”, “disagree”, “do not agree/do not disagree”, “agree” to “totally agree”. The respondents had to respond to the following statement: "Non-Western immigrants are a threat to the culture of the Netherlands".
4.5.3 Control variables
This study also took control variables into account for studying the conceptual model. Many of the control variables used are in line with other studies that also focused on the relationship between the relative immigrant size and anti-immigrant attitudes (see for example Savelkoul et al., 2010).
30 First of all, the study controlled for determinants of individual socio-economic conditions that could influence perceived ethnic threat as the literature argued by using questions about education, income and employment status (Scheepers et al., 2002; Schneider, 2008). Secondly, this study also added other control variables referring to individual characteristics that might influence the model of this research.
Education: Other studies showed this control variable has a negative influence on anti-immigrant attitudes (see for example Hjerm, 2009; Coenders & Scheepers, 2003). Education was measured using a single question. A higher score means a higher level of education of the respondent. The following question was used in this research: “What is the highest level of education you successfully completed?”
Income: The variable income was measured by asking the respondents the following question: “What is the total monthly disposable / net household income of your household?” This question had five answer categories ranging from “< 1500 euro”, “1500-2000 euro”, “2000-3000 euro”, “3000-4000 euro” to “4000 euro or more” per month. Also, an extra category “do not know” was added.
Employment status: This control variable was measured using two questions. Firstly, the respondents were asked: “Did you do any paid work in the last 7 days for one hour or longer? Secondly, respondents were asked: “How many hours a week do you have paid work?” The answers of the respondents were split into three categories: unemployed (0-11 hours a week), part-time (12-35 hours a week) and full-time (36+ hours a week).
Occupational status: This control variable was measured using one question: “Which of the descriptions applies the best to what you did in the last 7 days?” This question had seven answer possibilities “student”, “unemployed”, “retired”,
31 “volunteering’, “paid job”, “housewife/houseman” and “other”. Also the answer possibility “do not know” was added. The variable was coded into a dummy variable.
Marital status: This control variable was added to the model of this study by asking respondents to their marital status. The variable was coded into a dummy variable. Four answer categories were used ranging from “not married and never have been married”, “married”, “divorced” to “widow(er)”.
Religiosity: This study controlled for the variable religiosity, referring to the frequency people participate in religious services (Scheepers et al., 2002). The following question was asked to the respondents: “How often do you use religious services?” There were four answer categories ranging from “never /almost never”, “once a month”, “a few times a month”, to “one or more times a week”.
Gender: Gender is a frequently used control variable that will also be included in this study. The following question was used: “What is your gender?” The answer possibilities were “woman” or “men”.
Age: This variable is, like gender, also a frequently used control variable that will be included in this study is another frequently used control variable that will also be included in the model. This control variable was measured on a ratio scale. Respondents were asked in the structured survey: “What is your age?”
5. Results
This section shows the tests that were conducted for measuring the relationship between the relative size of immigrants in a district and the attitudes towards immigrants of Dutch natives. Firstly, the descriptive characteristics will be presented. Secondly, the results of the multiple regression analysis and the
32 study, but as well the control variables .The section ends with a discussion of the analysis.
5.1 Descriptive characteristics
The dataset that was used in this study checked for missing values and outliers. As most of the variables are measured ordinal, the lowest scores, highest scores, means, standard deviation and the amount of respondents of the key variables relative size of immigrants, contact with immigrant friends, shallow contact with immigrants, economic threat, cultural threat and anti-immigrant attitudes but as well the control variables are presented in Table 3.
33 Table 3: Descriptive statistics
Variables N Min. Max. Mean Std.
dev. The relative size of
immigrants 205 1 3 2.03 .810 Perceived economic threat 204 1 5 3.08 .601 Perceived cultural . Threat 205 1 5 2.64 .947 Contact with immigrant friends 2051 1 4 2.07 1.100 Shallow contact 205 1 7 5.72 1.349 Control variables: Education 204 1 7 4.65 1.677 Income 203 1 5 2.65 1.423 Employment status 196 1 3 2.20 .837 Gender 205 1 2 1.52 .501 Age 205 18 67 46.06 15.621 Religiosity 205 1 7 1.31 .902 Married 205 0 1 .61 .488 Not married 205 0 1 .29 .458 Divorced 205 0 1 .07` .253 Widowed 205 0 1 .02 .139 Student 205 0 1 .09 .284 Unemployed 205 0 1 .05 .216 Retired 205 0 1 .08 .269 Volunteer 205 0 1 .00 .070 Paid job 205 0 1 .48 .501 Housewife/Houseman 205 0 1 .13 .339
For the mediating variables “perceived economic threat” and “perceived cultural threat” a t-test was conducted to test whether the averages of perceived economic and cultural threat are statistically different. This was done as it could be argued both variables are theoretically closely linked with each other. The t-test was significant for these variables (t = 6.664, p < .01). The mean of economic threat is 3.082 and the mean of cultural threat is 2.64 indicating that people perceive immigrants more as an economic threat than as a cultural threat. On the other hand
34 the spread of perceived cultural threat was higher (SD = .947) than of perceived economic threat (SD = .601).
This study also presents a cross-tabulation between the relative size of immigrants in a local district and anti-immigrant attitudes to show briefly the association between these two variables. Table 4 shows how the scores of anti-immigrant attitudes (7-point scale) are divided over the three local districts differing from a low relative size of immigrants to a high relative size of immigrants (3-point scale). In addition, this study conducted a chi-square analysis to show whether the variables are statistically associated. The score of the chi-square shows that the relative size of immigrants in a local district is significantly associated with anti-immigrant attitudes of Dutch natives (x² = 19.909, p < .05). 6.664
Table 4: Cross-table
Variables Relative size of immigrants
1 2 3 Total Anti-immigrant attitudes 1 7 5 4 16 2 6 4 6 18 3 12 28 27 67 4 24 13 10 47 5 9 13 15 37 6 3 5 2 10 7 2 3 4 9 Total 63 71 70 204 Chi-square 19.909** Note: ** = p < .01, * = p < .05
5.2 Multiple regression analysis
In order to answer the research question of this study, a multiple regression analysis was conducted. The study expected that, based on the ethnic threat theory
35 and the intergroup contact theory, the relationship between the relative immigrant size of immigrants and anti-immigrant attitudes is mediated by perceived ethnic threat and intergroup contact.
The relative size of immigrants and anti-immigrant attitudes
First of all, a multiple regression analysis was conducted leaving out the mediating indicators of ethnic threat, intergroup contact and the control variables to examine the direct relationship between the independent variable “the relative size of immigrants” in a local district and “anti-immigrant attitudes” of Dutch natives. This is shown in Table 5. Model 1 does not show a significant relationship (F = .121, p > .05), which means the model is a not useful to predict anti-immigrant attitudes. The analysis shows the direct relationship between the relative size of immigrants and anti-immigrant attitudes is not significant (β = .025, p > .05).
In model 2 the control variables of this study were added to test the direct relationship between the independent variable “the relative size of immigrants” and the dependent variable “anti-immigrant attitudes”. The model showed a significant relationship (F = 2.600), p < .01), but the relationship between the relative size of immigrants and anti-immigrant attitudes did not show a significant relationship (β = .037, p > .05). The multiple regression analysis did show a significant relationship for the control variables education (β = -.270, p < .01), occupational status: student (β = -.251, p < .05) and occupational status: unemployed (β = -.169, p < .05). These variables are the main predictors of anti-immigrant attitudes in this model. This means, the higher the level of education of the respondents, the lower their level of anti-immigrant attitudes. Also, students and unemployed people have a lower level of anti-immigrant attitudes in comparison with other occupational states of respondents.
36 Table 5: Multiple regression analysis
Variables Model 1 Model 2
S.E. B S.E. B
The relative size of immigrants .064 .025 .066 .037
Perceived economic threat
Perceived cultural threat
Contact with immigrant friends Shallow contact Anti-immigrant attitudes Education .036 -.270** Income .044 -.034 Employment status .070 -.066 Gender .104 .088 Age .006 -.090 Religiosity .056 .042
Marital status – Not Married (ref.)
Marital status – Married .167 -.200
Marital status – Divorced .244 .000
Marital status – Widowed .398 -.101
Occ.status – Student .267 -.251*
Occ.status – Unemployed .260 -.169*
Occ.status – Retired .240 .067
Occ.status – Volunteer .718 -.037
Occ.status – Paid job .160 -.151
Occ.status – Housewife/Houseman .185 -.149 Constant 2.319** 3.108** R-Square .000 .188 Respondents 193 193 F-score .121 2.600 Significance F-score .729 .001 Note: ** = p < .01, * = p < .05 Occ. Status = occupational status
37 Mediating effect of perceived economic threat
The conceptual model of this study assumes that the independent variable “the relative size of immigrants” has an indirect influence on the dependent variable “anti-immigrant attitudes” via the mediating variables “perceived economic threat” and “perceived cultural threat”. The results of the multiple regression analysis are presented in Table 6. Firstly, Model 3 tested the direct influence of the independent variable “the relative size of immigrants” on the mediating variable “perceived economic threat”. This model showed a significant relationship (F = 2.600, p < .01), which means the model is useful to predict anti-immigrant attitudes. Based on this analysis, the study concludes that the relative size of immigrants is not a strong predictor of perceived economic threat (β = .117, p > .05). The control variables did not show a significant relationship.
Model 4 shows the interaction effect of the independent variable “the relative size of immigrants” and the mediating variable “perceived economic threat” on the dependent variable “anti-immigrant attitudes”. This model is significant (F = 3.663, p < .01), which means that this model is useful for explaining anti-immigrant attitudes. The multiple regression analysis did not show a significant find support for the interaction effect (β = .034, p > .05). This means that the independent variable “the relative size of immigrants in a local district” is also not indirect related to “anti-immigrant attitudes” via the mediating variable “perceived economic threat”. The control variable education was significantly related (β = -.242, p < .01).
Mediating effect of perceived cultural threat
In Model 5 this study tested the influence of the independent variable “the relative size of immigrants” on the mediating variable “perceived cultural threat”. This model is significant (F = 2.710, p < .01). .Based on this test, this study concludes that “the relative size of immigrants in a local district” has no significant influence on “perceived cultural threat” (β = .016, p > .05). The control variables
38 education (β = -.322, p < .01) and religiosity (β = .163, p < .05) are influencing this model.
Secondly, in model 6 this study tested the interaction effect of the independent variable “the relative size of immigrants” and the mediating variable “perceived cultural threat” on the dependent variable “anti-immigrant attitudes”. The regression model that was significant (F = 4.320, p < .01). The multiple regression analysis did not show support for the interaction effect (β = -.045, p > .05). This means the relationship between the relative size of immigrants and anti-immigrant attitudes is not mediated by perceived cultural threat. For the control variable occupational status: student (β = -.178, p < .10) a significant relationship was found.
