Reducing Prejudice through Promoting Cross-Group Friendships

Abstract

Around the globe, individuals are affected by exclusion, discrimination, and prejudice targeting individuals from racial, ethnic, and immigrant backgrounds as well as crimes based on gender, nationality, and culture (United Nations General Assembly, 2016). Unfortunately, children are often the targeted victims (Costello & Dilliard, 2019). What is not widely understood is that the intergroup biases underlying systemic racism start long before adulthood with children displaying notable signs of intergroup bias, sometimes before entering grade school. Intergroup bias refers to the tendency to evaluate members of one’s own group more favorably than someone not identified with one’s group and is typically associated with prejudicial attitudes. Children are both the victims and the perpetrators of bias. In this review, we provide evidence of how biases emerge in childhood, along with an analysis of the significant role of intergroup friendships on enhancing children’s wellbeing and reducing prejudice in childhood. The review focuses predominantly on the context of race, with the inclusion of several other categories, such as nationality and religion. Fostering positive cross-group friendships in childhood helps to address the negative long-term consequences of racism, discrimination, and prejudice that emerges in childhood and continues through to adulthood.

Over the past several decades, there has been a more explicit societal awareness of the pervasiveness of children and adults who have been affected by an increase in hate crimes targeting individuals from racial, ethnic, and immigrant backgrounds as well as crimes based on gender, nationality, and culture (United Nations General Assembly, 2016). This shift has prompted the United Nations to call for renewed actions combating “contemporary forms of racism, racial discrimination, xenophobia and related intolerance” (United Nations General Assembly, 2016). This call has become even more urgent in 2021 as nations continue to confront the heightened health disparities resulting from racial injustice in the context of a worldwide pandemic.

What is not widely understood is that intergroup biases underlying systemic racism start long before adulthood with children displaying notable signs of intergroup bias, sometimes before entering grade school. Intergroup bias refers to the tendency to evaluate members of one’s own group more favorably than someone not identified with one’s group (Dovidio et al. 2017; Tajfel & Turner, 1979), and is typically associated with prejudicial attitudes. Intergroup biases, however, are not necessarily the same as ingroup preferences given that the latter refers to enhancing one’s own group without necessarily derogating the out-group. Research has shown that children as young as 3- to 5-years old demonstrate ingroup preference toward peers of their same gender and race (Cvencek et al., 2011; Hilliard & Liben, 2010; Renno & Shutts, 2015). Moreover, children in this age range also prefer to become friends with peers who natively speak their language over peers who speak with foreign accents (Kinzler et al., 2009).

Ingroup preferences, however, can turn into biases, particularly without experiences that enable children to understand what makes biases wrong and unfair (Rutland & Killen, 2015). What has been well documented is that biases (Dunham et al., 2011 McGlothlin & Killen, 2010; Rizzo et al,. 2021), exclusionary behavior based on race (Cooley et al., 2019), implicit racial bias (González et al., 2017) and the propensity to form racial categories about people based on verbal cues (Roberts & Gelman, 2017) emerge during childhood.

This body of research has shown that by the age of 4 years children have knowledge about different social groups and show preferences for some social groups over others (Nesdale et al., 2007). Furthermore, much of the research suggests that intergroup bias increases throughout childhood (Raabe & Beelmann, 2011), and into adulthood (Henry & Sears, 2009). Experiencing social exclusion and discrimination in childhood, moreover, can incur compromised well-being (Neblett et al., 2008; Russell et al., 2012; Yip, 2014), stress and anxiety (Fisher et al., 2000; Neblett & Carter, 2012), sleep disorders (Yip, 2015), and low academic achievement (Alfaro et al., 2006; Benner & Graham, 2011). A meta-analysis of the discrimination literature Benner et al. (2018) indicated that successfully reducing or eliminating intergroup biases helps to avoid these long-term negative consequences.

A recent field of psychological science, referred to as developmental intergroup research, has documented the presence of biases, prejudicial attitudes, and exclusionary judgments and behavior in childhood, and how it emerges, changes, and develops. Overall, the evidence points to the expectation that childhood is an ideal timeframe for reducing and combating intergroup bias. Understanding the development of children’s intergroup biases has been shown to be an effective way to change attitudes, which are forming during the childhood years and have not yet become ingrained patterns of behavior and thought. In fact, research has shown that intergroup attitudes are more malleable in childhood than in adulthood (Lee et al., 2017). A central goal for developmental intergroup research is therefore to understand and document what factors help to reduce prejudice in childhood.

One of the most significant factors that contributes to a reduction in prejudice, bias, and exclusionary attitudes and behavior is cross-race/ethnic friendships (Tropp & Pettigrew, 2005). Friendship is a fundamental form of peer relationship that provides a foundation for humans to develop a social understanding of the world (Piaget, 1932; Rubin et al., 2006). The experience of a friendship with someone from a different race, ethnicity, or nationality may enable one to reject stereotypic associations based on these group categories. This might come about when there is a recognition that such associations do not apply to everyone, and that such generalizations about others create psychological harm or unfair treatment. Thus, studying cross-group friendship has been theorized to be a unique and significant form of experience that helps to reduce bias and discrimination.

Cross-group friendships have long been of interest to developmental psychologists for several reasons. The first reason is historical. In many places around the globe, catastrophic events led psychologists to examine the underlying psychological mechanisms that led to intergroup conflict including factors that create partial solutions, such as ways to promote friendships across racial, ethnic, religious, cultural, national, and gender lines. As an example, social psychologists focused on the possible benefits of cross-group friendships that resulted after WWII, Apartheid in South Africa, as well as the Jewish-Arab conflict in the Middle East (for reviews, see Hewstone & Swart, 2011; Rutland & Killen, 2015; Tropp & Pettigrew, 2005; Verkuyten, 2011; Voci & Hewstone, 2003). Social psychology research, stemming from Allport’s theory, pointed to the conditions that must be met, outlined in his classic book (Allport, 1954), to reduce prejudice: common goals, cooperation, equal status, and authority support (Pettigrew & Tropp, 2006). Because cross-group friendships were presumed to satisfy the first three of these four conditions, friendship potential was thought to be a fifth condition of effective contact (Pettigrew, 1998).

A second reason for the focus on friendships as a way to reduce intergroup bias in childhood stemmed from a major historical event in the U.S. that occurred with the end of legalized racial segregation of the schools. With the landmark U.S. Brown v Board of Education Supreme Court decision in 1954, lawyers called upon developmental researchers who argued that greater interracial friendliness could be considered one indicator of successful school integration (St. John & Lewis, 1975). Since the 1970s, then, developmental scientists have been asking whether children prefer friends from their same racial/ethnic group, the conditions under which they become receptive to cross-racial/ethnic friendships, and what, if any, are the benefits of crossing racial/ethnic boundaries in friendship choices.

Cross-group friendships in schools have been studied at virtually all developmental periods from preschool to adolescence (see Graham & Echols, 2018). One robust finding from this literature is that students from all racial/ethnic groups and developmental periods studied show a preference for same race/ethnicity peers over different race/ethnicity peers. This pattern of ingroup preference is not surprising since one of the major determinants of friendship selection is homophily, or the principle that people are more likely to seek contact with similar than dissimilar others (McPherson et al., 2001; Ramiah, et al., 2015). Race/ethnicity is one of the most salient dimensions of similarity. Research taking a close look at how children interpret “similarity,” however, reveals that for some children similarity is about shared interests and values, not one’s actual physical appearance (Hitti & Killen, 2015; McGlothlin & Killen, 2005). What contributes to the finding that some children view similarity as about race, and other children view similarity as about shared interests has become a core focus of developmental intergroup research.

A core feature of the social science argument in support of Brown v Board of Education was that closer contact between members of different races (i.e., through school desegregation) promotes positive racial attitudes, assuming that Allport’s (1954) conditions of effective contact – that is, equal status, cooperation, common goals, support from authority figures, and intergroup friendships-- were satisfied (Pettigrew, 1998; Tropp & Pettigrew, 2005). Thus, just as the principle of homophily helps explain students’ preference for same-race peers, the principle of propinquity, or sharing the same space, helps explain the conditions under which youth are likely to form cross-race/ethnic friendships (Mouw & Entwisle, 2006). Students have more opportunities to be friends with peers who are readily available to them, such as when they reside in the same classroom or school. A series of studies conducted by Hallinan and colleagues during the 1980s with Black and White samples of 4^th-7^th graders (e.g., Hallinan & Smith, 1985; Hallinan & Williams, 1987) supported the propinquity hypothesis: as the proportion of numerical minority group students (Black or White) increased in a classroom, numerical majority group students (White or Black) were more likely to befriend them.

More recent studies of cross-race/ethnic friendships of older youth have benefited from the much larger and ethnically diverse samples available in national panel studies such as the U.S. National Longitudinal Study of Adolescent to Adult Health (Harris & Udry, 2014). Like the earlier and smaller scale research with younger children, studies report that increasing racial/ethnic diversity of middle schools and high schools leads to more cross-race/ethnic friendships (Kao et al., 2019;, Moody, 2001; Mouw & Entwisle, 2006). However, the rate of increase does not keep pace with the opportunity for such friendships based on racial/ethnic diversity. At moderate levels of school diversity with two racial/ethnic groups of about equal size, the likelihood of befriending a peer from another race/ethnicity actually decreases (Aboud, et al., 2003; Moody, 2001). There are constraints on who benefits from diversity (Brown & Juvonen, 2018; Brenick et al., 2018) and the school organizational and instructional practices that may limit students’ likelihood to form cross-group friendships even in racially/ethnically diverse schools are discussed in more detail later in this article.

Having laid the historical, theoretical, and empirical groundwork for studying cross-race/ethnic friendships as a context for reducing prejudice in youth, the remainder of this article is organized as follows. First we review research on the degree to which cross-race/ethnic friendships are indeed related to children’s improved intergroup attitudes as Allport’s contact theory would suggest. Second, we turn to mechanisms, or the psychological processes that might explain friendship effects on intergroup attitudes. Historically, most research on friendships and intergroup attitudes has been conducted from the perspective of high-status majority groups (e.g., White individuals). In the third section we address this asymmetry by considering the impact of cross-group friendships on racial/ethnic minoritized groups. Most cross-group friendship research studies children and adolescents in school settings. The fourth section considers school factors that may limit the formation of cross-race/ethnic friendships even in diverse schools where propinquity is presumably high. In the next sections we consider particular topics in the intergroup friendship literature that have received insufficient attention such as the role of biracial youth, intergroup friendships in cross-national contexts, the importance of friendships as indirect and extended contact, and intervention strategies for building cross-race/ethnic friendships. We conclude the article with suggestions for future research.

Do Cross-Race/Ethnic Friendships Reduce Prejudice?

Over the past several decades, studies with samples from childhood to adulthood have been conducted to examine whether and how interracial contact improves intergroup attitudes. In two meta-analyses of this vast literature, Pettigrew and Tropp (2006, 2008) concluded that contact does indeed reduce prejudice. Focusing particularly on friendships, a more recent meta-analysis reported that interracial friendships were more strongly related to better intergroup attitudes than mere contact that lacked the closeness of a friendship (Davies et al., 2011).

Recent studies with youth using peer nominations to determine interracial friendships reported improved attitudes about the group to which the friend belongs (e.g., Chen & Graham, 2015; Feddes et al., 2009; Kelleghan et al., 2019; Rastogi & Juvonen, 2019; Rivas-Drake et al., 2019; Titzmann et al., 2015). For example, Chen and Graham (2015) found that Asian American 6^th grade students reported more positive attitudes toward Whites, Latinx, and Black Americans respectively when friendships with members of those groups exceeded chance. In a novel analysis, Titzmann et al. (2015) assessed whether German adolescents had friendships with ethnic German immigrants over a 3-year period. Native Germans who gained an immigrant friend over that time reported a greater decrease in prejudice toward ethnic immigrants than youth who maintained, never had, or lost a friendship, suggesting that the intentional act of forming a new cross-group friendship can have a greater impact on changes in intergroup attitudes. Examining friendships between Black and Latinx youth in middle school, Rastogi and Juvonen (2019) found that stable cross-race/ethnic friendships were a significant predictor of positive intergroup attitudes over and above the actual number of friends from the target outgroup.

The U.S studies cited above are noteworthy in that they move beyond the White – ethnic minority divide that characterizes earlier research. These studies examine cross-race/ethnic friendships and attitudes of Whites toward racial/ethnic minoritized groups, minoritized groups toward Whites, and minoritized racial/ethnic groups toward each other.

Recent longitudinal studies with adolescent samples have attempted to address whether cross-group friendships lead to improved intergroup attitudes (socialization) or whether more positive attitudes increase one’s receptivity to crossing racial boundaries in friend choice (selection effects) (Binder et al., 2009; Rivas-Drake et al., 2019; Swart et al., 2011). The best evidence thus far supports bi-directional effects. Whereas youth who embrace interracial friendships have more positive outgroup attitudes over time, pre-existing attitudes also influence the extent to which individuals take up the opportunity to have interracial friends. Thus, both socialization and selection effects appear to be operating and it is difficult to determine “which comes first” even with a new generation of advanced statistical techniques.

Cross-Race Friendships and Decisions About Exclusion and Inclusion

In the research described above, attitudes are captured by preferences or general liking of ingroups and outgroups Expanding our understanding of intergroup attitudes as a function of cross-group friendships, one promising line of research has probed more deeply into how children and adolescents make decisions about whom to include or exclude in interracial and same-race contexts. This research has also demonstrated the benefits of cross-race friendships for children of different racial/ethnic groups. In one set of studies, participants attending racially homogeneous or heterogeneous schools were asked to evaluate scenarios about everyday interactions (during lunch, at home, at a school dance) in which race-based exclusion of peers took place (Crystal et al., 2008; Edmonds & Killen, 2009; Ruck, et al., 2014). Individuals with high reports of cross-race friendships were more likely to perceive race-based exclusion as wrong than were children and adolescents with low contact (Crystal et al., 2008; Ruck et al., 2011, 2014). Further, White children and adolescents reporting cross-race friendships in racially heterogeneous schools were less likely to use stereotypes to explain racial discomfort, and more likely to expect that racial exclusion occurs than were participants enrolled in low-diversity schools (Killen et al., 2010).

In a study with adolescents on more nuanced and developmentally salient indicators of inclusion and exclusion, Edmonds and Killen (2009) examined adolescents’ perceptions of parental messages about cross-race friendships, including friendship and dating. Adolescents’ intergroup contact and perceptions of parental attitudes differed according to the intimacy of the relationships, and the types of messages parents communicated. Adolescents who chose to limit their intimacy with cross-race relationships were influenced by discouraging messages from their parents about inviting their cross-race friendship home or about who they should date (whether it would be permissible to date someone from a different racial background). According to adolescents, parents conveyed more negative messages to their adolescents about dating than about friendship and this appeared to influence their own choices on inclusion of friends and dates from different racial backgrounds. This study provided a window into a significant source of influence for adolescents’ decision making about whom to date, and whether to include cross-race relationships when socializing in their own home. An important aspect of this study was the focus on intimate relationships, a topic rarely studied in developmental science despite the fact that cross-race marriage and dating has steadily increased in the U.S. over the past 45 years, from 3% to 17% (Kennedy, 2003; Livingstone & Brown, 2017).

Cross-Race Friendships in Early Childhood

Although most research on the relations between cross-race/ethnic friendships and intergroup attitudes reviewed in this article has focused on middle childhood and adolescence, a growing empirical literature with much younger children has examined the longitudinal implications for cross-race/ethnic friendships and racial/ethnic biases in early childhood. Gaias et al. (2018) investigated whether diversity in preschool predicted racial/ethnic bias and cross-race/ethnic friendships in first or third grade. Findings revealed stability in cross-race/ethnic friendships and biases over the elementary school years, with exposure to diversity in preschool associated with more cross-race/ethnic friendships in first grade and lower levels of racial/ethnic bias. Related research has examined young children’s use of race when predicting other children’s friendship patterns and demonstrated that children expect White and Black peers to have same-race friendships and that multiracial children would have both White and Black friends, revealing same-race bias expectations for friendship as early as 4 – 6 years of age (Roberts et al., 2017).

A recent study examined the developmental mechanisms underlying the emergence of racial bias in 4 year-old children by examining beliefs about interracial friendships and beliefs about racial inequalities (Rizzo et al., 2021). The researchers found that 4 year olds’ beliefs that their parents and peers do not value interracial friendships predicted increased racial bias. This study reveals the importance not only of interracial friendships but the belief that parents and peers support these types of relationships, even as early as 4 years of age (Rizzo et al., 2021).

In sum, even preschool-aged children are aware of racial categories (Rizzo et al., 2021; Shutts, 2015), can select friends based on race (Romero et al., 2009), and reliably report intergroup attitudes (Dunham et al., 2013). Moreover, age-appropriate novel measures have been created to examine these processes. At what age, we can therefore ask, do relations between cross-race friendships and intergroup attitudes actually impact one another? More longitudinal research on within-person change is needed to answer this important developmental question.

What Explains Cross Race/Ethnic Friendship Effects?

Another line of research has begun to uncover the processes, or mediating mechanisms, that might explain cross-race/ethnic friendship effects on intergroup attitudes. Reductions in anxiety about interacting with outgroup members, increased perspective taking, increased moral reasons about the unfairness of social exclusion, and peer group norms in support of contact appear to have explanatory power (Levy et al., 2016; Page-Gould et al., 2014; Turner & Cameron, 2016). One general conclusion is that cross-group friendships are important due to the reciprocity and intimacy involved in having a close friendship which, in turn, provides the foundation for rejecting stereotypic expectations that come from societal messages and group norms. Group affiliation is a basic human need and while there are many positive benefits from affiliating with groups, group norms about loyalty often discourage making friends with individuals from groups that are perceived to be the outgroup (Dovidio & Gaertner, 2000; Rutland et al., 2005). An effective way to change group norms about loyalty is to expand the concept of who is included in the group, and friendships provide a pathway for effecting this change.

There are other known psychosocial benefits of cross-race/ethnic friendships that might also help explain their positive effects on intergroup attitudes. We know that having at least one cross-race/ethnic friend is associated with greater social competence. For example, Lease and Blake (2005) found that elementary age children who were members of the majority racial group in their classroom (Black or White) with a reciprocated interracial friendship were perceived as more popular, self-confident, and stronger leaders than their majority group peers without an interracial friend. Better social competence was also documented in research by Kawabata and Crick (2008). The 4^th grade students from Black, White, Asian American, and Latinx backgrounds who reported cross-ethnic friends in that study were rated by their teachers as possessing more leadership skills and being more relationally inclusive. Further, cross-race/ethnic friendships were associated with decreases in relational victimization and increases in peer support (beyond the contribution of same-race/ethnic friendships to these variables), indicating that cross-race/ethnic friendships are beneficial for both majority-group and minority-group children (Kawabata & Crick, 2011). In sum, these studies tell us that children who are willing to cross racial/ethnic boundaries in their friendship choices enjoy reputations as socially competent, sensitive, and influential in the peer group, all of which can be precursors of positive intergroup attitudes.

Cross-race/ethnic friendships also foster feelings of safety and less vulnerability (Bagci et al., 2014; Graham et al., 2014; Kawabata & Crick, 2015; Liu et al., 2020). For example, Graham et al. (2014) found that students attending more racially/ethnically diverse urban middle schools felt less bullied and safer in the school’s unsupervised spaces (e.g., restrooms and stairwells) when they had more cross-race/ethnic friends. In racially/ethnically diverse schools where availability (propinquity) is higher, having more friends from different racial/ethnic groups can help ward off potential bullies from those racial/ethnic outgroups. When cross-race/ethnic friends serve this protective function, it seems likely that their befrienders would have more positive attitudes toward the groups to which these friends belong, although the needed empirical studies to test this hypothesis have not yet been carried out. Overall, research has shown that cross-race/ethnic friendships can mitigate the stress students experience when encountering intergroup contexts in daily life.

Conditions that Contribute to Positive Outcomes for Intergroup Contact

As indicated earlier, intergroup contact theory recognizes that certain conditions need to be met for contact to have positive outcomes for all individuals (Pettigrew & Tropp, 2005). Social interactions that do not meet the goals of intergroup contact have the potential to actually increase prejudice or bias, unfortunately, such as when U.S. schools in the 1970s were legally mandated to desegregate. Authorities in the communities were not on board, there was a lack of common goals, unequal status among children and families, and few opportunities for cross-race friendships. The result was a series of riots, violence, and ostracism that followed many of the court orders, generating unsafe conditions for children in schools, including harassment, victimization, and psychological harm (Frankenberg & Orfield, 2007). Similarly, efforts at multiculturalism in Europe between individuals claiming a cultural heritage and immigrants who are perceived as members of outgroups have been difficult to achieve. When authorities are not in agreement with the aims of mutual respect, children do not report cross-group friendships, and newcomers feel they are treated as a threat to cultural unity (Verkuyten, 2011). However, when schools and communities work to create the conditions under which cross-group friendship can form and flourish, cross-group friendships can be an effective tool to prevent the possible negative outcomes of intergroup contact.

Do Cross-Group Friendships Benefit Minority Status Groups?

An important issue that has been debated in the literature is to what extent children from minority status groups benefit from cross-group friendships. Most of the original research on cross-group friendships focused on the numerical majority status groups, and often this has meant White-European-American, Anglo-British, or White European samples. One of the justifications for this focus was based on previous research demonstrating that majority status groups are those that are most likely to hold biases about other groups, which are strategies for maintaining social hierarchies. Thus, determining whether the majority group’s friendships with peers of different statuses reduced bias was necessary. However, understanding a complex phenomenon such as prejudice necessitates studying all perspectives, including the viewpoints of numerical minority status groups. This entails sampling minority status children and adolescents about their experiences with cross-group friendships, and whether this reduces biases that they might hold about majority status groups as well as other minority status peers.

In one program of research by McGlothlin and colleagues, racial and ethnic minority children who reported cross-race friendships displayed minimal bias (Margie et al., 2005; McGlothlin & Killen, 2010). However, majority White children attending homogeneous schools, with low reports of cross-race friendships, assigned more negative intent to a Black minority child than to a White majority child in an ambiguous situation task, expecting that a Black minority child would push rather than help a peer who fell off a swing (McGlothlin & Killen, 2006). This highlights how attending a racially heterogenous school reduces negative intergroup attitudes for White children, such as racial-outgroup attribution bias, and increases positive intergroup attitudes for Black children, such as minimal outgroup attribution bias. Further, Ruck et al., (2010) found support for U.S. ethnic minority children’s and adolescents’ beneficial outcomes for cross-group friendships in predominantly racial/ethnic minority schools; racial/ethnic minority students who reported cross-race/ethnic friendships were more likely to evaluate racial/ethnic exclusion as wrong and unfair. A study by McGill et al. (2012) found that intra- and interracial best friendships were related to social and emotional wellbeing in middle school with ethnic/racial minority students, indicating another positive benefit of these experiences for this population.

In social contexts where lasting intergroup friendships are possible and present, there are unique positive developmental outcomes for minority status group members. While cross-group friendships may reduce bias or prejudice in high status groups, different contexts lead to different types of positive benefits for individuals from majority and minority status groups. For example, more recently, it has been shown that increases in socioemotional well-being, a sense of safety at school, and a greater sense of the recognition given to one’s own group by the majority (termed public and private regard) reflect positive outcomes for students from disadvantaged groups, such as ethnic minority groups, which are different, in some contexts, than the outcomes for majority status groups (Brown, 2017; Graham et al., 2014; Rivas-Drake et al., 2019; Rivas-Drake & Umaña-Taylor, 2019). Unlike homogeneous schools, heterogeneous school environments provide opportunities for cross-group friendships and presumably a sense of safety and wellbeing for all students. Further, cross-ethnic friendships have been shown to be related to better intergroup attitudes among Asian American adolescents, and particularly the behavioral dimension of attitudes (Chen & Graham, 2015).

Yet heterogeneity, or a racially/ethnically diverse school environment, does not guarantee that cross-race/ethnic friendships will be beneficial for numerical minority groups. Research shows that it depends on who those cross-race/ethnic friends are and the peer ecology in which those friendships unfold. For example, in a study of Turkish ethnic minority adolescents in Germany, friendships with native German classmates (the numerical majority) exacerbated the negative effects of perceived discrimination such that Turkish youth who experienced discrimination and had native German friends were especially likely to feel depressed and report negative socioemotional well-being (Brenick et al., 2018). The negative effect of majority group cross-race/ethnic friends was especially apparent when classroom peer norms did not support more inclusive friendships. Moreover, Black adolescents in the U.S. with close cross-race White friends in the absence of close same-race friends who also face the challenges of being a numerical minority group can experience declines in well-being over time (McGill et al., 2012). Such findings remind us that the function of cross-race/ethnic friendships will vary depending on whether one’s own group is a numerical majority or minority group and where that group resides on a social status hierarchy.

School Factors that Limit Cross-Ethnic Friendships

In the previous sections, we focused on the benefits (and challenges) of cross-race/ethnic friendship for intergroup attitudes and psychosocial adjustment. In this section, we consider school structural factors and organizational practices that might interfere with the formation of those friendships. When these factors operate to recreate a segregated environment along racial lines, many students do not have quality contact with cross-race/ethnic peers even when those peers are available (propinquity) in racially/ethnically diverse schools.

Academic Tracking

Instructional policies such as academic tracking and ability grouping that limit the mixing opportunities of students can undermine the formation of cross-ethnic friendships. Because Black and Latinx students are more likely to be placed in lower ability tracks, whereas White and Asian students are more likely to be placed in higher ability tracks (e.g., Mickelson, 2015; Oakes, 2005), tracking can re-segregate students by limiting cross-ethnic exposure even in ethnically diverse schools. Studies both older and more recent have documented the negative effects of academic tracking on cross-ethnic friendships (e.g. Moody, 2001; Schofield & Sagar, 1977; Stearns, 2004). Furthermore, tracking appears to particularly inhibit the willingness of White students to form friendships with Black students and for Asian students to befriend anyone else but Whites (Hamm et al., 2005). In one recent study only high-achieving Black and Latinx students were befriended by White and Asian peers because the four groups were more likely to share the same academic classes (Chen et al., 2020).

Recent research on course-taking patterns and friendships provides insight into how racialized academic tracking might inhibit the formation of cross-ethnic friendships. For example, Frank et al., (2013) reported that new friendships among high school students were more likely to form within clusters of students who take sets of courses together (e.g., Algebra II, Chemistry, AP History). Because tracking even in a few core courses often dictates the remainder of students’ schedules, there is a greater likelihood that students who are ability grouped for some courses will be together for others as well. Clustering students around shared course-taking makes it easier for students to find new friends locally, while reducing their motivation to search more broadly (across track, race/ethnicity) in seeking these fresh ties.

Numerical Size

Just as tracking can limit the willingness of students to make friends with peers from different racial/ethnic groups even in diverse schools, so too can being a small numerical minority. Consider the following two ethnically diverse schools. School A is 26% White, 26% Latinx, 26% Asian, and 20% Black. School B is 28% White, 28% Latinx, 28% Asian, and 8% Black. Both of these schools are similarly diverse according to the standard metrics for calculating diversity (Budescu & Budescu, 2012). What is different about these two identically diverse schools is that the size of the Black population is much smaller in School B than School A (20% versus 8%).

Being a small numerical minority even in an ethnically diverse school influences the likelihood that students will cross racial boundaries to form friendships. Using Add Health data, Quillian and Campbell (2003) found that Latinx and Asian high school students were least likely to form cross-ethnic friendships when the size of their own groups was 10 percent or less. Wilson and Rodkin (2011) reported similar findings for Black 3^rd to 5^th graders. Numerical minority status in some cases promotes “hunkering down” (Putnam, 2007), or turning inward to find solidarity with same-race peers (Kogachi & Graham, 2020).

Unstructured Time and Group Norms

While the formal organization and structure of schooling can affect the opportunity for contact and friendship formation, informal ways that students organize themselves in peer groups can also limit opportunities to develop cross-ethnic friendships. Naturally occurring cross-ethnic interactions in many everyday contexts like the school cafeteria, playground, locker room, dance floor, or on the school bus are places where group dynamics are enacted to influence who children select as friends. For example, a small but growing literature is utilizing techniques of micro-ecological analyses of students’ everyday practices and routines (Dixon et al., 2008) to examine seating patterns of ethnically diverse students in school cafeterias (e.g., Clack et al., 2005; Echols et al., 2014; Ramiah et al., 2015). A common theme in these studies is that students of all ages studied prefer peers of their own ethnic group when given the choice of where to sit (see Dixon et al., 2020 for a recent extension to unstructured spaces among adults in Northern Ireland).

Unstructured spaces like the school cafeteria are the breeding ground for the development of peer group norms about inclusion and exclusion, which we know can have an effect on children’s friendship preferences. For example, Hitti and Killen (2015) investigated group norms, individual characteristics, and stereotypes that contributed to intergroup exclusion based on ethnic membership. They found that non-Arab American children expected their own group to be inclusive and invite Arab-American peers to join them; but they expected that Arab groups would be exclusive, preferring only to be with other Arab Americans. This asymmetry in group-level expectations can perpetuate ethnic segregation given that when youth expect members of an outgroup to be exclusive and anticipate social rejection, they are less likely to initiate contact.

To summarize, the ways schools organize students and students organize themselves can create fluid or rigid boundaries between different racial/ethnic groups. Instructional practices such as academic tracking, when racialized, can limit the mixing opportunities of students, and therefore undermine their willingness or ability to cross racial/ethnic boundaries to form new friendships. In addition, students who are small numerical minorities even in diverse schools might be especially reluctant to cross racial boundaries to seek out a new friend. And the informal ways that students organize themselves in peer groups during unstructured time in school often determine how rigid or fluid racial/ethnic group boundaries are. Finally, peer norms, both ingroup norms and perceived outgroup norms, convey messages to young people about the legitimacy of cross-ethnic relationships as well as what it means to be a part of their own group.

What About Biracial and Multiracial Youth?

The percentage of people who identify as biracial or multiracial is increasing dramatically in the U.S. In 2013, about 1 in 10 children were born to mixed-race parents (up from 1 in 100 in 1970) and these percentages are expected to triple by 2060 (Parker et al., 2015). Individuals who identify as multiracial are the fastest growing racial/ethnic group in this country, with a projected increase of 36 percent in the 10 years between the 2010 and 2020 census (Mather et al., 2019). Despite these rapidly shifting demographics, biracial/multiracial youth are not well represented in the cross-race friendship literature, which is not surprising since multiracial populations in general have been understudied in psychological research (Charmaraman et al., 2014; Nishina & Witkow, 2020). Studying interracial friendships of biracial youth is particularly challenging because of a lack of agreement in the literature on what constitutes a same-race versus cross-race friendship. Consider, for example, a biracial Asian-Latinx student who chooses a monoracial Asian classmate as a friend. Is this a same-race friend because they share part of the same racial ancestry? Or must the match be identical (another Asian-Latinx peer) to count as same-race? These are complex question with no definitive answers, which partly explains why biracial and multiracial youth have largely been excluded from the interracial friendship literature (see Doyle & Kao, 2007 for a notable exception).

In recent years, researchers have focused more on the functions of biracial youth in cross-group friendship formation rather than particular friendships per se. In one novel study taking a social network approach, Quillian and Redd (2009) suggested that biracial youth often take on a bridging function: by choosing friends from either monoracial group, they have the potential to bring these two racial groups together. Building on this work, Echols and Graham (2020) examined the friendship networks of monoracial and biracial 6^th grade students attending diverse middle schools. The authors found that monoracial peers from different racial/ethnic groups were more likely to form a cross-race friendship when they had a mutual biracial friend. By befriending each, the biracial peer in essence “brokered” the cross-racial friendship. If biracial youth are indeed well-connected in the peer network, with direct access to monoracial peers of different racial/ethnic groups, then they could be an important source of information about intergroup attitudes. Although not directly focused on biracial youth, research by Rivas-Drake and colleagues (2019) similarly demonstrated that peer networks shape cross-group friendships in diverse schools. Sampling 6^th – 8^th graders, their study found that students with more positive intergroup contact attitudes were most likely to be friends with similarly minded students (Rivas-Drake, et al,. 2019).

Further, researchers have demonstrated that exposure to biracial individuals significantly reduces endorsement of colorblindness as a racial ideology among White individuals (Gaither et al., 2019), providing additional support for the unique advantages of friendships with individuals whose heritage combines at least two monoracial identities. Moreover, including biracial and multiracial participants in research studies on cross-race friendships will provide novel information about the role of heterogeneous cultural contexts on how children think about racial categorization (Pauker et al., 2016). We encourage much more research on biracial and multiracial youth in the friendship literature, especially the unique role that they play in promoting cross-race/ethnic friendships and, by implication, improved intergroup attitudes. As the fastest growing racial/ethnic group reflecting the diversity of U.S. and worldwide, multiracial youth can provide new insight into racial dynamics, including the fluidity versus rigidity of crossing racial boundaries.

Cross-National Contexts for Intergroup Friendships

In Europe, research on intergroup contact has focused on children’s friendships between individuals identifying with the dominant nationality of a country and individuals from families that have migrated due to war-torn conditions at home or a historical conflict. For example, research has examined whether cross-group friendships reduce bias in Germany between German and Turkish children (Brenick et al., 2018; Feddes et al., 2009; Jugert et al., 2013). In Germany, the Turkish population is the largest ethnic minority group, and research suggests this ethnic group experiences high levels of discrimination and social exclusion (Wagner et al., 2003). Already in elementary school, children with a Turkish migration background have been shown to perform worse than their German peers (Feddes et al., 2009). This suggests that Turkish children hold a lower social status position compared to German children.

A study by Jugert et. al (2011) investigated mutual friendships between native German and Turkish preadolescents (mean age: 10 years) during their first year in an ethnically heterogeneous school. Participants in this study completed questionnaires at the beginning, middle, and end of the school year. Among both German and Turkish children, preference for same-ethnic friendships decreased over time. Variability in friendships was predicted by ingroup preference norms for German children, and over time, classroom identification increasingly reduced preference for same-ethnic friendships among Turkish children. Results from these studies highlight the potential for group norms to impact children’s friendship choices as distinct from larger societal systems, and the positive role that cross-group friendships play on intergroup attitudes for children between 7 and 11 years of age.

Understanding group norms is important given that majority group members expressing negative bias towards minority group members occurs early in development. Children’s biases were examined in a study of young Anglo-British children in the U.K.. Children ages 3–5 years had more racial bias towards African Caribbean-British compared to Asian-British outgroups (Rutland et al., 2005). Yet, children in racially mixed areas did not show discrimination in favor of the White ingroup, similar to the findings from McGlothlin and Killen (2006) in the U.S. with slightly older children (7–10 years old). Rutland et al. (2005) concluded that intergroup bias among young children may be reduced through the facilitation of intergroup contact. Much research in Europe has focused on the role that multiethnic contexts play in enabling minority status children to make friends with majority status peers, and to feel part of the school community (Bagci et al., 2014), although as noted earlier, positive outcomes do not always accompany these friendships (Brenick et al., 2018).

Extended or Indirect Contact

Cross-race friendships are not possible in all social contexts, such as communities, towns and entire regions that are racially/ethnically homogenous. One way that intergroup contact through friendships can be promoted is by what has been referred to as extended or indirect contact (Turner & Cameron, 2016; Munniskma et al., 2013; Munniskma et al., 2015). Direct contact refers to actual friendships, extended contact refers to reading stories about someone like yourself who is friends with another child perceived to be an outgroup member (e.g., different race, religion, nationality), and indirect contact refers to witnessing ingroup peers who are friends with others identified as the outgroup (Cameron et al., 2006; Cameron et al., 2007; Turner & Cameron, 2016; Wright et al., 1997).

Extended contact can help address prejudice in both adolescents (Liebkind & McAlister, 1999) and young children (Cameron & Rutland, 2008). Cameron and colleagues, for example, tested extended contact interventions for children as young as five years (see Cameron & Rutland, 2008). Children were shown examples of intergroup friendships through reading illustrated stories that portrayed friendships between ingroup and outgroup members (e.g., White English children and non-White refugee children). It was shown that the extended contact intervention was effective in improving attitudes towards outgroups amongst 5–11 years old children and across a variety of stigmatized outgroups, including the disabled, non-White refugees and South East Asian British.

In another extended contact study focusing on White English children’s attitudes towards Black refugees, Cameron et al. (2006) manipulated whether the social categories of those in the stories were mentioned (or not) and whether individual characteristics were emphasized (i.e., decategorization approach). The decategorization approach (Brewer & Miller, 1984) contends that in order for positive attitude change to occur during cross-group contact, the outgroup member should not be seen as being a member of the outgroup but instead should be treated as an individual. In other stories, the superordinate (i.e., school) category membership of the story characters was stressed (i.e., common ingroup identity approach). Gaertner and Dovidio (2000, 2005) have proposed that contact is most effective when a common ingroup identity is made salient. As an example, emphasizing a superordinate category (such as “we all belong to the same school” for White English and Black refugee students) creates positive attitudes by ingroup members toward the “new” ingroup members. Findings showed that the extended contact conditions compared to a control condition improved White children’s attitudes toward outgroups, including their willingness to befriend someone from those outgroups.

Direct intergroup contact through actual friendships is relatively more effective for promoting positive attitudes and reducing exclusion than is indirect or extended contact (Cameron et al., 2011; Rutland & Killen, 2015). Thus, these findings indicate that extended contact should not be seen as a replacement for direct contact, rather these types of indirect contact interventions should be seen as an alternative to direct contact or in some cases as the preliminary stage for contact when making cross-group friends through direct contact is not readily available.

Interventions to Build Cross-Race/ethnicity Friendships

Cross-race/ethnicity friendships appear to decline with age, although the needed within-person developmental research to systematically test these patterns has not been conducted. Given the benefits of cross-group friendships for challenging prejudicial attitudes and behavior, it is incumbent upon developmental scientists to understand whether it is feasible to promote cross-group friendships from childhood to adulthood and at what age should these efforts begin. However, research specifically aimed at building or promoting cross-group friendships in childhood and adolescents is surprisingly rare. A small literature called befriending interventions trains prosocial youth to come to the aid of their bullied peers (e.g., McElearney et al., 2008; Menesini et al., 2003) and encourages friendships along with effective communication, conflict resolution, and other types of positive peer support (Miller et al., 2017). But these interventions were not designed to intentionally build or measure friendships.

Experimental methods from social psychology might be a guide to friendship-building interventions. In one often-cited study, Page-Gould et. al. (2008) experimentally created same-race or cross-race friendships using a Fast Friends procedure (Aron et al., 1997) in which participants are led through a series of questions designed to increase self-disclosure with the result that participants feel closer to one another at the end of the procedure. College-age participants induced to form a fast friendship with someone of a different race showed improved attitudes toward that friend’s racial group (see also Kende et al., 2017 for a more recent application).

Building on the Fast Friends paradigm, Echols and Ivanich (2021) recently created an intervention to build new same-and cross-race/ethnic friendships of sixth grade students that lasted until the end of the school year. The transition to middle school may be a particularly appropriate time for such interventions since this is the developmental period when preference for cross-race/ethnic friends begin to decline. From a developmental intergroup perspective the time seems right to adapt some of the experimental procedures from social psychology research with adults to build stable cross-race/ethnicity friendships for youth in the K-12 context.

In research with elementary age children and building on principles from the contact hypothesis, a program titled Developing Inclusive Youth (DIY) was developed to enable children to promote intergroup friendships and improve intergroup attitudes (Killen et al., 2021). A multisite randomized control trial (RCT) was implemented to test the effectiveness of this program in 48 classrooms across 6 schools: 24 DIY and 24 Business-As-Usual (BAU) control groups, with children in 3^rd, 4^th, and 5^th grades, ages 8 – 11 years (N= 983). Participants engaged with an animated, interactive web-based curriculum tool that focused on peer social exclusion scenarios for 15 minutes, and then participated in a teacher-led classroom discussion for 30 minutes, once a week, for 8 weeks. Children interacted with the web-based tool by making choices, attributing emotional states, evaluating the peers’ decisions, and selecting reasons for their responses. The teacher-led class discussions, designed to create a safe space, focused on the students’ reactions to the scenarios and whether they had experienced similar events at school or home.

The web-based curriculum tool provided two forms of intergroup contact in the form of building cross-group friendships. First, an indirect intergroup contact experience was created when students watched scenarios in which someone like them became friends with a peer from a different diverse backgrounds (race, ethnicity, gender, immigrant status, and wealth status). The classroom discussions were a form of direct intergroup contact in that the teacher created a safe space which emphasized mutual respect within the peer group while children shared experiences of social exclusion and worked on solutions for perceived peer conflicts.

Controlling for pretest levels, children in the program were more likely to view interracial and same-race peer exclusion as wrong, associate positive traits with peers of different racial, ethnic, and gender backgrounds, and express a desire to play with peers from different backgrounds than were children in the control group. This program focused on building intergroup friendships across a range of social identities, including race, ethnicity, gender, immigrant status, and wealth status (Killen et al., 2021). It was conceived as a first step towards creating a school-based program designed to promote intergroup friendships in childhood.

Summary and Conclusions

Research on how cross-group friendships are related to children’s and adolescents’ prejudicial attitudes and biases has been conducted in different parts of the globe and reflects both the histories of intergroup relationships and current events in the world. Additionally, further areas of research have explored children’s moral development, specifically how having cross-race/ethnic friendships may help children and adolescents determine that racial exclusion is unfair. Moreover, other research has focused on school and whether authority figures support cross-group friendships through their instructional practices and whether cross-group friendships in school are related to social competence and perceived vulnerability. Thus, research has expanded to examine a larger array of outcomes than focusing solely on reducing prejudice or bias.

Although we highlight cross-race/ethnic friendships in this article as precursors to better intergroup attitudes in youth, we do not wish to leave the impression that same-race friendships should be discouraged or that they are inversely related to positive perceptions of outgroup members. Same-race friendships are important for emotional well-being, support, and the nurturing of a strong racial identity, especially during adolescence (e.g., Graham et al. 2014). And a positive ingroup identity has been related to better intergroup attitudes (e.g., Whitehead et al., 2009). Children benefit from being members of groups that provide support, safety, and protection (Rivas-Drake & Umaña-Taylor, 2019). Thus same race/ethnicity friendships can set in motion a process whereby youth develop stronger connections to who they are (racial identity) which in turns gives them the confidence to both embrace and accept outgroup members. More research is needed on the role of both same- and cross-race friendships for improved intergroup attitudes (Killen et al., 2017).

We noted that children of all ages show preference for same race/ethnic friends, which is consistent with homophily as a principle of friendship formation. But we also know that as children get older they understand that friendship is based on shared values, interests, and personality traits, not just strictly observable features such as gender, race, or ethnicity. For example, with age, children define similarity between peers in terms of shared interests over same-race preferences when confronted with pairs of same-race and cross-race dyads that do and do not share interests. McGlothlin and Killen (2005) found that children will rate an interracial peer dyad that shares sports interests as more similar than a same-race peer dyad that does not share sports interests, and they are more likely to do so when they attend racially/ethnically diverse schools rather than homogenous schools. Thus, one area for future research is to help children and adolescents understand what it means to share values and interests, and to expand the opportunities to make friends with a broader range of peers than they would otherwise (Hitti & Killen, 2015; Nesdale, et al., 2007).

The current article focused primarily on race/ethnicity as the social identity shaping cross-group friendships. This focus reflects the current state of developmental science. We believe that studying friendships and their consequences should embrace multiple social identities in addition to race/ethnicity. Whether youth are willing to cross group boundaries in the formation of friendships will partly depend on their gender, social class, nationality, immigrant status, country of origin, wealth status, and increasingly their sexual orientation. Incorporating one or more of these social identities in friendship research requires an intersectional approach, recognizing that individuals belong to multiple social categories simultaneously (e.g., Rosenthal, 2016). These categories combined in yet unknown ways to influence how youth perceive and are treated by others. Because multiple social identities can more easily become stigmatized identities for youth of color, an intersectional approach to cross-race/ethnicity friendships is particularly important (Rogers et al., 2015).

We began this article by situating the need to study cross-group friendships within the current context of increasing intergroup tensions and divisiveness worldwide. This is a critical time to be studying policies and practices that promote tolerance for young people who differ from one another along multiple social identities. Fostering cross-group friendships can be a powerful antidote to the divisiveness that appears to have gripped the world as we know it. Developmental science provides both a window into the origins of bias but also the mechanisms and forms of relationships that appear to reduce bias and promote positive relationships among children and adolescents from different backgrounds. Research studying the conditions that foster friendships and help create inclusive school environments is both urgent and timely given the context of intergroup relationships in the second decade of the 21^st century.