‘I’m Not Swedish Swedish’: Self-Appraised National and Ethnic Identification among Migrant-Descendants in Sweden

Abstract

As a country of high migration, Sweden presents an interesting case for the study of belongingness. For the children of migrants, ethnic and national identification, as well as ascriptive identity, can pose challenges to feelings of belongingness, which is an essential element for positive mental health. In this article, survey data were collected from 626 Swedes whose parents were born in the following countries: Somalia, Poland, Vietnam, and Turkey. The results show that Poles significantly felt they received more reflective appraisals of ascription than any other group. However, despite not feeling as if they were being ascribed as Swedish, most group members (regardless of ethnic origin) had high feelings of belongingness to Sweden. Overall, individuals who felt that being Swedish was important for their identity indicated the highest feelings of belongingness. Further, individuals across groups showed a positive correlation between their national identification and ethnic identification, indicating a feeling of membership to both. These results mirror previous research in Sweden where individuals’ ethnic and national identities were positively correlated. The ability to inhabit multiple identities as a member of different groups is the choice of an individual within a pluralistic society. Multiple memberships between groups need not be contradictory but rather an expression of different spheres of inhabitance.

1. Introduction

In this study, descendants of migrants from different ethnic groups in Sweden were asked about their sense of belongingness and whether their feelings about their perceived identity, or the importance of their ethnic and Swedish identity influenced this. A multitude of different factors play a role in the feelings of belongingness for an individual in society, especially for the child of a migrant. This article contributes to the ethnic identity literature, in the Swedish context, by adding reflected appraisal of ascription to the model of predicted belongingness. By doing so, the social process of understanding who is seen as belonging within Swedish society can be studied. This article will focus on the self-reporting of the feeling of belongingness to the nation of Sweden rather than belongingness to the feeling of being Swedish. Measurement in this way allows for the individuals to introspect how they fit into a broader supranational context rather than the ethnic national group (Simonsen 2016). Further, this article begins from the stand point that all of the individuals included in the analysis are Swedish, and have the right to claim Swedishness. By examining ethnic and national identity together with ascription, a more nuanced discussion about what Swedishness is and what it is becoming can be had.

2. Background

Sweden presents an interesting case to examine given its dark history of racial biology and current status as a leader in humanitarian migration. In 1912, the Swedish State Institute for Race Biology (SIRB) was founded (Ericsson 2021). Under the direction of Herman Lundborg, a known right-wing activist, until 1935, the institute focused on racial science and eugenics. His successor, the anti-fascist Gunnar Dahlberg, helped shift the focus of the organization towards medical genetics. However, racial science still remained as active area of inquiry at SIRB until at least 1960. SIRB conducted studies which served a scientific interest in the endorsement of a racial order based on the intrinsic, immutable differences between human populations (ibid.). The preservation of the idea of Sweden as a white nation, and one that has always been white (Hubinette and Lundström 2014), is in part situated in the history of racial biology and the fear of miscegenation. A number of different policies during the first half of the 1900s aimed to help maintain the purity of the Nordic or Swedish race. This included the forced sterilizations of over 60,000 individuals from 1934 to 1975, most of whom were considered to be ‘lower stock’ such as the Roma or Travellers (Ericsson 2021; Hubinette and Lundström 2014). Practices such as these mirrored the desire for a homogenization of the population (ibid.).

In what feels like a contrast to this, Sweden’s migrant population has grown significantly since the second World War, with the highest rates of migration occurring within the last decade. After the dissolution of SIRB, Sweden began gaining traction as an active voice against racism (Hubinette and Lundström 2014). Other social movements in Sweden at this time, such as the women’s movement, helped diffuse the historical focus on race (ibid.). Building off the need for more tailored policies regarding the labor migrants from previous decades, newer integration policies were pushed forward. In 1975, integration policies were formulated under a specific assumption; that those persons coming to Sweden would stay permanently. These policies underscored three main objectives; equality, partnership, and freedom of choice. Equality gave migrants rights to housing, schooling, and other basic rights as well as access to the welfare state, positioning them as similar to Swedes as possible. Partnership implied the expectation of active participation in politics and a mutual workmanship between immigrant groups and natives. Finally, freedom of choice allowed for migrants to determine whether or not they wanted to culturally integrate into Swedish society or retain their own ethnic identity (Emilsson 2016; Westin 2003). Of considerable interest for this work is the freedom of choice tenet. While the policy was overarchingly multicultural, it placed the burden of cultural integration on migrants (Westin 2003). Migrants were presumed to develop into new national minorities (Soininen 1999). However, labor migration from outside Scandinavia had all but stopped when the policy came into effect. Instead, a new type of migrant was arriving in Sweden; refugees.

Multicultural immigration policies were broken from in 1986. Groups known to have been residing in Sweden for a long time, such as the indigenous Sami and certain groups of Finns, were deemed as minorities rather than immigrant groups (Emilsson 2016). Although the social and political guarantees of previous integration bills remained largely the same, the freedom of choice tenet became more challenged. Mainly, immigrant groups should assimilate to basic norms. This policy was later criticized for placing a cultural chasm between Swedes and immigrants, which in turn reinforced social stratification and boundaries between groups. In 1998, a new policy was introduced stating that both Swedes and ‘others’ should adapt to each other, de-emphasizing differences between these populations. The policy formally demanded equal rights, opportunities, and responsibilities between all persons in Sweden, despite ethnic or cultural background (Emilsson 2016).

Although Sweden has been found to be generally more accepting of newcomers than other EU nations, since 1991 anti-migrant attitudes have increased in Sweden, coloring the lives of migrants and their descendants (Rydgren and Tyrberg 2020). In tandem with the rise of migrants, right-wing political parties have also increased (Hellström and Bevelander 2018). In 2010, the Swedish Democrats party (SD), who espouse anti-migrant and anti-Semitic rhetoric, were able to secure enough votes to qualify for parliament participation. Attitudinal research within Sweden has shown an increase in anti-migrant attitudes allowing for the election of a party such as SD (ibid.). The white majority is often seen through media representations as being the victims of violent incidents involving ‘immigrants’ or ‘foreigners’ (Hubinette and Lundström 2011), portraying migrants and their children as aggressive ‘others’.

According to the Central Statistics Bureau of Sweden (SCB), 17.6% of the population under the age of 35 is foreign born (SCB 2020). As of 2020, for individuals under the age of 35, 11.4% had two foreign-born parents and 10.3% had one parent that was foreign born (ibid.). Considering the changing demographics and history of Sweden, it is important to study how individuals of non-Swedish backgrounds feel about their society. What shapes an individual’s feeling of belongingness? Prior research suggests that this may be affected by ascribed identity from others, feelings of national identity, and the importance of continued ethnic identification with ancestral country identities, especially for so-called ‘second generation’ individuals (Verkuyten et al. 2019).

This article will begin by presenting a framework for understanding identity and belongingness, as well as previous research conducted in Sweden. Then, the data will be introduced. The data used come from the Governing Citizenship in Scandinavia (GovCit) project, which examined attitudes, political opinions, and citizenship queries in different Scandinavian countries. Results and data reported in this article will focus exclusively on Sweden and specifically persons who are officially registered as having parents born in the following nations: Poland, Somalia, Vietnam, and Turkey. The article will conclude with a discussion of the results and implications for further research.

2.1. Belongingness and Identity

Belongingness is one of our most basic needs (Maslow 1954)—it legitimizes our place in society. Personal familiarity with a system or environment, and believing oneself to be an integral part of those systems, is to feel a sense of belonging (Hagerty et al. 1992). This can feel like security or acceptance and can increase someone’s ability to ‘function well’. The question of who feels that they belong is not only for immigrants but for society as a whole, as these questions are community bound (Ali et al. 2006). As humans are social creatures who wish to engage in group life (Fiske 2018), an individual’s feeling of belongingness to others, and others’ perceptions of the individual, have been the focus of social psychology for a long time. This is unsurprising given the real benefits that connectedness to others holds. In order to reap these benefits fully, individuals both consciously and unconsciously adhere and adapt themselves social norms and roles within group structures. The molding of one’s identity in relation to groups and within groups facilitates connectivity and, hopefully, belongingness (ibid.). How an individual belongs in a society is a negotiation between the identity they present, what they are ascribed, and what is allowed by that society (Schlenker 1985).

Identity is polysemic. Used so broadly and in so many different contexts, e.g., sociology, psychology, attitudinal studies, and ethnic relations, identity has become an academic ‘buzzword’ (Verkuyten et al. 2019). Following the work of Verkuyten et al. (2019), this paper acknowledges two separate types of identity that are relevant to the overall work; group membership as a form of social identity (Tajfel and Turner 1979) and the internalized, inner structure of cultural identity. Social identity theory (SIT) describes the ‘me with we’, or how the external process of being part of a group involves the individual (Tajfel and Turner 1979; Verkuyten et al. 2019). SIT refers to an individual’s assertation of group belongingness based on their self-concept (Tajfel and Turner 1979). Additionally, SIT proposes that individuals understand groups via comparison with relevant out-group members and therefore, is a social process (ibid.). In contrast, an individual’s cultural identity is the ‘we in me’; internalized processes and reflections of cultural membership as a part of an individual’s sense of self (Verkuyten et al. 2019). Despite holding the title of being part of a certain group, it is still possible for individuals to lack a sense of inclusion or belongingness to that group. A salient example of this would be identification to the nation state in Sweden, where true membership and inclusion is subject to unattainable criteria, such as race and ancestry (Simonsen 2016).

National identity as a social identity accounts for not only the self-ascribed label of a nation but to a greater, more all encompassing feeling of belongingness and acceptance within broader society (Phinney and Devich-Navarro 1997). Further, the idea of a national identity is entrenched in images of who does and does not belong to that body politic (Anderson 1991; Simonsen 2016). Therefore, the nation becomes ‘a stable and coherent object’ through which individuals conceptualize relationships to the feeling of belonging or, conversely, exclusion (Simonsen 2019). The nation state is present in our daily lives; evident in the news we watch, the language we speak, the values and history we are taught, and in the sports and athletes we support (ibid.). Previous research has lacked an exploration of what Simonsen (2016) has called the ‘affective dimension of integration’, meaning the degree to which migrants and their descendants feel belongingness to and identity with the nation state.

Conversely, the retention of ethnic identity within a host society has been studied a considerable amount. There have been many definitions of ethnicity and ethnic identity (see Ashmore et al. 2004 for review). Identifying and belonging to ethnic groups is an important and central way that humans have chosen to define themselves (Tsai et al. 2002). Ethnicity can be essentialist, or linked to ‘ancestral heritage’ (Phinney 2005, p. 188). Ethnic identity can be defined as one’s self-label (Rumbaut 1994). However, a more complex definition stems from ‘common ancestry and the sharing of one or more of the following elements: culture, religion, language, kinship, and place of origin’ (Phinney et al. 2001, p. 496). Ethnicity and race are related concepts (Verkuyten 2018) and in the Swedish context ethnicity is often used as a proxy term for race (Osanami Törngren 2015). Theorists, such as Berry (1997), for example, have emphasized that ethnic self-identification needs a minimum of two ethnic reference groups: an in-group and an out-group. Commonly, the ethnic reference groups utilized are that of the majority ethnic group in society and the ancestral ethnic group of the individual (Noels et al. 2010). Ethnic identity is not only formulated via in-group and out-group; rather, co-ethnics are often prominent in the shaping of these identities (Verkuyten 2018). An individual within this perspective can identify anywhere along the spectrum between both reference groups, but also with neither or both (Noels et al. 2010).

Ethnic identity is not the same as one’s ascribed identity or reflected appraisal. Reflected appraisal is a psychological term meaning one’s perception of how others see them (Mead 1934; Cooley 1902). This is distinct from how the individual perceives themselves, called self-appraisals, which also differ from actual appraisals, or how others actually perceive them (Fiske 2018). Instead, reflected appraisal is how individuals perceive and experience how they are seen. Ascribed identity is the application of this same process; it is the reporting of the identity one believes others see them as (Brubaker and Cooper 2000; Jenkins 2014). Reflected appraisals endorse and restrict identity, because there might be a gap between what you claim and what you are ascribed. Indeed, ethnic identity is the interaction between reflected appraisals and chosen identity (Verkuyten 2018). An individual can experience a discrepancy between their chosen identification and the one ascribed to them, especially if there are limited ethnic options for them to claim (Waters 1990).

In this paper, I conceptualize the reflection of whether or not individuals perceive that they are seen as belonging to an ethnic group as the reflected appraisal of ascription. This process could be represented in the question, do you feel that people look at you as Swedish? Research on ethnic identification has often been concerned with the ‘situated variable experience’ (Noels et al. 2010) or the ascription of identity and the social context this occurs in. This nesting of identification within the broader milieu presents different social categories for the individual to inhabit, which in turn also affect an individual’s self-identification. Reflected appraisals and ascription need not be the greater influencer of ethnic identification, or feelings of inclusion or exclusion. Certainly, individuals orient themselves within the situations they are in, as they see fit or as they are able. This is because identifications can be situation driven, meaning that an individual might find it needed to embody a specific identification in different situations or feel pressure from others to do so.

The language of identity is an expression of social representations (Hall 1997). Identity labels are applications of the social identities we feel exist in the world around us. As individuals, we have mental representations of things, people, and even places which are conceptualized via their relationship to other things. Language helps us as individuals articulate these meanings based our own interpretations of how different social representations relate to one another and ourselves. We use the language of social representations to confirm our membership to the groups in which we inhabit and those we do not. Additionally, we confirm and deny others’ identity by assigning identity labels to them (ibid.). In this article, participants were able to indicate the importance of ethnic and national identities to themselves, which is a reflection of how they feel that they socially represent these identities. In contrast, data within this work also explore how these individuals feel that others assign the label of ‘Swedishness’ to them. In other words, the reflected appraisal of ascription is an operationalization of the social representation of what it means to be perceived as ‘Swedish’.

2.2. Studies of Identity in Sweden

The dichotomy of being considered Swedish or not can be partially conceptualized in two considerations; the idea of Swedish nativity and the administrative practice of categorization carried out by the Central Statistics Bureau (SCB). Neither racial or ethnic identity is recorded in Sweden, therefore nativity becomes the categorical mechanism (SCB 2020). This can become problematic for individuals when a gap between ethnic and national identification emerges. Within official population statistics, individuals born in Sweden whose parents were born abroad are statistically classified as ‘having a foreign background’ (SCB 2020). However, the administrative distinction of ‘having a foreign background’ disappears with those whom have two parents born in Sweden, such as third generation individuals (ibid.). This can create further concept blurriness; individuals might not necessarily be considered to be Swedish in their daily lives, but are administratively seen as Swedish.

Countries such as the United States, Australia, and Canada allow for the presence of hyphenated identities (i.e., Indian-American). This type of labeling allows for an individual to express relations to the nation state as well as their ethnic group (Simonsen 2016). In Western Europe, hyphenated identities can be seen as a contrast to national belonging, rather than inclusion to both groups (Verkuyten and Martinovic 2012). Identifications such as these become more tricky and less common in the Swedish context (Behtoui 2019), other than a few notable exceptions, such as the Afro Swedes. Often in Sweden, it has been found that hyphenated identities are not validated by wider society. Instead, individuals could be categorized as ‘invandrare’ or ‘immigrant’ which extends into migrants’ descendants, creating an either/or belongingness dichotomy. Previous research has claimed that persons identified as ‘invandrare’ can be seen as not assimilated enough, or portraying the antithesis of ‘Swedishness’ (ibid.). Other researchers have shown that the word ‘immigrant’ connotes ‘a representation of social problems’ (Trondman 2006, p. 433). Therefore, being an ‘invandrare’ becomes a social category of its own. Despite these studies, there is a lack of research done on how ascription interacts with the claiming of Swedish identity, and whether or not feelings of belongingness are affected by this.

Moinian (2009) conducted qualitative research on individuals with foreign backgrounds between the ages of 12 and 16 in Sweden, in regard to how they self-identify. Here, it was found that most participants expressed a fluidity that was context dependent. Behtoui (2019) also studied adolescents with migrant backgrounds in Sweden and found that individuals’ identities are fluid yet contextualized in different societal spheres, such as with family, friends, or at school. However, studies have also shown that individuals feel that they are ascribed an identification which felt non-negotiable (e.g., Moinian 2009). Perhaps as a consequence of such ascription, other studies have found that descendants of migrants choose the identification of invandrare and use this as a monolith (e.g., Runfors 2016). This categorization allows for an individual to bypass ethnic or racial labeling, or even hyphenated identity, and arises from being seen as othered (ibid.). This can partially be explained by Sweden’s multiculturalist approach which allowed immigrants to retain their home languages and culture (Westin 2003). This strategy, however, has previously been accredited with segregating the population in ‘Swedish’ vs. ‘non-Swedish’ as migrant groups and their descendants are never seen as fully integrated (Scuzzarello 2015).

2.3. Somalians, Turks, Vietnamese, and Poles in Sweden

For this study, individuals whose parents were born in Somalia, Poland, Turkey, and Vietnam will be comparatively examined. In order to glean a better understanding and context for this, each group will be briefly presented below. However, it is important to clarify that the migration patterns mentioned below may not be applicable to all the participants included in this study. Rather, this is meant to serve as a framework for the most notable migration patterns for each group.

Somalian migration to Sweden began in the late 1980s and still continues today. Previous studies have shown that Somalian is spoken in home environments, which aids in creating stronger ethnic identity (Palm et al. 2019). First-generation Somalian youth in Sweden have been found to report feelings of social exclusion and difficulties in inclusion. Many of these issues were linked to the direct experience of migration; however, similar issues were found in other studies comparing Somalis in the UK and Sweden (Scuzzarello and Carlson 2019). Several studies have reinforced previous findings that hyphenated identities are less actualized in Sweden. For example, in Osman et al. (2020), participants corrected an earlier version of the article draft where they were referred to as ‘Somali-Swedish adolescents’, affirming that they would rather be referred to as ‘Somali’ despite living in Sweden. Other previous research has highlighted that for individuals with Somali backgrounds in Sweden, identification becomes binary; one is a Somali in Sweden and a Swede in Somalia (Scuzzarello and Carlson 2019). This is common finding within studies of ethnic identity (see Verkuyten 2018 for a review).

Many Poles within Sweden came after the EU expansion in 2004 (Scuzzarello 2015) or as guest laborers in the 1990s. Within Swedish discourse, Poles are often seen as more integrated into society than other groups due to participation in the workforce and low levels of criminality. As previous research has pointed out, perceptions of integration about different migrant groups do not always map to how migrants or their descendants’ actually feel within society (Scuzzarello 2015). Investigations into the feeling of belongingness within the Polish community in Sweden has shown that Poles mirror other groups’ feelings of exclusion from Swedish society (Scuzzarello 2015; Shmulyar Gréen et al. 2021). Ethnic identity in Poles tends to be centered around religious affiliation, with churches and religious associations taking on a role as social hubs as well as places of worship (Shmulyar Gréen et al. 2021). This has been found to be particularly true for more newly arrived Polish migrants (ibid.). Sweden is a notoriously secular country and identities which emphasize religion can be seen as an out-group since religiosity is often viewed as regressive or unmodern (Sjöborg 2015).

Turks within Sweden today are either descendants of labor migrants from the 1960s, some of whom were Kurds, or Assyrian refugees from the mid 1970s (Vedder and Virta 2005). Due to nativity and not ethnicity, religion, or language being recorded by the Swedish state, all three of these groups would be identified as Turks in official statistics (Westin 2003). For clarity in this work, I will also refer to them as Turks. Like many other groups in Sweden, Turks have been othered. Often seen as poor scholastic achievers and prone to behavioral problems within schools, Turkish men are particularly stereotyped. Previous research has indicated that identifying as Swedish negatively predicts positive adjustment for second-generation Turks (ibid.). This means that higher national identification does not necessarily translate into a higher feeling of belongingness in society.

Vietnamese migrants came to Sweden at the end of 1975 as political refugees (Sam and Virta 2001). Previous studies have found that when comparing self esteem scores between descendants of migrants from Chile, Turkey, and Vietnam and natives in Sweden and Norway, the Vietnamese group scored the lowest in both countries. Researchers have theorized that this is due to ‘their Confucian child upbringing where modesty may be deemed as a virtue’ (ibid., p. 375). However, other studies have shown that Vietnamese students in Stockholm reported higher levels of school adjustment than other migrants and natives (Virta and Westin 1999). These studies are quite old and to the best of my knowledge, the Vietnamese within Sweden are not studied as frequently as other groups, such as, for example, Somalians.

A uniting finding is that individuals from all four groups report feelings of exclusion in Sweden. In particular, studies on these groups point to how these individuals feel shut out of Swedishness (see Scuzzarello and Carlson 2019; Odmalm 2005) and others (see Hubinette and Lundström 2014) have suggested that this could be due the strong ties between national identification and ethnicity (i.e., whiteness). Further, ‘diversity’ in Sweden has been dichotomized into ‘invandrare’ (immigrants) and natives (Scuzzarello 2015; Behtoui 2019). Ethnic homogeneity is still a primary component for national belonging (Scuzzarello 2015), despite inclusive welfare state institutions and multiculturalist policies. This idea of ethnic versus civic national identity in Sweden could be said to have its roots in the racial biological practices at SIRB, where the idea of a pure, ethnically Swedish ‘race’ was promoted and studied for several decades.

How strongly does the reflected appraisal of ascription play a role in feeling a sense of belongingness? Do individuals from different ethnic groups in Sweden feel that being Swedish is important to their identity? For the descendants of migrants, identity is a complex and multifaceted process. National and ethnic identity may not go hand-in-hand. This article seeks to examine how individuals process their ascribed Swedishness, feelings of belongingness, and their identifications, both national and ethnic. By examining these different variables together, a more contextualized, richer description of different ethnic groups in Sweden can be conceptualized. This article is interested in the combination of these factors together, in order to paint a clearer picture of how much or to what degree they are of importance. Therefore, this article aims to explicate the following hypothesis:

• As descendants of migrants are not a monolith, groups with different ethnic backgrounds will significantly differ in all four variables of interest; ascription, importance of ethnic and national identity, and belongingness;
• Ascription will make an independent, statistically significant contribution to feelings of belongingness when controlling for the association of ethnic and national identity at the same time.

3. Materials and Methods

The data used in this article come from the GovCit project, which aimed to measure citizenship, participation, and belonging within Scandinavia of young adults of the age 20–35. The GovCit project was carried out in Denmark, Norway, and Sweden (Erdal et al. 2019). The data for Sweden was collected throughout 2018. The original survey included 69 questions which related to different facets of social trust, political participation, citizenship, and other integration concepts. All items were self-reported, and participants were asked to describe experiences and feelings regarding different inter-personal and societal issues. Care was taken during the development of the questions so that they were clear, direct, and sensitive given the nature of the survey topics (ibid.). All questions, in the Swedish sample, were in Swedish. For a review of the project in total, please see Erdal et al. (2019).

Data was collected via random sampling, both online and by telephone, and representativeness of the population was prioritized (Erdal et al. 2019). This means that the ethnic groups used within this article constitute some of the largest within Sweden (SCB 2020). Register data showing participants with two foreign-born parents from the central statistics bureau in Sweden (SCB) was used to create the groups presented in this article. The groups utilized within the analysis are consistent with the methods used by the Official Statistics Bureau in Sweden and therefore present a typical construction of how these individuals are administratively categorized.

3.1. Demographic Information

The sample (n = 626) consisted of four groups; Somalians (n = 103), Polish (n = 233), Turkish (n = 120), and Vietnamese (n = 170). All individuals included in the analysis were considered to be part of the so-called second generation. The sample was reasonably spilt between men (n = 425) and women (n = 474). The average age across all groups was 27.54 (sd = 4.78). Of the total respondents, 95.5% had citizenship within Sweden.

3.2. Variables

The GovCit survey consisted of 69 items relating to citizenship, political participation, social trust, and other concepts. The survey questions used in this article utilized a Likert scale of 1–10, with ten representing total agreement with the statement. Each item was determined to represent a different conceptual element of analysis. Therefore, the items were not considered for scale or index creation. Items of interest for this work are listed below in italics with their operationalizations following:

• I feel that most people look at me as Swedish. This question asked the individual if they perceived whether or not others understood them to be Swedish. It is not only asking about the individual’s reflected appraisal, but the reflected appraisal of ascription as I defined earlier. For simplicity, however, I will use the term ascription throughout the rest of this paper;
• How much do you feel you belong to Sweden? Interestingly, this item asks about belonging to the nation state rather than within it. It does not ask participants if they feel included in society, but rather if they feel that they belong to the nation. Similar to Simonsen (2016), despite the subjectivity of the question, it was operationalized to represent belonginess because it ‘captures the connotations of ties, relatedness and connectivity that are central to the theoretical belonging concept’ (p. 1159). I will refer to this variable as belongingness for the rest of the work;
• How important is it to be a part of your ethnic in-group for your identity? This item is fairly straightforward in representing the concept of the importance of ethnic in-group membership. Throughout the rest of the text, importance of ethnic in-group membership, ethnic identification, and ethnic identity will be used interchangeably;
• How important is it for your identity to be Swedish? National identity can mean a lot of things; patriotism, national pride, or nationalism to name a few. This question was seen to assess the importance of integrating the label of ‘Swede’ into an individual’s social identity, rather than the feelings associated with that labeling. A person denied the ascription of being ‘Swedish’ could still find it important for their own self labeling to be considered ‘Swedish’. Therefore, this item represents an internalized process of national identification as a social identity and will be referred to as such for the rest of the paper.

In addition, groups created in the analysis were done so via register data. This means that official statistics from the Central Statistics Bureau (SCB) indicated where a participant’s parents had been born, even though the participants themselves were born in Sweden. In order to increase the fluidity in reporting results, participants included in this analysis will be identified by these countries. Therefore, persons whose parents were born in Poland will be referred to as ‘Poles’, for example.

4. Results

4.1. Descriptive Results

Each variable was examined per group for descriptive results (see

Table 1. Descriptive table for the four variables, per subgroup.

	Ascription	Belongingness	Ethnic Identity	National Identity
Somalian (n)	103	100	100	99
Mean	2.54	6.15	7.59	6.77
SD	3	3.06	2.63	3.26
SE of M	0.30	0.31	0.26	0.33
Poland (n)	229	232	232	230
Mean	8.01	7.94	4.81	5.81
SD	2.34	2.35	3.32	3.46
SE of M	0.15	0.15	0.22	0.23
Turkey (n)	117	118	118	117
Mean	3.92	7.96	6.39	7.37
SD	3.49	2.29	3.6	2.93
SE of M	0.32	0.21	0.33	0.27
Vietnam (n)	169	169	167	167
Mean	5.14	8.05	4.98	6.52
SD	2.96	2.06	3.61	3.20
SE of M	0.23	0.08	0.12	0.11

below). The highest mean score for ascription was 8.01 (N = 229) for the Polish group. The lowest was for Somalians (m = 2.54). Belongingness had a small range of scores. Vietnamese (m = 8.05) reported the highest scores of belongingness, whereas Somalians (m = 6.15) felt the lowest levels of belongingness.

Somalis had the highest mean scores (m = 7.59) for ethnic identification. Turks had high scores as well (Turks = 6.39). Interestingly, Vietnamese and Poles reported the lowest scores (m = 4.98; m = 4.81). Although Turks had the highest national identification scores (m = 7.37), all other groups except for Poles (m = 5.81) scored similarly (see Table 1 and Figure 1 below). Therefore, the range of scores was not very large.

In order to assess the relations between the variables, Pearson’s correlations were run on all variables, for all subgroups (see

Table 2. Pearson’s correlations for all variables per subgroup.

Group		Ascription	Belongingness	Ethnic ID
Somalia	Belongingness	0.405 **
	Ethnic ID	0.055	0.071
	National ID	0.317 **	0.500 **	0.377 **
Poland	Belongingness	0.280 **
	Ethnic ID	−0.068	0.012
	National ID	0.067	0.397 **	0.361 **
Turkey	Belongingness	0.380 **
	Ethnic ID	−0.116	−0.008
	National ID	0.196 *	0.387 **	0.328 **
Vietnam	Belongingness	0.384 **
	Ethnic ID	0.129	0.017
	National ID	0.051	0.339 **	0.319 **

** = 0.01 significance level, * = 0.05 significance level.

). Ascription and national identification related positively to belongingness for all groups. The ethnic and national identifications were also positively related. However, ascription was only related to national identifications for Turks and Somalis.

4.2. Hypothesis One

A one-way between groups multivariate analysis of variance (MANOVA) was performed to investigate group differences between each of the four variables. MANOVAs are beneficial as they allow for univariate and post-hoc results to be examined as well. There was a statistically significant difference between all groups in the combined dependent variables: F (4, 596) = 31.53, p = 0.000; Wilks’ Lambda = 0.57; partial eta squared = 0.173. This result shows a small effect size. Univariate results per variable were also examined. In order to reduce the chance of a type one error, a Bonferroni adjustment was applied and a new alpha level of 0.0125 was used. Even with this adjustment, all dependent variables were found to have significant differences: ascription, F (4, 599) = 18.94, p = 0.000, partial eta squared = 0.353; ethnic identification, F (4, 599) = 8.5, p = 0.000, partial eta squared = 0.090; national identification, F (4, 599) = 2.9, p = 0.001, partial eta squared = 0.028; and belongingness, F (4, 599) = 9.23, p = 0.000, partial eta squared = 0.07. Effect sizes obtained in the individual ANOVAs show the magnitude of difference between groups in each of the variables separately. National identification had a small effect size, whereas belongingness and ethnic identification were found to have moderate ones. However, ascription had a large effect size (partial eta squared = 0.291). The results found warranted post-hoc analyses, which were compared per variable.

Somalians were found to have a significantly lower feelings of belongingness and this was found to be in comparison to all other groups. Post-hoc comparisons in ethnic identity showed a few differences between groups. These findings are summarized in

Table 3. Post-hoc comparisons for ethnic identity.

Parent’s Country of Birth		Mean Difference	Std. Error	Sig.
Somalia	Poland	2.76	0.41	0.00
	Vietnam	2.62	0.43	0.00
Poland	Turkey	−1.65	0.39	0.00
Turkey	Vietnam	1.51	0.41	0.01

below:

National identification was also subjected to post-hoc comparison. However, with the Bonferroni adjustment, there was only one significant group difference found. This was between Poles and Turks, with Turks indicating a higher importance of national identity to their personalities.

However, many groups were found to have significant differences in feelings of ascription. Namely, Poles differed significantly to all other groups. See

Table 4. Post-hoc comparisons for the variable Ascription.

Parent’s Country of Birth		Mean Difference	Std. Error	Sig.
Poland	Turkey	4.08	0.33	0.00
	Vietnam	2.88	0.29	0.00
	Somalia	5.53	0.34	0.00
Somalia	Turkey	−1.44	0.39	0.01
	Vietnam	−2.65	0.36	0.00
Turkey	Vietnam	−1.21	0.34	0.03

below:

4.3. Hypothesis Two

Hierarchical multiple regression was used to assess the ability of ascription, ethnic identity, and national identity to predict belongingness. In order to best glean the independent contribution of ascription in the model, this variable was entered in a second step. As my hypotheses concerned how each different group formulated their own belongingness, rather than predicting group membership, I ran four separate hierarchical multiple regressions, one for each group. In addition to this, I ran a separate analysis on the entire population, in order to have a point of comparison for the different groups. All regressions were tested for multicollinearity, normality, and outliers, none of which were concerns. See

Table 5. Results from multiple regressions on belongingness per group, with R² change from ascription.

	Ethnic	National	Ascription	Adj. R2	R2 Change
Somalia 1	−0.137	0.551 **		0.250 **
2	−0.117	0.459 **	0.266 *	0.307 *	0.063 *
Poland 1	−0.151 *	0.452 **		0.170 **
2	−0.125 *	0.426 **	0.243 **	0.225 **	0.058 **
Turkey 1	−0.152	0.437 **		0.171 **
2	0.091	0.359 **	0.299 **	0.233 **	0.083 **
Vietnam 1	−0.101	0.371 **		0.113 **
2	−0.149 *	0.367 **	0.384 **	0.269 **	0.145 **

** = 0.05 significance level and * = 0.10 level.

below for a summary of findings:

One of the most striking results is that ethnic identity did not make a significantly unique contribution to explaining belongingness, other than for Poles and the final model for Vietnamese. For these groups, the beta values reached a 0.05 significance level, yet were still the lowest unique contributors to the models. For Somalians, Poles, and Turks, national identification made a higher, and statistically significant, unique contribution to explaining belongingness in the final models than either ethnic identification or ascription. For the Vietnamese group, although ascription made the strongest significantly unique contribution, this was comparable to national identification. Overall, ascription was a powerful contributor for all models. For Poles, ascription explained an additional 5.8% of the variance in belongingness, the lowest for all groups. Vietnamese had high results, with ascription explaining 14.5%.

Table 6. Change from ascription.

	Total Population
	1	2
Ethnic	−0.209 **	−0.146 **
National	0.444 **	0.416 **
Ascription		0.343 **
Adj. R2	0.172 **	0.285 **
R2 Change		0.114 **

** = 0.05 significance level.

below shows the results for the same modeling over the whole population.

For the entire population, all variables within the model made unique, significant contributions, with national identification contributing the most in both model one and two. The final model was significant, with ascription explaining an additional 11.4% of the variance in belongingness. This result is similar to ascription’s R² change in the individual regressions, as well.

5. Discussion

All individuals possess multiple, different identities. The study of identity within migrants specifically and their descendants can sometimes underlie an assumption of questioning loyalties between the ethnic and national group, or even the compatibility between these (Verkuyten et al. 2019). This article sought to examine the interconnection between the importance of ethnic and national identification, and reflected appraisals of ascription on feelings of belongingness in different ‘second generation’ migrant groups in Sweden. Overall, the results show that all of the four different groups present in the analysis felt that they more than somewhat belonged to Sweden. Somalians were the only group found to have significantly different belongingness scores, but still indicated a mean of 6.15, showing a relatively high score. As SIT presupposes that individuals orient themselves via comparison, this result could indicate the presence of multiple, different out-groups within society that are accepted. Given that diversity in Sweden can be thought of in terms of Swedes and ‘immigrants’, the amalgamation of all migrant groups into one ‘non-ethnically Swedish’ entity could provide a sense of comradery. These individuals might not be considered ethnically Swedish or feel that they belong to the label of ‘Swedishness’ but perhaps they do feel that they belong within the nation of Sweden. This could partially be due to the exclusivity of the ethno-national ties; only ethnically Swedish persons can be considered Swedish. This result adds to the literature a possible mechanism out-groups use to cope within exclusive societies.

The importance of ethnic identification was found to have some group differences. Here, Somalians were found to have significantly higher ethnic identity scores than Poles and Vietnamese. Additionally, Poles and Vietnamese reported the lowest scores for the importance of ethnic identity and because of this, there was a significant difference found between these groups and Turks. Although a one-way analysis of variance showed that the groups differed significantly on the importance of national identification, the effect size was not large. Post-hoc comparisons showed that only Poles and Turks significantly differed in this variable, with Poles reporting higher scores.

Poles were found to have reported significantly higher ascriptions than all other groups. Polish persons seem to occupy a similar sphere of inhabitance to Swedes; they experience a higher degree of reflected appraisals as Swedes while other groups do not experience this to the same extent. Phenotypically, Poles would be the most likely to possess stereotypical Swedish traits, although other inclusion factors such as a Swedish name might be missing (Behtoui 2019). Examining the variance between groups, Turks and Vietnamese indicated similar levels of ascription to one another. This result can also point towards the invandrare monolith (Runfors 2016), where these groups are considered to be most similar to one another and seen to be the same type of ‘otherness’. Somalians were significantly less ascribed than all other groups. Results such as these indicate a hierarchy of whom is most likely to experience ascription; in the top tier, those who are closer to whiteness, and on the bottom, those whom are farthest away from it. This is reflected in the results found in this study; Poles reported the highest rates of ascription, then Turks and Vietnamese, and finally, Somalians. This could be in part due to the entanglement between whiteness and Swedishness, which would suggest that a non-white person would be less likely to be ascribed as a Swede (Hubinette and Lundström 2014).

When examined together, these results signal that high feelings of belongingness can still be felt by the individual even with a lack of ascription. In other words, whether you believe that you are ascribed or not need not be a barrier to asserting Swedish identity. This is in line with previous research showing that the strict dichotomy between ethnic and other types of identifications for migrants and their decedent’s needs to be extended (van de Vijver et al. 2015). Results such as these can also point to the complexity of environments where identifications are now developing for these individuals (ibid.). As Sweden becomes more diverse, there is a need for more nuanced language around different types of identities, such as hyphenated ones. Although individuals may still be ascribed as ‘non-Swedish’ in the future, national identification may continue to be important for their self-concept.

The relationships between the importance of national identification and ethnic identification were explored by using Pearson’s correlation coefficient for each sub-group. There was a medium, positive correlation found for all. This means that for all groups, greater feelings of national identification are associated with greater feelings of ethnic identification. This could mean that those persons who see an importance in affixing a label to their identification do so across a variety of different social categories. Ascription was only found to have a positive relation to national identification for Turks and Somalians. All groups were also found to have positive correlations between ascription and the importance of national identification to belongingness. This indicates that the more an individual was ascribed to be Swedish and the more important Swedish identification was for their identity, the more they felt that they belonged in Sweden. Interestingly, there was no relation between ascription and the importance of ethnic identification or between ethnic identity and belongingness. Additionally, regression analyses showed that ethnic identification did not add to the prediction of feelings of belongingness. Ascription did add significantly for all group models, explaining up to 14.5% of the variance in belongingness for Vietnamese. However, the strongest predictor for most groups was national identification.

Previous research has indeed shown a connection between strong ethnic identity and lower levels of integration or even acculturation (Berry 1997). Additionally, migrants and their descendants have been found to ethnically self-identify, in lieu of national or majority identification despite being culturally adjusted to the norms and values of the host society (Verkuyten et al. 2019). Perhaps due to the restrictiveness of ethno-Swedishness, different ethnic groups in Sweden have to maintain other ethnic identities. The connection between national identification and belongingness within society has also been noted by previous research (Phinney and Devich-Navarro 1997). It is unsurprising that individuals whom felt that being Swedish was important to their identity also felt that they belonged within that same supra-nation context. For these individuals, the ‘we in me’ or the internalized, inner structure of cultural identification could be both Swedish identification as well as ethnic identification. Rather than just holding a title of membership, or the ‘me in we’, individuals here found Swedishness to be important to their identity despite a lack of ascription.

Descendants of migrants are often seen as living in the middle, wedged between their parent’s heritage and their current societies (cf. Scuzzarello and Carlson 2019). Therefore, these individuals come to represent both the blooming heterogenous societies they inhabit and their ancestral backgrounds. These identities can be sometimes in conflict with one another. The social representation of a migrant in the Swedish context is often racialized (see Hubinette and Lundström 2014). There is often space, then, for a civic identity, one that is politically involved and has knowledge of laws and institutions (Scuzzarello 2015). This civic identity stands in contrast to the ethno-national identity of Swedishness, one which is rooted in whiteness (Hubinette and Lundström 2014). Belongingness to the nation then is comprised of both these identities. Denial of ethno-national identity in Sweden is operationalized through the usage of the term ‘invandrare’ or immigrant. This can be seen in this work, where Poles were far more likely to be ascribed as Swedish than other groups in the analysis.

When we define the other, we also define ourselves. When we ask, does this person belong? We are also asking what does it mean to belong to ‘us’? There are always inherent disagreements about the distinction between the in-group and the out-group as well as the definitions of certain labels, which invariably change depending on who is asked (Simonsen 2019). Individuals possess many different identities, as well as different identities in different contexts. Specifically, immigrants and their children can often face the complexity of inhabiting multiple in-groups. Identity is the negotiation between how an individual perceives themselves versus how they are perceived by others. For immigrants and their children, the so-called second generation, this interaction involves how an individual comprehends their association with their ethnic and national groups and how they feel about those groupings (Verkuyten et al. 2019). This involves both the titles of group membership that they possess as well as more indelible feelings of associateship. Additionally, how other co-ethnics and members of the supranational majority group view these individuals, as well as the interplay between this and other social identities, such as religious affiliation, also impact the development of multiple identities (Verkuyten et al. 2019; Verkuyten 2018).

6. Conclusions

This article aimed to examine specific groups in Sweden and their interconnectivity between ethnic and national identification, ascription, and belongingness. It is important to ask how tied and connected to the nation state different members of society are, as this relates to individuals’ expectations of what the nation should provide for them as citizens. For example, belongingness to the nation state has been linked to greater political participation (Scuzzarello 2015). As well, integration and discrimination policies would do well to incorporate the lived experience of the non-ethnic majority group. Further, different groups within societies deserve the opportunity to be heard and understood through their own words and experiences, rather than via the lens of the majority. These findings reveal that under-represented, non-majority individuals feel that being Swedish is central to their identity, specifically in their sense of belongingness and national identity.

Although participants were not asked how they ethnically self-identify, they were able to indicate the importance of their ethnic and national identification. Similar to Behtoui (2019), future research could explore more specifically what specific factors increase belongingness and ethnic and national identification. This article, in part, reproduces the same categories it seeks to critically examine by creating groups of individuals based on their parent’s countries of origin. Whether or not participants ethnically identify with the country they are meant to represent in the data cannot be said with certainty. Future studies could ask participants to self-report their identification. This would be similar to Behtoui (2019), who asked participants to indicate on a 1–5 Likert scale if they identified as Swedish, hyphenated Swedish, or as an ethnic or religious identity. Additionally, there is a high degree of subjectivity when measuring both ethnic and national identity. What it means for one group to have high or strong ethnic identification might be the same for another (Tsai et al. 2002). Therefore, these factors would be better explored via long answer format or interviews.

Future research should focus on the ever-evolving nature of identifications felt by individuals. Self-reporting was used in this article and variables were represented by singular survey items which utilized Likert scales. The subjects discussed in this article are complex and singular survey items cannot account for all these complexities. Therefore, interviews and/or more complex multifaceted validated measures of ethnicity, identity, and belonging should be utilized to gain a more holistic picture of how individuals negotiate different facets of their own identification and how this in turn affects feelings of belongingness. As these results also show, the relation between national and ethnic identity, ascription, and belongingness is not linear. Further, national and ethnic identification need not be opposing forces, claiming or pulling the individual in one direction or the other.

Multiple identifications across ethnic and national identities is a reality for many migrants’ descendants. These activations or switches can be temporally based; over the lifetime they can change. The malleability of identification and its ties to culture, ethnicity, race, and nation-state remain complex. The ways in which individuals within societies feel connected and situated should be focused on, outside of binary models which dichotomize ethnic and national identities, which oversimplify. National belongingness can exist for individuals who are not ascribed majority identity status. Additionally, national and ethnic identification need not be a one-to-one relation; rather, these can create a Venn diagram of inter-related facets (van de Vijver et al. 2015).