Understanding causes of adolescents' ascriptions of peers' dual ethnic and national belonging

Descendants of migrants often feel belonging to the ethnic minority group of their parents and the national group (i.e., dual identifiers). If second-generation migrants are recognized as members of both groups, they can create social bridges and improve intergroup relations. However, people often ascribe them only minority belonging, thereby not recognizing dual belonging. Following the social identity approach, we examined who ascribes dual belonging to second-generation migrants and why they do so. We focus on the role of national self-identification, and endorsing a civic or ethnic conception of nationhood on the ascriptions that national majority members make to their second-generation peers. We collected sociometric data from secondary vocational school classes in the Netherlands ( N students = 397, N classes = 36). Findings replicated that many second-generation migrants self-identified as dual (69%), but only a small number of dual ascriptions were given to them (15%). Although we found no effects for the hypothesized relations on ascribing only minority rather than dual belonging, national majority members who had a stronger national identification displayed a stronger tendency to ascribe only national rather than dual belonging.

We collected sociometric data from ethnically diverse secondary vocational school classes in the Netherlands that capture adolescents' ethnic/national ascriptions of peers in classrooms.As adolescents grow up in a society characterized by unprecedented ethnic diversity, school classes are a particularly relevant context where they encounter diversity and regularly interact with peers.Our study advances existing methodology in ethnic identity research, as ascriptions made by others are rarely measured (for an exception with only one minority group, see Boda & Néray, 2015) and existing research often relied on racial assessments from the interviewer (Harris & Sim, 2002;Saperstein & Penner, 2012) or assessments made in lab settings (e.g., Chen & Hamilton, 2012).Our data allow us to descriptively examine whether participants, based on their own ethnic self-identification, differ in giving ascriptions.It also provides us with the opportunity to investigate why national majority members do or do not ascribe dual belonging to their second-generation peers.

| ASCRIPTIONS OF ETHNIC AND NATIONAL BELONGING
Extensive research has been conducted on how migrants and their descendants might negotiate and combine ethnic minority and national identity (e.g., Benet-Martínez & Hong, 2014;Berry et al., 2006;Verkuyten, 2018).While various types and combinations of ethnic identities are possible (e.g., with multiple groups, differing in strength), in this study, we focus on dual belonging, defined as self-identification with both the ethnic minority and national group.
We study dual belonging in contrast with self-identification with only the ethnic minority or national group.
Scholars have argued that an ethnic identity is (re)constructed through a process of social ascriptions by oneself and others.This process involves which social groups (e.g., national, ethnic) you think you belong to (i.e., internal ascription or self-identification) and which groups others think you belong to (i.e., external ascription or categorization) (Barth., 1969;Verkuyten, 2018).Hence, as people categorize themselves into social groups, they also routinely categorize others as members of social groups (ascribe belonging) (Turner, 1985).In turn, these ascriptions inform attitudinal and behavioural expectations (Crisp & Hewstone, 2007).External ascriptions of belonging might be especially important for social interactions and intergroup relations, as people use these ascriptions to distinguish between ingroup and outgroup members (Dovidio & Gaertner, 2010).In fact, theoretical ideas on how dual identifiers might positively affect interethnic relations, for example, by blurring group boundaries and promoting inclusive social categories (Love & Levy, 2019), largely depend on the recognition of dual identifiers by others as belonging to both the in-and outgroup.Recent experimental research has shown that, for those prone to prejudice, being made aware of dual identifiers in one's environment reduced prejudice (Levy, 2019).Ascribing others a dual ethnic belonging might thus be beneficial for interethnic relations.

| CAUSES OF DUAL ASCRIPTIONS: SELF-IDENTIFICATION
Whereas studies on multiethnic or dual ethnic categorization in the European context are scarce (but see Boda & Néray, 2015), various studies on racial categorization have been conducted in the United States (e.g., Chen et al., 2014;Feliciano, 2016;Gaither et al., 2016;Kteily et al., 2014;Pauker et al., 2018).This literature shows that people rely on phenotype (e.g., skin colour) to determine someone's race, that they easily categorize faces in discrete binary categories (e.g., Black, White), and are prone to categorize Black/White biracial individuals as Black.People also take longer when categorizing others as biracial (e.g., Chen & Hamilton, 2012;Ho et al., 2011).Thus, in line with work on cross-categorization (e.g., Crisp & Hewstone, 2007), people are capable of simultaneously categorizing multiple social identities, but this process might be cognitively more challenging.Although it is unclear how well research on biracial individuals applies to descendants of migrants (because their dual belonging is not typically accompanied by physical markers), it seems likely that ascribing dual belonging also requires more flexible ways of dealing with group boundaries.
Dual identifiers deal with complex social categorizations for themselves, and might be more ready to also see others in complex ways (Love & Levy, 2019;Roccas & Brewer, 2002).For them, the social category of dual belonging might be more accessible, thereby facilitating ascriptions of dual belonging to others.SCT states that the extent to which a particular social identity becomes salient depends on the accessibility of that social identity for an individual (Turner, 1985).Accessibility, or perceiver readiness, refers to relevant individual aspects that a person brings to the situation, like the strength to which an individual identifies with a certain group or personality traits, that can make a social identity more salient.The theory furthermore proposes that when a shared social identity is salient, individuals tend to see themselves as part of that social group ('we') rather than as an individual person ('I') (Turner, 1985).In this process, they tend to highlight similarities between ingroup members, and exaggerate difference with outgroup members.Hence, one's own dual self-identification can make the relevant social categories more accessible, and make it more likely to recognize and ascribe dual belonging to others.Accordingly, we expect that individuals who self-identify as dual are more likely to ascribe dual belonging to second-generation migrants than those who self-identify only as member of a minority or the national group (H1).

| ASCRIBING DUAL BELONGING: GROUP BOUNDARIES
In the European context, research has shown that people with a migration background and a dual identity often feel that their national group membership, and by extension their dual identity, are not recognized by national majority members (e.g., Cárdenas et al., 2021;Slootman, 2018a).According to social identity theory, national majority members could be motivated to exclude people with a migration background from the national group because they have a desire to keep national group membership selective and maintain rigid group boundaries (Brewer, 1991;Tajfel, 2010).
Strongly identifying national majority members, who typically occupy a high status position, might in particular have a desire to maintain their socially dominant position in society (Gaither et al., 2016).Dual membership can raise questions of group loyalty, and divided loyalties can be perceived as a threat to the position of the dominant group (Kunst et al., 2019).Indeed, studies have found that strong identifiers (vs.weak identifiers) are more likely to protect their ingroup (Castano et al., 2002;Gaither et al., 2016).Gaither et al. (2016) found that White participants (in the United States) with higher levels of racial self-identification excluded more racially ambiguous Black/White targets from their ingroup.Hence, we expect that the stronger national majority individuals self-identify with the national group, the more likely they are to ascribe only minority belonging, compared to dual belonging, to second-generation migrants (H2).
Whether national majority members exclude those with a migration background from the national category and do not ascribe them dual belonging could also depend on majority members' conceptions of what a national group member is or should be.SCT (Turner, 1985) argues that group members use group norms, values and characteristics, as criteria for who belongs to the group and who is excluded.As inclusion criteria for national group membership, people can endorse a more ethnic or civic conception of nationhood (e.g., Feeney et al., 2020;Reijerse et al., 2013;Yogeeswaran & Dasgupta, 2014).An ethnic conception of nationhood is the belief that nationhood is based on shared ancestry and descent.It is related to the principle of jus sanguinis (right of blood) and means that one's nationality is determined by the nationality of one's (grand)parents.Hence, an ethnic conception excludes migrants and their descendants from national group membership because of their migration background.On the contrary, a civic conception defines nationhood based on citizenship and one's commitment to the nation's institutions (Feeney et al., 2020;Yogeeswaran & Dasgupta, 2014).Following this conception, anyone who embraces certain values and institutions is considered a part of the nation.In this way, a civic conception is more liberal and inclusive, as migrants and their descendants can be a part of the national group.Therefore, we expect that the stronger national majority individuals endorse an ethnic conception of nationhood, the more likely they are to ascribe only minority belonging, compared to dual belonging, to second-generation migrants (H3a).On the contrary, we expect that the stronger national majority individuals endorse a civic conception of nationhood, the less likely they are to ascribe only minority belonging, to compared dual belonging, to second-generation migrants (H3b).

| Data and participants
The study was approved by the ethical committee of the faculty of social and behavioural sciences of Utrecht University (23-0271) and preregistered on the Open Science Framework.iii Cross-sectional data were collected with an online survey, programmed in Qualtrics, which students completed in class on their smartphone.Participants were students from five secondary vocational schools (SVE) with ethnically diverse student bodies in the Netherlands.In 2022, in the Netherlands, students who attended SVE were at least 15 years and 82% was 22 years or younger (van Onderwijs, Cultuur en Wetenschap, 2022).In total, 518 students started the survey (N classes = 42).We excluded those who did not complete the survey (N = 58), and six classes in which no second-generation migrants (N = 24) or national majority members (N = 20) were present.Additionally, participants were listwise deleted because they had not indicated their birth country (N = 2), or did not give any ascriptions to second-generation peers (N = 17).
In the final sample of observers (N = 397, N classes = 36), 53.40% was female and 45.59% male.Following the definition of Statistics Netherlands (CBS, 2023), we classified observers based on their country of birth and the country of birth of their parents.In our sample, 33.75% participants were second-generation migrants (N = 134), they were born in the Netherlands but had at least one parent born abroad.Next, 15.11% were first-generation migrants (N = 60), they were born abroad themselves, and 51.13% were national majority members (N = 203), they were born in the Netherlands and both their parents were born in the Netherlands.iv Of those with a migration background, 13.92% had a Turkish background, 16.49% a Moroccan, 11.86% a Surinamese and 9.28% an Antillean background.Moreover, 38.14% filled in another origin, resulting in an additional 55 origin countries.v Secondary vocation education has several levels of training, participants attended level one: assistant training (7.81%), two: basic vocational training (12.85%), or three: professional training (79.35%).In total, 1,529 ascriptions were given to 137 targets (second-generation migrants).Three second-generation migrants were only included as targets and not observers because they received ascriptions, but did not give any to second-generation peers.

| Procedure
Teachers were approached by the researchers and asked if they wanted to collect data.Those who agreed to participate were informed about the goal and the procedure of the study, and received instructions from the researchers to collect the data themselves.All students who were present in the class completed the survey simultaneously, but independently and in silence.Prior to administering the survey, participants received an information letter and watched a video, informing them of the purpose of the study and their rights.They could stop with the survey at any time.Participants did not receive payment, but had a chance of winning a gift card of 10 euro at the end of the survey.The survey took on average 10-15 min and was completed on students' smartphones.On the first page of the survey, participants were informed of the study and by starting the survey they gave their informed consent.They were informed that they could skip any question and could stop with the survey at any time.By giving consent, they agreed that their participation was voluntary and that their data would be used for scientific research.In line with the general data protection regulation (GDPR), participants remained anonymous in the survey.

| Measures
Ascriptions of ethnic belonging.Teachers created a score sheet in which each student in the class received a number (e.g., 1 = Tim, 2 = Ibrahim).This score sheet was shown on a screen during the survey, and allowed students to give and receive ascriptions without providing their names.As participants also indicated which number was theirs, we obtained the information about whom the ascriptions were made.Then participants were asked 'To which culture or cultures do you think classmate number X belongs?You can choose multiple cultures'.They could select: Dutch (1), Turkish (2), Moroccan (3), Antillean (4), Surinamese (5), and/or other (6).We gave these options, because the four ethnicities are the largest ethnic minority groups in the Netherlands (CBS Statline, 2022).This question was repeated for all peers in the classroom, thus for number 1 to X (depending on the number of students in a class).Our dependent variable was based on the ascriptions of ethnicity given to second-generation migrants.The answers were recoded into three categories (dual, only minority and only national).Dual belonging was coded when both Dutch (1) and at least one ethnic minority group (2-6) were selected.Only minority was coded when only one or more ethnic minority groups were selected (2-6), and only national was coded when only Dutch was selected (1).It was treated as a nominal variable.
Strength of Dutch self-identification.Participants were asked about their self-identification as Dutch.They were asked 'How strongly do you feel Dutch?' (1 = not strong at all, 2 = not strong, 3 = moderately strong, 4 = strong, 5 = very strong).Strength of Dutch self-identification was treated as an ordinal variable.
Self-identification in categories.If participants indicated that their family also had another heritage country, they were asked 'How strongly do you feel belonging to people from your heritage country?' (1 = not strong at all, 2 = not strong, 3 = moderately strong, 4 = strong, 5 = very strong).Together with Dutch self-identification, these two questions were used to categorize participants into self-identification categories as only national (0/1), only minority (0/1) or dual (0/1).Someone was considered self-identifying as only national, if they scored three or higher on the Dutch self-identification question and two or lower on the minority self-identification question.In contrast, someone was considered self-identifying as only minority if they scored three or higher on the minority question and two or lower on the Dutch question.Lastly, someone was considered a dual identifier when they scored a three or higher on both questions.Some participants (N = 13) could not be categorized in any of these three self-identification groups because they scored lower than three on Dutch self-identification and, if applicable, on minority identification.They were given a missing value for 'self-identification in categories'.
Ethnic conception of nationhood.Participants were asked to what extent they agreed with three statements, taken from earlier research (e.g., Verkuyten & Martinovic, 2015), namely: 'A real Dutch person has Dutch ancestors', 'A real Dutch person is someone who is of Dutch origin', and 'A real Dutch person is someone who passes Dutch culture on to the next generation' (1 = totally disagree, 2 = disagree, 3 = neither disagree, nor agree, 4 = agree, 5 = totally agree).The three items had a reliability of α = .72.A mean score was created based on the three items (or on less items vi ) and a higher score indicated a stronger endorsement of an ethnic conception.Seven participants had a missing value on this variable.
Civic conception of nationhood.Participants were asked to what extent they agreed with the three statements, taken from earlier research (e.g., Verkuyten & Martinovic, 2015), namely: 'Anyone who legally resides in the Netherlands is a real Dutch person', 'A real Dutch person is anyone in possession of a Dutch passport', and 'Anyone who respects Dutch democracy and laws is a real Dutch person', using the same 1-5 answer scale as for ethnic conception.The three items had a reliability of α = .62.A mean score was created based on the three items (or on less items) and a higher score indicated a stronger endorsement of a civic conception.Six participants had a missing value on this variable.
Gender (recoded into 0 = female, 1 = male) and level of secondary vocational education (interval, level 1-4) were used as control variables.Gender was included because previous research has shown that women were more likely to classify others in complex ways (i.e., as multiracial) than men (Feliciano, 2016;Harris, 2002).Those who had selected 'other' for gender were set to missing (N = 4).In similar vein, level of education was added as a control variable because it reflects someone's cognitive style, which has been argued to be an antecedent of complex classification (Miller et al., 2009).

| Plan of analysis
Two separate two-level multinominal logistic regression models were estimated in Mplus (version 8.8).To deal with missing values on our independent variables full information maximum likelihood (FIML) was performed.For the model to test H1, all observers were included (N = 397).For the model to test H2 and H3a/b, only national majority members were selected as observers (N = 203).The nominal dependent variable was the ascription (dual, only minority or only national) given to second-generation migrants, which was nested in the observers.In the models, dual ascription was treated as the reference category.Contrary to our expectation, it is not possible with multinominal logistic regression to have a three-level multilevel model (ascriptions cross-nested in observers and targets, nested in school classes).We thus deviated from the preregistration by only accounting for the nesting of ascriptions in observers.

| Descriptive results
Based on the self-identification categories, 36.20%self-identified as dual (N = 139), 8.85% as only minority (N = 34), and 54.95% as only national (N = 211).Of all second-generation migrants (those born in the Netherlands with at least one parent born abroad), 68.99% self-identified as dual (N = 89), 17.83% as only minority (N = 23) and 13.18% as only national (N = 17).Of all the ascriptions that second-generation students received (N = 1,529), 15.11% was a dual ascription, 65.66% was an only minority ascription and 19.23% was an only national ascription.
Descriptive results showed that national majority members scored significantly higher than the midpoint of the  A1 in Supporting Information).This means that on average, national majority members in our sample scored high on strength of Dutch self-identification, and ethnic and civic conception of nationhood.

| Who ascribes dual belonging?
We expected that individuals who self-identified as dual were more likely to ascribe dual belonging (H1).Figure 1 shows which ascriptions were given to second-generation peers based on the self-identification of the observer.vii Overall, only a small percentage of ascriptions were dual (15.07%), and those who self-identified as only national most often ascribed dual belonging (Figure 1).Dual identifiers gave significantly fewer dual ascriptions than those who self-identified as only national (χ 2 (1) 14.84, p < .001),but did not significantly differ in these ascriptions compared to those self-identifying as only minority (χ 2 (1) = .15,p = .698).
Results from the two-level multinominal regression also showed that dual self-identification was a positive significant predictor for ascribing only minority compared to dual belonging (B = .508,SE = .243,p = .036),and a positive but nonsignificant predictor for ascribing only national compared to dual ascriptions (B = .407,SE = .260,p = .117;see Table A2 in Supporting Information for the complete model results).Hence, individuals who selfidentified as dual were not more likely to ascribe dual belonging, rejecting our hypothesis 1.Additionally, those who self-identified as dual (χ 2 (1) = 9.790, p = .002)and only minority (χ 2 (1) = 6.460, p = .011)gave significantly more only minority ascriptions compared to those who self-identified with only the national group.Self-identified duals did not give significantly fewer only national ascriptions compared to self-identified only nationals (χ 2 (1) = .028,p = .866).

| Predictors of dual ascriptions among National Majority Members
For hypotheses 2, 3a and 3b, the sample was reduced to only include national majority members (N classes = 36, N ascriptions = 648, N observers = 203, N targets = 126).Table 1 shows the result of multinominal logistic regression models in which strength of Dutch self-identification, ethnic and civic conception were added as predictors at the F I G U R E 1 Percentage (%) of ascriptions given to second-generation peers based on observers' self-identification (N ascriptions = 1,453, N observers = 384, N targets = 137).Nselfid = Number of observers that had this self-identification, Nascr = Number of ascriptions given by observers that had this self-identification.Those who could not be classified into one of these three self-identification categories (N = 13) were omitted.
T A B L E 1 Results for two-level multinominal logistic regression models with only minority (1) and only national (1) ascriptions contrasted to dual ascriptions as reference category (0) (N ascriptions = 648, N observers = 203, N targets = 126).observer level, and gender and level of education were added as control variables (see Table A4 in Supporting Information for the null model).We expected that national majority individuals with a stronger Dutch self-identification (H2) and those who endorsed an ethnic conception of nationhood more, would be more likely to ascribe only minority, compared to dual belonging (H3a).In contrast, we expected that those who endorsed a civic conception of nationhood more, would be less likely to ascribe only minority, compared to dual belonging (H3b).Model A shows that the effects were in the hypothesized direction for predicting only minority compared to dual ascriptions, but were all non-significant (Table 1).Therefore, hypothesis 2, 3a and 3b could not be confirmed.
Although we did not hypothesize about the contrast between ascribing national versus dual, our findings showed that the strength of Dutch self-identification was positively associated with ascriptions of national instead of dual belonging (Model B, Table 1).Those who had a stronger Dutch self-identification and those with a higher level of education were significantly more likely to ascribe only national compared to dual belonging to secondgeneration classmates.

| DISCUSSION
The aim of this study was to examine which ethnic and national ascriptions (dual, only minority or only national) adolescents gave their peers with a second-generation migration background and why they did so.First, we investigated whether dual self-identification could predict ascribing dual belonging to peers.Second, we focused on whether (the strength of) Dutch self-identification, and endorsing a civic or ethnic conception of nationhood explained why national majority members did or did not ascribe second-generation migrants a dual belonging.
Our findings showed that although 69% of the second-generation migrant self-identified as dual, only a small number of dual ascriptions were given (15%).Most second-generation migrants were ascribed only minority belonging.Surprisingly, dual self-identifiers were not more likely to ascribe dual belonging.In fact, those who self-identified as only national gave significantly more dual ascriptions both compared to pupils who self-identified as dual or as only minority.Thus, incongruent with SCT (Turner, 1985), the dual category might not be readily used in ascriptions by those with a dual identity themselves.Research on social identity complexity has addressed that some individuals manage to successfully integrate multiple social groups in their mental representations (Roccas & Brewer, 2002).
However, in our sample, this ability did not seem to generalize to the social categorization of other peers.Nonetheless, research has also shown that people are more accurate in recognizing members of their own ethnic group (e.g., Vingilis-Jaremko et al., 2020).Therefore, future research, with a bigger sample, should investigate the possibility that dual identifiers more often ascribe dual identities to peers with the same ethnic background rather than to all peers with a migration background.
Furthermore, we found no support for our hypothesis that national majority members with stronger Dutch selfidentification were more likely to ascribe only minority than dual belonging to their second-generation peers.Instead, majority members with stronger Dutch self-identification were more likely to see second-generation peers as part of only their national ingroup rather than ascribing them dual belonging.This finding aligns with research suggesting that dual belonging could threaten national majority members by raising questions of group loyalty (Kunst et al., 2019).Strongly identifying national majority members might try to protect their ingroup and maintain rigid group boundaries by not recognizing dual belonging, rather than excluding those with a migration background from the national group (e.g., Gaither et al., 2016;Tajfel, 2010).This is also consistent with acculturation literature showing that national majority members generally prefer migrants to adopt the host culture (i.e., assimilate) rather than combining this with maintaining their heritage culture (i.e., integrate) (e.g., Roblain et al., 2016).
Contrary to our expectations, we found that national majority's endorsement of an ethnic or civic conception did not predict minority only ascriptions compared to dual ascriptions.This suggests that, in our sample, stronger endorsement of a civic conception did not result in more inclusive ascriptions to those with a migration background (e.g., dual), neither did stronger endorsement of an ethnic conception lead to more exclusive ascriptions (e.g., only minority).Hence, other factors might be more influential for ascriptions of belonging given to second-generation peers.
Overall, this study enhances our understanding of ascriptions of dual belonging to second-generation migrants.
Foremost, we advanced existing methodologies by also measuring dual ascriptions given by/to peers in classes (For other research with ascriptions of one group, see Boda & Néray, 2015).Our results showed that dual membership was rarely ascribed; people were more likely to categorize others into one group rather than categorizing them in more complex dual ways.In line with previous research on the experiences of identity denial of bicultural people (e.g., Cárdenas et al., 2021), second-generation students were mostly ascribed belonging to their ethnic minority group.Surprisingly, majority members most often ascribed (also) Dutch belonging, and those who more strongly identified with the Netherlands were more likely to ascribe only national belonging compared to dual belonging.The latter could be a way for majority members to maintain rigid group boundaries.Another explanation could be that targets might 'show' one ethnic/national identity more, or that the school context makes a certain social identity more salient.Research has shown that dual identifiers adjust the expression of their identity based on the person they are interacting with, and that they sometimes decide to conceal or highlight one social identity (e.g., Slootman, 2018a;Van De Weerd, 2020).Moreover, given the high ethnic diversity of our school context, the ethnic minority background of second-generation participants may have been more salient to most observers, and consequently more ascribed.

| Limitations and future research
A few limitations must be addressed.First, the generalizability of our findings is limited due to our selective sample, which depended largely on schools' and teachers' willingness to participate.Future research should include a larger sample from different schools and levels of education.Second, it is currently statistically not possible to specify cross-nested multinominal logistic regression models.By controlling only for the nestedness of the ascriptions in observers but not in targets of the ascriptions, we may underestimate the standard errors of our coefficients and, because we could not account for target level variables, we may have overestimated the effect of our observer variables.
Third, dual belonging had to be indicated by nominating two separate ethnic/national groups (e.g., Turkish and Dutch) but some participants may have hyphenated belongings (e.g., Turkish-Dutch) in mind (e.g., Verkuyten et al., 2019).This could have led to an underestimation of ascriptions of dual belonging.Future work could investigate if offering such answer options leads to more ascriptions of dual belonging.Although the current findings counter the suggestion that observers mainly picked the first answer option, future work could randomize the order to avoid potential priming or fatiguing-effects.Another suggestion would be to experiment with the wording of the question on ascriptions of belonging, as the current reference to 'culture' may carry different connotations than 'ethnicity' or 'country'.Additionally, because we find a large discrepancy between self-identification and ascribed identities, future research could further examine this mismatch and its consequences.

| Conclusion
Our findings showed that many Dutch adolescent descendants of migrants self-identify with both their ethnic minority group as well as the national group.However, dual belonging was rarely ascribed to them, neither by pupils who self-identified as national majority members, nor by those self-identifying as dual or minority member.Instead, ascriptions of only minority belonging were most common.This discrepancy between self-identification and ascriptions could negatively affect dual identifiers' psychological health and social integration.Research has shown that when individuals feel that their dual identity is not recognized, this can lead to feelings of exclusion, identity conflict (e.g., Albuja et al., 2019) and rejection of the national culture (Jasinskaja-Lahti et al., 2009).
Although we found no effects for the hypothesized relations predicting ascribing only minority compared to dual belonging, our results showed that a stronger national identification among national majority members related to a higher tendency to ascribe only national rather than dual belonging.This indicates that high identifying national majority members might try to maintain rigid group boundaries by not endorsing dual belonging, and instead prefer second-generation migrants to assimilate into the national majority group.While bicultural individuals are increasingly present in Dutch society, our findings suggest that most people are not aware of them in their environment.
On a societal level, not recognizing others' dual belonging could hinder positive developments for interethnic relations that go beyond the dichotomy of in-and outgroup.While research has shown that being aware of dual identifiers improves attitudes toward the associated minority group (Levy, 2019;Levy et al., 2023), our findings raise the question whether dual identifiers could act as 'gateway groups' to improve intergroup relations (Love & Levy, 2019).
We found that others' dual belonging is often not seen and, more specifically, seems not to be endorsed by neither the national majority group nor by the minority group.Accordingly, dual identifiers' potential to foster positive intergroup relations may be rather limited.