deltie.net archive

Extracted abstract

When is migrants' socio-economic attainment associated with enhanced national belonging to their residence country? Drawing on a large-scale survey, we compare migrants from the same 10 origin countries across Denmark, Germany, and the Netherlands. We exploit this double comparison across origin groups and residence countries to contextualize mixed findings of positive, negative, and null associations between migrants' socio-economic attainment and national belonging in earlier research. We introduce the notion of ethnic boundaries to situate when migrants' socio-economic attainment and belonging are positively or negatively associated. We examine how migrants' socio-economic attainment interacts with contextual variation in ethnic boundaries and with individual-level variation in perceptions of ethnic boundaries based on perceived group discrimination. Multi-group structural equation models show that migrants' socio-economic attainment is often decoupled from national belonging. However, they also reveal crucial contextual variation as same-origin migrants can succeed socioeconomically with or without feeling belonging in different residence countries. At the individual level, perceived group discrimination conditions this association, so that socio-economic attainment translates into more national belonging only when perceptions of discrimination are low. Our comparative findings of contingent national belonging thereby challenge existing assumptions that associations between socio-economic attainment and national belonging are linear and that ethnic boundaries in European migration contexts are rigid.

Source: MIFARE 2015, control variables are included but not presented. ***P < 0.001, **P < 0.01, *P < 0.05, ∼P < 0.1.

Introduction

Today, public debates over migrants' integration in Western Europe highlight issues of migrants' (lack of) national belonging and the threat this poses to social cohesion and liberal democracy (Verkuyten, 2016) . Much research has focused on national belonging gaps where migrants on average experience less belonging to their residence country than the native population (Fleischmann and Phalet, 2018) . Looking across origin groups and residence countries, however, there are large unexplained differences in national belonging among migrant populations too. Figure 1 shows average levels of national belonging for 10 origin groups in Denmark, Germany, and the Netherlands. Drawing on these comparative data, our study heeds the call to study this contextual variation in national belonging (De Vroome, Verkuyten and Martinovic, 2014) . To explain such gaps in national belonging, public debates are often quick to point to migrants' 'failed integration' in the residence country (Sniderman, Hagendoorn and Hagendoorn, 2007) . This discourse resonates with classic assimilation theories, which have typically advanced migrants' socio-economic attainment as a litmus test of successful integration. Extending classic (Gordon, 1964) and new assimilation theories (Alba and Nee, 1997) to European migration contexts (Alba and Foner, 2015) , socio-economic attainment is proposed as the royal highway to full participation in structures and institutions of the residence country, which arguably increases allegiance to mainstream norms and values, stimulate positive contacts with the native population, and entail the feeling of being valued members of society (Hagendoorn, Veenman and Vollebergh, 2003) , contributing to migrants' sense of belonging to the residence country.

However, in spite of much evidence that migrants report more national belonging at higher socioeconomic attainment levels (De Vroome et al., 2011, De Vroome, Verkuyten and Martinovic 2014) , there are also repeated findings of no significant associations, which we will refer to as dissociations (Tolsma, Lubbers and Gijsberts, 2012; Alba and Foner, 2015) . Some studies even report negative associations, when socio-economically more successful migrants experience less belonging to the residence country than migrants with lower socio-economic attainments (Ten Teije, Coenders and Verkuyten, 2013; Verkuyten, 2016; Geurts, Lubbers and Spierings, 2020) . This latter finding has been dubbed 'the integration paradox' and directly challenges classic straight line and new assimilation theories (Schaeffer and Kas, 2023) . Previous research argues that such an integration paradox can be understood via mechanisms of exposure to discrimination (in society and/or personally) and rising expectations that are not met in reality, which result in disidentifying with the residence country (Verkuyten, 2016; Geurts, Lubbers and Spierings, 2020) .

This study aims to situate these mixed findings of positive, negative, or null associations between migrants' socio-economic attainment and national belonging through introducing (perceptions of) ethnic boundaries as moderating condition of the attainment-belonging linkage. Specifically, we have two research aims. First, we study how migrants' socio-economic attainment and national belonging are associated across intergroup contexts, meaning migrants' origin group, referring to migrants' country of origin, and residence country. We hereby propose that contextual variation in ethnic boundaries may account for variable associations across specific origin groups and residence countries. Ethnic boundaries are social constructions that define group membership and imply differential worth and access to resources and opportunities in unequal intergroup relations (Lamont and Molnár, 2002) . We define ethnic boundaries broadly in terms of ethnoreligious markers of group membership and status, which jointly (re)produce so-called 'ethnic hierarchies' of differential valuation and treatment of specific origin groups in specific residence countries (Wimmer, 2008) .

Ethnic boundaries are politically contested and open to social change. As specific ethnic markers (such as parentage, language, or religiosity) are made salient and consequential in particular intergroup contexts, ethnic boundaries vary across origin groups and residence countries. Thus, intergroup contexts define ethnic boundaries in ways that exclude specific groups from the national identity of a specific residence country (Alba, 2005) . Alba (2005) distinguishes between bright boundaries, such as boundaries excluding Muslims in the absence of religious accommodation, which are typically hard to cross, from so-called blurred boundaries, which are for instance manifested by more inclusive integration policies that are more permeable and as such may prevent or unmake certain distinctions between migrants and natives (Phalet et al., 2013) .

SUCCEEDING WITHOUT BELONGING?

We argue that these contextual differences in ethnic boundaries may not only directly affect migrants' levels of national belonging, so that migrants feel less belonging in intergroup contexts with bright boundaries, but that these contextual differences may also moderate the association of migrants' socio-economic attainment with their national belonging. For instance, we explore whether more inclusive circumstances enable a positive attainment-belonging linkage as such circumstances may allow for assimilation theory to play out (Diehl, Liebau and Mühlau, 2021; Schaeffer and Kas, 2023) . We thus expect variation across origin groups and residence countries in the association between migrants' socio-economic attainment and national belonging due to contextual variation in ethnic boundaries. As such, we address the following question: (i) To what extent does the association between migrants' socio-economic attainment and their national belonging vary across origin groups and residence countries? For our purposes, we define socio-economic attainment broadly, including employment status and income in the residence country-as well as educational qualifications obtained in both origin and residence countries.

Second, there is much evidence that migrants report less national belonging when they personally experience discrimination or when they perceive intergroup relations as discriminatory, unfair or impermeable for members of their ethnic group (Verkuyten and Martinovic, 2012; De Vroome, Martinovic and Verkuyten, 2014) . In addition, we know that there are differences in perceived discrimination across migrants' socio-economic attainment, so that migrants tend to perceive more, not less, discrimination at higher attainment levels (Steinmann, 2019; Schaeffer and Kas, 2023) . Previous research has therefore extensively studied a possible mediating role of perceived discrimination in (negative) associations between socio-economic attainment and national belonging (Verkuyten, 2016; Geurts, Lubbers and Spierings, 2020) . Importantly however, this mediating mechanism does not explain mixed findings in previous research. We therefore introduce perceived group discrimination, as indicative of migrants' perceptions of ethnic boundaries, as a hypothetical moderator to explain why for some migrants higher socio-economic attainment translates into more national belonging, and for others, it does not. We propose parallel mechanisms at the level of individual perceptions and at the contextual level to argue how individual differences in perceived discrimination and country differences in actual ethnic boundaries may affect the association between attainment and belonging. In short, we argue that more inclusive circumstances enable a positive association between migrants' socio-economic attainment and national belonging, whereas no or negative associations may result from more exclusive circumstances that trigger mechanisms as theorized by the integration paradox literature. In exploring how perceived discrimination moderates the association-belonging linkage, we answer the following question: (ii) To what extent is the association between migrants' socio-economic attainment and their national belonging contingent on migrants' perceived group discrimination?

We address our research questions using a double comparison across the same 10 origin groups in Denmark, Germany, and the Netherlands. These Western European liberal democracies are similar in many ways, yet differ in their institutional arrangements and current policies of migrant integration. To enable simultaneous comparisons across origin groups and residence countries, we draw on 'Migrants' Welfare State Attitudes' (MIFARE) data including first-generation migrants from Eastern and Western Europe, the United States, and Asia (Bekhuis et al., 2018) , which were collected at the end of 2015 and beginning of 2016. By building on a cross-national survey specifically targeted at immigrants and sampling the same-origin groups across countries, we overcome problems of diverging compositions of the migrant population and underrepresentation of (specific) migrants in earlier research when the questionnaire is only offered in the language of the residence country (Ersanilli and Koopmans, 2011) . The MIFARE data moreover include, next to more well-known origin groups, smaller and less well-researched groups such as Japanese, Romanian, and Spanish migrants. We thus exploit the unique comparative design of the MIFARE data to examine whether and how (perceived) ethnic boundaries condition the attainment-belonging linkage across the same-origin groups in three different countries (Tuppat and Gerhards, 2020) .

Comparing countries as socio-political contexts

Differences in ethnic boundaries in Denmark, Germany, and the Netherlands While similar in many ways, Denmark, Germany, and the Netherlands differ in their approaches to accommodate migrants' integration, which have varied in the past and still varied at the time of data collection (2015). In line with previous research, we argue that more inclusive versus restrictive integration policies are a formal manifestation of ethnic boundaries in terms of access to legal or social rights and services, for example (Simonsen, 2016) . In addition, we will discuss informal manifestations of ethnic boundaries along the lines of native populations' attitudes towards migrants.

Traditionally, the Netherlands is characterized as a multiculturalist country, Germany as an assimilationist country, and Denmark as an even stronger assimilationist country (Ersanilli, 2010) . This changed slightly in the last decades, where a more assimilationist shift with more emphasis on individual responsibility took place in the Netherlands, a trend that continued when the radical right politicized integration policy and the traditional commitment to equal opportunities was abandoned, resulting in a large drop in the Dutch Migrant Integration Policy Index (MIPEX 1 ) score, see Supplementary Table A1 in Supplementary Appendix (Huddleston et al., 2015) . Despite the drop in their overall MIPEX score, signalling more restrictive integration policies, the Netherlands nevertheless remains close to the multicultural side of the spectrum when compared with other Western European countries (Cebolla-Boado and Finotelli, 2015) .

Looking at Denmark, it is often described as the odd one out when comparing it to other Nordic countries that in general have more open and tolerant integration policies. Denmark differs with its more assimilationist policies, with, for example, limited legal equality (Gudrun Jensen, 2008; Cebolla-Boado and Finotelli, 2015) . Since 2001, the People's Party has expressed support to the centre-right minority government in return for more immigration and integration restrictions, resulting in assimilationist policies, which is seemingly reflected in the lower overall MIPEX score in 2007 (44 of 100). Slowly but surely, this score has increased towards 2015, likely due to the centre-left coalition in 2011 introducing more inclusive policies, resulting in a score of 59 in 2014. 2 We see a similar trend when looking at Germany, which shows a steady increase in their MIPEX score towards more inclusive integration policies (from 53 in 2007 to 61 in 2014). Traditionally, German-speaking countries have been defined as segregationist and not favouring the inclusion of immigrants (Joppke, 2007; Ersanilli and Saharso, 2011) . More recently, Germany has taken the lead to facilitate and support the recognition of foreign qualifications and skills with the 2012 Recognition Act, and it is concluded that more pro-active integration initiatives on the local (instead of national) levels make for a more inclusive approach (Cebolla-Boado and Finotelli, 2015) , which fits the upward trend in their MIPEX score.

All in all, Denmark and Germany's integration policies became more inclusive between 2007 and 2014, resulting in a great increase in their general MIPEX score around the time of data collection (59 and 61 of 100, respectively). Around 2015, they score similarly to the Netherlands when looking at this overall score, which dropped to 60 in 2015. However, in specific policy areas, differences between these residence countries remain. When looking at anti-discrimination measurements as an active way to prevent and diminish ethnic boundaries, the Netherlands, Denmark, and Germany score 73, 50, and 58, respectively, in 2014. The Netherlands is known for its relatively strong antidiscrimination laws (Huddleston et al., 2015) , although we see-similar in the case of the general MIPEX score-that its score has dropped between 2007 and 2014. In Denmark and Germany, anti-discrimination policies have become more inclusive towards 2015, yet quite a big gap remains with the Netherlands, which despite a downward trend is most inclusive when it comes to anti-discrimination initiatives and formal inclusion of migrants as such.

In addition to the policy as manifestation of ethnic boundaries, more or less negative public attitudes towards migrants among the native population informally reinforce or attenuate ethnic boundaries too (Simonsen, 2016) . Across Western Europe, public attitudes towards migration have become less inclusive in recent times, with German attitudes being most exclusionist and Dutch attitudes least exclusionist (cf. ESS, 2014 and Special Eurobarometer 469, 2017) . In Germany, 30.4 per cent of the majority population find that immigrants make Germany a worse place to live. In Denmark and the Netherlands, this percentage is 23.3 and 23.7, respectively (based on data from the European Social Survey collected in 2014). This is in line with the Special Eurobarometer 469 data (2017), showing that 76 and 64 per cent of the majority population in the Netherlands and Denmark, respectively, reported feeling totally comfortable with having an immigrant as their neighbour, whereas this was only 36 per cent in Germany.

Combining the formal and informal manifestations of ethnic boundaries across the three included countries, we conclude that the Netherlands is most inclusive both formally and informally around the time of data collection. Germany and Denmark are similar when looking at the inclusiveness of their integration policies, whereas Denmark is substantially more inclusive when looking at public attitudes towards immigrants.

Migration from 10 origin countries

This study includes migrants from a range of 10 origin countries who are likely to have migrated for different reasons and differ in various respects. Although we refrain from extensively discussing each specific origin group in their respective residence country, as hypothesizing about these differences is beyond the scope of this article, general migration flows and patterns can be distinguished. In all three countries, large-scale immigration goes back historically to intense labour recruitment after World War II from origin countries such as Turkey and Spain (Salentin and Schmeets, 2017) . Since the early 1970s, they sharply reduced labour migration SUCCEEDING WITHOUT BELONGING? while enabling family reunification (Hansen, 2003) . More recently, they witnessed continuing immigration and a diversification of migrants' origins and motives (Vertovec, 2007) , including immigrants from EU countries including Romania and Bulgaria (Salentin and Schmeets, 2017) .

At the same time, we realize and stress that although these migration flows to the three residence countries may show similar patterns, there may be great variation within these groups across residence countries. For instance, when looking at migration from Poland and Russia, which in Germany may consist of 'late resettlers' (Spätaussiedler, see Will, 2019) and in that respect likely differ from migrants from Poland in the Netherlands who often migrate with a labour motive (Peperkamp, 2018) . Similarly, we expect variation across origin groups' experiences in the residence countries, for example because first-generation migrants from Turkey may find a large established community of people with a Turkish migration background in Germany and the Netherlands, which would not typically be the case for, for instance, migrants from Japan and the Unites States (see Supplementary Table A2 in the Supplementary Appendix). In addition, we see that some groups have more temporary intentions to stay and/or mainly move for educational purposes (see, for example, in Germany the share of students from China versus the share of students from Great Britain in Supplementary Table A2 ).

These differences among and within groups are likely to affect experiences of exclusion and belonging in the residence country, which is underlined by previous research showing how migrants' intention to stay, migration motives, and (existing) ties to the residence country affect their sense of belonging (Wachter and Fleischmann, 2018; Geurts and Lubbers, 2023) . The question remains to what extent different origin groups in various residence countries perceive and are exposed to different ethnic boundaries, and how such differences in perceptions or actual boundaries that are in place might condition the association between migrants' socio-economic attainment and national belonging. In the next paragraph, we will theorize how the ethnic boundaries present in residence countries may affect the association at hand. In addition, we will empirically explore how (perceptions of) ethnic boundaries affect the attainment-belonging linkage across origin groups and residence countries.

Theoretical expectations

Socio-economic attainment and national belonging

According to the classical definition, assimilation refers to the trend of increasingly similar life chances and life styles between migrants and natives, in which classic assimilation theories propose that incorporation in one life domain will spill over into other domains as well. 3 By implication, participation in the structures and institutions of the residence country (such as the education system or the labour market) is seen as necessary and sufficient for migrants' national belonging (Gordon, 1964) . However, classic or linear conceptions of assimilation have been challenged by segmented assimilation theory (Portes and Zhou, 1993) , revised by new assimilation theory (Alba and Nee, 1997), and further qualified in European countries, which are 'less likely to fold the newcomers into the national whole, and tend to imagine their societies as derived from a core majority population that has occupied the national territory since time immemorial' (Alba and Foner, 2015) . Nevertheless, one basic tenet of assimilation theories has been that upward socio-economic mobility is a powerful driver of migrants' becoming part of a new society. A higher socio-economic attainment arguably increases exposure and allegiance to mainstream norms and values, stimulate positive contacts with the native population, and entail the feeling of being valued members of society (Hagendoorn, Veenman and Vollebergh, 2003) , contributing to migrants' sense of belonging to the residence country.

Pre-immigration socio-economic factors are also found to play a role (Zimmermann, Zimmermann, and Constant, 2007) , so that a higher education obtained in the origin country facilitates national belonging (Fokkema and De Haas, 2015) . Along those lines, highskilled migrants receive a warmer welcome (Turper et al., 2015) , which should facilitate feeling included in society. Reasoning from assimilation theories and research, therefore, we expect that (H1a) migrants with higher socio-economic attainment will experience more national belonging.

Although a positive association can be derived from assimilation theories, these theories have not gone unchallenged (Alba and Foner, 2015) . Several studies report no significant association between migrants' socio-economic attainment and their national belonging (De Vroome et al., 2011, De Vroome, Verkuyten and Martinovic, 2014; Tolsma, Lubbers and Gijsberts, 2012) . Some studies even report reversed or negative associations in line with an integration paradox, so that socio-economically successful migrants experience less national belonging than those with lower socio-economic attainment (Verkuyten, 2016; Geurts, Lubbers and Spierings, 2020 ). One explanation is that highly educated migrants are more sensitive to and/ or aware of group discrimination against their origin group or ethnic minorities in general (Tolsma, Lubbers and Gijsberts, 2012; Verkuyten, 2016; Steinmann, 2019) . Also when looking at opportunity structures, migrants with a higher socio-economic attainment are more likely to share time with natives via exposure to certain societal spheres, including living in certain neighbourhoods or being employed in higher qualified jobs (Rollock et al., 2011; Van Doorn et al., 2013) . As such, the risk of getting exposed to discriminatory beliefs and behaviour, towards migrants personally or in general, increases. Those with higher socio-economic attainment levels may face more personal resistance in intergroup relations, to the extent that natives perceive upwardly mobile migrants as more of a threat (Helbling, 2011) . In line with the well-established 'rejection-disidentification' mechanism, both migrants' personal experiences with discrimination and general perceptions of discrimination towards ethnic minorities predict a weaker sense of national belonging (Jasinskaja-Lathi, Liebkind and Solheim, 2009; Fleischmann and Phalet, 2016; Bobowik et al., 2017) .

A related explanation is that migrants who are socio-economically better off would have higher expectations from the residence country as a return on their (educational and economic) investments, resulting in disappointment and frustration when these expectations are not met (Verkuyten, 2016) . The same reasoning also applies to migrants who obtained their highest education level in their origin country, where unmet expectations undermine their sense of belonging in the residence country (Diehl, Fischer-Neumann and Mühlau, 2016; Geurts, Davids and Spierings, 2021) .

Based on the above arguments and findings, we therefore propose a competing hypothesis: (H1b) migrants with higher socio-economic attainment will experience less national belonging.

Socio-economic attainment and national belonging: (perceived) ethnic boundaries as moderators

In light of mixed evidence of positive, negative, and null associations between migrants' socio-economic attainment and national belonging in European countries (Schaeffer and Kas, 2023) , we draw on ethnic boundaries as a framework for our comparative research, which highlights 'the power of wider cultural notions of membership to constrain individual identity choices' in particular societies or intergroup settings (Simonsen, 2016) . Ethnic boundaries are social constructions that define group membership and explain why migrants' sense of national belonging differs across intergroup contexts, so that bright ethnic boundaries undermine and blurred ethnic boundaries enable migrants' belonging in the residence country (Wimmer, 2008; Ersanilli and Saharso, 2011; Simonsen, 2016) . Our study newly articulates the moderating influence of (perceived) ethnic boundaries in situating previously mixed findings of the association between migrants' socio-economic attainment and national belonging. We acknowledge that such perceptions can also mediate the association, i.e., as part of integration paradox mechanisms, which also will be acknowledged in the analyses.

We build on Schaeffer and Kas (2023) in addressing moderation, who argue that an integration paradox depends on the socio-political context of reception and associated levels of prejudice among the majority population (Schlueter, Meuleman and Davidov, 2013) . To situate and grasp mixed findings and opposing associations between socio-economic attainment and national belonging, our study explores how contextual and individual variation in (perceived) ethnic boundaries condition the association between migrants' socioeconomic attainment and their national belonging.

We argue that only in inclusive circumstances in which ethnic boundaries are more blurry, migrants' socio-economic attainment can translate more readily into a greater sense of belonging to the residence country. These circumstances enable a positive association between migrants' socio-economic attainment and national belonging as they allow for classic assimilation theory to play out. It is under these more inclusive circumstances that it is likely that socioeconomic attainment results in more positive contact with natives, exposure to norms and values, and feeling like a valued member of society, which translates into more national belonging accordingly.

Summing up, we expect that (H2) a positive association between migrants' socio-economic attainment and national belonging is most likely in countries with more (formal and informal) inclusive ethnic boundaries, such as the Netherlands. We thus also expect that if countries have more bright and exclusive ethnic boundaries, arguably Denmark and Germany compared to the Netherlands, a so-called integration paradox is more likely. These exclusive circumstances may interfere with the positive spillover effects of migrants' socio-economic attainment for their national belonging, as bright ethnic boundaries may instead augment integration paradox mechanisms of disappointment and awareness of discrimination (Jaskulowski and Pawlak, 2020; Geurts, Davids and Spierings, 2021) . We thus argue that such (integration paradox) mechanisms are activated under more exclusive circumstances, resulting in a negative association between migrants' socio-economic attainment and national belonging (Schaeffer and Kas, 2023) .

We moreover expect that ethnic boundaries not only vary between residence countries but also differ across origin groups (within and across residence countries). Previous research pointing to a so-called 'ethnic hierarchy' supports the notion that certain groups face more rigid ethnic boundaries, apparent in more SUCCEEDING WITHOUT BELONGING? experiences of discrimination (Lamont and Molnár, 2002; Phalet et al., 2013; Diehl et al., 2021) . Bright boundaries are more likely for origin groups seen as more distant (Hagendoorn, 1995) and more identifiable as such (based on, i.e., ethnicity and religiosity) (Schaeffer and Kas, 2023) . This is underlined by the finding that a discrimination paradox is more likely for migrants who are more identifiable based on race and religion (Flores, 2015) and for migrants who have a first name uncommon among the majority population (Tuppat and Gerhads, 2020) ; in our data, this can refer to migrants from Turkey, China, Japan, and the Philippines.

In our analyses, we will therefore explore possible differences in associations across origin groups and residence countries, which we will do without formulating concrete hypotheses as knowledge on how specific ethnic boundaries vary across these origin groups within specific residence countries is limited.

Our second major research aim is to test how individual perceptions of ethnic boundaries condition the link between migrants' socio-economic attainment and sense of national belonging, as a hypothetical micro-mechanism of the moderation argument presented above. In particular, we assess perceived discrimination of one's origin group as a proxy for these perceptions. 4 We already know that migrants' experiences of relative deprivation or unequal treatment of their group hamper their motivation and opportunities to feel part of the residence country (Verkuyten and Martinovic, 2012; Diehl, Fischer-Neumann and Mühlau, 2016; Fleischmann and Phalet, 2016) . We in addition expect that such perceptions of origin group discrimination condition how migrants' socio-economic attainment and national belonging are related. When little discrimination is perceived, the classic assimilation theory can play out as those with higher socio-economic attainment have more positive contact with natives, are exposed to norms and values, and feel a valued member of society, resulting in more belonging. Accordingly, we hypothesize that (H3) a positive association between migrants' socio-economic attainment and national belonging is most likely when perceptions of group discrimination are low. At the same time, this hypothesis implies that when perceptions of group discrimination are higher, a negative association between migrants' socio-economic attainment and national belonging is more likely, as these perceptions augment the integration paradox mechanisms. So, these circumstances are likely to enhance that (especially) migrants with a higher socio-economic attainment are hampered by discriminatory beliefs in society and unmet expectations, resulting in a lower sense of national belonging to the residence country.

Data and operationalization

To test the hypotheses, we make use of the MIFARE survey, containing data of 10 first-generation migrant groups in Denmark, Germany, and the Netherlands. The origin groups included in each residence country migrated from the Philippines, Japan, China (excluding Hong Kong), Poland, Russia, Spain, Great Britain, Romania, Turkey, and the United States. Migrants from the Philippines were not sampled in Germany as the group was too small (Bekhuis et al., 2018) . Data were collected between 1 December 2015 and 25 January 2016.

In Denmark and the Netherlands, the sampling was done by national statistics agencies using the Civil Registration System. In Germany, sampled municipalities were contacted. Representative samples were drawn based on the distribution of these migrant groups within the respective country, upon which migrants' origin country was also determined. A gross sample of around 1,000 was drawn for each origin group, with larger samples among non-Western groups due to possible higher non-responses. The response rate was 29 per cent with a total of 10,162 responses received, including a total of 8,807 of migrant respondents analysed for this study. In general, the response rates and total number of responses were lower in Germany, supposedly due to fact that the sampling was not based on a register. Background information of the origin groups can be found in Supplementary Table A2 of Supplementary Appendix.

Samples of migrants between the age of 18 and 75 years old were invited to participate in a written or online survey. Only migrants who migrated at age 16 or older and who had been in the residence country for at least 1 year were eligible. Invitations and questionnaires were translated into migrants' mother tongue, with two thirds preferring their mother tongue over the national-language version of the survey. On average, the questionnaire took 25 minutes. To further boost response rates, an incentive of 10 euros was offered.

Dependent measure

Migrants' national belonging was measured using the item: How strong, would you say, is your sense of belonging to people from [the residence country]? Answer categories were (0) not at all, (1) weak, (2) moderate, (3) close, and (4) very close. A similar item was used in previous studies (Fokkema and De Haas, 2015; Fleischmann and Phalet, 2018) . Missing values (2.1 per cent) were listwise deleted.

Independent variables

In line with earlier work, we used multiple measures of migrants' socio-economic attainment including their educational qualifications in both origin and residence countries, and their employment and earnings in the residence country (De Vroome et al., 2011; Geurts, Lubbers and Spierings, 2020) .

To measure educational attainments, we used the highest completed level of education in the origin country and in the residence country separately. In view of real constraints on cross-border recognition of qualifications and transferability of human capital, we followed De Vroome et al. (2011) and take into account where one's education was obtained. Both measures of educational attainment were standardized into the International Standard Classification of Education (ISCED) scale of 2011, which ranges from (0) no formal education to (6) upper-level tertiary education. We treated these two measures of education as continuous scales. 5 By including both measures simultaneously in the analyses, we elucidate the unique contributions of origin and residence country qualifications, and we take into account the (differential) selectivity of migration from different origin countries to each residence country.

To assess economic attainment proper, we supplemented migrants' employment status as an indicator of labour market participation with their household income as a measure of economic resources or wealth. Employment status refers to respondents' current main activity: (0) unemployed, (1) employed, (2) non-employed, and (3) in education. 6 The non-employed category includes those who were retired, looked after the home, were permanently sick, or unable to work as distinct from the unemployed proper (out of work and looking for paid work). These four categories were included as dummy variables with unemployed as reference category. Household income (after tax and compulsory deductions) refers to self-reported monthly totals ranging from (0) less than 600 euro to (10) 5,000 euros or more, which is adjusted for household size following the square root scale.

To gauge individual perceptions of ethnic boundaries, respondents were asked to rate the frequency of perceived group discrimination: Some say that migrants are being discriminated against in [residence country]. Others say this is not the case. How often do you think that migrants from [country of origin] are discriminated against in [residence country]? Answer categories are (0) never, (1) almost never, (2) sometimes, (3) often, and (4) very often.

Control variables

We control the associations of interest for numerous factors. Migrants' duration of stay is measured by the number of years between migration and survey years, ranging from 1 to 60 years (excluding 0.01 per cent outliers above 60). Migrants' age at migration and gender, (0) man, (1) woman, were included, and also their intention to stay: How long do you plan to stay in [residence country]? Answer categories were recoded into dummy variables: (0) 3 years or less (including the original categories '1 year or less' and 'more than 1 year, but not more than 3 years'), (1) more than 3, but less than 10 years (including original categories '3 years or more, but not more than 5 years' and '5 years or more, but less than 10 years'), (2) 10 years or more, and (3) don't know.

To replace missing values on all predictor or control variables, a maximum likelihood approach was used, which uses the available information in the specified model to provide a maximum likelihood estimation (Enders, 2001) , resulting in a final sample of 8,619 respondents (in case of listwise deletion, the sample size would have been 6,214). Table 1 presents descriptive statistics by residence country, indicating percentages of missing values.

Analytical strategy

Multiple-group structural equation modelling was applied to compare and test the associations between indicators of socio-economic attainment and national belonging across the several origin groups in specific residence countries. This way of modelling offers the advantage of conducting multiple-group tests of associations: it allows us to estimate the main associations for every combination of origin group and residence country to test our hypotheses instead of separately analysing the associations per origin group, residence country and including interaction terms to test for significance. This method moreover allows us to identify and group origin groups based on similar patterns in the associations that are tested within residence countries, as we expect H2 to possibly vary across origin groups. As such, it paints a bigger picture on whether and to what extent the associations between migrants' socio-economic attainment and national belonging vary across origin groups and residence countries. Doing so improves upon grouping origin groups on forehand based on expected similarities due to, for example, migration motive, religion, or region of birth, and instead, our results paint a picture in which origin groups in each residence country show similar patterns in the specific associations central to this study (which, due to adding the perceived discrimination variable, varies between results presented in Tables 2 and 3 ).

To rigorously test variable associations (path coefficients) across origin groups in each residence country, we start from baseline models freely estimating all coefficients for each origin group in each residence country (unconstrained effects). These results are presented in Figures 2 3 4 5 and are in line with the results that OLS regression analyses offer (presented in Supplementary Table 3 ). Next, we compare the relative model fit of the model without imposing (partial) equality constraints-the unconstrained model-to the model with such constraints imposed based on similar associations per socio-economic attainment indicators. If the model fit increases significantly by imposing equality constraints, we set (similar) coefficients equal across origin groups within residence countries (constrained effects). Thus, our results present both unconstrained effects from each origin group across residence countries, as well as constrained effects where certain origin groups within a residence country are merged, based on their similarity in association patterns, to improve model fit. We select and present the most parsimonious models with equality constraints except when the fit of the constrained model deteriorates significantly. When discussing the unconstrained results, we will be cautious in labelling non-significant associations as dissociations, as the limited number of respondents for each origin group in each residence may drive these findings. When presenting results, we therefore discuss both significance levels and effect sizes based on the confidence intervals presented in the figures.

To test H1 and H2, we include all measures of socio-economic attainment and controls (unconstrained models presented in Figures 2 3 4 5 , constrained model in Table 2 ). Next, we add perceived group discrimination (presented in Table 3 ) and subsequently test interactions with socio-economic measures one by one to test Hypothesis 3. Tables 2 and 3 thus show constrained path coefficients in separate multi-group models for Denmark, Germany, and the Netherlands in the same table, presenting the association patterns for the same-origin group in each country in parallel. For reasons of readability, we only present results testing our hypotheses, effects of the control variables are presented in Supplementary Table A4 . Significant interactions of perceived group discrimination with socio-economic measures are described and shown in separate graphs for specific origin groups by residence country combinations and presented in Supplementary Table A7 in Supplementary Appendix for reasons of readability. Table A1 gives an overview of model fit indices. All presented models show good fit as evident from comparative fit indices above 0.95, Tucker-Lewis indices above 0.92, and root mean square errors of approximation below 0.04. Explained variance varies across origin groups and residence countries, with the least variance explained for migrants from the Philippines and Turkey and most explained for migrants from Poland.

Results

Contextual variation across origin groups and residence countries

Results (Figures 2 3 4 5 and Table 2 ) suggest that national belonging relates differently to specific aspects of socio-economic attainment across origin groups and residence countries.

Starting with educational attainment in the origin country, higher pre-migration qualifications are most often not significantly related or even negatively associated with national belonging (in line with H1b). For instance, migrants from Spain with higher education report less belonging in all three countries. Exceptions are positive associations for migrants from Russia and Turkey in Denmark and migrants from Turkey in Germany, who report more belonging with higher education (supporting H1a). With the exception of Spanish migrants, associations typically also vary within origin groups across residence countries. For migrants from Russia, for instance, their Russian educational qualifications are positively associated with belonging in Denmark, yet negatively associated with belonging in Germany.

Most significant negative associations are found in Denmark and Germany, although not only for origin groups that have more visible 'markers of difference'. In the Netherlands, we found significant negative associations for only three groups, migrants from Spain, Downloaded from https://academic.oup.com/esr/article/41/1/17/7634658 by guest on 02 May 2025 29 SUCCEEDING WITHOUT BELONGING? Figure 4 Unconstrained multi-group SEM analyses of migrants' employment (versus unemployment) on sense of belonging to the residence country (N Denmark = 3,157, N Germany = 2,318, N The Netherlands = 3,144) Note: Line with triangle is Germany (GR), line with square is the Netherlands (NL) and line with diamond is Denmark (DK). Source: MIFARE 2015, control variables are included but not presented. Source: MIFARE 2015, control variables are included but not presented Russia, and Romania. In the unconstrained model, we find a lot of associations that do not reach statistical significance, where most show a small negative association with national belonging.

For educational attainment in the residence country (after taking into account the origin country education), associations with belonging are mostly positive and significant in Denmark and the Netherlands (in line with H1a), except for a negative association for Chinese migrants in the Netherlands (in line with H1b). With respect to the Netherlands, this partly confirms H2 as a positive association is more likely in a more inclusive context. In Denmark, positive associations are found for all origin groups, most also significant in the unconstrained model. The negative association for migrants from China in the Netherlands is seemingly in line with a notion that an integration paradox is more likely for those with visible 'markers of differences' (Flores, 2015) . However, such findings on group differences should be qualified as no evidence of a similar paradox is found for other racialized origin groups in the Netherlands nor for migrants from China in the other two residence countries. In contrast, educational attainment in Germany is only positively and significantly related to belonging among migrants from Japan and Russia. Among the other origin groups, effects are small and range between -0.05 and .06. This seemingly confirms H2, suggesting that a lack of positive associations is most likely in Germany, the residence country where formal and informal manifestations of ethnic boundaries are most bright. Overall, the association between post-migration educational attainment and migrants' national belonging thus differs mainly across residence countries, and less so across origin groups.

Turning to employment status, we again find positive associations with belonging for most origin groups in the Netherlands and Denmark. This supports H1a and H2 to a certain extent, suggesting that a positive association is most likely in the Netherlands as a more inclusive residence country, except again for a strong negative association among migrants from China. In Germany, however, there are less significant associations found and most associations are small (between -0.05 and 0.05), except for significant positive associations among migrants from Russia and Japan. This again supports our hypothesis (H2) that a positive association is more likely in more inclusive contexts compared with contexts with more bright (informal) ethnic boundaries, like Germany.

Finally, household income (net of education and employment) positively relates to national belonging for all groups in Germany (except for migrants from Turkey), as well as for migrants from China, Great Britain, Poland, and Spain in Denmark and migrants from China and Poland in the Netherlands (in line with H1a). In general, the effects are small. Nevertheless, the most positive associations are found in Germany, which is not in line with H2. Conversely, the near absence of positive associations in the Netherlands, as most inclusive context, is equally unexpected and opposes H2. Looking at the various origin groups, we find no clear patterns between those who are more identifiable as ethno-religious minority and those who are not, as we find positive and negative associations for both across the board.

Summing up, how migrants' socio-economic attainment relates to their national belonging varies significantly across residence countries as well as-to a lesser extent-across origin groups. While many associations are not significant, there is evidence on both sides of the ledger, so that socio-economic attainment positively predicts belonging in some intergroup contexts (supporting H1a) and negatively in others (supporting H1b). For most indicators of socio-economic attainment, except for income, a positive association is more likely in a more inclusive context such as the Netherlands. There is less clear evidence of origin group differences within countries, as we do not find that groups who are more identifiable as ethnoreligious minority are more likely to evince an integration paradox.

Previous research has stressed the role of a possible mismatch between educational investments and economic success in explaining (reduced) national belonging among migrants (Geurts, Lubbers and Spierings, 2020; Diehl et al., 2021) . As the data used did not contain migrants' occupational status nor their feeling of relative deprivation, such discrepancies within migrants' socio-economic attainment could not be addressed optimally. Nevertheless, Supplementary Table A5 in Supplementary Appendix presents additional analyses to explore the role of such a mismatch in line with Geurts, Lubbers and Spierings' (2020) approach. While realizing the limitations of measuring a mismatch indirectly as such, we have presented moderation effects to indicate to what extent the effect of pre-migration educational qualifications changes depending on one's economic attainment (household income and employment). We find limited support that a mismatch is negatively associated with national belonging, which we only found among migrants from Russia in the Netherlands and among migrants from China and the United States in Denmark.

Perceived group discrimination as moderating condition

To explain the association between socio-economic attainment and national belonging, Table 3 first adds perceived group discrimination. General patterns described above hold when perceived discrimination is added, although a few associations turn

Germany China and Russia -0.004 0.052 All origin groups -0.182*** 0.025 The Netherlands Russia and Romania -0.058 0.035 All other origin groups -0.147*** 0.021 Source: MIFARE 2015, control variables are included but not presented. ***P < 0.001, **P < 0.01, *P < 0.05, ∼ P < 0.1. Table 3. Continued insignificant when perceive discrimination, suggesting that perceived discrimination mediates the effect of socioeconomic attainment on national belonging for some, supporting previous research. Supplementary Table A6 in Supplementary Appendix however illustrates that significant associations between socio-economic attainment and perceived discrimination are limited, with examples of both positive and negative associations, suggesting mediation by perceived group discrimination is not apparent for all groups.

In line with the rejection-disidentification model, we do find that those who perceive more discrimination towards their origin group experience less national belonging (Bobowik et al., 2017) . We find this for most groups in all three countries, and perceived discrimination never significantly improves national belonging. Thus, regardless of socio-economic attainment, migrants who perceive more discrimination against their origin group tend to experience less national belonging.

The hypothesized interaction of socio-economic attainment with perceived group discrimination was tested in models presented in Supplementary Table A7 for each origin group and residence country combination. For reasons of readability, we present significant interactions for specific combinations of origin groups with residence countries in figures to exemplify similar interaction patterns among other origin groups and countries. With respect to pre-migration educational attainment, we find significant interaction patterns for migrants from the Philippines, Romania, Great Britain, and United States in Denmark, migrants from Japan in Germany, and migrants from Poland and Great Britain in the Netherlands. In Figure 6 , among migrants from Romania in Denmark, pre-migration education negatively predicts belonging at higher levels of perceived discrimination, whereas education and belonging are positively associated at low discrimination levels, supporting H3. The only significant exception is found in the case of migrants from Great Britain in Germany, where never perceiving group discrimination augments a negative association between socio-economic attainment and national belonging, and when discrimination is perceived, this association turns positive.

With respect to educational attainment after migration, we again find similar significant interactions in several intergroup contexts: migrants from Japan, Turkey, SUCCEEDING WITHOUT BELONGING?

and the United States in Denmark, migrants from Japan and Spain in Germany, and migrants from China and Philippines in the Netherlands. Significant interactions are in line with H3, so that higher socio-economic attainment predicts more national belonging only when perceived discrimination is low. Figure 7 about migrants from Japan in Germany shows that education in Germany positively predicts national belonging in the absence of perceived discrimination against Japanese migrants. Again, one significant exception is found in the case of migrants from Great Britain in the Netherlands, where there is seemingly no effect when discrimination is never perceived, but a positive association when more group discrimination is perceived.

We find no significant interactions of perceived discrimination with migrants' employment status. For household income, however, we do find some support for our hypothesis among migrants from Romania, Russia, Spain, and Poland in Denmark and among migrants from Great Britain in the Netherlands. In Figure 8 , this interaction pattern is shown for migrants from Great Britain in the Netherlands.

Summing up, migrants' perceptions of discrimination can moderate how socio-economic success relates to their national belonging. Although many tested interactions fall short of significance, almost all significant interactions do show the hypothesized pattern. Typically, only when migrants perceive little or no discrimination against their origin group, they can derive an enhanced sense of belonging from their socioeconomic attainment. When more group discrimination is perceived, in contrast, socio-economic success is more likely decoupled from, or even inversely related to, belonging. Our results thus provide some evidence for the proposition that perceived ethnic boundaries moderate the attainment-belonging association, although support is not significant in all intergroup contexts, and never significant for employment status.

Conclusion and discussion

This article examined the role of migrants' socioeconomic attainment in their sense of national belonging, taking a double comparative perspective across 10 origin groups in three Western-European residence countries. This design adds to earlier comparative research by surveying the same-origin groups in three different countries. To elucidate mixed results of positive and negative associations and dissociations between migrants' socio-economic attainment and their belonging to the residence country, we explore the role of ethnic boundaries and the perceptions thereof as moderating conditions of this association.

Our first research aim was to map contextual variation in the association between socio-economic attainment and national belonging across origin groups and residence countries. Drawing on comparative survey data, we indeed established contextual variationin line with the notion that ethnic boundaries differ across specific origin groups and residence countries. We find evidence of both positive and negative associations between migrants' socio-economic attainment and their belonging, whereas many associations did not reach statistical significance. Importantly, there was meaningful variation between the three residence countries, so that, for instance, post-migration educational investments more often enhance national belonging in the Netherlands than in Germany, where informal manifestations of ethnic boundaries are more exclusive.

There was further meaningful variation between origin groups, which was almost never uniform across the Note: Household income ranges from 0 (less than 600 euros) to 10 (5000 euros or more).

Source: MIFARE 2015. three residence countries. Previous comparative research has either focused exclusively on residence countries as comparative cases, so that integration policies and public attitudes may define more or less rigid ethnic boundaries excluding migrants, without distinguishing further between origin groups (Alba and Foner, 2015) . Other research mainly compared origin groups within one single country (with some exceptions, see Van Tubergen, Maas and Flap, 2004) . Our comparative evidence elucidates the interplay of origin group and residence country differences, with real consequences for how migrants' socio-economic attainment and national belonging relate. We find that this association varies across origin groups and residence countries, suggesting that ethnic boundaries thus differ across specific intergroup contexts. We conclude that negative associations between migrants' socioeconomic attainment and belonging are not simply more likely for certain origin groups that are more disadvantaged or discriminated against based on religious or ethnic grounds as salient 'markers of differences', such as migrants from Turkey or East-Asia. Conversely, we found evidence of a so-called integration paradox for migrants from Spain, who face less rigid ethnic boundaries in intergroup relations with natives as a relatively 'higher ranked' origin group. For migrants from Turkey, in contrast, who are typically at the bottom end of the so-called 'ethnic hierarchy' in terms of (religious) intergroup distance, there is hardly any evidence of an integration paradox. These comparative findings therefore challenge previous research stressing rigid ethnic and religious group boundaries in European migration contexts (Trittler, 2017) .

These descriptive findings call for more comparative research at the intersection of distinct origin groups with receiving societies to better understand differences across these intergroup contexts and to delve deeper into the underlying mechanisms of boundary-blurring processes and its implications for the association between socio-economic attainment and national belonging. A multi-level analysis including origin group and residence country characteristics and a larger sample size would allow for such a more extensive exploration. We moreover recommend future research to explore the importance of migrants' self-identification, as we foresee that this shapes how one is affected by exposure to group discrimination.

Our second research aim was to assess whether perceived ethnic boundaries, indicated by perceptions of group discrimination, may help us situate the variable role of migrants' socio-economic attainment for national belonging. Our multi-group models replicate earlier findings that perceiving more group discrimination hampers migrants' national belonging (Diehl, Fischer-Neumann and Mühlau, 2016; Jasinskaja-Lahti et al., 2009) , regardless of their socio-economic attainment. Moreover, we find some evidence that the association of socioeconomic attainment with national belonging can indeed be contingent on such perceptions of group discrimination. For some origin groups, only when little discrimination is perceived, higher socioeconomic attainment translates into more national belonging. This seems to apply especially to migrants from Eastern Europe, typically ranked lower in the ethnic hierarchy based on their socioeconomic status and to migrants from East-Asia, generally more racialized and identifiable as ethnoreligious minority. We were limited in our discrimination measurement, and recommend future research to also explore the role of personal discrimination and its relative importance next to group discrimination.

Introducing an ethnic boundaries perspective to assimilation and integration paradox theories thus seems worthwhile to make sense of the considerable contextual variation in the association or lack thereof between migrants' socio-economic attainment and national belonging in Western Europe. (Perceptions of) ethnic boundaries clearly matter for whether and how migrants' socio-economic success translates to national belonging-with important consequences for social cohesion and liberal democracy.

Another contribution of this article is the inclusion of multiple indicators of migrants' socioeconomic attainment, which are not necessarily aligned or equally important across contexts, from migrants' educational qualifications before and after migration to their labour market participation and earnings (as recommended by De Vroome et al., 2011) . Seemingly different mechanisms are at play, as some aspects of socio-economic attainment are not significantly associated with belonging (being employed), whereas others tend to enable (education in the residence country) or hamper (education in the origin country) national belonging. Possible explanations for differential associations could be that higher educational qualifications in the origin country entail higher expected returns in the residence country, which do not always materialize (Geurts, Davids and Spierings, 2021) , whereas educational qualifications in the residence country may enable more immediate and substantial returns on investments (Turper et al., 2015) . Distinguishing between these indicators also sheds new light on mixed results in earlier research, pinpointing which aspects of migrants' socioeconomic attainment do and do not relate to national belonging. In future studies on migrants' sense of belonging, we recommend to address the relativity of postmigration attainments and assess the so-called mismatch and related experiences of relative deprivation more directly (Diehl et al., 2021) .

SUCCEEDING WITHOUT BELONGING?

Reflecting on this study's approach and findings, we set out to contextualize the (lack of) association between migrants' socio-economic attainment and national belonging in Western Europe. Our findings show that mixed results in previous research can be understood by looking at variation across origin groups and residence countries, which show that positive and negative associations exist across such intergroup contexts. In this way, these results challenge the assumption, present in classic assimilation paradigms, that there is a linear association between socio-economic attainment and national belonging. Instead, our findings shift focus to the specific intergroup context of migrants by articulating and illustrating that (perceived) ethnic boundaries condition how migrants' socio-economic attainment links to national belonging and that these associations therefore vary across (combinations of) origin groups and residence countries. Future comparative research on migrants' national belonging, and supposedly other outcomes, therefore benefit from addressing contextual variation across origin groups and residence countries to pinpoint origins of variation among groups. Importantly, we conclude that perceiving group discrimination interferes with possible belonging benefits of migrants' socio-economic attainment, so that only when perceived group discrimination is low, migrants' socio-economic attainment translates into a greater sense of national belonging.

Notes

1. This score is a tool that measures national integration policies based on 167 policy indicators to provide a multidimensional picture of migrants' opportunities to participate in society (Huddleston et al., 2015) . A higher score indicates more inclusive policies. 2. For this article, we use survey data collected in 2015, which is why we describe the context up until then. If we would look at the current status quo, we see that Denmark did not continue this trend. 3. Within the European context, integration is used as an umbrella term referring to increasing similarities in migrant and native outcomes across socio-economic, sociopolitical, and socio-cultural domains (Alba & Foner, 2015) . 4. Unfortunately, the data did not allow us to capture experiences of personal discrimination, which we will reflect upon in the conclusion. 5. For all ordinal and interval items, deviation from linearity was checked and the operationalization was decided upon accordingly. 6. Additional analyses without students did not yield different results.

References

1. The settlement country and ethnic identification of children of Turkish immigrants in Germany, France, and the Netherlands: what role do national integration policies play? E Ersanilli S Saharso International Migration Review 45 2011
2. Identity Conflict or Compatibility: A Comparison of Muslim Minorities in Five European Cities Fenella Fleischmann Karen Phalet 10.1111/pops.12278 Political Psychology Political Psychology 0162-895X 1467-9221 37 4 2016 Wiley
3. Religion and National Identification in Europe: Comparing Muslim Youth in Belgium, England, Germany, the Netherlands, and Sweden Fenella Fleischmann Karen Phalet 10.1177/0022022117741988 Journal of Cross-Cultural Psychology Journal of Cross-Cultural Psychology 0022-0221 1552-5422 49 1 2018 SAGE Publications England, Germany, the Netherlands, and Sweden
4. The resurgence of race in Spain: perceptions of discrimination among immigrants R D Flores Social Forces 94 2015
5. Pre‐ and Post‐Migration Determinants of Socio‐Cultural Integration of African Immigrants in Italy and Spain Tineke Fokkema Hein De Haas 10.1111/j.1468-2435.2011.00687.x International Migration International Migration 0020-7985 1468-2435 53 6 2015 Wiley
6. The good and the bad: do immigrants' positive and negative evaluations of life after migration go hand in hand N Geurts M Lubbers Journal of International Migration and Integration 24 2023
7. Structural position and relative deprivation among recent migrants: a longitudinal take on the integration paradox N Geurts M Lubbers N Spierings Journal of Ethnic and Migration Studies 46 2020
8. The lived experience of an integration paradox: why high-skilled migrants from Turkey experience little national belonging in the Netherlands N Geurts T Davids N Spierings Journal of Ethnic and Migration Studies 47 2021
9. M M Gordon Assimilation in American Life: The Role of Race, Religion, and National Origins New York Oxford University Press 1964
10. To be 'Danish', becoming 'Muslim': Contestations of national identity? Gudrun Jensen T Journal of Ethnic and Migration Studies 34 2008
11. Intergroup biases in multiple group systems: the perception of ethnic hierarchies L Hagendoorn European Review of Social Psychology 6 1995
12. Integrating immigrants in the Netherlands: cultural versus socio-economic integration L Hagendoorn J Veenman W Vollebergh 2003 Ashgate Aldershot
13. R Hansen Migration to Europe since 1945: Its history and its lessons. The Political Quarterly 2003 74
14. Why Swiss-Germans dislike Germans: opposition to culturally similar and highly skilled immigrants M Helbling 2011 European Societies 13
15. T Huddleston Ö Bilgili A Joki Z Vankove Migrant Integration Policy Index Barcelona/Brussels CIDOB and MPG 2015. 2015
16. To identify or not to identify: national disidentification as an alternative reaction to perceived ethnic discrimination I Jasinskaja-Lathi K Liebkind E Solheim Applied Psychology 58 2009
17. Migration and lived experiences of racism: the case of high-skilled migrants in Wrocław K Jaskulowski M Pawlak International Migration Review 54 2020
18. Beyond national models: Civic integration policies for immigrants in Western Europe C Joppke West European Politics 30 2007
19. The study of boundaries in the social sciences M Lamont V Molnár Annual Review of Sociology 28 2002
20. Dutch don't dance'-leisure experiences and sense of belonging among Polish migrants in the Netherlands E Peperkamp Leisure Studies 37 2018
21. K Phalet The making and unmaking of religious boundaries: comparing T and M Muslims in European cities 2013 1
22. The new second generation: segmented assimilation and its variants A Portes M Zhou The Annals of the American Academy of Political and Social Science 1993 530
23. The public identities of the black middle classes: Managing race in public spaces N Rollock Sociology 45 2011
24. Sampling immigrants in the Netherlands and Germany K Salentin H Schmeets Comparative migration Studies 5 2017
25. The integration paradox: a review and meta-analysis of the complex relationship between integration and reports of discrimination M Schaeffer J Kas 10.1177/01979183231170809 International Migration Review 2023
26. Immigrant integration policies and perceived group threat: a multilevel study of 27 Western and Eastern European countries E Schlueter B Meuleman E Davidov Social Science Research 42 2013
27. How the host nation's boundary drawing affects immigrants' belonging K Simonsen Journal of Ethnic and Migration Studies 42 2016
28. P Sniderman L Hagendoorn A Hagendoorn When Ways of Life Collide: Multiculturalism and its Discontents in the Netherlands Princeton Princeton University Press 2007
29. The paradox of integration: why do higher educated new immigrants perceive more discrimination in Germany J Steinmann Journal of Ethnic and Migration Studies 45 2019
30. The Paradox of Integration Irene Ten Teije Marcel Coenders Maykel Verkuyten 10.1027/1864-9335/a000113 Social Psychology Social Psychology 1864-9335 2151-2590 44 4 2013 Hogrefe Publishing Group
31. Education and cultural integration among ethnic minorities and natives in The Netherlands: a test of the integration paradox J Tolsma M Lubbers M Gijsberts Journal of Ethnic and Migration Studies 38 2012
32. Explaining Differences in the Salience of Religion as a Symbolic Boundary of National Belonging in Europe Sabine Trittler 10.1093/esr/jcx069 European Sociological Review 0266-7215 1468-2672 33 5 2017 Oxford University Press (OUP)
33. Immigrants' first names and perceived discrimination: a contribution to understanding the integration Paradox J Tuppat J Gerhads European Sociological Review 37 2020
34. Who is Less Welcome?: The Impact of Individuating Cues on Attitudes towards Immigrants Sedef Turper Shanto Iyengar Kees Aarts Minna Van Gerven 10.1080/1369183x.2014.912941 Journal of Ethnic and Migration Studies Journal of Ethnic and Migration Studies 1369-183X 1469-9451 41 2 2015 Informa UK Limited Journal of Ethnic and Migration Studies
35. Explaining the integration paradox among small immigrant groups in the Netherlands M Van Doorn P Scheepers J Dagevos Journal of International Migration and Integration 14 2013
36. The economic incorporation of immigrants in 18 western societies: origin, destination, and community effects F Van Tubergen I Maas H Flap American Sociological Review 69 2004
37. SUCCEEDING WITHOUT BELONGING?
38. The integration paradox M Verkuyten The American Behavioral Scientist 60 2016
39. Immigrants' national identification: meanings, determinants, and consequences M Verkuyten B Martinovic Social Issues and Policy Review 6 2012
40. Super-diversity and its implications S Vertovec Ethnic and Racial Studies 30 2007
41. Settlement intentions and immigrant integration: the case of recently arrived EU-immigrants in the Netherlands G G Wachter F Fleischmann International Migration 56 2018
42. The German statistical category 'migration background': historical roots, revisions and shortcomings A K Will Ethnicities 19 2019
43. Elementary strategies of ethnic boundary making A Wimmer Ethnic and Racial Studies 31 2008
44. Ethnic self-identification of first-generation immigrants L Zimmermann K F Zimmermann A Constant International Migration Review 41 2007