Understanding Intergroup Relations in Childhood and Adolescence

Abstract

There are various theoretical approaches for understanding intergroup biases among children and adolescents. This article focuses on the social identity approach and argues that existing research will benefit by more fully considering the implications of this approach for examining intergroup relations among youngsters. These implications include (a) the importance of self-categorization, (b) the role of self-stereotyping and group identification, (c) the relevance of shared understandings and developing ingroup consensus, and (d) the importance of coordinated action for positive and negative intergroup relations. These implications of the social identity approach suggest several avenues for investigating children’s and adolescents’ intergroup relations that have not been fully appreciated in the existing literature. However, there are also limitations to the social identity approach for the developmental understanding and some of these are discussed.

“many of the ‘individual’ theories start from general descriptions of psychological processes which are assumed to operate in individuals in a way which is independent of the effects of social interaction and social context. …In contrast, ‘social psychological’ theories start from individuals in groups rather than individuals tout court. …they stress the need to take into account the fact that inter-group behaviour is displayed in situations in which we are not dealing with random collections of individuals who somehow come to act in unison because they all happen to be in a similar psychological state” (Tajfel, 1978, p. 40).

Children form conceptually rich social categories, define themselves and others in group terms, and their group memberships become important sources of influence on their attitudes and behavior (Liberman et al., 2017; Rhodes & Baron, 2019). The way in which children perceive and evaluate the social world does not only depend on their personal dispositions, abilities and desires but also on their understanding of the groups to which they belong. Children act as individual persons but also as group members whereby shared perceptions, evaluations, and goals become self-relevant and group differentiations are made. Intergroup distinctions between “us” (ingroup) and “them” (outgroup) are common and often relevant and functional, but can also lead to biases in which the ingroup is preferred over the outgroup and outgroup members are sometimes disliked, hatred, and rejected. However, these negative intergroup reactions are not generic products of group distinctions and also not the result of the aggregation of individual psychologies. Tajfel, who initiated the social identity approach, argues in the quote above that intergroup relations do not involve a situation in which a random collection of individuals “somehow come to act in unison because they all happen to be in a similar psychological state.” Rather, intergroup relations involve shared understandings about who “we” are and the related ways in which people think and feel about themselves as members of a social group.

The aim of the current paper is to demonstrate that the social identity approach, which includes social identity theory and self-categorization theory (Reicher et al., 2010; Tajfel & Turner, 1979; Turner et al., 1987), offers a broad and coherent theoretical framework for understanding and examining intergroup relations among children and adolescents (see also Verkuyten, 2021). This social psychological approach is widely adopted in research on children’s intergroup biases but is typically only used to predict that children will show ingroup favoritism and outgroup dislike because of their need for a positive group identity. However, the social identity approach has much more to offer because it draws attention to questions and processes that are often overlooked in research on intergroup biases among children and adolescents. For example, the approach makes it possible to develop connections with children’s historical and political understanding of society and the social developmental literature more generally.

Furthermore, although the social identity approach is rooted in social psychology, developmental social psychologists have provided developmental specifications to this approach (e.g., Killen & Rutland, 2011; Verkuyten, 2016). A prominent example is the social identity development model (Nesdale, 2017). According to this model, very young children become aware of socially relevant group distinctions and begin to categorize themselves and others in social groups. They develop an understanding of which groups are better off and more highly regarded than others and gradually develop a sense of emotional belonging and identification with their group. Initially, children’s main interest is in their meaningful ingroups and they form preferences for these groups. However, ingroup preference can turn into a focus on outgroups and in outgroup negativity, for example, when children feel threatened or live in a context that fosters social division, promotes competition, or communicates negative norms about outgroups.

In this article, I will not discuss the social identity development model and the related empirical research (see Nesdale, 2017), but rather go beyond this model by using aspects of the social identity approach for discussing several key processes involved in children’s and adolescents’ intergroup relations. The social identity approach suggests a number of important avenues for investigating these relations, but also has limitations when applied to the developmental context (see also Barrett & Davis, 2008; Bennett & Sani, 2004).

I will first give a short overview of the main points of the social identity approach which. Specifically, I will focus on three features of this approach that are discussed in various sections of the paper and that are particularly useful for examining children’s intergroup relations. This is followed by a discussion of the importance of social categorization and self-categorization. Then I will elaborate on the process of self-stereotyping and the role of group identification and group identity content for understanding intergroup bias. Subsequently I will focus on group identity as a basis of shared evaluations of outgroups and for developing consensual views. This is followed by a discussing of the importance of group identity for children and adolescents acting together in discriminating against outgroup members, but also for trying to change the social world in the direction of social justice. In the final section, I will discuss two limitations of the social identity approach for understanding the development of intergroup biases.

The Social Identity Approach

The social identity approach emphasizes that the way in which psychological processes play out is dependent upon how the social world is structured. The focus is on addressing how psychological processes interact with social, cultural, and political processes in the explanation of people’s attitudes and behaviors. Specifically, the interest is in the ways in which people define themselves and others as members of social categories and groups. At least three features of this approach are useful for understanding children’s intergroup relations and I will discuss and developed these three further in the various sections of the paper.

First, the social identity approach focuses on the processes involved in making categorical distinctions with the related self-understandings. Social categorization provides structure and order to the social world and forms the basis for developing an understanding of oneself as a group member (“who and what am I”). Already infants are motivated to understand the categorial distinctions that are relevant in their social environment (e.g., language, accent, gender, and race), experience that they are categorized by others, and learn to categorize themselves (see Rhodes & Baron, 2019). For the social identity approach, an understanding of group processes starts with the cognitive act of self-categorization. Self-categorization implies that children define themselves as a group member, making the normative group understandings self-relevant. Self-categorization (e.g., “I am a member of a particular ethnic group”) involves a process of self-stereotyping in which one starts to understand oneself in terms of the group norms, beliefs, and values (“my ethnic group values tradition and respect for older generations”) associated with the particular group identity (as an ethnic group member “I value traditions and treat older generations with respect”).

Second, the cognitive act of self-categorization does not necessarily imply that one identifies with a particular category or group. Self-categorization can exist without feelings of belonging or the group membership having emotional significance. A child may recognize and accept a group as self-defining, but he or she does not have to consider this definition personally important or be emotionally involved. However, according the social identity approach, group identities often do provide a sense of ingroup belonging (“this is where I belong”) with the related motivation to show intergroup bias. Because of their need for a positive group self, people tend to seek positive ingroup distinctiveness along valued dimensions of intergroup comparison. They try to enhance their sense of self-esteem by positively differentiating their ingroup from relevant outgroups on those dimensions that matter to their group. There is empirical evidence for this self-esteem proposition in research on children’s intergroup bias (e.g., Verkuyten, 2007). However, according to the social identity approach and the related social identity developmental theory (Nesdale, 2017), such a bias is by no means an automatic product of group distinctions. In addition to the strength of group identification, ingroup bias depends also on the cultural meanings of the group identity. The same group distinction (gender, ethnicity, and race) can have quite different meanings in different social and cultural contexts and these meanings provide the direction for how to act toward others (“what I am meant to do”). Positively differentiating one’s ingroup from relevant outgroups can also be done by trying to be more cooperative, supportive, and prosocial toward others and prosocial group norms can instigate more positive outgroup attitudes (e.g., Nesdale & Lawson, 2011) and stimulate children’s helping behavior (e.g., Sierksma et al., 2014).

Third, according to the social identity approach, group identity is shared with others and provides a basis for coordinated social behavior. Self-categorization is “the cognitive mechanism that makes group behavior possible” (Turner, 1982, p. 21). With the process of self-stereotyping people start to think, feel and behave as group members with shared perceptions, understandings and goals (“we”; e.g., collective understandings, collective interests, and collective standing) rather than as unique individuals (“I”; e.g., personal understandings, personal interests, and personal standing). Further, group identity forms the basis of social influence because there is the expectation of agreement with ingroup members who are also considered more knowledgeable about what the group means and stands for. And in situations in which the ingroup stereotype or its implications for how to act remain unclear, ingroup members tend to be actively involved in developing shared understandings. According to the social identity approach, such an understanding is important for coordinating the behavior of ingroup members and thus for social collaboration, in either the direction of positive or negative intergroup relations.

Categorization and Self-Understanding

For the social identity approach, the cognitive act of categorizing oneself as a member of a particular group is the starting point for understanding intergroup relations. Self-categorization derives from the organization of the social world which means that children first of all have to perceive and recognize the categorical distinctions that are common in their social environment.

Social Categorization

Already prelinguistic infants are capable of recognizing categorial distinctions among humans. They demonstrate a preference to look at own-ethnicity faces and listen to own-language speakers, and prefer those others who share trivial, but not arbitrary, similarities to themselves (Kelly et al., 2005; Kinzler, 2021; Mahajan & Wynn, 2012). From a very young age, children are alert to the categorial distinctions that are meaningful and used in their social environment and quickly familiarize themselves to what they are exposed to. For example, infants use language and accents as social markers and intuitively use native speakers as a particularly good source of culturally relevant information (Kinzler, 2021). This sensitivity to categorizations is critical for making sense of the social world in which infants find themselves and need to learn to function. This early sensitivity to social distinctions gradually develops into an awareness of particular categories (gender, ethnicity, race, language, and age) and the ability to use them for imposing order onto the social world and to demonstrate ingroup preference (Liberman et al., 2017; Rhodes & Baron, 2019).

However, a focus on and preference for the ingroup which is familiar and provides relevant cultural information does not have to imply outgroup dislike. Children can view their ingroup in a relatively positive way (ingroup favoritism) without derogating outgroups (outgroup dislike). They tend to be positive about their ingroup but often do not dislike outgroups (Cameron et al., 2001; Nesdale, 2017). A pro-group orientation provides a sense of familiarity and security and thus has positive psychological implications for group members and for the functioning of social groups. To be sure, such an orientation does have social implications in peer preferences and exclusive play, but these implications differ from outgroup dislike in group focus and underlying processes. However, negativity toward outgroups does occur, is sometimes quite strong (e.g., Verkuyten & Thijs, 2010), and seems to start to develop around the age of six (Buttelmann & Böhm, 2014; Cameron et al., 2001). Furthermore, it can develop earlier in deeply divided societies, such as Israel, Kosovo, and Northern Ireland in which children learn from very early on about meaningful group distinctions and the related hostilities (Connolly et al., 2009; Nasie et al., 2016).

Although social categorization in “us and them” leads to ingroup preference among children, it is not clear what happens in more complex intergroup settings. The great majority of research on children’s intergroup attitudes uses a binary framework and in doing so tends to ignore settings in which there are multiple groups on a single categorization dimension or settings in which multiple categorization dimensions are used. Related to the former, experimental research among children that uses a third group has found only signs of ingroup preference within a competitive setting (e.g., Hartstone & Augoustinos, 1995 Augoustinos & Rosewarne, 2001; Spielman, 2000). Children’s understanding and evaluation of intergroup differences can be different in a tripartite group setting compared to a binary one.

In relation to multiple categorization dimensions, there is evidence that using single categories independently of each other (e.g., only race, or gender) for making distinctions between people is probably not very common in many real-life situations. For example, older children have been found to favor using several category features simultaneously for judging contemporaries (e.g., Verkuyten et al., 1995). Further, experimental work combining two categorization dimensions has found more complex patterns of intergroup differentiation among early adolescents (e.g., Brewer et al., 1987; Vanbeselaere, 1996), also with lower ingroup preference (Vanbeselaere, 1987). The typical situation examined in crossed categorization research is a fourfold group system in which two categorization dimensions (e.g., ethnicity and gender) are crossed leading to a double ingroup (same ethnicity, same gender), two single ingroups, and a double outgroup (different ethnicity, different gender). In agreement with the social identity approach, research tends to find an additive pattern of intergroup bias with double ingroups being evaluated more favorably than single ingroups, who are rated more favorably than double outgroups. However, the pattern found depends, for example, on whether both category dimensions are equally salient and socially meaningful (Crisp & Hewstone, 2007). For example, among children in Hong Kong, a hierarchical pattern was found in which gender was the primary category and ethnicity the subordinate dimensions (Brewer et al., 1987). The relative higher importance and earlier emergence of gender-based preferences is also found in other research (Shutts, 2015; Verkuyten & Kinket, 1999) and suggests that being of the same gender is a more primary and stable determinant of children’s social interactions than ethnicity.

Self-Categorization

Social categorization allows children to navigate their complex social world in a meaningful way and to make sense of themselves. Children tend to place themselves in existing social categories and give themselves a name or a label (“I am a girl,” “I am a Mexican, German”). The ability to self-categorize develops early in life. Children quickly learn to which category they belong and to label themselves accordingly which is a sufficient precondition for ingroup preference (Bennett et al., 1998; Dunham, 2018). Children are motivated to discover their own category memberships and develop a self-understanding based on the features commonly associated with these categories. Early research demonstrated that children as young as three self-identify with a doll appropriate to their racial group (Clark & Clark, 1965) and later studies have found that self-categorization does occur around this age (Dunham, 2018). This process of self-categorization implies an ingroup–outgroup distinction that structures the social world in a subjective relevant and meaningful way. For example, research has found that ethnic self-categorization matters for friendship choices among adolescents in ethnically diverse schools (Jugert et al., 2018).

However, according to the social identity approach, self-categorization is a context dependent process. People always belong to multiple groups and these can become salient and relevant depending on the circumstances. For example, Nishina et al. (2010) found that ethnic self-categorization remained stable for only about 60% of a sample of middle school students in the US, and a study in Germany found that between 20 and 60% of adolescents’ ethnic self-categorizations varied within individuals across time (Jugert et al., 2018). These over-time changes in ethnic self-categorization do not only represent intrapersonal developments but can also reflect changing societal and political circumstances, such as changing rates of immigration and the related public and political debates (Geerlings et al., 2015).

Self-categorization influences the perception (attention, encoding, and memory) and evaluation of the social world (Xiao et al., 2016). Events that have to do with “our” group are considered self-relevant and therefore grab our attention and alter information processing when the particular group identity is psychologically salient. When a child defines herself in a particular situation as German, then what counts in that situation is the fate and standing of Germans as a whole and of her as a German person, rather than her personal fate and standing. She is concerned about Germans as a group and her feelings of pride, shame, or threat do not relate to her personally but to the achievements, standing, and interests of the Germans.

Experimental research with novel, ad-hoc groups, shows that children as young as 4–5 years of age tend to explicitly and implicitly favor their ingroup over outgroups in situations in which they cannot favor themselves personally, and also encode incoming information in an ingroup biased way (e.g., Dunham et al., 2011). This connection to the group does not only involve positive ingroup outcomes and feelings of group pride and satisfaction but can also imply the negative psychological implications of recognizing that one’s minority ingroup is being discriminated against, even when oneself does not face discrimination (e.g., Verkuyten, 1998). Further, around the age of five or 6 years, children start to accept collective responsibility for the wrongdoing or damage caused by ingroup peers. For example, children do not only want to apologize for breaking a window themselves but also when an ingroup member did so (Bennett & Sani, 2008). Furthermore, young children who have no personal involvement in the harm caused by actions are more likely to accept responsibility for negative actions of ingroup than of outgroup members (Over et al., 2016). When a transgressor belongs to their own group, children display more signs of guilt and are more likely to say that an apology is needed and that they themselves should try to repair the broken object.

Self-categorization is important for understanding intergroup relations but children can categorize themselves in multiple and changing ways. For example, ethnic self-categorization depends on the presence of ethnically relevant cues and being in a situation in which one is a numerical minority (e.g., Yip & Fuligni, 2002). However, self-categorization is not only a function of context because defining oneself in a particular situation, for example, in ethnic group terms is more common among some individuals than others. According to the social identity approach, there are individual differences in the subjective tendency or “readiness” to view oneself and the social world in ethnic terms. The readiness to think about oneself and others in ethnic terms depends on individual experiences, expectations, beliefs, commitments, and needs. And it can also depend on the way in which multiple identities are cognitively represented.

There are several theoretical models about the ways in which multiple identities gradually become, and are, cognitively organized and integrated into a sense of self (e.g., Amiot et al., 2007; Roccas & Brewer, 2002). These models are not fully the same, but also overlap somewhat, for example, in conceptualizing intersectionality as one form of multiple identity integration and proposing other forms of combination, such as blended, merged, and additive integration. Roccas and Brewer (2002) argue that the different forms of multiple identity organization can be placed on a continuum in terms of their cognitive complexity. Social identity complexity refers to individual differences in the way in which different ingroup memberships are subjectively differentiated and integrated. A complex identity structure implies that an individual accepts and acknowledges the distinctive memberships of their various ingroups (e.g., distinct ethnic, religious, and professional identities). Alternatively, individuals with a relatively simplified structure perceive a strong overlap and interrelation among their identities. Older adolescents tend to have higher social identity complexity than younger adolescents (Van Dommelen, 2014).

Low social identity complexity means that multiple identities are embedded in a single ingroup understanding, making an individual who is an outgroup (ingroup) member on one dimension (ethnicity) also an outgroup (ingroup) member on another dimension (religion). This lack of cross-cutting identities strengthens ingroup versus outgroup distinctions and thereby the distancing from outgroup members. Low identity complexity would increase the importance of the ingroup in intergroup comparisons and the motivation to favor one’s own group compared to outgroups. There is empirical evidence for the proposition that lower social identity complexity is associated with lower openness, less outgroup trust and more intergroup bias (Schmid et al., 2009; Xiao et al., 2016). For example, among ethnically diverse samples of young adolescents (12–14 years) in the United States, high social identity complexity has been found to be related to positive intergroup attitudes, both cross-sectionally and longitudinally (Knifsend & Junoven, 2013), and lower social identity complexity is associated with higher social distance from ethnic outgroups (Knifsend & Juvonen, 2014). Similar findings have been found among adolescents in Belgium and Australia (Van Dommelen et al., 2015).

Self-Stereotyping and Group Identification

Self-categorization involves giving oneself a name or a label (“I am a Mexican, a German”) and each label carries different meanings. Changing labels means taking on a different set of expectations: “a person by another name will act according to that other name” (Foote, 1951, p. 17). According to the social identity approach, there is a process of self-stereotyping in which the socially defined meanings and expectations become self-relevant. Additionally, children will gradually develop an affective sense of group belonging and commitment and identify with their group.

Both self-stereotyping and group identification are important processes for understanding children’s and adolescents’ intergroup relations. Self-stereotyping is especially relevant for the nature and direction for how children will behave toward others: when, in a particular situation, children see themselves as member of a group, they will think, feel, and act in terms of the norms, values, and beliefs that are considered to characterize the group. Group identification is especially important for the intensity of children’s behavior: higher identifiers are more strongly involved and committed to their group and therefore are likely to react more strongly toward outgroups.

Self-stereotyping and identification

According to the social identity approach, the act of defining oneself as a category member implies that the socially defined category understandings become self-relevant. A process of self-stereotyping occurs which involves an understanding of the self away from personal features and characteristics (“I am myself”), toward the relevant group-based features and characteristics instead (“I am like my group”). With self-stereotyping, the child comes to think about themselves in terms of the stereotypical beliefs and behavioral expectations that are commonly associated with what it means to be, for example, black or white or an ethnic or religious group member.

With age children develop an increasing understanding of group differences and what characterizes various groups, including their own (Abrams & Rutland, 2008). This is important because the process of self-stereotyping requires such an understanding that makes stereotyping oneself in terms of typical ingroup attributes and characteristics possible. This understanding depends, in turn, on children’s cognitive capabilities, social experiences, and the information provided by the social surrounding. For example, young children tend to differentiate less between themselves and others, have less advanced perspective taking abilities, and assume that other people see, hear, and feel the same as the child does (“the group is like me”).

From middle childhood on, children become less self-centered and increasingly interested in group differences, develop perspective taking abilities, and have more experiences with groups in various situations. As a result, they become more sensitive to group norms (e.g., Abrams & Rutland, 2008), develop relatively abstract understandings of intergroup differences (Karcher & Fischer, 2004), and apply ingroup stereotypes to themselves (“I am like my group”). Research on gender categories, for example, has shown that in middle childhood, children are able to gender self-stereotype and do so differently depending on which ingroup stereotypes make most sense in a particular comparative intergroup context (David et al., 2004; Sani & Bennett, 2004). Further, experimental research among older children has found evidence for self-stereotyping differences depending on the particular ethnic identity that is salient (Verkuyten & Pouliasi, 2002), and that the mental activation of ingroup stereotypes affects children’s performance in quantitative tasks (Ambady et al., 2001).

While self-categorization is the starting point of group processes, the social identity approach does not suggest that cognition is all there is. Self-labelling as a member of a particular group typically develops into a subjective identification with that group. Identification with broad categories such as race, ethnicity and nationality tends to develop in middle childhood (e.g., Ruble et al., 2004; Sani & Bennett, 2004) and this can have implications for outgroup attitudes. A stronger sense of ingroup belonging and attachment implies a higher dependency on the group for a positive sense of self, a higher sensitivity to anything that concerns and could harm the group, and a stronger tendency to see the world in the related group terms. Group identification plays an important role in children’s intergroup relations and can be a driver of outgroup negativity (Nesdale, 2007).

Identity Content

Children with stronger ingroup identification tend to have more positive ingroup attitudes, and sometimes also more negative outgroup attitudes, especially in competitive and threatening contexts (e.g., Nesdale et al., 2005; Verkuyten & Thijs, 2019). However, the norms and beliefs of a particular group identity can differ substantially across situations, societal contexts, and cultures. The social identity approach argues that an understanding of the relation between group identification and intergroup behavior requires attention to identity content.

The process of self-stereotyping means that children think, feel and act in terms of the meanings that are considered to characterizes the ingroup. Children understand themselves (“who I am”) in terms of how they understand their ingroup (“who we are”) and this has implications for how they perceive and behave toward others. Establishing and maintaining a distinctive and positive social identity does not have to lead to outgroup negativity. The direction for how to act toward others is provided by the norms and beliefs that characterize the ingroup, and these can also be cooperative, supportive and prosocial (“we are helpful,” “we are fair”). This means that the outgroup implications of group identification depends on the way in which children understand their ingroup. What it means to be a member of an ethnic, racial, religious, or national group does not have to mean the same thing in different local settings and different cultural and political context. As a result, the way children understand their group identity can differ, and this matters for their feelings and actions toward outgroup members.

For example, cross-national research has found that ingroup identification is typically related to positive ingroup evaluation but not necessarily to outgroup attitudes (Barrett & Davis, 2008). Also, research in multicultural Mauritius has found that adolescents with higher national identification more strongly endorse the national norm of multiculturalism which in turn is associated with lower outgroup social distance (Ng-Tseung Wong & Verkuyten, 2015). In contrast, in the US context, negative attitudes toward immigrant groups have been found to be most evident among children who have a strong view about who is a typical national (Brown, 2011). Further, longitudinal research has found that peers’ tolerance leads to higher tolerance of immigrants, while peers’ xenophobia predicts increases in adolescents’ xenophobia (Van Zalk et al., 2015). Additionally, (quasi-)experimental research has found that prosocial and inclusive ingroup norms can increase children’s ethnic outgroup liking (Nesdale et al., 2005), their positive behavioral tendency toward an outgroup (Cameron et al., 2001), and can stimulate outgroup support and cooperation (House, 2018). In contrast, outgroup dislike occurs when the ingroup has an explicit norm of outgroup rejection versus acceptance (Nesdale et al., 2005). Supporting the social identity approach, these findings indicate that intergroup biases do not only depend on the strength of ingroup identification but also on the way in which the ingroup is understood. Intergroup bias is not only about “who they are” but also about “who we are.”

Shared Understandings and Social Influence

There are many situations in which children are not concerned with their personal situation and circumstances, but rather with the category or group to which they belong. Self-related terms apply to oneself as an individual (“I”), but also to the category or group to which one belongs (“we”). According to the social identity approach, group identity provides a shared understanding of the social world which results in ingroup members evaluating outgroups in similar ways.

However, those who identify together do not always agree. Developing a shared understanding involves processes of communication and social influence, for example, when children in interaction with each other develop such understandings. This means that research should also focus on the ways in which children in social interactions develop shared intergroup understandings.

Shared Understandings

Infants expect group members to behave in similar ways (Powell & Spelke, 2013), to share preferences and traits (Diesendruck & HaLevi, 2006), and to help one another (Jin & Baillargeon, 2017). And young children (4 years and older) generalize social norms across members of the same category (Kalish. 2012), negatively evaluate people who do not follow their ingroup norms and conventions (Schmidt et al., 2012), and use information about how a group is for making inference about how individual group members should be (Roberts et al., 2020).

A group identity gives a shared interpretation of the social world. Self-categorization and identification with a particular group provide a “lens” through which one sees the world and makes sense of it. Social understandings are tied to the social groups we belong to and tend to be shared among group members. For example, children as young as 4 years of age can be aware of the prevalent cultural stereotypes about social groups (Bar-Tal & Teichman, 2005). This awareness develops further in middle childhood (e.g., McKown & Weinstein, 2003) with children using cultural stereotypes for thinking about themselves and others, and for differentiating between multiple outgroups.

Much of the research on children’s outgroup attitudes focuses on binary racial, ethnic, national, or religious group distinctions. The focus is on individual differences in attitudes toward one particular (minority) outgroup. However, in stratified and multi-ethnic, multi-religious societies the intergroup context is a series of groups about which there are common stereotypical understandings and shared social distance hierarchies that reflect the relative societal positions of the minority groups (e.g., Barrett & Davis, 2008). In spite of individual differences in attitudes, there tend to be consensus among ingroup members about the evaluation of multiple outgroups.

Different groups enjoy varying degrees of social acceptability so that some outgroups are kept at a greater social distance than others. There is evidence for such a sequence in implicit evaluations and in social distances among majority group members in different countries. For example, within the US, there is a social hierarchy in implicit evaluations by race (Whites > Asians > Blacks > Hispanics) and religion (Christianity > Judaism > Hinduism or Buddhism > Islam) (Axt et al., 2014). These sorts of ethnic and religious social hierarchies are also found among children and adolescents in countries such as Australia (Augoustinos & Rosewarne, 2001), Spain (Enesco et al., 2005), Britain (Barrett & Short, 1992), and the Netherlands (Bruijn et al., 2020; Verkuyten & Thijs, 2010).

With Guttman scalogram analysis, it is possible to systematically test in how far youngsters have similar social distances toward different outgroups and whether they follow a similar unidimensional hierarchical pattern of outgroup distances, regardless of individual differences in attitudes. Using this type of analysis, research among early adolescents (10–12 years) and middle adolescents (14–16 years) in the Netherlands has found such a pattern for ethnic outgroup distances indicating consensus on the ethnic hierarchy (Verkuyten et al., 1996; Verkuyten & Kinket, 2000). Adolescents shared a similar pattern of distances, independent of age, gender and also level of prejudice. Adolescents low on prejudice indicated overall lower social distances toward the different ethnic outgroups but had a similar cumulative pattern of preferences as those high on prejudice. This indicates that the ethnic hierarchy is a scheme of social preferences that is shared by ingroup members, independently of adolescents’ own attitudes.

However, even when a scheme of preferences and group stereotypes is shared, there are still differences between ingroup members in how they understand and use these preferences and stereotypes. For example, one’s knowledge about cultural stereotypes can correspond less or more strongly with personal group beliefs and the personal endorsement of stereotypes. Correspondence is more likely for younger children (5–6) than for older children (8–9) but can also differ between children of the same age (Augoustinos & Rosewarne, 2001; Degner & Wentura, 2010). Personal belief and stereotype endorsement depends on children’s general categorization skills, experiences with outgroup members, and the relative importance that particular group distinctions have in their immediate contexts and the broader environment.

Social Influence

Children’s attitudes toward ethnic, racial, and other groups are affected by experimentally induced (e.g., McGuire et al., 2015; Nesdale & Dalton, 2011) as well as perceived peer group norms (e.g., Brenick & Romano, 2016; Jugert et al., 2011). These influences are often understood in terms of normative compliance whereby children want to gain or maintain social approval and avoid disapproval (Rutland, 2004). However, from the social identity approach, group influence is less about normative compliance and more about informational influence (Turner, 1991). Group members tend to trust their ingroup which as a reference group is considered to provide valid and relevant information about the world. Children follow and rely on ingroup members for developing an understanding of social reality and not only for concerns about social sanctions and compliance. Infants and children prefer to learn from ingroup, compared to outgroup, members who are seen as providing relevant cultural information (Begus et al., 2016) and are considered more trustworthy (Chen et al., 2013).

Group identity provides a shared understanding of social reality, but this does not imply that those who identify with each other automatically agree. At a young age (around 4–5 years), children become aware of socially meaningful category distinctions and the related social status differences. However, it is not always clear what these distinctions mean and group members can disagree on the meanings. Disagreement and debate about group distinctions is always possible and also occurs, for example, in peer groups. Already pre-schoolers recognize and follow social norms, but also collectively create and change them (Schmidt & Rakoczy, 2018). Children are active participants who interact with each other and deliberate and negotiate in quite sophisticated ways about, for example, ethnic, racial and cultural group differences (Moore, 2002; Van Ausdale & Feagan, 2002).

Yet, group identity is an important basis for mutual influence and developing consensus. When people think about themselves as belonging to the same ethnic category, there is an expectation of agreement and a motivation to reach consensus on the meanings and implications of the group identity (Haslam et al., 1999). When learning new information, young children look to ingroup, rather than outgroup, members (Begus et al., 2016) and they seek, accept, and endorse consensual information from members of their own group, rather than from a different group (Chen et al., 2013). A common identity means that one is similar in one way or another and belong together, and this leads group members to seek agreement and try to create consensus. This means that is important to examine the ways in which children understand social reality and collectively define it. According to the social identity approach, ingroup preference and outgroup dislike are shared products of social processes of influence.

Research shows that the language that parents use for social groups matters for children’s understanding of these groups (e.g., Moty & Rhodes, 2021). Yet, children do not simply adopt their parents’ views and also talk among each other and try to convince each other about the meaning of group differences. For example, children use various discursive devices to present their outgroup judgments as factual and accurate features of the social world rather than the result of their own perceptions, attitudes, and desires (Verkuyten & Steenhuis, 2005). In this way, the rejection of outgroup members as potential friends can be made empirically grounded (“they really are mean, dirty”) and thereby reasonable and acceptable. An explanation for ethnic peer exclusion is more convincing when it is presented as based on actual cultural, religious, or linguistic differences rather than on stereotypical views.

Research that emphasizes the role of social interaction in negotiating race and ethnic group differences shows that the interactions of very young children can be highly racialized (e.g., Connolly, 1998; Van Ausdale & Feagan, 2002). However, other research indicates that children also use more inclusive racial and ethnic discourses (Carrington & Short, 1995; Howard & Gill, 2001), and that there are multiple understandings. For example, research has found that aspects of 8- to 11-year-old children’s talk (e.g., racialized) poses a barrier to the development of an inclusive Welsh citizenship for ethnic minority children, but that there are also aspects of talk (e.g. equality, similarity) that supports inclusiveness (Scourfield & Davies, 2005). Other research on social interactions shows that older children establish and use, but also challenge and disrupt, race category membership in peer relations and renegotiate group boundaries of inclusion and exclusion (Moore, 2002).

Research on children’s outgroup attitudes has mostly ignored the importance of communication processes and how peers in interaction with each other develop shared understandings. Much of the existing research is interested in understanding and changing children’s hearts and minds. This research tends to create an experimental or survey setting “of silence” in which children do not talk to each other and do not hear what others have to say. However, for understanding intergroup relations in children’s everyday lives, it is also important to examine how group characterizations and understandings are actually discussed, fabricated and challenged in interaction. The active and flexible ways in which children in peer interactions present, negotiate and collectively define group identities and meaningful group differences should also be studied because this can improve our understanding of how children actually live with group differences (e.g., Connolly, 2001; Edwards & Stokoe, 2004; Van Ausdale & Feagan, 2002).

Coordinated Action and Societal Context

According to the social identity approach, a shared sense of “us” transforms individual relationships in collective ones because people see each other as belonging to the same group and start to act on the basis of the shared understandings, beliefs, and norms that define who “we” are and what counts for “us.” Thus, a sense of “us” is the psychological mechanism that makes group behavior possible because it provides unity and common direction in perception and behavior.

Children may form a “we” and the ability to act on the basis of this allows to coordinate their behavior with each other. For example, when two groups of children are playing a game of soccer they will see themselves and others as representatives of their respective teams and not as unique individuals. A sense of “us” makes it possible to try to achieve group identity-related goals, and these goals can be very different depending on the values, norms, and beliefs of the specific group. For instance, a group of children can try to raise money to help the less fortunate or work together to welcome and support refugees (e.g., “Children helping other children”), but they can also be involved in claiming and marking “our” territory and playground or in drawing group boundaries for excluding outsiders and newcomers (Sherif, 1956). Further, clique building is a strong element of youth culture and there are, for example, politically extreme right-wing and also left-wing youth cliques that define themselves in opposition to their ideological “rivals” (Steinbach & Gissendanner, 2003).

Most of the research on children’s and adolescent’s intergroup biases focuses on individual perceptions, feelings and beliefs, and finding effective ways for trying to change the negative attitudes. Yet, bias, exclusion and discrimination does not happen in a vacuum but involve social processes that result from interactions within and between groups. Peer groups often serve to justify and stimulate negative behaviors that tend not to occur otherwise, such as among hate groups, aggressive subcultures, and street gangs (Levin & McDevitt, 1993). Forms of interaction in peer groups are especially relevant for youth who in public places often congregate together in homogeneous cliques where a sense of togetherness and meaning can be found. For example, the typical xenophobic youth is a clique member who defines themselves as right-wing (e.g., Watt, 2001). The drawing and marking of territorial and of ethnic or racial group boundaries can be a daily process that fuels intergroup negativity (Kintrea et al., 2008).

However, in many countries, there are also anti-racist youth initiatives, such as SOS Racism in France (“Touche pas à mon pote”), Youth Fight Against Racism in Britain, and Youth against Racism in Sweden (“Ungdom mot rasism”), Germany (“Jugend gegen Rassismus”), and Belgium (“Jongeren tegen racism”). Collective action is also an important strategy for challenging discriminatory practices and trying to increase social justice in the context of school or local neighborhood, and on a regional or national level. For example, youth can be actually and virtually (social media) involved in social justice movements (Cohen & Kahne, 2012).

These different forms of collective action are complex phenomenon in which, for example, network dynamics, local conditions, and broader institutional, economic, and political circumstances play a role. The contribution of the social identity approach is that it points at the importance of the development of a shared sense of “us” and how this plays out in the particular sociocultural context. According to the approach, forms of collective action depend on the way in which the societal context is perceived (Tajfel & Turner, 1979). Specifically, collective action among youth is more likely when the intergroup structure is considered unjust or illegitimate and there is the belief that the status quo can be changed (Bettencourt et al., 2001; Verkuyten & Reijerse, 2008). For example, in a research among adolescents in Macedonia, it was found that perceived social injustice toward the Roma minority was associated with support for Roma empowerment (Kamberi et al., 2017). Alternatively, perceived threats and social injustices as well as perceived “reversed discrimination” can underlie the engagement of majority youth in right-wing extremist groups and their involvement in defending “our” neighborhood (Kintrea et al., 2008).

From a developmental perspective, this raises important questions about how an understanding of the stability and legitimacy of the prevailing societal structure is gained. For example, although older children have a sense of societal inequalities and their legitimacy (Emler & Dickenson, 2005), they tend to hold views that are consistent with the status quo and, for example, do not readily perceive institutional discrimination (Spears Brown & Bigler, 2005). Gradually during adolescence a more sophisticated societal and political understanding develops, including beliefs and lay theories about unfairness and inequalities between groups (Sears & Levy, 2003). For example, perceptions of racial disparities and attributions of these disparities to racism has been found to predict American adolescents’ support of affirmative action and school desegregation policies, especially among older adolescents (Hughes & Bigler, 2011). Older adolescents are more likely to understand the institutional creation of racial disparities and therefore can view race-conscious policies as an appropriate method of addressing these disparities.

Alternatively, perceived threats and unfair treatments of majorities might lead to opposition toward minority and immigrant policies. Furthermore, lay theories can justify youngsters’ negative outgroup attitudes and exclusionary behavior, as has been found for protestant work ethic beliefs (Levy et al., 2008), and shared conflict beliefs (Nasie et al., 2016). Further, research among adolescents in several countries shows that a more civic understanding of national belonging is associated with more positive attitudes toward immigrants and minority groups, whereas an ethnic understanding is associated with more negative attitudes (Reijerse et al., 2013). This latter understanding can be linked to the belief that a place belongs to the indigenous majority group because they were here first and therefore that it is acceptable to exclude newcomers. This notion of primo-occupancy with the related ownership feelings (“this is ours”) has been found to be associated with Dutch majority group children’s (grades 4–6) negative attitudes toward immigrants (Verkuyten & Thijs, 2019).

The social identity approach makes it possible to connect more systematically the research on intergroup biases with the extensive literature on social development and children’s historical and political understanding of society (Patterson et al., 2019; Sears & Levy, 2003). Experimental research, for example, has found that history lessons about racism make European American children (6–11 years) to more strongly value racial fairness and therefore to be less negative toward African Americans (Hughes et al., 2007). However, historical narratives can also justify adolescents’ negative outgroup reactions, such as in Northern Ireland and Israel (Bar-Tal & Teichman, 2005).

It is also important to recognize that there can be an intergroup dynamic of conflict between, for example, youth scenes with negative contacts leading to increasing hostility. For example, local acts of violence between neo-Nazis and anti-racists have been shown to be part of a conflict dynamics between these two youth groups. These dynamics involve processes of co-radicalization (Pratt, 2015) and polarization within and between these two groups that were based on the shared notion of “one for all and all for one” (Bjørgo, 2005). Further, Islamic extremist youth can use right-wing extremism to claim that the West is hostile and violent toward Islam, and right-wing extremist use Islamic extremism to argue that Islam is incompatible with the West (Holtz et al., 2015). Another example, is that the discrimination that minority youth face can lead to stronger minority group identification with the related visible cultural enactments (e.g., in clothing, style, and preferences). This enactment publicly expresses and affirms one’s minority identity. In turn, majority group members can react more negatively toward these identity enactments because they see them as threatening their cultural identity and worldview (Kaiser & Pratt-Hyatt, 2009) leading to the fear that the minority culture will override one’s own way of life and thereby the prevailing status arrangements in society (Verkuyten, 2018).

Limitations of the Social Identity Approach

The social identity approach provides systematic avenues for understanding and examining children’s and adolescents’ intergroup biases, but also has limitations. Specifically, there are various developmental processes and factors that are not considered and that future work could try to relate to social identity processes. I will discuss two of these limitations: identity development and social influences.

Identity Development

A first limitation is that in examining intergroup relations, the social identity approach is not concerned with important developmental changes in children’s cognitive skills (e.g., multiple classification; Aboud, 2005), affective abilities (e.g., perspective taking, empathy; Eisenberg, 2005), understanding of group dynamics (Abrams & Rutland, 2008), and their sense of identity. Concerning the latter, the social identity approach tends to focus on the situational process of self-categorization and individual differences in the degree of group identification and the approach has very little to say about the development of an inner sense of group self. Developmental work makes a useful contribution here, for example, by conceptualizing and examining how processes of identity exploration are involved in developing a sense of identity. This work suggests that a secure identity is the result of an exploration process whereby adolescents try to understand how, for example, their ethnic or racial group membership impacts their life, and establish a clear sense of ingroup belonging and commitment based on that understanding (Phinney, 1992; Umaña-Taylor et al., 2014). A secure group identity would encourage positive intergroup relations, if it is based on identity exploration (Phinney et al., 1997).

However, whereas many developmental studies have examined the beneficial effects of identity exploration for adolescents’ well-being, sociocultural adjustment, and health (Rivas-Drake et al., 2014), there is little empirical evidence for the expected relevance for outgroup acceptance. Very few studies have examined whether exploration-based ingroup commitments do indeed make youth more open toward others. One example is a study among Latino and Asian American adolescents that found more positive attitudes toward ethnic outgroups among those with high ethnic ingroup commitment and exploration, whereas those with high commitment and low exploration were more negative toward outgroups (Phinney et al., 2007). Another study among Latino, Asian, and European American adolescents found that ethnic identity exploration promotes positive outgroup attitudes via a secure attachment to one’s ethnic identity (Whitehead et al., 2009). Further, Turkish-German ethnic minority adolescents showed less ingroup bias when they identified strongly as Turkish and exploration was high, but more ingroup bias when they identified strongly and exploration was low (Spiegler et al., 2016).

A research among German majority members further examined the importance of national identity exploration for the association between national identification and outgroup attitude (Spiegler et al., 2021). With five experimental studies among late adolescents and emerging adults it was found that strong national identifiers with high identity exploration have more positive attitudes toward immigrants than strong identifiers with low exploration. Furthermore, identity exploration was found to enable strong identifiers to oppose exclusionary ingroup norms and that exploration was associated with more positive attitudes via lower feelings of outgroup threat and an more nuanced perspective on the ingroup culture.

Social Influences

A second limitation, is that the social identity approach argues for the importance of interaction and communication for developing shared group understandings, but does not discuss how exactly this is achieved. The focus is on shared identity as a basis for social influence but children have multiple identities and there are many sources of influence on children’s intergroup biases. For example, not only the role of peers but also of parents, teachers, school curriculum, and the mass media are important to consider. Through their discourses and practices, parents and teachers communicate beliefs about group differences and the mass media influences the representational content that children are exposed to. There are various potential sources of influence that also can contradict each other and that need to be examined in comprehensive ways.

Furthermore, children do not passively absorb information of the different sources. They can be faced with social environments that are unclear or ambiguous and with groups that lack clarity in their identity content (Van Veelen et al., 2016). This makes, for example, self-stereotyping (“I am like my group”) more difficult and social projection whereby personal self-attributes are projected onto the group more likely (“my group is like me”). Social projection is not only more likely among younger (self-centric) children who tend to assumes that other people see the world in the same as they do, but also, for example, with increasing cultural diversity and rapid cultural changes in which groups are heterogenous and ambiguously defined (Thijs & Verkuyten, 2016; Thijs & Zee, 2019).

Conclusions

The social identity approach is increasingly being used for understanding intergroup relations among children and adolescents, in particular for predicting that youngsters will show ingroup bias for creating or maintaining a positive group identity with the related positive self-esteem. However, the approach has a broader relevance because it draws attention to several processes and issues that are often overlooked in empirical research with children and adolescents. The social identity approach goes beyond individual-level abilities and dispositions (personal self) by focusing on group-level psychology (group self) and linking psychological processes with societal circumstances (Reicher et al., 2010).

The emphasis is on the processes of self-categorization, self-stereotyping, and group identification that serve to unite and shape the thoughts, feelings, and actions of those who belong to the same group. Intergroup biases are not only a question of personal attitudes that are mainly studied by developmental researchers but also of shared understandings about group differences and group belonging. The social identity approach does not propose a generic ingroup bias but rather emphasizes that the way in which psychological processes play out is dependent upon the local and broader societal context. For example, it is important to know that children strive for a positive group identity, but that tells us little about how this is achieved. There are many ways in which one’s ingroup can stand out positively, including by being more cooperative, supportive, and helpful. Intergroup biases can take different forms in different local contexts (e.g. schools and neighborhoods), and children’s understanding of group differences in deeply divided (e.g., Israel and Northern Ireland) or strongly racialized (U.S.) societies is likely to differ from their understandings in more egalitarian and multicultural nations (e.g., Mauritius and New Zealand).

Research on children’s and adolescents’ intergroup relations has been on-going for several decades, is increasingly conducted outside the U.S. and continues to grow. Questions related to children’s group biases are obviously of great importance, theoretically and practically. There is a substantial body of empirical knowledge and there are various theoretical approaches that make different contributions to our understanding, and for addressing intergroup biases among youngsters across the globe. The social identity approach is quite influential in the field because it has proven to be very helpful for researchers to explain and understand some central features of intergroup relations in a wide range of settings. The approach has much more to offer than the well-known prediction that children will tend to show ingroup bias because of their need for a positive group identity. The theoretical framework is broader and richer in its focus on cognitive and motivational processes and how these operate and play out in the social, cultural, and political context in which children find themselves. The social identity approach, like all approaches, has limitations when applied to the developmental context but also suggests a number of useful and important avenues for further investigating and understanding children’s and adolescents’ biases within the increasingly diverse world in which they grow up.